[v8,2/3] CMDQ: Mediatek CMDQ driver

Commit Message

[e-mail] May 30, 2016, 3:19 a.m. UTC

This patch is first version of Mediatek Command Queue(CMDQ) driver. The
CMDQ is used to help read/write registers with critical time limitation,
such as updating display configuration during the vblank. It controls
Global Command Engine (GCE) hardware to achieve this requirement.
Currently, CMDQ only supports display related hardwares, but we expect
it can be extended to other hardwares for future requirements.

Signed-off-by: HS Liao <[e-mail]>
Signed-off-by: CK Hu <[e-mail]>
---
 drivers/soc/mediatek/Kconfig    |  10 +
 drivers/soc/mediatek/Makefile   |   1 +
 drivers/soc/mediatek/mtk-cmdq.c | 943 ++++++++++++++++++++++++++++++++++++++++
 include/soc/mediatek/cmdq.h     | 197 +++++++++
 4 files changed, 1151 insertions(+)
 create mode 100644 drivers/soc/mediatek/mtk-cmdq.c
 create mode 100644 include/soc/mediatek/cmdq.h

Comments

CK Hu (胡俊光) May 30, 2016, 6:49 a.m. UTC | #1

Hi, HS:

Some comments inline.

On Mon, 2016-05-30 at 11:19 +0800, HS Liao wrote:
> This patch is first version of Mediatek Command Queue(CMDQ) driver. The
> CMDQ is used to help read/write registers with critical time limitation,
> such as updating display configuration during the vblank. It controls
> Global Command Engine (GCE) hardware to achieve this requirement.
> Currently, CMDQ only supports display related hardwares, but we expect
> it can be extended to other hardwares for future requirements.
> 
> Signed-off-by: HS Liao <[e-mail]>
> Signed-off-by: CK Hu <[e-mail]>
> ---
>  drivers/soc/mediatek/Kconfig    |  10 +
>  drivers/soc/mediatek/Makefile   |   1 +
>  drivers/soc/mediatek/mtk-cmdq.c | 943 ++++++++++++++++++++++++++++++++++++++++
>  include/soc/mediatek/cmdq.h     | 197 +++++++++
>  4 files changed, 1151 insertions(+)
>  create mode 100644 drivers/soc/mediatek/mtk-cmdq.c
>  create mode 100644 include/soc/mediatek/cmdq.h
> 
> diff --git a/drivers/soc/mediatek/Kconfig b/drivers/soc/mediatek/Kconfig
> index 0a4ea80..c4ad75c 100644
> --- a/drivers/soc/mediatek/Kconfig
> +++ b/drivers/soc/mediatek/Kconfig
> @@ -1,6 +1,16 @@
>  #
>  # MediaTek SoC drivers
>  #
> +config MTK_CMDQ
> +	bool "MediaTek CMDQ Support"
> +	depends on ARCH_MEDIATEK || COMPILE_TEST
> +	select MTK_INFRACFG
> +	help
> +	  Say yes here to add support for the MediaTek Command Queue (CMDQ)
> +	  driver. The CMDQ is used to help read/write registers with critical
> +	  time limitation, such as updating display configuration during the
> +	  vblank.
> +
>  config MTK_INFRACFG
>  	bool "MediaTek INFRACFG Support"
>  	depends on ARCH_MEDIATEK || COMPILE_TEST
> diff --git a/drivers/soc/mediatek/Makefile b/drivers/soc/mediatek/Makefile
> index 12998b0..f7397ef 100644
> --- a/drivers/soc/mediatek/Makefile
> +++ b/drivers/soc/mediatek/Makefile
> @@ -1,3 +1,4 @@
> +obj-$(CONFIG_MTK_CMDQ) += mtk-cmdq.o
>  obj-$(CONFIG_MTK_INFRACFG) += mtk-infracfg.o
>  obj-$(CONFIG_MTK_PMIC_WRAP) += mtk-pmic-wrap.o
>  obj-$(CONFIG_MTK_SCPSYS) += mtk-scpsys.o
> diff --git a/drivers/soc/mediatek/mtk-cmdq.c b/drivers/soc/mediatek/mtk-cmdq.c
> new file mode 100644
> index 0000000..e9d6e1c
> --- /dev/null
> +++ b/drivers/soc/mediatek/mtk-cmdq.c
> @@ -0,0 +1,943 @@
> +/*
> + * Copyright (c) 2015 MediaTek Inc.
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License version 2 as
> + * published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> + * GNU General Public License for more details.
> + */
> +
> +#include <linux/clk.h>
> +#include <linux/clk-provider.h>
> +#include <linux/completion.h>
> +#include <linux/dma-mapping.h>
> +#include <linux/errno.h>
> +#include <linux/interrupt.h>
> +#include <linux/iopoll.h>
> +#include <linux/kernel.h>
> +#include <linux/kthread.h>
> +#include <linux/module.h>
> +#include <linux/mutex.h>
> +#include <linux/of_address.h>
> +#include <linux/of_irq.h>
> +#include <linux/platform_device.h>
> +#include <linux/slab.h>
> +#include <linux/spinlock.h>
> +#include <linux/suspend.h>
> +#include <linux/workqueue.h>
> +#include <soc/mediatek/cmdq.h>
> +
> +#define CMDQ_INITIAL_CMD_BLOCK_SIZE	PAGE_SIZE
> +#define CMDQ_INST_SIZE			8 /* instruction is 64-bit */
> +#define CMDQ_TIMEOUT_MS			1000
> +#define CMDQ_IRQ_MASK			0xffff
> +#define CMDQ_DRIVER_DEVICE_NAME		"mtk_cmdq"
> +#define CMDQ_CLK_NAME			"gce"
> +
> +#define CMDQ_CURR_IRQ_STATUS		0x010
> +#define CMDQ_CURR_LOADED_THR		0x018
> +#define CMDQ_THR_SLOT_CYCLES		0x030
> +
> +#define CMDQ_THR_BASE			0x100
> +#define CMDQ_THR_SHIFT			0x080
> +#define CMDQ_THR_WARM_RESET		0x00
> +#define CMDQ_THR_ENABLE_TASK		0x04
> +#define CMDQ_THR_SUSPEND_TASK		0x08
> +#define CMDQ_THR_CURR_STATUS		0x0c
> +#define CMDQ_THR_IRQ_STATUS		0x10
> +#define CMDQ_THR_IRQ_ENABLE		0x14
> +#define CMDQ_THR_CURR_ADDR		0x20
> +#define CMDQ_THR_END_ADDR		0x24
> +#define CMDQ_THR_CFG			0x40
> +
> +#define CMDQ_THR_ENABLED		0x1
> +#define CMDQ_THR_DISABLED		0x0
> +#define CMDQ_THR_SUSPEND		0x1
> +#define CMDQ_THR_RESUME			0x0
> +#define CMDQ_THR_STATUS_SUSPENDED	BIT(1)
> +#define CMDQ_THR_DO_WARM_RESET		BIT(0)
> +#define CMDQ_THR_ACTIVE_SLOT_CYCLES	0x3200
> +#define CMDQ_THR_PRIORITY		3
> +#define CMDQ_THR_IRQ_DONE		0x1
> +#define CMDQ_THR_IRQ_ERROR		0x12
> +#define CMDQ_THR_IRQ_EN			0x13 /* done + error */
> +#define CMDQ_THR_IRQ_MASK		0x13
> +#define CMDQ_THR_EXECUTING		BIT(31)
> +
> +#define CMDQ_ARG_A_WRITE_MASK		0xffff
> +#define CMDQ_SUBSYS_MASK		0x1f
> +#define CMDQ_OP_CODE_MASK		0xff000000
> +
> +#define CMDQ_OP_CODE_SHIFT		24
> +#define CMDQ_SUBSYS_SHIFT		16
> +
> +#define CMDQ_WRITE_ENABLE_MASK		BIT(0)
> +#define CMDQ_JUMP_BY_OFFSET		0x10000000
> +#define CMDQ_JUMP_BY_PA			0x10000001
> +#define CMDQ_JUMP_PASS			CMDQ_INST_SIZE
> +#define CMDQ_WFE_UPDATE			BIT(31)
> +#define CMDQ_WFE_WAIT			BIT(15)
> +#define CMDQ_WFE_WAIT_VALUE		0x1
> +#define CMDQ_EOC_IRQ_EN			BIT(0)
> +
> +enum cmdq_thread_index {
> +	CMDQ_THR_DISP_MAIN_IDX,	/* main */
> +	CMDQ_THR_DISP_SUB_IDX,	/* sub */
> +	CMDQ_THR_DISP_MISC_IDX,	/* misc */
> +	CMDQ_THR_MAX_COUNT,	/* max */
> +};
> +
> +/*
> + * CMDQ_CODE_MOVE:
> + *   move value into internal register as mask
> + *   format: op mask
> + * CMDQ_CODE_WRITE:
> + *   write value into target register
> + *   format: op subsys address value
> + * CMDQ_CODE_JUMP:
> + *   jump by offset
> + *   format: op offset
> + * CMDQ_CODE_WFE:
> + *   wait for event and clear
> + *   it is just clear if no wait
> + *   format: [wait]  op event update:1 to_wait:1 wait:1
> + *           [clear] op event update:1 to_wait:0 wait:0
> + * CMDQ_CODE_EOC:
> + *   end of command
> + *   format: op irq_flag
> + */
> +enum cmdq_code {
> +	CMDQ_CODE_MOVE = 0x02,
> +	CMDQ_CODE_WRITE = 0x04,
> +	CMDQ_CODE_JUMP = 0x10,
> +	CMDQ_CODE_WFE = 0x20,
> +	CMDQ_CODE_EOC = 0x40,
> +};
> +
> +enum cmdq_task_state {
> +	TASK_STATE_BUSY, /* running on a GCE thread */
> +	TASK_STATE_ERROR,
> +	TASK_STATE_DONE,
> +};
> +
> +struct cmdq_task_cb {
> +	cmdq_async_flush_cb	cb;
> +	void			*data;
> +};
> +
> +struct cmdq_thread {
> +	void __iomem		*base;
> +	struct list_head	task_busy_list;
> +	wait_queue_head_t	wait_task_done;
> +};
> +
> +struct cmdq_task {
> +	struct cmdq		*cmdq;
> +	struct list_head	list_entry;
> +	enum cmdq_task_state	task_state;
> +	void			*va_base;
> +	dma_addr_t		pa_base;
> +	u64			engine_flag;
> +	size_t			command_size;
> +	u32			num_cmd;
> +	struct cmdq_thread	*thread;
> +	struct cmdq_task_cb	cb;
> +	struct work_struct	release_work;
> +};
> +
> +struct cmdq {
> +	struct device		*dev;
> +	void __iomem		*base;
> +	u32			irq;
> +	struct workqueue_struct	*task_release_wq;
> +	struct cmdq_thread	thread[CMDQ_THR_MAX_COUNT];
> +	struct mutex		task_mutex;	/* for task */
> +	spinlock_t		exec_lock;	/* for exec */
> +	struct clk		*clock;
> +	bool			suspended;
> +};
> +
> +struct cmdq_subsys {
> +	u32	base;
> +	int	id;
> +};
> +
> +static const struct cmdq_subsys gce_subsys[] = {
> +	{0x1400, 1},
> +	{0x1401, 2},
> +	{0x1402, 3},
> +};
> +
> +static int cmdq_subsys_base_to_id(u32 base)
> +{
> +	int i;
> +
> +	for (i = 0; i < ARRAY_SIZE(gce_subsys); i++)
> +		if (gce_subsys[i].base == base)
> +			return gce_subsys[i].id;
> +	return -EFAULT;
> +}
> +
> +static int cmdq_eng_get_thread(u64 flag)
> +{
> +	if (flag & BIT_ULL(CMDQ_ENG_DISP_DSI0))
> +		return CMDQ_THR_DISP_MAIN_IDX;
> +	else if (flag & BIT_ULL(CMDQ_ENG_DISP_DPI0))
> +		return CMDQ_THR_DISP_SUB_IDX;
> +	else
> +		return CMDQ_THR_DISP_MISC_IDX;
> +}
> +
> +static void cmdq_task_release(struct cmdq_task *task)
> +{
> +	struct cmdq *cmdq = task->cmdq;
> +
> +	dma_free_coherent(cmdq->dev, task->command_size, task->va_base,
> +			  task->pa_base);
> +	kfree(task);
> +}
> +
> +static struct cmdq_task *cmdq_task_acquire(struct cmdq_rec *rec,
> +					   struct cmdq_task_cb cb)
> +{
> +	struct cmdq *cmdq = rec->cmdq;
> +	struct device *dev = cmdq->dev;
> +	struct cmdq_task *task;
> +
> +	task = kzalloc(sizeof(*task), GFP_KERNEL);
> +	INIT_LIST_HEAD(&task->list_entry);
> +	task->va_base = dma_alloc_coherent(dev, rec->command_size,
> +					   &task->pa_base, GFP_KERNEL);
> +	if (!task->va_base) {
> +		dev_err(dev, "allocate command buffer failed\n");
> +		kfree(task);
> +		return NULL;
> +	}
> +
> +	task->cmdq = cmdq;
> +	task->command_size = rec->command_size;
> +	task->engine_flag = rec->engine_flag;
> +	task->task_state = TASK_STATE_BUSY;
> +	task->cb = cb;
> +	memcpy(task->va_base, rec->buf, rec->command_size);
> +	task->num_cmd = task->command_size / CMDQ_INST_SIZE;
> +	return task;
> +}
> +
> +static void cmdq_thread_writel(struct cmdq_thread *thread, u32 value,
> +			       u32 offset)
> +{
> +	writel(value, thread->base + offset);
> +}
> +
> +static u32 cmdq_thread_readl(struct cmdq_thread *thread, u32 offset)
> +{
> +	return readl(thread->base + offset);
> +}
> +
> +static int cmdq_thread_suspend(struct cmdq *cmdq, struct cmdq_thread *thread)
> +{
> +	u32 status;
> +
> +	cmdq_thread_writel(thread, CMDQ_THR_SUSPEND, CMDQ_THR_SUSPEND_TASK);
> +
> +	/* If already disabled, treat as suspended successful. */
> +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> +	      CMDQ_THR_ENABLED))
> +		return 0;
> +
> +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_CURR_STATUS,
> +			status, status & CMDQ_THR_STATUS_SUSPENDED, 0, 10)) {
> +		dev_err(cmdq->dev, "suspend GCE thread 0x%x failed\n",
> +			(u32)(thread->base - cmdq->base));
> +		return -EFAULT;
> +	}
> +
> +	return 0;
> +}
> +
> +static void cmdq_thread_resume(struct cmdq_thread *thread)
> +{
> +	cmdq_thread_writel(thread, CMDQ_THR_RESUME, CMDQ_THR_SUSPEND_TASK);
> +}
> +
> +static int cmdq_thread_reset(struct cmdq *cmdq, struct cmdq_thread *thread)
> +{
> +	u32 warm_reset;
> +
> +	cmdq_thread_writel(thread, CMDQ_THR_DO_WARM_RESET, CMDQ_THR_WARM_RESET);
> +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_WARM_RESET,
> +			warm_reset, !(warm_reset & CMDQ_THR_DO_WARM_RESET),
> +			0, 10)) {
> +		dev_err(cmdq->dev, "reset GCE thread 0x%x failed\n",
> +			(u32)(thread->base - cmdq->base));
> +		return -EFAULT;
> +	}
> +	writel(CMDQ_THR_ACTIVE_SLOT_CYCLES, cmdq->base + CMDQ_THR_SLOT_CYCLES);
> +	return 0;
> +}
> +
> +static void cmdq_thread_disable(struct cmdq *cmdq, struct cmdq_thread *thread)
> +{
> +	cmdq_thread_reset(cmdq, thread);
> +	cmdq_thread_writel(thread, CMDQ_THR_DISABLED, CMDQ_THR_ENABLE_TASK);
> +}
> +
> +/* notify GCE to re-fetch commands by setting GCE thread PC */
> +static void cmdq_thread_invalidate_fetched_data(struct cmdq_thread *thread)
> +{
> +	cmdq_thread_writel(thread,
> +			   cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR),
> +			   CMDQ_THR_CURR_ADDR);
> +}
> +
> +static void cmdq_task_insert_into_thread(struct cmdq_task *task)
> +{
> +	struct cmdq_thread *thread = task->thread;
> +	struct cmdq_task *prev_task = list_last_entry(
> +			&thread->task_busy_list, typeof(*task), list_entry);
> +	u64 *prev_task_base = prev_task->va_base;
> +
> +	/* let previous task jump to this task */
> +	prev_task_base[prev_task->num_cmd - 1] = (u64)CMDQ_JUMP_BY_PA << 32 |
> +						 task->pa_base;
> +
> +	cmdq_thread_invalidate_fetched_data(thread);
> +}
> +
> +/* we assume tasks in the same display GCE thread are waiting the same event. */
> +static void cmdq_task_remove_wfe(struct cmdq_task *task)
> +{
> +	u32 wfe_option = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
> +	u32 wfe_op = CMDQ_CODE_WFE << CMDQ_OP_CODE_SHIFT;
> +	u32 *base = task->va_base;
> +	u32 num_cmd = task->num_cmd << 1;
> +	int i;
> +
> +	for (i = 0; i < num_cmd; i += 2)
> +		if (base[i] == wfe_option &&
> +		    (base[i + 1] & CMDQ_OP_CODE_MASK) == wfe_op) {
> +			base[i] = CMDQ_JUMP_PASS;
> +			base[i + 1] = CMDQ_JUMP_BY_OFFSET;
> +		}
> +}
> +
> +static void cmdq_task_exec(struct cmdq_task *task, struct cmdq_thread *thread)
> +{
> +	struct cmdq *cmdq = task->cmdq;
> +	unsigned long flags;
> +	unsigned long curr_pa, end_pa;
> +
> +	WARN_ON(clk_prepare_enable(cmdq->clock) < 0);
> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
> +	task->thread = thread;
> +	task->task_state = TASK_STATE_BUSY;
> +	if (list_empty(&thread->task_busy_list)) {
> +		WARN_ON(cmdq_thread_reset(cmdq, thread) < 0);
> +
> +		cmdq_thread_writel(thread, task->pa_base, CMDQ_THR_CURR_ADDR);
> +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
> +				   CMDQ_THR_END_ADDR);
> +		cmdq_thread_writel(thread, CMDQ_THR_PRIORITY, CMDQ_THR_CFG);
> +		cmdq_thread_writel(thread, CMDQ_THR_IRQ_EN,
> +				   CMDQ_THR_IRQ_ENABLE);
> +
> +		cmdq_thread_writel(thread, CMDQ_THR_ENABLED,
> +				   CMDQ_THR_ENABLE_TASK);
> +	} else {
> +		WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
> +
> +		/*
> +		 * check boundary condition
> +		 * PC = END - 8, EOC is executed
> +		 * PC = END, all CMDs are executed
> +		 */
> +		curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> +		end_pa = cmdq_thread_readl(thread, CMDQ_THR_END_ADDR);
> +		if (curr_pa == end_pa - 8 || curr_pa == end_pa) {
> +			/* set to this task directly */
> +			cmdq_thread_writel(thread, task->pa_base,
> +					   CMDQ_THR_CURR_ADDR);
> +		} else {
> +			cmdq_task_insert_into_thread(task);
> +
> +			if (thread == &cmdq->thread[CMDQ_THR_DISP_MAIN_IDX] ||
> +			    thread == &cmdq->thread[CMDQ_THR_DISP_SUB_IDX])
> +				cmdq_task_remove_wfe(task);
> +
> +			smp_mb(); /* modify jump before enable thread */
> +		}
> +
> +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
> +				   CMDQ_THR_END_ADDR);
> +		cmdq_thread_resume(thread);
> +	}
> +	list_move_tail(&task->list_entry, &thread->task_busy_list);
> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> +}
> +
> +static void cmdq_handle_error_done(struct cmdq *cmdq,
> +				   struct cmdq_thread *thread, u32 irq_flag)
> +{
> +	struct cmdq_task *task, *tmp, *curr_task = NULL;
> +	u32 curr_pa;
> +	struct cmdq_cb_data cmdq_cb_data;
> +	bool err;
> +
> +	if (irq_flag & CMDQ_THR_IRQ_ERROR)
> +		err = true;
> +	else if (irq_flag & CMDQ_THR_IRQ_DONE)
> +		err = false;
> +	else
> +		return;
> +
> +	curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> +
> +	list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
> +				 list_entry) {
> +		if (curr_pa >= task->pa_base &&
> +		    curr_pa < (task->pa_base + task->command_size))
> +			curr_task = task;
> +		if (task->cb.cb) {
> +			cmdq_cb_data.err = curr_task ? err : false;
> +			cmdq_cb_data.data = task->cb.data;
> +			task->cb.cb(cmdq_cb_data);
> +		}
> +		task->task_state = (curr_task && err) ? TASK_STATE_ERROR :
> +				TASK_STATE_DONE;
> +		list_del(&task->list_entry);
> +		if (curr_task)
> +			break;
> +	}
> +
> +	wake_up(&thread->wait_task_done);
> +}
> +
> +static void cmdq_thread_irq_handler(struct cmdq *cmdq, int tid)
> +{
> +	struct cmdq_thread *thread = &cmdq->thread[tid];
> +	unsigned long flags = 0L;
> +	u32 irq_flag;
> +
> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
> +
> +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
> +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
> +
> +	/*
> +	 * Another CPU core could run "release task" right before we acquire
> +	 * the spin lock, and thus reset / disable this GCE thread, so we
> +	 * need to check the enable bit of this GCE thread.
> +	 */
> +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> +	      CMDQ_THR_ENABLED))
> +		irq_flag = 0;

These three statement (clear irq flag and detect thread enable) can be
moved into cmdq_handle_error_done() and the duplicated part in
cmdq_task_handle_error_result() can be removed. Even though
cmdq_task_handle_error_result() need not check thread enable, it would
not influence the flow.

> +
> +	cmdq_handle_error_done(cmdq, thread, irq_flag);
> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> +}
> +
> +static irqreturn_t cmdq_irq_handler(int irq, void *dev)
> +{
> +	struct cmdq *cmdq = dev;
> +	u32 irq_status;
> +	int i;
> +
> +	irq_status = readl(cmdq->base + CMDQ_CURR_IRQ_STATUS);
> +	irq_status &= CMDQ_IRQ_MASK;
> +	irq_status ^= CMDQ_IRQ_MASK;
> +
> +	if (!irq_status)
> +		return IRQ_NONE;
> +
> +	while (irq_status) {
> +		i = ffs(irq_status) - 1;
> +		irq_status &= ~BIT(i);
> +		cmdq_thread_irq_handler(cmdq, i);
> +	}
> +
> +	return IRQ_HANDLED;
> +}
> +
> +static int cmdq_task_handle_error_result(struct cmdq_task *task)
> +{
> +	struct cmdq *cmdq = task->cmdq;
> +	struct device *dev = cmdq->dev;
> +	struct cmdq_thread *thread = task->thread;
> +	struct cmdq_task *next_task, *prev_task;
> +	u32 irq_flag;
> +
> +	/* suspend GCE thread to ensure consistency */
> +	WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
> +
> +	/* ISR has handled this error task */
> +	if (task->task_state == TASK_STATE_ERROR) {
> +		next_task = list_first_entry_or_null(&thread->task_busy_list,
> +				struct cmdq_task, list_entry);
> +		if (next_task) /* move to next task */
> +			cmdq_thread_writel(thread, next_task->pa_base,
> +					   CMDQ_THR_CURR_ADDR);
> +		cmdq_thread_resume(thread);
> +		return -ECANCELED;
> +	}
> +
> +	/*
> +	 * Save next_task and prev_task in advance
> +	 * since cmdq_handle_error_done will remove list_entry.
> +	 */
> +	next_task = prev_task = NULL;
> +	if (task->list_entry.next != &thread->task_busy_list)
> +		next_task = list_next_entry(task, list_entry);
> +	if (task->list_entry.prev != &thread->task_busy_list)
> +		prev_task = list_prev_entry(task, list_entry);

I think there is always no previous task because task order in
thread->task_busy_list is the same as in cmdq->task_release_wq. So each
task processed by release work should be the first item in
thread->task_busy_list or be removed from thread->task_busy_list.

> +
> +	/*
> +	 * Although IRQ is disabled, GCE continues to execute.
> +	 * It may have pending IRQ before GCE thread is suspended,
> +	 * so check this condition again.
> +	 */
> +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
> +	cmdq_handle_error_done(cmdq, thread, irq_flag);
> +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
> +
> +	if (task->task_state == TASK_STATE_DONE) {
> +		cmdq_thread_resume(thread);
> +		return 0;
> +	}
> +
> +	if (task->task_state == TASK_STATE_ERROR) {
> +		dev_err(dev, "task 0x%p error\n", task);
> +		if (next_task) /* move to next task */
> +			cmdq_thread_writel(thread, next_task->pa_base,
> +					   CMDQ_THR_CURR_ADDR);
> +		cmdq_thread_resume(thread);
> +		return -ECANCELED;
> +	}
> +
> +	/* Task is running, so we force to remove it. */
> +	dev_err(dev, "task 0x%p timeout or killed\n", task);

No 'kill' state.

> +	task->task_state = TASK_STATE_ERROR;
> +
> +	if (prev_task) {
> +		u64 *prev_va = prev_task->va_base;
> +		u64 *curr_va = task->va_base;
> +
> +		/* copy JUMP instruction */
> +		prev_va[prev_task->num_cmd - 1] = curr_va[task->num_cmd - 1];
> +
> +		cmdq_thread_invalidate_fetched_data(thread);
> +	} else if (next_task) { /* move to next task */
> +		cmdq_thread_writel(thread, next_task->pa_base,
> +				   CMDQ_THR_CURR_ADDR);
> +	}

Just one statement in 'else if' part. So remove the parentheses.

> +
> +	list_del(&task->list_entry);
> +	cmdq_thread_resume(thread);
> +
> +	/* call cb here to prevent lock */

I think these three statements should be place together and there should
be no this comment.

1. Change task state to DONE or ERROR.
2. Callback
3. Remove task from task_busy_list.

> +	if (task->cb.cb) {
> +		struct cmdq_cb_data cmdq_cb_data;
> +
> +		cmdq_cb_data.err = true;
> +		cmdq_cb_data.data = task->cb.data;
> +		task->cb.cb(cmdq_cb_data);
> +	}
> +
> +	return -ECANCELED;
> +}
> +
> +static void cmdq_task_wait_release_work(struct work_struct *work_item)
> +{
> +	struct cmdq_task *task = container_of(work_item, struct cmdq_task,
> +					      release_work);
> +	struct cmdq *cmdq = task->cmdq;
> +	struct cmdq_thread *thread = task->thread;
> +	unsigned long flags;
> +
> +	wait_event_timeout(thread->wait_task_done,
> +			   task->task_state != TASK_STATE_BUSY,
> +			   msecs_to_jiffies(CMDQ_TIMEOUT_MS));
> +
> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
> +	if (task->task_state != TASK_STATE_DONE)
> +		cmdq_task_handle_error_result(task);
> +	if (list_empty(&thread->task_busy_list))
> +		cmdq_thread_disable(cmdq, thread);
> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> +
> +	/* release regardless of success or not */
> +	clk_disable_unprepare(cmdq->clock);
> +	cmdq_task_release(task);
> +}
> +
> +static void cmdq_task_wait_release_schedule(struct cmdq_task *task)
> +{
> +	struct cmdq *cmdq = task->cmdq;
> +
> +	INIT_WORK(&task->release_work, cmdq_task_wait_release_work);
> +	queue_work(cmdq->task_release_wq, &task->release_work);
> +}
> +
> +static int cmdq_rec_realloc_cmd_buffer(struct cmdq_rec *rec, size_t size)
> +{
> +	void *new_buf;
> +
> +	new_buf = krealloc(rec->buf, size, GFP_KERNEL | __GFP_ZERO);
> +	if (!new_buf)
> +		return -ENOMEM;
> +	rec->buf = new_buf;
> +	rec->buf_size = size;
> +	return 0;
> +}
> +
> +struct cmdq_base *cmdq_register_device(struct device *dev)
> +{
> +	struct cmdq_base *cmdq_base;
> +	struct resource res;
> +	int subsys;
> +	u32 base;
> +
> +	if (of_address_to_resource(dev->of_node, 0, &res))
> +		return NULL;
> +	base = (u32)res.start;
> +
> +	subsys = cmdq_subsys_base_to_id(base >> 16);
> +	if (subsys < 0)
> +		return NULL;
> +
> +	cmdq_base = devm_kmalloc(dev, sizeof(*cmdq_base), GFP_KERNEL);
> +	if (!cmdq_base)
> +		return NULL;
> +	cmdq_base->subsys = subsys;
> +	cmdq_base->base = base;
> +
> +	return cmdq_base;
> +}
> +EXPORT_SYMBOL(cmdq_register_device);
> +
> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
> +		    struct cmdq_rec **rec_ptr)
> +{
> +	struct cmdq_rec *rec;
> +	int err;
> +
> +	rec = kzalloc(sizeof(*rec), GFP_KERNEL);
> +	if (!rec)
> +		return -ENOMEM;
> +	rec->cmdq = dev_get_drvdata(dev);
> +	rec->engine_flag = engine_flag;
> +	err = cmdq_rec_realloc_cmd_buffer(rec, CMDQ_INITIAL_CMD_BLOCK_SIZE);
> +	if (err < 0) {
> +		kfree(rec);
> +		return err;
> +	}
> +	*rec_ptr = rec;
> +	return 0;
> +}
> +EXPORT_SYMBOL(cmdq_rec_create);
> +
> +static int cmdq_rec_append_command(struct cmdq_rec *rec, enum cmdq_code code,
> +				   u32 arg_a, u32 arg_b)
> +{
> +	u64 *cmd_ptr;
> +	int err;
> +
> +	if (WARN_ON(rec->finalized))
> +		return -EBUSY;
> +	if (code < CMDQ_CODE_MOVE || code > CMDQ_CODE_EOC)
> +		return -EINVAL;
> +	if (unlikely(rec->command_size + CMDQ_INST_SIZE > rec->buf_size)) {
> +		err = cmdq_rec_realloc_cmd_buffer(rec, rec->buf_size * 2);
> +		if (err < 0)
> +			return err;
> +	}
> +	cmd_ptr = rec->buf + rec->command_size;
> +	(*cmd_ptr) = (u64)((code << CMDQ_OP_CODE_SHIFT) | arg_a) << 32 | arg_b;
> +	rec->command_size += CMDQ_INST_SIZE;
> +	return 0;
> +}
> +
> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value, struct cmdq_base *base,
> +		   u32 offset)
> +{
> +	u32 arg_a = ((base->base + offset) & CMDQ_ARG_A_WRITE_MASK) |
> +		    ((base->subsys & CMDQ_SUBSYS_MASK) << CMDQ_SUBSYS_SHIFT);
> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WRITE, arg_a, value);
> +}
> +EXPORT_SYMBOL(cmdq_rec_write);
> +
> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
> +			struct cmdq_base *base, u32 offset, u32 mask)
> +{
> +	u32 offset_mask = offset;
> +	int err;
> +
> +	if (mask != 0xffffffff) {
> +		err = cmdq_rec_append_command(rec, CMDQ_CODE_MOVE, 0, ~mask);
> +		if (err < 0)
> +			return err;
> +		offset_mask |= CMDQ_WRITE_ENABLE_MASK;
> +	}
> +	return cmdq_rec_write(rec, value, base, offset_mask);
> +}
> +EXPORT_SYMBOL(cmdq_rec_write_mask);
> +
> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event)
> +{
> +	u32 arg_b;
> +
> +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
> +		return -EINVAL;
> +
> +	/*
> +	 * bit 0-11: wait value
> +	 * bit 15: 1 - wait, 0 - no wait
> +	 * bit 16-27: update value
> +	 * bit 31: 1 - update, 0 - no update
> +	 */
> +	arg_b = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event, arg_b);
> +}
> +EXPORT_SYMBOL(cmdq_rec_wfe);
> +
> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event)
> +{
> +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
> +		return -EINVAL;
> +
> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event,
> +				       CMDQ_WFE_UPDATE);
> +}
> +EXPORT_SYMBOL(cmdq_rec_clear_event);
> +
> +static int cmdq_rec_finalize(struct cmdq_rec *rec)
> +{
> +	int err;
> +
> +	if (rec->finalized)
> +		return 0;
> +
> +	/* insert EOC and generate IRQ for each command iteration */
> +	err = cmdq_rec_append_command(rec, CMDQ_CODE_EOC, 0, CMDQ_EOC_IRQ_EN);
> +	if (err < 0)
> +		return err;
> +
> +	/* JUMP to end */
> +	err = cmdq_rec_append_command(rec, CMDQ_CODE_JUMP, 0, CMDQ_JUMP_PASS);
> +	if (err < 0)
> +		return err;
> +
> +	rec->finalized = true;
> +	return 0;
> +}
> +
> +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
> +			 void *data)
> +{
> +	struct cmdq *cmdq = rec->cmdq;
> +	struct cmdq_task *task;
> +	struct cmdq_task_cb task_cb;
> +	struct cmdq_thread *thread;
> +	int err;
> +
> +	mutex_lock(&cmdq->task_mutex);
> +	if (cmdq->suspended) {
> +		dev_err(cmdq->dev, "%s is called after suspended\n", __func__);
> +		mutex_unlock(&cmdq->task_mutex);
> +		return -EPERM;
> +	}
> +
> +	err = cmdq_rec_finalize(rec);
> +	if (err < 0) {
> +		mutex_unlock(&cmdq->task_mutex);
> +		return err;
> +	}
> +
> +	task_cb.cb = cb;
> +	task_cb.data = data;
> +	task = cmdq_task_acquire(rec, task_cb);
> +	if (!task) {
> +		mutex_unlock(&cmdq->task_mutex);
> +		return -EFAULT;
> +	}
> +
> +	thread = &cmdq->thread[cmdq_eng_get_thread(task->engine_flag)];
> +	cmdq_task_exec(task, thread);
> +	cmdq_task_wait_release_schedule(task);
> +	mutex_unlock(&cmdq->task_mutex);
> +	return 0;
> +}
> +EXPORT_SYMBOL(cmdq_rec_flush_async);
> +
> +struct cmdq_flush_completion {
> +	struct completion cmplt;
> +	bool err;
> +};
> +
> +static int cmdq_rec_flush_cb(struct cmdq_cb_data data)
> +{
> +	struct cmdq_flush_completion *cmplt = data.data;
> +
> +	cmplt->err = data.err;
> +	complete(&cmplt->cmplt);
> +	return 0;
> +}
> +
> +int cmdq_rec_flush(struct cmdq_rec *rec)
> +{
> +	struct cmdq_flush_completion cmplt;
> +	int err;
> +
> +	init_completion(&cmplt.cmplt);
> +	err = cmdq_rec_flush_async(rec, cmdq_rec_flush_cb, &cmplt);
> +	if (err < 0)
> +		return err;
> +	wait_for_completion(&cmplt.cmplt);
> +	return cmplt.err ? -EFAULT : 0;
> +}
> +EXPORT_SYMBOL(cmdq_rec_flush);
> +
> +void cmdq_rec_destroy(struct cmdq_rec *rec)
> +{
> +	kfree(rec->buf);
> +	kfree(rec);
> +}
> +EXPORT_SYMBOL(cmdq_rec_destroy);
> +
> +static bool cmdq_task_is_empty(struct cmdq *cmdq)
> +{
> +	struct cmdq_thread *thread;
> +	int i;
> +
> +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
> +		thread = &cmdq->thread[i];
> +		if (!list_empty(&thread->task_busy_list))
> +			return false;
> +	}
> +	return true;
> +}
> +
> +static int cmdq_suspend(struct device *dev)
> +{
> +	struct cmdq *cmdq = dev_get_drvdata(dev);
> +	u32 exec_threads;
> +
> +	mutex_lock(&cmdq->task_mutex);
> +	cmdq->suspended = true;
> +	mutex_unlock(&cmdq->task_mutex);
> +
> +	exec_threads = readl(cmdq->base + CMDQ_CURR_LOADED_THR);
> +	if ((exec_threads & CMDQ_THR_EXECUTING) && !cmdq_task_is_empty(cmdq)) {
> +		dev_err(dev, "wait active tasks timeout.\n");
> +		flush_workqueue(cmdq->task_release_wq);
> +	}

Maybe it need not to check exec_threads, it may be written as:

	if (!cmdq_task_is_empty(cmdq))
		dev_err(dev, "wait active tasks timeout.\n");

	flush_workqueue(cmdq->task_release_wq);

Even though task_busy_list is empty, that does not mean every task in
cmdq->task_release_wq is finished. So always do flush
cmdq->task_release_wq.

> +	return 0;
> +}
> +
> +static int cmdq_resume(struct device *dev)
> +{
> +	struct cmdq *cmdq = dev_get_drvdata(dev);
> +
> +	cmdq->suspended = false;
> +	return 0;
> +}
> +
> +static int cmdq_remove(struct platform_device *pdev)
> +{
> +	struct cmdq *cmdq = platform_get_drvdata(pdev);
> +
> +	destroy_workqueue(cmdq->task_release_wq);
> +	cmdq->task_release_wq = NULL;
> +	return 0;
> +}
> +
> +static int cmdq_probe(struct platform_device *pdev)
> +{
> +	struct device *dev = &pdev->dev;
> +	struct device_node *node = dev->of_node;
> +	struct resource *res;
> +	struct cmdq *cmdq;
> +	int err, i;
> +
> +	cmdq = devm_kzalloc(dev, sizeof(*cmdq), GFP_KERNEL);
> +	if (!cmdq)
> +		return -ENOMEM;
> +	cmdq->dev = dev;
> +
> +	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> +	cmdq->base = devm_ioremap_resource(dev, res);
> +	if (IS_ERR(cmdq->base)) {
> +		dev_err(dev, "failed to ioremap gce\n");
> +		return PTR_ERR(cmdq->base);
> +	}
> +
> +	cmdq->irq = irq_of_parse_and_map(node, 0);
> +	if (!cmdq->irq) {
> +		dev_err(dev, "failed to get irq\n");
> +		return -EINVAL;
> +	}
> +
> +	dev_dbg(dev, "cmdq device: addr:0x%p, va:0x%p, irq:%d\n",
> +		dev, cmdq->base, cmdq->irq);
> +
> +	mutex_init(&cmdq->task_mutex);
> +	spin_lock_init(&cmdq->exec_lock);
> +	cmdq->task_release_wq = alloc_ordered_workqueue(
> +			"%s", WQ_MEM_RECLAIM | WQ_HIGHPRI,
> +			"cmdq_task_wait_release");
> +
> +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
> +		cmdq->thread[i].base = cmdq->base + CMDQ_THR_BASE +
> +				CMDQ_THR_SHIFT * i;
> +		init_waitqueue_head(&cmdq->thread[i].wait_task_done);
> +		INIT_LIST_HEAD(&cmdq->thread[i].task_busy_list);
> +	}
> +
> +	platform_set_drvdata(pdev, cmdq);
> +
> +	err = devm_request_irq(dev, cmdq->irq, cmdq_irq_handler, IRQF_SHARED,
> +			       CMDQ_DRIVER_DEVICE_NAME, cmdq);
> +	if (err < 0) {
> +		dev_err(dev, "failed to register ISR (%d)\n", err);
> +		goto fail;
> +	}
> +
> +	cmdq->clock = devm_clk_get(dev, CMDQ_CLK_NAME);
> +	if (IS_ERR(cmdq->clock)) {
> +		dev_err(dev, "failed to get clk:%s\n", CMDQ_CLK_NAME);
> +		err = PTR_ERR(cmdq->clock);
> +		goto fail;
> +	}
> +	return 0;
> +
> +fail:
> +	cmdq_remove(pdev);
> +	return err;
> +}
> +
> +static const struct dev_pm_ops cmdq_pm_ops = {
> +	.suspend = cmdq_suspend,
> +	.resume = cmdq_resume,
> +};
> +
> +static const struct of_device_id cmdq_of_ids[] = {
> +	{.compatible = "mediatek,mt8173-gce",},
> +	{}
> +};
> +
> +static struct platform_driver cmdq_drv = {
> +	.probe = cmdq_probe,
> +	.remove = cmdq_remove,
> +	.driver = {
> +		.name = CMDQ_DRIVER_DEVICE_NAME,
> +		.owner = THIS_MODULE,
> +		.pm = &cmdq_pm_ops,
> +		.of_match_table = cmdq_of_ids,
> +	}
> +};
> +
> +builtin_platform_driver(cmdq_drv);
> diff --git a/include/soc/mediatek/cmdq.h b/include/soc/mediatek/cmdq.h
> new file mode 100644
> index 0000000..60eef3d
> --- /dev/null
> +++ b/include/soc/mediatek/cmdq.h
> @@ -0,0 +1,197 @@
> +/*
> + * Copyright (c) 2015 MediaTek Inc.
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License version 2 as
> + * published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> + * GNU General Public License for more details.
> + */
> +
> +#ifndef __MTK_CMDQ_H__
> +#define __MTK_CMDQ_H__
> +
> +#include <linux/platform_device.h>
> +#include <linux/types.h>
> +
> +enum cmdq_eng {
> +	CMDQ_ENG_DISP_AAL,
> +	CMDQ_ENG_DISP_COLOR0,
> +	CMDQ_ENG_DISP_COLOR1,
> +	CMDQ_ENG_DISP_DPI0,
> +	CMDQ_ENG_DISP_DSI0,
> +	CMDQ_ENG_DISP_DSI1,
> +	CMDQ_ENG_DISP_GAMMA,
> +	CMDQ_ENG_DISP_OD,
> +	CMDQ_ENG_DISP_OVL0,
> +	CMDQ_ENG_DISP_OVL1,
> +	CMDQ_ENG_DISP_PWM0,
> +	CMDQ_ENG_DISP_PWM1,
> +	CMDQ_ENG_DISP_RDMA0,
> +	CMDQ_ENG_DISP_RDMA1,
> +	CMDQ_ENG_DISP_RDMA2,
> +	CMDQ_ENG_DISP_UFOE,
> +	CMDQ_ENG_DISP_WDMA0,
> +	CMDQ_ENG_DISP_WDMA1,
> +	CMDQ_ENG_MAX,
> +};
> +
> +/* events for CMDQ and display */
> +enum cmdq_event {
> +	/* Display start of frame(SOF) events */
> +	CMDQ_EVENT_DISP_OVL0_SOF = 11,
> +	CMDQ_EVENT_DISP_OVL1_SOF = 12,
> +	CMDQ_EVENT_DISP_RDMA0_SOF = 13,
> +	CMDQ_EVENT_DISP_RDMA1_SOF = 14,
> +	CMDQ_EVENT_DISP_RDMA2_SOF = 15,
> +	CMDQ_EVENT_DISP_WDMA0_SOF = 16,
> +	CMDQ_EVENT_DISP_WDMA1_SOF = 17,
> +	/* Display end of frame(EOF) events */
> +	CMDQ_EVENT_DISP_OVL0_EOF = 39,
> +	CMDQ_EVENT_DISP_OVL1_EOF = 40,
> +	CMDQ_EVENT_DISP_RDMA0_EOF = 41,
> +	CMDQ_EVENT_DISP_RDMA1_EOF = 42,
> +	CMDQ_EVENT_DISP_RDMA2_EOF = 43,
> +	CMDQ_EVENT_DISP_WDMA0_EOF = 44,
> +	CMDQ_EVENT_DISP_WDMA1_EOF = 45,
> +	/* Mutex end of frame(EOF) events */
> +	CMDQ_EVENT_MUTEX0_STREAM_EOF = 53,
> +	CMDQ_EVENT_MUTEX1_STREAM_EOF = 54,
> +	CMDQ_EVENT_MUTEX2_STREAM_EOF = 55,
> +	CMDQ_EVENT_MUTEX3_STREAM_EOF = 56,
> +	CMDQ_EVENT_MUTEX4_STREAM_EOF = 57,
> +	/* Display underrun events */
> +	CMDQ_EVENT_DISP_RDMA0_UNDERRUN = 63,
> +	CMDQ_EVENT_DISP_RDMA1_UNDERRUN = 64,
> +	CMDQ_EVENT_DISP_RDMA2_UNDERRUN = 65,
> +	/* Keep this at the end of HW events */
> +	CMDQ_MAX_HW_EVENT_COUNT = 260,
> +};
> +
> +struct cmdq_cb_data {
> +	bool	err;
> +	void	*data;
> +};
> +
> +typedef int (*cmdq_async_flush_cb)(struct cmdq_cb_data data);
> +
> +struct cmdq_task;
> +struct cmdq;
> +
> +struct cmdq_rec {
> +	struct cmdq		*cmdq;
> +	u64			engine_flag;
> +	size_t			command_size;
> +	void			*buf;
> +	size_t			buf_size;
> +	bool			finalized;
> +};
> +
> +struct cmdq_base {
> +	int	subsys;
> +	u32	base;
> +};
> +
> +/**
> + * cmdq_register_device() - register device which needs CMDQ
> + * @dev:		device
> + *
> + * Return: cmdq_base pointer or NULL for failed
> + */
> +struct cmdq_base *cmdq_register_device(struct device *dev);
> +
> +/**
> + * cmdq_rec_create() - create command queue record
> + * @dev:		device
> + * @engine_flag:	command queue engine flag
> + * @rec_ptr:		command queue record pointer to retrieve cmdq_rec
> + *
> + * Return: 0 for success; else the error code is returned
> + */
> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
> +		    struct cmdq_rec **rec_ptr);
> +
> +/**
> + * cmdq_rec_write() - append write command to the command queue record
> + * @rec:	the command queue record
> + * @value:	the specified target register value
> + * @base:	the command queue base
> + * @offset:	register offset from module base
> + *
> + * Return: 0 for success; else the error code is returned
> + */
> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value,
> +		   struct cmdq_base *base, u32 offset);
> +
> +/**
> + * cmdq_rec_write_mask() - append write command with mask to the command
> + *			   queue record
> + * @rec:	the command queue record
> + * @value:	the specified target register value
> + * @base:	the command queue base
> + * @offset:	register offset from module base
> + * @mask:	the specified target register mask
> + *
> + * Return: 0 for success; else the error code is returned
> + */
> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
> +			struct cmdq_base *base, u32 offset, u32 mask);
> +
> +/**
> + * cmdq_rec_wfe() - append wait for event command to the command queue record
> + * @rec:	the command queue record
> + * @event:	the desired event type to "wait and CLEAR"
> + *
> + * Return: 0 for success; else the error code is returned
> + */
> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event);
> +
> +/**
> + * cmdq_rec_clear_event() - append clear event command to the command queue
> + *			    record
> + * @rec:	the command queue record
> + * @event:	the desired event to be cleared
> + *
> + * Return: 0 for success; else the error code is returned
> + */
> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event);
> +
> +/**
> + * cmdq_rec_flush() - trigger CMDQ to execute the recorded commands
> + * @rec:	the command queue record
> + *
> + * Return: 0 for success; else the error code is returned
> + *
> + * Trigger CMDQ to execute the recorded commands. Note that this is a
> + * synchronous flush function. When the function returned, the recorded
> + * commands have been done.
> + */
> +int cmdq_rec_flush(struct cmdq_rec *rec);
> +
> +/**
> + * cmdq_rec_flush_async() - trigger CMDQ to asynchronously execute the recorded
> + *			    commands and call back after ISR is finished
> + * @rec:	the command queue record
> + * @cb:		called in the end of CMDQ ISR
> + * @data:	this data will pass back to cb
> + *
> + * Return: 0 for success; else the error code is returned
> + *
> + * Trigger CMDQ to asynchronously execute the recorded commands and call back
> + * after ISR is finished. Note that this is an ASYNC function. When the function
> + * returned, it may or may not be finished. The ISR callback function is called
> + * in the end of ISR.
> + */
> +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
> +			 void *data);
> +
> +/**
> + * cmdq_rec_destroy() - destroy command queue record
> + * @rec:	the command queue record
> + */
> +void cmdq_rec_destroy(struct cmdq_rec *rec);
> +
> +#endif	/* __MTK_CMDQ_H__ */

Regards,
CK

[e-mail] May 30, 2016, 9:38 a.m. UTC | #2

Hi CK,

Reply inline.

On Mon, 2016-05-30 at 14:49 +0800, CK Hu wrote:
> Hi, HS:
> 
> Some comments inline.
> 
> On Mon, 2016-05-30 at 11:19 +0800, HS Liao wrote:
...
> > +static void cmdq_thread_irq_handler(struct cmdq *cmdq, int tid)
> > +{
> > +	struct cmdq_thread *thread = &cmdq->thread[tid];
> > +	unsigned long flags = 0L;
> > +	u32 irq_flag;
> > +
> > +	spin_lock_irqsave(&cmdq->exec_lock, flags);
> > +
> > +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
> > +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
> > +
> > +	/*
> > +	 * Another CPU core could run "release task" right before we acquire
> > +	 * the spin lock, and thus reset / disable this GCE thread, so we
> > +	 * need to check the enable bit of this GCE thread.
> > +	 */
> > +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> > +	      CMDQ_THR_ENABLED))
> > +		irq_flag = 0;
> 
> These three statement (clear irq flag and detect thread enable) can be
> moved into cmdq_handle_error_done() and the duplicated part in
> cmdq_task_handle_error_result() can be removed. Even though
> cmdq_task_handle_error_result() need not check thread enable, it would
> not influence the flow.

Will do.

> > +
> > +	cmdq_handle_error_done(cmdq, thread, irq_flag);
> > +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> > +}
...
> > +static int cmdq_task_handle_error_result(struct cmdq_task *task)
> > +{
> > +	struct cmdq *cmdq = task->cmdq;
> > +	struct device *dev = cmdq->dev;
> > +	struct cmdq_thread *thread = task->thread;
> > +	struct cmdq_task *next_task, *prev_task;
> > +	u32 irq_flag;
> > +
> > +	/* suspend GCE thread to ensure consistency */
> > +	WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
> > +
> > +	/* ISR has handled this error task */
> > +	if (task->task_state == TASK_STATE_ERROR) {
> > +		next_task = list_first_entry_or_null(&thread->task_busy_list,
> > +				struct cmdq_task, list_entry);
> > +		if (next_task) /* move to next task */
> > +			cmdq_thread_writel(thread, next_task->pa_base,
> > +					   CMDQ_THR_CURR_ADDR);
> > +		cmdq_thread_resume(thread);
> > +		return -ECANCELED;
> > +	}
> > +
> > +	/*
> > +	 * Save next_task and prev_task in advance
> > +	 * since cmdq_handle_error_done will remove list_entry.
> > +	 */
> > +	next_task = prev_task = NULL;
> > +	if (task->list_entry.next != &thread->task_busy_list)
> > +		next_task = list_next_entry(task, list_entry);
> > +	if (task->list_entry.prev != &thread->task_busy_list)
> > +		prev_task = list_prev_entry(task, list_entry);
> 
> I think there is always no previous task because task order in
> thread->task_busy_list is the same as in cmdq->task_release_wq. So each
> task processed by release work should be the first item in
> thread->task_busy_list or be removed from thread->task_busy_list.

Will remove previous task.

> > +
> > +	/*
> > +	 * Although IRQ is disabled, GCE continues to execute.
> > +	 * It may have pending IRQ before GCE thread is suspended,
> > +	 * so check this condition again.
> > +	 */
> > +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
> > +	cmdq_handle_error_done(cmdq, thread, irq_flag);
> > +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
> > +
> > +	if (task->task_state == TASK_STATE_DONE) {
> > +		cmdq_thread_resume(thread);
> > +		return 0;
> > +	}
> > +
> > +	if (task->task_state == TASK_STATE_ERROR) {
> > +		dev_err(dev, "task 0x%p error\n", task);
> > +		if (next_task) /* move to next task */
> > +			cmdq_thread_writel(thread, next_task->pa_base,
> > +					   CMDQ_THR_CURR_ADDR);
> > +		cmdq_thread_resume(thread);
> > +		return -ECANCELED;
> > +	}
> > +
> > +	/* Task is running, so we force to remove it. */
> > +	dev_err(dev, "task 0x%p timeout or killed\n", task);
> 
> No 'kill' state.

From CMDQ v8, there is no killed task.
I will rewrite output message.

> > +	task->task_state = TASK_STATE_ERROR;
> > +
> > +	if (prev_task) {
> > +		u64 *prev_va = prev_task->va_base;
> > +		u64 *curr_va = task->va_base;
> > +
> > +		/* copy JUMP instruction */
> > +		prev_va[prev_task->num_cmd - 1] = curr_va[task->num_cmd - 1];
> > +
> > +		cmdq_thread_invalidate_fetched_data(thread);
> > +	} else if (next_task) { /* move to next task */
> > +		cmdq_thread_writel(thread, next_task->pa_base,
> > +				   CMDQ_THR_CURR_ADDR);
> > +	}
> 
> Just one statement in 'else if' part. So remove the parentheses.

I think we should keep this parentheses.

Quote from CodingStyle document:

This does not apply if only one branch of a conditional statement is a
single
statement; in the latter case use braces in both branches:

if (condition) {
        do_this();
        do_that();
} else {
        otherwise();
}

> > +
> > +	list_del(&task->list_entry);
> > +	cmdq_thread_resume(thread);
> > +
> > +	/* call cb here to prevent lock */
> 
> I think these three statements should be place together and there should
> be no this comment.
> 
> 1. Change task state to DONE or ERROR.
> 2. Callback
> 3. Remove task from task_busy_list.

I will put these three statements together and wrap up them as a
function.

> > +	if (task->cb.cb) {
> > +		struct cmdq_cb_data cmdq_cb_data;
> > +
> > +		cmdq_cb_data.err = true;
> > +		cmdq_cb_data.data = task->cb.data;
> > +		task->cb.cb(cmdq_cb_data);
> > +	}
> > +
> > +	return -ECANCELED;
> > +}
...
> > +static int cmdq_suspend(struct device *dev)
> > +{
> > +	struct cmdq *cmdq = dev_get_drvdata(dev);
> > +	u32 exec_threads;
> > +
> > +	mutex_lock(&cmdq->task_mutex);
> > +	cmdq->suspended = true;
> > +	mutex_unlock(&cmdq->task_mutex);
> > +
> > +	exec_threads = readl(cmdq->base + CMDQ_CURR_LOADED_THR);
> > +	if ((exec_threads & CMDQ_THR_EXECUTING) && !cmdq_task_is_empty(cmdq)) {
> > +		dev_err(dev, "wait active tasks timeout.\n");
> > +		flush_workqueue(cmdq->task_release_wq);
> > +	}
> 
> Maybe it need not to check exec_threads, it may be written as:
> 
> 	if (!cmdq_task_is_empty(cmdq))
> 		dev_err(dev, "wait active tasks timeout.\n");
> 
> 	flush_workqueue(cmdq->task_release_wq);
> 
> Even though task_busy_list is empty, that does not mean every task in
> cmdq->task_release_wq is finished. So always do flush
> cmdq->task_release_wq.

Will do.

> > +	return 0;
> > +}
> > +
...
> 
> Regards,
> CK
> 

Thanks,
HS

Matthias Brugger May 30, 2016, 3:31 p.m. UTC | #3

On 30/05/16 05:19, HS Liao wrote:
> This patch is first version of Mediatek Command Queue(CMDQ) driver. The
> CMDQ is used to help read/write registers with critical time limitation,
> such as updating display configuration during the vblank. It controls
> Global Command Engine (GCE) hardware to achieve this requirement.
> Currently, CMDQ only supports display related hardwares, but we expect
> it can be extended to other hardwares for future requirements.
>
> Signed-off-by: HS Liao <[e-mail]>
> Signed-off-by: CK Hu <[e-mail]>
> ---
>   drivers/soc/mediatek/Kconfig    |  10 +
>   drivers/soc/mediatek/Makefile   |   1 +
>   drivers/soc/mediatek/mtk-cmdq.c | 943 ++++++++++++++++++++++++++++++++++++++++
>   include/soc/mediatek/cmdq.h     | 197 +++++++++
>   4 files changed, 1151 insertions(+)
>   create mode 100644 drivers/soc/mediatek/mtk-cmdq.c
>   create mode 100644 include/soc/mediatek/cmdq.h
>
> diff --git a/drivers/soc/mediatek/Kconfig b/drivers/soc/mediatek/Kconfig
> index 0a4ea80..c4ad75c 100644
> --- a/drivers/soc/mediatek/Kconfig
> +++ b/drivers/soc/mediatek/Kconfig
> @@ -1,6 +1,16 @@
>   #
>   # MediaTek SoC drivers
>   #
> +config MTK_CMDQ
> +	bool "MediaTek CMDQ Support"
> +	depends on ARCH_MEDIATEK || COMPILE_TEST
> +	select MTK_INFRACFG
> +	help
> +	  Say yes here to add support for the MediaTek Command Queue (CMDQ)
> +	  driver. The CMDQ is used to help read/write registers with critical
> +	  time limitation, such as updating display configuration during the
> +	  vblank.
> +
>   config MTK_INFRACFG
>   	bool "MediaTek INFRACFG Support"
>   	depends on ARCH_MEDIATEK || COMPILE_TEST
> diff --git a/drivers/soc/mediatek/Makefile b/drivers/soc/mediatek/Makefile
> index 12998b0..f7397ef 100644
> --- a/drivers/soc/mediatek/Makefile
> +++ b/drivers/soc/mediatek/Makefile
> @@ -1,3 +1,4 @@
> +obj-$(CONFIG_MTK_CMDQ) += mtk-cmdq.o
>   obj-$(CONFIG_MTK_INFRACFG) += mtk-infracfg.o
>   obj-$(CONFIG_MTK_PMIC_WRAP) += mtk-pmic-wrap.o
>   obj-$(CONFIG_MTK_SCPSYS) += mtk-scpsys.o
> diff --git a/drivers/soc/mediatek/mtk-cmdq.c b/drivers/soc/mediatek/mtk-cmdq.c
> new file mode 100644
> index 0000000..e9d6e1c
> --- /dev/null
> +++ b/drivers/soc/mediatek/mtk-cmdq.c
> @@ -0,0 +1,943 @@
> +/*
> + * Copyright (c) 2015 MediaTek Inc.
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License version 2 as
> + * published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> + * GNU General Public License for more details.
> + */
> +
> +#include <linux/clk.h>
> +#include <linux/clk-provider.h>
> +#include <linux/completion.h>
> +#include <linux/dma-mapping.h>
> +#include <linux/errno.h>
> +#include <linux/interrupt.h>
> +#include <linux/iopoll.h>
> +#include <linux/kernel.h>
> +#include <linux/kthread.h>
> +#include <linux/module.h>
> +#include <linux/mutex.h>
> +#include <linux/of_address.h>
> +#include <linux/of_irq.h>
> +#include <linux/platform_device.h>
> +#include <linux/slab.h>
> +#include <linux/spinlock.h>
> +#include <linux/suspend.h>
> +#include <linux/workqueue.h>
> +#include <soc/mediatek/cmdq.h>
> +
> +#define CMDQ_INITIAL_CMD_BLOCK_SIZE	PAGE_SIZE
> +#define CMDQ_INST_SIZE			8 /* instruction is 64-bit */
> +#define CMDQ_TIMEOUT_MS			1000
> +#define CMDQ_IRQ_MASK			0xffff
> +#define CMDQ_DRIVER_DEVICE_NAME		"mtk_cmdq"
> +#define CMDQ_CLK_NAME			"gce"

We can put the names in directly to un-bloat the defines.

> +
> +#define CMDQ_CURR_IRQ_STATUS		0x010
> +#define CMDQ_CURR_LOADED_THR		0x018
> +#define CMDQ_THR_SLOT_CYCLES		0x030
> +
> +#define CMDQ_THR_BASE			0x100
> +#define CMDQ_THR_SHIFT			0x080

Wouldn't be CMDQ_THR_SIZE more accurate?

> +#define CMDQ_THR_WARM_RESET		0x00
> +#define CMDQ_THR_ENABLE_TASK		0x04
> +#define CMDQ_THR_SUSPEND_TASK		0x08
> +#define CMDQ_THR_CURR_STATUS		0x0c
> +#define CMDQ_THR_IRQ_STATUS		0x10
> +#define CMDQ_THR_IRQ_ENABLE		0x14
> +#define CMDQ_THR_CURR_ADDR		0x20
> +#define CMDQ_THR_END_ADDR		0x24
> +#define CMDQ_THR_CFG			0x40
> +
> +#define CMDQ_THR_ENABLED		0x1
> +#define CMDQ_THR_DISABLED		0x0
> +#define CMDQ_THR_SUSPEND		0x1
> +#define CMDQ_THR_RESUME			0x0
> +#define CMDQ_THR_STATUS_SUSPENDED	BIT(1)
> +#define CMDQ_THR_DO_WARM_RESET		BIT(0)
> +#define CMDQ_THR_ACTIVE_SLOT_CYCLES	0x3200
> +#define CMDQ_THR_PRIORITY		3
> +#define CMDQ_THR_IRQ_DONE		0x1
> +#define CMDQ_THR_IRQ_ERROR		0x12
> +#define CMDQ_THR_IRQ_EN			0x13 /* done + error */

#define CMDQ_THR_IRQ_EN (CMDQ_THR_IRQ_ERROR | CMDQ_THR_IRQ_DONE)

> +#define CMDQ_THR_IRQ_MASK		0x13

never used.

> +#define CMDQ_THR_EXECUTING		BIT(31)
> +
> +#define CMDQ_ARG_A_WRITE_MASK		0xffff
> +#define CMDQ_SUBSYS_MASK		0x1f
> +#define CMDQ_OP_CODE_MASK		0xff000000
> +
> +#define CMDQ_OP_CODE_SHIFT		24

Couldn't we connect the mask with the shift, or aren't they related?

#define CMDQ_OP_CODE_MASK (0xff << CMDQ_OP_CODE_SHIFT)

> +#define CMDQ_SUBSYS_SHIFT		16
> +
> +#define CMDQ_WRITE_ENABLE_MASK		BIT(0)
> +#define CMDQ_JUMP_BY_OFFSET		0x10000000
> +#define CMDQ_JUMP_BY_PA			0x10000001
> +#define CMDQ_JUMP_PASS			CMDQ_INST_SIZE
> +#define CMDQ_WFE_UPDATE			BIT(31)
> +#define CMDQ_WFE_WAIT			BIT(15)
> +#define CMDQ_WFE_WAIT_VALUE		0x1
> +#define CMDQ_EOC_IRQ_EN			BIT(0)
> +
> +enum cmdq_thread_index {
> +	CMDQ_THR_DISP_MAIN_IDX,	/* main */
> +	CMDQ_THR_DISP_SUB_IDX,	/* sub */
> +	CMDQ_THR_DISP_MISC_IDX,	/* misc */
> +	CMDQ_THR_MAX_COUNT,	/* max */
> +};
> +
> +/*
> + * CMDQ_CODE_MOVE:
> + *   move value into internal register as mask
> + *   format: op mask
> + * CMDQ_CODE_WRITE:
> + *   write value into target register
> + *   format: op subsys address value
> + * CMDQ_CODE_JUMP:
> + *   jump by offset
> + *   format: op offset
> + * CMDQ_CODE_WFE:
> + *   wait for event and clear
> + *   it is just clear if no wait
> + *   format: [wait]  op event update:1 to_wait:1 wait:1
> + *           [clear] op event update:1 to_wait:0 wait:0
> + * CMDQ_CODE_EOC:
> + *   end of command
> + *   format: op irq_flag
> + */

I think we need more documentation of how this command queue engine is 
working. If not, I think it will be really complicated to understand how 
to use this.

> +enum cmdq_code {
> +	CMDQ_CODE_MOVE = 0x02,
> +	CMDQ_CODE_WRITE = 0x04,
> +	CMDQ_CODE_JUMP = 0x10,
> +	CMDQ_CODE_WFE = 0x20,
> +	CMDQ_CODE_EOC = 0x40,
> +};
> +
> +enum cmdq_task_state {
> +	TASK_STATE_BUSY, /* running on a GCE thread */
> +	TASK_STATE_ERROR,
> +	TASK_STATE_DONE,
> +};
> +
> +struct cmdq_task_cb {
> +	cmdq_async_flush_cb	cb;
> +	void			*data;
> +};
> +
> +struct cmdq_thread {
> +	void __iomem		*base;
> +	struct list_head	task_busy_list;
> +	wait_queue_head_t	wait_task_done;
> +};
> +
> +struct cmdq_task {
> +	struct cmdq		*cmdq;
> +	struct list_head	list_entry;
> +	enum cmdq_task_state	task_state;
> +	void			*va_base;
> +	dma_addr_t		pa_base;
> +	u64			engine_flag;
> +	size_t			command_size;
> +	u32			num_cmd;

num_cmd is directly connected to command_size. I prefer to just keep 
num_cmd and calculate the command_size where necessary.

> +	struct cmdq_thread	*thread;
> +	struct cmdq_task_cb	cb;
> +	struct work_struct	release_work;
> +};
> +
> +struct cmdq {
> +	struct device		*dev;
> +	void __iomem		*base;
> +	u32			irq;
> +	struct workqueue_struct	*task_release_wq;
> +	struct cmdq_thread	thread[CMDQ_THR_MAX_COUNT];
> +	struct mutex		task_mutex;	/* for task */
> +	spinlock_t		exec_lock;	/* for exec */
> +	struct clk		*clock;
> +	bool			suspended;
> +};
> +
> +struct cmdq_subsys {
> +	u32	base;
> +	int	id;
> +};
> +
> +static const struct cmdq_subsys gce_subsys[] = {
> +	{0x1400, 1},
> +	{0x1401, 2},
> +	{0x1402, 3},
> +};
> +
> +static int cmdq_subsys_base_to_id(u32 base)
> +{
> +	int i;
> +
> +	for (i = 0; i < ARRAY_SIZE(gce_subsys); i++)
> +		if (gce_subsys[i].base == base)
> +			return gce_subsys[i].id;
> +	return -EFAULT;
> +}
> +
> +static int cmdq_eng_get_thread(u64 flag)
> +{
> +	if (flag & BIT_ULL(CMDQ_ENG_DISP_DSI0))
> +		return CMDQ_THR_DISP_MAIN_IDX;
> +	else if (flag & BIT_ULL(CMDQ_ENG_DISP_DPI0))
> +		return CMDQ_THR_DISP_SUB_IDX;
> +	else
> +		return CMDQ_THR_DISP_MISC_IDX;
> +}
> +
> +static void cmdq_task_release(struct cmdq_task *task)
> +{
> +	struct cmdq *cmdq = task->cmdq;
> +
> +	dma_free_coherent(cmdq->dev, task->command_size, task->va_base,
> +			  task->pa_base);
> +	kfree(task);
> +}
> +
> +static struct cmdq_task *cmdq_task_acquire(struct cmdq_rec *rec,
> +					   struct cmdq_task_cb cb)
> +{
> +	struct cmdq *cmdq = rec->cmdq;
> +	struct device *dev = cmdq->dev;
> +	struct cmdq_task *task;
> +
> +	task = kzalloc(sizeof(*task), GFP_KERNEL);
> +	INIT_LIST_HEAD(&task->list_entry);
> +	task->va_base = dma_alloc_coherent(dev, rec->command_size,
> +					   &task->pa_base, GFP_KERNEL);
> +	if (!task->va_base) {
> +		dev_err(dev, "allocate command buffer failed\n");
> +		kfree(task);
> +		return NULL;
> +	}
> +
> +	task->cmdq = cmdq;
> +	task->command_size = rec->command_size;
> +	task->engine_flag = rec->engine_flag;
> +	task->task_state = TASK_STATE_BUSY;
> +	task->cb = cb;
> +	memcpy(task->va_base, rec->buf, rec->command_size);
> +	task->num_cmd = task->command_size / CMDQ_INST_SIZE;
> +	return task;
> +}
> +
> +static void cmdq_thread_writel(struct cmdq_thread *thread, u32 value,
> +			       u32 offset)
> +{
> +	writel(value, thread->base + offset);
> +}
> +
> +static u32 cmdq_thread_readl(struct cmdq_thread *thread, u32 offset)
> +{
> +	return readl(thread->base + offset);
> +}

We can get rid of cmdq_thread_readl/writel.

> +
> +static int cmdq_thread_suspend(struct cmdq *cmdq, struct cmdq_thread *thread)
> +{
> +	u32 status;
> +
> +	cmdq_thread_writel(thread, CMDQ_THR_SUSPEND, CMDQ_THR_SUSPEND_TASK);
> +
> +	/* If already disabled, treat as suspended successful. */
> +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> +	      CMDQ_THR_ENABLED))
> +		return 0;
> +
> +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_CURR_STATUS,
> +			status, status & CMDQ_THR_STATUS_SUSPENDED, 0, 10)) {
> +		dev_err(cmdq->dev, "suspend GCE thread 0x%x failed\n",
> +			(u32)(thread->base - cmdq->base));
> +		return -EFAULT;
> +	}
> +
> +	return 0;
> +}
> +
> +static void cmdq_thread_resume(struct cmdq_thread *thread)
> +{
> +	cmdq_thread_writel(thread, CMDQ_THR_RESUME, CMDQ_THR_SUSPEND_TASK);
> +}
> +
> +static int cmdq_thread_reset(struct cmdq *cmdq, struct cmdq_thread *thread)
> +{
> +	u32 warm_reset;
> +
> +	cmdq_thread_writel(thread, CMDQ_THR_DO_WARM_RESET, CMDQ_THR_WARM_RESET);
> +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_WARM_RESET,
> +			warm_reset, !(warm_reset & CMDQ_THR_DO_WARM_RESET),
> +			0, 10)) {
> +		dev_err(cmdq->dev, "reset GCE thread 0x%x failed\n",
> +			(u32)(thread->base - cmdq->base));
> +		return -EFAULT;
> +	}
> +	writel(CMDQ_THR_ACTIVE_SLOT_CYCLES, cmdq->base + CMDQ_THR_SLOT_CYCLES);
> +	return 0;
> +}
> +
> +static void cmdq_thread_disable(struct cmdq *cmdq, struct cmdq_thread *thread)
> +{
> +	cmdq_thread_reset(cmdq, thread);
> +	cmdq_thread_writel(thread, CMDQ_THR_DISABLED, CMDQ_THR_ENABLE_TASK);
> +}
> +
> +/* notify GCE to re-fetch commands by setting GCE thread PC */
> +static void cmdq_thread_invalidate_fetched_data(struct cmdq_thread *thread)
> +{
> +	cmdq_thread_writel(thread,
> +			   cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR),
> +			   CMDQ_THR_CURR_ADDR);
> +}
> +
> +static void cmdq_task_insert_into_thread(struct cmdq_task *task)
> +{
> +	struct cmdq_thread *thread = task->thread;
> +	struct cmdq_task *prev_task = list_last_entry(
> +			&thread->task_busy_list, typeof(*task), list_entry);
> +	u64 *prev_task_base = prev_task->va_base;
> +
> +	/* let previous task jump to this task */
> +	prev_task_base[prev_task->num_cmd - 1] = (u64)CMDQ_JUMP_BY_PA << 32 |
> +						 task->pa_base;
> +
> +	cmdq_thread_invalidate_fetched_data(thread);
> +}
> +
> +/* we assume tasks in the same display GCE thread are waiting the same event. */
> +static void cmdq_task_remove_wfe(struct cmdq_task *task)
> +{
> +	u32 wfe_option = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
> +	u32 wfe_op = CMDQ_CODE_WFE << CMDQ_OP_CODE_SHIFT;
> +	u32 *base = task->va_base;
> +	u32 num_cmd = task->num_cmd << 1;
> +	int i;
> +
> +	for (i = 0; i < num_cmd; i += 2)
> +		if (base[i] == wfe_option &&
> +		    (base[i + 1] & CMDQ_OP_CODE_MASK) == wfe_op) {
> +			base[i] = CMDQ_JUMP_PASS;
> +			base[i + 1] = CMDQ_JUMP_BY_OFFSET;
> +		}

After using the command buffer as a void pointer a u64 pointer, we now 
reference to it as u32. I would prefer to explain here, how the command 
looks like we are searching for and use a for loop passing task->num_cmd 
instead.

> +}
> +
> +static void cmdq_task_exec(struct cmdq_task *task, struct cmdq_thread *thread)
> +{
> +	struct cmdq *cmdq = task->cmdq;
> +	unsigned long flags;
> +	unsigned long curr_pa, end_pa;
> +
> +	WARN_ON(clk_prepare_enable(cmdq->clock) < 0);
> +	spin_lock_irqsave(&cmdq->exec_lock, flags);

cmdq_task_exec is called with cmdq->task_mutex held, so why do we need 
the spin_lock here? Can't we just use one of the two?

> +	task->thread = thread;
> +	task->task_state = TASK_STATE_BUSY;

That was already set in cmdq_task_acquire, why do we need to set it here 
again?

> +	if (list_empty(&thread->task_busy_list)) {
> +		WARN_ON(cmdq_thread_reset(cmdq, thread) < 0);
> +
> +		cmdq_thread_writel(thread, task->pa_base, CMDQ_THR_CURR_ADDR);
> +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
> +				   CMDQ_THR_END_ADDR);
> +		cmdq_thread_writel(thread, CMDQ_THR_PRIORITY, CMDQ_THR_CFG);
> +		cmdq_thread_writel(thread, CMDQ_THR_IRQ_EN,
> +				   CMDQ_THR_IRQ_ENABLE);
> +
> +		cmdq_thread_writel(thread, CMDQ_THR_ENABLED,
> +				   CMDQ_THR_ENABLE_TASK);
> +	} else {
> +		WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
> +
> +		/*
> +		 * check boundary condition
> +		 * PC = END - 8, EOC is executed
> +		 * PC = END, all CMDs are executed
> +		 */
> +		curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> +		end_pa = cmdq_thread_readl(thread, CMDQ_THR_END_ADDR);
> +		if (curr_pa == end_pa - 8 || curr_pa == end_pa) {

8 refers to CMDQ_INST_SIZE, right?

> +			/* set to this task directly */
> +			cmdq_thread_writel(thread, task->pa_base,
> +					   CMDQ_THR_CURR_ADDR);
> +		} else {
> +			cmdq_task_insert_into_thread(task);
> +
> +			if (thread == &cmdq->thread[CMDQ_THR_DISP_MAIN_IDX] ||
> +			    thread == &cmdq->thread[CMDQ_THR_DISP_SUB_IDX])
> +				cmdq_task_remove_wfe(task);

We could do this check using the task->engine_flag, I suppose that's 
easier to undestand then.

> +
> +			smp_mb(); /* modify jump before enable thread */
> +		}
> +
> +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
> +				   CMDQ_THR_END_ADDR);
> +		cmdq_thread_resume(thread);
> +	}
> +	list_move_tail(&task->list_entry, &thread->task_busy_list);
> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> +}
> +
> +static void cmdq_handle_error_done(struct cmdq *cmdq,
> +				   struct cmdq_thread *thread, u32 irq_flag)
> +{
> +	struct cmdq_task *task, *tmp, *curr_task = NULL;
> +	u32 curr_pa;
> +	struct cmdq_cb_data cmdq_cb_data;
> +	bool err;
> +
> +	if (irq_flag & CMDQ_THR_IRQ_ERROR)
> +		err = true;
> +	else if (irq_flag & CMDQ_THR_IRQ_DONE)
> +		err = false;
> +	else
> +		return;
> +
> +	curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> +
> +	list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
> +				 list_entry) {
> +		if (curr_pa >= task->pa_base &&
> +		    curr_pa < (task->pa_base + task->command_size))

What are you checking here? It seems as if you make some implcit 
assumptions about pa_base and the order of execution of commands in the 
thread. Is it save to do so? Does dma_alloc_coherent give any guarantees 
about dma_handle?

> +			curr_task = task;
> +		if (task->cb.cb) {
> +			cmdq_cb_data.err = curr_task ? err : false;
> +			cmdq_cb_data.data = task->cb.data;
> +			task->cb.cb(cmdq_cb_data);
> +		}
> +		task->task_state = (curr_task && err) ? TASK_STATE_ERROR :
> +				TASK_STATE_DONE;
> +		list_del(&task->list_entry);
> +		if (curr_task)
> +			break;
> +	}
> +
> +	wake_up(&thread->wait_task_done);
> +}
> +
> +static void cmdq_thread_irq_handler(struct cmdq *cmdq, int tid)
> +{
> +	struct cmdq_thread *thread = &cmdq->thread[tid];
> +	unsigned long flags = 0L;
> +	u32 irq_flag;
> +
> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
> +
> +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
> +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
> +
> +	/*
> +	 * Another CPU core could run "release task" right before we acquire
> +	 * the spin lock, and thus reset / disable this GCE thread, so we
> +	 * need to check the enable bit of this GCE thread.
> +	 */
> +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> +	      CMDQ_THR_ENABLED))
> +		irq_flag = 0;

cmdq_handle_error_done just retuns in this case. Programming this way 
just makes things confusing. What about:

if (cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
	      CMDQ_THR_ENABLED)
	cmdq_handle_error_done(cmdq, thread, irq_flag);
else
		irq_flag = 0;

spin_unlock_irqrestore(&cmdq->exec_lock, flags);

> +
> +	cmdq_handle_error_done(cmdq, thread, irq_flag);
> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> +}
> +
> +static irqreturn_t cmdq_irq_handler(int irq, void *dev)
> +{
> +	struct cmdq *cmdq = dev;
> +	u32 irq_status;
> +	int i;
> +
> +	irq_status = readl(cmdq->base + CMDQ_CURR_IRQ_STATUS);
> +	irq_status &= CMDQ_IRQ_MASK;
> +	irq_status ^= CMDQ_IRQ_MASK;

irq_status can be much bigger then 3, which is the number of threads in 
the system (CMDQ_THR_MAX_COUNT). So why we use this mask here isn't 
clear to me.

> +
> +	if (!irq_status)
> +		return IRQ_NONE;
> +
> +	while (irq_status) {
> +		i = ffs(irq_status) - 1;
> +		irq_status &= ~BIT(i);
> +		cmdq_thread_irq_handler(cmdq, i);
> +	}

Can you explain how the irq status register looks like, that would it 
make much easier to understand what happens here.

> +
> +	return IRQ_HANDLED;	
> +}
> +
> +static int cmdq_task_handle_error_result(struct cmdq_task *task)

We never check the return values, why do we have them?

> +{
> +	struct cmdq *cmdq = task->cmdq;
> +	struct device *dev = cmdq->dev;
> +	struct cmdq_thread *thread = task->thread;
> +	struct cmdq_task *next_task, *prev_task;
> +	u32 irq_flag;
> +
> +	/* suspend GCE thread to ensure consistency */
> +	WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
> +
> +	/* ISR has handled this error task */
> +	if (task->task_state == TASK_STATE_ERROR) {
> +		next_task = list_first_entry_or_null(&thread->task_busy_list,
> +				struct cmdq_task, list_entry);
> +		if (next_task) /* move to next task */
> +			cmdq_thread_writel(thread, next_task->pa_base,
> +					   CMDQ_THR_CURR_ADDR);

We have to do this, as we suppose that the thread did not reach the jump 
instruction we put into it's command queue, right?

> +		cmdq_thread_resume(thread);
> +		return -ECANCELED;
> +	}
> +

if task_state != ERROR and != DONE. This means that the timeout of 
task_release_wq has timed out, right?

> +	/*
> +	 * Save next_task and prev_task in advance
> +	 * since cmdq_handle_error_done will remove list_entry.
> +	 */
> +	next_task = prev_task = NULL;
> +	if (task->list_entry.next != &thread->task_busy_list)
> +		next_task = list_next_entry(task, list_entry);
> +	if (task->list_entry.prev != &thread->task_busy_list)
> +		prev_task = list_prev_entry(task, list_entry);
> +
> +	/*
> +	 * Although IRQ is disabled, GCE continues to execute.
> +	 * It may have pending IRQ before GCE thread is suspended,
> +	 * so check this condition again.
> +	 */

The first thing we did in this function was suspending the thread. Why 
do we need this then?
To be honest this whole functions looks really like a design error. We 
have to sperate the states much clearer so that it is possible to 
understand what is happening in the GCE. Isn't it for example posible to 
have worker queues for timed out tasks and tasks with an error? I'm not 
sure how to do this, actually I'm not sure if I really understood how 
this is supposed to work.

How much do we win, when we patch the thread command queue for every 
task we add, instead of just taking one task after another from the 
task_busy_list?

> +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
> +	cmdq_handle_error_done(cmdq, thread, irq_flag);
> +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
> +
> +	if (task->task_state == TASK_STATE_DONE) {
> +		cmdq_thread_resume(thread);
> +		return 0;
> +	}
> +
> +	if (task->task_state == TASK_STATE_ERROR) {
> +		dev_err(dev, "task 0x%p error\n", task);
> +		if (next_task) /* move to next task */
> +			cmdq_thread_writel(thread, next_task->pa_base,
> +					   CMDQ_THR_CURR_ADDR);
> +		cmdq_thread_resume(thread);
> +		return -ECANCELED;
> +	}
> +
> +	/* Task is running, so we force to remove it. */
> +	dev_err(dev, "task 0x%p timeout or killed\n", task);
> +	task->task_state = TASK_STATE_ERROR;
> +
> +	if (prev_task) {
> +		u64 *prev_va = prev_task->va_base;
> +		u64 *curr_va = task->va_base;
> +
> +		/* copy JUMP instruction */
> +		prev_va[prev_task->num_cmd - 1] = curr_va[task->num_cmd - 1];
> +
> +		cmdq_thread_invalidate_fetched_data(thread);
> +	} else if (next_task) { /* move to next task */
> +		cmdq_thread_writel(thread, next_task->pa_base,
> +				   CMDQ_THR_CURR_ADDR);
> +	}
> +
> +	list_del(&task->list_entry);
> +	cmdq_thread_resume(thread);
> +
> +	/* call cb here to prevent lock */
> +	if (task->cb.cb) {
> +		struct cmdq_cb_data cmdq_cb_data;
> +
> +		cmdq_cb_data.err = true;
> +		cmdq_cb_data.data = task->cb.data;
> +		task->cb.cb(cmdq_cb_data);
> +	}
> +
> +	return -ECANCELED;
> +}
> +
> +static void cmdq_task_wait_release_work(struct work_struct *work_item)
> +{
> +	struct cmdq_task *task = container_of(work_item, struct cmdq_task,
> +					      release_work);
> +	struct cmdq *cmdq = task->cmdq;
> +	struct cmdq_thread *thread = task->thread;
> +	unsigned long flags;
> +
> +	wait_event_timeout(thread->wait_task_done,
> +			   task->task_state != TASK_STATE_BUSY,
> +			   msecs_to_jiffies(CMDQ_TIMEOUT_MS));
> +
> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
> +	if (task->task_state != TASK_STATE_DONE)
> +		cmdq_task_handle_error_result(task);
> +	if (list_empty(&thread->task_busy_list))
> +		cmdq_thread_disable(cmdq, thread);
> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> +
> +	/* release regardless of success or not */
> +	clk_disable_unprepare(cmdq->clock);
> +	cmdq_task_release(task);
> +}
> +
> +static void cmdq_task_wait_release_schedule(struct cmdq_task *task)
> +{
> +	struct cmdq *cmdq = task->cmdq;
> +
> +	INIT_WORK(&task->release_work, cmdq_task_wait_release_work);
> +	queue_work(cmdq->task_release_wq, &task->release_work);
> +}
> +
> +static int cmdq_rec_realloc_cmd_buffer(struct cmdq_rec *rec, size_t size)
> +{
> +	void *new_buf;
> +
> +	new_buf = krealloc(rec->buf, size, GFP_KERNEL | __GFP_ZERO);
> +	if (!new_buf)
> +		return -ENOMEM;
> +	rec->buf = new_buf;
> +	rec->buf_size = size;
> +	return 0;
> +}
> +
> +struct cmdq_base *cmdq_register_device(struct device *dev)
> +{
> +	struct cmdq_base *cmdq_base;
> +	struct resource res;
> +	int subsys;
> +	u32 base;
> +
> +	if (of_address_to_resource(dev->of_node, 0, &res))
> +		return NULL;
> +	base = (u32)res.start;
> +
> +	subsys = cmdq_subsys_base_to_id(base >> 16);
> +	if (subsys < 0)
> +		return NULL;
> +
> +	cmdq_base = devm_kmalloc(dev, sizeof(*cmdq_base), GFP_KERNEL);
> +	if (!cmdq_base)
> +		return NULL;
> +	cmdq_base->subsys = subsys;
> +	cmdq_base->base = base;
> +
> +	return cmdq_base;
> +}
> +EXPORT_SYMBOL(cmdq_register_device);
> +
> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
> +		    struct cmdq_rec **rec_ptr)
> +{
> +	struct cmdq_rec *rec;
> +	int err;
> +
> +	rec = kzalloc(sizeof(*rec), GFP_KERNEL);
> +	if (!rec)
> +		return -ENOMEM;
> +	rec->cmdq = dev_get_drvdata(dev);
> +	rec->engine_flag = engine_flag;
> +	err = cmdq_rec_realloc_cmd_buffer(rec, CMDQ_INITIAL_CMD_BLOCK_SIZE);

Just pass PAGE_SIZE here, no need for CMDQ_INITIAL_CMD_BLOCK_SIZE.

> +	if (err < 0) {
> +		kfree(rec);
> +		return err;
> +	}
> +	*rec_ptr = rec;
> +	return 0;
> +}
> +EXPORT_SYMBOL(cmdq_rec_create);
> +
> +static int cmdq_rec_append_command(struct cmdq_rec *rec, enum cmdq_code code,
> +				   u32 arg_a, u32 arg_b)
> +{
> +	u64 *cmd_ptr;
> +	int err;
> +
> +	if (WARN_ON(rec->finalized))
> +		return -EBUSY;
> +	if (code < CMDQ_CODE_MOVE || code > CMDQ_CODE_EOC)
> +		return -EINVAL;

cmdq_rec_append_command is just called from inside this driver and code 
is a enum. We can expect it to be correct, no need for this check.

> +	if (unlikely(rec->command_size + CMDQ_INST_SIZE > rec->buf_size)) {

command_size is the offset into the buffer to which a new command is 
written, so this name is highly confusing. I wonder if this would be 
easier to understand if we redefine command_size to something like the 
number of commands and divide/multiply CMDQ_INST_SIZE where this is needed.

> +		err = cmdq_rec_realloc_cmd_buffer(rec, rec->buf_size * 2);
> +		if (err < 0)
> +			return err;
> +	}
> +	cmd_ptr = rec->buf + rec->command_size;
> +	(*cmd_ptr) = (u64)((code << CMDQ_OP_CODE_SHIFT) | arg_a) << 32 | arg_b;
> +	rec->command_size += CMDQ_INST_SIZE;
> +	return 0;
> +}
> +
> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value, struct cmdq_base *base,
> +		   u32 offset)
> +{
> +	u32 arg_a = ((base->base + offset) & CMDQ_ARG_A_WRITE_MASK) |
> +		    ((base->subsys & CMDQ_SUBSYS_MASK) << CMDQ_SUBSYS_SHIFT);

base->subsys is the id from gce_sybsys, so we can expect it to be 
correct, no need to mask with CMDQ_SUBSYS_MASK.

> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WRITE, arg_a, value);
> +}
> +EXPORT_SYMBOL(cmdq_rec_write);
> +
> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
> +			struct cmdq_base *base, u32 offset, u32 mask)
> +{
> +	u32 offset_mask = offset;
> +	int err;
> +
> +	if (mask != 0xffffffff) {
> +		err = cmdq_rec_append_command(rec, CMDQ_CODE_MOVE, 0, ~mask);
> +		if (err < 0)
> +			return err;
> +		offset_mask |= CMDQ_WRITE_ENABLE_MASK;
> +	}
> +	return cmdq_rec_write(rec, value, base, offset_mask);
> +}
> +EXPORT_SYMBOL(cmdq_rec_write_mask);
> +
> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event)
> +{
> +	u32 arg_b;
> +
> +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
> +		return -EINVAL;
> +
> +	/*
> +	 * bit 0-11: wait value
> +	 * bit 15: 1 - wait, 0 - no wait
> +	 * bit 16-27: update value
> +	 * bit 31: 1 - update, 0 - no update
> +	 */

I don't understand this comments. What are they for?

> +	arg_b = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event, arg_b);
> +}
> +EXPORT_SYMBOL(cmdq_rec_wfe);
> +
> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event)
> +{
> +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
> +		return -EINVAL;
> +
> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event,
> +				       CMDQ_WFE_UPDATE);
> +}
> +EXPORT_SYMBOL(cmdq_rec_clear_event);
> +
> +static int cmdq_rec_finalize(struct cmdq_rec *rec)
> +{
> +	int err;
> +
> +	if (rec->finalized)
> +		return 0;
> +
> +	/* insert EOC and generate IRQ for each command iteration */
> +	err = cmdq_rec_append_command(rec, CMDQ_CODE_EOC, 0, CMDQ_EOC_IRQ_EN);
> +	if (err < 0)
> +		return err;
> +
> +	/* JUMP to end */
> +	err = cmdq_rec_append_command(rec, CMDQ_CODE_JUMP, 0, CMDQ_JUMP_PASS);
> +	if (err < 0)
> +		return err;
> +

Does this need to be atomic?
What happens if after CODE_EOC and before CODE_JUMP some 
write/read/event gets added?
What happens if more commands get added to the queue after CODE_JUMP, 
but before finalized is set to true. Why don't you use atomic functions 
to access finalized?

> +	rec->finalized = true;
> +	return 0;
> +}
> +
> +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
> +			 void *data)
> +{
> +	struct cmdq *cmdq = rec->cmdq;
> +	struct cmdq_task *task;
> +	struct cmdq_task_cb task_cb;
> +	struct cmdq_thread *thread;
> +	int err;
> +
> +	mutex_lock(&cmdq->task_mutex);
> +	if (cmdq->suspended) {
> +		dev_err(cmdq->dev, "%s is called after suspended\n", __func__);
> +		mutex_unlock(&cmdq->task_mutex);
> +		return -EPERM;
> +	}
> +
> +	err = cmdq_rec_finalize(rec);
> +	if (err < 0) {
> +		mutex_unlock(&cmdq->task_mutex);
> +		return err;
> +	}
> +
> +	task_cb.cb = cb;
> +	task_cb.data = data;
> +	task = cmdq_task_acquire(rec, task_cb);
> +	if (!task) {
> +		mutex_unlock(&cmdq->task_mutex);
> +		return -EFAULT;
> +	}
> +
> +	thread = &cmdq->thread[cmdq_eng_get_thread(task->engine_flag)];
> +	cmdq_task_exec(task, thread);
> +	cmdq_task_wait_release_schedule(task);
> +	mutex_unlock(&cmdq->task_mutex);
> +	return 0;
> +}
> +EXPORT_SYMBOL(cmdq_rec_flush_async);
> +
> +struct cmdq_flush_completion {
> +	struct completion cmplt;
> +	bool err;
> +};
> +
> +static int cmdq_rec_flush_cb(struct cmdq_cb_data data)
> +{
> +	struct cmdq_flush_completion *cmplt = data.data;
> +
> +	cmplt->err = data.err;
> +	complete(&cmplt->cmplt);
> +	return 0;
> +}
> +
> +int cmdq_rec_flush(struct cmdq_rec *rec)
> +{
> +	struct cmdq_flush_completion cmplt;
> +	int err;
> +
> +	init_completion(&cmplt.cmplt);
> +	err = cmdq_rec_flush_async(rec, cmdq_rec_flush_cb, &cmplt);
> +	if (err < 0)
> +		return err;
> +	wait_for_completion(&cmplt.cmplt);
> +	return cmplt.err ? -EFAULT : 0;
> +}
> +EXPORT_SYMBOL(cmdq_rec_flush);
> +
> +void cmdq_rec_destroy(struct cmdq_rec *rec)
> +{
> +	kfree(rec->buf);
> +	kfree(rec);
> +}
> +EXPORT_SYMBOL(cmdq_rec_destroy);
> +
> +static bool cmdq_task_is_empty(struct cmdq *cmdq)
> +{
> +	struct cmdq_thread *thread;
> +	int i;
> +
> +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
> +		thread = &cmdq->thread[i];
> +		if (!list_empty(&thread->task_busy_list))
> +			return false;
> +	}
> +	return true;
> +}
> +
> +static int cmdq_suspend(struct device *dev)
> +{
> +	struct cmdq *cmdq = dev_get_drvdata(dev);
> +	u32 exec_threads;
> +
> +	mutex_lock(&cmdq->task_mutex);
> +	cmdq->suspended = true;
> +	mutex_unlock(&cmdq->task_mutex);
> +
> +	exec_threads = readl(cmdq->base + CMDQ_CURR_LOADED_THR);
> +	if ((exec_threads & CMDQ_THR_EXECUTING) && !cmdq_task_is_empty(cmdq)) {
> +		dev_err(dev, "wait active tasks timeout.\n");
> +		flush_workqueue(cmdq->task_release_wq);
> +	}
> +	return 0;
> +}
> +
> +static int cmdq_resume(struct device *dev)
> +{
> +	struct cmdq *cmdq = dev_get_drvdata(dev);
> +
> +	cmdq->suspended = false;
> +	return 0;
> +}
> +
> +static int cmdq_remove(struct platform_device *pdev)
> +{
> +	struct cmdq *cmdq = platform_get_drvdata(pdev);
> +
> +	destroy_workqueue(cmdq->task_release_wq);
> +	cmdq->task_release_wq = NULL;
> +	return 0;
> +}
> +
> +static int cmdq_probe(struct platform_device *pdev)
> +{
> +	struct device *dev = &pdev->dev;
> +	struct device_node *node = dev->of_node;
> +	struct resource *res;
> +	struct cmdq *cmdq;
> +	int err, i;
> +
> +	cmdq = devm_kzalloc(dev, sizeof(*cmdq), GFP_KERNEL);
> +	if (!cmdq)
> +		return -ENOMEM;
> +	cmdq->dev = dev;
> +
> +	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> +	cmdq->base = devm_ioremap_resource(dev, res);
> +	if (IS_ERR(cmdq->base)) {
> +		dev_err(dev, "failed to ioremap gce\n");
> +		return PTR_ERR(cmdq->base);
> +	}
> +
> +	cmdq->irq = irq_of_parse_and_map(node, 0);
> +	if (!cmdq->irq) {
> +		dev_err(dev, "failed to get irq\n");
> +		return -EINVAL;
> +	}
> +
> +	dev_dbg(dev, "cmdq device: addr:0x%p, va:0x%p, irq:%d\n",
> +		dev, cmdq->base, cmdq->irq);
> +
> +	mutex_init(&cmdq->task_mutex);
> +	spin_lock_init(&cmdq->exec_lock);
> +	cmdq->task_release_wq = alloc_ordered_workqueue(
> +			"%s", WQ_MEM_RECLAIM | WQ_HIGHPRI,
> +			"cmdq_task_wait_release");
> +
> +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
> +		cmdq->thread[i].base = cmdq->base + CMDQ_THR_BASE +
> +				CMDQ_THR_SHIFT * i;
> +		init_waitqueue_head(&cmdq->thread[i].wait_task_done);
> +		INIT_LIST_HEAD(&cmdq->thread[i].task_busy_list);
> +	}
> +
> +	platform_set_drvdata(pdev, cmdq);
> +
> +	err = devm_request_irq(dev, cmdq->irq, cmdq_irq_handler, IRQF_SHARED,
> +			       CMDQ_DRIVER_DEVICE_NAME, cmdq);
> +	if (err < 0) {
> +		dev_err(dev, "failed to register ISR (%d)\n", err);
> +		goto fail;
> +	}
> +
> +	cmdq->clock = devm_clk_get(dev, CMDQ_CLK_NAME);
> +	if (IS_ERR(cmdq->clock)) {
> +		dev_err(dev, "failed to get clk:%s\n", CMDQ_CLK_NAME);
> +		err = PTR_ERR(cmdq->clock);
> +		goto fail;
> +	}
> +	return 0;
> +
> +fail:
> +	cmdq_remove(pdev);
> +	return err;
> +}
> +
> +static const struct dev_pm_ops cmdq_pm_ops = {
> +	.suspend = cmdq_suspend,
> +	.resume = cmdq_resume,
> +};
> +
> +static const struct of_device_id cmdq_of_ids[] = {
> +	{.compatible = "mediatek,mt8173-gce",},
> +	{}
> +};
> +
> +static struct platform_driver cmdq_drv = {
> +	.probe = cmdq_probe,
> +	.remove = cmdq_remove,
> +	.driver = {
> +		.name = CMDQ_DRIVER_DEVICE_NAME,
> +		.owner = THIS_MODULE,
> +		.pm = &cmdq_pm_ops,
> +		.of_match_table = cmdq_of_ids,
> +	}
> +};
> +
> +builtin_platform_driver(cmdq_drv);
> diff --git a/include/soc/mediatek/cmdq.h b/include/soc/mediatek/cmdq.h
> new file mode 100644
> index 0000000..60eef3d
> --- /dev/null
> +++ b/include/soc/mediatek/cmdq.h
> @@ -0,0 +1,197 @@
> +/*
> + * Copyright (c) 2015 MediaTek Inc.
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License version 2 as
> + * published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> + * GNU General Public License for more details.
> + */
> +
> +#ifndef __MTK_CMDQ_H__
> +#define __MTK_CMDQ_H__
> +
> +#include <linux/platform_device.h>
> +#include <linux/types.h>
> +
> +enum cmdq_eng {
> +	CMDQ_ENG_DISP_AAL,
> +	CMDQ_ENG_DISP_COLOR0,
> +	CMDQ_ENG_DISP_COLOR1,
> +	CMDQ_ENG_DISP_DPI0,
> +	CMDQ_ENG_DISP_DSI0,
> +	CMDQ_ENG_DISP_DSI1,
> +	CMDQ_ENG_DISP_GAMMA,
> +	CMDQ_ENG_DISP_OD,
> +	CMDQ_ENG_DISP_OVL0,
> +	CMDQ_ENG_DISP_OVL1,
> +	CMDQ_ENG_DISP_PWM0,
> +	CMDQ_ENG_DISP_PWM1,
> +	CMDQ_ENG_DISP_RDMA0,
> +	CMDQ_ENG_DISP_RDMA1,
> +	CMDQ_ENG_DISP_RDMA2,
> +	CMDQ_ENG_DISP_UFOE,
> +	CMDQ_ENG_DISP_WDMA0,
> +	CMDQ_ENG_DISP_WDMA1,
> +	CMDQ_ENG_MAX,
> +};
> +
> +/* events for CMDQ and display */
> +enum cmdq_event {
> +	/* Display start of frame(SOF) events */
> +	CMDQ_EVENT_DISP_OVL0_SOF = 11,
> +	CMDQ_EVENT_DISP_OVL1_SOF = 12,
> +	CMDQ_EVENT_DISP_RDMA0_SOF = 13,
> +	CMDQ_EVENT_DISP_RDMA1_SOF = 14,
> +	CMDQ_EVENT_DISP_RDMA2_SOF = 15,
> +	CMDQ_EVENT_DISP_WDMA0_SOF = 16,
> +	CMDQ_EVENT_DISP_WDMA1_SOF = 17,
> +	/* Display end of frame(EOF) events */
> +	CMDQ_EVENT_DISP_OVL0_EOF = 39,
> +	CMDQ_EVENT_DISP_OVL1_EOF = 40,
> +	CMDQ_EVENT_DISP_RDMA0_EOF = 41,
> +	CMDQ_EVENT_DISP_RDMA1_EOF = 42,
> +	CMDQ_EVENT_DISP_RDMA2_EOF = 43,
> +	CMDQ_EVENT_DISP_WDMA0_EOF = 44,
> +	CMDQ_EVENT_DISP_WDMA1_EOF = 45,
> +	/* Mutex end of frame(EOF) events */
> +	CMDQ_EVENT_MUTEX0_STREAM_EOF = 53,
> +	CMDQ_EVENT_MUTEX1_STREAM_EOF = 54,
> +	CMDQ_EVENT_MUTEX2_STREAM_EOF = 55,
> +	CMDQ_EVENT_MUTEX3_STREAM_EOF = 56,
> +	CMDQ_EVENT_MUTEX4_STREAM_EOF = 57,
> +	/* Display underrun events */
> +	CMDQ_EVENT_DISP_RDMA0_UNDERRUN = 63,
> +	CMDQ_EVENT_DISP_RDMA1_UNDERRUN = 64,
> +	CMDQ_EVENT_DISP_RDMA2_UNDERRUN = 65,
> +	/* Keep this at the end of HW events */
> +	CMDQ_MAX_HW_EVENT_COUNT = 260,
> +};
> +
> +struct cmdq_cb_data {
> +	bool	err;
> +	void	*data;
> +};
> +
> +typedef int (*cmdq_async_flush_cb)(struct cmdq_cb_data data);
> +
> +struct cmdq_task;
> +struct cmdq;
> +
> +struct cmdq_rec {
> +	struct cmdq		*cmdq;
> +	u64			engine_flag;
> +	size_t			command_size;
> +	void			*buf;
> +	size_t			buf_size;
> +	bool			finalized;
> +};
> +
> +struct cmdq_base {
> +	int	subsys;
> +	u32	base;

subsys can always be calculated via cmdq_subsys_base_to_id(base >> 16) 
so we can get rid of the struct, right?

> +};
> +
> +/**
> + * cmdq_register_device() - register device which needs CMDQ
> + * @dev:		device
> + *
> + * Return: cmdq_base pointer or NULL for failed
> + */
> +struct cmdq_base *cmdq_register_device(struct device *dev);
> +
> +/**
> + * cmdq_rec_create() - create command queue record
> + * @dev:		device
> + * @engine_flag:	command queue engine flag
> + * @rec_ptr:		command queue record pointer to retrieve cmdq_rec
> + *
> + * Return: 0 for success; else the error code is returned
> + */
> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
> +		    struct cmdq_rec **rec_ptr);
> +
> +/**
> + * cmdq_rec_write() - append write command to the command queue record
> + * @rec:	the command queue record
> + * @value:	the specified target register value
> + * @base:	the command queue base
> + * @offset:	register offset from module base
> + *
> + * Return: 0 for success; else the error code is returned
> + */
> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value,
> +		   struct cmdq_base *base, u32 offset);
> +
> +/**
> + * cmdq_rec_write_mask() - append write command with mask to the command
> + *			   queue record
> + * @rec:	the command queue record
> + * @value:	the specified target register value
> + * @base:	the command queue base
> + * @offset:	register offset from module base
> + * @mask:	the specified target register mask
> + *
> + * Return: 0 for success; else the error code is returned
> + */
> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
> +			struct cmdq_base *base, u32 offset, u32 mask);
> +
> +/**
> + * cmdq_rec_wfe() - append wait for event command to the command queue reco	rd

reco rd -> record

Regards,
Matthias

> + * @rec:	the command queue record
> + * @event:	the desired event type to "wait and CLEAR"
> + *
> + * Return: 0 for success; else the error code is returned
> + */
> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event);
> +
> +/**
> + * cmdq_rec_clear_event() - append clear event command to the command queue
> + *			    record
> + * @rec:	the command queue record
> + * @event:	the desired event to be cleared
> + *
> + * Return: 0 for success; else the error code is returned
> + */
> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event);
> +
> +/**
> + * cmdq_rec_flush() - trigger CMDQ to execute the recorded commands
> + * @rec:	the command queue record
> + *
> + * Return: 0 for success; else the error code is returned
> + *
> + * Trigger CMDQ to execute the recorded commands. Note that this is a
> + * synchronous flush function. When the function returned, the recorded
> + * commands have been done.
> + */
> +int cmdq_rec_flush(struct cmdq_rec *rec);
> +
> +/**
> + * cmdq_rec_flush_async() - trigger CMDQ to asynchronously execute the recorded
> + *			    commands and call back after ISR is finished
> + * @rec:	the command queue record
> + * @cb:		called in the end of CMDQ ISR
> + * @data:	this data will pass back to cb
> + *
> + * Return: 0 for success; else the error code is returned
> + *
> + * Trigger CMDQ to asynchronously execute the recorded commands and call back
> + * after ISR is finished. Note that this is an ASYNC function. When the function
> + * returned, it may or may not be finished. The ISR callback function is called
> + * in the end of ISR.
> + */
> +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
> +			 void *data);
> +
> +/**
> + * cmdq_rec_destroy() - destroy command queue record
> + * @rec:	the command queue record
> + */
> +void cmdq_rec_destroy(struct cmdq_rec *rec);
> +
> +#endif	/* __MTK_CMDQ_H__ */
>

[e-mail] May 31, 2016, 8:36 a.m. UTC | #4

Hi Mathias,

Please see my inline reply.

On Mon, 2016-05-30 at 17:31 +0200, Matthias Brugger wrote:
> 
> On 30/05/16 05:19, HS Liao wrote:
> > This patch is first version of Mediatek Command Queue(CMDQ) driver. The
> > CMDQ is used to help read/write registers with critical time limitation,
> > such as updating display configuration during the vblank. It controls
> > Global Command Engine (GCE) hardware to achieve this requirement.
> > Currently, CMDQ only supports display related hardwares, but we expect
> > it can be extended to other hardwares for future requirements.
> >
> > Signed-off-by: HS Liao <[e-mail]>
> > Signed-off-by: CK Hu <[e-mail]>
> > ---
> >   drivers/soc/mediatek/Kconfig    |  10 +
> >   drivers/soc/mediatek/Makefile   |   1 +
> >   drivers/soc/mediatek/mtk-cmdq.c | 943 ++++++++++++++++++++++++++++++++++++++++
> >   include/soc/mediatek/cmdq.h     | 197 +++++++++
> >   4 files changed, 1151 insertions(+)
> >   create mode 100644 drivers/soc/mediatek/mtk-cmdq.c
> >   create mode 100644 include/soc/mediatek/cmdq.h
> >
> > diff --git a/drivers/soc/mediatek/Kconfig b/drivers/soc/mediatek/Kconfig
> > index 0a4ea80..c4ad75c 100644
> > --- a/drivers/soc/mediatek/Kconfig
> > +++ b/drivers/soc/mediatek/Kconfig
> > @@ -1,6 +1,16 @@
> >   #
> >   # MediaTek SoC drivers
> >   #
> > +config MTK_CMDQ
> > +	bool "MediaTek CMDQ Support"
> > +	depends on ARCH_MEDIATEK || COMPILE_TEST
> > +	select MTK_INFRACFG
> > +	help
> > +	  Say yes here to add support for the MediaTek Command Queue (CMDQ)
> > +	  driver. The CMDQ is used to help read/write registers with critical
> > +	  time limitation, such as updating display configuration during the
> > +	  vblank.
> > +
> >   config MTK_INFRACFG
> >   	bool "MediaTek INFRACFG Support"
> >   	depends on ARCH_MEDIATEK || COMPILE_TEST
> > diff --git a/drivers/soc/mediatek/Makefile b/drivers/soc/mediatek/Makefile
> > index 12998b0..f7397ef 100644
> > --- a/drivers/soc/mediatek/Makefile
> > +++ b/drivers/soc/mediatek/Makefile
> > @@ -1,3 +1,4 @@
> > +obj-$(CONFIG_MTK_CMDQ) += mtk-cmdq.o
> >   obj-$(CONFIG_MTK_INFRACFG) += mtk-infracfg.o
> >   obj-$(CONFIG_MTK_PMIC_WRAP) += mtk-pmic-wrap.o
> >   obj-$(CONFIG_MTK_SCPSYS) += mtk-scpsys.o
> > diff --git a/drivers/soc/mediatek/mtk-cmdq.c b/drivers/soc/mediatek/mtk-cmdq.c
> > new file mode 100644
> > index 0000000..e9d6e1c
> > --- /dev/null
> > +++ b/drivers/soc/mediatek/mtk-cmdq.c
> > @@ -0,0 +1,943 @@
> > +/*
> > + * Copyright (c) 2015 MediaTek Inc.
> > + *
> > + * This program is free software; you can redistribute it and/or modify
> > + * it under the terms of the GNU General Public License version 2 as
> > + * published by the Free Software Foundation.
> > + *
> > + * This program is distributed in the hope that it will be useful,
> > + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> > + * GNU General Public License for more details.
> > + */
> > +
> > +#include <linux/clk.h>
> > +#include <linux/clk-provider.h>
> > +#include <linux/completion.h>
> > +#include <linux/dma-mapping.h>
> > +#include <linux/errno.h>
> > +#include <linux/interrupt.h>
> > +#include <linux/iopoll.h>
> > +#include <linux/kernel.h>
> > +#include <linux/kthread.h>
> > +#include <linux/module.h>
> > +#include <linux/mutex.h>
> > +#include <linux/of_address.h>
> > +#include <linux/of_irq.h>
> > +#include <linux/platform_device.h>
> > +#include <linux/slab.h>
> > +#include <linux/spinlock.h>
> > +#include <linux/suspend.h>
> > +#include <linux/workqueue.h>
> > +#include <soc/mediatek/cmdq.h>
> > +
> > +#define CMDQ_INITIAL_CMD_BLOCK_SIZE	PAGE_SIZE
> > +#define CMDQ_INST_SIZE			8 /* instruction is 64-bit */
> > +#define CMDQ_TIMEOUT_MS			1000
> > +#define CMDQ_IRQ_MASK			0xffff
> > +#define CMDQ_DRIVER_DEVICE_NAME		"mtk_cmdq"
> > +#define CMDQ_CLK_NAME			"gce"
> 
> We can put the names in directly to un-bloat the defines.

I will use the names directly and remove defines.

> > +
> > +#define CMDQ_CURR_IRQ_STATUS		0x010
> > +#define CMDQ_CURR_LOADED_THR		0x018
> > +#define CMDQ_THR_SLOT_CYCLES		0x030
> > +
> > +#define CMDQ_THR_BASE			0x100
> > +#define CMDQ_THR_SHIFT			0x080
> 
> Wouldn't be CMDQ_THR_SIZE more accurate?

Will rename it.

> > +#define CMDQ_THR_WARM_RESET		0x00
> > +#define CMDQ_THR_ENABLE_TASK		0x04
> > +#define CMDQ_THR_SUSPEND_TASK		0x08
> > +#define CMDQ_THR_CURR_STATUS		0x0c
> > +#define CMDQ_THR_IRQ_STATUS		0x10
> > +#define CMDQ_THR_IRQ_ENABLE		0x14
> > +#define CMDQ_THR_CURR_ADDR		0x20
> > +#define CMDQ_THR_END_ADDR		0x24
> > +#define CMDQ_THR_CFG			0x40
> > +
> > +#define CMDQ_THR_ENABLED		0x1
> > +#define CMDQ_THR_DISABLED		0x0
> > +#define CMDQ_THR_SUSPEND		0x1
> > +#define CMDQ_THR_RESUME			0x0
> > +#define CMDQ_THR_STATUS_SUSPENDED	BIT(1)
> > +#define CMDQ_THR_DO_WARM_RESET		BIT(0)
> > +#define CMDQ_THR_ACTIVE_SLOT_CYCLES	0x3200
> > +#define CMDQ_THR_PRIORITY		3
> > +#define CMDQ_THR_IRQ_DONE		0x1
> > +#define CMDQ_THR_IRQ_ERROR		0x12
> > +#define CMDQ_THR_IRQ_EN			0x13 /* done + error */
> 
> #define CMDQ_THR_IRQ_EN (CMDQ_THR_IRQ_ERROR | CMDQ_THR_IRQ_DONE)

Will do.

> > +#define CMDQ_THR_IRQ_MASK		0x13
> 
> never used.

Will remove.

> > +#define CMDQ_THR_EXECUTING		BIT(31)
> > +
> > +#define CMDQ_ARG_A_WRITE_MASK		0xffff
> > +#define CMDQ_SUBSYS_MASK		0x1f
> > +#define CMDQ_OP_CODE_MASK		0xff000000
> > +
> > +#define CMDQ_OP_CODE_SHIFT		24
> 
> Couldn't we connect the mask with the shift, or aren't they related?
> 
> #define CMDQ_OP_CODE_MASK (0xff << CMDQ_OP_CODE_SHIFT)

Will do.

> > +#define CMDQ_SUBSYS_SHIFT		16
> > +
> > +#define CMDQ_WRITE_ENABLE_MASK		BIT(0)
> > +#define CMDQ_JUMP_BY_OFFSET		0x10000000
> > +#define CMDQ_JUMP_BY_PA			0x10000001
> > +#define CMDQ_JUMP_PASS			CMDQ_INST_SIZE
> > +#define CMDQ_WFE_UPDATE			BIT(31)
> > +#define CMDQ_WFE_WAIT			BIT(15)
> > +#define CMDQ_WFE_WAIT_VALUE		0x1
> > +#define CMDQ_EOC_IRQ_EN			BIT(0)
> > +
> > +enum cmdq_thread_index {
> > +	CMDQ_THR_DISP_MAIN_IDX,	/* main */
> > +	CMDQ_THR_DISP_SUB_IDX,	/* sub */
> > +	CMDQ_THR_DISP_MISC_IDX,	/* misc */
> > +	CMDQ_THR_MAX_COUNT,	/* max */
> > +};
> > +
> > +/*
> > + * CMDQ_CODE_MOVE:
> > + *   move value into internal register as mask
> > + *   format: op mask
> > + * CMDQ_CODE_WRITE:
> > + *   write value into target register
> > + *   format: op subsys address value
> > + * CMDQ_CODE_JUMP:
> > + *   jump by offset
> > + *   format: op offset
> > + * CMDQ_CODE_WFE:
> > + *   wait for event and clear
> > + *   it is just clear if no wait
> > + *   format: [wait]  op event update:1 to_wait:1 wait:1
> > + *           [clear] op event update:1 to_wait:0 wait:0
> > + * CMDQ_CODE_EOC:
> > + *   end of command
> > + *   format: op irq_flag
> > + */
> 
> I think we need more documentation of how this command queue engine is 
> working. If not, I think it will be really complicated to understand how 
> to use this.
> 
> > +enum cmdq_code {
> > +	CMDQ_CODE_MOVE = 0x02,
> > +	CMDQ_CODE_WRITE = 0x04,
> > +	CMDQ_CODE_JUMP = 0x10,
> > +	CMDQ_CODE_WFE = 0x20,
> > +	CMDQ_CODE_EOC = 0x40,
> > +};
> > +
> > +enum cmdq_task_state {
> > +	TASK_STATE_BUSY, /* running on a GCE thread */
> > +	TASK_STATE_ERROR,
> > +	TASK_STATE_DONE,
> > +};
> > +
> > +struct cmdq_task_cb {
> > +	cmdq_async_flush_cb	cb;
> > +	void			*data;
> > +};
> > +
> > +struct cmdq_thread {
> > +	void __iomem		*base;
> > +	struct list_head	task_busy_list;
> > +	wait_queue_head_t	wait_task_done;
> > +};
> > +
> > +struct cmdq_task {
> > +	struct cmdq		*cmdq;
> > +	struct list_head	list_entry;
> > +	enum cmdq_task_state	task_state;
> > +	void			*va_base;
> > +	dma_addr_t		pa_base;
> > +	u64			engine_flag;
> > +	size_t			command_size;
> > +	u32			num_cmd;
> 
> num_cmd is directly connected to command_size. I prefer to just keep 
> num_cmd and calculate the command_size where necessary.

After I trace code, I prefer to keep command_size and calculate num_cmd
where necessary. What do you think?

> > +	struct cmdq_thread	*thread;
> > +	struct cmdq_task_cb	cb;
> > +	struct work_struct	release_work;
> > +};
> > +
> > +struct cmdq {
> > +	struct device		*dev;
> > +	void __iomem		*base;
> > +	u32			irq;
> > +	struct workqueue_struct	*task_release_wq;
> > +	struct cmdq_thread	thread[CMDQ_THR_MAX_COUNT];
> > +	struct mutex		task_mutex;	/* for task */
> > +	spinlock_t		exec_lock;	/* for exec */
> > +	struct clk		*clock;
> > +	bool			suspended;
> > +};
> > +
> > +struct cmdq_subsys {
> > +	u32	base;
> > +	int	id;
> > +};
> > +
> > +static const struct cmdq_subsys gce_subsys[] = {
> > +	{0x1400, 1},
> > +	{0x1401, 2},
> > +	{0x1402, 3},
> > +};
> > +
> > +static int cmdq_subsys_base_to_id(u32 base)
> > +{
> > +	int i;
> > +
> > +	for (i = 0; i < ARRAY_SIZE(gce_subsys); i++)
> > +		if (gce_subsys[i].base == base)
> > +			return gce_subsys[i].id;
> > +	return -EFAULT;
> > +}
> > +
> > +static int cmdq_eng_get_thread(u64 flag)
> > +{
> > +	if (flag & BIT_ULL(CMDQ_ENG_DISP_DSI0))
> > +		return CMDQ_THR_DISP_MAIN_IDX;
> > +	else if (flag & BIT_ULL(CMDQ_ENG_DISP_DPI0))
> > +		return CMDQ_THR_DISP_SUB_IDX;
> > +	else
> > +		return CMDQ_THR_DISP_MISC_IDX;
> > +}
> > +
> > +static void cmdq_task_release(struct cmdq_task *task)
> > +{
> > +	struct cmdq *cmdq = task->cmdq;
> > +
> > +	dma_free_coherent(cmdq->dev, task->command_size, task->va_base,
> > +			  task->pa_base);
> > +	kfree(task);
> > +}
> > +
> > +static struct cmdq_task *cmdq_task_acquire(struct cmdq_rec *rec,
> > +					   struct cmdq_task_cb cb)
> > +{
> > +	struct cmdq *cmdq = rec->cmdq;
> > +	struct device *dev = cmdq->dev;
> > +	struct cmdq_task *task;
> > +
> > +	task = kzalloc(sizeof(*task), GFP_KERNEL);
> > +	INIT_LIST_HEAD(&task->list_entry);
> > +	task->va_base = dma_alloc_coherent(dev, rec->command_size,
> > +					   &task->pa_base, GFP_KERNEL);
> > +	if (!task->va_base) {
> > +		dev_err(dev, "allocate command buffer failed\n");
> > +		kfree(task);
> > +		return NULL;
> > +	}
> > +
> > +	task->cmdq = cmdq;
> > +	task->command_size = rec->command_size;
> > +	task->engine_flag = rec->engine_flag;
> > +	task->task_state = TASK_STATE_BUSY;
> > +	task->cb = cb;
> > +	memcpy(task->va_base, rec->buf, rec->command_size);
> > +	task->num_cmd = task->command_size / CMDQ_INST_SIZE;
> > +	return task;
> > +}
> > +
> > +static void cmdq_thread_writel(struct cmdq_thread *thread, u32 value,
> > +			       u32 offset)
> > +{
> > +	writel(value, thread->base + offset);
> > +}
> > +
> > +static u32 cmdq_thread_readl(struct cmdq_thread *thread, u32 offset)
> > +{
> > +	return readl(thread->base + offset);
> > +}
> 
> We can get rid of cmdq_thread_readl/writel.

Will do.

> > +
> > +static int cmdq_thread_suspend(struct cmdq *cmdq, struct cmdq_thread *thread)
> > +{
> > +	u32 status;
> > +
> > +	cmdq_thread_writel(thread, CMDQ_THR_SUSPEND, CMDQ_THR_SUSPEND_TASK);
> > +
> > +	/* If already disabled, treat as suspended successful. */
> > +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> > +	      CMDQ_THR_ENABLED))
> > +		return 0;
> > +
> > +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_CURR_STATUS,
> > +			status, status & CMDQ_THR_STATUS_SUSPENDED, 0, 10)) {
> > +		dev_err(cmdq->dev, "suspend GCE thread 0x%x failed\n",
> > +			(u32)(thread->base - cmdq->base));
> > +		return -EFAULT;
> > +	}
> > +
> > +	return 0;
> > +}
> > +
> > +static void cmdq_thread_resume(struct cmdq_thread *thread)
> > +{
> > +	cmdq_thread_writel(thread, CMDQ_THR_RESUME, CMDQ_THR_SUSPEND_TASK);
> > +}
> > +
> > +static int cmdq_thread_reset(struct cmdq *cmdq, struct cmdq_thread *thread)
> > +{
> > +	u32 warm_reset;
> > +
> > +	cmdq_thread_writel(thread, CMDQ_THR_DO_WARM_RESET, CMDQ_THR_WARM_RESET);
> > +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_WARM_RESET,
> > +			warm_reset, !(warm_reset & CMDQ_THR_DO_WARM_RESET),
> > +			0, 10)) {
> > +		dev_err(cmdq->dev, "reset GCE thread 0x%x failed\n",
> > +			(u32)(thread->base - cmdq->base));
> > +		return -EFAULT;
> > +	}
> > +	writel(CMDQ_THR_ACTIVE_SLOT_CYCLES, cmdq->base + CMDQ_THR_SLOT_CYCLES);
> > +	return 0;
> > +}
> > +
> > +static void cmdq_thread_disable(struct cmdq *cmdq, struct cmdq_thread *thread)
> > +{
> > +	cmdq_thread_reset(cmdq, thread);
> > +	cmdq_thread_writel(thread, CMDQ_THR_DISABLED, CMDQ_THR_ENABLE_TASK);
> > +}
> > +
> > +/* notify GCE to re-fetch commands by setting GCE thread PC */
> > +static void cmdq_thread_invalidate_fetched_data(struct cmdq_thread *thread)
> > +{
> > +	cmdq_thread_writel(thread,
> > +			   cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR),
> > +			   CMDQ_THR_CURR_ADDR);
> > +}
> > +
> > +static void cmdq_task_insert_into_thread(struct cmdq_task *task)
> > +{
> > +	struct cmdq_thread *thread = task->thread;
> > +	struct cmdq_task *prev_task = list_last_entry(
> > +			&thread->task_busy_list, typeof(*task), list_entry);
> > +	u64 *prev_task_base = prev_task->va_base;
> > +
> > +	/* let previous task jump to this task */
> > +	prev_task_base[prev_task->num_cmd - 1] = (u64)CMDQ_JUMP_BY_PA << 32 |
> > +						 task->pa_base;
> > +
> > +	cmdq_thread_invalidate_fetched_data(thread);
> > +}
> > +
> > +/* we assume tasks in the same display GCE thread are waiting the same event. */
> > +static void cmdq_task_remove_wfe(struct cmdq_task *task)
> > +{
> > +	u32 wfe_option = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
> > +	u32 wfe_op = CMDQ_CODE_WFE << CMDQ_OP_CODE_SHIFT;
> > +	u32 *base = task->va_base;
> > +	u32 num_cmd = task->num_cmd << 1;
> > +	int i;
> > +
> > +	for (i = 0; i < num_cmd; i += 2)
> > +		if (base[i] == wfe_option &&
> > +		    (base[i + 1] & CMDQ_OP_CODE_MASK) == wfe_op) {
> > +			base[i] = CMDQ_JUMP_PASS;
> > +			base[i + 1] = CMDQ_JUMP_BY_OFFSET;
> > +		}
> 
> After using the command buffer as a void pointer a u64 pointer, we now 
> reference to it as u32. I would prefer to explain here, how the command 
> looks like we are searching for and use a for loop passing task->num_cmd 
> instead.

Will use u64* to rewrite the above code.

> > +}
> > +
> > +static void cmdq_task_exec(struct cmdq_task *task, struct cmdq_thread *thread)
> > +{
> > +	struct cmdq *cmdq = task->cmdq;
> > +	unsigned long flags;
> > +	unsigned long curr_pa, end_pa;
> > +
> > +	WARN_ON(clk_prepare_enable(cmdq->clock) < 0);
> > +	spin_lock_irqsave(&cmdq->exec_lock, flags);
> 
> cmdq_task_exec is called with cmdq->task_mutex held, so why do we need 
> the spin_lock here? Can't we just use one of the two?

We can drop task_mutex, but we will get some side effects.
1. exec_lock needs to include more code, but I think it is not good for
   spinlock.
2. In cmdq_rec_flush_async(), task_mutex needs to protect
   (1) cmdq->suspended, (2) cmdq_task_exec(), and
   (3) cmdq_task_wait_release_schedule().
   If we drop task_mutex, we have to put cmdq->suspended if condition
   just before cmdq_task_exec() and inside exec_lock, and we have to
   release task and its command buffer if error. This will let flow
   become more complex and enlarge code size.

What do you think?

> > +	task->thread = thread;
> > +	task->task_state = TASK_STATE_BUSY;
> 
> That was already set in cmdq_task_acquire, why do we need to set it here 
> again?

Will drop it.

> > +	if (list_empty(&thread->task_busy_list)) {
> > +		WARN_ON(cmdq_thread_reset(cmdq, thread) < 0);
> > +
> > +		cmdq_thread_writel(thread, task->pa_base, CMDQ_THR_CURR_ADDR);
> > +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
> > +				   CMDQ_THR_END_ADDR);
> > +		cmdq_thread_writel(thread, CMDQ_THR_PRIORITY, CMDQ_THR_CFG);
> > +		cmdq_thread_writel(thread, CMDQ_THR_IRQ_EN,
> > +				   CMDQ_THR_IRQ_ENABLE);
> > +
> > +		cmdq_thread_writel(thread, CMDQ_THR_ENABLED,
> > +				   CMDQ_THR_ENABLE_TASK);
> > +	} else {
> > +		WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
> > +
> > +		/*
> > +		 * check boundary condition
> > +		 * PC = END - 8, EOC is executed
> > +		 * PC = END, all CMDs are executed
> > +		 */
> > +		curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> > +		end_pa = cmdq_thread_readl(thread, CMDQ_THR_END_ADDR);
> > +		if (curr_pa == end_pa - 8 || curr_pa == end_pa) {
> 
> 8 refers to CMDQ_INST_SIZE, right?

Yes, I will use CMDQ_INST_SIZE.

> > +			/* set to this task directly */
> > +			cmdq_thread_writel(thread, task->pa_base,
> > +					   CMDQ_THR_CURR_ADDR);
> > +		} else {
> > +			cmdq_task_insert_into_thread(task);
> > +
> > +			if (thread == &cmdq->thread[CMDQ_THR_DISP_MAIN_IDX] ||
> > +			    thread == &cmdq->thread[CMDQ_THR_DISP_SUB_IDX])
> > +				cmdq_task_remove_wfe(task);
> 
> We could do this check using the task->engine_flag, I suppose that's 
> easier to undestand then.

Will use task->engine_flag.

> > +
> > +			smp_mb(); /* modify jump before enable thread */
> > +		}
> > +
> > +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
> > +				   CMDQ_THR_END_ADDR);
> > +		cmdq_thread_resume(thread);
> > +	}
> > +	list_move_tail(&task->list_entry, &thread->task_busy_list);
> > +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> > +}
> > +
> > +static void cmdq_handle_error_done(struct cmdq *cmdq,
> > +				   struct cmdq_thread *thread, u32 irq_flag)
> > +{
> > +	struct cmdq_task *task, *tmp, *curr_task = NULL;
> > +	u32 curr_pa;
> > +	struct cmdq_cb_data cmdq_cb_data;
> > +	bool err;
> > +
> > +	if (irq_flag & CMDQ_THR_IRQ_ERROR)
> > +		err = true;
> > +	else if (irq_flag & CMDQ_THR_IRQ_DONE)
> > +		err = false;
> > +	else
> > +		return;
> > +
> > +	curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> > +
> > +	list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
> > +				 list_entry) {
> > +		if (curr_pa >= task->pa_base &&
> > +		    curr_pa < (task->pa_base + task->command_size))
> 
> What are you checking here? It seems as if you make some implcit 
> assumptions about pa_base and the order of execution of commands in the 
> thread. Is it save to do so? Does dma_alloc_coherent give any guarantees 
> about dma_handle?

1. Check what is the current running task in this GCE thread.
2. Yes.
3. Yes, CMDQ doesn't use iommu, so physical address is continuous.

> > +			curr_task = task;
> > +		if (task->cb.cb) {
> > +			cmdq_cb_data.err = curr_task ? err : false;
> > +			cmdq_cb_data.data = task->cb.data;
> > +			task->cb.cb(cmdq_cb_data);
> > +		}
> > +		task->task_state = (curr_task && err) ? TASK_STATE_ERROR :
> > +				TASK_STATE_DONE;
> > +		list_del(&task->list_entry);
> > +		if (curr_task)
> > +			break;
> > +	}
> > +
> > +	wake_up(&thread->wait_task_done);
> > +}
> > +
> > +static void cmdq_thread_irq_handler(struct cmdq *cmdq, int tid)
> > +{
> > +	struct cmdq_thread *thread = &cmdq->thread[tid];
> > +	unsigned long flags = 0L;
> > +	u32 irq_flag;
> > +
> > +	spin_lock_irqsave(&cmdq->exec_lock, flags);
> > +
> > +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
> > +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
> > +
> > +	/*
> > +	 * Another CPU core could run "release task" right before we acquire
> > +	 * the spin lock, and thus reset / disable this GCE thread, so we
> > +	 * need to check the enable bit of this GCE thread.
> > +	 */
> > +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> > +	      CMDQ_THR_ENABLED))
> > +		irq_flag = 0;
> 
> cmdq_handle_error_done just retuns in this case. Programming this way 
> just makes things confusing. What about:
> 
> if (cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> 	      CMDQ_THR_ENABLED)
> 	cmdq_handle_error_done(cmdq, thread, irq_flag);
> else
> 		irq_flag = 0;
> 
> spin_unlock_irqrestore(&cmdq->exec_lock, flags);

We still need to clear irq_flag if GCE thread is disabled.
So, I think we can just return here.

	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
	      CMDQ_THR_ENABLED))
		return;
 
What do you think?

> > +
> > +	cmdq_handle_error_done(cmdq, thread, irq_flag);
> > +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> > +}
> > +
> > +static irqreturn_t cmdq_irq_handler(int irq, void *dev)
> > +{
> > +	struct cmdq *cmdq = dev;
> > +	u32 irq_status;
> > +	int i;
> > +
> > +	irq_status = readl(cmdq->base + CMDQ_CURR_IRQ_STATUS);
> > +	irq_status &= CMDQ_IRQ_MASK;
> > +	irq_status ^= CMDQ_IRQ_MASK;
> 
> irq_status can be much bigger then 3, which is the number of threads in 
> the system (CMDQ_THR_MAX_COUNT). So why we use this mask here isn't 
> clear to me.

Our GCE hardware has 16 threads, but we only use 3 threads currently.

> > +
> > +	if (!irq_status)
> > +		return IRQ_NONE;
> > +
> > +	while (irq_status) {
> > +		i = ffs(irq_status) - 1;
> > +		irq_status &= ~BIT(i);
> > +		cmdq_thread_irq_handler(cmdq, i);
> > +	}
> 
> Can you explain how the irq status register looks like, that would it 
> make much easier to understand what happens here.

Bit 0 ~ 15 of irq status register represents GCE thread 0 ~ 15
interrupt. 0 means asserting interrupt; 1 means no interrupt.

> > +
> > +	return IRQ_HANDLED;	
> > +}
> > +
> > +static int cmdq_task_handle_error_result(struct cmdq_task *task)
> 
> We never check the return values, why do we have them?

Will drop return value.

> > +{
> > +	struct cmdq *cmdq = task->cmdq;
> > +	struct device *dev = cmdq->dev;
> > +	struct cmdq_thread *thread = task->thread;
> > +	struct cmdq_task *next_task, *prev_task;
> > +	u32 irq_flag;
> > +
> > +	/* suspend GCE thread to ensure consistency */
> > +	WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
> > +
> > +	/* ISR has handled this error task */
> > +	if (task->task_state == TASK_STATE_ERROR) {
> > +		next_task = list_first_entry_or_null(&thread->task_busy_list,
> > +				struct cmdq_task, list_entry);
> > +		if (next_task) /* move to next task */
> > +			cmdq_thread_writel(thread, next_task->pa_base,
> > +					   CMDQ_THR_CURR_ADDR);
> 
> We have to do this, as we suppose that the thread did not reach the jump 
> instruction we put into it's command queue, right?

Yes.

> > +		cmdq_thread_resume(thread);
> > +		return -ECANCELED;
> > +	}
> > +
> 
> if task_state != ERROR and != DONE. This means that the timeout of 
> task_release_wq has timed out, right?

Yes.

> > +	/*
> > +	 * Save next_task and prev_task in advance
> > +	 * since cmdq_handle_error_done will remove list_entry.
> > +	 */
> > +	next_task = prev_task = NULL;
> > +	if (task->list_entry.next != &thread->task_busy_list)
> > +		next_task = list_next_entry(task, list_entry);
> > +	if (task->list_entry.prev != &thread->task_busy_list)
> > +		prev_task = list_prev_entry(task, list_entry);
> > +
> > +	/*
> > +	 * Although IRQ is disabled, GCE continues to execute.
> > +	 * It may have pending IRQ before GCE thread is suspended,
> > +	 * so check this condition again.
> > +	 */
> 
> The first thing we did in this function was suspending the thread. Why 
> do we need this then?

Because timeout is CPU timeout not GCE timeout, GCE could just finish
this task before the GCE thread is suspended.

> To be honest this whole functions looks really like a design error. We 
> have to sperate the states much clearer so that it is possible to 
> understand what is happening in the GCE. Isn't it for example posible to 
> have worker queues for timed out tasks and tasks with an error? I'm not 
> sure how to do this, actually I'm not sure if I really understood how 
> this is supposed to work.

GCE doesn't have timeout. The timeout is decided and controlled by CPU,
so we check timeout in release work.
For error and done, they are easy to check by register, and we have
already created release work for timeout. So, I don't think we need to
create work queue for each case.

What do you think?

> How much do we win, when we patch the thread command queue for every 
> task we add, instead of just taking one task after another from the 
> task_busy_list?

GCE is used to help read/write registers with critical time limitation.
Sometimes, client may ask to process multiple tasks in a short period
of time, e.g. display flush multiple tasks for next vblank. So, CMDQ
shouldn't limit to process one task after another from the
task_busy_list. Currently, when interrupt or timeout, we will check
how many tasks are done, and which one is error or timeout.

> > +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
> > +	cmdq_handle_error_done(cmdq, thread, irq_flag);
> > +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
> > +
> > +	if (task->task_state == TASK_STATE_DONE) {
> > +		cmdq_thread_resume(thread);
> > +		return 0;
> > +	}
> > +
> > +	if (task->task_state == TASK_STATE_ERROR) {
> > +		dev_err(dev, "task 0x%p error\n", task);
> > +		if (next_task) /* move to next task */
> > +			cmdq_thread_writel(thread, next_task->pa_base,
> > +					   CMDQ_THR_CURR_ADDR);
> > +		cmdq_thread_resume(thread);
> > +		return -ECANCELED;
> > +	}
> > +
> > +	/* Task is running, so we force to remove it. */
> > +	dev_err(dev, "task 0x%p timeout or killed\n", task);
> > +	task->task_state = TASK_STATE_ERROR;
> > +
> > +	if (prev_task) {
> > +		u64 *prev_va = prev_task->va_base;
> > +		u64 *curr_va = task->va_base;
> > +
> > +		/* copy JUMP instruction */
> > +		prev_va[prev_task->num_cmd - 1] = curr_va[task->num_cmd - 1];
> > +
> > +		cmdq_thread_invalidate_fetched_data(thread);
> > +	} else if (next_task) { /* move to next task */
> > +		cmdq_thread_writel(thread, next_task->pa_base,
> > +				   CMDQ_THR_CURR_ADDR);
> > +	}
> > +
> > +	list_del(&task->list_entry);
> > +	cmdq_thread_resume(thread);
> > +
> > +	/* call cb here to prevent lock */
> > +	if (task->cb.cb) {
> > +		struct cmdq_cb_data cmdq_cb_data;
> > +
> > +		cmdq_cb_data.err = true;
> > +		cmdq_cb_data.data = task->cb.data;
> > +		task->cb.cb(cmdq_cb_data);
> > +	}
> > +
> > +	return -ECANCELED;
> > +}
> > +
> > +static void cmdq_task_wait_release_work(struct work_struct *work_item)
> > +{
> > +	struct cmdq_task *task = container_of(work_item, struct cmdq_task,
> > +					      release_work);
> > +	struct cmdq *cmdq = task->cmdq;
> > +	struct cmdq_thread *thread = task->thread;
> > +	unsigned long flags;
> > +
> > +	wait_event_timeout(thread->wait_task_done,
> > +			   task->task_state != TASK_STATE_BUSY,
> > +			   msecs_to_jiffies(CMDQ_TIMEOUT_MS));
> > +
> > +	spin_lock_irqsave(&cmdq->exec_lock, flags);
> > +	if (task->task_state != TASK_STATE_DONE)
> > +		cmdq_task_handle_error_result(task);
> > +	if (list_empty(&thread->task_busy_list))
> > +		cmdq_thread_disable(cmdq, thread);
> > +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> > +
> > +	/* release regardless of success or not */
> > +	clk_disable_unprepare(cmdq->clock);
> > +	cmdq_task_release(task);
> > +}
> > +
> > +static void cmdq_task_wait_release_schedule(struct cmdq_task *task)
> > +{
> > +	struct cmdq *cmdq = task->cmdq;
> > +
> > +	INIT_WORK(&task->release_work, cmdq_task_wait_release_work);
> > +	queue_work(cmdq->task_release_wq, &task->release_work);
> > +}
> > +
> > +static int cmdq_rec_realloc_cmd_buffer(struct cmdq_rec *rec, size_t size)
> > +{
> > +	void *new_buf;
> > +
> > +	new_buf = krealloc(rec->buf, size, GFP_KERNEL | __GFP_ZERO);
> > +	if (!new_buf)
> > +		return -ENOMEM;
> > +	rec->buf = new_buf;
> > +	rec->buf_size = size;
> > +	return 0;
> > +}
> > +
> > +struct cmdq_base *cmdq_register_device(struct device *dev)
> > +{
> > +	struct cmdq_base *cmdq_base;
> > +	struct resource res;
> > +	int subsys;
> > +	u32 base;
> > +
> > +	if (of_address_to_resource(dev->of_node, 0, &res))
> > +		return NULL;
> > +	base = (u32)res.start;
> > +
> > +	subsys = cmdq_subsys_base_to_id(base >> 16);
> > +	if (subsys < 0)
> > +		return NULL;
> > +
> > +	cmdq_base = devm_kmalloc(dev, sizeof(*cmdq_base), GFP_KERNEL);
> > +	if (!cmdq_base)
> > +		return NULL;
> > +	cmdq_base->subsys = subsys;
> > +	cmdq_base->base = base;
> > +
> > +	return cmdq_base;
> > +}
> > +EXPORT_SYMBOL(cmdq_register_device);
> > +
> > +int cmdq_rec_create(struct device *dev, u64 engine_flag,
> > +		    struct cmdq_rec **rec_ptr)
> > +{
> > +	struct cmdq_rec *rec;
> > +	int err;
> > +
> > +	rec = kzalloc(sizeof(*rec), GFP_KERNEL);
> > +	if (!rec)
> > +		return -ENOMEM;
> > +	rec->cmdq = dev_get_drvdata(dev);
> > +	rec->engine_flag = engine_flag;
> > +	err = cmdq_rec_realloc_cmd_buffer(rec, CMDQ_INITIAL_CMD_BLOCK_SIZE);
> 
> Just pass PAGE_SIZE here, no need for CMDQ_INITIAL_CMD_BLOCK_SIZE.

Will do.

> > +	if (err < 0) {
> > +		kfree(rec);
> > +		return err;
> > +	}
> > +	*rec_ptr = rec;
> > +	return 0;
> > +}
> > +EXPORT_SYMBOL(cmdq_rec_create);
> > +
> > +static int cmdq_rec_append_command(struct cmdq_rec *rec, enum cmdq_code code,
> > +				   u32 arg_a, u32 arg_b)
> > +{
> > +	u64 *cmd_ptr;
> > +	int err;
> > +
> > +	if (WARN_ON(rec->finalized))
> > +		return -EBUSY;
> > +	if (code < CMDQ_CODE_MOVE || code > CMDQ_CODE_EOC)
> > +		return -EINVAL;
> 
> cmdq_rec_append_command is just called from inside this driver and code 
> is a enum. We can expect it to be correct, no need for this check.

Will drop this check.

> > +	if (unlikely(rec->command_size + CMDQ_INST_SIZE > rec->buf_size)) {
> 
> command_size is the offset into the buffer to which a new command is 
> written, so this name is highly confusing. I wonder if this would be 
> easier to understand if we redefine command_size to something like the 
> number of commands and divide/multiply CMDQ_INST_SIZE where this is needed.

I can rename command_size to cmd_buf_size and calculate num_cmd by
dividing CMDQ_INST_SIZE.
What do you think?

> > +		err = cmdq_rec_realloc_cmd_buffer(rec, rec->buf_size * 2);
> > +		if (err < 0)
> > +			return err;
> > +	}
> > +	cmd_ptr = rec->buf + rec->command_size;
> > +	(*cmd_ptr) = (u64)((code << CMDQ_OP_CODE_SHIFT) | arg_a) << 32 | arg_b;
> > +	rec->command_size += CMDQ_INST_SIZE;
> > +	return 0;
> > +}
> > +
> > +int cmdq_rec_write(struct cmdq_rec *rec, u32 value, struct cmdq_base *base,
> > +		   u32 offset)
> > +{
> > +	u32 arg_a = ((base->base + offset) & CMDQ_ARG_A_WRITE_MASK) |
> > +		    ((base->subsys & CMDQ_SUBSYS_MASK) << CMDQ_SUBSYS_SHIFT);
> 
> base->subsys is the id from gce_sybsys, so we can expect it to be 
> correct, no need to mask with CMDQ_SUBSYS_MASK.

Will drop it.

> > +	return cmdq_rec_append_command(rec, CMDQ_CODE_WRITE, arg_a, value);
> > +}
> > +EXPORT_SYMBOL(cmdq_rec_write);
> > +
> > +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
> > +			struct cmdq_base *base, u32 offset, u32 mask)
> > +{
> > +	u32 offset_mask = offset;
> > +	int err;
> > +
> > +	if (mask != 0xffffffff) {
> > +		err = cmdq_rec_append_command(rec, CMDQ_CODE_MOVE, 0, ~mask);
> > +		if (err < 0)
> > +			return err;
> > +		offset_mask |= CMDQ_WRITE_ENABLE_MASK;
> > +	}
> > +	return cmdq_rec_write(rec, value, base, offset_mask);
> > +}
> > +EXPORT_SYMBOL(cmdq_rec_write_mask);
> > +
> > +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event)
> > +{
> > +	u32 arg_b;
> > +
> > +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
> > +		return -EINVAL;
> > +
> > +	/*
> > +	 * bit 0-11: wait value
> > +	 * bit 15: 1 - wait, 0 - no wait
> > +	 * bit 16-27: update value
> > +	 * bit 31: 1 - update, 0 - no update
> > +	 */
> 
> I don't understand this comments. What are they for?

This is for WFE command. I will comment it.

> > +	arg_b = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
> > +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event, arg_b);
> > +}
> > +EXPORT_SYMBOL(cmdq_rec_wfe);
> > +
> > +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event)
> > +{
> > +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
> > +		return -EINVAL;
> > +
> > +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event,
> > +				       CMDQ_WFE_UPDATE);
> > +}
> > +EXPORT_SYMBOL(cmdq_rec_clear_event);
> > +
> > +static int cmdq_rec_finalize(struct cmdq_rec *rec)
> > +{
> > +	int err;
> > +
> > +	if (rec->finalized)
> > +		return 0;
> > +
> > +	/* insert EOC and generate IRQ for each command iteration */
> > +	err = cmdq_rec_append_command(rec, CMDQ_CODE_EOC, 0, CMDQ_EOC_IRQ_EN);
> > +	if (err < 0)
> > +		return err;
> > +
> > +	/* JUMP to end */
> > +	err = cmdq_rec_append_command(rec, CMDQ_CODE_JUMP, 0, CMDQ_JUMP_PASS);
> > +	if (err < 0)
> > +		return err;
> > +
> 
> Does this need to be atomic?
> What happens if after CODE_EOC and before CODE_JUMP some 
> write/read/event gets added?
> What happens if more commands get added to the queue after CODE_JUMP, 
> but before finalized is set to true. Why don't you use atomic functions 
> to access finalized?

Since cmdq_rec doesn't guarantee thread safe, mutex is needed when
client uses cmdq_rec.

> > +	rec->finalized = true;
> > +	return 0;
> > +}
> > +
> > +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
> > +			 void *data)
> > +{
> > +	struct cmdq *cmdq = rec->cmdq;
> > +	struct cmdq_task *task;
> > +	struct cmdq_task_cb task_cb;
> > +	struct cmdq_thread *thread;
> > +	int err;
> > +
> > +	mutex_lock(&cmdq->task_mutex);
> > +	if (cmdq->suspended) {
> > +		dev_err(cmdq->dev, "%s is called after suspended\n", __func__);
> > +		mutex_unlock(&cmdq->task_mutex);
> > +		return -EPERM;
> > +	}
> > +
> > +	err = cmdq_rec_finalize(rec);
> > +	if (err < 0) {
> > +		mutex_unlock(&cmdq->task_mutex);
> > +		return err;
> > +	}
> > +
> > +	task_cb.cb = cb;
> > +	task_cb.data = data;
> > +	task = cmdq_task_acquire(rec, task_cb);
> > +	if (!task) {
> > +		mutex_unlock(&cmdq->task_mutex);
> > +		return -EFAULT;
> > +	}
> > +
> > +	thread = &cmdq->thread[cmdq_eng_get_thread(task->engine_flag)];
> > +	cmdq_task_exec(task, thread);
> > +	cmdq_task_wait_release_schedule(task);
> > +	mutex_unlock(&cmdq->task_mutex);
> > +	return 0;
> > +}
> > +EXPORT_SYMBOL(cmdq_rec_flush_async);
> > +
> > +struct cmdq_flush_completion {
> > +	struct completion cmplt;
> > +	bool err;
> > +};
> > +
> > +static int cmdq_rec_flush_cb(struct cmdq_cb_data data)
> > +{
> > +	struct cmdq_flush_completion *cmplt = data.data;
> > +
> > +	cmplt->err = data.err;
> > +	complete(&cmplt->cmplt);
> > +	return 0;
> > +}
> > +
> > +int cmdq_rec_flush(struct cmdq_rec *rec)
> > +{
> > +	struct cmdq_flush_completion cmplt;
> > +	int err;
> > +
> > +	init_completion(&cmplt.cmplt);
> > +	err = cmdq_rec_flush_async(rec, cmdq_rec_flush_cb, &cmplt);
> > +	if (err < 0)
> > +		return err;
> > +	wait_for_completion(&cmplt.cmplt);
> > +	return cmplt.err ? -EFAULT : 0;
> > +}
> > +EXPORT_SYMBOL(cmdq_rec_flush);
> > +
> > +void cmdq_rec_destroy(struct cmdq_rec *rec)
> > +{
> > +	kfree(rec->buf);
> > +	kfree(rec);
> > +}
> > +EXPORT_SYMBOL(cmdq_rec_destroy);
> > +
> > +static bool cmdq_task_is_empty(struct cmdq *cmdq)
> > +{
> > +	struct cmdq_thread *thread;
> > +	int i;
> > +
> > +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
> > +		thread = &cmdq->thread[i];
> > +		if (!list_empty(&thread->task_busy_list))
> > +			return false;
> > +	}
> > +	return true;
> > +}
> > +
> > +static int cmdq_suspend(struct device *dev)
> > +{
> > +	struct cmdq *cmdq = dev_get_drvdata(dev);
> > +	u32 exec_threads;
> > +
> > +	mutex_lock(&cmdq->task_mutex);
> > +	cmdq->suspended = true;
> > +	mutex_unlock(&cmdq->task_mutex);
> > +
> > +	exec_threads = readl(cmdq->base + CMDQ_CURR_LOADED_THR);
> > +	if ((exec_threads & CMDQ_THR_EXECUTING) && !cmdq_task_is_empty(cmdq)) {
> > +		dev_err(dev, "wait active tasks timeout.\n");
> > +		flush_workqueue(cmdq->task_release_wq);
> > +	}
> > +	return 0;
> > +}
> > +
> > +static int cmdq_resume(struct device *dev)
> > +{
> > +	struct cmdq *cmdq = dev_get_drvdata(dev);
> > +
> > +	cmdq->suspended = false;
> > +	return 0;
> > +}
> > +
> > +static int cmdq_remove(struct platform_device *pdev)
> > +{
> > +	struct cmdq *cmdq = platform_get_drvdata(pdev);
> > +
> > +	destroy_workqueue(cmdq->task_release_wq);
> > +	cmdq->task_release_wq = NULL;
> > +	return 0;
> > +}
> > +
> > +static int cmdq_probe(struct platform_device *pdev)
> > +{
> > +	struct device *dev = &pdev->dev;
> > +	struct device_node *node = dev->of_node;
> > +	struct resource *res;
> > +	struct cmdq *cmdq;
> > +	int err, i;
> > +
> > +	cmdq = devm_kzalloc(dev, sizeof(*cmdq), GFP_KERNEL);
> > +	if (!cmdq)
> > +		return -ENOMEM;
> > +	cmdq->dev = dev;
> > +
> > +	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> > +	cmdq->base = devm_ioremap_resource(dev, res);
> > +	if (IS_ERR(cmdq->base)) {
> > +		dev_err(dev, "failed to ioremap gce\n");
> > +		return PTR_ERR(cmdq->base);
> > +	}
> > +
> > +	cmdq->irq = irq_of_parse_and_map(node, 0);
> > +	if (!cmdq->irq) {
> > +		dev_err(dev, "failed to get irq\n");
> > +		return -EINVAL;
> > +	}
> > +
> > +	dev_dbg(dev, "cmdq device: addr:0x%p, va:0x%p, irq:%d\n",
> > +		dev, cmdq->base, cmdq->irq);
> > +
> > +	mutex_init(&cmdq->task_mutex);
> > +	spin_lock_init(&cmdq->exec_lock);
> > +	cmdq->task_release_wq = alloc_ordered_workqueue(
> > +			"%s", WQ_MEM_RECLAIM | WQ_HIGHPRI,
> > +			"cmdq_task_wait_release");
> > +
> > +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
> > +		cmdq->thread[i].base = cmdq->base + CMDQ_THR_BASE +
> > +				CMDQ_THR_SHIFT * i;
> > +		init_waitqueue_head(&cmdq->thread[i].wait_task_done);
> > +		INIT_LIST_HEAD(&cmdq->thread[i].task_busy_list);
> > +	}
> > +
> > +	platform_set_drvdata(pdev, cmdq);
> > +
> > +	err = devm_request_irq(dev, cmdq->irq, cmdq_irq_handler, IRQF_SHARED,
> > +			       CMDQ_DRIVER_DEVICE_NAME, cmdq);
> > +	if (err < 0) {
> > +		dev_err(dev, "failed to register ISR (%d)\n", err);
> > +		goto fail;
> > +	}
> > +
> > +	cmdq->clock = devm_clk_get(dev, CMDQ_CLK_NAME);
> > +	if (IS_ERR(cmdq->clock)) {
> > +		dev_err(dev, "failed to get clk:%s\n", CMDQ_CLK_NAME);
> > +		err = PTR_ERR(cmdq->clock);
> > +		goto fail;
> > +	}
> > +	return 0;
> > +
> > +fail:
> > +	cmdq_remove(pdev);
> > +	return err;
> > +}
> > +
> > +static const struct dev_pm_ops cmdq_pm_ops = {
> > +	.suspend = cmdq_suspend,
> > +	.resume = cmdq_resume,
> > +};
> > +
> > +static const struct of_device_id cmdq_of_ids[] = {
> > +	{.compatible = "mediatek,mt8173-gce",},
> > +	{}
> > +};
> > +
> > +static struct platform_driver cmdq_drv = {
> > +	.probe = cmdq_probe,
> > +	.remove = cmdq_remove,
> > +	.driver = {
> > +		.name = CMDQ_DRIVER_DEVICE_NAME,
> > +		.owner = THIS_MODULE,
> > +		.pm = &cmdq_pm_ops,
> > +		.of_match_table = cmdq_of_ids,
> > +	}
> > +};
> > +
> > +builtin_platform_driver(cmdq_drv);
> > diff --git a/include/soc/mediatek/cmdq.h b/include/soc/mediatek/cmdq.h
> > new file mode 100644
> > index 0000000..60eef3d
> > --- /dev/null
> > +++ b/include/soc/mediatek/cmdq.h
> > @@ -0,0 +1,197 @@
> > +/*
> > + * Copyright (c) 2015 MediaTek Inc.
> > + *
> > + * This program is free software; you can redistribute it and/or modify
> > + * it under the terms of the GNU General Public License version 2 as
> > + * published by the Free Software Foundation.
> > + *
> > + * This program is distributed in the hope that it will be useful,
> > + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> > + * GNU General Public License for more details.
> > + */
> > +
> > +#ifndef __MTK_CMDQ_H__
> > +#define __MTK_CMDQ_H__
> > +
> > +#include <linux/platform_device.h>
> > +#include <linux/types.h>
> > +
> > +enum cmdq_eng {
> > +	CMDQ_ENG_DISP_AAL,
> > +	CMDQ_ENG_DISP_COLOR0,
> > +	CMDQ_ENG_DISP_COLOR1,
> > +	CMDQ_ENG_DISP_DPI0,
> > +	CMDQ_ENG_DISP_DSI0,
> > +	CMDQ_ENG_DISP_DSI1,
> > +	CMDQ_ENG_DISP_GAMMA,
> > +	CMDQ_ENG_DISP_OD,
> > +	CMDQ_ENG_DISP_OVL0,
> > +	CMDQ_ENG_DISP_OVL1,
> > +	CMDQ_ENG_DISP_PWM0,
> > +	CMDQ_ENG_DISP_PWM1,
> > +	CMDQ_ENG_DISP_RDMA0,
> > +	CMDQ_ENG_DISP_RDMA1,
> > +	CMDQ_ENG_DISP_RDMA2,
> > +	CMDQ_ENG_DISP_UFOE,
> > +	CMDQ_ENG_DISP_WDMA0,
> > +	CMDQ_ENG_DISP_WDMA1,
> > +	CMDQ_ENG_MAX,
> > +};
> > +
> > +/* events for CMDQ and display */
> > +enum cmdq_event {
> > +	/* Display start of frame(SOF) events */
> > +	CMDQ_EVENT_DISP_OVL0_SOF = 11,
> > +	CMDQ_EVENT_DISP_OVL1_SOF = 12,
> > +	CMDQ_EVENT_DISP_RDMA0_SOF = 13,
> > +	CMDQ_EVENT_DISP_RDMA1_SOF = 14,
> > +	CMDQ_EVENT_DISP_RDMA2_SOF = 15,
> > +	CMDQ_EVENT_DISP_WDMA0_SOF = 16,
> > +	CMDQ_EVENT_DISP_WDMA1_SOF = 17,
> > +	/* Display end of frame(EOF) events */
> > +	CMDQ_EVENT_DISP_OVL0_EOF = 39,
> > +	CMDQ_EVENT_DISP_OVL1_EOF = 40,
> > +	CMDQ_EVENT_DISP_RDMA0_EOF = 41,
> > +	CMDQ_EVENT_DISP_RDMA1_EOF = 42,
> > +	CMDQ_EVENT_DISP_RDMA2_EOF = 43,
> > +	CMDQ_EVENT_DISP_WDMA0_EOF = 44,
> > +	CMDQ_EVENT_DISP_WDMA1_EOF = 45,
> > +	/* Mutex end of frame(EOF) events */
> > +	CMDQ_EVENT_MUTEX0_STREAM_EOF = 53,
> > +	CMDQ_EVENT_MUTEX1_STREAM_EOF = 54,
> > +	CMDQ_EVENT_MUTEX2_STREAM_EOF = 55,
> > +	CMDQ_EVENT_MUTEX3_STREAM_EOF = 56,
> > +	CMDQ_EVENT_MUTEX4_STREAM_EOF = 57,
> > +	/* Display underrun events */
> > +	CMDQ_EVENT_DISP_RDMA0_UNDERRUN = 63,
> > +	CMDQ_EVENT_DISP_RDMA1_UNDERRUN = 64,
> > +	CMDQ_EVENT_DISP_RDMA2_UNDERRUN = 65,
> > +	/* Keep this at the end of HW events */
> > +	CMDQ_MAX_HW_EVENT_COUNT = 260,
> > +};
> > +
> > +struct cmdq_cb_data {
> > +	bool	err;
> > +	void	*data;
> > +};
> > +
> > +typedef int (*cmdq_async_flush_cb)(struct cmdq_cb_data data);
> > +
> > +struct cmdq_task;
> > +struct cmdq;
> > +
> > +struct cmdq_rec {
> > +	struct cmdq		*cmdq;
> > +	u64			engine_flag;
> > +	size_t			command_size;
> > +	void			*buf;
> > +	size_t			buf_size;
> > +	bool			finalized;
> > +};
> > +
> > +struct cmdq_base {
> > +	int	subsys;
> > +	u32	base;
> 
> subsys can always be calculated via cmdq_subsys_base_to_id(base >> 16) 
> so we can get rid of the struct, right?

Current subsys method is based on previous comment from Daniel Kurtz.
Please take a look of our previous discussion.
http://lists.infradead.org/pipermail/linux-mediatek/2016-February/004483.html
Thanks.

> > +};
> > +
> > +/**
> > + * cmdq_register_device() - register device which needs CMDQ
> > + * @dev:		device
> > + *
> > + * Return: cmdq_base pointer or NULL for failed
> > + */
> > +struct cmdq_base *cmdq_register_device(struct device *dev);
> > +
> > +/**
> > + * cmdq_rec_create() - create command queue record
> > + * @dev:		device
> > + * @engine_flag:	command queue engine flag
> > + * @rec_ptr:		command queue record pointer to retrieve cmdq_rec
> > + *
> > + * Return: 0 for success; else the error code is returned
> > + */
> > +int cmdq_rec_create(struct device *dev, u64 engine_flag,
> > +		    struct cmdq_rec **rec_ptr);
> > +
> > +/**
> > + * cmdq_rec_write() - append write command to the command queue record
> > + * @rec:	the command queue record
> > + * @value:	the specified target register value
> > + * @base:	the command queue base
> > + * @offset:	register offset from module base
> > + *
> > + * Return: 0 for success; else the error code is returned
> > + */
> > +int cmdq_rec_write(struct cmdq_rec *rec, u32 value,
> > +		   struct cmdq_base *base, u32 offset);
> > +
> > +/**
> > + * cmdq_rec_write_mask() - append write command with mask to the command
> > + *			   queue record
> > + * @rec:	the command queue record
> > + * @value:	the specified target register value
> > + * @base:	the command queue base
> > + * @offset:	register offset from module base
> > + * @mask:	the specified target register mask
> > + *
> > + * Return: 0 for success; else the error code is returned
> > + */
> > +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
> > +			struct cmdq_base *base, u32 offset, u32 mask);
> > +
> > +/**
> > + * cmdq_rec_wfe() - append wait for event command to the command queue reco	rd
> 
> reco rd -> record

Will fix it.

> Regards,
> Matthias
> 
> > + * @rec:	the command queue record
> > + * @event:	the desired event type to "wait and CLEAR"
> > + *
> > + * Return: 0 for success; else the error code is returned
> > + */
> > +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event);
> > +
> > +/**
> > + * cmdq_rec_clear_event() - append clear event command to the command queue
> > + *			    record
> > + * @rec:	the command queue record
> > + * @event:	the desired event to be cleared
> > + *
> > + * Return: 0 for success; else the error code is returned
> > + */
> > +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event);
> > +
> > +/**
> > + * cmdq_rec_flush() - trigger CMDQ to execute the recorded commands
> > + * @rec:	the command queue record
> > + *
> > + * Return: 0 for success; else the error code is returned
> > + *
> > + * Trigger CMDQ to execute the recorded commands. Note that this is a
> > + * synchronous flush function. When the function returned, the recorded
> > + * commands have been done.
> > + */
> > +int cmdq_rec_flush(struct cmdq_rec *rec);
> > +
> > +/**
> > + * cmdq_rec_flush_async() - trigger CMDQ to asynchronously execute the recorded
> > + *			    commands and call back after ISR is finished
> > + * @rec:	the command queue record
> > + * @cb:		called in the end of CMDQ ISR
> > + * @data:	this data will pass back to cb
> > + *
> > + * Return: 0 for success; else the error code is returned
> > + *
> > + * Trigger CMDQ to asynchronously execute the recorded commands and call back
> > + * after ISR is finished. Note that this is an ASYNC function. When the function
> > + * returned, it may or may not be finished. The ISR callback function is called
> > + * in the end of ISR.
> > + */
> > +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
> > +			 void *data);
> > +
> > +/**
> > + * cmdq_rec_destroy() - destroy command queue record
> > + * @rec:	the command queue record
> > + */
> > +void cmdq_rec_destroy(struct cmdq_rec *rec);
> > +
> > +#endif	/* __MTK_CMDQ_H__ */
> >

Thanks,
HS

Matthias Brugger May 31, 2016, 8:04 p.m. UTC | #5

On 31/05/16 10:36, Horng-Shyang Liao wrote:
> Hi Mathias,
>
> Please see my inline reply.
>
> On Mon, 2016-05-30 at 17:31 +0200, Matthias Brugger wrote:
>>
>> On 30/05/16 05:19, HS Liao wrote:
>>> This patch is first version of Mediatek Command Queue(CMDQ) driver. The
>>> CMDQ is used to help read/write registers with critical time limitation,
>>> such as updating display configuration during the vblank. It controls
>>> Global Command Engine (GCE) hardware to achieve this requirement.
>>> Currently, CMDQ only supports display related hardwares, but we expect
>>> it can be extended to other hardwares for future requirements.
>>>
>>> Signed-off-by: HS Liao <[e-mail]>
>>> Signed-off-by: CK Hu <[e-mail]>
>>> ---
>>>    drivers/soc/mediatek/Kconfig    |  10 +
>>>    drivers/soc/mediatek/Makefile   |   1 +
>>>    drivers/soc/mediatek/mtk-cmdq.c | 943 ++++++++++++++++++++++++++++++++++++++++
>>>    include/soc/mediatek/cmdq.h     | 197 +++++++++
>>>    4 files changed, 1151 insertions(+)
>>>    create mode 100644 drivers/soc/mediatek/mtk-cmdq.c
>>>    create mode 100644 include/soc/mediatek/cmdq.h
>>>
>>> diff --git a/drivers/soc/mediatek/Kconfig b/drivers/soc/mediatek/Kconfig
>>> index 0a4ea80..c4ad75c 100644
>>> --- a/drivers/soc/mediatek/Kconfig
>>> +++ b/drivers/soc/mediatek/Kconfig
>>> @@ -1,6 +1,16 @@
>>>    #
>>>    # MediaTek SoC drivers
>>>    #
>>> +config MTK_CMDQ
>>> +	bool "MediaTek CMDQ Support"
>>> +	depends on ARCH_MEDIATEK || COMPILE_TEST

depends on ARM64 ?

>>> +	select MTK_INFRACFG
>>> +	help
>>> +	  Say yes here to add support for the MediaTek Command Queue (CMDQ)
>>> +	  driver. The CMDQ is used to help read/write registers with critical
>>> +	  time limitation, such as updating display configuration during the
>>> +	  vblank.
>>> +
>>>    config MTK_INFRACFG
>>>    	bool "MediaTek INFRACFG Support"
>>>    	depends on ARCH_MEDIATEK || COMPILE_TEST
>>> diff --git a/drivers/soc/mediatek/Makefile b/drivers/soc/mediatek/Makefile
>>> index 12998b0..f7397ef 100644
>>> --- a/drivers/soc/mediatek/Makefile
>>> +++ b/drivers/soc/mediatek/Makefile
>>> @@ -1,3 +1,4 @@
>>> +obj-$(CONFIG_MTK_CMDQ) += mtk-cmdq.o
>>>    obj-$(CONFIG_MTK_INFRACFG) += mtk-infracfg.o
>>>    obj-$(CONFIG_MTK_PMIC_WRAP) += mtk-pmic-wrap.o
>>>    obj-$(CONFIG_MTK_SCPSYS) += mtk-scpsys.o
>>> diff --git a/drivers/soc/mediatek/mtk-cmdq.c b/drivers/soc/mediatek/mtk-cmdq.c
>>> new file mode 100644
>>> index 0000000..e9d6e1c
>>> --- /dev/null
>>> +++ b/drivers/soc/mediatek/mtk-cmdq.c
>>> @@ -0,0 +1,943 @@
>>> +/*
>>> + * Copyright (c) 2015 MediaTek Inc.
>>> + *
>>> + * This program is free software; you can redistribute it and/or modify
>>> + * it under the terms of the GNU General Public License version 2 as
>>> + * published by the Free Software Foundation.
>>> + *
>>> + * This program is distributed in the hope that it will be useful,
>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
>>> + * GNU General Public License for more details.
>>> + */
>>> +
>>> +#include <linux/clk.h>
>>> +#include <linux/clk-provider.h>
>>> +#include <linux/completion.h>
>>> +#include <linux/dma-mapping.h>
>>> +#include <linux/errno.h>
>>> +#include <linux/interrupt.h>
>>> +#include <linux/iopoll.h>
>>> +#include <linux/kernel.h>
>>> +#include <linux/kthread.h>
>>> +#include <linux/module.h>
>>> +#include <linux/mutex.h>
>>> +#include <linux/of_address.h>
>>> +#include <linux/of_irq.h>
>>> +#include <linux/platform_device.h>
>>> +#include <linux/slab.h>
>>> +#include <linux/spinlock.h>
>>> +#include <linux/suspend.h>
>>> +#include <linux/workqueue.h>
>>> +#include <soc/mediatek/cmdq.h>
>>> +
>>> +#define CMDQ_INITIAL_CMD_BLOCK_SIZE	PAGE_SIZE
>>> +#define CMDQ_INST_SIZE			8 /* instruction is 64-bit */
>>> +#define CMDQ_TIMEOUT_MS			1000
>>> +#define CMDQ_IRQ_MASK			0xffff
>>> +#define CMDQ_DRIVER_DEVICE_NAME		"mtk_cmdq"
>>> +#define CMDQ_CLK_NAME			"gce"
>>
>> We can put the names in directly to un-bloat the defines.
>
> I will use the names directly and remove defines.
>
>>> +
>>> +#define CMDQ_CURR_IRQ_STATUS		0x010
>>> +#define CMDQ_CURR_LOADED_THR		0x018
>>> +#define CMDQ_THR_SLOT_CYCLES		0x030
>>> +
>>> +#define CMDQ_THR_BASE			0x100
>>> +#define CMDQ_THR_SHIFT			0x080
>>
>> Wouldn't be CMDQ_THR_SIZE more accurate?
>
> Will rename it.
>
>>> +#define CMDQ_THR_WARM_RESET		0x00
>>> +#define CMDQ_THR_ENABLE_TASK		0x04
>>> +#define CMDQ_THR_SUSPEND_TASK		0x08
>>> +#define CMDQ_THR_CURR_STATUS		0x0c
>>> +#define CMDQ_THR_IRQ_STATUS		0x10
>>> +#define CMDQ_THR_IRQ_ENABLE		0x14
>>> +#define CMDQ_THR_CURR_ADDR		0x20
>>> +#define CMDQ_THR_END_ADDR		0x24
>>> +#define CMDQ_THR_CFG			0x40
>>> +
>>> +#define CMDQ_THR_ENABLED		0x1
>>> +#define CMDQ_THR_DISABLED		0x0
>>> +#define CMDQ_THR_SUSPEND		0x1
>>> +#define CMDQ_THR_RESUME			0x0
>>> +#define CMDQ_THR_STATUS_SUSPENDED	BIT(1)
>>> +#define CMDQ_THR_DO_WARM_RESET		BIT(0)
>>> +#define CMDQ_THR_ACTIVE_SLOT_CYCLES	0x3200
>>> +#define CMDQ_THR_PRIORITY		3
>>> +#define CMDQ_THR_IRQ_DONE		0x1
>>> +#define CMDQ_THR_IRQ_ERROR		0x12
>>> +#define CMDQ_THR_IRQ_EN			0x13 /* done + error */
>>
>> #define CMDQ_THR_IRQ_EN (CMDQ_THR_IRQ_ERROR | CMDQ_THR_IRQ_DONE)
>
> Will do.
>
>>> +#define CMDQ_THR_IRQ_MASK		0x13
>>
>> never used.
>
> Will remove.
>
>>> +#define CMDQ_THR_EXECUTING		BIT(31)
>>> +
>>> +#define CMDQ_ARG_A_WRITE_MASK		0xffff
>>> +#define CMDQ_SUBSYS_MASK		0x1f
>>> +#define CMDQ_OP_CODE_MASK		0xff000000
>>> +
>>> +#define CMDQ_OP_CODE_SHIFT		24
>>
>> Couldn't we connect the mask with the shift, or aren't they related?
>>
>> #define CMDQ_OP_CODE_MASK (0xff << CMDQ_OP_CODE_SHIFT)
>
> Will do.
>
>>> +#define CMDQ_SUBSYS_SHIFT		16
>>> +
>>> +#define CMDQ_WRITE_ENABLE_MASK		BIT(0)
>>> +#define CMDQ_JUMP_BY_OFFSET		0x10000000
>>> +#define CMDQ_JUMP_BY_PA			0x10000001
>>> +#define CMDQ_JUMP_PASS			CMDQ_INST_SIZE
>>> +#define CMDQ_WFE_UPDATE			BIT(31)
>>> +#define CMDQ_WFE_WAIT			BIT(15)
>>> +#define CMDQ_WFE_WAIT_VALUE		0x1
>>> +#define CMDQ_EOC_IRQ_EN			BIT(0)
>>> +
>>> +enum cmdq_thread_index {
>>> +	CMDQ_THR_DISP_MAIN_IDX,	/* main */
>>> +	CMDQ_THR_DISP_SUB_IDX,	/* sub */
>>> +	CMDQ_THR_DISP_MISC_IDX,	/* misc */
>>> +	CMDQ_THR_MAX_COUNT,	/* max */
>>> +};
>>> +
>>> +/*
>>> + * CMDQ_CODE_MOVE:
>>> + *   move value into internal register as mask
>>> + *   format: op mask
>>> + * CMDQ_CODE_WRITE:
>>> + *   write value into target register
>>> + *   format: op subsys address value
>>> + * CMDQ_CODE_JUMP:
>>> + *   jump by offset
>>> + *   format: op offset
>>> + * CMDQ_CODE_WFE:
>>> + *   wait for event and clear
>>> + *   it is just clear if no wait
>>> + *   format: [wait]  op event update:1 to_wait:1 wait:1
>>> + *           [clear] op event update:1 to_wait:0 wait:0
>>> + * CMDQ_CODE_EOC:
>>> + *   end of command
>>> + *   format: op irq_flag
>>> + */
>>
>> I think we need more documentation of how this command queue engine is
>> working. If not, I think it will be really complicated to understand how
>> to use this.
>>
>>> +enum cmdq_code {
>>> +	CMDQ_CODE_MOVE = 0x02,
>>> +	CMDQ_CODE_WRITE = 0x04,
>>> +	CMDQ_CODE_JUMP = 0x10,
>>> +	CMDQ_CODE_WFE = 0x20,
>>> +	CMDQ_CODE_EOC = 0x40,
>>> +};
>>> +
>>> +enum cmdq_task_state {
>>> +	TASK_STATE_BUSY, /* running on a GCE thread */
>>> +	TASK_STATE_ERROR,
>>> +	TASK_STATE_DONE,
>>> +};
>>> +
>>> +struct cmdq_task_cb {
>>> +	cmdq_async_flush_cb	cb;
>>> +	void			*data;
>>> +};
>>> +
>>> +struct cmdq_thread {
>>> +	void __iomem		*base;
>>> +	struct list_head	task_busy_list;
>>> +	wait_queue_head_t	wait_task_done;
>>> +};
>>> +
>>> +struct cmdq_task {
>>> +	struct cmdq		*cmdq;
>>> +	struct list_head	list_entry;
>>> +	enum cmdq_task_state	task_state;
>>> +	void			*va_base;
>>> +	dma_addr_t		pa_base;
>>> +	u64			engine_flag;
>>> +	size_t			command_size;
>>> +	u32			num_cmd;
>>
>> num_cmd is directly connected to command_size. I prefer to just keep
>> num_cmd and calculate the command_size where necessary.
>
> After I trace code, I prefer to keep command_size and calculate num_cmd
> where necessary. What do you think?
>

I suppose you prefer this, as you are writing to the GCE depending on 
the command_size. I think it is worth to create a macro for the 
calculation of the number of commands, to make the code more readable.
Would be nice if you would just pass cmdq_task to it and it would return 
the number. Just as an idea.

>>> +	struct cmdq_thread	*thread;
>>> +	struct cmdq_task_cb	cb;
>>> +	struct work_struct	release_work;
>>> +};
>>> +
>>> +struct cmdq {
>>> +	struct device		*dev;
>>> +	void __iomem		*base;
>>> +	u32			irq;
>>> +	struct workqueue_struct	*task_release_wq;
>>> +	struct cmdq_thread	thread[CMDQ_THR_MAX_COUNT];
>>> +	struct mutex		task_mutex;	/* for task */
>>> +	spinlock_t		exec_lock;	/* for exec */
>>> +	struct clk		*clock;
>>> +	bool			suspended;
>>> +};
>>> +
>>> +struct cmdq_subsys {
>>> +	u32	base;
>>> +	int	id;
>>> +};
>>> +
>>> +static const struct cmdq_subsys gce_subsys[] = {
>>> +	{0x1400, 1},
>>> +	{0x1401, 2},
>>> +	{0x1402, 3},
>>> +};
>>> +
>>> +static int cmdq_subsys_base_to_id(u32 base)
>>> +{
>>> +	int i;
>>> +
>>> +	for (i = 0; i < ARRAY_SIZE(gce_subsys); i++)
>>> +		if (gce_subsys[i].base == base)
>>> +			return gce_subsys[i].id;
>>> +	return -EFAULT;
>>> +}
>>> +
>>> +static int cmdq_eng_get_thread(u64 flag)
>>> +{
>>> +	if (flag & BIT_ULL(CMDQ_ENG_DISP_DSI0))
>>> +		return CMDQ_THR_DISP_MAIN_IDX;
>>> +	else if (flag & BIT_ULL(CMDQ_ENG_DISP_DPI0))
>>> +		return CMDQ_THR_DISP_SUB_IDX;
>>> +	else
>>> +		return CMDQ_THR_DISP_MISC_IDX;
>>> +}
>>> +
>>> +static void cmdq_task_release(struct cmdq_task *task)
>>> +{
>>> +	struct cmdq *cmdq = task->cmdq;
>>> +
>>> +	dma_free_coherent(cmdq->dev, task->command_size, task->va_base,
>>> +			  task->pa_base);
>>> +	kfree(task);
>>> +}
>>> +
>>> +static struct cmdq_task *cmdq_task_acquire(struct cmdq_rec *rec,
>>> +					   struct cmdq_task_cb cb)
>>> +{
>>> +	struct cmdq *cmdq = rec->cmdq;
>>> +	struct device *dev = cmdq->dev;
>>> +	struct cmdq_task *task;
>>> +
>>> +	task = kzalloc(sizeof(*task), GFP_KERNEL);
>>> +	INIT_LIST_HEAD(&task->list_entry);
>>> +	task->va_base = dma_alloc_coherent(dev, rec->command_size,
>>> +					   &task->pa_base, GFP_KERNEL);
>>> +	if (!task->va_base) {
>>> +		dev_err(dev, "allocate command buffer failed\n");
>>> +		kfree(task);
>>> +		return NULL;
>>> +	}
>>> +
>>> +	task->cmdq = cmdq;
>>> +	task->command_size = rec->command_size;
>>> +	task->engine_flag = rec->engine_flag;
>>> +	task->task_state = TASK_STATE_BUSY;
>>> +	task->cb = cb;
>>> +	memcpy(task->va_base, rec->buf, rec->command_size);
>>> +	task->num_cmd = task->command_size / CMDQ_INST_SIZE;
>>> +	return task;
>>> +}
>>> +
>>> +static void cmdq_thread_writel(struct cmdq_thread *thread, u32 value,
>>> +			       u32 offset)
>>> +{
>>> +	writel(value, thread->base + offset);
>>> +}
>>> +
>>> +static u32 cmdq_thread_readl(struct cmdq_thread *thread, u32 offset)
>>> +{
>>> +	return readl(thread->base + offset);
>>> +}
>>
>> We can get rid of cmdq_thread_readl/writel.
>
> Will do.
>
>>> +
>>> +static int cmdq_thread_suspend(struct cmdq *cmdq, struct cmdq_thread *thread)
>>> +{
>>> +	u32 status;
>>> +
>>> +	cmdq_thread_writel(thread, CMDQ_THR_SUSPEND, CMDQ_THR_SUSPEND_TASK);
>>> +
>>> +	/* If already disabled, treat as suspended successful. */
>>> +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
>>> +	      CMDQ_THR_ENABLED))
>>> +		return 0;
>>> +
>>> +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_CURR_STATUS,
>>> +			status, status & CMDQ_THR_STATUS_SUSPENDED, 0, 10)) {
>>> +		dev_err(cmdq->dev, "suspend GCE thread 0x%x failed\n",
>>> +			(u32)(thread->base - cmdq->base));
>>> +		return -EFAULT;
>>> +	}
>>> +
>>> +	return 0;
>>> +}
>>> +
>>> +static void cmdq_thread_resume(struct cmdq_thread *thread)
>>> +{
>>> +	cmdq_thread_writel(thread, CMDQ_THR_RESUME, CMDQ_THR_SUSPEND_TASK);
>>> +}
>>> +
>>> +static int cmdq_thread_reset(struct cmdq *cmdq, struct cmdq_thread *thread)
>>> +{
>>> +	u32 warm_reset;
>>> +
>>> +	cmdq_thread_writel(thread, CMDQ_THR_DO_WARM_RESET, CMDQ_THR_WARM_RESET);
>>> +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_WARM_RESET,
>>> +			warm_reset, !(warm_reset & CMDQ_THR_DO_WARM_RESET),
>>> +			0, 10)) {
>>> +		dev_err(cmdq->dev, "reset GCE thread 0x%x failed\n",
>>> +			(u32)(thread->base - cmdq->base));
>>> +		return -EFAULT;
>>> +	}
>>> +	writel(CMDQ_THR_ACTIVE_SLOT_CYCLES, cmdq->base + CMDQ_THR_SLOT_CYCLES);
>>> +	return 0;
>>> +}
>>> +
>>> +static void cmdq_thread_disable(struct cmdq *cmdq, struct cmdq_thread *thread)
>>> +{
>>> +	cmdq_thread_reset(cmdq, thread);
>>> +	cmdq_thread_writel(thread, CMDQ_THR_DISABLED, CMDQ_THR_ENABLE_TASK);
>>> +}
>>> +
>>> +/* notify GCE to re-fetch commands by setting GCE thread PC */
>>> +static void cmdq_thread_invalidate_fetched_data(struct cmdq_thread *thread)
>>> +{
>>> +	cmdq_thread_writel(thread,
>>> +			   cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR),
>>> +			   CMDQ_THR_CURR_ADDR);
>>> +}
>>> +
>>> +static void cmdq_task_insert_into_thread(struct cmdq_task *task)
>>> +{
>>> +	struct cmdq_thread *thread = task->thread;
>>> +	struct cmdq_task *prev_task = list_last_entry(
>>> +			&thread->task_busy_list, typeof(*task), list_entry);
>>> +	u64 *prev_task_base = prev_task->va_base;
>>> +
>>> +	/* let previous task jump to this task */
>>> +	prev_task_base[prev_task->num_cmd - 1] = (u64)CMDQ_JUMP_BY_PA << 32 |
>>> +						 task->pa_base;
>>> +
>>> +	cmdq_thread_invalidate_fetched_data(thread);
>>> +}
>>> +
>>> +/* we assume tasks in the same display GCE thread are waiting the same event. */
>>> +static void cmdq_task_remove_wfe(struct cmdq_task *task)
>>> +{
>>> +	u32 wfe_option = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
>>> +	u32 wfe_op = CMDQ_CODE_WFE << CMDQ_OP_CODE_SHIFT;
>>> +	u32 *base = task->va_base;
>>> +	u32 num_cmd = task->num_cmd << 1;
>>> +	int i;
>>> +
>>> +	for (i = 0; i < num_cmd; i += 2)
>>> +		if (base[i] == wfe_option &&
>>> +		    (base[i + 1] & CMDQ_OP_CODE_MASK) == wfe_op) {
>>> +			base[i] = CMDQ_JUMP_PASS;
>>> +			base[i + 1] = CMDQ_JUMP_BY_OFFSET;
>>> +		}
>>
>> After using the command buffer as a void pointer a u64 pointer, we now
>> reference to it as u32. I would prefer to explain here, how the command
>> looks like we are searching for and use a for loop passing task->num_cmd
>> instead.
>
> Will use u64* to rewrite the above code.
>
>>> +}
>>> +
>>> +static void cmdq_task_exec(struct cmdq_task *task, struct cmdq_thread *thread)
>>> +{
>>> +	struct cmdq *cmdq = task->cmdq;
>>> +	unsigned long flags;
>>> +	unsigned long curr_pa, end_pa;
>>> +
>>> +	WARN_ON(clk_prepare_enable(cmdq->clock) < 0);
>>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
>>
>> cmdq_task_exec is called with cmdq->task_mutex held, so why do we need
>> the spin_lock here? Can't we just use one of the two?
>
> We can drop task_mutex, but we will get some side effects.
> 1. exec_lock needs to include more code, but I think it is not good for
>     spinlock.
> 2. In cmdq_rec_flush_async(), task_mutex needs to protect
>     (1) cmdq->suspended, (2) cmdq_task_exec(), and
>     (3) cmdq_task_wait_release_schedule().
>     If we drop task_mutex, we have to put cmdq->suspended if condition
>     just before cmdq_task_exec() and inside exec_lock, and we have to
>     release task and its command buffer if error. This will let flow
>     become more complex and enlarge code size.
>
> What do you think?

Why do you need to protect cmdq_task_wait_release_schedule? We don't 
care about the order of the workqueue elements, do we?

As far as I understand you would need to protect cmdq_task_acquire as 
well, to "ensure" continously growing pa_base. More on that below.

>
>>> +	task->thread = thread;
>>> +	task->task_state = TASK_STATE_BUSY;
>>
>> That was already set in cmdq_task_acquire, why do we need to set it here
>> again?
>
> Will drop it.
>

Yeah, but I think it makes more sense to drop it in cmdq_task_acquire 
instead.

>>> +	if (list_empty(&thread->task_busy_list)) {
>>> +		WARN_ON(cmdq_thread_reset(cmdq, thread) < 0);
>>> +
>>> +		cmdq_thread_writel(thread, task->pa_base, CMDQ_THR_CURR_ADDR);
>>> +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
>>> +				   CMDQ_THR_END_ADDR);
>>> +		cmdq_thread_writel(thread, CMDQ_THR_PRIORITY, CMDQ_THR_CFG);
>>> +		cmdq_thread_writel(thread, CMDQ_THR_IRQ_EN,
>>> +				   CMDQ_THR_IRQ_ENABLE);
>>> +
>>> +		cmdq_thread_writel(thread, CMDQ_THR_ENABLED,
>>> +				   CMDQ_THR_ENABLE_TASK);
>>> +	} else {
>>> +		WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
>>> +
>>> +		/*
>>> +		 * check boundary condition
>>> +		 * PC = END - 8, EOC is executed
>>> +		 * PC = END, all CMDs are executed
>>> +		 */
>>> +		curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
>>> +		end_pa = cmdq_thread_readl(thread, CMDQ_THR_END_ADDR);
>>> +		if (curr_pa == end_pa - 8 || curr_pa == end_pa) {
>>
>> 8 refers to CMDQ_INST_SIZE, right?
>
> Yes, I will use CMDQ_INST_SIZE.
>
>>> +			/* set to this task directly */
>>> +			cmdq_thread_writel(thread, task->pa_base,
>>> +					   CMDQ_THR_CURR_ADDR);
>>> +		} else {
>>> +			cmdq_task_insert_into_thread(task);
>>> +
>>> +			if (thread == &cmdq->thread[CMDQ_THR_DISP_MAIN_IDX] ||
>>> +			    thread == &cmdq->thread[CMDQ_THR_DISP_SUB_IDX])
>>> +				cmdq_task_remove_wfe(task);
>>
>> We could do this check using the task->engine_flag, I suppose that's
>> easier to undestand then.
>
> Will use task->engine_flag.
>
>>> +
>>> +			smp_mb(); /* modify jump before enable thread */
>>> +		}
>>> +
>>> +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
>>> +				   CMDQ_THR_END_ADDR);
>>> +		cmdq_thread_resume(thread);
>>> +	}
>>> +	list_move_tail(&task->list_entry, &thread->task_busy_list);
>>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>> +}
>>> +
>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
>>> +				   struct cmdq_thread *thread, u32 irq_flag)
>>> +{
>>> +	struct cmdq_task *task, *tmp, *curr_task = NULL;
>>> +	u32 curr_pa;
>>> +	struct cmdq_cb_data cmdq_cb_data;
>>> +	bool err;
>>> +
>>> +	if (irq_flag & CMDQ_THR_IRQ_ERROR)
>>> +		err = true;
>>> +	else if (irq_flag & CMDQ_THR_IRQ_DONE)
>>> +		err = false;
>>> +	else
>>> +		return;
>>> +
>>> +	curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
>>> +
>>> +	list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
>>> +				 list_entry) {
>>> +		if (curr_pa >= task->pa_base &&
>>> +		    curr_pa < (task->pa_base + task->command_size))
>>
>> What are you checking here? It seems as if you make some implcit
>> assumptions about pa_base and the order of execution of commands in the
>> thread. Is it save to do so? Does dma_alloc_coherent give any guarantees
>> about dma_handle?
>
> 1. Check what is the current running task in this GCE thread.
> 2. Yes.
> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
>

Yes, physical addresses might be continous, but AFAIK there is no 
guarantee that the dma_handle address is steadily growing, when calling 
dma_alloc_coherent. And if I understand the code correctly, you use this 
assumption to decide if the task picked from task_busy_list is currently 
executing. So I think this mecanism is not working.

In which cases does the HW thread raise an interrupt.
In case of error. When does CMDQ_THR_IRQ_DONE get raised?

>>> +			curr_task = task;
>>> +		if (task->cb.cb) {
>>> +			cmdq_cb_data.err = curr_task ? err : false;
>>> +			cmdq_cb_data.data = task->cb.data;
>>> +			task->cb.cb(cmdq_cb_data);
>>> +		}
>>> +		task->task_state = (curr_task && err) ? TASK_STATE_ERROR :
>>> +				TASK_STATE_DONE;
>>> +		list_del(&task->list_entry);
>>> +		if (curr_task)
>>> +			break;
>>> +	}
>>> +
>>> +	wake_up(&thread->wait_task_done);
>>> +}
>>> +
>>> +static void cmdq_thread_irq_handler(struct cmdq *cmdq, int tid)
>>> +{
>>> +	struct cmdq_thread *thread = &cmdq->thread[tid];
>>> +	unsigned long flags = 0L;
>>> +	u32 irq_flag;
>>> +
>>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
>>> +
>>> +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
>>> +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
>>> +
>>> +	/*
>>> +	 * Another CPU core could run "release task" right before we acquire
>>> +	 * the spin lock, and thus reset / disable this GCE thread, so we
>>> +	 * need to check the enable bit of this GCE thread.
>>> +	 */
>>> +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
>>> +	      CMDQ_THR_ENABLED))
>>> +		irq_flag = 0;
>>
>> cmdq_handle_error_done just retuns in this case. Programming this way
>> just makes things confusing. What about:
>>
>> if (cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
>> 	      CMDQ_THR_ENABLED)
>> 	cmdq_handle_error_done(cmdq, thread, irq_flag);
>> else
>> 		irq_flag = 0;
>>
>> spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>
> We still need to clear irq_flag if GCE thread is disabled.
> So, I think we can just return here.
>
> 	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> 	      CMDQ_THR_ENABLED))
> 		return;
>
> What do you think?
>

No, you can't just return, you need to unlock the spinlock.
Anyway I would prefer it the other way round, as I put it in my last 
mail. Just delete the else branch, we don't need to set irq_flag to zero.

>>> +
>>> +	cmdq_handle_error_done(cmdq, thread, irq_flag);
>>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>> +}
>>> +
>>> +static irqreturn_t cmdq_irq_handler(int irq, void *dev)
>>> +{
>>> +	struct cmdq *cmdq = dev;
>>> +	u32 irq_status;
>>> +	int i;
>>> +
>>> +	irq_status = readl(cmdq->base + CMDQ_CURR_IRQ_STATUS);
>>> +	irq_status &= CMDQ_IRQ_MASK;
>>> +	irq_status ^= CMDQ_IRQ_MASK;
>>
>> irq_status can be much bigger then 3, which is the number of threads in
>> the system (CMDQ_THR_MAX_COUNT). So why we use this mask here isn't
>> clear to me.
>
> Our GCE hardware has 16 threads, but we only use 3 threads currently.
>

Ok, but please use bitops here.

>>> +
>>> +	if (!irq_status)
>>> +		return IRQ_NONE;
>>> +
>>> +	while (irq_status) {
>>> +		i = ffs(irq_status) - 1;
>>> +		irq_status &= ~BIT(i);
>>> +		cmdq_thread_irq_handler(cmdq, i);
>>> +	}
>>
>> Can you explain how the irq status register looks like, that would it
>> make much easier to understand what happens here.
>
> Bit 0 ~ 15 of irq status register represents GCE thread 0 ~ 15
> interrupt. 0 means asserting interrupt; 1 means no interrupt.
>

Thanks, that helped. :)

>>> +
>>> +	return IRQ_HANDLED;	
>>> +}
>>> +
>>> +static int cmdq_task_handle_error_result(struct cmdq_task *task)
>>
>> We never check the return values, why do we have them?
>
> Will drop return value.
>
>>> +{
>>> +	struct cmdq *cmdq = task->cmdq;
>>> +	struct device *dev = cmdq->dev;
>>> +	struct cmdq_thread *thread = task->thread;
>>> +	struct cmdq_task *next_task, *prev_task;
>>> +	u32 irq_flag;
>>> +
>>> +	/* suspend GCE thread to ensure consistency */
>>> +	WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
>>> +
>>> +	/* ISR has handled this error task */
>>> +	if (task->task_state == TASK_STATE_ERROR) {
>>> +		next_task = list_first_entry_or_null(&thread->task_busy_list,
>>> +				struct cmdq_task, list_entry);
>>> +		if (next_task) /* move to next task */
>>> +			cmdq_thread_writel(thread, next_task->pa_base,
>>> +					   CMDQ_THR_CURR_ADDR);
>>
>> We have to do this, as we suppose that the thread did not reach the jump
>> instruction we put into it's command queue, right?
>
> Yes.
>

So this should then go into it's own function. In wait_release_work, 
something like this:

if(task->task_state == TASK_STATE_ERROR)
      cmdq_task_handle_error(task)

>>> +		cmdq_thread_resume(thread);
>>> +		return -ECANCELED;
>>> +	}
>>> +
>>
>> if task_state != ERROR and != DONE. This means that the timeout of
>> task_release_wq has timed out, right?
>
> Yes.
>
>>> +	/*
>>> +	 * Save next_task and prev_task in advance
>>> +	 * since cmdq_handle_error_done will remove list_entry.
>>> +	 */
>>> +	next_task = prev_task = NULL;
>>> +	if (task->list_entry.next != &thread->task_busy_list)
>>> +		next_task = list_next_entry(task, list_entry);
>>> +	if (task->list_entry.prev != &thread->task_busy_list)
>>> +		prev_task = list_prev_entry(task, list_entry);
>>> +
>>> +	/*
>>> +	 * Although IRQ is disabled, GCE continues to execute.
>>> +	 * It may have pending IRQ before GCE thread is suspended,
>>> +	 * so check this condition again.
>>> +	 */
>>
>> The first thing we did in this function was suspending the thread. Why
>> do we need this then?
>
> Because timeout is CPU timeout not GCE timeout, GCE could just finish
> this task before the GCE thread is suspended.
>

What are the reasons for a timeout? An error has happend, or the task is 
still executing.

>> To be honest this whole functions looks really like a design error. We
>> have to sperate the states much clearer so that it is possible to
>> understand what is happening in the GCE. Isn't it for example posible to
>> have worker queues for timed out tasks and tasks with an error? I'm not
>> sure how to do this, actually I'm not sure if I really understood how
>> this is supposed to work.
>
> GCE doesn't have timeout. The timeout is decided and controlled by CPU,
> so we check timeout in release work.
> For error and done, they are easy to check by register, and we have
> already created release work for timeout. So, I don't think we need to
> create work queue for each case.
>
> What do you think?
>

I think, if we find in here, that the irq_flag is set, then the the 
interrupt handler was triggered and is spinning the spinlock. If this is 
not the case, we have a timeout and we handle this.

>> How much do we win, when we patch the thread command queue for every
>> task we add, instead of just taking one task after another from the
>> task_busy_list?
>
> GCE is used to help read/write registers with critical time limitation.
> Sometimes, client may ask to process multiple tasks in a short period
> of time, e.g. display flush multiple tasks for next vblank. So, CMDQ
> shouldn't limit to process one task after another from the
> task_busy_list. Currently, when interrupt or timeout, we will check
> how many tasks are done, and which one is error or timeout.
>

So I suppose the device driver who use this are interested in throughput 
and not in latency. The callback of every

>>> +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
>>> +	cmdq_handle_error_done(cmdq, thread, irq_flag);
>>> +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
>>> +
>>> +	if (task->task_state == TASK_STATE_DONE) {
>>> +		cmdq_thread_resume(thread);
>>> +		return 0;
>>> +	}
>>> +
>>> +	if (task->task_state == TASK_STATE_ERROR) {
>>> +		dev_err(dev, "task 0x%p error\n", task);
>>> +		if (next_task) /* move to next task */
>>> +			cmdq_thread_writel(thread, next_task->pa_base,
>>> +					   CMDQ_THR_CURR_ADDR);
>>> +		cmdq_thread_resume(thread);
>>> +		return -ECANCELED;
>>> +	}
>>> +
>>> +	/* Task is running, so we force to remove it. */
>>> +	dev_err(dev, "task 0x%p timeout or killed\n", task);
>>> +	task->task_state = TASK_STATE_ERROR;
>>> +
>>> +	if (prev_task) {
>>> +		u64 *prev_va = prev_task->va_base;
>>> +		u64 *curr_va = task->va_base;
>>> +
>>> +		/* copy JUMP instruction */
>>> +		prev_va[prev_task->num_cmd - 1] = curr_va[task->num_cmd - 1];
>>> +
>>> +		cmdq_thread_invalidate_fetched_data(thread);
>>> +	} else if (next_task) { /* move to next task */
>>> +		cmdq_thread_writel(thread, next_task->pa_base,
>>> +				   CMDQ_THR_CURR_ADDR);
>>> +	}
>>> +
>>> +	list_del(&task->list_entry);
>>> +	cmdq_thread_resume(thread);
>>> +
>>> +	/* call cb here to prevent lock */
>>> +	if (task->cb.cb) {
>>> +		struct cmdq_cb_data cmdq_cb_data;
>>> +
>>> +		cmdq_cb_data.err = true;
>>> +		cmdq_cb_data.data = task->cb.data;
>>> +		task->cb.cb(cmdq_cb_data);
>>> +	}
>>> +
>>> +	return -ECANCELED;
>>> +}
>>> +
>>> +static void cmdq_task_wait_release_work(struct work_struct *work_item)
>>> +{
>>> +	struct cmdq_task *task = container_of(work_item, struct cmdq_task,
>>> +					      release_work);
>>> +	struct cmdq *cmdq = task->cmdq;
>>> +	struct cmdq_thread *thread = task->thread;
>>> +	unsigned long flags;
>>> +
>>> +	wait_event_timeout(thread->wait_task_done,
>>> +			   task->task_state != TASK_STATE_BUSY,
>>> +			   msecs_to_jiffies(CMDQ_TIMEOUT_MS));
>>> +
>>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
>>> +	if (task->task_state != TASK_STATE_DONE)
>>> +		cmdq_task_handle_error_result(task);
>>> +	if (list_empty(&thread->task_busy_list))
>>> +		cmdq_thread_disable(cmdq, thread);
>>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>> +
>>> +	/* release regardless of success or not */
>>> +	clk_disable_unprepare(cmdq->clock);
>>> +	cmdq_task_release(task);
>>> +}
>>> +
>>> +static void cmdq_task_wait_release_schedule(struct cmdq_task *task)
>>> +{
>>> +	struct cmdq *cmdq = task->cmdq;
>>> +
>>> +	INIT_WORK(&task->release_work, cmdq_task_wait_release_work);
>>> +	queue_work(cmdq->task_release_wq, &task->release_work);
>>> +}
>>> +
>>> +static int cmdq_rec_realloc_cmd_buffer(struct cmdq_rec *rec, size_t size)
>>> +{
>>> +	void *new_buf;
>>> +
>>> +	new_buf = krealloc(rec->buf, size, GFP_KERNEL | __GFP_ZERO);
>>> +	if (!new_buf)
>>> +		return -ENOMEM;
>>> +	rec->buf = new_buf;
>>> +	rec->buf_size = size;
>>> +	return 0;
>>> +}
>>> +
>>> +struct cmdq_base *cmdq_register_device(struct device *dev)
>>> +{
>>> +	struct cmdq_base *cmdq_base;
>>> +	struct resource res;
>>> +	int subsys;
>>> +	u32 base;
>>> +
>>> +	if (of_address_to_resource(dev->of_node, 0, &res))
>>> +		return NULL;
>>> +	base = (u32)res.start;
>>> +
>>> +	subsys = cmdq_subsys_base_to_id(base >> 16);
>>> +	if (subsys < 0)
>>> +		return NULL;
>>> +
>>> +	cmdq_base = devm_kmalloc(dev, sizeof(*cmdq_base), GFP_KERNEL);
>>> +	if (!cmdq_base)
>>> +		return NULL;
>>> +	cmdq_base->subsys = subsys;
>>> +	cmdq_base->base = base;
>>> +
>>> +	return cmdq_base;
>>> +}
>>> +EXPORT_SYMBOL(cmdq_register_device);
>>> +
>>> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
>>> +		    struct cmdq_rec **rec_ptr)
>>> +{
>>> +	struct cmdq_rec *rec;
>>> +	int err;
>>> +
>>> +	rec = kzalloc(sizeof(*rec), GFP_KERNEL);
>>> +	if (!rec)
>>> +		return -ENOMEM;
>>> +	rec->cmdq = dev_get_drvdata(dev);
>>> +	rec->engine_flag = engine_flag;
>>> +	err = cmdq_rec_realloc_cmd_buffer(rec, CMDQ_INITIAL_CMD_BLOCK_SIZE);
>>
>> Just pass PAGE_SIZE here, no need for CMDQ_INITIAL_CMD_BLOCK_SIZE.
>
> Will do.
>
>>> +	if (err < 0) {
>>> +		kfree(rec);
>>> +		return err;
>>> +	}
>>> +	*rec_ptr = rec;
>>> +	return 0;
>>> +}
>>> +EXPORT_SYMBOL(cmdq_rec_create);
>>> +
>>> +static int cmdq_rec_append_command(struct cmdq_rec *rec, enum cmdq_code code,
>>> +				   u32 arg_a, u32 arg_b)
>>> +{
>>> +	u64 *cmd_ptr;
>>> +	int err;
>>> +
>>> +	if (WARN_ON(rec->finalized))
>>> +		return -EBUSY;
>>> +	if (code < CMDQ_CODE_MOVE || code > CMDQ_CODE_EOC)
>>> +		return -EINVAL;
>>
>> cmdq_rec_append_command is just called from inside this driver and code
>> is a enum. We can expect it to be correct, no need for this check.
>
> Will drop this check.
>
>>> +	if (unlikely(rec->command_size + CMDQ_INST_SIZE > rec->buf_size)) {
>>
>> command_size is the offset into the buffer to which a new command is
>> written, so this name is highly confusing. I wonder if this would be
>> easier to understand if we redefine command_size to something like the
>> number of commands and divide/multiply CMDQ_INST_SIZE where this is needed.
>
> I can rename command_size to cmd_buf_size and calculate num_cmd by
> dividing CMDQ_INST_SIZE.
> What do you think?
>
>>> +		err = cmdq_rec_realloc_cmd_buffer(rec, rec->buf_size * 2);
>>> +		if (err < 0)
>>> +			return err;
>>> +	}
>>> +	cmd_ptr = rec->buf + rec->command_size;
>>> +	(*cmd_ptr) = (u64)((code << CMDQ_OP_CODE_SHIFT) | arg_a) << 32 | arg_b;
>>> +	rec->command_size += CMDQ_INST_SIZE;
>>> +	return 0;
>>> +}
>>> +
>>> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value, struct cmdq_base *base,
>>> +		   u32 offset)
>>> +{
>>> +	u32 arg_a = ((base->base + offset) & CMDQ_ARG_A_WRITE_MASK) |
>>> +		    ((base->subsys & CMDQ_SUBSYS_MASK) << CMDQ_SUBSYS_SHIFT);
>>
>> base->subsys is the id from gce_sybsys, so we can expect it to be
>> correct, no need to mask with CMDQ_SUBSYS_MASK.
>
> Will drop it.
>
>>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WRITE, arg_a, value);
>>> +}
>>> +EXPORT_SYMBOL(cmdq_rec_write);
>>> +
>>> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
>>> +			struct cmdq_base *base, u32 offset, u32 mask)
>>> +{
>>> +	u32 offset_mask = offset;
>>> +	int err;
>>> +
>>> +	if (mask != 0xffffffff) {
>>> +		err = cmdq_rec_append_command(rec, CMDQ_CODE_MOVE, 0, ~mask);
>>> +		if (err < 0)
>>> +			return err;
>>> +		offset_mask |= CMDQ_WRITE_ENABLE_MASK;
>>> +	}
>>> +	return cmdq_rec_write(rec, value, base, offset_mask);
>>> +}
>>> +EXPORT_SYMBOL(cmdq_rec_write_mask);
>>> +
>>> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event)
>>> +{
>>> +	u32 arg_b;
>>> +
>>> +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
>>> +		return -EINVAL;
>>> +
>>> +	/*
>>> +	 * bit 0-11: wait value
>>> +	 * bit 15: 1 - wait, 0 - no wait
>>> +	 * bit 16-27: update value
>>> +	 * bit 31: 1 - update, 0 - no update
>>> +	 */
>>
>> I don't understand this comments. What are they for?
>
> This is for WFE command. I will comment it.
>
>>> +	arg_b = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
>>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event, arg_b);
>>> +}
>>> +EXPORT_SYMBOL(cmdq_rec_wfe);
>>> +
>>> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event)
>>> +{
>>> +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
>>> +		return -EINVAL;
>>> +
>>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event,
>>> +				       CMDQ_WFE_UPDATE);
>>> +}
>>> +EXPORT_SYMBOL(cmdq_rec_clear_event);
>>> +
>>> +static int cmdq_rec_finalize(struct cmdq_rec *rec)
>>> +{
>>> +	int err;
>>> +
>>> +	if (rec->finalized)
>>> +		return 0;
>>> +
>>> +	/* insert EOC and generate IRQ for each command iteration */
>>> +	err = cmdq_rec_append_command(rec, CMDQ_CODE_EOC, 0, CMDQ_EOC_IRQ_EN);
>>> +	if (err < 0)
>>> +		return err;
>>> +
>>> +	/* JUMP to end */
>>> +	err = cmdq_rec_append_command(rec, CMDQ_CODE_JUMP, 0, CMDQ_JUMP_PASS);
>>> +	if (err < 0)
>>> +		return err;
>>> +
>>
>> Does this need to be atomic?
>> What happens if after CODE_EOC and before CODE_JUMP some
>> write/read/event gets added?
>> What happens if more commands get added to the queue after CODE_JUMP,
>> but before finalized is set to true. Why don't you use atomic functions
>> to access finalized?
>
> Since cmdq_rec doesn't guarantee thread safe, mutex is needed when
> client uses cmdq_rec.
>

Well I think that rec->finalized tries to implement this, but might 
fail, if two kernel threads work on the same cmdq_rec.

>>> +	rec->finalized = true;
>>> +	return 0;
>>> +}
>>> +
>>> +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
>>> +			 void *data)
>>> +{
>>> +	struct cmdq *cmdq = rec->cmdq;
>>> +	struct cmdq_task *task;
>>> +	struct cmdq_task_cb task_cb;
>>> +	struct cmdq_thread *thread;
>>> +	int err;
>>> +
>>> +	mutex_lock(&cmdq->task_mutex);
>>> +	if (cmdq->suspended) {
>>> +		dev_err(cmdq->dev, "%s is called after suspended\n", __func__);
>>> +		mutex_unlock(&cmdq->task_mutex);
>>> +		return -EPERM;
>>> +	}
>>> +
>>> +	err = cmdq_rec_finalize(rec);
>>> +	if (err < 0) {
>>> +		mutex_unlock(&cmdq->task_mutex);
>>> +		return err;
>>> +	}
>>> +
>>> +	task_cb.cb = cb;
>>> +	task_cb.data = data;
>>> +	task = cmdq_task_acquire(rec, task_cb);
>>> +	if (!task) {
>>> +		mutex_unlock(&cmdq->task_mutex);
>>> +		return -EFAULT;
>>> +	}
>>> +
>>> +	thread = &cmdq->thread[cmdq_eng_get_thread(task->engine_flag)];
>>> +	cmdq_task_exec(task, thread);
>>> +	cmdq_task_wait_release_schedule(task);
>>> +	mutex_unlock(&cmdq->task_mutex);
>>> +	return 0;
>>> +}
>>> +EXPORT_SYMBOL(cmdq_rec_flush_async);
>>> +
>>> +struct cmdq_flush_completion {
>>> +	struct completion cmplt;
>>> +	bool err;
>>> +};
>>> +
>>> +static int cmdq_rec_flush_cb(struct cmdq_cb_data data)
>>> +{
>>> +	struct cmdq_flush_completion *cmplt = data.data;
>>> +
>>> +	cmplt->err = data.err;
>>> +	complete(&cmplt->cmplt);
>>> +	return 0;
>>> +}
>>> +
>>> +int cmdq_rec_flush(struct cmdq_rec *rec)
>>> +{
>>> +	struct cmdq_flush_completion cmplt;
>>> +	int err;
>>> +
>>> +	init_completion(&cmplt.cmplt);
>>> +	err = cmdq_rec_flush_async(rec, cmdq_rec_flush_cb, &cmplt);
>>> +	if (err < 0)
>>> +		return err;
>>> +	wait_for_completion(&cmplt.cmplt);
>>> +	return cmplt.err ? -EFAULT : 0;
>>> +}
>>> +EXPORT_SYMBOL(cmdq_rec_flush);
>>> +
>>> +void cmdq_rec_destroy(struct cmdq_rec *rec)
>>> +{
>>> +	kfree(rec->buf);
>>> +	kfree(rec);
>>> +}
>>> +EXPORT_SYMBOL(cmdq_rec_destroy);
>>> +
>>> +static bool cmdq_task_is_empty(struct cmdq *cmdq)
>>> +{
>>> +	struct cmdq_thread *thread;
>>> +	int i;
>>> +
>>> +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
>>> +		thread = &cmdq->thread[i];
>>> +		if (!list_empty(&thread->task_busy_list))
>>> +			return false;
>>> +	}
>>> +	return true;
>>> +}
>>> +
>>> +static int cmdq_suspend(struct device *dev)
>>> +{
>>> +	struct cmdq *cmdq = dev_get_drvdata(dev);
>>> +	u32 exec_threads;
>>> +
>>> +	mutex_lock(&cmdq->task_mutex);
>>> +	cmdq->suspended = true;
>>> +	mutex_unlock(&cmdq->task_mutex);
>>> +
>>> +	exec_threads = readl(cmdq->base + CMDQ_CURR_LOADED_THR);
>>> +	if ((exec_threads & CMDQ_THR_EXECUTING) && !cmdq_task_is_empty(cmdq)) {
>>> +		dev_err(dev, "wait active tasks timeout.\n");
>>> +		flush_workqueue(cmdq->task_release_wq);
>>> +	}
>>> +	return 0;
>>> +}
>>> +
>>> +static int cmdq_resume(struct device *dev)
>>> +{
>>> +	struct cmdq *cmdq = dev_get_drvdata(dev);
>>> +
>>> +	cmdq->suspended = false;
>>> +	return 0;
>>> +}
>>> +
>>> +static int cmdq_remove(struct platform_device *pdev)
>>> +{
>>> +	struct cmdq *cmdq = platform_get_drvdata(pdev);
>>> +
>>> +	destroy_workqueue(cmdq->task_release_wq);
>>> +	cmdq->task_release_wq = NULL;
>>> +	return 0;
>>> +}
>>> +
>>> +static int cmdq_probe(struct platform_device *pdev)
>>> +{
>>> +	struct device *dev = &pdev->dev;
>>> +	struct device_node *node = dev->of_node;
>>> +	struct resource *res;
>>> +	struct cmdq *cmdq;
>>> +	int err, i;
>>> +
>>> +	cmdq = devm_kzalloc(dev, sizeof(*cmdq), GFP_KERNEL);
>>> +	if (!cmdq)
>>> +		return -ENOMEM;
>>> +	cmdq->dev = dev;
>>> +
>>> +	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
>>> +	cmdq->base = devm_ioremap_resource(dev, res);
>>> +	if (IS_ERR(cmdq->base)) {
>>> +		dev_err(dev, "failed to ioremap gce\n");
>>> +		return PTR_ERR(cmdq->base);
>>> +	}
>>> +
>>> +	cmdq->irq = irq_of_parse_and_map(node, 0);
>>> +	if (!cmdq->irq) {
>>> +		dev_err(dev, "failed to get irq\n");
>>> +		return -EINVAL;
>>> +	}
>>> +
>>> +	dev_dbg(dev, "cmdq device: addr:0x%p, va:0x%p, irq:%d\n",
>>> +		dev, cmdq->base, cmdq->irq);
>>> +
>>> +	mutex_init(&cmdq->task_mutex);
>>> +	spin_lock_init(&cmdq->exec_lock);
>>> +	cmdq->task_release_wq = alloc_ordered_workqueue(
>>> +			"%s", WQ_MEM_RECLAIM | WQ_HIGHPRI,
>>> +			"cmdq_task_wait_release");
>>> +
>>> +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
>>> +		cmdq->thread[i].base = cmdq->base + CMDQ_THR_BASE +
>>> +				CMDQ_THR_SHIFT * i;
>>> +		init_waitqueue_head(&cmdq->thread[i].wait_task_done);
>>> +		INIT_LIST_HEAD(&cmdq->thread[i].task_busy_list);
>>> +	}
>>> +
>>> +	platform_set_drvdata(pdev, cmdq);
>>> +
>>> +	err = devm_request_irq(dev, cmdq->irq, cmdq_irq_handler, IRQF_SHARED,
>>> +			       CMDQ_DRIVER_DEVICE_NAME, cmdq);
>>> +	if (err < 0) {
>>> +		dev_err(dev, "failed to register ISR (%d)\n", err);
>>> +		goto fail;
>>> +	}
>>> +
>>> +	cmdq->clock = devm_clk_get(dev, CMDQ_CLK_NAME);
>>> +	if (IS_ERR(cmdq->clock)) {
>>> +		dev_err(dev, "failed to get clk:%s\n", CMDQ_CLK_NAME);
>>> +		err = PTR_ERR(cmdq->clock);
>>> +		goto fail;
>>> +	}
>>> +	return 0;
>>> +
>>> +fail:
>>> +	cmdq_remove(pdev);
>>> +	return err;
>>> +}
>>> +
>>> +static const struct dev_pm_ops cmdq_pm_ops = {
>>> +	.suspend = cmdq_suspend,
>>> +	.resume = cmdq_resume,
>>> +};
>>> +
>>> +static const struct of_device_id cmdq_of_ids[] = {
>>> +	{.compatible = "mediatek,mt8173-gce",},
>>> +	{}
>>> +};
>>> +
>>> +static struct platform_driver cmdq_drv = {
>>> +	.probe = cmdq_probe,
>>> +	.remove = cmdq_remove,
>>> +	.driver = {
>>> +		.name = CMDQ_DRIVER_DEVICE_NAME,
>>> +		.owner = THIS_MODULE,
>>> +		.pm = &cmdq_pm_ops,
>>> +		.of_match_table = cmdq_of_ids,
>>> +	}
>>> +};
>>> +
>>> +builtin_platform_driver(cmdq_drv);
>>> diff --git a/include/soc/mediatek/cmdq.h b/include/soc/mediatek/cmdq.h
>>> new file mode 100644
>>> index 0000000..60eef3d
>>> --- /dev/null
>>> +++ b/include/soc/mediatek/cmdq.h
>>> @@ -0,0 +1,197 @@
>>> +/*
>>> + * Copyright (c) 2015 MediaTek Inc.
>>> + *
>>> + * This program is free software; you can redistribute it and/or modify
>>> + * it under the terms of the GNU General Public License version 2 as
>>> + * published by the Free Software Foundation.
>>> + *
>>> + * This program is distributed in the hope that it will be useful,
>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
>>> + * GNU General Public License for more details.
>>> + */
>>> +
>>> +#ifndef __MTK_CMDQ_H__
>>> +#define __MTK_CMDQ_H__
>>> +
>>> +#include <linux/platform_device.h>
>>> +#include <linux/types.h>
>>> +
>>> +enum cmdq_eng {
>>> +	CMDQ_ENG_DISP_AAL,
>>> +	CMDQ_ENG_DISP_COLOR0,
>>> +	CMDQ_ENG_DISP_COLOR1,
>>> +	CMDQ_ENG_DISP_DPI0,
>>> +	CMDQ_ENG_DISP_DSI0,
>>> +	CMDQ_ENG_DISP_DSI1,
>>> +	CMDQ_ENG_DISP_GAMMA,
>>> +	CMDQ_ENG_DISP_OD,
>>> +	CMDQ_ENG_DISP_OVL0,
>>> +	CMDQ_ENG_DISP_OVL1,
>>> +	CMDQ_ENG_DISP_PWM0,
>>> +	CMDQ_ENG_DISP_PWM1,
>>> +	CMDQ_ENG_DISP_RDMA0,
>>> +	CMDQ_ENG_DISP_RDMA1,
>>> +	CMDQ_ENG_DISP_RDMA2,
>>> +	CMDQ_ENG_DISP_UFOE,
>>> +	CMDQ_ENG_DISP_WDMA0,
>>> +	CMDQ_ENG_DISP_WDMA1,
>>> +	CMDQ_ENG_MAX,
>>> +};
>>> +
>>> +/* events for CMDQ and display */
>>> +enum cmdq_event {
>>> +	/* Display start of frame(SOF) events */
>>> +	CMDQ_EVENT_DISP_OVL0_SOF = 11,
>>> +	CMDQ_EVENT_DISP_OVL1_SOF = 12,
>>> +	CMDQ_EVENT_DISP_RDMA0_SOF = 13,
>>> +	CMDQ_EVENT_DISP_RDMA1_SOF = 14,
>>> +	CMDQ_EVENT_DISP_RDMA2_SOF = 15,
>>> +	CMDQ_EVENT_DISP_WDMA0_SOF = 16,
>>> +	CMDQ_EVENT_DISP_WDMA1_SOF = 17,
>>> +	/* Display end of frame(EOF) events */
>>> +	CMDQ_EVENT_DISP_OVL0_EOF = 39,
>>> +	CMDQ_EVENT_DISP_OVL1_EOF = 40,
>>> +	CMDQ_EVENT_DISP_RDMA0_EOF = 41,
>>> +	CMDQ_EVENT_DISP_RDMA1_EOF = 42,
>>> +	CMDQ_EVENT_DISP_RDMA2_EOF = 43,
>>> +	CMDQ_EVENT_DISP_WDMA0_EOF = 44,
>>> +	CMDQ_EVENT_DISP_WDMA1_EOF = 45,
>>> +	/* Mutex end of frame(EOF) events */
>>> +	CMDQ_EVENT_MUTEX0_STREAM_EOF = 53,
>>> +	CMDQ_EVENT_MUTEX1_STREAM_EOF = 54,
>>> +	CMDQ_EVENT_MUTEX2_STREAM_EOF = 55,
>>> +	CMDQ_EVENT_MUTEX3_STREAM_EOF = 56,
>>> +	CMDQ_EVENT_MUTEX4_STREAM_EOF = 57,
>>> +	/* Display underrun events */
>>> +	CMDQ_EVENT_DISP_RDMA0_UNDERRUN = 63,
>>> +	CMDQ_EVENT_DISP_RDMA1_UNDERRUN = 64,
>>> +	CMDQ_EVENT_DISP_RDMA2_UNDERRUN = 65,
>>> +	/* Keep this at the end of HW events */
>>> +	CMDQ_MAX_HW_EVENT_COUNT = 260,
>>> +};
>>> +
>>> +struct cmdq_cb_data {
>>> +	bool	err;
>>> +	void	*data;
>>> +};
>>> +
>>> +typedef int (*cmdq_async_flush_cb)(struct cmdq_cb_data data);
>>> +
>>> +struct cmdq_task;
>>> +struct cmdq;
>>> +
>>> +struct cmdq_rec {
>>> +	struct cmdq		*cmdq;
>>> +	u64			engine_flag;
>>> +	size_t			command_size;
>>> +	void			*buf;
>>> +	size_t			buf_size;
>>> +	bool			finalized;
>>> +};

Why do we need cmdq_rec at all? Can't we just use the cmdq_task for that 
and this way make the driver less complex?

>>> +
>>> +struct cmdq_base {
>>> +	int	subsys;
>>> +	u32	base;
>>
>> subsys can always be calculated via cmdq_subsys_base_to_id(base >> 16)
>> so we can get rid of the struct, right?
>
> Current subsys method is based on previous comment from Daniel Kurtz.
> Please take a look of our previous discussion.
> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/004483.html
> Thanks.
>

I have to look deeper into this, but from what I read, the proposal from 
Daniel [1] seems good to me.

[1] https://patchwork.kernel.org/patch/8068311/

>>> +};
>>> +
>>> +/**
>>> + * cmdq_register_device() - register device which needs CMDQ
>>> + * @dev:		device
>>> + *
>>> + * Return: cmdq_base pointer or NULL for failed
>>> + */
>>> +struct cmdq_base *cmdq_register_device(struct device *dev);
>>> +
>>> +/**
>>> + * cmdq_rec_create() - create command queue record
>>> + * @dev:		device
>>> + * @engine_flag:	command queue engine flag
>>> + * @rec_ptr:		command queue record pointer to retrieve cmdq_rec
>>> + *
>>> + * Return: 0 for success; else the error code is returned
>>> + */
>>> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
>>> +		    struct cmdq_rec **rec_ptr);
>>> +
>>> +/**
>>> + * cmdq_rec_write() - append write command to the command queue record
>>> + * @rec:	the command queue record
>>> + * @value:	the specified target register value
>>> + * @base:	the command queue base
>>> + * @offset:	register offset from module base
>>> + *
>>> + * Return: 0 for success; else the error code is returned
>>> + */
>>> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value,
>>> +		   struct cmdq_base *base, u32 offset);
>>> +
>>> +/**
>>> + * cmdq_rec_write_mask() - append write command with mask to the command
>>> + *			   queue record
>>> + * @rec:	the command queue record
>>> + * @value:	the specified target register value
>>> + * @base:	the command queue base
>>> + * @offset:	register offset from module base
>>> + * @mask:	the specified target register mask
>>> + *
>>> + * Return: 0 for success; else the error code is returned
>>> + */
>>> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
>>> +			struct cmdq_base *base, u32 offset, u32 mask);
>>> +
>>> +/**
>>> + * cmdq_rec_wfe() - append wait for event command to the command queue reco	rd
>>
>> reco rd -> record
>
> Will fix it.
>
>> Regards,
>> Matthias
>>
>>> + * @rec:	the command queue record
>>> + * @event:	the desired event type to "wait and CLEAR"
>>> + *
>>> + * Return: 0 for success; else the error code is returned
>>> + */
>>> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event);
>>> +
>>> +/**
>>> + * cmdq_rec_clear_event() - append clear event command to the command queue
>>> + *			    record
>>> + * @rec:	the command queue record
>>> + * @event:	the desired event to be cleared
>>> + *
>>> + * Return: 0 for success; else the error code is returned
>>> + */
>>> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event);
>>> +
>>> +/**
>>> + * cmdq_rec_flush() - trigger CMDQ to execute the recorded commands
>>> + * @rec:	the command queue record
>>> + *
>>> + * Return: 0 for success; else the error code is returned
>>> + *
>>> + * Trigger CMDQ to execute the recorded commands. Note that this is a
>>> + * synchronous flush function. When the function returned, the recorded
>>> + * commands have been done.
>>> + */
>>> +int cmdq_rec_flush(struct cmdq_rec *rec);
>>> +
>>> +/**
>>> + * cmdq_rec_flush_async() - trigger CMDQ to asynchronously execute the recorded
>>> + *			    commands and call back after ISR is finished
>>> + * @rec:	the command queue record
>>> + * @cb:		called in the end of CMDQ ISR
>>> + * @data:	this data will pass back to cb
>>> + *
>>> + * Return: 0 for success; else the error code is returned
>>> + *
>>> + * Trigger CMDQ to asynchronously execute the recorded commands and call back
>>> + * after ISR is finished. Note that this is an ASYNC function. When the function
>>> + * returned, it may or may not be finished. The ISR callback function is called
>>> + * in the end of ISR.

"The callback is called from the ISR."

Regards,
Matthias

>>> + */
>>> +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
>>> +			 void *data);
>>> +
>>> +/**
>>> + * cmdq_rec_destroy() - destroy command queue record
>>> + * @rec:	the command queue record
>>> + */
>>> +void cmdq_rec_destroy(struct cmdq_rec *rec);
>>> +
>>> +#endif	/* __MTK_CMDQ_H__ */
>>>
>
> Thanks,
> HS
>

[e-mail] June 1, 2016, 9:57 a.m. UTC | #6

Hi Mathias,

Please see my inline reply.

On Tue, 2016-05-31 at 22:04 +0200, Matthias Brugger wrote:
> 
> On 31/05/16 10:36, Horng-Shyang Liao wrote:
> > Hi Mathias,
> >
> > Please see my inline reply.
> >
> > On Mon, 2016-05-30 at 17:31 +0200, Matthias Brugger wrote:
> >>
> >> On 30/05/16 05:19, HS Liao wrote:
> >>> This patch is first version of Mediatek Command Queue(CMDQ) driver. The
> >>> CMDQ is used to help read/write registers with critical time limitation,
> >>> such as updating display configuration during the vblank. It controls
> >>> Global Command Engine (GCE) hardware to achieve this requirement.
> >>> Currently, CMDQ only supports display related hardwares, but we expect
> >>> it can be extended to other hardwares for future requirements.
> >>>
> >>> Signed-off-by: HS Liao <[e-mail]>
> >>> Signed-off-by: CK Hu <[e-mail]>
> >>> ---
> >>>    drivers/soc/mediatek/Kconfig    |  10 +
> >>>    drivers/soc/mediatek/Makefile   |   1 +
> >>>    drivers/soc/mediatek/mtk-cmdq.c | 943 ++++++++++++++++++++++++++++++++++++++++
> >>>    include/soc/mediatek/cmdq.h     | 197 +++++++++
> >>>    4 files changed, 1151 insertions(+)
> >>>    create mode 100644 drivers/soc/mediatek/mtk-cmdq.c
> >>>    create mode 100644 include/soc/mediatek/cmdq.h
> >>>
> >>> diff --git a/drivers/soc/mediatek/Kconfig b/drivers/soc/mediatek/Kconfig
> >>> index 0a4ea80..c4ad75c 100644
> >>> --- a/drivers/soc/mediatek/Kconfig
> >>> +++ b/drivers/soc/mediatek/Kconfig
> >>> @@ -1,6 +1,16 @@
> >>>    #
> >>>    # MediaTek SoC drivers
> >>>    #
> >>> +config MTK_CMDQ
> >>> +	bool "MediaTek CMDQ Support"
> >>> +	depends on ARCH_MEDIATEK || COMPILE_TEST
> 
> depends on ARM64 ?

Will add ARM64.

> >>> +	select MTK_INFRACFG
> >>> +	help
> >>> +	  Say yes here to add support for the MediaTek Command Queue (CMDQ)
> >>> +	  driver. The CMDQ is used to help read/write registers with critical
> >>> +	  time limitation, such as updating display configuration during the
> >>> +	  vblank.
> >>> +
> >>>    config MTK_INFRACFG
> >>>    	bool "MediaTek INFRACFG Support"
> >>>    	depends on ARCH_MEDIATEK || COMPILE_TEST
> >>> diff --git a/drivers/soc/mediatek/Makefile b/drivers/soc/mediatek/Makefile
> >>> index 12998b0..f7397ef 100644
> >>> --- a/drivers/soc/mediatek/Makefile
> >>> +++ b/drivers/soc/mediatek/Makefile
> >>> @@ -1,3 +1,4 @@
> >>> +obj-$(CONFIG_MTK_CMDQ) += mtk-cmdq.o
> >>>    obj-$(CONFIG_MTK_INFRACFG) += mtk-infracfg.o
> >>>    obj-$(CONFIG_MTK_PMIC_WRAP) += mtk-pmic-wrap.o
> >>>    obj-$(CONFIG_MTK_SCPSYS) += mtk-scpsys.o
> >>> diff --git a/drivers/soc/mediatek/mtk-cmdq.c b/drivers/soc/mediatek/mtk-cmdq.c
> >>> new file mode 100644
> >>> index 0000000..e9d6e1c
> >>> --- /dev/null
> >>> +++ b/drivers/soc/mediatek/mtk-cmdq.c
> >>> @@ -0,0 +1,943 @@
> >>> +/*
> >>> + * Copyright (c) 2015 MediaTek Inc.
> >>> + *
> >>> + * This program is free software; you can redistribute it and/or modify
> >>> + * it under the terms of the GNU General Public License version 2 as
> >>> + * published by the Free Software Foundation.
> >>> + *
> >>> + * This program is distributed in the hope that it will be useful,
> >>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> >>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> >>> + * GNU General Public License for more details.
> >>> + */
> >>> +
> >>> +#include <linux/clk.h>
> >>> +#include <linux/clk-provider.h>
> >>> +#include <linux/completion.h>
> >>> +#include <linux/dma-mapping.h>
> >>> +#include <linux/errno.h>
> >>> +#include <linux/interrupt.h>
> >>> +#include <linux/iopoll.h>
> >>> +#include <linux/kernel.h>
> >>> +#include <linux/kthread.h>
> >>> +#include <linux/module.h>
> >>> +#include <linux/mutex.h>
> >>> +#include <linux/of_address.h>
> >>> +#include <linux/of_irq.h>
> >>> +#include <linux/platform_device.h>
> >>> +#include <linux/slab.h>
> >>> +#include <linux/spinlock.h>
> >>> +#include <linux/suspend.h>
> >>> +#include <linux/workqueue.h>
> >>> +#include <soc/mediatek/cmdq.h>
> >>> +
> >>> +#define CMDQ_INITIAL_CMD_BLOCK_SIZE	PAGE_SIZE
> >>> +#define CMDQ_INST_SIZE			8 /* instruction is 64-bit */
> >>> +#define CMDQ_TIMEOUT_MS			1000
> >>> +#define CMDQ_IRQ_MASK			0xffff
> >>> +#define CMDQ_DRIVER_DEVICE_NAME		"mtk_cmdq"
> >>> +#define CMDQ_CLK_NAME			"gce"
> >>
> >> We can put the names in directly to un-bloat the defines.
> >
> > I will use the names directly and remove defines.
> >
> >>> +
> >>> +#define CMDQ_CURR_IRQ_STATUS		0x010
> >>> +#define CMDQ_CURR_LOADED_THR		0x018
> >>> +#define CMDQ_THR_SLOT_CYCLES		0x030
> >>> +
> >>> +#define CMDQ_THR_BASE			0x100
> >>> +#define CMDQ_THR_SHIFT			0x080
> >>
> >> Wouldn't be CMDQ_THR_SIZE more accurate?
> >
> > Will rename it.
> >
> >>> +#define CMDQ_THR_WARM_RESET		0x00
> >>> +#define CMDQ_THR_ENABLE_TASK		0x04
> >>> +#define CMDQ_THR_SUSPEND_TASK		0x08
> >>> +#define CMDQ_THR_CURR_STATUS		0x0c
> >>> +#define CMDQ_THR_IRQ_STATUS		0x10
> >>> +#define CMDQ_THR_IRQ_ENABLE		0x14
> >>> +#define CMDQ_THR_CURR_ADDR		0x20
> >>> +#define CMDQ_THR_END_ADDR		0x24
> >>> +#define CMDQ_THR_CFG			0x40
> >>> +
> >>> +#define CMDQ_THR_ENABLED		0x1
> >>> +#define CMDQ_THR_DISABLED		0x0
> >>> +#define CMDQ_THR_SUSPEND		0x1
> >>> +#define CMDQ_THR_RESUME			0x0
> >>> +#define CMDQ_THR_STATUS_SUSPENDED	BIT(1)
> >>> +#define CMDQ_THR_DO_WARM_RESET		BIT(0)
> >>> +#define CMDQ_THR_ACTIVE_SLOT_CYCLES	0x3200
> >>> +#define CMDQ_THR_PRIORITY		3
> >>> +#define CMDQ_THR_IRQ_DONE		0x1
> >>> +#define CMDQ_THR_IRQ_ERROR		0x12
> >>> +#define CMDQ_THR_IRQ_EN			0x13 /* done + error */
> >>
> >> #define CMDQ_THR_IRQ_EN (CMDQ_THR_IRQ_ERROR | CMDQ_THR_IRQ_DONE)
> >
> > Will do.
> >
> >>> +#define CMDQ_THR_IRQ_MASK		0x13
> >>
> >> never used.
> >
> > Will remove.
> >
> >>> +#define CMDQ_THR_EXECUTING		BIT(31)
> >>> +
> >>> +#define CMDQ_ARG_A_WRITE_MASK		0xffff
> >>> +#define CMDQ_SUBSYS_MASK		0x1f
> >>> +#define CMDQ_OP_CODE_MASK		0xff000000
> >>> +
> >>> +#define CMDQ_OP_CODE_SHIFT		24
> >>
> >> Couldn't we connect the mask with the shift, or aren't they related?
> >>
> >> #define CMDQ_OP_CODE_MASK (0xff << CMDQ_OP_CODE_SHIFT)
> >
> > Will do.
> >
> >>> +#define CMDQ_SUBSYS_SHIFT		16
> >>> +
> >>> +#define CMDQ_WRITE_ENABLE_MASK		BIT(0)
> >>> +#define CMDQ_JUMP_BY_OFFSET		0x10000000
> >>> +#define CMDQ_JUMP_BY_PA			0x10000001
> >>> +#define CMDQ_JUMP_PASS			CMDQ_INST_SIZE
> >>> +#define CMDQ_WFE_UPDATE			BIT(31)
> >>> +#define CMDQ_WFE_WAIT			BIT(15)
> >>> +#define CMDQ_WFE_WAIT_VALUE		0x1
> >>> +#define CMDQ_EOC_IRQ_EN			BIT(0)
> >>> +
> >>> +enum cmdq_thread_index {
> >>> +	CMDQ_THR_DISP_MAIN_IDX,	/* main */
> >>> +	CMDQ_THR_DISP_SUB_IDX,	/* sub */
> >>> +	CMDQ_THR_DISP_MISC_IDX,	/* misc */
> >>> +	CMDQ_THR_MAX_COUNT,	/* max */
> >>> +};
> >>> +
> >>> +/*
> >>> + * CMDQ_CODE_MOVE:
> >>> + *   move value into internal register as mask
> >>> + *   format: op mask
> >>> + * CMDQ_CODE_WRITE:
> >>> + *   write value into target register
> >>> + *   format: op subsys address value
> >>> + * CMDQ_CODE_JUMP:
> >>> + *   jump by offset
> >>> + *   format: op offset
> >>> + * CMDQ_CODE_WFE:
> >>> + *   wait for event and clear
> >>> + *   it is just clear if no wait
> >>> + *   format: [wait]  op event update:1 to_wait:1 wait:1
> >>> + *           [clear] op event update:1 to_wait:0 wait:0
> >>> + * CMDQ_CODE_EOC:
> >>> + *   end of command
> >>> + *   format: op irq_flag
> >>> + */
> >>
> >> I think we need more documentation of how this command queue engine is
> >> working. If not, I think it will be really complicated to understand how
> >> to use this.
> >>
> >>> +enum cmdq_code {
> >>> +	CMDQ_CODE_MOVE = 0x02,
> >>> +	CMDQ_CODE_WRITE = 0x04,
> >>> +	CMDQ_CODE_JUMP = 0x10,
> >>> +	CMDQ_CODE_WFE = 0x20,
> >>> +	CMDQ_CODE_EOC = 0x40,
> >>> +};
> >>> +
> >>> +enum cmdq_task_state {
> >>> +	TASK_STATE_BUSY, /* running on a GCE thread */
> >>> +	TASK_STATE_ERROR,
> >>> +	TASK_STATE_DONE,
> >>> +};
> >>> +
> >>> +struct cmdq_task_cb {
> >>> +	cmdq_async_flush_cb	cb;
> >>> +	void			*data;
> >>> +};
> >>> +
> >>> +struct cmdq_thread {
> >>> +	void __iomem		*base;
> >>> +	struct list_head	task_busy_list;
> >>> +	wait_queue_head_t	wait_task_done;
> >>> +};
> >>> +
> >>> +struct cmdq_task {
> >>> +	struct cmdq		*cmdq;
> >>> +	struct list_head	list_entry;
> >>> +	enum cmdq_task_state	task_state;
> >>> +	void			*va_base;
> >>> +	dma_addr_t		pa_base;
> >>> +	u64			engine_flag;
> >>> +	size_t			command_size;
> >>> +	u32			num_cmd;
> >>
> >> num_cmd is directly connected to command_size. I prefer to just keep
> >> num_cmd and calculate the command_size where necessary.
> >
> > After I trace code, I prefer to keep command_size and calculate num_cmd
> > where necessary. What do you think?
> >
> 
> I suppose you prefer this, as you are writing to the GCE depending on 
> the command_size. I think it is worth to create a macro for the 
> calculation of the number of commands, to make the code more readable.
> Would be nice if you would just pass cmdq_task to it and it would return 
> the number. Just as an idea.

Will add macro.

> >>> +	struct cmdq_thread	*thread;
> >>> +	struct cmdq_task_cb	cb;
> >>> +	struct work_struct	release_work;
> >>> +};
> >>> +
> >>> +struct cmdq {
> >>> +	struct device		*dev;
> >>> +	void __iomem		*base;
> >>> +	u32			irq;
> >>> +	struct workqueue_struct	*task_release_wq;
> >>> +	struct cmdq_thread	thread[CMDQ_THR_MAX_COUNT];
> >>> +	struct mutex		task_mutex;	/* for task */
> >>> +	spinlock_t		exec_lock;	/* for exec */
> >>> +	struct clk		*clock;
> >>> +	bool			suspended;
> >>> +};
> >>> +
> >>> +struct cmdq_subsys {
> >>> +	u32	base;
> >>> +	int	id;
> >>> +};
> >>> +
> >>> +static const struct cmdq_subsys gce_subsys[] = {
> >>> +	{0x1400, 1},
> >>> +	{0x1401, 2},
> >>> +	{0x1402, 3},
> >>> +};
> >>> +
> >>> +static int cmdq_subsys_base_to_id(u32 base)
> >>> +{
> >>> +	int i;
> >>> +
> >>> +	for (i = 0; i < ARRAY_SIZE(gce_subsys); i++)
> >>> +		if (gce_subsys[i].base == base)
> >>> +			return gce_subsys[i].id;
> >>> +	return -EFAULT;
> >>> +}
> >>> +
> >>> +static int cmdq_eng_get_thread(u64 flag)
> >>> +{
> >>> +	if (flag & BIT_ULL(CMDQ_ENG_DISP_DSI0))
> >>> +		return CMDQ_THR_DISP_MAIN_IDX;
> >>> +	else if (flag & BIT_ULL(CMDQ_ENG_DISP_DPI0))
> >>> +		return CMDQ_THR_DISP_SUB_IDX;
> >>> +	else
> >>> +		return CMDQ_THR_DISP_MISC_IDX;
> >>> +}
> >>> +
> >>> +static void cmdq_task_release(struct cmdq_task *task)
> >>> +{
> >>> +	struct cmdq *cmdq = task->cmdq;
> >>> +
> >>> +	dma_free_coherent(cmdq->dev, task->command_size, task->va_base,
> >>> +			  task->pa_base);
> >>> +	kfree(task);
> >>> +}
> >>> +
> >>> +static struct cmdq_task *cmdq_task_acquire(struct cmdq_rec *rec,
> >>> +					   struct cmdq_task_cb cb)
> >>> +{
> >>> +	struct cmdq *cmdq = rec->cmdq;
> >>> +	struct device *dev = cmdq->dev;
> >>> +	struct cmdq_task *task;
> >>> +
> >>> +	task = kzalloc(sizeof(*task), GFP_KERNEL);
> >>> +	INIT_LIST_HEAD(&task->list_entry);
> >>> +	task->va_base = dma_alloc_coherent(dev, rec->command_size,
> >>> +					   &task->pa_base, GFP_KERNEL);
> >>> +	if (!task->va_base) {
> >>> +		dev_err(dev, "allocate command buffer failed\n");
> >>> +		kfree(task);
> >>> +		return NULL;
> >>> +	}
> >>> +
> >>> +	task->cmdq = cmdq;
> >>> +	task->command_size = rec->command_size;
> >>> +	task->engine_flag = rec->engine_flag;
> >>> +	task->task_state = TASK_STATE_BUSY;
> >>> +	task->cb = cb;
> >>> +	memcpy(task->va_base, rec->buf, rec->command_size);
> >>> +	task->num_cmd = task->command_size / CMDQ_INST_SIZE;
> >>> +	return task;
> >>> +}
> >>> +
> >>> +static void cmdq_thread_writel(struct cmdq_thread *thread, u32 value,
> >>> +			       u32 offset)
> >>> +{
> >>> +	writel(value, thread->base + offset);
> >>> +}
> >>> +
> >>> +static u32 cmdq_thread_readl(struct cmdq_thread *thread, u32 offset)
> >>> +{
> >>> +	return readl(thread->base + offset);
> >>> +}
> >>
> >> We can get rid of cmdq_thread_readl/writel.
> >
> > Will do.
> >
> >>> +
> >>> +static int cmdq_thread_suspend(struct cmdq *cmdq, struct cmdq_thread *thread)
> >>> +{
> >>> +	u32 status;
> >>> +
> >>> +	cmdq_thread_writel(thread, CMDQ_THR_SUSPEND, CMDQ_THR_SUSPEND_TASK);
> >>> +
> >>> +	/* If already disabled, treat as suspended successful. */
> >>> +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> >>> +	      CMDQ_THR_ENABLED))
> >>> +		return 0;
> >>> +
> >>> +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_CURR_STATUS,
> >>> +			status, status & CMDQ_THR_STATUS_SUSPENDED, 0, 10)) {
> >>> +		dev_err(cmdq->dev, "suspend GCE thread 0x%x failed\n",
> >>> +			(u32)(thread->base - cmdq->base));
> >>> +		return -EFAULT;
> >>> +	}
> >>> +
> >>> +	return 0;
> >>> +}
> >>> +
> >>> +static void cmdq_thread_resume(struct cmdq_thread *thread)
> >>> +{
> >>> +	cmdq_thread_writel(thread, CMDQ_THR_RESUME, CMDQ_THR_SUSPEND_TASK);
> >>> +}
> >>> +
> >>> +static int cmdq_thread_reset(struct cmdq *cmdq, struct cmdq_thread *thread)
> >>> +{
> >>> +	u32 warm_reset;
> >>> +
> >>> +	cmdq_thread_writel(thread, CMDQ_THR_DO_WARM_RESET, CMDQ_THR_WARM_RESET);
> >>> +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_WARM_RESET,
> >>> +			warm_reset, !(warm_reset & CMDQ_THR_DO_WARM_RESET),
> >>> +			0, 10)) {
> >>> +		dev_err(cmdq->dev, "reset GCE thread 0x%x failed\n",
> >>> +			(u32)(thread->base - cmdq->base));
> >>> +		return -EFAULT;
> >>> +	}
> >>> +	writel(CMDQ_THR_ACTIVE_SLOT_CYCLES, cmdq->base + CMDQ_THR_SLOT_CYCLES);
> >>> +	return 0;
> >>> +}
> >>> +
> >>> +static void cmdq_thread_disable(struct cmdq *cmdq, struct cmdq_thread *thread)
> >>> +{
> >>> +	cmdq_thread_reset(cmdq, thread);
> >>> +	cmdq_thread_writel(thread, CMDQ_THR_DISABLED, CMDQ_THR_ENABLE_TASK);
> >>> +}
> >>> +
> >>> +/* notify GCE to re-fetch commands by setting GCE thread PC */
> >>> +static void cmdq_thread_invalidate_fetched_data(struct cmdq_thread *thread)
> >>> +{
> >>> +	cmdq_thread_writel(thread,
> >>> +			   cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR),
> >>> +			   CMDQ_THR_CURR_ADDR);
> >>> +}
> >>> +
> >>> +static void cmdq_task_insert_into_thread(struct cmdq_task *task)
> >>> +{
> >>> +	struct cmdq_thread *thread = task->thread;
> >>> +	struct cmdq_task *prev_task = list_last_entry(
> >>> +			&thread->task_busy_list, typeof(*task), list_entry);
> >>> +	u64 *prev_task_base = prev_task->va_base;
> >>> +
> >>> +	/* let previous task jump to this task */
> >>> +	prev_task_base[prev_task->num_cmd - 1] = (u64)CMDQ_JUMP_BY_PA << 32 |
> >>> +						 task->pa_base;
> >>> +
> >>> +	cmdq_thread_invalidate_fetched_data(thread);
> >>> +}
> >>> +
> >>> +/* we assume tasks in the same display GCE thread are waiting the same event. */
> >>> +static void cmdq_task_remove_wfe(struct cmdq_task *task)
> >>> +{
> >>> +	u32 wfe_option = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
> >>> +	u32 wfe_op = CMDQ_CODE_WFE << CMDQ_OP_CODE_SHIFT;
> >>> +	u32 *base = task->va_base;
> >>> +	u32 num_cmd = task->num_cmd << 1;
> >>> +	int i;
> >>> +
> >>> +	for (i = 0; i < num_cmd; i += 2)
> >>> +		if (base[i] == wfe_option &&
> >>> +		    (base[i + 1] & CMDQ_OP_CODE_MASK) == wfe_op) {
> >>> +			base[i] = CMDQ_JUMP_PASS;
> >>> +			base[i + 1] = CMDQ_JUMP_BY_OFFSET;
> >>> +		}
> >>
> >> After using the command buffer as a void pointer a u64 pointer, we now
> >> reference to it as u32. I would prefer to explain here, how the command
> >> looks like we are searching for and use a for loop passing task->num_cmd
> >> instead.
> >
> > Will use u64* to rewrite the above code.
> >
> >>> +}
> >>> +
> >>> +static void cmdq_task_exec(struct cmdq_task *task, struct cmdq_thread *thread)
> >>> +{
> >>> +	struct cmdq *cmdq = task->cmdq;
> >>> +	unsigned long flags;
> >>> +	unsigned long curr_pa, end_pa;
> >>> +
> >>> +	WARN_ON(clk_prepare_enable(cmdq->clock) < 0);
> >>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
> >>
> >> cmdq_task_exec is called with cmdq->task_mutex held, so why do we need
> >> the spin_lock here? Can't we just use one of the two?
> >
> > We can drop task_mutex, but we will get some side effects.
> > 1. exec_lock needs to include more code, but I think it is not good for
> >     spinlock.
> > 2. In cmdq_rec_flush_async(), task_mutex needs to protect
> >     (1) cmdq->suspended, (2) cmdq_task_exec(), and
> >     (3) cmdq_task_wait_release_schedule().
> >     If we drop task_mutex, we have to put cmdq->suspended if condition
> >     just before cmdq_task_exec() and inside exec_lock, and we have to
> >     release task and its command buffer if error. This will let flow
> >     become more complex and enlarge code size.
> >
> > What do you think?
> 
> Why do you need to protect cmdq_task_wait_release_schedule? We don't 
> care about the order of the workqueue elements, do we?

According to CK's comment, we have to ensure order to remove previous
task.
http://lists.infradead.org/pipermail/linux-mediatek/2016-May/005547.html

> As far as I understand you would need to protect cmdq_task_acquire as 
> well, to "ensure" continously growing pa_base. More on that below.

We need to ensure continuous physical addresses in a task, but we don't
need to ensure continuous physical addresses between tasks.
So, I think it is unnecessary to protect by mutex or spinlock.

> >
> >>> +	task->thread = thread;
> >>> +	task->task_state = TASK_STATE_BUSY;
> >>
> >> That was already set in cmdq_task_acquire, why do we need to set it here
> >> again?
> >
> > Will drop it.
> >
> 
> Yeah, but I think it makes more sense to drop it in cmdq_task_acquire 
> instead.

Will drop it in cmdq_task_acquire instead.

> >>> +	if (list_empty(&thread->task_busy_list)) {
> >>> +		WARN_ON(cmdq_thread_reset(cmdq, thread) < 0);
> >>> +
> >>> +		cmdq_thread_writel(thread, task->pa_base, CMDQ_THR_CURR_ADDR);
> >>> +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
> >>> +				   CMDQ_THR_END_ADDR);
> >>> +		cmdq_thread_writel(thread, CMDQ_THR_PRIORITY, CMDQ_THR_CFG);
> >>> +		cmdq_thread_writel(thread, CMDQ_THR_IRQ_EN,
> >>> +				   CMDQ_THR_IRQ_ENABLE);
> >>> +
> >>> +		cmdq_thread_writel(thread, CMDQ_THR_ENABLED,
> >>> +				   CMDQ_THR_ENABLE_TASK);
> >>> +	} else {
> >>> +		WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
> >>> +
> >>> +		/*
> >>> +		 * check boundary condition
> >>> +		 * PC = END - 8, EOC is executed
> >>> +		 * PC = END, all CMDs are executed
> >>> +		 */
> >>> +		curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> >>> +		end_pa = cmdq_thread_readl(thread, CMDQ_THR_END_ADDR);
> >>> +		if (curr_pa == end_pa - 8 || curr_pa == end_pa) {
> >>
> >> 8 refers to CMDQ_INST_SIZE, right?
> >
> > Yes, I will use CMDQ_INST_SIZE.
> >
> >>> +			/* set to this task directly */
> >>> +			cmdq_thread_writel(thread, task->pa_base,
> >>> +					   CMDQ_THR_CURR_ADDR);
> >>> +		} else {
> >>> +			cmdq_task_insert_into_thread(task);
> >>> +
> >>> +			if (thread == &cmdq->thread[CMDQ_THR_DISP_MAIN_IDX] ||
> >>> +			    thread == &cmdq->thread[CMDQ_THR_DISP_SUB_IDX])
> >>> +				cmdq_task_remove_wfe(task);
> >>
> >> We could do this check using the task->engine_flag, I suppose that's
> >> easier to undestand then.
> >
> > Will use task->engine_flag.
> >
> >>> +
> >>> +			smp_mb(); /* modify jump before enable thread */
> >>> +		}
> >>> +
> >>> +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
> >>> +				   CMDQ_THR_END_ADDR);
> >>> +		cmdq_thread_resume(thread);
> >>> +	}
> >>> +	list_move_tail(&task->list_entry, &thread->task_busy_list);
> >>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>> +}
> >>> +
> >>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
> >>> +				   struct cmdq_thread *thread, u32 irq_flag)
> >>> +{
> >>> +	struct cmdq_task *task, *tmp, *curr_task = NULL;
> >>> +	u32 curr_pa;
> >>> +	struct cmdq_cb_data cmdq_cb_data;
> >>> +	bool err;
> >>> +
> >>> +	if (irq_flag & CMDQ_THR_IRQ_ERROR)
> >>> +		err = true;
> >>> +	else if (irq_flag & CMDQ_THR_IRQ_DONE)
> >>> +		err = false;
> >>> +	else
> >>> +		return;
> >>> +
> >>> +	curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> >>> +
> >>> +	list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
> >>> +				 list_entry) {
> >>> +		if (curr_pa >= task->pa_base &&
> >>> +		    curr_pa < (task->pa_base + task->command_size))
> >>
> >> What are you checking here? It seems as if you make some implcit
> >> assumptions about pa_base and the order of execution of commands in the
> >> thread. Is it save to do so? Does dma_alloc_coherent give any guarantees
> >> about dma_handle?
> >
> > 1. Check what is the current running task in this GCE thread.
> > 2. Yes.
> > 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
> >
> 
> Yes, physical addresses might be continous, but AFAIK there is no 
> guarantee that the dma_handle address is steadily growing, when calling 
> dma_alloc_coherent. And if I understand the code correctly, you use this 
> assumption to decide if the task picked from task_busy_list is currently 
> executing. So I think this mecanism is not working.

I don't use dma_handle address, and just use physical addresses.
From CPU's point of view, tasks are linked by the busy list.
From GCE's point of view, tasks are linked by the JUMP command.

> In which cases does the HW thread raise an interrupt.
> In case of error. When does CMDQ_THR_IRQ_DONE get raised?

GCE will raise interrupt if any task is done or error.
However, GCE is fast, so CPU may get multiple done tasks
when it is running ISR.

In case of error, that GCE thread will pause and raise interrupt.
So, CPU may get multiple done tasks and one error task.

> >>> +			curr_task = task;
> >>> +		if (task->cb.cb) {
> >>> +			cmdq_cb_data.err = curr_task ? err : false;
> >>> +			cmdq_cb_data.data = task->cb.data;
> >>> +			task->cb.cb(cmdq_cb_data);
> >>> +		}
> >>> +		task->task_state = (curr_task && err) ? TASK_STATE_ERROR :
> >>> +				TASK_STATE_DONE;
> >>> +		list_del(&task->list_entry);
> >>> +		if (curr_task)
> >>> +			break;
> >>> +	}
> >>> +
> >>> +	wake_up(&thread->wait_task_done);
> >>> +}
> >>> +
> >>> +static void cmdq_thread_irq_handler(struct cmdq *cmdq, int tid)
> >>> +{
> >>> +	struct cmdq_thread *thread = &cmdq->thread[tid];
> >>> +	unsigned long flags = 0L;
> >>> +	u32 irq_flag;
> >>> +
> >>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
> >>> +
> >>> +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
> >>> +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
> >>> +
> >>> +	/*
> >>> +	 * Another CPU core could run "release task" right before we acquire
> >>> +	 * the spin lock, and thus reset / disable this GCE thread, so we
> >>> +	 * need to check the enable bit of this GCE thread.
> >>> +	 */
> >>> +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> >>> +	      CMDQ_THR_ENABLED))
> >>> +		irq_flag = 0;
> >>
> >> cmdq_handle_error_done just retuns in this case. Programming this way
> >> just makes things confusing. What about:
> >>
> >> if (cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> >> 	      CMDQ_THR_ENABLED)
> >> 	cmdq_handle_error_done(cmdq, thread, irq_flag);
> >> else
> >> 		irq_flag = 0;
> >>
> >> spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >
> > We still need to clear irq_flag if GCE thread is disabled.
> > So, I think we can just return here.
> >
> > 	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> > 	      CMDQ_THR_ENABLED))
> > 		return;
> >
> > What do you think?
> >
> 
> No, you can't just return, you need to unlock the spinlock.
> Anyway I would prefer it the other way round, as I put it in my last 
> mail. Just delete the else branch, we don't need to set irq_flag to zero.

Sorry for my previous wrong reply since I merge your comment
and CK's comment.
http://lists.infradead.org/pipermail/linux-mediatek/2016-May/005547.html
So, I will put this if condition into cmdq_task_handle_error_result()
and then just return it if GCE thread is disabled.

> >>> +
> >>> +	cmdq_handle_error_done(cmdq, thread, irq_flag);
> >>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>> +}
> >>> +
> >>> +static irqreturn_t cmdq_irq_handler(int irq, void *dev)
> >>> +{
> >>> +	struct cmdq *cmdq = dev;
> >>> +	u32 irq_status;
> >>> +	int i;
> >>> +
> >>> +	irq_status = readl(cmdq->base + CMDQ_CURR_IRQ_STATUS);
> >>> +	irq_status &= CMDQ_IRQ_MASK;
> >>> +	irq_status ^= CMDQ_IRQ_MASK;
> >>
> >> irq_status can be much bigger then 3, which is the number of threads in
> >> the system (CMDQ_THR_MAX_COUNT). So why we use this mask here isn't
> >> clear to me.
> >
> > Our GCE hardware has 16 threads, but we only use 3 threads currently.
> >
> 
> Ok, but please use bitops here.

Will use bitops.

> >>> +
> >>> +	if (!irq_status)
> >>> +		return IRQ_NONE;
> >>> +
> >>> +	while (irq_status) {
> >>> +		i = ffs(irq_status) - 1;
> >>> +		irq_status &= ~BIT(i);
> >>> +		cmdq_thread_irq_handler(cmdq, i);
> >>> +	}
> >>
> >> Can you explain how the irq status register looks like, that would it
> >> make much easier to understand what happens here.
> >
> > Bit 0 ~ 15 of irq status register represents GCE thread 0 ~ 15
> > interrupt. 0 means asserting interrupt; 1 means no interrupt.
> >
> 
> Thanks, that helped. :)
> 
> >>> +
> >>> +	return IRQ_HANDLED;	
> >>> +}
> >>> +
> >>> +static int cmdq_task_handle_error_result(struct cmdq_task *task)
> >>
> >> We never check the return values, why do we have them?
> >
> > Will drop return value.
> >
> >>> +{
> >>> +	struct cmdq *cmdq = task->cmdq;
> >>> +	struct device *dev = cmdq->dev;
> >>> +	struct cmdq_thread *thread = task->thread;
> >>> +	struct cmdq_task *next_task, *prev_task;
> >>> +	u32 irq_flag;
> >>> +
> >>> +	/* suspend GCE thread to ensure consistency */
> >>> +	WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
> >>> +
> >>> +	/* ISR has handled this error task */
> >>> +	if (task->task_state == TASK_STATE_ERROR) {
> >>> +		next_task = list_first_entry_or_null(&thread->task_busy_list,
> >>> +				struct cmdq_task, list_entry);
> >>> +		if (next_task) /* move to next task */
> >>> +			cmdq_thread_writel(thread, next_task->pa_base,
> >>> +					   CMDQ_THR_CURR_ADDR);
> >>
> >> We have to do this, as we suppose that the thread did not reach the jump
> >> instruction we put into it's command queue, right?
> >
> > Yes.
> >
> 
> So this should then go into it's own function. In wait_release_work, 
> something like this:
> 
> if(task->task_state == TASK_STATE_ERROR)
>       cmdq_task_handle_error(task)

OK.
I will write new function cmdq_task_handle_error() to handle error case.

> >>> +		cmdq_thread_resume(thread);
> >>> +		return -ECANCELED;
> >>> +	}
> >>> +
> >>
> >> if task_state != ERROR and != DONE. This means that the timeout of
> >> task_release_wq has timed out, right?
> >
> > Yes.
> >
> >>> +	/*
> >>> +	 * Save next_task and prev_task in advance
> >>> +	 * since cmdq_handle_error_done will remove list_entry.
> >>> +	 */
> >>> +	next_task = prev_task = NULL;
> >>> +	if (task->list_entry.next != &thread->task_busy_list)
> >>> +		next_task = list_next_entry(task, list_entry);
> >>> +	if (task->list_entry.prev != &thread->task_busy_list)
> >>> +		prev_task = list_prev_entry(task, list_entry);
> >>> +
> >>> +	/*
> >>> +	 * Although IRQ is disabled, GCE continues to execute.
> >>> +	 * It may have pending IRQ before GCE thread is suspended,
> >>> +	 * so check this condition again.
> >>> +	 */
> >>
> >> The first thing we did in this function was suspending the thread. Why
> >> do we need this then?
> >
> > Because timeout is CPU timeout not GCE timeout, GCE could just finish
> > this task before the GCE thread is suspended.
> >
> 
> What are the reasons for a timeout? An error has happend, or the task is 
> still executing.

From GCE's point of view, this task is still executing.
But, it could be an error of client.
For example, task may never get event if display turn off hardware
before waiting for task to finish its work.

> >> To be honest this whole functions looks really like a design error. We
> >> have to sperate the states much clearer so that it is possible to
> >> understand what is happening in the GCE. Isn't it for example posible to
> >> have worker queues for timed out tasks and tasks with an error? I'm not
> >> sure how to do this, actually I'm not sure if I really understood how
> >> this is supposed to work.
> >
> > GCE doesn't have timeout. The timeout is decided and controlled by CPU,
> > so we check timeout in release work.
> > For error and done, they are easy to check by register, and we have
> > already created release work for timeout. So, I don't think we need to
> > create work queue for each case.
> >
> > What do you think?
> >
> 
> I think, if we find in here, that the irq_flag is set, then the the 
> interrupt handler was triggered and is spinning the spinlock. If this is 
> not the case, we have a timeout and we handle this.

I will write another function to handle error, and handle timeout here.

> >> How much do we win, when we patch the thread command queue for every
> >> task we add, instead of just taking one task after another from the
> >> task_busy_list?
> >
> > GCE is used to help read/write registers with critical time limitation.
> > Sometimes, client may ask to process multiple tasks in a short period
> > of time, e.g. display flush multiple tasks for next vblank. So, CMDQ
> > shouldn't limit to process one task after another from the
> > task_busy_list. Currently, when interrupt or timeout, we will check
> > how many tasks are done, and which one is error or timeout.
> >
> 
> So I suppose the device driver who use this are interested in throughput 
> and not in latency. The callback of every
> 
> >>> +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
> >>> +	cmdq_handle_error_done(cmdq, thread, irq_flag);
> >>> +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
> >>> +
> >>> +	if (task->task_state == TASK_STATE_DONE) {
> >>> +		cmdq_thread_resume(thread);
> >>> +		return 0;
> >>> +	}
> >>> +
> >>> +	if (task->task_state == TASK_STATE_ERROR) {
> >>> +		dev_err(dev, "task 0x%p error\n", task);
> >>> +		if (next_task) /* move to next task */
> >>> +			cmdq_thread_writel(thread, next_task->pa_base,
> >>> +					   CMDQ_THR_CURR_ADDR);
> >>> +		cmdq_thread_resume(thread);
> >>> +		return -ECANCELED;
> >>> +	}
> >>> +
> >>> +	/* Task is running, so we force to remove it. */
> >>> +	dev_err(dev, "task 0x%p timeout or killed\n", task);
> >>> +	task->task_state = TASK_STATE_ERROR;
> >>> +
> >>> +	if (prev_task) {
> >>> +		u64 *prev_va = prev_task->va_base;
> >>> +		u64 *curr_va = task->va_base;
> >>> +
> >>> +		/* copy JUMP instruction */
> >>> +		prev_va[prev_task->num_cmd - 1] = curr_va[task->num_cmd - 1];
> >>> +
> >>> +		cmdq_thread_invalidate_fetched_data(thread);
> >>> +	} else if (next_task) { /* move to next task */
> >>> +		cmdq_thread_writel(thread, next_task->pa_base,
> >>> +				   CMDQ_THR_CURR_ADDR);
> >>> +	}
> >>> +
> >>> +	list_del(&task->list_entry);
> >>> +	cmdq_thread_resume(thread);
> >>> +
> >>> +	/* call cb here to prevent lock */
> >>> +	if (task->cb.cb) {
> >>> +		struct cmdq_cb_data cmdq_cb_data;
> >>> +
> >>> +		cmdq_cb_data.err = true;
> >>> +		cmdq_cb_data.data = task->cb.data;
> >>> +		task->cb.cb(cmdq_cb_data);
> >>> +	}
> >>> +
> >>> +	return -ECANCELED;
> >>> +}
> >>> +
> >>> +static void cmdq_task_wait_release_work(struct work_struct *work_item)
> >>> +{
> >>> +	struct cmdq_task *task = container_of(work_item, struct cmdq_task,
> >>> +					      release_work);
> >>> +	struct cmdq *cmdq = task->cmdq;
> >>> +	struct cmdq_thread *thread = task->thread;
> >>> +	unsigned long flags;
> >>> +
> >>> +	wait_event_timeout(thread->wait_task_done,
> >>> +			   task->task_state != TASK_STATE_BUSY,
> >>> +			   msecs_to_jiffies(CMDQ_TIMEOUT_MS));
> >>> +
> >>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
> >>> +	if (task->task_state != TASK_STATE_DONE)
> >>> +		cmdq_task_handle_error_result(task);
> >>> +	if (list_empty(&thread->task_busy_list))
> >>> +		cmdq_thread_disable(cmdq, thread);
> >>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>> +
> >>> +	/* release regardless of success or not */
> >>> +	clk_disable_unprepare(cmdq->clock);
> >>> +	cmdq_task_release(task);
> >>> +}
> >>> +
> >>> +static void cmdq_task_wait_release_schedule(struct cmdq_task *task)
> >>> +{
> >>> +	struct cmdq *cmdq = task->cmdq;
> >>> +
> >>> +	INIT_WORK(&task->release_work, cmdq_task_wait_release_work);
> >>> +	queue_work(cmdq->task_release_wq, &task->release_work);
> >>> +}
> >>> +
> >>> +static int cmdq_rec_realloc_cmd_buffer(struct cmdq_rec *rec, size_t size)
> >>> +{
> >>> +	void *new_buf;
> >>> +
> >>> +	new_buf = krealloc(rec->buf, size, GFP_KERNEL | __GFP_ZERO);
> >>> +	if (!new_buf)
> >>> +		return -ENOMEM;
> >>> +	rec->buf = new_buf;
> >>> +	rec->buf_size = size;
> >>> +	return 0;
> >>> +}
> >>> +
> >>> +struct cmdq_base *cmdq_register_device(struct device *dev)
> >>> +{
> >>> +	struct cmdq_base *cmdq_base;
> >>> +	struct resource res;
> >>> +	int subsys;
> >>> +	u32 base;
> >>> +
> >>> +	if (of_address_to_resource(dev->of_node, 0, &res))
> >>> +		return NULL;
> >>> +	base = (u32)res.start;
> >>> +
> >>> +	subsys = cmdq_subsys_base_to_id(base >> 16);
> >>> +	if (subsys < 0)
> >>> +		return NULL;
> >>> +
> >>> +	cmdq_base = devm_kmalloc(dev, sizeof(*cmdq_base), GFP_KERNEL);
> >>> +	if (!cmdq_base)
> >>> +		return NULL;
> >>> +	cmdq_base->subsys = subsys;
> >>> +	cmdq_base->base = base;
> >>> +
> >>> +	return cmdq_base;
> >>> +}
> >>> +EXPORT_SYMBOL(cmdq_register_device);
> >>> +
> >>> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
> >>> +		    struct cmdq_rec **rec_ptr)
> >>> +{
> >>> +	struct cmdq_rec *rec;
> >>> +	int err;
> >>> +
> >>> +	rec = kzalloc(sizeof(*rec), GFP_KERNEL);
> >>> +	if (!rec)
> >>> +		return -ENOMEM;
> >>> +	rec->cmdq = dev_get_drvdata(dev);
> >>> +	rec->engine_flag = engine_flag;
> >>> +	err = cmdq_rec_realloc_cmd_buffer(rec, CMDQ_INITIAL_CMD_BLOCK_SIZE);
> >>
> >> Just pass PAGE_SIZE here, no need for CMDQ_INITIAL_CMD_BLOCK_SIZE.
> >
> > Will do.
> >
> >>> +	if (err < 0) {
> >>> +		kfree(rec);
> >>> +		return err;
> >>> +	}
> >>> +	*rec_ptr = rec;
> >>> +	return 0;
> >>> +}
> >>> +EXPORT_SYMBOL(cmdq_rec_create);
> >>> +
> >>> +static int cmdq_rec_append_command(struct cmdq_rec *rec, enum cmdq_code code,
> >>> +				   u32 arg_a, u32 arg_b)
> >>> +{
> >>> +	u64 *cmd_ptr;
> >>> +	int err;
> >>> +
> >>> +	if (WARN_ON(rec->finalized))
> >>> +		return -EBUSY;
> >>> +	if (code < CMDQ_CODE_MOVE || code > CMDQ_CODE_EOC)
> >>> +		return -EINVAL;
> >>
> >> cmdq_rec_append_command is just called from inside this driver and code
> >> is a enum. We can expect it to be correct, no need for this check.
> >
> > Will drop this check.
> >
> >>> +	if (unlikely(rec->command_size + CMDQ_INST_SIZE > rec->buf_size)) {
> >>
> >> command_size is the offset into the buffer to which a new command is
> >> written, so this name is highly confusing. I wonder if this would be
> >> easier to understand if we redefine command_size to something like the
> >> number of commands and divide/multiply CMDQ_INST_SIZE where this is needed.
> >
> > I can rename command_size to cmd_buf_size and calculate num_cmd by
> > dividing CMDQ_INST_SIZE.
> > What do you think?
> >
> >>> +		err = cmdq_rec_realloc_cmd_buffer(rec, rec->buf_size * 2);
> >>> +		if (err < 0)
> >>> +			return err;
> >>> +	}
> >>> +	cmd_ptr = rec->buf + rec->command_size;
> >>> +	(*cmd_ptr) = (u64)((code << CMDQ_OP_CODE_SHIFT) | arg_a) << 32 | arg_b;
> >>> +	rec->command_size += CMDQ_INST_SIZE;
> >>> +	return 0;
> >>> +}
> >>> +
> >>> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value, struct cmdq_base *base,
> >>> +		   u32 offset)
> >>> +{
> >>> +	u32 arg_a = ((base->base + offset) & CMDQ_ARG_A_WRITE_MASK) |
> >>> +		    ((base->subsys & CMDQ_SUBSYS_MASK) << CMDQ_SUBSYS_SHIFT);
> >>
> >> base->subsys is the id from gce_sybsys, so we can expect it to be
> >> correct, no need to mask with CMDQ_SUBSYS_MASK.
> >
> > Will drop it.
> >
> >>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WRITE, arg_a, value);
> >>> +}
> >>> +EXPORT_SYMBOL(cmdq_rec_write);
> >>> +
> >>> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
> >>> +			struct cmdq_base *base, u32 offset, u32 mask)
> >>> +{
> >>> +	u32 offset_mask = offset;
> >>> +	int err;
> >>> +
> >>> +	if (mask != 0xffffffff) {
> >>> +		err = cmdq_rec_append_command(rec, CMDQ_CODE_MOVE, 0, ~mask);
> >>> +		if (err < 0)
> >>> +			return err;
> >>> +		offset_mask |= CMDQ_WRITE_ENABLE_MASK;
> >>> +	}
> >>> +	return cmdq_rec_write(rec, value, base, offset_mask);
> >>> +}
> >>> +EXPORT_SYMBOL(cmdq_rec_write_mask);
> >>> +
> >>> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event)
> >>> +{
> >>> +	u32 arg_b;
> >>> +
> >>> +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
> >>> +		return -EINVAL;
> >>> +
> >>> +	/*
> >>> +	 * bit 0-11: wait value
> >>> +	 * bit 15: 1 - wait, 0 - no wait
> >>> +	 * bit 16-27: update value
> >>> +	 * bit 31: 1 - update, 0 - no update
> >>> +	 */
> >>
> >> I don't understand this comments. What are they for?
> >
> > This is for WFE command. I will comment it.
> >
> >>> +	arg_b = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
> >>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event, arg_b);
> >>> +}
> >>> +EXPORT_SYMBOL(cmdq_rec_wfe);
> >>> +
> >>> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event)
> >>> +{
> >>> +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
> >>> +		return -EINVAL;
> >>> +
> >>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event,
> >>> +				       CMDQ_WFE_UPDATE);
> >>> +}
> >>> +EXPORT_SYMBOL(cmdq_rec_clear_event);
> >>> +
> >>> +static int cmdq_rec_finalize(struct cmdq_rec *rec)
> >>> +{
> >>> +	int err;
> >>> +
> >>> +	if (rec->finalized)
> >>> +		return 0;
> >>> +
> >>> +	/* insert EOC and generate IRQ for each command iteration */
> >>> +	err = cmdq_rec_append_command(rec, CMDQ_CODE_EOC, 0, CMDQ_EOC_IRQ_EN);
> >>> +	if (err < 0)
> >>> +		return err;
> >>> +
> >>> +	/* JUMP to end */
> >>> +	err = cmdq_rec_append_command(rec, CMDQ_CODE_JUMP, 0, CMDQ_JUMP_PASS);
> >>> +	if (err < 0)
> >>> +		return err;
> >>> +
> >>
> >> Does this need to be atomic?
> >> What happens if after CODE_EOC and before CODE_JUMP some
> >> write/read/event gets added?
> >> What happens if more commands get added to the queue after CODE_JUMP,
> >> but before finalized is set to true. Why don't you use atomic functions
> >> to access finalized?
> >
> > Since cmdq_rec doesn't guarantee thread safe, mutex is needed when
> > client uses cmdq_rec.
> >
> 
> Well I think that rec->finalized tries to implement this, but might 
> fail, if two kernel threads work on the same cmdq_rec.
> 
> >>> +	rec->finalized = true;
> >>> +	return 0;
> >>> +}
> >>> +
> >>> +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
> >>> +			 void *data)
> >>> +{
> >>> +	struct cmdq *cmdq = rec->cmdq;
> >>> +	struct cmdq_task *task;
> >>> +	struct cmdq_task_cb task_cb;
> >>> +	struct cmdq_thread *thread;
> >>> +	int err;
> >>> +
> >>> +	mutex_lock(&cmdq->task_mutex);
> >>> +	if (cmdq->suspended) {
> >>> +		dev_err(cmdq->dev, "%s is called after suspended\n", __func__);
> >>> +		mutex_unlock(&cmdq->task_mutex);
> >>> +		return -EPERM;
> >>> +	}
> >>> +
> >>> +	err = cmdq_rec_finalize(rec);
> >>> +	if (err < 0) {
> >>> +		mutex_unlock(&cmdq->task_mutex);
> >>> +		return err;
> >>> +	}
> >>> +
> >>> +	task_cb.cb = cb;
> >>> +	task_cb.data = data;
> >>> +	task = cmdq_task_acquire(rec, task_cb);
> >>> +	if (!task) {
> >>> +		mutex_unlock(&cmdq->task_mutex);
> >>> +		return -EFAULT;
> >>> +	}
> >>> +
> >>> +	thread = &cmdq->thread[cmdq_eng_get_thread(task->engine_flag)];
> >>> +	cmdq_task_exec(task, thread);
> >>> +	cmdq_task_wait_release_schedule(task);
> >>> +	mutex_unlock(&cmdq->task_mutex);
> >>> +	return 0;
> >>> +}
> >>> +EXPORT_SYMBOL(cmdq_rec_flush_async);
> >>> +
> >>> +struct cmdq_flush_completion {
> >>> +	struct completion cmplt;
> >>> +	bool err;
> >>> +};
> >>> +
> >>> +static int cmdq_rec_flush_cb(struct cmdq_cb_data data)
> >>> +{
> >>> +	struct cmdq_flush_completion *cmplt = data.data;
> >>> +
> >>> +	cmplt->err = data.err;
> >>> +	complete(&cmplt->cmplt);
> >>> +	return 0;
> >>> +}
> >>> +
> >>> +int cmdq_rec_flush(struct cmdq_rec *rec)
> >>> +{
> >>> +	struct cmdq_flush_completion cmplt;
> >>> +	int err;
> >>> +
> >>> +	init_completion(&cmplt.cmplt);
> >>> +	err = cmdq_rec_flush_async(rec, cmdq_rec_flush_cb, &cmplt);
> >>> +	if (err < 0)
> >>> +		return err;
> >>> +	wait_for_completion(&cmplt.cmplt);
> >>> +	return cmplt.err ? -EFAULT : 0;
> >>> +}
> >>> +EXPORT_SYMBOL(cmdq_rec_flush);
> >>> +
> >>> +void cmdq_rec_destroy(struct cmdq_rec *rec)
> >>> +{
> >>> +	kfree(rec->buf);
> >>> +	kfree(rec);
> >>> +}
> >>> +EXPORT_SYMBOL(cmdq_rec_destroy);
> >>> +
> >>> +static bool cmdq_task_is_empty(struct cmdq *cmdq)
> >>> +{
> >>> +	struct cmdq_thread *thread;
> >>> +	int i;
> >>> +
> >>> +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
> >>> +		thread = &cmdq->thread[i];
> >>> +		if (!list_empty(&thread->task_busy_list))
> >>> +			return false;
> >>> +	}
> >>> +	return true;
> >>> +}
> >>> +
> >>> +static int cmdq_suspend(struct device *dev)
> >>> +{
> >>> +	struct cmdq *cmdq = dev_get_drvdata(dev);
> >>> +	u32 exec_threads;
> >>> +
> >>> +	mutex_lock(&cmdq->task_mutex);
> >>> +	cmdq->suspended = true;
> >>> +	mutex_unlock(&cmdq->task_mutex);
> >>> +
> >>> +	exec_threads = readl(cmdq->base + CMDQ_CURR_LOADED_THR);
> >>> +	if ((exec_threads & CMDQ_THR_EXECUTING) && !cmdq_task_is_empty(cmdq)) {
> >>> +		dev_err(dev, "wait active tasks timeout.\n");
> >>> +		flush_workqueue(cmdq->task_release_wq);
> >>> +	}
> >>> +	return 0;
> >>> +}
> >>> +
> >>> +static int cmdq_resume(struct device *dev)
> >>> +{
> >>> +	struct cmdq *cmdq = dev_get_drvdata(dev);
> >>> +
> >>> +	cmdq->suspended = false;
> >>> +	return 0;
> >>> +}
> >>> +
> >>> +static int cmdq_remove(struct platform_device *pdev)
> >>> +{
> >>> +	struct cmdq *cmdq = platform_get_drvdata(pdev);
> >>> +
> >>> +	destroy_workqueue(cmdq->task_release_wq);
> >>> +	cmdq->task_release_wq = NULL;
> >>> +	return 0;
> >>> +}
> >>> +
> >>> +static int cmdq_probe(struct platform_device *pdev)
> >>> +{
> >>> +	struct device *dev = &pdev->dev;
> >>> +	struct device_node *node = dev->of_node;
> >>> +	struct resource *res;
> >>> +	struct cmdq *cmdq;
> >>> +	int err, i;
> >>> +
> >>> +	cmdq = devm_kzalloc(dev, sizeof(*cmdq), GFP_KERNEL);
> >>> +	if (!cmdq)
> >>> +		return -ENOMEM;
> >>> +	cmdq->dev = dev;
> >>> +
> >>> +	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> >>> +	cmdq->base = devm_ioremap_resource(dev, res);
> >>> +	if (IS_ERR(cmdq->base)) {
> >>> +		dev_err(dev, "failed to ioremap gce\n");
> >>> +		return PTR_ERR(cmdq->base);
> >>> +	}
> >>> +
> >>> +	cmdq->irq = irq_of_parse_and_map(node, 0);
> >>> +	if (!cmdq->irq) {
> >>> +		dev_err(dev, "failed to get irq\n");
> >>> +		return -EINVAL;
> >>> +	}
> >>> +
> >>> +	dev_dbg(dev, "cmdq device: addr:0x%p, va:0x%p, irq:%d\n",
> >>> +		dev, cmdq->base, cmdq->irq);
> >>> +
> >>> +	mutex_init(&cmdq->task_mutex);
> >>> +	spin_lock_init(&cmdq->exec_lock);
> >>> +	cmdq->task_release_wq = alloc_ordered_workqueue(
> >>> +			"%s", WQ_MEM_RECLAIM | WQ_HIGHPRI,
> >>> +			"cmdq_task_wait_release");
> >>> +
> >>> +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
> >>> +		cmdq->thread[i].base = cmdq->base + CMDQ_THR_BASE +
> >>> +				CMDQ_THR_SHIFT * i;
> >>> +		init_waitqueue_head(&cmdq->thread[i].wait_task_done);
> >>> +		INIT_LIST_HEAD(&cmdq->thread[i].task_busy_list);
> >>> +	}
> >>> +
> >>> +	platform_set_drvdata(pdev, cmdq);
> >>> +
> >>> +	err = devm_request_irq(dev, cmdq->irq, cmdq_irq_handler, IRQF_SHARED,
> >>> +			       CMDQ_DRIVER_DEVICE_NAME, cmdq);
> >>> +	if (err < 0) {
> >>> +		dev_err(dev, "failed to register ISR (%d)\n", err);
> >>> +		goto fail;
> >>> +	}
> >>> +
> >>> +	cmdq->clock = devm_clk_get(dev, CMDQ_CLK_NAME);
> >>> +	if (IS_ERR(cmdq->clock)) {
> >>> +		dev_err(dev, "failed to get clk:%s\n", CMDQ_CLK_NAME);
> >>> +		err = PTR_ERR(cmdq->clock);
> >>> +		goto fail;
> >>> +	}
> >>> +	return 0;
> >>> +
> >>> +fail:
> >>> +	cmdq_remove(pdev);
> >>> +	return err;
> >>> +}
> >>> +
> >>> +static const struct dev_pm_ops cmdq_pm_ops = {
> >>> +	.suspend = cmdq_suspend,
> >>> +	.resume = cmdq_resume,
> >>> +};
> >>> +
> >>> +static const struct of_device_id cmdq_of_ids[] = {
> >>> +	{.compatible = "mediatek,mt8173-gce",},
> >>> +	{}
> >>> +};
> >>> +
> >>> +static struct platform_driver cmdq_drv = {
> >>> +	.probe = cmdq_probe,
> >>> +	.remove = cmdq_remove,
> >>> +	.driver = {
> >>> +		.name = CMDQ_DRIVER_DEVICE_NAME,
> >>> +		.owner = THIS_MODULE,
> >>> +		.pm = &cmdq_pm_ops,
> >>> +		.of_match_table = cmdq_of_ids,
> >>> +	}
> >>> +};
> >>> +
> >>> +builtin_platform_driver(cmdq_drv);
> >>> diff --git a/include/soc/mediatek/cmdq.h b/include/soc/mediatek/cmdq.h
> >>> new file mode 100644
> >>> index 0000000..60eef3d
> >>> --- /dev/null
> >>> +++ b/include/soc/mediatek/cmdq.h
> >>> @@ -0,0 +1,197 @@
> >>> +/*
> >>> + * Copyright (c) 2015 MediaTek Inc.
> >>> + *
> >>> + * This program is free software; you can redistribute it and/or modify
> >>> + * it under the terms of the GNU General Public License version 2 as
> >>> + * published by the Free Software Foundation.
> >>> + *
> >>> + * This program is distributed in the hope that it will be useful,
> >>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> >>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> >>> + * GNU General Public License for more details.
> >>> + */
> >>> +
> >>> +#ifndef __MTK_CMDQ_H__
> >>> +#define __MTK_CMDQ_H__
> >>> +
> >>> +#include <linux/platform_device.h>
> >>> +#include <linux/types.h>
> >>> +
> >>> +enum cmdq_eng {
> >>> +	CMDQ_ENG_DISP_AAL,
> >>> +	CMDQ_ENG_DISP_COLOR0,
> >>> +	CMDQ_ENG_DISP_COLOR1,
> >>> +	CMDQ_ENG_DISP_DPI0,
> >>> +	CMDQ_ENG_DISP_DSI0,
> >>> +	CMDQ_ENG_DISP_DSI1,
> >>> +	CMDQ_ENG_DISP_GAMMA,
> >>> +	CMDQ_ENG_DISP_OD,
> >>> +	CMDQ_ENG_DISP_OVL0,
> >>> +	CMDQ_ENG_DISP_OVL1,
> >>> +	CMDQ_ENG_DISP_PWM0,
> >>> +	CMDQ_ENG_DISP_PWM1,
> >>> +	CMDQ_ENG_DISP_RDMA0,
> >>> +	CMDQ_ENG_DISP_RDMA1,
> >>> +	CMDQ_ENG_DISP_RDMA2,
> >>> +	CMDQ_ENG_DISP_UFOE,
> >>> +	CMDQ_ENG_DISP_WDMA0,
> >>> +	CMDQ_ENG_DISP_WDMA1,
> >>> +	CMDQ_ENG_MAX,
> >>> +};
> >>> +
> >>> +/* events for CMDQ and display */
> >>> +enum cmdq_event {
> >>> +	/* Display start of frame(SOF) events */
> >>> +	CMDQ_EVENT_DISP_OVL0_SOF = 11,
> >>> +	CMDQ_EVENT_DISP_OVL1_SOF = 12,
> >>> +	CMDQ_EVENT_DISP_RDMA0_SOF = 13,
> >>> +	CMDQ_EVENT_DISP_RDMA1_SOF = 14,
> >>> +	CMDQ_EVENT_DISP_RDMA2_SOF = 15,
> >>> +	CMDQ_EVENT_DISP_WDMA0_SOF = 16,
> >>> +	CMDQ_EVENT_DISP_WDMA1_SOF = 17,
> >>> +	/* Display end of frame(EOF) events */
> >>> +	CMDQ_EVENT_DISP_OVL0_EOF = 39,
> >>> +	CMDQ_EVENT_DISP_OVL1_EOF = 40,
> >>> +	CMDQ_EVENT_DISP_RDMA0_EOF = 41,
> >>> +	CMDQ_EVENT_DISP_RDMA1_EOF = 42,
> >>> +	CMDQ_EVENT_DISP_RDMA2_EOF = 43,
> >>> +	CMDQ_EVENT_DISP_WDMA0_EOF = 44,
> >>> +	CMDQ_EVENT_DISP_WDMA1_EOF = 45,
> >>> +	/* Mutex end of frame(EOF) events */
> >>> +	CMDQ_EVENT_MUTEX0_STREAM_EOF = 53,
> >>> +	CMDQ_EVENT_MUTEX1_STREAM_EOF = 54,
> >>> +	CMDQ_EVENT_MUTEX2_STREAM_EOF = 55,
> >>> +	CMDQ_EVENT_MUTEX3_STREAM_EOF = 56,
> >>> +	CMDQ_EVENT_MUTEX4_STREAM_EOF = 57,
> >>> +	/* Display underrun events */
> >>> +	CMDQ_EVENT_DISP_RDMA0_UNDERRUN = 63,
> >>> +	CMDQ_EVENT_DISP_RDMA1_UNDERRUN = 64,
> >>> +	CMDQ_EVENT_DISP_RDMA2_UNDERRUN = 65,
> >>> +	/* Keep this at the end of HW events */
> >>> +	CMDQ_MAX_HW_EVENT_COUNT = 260,
> >>> +};
> >>> +
> >>> +struct cmdq_cb_data {
> >>> +	bool	err;
> >>> +	void	*data;
> >>> +};
> >>> +
> >>> +typedef int (*cmdq_async_flush_cb)(struct cmdq_cb_data data);
> >>> +
> >>> +struct cmdq_task;
> >>> +struct cmdq;
> >>> +
> >>> +struct cmdq_rec {
> >>> +	struct cmdq		*cmdq;
> >>> +	u64			engine_flag;
> >>> +	size_t			command_size;
> >>> +	void			*buf;
> >>> +	size_t			buf_size;
> >>> +	bool			finalized;
> >>> +};
> 
> Why do we need cmdq_rec at all? Can't we just use the cmdq_task for that 
> and this way make the driver less complex?

There are two main reasons for cmdq_rec.
1. It is slow to access dma too frequently.
   So, we append commands to cacheable memory, and then flush to dma.
2. cmdq_rec is not thread safe, but cmdq_task needs thread safe.
   If we merge them, we need to take care of some synchronization
   issues.

> >>> +
> >>> +struct cmdq_base {
> >>> +	int	subsys;
> >>> +	u32	base;
> >>
> >> subsys can always be calculated via cmdq_subsys_base_to_id(base >> 16)
> >> so we can get rid of the struct, right?
> >
> > Current subsys method is based on previous comment from Daniel Kurtz.
> > Please take a look of our previous discussion.
> > http://lists.infradead.org/pipermail/linux-mediatek/2016-February/004483.html
> > Thanks.
> >
> 
> I have to look deeper into this, but from what I read, the proposal from 
> Daniel [1] seems good to me.
> 
> [1] https://patchwork.kernel.org/patch/8068311/

The initial proposal has some problem, so please see the follow-up
discussions. Thanks.
http://lists.infradead.org/pipermail/linux-mediatek/2016-February/003972.html
http://lists.infradead.org/pipermail/linux-mediatek/2016-February/004483.html


> >>> +};
> >>> +
> >>> +/**
> >>> + * cmdq_register_device() - register device which needs CMDQ
> >>> + * @dev:		device
> >>> + *
> >>> + * Return: cmdq_base pointer or NULL for failed
> >>> + */
> >>> +struct cmdq_base *cmdq_register_device(struct device *dev);
> >>> +
> >>> +/**
> >>> + * cmdq_rec_create() - create command queue record
> >>> + * @dev:		device
> >>> + * @engine_flag:	command queue engine flag
> >>> + * @rec_ptr:		command queue record pointer to retrieve cmdq_rec
> >>> + *
> >>> + * Return: 0 for success; else the error code is returned
> >>> + */
> >>> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
> >>> +		    struct cmdq_rec **rec_ptr);
> >>> +
> >>> +/**
> >>> + * cmdq_rec_write() - append write command to the command queue record
> >>> + * @rec:	the command queue record
> >>> + * @value:	the specified target register value
> >>> + * @base:	the command queue base
> >>> + * @offset:	register offset from module base
> >>> + *
> >>> + * Return: 0 for success; else the error code is returned
> >>> + */
> >>> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value,
> >>> +		   struct cmdq_base *base, u32 offset);
> >>> +
> >>> +/**
> >>> + * cmdq_rec_write_mask() - append write command with mask to the command
> >>> + *			   queue record
> >>> + * @rec:	the command queue record
> >>> + * @value:	the specified target register value
> >>> + * @base:	the command queue base
> >>> + * @offset:	register offset from module base
> >>> + * @mask:	the specified target register mask
> >>> + *
> >>> + * Return: 0 for success; else the error code is returned
> >>> + */
> >>> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
> >>> +			struct cmdq_base *base, u32 offset, u32 mask);
> >>> +
> >>> +/**
> >>> + * cmdq_rec_wfe() - append wait for event command to the command queue reco	rd
> >>
> >> reco rd -> record
> >
> > Will fix it.
> >
> >> Regards,
> >> Matthias
> >>
> >>> + * @rec:	the command queue record
> >>> + * @event:	the desired event type to "wait and CLEAR"
> >>> + *
> >>> + * Return: 0 for success; else the error code is returned
> >>> + */
> >>> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event);
> >>> +
> >>> +/**
> >>> + * cmdq_rec_clear_event() - append clear event command to the command queue
> >>> + *			    record
> >>> + * @rec:	the command queue record
> >>> + * @event:	the desired event to be cleared
> >>> + *
> >>> + * Return: 0 for success; else the error code is returned
> >>> + */
> >>> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event);
> >>> +
> >>> +/**
> >>> + * cmdq_rec_flush() - trigger CMDQ to execute the recorded commands
> >>> + * @rec:	the command queue record
> >>> + *
> >>> + * Return: 0 for success; else the error code is returned
> >>> + *
> >>> + * Trigger CMDQ to execute the recorded commands. Note that this is a
> >>> + * synchronous flush function. When the function returned, the recorded
> >>> + * commands have been done.
> >>> + */
> >>> +int cmdq_rec_flush(struct cmdq_rec *rec);
> >>> +
> >>> +/**
> >>> + * cmdq_rec_flush_async() - trigger CMDQ to asynchronously execute the recorded
> >>> + *			    commands and call back after ISR is finished
> >>> + * @rec:	the command queue record
> >>> + * @cb:		called in the end of CMDQ ISR
> >>> + * @data:	this data will pass back to cb
> >>> + *
> >>> + * Return: 0 for success; else the error code is returned
> >>> + *
> >>> + * Trigger CMDQ to asynchronously execute the recorded commands and call back
> >>> + * after ISR is finished. Note that this is an ASYNC function. When the function
> >>> + * returned, it may or may not be finished. The ISR callback function is called
> >>> + * in the end of ISR.
> 
> "The callback is called from the ISR."
> 
> Regards,
> Matthias
> 
> >>> + */
> >>> +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
> >>> +			 void *data);
> >>> +
> >>> +/**
> >>> + * cmdq_rec_destroy() - destroy command queue record
> >>> + * @rec:	the command queue record
> >>> + */
> >>> +void cmdq_rec_destroy(struct cmdq_rec *rec);
> >>> +
> >>> +#endif	/* __MTK_CMDQ_H__ */
> >>>
> >
> > Thanks,
> > HS
> >

Thanks,
HS

Matthias Brugger June 2, 2016, 8:46 a.m. UTC | #7

On 01/06/16 11:57, Horng-Shyang Liao wrote:
> Hi Mathias,
>
> Please see my inline reply.
>
> On Tue, 2016-05-31 at 22:04 +0200, Matthias Brugger wrote:
>>
>> On 31/05/16 10:36, Horng-Shyang Liao wrote:
>>> Hi Mathias,
>>>
>>> Please see my inline reply.
>>>
>>> On Mon, 2016-05-30 at 17:31 +0200, Matthias Brugger wrote:
>>>>
>>>> On 30/05/16 05:19, HS Liao wrote:
>>>>> This patch is first version of Mediatek Command Queue(CMDQ) driver. The
>>>>> CMDQ is used to help read/write registers with critical time limitation,
>>>>> such as updating display configuration during the vblank. It controls
>>>>> Global Command Engine (GCE) hardware to achieve this requirement.
>>>>> Currently, CMDQ only supports display related hardwares, but we expect
>>>>> it can be extended to other hardwares for future requirements.
>>>>>
>>>>> Signed-off-by: HS Liao <[e-mail]>
>>>>> Signed-off-by: CK Hu <[e-mail]>
>>>>> ---
>>>>>     drivers/soc/mediatek/Kconfig    |  10 +
>>>>>     drivers/soc/mediatek/Makefile   |   1 +
>>>>>     drivers/soc/mediatek/mtk-cmdq.c | 943 ++++++++++++++++++++++++++++++++++++++++
>>>>>     include/soc/mediatek/cmdq.h     | 197 +++++++++
>>>>>     4 files changed, 1151 insertions(+)
>>>>>     create mode 100644 drivers/soc/mediatek/mtk-cmdq.c
>>>>>     create mode 100644 include/soc/mediatek/cmdq.h
>>>>>
>>>>> diff --git a/drivers/soc/mediatek/Kconfig b/drivers/soc/mediatek/Kconfig
>>>>> index 0a4ea80..c4ad75c 100644
>>>>> --- a/drivers/soc/mediatek/Kconfig
>>>>> +++ b/drivers/soc/mediatek/Kconfig
>>>>> @@ -1,6 +1,16 @@
>>>>>     #
>>>>>     # MediaTek SoC drivers
>>>>>     #
>>>>> +config MTK_CMDQ
>>>>> +	bool "MediaTek CMDQ Support"
>>>>> +	depends on ARCH_MEDIATEK || COMPILE_TEST
>>
>> depends on ARM64 ?
>
> Will add ARM64.
>
>>>>> +	select MTK_INFRACFG
>>>>> +	help
>>>>> +	  Say yes here to add support for the MediaTek Command Queue (CMDQ)
>>>>> +	  driver. The CMDQ is used to help read/write registers with critical
>>>>> +	  time limitation, such as updating display configuration during the
>>>>> +	  vblank.
>>>>> +
>>>>>     config MTK_INFRACFG
>>>>>     	bool "MediaTek INFRACFG Support"
>>>>>     	depends on ARCH_MEDIATEK || COMPILE_TEST
>>>>> diff --git a/drivers/soc/mediatek/Makefile b/drivers/soc/mediatek/Makefile
>>>>> index 12998b0..f7397ef 100644
>>>>> --- a/drivers/soc/mediatek/Makefile
>>>>> +++ b/drivers/soc/mediatek/Makefile
>>>>> @@ -1,3 +1,4 @@
>>>>> +obj-$(CONFIG_MTK_CMDQ) += mtk-cmdq.o
>>>>>     obj-$(CONFIG_MTK_INFRACFG) += mtk-infracfg.o
>>>>>     obj-$(CONFIG_MTK_PMIC_WRAP) += mtk-pmic-wrap.o
>>>>>     obj-$(CONFIG_MTK_SCPSYS) += mtk-scpsys.o
>>>>> diff --git a/drivers/soc/mediatek/mtk-cmdq.c b/drivers/soc/mediatek/mtk-cmdq.c
>>>>> new file mode 100644
>>>>> index 0000000..e9d6e1c
>>>>> --- /dev/null
>>>>> +++ b/drivers/soc/mediatek/mtk-cmdq.c
>>>>> @@ -0,0 +1,943 @@
>>>>> +/*
>>>>> + * Copyright (c) 2015 MediaTek Inc.
>>>>> + *
>>>>> + * This program is free software; you can redistribute it and/or modify
>>>>> + * it under the terms of the GNU General Public License version 2 as
>>>>> + * published by the Free Software Foundation.
>>>>> + *
>>>>> + * This program is distributed in the hope that it will be useful,
>>>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>>>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
>>>>> + * GNU General Public License for more details.
>>>>> + */
>>>>> +
>>>>> +#include <linux/clk.h>
>>>>> +#include <linux/clk-provider.h>
>>>>> +#include <linux/completion.h>
>>>>> +#include <linux/dma-mapping.h>
>>>>> +#include <linux/errno.h>
>>>>> +#include <linux/interrupt.h>
>>>>> +#include <linux/iopoll.h>
>>>>> +#include <linux/kernel.h>
>>>>> +#include <linux/kthread.h>
>>>>> +#include <linux/module.h>
>>>>> +#include <linux/mutex.h>
>>>>> +#include <linux/of_address.h>
>>>>> +#include <linux/of_irq.h>
>>>>> +#include <linux/platform_device.h>
>>>>> +#include <linux/slab.h>
>>>>> +#include <linux/spinlock.h>
>>>>> +#include <linux/suspend.h>
>>>>> +#include <linux/workqueue.h>
>>>>> +#include <soc/mediatek/cmdq.h>
>>>>> +
>>>>> +#define CMDQ_INITIAL_CMD_BLOCK_SIZE	PAGE_SIZE
>>>>> +#define CMDQ_INST_SIZE			8 /* instruction is 64-bit */
>>>>> +#define CMDQ_TIMEOUT_MS			1000
>>>>> +#define CMDQ_IRQ_MASK			0xffff
>>>>> +#define CMDQ_DRIVER_DEVICE_NAME		"mtk_cmdq"
>>>>> +#define CMDQ_CLK_NAME			"gce"
>>>>
>>>> We can put the names in directly to un-bloat the defines.
>>>
>>> I will use the names directly and remove defines.
>>>
>>>>> +
>>>>> +#define CMDQ_CURR_IRQ_STATUS		0x010
>>>>> +#define CMDQ_CURR_LOADED_THR		0x018
>>>>> +#define CMDQ_THR_SLOT_CYCLES		0x030
>>>>> +
>>>>> +#define CMDQ_THR_BASE			0x100
>>>>> +#define CMDQ_THR_SHIFT			0x080
>>>>
>>>> Wouldn't be CMDQ_THR_SIZE more accurate?
>>>
>>> Will rename it.
>>>
>>>>> +#define CMDQ_THR_WARM_RESET		0x00
>>>>> +#define CMDQ_THR_ENABLE_TASK		0x04
>>>>> +#define CMDQ_THR_SUSPEND_TASK		0x08
>>>>> +#define CMDQ_THR_CURR_STATUS		0x0c
>>>>> +#define CMDQ_THR_IRQ_STATUS		0x10
>>>>> +#define CMDQ_THR_IRQ_ENABLE		0x14
>>>>> +#define CMDQ_THR_CURR_ADDR		0x20
>>>>> +#define CMDQ_THR_END_ADDR		0x24
>>>>> +#define CMDQ_THR_CFG			0x40
>>>>> +
>>>>> +#define CMDQ_THR_ENABLED		0x1
>>>>> +#define CMDQ_THR_DISABLED		0x0
>>>>> +#define CMDQ_THR_SUSPEND		0x1
>>>>> +#define CMDQ_THR_RESUME			0x0
>>>>> +#define CMDQ_THR_STATUS_SUSPENDED	BIT(1)
>>>>> +#define CMDQ_THR_DO_WARM_RESET		BIT(0)
>>>>> +#define CMDQ_THR_ACTIVE_SLOT_CYCLES	0x3200
>>>>> +#define CMDQ_THR_PRIORITY		3
>>>>> +#define CMDQ_THR_IRQ_DONE		0x1
>>>>> +#define CMDQ_THR_IRQ_ERROR		0x12
>>>>> +#define CMDQ_THR_IRQ_EN			0x13 /* done + error */
>>>>
>>>> #define CMDQ_THR_IRQ_EN (CMDQ_THR_IRQ_ERROR | CMDQ_THR_IRQ_DONE)
>>>
>>> Will do.
>>>
>>>>> +#define CMDQ_THR_IRQ_MASK		0x13
>>>>
>>>> never used.
>>>
>>> Will remove.
>>>
>>>>> +#define CMDQ_THR_EXECUTING		BIT(31)
>>>>> +
>>>>> +#define CMDQ_ARG_A_WRITE_MASK		0xffff
>>>>> +#define CMDQ_SUBSYS_MASK		0x1f
>>>>> +#define CMDQ_OP_CODE_MASK		0xff000000
>>>>> +
>>>>> +#define CMDQ_OP_CODE_SHIFT		24
>>>>
>>>> Couldn't we connect the mask with the shift, or aren't they related?
>>>>
>>>> #define CMDQ_OP_CODE_MASK (0xff << CMDQ_OP_CODE_SHIFT)
>>>
>>> Will do.
>>>
>>>>> +#define CMDQ_SUBSYS_SHIFT		16
>>>>> +
>>>>> +#define CMDQ_WRITE_ENABLE_MASK		BIT(0)
>>>>> +#define CMDQ_JUMP_BY_OFFSET		0x10000000
>>>>> +#define CMDQ_JUMP_BY_PA			0x10000001
>>>>> +#define CMDQ_JUMP_PASS			CMDQ_INST_SIZE
>>>>> +#define CMDQ_WFE_UPDATE			BIT(31)
>>>>> +#define CMDQ_WFE_WAIT			BIT(15)
>>>>> +#define CMDQ_WFE_WAIT_VALUE		0x1
>>>>> +#define CMDQ_EOC_IRQ_EN			BIT(0)
>>>>> +
>>>>> +enum cmdq_thread_index {
>>>>> +	CMDQ_THR_DISP_MAIN_IDX,	/* main */
>>>>> +	CMDQ_THR_DISP_SUB_IDX,	/* sub */
>>>>> +	CMDQ_THR_DISP_MISC_IDX,	/* misc */
>>>>> +	CMDQ_THR_MAX_COUNT,	/* max */
>>>>> +};
>>>>> +
>>>>> +/*
>>>>> + * CMDQ_CODE_MOVE:
>>>>> + *   move value into internal register as mask
>>>>> + *   format: op mask
>>>>> + * CMDQ_CODE_WRITE:
>>>>> + *   write value into target register
>>>>> + *   format: op subsys address value
>>>>> + * CMDQ_CODE_JUMP:
>>>>> + *   jump by offset
>>>>> + *   format: op offset
>>>>> + * CMDQ_CODE_WFE:
>>>>> + *   wait for event and clear
>>>>> + *   it is just clear if no wait
>>>>> + *   format: [wait]  op event update:1 to_wait:1 wait:1
>>>>> + *           [clear] op event update:1 to_wait:0 wait:0
>>>>> + * CMDQ_CODE_EOC:
>>>>> + *   end of command
>>>>> + *   format: op irq_flag
>>>>> + */
>>>>
>>>> I think we need more documentation of how this command queue engine is
>>>> working. If not, I think it will be really complicated to understand how
>>>> to use this.
>>>>
>>>>> +enum cmdq_code {
>>>>> +	CMDQ_CODE_MOVE = 0x02,
>>>>> +	CMDQ_CODE_WRITE = 0x04,
>>>>> +	CMDQ_CODE_JUMP = 0x10,
>>>>> +	CMDQ_CODE_WFE = 0x20,
>>>>> +	CMDQ_CODE_EOC = 0x40,
>>>>> +};
>>>>> +
>>>>> +enum cmdq_task_state {
>>>>> +	TASK_STATE_BUSY, /* running on a GCE thread */
>>>>> +	TASK_STATE_ERROR,
>>>>> +	TASK_STATE_DONE,
>>>>> +};
>>>>> +
>>>>> +struct cmdq_task_cb {
>>>>> +	cmdq_async_flush_cb	cb;
>>>>> +	void			*data;
>>>>> +};
>>>>> +
>>>>> +struct cmdq_thread {
>>>>> +	void __iomem		*base;
>>>>> +	struct list_head	task_busy_list;
>>>>> +	wait_queue_head_t	wait_task_done;
>>>>> +};
>>>>> +
>>>>> +struct cmdq_task {
>>>>> +	struct cmdq		*cmdq;
>>>>> +	struct list_head	list_entry;
>>>>> +	enum cmdq_task_state	task_state;
>>>>> +	void			*va_base;
>>>>> +	dma_addr_t		pa_base;
>>>>> +	u64			engine_flag;
>>>>> +	size_t			command_size;
>>>>> +	u32			num_cmd;
>>>>
>>>> num_cmd is directly connected to command_size. I prefer to just keep
>>>> num_cmd and calculate the command_size where necessary.
>>>
>>> After I trace code, I prefer to keep command_size and calculate num_cmd
>>> where necessary. What do you think?
>>>
>>
>> I suppose you prefer this, as you are writing to the GCE depending on
>> the command_size. I think it is worth to create a macro for the
>> calculation of the number of commands, to make the code more readable.
>> Would be nice if you would just pass cmdq_task to it and it would return
>> the number. Just as an idea.
>
> Will add macro.
>
>>>>> +	struct cmdq_thread	*thread;
>>>>> +	struct cmdq_task_cb	cb;
>>>>> +	struct work_struct	release_work;
>>>>> +};
>>>>> +
>>>>> +struct cmdq {
>>>>> +	struct device		*dev;
>>>>> +	void __iomem		*base;
>>>>> +	u32			irq;
>>>>> +	struct workqueue_struct	*task_release_wq;
>>>>> +	struct cmdq_thread	thread[CMDQ_THR_MAX_COUNT];
>>>>> +	struct mutex		task_mutex;	/* for task */
>>>>> +	spinlock_t		exec_lock;	/* for exec */
>>>>> +	struct clk		*clock;
>>>>> +	bool			suspended;
>>>>> +};
>>>>> +
>>>>> +struct cmdq_subsys {
>>>>> +	u32	base;
>>>>> +	int	id;
>>>>> +};
>>>>> +
>>>>> +static const struct cmdq_subsys gce_subsys[] = {
>>>>> +	{0x1400, 1},
>>>>> +	{0x1401, 2},
>>>>> +	{0x1402, 3},
>>>>> +};
>>>>> +
>>>>> +static int cmdq_subsys_base_to_id(u32 base)
>>>>> +{
>>>>> +	int i;
>>>>> +
>>>>> +	for (i = 0; i < ARRAY_SIZE(gce_subsys); i++)
>>>>> +		if (gce_subsys[i].base == base)
>>>>> +			return gce_subsys[i].id;
>>>>> +	return -EFAULT;
>>>>> +}
>>>>> +
>>>>> +static int cmdq_eng_get_thread(u64 flag)
>>>>> +{
>>>>> +	if (flag & BIT_ULL(CMDQ_ENG_DISP_DSI0))
>>>>> +		return CMDQ_THR_DISP_MAIN_IDX;
>>>>> +	else if (flag & BIT_ULL(CMDQ_ENG_DISP_DPI0))
>>>>> +		return CMDQ_THR_DISP_SUB_IDX;
>>>>> +	else
>>>>> +		return CMDQ_THR_DISP_MISC_IDX;
>>>>> +}
>>>>> +
>>>>> +static void cmdq_task_release(struct cmdq_task *task)
>>>>> +{
>>>>> +	struct cmdq *cmdq = task->cmdq;
>>>>> +
>>>>> +	dma_free_coherent(cmdq->dev, task->command_size, task->va_base,
>>>>> +			  task->pa_base);
>>>>> +	kfree(task);
>>>>> +}
>>>>> +
>>>>> +static struct cmdq_task *cmdq_task_acquire(struct cmdq_rec *rec,
>>>>> +					   struct cmdq_task_cb cb)
>>>>> +{
>>>>> +	struct cmdq *cmdq = rec->cmdq;
>>>>> +	struct device *dev = cmdq->dev;
>>>>> +	struct cmdq_task *task;
>>>>> +
>>>>> +	task = kzalloc(sizeof(*task), GFP_KERNEL);
>>>>> +	INIT_LIST_HEAD(&task->list_entry);
>>>>> +	task->va_base = dma_alloc_coherent(dev, rec->command_size,
>>>>> +					   &task->pa_base, GFP_KERNEL);
>>>>> +	if (!task->va_base) {
>>>>> +		dev_err(dev, "allocate command buffer failed\n");
>>>>> +		kfree(task);
>>>>> +		return NULL;
>>>>> +	}
>>>>> +
>>>>> +	task->cmdq = cmdq;
>>>>> +	task->command_size = rec->command_size;
>>>>> +	task->engine_flag = rec->engine_flag;
>>>>> +	task->task_state = TASK_STATE_BUSY;
>>>>> +	task->cb = cb;
>>>>> +	memcpy(task->va_base, rec->buf, rec->command_size);
>>>>> +	task->num_cmd = task->command_size / CMDQ_INST_SIZE;
>>>>> +	return task;
>>>>> +}
>>>>> +
>>>>> +static void cmdq_thread_writel(struct cmdq_thread *thread, u32 value,
>>>>> +			       u32 offset)
>>>>> +{
>>>>> +	writel(value, thread->base + offset);
>>>>> +}
>>>>> +
>>>>> +static u32 cmdq_thread_readl(struct cmdq_thread *thread, u32 offset)
>>>>> +{
>>>>> +	return readl(thread->base + offset);
>>>>> +}
>>>>
>>>> We can get rid of cmdq_thread_readl/writel.
>>>
>>> Will do.
>>>
>>>>> +
>>>>> +static int cmdq_thread_suspend(struct cmdq *cmdq, struct cmdq_thread *thread)
>>>>> +{
>>>>> +	u32 status;
>>>>> +
>>>>> +	cmdq_thread_writel(thread, CMDQ_THR_SUSPEND, CMDQ_THR_SUSPEND_TASK);
>>>>> +
>>>>> +	/* If already disabled, treat as suspended successful. */
>>>>> +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
>>>>> +	      CMDQ_THR_ENABLED))
>>>>> +		return 0;
>>>>> +
>>>>> +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_CURR_STATUS,
>>>>> +			status, status & CMDQ_THR_STATUS_SUSPENDED, 0, 10)) {
>>>>> +		dev_err(cmdq->dev, "suspend GCE thread 0x%x failed\n",
>>>>> +			(u32)(thread->base - cmdq->base));
>>>>> +		return -EFAULT;
>>>>> +	}
>>>>> +
>>>>> +	return 0;
>>>>> +}
>>>>> +
>>>>> +static void cmdq_thread_resume(struct cmdq_thread *thread)
>>>>> +{
>>>>> +	cmdq_thread_writel(thread, CMDQ_THR_RESUME, CMDQ_THR_SUSPEND_TASK);
>>>>> +}
>>>>> +
>>>>> +static int cmdq_thread_reset(struct cmdq *cmdq, struct cmdq_thread *thread)
>>>>> +{
>>>>> +	u32 warm_reset;
>>>>> +
>>>>> +	cmdq_thread_writel(thread, CMDQ_THR_DO_WARM_RESET, CMDQ_THR_WARM_RESET);
>>>>> +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_WARM_RESET,
>>>>> +			warm_reset, !(warm_reset & CMDQ_THR_DO_WARM_RESET),
>>>>> +			0, 10)) {
>>>>> +		dev_err(cmdq->dev, "reset GCE thread 0x%x failed\n",
>>>>> +			(u32)(thread->base - cmdq->base));
>>>>> +		return -EFAULT;
>>>>> +	}
>>>>> +	writel(CMDQ_THR_ACTIVE_SLOT_CYCLES, cmdq->base + CMDQ_THR_SLOT_CYCLES);
>>>>> +	return 0;
>>>>> +}
>>>>> +
>>>>> +static void cmdq_thread_disable(struct cmdq *cmdq, struct cmdq_thread *thread)
>>>>> +{
>>>>> +	cmdq_thread_reset(cmdq, thread);
>>>>> +	cmdq_thread_writel(thread, CMDQ_THR_DISABLED, CMDQ_THR_ENABLE_TASK);
>>>>> +}
>>>>> +
>>>>> +/* notify GCE to re-fetch commands by setting GCE thread PC */
>>>>> +static void cmdq_thread_invalidate_fetched_data(struct cmdq_thread *thread)
>>>>> +{
>>>>> +	cmdq_thread_writel(thread,
>>>>> +			   cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR),
>>>>> +			   CMDQ_THR_CURR_ADDR);
>>>>> +}
>>>>> +
>>>>> +static void cmdq_task_insert_into_thread(struct cmdq_task *task)
>>>>> +{
>>>>> +	struct cmdq_thread *thread = task->thread;
>>>>> +	struct cmdq_task *prev_task = list_last_entry(
>>>>> +			&thread->task_busy_list, typeof(*task), list_entry);
>>>>> +	u64 *prev_task_base = prev_task->va_base;
>>>>> +
>>>>> +	/* let previous task jump to this task */
>>>>> +	prev_task_base[prev_task->num_cmd - 1] = (u64)CMDQ_JUMP_BY_PA << 32 |
>>>>> +						 task->pa_base;
>>>>> +
>>>>> +	cmdq_thread_invalidate_fetched_data(thread);
>>>>> +}
>>>>> +
>>>>> +/* we assume tasks in the same display GCE thread are waiting the same event. */
>>>>> +static void cmdq_task_remove_wfe(struct cmdq_task *task)
>>>>> +{
>>>>> +	u32 wfe_option = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
>>>>> +	u32 wfe_op = CMDQ_CODE_WFE << CMDQ_OP_CODE_SHIFT;
>>>>> +	u32 *base = task->va_base;
>>>>> +	u32 num_cmd = task->num_cmd << 1;
>>>>> +	int i;
>>>>> +
>>>>> +	for (i = 0; i < num_cmd; i += 2)
>>>>> +		if (base[i] == wfe_option &&
>>>>> +		    (base[i + 1] & CMDQ_OP_CODE_MASK) == wfe_op) {
>>>>> +			base[i] = CMDQ_JUMP_PASS;
>>>>> +			base[i + 1] = CMDQ_JUMP_BY_OFFSET;
>>>>> +		}
>>>>
>>>> After using the command buffer as a void pointer a u64 pointer, we now
>>>> reference to it as u32. I would prefer to explain here, how the command
>>>> looks like we are searching for and use a for loop passing task->num_cmd
>>>> instead.
>>>
>>> Will use u64* to rewrite the above code.
>>>
>>>>> +}
>>>>> +
>>>>> +static void cmdq_task_exec(struct cmdq_task *task, struct cmdq_thread *thread)
>>>>> +{
>>>>> +	struct cmdq *cmdq = task->cmdq;
>>>>> +	unsigned long flags;
>>>>> +	unsigned long curr_pa, end_pa;
>>>>> +
>>>>> +	WARN_ON(clk_prepare_enable(cmdq->clock) < 0);
>>>>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
>>>>
>>>> cmdq_task_exec is called with cmdq->task_mutex held, so why do we need
>>>> the spin_lock here? Can't we just use one of the two?
>>>
>>> We can drop task_mutex, but we will get some side effects.
>>> 1. exec_lock needs to include more code, but I think it is not good for
>>>      spinlock.
>>> 2. In cmdq_rec_flush_async(), task_mutex needs to protect
>>>      (1) cmdq->suspended, (2) cmdq_task_exec(), and
>>>      (3) cmdq_task_wait_release_schedule().
>>>      If we drop task_mutex, we have to put cmdq->suspended if condition
>>>      just before cmdq_task_exec() and inside exec_lock, and we have to
>>>      release task and its command buffer if error. This will let flow
>>>      become more complex and enlarge code size.
>>>
>>> What do you think?
>>
>> Why do you need to protect cmdq_task_wait_release_schedule? We don't
>> care about the order of the workqueue elements, do we?
>
> According to CK's comment, we have to ensure order to remove previous
> task.
> http://lists.infradead.org/pipermail/linux-mediatek/2016-May/005547.html
>
>> As far as I understand you would need to protect cmdq_task_acquire as
>> well, to "ensure" continously growing pa_base. More on that below.
>
> We need to ensure continuous physical addresses in a task, but we don't
> need to ensure continuous physical addresses between tasks.
> So, I think it is unnecessary to protect by mutex or spinlock.
>

True, I didn't get that.

>>>
>>>>> +	task->thread = thread;
>>>>> +	task->task_state = TASK_STATE_BUSY;
>>>>
>>>> That was already set in cmdq_task_acquire, why do we need to set it here
>>>> again?
>>>
>>> Will drop it.
>>>
>>
>> Yeah, but I think it makes more sense to drop it in cmdq_task_acquire
>> instead.
>
> Will drop it in cmdq_task_acquire instead.
>
>>>>> +	if (list_empty(&thread->task_busy_list)) {
>>>>> +		WARN_ON(cmdq_thread_reset(cmdq, thread) < 0);
>>>>> +
>>>>> +		cmdq_thread_writel(thread, task->pa_base, CMDQ_THR_CURR_ADDR);
>>>>> +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
>>>>> +				   CMDQ_THR_END_ADDR);
>>>>> +		cmdq_thread_writel(thread, CMDQ_THR_PRIORITY, CMDQ_THR_CFG);
>>>>> +		cmdq_thread_writel(thread, CMDQ_THR_IRQ_EN,
>>>>> +				   CMDQ_THR_IRQ_ENABLE);
>>>>> +
>>>>> +		cmdq_thread_writel(thread, CMDQ_THR_ENABLED,
>>>>> +				   CMDQ_THR_ENABLE_TASK);
>>>>> +	} else {
>>>>> +		WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
>>>>> +
>>>>> +		/*
>>>>> +		 * check boundary condition
>>>>> +		 * PC = END - 8, EOC is executed
>>>>> +		 * PC = END, all CMDs are executed
>>>>> +		 */
>>>>> +		curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
>>>>> +		end_pa = cmdq_thread_readl(thread, CMDQ_THR_END_ADDR);
>>>>> +		if (curr_pa == end_pa - 8 || curr_pa == end_pa) {
>>>>
>>>> 8 refers to CMDQ_INST_SIZE, right?
>>>
>>> Yes, I will use CMDQ_INST_SIZE.
>>>
>>>>> +			/* set to this task directly */
>>>>> +			cmdq_thread_writel(thread, task->pa_base,
>>>>> +					   CMDQ_THR_CURR_ADDR);
>>>>> +		} else {
>>>>> +			cmdq_task_insert_into_thread(task);
>>>>> +
>>>>> +			if (thread == &cmdq->thread[CMDQ_THR_DISP_MAIN_IDX] ||
>>>>> +			    thread == &cmdq->thread[CMDQ_THR_DISP_SUB_IDX])
>>>>> +				cmdq_task_remove_wfe(task);
>>>>
>>>> We could do this check using the task->engine_flag, I suppose that's
>>>> easier to undestand then.
>>>
>>> Will use task->engine_flag.
>>>
>>>>> +
>>>>> +			smp_mb(); /* modify jump before enable thread */
>>>>> +		}
>>>>> +
>>>>> +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
>>>>> +				   CMDQ_THR_END_ADDR);
>>>>> +		cmdq_thread_resume(thread);
>>>>> +	}
>>>>> +	list_move_tail(&task->list_entry, &thread->task_busy_list);
>>>>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>> +}
>>>>> +
>>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
>>>>> +				   struct cmdq_thread *thread, u32 irq_flag)
>>>>> +{
>>>>> +	struct cmdq_task *task, *tmp, *curr_task = NULL;
>>>>> +	u32 curr_pa;
>>>>> +	struct cmdq_cb_data cmdq_cb_data;
>>>>> +	bool err;
>>>>> +
>>>>> +	if (irq_flag & CMDQ_THR_IRQ_ERROR)
>>>>> +		err = true;
>>>>> +	else if (irq_flag & CMDQ_THR_IRQ_DONE)
>>>>> +		err = false;
>>>>> +	else
>>>>> +		return;
>>>>> +
>>>>> +	curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
>>>>> +
>>>>> +	list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
>>>>> +				 list_entry) {
>>>>> +		if (curr_pa >= task->pa_base &&
>>>>> +		    curr_pa < (task->pa_base + task->command_size))
>>>>
>>>> What are you checking here? It seems as if you make some implcit
>>>> assumptions about pa_base and the order of execution of commands in the
>>>> thread. Is it save to do so? Does dma_alloc_coherent give any guarantees
>>>> about dma_handle?
>>>
>>> 1. Check what is the current running task in this GCE thread.
>>> 2. Yes.
>>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
>>>
>>
>> Yes, physical addresses might be continous, but AFAIK there is no
>> guarantee that the dma_handle address is steadily growing, when calling
>> dma_alloc_coherent. And if I understand the code correctly, you use this
>> assumption to decide if the task picked from task_busy_list is currently
>> executing. So I think this mecanism is not working.
>
> I don't use dma_handle address, and just use physical addresses.
>  From CPU's point of view, tasks are linked by the busy list.
>  From GCE's point of view, tasks are linked by the JUMP command.
>
>> In which cases does the HW thread raise an interrupt.
>> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
>
> GCE will raise interrupt if any task is done or error.
> However, GCE is fast, so CPU may get multiple done tasks
> when it is running ISR.
>
> In case of error, that GCE thread will pause and raise interrupt.
> So, CPU may get multiple done tasks and one error task.
>

I think we should reimplement the ISR mechanism. Can't we just read 
CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave 
cmdq_handle_error_done to the thread_fn? You will need to pass 
information from the handler to thread_fn, but that shouldn't be an 
issue. AFAIK interrupts are disabled in the handler, so we should stay 
there as short as possible. Traversing task_busy_list is expensive, so 
we need to do it in a thread context.

I keep thinking about how to get rid of the two data structures, 
task_busy_list and the task_release_wq. We need the latter for the only 
sake of getting a timeout.

Did you have a look on how the mailbox framework handles this?
By the way, what is the reason to not implement the whole driver as a 
mailbox controller? For me, this driver looks like a good fit.


>>>>> +			curr_task = task;
>>>>> +		if (task->cb.cb) {
>>>>> +			cmdq_cb_data.err = curr_task ? err : false;
>>>>> +			cmdq_cb_data.data = task->cb.data;
>>>>> +			task->cb.cb(cmdq_cb_data);
>>>>> +		}
>>>>> +		task->task_state = (curr_task && err) ? TASK_STATE_ERROR :
>>>>> +				TASK_STATE_DONE;
>>>>> +		list_del(&task->list_entry);
>>>>> +		if (curr_task)
>>>>> +			break;
>>>>> +	}
>>>>> +
>>>>> +	wake_up(&thread->wait_task_done);
>>>>> +}
>>>>> +
>>>>> +static void cmdq_thread_irq_handler(struct cmdq *cmdq, int tid)
>>>>> +{
>>>>> +	struct cmdq_thread *thread = &cmdq->thread[tid];
>>>>> +	unsigned long flags = 0L;
>>>>> +	u32 irq_flag;
>>>>> +
>>>>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
>>>>> +
>>>>> +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
>>>>> +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
>>>>> +
>>>>> +	/*
>>>>> +	 * Another CPU core could run "release task" right before we acquire
>>>>> +	 * the spin lock, and thus reset / disable this GCE thread, so we
>>>>> +	 * need to check the enable bit of this GCE thread.
>>>>> +	 */
>>>>> +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
>>>>> +	      CMDQ_THR_ENABLED))
>>>>> +		irq_flag = 0;
>>>>
>>>> cmdq_handle_error_done just retuns in this case. Programming this way
>>>> just makes things confusing. What about:
>>>>
>>>> if (cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
>>>> 	      CMDQ_THR_ENABLED)
>>>> 	cmdq_handle_error_done(cmdq, thread, irq_flag);
>>>> else
>>>> 		irq_flag = 0;
>>>>
>>>> spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>
>>> We still need to clear irq_flag if GCE thread is disabled.
>>> So, I think we can just return here.
>>>
>>> 	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
>>> 	      CMDQ_THR_ENABLED))
>>> 		return;
>>>
>>> What do you think?
>>>
>>
>> No, you can't just return, you need to unlock the spinlock.
>> Anyway I would prefer it the other way round, as I put it in my last
>> mail. Just delete the else branch, we don't need to set irq_flag to zero.
>
> Sorry for my previous wrong reply since I merge your comment
> and CK's comment.
> http://lists.infradead.org/pipermail/linux-mediatek/2016-May/005547.html
> So, I will put this if condition into cmdq_task_handle_error_result()
> and then just return it if GCE thread is disabled.
>

You mean in cmdq_task_handle_done
We should rename this functions to not create confusion.

Regards,
Matthias

>>>>> +
>>>>> +	cmdq_handle_error_done(cmdq, thread, irq_flag);
>>>>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>> +}
>>>>> +
>>>>> +static irqreturn_t cmdq_irq_handler(int irq, void *dev)
>>>>> +{
>>>>> +	struct cmdq *cmdq = dev;
>>>>> +	u32 irq_status;
>>>>> +	int i;
>>>>> +
>>>>> +	irq_status = readl(cmdq->base + CMDQ_CURR_IRQ_STATUS);
>>>>> +	irq_status &= CMDQ_IRQ_MASK;
>>>>> +	irq_status ^= CMDQ_IRQ_MASK;
>>>>
>>>> irq_status can be much bigger then 3, which is the number of threads in
>>>> the system (CMDQ_THR_MAX_COUNT). So why we use this mask here isn't
>>>> clear to me.
>>>
>>> Our GCE hardware has 16 threads, but we only use 3 threads currently.
>>>
>>
>> Ok, but please use bitops here.
>
> Will use bitops.
>
>>>>> +
>>>>> +	if (!irq_status)
>>>>> +		return IRQ_NONE;
>>>>> +
>>>>> +	while (irq_status) {
>>>>> +		i = ffs(irq_status) - 1;
>>>>> +		irq_status &= ~BIT(i);
>>>>> +		cmdq_thread_irq_handler(cmdq, i);
>>>>> +	}
>>>>
>>>> Can you explain how the irq status register looks like, that would it
>>>> make much easier to understand what happens here.
>>>
>>> Bit 0 ~ 15 of irq status register represents GCE thread 0 ~ 15
>>> interrupt. 0 means asserting interrupt; 1 means no interrupt.
>>>
>>
>> Thanks, that helped. :)
>>
>>>>> +
>>>>> +	return IRQ_HANDLED;	
>>>>> +}
>>>>> +
>>>>> +static int cmdq_task_handle_error_result(struct cmdq_task *task)
>>>>
>>>> We never check the return values, why do we have them?
>>>
>>> Will drop return value.
>>>
>>>>> +{
>>>>> +	struct cmdq *cmdq = task->cmdq;
>>>>> +	struct device *dev = cmdq->dev;
>>>>> +	struct cmdq_thread *thread = task->thread;
>>>>> +	struct cmdq_task *next_task, *prev_task;
>>>>> +	u32 irq_flag;
>>>>> +
>>>>> +	/* suspend GCE thread to ensure consistency */
>>>>> +	WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
>>>>> +
>>>>> +	/* ISR has handled this error task */
>>>>> +	if (task->task_state == TASK_STATE_ERROR) {
>>>>> +		next_task = list_first_entry_or_null(&thread->task_busy_list,
>>>>> +				struct cmdq_task, list_entry);
>>>>> +		if (next_task) /* move to next task */
>>>>> +			cmdq_thread_writel(thread, next_task->pa_base,
>>>>> +					   CMDQ_THR_CURR_ADDR);
>>>>
>>>> We have to do this, as we suppose that the thread did not reach the jump
>>>> instruction we put into it's command queue, right?
>>>
>>> Yes.
>>>
>>
>> So this should then go into it's own function. In wait_release_work,
>> something like this:
>>
>> if(task->task_state == TASK_STATE_ERROR)
>>        cmdq_task_handle_error(task)
>
> OK.
> I will write new function cmdq_task_handle_error() to handle error case.
>
>>>>> +		cmdq_thread_resume(thread);
>>>>> +		return -ECANCELED;
>>>>> +	}
>>>>> +
>>>>
>>>> if task_state != ERROR and != DONE. This means that the timeout of
>>>> task_release_wq has timed out, right?
>>>
>>> Yes.
>>>
>>>>> +	/*
>>>>> +	 * Save next_task and prev_task in advance
>>>>> +	 * since cmdq_handle_error_done will remove list_entry.
>>>>> +	 */
>>>>> +	next_task = prev_task = NULL;
>>>>> +	if (task->list_entry.next != &thread->task_busy_list)
>>>>> +		next_task = list_next_entry(task, list_entry);
>>>>> +	if (task->list_entry.prev != &thread->task_busy_list)
>>>>> +		prev_task = list_prev_entry(task, list_entry);
>>>>> +
>>>>> +	/*
>>>>> +	 * Although IRQ is disabled, GCE continues to execute.
>>>>> +	 * It may have pending IRQ before GCE thread is suspended,
>>>>> +	 * so check this condition again.
>>>>> +	 */
>>>>
>>>> The first thing we did in this function was suspending the thread. Why
>>>> do we need this then?
>>>
>>> Because timeout is CPU timeout not GCE timeout, GCE could just finish
>>> this task before the GCE thread is suspended.
>>>
>>
>> What are the reasons for a timeout? An error has happend, or the task is
>> still executing.
>
>  From GCE's point of view, this task is still executing.
> But, it could be an error of client.
> For example, task may never get event if display turn off hardware
> before waiting for task to finish its work.
>
>>>> To be honest this whole functions looks really like a design error. We
>>>> have to sperate the states much clearer so that it is possible to
>>>> understand what is happening in the GCE. Isn't it for example posible to
>>>> have worker queues for timed out tasks and tasks with an error? I'm not
>>>> sure how to do this, actually I'm not sure if I really understood how
>>>> this is supposed to work.
>>>
>>> GCE doesn't have timeout. The timeout is decided and controlled by CPU,
>>> so we check timeout in release work.
>>> For error and done, they are easy to check by register, and we have
>>> already created release work for timeout. So, I don't think we need to
>>> create work queue for each case.
>>>
>>> What do you think?
>>>
>>
>> I think, if we find in here, that the irq_flag is set, then the the
>> interrupt handler was triggered and is spinning the spinlock. If this is
>> not the case, we have a timeout and we handle this.
>
> I will write another function to handle error, and handle timeout here.
>
>>>> How much do we win, when we patch the thread command queue for every
>>>> task we add, instead of just taking one task after another from the
>>>> task_busy_list?
>>>
>>> GCE is used to help read/write registers with critical time limitation.
>>> Sometimes, client may ask to process multiple tasks in a short period
>>> of time, e.g. display flush multiple tasks for next vblank. So, CMDQ
>>> shouldn't limit to process one task after another from the
>>> task_busy_list. Currently, when interrupt or timeout, we will check
>>> how many tasks are done, and which one is error or timeout.
>>>
>>
>> So I suppose the device driver who use this are interested in throughput
>> and not in latency. The callback of every
>>
>>>>> +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
>>>>> +	cmdq_handle_error_done(cmdq, thread, irq_flag);
>>>>> +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
>>>>> +
>>>>> +	if (task->task_state == TASK_STATE_DONE) {
>>>>> +		cmdq_thread_resume(thread);
>>>>> +		return 0;
>>>>> +	}
>>>>> +
>>>>> +	if (task->task_state == TASK_STATE_ERROR) {
>>>>> +		dev_err(dev, "task 0x%p error\n", task);
>>>>> +		if (next_task) /* move to next task */
>>>>> +			cmdq_thread_writel(thread, next_task->pa_base,
>>>>> +					   CMDQ_THR_CURR_ADDR);
>>>>> +		cmdq_thread_resume(thread);
>>>>> +		return -ECANCELED;
>>>>> +	}
>>>>> +
>>>>> +	/* Task is running, so we force to remove it. */
>>>>> +	dev_err(dev, "task 0x%p timeout or killed\n", task);
>>>>> +	task->task_state = TASK_STATE_ERROR;
>>>>> +
>>>>> +	if (prev_task) {
>>>>> +		u64 *prev_va = prev_task->va_base;
>>>>> +		u64 *curr_va = task->va_base;
>>>>> +
>>>>> +		/* copy JUMP instruction */
>>>>> +		prev_va[prev_task->num_cmd - 1] = curr_va[task->num_cmd - 1];
>>>>> +
>>>>> +		cmdq_thread_invalidate_fetched_data(thread);
>>>>> +	} else if (next_task) { /* move to next task */
>>>>> +		cmdq_thread_writel(thread, next_task->pa_base,
>>>>> +				   CMDQ_THR_CURR_ADDR);
>>>>> +	}
>>>>> +
>>>>> +	list_del(&task->list_entry);
>>>>> +	cmdq_thread_resume(thread);
>>>>> +
>>>>> +	/* call cb here to prevent lock */
>>>>> +	if (task->cb.cb) {
>>>>> +		struct cmdq_cb_data cmdq_cb_data;
>>>>> +
>>>>> +		cmdq_cb_data.err = true;
>>>>> +		cmdq_cb_data.data = task->cb.data;
>>>>> +		task->cb.cb(cmdq_cb_data);
>>>>> +	}
>>>>> +
>>>>> +	return -ECANCELED;
>>>>> +}
>>>>> +
>>>>> +static void cmdq_task_wait_release_work(struct work_struct *work_item)
>>>>> +{
>>>>> +	struct cmdq_task *task = container_of(work_item, struct cmdq_task,
>>>>> +					      release_work);
>>>>> +	struct cmdq *cmdq = task->cmdq;
>>>>> +	struct cmdq_thread *thread = task->thread;
>>>>> +	unsigned long flags;
>>>>> +
>>>>> +	wait_event_timeout(thread->wait_task_done,
>>>>> +			   task->task_state != TASK_STATE_BUSY,
>>>>> +			   msecs_to_jiffies(CMDQ_TIMEOUT_MS));
>>>>> +
>>>>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
>>>>> +	if (task->task_state != TASK_STATE_DONE)
>>>>> +		cmdq_task_handle_error_result(task);
>>>>> +	if (list_empty(&thread->task_busy_list))
>>>>> +		cmdq_thread_disable(cmdq, thread);
>>>>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>> +
>>>>> +	/* release regardless of success or not */
>>>>> +	clk_disable_unprepare(cmdq->clock);
>>>>> +	cmdq_task_release(task);
>>>>> +}
>>>>> +
>>>>> +static void cmdq_task_wait_release_schedule(struct cmdq_task *task)
>>>>> +{
>>>>> +	struct cmdq *cmdq = task->cmdq;
>>>>> +
>>>>> +	INIT_WORK(&task->release_work, cmdq_task_wait_release_work);
>>>>> +	queue_work(cmdq->task_release_wq, &task->release_work);
>>>>> +}
>>>>> +
>>>>> +static int cmdq_rec_realloc_cmd_buffer(struct cmdq_rec *rec, size_t size)
>>>>> +{
>>>>> +	void *new_buf;
>>>>> +
>>>>> +	new_buf = krealloc(rec->buf, size, GFP_KERNEL | __GFP_ZERO);
>>>>> +	if (!new_buf)
>>>>> +		return -ENOMEM;
>>>>> +	rec->buf = new_buf;
>>>>> +	rec->buf_size = size;
>>>>> +	return 0;
>>>>> +}
>>>>> +
>>>>> +struct cmdq_base *cmdq_register_device(struct device *dev)
>>>>> +{
>>>>> +	struct cmdq_base *cmdq_base;
>>>>> +	struct resource res;
>>>>> +	int subsys;
>>>>> +	u32 base;
>>>>> +
>>>>> +	if (of_address_to_resource(dev->of_node, 0, &res))
>>>>> +		return NULL;
>>>>> +	base = (u32)res.start;
>>>>> +
>>>>> +	subsys = cmdq_subsys_base_to_id(base >> 16);
>>>>> +	if (subsys < 0)
>>>>> +		return NULL;
>>>>> +
>>>>> +	cmdq_base = devm_kmalloc(dev, sizeof(*cmdq_base), GFP_KERNEL);
>>>>> +	if (!cmdq_base)
>>>>> +		return NULL;
>>>>> +	cmdq_base->subsys = subsys;
>>>>> +	cmdq_base->base = base;
>>>>> +
>>>>> +	return cmdq_base;
>>>>> +}
>>>>> +EXPORT_SYMBOL(cmdq_register_device);
>>>>> +
>>>>> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
>>>>> +		    struct cmdq_rec **rec_ptr)
>>>>> +{
>>>>> +	struct cmdq_rec *rec;
>>>>> +	int err;
>>>>> +
>>>>> +	rec = kzalloc(sizeof(*rec), GFP_KERNEL);
>>>>> +	if (!rec)
>>>>> +		return -ENOMEM;
>>>>> +	rec->cmdq = dev_get_drvdata(dev);
>>>>> +	rec->engine_flag = engine_flag;
>>>>> +	err = cmdq_rec_realloc_cmd_buffer(rec, CMDQ_INITIAL_CMD_BLOCK_SIZE);
>>>>
>>>> Just pass PAGE_SIZE here, no need for CMDQ_INITIAL_CMD_BLOCK_SIZE.
>>>
>>> Will do.
>>>
>>>>> +	if (err < 0) {
>>>>> +		kfree(rec);
>>>>> +		return err;
>>>>> +	}
>>>>> +	*rec_ptr = rec;
>>>>> +	return 0;
>>>>> +}
>>>>> +EXPORT_SYMBOL(cmdq_rec_create);
>>>>> +
>>>>> +static int cmdq_rec_append_command(struct cmdq_rec *rec, enum cmdq_code code,
>>>>> +				   u32 arg_a, u32 arg_b)
>>>>> +{
>>>>> +	u64 *cmd_ptr;
>>>>> +	int err;
>>>>> +
>>>>> +	if (WARN_ON(rec->finalized))
>>>>> +		return -EBUSY;
>>>>> +	if (code < CMDQ_CODE_MOVE || code > CMDQ_CODE_EOC)
>>>>> +		return -EINVAL;
>>>>
>>>> cmdq_rec_append_command is just called from inside this driver and code
>>>> is a enum. We can expect it to be correct, no need for this check.
>>>
>>> Will drop this check.
>>>
>>>>> +	if (unlikely(rec->command_size + CMDQ_INST_SIZE > rec->buf_size)) {
>>>>
>>>> command_size is the offset into the buffer to which a new command is
>>>> written, so this name is highly confusing. I wonder if this would be
>>>> easier to understand if we redefine command_size to something like the
>>>> number of commands and divide/multiply CMDQ_INST_SIZE where this is needed.
>>>
>>> I can rename command_size to cmd_buf_size and calculate num_cmd by
>>> dividing CMDQ_INST_SIZE.
>>> What do you think?
>>>
>>>>> +		err = cmdq_rec_realloc_cmd_buffer(rec, rec->buf_size * 2);
>>>>> +		if (err < 0)
>>>>> +			return err;
>>>>> +	}
>>>>> +	cmd_ptr = rec->buf + rec->command_size;
>>>>> +	(*cmd_ptr) = (u64)((code << CMDQ_OP_CODE_SHIFT) | arg_a) << 32 | arg_b;
>>>>> +	rec->command_size += CMDQ_INST_SIZE;
>>>>> +	return 0;
>>>>> +}
>>>>> +
>>>>> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value, struct cmdq_base *base,
>>>>> +		   u32 offset)
>>>>> +{
>>>>> +	u32 arg_a = ((base->base + offset) & CMDQ_ARG_A_WRITE_MASK) |
>>>>> +		    ((base->subsys & CMDQ_SUBSYS_MASK) << CMDQ_SUBSYS_SHIFT);
>>>>
>>>> base->subsys is the id from gce_sybsys, so we can expect it to be
>>>> correct, no need to mask with CMDQ_SUBSYS_MASK.
>>>
>>> Will drop it.
>>>
>>>>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WRITE, arg_a, value);
>>>>> +}
>>>>> +EXPORT_SYMBOL(cmdq_rec_write);
>>>>> +
>>>>> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
>>>>> +			struct cmdq_base *base, u32 offset, u32 mask)
>>>>> +{
>>>>> +	u32 offset_mask = offset;
>>>>> +	int err;
>>>>> +
>>>>> +	if (mask != 0xffffffff) {
>>>>> +		err = cmdq_rec_append_command(rec, CMDQ_CODE_MOVE, 0, ~mask);
>>>>> +		if (err < 0)
>>>>> +			return err;
>>>>> +		offset_mask |= CMDQ_WRITE_ENABLE_MASK;
>>>>> +	}
>>>>> +	return cmdq_rec_write(rec, value, base, offset_mask);
>>>>> +}
>>>>> +EXPORT_SYMBOL(cmdq_rec_write_mask);
>>>>> +
>>>>> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event)
>>>>> +{
>>>>> +	u32 arg_b;
>>>>> +
>>>>> +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
>>>>> +		return -EINVAL;
>>>>> +
>>>>> +	/*
>>>>> +	 * bit 0-11: wait value
>>>>> +	 * bit 15: 1 - wait, 0 - no wait
>>>>> +	 * bit 16-27: update value
>>>>> +	 * bit 31: 1 - update, 0 - no update
>>>>> +	 */
>>>>
>>>> I don't understand this comments. What are they for?
>>>
>>> This is for WFE command. I will comment it.
>>>
>>>>> +	arg_b = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
>>>>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event, arg_b);
>>>>> +}
>>>>> +EXPORT_SYMBOL(cmdq_rec_wfe);
>>>>> +
>>>>> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event)
>>>>> +{
>>>>> +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
>>>>> +		return -EINVAL;
>>>>> +
>>>>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event,
>>>>> +				       CMDQ_WFE_UPDATE);
>>>>> +}
>>>>> +EXPORT_SYMBOL(cmdq_rec_clear_event);
>>>>> +
>>>>> +static int cmdq_rec_finalize(struct cmdq_rec *rec)
>>>>> +{
>>>>> +	int err;
>>>>> +
>>>>> +	if (rec->finalized)
>>>>> +		return 0;
>>>>> +
>>>>> +	/* insert EOC and generate IRQ for each command iteration */
>>>>> +	err = cmdq_rec_append_command(rec, CMDQ_CODE_EOC, 0, CMDQ_EOC_IRQ_EN);
>>>>> +	if (err < 0)
>>>>> +		return err;
>>>>> +
>>>>> +	/* JUMP to end */
>>>>> +	err = cmdq_rec_append_command(rec, CMDQ_CODE_JUMP, 0, CMDQ_JUMP_PASS);
>>>>> +	if (err < 0)
>>>>> +		return err;
>>>>> +
>>>>
>>>> Does this need to be atomic?
>>>> What happens if after CODE_EOC and before CODE_JUMP some
>>>> write/read/event gets added?
>>>> What happens if more commands get added to the queue after CODE_JUMP,
>>>> but before finalized is set to true. Why don't you use atomic functions
>>>> to access finalized?
>>>
>>> Since cmdq_rec doesn't guarantee thread safe, mutex is needed when
>>> client uses cmdq_rec.
>>>
>>
>> Well I think that rec->finalized tries to implement this, but might
>> fail, if two kernel threads work on the same cmdq_rec.
>>
>>>>> +	rec->finalized = true;
>>>>> +	return 0;
>>>>> +}
>>>>> +
>>>>> +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
>>>>> +			 void *data)
>>>>> +{
>>>>> +	struct cmdq *cmdq = rec->cmdq;
>>>>> +	struct cmdq_task *task;
>>>>> +	struct cmdq_task_cb task_cb;
>>>>> +	struct cmdq_thread *thread;
>>>>> +	int err;
>>>>> +
>>>>> +	mutex_lock(&cmdq->task_mutex);
>>>>> +	if (cmdq->suspended) {
>>>>> +		dev_err(cmdq->dev, "%s is called after suspended\n", __func__);
>>>>> +		mutex_unlock(&cmdq->task_mutex);
>>>>> +		return -EPERM;
>>>>> +	}
>>>>> +
>>>>> +	err = cmdq_rec_finalize(rec);
>>>>> +	if (err < 0) {
>>>>> +		mutex_unlock(&cmdq->task_mutex);
>>>>> +		return err;
>>>>> +	}
>>>>> +
>>>>> +	task_cb.cb = cb;
>>>>> +	task_cb.data = data;
>>>>> +	task = cmdq_task_acquire(rec, task_cb);
>>>>> +	if (!task) {
>>>>> +		mutex_unlock(&cmdq->task_mutex);
>>>>> +		return -EFAULT;
>>>>> +	}
>>>>> +
>>>>> +	thread = &cmdq->thread[cmdq_eng_get_thread(task->engine_flag)];
>>>>> +	cmdq_task_exec(task, thread);
>>>>> +	cmdq_task_wait_release_schedule(task);
>>>>> +	mutex_unlock(&cmdq->task_mutex);
>>>>> +	return 0;
>>>>> +}
>>>>> +EXPORT_SYMBOL(cmdq_rec_flush_async);
>>>>> +
>>>>> +struct cmdq_flush_completion {
>>>>> +	struct completion cmplt;
>>>>> +	bool err;
>>>>> +};
>>>>> +
>>>>> +static int cmdq_rec_flush_cb(struct cmdq_cb_data data)
>>>>> +{
>>>>> +	struct cmdq_flush_completion *cmplt = data.data;
>>>>> +
>>>>> +	cmplt->err = data.err;
>>>>> +	complete(&cmplt->cmplt);
>>>>> +	return 0;
>>>>> +}
>>>>> +
>>>>> +int cmdq_rec_flush(struct cmdq_rec *rec)
>>>>> +{
>>>>> +	struct cmdq_flush_completion cmplt;
>>>>> +	int err;
>>>>> +
>>>>> +	init_completion(&cmplt.cmplt);
>>>>> +	err = cmdq_rec_flush_async(rec, cmdq_rec_flush_cb, &cmplt);
>>>>> +	if (err < 0)
>>>>> +		return err;
>>>>> +	wait_for_completion(&cmplt.cmplt);
>>>>> +	return cmplt.err ? -EFAULT : 0;
>>>>> +}
>>>>> +EXPORT_SYMBOL(cmdq_rec_flush);
>>>>> +
>>>>> +void cmdq_rec_destroy(struct cmdq_rec *rec)
>>>>> +{
>>>>> +	kfree(rec->buf);
>>>>> +	kfree(rec);
>>>>> +}
>>>>> +EXPORT_SYMBOL(cmdq_rec_destroy);
>>>>> +
>>>>> +static bool cmdq_task_is_empty(struct cmdq *cmdq)
>>>>> +{
>>>>> +	struct cmdq_thread *thread;
>>>>> +	int i;
>>>>> +
>>>>> +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
>>>>> +		thread = &cmdq->thread[i];
>>>>> +		if (!list_empty(&thread->task_busy_list))
>>>>> +			return false;
>>>>> +	}
>>>>> +	return true;
>>>>> +}
>>>>> +
>>>>> +static int cmdq_suspend(struct device *dev)
>>>>> +{
>>>>> +	struct cmdq *cmdq = dev_get_drvdata(dev);
>>>>> +	u32 exec_threads;
>>>>> +
>>>>> +	mutex_lock(&cmdq->task_mutex);
>>>>> +	cmdq->suspended = true;
>>>>> +	mutex_unlock(&cmdq->task_mutex);
>>>>> +
>>>>> +	exec_threads = readl(cmdq->base + CMDQ_CURR_LOADED_THR);
>>>>> +	if ((exec_threads & CMDQ_THR_EXECUTING) && !cmdq_task_is_empty(cmdq)) {
>>>>> +		dev_err(dev, "wait active tasks timeout.\n");
>>>>> +		flush_workqueue(cmdq->task_release_wq);
>>>>> +	}
>>>>> +	return 0;
>>>>> +}
>>>>> +
>>>>> +static int cmdq_resume(struct device *dev)
>>>>> +{
>>>>> +	struct cmdq *cmdq = dev_get_drvdata(dev);
>>>>> +
>>>>> +	cmdq->suspended = false;
>>>>> +	return 0;
>>>>> +}
>>>>> +
>>>>> +static int cmdq_remove(struct platform_device *pdev)
>>>>> +{
>>>>> +	struct cmdq *cmdq = platform_get_drvdata(pdev);
>>>>> +
>>>>> +	destroy_workqueue(cmdq->task_release_wq);
>>>>> +	cmdq->task_release_wq = NULL;
>>>>> +	return 0;
>>>>> +}
>>>>> +
>>>>> +static int cmdq_probe(struct platform_device *pdev)
>>>>> +{
>>>>> +	struct device *dev = &pdev->dev;
>>>>> +	struct device_node *node = dev->of_node;
>>>>> +	struct resource *res;
>>>>> +	struct cmdq *cmdq;
>>>>> +	int err, i;
>>>>> +
>>>>> +	cmdq = devm_kzalloc(dev, sizeof(*cmdq), GFP_KERNEL);
>>>>> +	if (!cmdq)
>>>>> +		return -ENOMEM;
>>>>> +	cmdq->dev = dev;
>>>>> +
>>>>> +	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
>>>>> +	cmdq->base = devm_ioremap_resource(dev, res);
>>>>> +	if (IS_ERR(cmdq->base)) {
>>>>> +		dev_err(dev, "failed to ioremap gce\n");
>>>>> +		return PTR_ERR(cmdq->base);
>>>>> +	}
>>>>> +
>>>>> +	cmdq->irq = irq_of_parse_and_map(node, 0);
>>>>> +	if (!cmdq->irq) {
>>>>> +		dev_err(dev, "failed to get irq\n");
>>>>> +		return -EINVAL;
>>>>> +	}
>>>>> +
>>>>> +	dev_dbg(dev, "cmdq device: addr:0x%p, va:0x%p, irq:%d\n",
>>>>> +		dev, cmdq->base, cmdq->irq);
>>>>> +
>>>>> +	mutex_init(&cmdq->task_mutex);
>>>>> +	spin_lock_init(&cmdq->exec_lock);
>>>>> +	cmdq->task_release_wq = alloc_ordered_workqueue(
>>>>> +			"%s", WQ_MEM_RECLAIM | WQ_HIGHPRI,
>>>>> +			"cmdq_task_wait_release");
>>>>> +
>>>>> +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
>>>>> +		cmdq->thread[i].base = cmdq->base + CMDQ_THR_BASE +
>>>>> +				CMDQ_THR_SHIFT * i;
>>>>> +		init_waitqueue_head(&cmdq->thread[i].wait_task_done);
>>>>> +		INIT_LIST_HEAD(&cmdq->thread[i].task_busy_list);
>>>>> +	}
>>>>> +
>>>>> +	platform_set_drvdata(pdev, cmdq);
>>>>> +
>>>>> +	err = devm_request_irq(dev, cmdq->irq, cmdq_irq_handler, IRQF_SHARED,
>>>>> +			       CMDQ_DRIVER_DEVICE_NAME, cmdq);
>>>>> +	if (err < 0) {
>>>>> +		dev_err(dev, "failed to register ISR (%d)\n", err);
>>>>> +		goto fail;
>>>>> +	}
>>>>> +
>>>>> +	cmdq->clock = devm_clk_get(dev, CMDQ_CLK_NAME);
>>>>> +	if (IS_ERR(cmdq->clock)) {
>>>>> +		dev_err(dev, "failed to get clk:%s\n", CMDQ_CLK_NAME);
>>>>> +		err = PTR_ERR(cmdq->clock);
>>>>> +		goto fail;
>>>>> +	}
>>>>> +	return 0;
>>>>> +
>>>>> +fail:
>>>>> +	cmdq_remove(pdev);
>>>>> +	return err;
>>>>> +}
>>>>> +
>>>>> +static const struct dev_pm_ops cmdq_pm_ops = {
>>>>> +	.suspend = cmdq_suspend,
>>>>> +	.resume = cmdq_resume,
>>>>> +};
>>>>> +
>>>>> +static const struct of_device_id cmdq_of_ids[] = {
>>>>> +	{.compatible = "mediatek,mt8173-gce",},
>>>>> +	{}
>>>>> +};
>>>>> +
>>>>> +static struct platform_driver cmdq_drv = {
>>>>> +	.probe = cmdq_probe,
>>>>> +	.remove = cmdq_remove,
>>>>> +	.driver = {
>>>>> +		.name = CMDQ_DRIVER_DEVICE_NAME,
>>>>> +		.owner = THIS_MODULE,
>>>>> +		.pm = &cmdq_pm_ops,
>>>>> +		.of_match_table = cmdq_of_ids,
>>>>> +	}
>>>>> +};
>>>>> +
>>>>> +builtin_platform_driver(cmdq_drv);
>>>>> diff --git a/include/soc/mediatek/cmdq.h b/include/soc/mediatek/cmdq.h
>>>>> new file mode 100644
>>>>> index 0000000..60eef3d
>>>>> --- /dev/null
>>>>> +++ b/include/soc/mediatek/cmdq.h
>>>>> @@ -0,0 +1,197 @@
>>>>> +/*
>>>>> + * Copyright (c) 2015 MediaTek Inc.
>>>>> + *
>>>>> + * This program is free software; you can redistribute it and/or modify
>>>>> + * it under the terms of the GNU General Public License version 2 as
>>>>> + * published by the Free Software Foundation.
>>>>> + *
>>>>> + * This program is distributed in the hope that it will be useful,
>>>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>>>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
>>>>> + * GNU General Public License for more details.
>>>>> + */
>>>>> +
>>>>> +#ifndef __MTK_CMDQ_H__
>>>>> +#define __MTK_CMDQ_H__
>>>>> +
>>>>> +#include <linux/platform_device.h>
>>>>> +#include <linux/types.h>
>>>>> +
>>>>> +enum cmdq_eng {
>>>>> +	CMDQ_ENG_DISP_AAL,
>>>>> +	CMDQ_ENG_DISP_COLOR0,
>>>>> +	CMDQ_ENG_DISP_COLOR1,
>>>>> +	CMDQ_ENG_DISP_DPI0,
>>>>> +	CMDQ_ENG_DISP_DSI0,
>>>>> +	CMDQ_ENG_DISP_DSI1,
>>>>> +	CMDQ_ENG_DISP_GAMMA,
>>>>> +	CMDQ_ENG_DISP_OD,
>>>>> +	CMDQ_ENG_DISP_OVL0,
>>>>> +	CMDQ_ENG_DISP_OVL1,
>>>>> +	CMDQ_ENG_DISP_PWM0,
>>>>> +	CMDQ_ENG_DISP_PWM1,
>>>>> +	CMDQ_ENG_DISP_RDMA0,
>>>>> +	CMDQ_ENG_DISP_RDMA1,
>>>>> +	CMDQ_ENG_DISP_RDMA2,
>>>>> +	CMDQ_ENG_DISP_UFOE,
>>>>> +	CMDQ_ENG_DISP_WDMA0,
>>>>> +	CMDQ_ENG_DISP_WDMA1,
>>>>> +	CMDQ_ENG_MAX,
>>>>> +};
>>>>> +
>>>>> +/* events for CMDQ and display */
>>>>> +enum cmdq_event {
>>>>> +	/* Display start of frame(SOF) events */
>>>>> +	CMDQ_EVENT_DISP_OVL0_SOF = 11,
>>>>> +	CMDQ_EVENT_DISP_OVL1_SOF = 12,
>>>>> +	CMDQ_EVENT_DISP_RDMA0_SOF = 13,
>>>>> +	CMDQ_EVENT_DISP_RDMA1_SOF = 14,
>>>>> +	CMDQ_EVENT_DISP_RDMA2_SOF = 15,
>>>>> +	CMDQ_EVENT_DISP_WDMA0_SOF = 16,
>>>>> +	CMDQ_EVENT_DISP_WDMA1_SOF = 17,
>>>>> +	/* Display end of frame(EOF) events */
>>>>> +	CMDQ_EVENT_DISP_OVL0_EOF = 39,
>>>>> +	CMDQ_EVENT_DISP_OVL1_EOF = 40,
>>>>> +	CMDQ_EVENT_DISP_RDMA0_EOF = 41,
>>>>> +	CMDQ_EVENT_DISP_RDMA1_EOF = 42,
>>>>> +	CMDQ_EVENT_DISP_RDMA2_EOF = 43,
>>>>> +	CMDQ_EVENT_DISP_WDMA0_EOF = 44,
>>>>> +	CMDQ_EVENT_DISP_WDMA1_EOF = 45,
>>>>> +	/* Mutex end of frame(EOF) events */
>>>>> +	CMDQ_EVENT_MUTEX0_STREAM_EOF = 53,
>>>>> +	CMDQ_EVENT_MUTEX1_STREAM_EOF = 54,
>>>>> +	CMDQ_EVENT_MUTEX2_STREAM_EOF = 55,
>>>>> +	CMDQ_EVENT_MUTEX3_STREAM_EOF = 56,
>>>>> +	CMDQ_EVENT_MUTEX4_STREAM_EOF = 57,
>>>>> +	/* Display underrun events */
>>>>> +	CMDQ_EVENT_DISP_RDMA0_UNDERRUN = 63,
>>>>> +	CMDQ_EVENT_DISP_RDMA1_UNDERRUN = 64,
>>>>> +	CMDQ_EVENT_DISP_RDMA2_UNDERRUN = 65,
>>>>> +	/* Keep this at the end of HW events */
>>>>> +	CMDQ_MAX_HW_EVENT_COUNT = 260,
>>>>> +};
>>>>> +
>>>>> +struct cmdq_cb_data {
>>>>> +	bool	err;
>>>>> +	void	*data;
>>>>> +};
>>>>> +
>>>>> +typedef int (*cmdq_async_flush_cb)(struct cmdq_cb_data data);
>>>>> +
>>>>> +struct cmdq_task;
>>>>> +struct cmdq;
>>>>> +
>>>>> +struct cmdq_rec {
>>>>> +	struct cmdq		*cmdq;
>>>>> +	u64			engine_flag;
>>>>> +	size_t			command_size;
>>>>> +	void			*buf;
>>>>> +	size_t			buf_size;
>>>>> +	bool			finalized;
>>>>> +};
>>
>> Why do we need cmdq_rec at all? Can't we just use the cmdq_task for that
>> and this way make the driver less complex?
>
> There are two main reasons for cmdq_rec.
> 1. It is slow to access dma too frequently.
>     So, we append commands to cacheable memory, and then flush to dma.
> 2. cmdq_rec is not thread safe, but cmdq_task needs thread safe.
>     If we merge them, we need to take care of some synchronization
>     issues.
>
>>>>> +
>>>>> +struct cmdq_base {
>>>>> +	int	subsys;
>>>>> +	u32	base;
>>>>
>>>> subsys can always be calculated via cmdq_subsys_base_to_id(base >> 16)
>>>> so we can get rid of the struct, right?
>>>
>>> Current subsys method is based on previous comment from Daniel Kurtz.
>>> Please take a look of our previous discussion.
>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/004483.html
>>> Thanks.
>>>
>>
>> I have to look deeper into this, but from what I read, the proposal from
>> Daniel [1] seems good to me.
>>
>> [1] https://patchwork.kernel.org/patch/8068311/
>
> The initial proposal has some problem, so please see the follow-up
> discussions. Thanks.
> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/003972.html
> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/004483.html
>
>
>>>>> +};
>>>>> +
>>>>> +/**
>>>>> + * cmdq_register_device() - register device which needs CMDQ
>>>>> + * @dev:		device
>>>>> + *
>>>>> + * Return: cmdq_base pointer or NULL for failed
>>>>> + */
>>>>> +struct cmdq_base *cmdq_register_device(struct device *dev);
>>>>> +
>>>>> +/**
>>>>> + * cmdq_rec_create() - create command queue record
>>>>> + * @dev:		device
>>>>> + * @engine_flag:	command queue engine flag
>>>>> + * @rec_ptr:		command queue record pointer to retrieve cmdq_rec
>>>>> + *
>>>>> + * Return: 0 for success; else the error code is returned
>>>>> + */
>>>>> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
>>>>> +		    struct cmdq_rec **rec_ptr);
>>>>> +
>>>>> +/**
>>>>> + * cmdq_rec_write() - append write command to the command queue record
>>>>> + * @rec:	the command queue record
>>>>> + * @value:	the specified target register value
>>>>> + * @base:	the command queue base
>>>>> + * @offset:	register offset from module base
>>>>> + *
>>>>> + * Return: 0 for success; else the error code is returned
>>>>> + */
>>>>> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value,
>>>>> +		   struct cmdq_base *base, u32 offset);
>>>>> +
>>>>> +/**
>>>>> + * cmdq_rec_write_mask() - append write command with mask to the command
>>>>> + *			   queue record
>>>>> + * @rec:	the command queue record
>>>>> + * @value:	the specified target register value
>>>>> + * @base:	the command queue base
>>>>> + * @offset:	register offset from module base
>>>>> + * @mask:	the specified target register mask
>>>>> + *
>>>>> + * Return: 0 for success; else the error code is returned
>>>>> + */
>>>>> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
>>>>> +			struct cmdq_base *base, u32 offset, u32 mask);
>>>>> +
>>>>> +/**
>>>>> + * cmdq_rec_wfe() - append wait for event command to the command queue reco	rd
>>>>
>>>> reco rd -> record
>>>
>>> Will fix it.
>>>
>>>> Regards,
>>>> Matthias
>>>>
>>>>> + * @rec:	the command queue record
>>>>> + * @event:	the desired event type to "wait and CLEAR"
>>>>> + *
>>>>> + * Return: 0 for success; else the error code is returned
>>>>> + */
>>>>> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event);
>>>>> +
>>>>> +/**
>>>>> + * cmdq_rec_clear_event() - append clear event command to the command queue
>>>>> + *			    record
>>>>> + * @rec:	the command queue record
>>>>> + * @event:	the desired event to be cleared
>>>>> + *
>>>>> + * Return: 0 for success; else the error code is returned
>>>>> + */
>>>>> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event);
>>>>> +
>>>>> +/**
>>>>> + * cmdq_rec_flush() - trigger CMDQ to execute the recorded commands
>>>>> + * @rec:	the command queue record
>>>>> + *
>>>>> + * Return: 0 for success; else the error code is returned
>>>>> + *
>>>>> + * Trigger CMDQ to execute the recorded commands. Note that this is a
>>>>> + * synchronous flush function. When the function returned, the recorded
>>>>> + * commands have been done.
>>>>> + */
>>>>> +int cmdq_rec_flush(struct cmdq_rec *rec);
>>>>> +
>>>>> +/**
>>>>> + * cmdq_rec_flush_async() - trigger CMDQ to asynchronously execute the recorded
>>>>> + *			    commands and call back after ISR is finished
>>>>> + * @rec:	the command queue record
>>>>> + * @cb:		called in the end of CMDQ ISR
>>>>> + * @data:	this data will pass back to cb
>>>>> + *
>>>>> + * Return: 0 for success; else the error code is returned
>>>>> + *
>>>>> + * Trigger CMDQ to asynchronously execute the recorded commands and call back
>>>>> + * after ISR is finished. Note that this is an ASYNC function. When the function
>>>>> + * returned, it may or may not be finished. The ISR callback function is called
>>>>> + * in the end of ISR.
>>
>> "The callback is called from the ISR."
>>
>> Regards,
>> Matthias
>>
>>>>> + */
>>>>> +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
>>>>> +			 void *data);
>>>>> +
>>>>> +/**
>>>>> + * cmdq_rec_destroy() - destroy command queue record
>>>>> + * @rec:	the command queue record
>>>>> + */
>>>>> +void cmdq_rec_destroy(struct cmdq_rec *rec);
>>>>> +
>>>>> +#endif	/* __MTK_CMDQ_H__ */
>>>>>
>>>
>>> Thanks,
>>> HS
>>>
>
> Thanks,
> HS
>

[e-mail] June 3, 2016, 6:12 a.m. UTC | #8

Hi Mathias,

Please see my inline reply.

On Thu, 2016-06-02 at 10:46 +0200, Matthias Brugger wrote:
> 
> On 01/06/16 11:57, Horng-Shyang Liao wrote:
> > Hi Mathias,
> >
> > Please see my inline reply.
> >
> > On Tue, 2016-05-31 at 22:04 +0200, Matthias Brugger wrote:
> >>
> >> On 31/05/16 10:36, Horng-Shyang Liao wrote:
> >>> Hi Mathias,
> >>>
> >>> Please see my inline reply.
> >>>
> >>> On Mon, 2016-05-30 at 17:31 +0200, Matthias Brugger wrote:
> >>>>
> >>>> On 30/05/16 05:19, HS Liao wrote:
> >>>>> This patch is first version of Mediatek Command Queue(CMDQ) driver. The
> >>>>> CMDQ is used to help read/write registers with critical time limitation,
> >>>>> such as updating display configuration during the vblank. It controls
> >>>>> Global Command Engine (GCE) hardware to achieve this requirement.
> >>>>> Currently, CMDQ only supports display related hardwares, but we expect
> >>>>> it can be extended to other hardwares for future requirements.
> >>>>>
> >>>>> Signed-off-by: HS Liao <[e-mail]>
> >>>>> Signed-off-by: CK Hu <[e-mail]>
> >>>>> ---
> >>>>>     drivers/soc/mediatek/Kconfig    |  10 +
> >>>>>     drivers/soc/mediatek/Makefile   |   1 +
> >>>>>     drivers/soc/mediatek/mtk-cmdq.c | 943 ++++++++++++++++++++++++++++++++++++++++
> >>>>>     include/soc/mediatek/cmdq.h     | 197 +++++++++
> >>>>>     4 files changed, 1151 insertions(+)
> >>>>>     create mode 100644 drivers/soc/mediatek/mtk-cmdq.c
> >>>>>     create mode 100644 include/soc/mediatek/cmdq.h
> >>>>>
> >>>>> diff --git a/drivers/soc/mediatek/Kconfig b/drivers/soc/mediatek/Kconfig
> >>>>> index 0a4ea80..c4ad75c 100644
> >>>>> --- a/drivers/soc/mediatek/Kconfig
> >>>>> +++ b/drivers/soc/mediatek/Kconfig
> >>>>> @@ -1,6 +1,16 @@
> >>>>>     #
> >>>>>     # MediaTek SoC drivers
> >>>>>     #
> >>>>> +config MTK_CMDQ
> >>>>> +	bool "MediaTek CMDQ Support"
> >>>>> +	depends on ARCH_MEDIATEK || COMPILE_TEST
> >>
> >> depends on ARM64 ?
> >
> > Will add ARM64.
> >
> >>>>> +	select MTK_INFRACFG
> >>>>> +	help
> >>>>> +	  Say yes here to add support for the MediaTek Command Queue (CMDQ)
> >>>>> +	  driver. The CMDQ is used to help read/write registers with critical
> >>>>> +	  time limitation, such as updating display configuration during the
> >>>>> +	  vblank.
> >>>>> +
> >>>>>     config MTK_INFRACFG
> >>>>>     	bool "MediaTek INFRACFG Support"
> >>>>>     	depends on ARCH_MEDIATEK || COMPILE_TEST
> >>>>> diff --git a/drivers/soc/mediatek/Makefile b/drivers/soc/mediatek/Makefile
> >>>>> index 12998b0..f7397ef 100644
> >>>>> --- a/drivers/soc/mediatek/Makefile
> >>>>> +++ b/drivers/soc/mediatek/Makefile
> >>>>> @@ -1,3 +1,4 @@
> >>>>> +obj-$(CONFIG_MTK_CMDQ) += mtk-cmdq.o
> >>>>>     obj-$(CONFIG_MTK_INFRACFG) += mtk-infracfg.o
> >>>>>     obj-$(CONFIG_MTK_PMIC_WRAP) += mtk-pmic-wrap.o
> >>>>>     obj-$(CONFIG_MTK_SCPSYS) += mtk-scpsys.o
> >>>>> diff --git a/drivers/soc/mediatek/mtk-cmdq.c b/drivers/soc/mediatek/mtk-cmdq.c
> >>>>> new file mode 100644
> >>>>> index 0000000..e9d6e1c
> >>>>> --- /dev/null
> >>>>> +++ b/drivers/soc/mediatek/mtk-cmdq.c
> >>>>> @@ -0,0 +1,943 @@
> >>>>> +/*
> >>>>> + * Copyright (c) 2015 MediaTek Inc.
> >>>>> + *
> >>>>> + * This program is free software; you can redistribute it and/or modify
> >>>>> + * it under the terms of the GNU General Public License version 2 as
> >>>>> + * published by the Free Software Foundation.
> >>>>> + *
> >>>>> + * This program is distributed in the hope that it will be useful,
> >>>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> >>>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> >>>>> + * GNU General Public License for more details.
> >>>>> + */
> >>>>> +
> >>>>> +#include <linux/clk.h>
> >>>>> +#include <linux/clk-provider.h>
> >>>>> +#include <linux/completion.h>
> >>>>> +#include <linux/dma-mapping.h>
> >>>>> +#include <linux/errno.h>
> >>>>> +#include <linux/interrupt.h>
> >>>>> +#include <linux/iopoll.h>
> >>>>> +#include <linux/kernel.h>
> >>>>> +#include <linux/kthread.h>
> >>>>> +#include <linux/module.h>
> >>>>> +#include <linux/mutex.h>
> >>>>> +#include <linux/of_address.h>
> >>>>> +#include <linux/of_irq.h>
> >>>>> +#include <linux/platform_device.h>
> >>>>> +#include <linux/slab.h>
> >>>>> +#include <linux/spinlock.h>
> >>>>> +#include <linux/suspend.h>
> >>>>> +#include <linux/workqueue.h>
> >>>>> +#include <soc/mediatek/cmdq.h>
> >>>>> +
> >>>>> +#define CMDQ_INITIAL_CMD_BLOCK_SIZE	PAGE_SIZE
> >>>>> +#define CMDQ_INST_SIZE			8 /* instruction is 64-bit */
> >>>>> +#define CMDQ_TIMEOUT_MS			1000
> >>>>> +#define CMDQ_IRQ_MASK			0xffff
> >>>>> +#define CMDQ_DRIVER_DEVICE_NAME		"mtk_cmdq"
> >>>>> +#define CMDQ_CLK_NAME			"gce"
> >>>>
> >>>> We can put the names in directly to un-bloat the defines.
> >>>
> >>> I will use the names directly and remove defines.
> >>>
> >>>>> +
> >>>>> +#define CMDQ_CURR_IRQ_STATUS		0x010
> >>>>> +#define CMDQ_CURR_LOADED_THR		0x018
> >>>>> +#define CMDQ_THR_SLOT_CYCLES		0x030
> >>>>> +
> >>>>> +#define CMDQ_THR_BASE			0x100
> >>>>> +#define CMDQ_THR_SHIFT			0x080
> >>>>
> >>>> Wouldn't be CMDQ_THR_SIZE more accurate?
> >>>
> >>> Will rename it.
> >>>
> >>>>> +#define CMDQ_THR_WARM_RESET		0x00
> >>>>> +#define CMDQ_THR_ENABLE_TASK		0x04
> >>>>> +#define CMDQ_THR_SUSPEND_TASK		0x08
> >>>>> +#define CMDQ_THR_CURR_STATUS		0x0c
> >>>>> +#define CMDQ_THR_IRQ_STATUS		0x10
> >>>>> +#define CMDQ_THR_IRQ_ENABLE		0x14
> >>>>> +#define CMDQ_THR_CURR_ADDR		0x20
> >>>>> +#define CMDQ_THR_END_ADDR		0x24
> >>>>> +#define CMDQ_THR_CFG			0x40
> >>>>> +
> >>>>> +#define CMDQ_THR_ENABLED		0x1
> >>>>> +#define CMDQ_THR_DISABLED		0x0
> >>>>> +#define CMDQ_THR_SUSPEND		0x1
> >>>>> +#define CMDQ_THR_RESUME			0x0
> >>>>> +#define CMDQ_THR_STATUS_SUSPENDED	BIT(1)
> >>>>> +#define CMDQ_THR_DO_WARM_RESET		BIT(0)
> >>>>> +#define CMDQ_THR_ACTIVE_SLOT_CYCLES	0x3200
> >>>>> +#define CMDQ_THR_PRIORITY		3
> >>>>> +#define CMDQ_THR_IRQ_DONE		0x1
> >>>>> +#define CMDQ_THR_IRQ_ERROR		0x12
> >>>>> +#define CMDQ_THR_IRQ_EN			0x13 /* done + error */
> >>>>
> >>>> #define CMDQ_THR_IRQ_EN (CMDQ_THR_IRQ_ERROR | CMDQ_THR_IRQ_DONE)
> >>>
> >>> Will do.
> >>>
> >>>>> +#define CMDQ_THR_IRQ_MASK		0x13
> >>>>
> >>>> never used.
> >>>
> >>> Will remove.
> >>>
> >>>>> +#define CMDQ_THR_EXECUTING		BIT(31)
> >>>>> +
> >>>>> +#define CMDQ_ARG_A_WRITE_MASK		0xffff
> >>>>> +#define CMDQ_SUBSYS_MASK		0x1f
> >>>>> +#define CMDQ_OP_CODE_MASK		0xff000000
> >>>>> +
> >>>>> +#define CMDQ_OP_CODE_SHIFT		24
> >>>>
> >>>> Couldn't we connect the mask with the shift, or aren't they related?
> >>>>
> >>>> #define CMDQ_OP_CODE_MASK (0xff << CMDQ_OP_CODE_SHIFT)
> >>>
> >>> Will do.
> >>>
> >>>>> +#define CMDQ_SUBSYS_SHIFT		16
> >>>>> +
> >>>>> +#define CMDQ_WRITE_ENABLE_MASK		BIT(0)
> >>>>> +#define CMDQ_JUMP_BY_OFFSET		0x10000000
> >>>>> +#define CMDQ_JUMP_BY_PA			0x10000001
> >>>>> +#define CMDQ_JUMP_PASS			CMDQ_INST_SIZE
> >>>>> +#define CMDQ_WFE_UPDATE			BIT(31)
> >>>>> +#define CMDQ_WFE_WAIT			BIT(15)
> >>>>> +#define CMDQ_WFE_WAIT_VALUE		0x1
> >>>>> +#define CMDQ_EOC_IRQ_EN			BIT(0)
> >>>>> +
> >>>>> +enum cmdq_thread_index {
> >>>>> +	CMDQ_THR_DISP_MAIN_IDX,	/* main */
> >>>>> +	CMDQ_THR_DISP_SUB_IDX,	/* sub */
> >>>>> +	CMDQ_THR_DISP_MISC_IDX,	/* misc */
> >>>>> +	CMDQ_THR_MAX_COUNT,	/* max */
> >>>>> +};
> >>>>> +
> >>>>> +/*
> >>>>> + * CMDQ_CODE_MOVE:
> >>>>> + *   move value into internal register as mask
> >>>>> + *   format: op mask
> >>>>> + * CMDQ_CODE_WRITE:
> >>>>> + *   write value into target register
> >>>>> + *   format: op subsys address value
> >>>>> + * CMDQ_CODE_JUMP:
> >>>>> + *   jump by offset
> >>>>> + *   format: op offset
> >>>>> + * CMDQ_CODE_WFE:
> >>>>> + *   wait for event and clear
> >>>>> + *   it is just clear if no wait
> >>>>> + *   format: [wait]  op event update:1 to_wait:1 wait:1
> >>>>> + *           [clear] op event update:1 to_wait:0 wait:0
> >>>>> + * CMDQ_CODE_EOC:
> >>>>> + *   end of command
> >>>>> + *   format: op irq_flag
> >>>>> + */
> >>>>
> >>>> I think we need more documentation of how this command queue engine is
> >>>> working. If not, I think it will be really complicated to understand how
> >>>> to use this.
> >>>>
> >>>>> +enum cmdq_code {
> >>>>> +	CMDQ_CODE_MOVE = 0x02,
> >>>>> +	CMDQ_CODE_WRITE = 0x04,
> >>>>> +	CMDQ_CODE_JUMP = 0x10,
> >>>>> +	CMDQ_CODE_WFE = 0x20,
> >>>>> +	CMDQ_CODE_EOC = 0x40,
> >>>>> +};
> >>>>> +
> >>>>> +enum cmdq_task_state {
> >>>>> +	TASK_STATE_BUSY, /* running on a GCE thread */
> >>>>> +	TASK_STATE_ERROR,
> >>>>> +	TASK_STATE_DONE,
> >>>>> +};
> >>>>> +
> >>>>> +struct cmdq_task_cb {
> >>>>> +	cmdq_async_flush_cb	cb;
> >>>>> +	void			*data;
> >>>>> +};
> >>>>> +
> >>>>> +struct cmdq_thread {
> >>>>> +	void __iomem		*base;
> >>>>> +	struct list_head	task_busy_list;
> >>>>> +	wait_queue_head_t	wait_task_done;
> >>>>> +};
> >>>>> +
> >>>>> +struct cmdq_task {
> >>>>> +	struct cmdq		*cmdq;
> >>>>> +	struct list_head	list_entry;
> >>>>> +	enum cmdq_task_state	task_state;
> >>>>> +	void			*va_base;
> >>>>> +	dma_addr_t		pa_base;
> >>>>> +	u64			engine_flag;
> >>>>> +	size_t			command_size;
> >>>>> +	u32			num_cmd;
> >>>>
> >>>> num_cmd is directly connected to command_size. I prefer to just keep
> >>>> num_cmd and calculate the command_size where necessary.
> >>>
> >>> After I trace code, I prefer to keep command_size and calculate num_cmd
> >>> where necessary. What do you think?
> >>>
> >>
> >> I suppose you prefer this, as you are writing to the GCE depending on
> >> the command_size. I think it is worth to create a macro for the
> >> calculation of the number of commands, to make the code more readable.
> >> Would be nice if you would just pass cmdq_task to it and it would return
> >> the number. Just as an idea.
> >
> > Will add macro.
> >
> >>>>> +	struct cmdq_thread	*thread;
> >>>>> +	struct cmdq_task_cb	cb;
> >>>>> +	struct work_struct	release_work;
> >>>>> +};
> >>>>> +
> >>>>> +struct cmdq {
> >>>>> +	struct device		*dev;
> >>>>> +	void __iomem		*base;
> >>>>> +	u32			irq;
> >>>>> +	struct workqueue_struct	*task_release_wq;
> >>>>> +	struct cmdq_thread	thread[CMDQ_THR_MAX_COUNT];
> >>>>> +	struct mutex		task_mutex;	/* for task */
> >>>>> +	spinlock_t		exec_lock;	/* for exec */
> >>>>> +	struct clk		*clock;
> >>>>> +	bool			suspended;
> >>>>> +};
> >>>>> +
> >>>>> +struct cmdq_subsys {
> >>>>> +	u32	base;
> >>>>> +	int	id;
> >>>>> +};
> >>>>> +
> >>>>> +static const struct cmdq_subsys gce_subsys[] = {
> >>>>> +	{0x1400, 1},
> >>>>> +	{0x1401, 2},
> >>>>> +	{0x1402, 3},
> >>>>> +};
> >>>>> +
> >>>>> +static int cmdq_subsys_base_to_id(u32 base)
> >>>>> +{
> >>>>> +	int i;
> >>>>> +
> >>>>> +	for (i = 0; i < ARRAY_SIZE(gce_subsys); i++)
> >>>>> +		if (gce_subsys[i].base == base)
> >>>>> +			return gce_subsys[i].id;
> >>>>> +	return -EFAULT;
> >>>>> +}
> >>>>> +
> >>>>> +static int cmdq_eng_get_thread(u64 flag)
> >>>>> +{
> >>>>> +	if (flag & BIT_ULL(CMDQ_ENG_DISP_DSI0))
> >>>>> +		return CMDQ_THR_DISP_MAIN_IDX;
> >>>>> +	else if (flag & BIT_ULL(CMDQ_ENG_DISP_DPI0))
> >>>>> +		return CMDQ_THR_DISP_SUB_IDX;
> >>>>> +	else
> >>>>> +		return CMDQ_THR_DISP_MISC_IDX;
> >>>>> +}
> >>>>> +
> >>>>> +static void cmdq_task_release(struct cmdq_task *task)
> >>>>> +{
> >>>>> +	struct cmdq *cmdq = task->cmdq;
> >>>>> +
> >>>>> +	dma_free_coherent(cmdq->dev, task->command_size, task->va_base,
> >>>>> +			  task->pa_base);
> >>>>> +	kfree(task);
> >>>>> +}
> >>>>> +
> >>>>> +static struct cmdq_task *cmdq_task_acquire(struct cmdq_rec *rec,
> >>>>> +					   struct cmdq_task_cb cb)
> >>>>> +{
> >>>>> +	struct cmdq *cmdq = rec->cmdq;
> >>>>> +	struct device *dev = cmdq->dev;
> >>>>> +	struct cmdq_task *task;
> >>>>> +
> >>>>> +	task = kzalloc(sizeof(*task), GFP_KERNEL);
> >>>>> +	INIT_LIST_HEAD(&task->list_entry);
> >>>>> +	task->va_base = dma_alloc_coherent(dev, rec->command_size,
> >>>>> +					   &task->pa_base, GFP_KERNEL);
> >>>>> +	if (!task->va_base) {
> >>>>> +		dev_err(dev, "allocate command buffer failed\n");
> >>>>> +		kfree(task);
> >>>>> +		return NULL;
> >>>>> +	}
> >>>>> +
> >>>>> +	task->cmdq = cmdq;
> >>>>> +	task->command_size = rec->command_size;
> >>>>> +	task->engine_flag = rec->engine_flag;
> >>>>> +	task->task_state = TASK_STATE_BUSY;
> >>>>> +	task->cb = cb;
> >>>>> +	memcpy(task->va_base, rec->buf, rec->command_size);
> >>>>> +	task->num_cmd = task->command_size / CMDQ_INST_SIZE;
> >>>>> +	return task;
> >>>>> +}
> >>>>> +
> >>>>> +static void cmdq_thread_writel(struct cmdq_thread *thread, u32 value,
> >>>>> +			       u32 offset)
> >>>>> +{
> >>>>> +	writel(value, thread->base + offset);
> >>>>> +}
> >>>>> +
> >>>>> +static u32 cmdq_thread_readl(struct cmdq_thread *thread, u32 offset)
> >>>>> +{
> >>>>> +	return readl(thread->base + offset);
> >>>>> +}
> >>>>
> >>>> We can get rid of cmdq_thread_readl/writel.
> >>>
> >>> Will do.
> >>>
> >>>>> +
> >>>>> +static int cmdq_thread_suspend(struct cmdq *cmdq, struct cmdq_thread *thread)
> >>>>> +{
> >>>>> +	u32 status;
> >>>>> +
> >>>>> +	cmdq_thread_writel(thread, CMDQ_THR_SUSPEND, CMDQ_THR_SUSPEND_TASK);
> >>>>> +
> >>>>> +	/* If already disabled, treat as suspended successful. */
> >>>>> +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> >>>>> +	      CMDQ_THR_ENABLED))
> >>>>> +		return 0;
> >>>>> +
> >>>>> +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_CURR_STATUS,
> >>>>> +			status, status & CMDQ_THR_STATUS_SUSPENDED, 0, 10)) {
> >>>>> +		dev_err(cmdq->dev, "suspend GCE thread 0x%x failed\n",
> >>>>> +			(u32)(thread->base - cmdq->base));
> >>>>> +		return -EFAULT;
> >>>>> +	}
> >>>>> +
> >>>>> +	return 0;
> >>>>> +}
> >>>>> +
> >>>>> +static void cmdq_thread_resume(struct cmdq_thread *thread)
> >>>>> +{
> >>>>> +	cmdq_thread_writel(thread, CMDQ_THR_RESUME, CMDQ_THR_SUSPEND_TASK);
> >>>>> +}
> >>>>> +
> >>>>> +static int cmdq_thread_reset(struct cmdq *cmdq, struct cmdq_thread *thread)
> >>>>> +{
> >>>>> +	u32 warm_reset;
> >>>>> +
> >>>>> +	cmdq_thread_writel(thread, CMDQ_THR_DO_WARM_RESET, CMDQ_THR_WARM_RESET);
> >>>>> +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_WARM_RESET,
> >>>>> +			warm_reset, !(warm_reset & CMDQ_THR_DO_WARM_RESET),
> >>>>> +			0, 10)) {
> >>>>> +		dev_err(cmdq->dev, "reset GCE thread 0x%x failed\n",
> >>>>> +			(u32)(thread->base - cmdq->base));
> >>>>> +		return -EFAULT;
> >>>>> +	}
> >>>>> +	writel(CMDQ_THR_ACTIVE_SLOT_CYCLES, cmdq->base + CMDQ_THR_SLOT_CYCLES);
> >>>>> +	return 0;
> >>>>> +}
> >>>>> +
> >>>>> +static void cmdq_thread_disable(struct cmdq *cmdq, struct cmdq_thread *thread)
> >>>>> +{
> >>>>> +	cmdq_thread_reset(cmdq, thread);
> >>>>> +	cmdq_thread_writel(thread, CMDQ_THR_DISABLED, CMDQ_THR_ENABLE_TASK);
> >>>>> +}
> >>>>> +
> >>>>> +/* notify GCE to re-fetch commands by setting GCE thread PC */
> >>>>> +static void cmdq_thread_invalidate_fetched_data(struct cmdq_thread *thread)
> >>>>> +{
> >>>>> +	cmdq_thread_writel(thread,
> >>>>> +			   cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR),
> >>>>> +			   CMDQ_THR_CURR_ADDR);
> >>>>> +}
> >>>>> +
> >>>>> +static void cmdq_task_insert_into_thread(struct cmdq_task *task)
> >>>>> +{
> >>>>> +	struct cmdq_thread *thread = task->thread;
> >>>>> +	struct cmdq_task *prev_task = list_last_entry(
> >>>>> +			&thread->task_busy_list, typeof(*task), list_entry);
> >>>>> +	u64 *prev_task_base = prev_task->va_base;
> >>>>> +
> >>>>> +	/* let previous task jump to this task */
> >>>>> +	prev_task_base[prev_task->num_cmd - 1] = (u64)CMDQ_JUMP_BY_PA << 32 |
> >>>>> +						 task->pa_base;
> >>>>> +
> >>>>> +	cmdq_thread_invalidate_fetched_data(thread);
> >>>>> +}
> >>>>> +
> >>>>> +/* we assume tasks in the same display GCE thread are waiting the same event. */
> >>>>> +static void cmdq_task_remove_wfe(struct cmdq_task *task)
> >>>>> +{
> >>>>> +	u32 wfe_option = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
> >>>>> +	u32 wfe_op = CMDQ_CODE_WFE << CMDQ_OP_CODE_SHIFT;
> >>>>> +	u32 *base = task->va_base;
> >>>>> +	u32 num_cmd = task->num_cmd << 1;
> >>>>> +	int i;
> >>>>> +
> >>>>> +	for (i = 0; i < num_cmd; i += 2)
> >>>>> +		if (base[i] == wfe_option &&
> >>>>> +		    (base[i + 1] & CMDQ_OP_CODE_MASK) == wfe_op) {
> >>>>> +			base[i] = CMDQ_JUMP_PASS;
> >>>>> +			base[i + 1] = CMDQ_JUMP_BY_OFFSET;
> >>>>> +		}
> >>>>
> >>>> After using the command buffer as a void pointer a u64 pointer, we now
> >>>> reference to it as u32. I would prefer to explain here, how the command
> >>>> looks like we are searching for and use a for loop passing task->num_cmd
> >>>> instead.
> >>>
> >>> Will use u64* to rewrite the above code.
> >>>
> >>>>> +}
> >>>>> +
> >>>>> +static void cmdq_task_exec(struct cmdq_task *task, struct cmdq_thread *thread)
> >>>>> +{
> >>>>> +	struct cmdq *cmdq = task->cmdq;
> >>>>> +	unsigned long flags;
> >>>>> +	unsigned long curr_pa, end_pa;
> >>>>> +
> >>>>> +	WARN_ON(clk_prepare_enable(cmdq->clock) < 0);
> >>>>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
> >>>>
> >>>> cmdq_task_exec is called with cmdq->task_mutex held, so why do we need
> >>>> the spin_lock here? Can't we just use one of the two?
> >>>
> >>> We can drop task_mutex, but we will get some side effects.
> >>> 1. exec_lock needs to include more code, but I think it is not good for
> >>>      spinlock.
> >>> 2. In cmdq_rec_flush_async(), task_mutex needs to protect
> >>>      (1) cmdq->suspended, (2) cmdq_task_exec(), and
> >>>      (3) cmdq_task_wait_release_schedule().
> >>>      If we drop task_mutex, we have to put cmdq->suspended if condition
> >>>      just before cmdq_task_exec() and inside exec_lock, and we have to
> >>>      release task and its command buffer if error. This will let flow
> >>>      become more complex and enlarge code size.
> >>>
> >>> What do you think?
> >>
> >> Why do you need to protect cmdq_task_wait_release_schedule? We don't
> >> care about the order of the workqueue elements, do we?
> >
> > According to CK's comment, we have to ensure order to remove previous
> > task.
> > http://lists.infradead.org/pipermail/linux-mediatek/2016-May/005547.html
> >
> >> As far as I understand you would need to protect cmdq_task_acquire as
> >> well, to "ensure" continously growing pa_base. More on that below.
> >
> > We need to ensure continuous physical addresses in a task, but we don't
> > need to ensure continuous physical addresses between tasks.
> > So, I think it is unnecessary to protect by mutex or spinlock.
> >
> 
> True, I didn't get that.
> 
> >>>
> >>>>> +	task->thread = thread;
> >>>>> +	task->task_state = TASK_STATE_BUSY;
> >>>>
> >>>> That was already set in cmdq_task_acquire, why do we need to set it here
> >>>> again?
> >>>
> >>> Will drop it.
> >>>
> >>
> >> Yeah, but I think it makes more sense to drop it in cmdq_task_acquire
> >> instead.
> >
> > Will drop it in cmdq_task_acquire instead.
> >
> >>>>> +	if (list_empty(&thread->task_busy_list)) {
> >>>>> +		WARN_ON(cmdq_thread_reset(cmdq, thread) < 0);
> >>>>> +
> >>>>> +		cmdq_thread_writel(thread, task->pa_base, CMDQ_THR_CURR_ADDR);
> >>>>> +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
> >>>>> +				   CMDQ_THR_END_ADDR);
> >>>>> +		cmdq_thread_writel(thread, CMDQ_THR_PRIORITY, CMDQ_THR_CFG);
> >>>>> +		cmdq_thread_writel(thread, CMDQ_THR_IRQ_EN,
> >>>>> +				   CMDQ_THR_IRQ_ENABLE);
> >>>>> +
> >>>>> +		cmdq_thread_writel(thread, CMDQ_THR_ENABLED,
> >>>>> +				   CMDQ_THR_ENABLE_TASK);
> >>>>> +	} else {
> >>>>> +		WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
> >>>>> +
> >>>>> +		/*
> >>>>> +		 * check boundary condition
> >>>>> +		 * PC = END - 8, EOC is executed
> >>>>> +		 * PC = END, all CMDs are executed
> >>>>> +		 */
> >>>>> +		curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> >>>>> +		end_pa = cmdq_thread_readl(thread, CMDQ_THR_END_ADDR);
> >>>>> +		if (curr_pa == end_pa - 8 || curr_pa == end_pa) {
> >>>>
> >>>> 8 refers to CMDQ_INST_SIZE, right?
> >>>
> >>> Yes, I will use CMDQ_INST_SIZE.
> >>>
> >>>>> +			/* set to this task directly */
> >>>>> +			cmdq_thread_writel(thread, task->pa_base,
> >>>>> +					   CMDQ_THR_CURR_ADDR);
> >>>>> +		} else {
> >>>>> +			cmdq_task_insert_into_thread(task);
> >>>>> +
> >>>>> +			if (thread == &cmdq->thread[CMDQ_THR_DISP_MAIN_IDX] ||
> >>>>> +			    thread == &cmdq->thread[CMDQ_THR_DISP_SUB_IDX])
> >>>>> +				cmdq_task_remove_wfe(task);
> >>>>
> >>>> We could do this check using the task->engine_flag, I suppose that's
> >>>> easier to undestand then.
> >>>
> >>> Will use task->engine_flag.
> >>>
> >>>>> +
> >>>>> +			smp_mb(); /* modify jump before enable thread */
> >>>>> +		}
> >>>>> +
> >>>>> +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
> >>>>> +				   CMDQ_THR_END_ADDR);
> >>>>> +		cmdq_thread_resume(thread);
> >>>>> +	}
> >>>>> +	list_move_tail(&task->list_entry, &thread->task_busy_list);
> >>>>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>>>> +}
> >>>>> +
> >>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
> >>>>> +				   struct cmdq_thread *thread, u32 irq_flag)
> >>>>> +{
> >>>>> +	struct cmdq_task *task, *tmp, *curr_task = NULL;
> >>>>> +	u32 curr_pa;
> >>>>> +	struct cmdq_cb_data cmdq_cb_data;
> >>>>> +	bool err;
> >>>>> +
> >>>>> +	if (irq_flag & CMDQ_THR_IRQ_ERROR)
> >>>>> +		err = true;
> >>>>> +	else if (irq_flag & CMDQ_THR_IRQ_DONE)
> >>>>> +		err = false;
> >>>>> +	else
> >>>>> +		return;
> >>>>> +
> >>>>> +	curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> >>>>> +
> >>>>> +	list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
> >>>>> +				 list_entry) {
> >>>>> +		if (curr_pa >= task->pa_base &&
> >>>>> +		    curr_pa < (task->pa_base + task->command_size))
> >>>>
> >>>> What are you checking here? It seems as if you make some implcit
> >>>> assumptions about pa_base and the order of execution of commands in the
> >>>> thread. Is it save to do so? Does dma_alloc_coherent give any guarantees
> >>>> about dma_handle?
> >>>
> >>> 1. Check what is the current running task in this GCE thread.
> >>> 2. Yes.
> >>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
> >>>
> >>
> >> Yes, physical addresses might be continous, but AFAIK there is no
> >> guarantee that the dma_handle address is steadily growing, when calling
> >> dma_alloc_coherent. And if I understand the code correctly, you use this
> >> assumption to decide if the task picked from task_busy_list is currently
> >> executing. So I think this mecanism is not working.
> >
> > I don't use dma_handle address, and just use physical addresses.
> >  From CPU's point of view, tasks are linked by the busy list.
> >  From GCE's point of view, tasks are linked by the JUMP command.
> >
> >> In which cases does the HW thread raise an interrupt.
> >> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
> >
> > GCE will raise interrupt if any task is done or error.
> > However, GCE is fast, so CPU may get multiple done tasks
> > when it is running ISR.
> >
> > In case of error, that GCE thread will pause and raise interrupt.
> > So, CPU may get multiple done tasks and one error task.
> >
> 
> I think we should reimplement the ISR mechanism. Can't we just read 
> CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave 
> cmdq_handle_error_done to the thread_fn? You will need to pass 
> information from the handler to thread_fn, but that shouldn't be an 
> issue. AFAIK interrupts are disabled in the handler, so we should stay 
> there as short as possible. Traversing task_busy_list is expensive, so 
> we need to do it in a thread context.

Actually, our initial implementation is similar to your suggestion,
but display needs CMDQ to return callback function very precisely,
else display will drop frame.
For display, CMDQ interrupt will be raised every 16 ~ 17 ms,
and CMDQ needs to call callback function in ISR.
If we defer callback to workqueue, the time interval may be larger than
32 ms.sometimes.

> I keep thinking about how to get rid of the two data structures, 
> task_busy_list and the task_release_wq. We need the latter for the only 
> sake of getting a timeout.
> 
> Did you have a look on how the mailbox framework handles this?
> By the way, what is the reason to not implement the whole driver as a 
> mailbox controller? For me, this driver looks like a good fit.

CMDQ needs to encode commands for GCE hardware. We think this behavior
should be put in CMDQ driver, and client just call CMDQ functions.
Therefore, if we want to use mailbox framework, cmdq_rec must be
mailbox client, and the others must be mailbox controller.

However, if we use mailbox controller, CMDQ driver still needs to
control busy list for each GCE thread, and use workqueue to handle
timeout tasks.

The only thing that we can borrow from mailbox framework is the send
(CMDQ flush) and receive (CMDQ callback) interface, However, we don't
think we can gain many benefits from it, and we have some overheads to
conform to mailbox interface.


> 
> >>>>> +			curr_task = task;
> >>>>> +		if (task->cb.cb) {
> >>>>> +			cmdq_cb_data.err = curr_task ? err : false;
> >>>>> +			cmdq_cb_data.data = task->cb.data;
> >>>>> +			task->cb.cb(cmdq_cb_data);
> >>>>> +		}
> >>>>> +		task->task_state = (curr_task && err) ? TASK_STATE_ERROR :
> >>>>> +				TASK_STATE_DONE;
> >>>>> +		list_del(&task->list_entry);
> >>>>> +		if (curr_task)
> >>>>> +			break;
> >>>>> +	}
> >>>>> +
> >>>>> +	wake_up(&thread->wait_task_done);
> >>>>> +}
> >>>>> +
> >>>>> +static void cmdq_thread_irq_handler(struct cmdq *cmdq, int tid)
> >>>>> +{
> >>>>> +	struct cmdq_thread *thread = &cmdq->thread[tid];
> >>>>> +	unsigned long flags = 0L;
> >>>>> +	u32 irq_flag;
> >>>>> +
> >>>>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
> >>>>> +
> >>>>> +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
> >>>>> +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
> >>>>> +
> >>>>> +	/*
> >>>>> +	 * Another CPU core could run "release task" right before we acquire
> >>>>> +	 * the spin lock, and thus reset / disable this GCE thread, so we
> >>>>> +	 * need to check the enable bit of this GCE thread.
> >>>>> +	 */
> >>>>> +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> >>>>> +	      CMDQ_THR_ENABLED))
> >>>>> +		irq_flag = 0;
> >>>>
> >>>> cmdq_handle_error_done just retuns in this case. Programming this way
> >>>> just makes things confusing. What about:
> >>>>
> >>>> if (cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> >>>> 	      CMDQ_THR_ENABLED)
> >>>> 	cmdq_handle_error_done(cmdq, thread, irq_flag);
> >>>> else
> >>>> 		irq_flag = 0;
> >>>>
> >>>> spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>>
> >>> We still need to clear irq_flag if GCE thread is disabled.
> >>> So, I think we can just return here.
> >>>
> >>> 	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> >>> 	      CMDQ_THR_ENABLED))
> >>> 		return;
> >>>
> >>> What do you think?
> >>>
> >>
> >> No, you can't just return, you need to unlock the spinlock.
> >> Anyway I would prefer it the other way round, as I put it in my last
> >> mail. Just delete the else branch, we don't need to set irq_flag to zero.
> >
> > Sorry for my previous wrong reply since I merge your comment
> > and CK's comment.
> > http://lists.infradead.org/pipermail/linux-mediatek/2016-May/005547.html
> > So, I will put this if condition into cmdq_task_handle_error_result()
> > and then just return it if GCE thread is disabled.
> >
> 
> You mean in cmdq_task_handle_done
> We should rename this functions to not create confusion.

Sorry again. I mean in cmdq_handle_error_done().
This function handles both done and error.

I agree the function name looks confusion.
I think it can be renamed to cmdq_thread_irq_handler()
since it is used to handle irq for GCE thread.

> Regards,
> Matthias

Thanks,
HS

> >>>>> +
> >>>>> +	cmdq_handle_error_done(cmdq, thread, irq_flag);
> >>>>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>>>> +}
> >>>>> +
> >>>>> +static irqreturn_t cmdq_irq_handler(int irq, void *dev)
> >>>>> +{
> >>>>> +	struct cmdq *cmdq = dev;
> >>>>> +	u32 irq_status;
> >>>>> +	int i;
> >>>>> +
> >>>>> +	irq_status = readl(cmdq->base + CMDQ_CURR_IRQ_STATUS);
> >>>>> +	irq_status &= CMDQ_IRQ_MASK;
> >>>>> +	irq_status ^= CMDQ_IRQ_MASK;
> >>>>
> >>>> irq_status can be much bigger then 3, which is the number of threads in
> >>>> the system (CMDQ_THR_MAX_COUNT). So why we use this mask here isn't
> >>>> clear to me.
> >>>
> >>> Our GCE hardware has 16 threads, but we only use 3 threads currently.
> >>>
> >>
> >> Ok, but please use bitops here.
> >
> > Will use bitops.
> >
> >>>>> +
> >>>>> +	if (!irq_status)
> >>>>> +		return IRQ_NONE;
> >>>>> +
> >>>>> +	while (irq_status) {
> >>>>> +		i = ffs(irq_status) - 1;
> >>>>> +		irq_status &= ~BIT(i);
> >>>>> +		cmdq_thread_irq_handler(cmdq, i);
> >>>>> +	}
> >>>>
> >>>> Can you explain how the irq status register looks like, that would it
> >>>> make much easier to understand what happens here.
> >>>
> >>> Bit 0 ~ 15 of irq status register represents GCE thread 0 ~ 15
> >>> interrupt. 0 means asserting interrupt; 1 means no interrupt.
> >>>
> >>
> >> Thanks, that helped. :)
> >>
> >>>>> +
> >>>>> +	return IRQ_HANDLED;	
> >>>>> +}
> >>>>> +
> >>>>> +static int cmdq_task_handle_error_result(struct cmdq_task *task)
> >>>>
> >>>> We never check the return values, why do we have them?
> >>>
> >>> Will drop return value.
> >>>
> >>>>> +{
> >>>>> +	struct cmdq *cmdq = task->cmdq;
> >>>>> +	struct device *dev = cmdq->dev;
> >>>>> +	struct cmdq_thread *thread = task->thread;
> >>>>> +	struct cmdq_task *next_task, *prev_task;
> >>>>> +	u32 irq_flag;
> >>>>> +
> >>>>> +	/* suspend GCE thread to ensure consistency */
> >>>>> +	WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
> >>>>> +
> >>>>> +	/* ISR has handled this error task */
> >>>>> +	if (task->task_state == TASK_STATE_ERROR) {
> >>>>> +		next_task = list_first_entry_or_null(&thread->task_busy_list,
> >>>>> +				struct cmdq_task, list_entry);
> >>>>> +		if (next_task) /* move to next task */
> >>>>> +			cmdq_thread_writel(thread, next_task->pa_base,
> >>>>> +					   CMDQ_THR_CURR_ADDR);
> >>>>
> >>>> We have to do this, as we suppose that the thread did not reach the jump
> >>>> instruction we put into it's command queue, right?
> >>>
> >>> Yes.
> >>>
> >>
> >> So this should then go into it's own function. In wait_release_work,
> >> something like this:
> >>
> >> if(task->task_state == TASK_STATE_ERROR)
> >>        cmdq_task_handle_error(task)
> >
> > OK.
> > I will write new function cmdq_task_handle_error() to handle error case.
> >
> >>>>> +		cmdq_thread_resume(thread);
> >>>>> +		return -ECANCELED;
> >>>>> +	}
> >>>>> +
> >>>>
> >>>> if task_state != ERROR and != DONE. This means that the timeout of
> >>>> task_release_wq has timed out, right?
> >>>
> >>> Yes.
> >>>
> >>>>> +	/*
> >>>>> +	 * Save next_task and prev_task in advance
> >>>>> +	 * since cmdq_handle_error_done will remove list_entry.
> >>>>> +	 */
> >>>>> +	next_task = prev_task = NULL;
> >>>>> +	if (task->list_entry.next != &thread->task_busy_list)
> >>>>> +		next_task = list_next_entry(task, list_entry);
> >>>>> +	if (task->list_entry.prev != &thread->task_busy_list)
> >>>>> +		prev_task = list_prev_entry(task, list_entry);
> >>>>> +
> >>>>> +	/*
> >>>>> +	 * Although IRQ is disabled, GCE continues to execute.
> >>>>> +	 * It may have pending IRQ before GCE thread is suspended,
> >>>>> +	 * so check this condition again.
> >>>>> +	 */
> >>>>
> >>>> The first thing we did in this function was suspending the thread. Why
> >>>> do we need this then?
> >>>
> >>> Because timeout is CPU timeout not GCE timeout, GCE could just finish
> >>> this task before the GCE thread is suspended.
> >>>
> >>
> >> What are the reasons for a timeout? An error has happend, or the task is
> >> still executing.
> >
> >  From GCE's point of view, this task is still executing.
> > But, it could be an error of client.
> > For example, task may never get event if display turn off hardware
> > before waiting for task to finish its work.
> >
> >>>> To be honest this whole functions looks really like a design error. We
> >>>> have to sperate the states much clearer so that it is possible to
> >>>> understand what is happening in the GCE. Isn't it for example posible to
> >>>> have worker queues for timed out tasks and tasks with an error? I'm not
> >>>> sure how to do this, actually I'm not sure if I really understood how
> >>>> this is supposed to work.
> >>>
> >>> GCE doesn't have timeout. The timeout is decided and controlled by CPU,
> >>> so we check timeout in release work.
> >>> For error and done, they are easy to check by register, and we have
> >>> already created release work for timeout. So, I don't think we need to
> >>> create work queue for each case.
> >>>
> >>> What do you think?
> >>>
> >>
> >> I think, if we find in here, that the irq_flag is set, then the the
> >> interrupt handler was triggered and is spinning the spinlock. If this is
> >> not the case, we have a timeout and we handle this.
> >
> > I will write another function to handle error, and handle timeout here.
> >
> >>>> How much do we win, when we patch the thread command queue for every
> >>>> task we add, instead of just taking one task after another from the
> >>>> task_busy_list?
> >>>
> >>> GCE is used to help read/write registers with critical time limitation.
> >>> Sometimes, client may ask to process multiple tasks in a short period
> >>> of time, e.g. display flush multiple tasks for next vblank. So, CMDQ
> >>> shouldn't limit to process one task after another from the
> >>> task_busy_list. Currently, when interrupt or timeout, we will check
> >>> how many tasks are done, and which one is error or timeout.
> >>>
> >>
> >> So I suppose the device driver who use this are interested in throughput
> >> and not in latency. The callback of every
> >>
> >>>>> +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
> >>>>> +	cmdq_handle_error_done(cmdq, thread, irq_flag);
> >>>>> +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
> >>>>> +
> >>>>> +	if (task->task_state == TASK_STATE_DONE) {
> >>>>> +		cmdq_thread_resume(thread);
> >>>>> +		return 0;
> >>>>> +	}
> >>>>> +
> >>>>> +	if (task->task_state == TASK_STATE_ERROR) {
> >>>>> +		dev_err(dev, "task 0x%p error\n", task);
> >>>>> +		if (next_task) /* move to next task */
> >>>>> +			cmdq_thread_writel(thread, next_task->pa_base,
> >>>>> +					   CMDQ_THR_CURR_ADDR);
> >>>>> +		cmdq_thread_resume(thread);
> >>>>> +		return -ECANCELED;
> >>>>> +	}
> >>>>> +
> >>>>> +	/* Task is running, so we force to remove it. */
> >>>>> +	dev_err(dev, "task 0x%p timeout or killed\n", task);
> >>>>> +	task->task_state = TASK_STATE_ERROR;
> >>>>> +
> >>>>> +	if (prev_task) {
> >>>>> +		u64 *prev_va = prev_task->va_base;
> >>>>> +		u64 *curr_va = task->va_base;
> >>>>> +
> >>>>> +		/* copy JUMP instruction */
> >>>>> +		prev_va[prev_task->num_cmd - 1] = curr_va[task->num_cmd - 1];
> >>>>> +
> >>>>> +		cmdq_thread_invalidate_fetched_data(thread);
> >>>>> +	} else if (next_task) { /* move to next task */
> >>>>> +		cmdq_thread_writel(thread, next_task->pa_base,
> >>>>> +				   CMDQ_THR_CURR_ADDR);
> >>>>> +	}
> >>>>> +
> >>>>> +	list_del(&task->list_entry);
> >>>>> +	cmdq_thread_resume(thread);
> >>>>> +
> >>>>> +	/* call cb here to prevent lock */
> >>>>> +	if (task->cb.cb) {
> >>>>> +		struct cmdq_cb_data cmdq_cb_data;
> >>>>> +
> >>>>> +		cmdq_cb_data.err = true;
> >>>>> +		cmdq_cb_data.data = task->cb.data;
> >>>>> +		task->cb.cb(cmdq_cb_data);
> >>>>> +	}
> >>>>> +
> >>>>> +	return -ECANCELED;
> >>>>> +}
> >>>>> +
> >>>>> +static void cmdq_task_wait_release_work(struct work_struct *work_item)
> >>>>> +{
> >>>>> +	struct cmdq_task *task = container_of(work_item, struct cmdq_task,
> >>>>> +					      release_work);
> >>>>> +	struct cmdq *cmdq = task->cmdq;
> >>>>> +	struct cmdq_thread *thread = task->thread;
> >>>>> +	unsigned long flags;
> >>>>> +
> >>>>> +	wait_event_timeout(thread->wait_task_done,
> >>>>> +			   task->task_state != TASK_STATE_BUSY,
> >>>>> +			   msecs_to_jiffies(CMDQ_TIMEOUT_MS));
> >>>>> +
> >>>>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
> >>>>> +	if (task->task_state != TASK_STATE_DONE)
> >>>>> +		cmdq_task_handle_error_result(task);
> >>>>> +	if (list_empty(&thread->task_busy_list))
> >>>>> +		cmdq_thread_disable(cmdq, thread);
> >>>>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>>>> +
> >>>>> +	/* release regardless of success or not */
> >>>>> +	clk_disable_unprepare(cmdq->clock);
> >>>>> +	cmdq_task_release(task);
> >>>>> +}
> >>>>> +
> >>>>> +static void cmdq_task_wait_release_schedule(struct cmdq_task *task)
> >>>>> +{
> >>>>> +	struct cmdq *cmdq = task->cmdq;
> >>>>> +
> >>>>> +	INIT_WORK(&task->release_work, cmdq_task_wait_release_work);
> >>>>> +	queue_work(cmdq->task_release_wq, &task->release_work);
> >>>>> +}
> >>>>> +
> >>>>> +static int cmdq_rec_realloc_cmd_buffer(struct cmdq_rec *rec, size_t size)
> >>>>> +{
> >>>>> +	void *new_buf;
> >>>>> +
> >>>>> +	new_buf = krealloc(rec->buf, size, GFP_KERNEL | __GFP_ZERO);
> >>>>> +	if (!new_buf)
> >>>>> +		return -ENOMEM;
> >>>>> +	rec->buf = new_buf;
> >>>>> +	rec->buf_size = size;
> >>>>> +	return 0;
> >>>>> +}
> >>>>> +
> >>>>> +struct cmdq_base *cmdq_register_device(struct device *dev)
> >>>>> +{
> >>>>> +	struct cmdq_base *cmdq_base;
> >>>>> +	struct resource res;
> >>>>> +	int subsys;
> >>>>> +	u32 base;
> >>>>> +
> >>>>> +	if (of_address_to_resource(dev->of_node, 0, &res))
> >>>>> +		return NULL;
> >>>>> +	base = (u32)res.start;
> >>>>> +
> >>>>> +	subsys = cmdq_subsys_base_to_id(base >> 16);
> >>>>> +	if (subsys < 0)
> >>>>> +		return NULL;
> >>>>> +
> >>>>> +	cmdq_base = devm_kmalloc(dev, sizeof(*cmdq_base), GFP_KERNEL);
> >>>>> +	if (!cmdq_base)
> >>>>> +		return NULL;
> >>>>> +	cmdq_base->subsys = subsys;
> >>>>> +	cmdq_base->base = base;
> >>>>> +
> >>>>> +	return cmdq_base;
> >>>>> +}
> >>>>> +EXPORT_SYMBOL(cmdq_register_device);
> >>>>> +
> >>>>> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
> >>>>> +		    struct cmdq_rec **rec_ptr)
> >>>>> +{
> >>>>> +	struct cmdq_rec *rec;
> >>>>> +	int err;
> >>>>> +
> >>>>> +	rec = kzalloc(sizeof(*rec), GFP_KERNEL);
> >>>>> +	if (!rec)
> >>>>> +		return -ENOMEM;
> >>>>> +	rec->cmdq = dev_get_drvdata(dev);
> >>>>> +	rec->engine_flag = engine_flag;
> >>>>> +	err = cmdq_rec_realloc_cmd_buffer(rec, CMDQ_INITIAL_CMD_BLOCK_SIZE);
> >>>>
> >>>> Just pass PAGE_SIZE here, no need for CMDQ_INITIAL_CMD_BLOCK_SIZE.
> >>>
> >>> Will do.
> >>>
> >>>>> +	if (err < 0) {
> >>>>> +		kfree(rec);
> >>>>> +		return err;
> >>>>> +	}
> >>>>> +	*rec_ptr = rec;
> >>>>> +	return 0;
> >>>>> +}
> >>>>> +EXPORT_SYMBOL(cmdq_rec_create);
> >>>>> +
> >>>>> +static int cmdq_rec_append_command(struct cmdq_rec *rec, enum cmdq_code code,
> >>>>> +				   u32 arg_a, u32 arg_b)
> >>>>> +{
> >>>>> +	u64 *cmd_ptr;
> >>>>> +	int err;
> >>>>> +
> >>>>> +	if (WARN_ON(rec->finalized))
> >>>>> +		return -EBUSY;
> >>>>> +	if (code < CMDQ_CODE_MOVE || code > CMDQ_CODE_EOC)
> >>>>> +		return -EINVAL;
> >>>>
> >>>> cmdq_rec_append_command is just called from inside this driver and code
> >>>> is a enum. We can expect it to be correct, no need for this check.
> >>>
> >>> Will drop this check.
> >>>
> >>>>> +	if (unlikely(rec->command_size + CMDQ_INST_SIZE > rec->buf_size)) {
> >>>>
> >>>> command_size is the offset into the buffer to which a new command is
> >>>> written, so this name is highly confusing. I wonder if this would be
> >>>> easier to understand if we redefine command_size to something like the
> >>>> number of commands and divide/multiply CMDQ_INST_SIZE where this is needed.
> >>>
> >>> I can rename command_size to cmd_buf_size and calculate num_cmd by
> >>> dividing CMDQ_INST_SIZE.
> >>> What do you think?
> >>>
> >>>>> +		err = cmdq_rec_realloc_cmd_buffer(rec, rec->buf_size * 2);
> >>>>> +		if (err < 0)
> >>>>> +			return err;
> >>>>> +	}
> >>>>> +	cmd_ptr = rec->buf + rec->command_size;
> >>>>> +	(*cmd_ptr) = (u64)((code << CMDQ_OP_CODE_SHIFT) | arg_a) << 32 | arg_b;
> >>>>> +	rec->command_size += CMDQ_INST_SIZE;
> >>>>> +	return 0;
> >>>>> +}
> >>>>> +
> >>>>> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value, struct cmdq_base *base,
> >>>>> +		   u32 offset)
> >>>>> +{
> >>>>> +	u32 arg_a = ((base->base + offset) & CMDQ_ARG_A_WRITE_MASK) |
> >>>>> +		    ((base->subsys & CMDQ_SUBSYS_MASK) << CMDQ_SUBSYS_SHIFT);
> >>>>
> >>>> base->subsys is the id from gce_sybsys, so we can expect it to be
> >>>> correct, no need to mask with CMDQ_SUBSYS_MASK.
> >>>
> >>> Will drop it.
> >>>
> >>>>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WRITE, arg_a, value);
> >>>>> +}
> >>>>> +EXPORT_SYMBOL(cmdq_rec_write);
> >>>>> +
> >>>>> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
> >>>>> +			struct cmdq_base *base, u32 offset, u32 mask)
> >>>>> +{
> >>>>> +	u32 offset_mask = offset;
> >>>>> +	int err;
> >>>>> +
> >>>>> +	if (mask != 0xffffffff) {
> >>>>> +		err = cmdq_rec_append_command(rec, CMDQ_CODE_MOVE, 0, ~mask);
> >>>>> +		if (err < 0)
> >>>>> +			return err;
> >>>>> +		offset_mask |= CMDQ_WRITE_ENABLE_MASK;
> >>>>> +	}
> >>>>> +	return cmdq_rec_write(rec, value, base, offset_mask);
> >>>>> +}
> >>>>> +EXPORT_SYMBOL(cmdq_rec_write_mask);
> >>>>> +
> >>>>> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event)
> >>>>> +{
> >>>>> +	u32 arg_b;
> >>>>> +
> >>>>> +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
> >>>>> +		return -EINVAL;
> >>>>> +
> >>>>> +	/*
> >>>>> +	 * bit 0-11: wait value
> >>>>> +	 * bit 15: 1 - wait, 0 - no wait
> >>>>> +	 * bit 16-27: update value
> >>>>> +	 * bit 31: 1 - update, 0 - no update
> >>>>> +	 */
> >>>>
> >>>> I don't understand this comments. What are they for?
> >>>
> >>> This is for WFE command. I will comment it.
> >>>
> >>>>> +	arg_b = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
> >>>>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event, arg_b);
> >>>>> +}
> >>>>> +EXPORT_SYMBOL(cmdq_rec_wfe);
> >>>>> +
> >>>>> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event)
> >>>>> +{
> >>>>> +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
> >>>>> +		return -EINVAL;
> >>>>> +
> >>>>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event,
> >>>>> +				       CMDQ_WFE_UPDATE);
> >>>>> +}
> >>>>> +EXPORT_SYMBOL(cmdq_rec_clear_event);
> >>>>> +
> >>>>> +static int cmdq_rec_finalize(struct cmdq_rec *rec)
> >>>>> +{
> >>>>> +	int err;
> >>>>> +
> >>>>> +	if (rec->finalized)
> >>>>> +		return 0;
> >>>>> +
> >>>>> +	/* insert EOC and generate IRQ for each command iteration */
> >>>>> +	err = cmdq_rec_append_command(rec, CMDQ_CODE_EOC, 0, CMDQ_EOC_IRQ_EN);
> >>>>> +	if (err < 0)
> >>>>> +		return err;
> >>>>> +
> >>>>> +	/* JUMP to end */
> >>>>> +	err = cmdq_rec_append_command(rec, CMDQ_CODE_JUMP, 0, CMDQ_JUMP_PASS);
> >>>>> +	if (err < 0)
> >>>>> +		return err;
> >>>>> +
> >>>>
> >>>> Does this need to be atomic?
> >>>> What happens if after CODE_EOC and before CODE_JUMP some
> >>>> write/read/event gets added?
> >>>> What happens if more commands get added to the queue after CODE_JUMP,
> >>>> but before finalized is set to true. Why don't you use atomic functions
> >>>> to access finalized?
> >>>
> >>> Since cmdq_rec doesn't guarantee thread safe, mutex is needed when
> >>> client uses cmdq_rec.
> >>>
> >>
> >> Well I think that rec->finalized tries to implement this, but might
> >> fail, if two kernel threads work on the same cmdq_rec.
> >>
> >>>>> +	rec->finalized = true;
> >>>>> +	return 0;
> >>>>> +}
> >>>>> +
> >>>>> +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
> >>>>> +			 void *data)
> >>>>> +{
> >>>>> +	struct cmdq *cmdq = rec->cmdq;
> >>>>> +	struct cmdq_task *task;
> >>>>> +	struct cmdq_task_cb task_cb;
> >>>>> +	struct cmdq_thread *thread;
> >>>>> +	int err;
> >>>>> +
> >>>>> +	mutex_lock(&cmdq->task_mutex);
> >>>>> +	if (cmdq->suspended) {
> >>>>> +		dev_err(cmdq->dev, "%s is called after suspended\n", __func__);
> >>>>> +		mutex_unlock(&cmdq->task_mutex);
> >>>>> +		return -EPERM;
> >>>>> +	}
> >>>>> +
> >>>>> +	err = cmdq_rec_finalize(rec);
> >>>>> +	if (err < 0) {
> >>>>> +		mutex_unlock(&cmdq->task_mutex);
> >>>>> +		return err;
> >>>>> +	}
> >>>>> +
> >>>>> +	task_cb.cb = cb;
> >>>>> +	task_cb.data = data;
> >>>>> +	task = cmdq_task_acquire(rec, task_cb);
> >>>>> +	if (!task) {
> >>>>> +		mutex_unlock(&cmdq->task_mutex);
> >>>>> +		return -EFAULT;
> >>>>> +	}
> >>>>> +
> >>>>> +	thread = &cmdq->thread[cmdq_eng_get_thread(task->engine_flag)];
> >>>>> +	cmdq_task_exec(task, thread);
> >>>>> +	cmdq_task_wait_release_schedule(task);
> >>>>> +	mutex_unlock(&cmdq->task_mutex);
> >>>>> +	return 0;
> >>>>> +}
> >>>>> +EXPORT_SYMBOL(cmdq_rec_flush_async);
> >>>>> +
> >>>>> +struct cmdq_flush_completion {
> >>>>> +	struct completion cmplt;
> >>>>> +	bool err;
> >>>>> +};
> >>>>> +
> >>>>> +static int cmdq_rec_flush_cb(struct cmdq_cb_data data)
> >>>>> +{
> >>>>> +	struct cmdq_flush_completion *cmplt = data.data;
> >>>>> +
> >>>>> +	cmplt->err = data.err;
> >>>>> +	complete(&cmplt->cmplt);
> >>>>> +	return 0;
> >>>>> +}
> >>>>> +
> >>>>> +int cmdq_rec_flush(struct cmdq_rec *rec)
> >>>>> +{
> >>>>> +	struct cmdq_flush_completion cmplt;
> >>>>> +	int err;
> >>>>> +
> >>>>> +	init_completion(&cmplt.cmplt);
> >>>>> +	err = cmdq_rec_flush_async(rec, cmdq_rec_flush_cb, &cmplt);
> >>>>> +	if (err < 0)
> >>>>> +		return err;
> >>>>> +	wait_for_completion(&cmplt.cmplt);
> >>>>> +	return cmplt.err ? -EFAULT : 0;
> >>>>> +}
> >>>>> +EXPORT_SYMBOL(cmdq_rec_flush);
> >>>>> +
> >>>>> +void cmdq_rec_destroy(struct cmdq_rec *rec)
> >>>>> +{
> >>>>> +	kfree(rec->buf);
> >>>>> +	kfree(rec);
> >>>>> +}
> >>>>> +EXPORT_SYMBOL(cmdq_rec_destroy);
> >>>>> +
> >>>>> +static bool cmdq_task_is_empty(struct cmdq *cmdq)
> >>>>> +{
> >>>>> +	struct cmdq_thread *thread;
> >>>>> +	int i;
> >>>>> +
> >>>>> +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
> >>>>> +		thread = &cmdq->thread[i];
> >>>>> +		if (!list_empty(&thread->task_busy_list))
> >>>>> +			return false;
> >>>>> +	}
> >>>>> +	return true;
> >>>>> +}
> >>>>> +
> >>>>> +static int cmdq_suspend(struct device *dev)
> >>>>> +{
> >>>>> +	struct cmdq *cmdq = dev_get_drvdata(dev);
> >>>>> +	u32 exec_threads;
> >>>>> +
> >>>>> +	mutex_lock(&cmdq->task_mutex);
> >>>>> +	cmdq->suspended = true;
> >>>>> +	mutex_unlock(&cmdq->task_mutex);
> >>>>> +
> >>>>> +	exec_threads = readl(cmdq->base + CMDQ_CURR_LOADED_THR);
> >>>>> +	if ((exec_threads & CMDQ_THR_EXECUTING) && !cmdq_task_is_empty(cmdq)) {
> >>>>> +		dev_err(dev, "wait active tasks timeout.\n");
> >>>>> +		flush_workqueue(cmdq->task_release_wq);
> >>>>> +	}
> >>>>> +	return 0;
> >>>>> +}
> >>>>> +
> >>>>> +static int cmdq_resume(struct device *dev)
> >>>>> +{
> >>>>> +	struct cmdq *cmdq = dev_get_drvdata(dev);
> >>>>> +
> >>>>> +	cmdq->suspended = false;
> >>>>> +	return 0;
> >>>>> +}
> >>>>> +
> >>>>> +static int cmdq_remove(struct platform_device *pdev)
> >>>>> +{
> >>>>> +	struct cmdq *cmdq = platform_get_drvdata(pdev);
> >>>>> +
> >>>>> +	destroy_workqueue(cmdq->task_release_wq);
> >>>>> +	cmdq->task_release_wq = NULL;
> >>>>> +	return 0;
> >>>>> +}
> >>>>> +
> >>>>> +static int cmdq_probe(struct platform_device *pdev)
> >>>>> +{
> >>>>> +	struct device *dev = &pdev->dev;
> >>>>> +	struct device_node *node = dev->of_node;
> >>>>> +	struct resource *res;
> >>>>> +	struct cmdq *cmdq;
> >>>>> +	int err, i;
> >>>>> +
> >>>>> +	cmdq = devm_kzalloc(dev, sizeof(*cmdq), GFP_KERNEL);
> >>>>> +	if (!cmdq)
> >>>>> +		return -ENOMEM;
> >>>>> +	cmdq->dev = dev;
> >>>>> +
> >>>>> +	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> >>>>> +	cmdq->base = devm_ioremap_resource(dev, res);
> >>>>> +	if (IS_ERR(cmdq->base)) {
> >>>>> +		dev_err(dev, "failed to ioremap gce\n");
> >>>>> +		return PTR_ERR(cmdq->base);
> >>>>> +	}
> >>>>> +
> >>>>> +	cmdq->irq = irq_of_parse_and_map(node, 0);
> >>>>> +	if (!cmdq->irq) {
> >>>>> +		dev_err(dev, "failed to get irq\n");
> >>>>> +		return -EINVAL;
> >>>>> +	}
> >>>>> +
> >>>>> +	dev_dbg(dev, "cmdq device: addr:0x%p, va:0x%p, irq:%d\n",
> >>>>> +		dev, cmdq->base, cmdq->irq);
> >>>>> +
> >>>>> +	mutex_init(&cmdq->task_mutex);
> >>>>> +	spin_lock_init(&cmdq->exec_lock);
> >>>>> +	cmdq->task_release_wq = alloc_ordered_workqueue(
> >>>>> +			"%s", WQ_MEM_RECLAIM | WQ_HIGHPRI,
> >>>>> +			"cmdq_task_wait_release");
> >>>>> +
> >>>>> +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
> >>>>> +		cmdq->thread[i].base = cmdq->base + CMDQ_THR_BASE +
> >>>>> +				CMDQ_THR_SHIFT * i;
> >>>>> +		init_waitqueue_head(&cmdq->thread[i].wait_task_done);
> >>>>> +		INIT_LIST_HEAD(&cmdq->thread[i].task_busy_list);
> >>>>> +	}
> >>>>> +
> >>>>> +	platform_set_drvdata(pdev, cmdq);
> >>>>> +
> >>>>> +	err = devm_request_irq(dev, cmdq->irq, cmdq_irq_handler, IRQF_SHARED,
> >>>>> +			       CMDQ_DRIVER_DEVICE_NAME, cmdq);
> >>>>> +	if (err < 0) {
> >>>>> +		dev_err(dev, "failed to register ISR (%d)\n", err);
> >>>>> +		goto fail;
> >>>>> +	}
> >>>>> +
> >>>>> +	cmdq->clock = devm_clk_get(dev, CMDQ_CLK_NAME);
> >>>>> +	if (IS_ERR(cmdq->clock)) {
> >>>>> +		dev_err(dev, "failed to get clk:%s\n", CMDQ_CLK_NAME);
> >>>>> +		err = PTR_ERR(cmdq->clock);
> >>>>> +		goto fail;
> >>>>> +	}
> >>>>> +	return 0;
> >>>>> +
> >>>>> +fail:
> >>>>> +	cmdq_remove(pdev);
> >>>>> +	return err;
> >>>>> +}
> >>>>> +
> >>>>> +static const struct dev_pm_ops cmdq_pm_ops = {
> >>>>> +	.suspend = cmdq_suspend,
> >>>>> +	.resume = cmdq_resume,
> >>>>> +};
> >>>>> +
> >>>>> +static const struct of_device_id cmdq_of_ids[] = {
> >>>>> +	{.compatible = "mediatek,mt8173-gce",},
> >>>>> +	{}
> >>>>> +};
> >>>>> +
> >>>>> +static struct platform_driver cmdq_drv = {
> >>>>> +	.probe = cmdq_probe,
> >>>>> +	.remove = cmdq_remove,
> >>>>> +	.driver = {
> >>>>> +		.name = CMDQ_DRIVER_DEVICE_NAME,
> >>>>> +		.owner = THIS_MODULE,
> >>>>> +		.pm = &cmdq_pm_ops,
> >>>>> +		.of_match_table = cmdq_of_ids,
> >>>>> +	}
> >>>>> +};
> >>>>> +
> >>>>> +builtin_platform_driver(cmdq_drv);
> >>>>> diff --git a/include/soc/mediatek/cmdq.h b/include/soc/mediatek/cmdq.h
> >>>>> new file mode 100644
> >>>>> index 0000000..60eef3d
> >>>>> --- /dev/null
> >>>>> +++ b/include/soc/mediatek/cmdq.h
> >>>>> @@ -0,0 +1,197 @@
> >>>>> +/*
> >>>>> + * Copyright (c) 2015 MediaTek Inc.
> >>>>> + *
> >>>>> + * This program is free software; you can redistribute it and/or modify
> >>>>> + * it under the terms of the GNU General Public License version 2 as
> >>>>> + * published by the Free Software Foundation.
> >>>>> + *
> >>>>> + * This program is distributed in the hope that it will be useful,
> >>>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> >>>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> >>>>> + * GNU General Public License for more details.
> >>>>> + */
> >>>>> +
> >>>>> +#ifndef __MTK_CMDQ_H__
> >>>>> +#define __MTK_CMDQ_H__
> >>>>> +
> >>>>> +#include <linux/platform_device.h>
> >>>>> +#include <linux/types.h>
> >>>>> +
> >>>>> +enum cmdq_eng {
> >>>>> +	CMDQ_ENG_DISP_AAL,
> >>>>> +	CMDQ_ENG_DISP_COLOR0,
> >>>>> +	CMDQ_ENG_DISP_COLOR1,
> >>>>> +	CMDQ_ENG_DISP_DPI0,
> >>>>> +	CMDQ_ENG_DISP_DSI0,
> >>>>> +	CMDQ_ENG_DISP_DSI1,
> >>>>> +	CMDQ_ENG_DISP_GAMMA,
> >>>>> +	CMDQ_ENG_DISP_OD,
> >>>>> +	CMDQ_ENG_DISP_OVL0,
> >>>>> +	CMDQ_ENG_DISP_OVL1,
> >>>>> +	CMDQ_ENG_DISP_PWM0,
> >>>>> +	CMDQ_ENG_DISP_PWM1,
> >>>>> +	CMDQ_ENG_DISP_RDMA0,
> >>>>> +	CMDQ_ENG_DISP_RDMA1,
> >>>>> +	CMDQ_ENG_DISP_RDMA2,
> >>>>> +	CMDQ_ENG_DISP_UFOE,
> >>>>> +	CMDQ_ENG_DISP_WDMA0,
> >>>>> +	CMDQ_ENG_DISP_WDMA1,
> >>>>> +	CMDQ_ENG_MAX,
> >>>>> +};
> >>>>> +
> >>>>> +/* events for CMDQ and display */
> >>>>> +enum cmdq_event {
> >>>>> +	/* Display start of frame(SOF) events */
> >>>>> +	CMDQ_EVENT_DISP_OVL0_SOF = 11,
> >>>>> +	CMDQ_EVENT_DISP_OVL1_SOF = 12,
> >>>>> +	CMDQ_EVENT_DISP_RDMA0_SOF = 13,
> >>>>> +	CMDQ_EVENT_DISP_RDMA1_SOF = 14,
> >>>>> +	CMDQ_EVENT_DISP_RDMA2_SOF = 15,
> >>>>> +	CMDQ_EVENT_DISP_WDMA0_SOF = 16,
> >>>>> +	CMDQ_EVENT_DISP_WDMA1_SOF = 17,
> >>>>> +	/* Display end of frame(EOF) events */
> >>>>> +	CMDQ_EVENT_DISP_OVL0_EOF = 39,
> >>>>> +	CMDQ_EVENT_DISP_OVL1_EOF = 40,
> >>>>> +	CMDQ_EVENT_DISP_RDMA0_EOF = 41,
> >>>>> +	CMDQ_EVENT_DISP_RDMA1_EOF = 42,
> >>>>> +	CMDQ_EVENT_DISP_RDMA2_EOF = 43,
> >>>>> +	CMDQ_EVENT_DISP_WDMA0_EOF = 44,
> >>>>> +	CMDQ_EVENT_DISP_WDMA1_EOF = 45,
> >>>>> +	/* Mutex end of frame(EOF) events */
> >>>>> +	CMDQ_EVENT_MUTEX0_STREAM_EOF = 53,
> >>>>> +	CMDQ_EVENT_MUTEX1_STREAM_EOF = 54,
> >>>>> +	CMDQ_EVENT_MUTEX2_STREAM_EOF = 55,
> >>>>> +	CMDQ_EVENT_MUTEX3_STREAM_EOF = 56,
> >>>>> +	CMDQ_EVENT_MUTEX4_STREAM_EOF = 57,
> >>>>> +	/* Display underrun events */
> >>>>> +	CMDQ_EVENT_DISP_RDMA0_UNDERRUN = 63,
> >>>>> +	CMDQ_EVENT_DISP_RDMA1_UNDERRUN = 64,
> >>>>> +	CMDQ_EVENT_DISP_RDMA2_UNDERRUN = 65,
> >>>>> +	/* Keep this at the end of HW events */
> >>>>> +	CMDQ_MAX_HW_EVENT_COUNT = 260,
> >>>>> +};
> >>>>> +
> >>>>> +struct cmdq_cb_data {
> >>>>> +	bool	err;
> >>>>> +	void	*data;
> >>>>> +};
> >>>>> +
> >>>>> +typedef int (*cmdq_async_flush_cb)(struct cmdq_cb_data data);
> >>>>> +
> >>>>> +struct cmdq_task;
> >>>>> +struct cmdq;
> >>>>> +
> >>>>> +struct cmdq_rec {
> >>>>> +	struct cmdq		*cmdq;
> >>>>> +	u64			engine_flag;
> >>>>> +	size_t			command_size;
> >>>>> +	void			*buf;
> >>>>> +	size_t			buf_size;
> >>>>> +	bool			finalized;
> >>>>> +};
> >>
> >> Why do we need cmdq_rec at all? Can't we just use the cmdq_task for that
> >> and this way make the driver less complex?
> >
> > There are two main reasons for cmdq_rec.
> > 1. It is slow to access dma too frequently.
> >     So, we append commands to cacheable memory, and then flush to dma.
> > 2. cmdq_rec is not thread safe, but cmdq_task needs thread safe.
> >     If we merge them, we need to take care of some synchronization
> >     issues.
> >
> >>>>> +
> >>>>> +struct cmdq_base {
> >>>>> +	int	subsys;
> >>>>> +	u32	base;
> >>>>
> >>>> subsys can always be calculated via cmdq_subsys_base_to_id(base >> 16)
> >>>> so we can get rid of the struct, right?
> >>>
> >>> Current subsys method is based on previous comment from Daniel Kurtz.
> >>> Please take a look of our previous discussion.
> >>> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/004483.html
> >>> Thanks.
> >>>
> >>
> >> I have to look deeper into this, but from what I read, the proposal from
> >> Daniel [1] seems good to me.
> >>
> >> [1] https://patchwork.kernel.org/patch/8068311/
> >
> > The initial proposal has some problem, so please see the follow-up
> > discussions. Thanks.
> > http://lists.infradead.org/pipermail/linux-mediatek/2016-February/003972.html
> > http://lists.infradead.org/pipermail/linux-mediatek/2016-February/004483.html
> >
> >
> >>>>> +};
> >>>>> +
> >>>>> +/**
> >>>>> + * cmdq_register_device() - register device which needs CMDQ
> >>>>> + * @dev:		device
> >>>>> + *
> >>>>> + * Return: cmdq_base pointer or NULL for failed
> >>>>> + */
> >>>>> +struct cmdq_base *cmdq_register_device(struct device *dev);
> >>>>> +
> >>>>> +/**
> >>>>> + * cmdq_rec_create() - create command queue record
> >>>>> + * @dev:		device
> >>>>> + * @engine_flag:	command queue engine flag
> >>>>> + * @rec_ptr:		command queue record pointer to retrieve cmdq_rec
> >>>>> + *
> >>>>> + * Return: 0 for success; else the error code is returned
> >>>>> + */
> >>>>> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
> >>>>> +		    struct cmdq_rec **rec_ptr);
> >>>>> +
> >>>>> +/**
> >>>>> + * cmdq_rec_write() - append write command to the command queue record
> >>>>> + * @rec:	the command queue record
> >>>>> + * @value:	the specified target register value
> >>>>> + * @base:	the command queue base
> >>>>> + * @offset:	register offset from module base
> >>>>> + *
> >>>>> + * Return: 0 for success; else the error code is returned
> >>>>> + */
> >>>>> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value,
> >>>>> +		   struct cmdq_base *base, u32 offset);
> >>>>> +
> >>>>> +/**
> >>>>> + * cmdq_rec_write_mask() - append write command with mask to the command
> >>>>> + *			   queue record
> >>>>> + * @rec:	the command queue record
> >>>>> + * @value:	the specified target register value
> >>>>> + * @base:	the command queue base
> >>>>> + * @offset:	register offset from module base
> >>>>> + * @mask:	the specified target register mask
> >>>>> + *
> >>>>> + * Return: 0 for success; else the error code is returned
> >>>>> + */
> >>>>> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
> >>>>> +			struct cmdq_base *base, u32 offset, u32 mask);
> >>>>> +
> >>>>> +/**
> >>>>> + * cmdq_rec_wfe() - append wait for event command to the command queue reco	rd
> >>>>
> >>>> reco rd -> record
> >>>
> >>> Will fix it.
> >>>
> >>>> Regards,
> >>>> Matthias
> >>>>
> >>>>> + * @rec:	the command queue record
> >>>>> + * @event:	the desired event type to "wait and CLEAR"
> >>>>> + *
> >>>>> + * Return: 0 for success; else the error code is returned
> >>>>> + */
> >>>>> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event);
> >>>>> +
> >>>>> +/**
> >>>>> + * cmdq_rec_clear_event() - append clear event command to the command queue
> >>>>> + *			    record
> >>>>> + * @rec:	the command queue record
> >>>>> + * @event:	the desired event to be cleared
> >>>>> + *
> >>>>> + * Return: 0 for success; else the error code is returned
> >>>>> + */
> >>>>> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event);
> >>>>> +
> >>>>> +/**
> >>>>> + * cmdq_rec_flush() - trigger CMDQ to execute the recorded commands
> >>>>> + * @rec:	the command queue record
> >>>>> + *
> >>>>> + * Return: 0 for success; else the error code is returned
> >>>>> + *
> >>>>> + * Trigger CMDQ to execute the recorded commands. Note that this is a
> >>>>> + * synchronous flush function. When the function returned, the recorded
> >>>>> + * commands have been done.
> >>>>> + */
> >>>>> +int cmdq_rec_flush(struct cmdq_rec *rec);
> >>>>> +
> >>>>> +/**
> >>>>> + * cmdq_rec_flush_async() - trigger CMDQ to asynchronously execute the recorded
> >>>>> + *			    commands and call back after ISR is finished
> >>>>> + * @rec:	the command queue record
> >>>>> + * @cb:		called in the end of CMDQ ISR
> >>>>> + * @data:	this data will pass back to cb
> >>>>> + *
> >>>>> + * Return: 0 for success; else the error code is returned
> >>>>> + *
> >>>>> + * Trigger CMDQ to asynchronously execute the recorded commands and call back
> >>>>> + * after ISR is finished. Note that this is an ASYNC function. When the function
> >>>>> + * returned, it may or may not be finished. The ISR callback function is called
> >>>>> + * in the end of ISR.
> >>
> >> "The callback is called from the ISR."
> >>
> >> Regards,
> >> Matthias
> >>
> >>>>> + */
> >>>>> +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
> >>>>> +			 void *data);
> >>>>> +
> >>>>> +/**
> >>>>> + * cmdq_rec_destroy() - destroy command queue record
> >>>>> + * @rec:	the command queue record
> >>>>> + */
> >>>>> +void cmdq_rec_destroy(struct cmdq_rec *rec);
> >>>>> +
> >>>>> +#endif	/* __MTK_CMDQ_H__ */
> >>>>>
> >>>
> >>> Thanks,
> >>> HS
> >>>
> >
> > Thanks,
> > HS
> >

Matthias Brugger June 3, 2016, 11:18 a.m. UTC | #9

On 03/06/16 08:12, Horng-Shyang Liao wrote:
> Hi Mathias,
>
> Please see my inline reply.
>
> On Thu, 2016-06-02 at 10:46 +0200, Matthias Brugger wrote:
>>
>> On 01/06/16 11:57, Horng-Shyang Liao wrote:
>>> Hi Mathias,
>>>
>>> Please see my inline reply.
>>>
>>> On Tue, 2016-05-31 at 22:04 +0200, Matthias Brugger wrote:
>>>>
>>>> On 31/05/16 10:36, Horng-Shyang Liao wrote:
>>>>> Hi Mathias,
>>>>>
>>>>> Please see my inline reply.
>>>>>
>>>>> On Mon, 2016-05-30 at 17:31 +0200, Matthias Brugger wrote:
>>>>>>
>>>>>> On 30/05/16 05:19, HS Liao wrote:
>>>>>>> This patch is first version of Mediatek Command Queue(CMDQ) driver. The
>>>>>>> CMDQ is used to help read/write registers with critical time limitation,
>>>>>>> such as updating display configuration during the vblank. It controls
>>>>>>> Global Command Engine (GCE) hardware to achieve this requirement.
>>>>>>> Currently, CMDQ only supports display related hardwares, but we expect
>>>>>>> it can be extended to other hardwares for future requirements.
>>>>>>>
>>>>>>> Signed-off-by: HS Liao <[e-mail]>
>>>>>>> Signed-off-by: CK Hu <[e-mail]>
>>>>>>> ---
>>>>>>>      drivers/soc/mediatek/Kconfig    |  10 +
>>>>>>>      drivers/soc/mediatek/Makefile   |   1 +
>>>>>>>      drivers/soc/mediatek/mtk-cmdq.c | 943 ++++++++++++++++++++++++++++++++++++++++
>>>>>>>      include/soc/mediatek/cmdq.h     | 197 +++++++++
>>>>>>>      4 files changed, 1151 insertions(+)
>>>>>>>      create mode 100644 drivers/soc/mediatek/mtk-cmdq.c
>>>>>>>      create mode 100644 include/soc/mediatek/cmdq.h
>>>>>>>
>>>>>>> diff --git a/drivers/soc/mediatek/Kconfig b/drivers/soc/mediatek/Kconfig
>>>>>>> index 0a4ea80..c4ad75c 100644
>>>>>>> --- a/drivers/soc/mediatek/Kconfig
>>>>>>> +++ b/drivers/soc/mediatek/Kconfig
>>>>>>> @@ -1,6 +1,16 @@
>>>>>>>      #
>>>>>>>      # MediaTek SoC drivers
>>>>>>>      #
>>>>>>> +config MTK_CMDQ
>>>>>>> +	bool "MediaTek CMDQ Support"
>>>>>>> +	depends on ARCH_MEDIATEK || COMPILE_TEST
>>>>
>>>> depends on ARM64 ?
>>>
>>> Will add ARM64.
>>>
>>>>>>> +	select MTK_INFRACFG
>>>>>>> +	help
>>>>>>> +	  Say yes here to add support for the MediaTek Command Queue (CMDQ)
>>>>>>> +	  driver. The CMDQ is used to help read/write registers with critical
>>>>>>> +	  time limitation, such as updating display configuration during the
>>>>>>> +	  vblank.
>>>>>>> +
>>>>>>>      config MTK_INFRACFG
>>>>>>>      	bool "MediaTek INFRACFG Support"
>>>>>>>      	depends on ARCH_MEDIATEK || COMPILE_TEST
>>>>>>> diff --git a/drivers/soc/mediatek/Makefile b/drivers/soc/mediatek/Makefile
>>>>>>> index 12998b0..f7397ef 100644
>>>>>>> --- a/drivers/soc/mediatek/Makefile
>>>>>>> +++ b/drivers/soc/mediatek/Makefile
>>>>>>> @@ -1,3 +1,4 @@
>>>>>>> +obj-$(CONFIG_MTK_CMDQ) += mtk-cmdq.o
>>>>>>>      obj-$(CONFIG_MTK_INFRACFG) += mtk-infracfg.o
>>>>>>>      obj-$(CONFIG_MTK_PMIC_WRAP) += mtk-pmic-wrap.o
>>>>>>>      obj-$(CONFIG_MTK_SCPSYS) += mtk-scpsys.o
>>>>>>> diff --git a/drivers/soc/mediatek/mtk-cmdq.c b/drivers/soc/mediatek/mtk-cmdq.c
>>>>>>> new file mode 100644
>>>>>>> index 0000000..e9d6e1c
>>>>>>> --- /dev/null
>>>>>>> +++ b/drivers/soc/mediatek/mtk-cmdq.c
>>>>>>> @@ -0,0 +1,943 @@
>>>>>>> +/*
>>>>>>> + * Copyright (c) 2015 MediaTek Inc.
>>>>>>> + *
>>>>>>> + * This program is free software; you can redistribute it and/or modify
>>>>>>> + * it under the terms of the GNU General Public License version 2 as
>>>>>>> + * published by the Free Software Foundation.
>>>>>>> + *
>>>>>>> + * This program is distributed in the hope that it will be useful,
>>>>>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>>>>>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
>>>>>>> + * GNU General Public License for more details.
>>>>>>> + */
>>>>>>> +
>>>>>>> +#include <linux/clk.h>
>>>>>>> +#include <linux/clk-provider.h>
>>>>>>> +#include <linux/completion.h>
>>>>>>> +#include <linux/dma-mapping.h>
>>>>>>> +#include <linux/errno.h>
>>>>>>> +#include <linux/interrupt.h>
>>>>>>> +#include <linux/iopoll.h>
>>>>>>> +#include <linux/kernel.h>
>>>>>>> +#include <linux/kthread.h>
>>>>>>> +#include <linux/module.h>
>>>>>>> +#include <linux/mutex.h>
>>>>>>> +#include <linux/of_address.h>
>>>>>>> +#include <linux/of_irq.h>
>>>>>>> +#include <linux/platform_device.h>
>>>>>>> +#include <linux/slab.h>
>>>>>>> +#include <linux/spinlock.h>
>>>>>>> +#include <linux/suspend.h>
>>>>>>> +#include <linux/workqueue.h>
>>>>>>> +#include <soc/mediatek/cmdq.h>
>>>>>>> +
>>>>>>> +#define CMDQ_INITIAL_CMD_BLOCK_SIZE	PAGE_SIZE
>>>>>>> +#define CMDQ_INST_SIZE			8 /* instruction is 64-bit */
>>>>>>> +#define CMDQ_TIMEOUT_MS			1000
>>>>>>> +#define CMDQ_IRQ_MASK			0xffff
>>>>>>> +#define CMDQ_DRIVER_DEVICE_NAME		"mtk_cmdq"
>>>>>>> +#define CMDQ_CLK_NAME			"gce"
>>>>>>
>>>>>> We can put the names in directly to un-bloat the defines.
>>>>>
>>>>> I will use the names directly and remove defines.
>>>>>
>>>>>>> +
>>>>>>> +#define CMDQ_CURR_IRQ_STATUS		0x010
>>>>>>> +#define CMDQ_CURR_LOADED_THR		0x018
>>>>>>> +#define CMDQ_THR_SLOT_CYCLES		0x030
>>>>>>> +
>>>>>>> +#define CMDQ_THR_BASE			0x100
>>>>>>> +#define CMDQ_THR_SHIFT			0x080
>>>>>>
>>>>>> Wouldn't be CMDQ_THR_SIZE more accurate?
>>>>>
>>>>> Will rename it.
>>>>>
>>>>>>> +#define CMDQ_THR_WARM_RESET		0x00
>>>>>>> +#define CMDQ_THR_ENABLE_TASK		0x04
>>>>>>> +#define CMDQ_THR_SUSPEND_TASK		0x08
>>>>>>> +#define CMDQ_THR_CURR_STATUS		0x0c
>>>>>>> +#define CMDQ_THR_IRQ_STATUS		0x10
>>>>>>> +#define CMDQ_THR_IRQ_ENABLE		0x14
>>>>>>> +#define CMDQ_THR_CURR_ADDR		0x20
>>>>>>> +#define CMDQ_THR_END_ADDR		0x24
>>>>>>> +#define CMDQ_THR_CFG			0x40
>>>>>>> +
>>>>>>> +#define CMDQ_THR_ENABLED		0x1
>>>>>>> +#define CMDQ_THR_DISABLED		0x0
>>>>>>> +#define CMDQ_THR_SUSPEND		0x1
>>>>>>> +#define CMDQ_THR_RESUME			0x0
>>>>>>> +#define CMDQ_THR_STATUS_SUSPENDED	BIT(1)
>>>>>>> +#define CMDQ_THR_DO_WARM_RESET		BIT(0)
>>>>>>> +#define CMDQ_THR_ACTIVE_SLOT_CYCLES	0x3200
>>>>>>> +#define CMDQ_THR_PRIORITY		3
>>>>>>> +#define CMDQ_THR_IRQ_DONE		0x1
>>>>>>> +#define CMDQ_THR_IRQ_ERROR		0x12
>>>>>>> +#define CMDQ_THR_IRQ_EN			0x13 /* done + error */
>>>>>>
>>>>>> #define CMDQ_THR_IRQ_EN (CMDQ_THR_IRQ_ERROR | CMDQ_THR_IRQ_DONE)
>>>>>
>>>>> Will do.
>>>>>
>>>>>>> +#define CMDQ_THR_IRQ_MASK		0x13
>>>>>>
>>>>>> never used.
>>>>>
>>>>> Will remove.
>>>>>
>>>>>>> +#define CMDQ_THR_EXECUTING		BIT(31)
>>>>>>> +
>>>>>>> +#define CMDQ_ARG_A_WRITE_MASK		0xffff
>>>>>>> +#define CMDQ_SUBSYS_MASK		0x1f
>>>>>>> +#define CMDQ_OP_CODE_MASK		0xff000000
>>>>>>> +
>>>>>>> +#define CMDQ_OP_CODE_SHIFT		24
>>>>>>
>>>>>> Couldn't we connect the mask with the shift, or aren't they related?
>>>>>>
>>>>>> #define CMDQ_OP_CODE_MASK (0xff << CMDQ_OP_CODE_SHIFT)
>>>>>
>>>>> Will do.
>>>>>
>>>>>>> +#define CMDQ_SUBSYS_SHIFT		16
>>>>>>> +
>>>>>>> +#define CMDQ_WRITE_ENABLE_MASK		BIT(0)
>>>>>>> +#define CMDQ_JUMP_BY_OFFSET		0x10000000
>>>>>>> +#define CMDQ_JUMP_BY_PA			0x10000001
>>>>>>> +#define CMDQ_JUMP_PASS			CMDQ_INST_SIZE
>>>>>>> +#define CMDQ_WFE_UPDATE			BIT(31)
>>>>>>> +#define CMDQ_WFE_WAIT			BIT(15)
>>>>>>> +#define CMDQ_WFE_WAIT_VALUE		0x1
>>>>>>> +#define CMDQ_EOC_IRQ_EN			BIT(0)
>>>>>>> +
>>>>>>> +enum cmdq_thread_index {
>>>>>>> +	CMDQ_THR_DISP_MAIN_IDX,	/* main */
>>>>>>> +	CMDQ_THR_DISP_SUB_IDX,	/* sub */
>>>>>>> +	CMDQ_THR_DISP_MISC_IDX,	/* misc */
>>>>>>> +	CMDQ_THR_MAX_COUNT,	/* max */
>>>>>>> +};
>>>>>>> +
>>>>>>> +/*
>>>>>>> + * CMDQ_CODE_MOVE:
>>>>>>> + *   move value into internal register as mask
>>>>>>> + *   format: op mask
>>>>>>> + * CMDQ_CODE_WRITE:
>>>>>>> + *   write value into target register
>>>>>>> + *   format: op subsys address value
>>>>>>> + * CMDQ_CODE_JUMP:
>>>>>>> + *   jump by offset
>>>>>>> + *   format: op offset
>>>>>>> + * CMDQ_CODE_WFE:
>>>>>>> + *   wait for event and clear
>>>>>>> + *   it is just clear if no wait
>>>>>>> + *   format: [wait]  op event update:1 to_wait:1 wait:1
>>>>>>> + *           [clear] op event update:1 to_wait:0 wait:0
>>>>>>> + * CMDQ_CODE_EOC:
>>>>>>> + *   end of command
>>>>>>> + *   format: op irq_flag
>>>>>>> + */
>>>>>>
>>>>>> I think we need more documentation of how this command queue engine is
>>>>>> working. If not, I think it will be really complicated to understand how
>>>>>> to use this.
>>>>>>
>>>>>>> +enum cmdq_code {
>>>>>>> +	CMDQ_CODE_MOVE = 0x02,
>>>>>>> +	CMDQ_CODE_WRITE = 0x04,
>>>>>>> +	CMDQ_CODE_JUMP = 0x10,
>>>>>>> +	CMDQ_CODE_WFE = 0x20,
>>>>>>> +	CMDQ_CODE_EOC = 0x40,
>>>>>>> +};
>>>>>>> +
>>>>>>> +enum cmdq_task_state {
>>>>>>> +	TASK_STATE_BUSY, /* running on a GCE thread */
>>>>>>> +	TASK_STATE_ERROR,
>>>>>>> +	TASK_STATE_DONE,
>>>>>>> +};
>>>>>>> +
>>>>>>> +struct cmdq_task_cb {
>>>>>>> +	cmdq_async_flush_cb	cb;
>>>>>>> +	void			*data;
>>>>>>> +};
>>>>>>> +
>>>>>>> +struct cmdq_thread {
>>>>>>> +	void __iomem		*base;
>>>>>>> +	struct list_head	task_busy_list;
>>>>>>> +	wait_queue_head_t	wait_task_done;
>>>>>>> +};
>>>>>>> +
>>>>>>> +struct cmdq_task {
>>>>>>> +	struct cmdq		*cmdq;
>>>>>>> +	struct list_head	list_entry;
>>>>>>> +	enum cmdq_task_state	task_state;
>>>>>>> +	void			*va_base;
>>>>>>> +	dma_addr_t		pa_base;
>>>>>>> +	u64			engine_flag;
>>>>>>> +	size_t			command_size;
>>>>>>> +	u32			num_cmd;
>>>>>>
>>>>>> num_cmd is directly connected to command_size. I prefer to just keep
>>>>>> num_cmd and calculate the command_size where necessary.
>>>>>
>>>>> After I trace code, I prefer to keep command_size and calculate num_cmd
>>>>> where necessary. What do you think?
>>>>>
>>>>
>>>> I suppose you prefer this, as you are writing to the GCE depending on
>>>> the command_size. I think it is worth to create a macro for the
>>>> calculation of the number of commands, to make the code more readable.
>>>> Would be nice if you would just pass cmdq_task to it and it would return
>>>> the number. Just as an idea.
>>>
>>> Will add macro.
>>>
>>>>>>> +	struct cmdq_thread	*thread;
>>>>>>> +	struct cmdq_task_cb	cb;
>>>>>>> +	struct work_struct	release_work;
>>>>>>> +};
>>>>>>> +
>>>>>>> +struct cmdq {
>>>>>>> +	struct device		*dev;
>>>>>>> +	void __iomem		*base;
>>>>>>> +	u32			irq;
>>>>>>> +	struct workqueue_struct	*task_release_wq;
>>>>>>> +	struct cmdq_thread	thread[CMDQ_THR_MAX_COUNT];
>>>>>>> +	struct mutex		task_mutex;	/* for task */
>>>>>>> +	spinlock_t		exec_lock;	/* for exec */
>>>>>>> +	struct clk		*clock;
>>>>>>> +	bool			suspended;
>>>>>>> +};
>>>>>>> +
>>>>>>> +struct cmdq_subsys {
>>>>>>> +	u32	base;
>>>>>>> +	int	id;
>>>>>>> +};
>>>>>>> +
>>>>>>> +static const struct cmdq_subsys gce_subsys[] = {
>>>>>>> +	{0x1400, 1},
>>>>>>> +	{0x1401, 2},
>>>>>>> +	{0x1402, 3},
>>>>>>> +};
>>>>>>> +
>>>>>>> +static int cmdq_subsys_base_to_id(u32 base)
>>>>>>> +{
>>>>>>> +	int i;
>>>>>>> +
>>>>>>> +	for (i = 0; i < ARRAY_SIZE(gce_subsys); i++)
>>>>>>> +		if (gce_subsys[i].base == base)
>>>>>>> +			return gce_subsys[i].id;
>>>>>>> +	return -EFAULT;
>>>>>>> +}
>>>>>>> +
>>>>>>> +static int cmdq_eng_get_thread(u64 flag)
>>>>>>> +{
>>>>>>> +	if (flag & BIT_ULL(CMDQ_ENG_DISP_DSI0))
>>>>>>> +		return CMDQ_THR_DISP_MAIN_IDX;
>>>>>>> +	else if (flag & BIT_ULL(CMDQ_ENG_DISP_DPI0))
>>>>>>> +		return CMDQ_THR_DISP_SUB_IDX;
>>>>>>> +	else
>>>>>>> +		return CMDQ_THR_DISP_MISC_IDX;
>>>>>>> +}
>>>>>>> +
>>>>>>> +static void cmdq_task_release(struct cmdq_task *task)
>>>>>>> +{
>>>>>>> +	struct cmdq *cmdq = task->cmdq;
>>>>>>> +
>>>>>>> +	dma_free_coherent(cmdq->dev, task->command_size, task->va_base,
>>>>>>> +			  task->pa_base);
>>>>>>> +	kfree(task);
>>>>>>> +}
>>>>>>> +
>>>>>>> +static struct cmdq_task *cmdq_task_acquire(struct cmdq_rec *rec,
>>>>>>> +					   struct cmdq_task_cb cb)
>>>>>>> +{
>>>>>>> +	struct cmdq *cmdq = rec->cmdq;
>>>>>>> +	struct device *dev = cmdq->dev;
>>>>>>> +	struct cmdq_task *task;
>>>>>>> +
>>>>>>> +	task = kzalloc(sizeof(*task), GFP_KERNEL);
>>>>>>> +	INIT_LIST_HEAD(&task->list_entry);
>>>>>>> +	task->va_base = dma_alloc_coherent(dev, rec->command_size,
>>>>>>> +					   &task->pa_base, GFP_KERNEL);
>>>>>>> +	if (!task->va_base) {
>>>>>>> +		dev_err(dev, "allocate command buffer failed\n");
>>>>>>> +		kfree(task);
>>>>>>> +		return NULL;
>>>>>>> +	}
>>>>>>> +
>>>>>>> +	task->cmdq = cmdq;
>>>>>>> +	task->command_size = rec->command_size;
>>>>>>> +	task->engine_flag = rec->engine_flag;
>>>>>>> +	task->task_state = TASK_STATE_BUSY;
>>>>>>> +	task->cb = cb;
>>>>>>> +	memcpy(task->va_base, rec->buf, rec->command_size);
>>>>>>> +	task->num_cmd = task->command_size / CMDQ_INST_SIZE;
>>>>>>> +	return task;
>>>>>>> +}
>>>>>>> +
>>>>>>> +static void cmdq_thread_writel(struct cmdq_thread *thread, u32 value,
>>>>>>> +			       u32 offset)
>>>>>>> +{
>>>>>>> +	writel(value, thread->base + offset);
>>>>>>> +}
>>>>>>> +
>>>>>>> +static u32 cmdq_thread_readl(struct cmdq_thread *thread, u32 offset)
>>>>>>> +{
>>>>>>> +	return readl(thread->base + offset);
>>>>>>> +}
>>>>>>
>>>>>> We can get rid of cmdq_thread_readl/writel.
>>>>>
>>>>> Will do.
>>>>>
>>>>>>> +
>>>>>>> +static int cmdq_thread_suspend(struct cmdq *cmdq, struct cmdq_thread *thread)
>>>>>>> +{
>>>>>>> +	u32 status;
>>>>>>> +
>>>>>>> +	cmdq_thread_writel(thread, CMDQ_THR_SUSPEND, CMDQ_THR_SUSPEND_TASK);
>>>>>>> +
>>>>>>> +	/* If already disabled, treat as suspended successful. */
>>>>>>> +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
>>>>>>> +	      CMDQ_THR_ENABLED))
>>>>>>> +		return 0;
>>>>>>> +
>>>>>>> +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_CURR_STATUS,
>>>>>>> +			status, status & CMDQ_THR_STATUS_SUSPENDED, 0, 10)) {
>>>>>>> +		dev_err(cmdq->dev, "suspend GCE thread 0x%x failed\n",
>>>>>>> +			(u32)(thread->base - cmdq->base));
>>>>>>> +		return -EFAULT;
>>>>>>> +	}
>>>>>>> +
>>>>>>> +	return 0;
>>>>>>> +}
>>>>>>> +
>>>>>>> +static void cmdq_thread_resume(struct cmdq_thread *thread)
>>>>>>> +{
>>>>>>> +	cmdq_thread_writel(thread, CMDQ_THR_RESUME, CMDQ_THR_SUSPEND_TASK);
>>>>>>> +}
>>>>>>> +
>>>>>>> +static int cmdq_thread_reset(struct cmdq *cmdq, struct cmdq_thread *thread)
>>>>>>> +{
>>>>>>> +	u32 warm_reset;
>>>>>>> +
>>>>>>> +	cmdq_thread_writel(thread, CMDQ_THR_DO_WARM_RESET, CMDQ_THR_WARM_RESET);
>>>>>>> +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_WARM_RESET,
>>>>>>> +			warm_reset, !(warm_reset & CMDQ_THR_DO_WARM_RESET),
>>>>>>> +			0, 10)) {
>>>>>>> +		dev_err(cmdq->dev, "reset GCE thread 0x%x failed\n",
>>>>>>> +			(u32)(thread->base - cmdq->base));
>>>>>>> +		return -EFAULT;
>>>>>>> +	}
>>>>>>> +	writel(CMDQ_THR_ACTIVE_SLOT_CYCLES, cmdq->base + CMDQ_THR_SLOT_CYCLES);
>>>>>>> +	return 0;
>>>>>>> +}
>>>>>>> +
>>>>>>> +static void cmdq_thread_disable(struct cmdq *cmdq, struct cmdq_thread *thread)
>>>>>>> +{
>>>>>>> +	cmdq_thread_reset(cmdq, thread);
>>>>>>> +	cmdq_thread_writel(thread, CMDQ_THR_DISABLED, CMDQ_THR_ENABLE_TASK);
>>>>>>> +}
>>>>>>> +
>>>>>>> +/* notify GCE to re-fetch commands by setting GCE thread PC */
>>>>>>> +static void cmdq_thread_invalidate_fetched_data(struct cmdq_thread *thread)
>>>>>>> +{
>>>>>>> +	cmdq_thread_writel(thread,
>>>>>>> +			   cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR),
>>>>>>> +			   CMDQ_THR_CURR_ADDR);
>>>>>>> +}
>>>>>>> +
>>>>>>> +static void cmdq_task_insert_into_thread(struct cmdq_task *task)
>>>>>>> +{
>>>>>>> +	struct cmdq_thread *thread = task->thread;
>>>>>>> +	struct cmdq_task *prev_task = list_last_entry(
>>>>>>> +			&thread->task_busy_list, typeof(*task), list_entry);
>>>>>>> +	u64 *prev_task_base = prev_task->va_base;
>>>>>>> +
>>>>>>> +	/* let previous task jump to this task */
>>>>>>> +	prev_task_base[prev_task->num_cmd - 1] = (u64)CMDQ_JUMP_BY_PA << 32 |
>>>>>>> +						 task->pa_base;
>>>>>>> +
>>>>>>> +	cmdq_thread_invalidate_fetched_data(thread);
>>>>>>> +}
>>>>>>> +
>>>>>>> +/* we assume tasks in the same display GCE thread are waiting the same event. */
>>>>>>> +static void cmdq_task_remove_wfe(struct cmdq_task *task)
>>>>>>> +{
>>>>>>> +	u32 wfe_option = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
>>>>>>> +	u32 wfe_op = CMDQ_CODE_WFE << CMDQ_OP_CODE_SHIFT;
>>>>>>> +	u32 *base = task->va_base;
>>>>>>> +	u32 num_cmd = task->num_cmd << 1;
>>>>>>> +	int i;
>>>>>>> +
>>>>>>> +	for (i = 0; i < num_cmd; i += 2)
>>>>>>> +		if (base[i] == wfe_option &&
>>>>>>> +		    (base[i + 1] & CMDQ_OP_CODE_MASK) == wfe_op) {
>>>>>>> +			base[i] = CMDQ_JUMP_PASS;
>>>>>>> +			base[i + 1] = CMDQ_JUMP_BY_OFFSET;
>>>>>>> +		}
>>>>>>
>>>>>> After using the command buffer as a void pointer a u64 pointer, we now
>>>>>> reference to it as u32. I would prefer to explain here, how the command
>>>>>> looks like we are searching for and use a for loop passing task->num_cmd
>>>>>> instead.
>>>>>
>>>>> Will use u64* to rewrite the above code.
>>>>>
>>>>>>> +}
>>>>>>> +
>>>>>>> +static void cmdq_task_exec(struct cmdq_task *task, struct cmdq_thread *thread)
>>>>>>> +{
>>>>>>> +	struct cmdq *cmdq = task->cmdq;
>>>>>>> +	unsigned long flags;
>>>>>>> +	unsigned long curr_pa, end_pa;
>>>>>>> +
>>>>>>> +	WARN_ON(clk_prepare_enable(cmdq->clock) < 0);
>>>>>>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
>>>>>>
>>>>>> cmdq_task_exec is called with cmdq->task_mutex held, so why do we need
>>>>>> the spin_lock here? Can't we just use one of the two?
>>>>>
>>>>> We can drop task_mutex, but we will get some side effects.
>>>>> 1. exec_lock needs to include more code, but I think it is not good for
>>>>>       spinlock.
>>>>> 2. In cmdq_rec_flush_async(), task_mutex needs to protect
>>>>>       (1) cmdq->suspended, (2) cmdq_task_exec(), and
>>>>>       (3) cmdq_task_wait_release_schedule().
>>>>>       If we drop task_mutex, we have to put cmdq->suspended if condition
>>>>>       just before cmdq_task_exec() and inside exec_lock, and we have to
>>>>>       release task and its command buffer if error. This will let flow
>>>>>       become more complex and enlarge code size.
>>>>>
>>>>> What do you think?
>>>>
>>>> Why do you need to protect cmdq_task_wait_release_schedule? We don't
>>>> care about the order of the workqueue elements, do we?
>>>
>>> According to CK's comment, we have to ensure order to remove previous
>>> task.
>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-May/005547.html
>>>
>>>> As far as I understand you would need to protect cmdq_task_acquire as
>>>> well, to "ensure" continously growing pa_base. More on that below.
>>>
>>> We need to ensure continuous physical addresses in a task, but we don't
>>> need to ensure continuous physical addresses between tasks.
>>> So, I think it is unnecessary to protect by mutex or spinlock.
>>>
>>
>> True, I didn't get that.
>>
>>>>>
>>>>>>> +	task->thread = thread;
>>>>>>> +	task->task_state = TASK_STATE_BUSY;
>>>>>>
>>>>>> That was already set in cmdq_task_acquire, why do we need to set it here
>>>>>> again?
>>>>>
>>>>> Will drop it.
>>>>>
>>>>
>>>> Yeah, but I think it makes more sense to drop it in cmdq_task_acquire
>>>> instead.
>>>
>>> Will drop it in cmdq_task_acquire instead.
>>>
>>>>>>> +	if (list_empty(&thread->task_busy_list)) {
>>>>>>> +		WARN_ON(cmdq_thread_reset(cmdq, thread) < 0);
>>>>>>> +
>>>>>>> +		cmdq_thread_writel(thread, task->pa_base, CMDQ_THR_CURR_ADDR);
>>>>>>> +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
>>>>>>> +				   CMDQ_THR_END_ADDR);
>>>>>>> +		cmdq_thread_writel(thread, CMDQ_THR_PRIORITY, CMDQ_THR_CFG);
>>>>>>> +		cmdq_thread_writel(thread, CMDQ_THR_IRQ_EN,
>>>>>>> +				   CMDQ_THR_IRQ_ENABLE);
>>>>>>> +
>>>>>>> +		cmdq_thread_writel(thread, CMDQ_THR_ENABLED,
>>>>>>> +				   CMDQ_THR_ENABLE_TASK);
>>>>>>> +	} else {
>>>>>>> +		WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
>>>>>>> +
>>>>>>> +		/*
>>>>>>> +		 * check boundary condition
>>>>>>> +		 * PC = END - 8, EOC is executed
>>>>>>> +		 * PC = END, all CMDs are executed
>>>>>>> +		 */
>>>>>>> +		curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
>>>>>>> +		end_pa = cmdq_thread_readl(thread, CMDQ_THR_END_ADDR);
>>>>>>> +		if (curr_pa == end_pa - 8 || curr_pa == end_pa) {
>>>>>>
>>>>>> 8 refers to CMDQ_INST_SIZE, right?
>>>>>
>>>>> Yes, I will use CMDQ_INST_SIZE.
>>>>>
>>>>>>> +			/* set to this task directly */
>>>>>>> +			cmdq_thread_writel(thread, task->pa_base,
>>>>>>> +					   CMDQ_THR_CURR_ADDR);
>>>>>>> +		} else {
>>>>>>> +			cmdq_task_insert_into_thread(task);
>>>>>>> +
>>>>>>> +			if (thread == &cmdq->thread[CMDQ_THR_DISP_MAIN_IDX] ||
>>>>>>> +			    thread == &cmdq->thread[CMDQ_THR_DISP_SUB_IDX])
>>>>>>> +				cmdq_task_remove_wfe(task);
>>>>>>
>>>>>> We could do this check using the task->engine_flag, I suppose that's
>>>>>> easier to undestand then.
>>>>>
>>>>> Will use task->engine_flag.
>>>>>
>>>>>>> +
>>>>>>> +			smp_mb(); /* modify jump before enable thread */
>>>>>>> +		}
>>>>>>> +
>>>>>>> +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
>>>>>>> +				   CMDQ_THR_END_ADDR);
>>>>>>> +		cmdq_thread_resume(thread);
>>>>>>> +	}
>>>>>>> +	list_move_tail(&task->list_entry, &thread->task_busy_list);
>>>>>>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>>>> +}
>>>>>>> +
>>>>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
>>>>>>> +				   struct cmdq_thread *thread, u32 irq_flag)
>>>>>>> +{
>>>>>>> +	struct cmdq_task *task, *tmp, *curr_task = NULL;
>>>>>>> +	u32 curr_pa;
>>>>>>> +	struct cmdq_cb_data cmdq_cb_data;
>>>>>>> +	bool err;
>>>>>>> +
>>>>>>> +	if (irq_flag & CMDQ_THR_IRQ_ERROR)
>>>>>>> +		err = true;
>>>>>>> +	else if (irq_flag & CMDQ_THR_IRQ_DONE)
>>>>>>> +		err = false;
>>>>>>> +	else
>>>>>>> +		return;
>>>>>>> +
>>>>>>> +	curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
>>>>>>> +
>>>>>>> +	list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
>>>>>>> +				 list_entry) {
>>>>>>> +		if (curr_pa >= task->pa_base &&
>>>>>>> +		    curr_pa < (task->pa_base + task->command_size))
>>>>>>
>>>>>> What are you checking here? It seems as if you make some implcit
>>>>>> assumptions about pa_base and the order of execution of commands in the
>>>>>> thread. Is it save to do so? Does dma_alloc_coherent give any guarantees
>>>>>> about dma_handle?
>>>>>
>>>>> 1. Check what is the current running task in this GCE thread.
>>>>> 2. Yes.
>>>>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
>>>>>
>>>>
>>>> Yes, physical addresses might be continous, but AFAIK there is no
>>>> guarantee that the dma_handle address is steadily growing, when calling
>>>> dma_alloc_coherent. And if I understand the code correctly, you use this
>>>> assumption to decide if the task picked from task_busy_list is currently
>>>> executing. So I think this mecanism is not working.
>>>
>>> I don't use dma_handle address, and just use physical addresses.
>>>   From CPU's point of view, tasks are linked by the busy list.
>>>   From GCE's point of view, tasks are linked by the JUMP command.
>>>
>>>> In which cases does the HW thread raise an interrupt.
>>>> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
>>>
>>> GCE will raise interrupt if any task is done or error.
>>> However, GCE is fast, so CPU may get multiple done tasks
>>> when it is running ISR.
>>>
>>> In case of error, that GCE thread will pause and raise interrupt.
>>> So, CPU may get multiple done tasks and one error task.
>>>
>>
>> I think we should reimplement the ISR mechanism. Can't we just read
>> CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave
>> cmdq_handle_error_done to the thread_fn? You will need to pass
>> information from the handler to thread_fn, but that shouldn't be an
>> issue. AFAIK interrupts are disabled in the handler, so we should stay
>> there as short as possible. Traversing task_busy_list is expensive, so
>> we need to do it in a thread context.
>
> Actually, our initial implementation is similar to your suggestion,
> but display needs CMDQ to return callback function very precisely,
> else display will drop frame.
> For display, CMDQ interrupt will be raised every 16 ~ 17 ms,
> and CMDQ needs to call callback function in ISR.
> If we defer callback to workqueue, the time interval may be larger than
> 32 ms.sometimes.
>

I think the problem is, that you implemented the workqueue as a ordered 
workqueue, so there is no parallel processing. I'm still not sure why 
you need the workqueue to be ordered. Can you please explain.

>> I keep thinking about how to get rid of the two data structures,
>> task_busy_list and the task_release_wq. We need the latter for the only
>> sake of getting a timeout.
>>
>> Did you have a look on how the mailbox framework handles this?
>> By the way, what is the reason to not implement the whole driver as a
>> mailbox controller? For me, this driver looks like a good fit.
>
> CMDQ needs to encode commands for GCE hardware. We think this behavior
> should be put in CMDQ driver, and client just call CMDQ functions.
> Therefore, if we want to use mailbox framework, cmdq_rec must be
> mailbox client, and the others must be mailbox controller.
>

You mean the functions to fill the cmdq_rec and execute it?
I think this should be part of the driver.

Jassi, can you have a look on the interface this driver exports [0].
They are needed to actually create the message which will be send.
Could something like this be part of a mailbox driver?

[0] https://patchwork.kernel.org/patch/9140221/

> However, if we use mailbox controller, CMDQ driver still needs to
> control busy list for each GCE thread, and use workqueue to handle
> timeout tasks.
>

Let me summarize my ideas around this driver:
When we enter the ISR, we know that all task in task_busy_list before 
the entry which represents curr_task can be set to TASK_STATE_DONE.
The curr_task could be TASK_STATE_ERROR if the corresponding bit in the 
irq status registers is set.
Do we need to call the callback in the same order as the tasks got 
dispatched to the HW thread? If not, we could try to call all this 
callbacks in a multithreaded workqueue.

Regards,
Matthias

> The only thing that we can borrow from mailbox framework is the send
> (CMDQ flush) and receive (CMDQ callback) interface, However, we don't
> think we can gain many benefits from it, and we have some overheads to
> conform to mailbox interface.
>
>
>>
>>>>>>> +			curr_task = task;
>>>>>>> +		if (task->cb.cb) {
>>>>>>> +			cmdq_cb_data.err = curr_task ? err : false;
>>>>>>> +			cmdq_cb_data.data = task->cb.data;
>>>>>>> +			task->cb.cb(cmdq_cb_data);
>>>>>>> +		}
>>>>>>> +		task->task_state = (curr_task && err) ? TASK_STATE_ERROR :
>>>>>>> +				TASK_STATE_DONE;
>>>>>>> +		list_del(&task->list_entry);
>>>>>>> +		if (curr_task)
>>>>>>> +			break;
>>>>>>> +	}
>>>>>>> +
>>>>>>> +	wake_up(&thread->wait_task_done);
>>>>>>> +}
>>>>>>> +
>>>>>>> +static void cmdq_thread_irq_handler(struct cmdq *cmdq, int tid)
>>>>>>> +{
>>>>>>> +	struct cmdq_thread *thread = &cmdq->thread[tid];
>>>>>>> +	unsigned long flags = 0L;
>>>>>>> +	u32 irq_flag;
>>>>>>> +
>>>>>>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
>>>>>>> +
>>>>>>> +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
>>>>>>> +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
>>>>>>> +
>>>>>>> +	/*
>>>>>>> +	 * Another CPU core could run "release task" right before we acquire
>>>>>>> +	 * the spin lock, and thus reset / disable this GCE thread, so we
>>>>>>> +	 * need to check the enable bit of this GCE thread.
>>>>>>> +	 */
>>>>>>> +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
>>>>>>> +	      CMDQ_THR_ENABLED))
>>>>>>> +		irq_flag = 0;
>>>>>>
>>>>>> cmdq_handle_error_done just retuns in this case. Programming this way
>>>>>> just makes things confusing. What about:
>>>>>>
>>>>>> if (cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
>>>>>> 	      CMDQ_THR_ENABLED)
>>>>>> 	cmdq_handle_error_done(cmdq, thread, irq_flag);
>>>>>> else
>>>>>> 		irq_flag = 0;
>>>>>>
>>>>>> spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>>
>>>>> We still need to clear irq_flag if GCE thread is disabled.
>>>>> So, I think we can just return here.
>>>>>
>>>>> 	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
>>>>> 	      CMDQ_THR_ENABLED))
>>>>> 		return;
>>>>>
>>>>> What do you think?
>>>>>
>>>>
>>>> No, you can't just return, you need to unlock the spinlock.
>>>> Anyway I would prefer it the other way round, as I put it in my last
>>>> mail. Just delete the else branch, we don't need to set irq_flag to zero.
>>>
>>> Sorry for my previous wrong reply since I merge your comment
>>> and CK's comment.
>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-May/005547.html
>>> So, I will put this if condition into cmdq_task_handle_error_result()
>>> and then just return it if GCE thread is disabled.
>>>
>>
>> You mean in cmdq_task_handle_done
>> We should rename this functions to not create confusion.
>
> Sorry again. I mean in cmdq_handle_error_done().
> This function handles both done and error.
>
> I agree the function name looks confusion.
> I think it can be renamed to cmdq_thread_irq_handler()
> since it is used to handle irq for GCE thread.
>
>> Regards,
>> Matthias
>
> Thanks,
> HS
>
>>>>>>> +
>>>>>>> +	cmdq_handle_error_done(cmdq, thread, irq_flag);
>>>>>>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>>>> +}
>>>>>>> +
>>>>>>> +static irqreturn_t cmdq_irq_handler(int irq, void *dev)
>>>>>>> +{
>>>>>>> +	struct cmdq *cmdq = dev;
>>>>>>> +	u32 irq_status;
>>>>>>> +	int i;
>>>>>>> +
>>>>>>> +	irq_status = readl(cmdq->base + CMDQ_CURR_IRQ_STATUS);
>>>>>>> +	irq_status &= CMDQ_IRQ_MASK;
>>>>>>> +	irq_status ^= CMDQ_IRQ_MASK;
>>>>>>
>>>>>> irq_status can be much bigger then 3, which is the number of threads in
>>>>>> the system (CMDQ_THR_MAX_COUNT). So why we use this mask here isn't
>>>>>> clear to me.
>>>>>
>>>>> Our GCE hardware has 16 threads, but we only use 3 threads currently.
>>>>>
>>>>
>>>> Ok, but please use bitops here.
>>>
>>> Will use bitops.
>>>
>>>>>>> +
>>>>>>> +	if (!irq_status)
>>>>>>> +		return IRQ_NONE;
>>>>>>> +
>>>>>>> +	while (irq_status) {
>>>>>>> +		i = ffs(irq_status) - 1;
>>>>>>> +		irq_status &= ~BIT(i);
>>>>>>> +		cmdq_thread_irq_handler(cmdq, i);
>>>>>>> +	}
>>>>>>
>>>>>> Can you explain how the irq status register looks like, that would it
>>>>>> make much easier to understand what happens here.
>>>>>
>>>>> Bit 0 ~ 15 of irq status register represents GCE thread 0 ~ 15
>>>>> interrupt. 0 means asserting interrupt; 1 means no interrupt.
>>>>>
>>>>
>>>> Thanks, that helped. :)
>>>>
>>>>>>> +
>>>>>>> +	return IRQ_HANDLED;	
>>>>>>> +}
>>>>>>> +
>>>>>>> +static int cmdq_task_handle_error_result(struct cmdq_task *task)
>>>>>>
>>>>>> We never check the return values, why do we have them?
>>>>>
>>>>> Will drop return value.
>>>>>
>>>>>>> +{
>>>>>>> +	struct cmdq *cmdq = task->cmdq;
>>>>>>> +	struct device *dev = cmdq->dev;
>>>>>>> +	struct cmdq_thread *thread = task->thread;
>>>>>>> +	struct cmdq_task *next_task, *prev_task;
>>>>>>> +	u32 irq_flag;
>>>>>>> +
>>>>>>> +	/* suspend GCE thread to ensure consistency */
>>>>>>> +	WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
>>>>>>> +
>>>>>>> +	/* ISR has handled this error task */
>>>>>>> +	if (task->task_state == TASK_STATE_ERROR) {
>>>>>>> +		next_task = list_first_entry_or_null(&thread->task_busy_list,
>>>>>>> +				struct cmdq_task, list_entry);
>>>>>>> +		if (next_task) /* move to next task */
>>>>>>> +			cmdq_thread_writel(thread, next_task->pa_base,
>>>>>>> +					   CMDQ_THR_CURR_ADDR);
>>>>>>
>>>>>> We have to do this, as we suppose that the thread did not reach the jump
>>>>>> instruction we put into it's command queue, right?
>>>>>
>>>>> Yes.
>>>>>
>>>>
>>>> So this should then go into it's own function. In wait_release_work,
>>>> something like this:
>>>>
>>>> if(task->task_state == TASK_STATE_ERROR)
>>>>         cmdq_task_handle_error(task)
>>>
>>> OK.
>>> I will write new function cmdq_task_handle_error() to handle error case.
>>>
>>>>>>> +		cmdq_thread_resume(thread);
>>>>>>> +		return -ECANCELED;
>>>>>>> +	}
>>>>>>> +
>>>>>>
>>>>>> if task_state != ERROR and != DONE. This means that the timeout of
>>>>>> task_release_wq has timed out, right?
>>>>>
>>>>> Yes.
>>>>>
>>>>>>> +	/*
>>>>>>> +	 * Save next_task and prev_task in advance
>>>>>>> +	 * since cmdq_handle_error_done will remove list_entry.
>>>>>>> +	 */
>>>>>>> +	next_task = prev_task = NULL;
>>>>>>> +	if (task->list_entry.next != &thread->task_busy_list)
>>>>>>> +		next_task = list_next_entry(task, list_entry);
>>>>>>> +	if (task->list_entry.prev != &thread->task_busy_list)
>>>>>>> +		prev_task = list_prev_entry(task, list_entry);
>>>>>>> +
>>>>>>> +	/*
>>>>>>> +	 * Although IRQ is disabled, GCE continues to execute.
>>>>>>> +	 * It may have pending IRQ before GCE thread is suspended,
>>>>>>> +	 * so check this condition again.
>>>>>>> +	 */
>>>>>>
>>>>>> The first thing we did in this function was suspending the thread. Why
>>>>>> do we need this then?
>>>>>
>>>>> Because timeout is CPU timeout not GCE timeout, GCE could just finish
>>>>> this task before the GCE thread is suspended.
>>>>>
>>>>
>>>> What are the reasons for a timeout? An error has happend, or the task is
>>>> still executing.
>>>
>>>   From GCE's point of view, this task is still executing.
>>> But, it could be an error of client.
>>> For example, task may never get event if display turn off hardware
>>> before waiting for task to finish its work.
>>>
>>>>>> To be honest this whole functions looks really like a design error. We
>>>>>> have to sperate the states much clearer so that it is possible to
>>>>>> understand what is happening in the GCE. Isn't it for example posible to
>>>>>> have worker queues for timed out tasks and tasks with an error? I'm not
>>>>>> sure how to do this, actually I'm not sure if I really understood how
>>>>>> this is supposed to work.
>>>>>
>>>>> GCE doesn't have timeout. The timeout is decided and controlled by CPU,
>>>>> so we check timeout in release work.
>>>>> For error and done, they are easy to check by register, and we have
>>>>> already created release work for timeout. So, I don't think we need to
>>>>> create work queue for each case.
>>>>>
>>>>> What do you think?
>>>>>
>>>>
>>>> I think, if we find in here, that the irq_flag is set, then the the
>>>> interrupt handler was triggered and is spinning the spinlock. If this is
>>>> not the case, we have a timeout and we handle this.
>>>
>>> I will write another function to handle error, and handle timeout here.
>>>
>>>>>> How much do we win, when we patch the thread command queue for every
>>>>>> task we add, instead of just taking one task after another from the
>>>>>> task_busy_list?
>>>>>
>>>>> GCE is used to help read/write registers with critical time limitation.
>>>>> Sometimes, client may ask to process multiple tasks in a short period
>>>>> of time, e.g. display flush multiple tasks for next vblank. So, CMDQ
>>>>> shouldn't limit to process one task after another from the
>>>>> task_busy_list. Currently, when interrupt or timeout, we will check
>>>>> how many tasks are done, and which one is error or timeout.
>>>>>
>>>>
>>>> So I suppose the device driver who use this are interested in throughput
>>>> and not in latency. The callback of every
>>>>
>>>>>>> +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
>>>>>>> +	cmdq_handle_error_done(cmdq, thread, irq_flag);
>>>>>>> +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
>>>>>>> +
>>>>>>> +	if (task->task_state == TASK_STATE_DONE) {
>>>>>>> +		cmdq_thread_resume(thread);
>>>>>>> +		return 0;
>>>>>>> +	}
>>>>>>> +
>>>>>>> +	if (task->task_state == TASK_STATE_ERROR) {
>>>>>>> +		dev_err(dev, "task 0x%p error\n", task);
>>>>>>> +		if (next_task) /* move to next task */
>>>>>>> +			cmdq_thread_writel(thread, next_task->pa_base,
>>>>>>> +					   CMDQ_THR_CURR_ADDR);
>>>>>>> +		cmdq_thread_resume(thread);
>>>>>>> +		return -ECANCELED;
>>>>>>> +	}
>>>>>>> +
>>>>>>> +	/* Task is running, so we force to remove it. */
>>>>>>> +	dev_err(dev, "task 0x%p timeout or killed\n", task);
>>>>>>> +	task->task_state = TASK_STATE_ERROR;
>>>>>>> +
>>>>>>> +	if (prev_task) {
>>>>>>> +		u64 *prev_va = prev_task->va_base;
>>>>>>> +		u64 *curr_va = task->va_base;
>>>>>>> +
>>>>>>> +		/* copy JUMP instruction */
>>>>>>> +		prev_va[prev_task->num_cmd - 1] = curr_va[task->num_cmd - 1];
>>>>>>> +
>>>>>>> +		cmdq_thread_invalidate_fetched_data(thread);
>>>>>>> +	} else if (next_task) { /* move to next task */
>>>>>>> +		cmdq_thread_writel(thread, next_task->pa_base,
>>>>>>> +				   CMDQ_THR_CURR_ADDR);
>>>>>>> +	}
>>>>>>> +
>>>>>>> +	list_del(&task->list_entry);
>>>>>>> +	cmdq_thread_resume(thread);
>>>>>>> +
>>>>>>> +	/* call cb here to prevent lock */
>>>>>>> +	if (task->cb.cb) {
>>>>>>> +		struct cmdq_cb_data cmdq_cb_data;
>>>>>>> +
>>>>>>> +		cmdq_cb_data.err = true;
>>>>>>> +		cmdq_cb_data.data = task->cb.data;
>>>>>>> +		task->cb.cb(cmdq_cb_data);
>>>>>>> +	}
>>>>>>> +
>>>>>>> +	return -ECANCELED;
>>>>>>> +}
>>>>>>> +
>>>>>>> +static void cmdq_task_wait_release_work(struct work_struct *work_item)
>>>>>>> +{
>>>>>>> +	struct cmdq_task *task = container_of(work_item, struct cmdq_task,
>>>>>>> +					      release_work);
>>>>>>> +	struct cmdq *cmdq = task->cmdq;
>>>>>>> +	struct cmdq_thread *thread = task->thread;
>>>>>>> +	unsigned long flags;
>>>>>>> +
>>>>>>> +	wait_event_timeout(thread->wait_task_done,
>>>>>>> +			   task->task_state != TASK_STATE_BUSY,
>>>>>>> +			   msecs_to_jiffies(CMDQ_TIMEOUT_MS));
>>>>>>> +
>>>>>>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
>>>>>>> +	if (task->task_state != TASK_STATE_DONE)
>>>>>>> +		cmdq_task_handle_error_result(task);
>>>>>>> +	if (list_empty(&thread->task_busy_list))
>>>>>>> +		cmdq_thread_disable(cmdq, thread);
>>>>>>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>>>> +
>>>>>>> +	/* release regardless of success or not */
>>>>>>> +	clk_disable_unprepare(cmdq->clock);
>>>>>>> +	cmdq_task_release(task);
>>>>>>> +}
>>>>>>> +
>>>>>>> +static void cmdq_task_wait_release_schedule(struct cmdq_task *task)
>>>>>>> +{
>>>>>>> +	struct cmdq *cmdq = task->cmdq;
>>>>>>> +
>>>>>>> +	INIT_WORK(&task->release_work, cmdq_task_wait_release_work);
>>>>>>> +	queue_work(cmdq->task_release_wq, &task->release_work);
>>>>>>> +}
>>>>>>> +
>>>>>>> +static int cmdq_rec_realloc_cmd_buffer(struct cmdq_rec *rec, size_t size)
>>>>>>> +{
>>>>>>> +	void *new_buf;
>>>>>>> +
>>>>>>> +	new_buf = krealloc(rec->buf, size, GFP_KERNEL | __GFP_ZERO);
>>>>>>> +	if (!new_buf)
>>>>>>> +		return -ENOMEM;
>>>>>>> +	rec->buf = new_buf;
>>>>>>> +	rec->buf_size = size;
>>>>>>> +	return 0;
>>>>>>> +}
>>>>>>> +
>>>>>>> +struct cmdq_base *cmdq_register_device(struct device *dev)
>>>>>>> +{
>>>>>>> +	struct cmdq_base *cmdq_base;
>>>>>>> +	struct resource res;
>>>>>>> +	int subsys;
>>>>>>> +	u32 base;
>>>>>>> +
>>>>>>> +	if (of_address_to_resource(dev->of_node, 0, &res))
>>>>>>> +		return NULL;
>>>>>>> +	base = (u32)res.start;
>>>>>>> +
>>>>>>> +	subsys = cmdq_subsys_base_to_id(base >> 16);
>>>>>>> +	if (subsys < 0)
>>>>>>> +		return NULL;
>>>>>>> +
>>>>>>> +	cmdq_base = devm_kmalloc(dev, sizeof(*cmdq_base), GFP_KERNEL);
>>>>>>> +	if (!cmdq_base)
>>>>>>> +		return NULL;
>>>>>>> +	cmdq_base->subsys = subsys;
>>>>>>> +	cmdq_base->base = base;
>>>>>>> +
>>>>>>> +	return cmdq_base;
>>>>>>> +}
>>>>>>> +EXPORT_SYMBOL(cmdq_register_device);
>>>>>>> +
>>>>>>> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
>>>>>>> +		    struct cmdq_rec **rec_ptr)
>>>>>>> +{
>>>>>>> +	struct cmdq_rec *rec;
>>>>>>> +	int err;
>>>>>>> +
>>>>>>> +	rec = kzalloc(sizeof(*rec), GFP_KERNEL);
>>>>>>> +	if (!rec)
>>>>>>> +		return -ENOMEM;
>>>>>>> +	rec->cmdq = dev_get_drvdata(dev);
>>>>>>> +	rec->engine_flag = engine_flag;
>>>>>>> +	err = cmdq_rec_realloc_cmd_buffer(rec, CMDQ_INITIAL_CMD_BLOCK_SIZE);
>>>>>>
>>>>>> Just pass PAGE_SIZE here, no need for CMDQ_INITIAL_CMD_BLOCK_SIZE.
>>>>>
>>>>> Will do.
>>>>>
>>>>>>> +	if (err < 0) {
>>>>>>> +		kfree(rec);
>>>>>>> +		return err;
>>>>>>> +	}
>>>>>>> +	*rec_ptr = rec;
>>>>>>> +	return 0;
>>>>>>> +}
>>>>>>> +EXPORT_SYMBOL(cmdq_rec_create);
>>>>>>> +
>>>>>>> +static int cmdq_rec_append_command(struct cmdq_rec *rec, enum cmdq_code code,
>>>>>>> +				   u32 arg_a, u32 arg_b)
>>>>>>> +{
>>>>>>> +	u64 *cmd_ptr;
>>>>>>> +	int err;
>>>>>>> +
>>>>>>> +	if (WARN_ON(rec->finalized))
>>>>>>> +		return -EBUSY;
>>>>>>> +	if (code < CMDQ_CODE_MOVE || code > CMDQ_CODE_EOC)
>>>>>>> +		return -EINVAL;
>>>>>>
>>>>>> cmdq_rec_append_command is just called from inside this driver and code
>>>>>> is a enum. We can expect it to be correct, no need for this check.
>>>>>
>>>>> Will drop this check.
>>>>>
>>>>>>> +	if (unlikely(rec->command_size + CMDQ_INST_SIZE > rec->buf_size)) {
>>>>>>
>>>>>> command_size is the offset into the buffer to which a new command is
>>>>>> written, so this name is highly confusing. I wonder if this would be
>>>>>> easier to understand if we redefine command_size to something like the
>>>>>> number of commands and divide/multiply CMDQ_INST_SIZE where this is needed.
>>>>>
>>>>> I can rename command_size to cmd_buf_size and calculate num_cmd by
>>>>> dividing CMDQ_INST_SIZE.
>>>>> What do you think?
>>>>>
>>>>>>> +		err = cmdq_rec_realloc_cmd_buffer(rec, rec->buf_size * 2);
>>>>>>> +		if (err < 0)
>>>>>>> +			return err;
>>>>>>> +	}
>>>>>>> +	cmd_ptr = rec->buf + rec->command_size;
>>>>>>> +	(*cmd_ptr) = (u64)((code << CMDQ_OP_CODE_SHIFT) | arg_a) << 32 | arg_b;
>>>>>>> +	rec->command_size += CMDQ_INST_SIZE;
>>>>>>> +	return 0;
>>>>>>> +}
>>>>>>> +
>>>>>>> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value, struct cmdq_base *base,
>>>>>>> +		   u32 offset)
>>>>>>> +{
>>>>>>> +	u32 arg_a = ((base->base + offset) & CMDQ_ARG_A_WRITE_MASK) |
>>>>>>> +		    ((base->subsys & CMDQ_SUBSYS_MASK) << CMDQ_SUBSYS_SHIFT);
>>>>>>
>>>>>> base->subsys is the id from gce_sybsys, so we can expect it to be
>>>>>> correct, no need to mask with CMDQ_SUBSYS_MASK.
>>>>>
>>>>> Will drop it.
>>>>>
>>>>>>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WRITE, arg_a, value);
>>>>>>> +}
>>>>>>> +EXPORT_SYMBOL(cmdq_rec_write);
>>>>>>> +
>>>>>>> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
>>>>>>> +			struct cmdq_base *base, u32 offset, u32 mask)
>>>>>>> +{
>>>>>>> +	u32 offset_mask = offset;
>>>>>>> +	int err;
>>>>>>> +
>>>>>>> +	if (mask != 0xffffffff) {
>>>>>>> +		err = cmdq_rec_append_command(rec, CMDQ_CODE_MOVE, 0, ~mask);
>>>>>>> +		if (err < 0)
>>>>>>> +			return err;
>>>>>>> +		offset_mask |= CMDQ_WRITE_ENABLE_MASK;
>>>>>>> +	}
>>>>>>> +	return cmdq_rec_write(rec, value, base, offset_mask);
>>>>>>> +}
>>>>>>> +EXPORT_SYMBOL(cmdq_rec_write_mask);
>>>>>>> +
>>>>>>> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event)
>>>>>>> +{
>>>>>>> +	u32 arg_b;
>>>>>>> +
>>>>>>> +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
>>>>>>> +		return -EINVAL;
>>>>>>> +
>>>>>>> +	/*
>>>>>>> +	 * bit 0-11: wait value
>>>>>>> +	 * bit 15: 1 - wait, 0 - no wait
>>>>>>> +	 * bit 16-27: update value
>>>>>>> +	 * bit 31: 1 - update, 0 - no update
>>>>>>> +	 */
>>>>>>
>>>>>> I don't understand this comments. What are they for?
>>>>>
>>>>> This is for WFE command. I will comment it.
>>>>>
>>>>>>> +	arg_b = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
>>>>>>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event, arg_b);
>>>>>>> +}
>>>>>>> +EXPORT_SYMBOL(cmdq_rec_wfe);
>>>>>>> +
>>>>>>> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event)
>>>>>>> +{
>>>>>>> +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
>>>>>>> +		return -EINVAL;
>>>>>>> +
>>>>>>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event,
>>>>>>> +				       CMDQ_WFE_UPDATE);
>>>>>>> +}
>>>>>>> +EXPORT_SYMBOL(cmdq_rec_clear_event);
>>>>>>> +
>>>>>>> +static int cmdq_rec_finalize(struct cmdq_rec *rec)
>>>>>>> +{
>>>>>>> +	int err;
>>>>>>> +
>>>>>>> +	if (rec->finalized)
>>>>>>> +		return 0;
>>>>>>> +
>>>>>>> +	/* insert EOC and generate IRQ for each command iteration */
>>>>>>> +	err = cmdq_rec_append_command(rec, CMDQ_CODE_EOC, 0, CMDQ_EOC_IRQ_EN);
>>>>>>> +	if (err < 0)
>>>>>>> +		return err;
>>>>>>> +
>>>>>>> +	/* JUMP to end */
>>>>>>> +	err = cmdq_rec_append_command(rec, CMDQ_CODE_JUMP, 0, CMDQ_JUMP_PASS);
>>>>>>> +	if (err < 0)
>>>>>>> +		return err;
>>>>>>> +
>>>>>>
>>>>>> Does this need to be atomic?
>>>>>> What happens if after CODE_EOC and before CODE_JUMP some
>>>>>> write/read/event gets added?
>>>>>> What happens if more commands get added to the queue after CODE_JUMP,
>>>>>> but before finalized is set to true. Why don't you use atomic functions
>>>>>> to access finalized?
>>>>>
>>>>> Since cmdq_rec doesn't guarantee thread safe, mutex is needed when
>>>>> client uses cmdq_rec.
>>>>>
>>>>
>>>> Well I think that rec->finalized tries to implement this, but might
>>>> fail, if two kernel threads work on the same cmdq_rec.
>>>>
>>>>>>> +	rec->finalized = true;
>>>>>>> +	return 0;
>>>>>>> +}
>>>>>>> +
>>>>>>> +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
>>>>>>> +			 void *data)
>>>>>>> +{
>>>>>>> +	struct cmdq *cmdq = rec->cmdq;
>>>>>>> +	struct cmdq_task *task;
>>>>>>> +	struct cmdq_task_cb task_cb;
>>>>>>> +	struct cmdq_thread *thread;
>>>>>>> +	int err;
>>>>>>> +
>>>>>>> +	mutex_lock(&cmdq->task_mutex);
>>>>>>> +	if (cmdq->suspended) {
>>>>>>> +		dev_err(cmdq->dev, "%s is called after suspended\n", __func__);
>>>>>>> +		mutex_unlock(&cmdq->task_mutex);
>>>>>>> +		return -EPERM;
>>>>>>> +	}
>>>>>>> +
>>>>>>> +	err = cmdq_rec_finalize(rec);
>>>>>>> +	if (err < 0) {
>>>>>>> +		mutex_unlock(&cmdq->task_mutex);
>>>>>>> +		return err;
>>>>>>> +	}
>>>>>>> +
>>>>>>> +	task_cb.cb = cb;
>>>>>>> +	task_cb.data = data;
>>>>>>> +	task = cmdq_task_acquire(rec, task_cb);
>>>>>>> +	if (!task) {
>>>>>>> +		mutex_unlock(&cmdq->task_mutex);
>>>>>>> +		return -EFAULT;
>>>>>>> +	}
>>>>>>> +
>>>>>>> +	thread = &cmdq->thread[cmdq_eng_get_thread(task->engine_flag)];
>>>>>>> +	cmdq_task_exec(task, thread);
>>>>>>> +	cmdq_task_wait_release_schedule(task);
>>>>>>> +	mutex_unlock(&cmdq->task_mutex);
>>>>>>> +	return 0;
>>>>>>> +}
>>>>>>> +EXPORT_SYMBOL(cmdq_rec_flush_async);
>>>>>>> +
>>>>>>> +struct cmdq_flush_completion {
>>>>>>> +	struct completion cmplt;
>>>>>>> +	bool err;
>>>>>>> +};
>>>>>>> +
>>>>>>> +static int cmdq_rec_flush_cb(struct cmdq_cb_data data)
>>>>>>> +{
>>>>>>> +	struct cmdq_flush_completion *cmplt = data.data;
>>>>>>> +
>>>>>>> +	cmplt->err = data.err;
>>>>>>> +	complete(&cmplt->cmplt);
>>>>>>> +	return 0;
>>>>>>> +}
>>>>>>> +
>>>>>>> +int cmdq_rec_flush(struct cmdq_rec *rec)
>>>>>>> +{
>>>>>>> +	struct cmdq_flush_completion cmplt;
>>>>>>> +	int err;
>>>>>>> +
>>>>>>> +	init_completion(&cmplt.cmplt);
>>>>>>> +	err = cmdq_rec_flush_async(rec, cmdq_rec_flush_cb, &cmplt);
>>>>>>> +	if (err < 0)
>>>>>>> +		return err;
>>>>>>> +	wait_for_completion(&cmplt.cmplt);
>>>>>>> +	return cmplt.err ? -EFAULT : 0;
>>>>>>> +}
>>>>>>> +EXPORT_SYMBOL(cmdq_rec_flush);
>>>>>>> +
>>>>>>> +void cmdq_rec_destroy(struct cmdq_rec *rec)
>>>>>>> +{
>>>>>>> +	kfree(rec->buf);
>>>>>>> +	kfree(rec);
>>>>>>> +}
>>>>>>> +EXPORT_SYMBOL(cmdq_rec_destroy);
>>>>>>> +
>>>>>>> +static bool cmdq_task_is_empty(struct cmdq *cmdq)
>>>>>>> +{
>>>>>>> +	struct cmdq_thread *thread;
>>>>>>> +	int i;
>>>>>>> +
>>>>>>> +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
>>>>>>> +		thread = &cmdq->thread[i];
>>>>>>> +		if (!list_empty(&thread->task_busy_list))
>>>>>>> +			return false;
>>>>>>> +	}
>>>>>>> +	return true;
>>>>>>> +}
>>>>>>> +
>>>>>>> +static int cmdq_suspend(struct device *dev)
>>>>>>> +{
>>>>>>> +	struct cmdq *cmdq = dev_get_drvdata(dev);
>>>>>>> +	u32 exec_threads;
>>>>>>> +
>>>>>>> +	mutex_lock(&cmdq->task_mutex);
>>>>>>> +	cmdq->suspended = true;
>>>>>>> +	mutex_unlock(&cmdq->task_mutex);
>>>>>>> +
>>>>>>> +	exec_threads = readl(cmdq->base + CMDQ_CURR_LOADED_THR);
>>>>>>> +	if ((exec_threads & CMDQ_THR_EXECUTING) && !cmdq_task_is_empty(cmdq)) {
>>>>>>> +		dev_err(dev, "wait active tasks timeout.\n");
>>>>>>> +		flush_workqueue(cmdq->task_release_wq);
>>>>>>> +	}
>>>>>>> +	return 0;
>>>>>>> +}
>>>>>>> +
>>>>>>> +static int cmdq_resume(struct device *dev)
>>>>>>> +{
>>>>>>> +	struct cmdq *cmdq = dev_get_drvdata(dev);
>>>>>>> +
>>>>>>> +	cmdq->suspended = false;
>>>>>>> +	return 0;
>>>>>>> +}
>>>>>>> +
>>>>>>> +static int cmdq_remove(struct platform_device *pdev)
>>>>>>> +{
>>>>>>> +	struct cmdq *cmdq = platform_get_drvdata(pdev);
>>>>>>> +
>>>>>>> +	destroy_workqueue(cmdq->task_release_wq);
>>>>>>> +	cmdq->task_release_wq = NULL;
>>>>>>> +	return 0;
>>>>>>> +}
>>>>>>> +
>>>>>>> +static int cmdq_probe(struct platform_device *pdev)
>>>>>>> +{
>>>>>>> +	struct device *dev = &pdev->dev;
>>>>>>> +	struct device_node *node = dev->of_node;
>>>>>>> +	struct resource *res;
>>>>>>> +	struct cmdq *cmdq;
>>>>>>> +	int err, i;
>>>>>>> +
>>>>>>> +	cmdq = devm_kzalloc(dev, sizeof(*cmdq), GFP_KERNEL);
>>>>>>> +	if (!cmdq)
>>>>>>> +		return -ENOMEM;
>>>>>>> +	cmdq->dev = dev;
>>>>>>> +
>>>>>>> +	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
>>>>>>> +	cmdq->base = devm_ioremap_resource(dev, res);
>>>>>>> +	if (IS_ERR(cmdq->base)) {
>>>>>>> +		dev_err(dev, "failed to ioremap gce\n");
>>>>>>> +		return PTR_ERR(cmdq->base);
>>>>>>> +	}
>>>>>>> +
>>>>>>> +	cmdq->irq = irq_of_parse_and_map(node, 0);
>>>>>>> +	if (!cmdq->irq) {
>>>>>>> +		dev_err(dev, "failed to get irq\n");
>>>>>>> +		return -EINVAL;
>>>>>>> +	}
>>>>>>> +
>>>>>>> +	dev_dbg(dev, "cmdq device: addr:0x%p, va:0x%p, irq:%d\n",
>>>>>>> +		dev, cmdq->base, cmdq->irq);
>>>>>>> +
>>>>>>> +	mutex_init(&cmdq->task_mutex);
>>>>>>> +	spin_lock_init(&cmdq->exec_lock);
>>>>>>> +	cmdq->task_release_wq = alloc_ordered_workqueue(
>>>>>>> +			"%s", WQ_MEM_RECLAIM | WQ_HIGHPRI,
>>>>>>> +			"cmdq_task_wait_release");
>>>>>>> +
>>>>>>> +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
>>>>>>> +		cmdq->thread[i].base = cmdq->base + CMDQ_THR_BASE +
>>>>>>> +				CMDQ_THR_SHIFT * i;
>>>>>>> +		init_waitqueue_head(&cmdq->thread[i].wait_task_done);
>>>>>>> +		INIT_LIST_HEAD(&cmdq->thread[i].task_busy_list);
>>>>>>> +	}
>>>>>>> +
>>>>>>> +	platform_set_drvdata(pdev, cmdq);
>>>>>>> +
>>>>>>> +	err = devm_request_irq(dev, cmdq->irq, cmdq_irq_handler, IRQF_SHARED,
>>>>>>> +			       CMDQ_DRIVER_DEVICE_NAME, cmdq);
>>>>>>> +	if (err < 0) {
>>>>>>> +		dev_err(dev, "failed to register ISR (%d)\n", err);
>>>>>>> +		goto fail;
>>>>>>> +	}
>>>>>>> +
>>>>>>> +	cmdq->clock = devm_clk_get(dev, CMDQ_CLK_NAME);
>>>>>>> +	if (IS_ERR(cmdq->clock)) {
>>>>>>> +		dev_err(dev, "failed to get clk:%s\n", CMDQ_CLK_NAME);
>>>>>>> +		err = PTR_ERR(cmdq->clock);
>>>>>>> +		goto fail;
>>>>>>> +	}
>>>>>>> +	return 0;
>>>>>>> +
>>>>>>> +fail:
>>>>>>> +	cmdq_remove(pdev);
>>>>>>> +	return err;
>>>>>>> +}
>>>>>>> +
>>>>>>> +static const struct dev_pm_ops cmdq_pm_ops = {
>>>>>>> +	.suspend = cmdq_suspend,
>>>>>>> +	.resume = cmdq_resume,
>>>>>>> +};
>>>>>>> +
>>>>>>> +static const struct of_device_id cmdq_of_ids[] = {
>>>>>>> +	{.compatible = "mediatek,mt8173-gce",},
>>>>>>> +	{}
>>>>>>> +};
>>>>>>> +
>>>>>>> +static struct platform_driver cmdq_drv = {
>>>>>>> +	.probe = cmdq_probe,
>>>>>>> +	.remove = cmdq_remove,
>>>>>>> +	.driver = {
>>>>>>> +		.name = CMDQ_DRIVER_DEVICE_NAME,
>>>>>>> +		.owner = THIS_MODULE,
>>>>>>> +		.pm = &cmdq_pm_ops,
>>>>>>> +		.of_match_table = cmdq_of_ids,
>>>>>>> +	}
>>>>>>> +};
>>>>>>> +
>>>>>>> +builtin_platform_driver(cmdq_drv);
>>>>>>> diff --git a/include/soc/mediatek/cmdq.h b/include/soc/mediatek/cmdq.h
>>>>>>> new file mode 100644
>>>>>>> index 0000000..60eef3d
>>>>>>> --- /dev/null
>>>>>>> +++ b/include/soc/mediatek/cmdq.h
>>>>>>> @@ -0,0 +1,197 @@
>>>>>>> +/*
>>>>>>> + * Copyright (c) 2015 MediaTek Inc.
>>>>>>> + *
>>>>>>> + * This program is free software; you can redistribute it and/or modify
>>>>>>> + * it under the terms of the GNU General Public License version 2 as
>>>>>>> + * published by the Free Software Foundation.
>>>>>>> + *
>>>>>>> + * This program is distributed in the hope that it will be useful,
>>>>>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>>>>>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
>>>>>>> + * GNU General Public License for more details.
>>>>>>> + */
>>>>>>> +
>>>>>>> +#ifndef __MTK_CMDQ_H__
>>>>>>> +#define __MTK_CMDQ_H__
>>>>>>> +
>>>>>>> +#include <linux/platform_device.h>
>>>>>>> +#include <linux/types.h>
>>>>>>> +
>>>>>>> +enum cmdq_eng {
>>>>>>> +	CMDQ_ENG_DISP_AAL,
>>>>>>> +	CMDQ_ENG_DISP_COLOR0,
>>>>>>> +	CMDQ_ENG_DISP_COLOR1,
>>>>>>> +	CMDQ_ENG_DISP_DPI0,
>>>>>>> +	CMDQ_ENG_DISP_DSI0,
>>>>>>> +	CMDQ_ENG_DISP_DSI1,
>>>>>>> +	CMDQ_ENG_DISP_GAMMA,
>>>>>>> +	CMDQ_ENG_DISP_OD,
>>>>>>> +	CMDQ_ENG_DISP_OVL0,
>>>>>>> +	CMDQ_ENG_DISP_OVL1,
>>>>>>> +	CMDQ_ENG_DISP_PWM0,
>>>>>>> +	CMDQ_ENG_DISP_PWM1,
>>>>>>> +	CMDQ_ENG_DISP_RDMA0,
>>>>>>> +	CMDQ_ENG_DISP_RDMA1,
>>>>>>> +	CMDQ_ENG_DISP_RDMA2,
>>>>>>> +	CMDQ_ENG_DISP_UFOE,
>>>>>>> +	CMDQ_ENG_DISP_WDMA0,
>>>>>>> +	CMDQ_ENG_DISP_WDMA1,
>>>>>>> +	CMDQ_ENG_MAX,
>>>>>>> +};
>>>>>>> +
>>>>>>> +/* events for CMDQ and display */
>>>>>>> +enum cmdq_event {
>>>>>>> +	/* Display start of frame(SOF) events */
>>>>>>> +	CMDQ_EVENT_DISP_OVL0_SOF = 11,
>>>>>>> +	CMDQ_EVENT_DISP_OVL1_SOF = 12,
>>>>>>> +	CMDQ_EVENT_DISP_RDMA0_SOF = 13,
>>>>>>> +	CMDQ_EVENT_DISP_RDMA1_SOF = 14,
>>>>>>> +	CMDQ_EVENT_DISP_RDMA2_SOF = 15,
>>>>>>> +	CMDQ_EVENT_DISP_WDMA0_SOF = 16,
>>>>>>> +	CMDQ_EVENT_DISP_WDMA1_SOF = 17,
>>>>>>> +	/* Display end of frame(EOF) events */
>>>>>>> +	CMDQ_EVENT_DISP_OVL0_EOF = 39,
>>>>>>> +	CMDQ_EVENT_DISP_OVL1_EOF = 40,
>>>>>>> +	CMDQ_EVENT_DISP_RDMA0_EOF = 41,
>>>>>>> +	CMDQ_EVENT_DISP_RDMA1_EOF = 42,
>>>>>>> +	CMDQ_EVENT_DISP_RDMA2_EOF = 43,
>>>>>>> +	CMDQ_EVENT_DISP_WDMA0_EOF = 44,
>>>>>>> +	CMDQ_EVENT_DISP_WDMA1_EOF = 45,
>>>>>>> +	/* Mutex end of frame(EOF) events */
>>>>>>> +	CMDQ_EVENT_MUTEX0_STREAM_EOF = 53,
>>>>>>> +	CMDQ_EVENT_MUTEX1_STREAM_EOF = 54,
>>>>>>> +	CMDQ_EVENT_MUTEX2_STREAM_EOF = 55,
>>>>>>> +	CMDQ_EVENT_MUTEX3_STREAM_EOF = 56,
>>>>>>> +	CMDQ_EVENT_MUTEX4_STREAM_EOF = 57,
>>>>>>> +	/* Display underrun events */
>>>>>>> +	CMDQ_EVENT_DISP_RDMA0_UNDERRUN = 63,
>>>>>>> +	CMDQ_EVENT_DISP_RDMA1_UNDERRUN = 64,
>>>>>>> +	CMDQ_EVENT_DISP_RDMA2_UNDERRUN = 65,
>>>>>>> +	/* Keep this at the end of HW events */
>>>>>>> +	CMDQ_MAX_HW_EVENT_COUNT = 260,
>>>>>>> +};
>>>>>>> +
>>>>>>> +struct cmdq_cb_data {
>>>>>>> +	bool	err;
>>>>>>> +	void	*data;
>>>>>>> +};
>>>>>>> +
>>>>>>> +typedef int (*cmdq_async_flush_cb)(struct cmdq_cb_data data);
>>>>>>> +
>>>>>>> +struct cmdq_task;
>>>>>>> +struct cmdq;
>>>>>>> +
>>>>>>> +struct cmdq_rec {
>>>>>>> +	struct cmdq		*cmdq;
>>>>>>> +	u64			engine_flag;
>>>>>>> +	size_t			command_size;
>>>>>>> +	void			*buf;
>>>>>>> +	size_t			buf_size;
>>>>>>> +	bool			finalized;
>>>>>>> +};
>>>>
>>>> Why do we need cmdq_rec at all? Can't we just use the cmdq_task for that
>>>> and this way make the driver less complex?
>>>
>>> There are two main reasons for cmdq_rec.
>>> 1. It is slow to access dma too frequently.
>>>      So, we append commands to cacheable memory, and then flush to dma.
>>> 2. cmdq_rec is not thread safe, but cmdq_task needs thread safe.
>>>      If we merge them, we need to take care of some synchronization
>>>      issues.
>>>
>>>>>>> +
>>>>>>> +struct cmdq_base {
>>>>>>> +	int	subsys;
>>>>>>> +	u32	base;
>>>>>>
>>>>>> subsys can always be calculated via cmdq_subsys_base_to_id(base >> 16)
>>>>>> so we can get rid of the struct, right?
>>>>>
>>>>> Current subsys method is based on previous comment from Daniel Kurtz.
>>>>> Please take a look of our previous discussion.
>>>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/004483.html
>>>>> Thanks.
>>>>>
>>>>
>>>> I have to look deeper into this, but from what I read, the proposal from
>>>> Daniel [1] seems good to me.
>>>>
>>>> [1] https://patchwork.kernel.org/patch/8068311/
>>>
>>> The initial proposal has some problem, so please see the follow-up
>>> discussions. Thanks.
>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/003972.html
>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/004483.html
>>>
>>>
>>>>>>> +};
>>>>>>> +
>>>>>>> +/**
>>>>>>> + * cmdq_register_device() - register device which needs CMDQ
>>>>>>> + * @dev:		device
>>>>>>> + *
>>>>>>> + * Return: cmdq_base pointer or NULL for failed
>>>>>>> + */
>>>>>>> +struct cmdq_base *cmdq_register_device(struct device *dev);
>>>>>>> +
>>>>>>> +/**
>>>>>>> + * cmdq_rec_create() - create command queue record
>>>>>>> + * @dev:		device
>>>>>>> + * @engine_flag:	command queue engine flag
>>>>>>> + * @rec_ptr:		command queue record pointer to retrieve cmdq_rec
>>>>>>> + *
>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>> + */
>>>>>>> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
>>>>>>> +		    struct cmdq_rec **rec_ptr);
>>>>>>> +
>>>>>>> +/**
>>>>>>> + * cmdq_rec_write() - append write command to the command queue record
>>>>>>> + * @rec:	the command queue record
>>>>>>> + * @value:	the specified target register value
>>>>>>> + * @base:	the command queue base
>>>>>>> + * @offset:	register offset from module base
>>>>>>> + *
>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>> + */
>>>>>>> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value,
>>>>>>> +		   struct cmdq_base *base, u32 offset);
>>>>>>> +
>>>>>>> +/**
>>>>>>> + * cmdq_rec_write_mask() - append write command with mask to the command
>>>>>>> + *			   queue record
>>>>>>> + * @rec:	the command queue record
>>>>>>> + * @value:	the specified target register value
>>>>>>> + * @base:	the command queue base
>>>>>>> + * @offset:	register offset from module base
>>>>>>> + * @mask:	the specified target register mask
>>>>>>> + *
>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>> + */
>>>>>>> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
>>>>>>> +			struct cmdq_base *base, u32 offset, u32 mask);
>>>>>>> +
>>>>>>> +/**
>>>>>>> + * cmdq_rec_wfe() - append wait for event command to the command queue reco	rd
>>>>>>
>>>>>> reco rd -> record
>>>>>
>>>>> Will fix it.
>>>>>
>>>>>> Regards,
>>>>>> Matthias
>>>>>>
>>>>>>> + * @rec:	the command queue record
>>>>>>> + * @event:	the desired event type to "wait and CLEAR"
>>>>>>> + *
>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>> + */
>>>>>>> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event);
>>>>>>> +
>>>>>>> +/**
>>>>>>> + * cmdq_rec_clear_event() - append clear event command to the command queue
>>>>>>> + *			    record
>>>>>>> + * @rec:	the command queue record
>>>>>>> + * @event:	the desired event to be cleared
>>>>>>> + *
>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>> + */
>>>>>>> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event);
>>>>>>> +
>>>>>>> +/**
>>>>>>> + * cmdq_rec_flush() - trigger CMDQ to execute the recorded commands
>>>>>>> + * @rec:	the command queue record
>>>>>>> + *
>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>> + *
>>>>>>> + * Trigger CMDQ to execute the recorded commands. Note that this is a
>>>>>>> + * synchronous flush function. When the function returned, the recorded
>>>>>>> + * commands have been done.
>>>>>>> + */
>>>>>>> +int cmdq_rec_flush(struct cmdq_rec *rec);
>>>>>>> +
>>>>>>> +/**
>>>>>>> + * cmdq_rec_flush_async() - trigger CMDQ to asynchronously execute the recorded
>>>>>>> + *			    commands and call back after ISR is finished
>>>>>>> + * @rec:	the command queue record
>>>>>>> + * @cb:		called in the end of CMDQ ISR
>>>>>>> + * @data:	this data will pass back to cb
>>>>>>> + *
>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>> + *
>>>>>>> + * Trigger CMDQ to asynchronously execute the recorded commands and call back
>>>>>>> + * after ISR is finished. Note that this is an ASYNC function. When the function
>>>>>>> + * returned, it may or may not be finished. The ISR callback function is called
>>>>>>> + * in the end of ISR.
>>>>
>>>> "The callback is called from the ISR."
>>>>
>>>> Regards,
>>>> Matthias
>>>>
>>>>>>> + */
>>>>>>> +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
>>>>>>> +			 void *data);
>>>>>>> +
>>>>>>> +/**
>>>>>>> + * cmdq_rec_destroy() - destroy command queue record
>>>>>>> + * @rec:	the command queue record
>>>>>>> + */
>>>>>>> +void cmdq_rec_destroy(struct cmdq_rec *rec);
>>>>>>> +
>>>>>>> +#endif	/* __MTK_CMDQ_H__ */
>>>>>>>
>>>>>
>>>>> Thanks,
>>>>> HS
>>>>>
>>>
>>> Thanks,
>>> HS
>>>
>
>

[e-mail] June 3, 2016, 12:13 p.m. UTC | #10

Hi Matthias,

On Fri, 2016-06-03 at 13:18 +0200, Matthias Brugger wrote:
> 
> On 03/06/16 08:12, Horng-Shyang Liao wrote:
> > Hi Mathias,
> >
> > Please see my inline reply.
> >
> > On Thu, 2016-06-02 at 10:46 +0200, Matthias Brugger wrote:
> >>
> >> On 01/06/16 11:57, Horng-Shyang Liao wrote:
> >>> Hi Mathias,
> >>>
> >>> Please see my inline reply.
> >>>
> >>> On Tue, 2016-05-31 at 22:04 +0200, Matthias Brugger wrote:
> >>>>
> >>>> On 31/05/16 10:36, Horng-Shyang Liao wrote:
> >>>>> Hi Mathias,
> >>>>>
> >>>>> Please see my inline reply.
> >>>>>
> >>>>> On Mon, 2016-05-30 at 17:31 +0200, Matthias Brugger wrote:
> >>>>>>
> >>>>>> On 30/05/16 05:19, HS Liao wrote:
> >>>>>>> This patch is first version of Mediatek Command Queue(CMDQ) driver. The
> >>>>>>> CMDQ is used to help read/write registers with critical time limitation,
> >>>>>>> such as updating display configuration during the vblank. It controls
> >>>>>>> Global Command Engine (GCE) hardware to achieve this requirement.
> >>>>>>> Currently, CMDQ only supports display related hardwares, but we expect
> >>>>>>> it can be extended to other hardwares for future requirements.
> >>>>>>>
> >>>>>>> Signed-off-by: HS Liao <[e-mail]>
> >>>>>>> Signed-off-by: CK Hu <[e-mail]>
> >>>>>>> ---
> >>>>>>>      drivers/soc/mediatek/Kconfig    |  10 +
> >>>>>>>      drivers/soc/mediatek/Makefile   |   1 +
> >>>>>>>      drivers/soc/mediatek/mtk-cmdq.c | 943 ++++++++++++++++++++++++++++++++++++++++
> >>>>>>>      include/soc/mediatek/cmdq.h     | 197 +++++++++
> >>>>>>>      4 files changed, 1151 insertions(+)
> >>>>>>>      create mode 100644 drivers/soc/mediatek/mtk-cmdq.c
> >>>>>>>      create mode 100644 include/soc/mediatek/cmdq.h
> >>>>>>>
> >>>>>>> diff --git a/drivers/soc/mediatek/Kconfig b/drivers/soc/mediatek/Kconfig
> >>>>>>> index 0a4ea80..c4ad75c 100644
> >>>>>>> --- a/drivers/soc/mediatek/Kconfig
> >>>>>>> +++ b/drivers/soc/mediatek/Kconfig
> >>>>>>> @@ -1,6 +1,16 @@
> >>>>>>>      #
> >>>>>>>      # MediaTek SoC drivers
> >>>>>>>      #
> >>>>>>> +config MTK_CMDQ
> >>>>>>> +	bool "MediaTek CMDQ Support"
> >>>>>>> +	depends on ARCH_MEDIATEK || COMPILE_TEST
> >>>>
> >>>> depends on ARM64 ?
> >>>
> >>> Will add ARM64.
> >>>
> >>>>>>> +	select MTK_INFRACFG
> >>>>>>> +	help
> >>>>>>> +	  Say yes here to add support for the MediaTek Command Queue (CMDQ)
> >>>>>>> +	  driver. The CMDQ is used to help read/write registers with critical
> >>>>>>> +	  time limitation, such as updating display configuration during the
> >>>>>>> +	  vblank.
> >>>>>>> +
> >>>>>>>      config MTK_INFRACFG
> >>>>>>>      	bool "MediaTek INFRACFG Support"
> >>>>>>>      	depends on ARCH_MEDIATEK || COMPILE_TEST
> >>>>>>> diff --git a/drivers/soc/mediatek/Makefile b/drivers/soc/mediatek/Makefile
> >>>>>>> index 12998b0..f7397ef 100644
> >>>>>>> --- a/drivers/soc/mediatek/Makefile
> >>>>>>> +++ b/drivers/soc/mediatek/Makefile
> >>>>>>> @@ -1,3 +1,4 @@
> >>>>>>> +obj-$(CONFIG_MTK_CMDQ) += mtk-cmdq.o
> >>>>>>>      obj-$(CONFIG_MTK_INFRACFG) += mtk-infracfg.o
> >>>>>>>      obj-$(CONFIG_MTK_PMIC_WRAP) += mtk-pmic-wrap.o
> >>>>>>>      obj-$(CONFIG_MTK_SCPSYS) += mtk-scpsys.o
> >>>>>>> diff --git a/drivers/soc/mediatek/mtk-cmdq.c b/drivers/soc/mediatek/mtk-cmdq.c
> >>>>>>> new file mode 100644
> >>>>>>> index 0000000..e9d6e1c
> >>>>>>> --- /dev/null
> >>>>>>> +++ b/drivers/soc/mediatek/mtk-cmdq.c
> >>>>>>> @@ -0,0 +1,943 @@
> >>>>>>> +/*
> >>>>>>> + * Copyright (c) 2015 MediaTek Inc.
> >>>>>>> + *
> >>>>>>> + * This program is free software; you can redistribute it and/or modify
> >>>>>>> + * it under the terms of the GNU General Public License version 2 as
> >>>>>>> + * published by the Free Software Foundation.
> >>>>>>> + *
> >>>>>>> + * This program is distributed in the hope that it will be useful,
> >>>>>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> >>>>>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> >>>>>>> + * GNU General Public License for more details.
> >>>>>>> + */
> >>>>>>> +
> >>>>>>> +#include <linux/clk.h>
> >>>>>>> +#include <linux/clk-provider.h>
> >>>>>>> +#include <linux/completion.h>
> >>>>>>> +#include <linux/dma-mapping.h>
> >>>>>>> +#include <linux/errno.h>
> >>>>>>> +#include <linux/interrupt.h>
> >>>>>>> +#include <linux/iopoll.h>
> >>>>>>> +#include <linux/kernel.h>
> >>>>>>> +#include <linux/kthread.h>
> >>>>>>> +#include <linux/module.h>
> >>>>>>> +#include <linux/mutex.h>
> >>>>>>> +#include <linux/of_address.h>
> >>>>>>> +#include <linux/of_irq.h>
> >>>>>>> +#include <linux/platform_device.h>
> >>>>>>> +#include <linux/slab.h>
> >>>>>>> +#include <linux/spinlock.h>
> >>>>>>> +#include <linux/suspend.h>
> >>>>>>> +#include <linux/workqueue.h>
> >>>>>>> +#include <soc/mediatek/cmdq.h>
> >>>>>>> +
> >>>>>>> +#define CMDQ_INITIAL_CMD_BLOCK_SIZE	PAGE_SIZE
> >>>>>>> +#define CMDQ_INST_SIZE			8 /* instruction is 64-bit */
> >>>>>>> +#define CMDQ_TIMEOUT_MS			1000
> >>>>>>> +#define CMDQ_IRQ_MASK			0xffff
> >>>>>>> +#define CMDQ_DRIVER_DEVICE_NAME		"mtk_cmdq"
> >>>>>>> +#define CMDQ_CLK_NAME			"gce"
> >>>>>>
> >>>>>> We can put the names in directly to un-bloat the defines.
> >>>>>
> >>>>> I will use the names directly and remove defines.
> >>>>>
> >>>>>>> +
> >>>>>>> +#define CMDQ_CURR_IRQ_STATUS		0x010
> >>>>>>> +#define CMDQ_CURR_LOADED_THR		0x018
> >>>>>>> +#define CMDQ_THR_SLOT_CYCLES		0x030
> >>>>>>> +
> >>>>>>> +#define CMDQ_THR_BASE			0x100
> >>>>>>> +#define CMDQ_THR_SHIFT			0x080
> >>>>>>
> >>>>>> Wouldn't be CMDQ_THR_SIZE more accurate?
> >>>>>
> >>>>> Will rename it.
> >>>>>
> >>>>>>> +#define CMDQ_THR_WARM_RESET		0x00
> >>>>>>> +#define CMDQ_THR_ENABLE_TASK		0x04
> >>>>>>> +#define CMDQ_THR_SUSPEND_TASK		0x08
> >>>>>>> +#define CMDQ_THR_CURR_STATUS		0x0c
> >>>>>>> +#define CMDQ_THR_IRQ_STATUS		0x10
> >>>>>>> +#define CMDQ_THR_IRQ_ENABLE		0x14
> >>>>>>> +#define CMDQ_THR_CURR_ADDR		0x20
> >>>>>>> +#define CMDQ_THR_END_ADDR		0x24
> >>>>>>> +#define CMDQ_THR_CFG			0x40
> >>>>>>> +
> >>>>>>> +#define CMDQ_THR_ENABLED		0x1
> >>>>>>> +#define CMDQ_THR_DISABLED		0x0
> >>>>>>> +#define CMDQ_THR_SUSPEND		0x1
> >>>>>>> +#define CMDQ_THR_RESUME			0x0
> >>>>>>> +#define CMDQ_THR_STATUS_SUSPENDED	BIT(1)
> >>>>>>> +#define CMDQ_THR_DO_WARM_RESET		BIT(0)
> >>>>>>> +#define CMDQ_THR_ACTIVE_SLOT_CYCLES	0x3200
> >>>>>>> +#define CMDQ_THR_PRIORITY		3
> >>>>>>> +#define CMDQ_THR_IRQ_DONE		0x1
> >>>>>>> +#define CMDQ_THR_IRQ_ERROR		0x12
> >>>>>>> +#define CMDQ_THR_IRQ_EN			0x13 /* done + error */
> >>>>>>
> >>>>>> #define CMDQ_THR_IRQ_EN (CMDQ_THR_IRQ_ERROR | CMDQ_THR_IRQ_DONE)
> >>>>>
> >>>>> Will do.
> >>>>>
> >>>>>>> +#define CMDQ_THR_IRQ_MASK		0x13
> >>>>>>
> >>>>>> never used.
> >>>>>
> >>>>> Will remove.
> >>>>>
> >>>>>>> +#define CMDQ_THR_EXECUTING		BIT(31)
> >>>>>>> +
> >>>>>>> +#define CMDQ_ARG_A_WRITE_MASK		0xffff
> >>>>>>> +#define CMDQ_SUBSYS_MASK		0x1f
> >>>>>>> +#define CMDQ_OP_CODE_MASK		0xff000000
> >>>>>>> +
> >>>>>>> +#define CMDQ_OP_CODE_SHIFT		24
> >>>>>>
> >>>>>> Couldn't we connect the mask with the shift, or aren't they related?
> >>>>>>
> >>>>>> #define CMDQ_OP_CODE_MASK (0xff << CMDQ_OP_CODE_SHIFT)
> >>>>>
> >>>>> Will do.
> >>>>>
> >>>>>>> +#define CMDQ_SUBSYS_SHIFT		16
> >>>>>>> +
> >>>>>>> +#define CMDQ_WRITE_ENABLE_MASK		BIT(0)
> >>>>>>> +#define CMDQ_JUMP_BY_OFFSET		0x10000000
> >>>>>>> +#define CMDQ_JUMP_BY_PA			0x10000001
> >>>>>>> +#define CMDQ_JUMP_PASS			CMDQ_INST_SIZE
> >>>>>>> +#define CMDQ_WFE_UPDATE			BIT(31)
> >>>>>>> +#define CMDQ_WFE_WAIT			BIT(15)
> >>>>>>> +#define CMDQ_WFE_WAIT_VALUE		0x1
> >>>>>>> +#define CMDQ_EOC_IRQ_EN			BIT(0)
> >>>>>>> +
> >>>>>>> +enum cmdq_thread_index {
> >>>>>>> +	CMDQ_THR_DISP_MAIN_IDX,	/* main */
> >>>>>>> +	CMDQ_THR_DISP_SUB_IDX,	/* sub */
> >>>>>>> +	CMDQ_THR_DISP_MISC_IDX,	/* misc */
> >>>>>>> +	CMDQ_THR_MAX_COUNT,	/* max */
> >>>>>>> +};
> >>>>>>> +
> >>>>>>> +/*
> >>>>>>> + * CMDQ_CODE_MOVE:
> >>>>>>> + *   move value into internal register as mask
> >>>>>>> + *   format: op mask
> >>>>>>> + * CMDQ_CODE_WRITE:
> >>>>>>> + *   write value into target register
> >>>>>>> + *   format: op subsys address value
> >>>>>>> + * CMDQ_CODE_JUMP:
> >>>>>>> + *   jump by offset
> >>>>>>> + *   format: op offset
> >>>>>>> + * CMDQ_CODE_WFE:
> >>>>>>> + *   wait for event and clear
> >>>>>>> + *   it is just clear if no wait
> >>>>>>> + *   format: [wait]  op event update:1 to_wait:1 wait:1
> >>>>>>> + *           [clear] op event update:1 to_wait:0 wait:0
> >>>>>>> + * CMDQ_CODE_EOC:
> >>>>>>> + *   end of command
> >>>>>>> + *   format: op irq_flag
> >>>>>>> + */
> >>>>>>
> >>>>>> I think we need more documentation of how this command queue engine is
> >>>>>> working. If not, I think it will be really complicated to understand how
> >>>>>> to use this.
> >>>>>>
> >>>>>>> +enum cmdq_code {
> >>>>>>> +	CMDQ_CODE_MOVE = 0x02,
> >>>>>>> +	CMDQ_CODE_WRITE = 0x04,
> >>>>>>> +	CMDQ_CODE_JUMP = 0x10,
> >>>>>>> +	CMDQ_CODE_WFE = 0x20,
> >>>>>>> +	CMDQ_CODE_EOC = 0x40,
> >>>>>>> +};
> >>>>>>> +
> >>>>>>> +enum cmdq_task_state {
> >>>>>>> +	TASK_STATE_BUSY, /* running on a GCE thread */
> >>>>>>> +	TASK_STATE_ERROR,
> >>>>>>> +	TASK_STATE_DONE,
> >>>>>>> +};
> >>>>>>> +
> >>>>>>> +struct cmdq_task_cb {
> >>>>>>> +	cmdq_async_flush_cb	cb;
> >>>>>>> +	void			*data;
> >>>>>>> +};
> >>>>>>> +
> >>>>>>> +struct cmdq_thread {
> >>>>>>> +	void __iomem		*base;
> >>>>>>> +	struct list_head	task_busy_list;
> >>>>>>> +	wait_queue_head_t	wait_task_done;
> >>>>>>> +};
> >>>>>>> +
> >>>>>>> +struct cmdq_task {
> >>>>>>> +	struct cmdq		*cmdq;
> >>>>>>> +	struct list_head	list_entry;
> >>>>>>> +	enum cmdq_task_state	task_state;
> >>>>>>> +	void			*va_base;
> >>>>>>> +	dma_addr_t		pa_base;
> >>>>>>> +	u64			engine_flag;
> >>>>>>> +	size_t			command_size;
> >>>>>>> +	u32			num_cmd;
> >>>>>>
> >>>>>> num_cmd is directly connected to command_size. I prefer to just keep
> >>>>>> num_cmd and calculate the command_size where necessary.
> >>>>>
> >>>>> After I trace code, I prefer to keep command_size and calculate num_cmd
> >>>>> where necessary. What do you think?
> >>>>>
> >>>>
> >>>> I suppose you prefer this, as you are writing to the GCE depending on
> >>>> the command_size. I think it is worth to create a macro for the
> >>>> calculation of the number of commands, to make the code more readable.
> >>>> Would be nice if you would just pass cmdq_task to it and it would return
> >>>> the number. Just as an idea.
> >>>
> >>> Will add macro.
> >>>
> >>>>>>> +	struct cmdq_thread	*thread;
> >>>>>>> +	struct cmdq_task_cb	cb;
> >>>>>>> +	struct work_struct	release_work;
> >>>>>>> +};
> >>>>>>> +
> >>>>>>> +struct cmdq {
> >>>>>>> +	struct device		*dev;
> >>>>>>> +	void __iomem		*base;
> >>>>>>> +	u32			irq;
> >>>>>>> +	struct workqueue_struct	*task_release_wq;
> >>>>>>> +	struct cmdq_thread	thread[CMDQ_THR_MAX_COUNT];
> >>>>>>> +	struct mutex		task_mutex;	/* for task */
> >>>>>>> +	spinlock_t		exec_lock;	/* for exec */
> >>>>>>> +	struct clk		*clock;
> >>>>>>> +	bool			suspended;
> >>>>>>> +};
> >>>>>>> +
> >>>>>>> +struct cmdq_subsys {
> >>>>>>> +	u32	base;
> >>>>>>> +	int	id;
> >>>>>>> +};
> >>>>>>> +
> >>>>>>> +static const struct cmdq_subsys gce_subsys[] = {
> >>>>>>> +	{0x1400, 1},
> >>>>>>> +	{0x1401, 2},
> >>>>>>> +	{0x1402, 3},
> >>>>>>> +};
> >>>>>>> +
> >>>>>>> +static int cmdq_subsys_base_to_id(u32 base)
> >>>>>>> +{
> >>>>>>> +	int i;
> >>>>>>> +
> >>>>>>> +	for (i = 0; i < ARRAY_SIZE(gce_subsys); i++)
> >>>>>>> +		if (gce_subsys[i].base == base)
> >>>>>>> +			return gce_subsys[i].id;
> >>>>>>> +	return -EFAULT;
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static int cmdq_eng_get_thread(u64 flag)
> >>>>>>> +{
> >>>>>>> +	if (flag & BIT_ULL(CMDQ_ENG_DISP_DSI0))
> >>>>>>> +		return CMDQ_THR_DISP_MAIN_IDX;
> >>>>>>> +	else if (flag & BIT_ULL(CMDQ_ENG_DISP_DPI0))
> >>>>>>> +		return CMDQ_THR_DISP_SUB_IDX;
> >>>>>>> +	else
> >>>>>>> +		return CMDQ_THR_DISP_MISC_IDX;
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static void cmdq_task_release(struct cmdq_task *task)
> >>>>>>> +{
> >>>>>>> +	struct cmdq *cmdq = task->cmdq;
> >>>>>>> +
> >>>>>>> +	dma_free_coherent(cmdq->dev, task->command_size, task->va_base,
> >>>>>>> +			  task->pa_base);
> >>>>>>> +	kfree(task);
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static struct cmdq_task *cmdq_task_acquire(struct cmdq_rec *rec,
> >>>>>>> +					   struct cmdq_task_cb cb)
> >>>>>>> +{
> >>>>>>> +	struct cmdq *cmdq = rec->cmdq;
> >>>>>>> +	struct device *dev = cmdq->dev;
> >>>>>>> +	struct cmdq_task *task;
> >>>>>>> +
> >>>>>>> +	task = kzalloc(sizeof(*task), GFP_KERNEL);
> >>>>>>> +	INIT_LIST_HEAD(&task->list_entry);
> >>>>>>> +	task->va_base = dma_alloc_coherent(dev, rec->command_size,
> >>>>>>> +					   &task->pa_base, GFP_KERNEL);
> >>>>>>> +	if (!task->va_base) {
> >>>>>>> +		dev_err(dev, "allocate command buffer failed\n");
> >>>>>>> +		kfree(task);
> >>>>>>> +		return NULL;
> >>>>>>> +	}
> >>>>>>> +
> >>>>>>> +	task->cmdq = cmdq;
> >>>>>>> +	task->command_size = rec->command_size;
> >>>>>>> +	task->engine_flag = rec->engine_flag;
> >>>>>>> +	task->task_state = TASK_STATE_BUSY;
> >>>>>>> +	task->cb = cb;
> >>>>>>> +	memcpy(task->va_base, rec->buf, rec->command_size);
> >>>>>>> +	task->num_cmd = task->command_size / CMDQ_INST_SIZE;
> >>>>>>> +	return task;
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static void cmdq_thread_writel(struct cmdq_thread *thread, u32 value,
> >>>>>>> +			       u32 offset)
> >>>>>>> +{
> >>>>>>> +	writel(value, thread->base + offset);
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static u32 cmdq_thread_readl(struct cmdq_thread *thread, u32 offset)
> >>>>>>> +{
> >>>>>>> +	return readl(thread->base + offset);
> >>>>>>> +}
> >>>>>>
> >>>>>> We can get rid of cmdq_thread_readl/writel.
> >>>>>
> >>>>> Will do.
> >>>>>
> >>>>>>> +
> >>>>>>> +static int cmdq_thread_suspend(struct cmdq *cmdq, struct cmdq_thread *thread)
> >>>>>>> +{
> >>>>>>> +	u32 status;
> >>>>>>> +
> >>>>>>> +	cmdq_thread_writel(thread, CMDQ_THR_SUSPEND, CMDQ_THR_SUSPEND_TASK);
> >>>>>>> +
> >>>>>>> +	/* If already disabled, treat as suspended successful. */
> >>>>>>> +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> >>>>>>> +	      CMDQ_THR_ENABLED))
> >>>>>>> +		return 0;
> >>>>>>> +
> >>>>>>> +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_CURR_STATUS,
> >>>>>>> +			status, status & CMDQ_THR_STATUS_SUSPENDED, 0, 10)) {
> >>>>>>> +		dev_err(cmdq->dev, "suspend GCE thread 0x%x failed\n",
> >>>>>>> +			(u32)(thread->base - cmdq->base));
> >>>>>>> +		return -EFAULT;
> >>>>>>> +	}
> >>>>>>> +
> >>>>>>> +	return 0;
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static void cmdq_thread_resume(struct cmdq_thread *thread)
> >>>>>>> +{
> >>>>>>> +	cmdq_thread_writel(thread, CMDQ_THR_RESUME, CMDQ_THR_SUSPEND_TASK);
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static int cmdq_thread_reset(struct cmdq *cmdq, struct cmdq_thread *thread)
> >>>>>>> +{
> >>>>>>> +	u32 warm_reset;
> >>>>>>> +
> >>>>>>> +	cmdq_thread_writel(thread, CMDQ_THR_DO_WARM_RESET, CMDQ_THR_WARM_RESET);
> >>>>>>> +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_WARM_RESET,
> >>>>>>> +			warm_reset, !(warm_reset & CMDQ_THR_DO_WARM_RESET),
> >>>>>>> +			0, 10)) {
> >>>>>>> +		dev_err(cmdq->dev, "reset GCE thread 0x%x failed\n",
> >>>>>>> +			(u32)(thread->base - cmdq->base));
> >>>>>>> +		return -EFAULT;
> >>>>>>> +	}
> >>>>>>> +	writel(CMDQ_THR_ACTIVE_SLOT_CYCLES, cmdq->base + CMDQ_THR_SLOT_CYCLES);
> >>>>>>> +	return 0;
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static void cmdq_thread_disable(struct cmdq *cmdq, struct cmdq_thread *thread)
> >>>>>>> +{
> >>>>>>> +	cmdq_thread_reset(cmdq, thread);
> >>>>>>> +	cmdq_thread_writel(thread, CMDQ_THR_DISABLED, CMDQ_THR_ENABLE_TASK);
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +/* notify GCE to re-fetch commands by setting GCE thread PC */
> >>>>>>> +static void cmdq_thread_invalidate_fetched_data(struct cmdq_thread *thread)
> >>>>>>> +{
> >>>>>>> +	cmdq_thread_writel(thread,
> >>>>>>> +			   cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR),
> >>>>>>> +			   CMDQ_THR_CURR_ADDR);
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static void cmdq_task_insert_into_thread(struct cmdq_task *task)
> >>>>>>> +{
> >>>>>>> +	struct cmdq_thread *thread = task->thread;
> >>>>>>> +	struct cmdq_task *prev_task = list_last_entry(
> >>>>>>> +			&thread->task_busy_list, typeof(*task), list_entry);
> >>>>>>> +	u64 *prev_task_base = prev_task->va_base;
> >>>>>>> +
> >>>>>>> +	/* let previous task jump to this task */
> >>>>>>> +	prev_task_base[prev_task->num_cmd - 1] = (u64)CMDQ_JUMP_BY_PA << 32 |
> >>>>>>> +						 task->pa_base;
> >>>>>>> +
> >>>>>>> +	cmdq_thread_invalidate_fetched_data(thread);
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +/* we assume tasks in the same display GCE thread are waiting the same event. */
> >>>>>>> +static void cmdq_task_remove_wfe(struct cmdq_task *task)
> >>>>>>> +{
> >>>>>>> +	u32 wfe_option = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
> >>>>>>> +	u32 wfe_op = CMDQ_CODE_WFE << CMDQ_OP_CODE_SHIFT;
> >>>>>>> +	u32 *base = task->va_base;
> >>>>>>> +	u32 num_cmd = task->num_cmd << 1;
> >>>>>>> +	int i;
> >>>>>>> +
> >>>>>>> +	for (i = 0; i < num_cmd; i += 2)
> >>>>>>> +		if (base[i] == wfe_option &&
> >>>>>>> +		    (base[i + 1] & CMDQ_OP_CODE_MASK) == wfe_op) {
> >>>>>>> +			base[i] = CMDQ_JUMP_PASS;
> >>>>>>> +			base[i + 1] = CMDQ_JUMP_BY_OFFSET;
> >>>>>>> +		}
> >>>>>>
> >>>>>> After using the command buffer as a void pointer a u64 pointer, we now
> >>>>>> reference to it as u32. I would prefer to explain here, how the command
> >>>>>> looks like we are searching for and use a for loop passing task->num_cmd
> >>>>>> instead.
> >>>>>
> >>>>> Will use u64* to rewrite the above code.
> >>>>>
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static void cmdq_task_exec(struct cmdq_task *task, struct cmdq_thread *thread)
> >>>>>>> +{
> >>>>>>> +	struct cmdq *cmdq = task->cmdq;
> >>>>>>> +	unsigned long flags;
> >>>>>>> +	unsigned long curr_pa, end_pa;
> >>>>>>> +
> >>>>>>> +	WARN_ON(clk_prepare_enable(cmdq->clock) < 0);
> >>>>>>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
> >>>>>>
> >>>>>> cmdq_task_exec is called with cmdq->task_mutex held, so why do we need
> >>>>>> the spin_lock here? Can't we just use one of the two?
> >>>>>
> >>>>> We can drop task_mutex, but we will get some side effects.
> >>>>> 1. exec_lock needs to include more code, but I think it is not good for
> >>>>>       spinlock.
> >>>>> 2. In cmdq_rec_flush_async(), task_mutex needs to protect
> >>>>>       (1) cmdq->suspended, (2) cmdq_task_exec(), and
> >>>>>       (3) cmdq_task_wait_release_schedule().
> >>>>>       If we drop task_mutex, we have to put cmdq->suspended if condition
> >>>>>       just before cmdq_task_exec() and inside exec_lock, and we have to
> >>>>>       release task and its command buffer if error. This will let flow
> >>>>>       become more complex and enlarge code size.
> >>>>>
> >>>>> What do you think?
> >>>>
> >>>> Why do you need to protect cmdq_task_wait_release_schedule? We don't
> >>>> care about the order of the workqueue elements, do we?
> >>>
> >>> According to CK's comment, we have to ensure order to remove previous
> >>> task.
> >>> http://lists.infradead.org/pipermail/linux-mediatek/2016-May/005547.html
> >>>
> >>>> As far as I understand you would need to protect cmdq_task_acquire as
> >>>> well, to "ensure" continously growing pa_base. More on that below.
> >>>
> >>> We need to ensure continuous physical addresses in a task, but we don't
> >>> need to ensure continuous physical addresses between tasks.
> >>> So, I think it is unnecessary to protect by mutex or spinlock.
> >>>
> >>
> >> True, I didn't get that.
> >>
> >>>>>
> >>>>>>> +	task->thread = thread;
> >>>>>>> +	task->task_state = TASK_STATE_BUSY;
> >>>>>>
> >>>>>> That was already set in cmdq_task_acquire, why do we need to set it here
> >>>>>> again?
> >>>>>
> >>>>> Will drop it.
> >>>>>
> >>>>
> >>>> Yeah, but I think it makes more sense to drop it in cmdq_task_acquire
> >>>> instead.
> >>>
> >>> Will drop it in cmdq_task_acquire instead.
> >>>
> >>>>>>> +	if (list_empty(&thread->task_busy_list)) {
> >>>>>>> +		WARN_ON(cmdq_thread_reset(cmdq, thread) < 0);
> >>>>>>> +
> >>>>>>> +		cmdq_thread_writel(thread, task->pa_base, CMDQ_THR_CURR_ADDR);
> >>>>>>> +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
> >>>>>>> +				   CMDQ_THR_END_ADDR);
> >>>>>>> +		cmdq_thread_writel(thread, CMDQ_THR_PRIORITY, CMDQ_THR_CFG);
> >>>>>>> +		cmdq_thread_writel(thread, CMDQ_THR_IRQ_EN,
> >>>>>>> +				   CMDQ_THR_IRQ_ENABLE);
> >>>>>>> +
> >>>>>>> +		cmdq_thread_writel(thread, CMDQ_THR_ENABLED,
> >>>>>>> +				   CMDQ_THR_ENABLE_TASK);
> >>>>>>> +	} else {
> >>>>>>> +		WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
> >>>>>>> +
> >>>>>>> +		/*
> >>>>>>> +		 * check boundary condition
> >>>>>>> +		 * PC = END - 8, EOC is executed
> >>>>>>> +		 * PC = END, all CMDs are executed
> >>>>>>> +		 */
> >>>>>>> +		curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> >>>>>>> +		end_pa = cmdq_thread_readl(thread, CMDQ_THR_END_ADDR);
> >>>>>>> +		if (curr_pa == end_pa - 8 || curr_pa == end_pa) {
> >>>>>>
> >>>>>> 8 refers to CMDQ_INST_SIZE, right?
> >>>>>
> >>>>> Yes, I will use CMDQ_INST_SIZE.
> >>>>>
> >>>>>>> +			/* set to this task directly */
> >>>>>>> +			cmdq_thread_writel(thread, task->pa_base,
> >>>>>>> +					   CMDQ_THR_CURR_ADDR);
> >>>>>>> +		} else {
> >>>>>>> +			cmdq_task_insert_into_thread(task);
> >>>>>>> +
> >>>>>>> +			if (thread == &cmdq->thread[CMDQ_THR_DISP_MAIN_IDX] ||
> >>>>>>> +			    thread == &cmdq->thread[CMDQ_THR_DISP_SUB_IDX])
> >>>>>>> +				cmdq_task_remove_wfe(task);
> >>>>>>
> >>>>>> We could do this check using the task->engine_flag, I suppose that's
> >>>>>> easier to undestand then.
> >>>>>
> >>>>> Will use task->engine_flag.
> >>>>>
> >>>>>>> +
> >>>>>>> +			smp_mb(); /* modify jump before enable thread */
> >>>>>>> +		}
> >>>>>>> +
> >>>>>>> +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
> >>>>>>> +				   CMDQ_THR_END_ADDR);
> >>>>>>> +		cmdq_thread_resume(thread);
> >>>>>>> +	}
> >>>>>>> +	list_move_tail(&task->list_entry, &thread->task_busy_list);
> >>>>>>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
> >>>>>>> +				   struct cmdq_thread *thread, u32 irq_flag)
> >>>>>>> +{
> >>>>>>> +	struct cmdq_task *task, *tmp, *curr_task = NULL;
> >>>>>>> +	u32 curr_pa;
> >>>>>>> +	struct cmdq_cb_data cmdq_cb_data;
> >>>>>>> +	bool err;
> >>>>>>> +
> >>>>>>> +	if (irq_flag & CMDQ_THR_IRQ_ERROR)
> >>>>>>> +		err = true;
> >>>>>>> +	else if (irq_flag & CMDQ_THR_IRQ_DONE)
> >>>>>>> +		err = false;
> >>>>>>> +	else
> >>>>>>> +		return;
> >>>>>>> +
> >>>>>>> +	curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> >>>>>>> +
> >>>>>>> +	list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
> >>>>>>> +				 list_entry) {
> >>>>>>> +		if (curr_pa >= task->pa_base &&
> >>>>>>> +		    curr_pa < (task->pa_base + task->command_size))
> >>>>>>
> >>>>>> What are you checking here? It seems as if you make some implcit
> >>>>>> assumptions about pa_base and the order of execution of commands in the
> >>>>>> thread. Is it save to do so? Does dma_alloc_coherent give any guarantees
> >>>>>> about dma_handle?
> >>>>>
> >>>>> 1. Check what is the current running task in this GCE thread.
> >>>>> 2. Yes.
> >>>>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
> >>>>>
> >>>>
> >>>> Yes, physical addresses might be continous, but AFAIK there is no
> >>>> guarantee that the dma_handle address is steadily growing, when calling
> >>>> dma_alloc_coherent. And if I understand the code correctly, you use this
> >>>> assumption to decide if the task picked from task_busy_list is currently
> >>>> executing. So I think this mecanism is not working.
> >>>
> >>> I don't use dma_handle address, and just use physical addresses.
> >>>   From CPU's point of view, tasks are linked by the busy list.
> >>>   From GCE's point of view, tasks are linked by the JUMP command.
> >>>
> >>>> In which cases does the HW thread raise an interrupt.
> >>>> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
> >>>
> >>> GCE will raise interrupt if any task is done or error.
> >>> However, GCE is fast, so CPU may get multiple done tasks
> >>> when it is running ISR.
> >>>
> >>> In case of error, that GCE thread will pause and raise interrupt.
> >>> So, CPU may get multiple done tasks and one error task.
> >>>
> >>
> >> I think we should reimplement the ISR mechanism. Can't we just read
> >> CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave
> >> cmdq_handle_error_done to the thread_fn? You will need to pass
> >> information from the handler to thread_fn, but that shouldn't be an
> >> issue. AFAIK interrupts are disabled in the handler, so we should stay
> >> there as short as possible. Traversing task_busy_list is expensive, so
> >> we need to do it in a thread context.
> >
> > Actually, our initial implementation is similar to your suggestion,
> > but display needs CMDQ to return callback function very precisely,
> > else display will drop frame.
> > For display, CMDQ interrupt will be raised every 16 ~ 17 ms,
> > and CMDQ needs to call callback function in ISR.
> > If we defer callback to workqueue, the time interval may be larger than
> > 32 ms.sometimes.
> >
> 
> I think the problem is, that you implemented the workqueue as a ordered 
> workqueue, so there is no parallel processing. I'm still not sure why 
> you need the workqueue to be ordered. Can you please explain.

The order should be kept.
Let me use mouse cursor as an example.
If task 1 means move mouse cursor to point A, task 2 means point B,
and task 3 means point C, our expected result is A -> B -> C.
If the order is not kept, the result could become A -> C -> B.

> >> I keep thinking about how to get rid of the two data structures,
> >> task_busy_list and the task_release_wq. We need the latter for the only
> >> sake of getting a timeout.
> >>
> >> Did you have a look on how the mailbox framework handles this?
> >> By the way, what is the reason to not implement the whole driver as a
> >> mailbox controller? For me, this driver looks like a good fit.
> >
> > CMDQ needs to encode commands for GCE hardware. We think this behavior
> > should be put in CMDQ driver, and client just call CMDQ functions.
> > Therefore, if we want to use mailbox framework, cmdq_rec must be
> > mailbox client, and the others must be mailbox controller.
> >
> 
> You mean the functions to fill the cmdq_rec and execute it?
> I think this should be part of the driver.

Yes.

> Jassi, can you have a look on the interface this driver exports [0].
> They are needed to actually create the message which will be send.
> Could something like this be part of a mailbox driver?
> 
> [0] https://patchwork.kernel.org/patch/9140221/
> 
> > However, if we use mailbox controller, CMDQ driver still needs to
> > control busy list for each GCE thread, and use workqueue to handle
> > timeout tasks.
> >
> 
> Let me summarize my ideas around this driver:
> When we enter the ISR, we know that all task in task_busy_list before 
> the entry which represents curr_task can be set to TASK_STATE_DONE.
> The curr_task could be TASK_STATE_ERROR if the corresponding bit in the 
> irq status registers is set.
> Do we need to call the callback in the same order as the tasks got 
> dispatched to the HW thread? If not, we could try to call all this 
> callbacks in a multithreaded workqueue.

Yes, we should keep order.

> Regards,
> Matthias

Thanks,
HS

> > The only thing that we can borrow from mailbox framework is the send
> > (CMDQ flush) and receive (CMDQ callback) interface, However, we don't
> > think we can gain many benefits from it, and we have some overheads to
> > conform to mailbox interface.
> >
> >
> >>
> >>>>>>> +			curr_task = task;
> >>>>>>> +		if (task->cb.cb) {
> >>>>>>> +			cmdq_cb_data.err = curr_task ? err : false;
> >>>>>>> +			cmdq_cb_data.data = task->cb.data;
> >>>>>>> +			task->cb.cb(cmdq_cb_data);
> >>>>>>> +		}
> >>>>>>> +		task->task_state = (curr_task && err) ? TASK_STATE_ERROR :
> >>>>>>> +				TASK_STATE_DONE;
> >>>>>>> +		list_del(&task->list_entry);
> >>>>>>> +		if (curr_task)
> >>>>>>> +			break;
> >>>>>>> +	}
> >>>>>>> +
> >>>>>>> +	wake_up(&thread->wait_task_done);
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static void cmdq_thread_irq_handler(struct cmdq *cmdq, int tid)
> >>>>>>> +{
> >>>>>>> +	struct cmdq_thread *thread = &cmdq->thread[tid];
> >>>>>>> +	unsigned long flags = 0L;
> >>>>>>> +	u32 irq_flag;
> >>>>>>> +
> >>>>>>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
> >>>>>>> +
> >>>>>>> +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
> >>>>>>> +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
> >>>>>>> +
> >>>>>>> +	/*
> >>>>>>> +	 * Another CPU core could run "release task" right before we acquire
> >>>>>>> +	 * the spin lock, and thus reset / disable this GCE thread, so we
> >>>>>>> +	 * need to check the enable bit of this GCE thread.
> >>>>>>> +	 */
> >>>>>>> +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> >>>>>>> +	      CMDQ_THR_ENABLED))
> >>>>>>> +		irq_flag = 0;
> >>>>>>
> >>>>>> cmdq_handle_error_done just retuns in this case. Programming this way
> >>>>>> just makes things confusing. What about:
> >>>>>>
> >>>>>> if (cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> >>>>>> 	      CMDQ_THR_ENABLED)
> >>>>>> 	cmdq_handle_error_done(cmdq, thread, irq_flag);
> >>>>>> else
> >>>>>> 		irq_flag = 0;
> >>>>>>
> >>>>>> spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>>>>
> >>>>> We still need to clear irq_flag if GCE thread is disabled.
> >>>>> So, I think we can just return here.
> >>>>>
> >>>>> 	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> >>>>> 	      CMDQ_THR_ENABLED))
> >>>>> 		return;
> >>>>>
> >>>>> What do you think?
> >>>>>
> >>>>
> >>>> No, you can't just return, you need to unlock the spinlock.
> >>>> Anyway I would prefer it the other way round, as I put it in my last
> >>>> mail. Just delete the else branch, we don't need to set irq_flag to zero.
> >>>
> >>> Sorry for my previous wrong reply since I merge your comment
> >>> and CK's comment.
> >>> http://lists.infradead.org/pipermail/linux-mediatek/2016-May/005547.html
> >>> So, I will put this if condition into cmdq_task_handle_error_result()
> >>> and then just return it if GCE thread is disabled.
> >>>
> >>
> >> You mean in cmdq_task_handle_done
> >> We should rename this functions to not create confusion.
> >
> > Sorry again. I mean in cmdq_handle_error_done().
> > This function handles both done and error.
> >
> > I agree the function name looks confusion.
> > I think it can be renamed to cmdq_thread_irq_handler()
> > since it is used to handle irq for GCE thread.
> >
> >> Regards,
> >> Matthias
> >
> > Thanks,
> > HS
> >
> >>>>>>> +
> >>>>>>> +	cmdq_handle_error_done(cmdq, thread, irq_flag);
> >>>>>>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static irqreturn_t cmdq_irq_handler(int irq, void *dev)
> >>>>>>> +{
> >>>>>>> +	struct cmdq *cmdq = dev;
> >>>>>>> +	u32 irq_status;
> >>>>>>> +	int i;
> >>>>>>> +
> >>>>>>> +	irq_status = readl(cmdq->base + CMDQ_CURR_IRQ_STATUS);
> >>>>>>> +	irq_status &= CMDQ_IRQ_MASK;
> >>>>>>> +	irq_status ^= CMDQ_IRQ_MASK;
> >>>>>>
> >>>>>> irq_status can be much bigger then 3, which is the number of threads in
> >>>>>> the system (CMDQ_THR_MAX_COUNT). So why we use this mask here isn't
> >>>>>> clear to me.
> >>>>>
> >>>>> Our GCE hardware has 16 threads, but we only use 3 threads currently.
> >>>>>
> >>>>
> >>>> Ok, but please use bitops here.
> >>>
> >>> Will use bitops.
> >>>
> >>>>>>> +
> >>>>>>> +	if (!irq_status)
> >>>>>>> +		return IRQ_NONE;
> >>>>>>> +
> >>>>>>> +	while (irq_status) {
> >>>>>>> +		i = ffs(irq_status) - 1;
> >>>>>>> +		irq_status &= ~BIT(i);
> >>>>>>> +		cmdq_thread_irq_handler(cmdq, i);
> >>>>>>> +	}
> >>>>>>
> >>>>>> Can you explain how the irq status register looks like, that would it
> >>>>>> make much easier to understand what happens here.
> >>>>>
> >>>>> Bit 0 ~ 15 of irq status register represents GCE thread 0 ~ 15
> >>>>> interrupt. 0 means asserting interrupt; 1 means no interrupt.
> >>>>>
> >>>>
> >>>> Thanks, that helped. :)
> >>>>
> >>>>>>> +
> >>>>>>> +	return IRQ_HANDLED;	
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static int cmdq_task_handle_error_result(struct cmdq_task *task)
> >>>>>>
> >>>>>> We never check the return values, why do we have them?
> >>>>>
> >>>>> Will drop return value.
> >>>>>
> >>>>>>> +{
> >>>>>>> +	struct cmdq *cmdq = task->cmdq;
> >>>>>>> +	struct device *dev = cmdq->dev;
> >>>>>>> +	struct cmdq_thread *thread = task->thread;
> >>>>>>> +	struct cmdq_task *next_task, *prev_task;
> >>>>>>> +	u32 irq_flag;
> >>>>>>> +
> >>>>>>> +	/* suspend GCE thread to ensure consistency */
> >>>>>>> +	WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
> >>>>>>> +
> >>>>>>> +	/* ISR has handled this error task */
> >>>>>>> +	if (task->task_state == TASK_STATE_ERROR) {
> >>>>>>> +		next_task = list_first_entry_or_null(&thread->task_busy_list,
> >>>>>>> +				struct cmdq_task, list_entry);
> >>>>>>> +		if (next_task) /* move to next task */
> >>>>>>> +			cmdq_thread_writel(thread, next_task->pa_base,
> >>>>>>> +					   CMDQ_THR_CURR_ADDR);
> >>>>>>
> >>>>>> We have to do this, as we suppose that the thread did not reach the jump
> >>>>>> instruction we put into it's command queue, right?
> >>>>>
> >>>>> Yes.
> >>>>>
> >>>>
> >>>> So this should then go into it's own function. In wait_release_work,
> >>>> something like this:
> >>>>
> >>>> if(task->task_state == TASK_STATE_ERROR)
> >>>>         cmdq_task_handle_error(task)
> >>>
> >>> OK.
> >>> I will write new function cmdq_task_handle_error() to handle error case.
> >>>
> >>>>>>> +		cmdq_thread_resume(thread);
> >>>>>>> +		return -ECANCELED;
> >>>>>>> +	}
> >>>>>>> +
> >>>>>>
> >>>>>> if task_state != ERROR and != DONE. This means that the timeout of
> >>>>>> task_release_wq has timed out, right?
> >>>>>
> >>>>> Yes.
> >>>>>
> >>>>>>> +	/*
> >>>>>>> +	 * Save next_task and prev_task in advance
> >>>>>>> +	 * since cmdq_handle_error_done will remove list_entry.
> >>>>>>> +	 */
> >>>>>>> +	next_task = prev_task = NULL;
> >>>>>>> +	if (task->list_entry.next != &thread->task_busy_list)
> >>>>>>> +		next_task = list_next_entry(task, list_entry);
> >>>>>>> +	if (task->list_entry.prev != &thread->task_busy_list)
> >>>>>>> +		prev_task = list_prev_entry(task, list_entry);
> >>>>>>> +
> >>>>>>> +	/*
> >>>>>>> +	 * Although IRQ is disabled, GCE continues to execute.
> >>>>>>> +	 * It may have pending IRQ before GCE thread is suspended,
> >>>>>>> +	 * so check this condition again.
> >>>>>>> +	 */
> >>>>>>
> >>>>>> The first thing we did in this function was suspending the thread. Why
> >>>>>> do we need this then?
> >>>>>
> >>>>> Because timeout is CPU timeout not GCE timeout, GCE could just finish
> >>>>> this task before the GCE thread is suspended.
> >>>>>
> >>>>
> >>>> What are the reasons for a timeout? An error has happend, or the task is
> >>>> still executing.
> >>>
> >>>   From GCE's point of view, this task is still executing.
> >>> But, it could be an error of client.
> >>> For example, task may never get event if display turn off hardware
> >>> before waiting for task to finish its work.
> >>>
> >>>>>> To be honest this whole functions looks really like a design error. We
> >>>>>> have to sperate the states much clearer so that it is possible to
> >>>>>> understand what is happening in the GCE. Isn't it for example posible to
> >>>>>> have worker queues for timed out tasks and tasks with an error? I'm not
> >>>>>> sure how to do this, actually I'm not sure if I really understood how
> >>>>>> this is supposed to work.
> >>>>>
> >>>>> GCE doesn't have timeout. The timeout is decided and controlled by CPU,
> >>>>> so we check timeout in release work.
> >>>>> For error and done, they are easy to check by register, and we have
> >>>>> already created release work for timeout. So, I don't think we need to
> >>>>> create work queue for each case.
> >>>>>
> >>>>> What do you think?
> >>>>>
> >>>>
> >>>> I think, if we find in here, that the irq_flag is set, then the the
> >>>> interrupt handler was triggered and is spinning the spinlock. If this is
> >>>> not the case, we have a timeout and we handle this.
> >>>
> >>> I will write another function to handle error, and handle timeout here.
> >>>
> >>>>>> How much do we win, when we patch the thread command queue for every
> >>>>>> task we add, instead of just taking one task after another from the
> >>>>>> task_busy_list?
> >>>>>
> >>>>> GCE is used to help read/write registers with critical time limitation.
> >>>>> Sometimes, client may ask to process multiple tasks in a short period
> >>>>> of time, e.g. display flush multiple tasks for next vblank. So, CMDQ
> >>>>> shouldn't limit to process one task after another from the
> >>>>> task_busy_list. Currently, when interrupt or timeout, we will check
> >>>>> how many tasks are done, and which one is error or timeout.
> >>>>>
> >>>>
> >>>> So I suppose the device driver who use this are interested in throughput
> >>>> and not in latency. The callback of every
> >>>>
> >>>>>>> +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
> >>>>>>> +	cmdq_handle_error_done(cmdq, thread, irq_flag);
> >>>>>>> +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
> >>>>>>> +
> >>>>>>> +	if (task->task_state == TASK_STATE_DONE) {
> >>>>>>> +		cmdq_thread_resume(thread);
> >>>>>>> +		return 0;
> >>>>>>> +	}
> >>>>>>> +
> >>>>>>> +	if (task->task_state == TASK_STATE_ERROR) {
> >>>>>>> +		dev_err(dev, "task 0x%p error\n", task);
> >>>>>>> +		if (next_task) /* move to next task */
> >>>>>>> +			cmdq_thread_writel(thread, next_task->pa_base,
> >>>>>>> +					   CMDQ_THR_CURR_ADDR);
> >>>>>>> +		cmdq_thread_resume(thread);
> >>>>>>> +		return -ECANCELED;
> >>>>>>> +	}
> >>>>>>> +
> >>>>>>> +	/* Task is running, so we force to remove it. */
> >>>>>>> +	dev_err(dev, "task 0x%p timeout or killed\n", task);
> >>>>>>> +	task->task_state = TASK_STATE_ERROR;
> >>>>>>> +
> >>>>>>> +	if (prev_task) {
> >>>>>>> +		u64 *prev_va = prev_task->va_base;
> >>>>>>> +		u64 *curr_va = task->va_base;
> >>>>>>> +
> >>>>>>> +		/* copy JUMP instruction */
> >>>>>>> +		prev_va[prev_task->num_cmd - 1] = curr_va[task->num_cmd - 1];
> >>>>>>> +
> >>>>>>> +		cmdq_thread_invalidate_fetched_data(thread);
> >>>>>>> +	} else if (next_task) { /* move to next task */
> >>>>>>> +		cmdq_thread_writel(thread, next_task->pa_base,
> >>>>>>> +				   CMDQ_THR_CURR_ADDR);
> >>>>>>> +	}
> >>>>>>> +
> >>>>>>> +	list_del(&task->list_entry);
> >>>>>>> +	cmdq_thread_resume(thread);
> >>>>>>> +
> >>>>>>> +	/* call cb here to prevent lock */
> >>>>>>> +	if (task->cb.cb) {
> >>>>>>> +		struct cmdq_cb_data cmdq_cb_data;
> >>>>>>> +
> >>>>>>> +		cmdq_cb_data.err = true;
> >>>>>>> +		cmdq_cb_data.data = task->cb.data;
> >>>>>>> +		task->cb.cb(cmdq_cb_data);
> >>>>>>> +	}
> >>>>>>> +
> >>>>>>> +	return -ECANCELED;
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static void cmdq_task_wait_release_work(struct work_struct *work_item)
> >>>>>>> +{
> >>>>>>> +	struct cmdq_task *task = container_of(work_item, struct cmdq_task,
> >>>>>>> +					      release_work);
> >>>>>>> +	struct cmdq *cmdq = task->cmdq;
> >>>>>>> +	struct cmdq_thread *thread = task->thread;
> >>>>>>> +	unsigned long flags;
> >>>>>>> +
> >>>>>>> +	wait_event_timeout(thread->wait_task_done,
> >>>>>>> +			   task->task_state != TASK_STATE_BUSY,
> >>>>>>> +			   msecs_to_jiffies(CMDQ_TIMEOUT_MS));
> >>>>>>> +
> >>>>>>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
> >>>>>>> +	if (task->task_state != TASK_STATE_DONE)
> >>>>>>> +		cmdq_task_handle_error_result(task);
> >>>>>>> +	if (list_empty(&thread->task_busy_list))
> >>>>>>> +		cmdq_thread_disable(cmdq, thread);
> >>>>>>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>>>>>> +
> >>>>>>> +	/* release regardless of success or not */
> >>>>>>> +	clk_disable_unprepare(cmdq->clock);
> >>>>>>> +	cmdq_task_release(task);
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static void cmdq_task_wait_release_schedule(struct cmdq_task *task)
> >>>>>>> +{
> >>>>>>> +	struct cmdq *cmdq = task->cmdq;
> >>>>>>> +
> >>>>>>> +	INIT_WORK(&task->release_work, cmdq_task_wait_release_work);
> >>>>>>> +	queue_work(cmdq->task_release_wq, &task->release_work);
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static int cmdq_rec_realloc_cmd_buffer(struct cmdq_rec *rec, size_t size)
> >>>>>>> +{
> >>>>>>> +	void *new_buf;
> >>>>>>> +
> >>>>>>> +	new_buf = krealloc(rec->buf, size, GFP_KERNEL | __GFP_ZERO);
> >>>>>>> +	if (!new_buf)
> >>>>>>> +		return -ENOMEM;
> >>>>>>> +	rec->buf = new_buf;
> >>>>>>> +	rec->buf_size = size;
> >>>>>>> +	return 0;
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +struct cmdq_base *cmdq_register_device(struct device *dev)
> >>>>>>> +{
> >>>>>>> +	struct cmdq_base *cmdq_base;
> >>>>>>> +	struct resource res;
> >>>>>>> +	int subsys;
> >>>>>>> +	u32 base;
> >>>>>>> +
> >>>>>>> +	if (of_address_to_resource(dev->of_node, 0, &res))
> >>>>>>> +		return NULL;
> >>>>>>> +	base = (u32)res.start;
> >>>>>>> +
> >>>>>>> +	subsys = cmdq_subsys_base_to_id(base >> 16);
> >>>>>>> +	if (subsys < 0)
> >>>>>>> +		return NULL;
> >>>>>>> +
> >>>>>>> +	cmdq_base = devm_kmalloc(dev, sizeof(*cmdq_base), GFP_KERNEL);
> >>>>>>> +	if (!cmdq_base)
> >>>>>>> +		return NULL;
> >>>>>>> +	cmdq_base->subsys = subsys;
> >>>>>>> +	cmdq_base->base = base;
> >>>>>>> +
> >>>>>>> +	return cmdq_base;
> >>>>>>> +}
> >>>>>>> +EXPORT_SYMBOL(cmdq_register_device);
> >>>>>>> +
> >>>>>>> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
> >>>>>>> +		    struct cmdq_rec **rec_ptr)
> >>>>>>> +{
> >>>>>>> +	struct cmdq_rec *rec;
> >>>>>>> +	int err;
> >>>>>>> +
> >>>>>>> +	rec = kzalloc(sizeof(*rec), GFP_KERNEL);
> >>>>>>> +	if (!rec)
> >>>>>>> +		return -ENOMEM;
> >>>>>>> +	rec->cmdq = dev_get_drvdata(dev);
> >>>>>>> +	rec->engine_flag = engine_flag;
> >>>>>>> +	err = cmdq_rec_realloc_cmd_buffer(rec, CMDQ_INITIAL_CMD_BLOCK_SIZE);
> >>>>>>
> >>>>>> Just pass PAGE_SIZE here, no need for CMDQ_INITIAL_CMD_BLOCK_SIZE.
> >>>>>
> >>>>> Will do.
> >>>>>
> >>>>>>> +	if (err < 0) {
> >>>>>>> +		kfree(rec);
> >>>>>>> +		return err;
> >>>>>>> +	}
> >>>>>>> +	*rec_ptr = rec;
> >>>>>>> +	return 0;
> >>>>>>> +}
> >>>>>>> +EXPORT_SYMBOL(cmdq_rec_create);
> >>>>>>> +
> >>>>>>> +static int cmdq_rec_append_command(struct cmdq_rec *rec, enum cmdq_code code,
> >>>>>>> +				   u32 arg_a, u32 arg_b)
> >>>>>>> +{
> >>>>>>> +	u64 *cmd_ptr;
> >>>>>>> +	int err;
> >>>>>>> +
> >>>>>>> +	if (WARN_ON(rec->finalized))
> >>>>>>> +		return -EBUSY;
> >>>>>>> +	if (code < CMDQ_CODE_MOVE || code > CMDQ_CODE_EOC)
> >>>>>>> +		return -EINVAL;
> >>>>>>
> >>>>>> cmdq_rec_append_command is just called from inside this driver and code
> >>>>>> is a enum. We can expect it to be correct, no need for this check.
> >>>>>
> >>>>> Will drop this check.
> >>>>>
> >>>>>>> +	if (unlikely(rec->command_size + CMDQ_INST_SIZE > rec->buf_size)) {
> >>>>>>
> >>>>>> command_size is the offset into the buffer to which a new command is
> >>>>>> written, so this name is highly confusing. I wonder if this would be
> >>>>>> easier to understand if we redefine command_size to something like the
> >>>>>> number of commands and divide/multiply CMDQ_INST_SIZE where this is needed.
> >>>>>
> >>>>> I can rename command_size to cmd_buf_size and calculate num_cmd by
> >>>>> dividing CMDQ_INST_SIZE.
> >>>>> What do you think?
> >>>>>
> >>>>>>> +		err = cmdq_rec_realloc_cmd_buffer(rec, rec->buf_size * 2);
> >>>>>>> +		if (err < 0)
> >>>>>>> +			return err;
> >>>>>>> +	}
> >>>>>>> +	cmd_ptr = rec->buf + rec->command_size;
> >>>>>>> +	(*cmd_ptr) = (u64)((code << CMDQ_OP_CODE_SHIFT) | arg_a) << 32 | arg_b;
> >>>>>>> +	rec->command_size += CMDQ_INST_SIZE;
> >>>>>>> +	return 0;
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value, struct cmdq_base *base,
> >>>>>>> +		   u32 offset)
> >>>>>>> +{
> >>>>>>> +	u32 arg_a = ((base->base + offset) & CMDQ_ARG_A_WRITE_MASK) |
> >>>>>>> +		    ((base->subsys & CMDQ_SUBSYS_MASK) << CMDQ_SUBSYS_SHIFT);
> >>>>>>
> >>>>>> base->subsys is the id from gce_sybsys, so we can expect it to be
> >>>>>> correct, no need to mask with CMDQ_SUBSYS_MASK.
> >>>>>
> >>>>> Will drop it.
> >>>>>
> >>>>>>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WRITE, arg_a, value);
> >>>>>>> +}
> >>>>>>> +EXPORT_SYMBOL(cmdq_rec_write);
> >>>>>>> +
> >>>>>>> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
> >>>>>>> +			struct cmdq_base *base, u32 offset, u32 mask)
> >>>>>>> +{
> >>>>>>> +	u32 offset_mask = offset;
> >>>>>>> +	int err;
> >>>>>>> +
> >>>>>>> +	if (mask != 0xffffffff) {
> >>>>>>> +		err = cmdq_rec_append_command(rec, CMDQ_CODE_MOVE, 0, ~mask);
> >>>>>>> +		if (err < 0)
> >>>>>>> +			return err;
> >>>>>>> +		offset_mask |= CMDQ_WRITE_ENABLE_MASK;
> >>>>>>> +	}
> >>>>>>> +	return cmdq_rec_write(rec, value, base, offset_mask);
> >>>>>>> +}
> >>>>>>> +EXPORT_SYMBOL(cmdq_rec_write_mask);
> >>>>>>> +
> >>>>>>> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event)
> >>>>>>> +{
> >>>>>>> +	u32 arg_b;
> >>>>>>> +
> >>>>>>> +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
> >>>>>>> +		return -EINVAL;
> >>>>>>> +
> >>>>>>> +	/*
> >>>>>>> +	 * bit 0-11: wait value
> >>>>>>> +	 * bit 15: 1 - wait, 0 - no wait
> >>>>>>> +	 * bit 16-27: update value
> >>>>>>> +	 * bit 31: 1 - update, 0 - no update
> >>>>>>> +	 */
> >>>>>>
> >>>>>> I don't understand this comments. What are they for?
> >>>>>
> >>>>> This is for WFE command. I will comment it.
> >>>>>
> >>>>>>> +	arg_b = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
> >>>>>>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event, arg_b);
> >>>>>>> +}
> >>>>>>> +EXPORT_SYMBOL(cmdq_rec_wfe);
> >>>>>>> +
> >>>>>>> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event)
> >>>>>>> +{
> >>>>>>> +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
> >>>>>>> +		return -EINVAL;
> >>>>>>> +
> >>>>>>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event,
> >>>>>>> +				       CMDQ_WFE_UPDATE);
> >>>>>>> +}
> >>>>>>> +EXPORT_SYMBOL(cmdq_rec_clear_event);
> >>>>>>> +
> >>>>>>> +static int cmdq_rec_finalize(struct cmdq_rec *rec)
> >>>>>>> +{
> >>>>>>> +	int err;
> >>>>>>> +
> >>>>>>> +	if (rec->finalized)
> >>>>>>> +		return 0;
> >>>>>>> +
> >>>>>>> +	/* insert EOC and generate IRQ for each command iteration */
> >>>>>>> +	err = cmdq_rec_append_command(rec, CMDQ_CODE_EOC, 0, CMDQ_EOC_IRQ_EN);
> >>>>>>> +	if (err < 0)
> >>>>>>> +		return err;
> >>>>>>> +
> >>>>>>> +	/* JUMP to end */
> >>>>>>> +	err = cmdq_rec_append_command(rec, CMDQ_CODE_JUMP, 0, CMDQ_JUMP_PASS);
> >>>>>>> +	if (err < 0)
> >>>>>>> +		return err;
> >>>>>>> +
> >>>>>>
> >>>>>> Does this need to be atomic?
> >>>>>> What happens if after CODE_EOC and before CODE_JUMP some
> >>>>>> write/read/event gets added?
> >>>>>> What happens if more commands get added to the queue after CODE_JUMP,
> >>>>>> but before finalized is set to true. Why don't you use atomic functions
> >>>>>> to access finalized?
> >>>>>
> >>>>> Since cmdq_rec doesn't guarantee thread safe, mutex is needed when
> >>>>> client uses cmdq_rec.
> >>>>>
> >>>>
> >>>> Well I think that rec->finalized tries to implement this, but might
> >>>> fail, if two kernel threads work on the same cmdq_rec.
> >>>>
> >>>>>>> +	rec->finalized = true;
> >>>>>>> +	return 0;
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
> >>>>>>> +			 void *data)
> >>>>>>> +{
> >>>>>>> +	struct cmdq *cmdq = rec->cmdq;
> >>>>>>> +	struct cmdq_task *task;
> >>>>>>> +	struct cmdq_task_cb task_cb;
> >>>>>>> +	struct cmdq_thread *thread;
> >>>>>>> +	int err;
> >>>>>>> +
> >>>>>>> +	mutex_lock(&cmdq->task_mutex);
> >>>>>>> +	if (cmdq->suspended) {
> >>>>>>> +		dev_err(cmdq->dev, "%s is called after suspended\n", __func__);
> >>>>>>> +		mutex_unlock(&cmdq->task_mutex);
> >>>>>>> +		return -EPERM;
> >>>>>>> +	}
> >>>>>>> +
> >>>>>>> +	err = cmdq_rec_finalize(rec);
> >>>>>>> +	if (err < 0) {
> >>>>>>> +		mutex_unlock(&cmdq->task_mutex);
> >>>>>>> +		return err;
> >>>>>>> +	}
> >>>>>>> +
> >>>>>>> +	task_cb.cb = cb;
> >>>>>>> +	task_cb.data = data;
> >>>>>>> +	task = cmdq_task_acquire(rec, task_cb);
> >>>>>>> +	if (!task) {
> >>>>>>> +		mutex_unlock(&cmdq->task_mutex);
> >>>>>>> +		return -EFAULT;
> >>>>>>> +	}
> >>>>>>> +
> >>>>>>> +	thread = &cmdq->thread[cmdq_eng_get_thread(task->engine_flag)];
> >>>>>>> +	cmdq_task_exec(task, thread);
> >>>>>>> +	cmdq_task_wait_release_schedule(task);
> >>>>>>> +	mutex_unlock(&cmdq->task_mutex);
> >>>>>>> +	return 0;
> >>>>>>> +}
> >>>>>>> +EXPORT_SYMBOL(cmdq_rec_flush_async);
> >>>>>>> +
> >>>>>>> +struct cmdq_flush_completion {
> >>>>>>> +	struct completion cmplt;
> >>>>>>> +	bool err;
> >>>>>>> +};
> >>>>>>> +
> >>>>>>> +static int cmdq_rec_flush_cb(struct cmdq_cb_data data)
> >>>>>>> +{
> >>>>>>> +	struct cmdq_flush_completion *cmplt = data.data;
> >>>>>>> +
> >>>>>>> +	cmplt->err = data.err;
> >>>>>>> +	complete(&cmplt->cmplt);
> >>>>>>> +	return 0;
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +int cmdq_rec_flush(struct cmdq_rec *rec)
> >>>>>>> +{
> >>>>>>> +	struct cmdq_flush_completion cmplt;
> >>>>>>> +	int err;
> >>>>>>> +
> >>>>>>> +	init_completion(&cmplt.cmplt);
> >>>>>>> +	err = cmdq_rec_flush_async(rec, cmdq_rec_flush_cb, &cmplt);
> >>>>>>> +	if (err < 0)
> >>>>>>> +		return err;
> >>>>>>> +	wait_for_completion(&cmplt.cmplt);
> >>>>>>> +	return cmplt.err ? -EFAULT : 0;
> >>>>>>> +}
> >>>>>>> +EXPORT_SYMBOL(cmdq_rec_flush);
> >>>>>>> +
> >>>>>>> +void cmdq_rec_destroy(struct cmdq_rec *rec)
> >>>>>>> +{
> >>>>>>> +	kfree(rec->buf);
> >>>>>>> +	kfree(rec);
> >>>>>>> +}
> >>>>>>> +EXPORT_SYMBOL(cmdq_rec_destroy);
> >>>>>>> +
> >>>>>>> +static bool cmdq_task_is_empty(struct cmdq *cmdq)
> >>>>>>> +{
> >>>>>>> +	struct cmdq_thread *thread;
> >>>>>>> +	int i;
> >>>>>>> +
> >>>>>>> +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
> >>>>>>> +		thread = &cmdq->thread[i];
> >>>>>>> +		if (!list_empty(&thread->task_busy_list))
> >>>>>>> +			return false;
> >>>>>>> +	}
> >>>>>>> +	return true;
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static int cmdq_suspend(struct device *dev)
> >>>>>>> +{
> >>>>>>> +	struct cmdq *cmdq = dev_get_drvdata(dev);
> >>>>>>> +	u32 exec_threads;
> >>>>>>> +
> >>>>>>> +	mutex_lock(&cmdq->task_mutex);
> >>>>>>> +	cmdq->suspended = true;
> >>>>>>> +	mutex_unlock(&cmdq->task_mutex);
> >>>>>>> +
> >>>>>>> +	exec_threads = readl(cmdq->base + CMDQ_CURR_LOADED_THR);
> >>>>>>> +	if ((exec_threads & CMDQ_THR_EXECUTING) && !cmdq_task_is_empty(cmdq)) {
> >>>>>>> +		dev_err(dev, "wait active tasks timeout.\n");
> >>>>>>> +		flush_workqueue(cmdq->task_release_wq);
> >>>>>>> +	}
> >>>>>>> +	return 0;
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static int cmdq_resume(struct device *dev)
> >>>>>>> +{
> >>>>>>> +	struct cmdq *cmdq = dev_get_drvdata(dev);
> >>>>>>> +
> >>>>>>> +	cmdq->suspended = false;
> >>>>>>> +	return 0;
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static int cmdq_remove(struct platform_device *pdev)
> >>>>>>> +{
> >>>>>>> +	struct cmdq *cmdq = platform_get_drvdata(pdev);
> >>>>>>> +
> >>>>>>> +	destroy_workqueue(cmdq->task_release_wq);
> >>>>>>> +	cmdq->task_release_wq = NULL;
> >>>>>>> +	return 0;
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static int cmdq_probe(struct platform_device *pdev)
> >>>>>>> +{
> >>>>>>> +	struct device *dev = &pdev->dev;
> >>>>>>> +	struct device_node *node = dev->of_node;
> >>>>>>> +	struct resource *res;
> >>>>>>> +	struct cmdq *cmdq;
> >>>>>>> +	int err, i;
> >>>>>>> +
> >>>>>>> +	cmdq = devm_kzalloc(dev, sizeof(*cmdq), GFP_KERNEL);
> >>>>>>> +	if (!cmdq)
> >>>>>>> +		return -ENOMEM;
> >>>>>>> +	cmdq->dev = dev;
> >>>>>>> +
> >>>>>>> +	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> >>>>>>> +	cmdq->base = devm_ioremap_resource(dev, res);
> >>>>>>> +	if (IS_ERR(cmdq->base)) {
> >>>>>>> +		dev_err(dev, "failed to ioremap gce\n");
> >>>>>>> +		return PTR_ERR(cmdq->base);
> >>>>>>> +	}
> >>>>>>> +
> >>>>>>> +	cmdq->irq = irq_of_parse_and_map(node, 0);
> >>>>>>> +	if (!cmdq->irq) {
> >>>>>>> +		dev_err(dev, "failed to get irq\n");
> >>>>>>> +		return -EINVAL;
> >>>>>>> +	}
> >>>>>>> +
> >>>>>>> +	dev_dbg(dev, "cmdq device: addr:0x%p, va:0x%p, irq:%d\n",
> >>>>>>> +		dev, cmdq->base, cmdq->irq);
> >>>>>>> +
> >>>>>>> +	mutex_init(&cmdq->task_mutex);
> >>>>>>> +	spin_lock_init(&cmdq->exec_lock);
> >>>>>>> +	cmdq->task_release_wq = alloc_ordered_workqueue(
> >>>>>>> +			"%s", WQ_MEM_RECLAIM | WQ_HIGHPRI,
> >>>>>>> +			"cmdq_task_wait_release");
> >>>>>>> +
> >>>>>>> +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
> >>>>>>> +		cmdq->thread[i].base = cmdq->base + CMDQ_THR_BASE +
> >>>>>>> +				CMDQ_THR_SHIFT * i;
> >>>>>>> +		init_waitqueue_head(&cmdq->thread[i].wait_task_done);
> >>>>>>> +		INIT_LIST_HEAD(&cmdq->thread[i].task_busy_list);
> >>>>>>> +	}
> >>>>>>> +
> >>>>>>> +	platform_set_drvdata(pdev, cmdq);
> >>>>>>> +
> >>>>>>> +	err = devm_request_irq(dev, cmdq->irq, cmdq_irq_handler, IRQF_SHARED,
> >>>>>>> +			       CMDQ_DRIVER_DEVICE_NAME, cmdq);
> >>>>>>> +	if (err < 0) {
> >>>>>>> +		dev_err(dev, "failed to register ISR (%d)\n", err);
> >>>>>>> +		goto fail;
> >>>>>>> +	}
> >>>>>>> +
> >>>>>>> +	cmdq->clock = devm_clk_get(dev, CMDQ_CLK_NAME);
> >>>>>>> +	if (IS_ERR(cmdq->clock)) {
> >>>>>>> +		dev_err(dev, "failed to get clk:%s\n", CMDQ_CLK_NAME);
> >>>>>>> +		err = PTR_ERR(cmdq->clock);
> >>>>>>> +		goto fail;
> >>>>>>> +	}
> >>>>>>> +	return 0;
> >>>>>>> +
> >>>>>>> +fail:
> >>>>>>> +	cmdq_remove(pdev);
> >>>>>>> +	return err;
> >>>>>>> +}
> >>>>>>> +
> >>>>>>> +static const struct dev_pm_ops cmdq_pm_ops = {
> >>>>>>> +	.suspend = cmdq_suspend,
> >>>>>>> +	.resume = cmdq_resume,
> >>>>>>> +};
> >>>>>>> +
> >>>>>>> +static const struct of_device_id cmdq_of_ids[] = {
> >>>>>>> +	{.compatible = "mediatek,mt8173-gce",},
> >>>>>>> +	{}
> >>>>>>> +};
> >>>>>>> +
> >>>>>>> +static struct platform_driver cmdq_drv = {
> >>>>>>> +	.probe = cmdq_probe,
> >>>>>>> +	.remove = cmdq_remove,
> >>>>>>> +	.driver = {
> >>>>>>> +		.name = CMDQ_DRIVER_DEVICE_NAME,
> >>>>>>> +		.owner = THIS_MODULE,
> >>>>>>> +		.pm = &cmdq_pm_ops,
> >>>>>>> +		.of_match_table = cmdq_of_ids,
> >>>>>>> +	}
> >>>>>>> +};
> >>>>>>> +
> >>>>>>> +builtin_platform_driver(cmdq_drv);
> >>>>>>> diff --git a/include/soc/mediatek/cmdq.h b/include/soc/mediatek/cmdq.h
> >>>>>>> new file mode 100644
> >>>>>>> index 0000000..60eef3d
> >>>>>>> --- /dev/null
> >>>>>>> +++ b/include/soc/mediatek/cmdq.h
> >>>>>>> @@ -0,0 +1,197 @@
> >>>>>>> +/*
> >>>>>>> + * Copyright (c) 2015 MediaTek Inc.
> >>>>>>> + *
> >>>>>>> + * This program is free software; you can redistribute it and/or modify
> >>>>>>> + * it under the terms of the GNU General Public License version 2 as
> >>>>>>> + * published by the Free Software Foundation.
> >>>>>>> + *
> >>>>>>> + * This program is distributed in the hope that it will be useful,
> >>>>>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> >>>>>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> >>>>>>> + * GNU General Public License for more details.
> >>>>>>> + */
> >>>>>>> +
> >>>>>>> +#ifndef __MTK_CMDQ_H__
> >>>>>>> +#define __MTK_CMDQ_H__
> >>>>>>> +
> >>>>>>> +#include <linux/platform_device.h>
> >>>>>>> +#include <linux/types.h>
> >>>>>>> +
> >>>>>>> +enum cmdq_eng {
> >>>>>>> +	CMDQ_ENG_DISP_AAL,
> >>>>>>> +	CMDQ_ENG_DISP_COLOR0,
> >>>>>>> +	CMDQ_ENG_DISP_COLOR1,
> >>>>>>> +	CMDQ_ENG_DISP_DPI0,
> >>>>>>> +	CMDQ_ENG_DISP_DSI0,
> >>>>>>> +	CMDQ_ENG_DISP_DSI1,
> >>>>>>> +	CMDQ_ENG_DISP_GAMMA,
> >>>>>>> +	CMDQ_ENG_DISP_OD,
> >>>>>>> +	CMDQ_ENG_DISP_OVL0,
> >>>>>>> +	CMDQ_ENG_DISP_OVL1,
> >>>>>>> +	CMDQ_ENG_DISP_PWM0,
> >>>>>>> +	CMDQ_ENG_DISP_PWM1,
> >>>>>>> +	CMDQ_ENG_DISP_RDMA0,
> >>>>>>> +	CMDQ_ENG_DISP_RDMA1,
> >>>>>>> +	CMDQ_ENG_DISP_RDMA2,
> >>>>>>> +	CMDQ_ENG_DISP_UFOE,
> >>>>>>> +	CMDQ_ENG_DISP_WDMA0,
> >>>>>>> +	CMDQ_ENG_DISP_WDMA1,
> >>>>>>> +	CMDQ_ENG_MAX,
> >>>>>>> +};
> >>>>>>> +
> >>>>>>> +/* events for CMDQ and display */
> >>>>>>> +enum cmdq_event {
> >>>>>>> +	/* Display start of frame(SOF) events */
> >>>>>>> +	CMDQ_EVENT_DISP_OVL0_SOF = 11,
> >>>>>>> +	CMDQ_EVENT_DISP_OVL1_SOF = 12,
> >>>>>>> +	CMDQ_EVENT_DISP_RDMA0_SOF = 13,
> >>>>>>> +	CMDQ_EVENT_DISP_RDMA1_SOF = 14,
> >>>>>>> +	CMDQ_EVENT_DISP_RDMA2_SOF = 15,
> >>>>>>> +	CMDQ_EVENT_DISP_WDMA0_SOF = 16,
> >>>>>>> +	CMDQ_EVENT_DISP_WDMA1_SOF = 17,
> >>>>>>> +	/* Display end of frame(EOF) events */
> >>>>>>> +	CMDQ_EVENT_DISP_OVL0_EOF = 39,
> >>>>>>> +	CMDQ_EVENT_DISP_OVL1_EOF = 40,
> >>>>>>> +	CMDQ_EVENT_DISP_RDMA0_EOF = 41,
> >>>>>>> +	CMDQ_EVENT_DISP_RDMA1_EOF = 42,
> >>>>>>> +	CMDQ_EVENT_DISP_RDMA2_EOF = 43,
> >>>>>>> +	CMDQ_EVENT_DISP_WDMA0_EOF = 44,
> >>>>>>> +	CMDQ_EVENT_DISP_WDMA1_EOF = 45,
> >>>>>>> +	/* Mutex end of frame(EOF) events */
> >>>>>>> +	CMDQ_EVENT_MUTEX0_STREAM_EOF = 53,
> >>>>>>> +	CMDQ_EVENT_MUTEX1_STREAM_EOF = 54,
> >>>>>>> +	CMDQ_EVENT_MUTEX2_STREAM_EOF = 55,
> >>>>>>> +	CMDQ_EVENT_MUTEX3_STREAM_EOF = 56,
> >>>>>>> +	CMDQ_EVENT_MUTEX4_STREAM_EOF = 57,
> >>>>>>> +	/* Display underrun events */
> >>>>>>> +	CMDQ_EVENT_DISP_RDMA0_UNDERRUN = 63,
> >>>>>>> +	CMDQ_EVENT_DISP_RDMA1_UNDERRUN = 64,
> >>>>>>> +	CMDQ_EVENT_DISP_RDMA2_UNDERRUN = 65,
> >>>>>>> +	/* Keep this at the end of HW events */
> >>>>>>> +	CMDQ_MAX_HW_EVENT_COUNT = 260,
> >>>>>>> +};
> >>>>>>> +
> >>>>>>> +struct cmdq_cb_data {
> >>>>>>> +	bool	err;
> >>>>>>> +	void	*data;
> >>>>>>> +};
> >>>>>>> +
> >>>>>>> +typedef int (*cmdq_async_flush_cb)(struct cmdq_cb_data data);
> >>>>>>> +
> >>>>>>> +struct cmdq_task;
> >>>>>>> +struct cmdq;
> >>>>>>> +
> >>>>>>> +struct cmdq_rec {
> >>>>>>> +	struct cmdq		*cmdq;
> >>>>>>> +	u64			engine_flag;
> >>>>>>> +	size_t			command_size;
> >>>>>>> +	void			*buf;
> >>>>>>> +	size_t			buf_size;
> >>>>>>> +	bool			finalized;
> >>>>>>> +};
> >>>>
> >>>> Why do we need cmdq_rec at all? Can't we just use the cmdq_task for that
> >>>> and this way make the driver less complex?
> >>>
> >>> There are two main reasons for cmdq_rec.
> >>> 1. It is slow to access dma too frequently.
> >>>      So, we append commands to cacheable memory, and then flush to dma.
> >>> 2. cmdq_rec is not thread safe, but cmdq_task needs thread safe.
> >>>      If we merge them, we need to take care of some synchronization
> >>>      issues.
> >>>
> >>>>>>> +
> >>>>>>> +struct cmdq_base {
> >>>>>>> +	int	subsys;
> >>>>>>> +	u32	base;
> >>>>>>
> >>>>>> subsys can always be calculated via cmdq_subsys_base_to_id(base >> 16)
> >>>>>> so we can get rid of the struct, right?
> >>>>>
> >>>>> Current subsys method is based on previous comment from Daniel Kurtz.
> >>>>> Please take a look of our previous discussion.
> >>>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/004483.html
> >>>>> Thanks.
> >>>>>
> >>>>
> >>>> I have to look deeper into this, but from what I read, the proposal from
> >>>> Daniel [1] seems good to me.
> >>>>
> >>>> [1] https://patchwork.kernel.org/patch/8068311/
> >>>
> >>> The initial proposal has some problem, so please see the follow-up
> >>> discussions. Thanks.
> >>> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/003972.html
> >>> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/004483.html
> >>>
> >>>
> >>>>>>> +};
> >>>>>>> +
> >>>>>>> +/**
> >>>>>>> + * cmdq_register_device() - register device which needs CMDQ
> >>>>>>> + * @dev:		device
> >>>>>>> + *
> >>>>>>> + * Return: cmdq_base pointer or NULL for failed
> >>>>>>> + */
> >>>>>>> +struct cmdq_base *cmdq_register_device(struct device *dev);
> >>>>>>> +
> >>>>>>> +/**
> >>>>>>> + * cmdq_rec_create() - create command queue record
> >>>>>>> + * @dev:		device
> >>>>>>> + * @engine_flag:	command queue engine flag
> >>>>>>> + * @rec_ptr:		command queue record pointer to retrieve cmdq_rec
> >>>>>>> + *
> >>>>>>> + * Return: 0 for success; else the error code is returned
> >>>>>>> + */
> >>>>>>> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
> >>>>>>> +		    struct cmdq_rec **rec_ptr);
> >>>>>>> +
> >>>>>>> +/**
> >>>>>>> + * cmdq_rec_write() - append write command to the command queue record
> >>>>>>> + * @rec:	the command queue record
> >>>>>>> + * @value:	the specified target register value
> >>>>>>> + * @base:	the command queue base
> >>>>>>> + * @offset:	register offset from module base
> >>>>>>> + *
> >>>>>>> + * Return: 0 for success; else the error code is returned
> >>>>>>> + */
> >>>>>>> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value,
> >>>>>>> +		   struct cmdq_base *base, u32 offset);
> >>>>>>> +
> >>>>>>> +/**
> >>>>>>> + * cmdq_rec_write_mask() - append write command with mask to the command
> >>>>>>> + *			   queue record
> >>>>>>> + * @rec:	the command queue record
> >>>>>>> + * @value:	the specified target register value
> >>>>>>> + * @base:	the command queue base
> >>>>>>> + * @offset:	register offset from module base
> >>>>>>> + * @mask:	the specified target register mask
> >>>>>>> + *
> >>>>>>> + * Return: 0 for success; else the error code is returned
> >>>>>>> + */
> >>>>>>> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
> >>>>>>> +			struct cmdq_base *base, u32 offset, u32 mask);
> >>>>>>> +
> >>>>>>> +/**
> >>>>>>> + * cmdq_rec_wfe() - append wait for event command to the command queue reco	rd
> >>>>>>
> >>>>>> reco rd -> record
> >>>>>
> >>>>> Will fix it.
> >>>>>
> >>>>>> Regards,
> >>>>>> Matthias
> >>>>>>
> >>>>>>> + * @rec:	the command queue record
> >>>>>>> + * @event:	the desired event type to "wait and CLEAR"
> >>>>>>> + *
> >>>>>>> + * Return: 0 for success; else the error code is returned
> >>>>>>> + */
> >>>>>>> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event);
> >>>>>>> +
> >>>>>>> +/**
> >>>>>>> + * cmdq_rec_clear_event() - append clear event command to the command queue
> >>>>>>> + *			    record
> >>>>>>> + * @rec:	the command queue record
> >>>>>>> + * @event:	the desired event to be cleared
> >>>>>>> + *
> >>>>>>> + * Return: 0 for success; else the error code is returned
> >>>>>>> + */
> >>>>>>> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event);
> >>>>>>> +
> >>>>>>> +/**
> >>>>>>> + * cmdq_rec_flush() - trigger CMDQ to execute the recorded commands
> >>>>>>> + * @rec:	the command queue record
> >>>>>>> + *
> >>>>>>> + * Return: 0 for success; else the error code is returned
> >>>>>>> + *
> >>>>>>> + * Trigger CMDQ to execute the recorded commands. Note that this is a
> >>>>>>> + * synchronous flush function. When the function returned, the recorded
> >>>>>>> + * commands have been done.
> >>>>>>> + */
> >>>>>>> +int cmdq_rec_flush(struct cmdq_rec *rec);
> >>>>>>> +
> >>>>>>> +/**
> >>>>>>> + * cmdq_rec_flush_async() - trigger CMDQ to asynchronously execute the recorded
> >>>>>>> + *			    commands and call back after ISR is finished
> >>>>>>> + * @rec:	the command queue record
> >>>>>>> + * @cb:		called in the end of CMDQ ISR
> >>>>>>> + * @data:	this data will pass back to cb
> >>>>>>> + *
> >>>>>>> + * Return: 0 for success; else the error code is returned
> >>>>>>> + *
> >>>>>>> + * Trigger CMDQ to asynchronously execute the recorded commands and call back
> >>>>>>> + * after ISR is finished. Note that this is an ASYNC function. When the function
> >>>>>>> + * returned, it may or may not be finished. The ISR callback function is called
> >>>>>>> + * in the end of ISR.
> >>>>
> >>>> "The callback is called from the ISR."
> >>>>
> >>>> Regards,
> >>>> Matthias
> >>>>
> >>>>>>> + */
> >>>>>>> +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
> >>>>>>> +			 void *data);
> >>>>>>> +
> >>>>>>> +/**
> >>>>>>> + * cmdq_rec_destroy() - destroy command queue record
> >>>>>>> + * @rec:	the command queue record
> >>>>>>> + */
> >>>>>>> +void cmdq_rec_destroy(struct cmdq_rec *rec);
> >>>>>>> +
> >>>>>>> +#endif	/* __MTK_CMDQ_H__ */
> >>>>>>>
> >>>>>
> >>>>> Thanks,
> >>>>> HS
> >>>>>
> >>>
> >>> Thanks,
> >>> HS
> >>>
> >
> >

Matthias Brugger June 3, 2016, 1:11 p.m. UTC | #11

On 03/06/16 14:13, Horng-Shyang Liao wrote:
> Hi Matthias,
>
> On Fri, 2016-06-03 at 13:18 +0200, Matthias Brugger wrote:
>>
>> On 03/06/16 08:12, Horng-Shyang Liao wrote:
>>> Hi Mathias,
>>>
>>> Please see my inline reply.
>>>
>>> On Thu, 2016-06-02 at 10:46 +0200, Matthias Brugger wrote:
>>>>
>>>> On 01/06/16 11:57, Horng-Shyang Liao wrote:
>>>>> Hi Mathias,
>>>>>
>>>>> Please see my inline reply.
>>>>>
>>>>> On Tue, 2016-05-31 at 22:04 +0200, Matthias Brugger wrote:
>>>>>>
>>>>>> On 31/05/16 10:36, Horng-Shyang Liao wrote:
>>>>>>> Hi Mathias,
>>>>>>>
>>>>>>> Please see my inline reply.
>>>>>>>
>>>>>>> On Mon, 2016-05-30 at 17:31 +0200, Matthias Brugger wrote:
>>>>>>>>
>>>>>>>> On 30/05/16 05:19, HS Liao wrote:
>>>>>>>>> This patch is first version of Mediatek Command Queue(CMDQ) driver. The
>>>>>>>>> CMDQ is used to help read/write registers with critical time limitation,
>>>>>>>>> such as updating display configuration during the vblank. It controls
>>>>>>>>> Global Command Engine (GCE) hardware to achieve this requirement.
>>>>>>>>> Currently, CMDQ only supports display related hardwares, but we expect
>>>>>>>>> it can be extended to other hardwares for future requirements.
>>>>>>>>>
>>>>>>>>> Signed-off-by: HS Liao <[e-mail]>
>>>>>>>>> Signed-off-by: CK Hu <[e-mail]>
>>>>>>>>> ---
>>>>>>>>>       drivers/soc/mediatek/Kconfig    |  10 +
>>>>>>>>>       drivers/soc/mediatek/Makefile   |   1 +
>>>>>>>>>       drivers/soc/mediatek/mtk-cmdq.c | 943 ++++++++++++++++++++++++++++++++++++++++
>>>>>>>>>       include/soc/mediatek/cmdq.h     | 197 +++++++++
>>>>>>>>>       4 files changed, 1151 insertions(+)
>>>>>>>>>       create mode 100644 drivers/soc/mediatek/mtk-cmdq.c
>>>>>>>>>       create mode 100644 include/soc/mediatek/cmdq.h
>>>>>>>>>
>>>>>>>>> diff --git a/drivers/soc/mediatek/Kconfig b/drivers/soc/mediatek/Kconfig
>>>>>>>>> index 0a4ea80..c4ad75c 100644
>>>>>>>>> --- a/drivers/soc/mediatek/Kconfig
>>>>>>>>> +++ b/drivers/soc/mediatek/Kconfig
>>>>>>>>> @@ -1,6 +1,16 @@
>>>>>>>>>       #
>>>>>>>>>       # MediaTek SoC drivers
>>>>>>>>>       #
>>>>>>>>> +config MTK_CMDQ
>>>>>>>>> +	bool "MediaTek CMDQ Support"
>>>>>>>>> +	depends on ARCH_MEDIATEK || COMPILE_TEST
>>>>>>
>>>>>> depends on ARM64 ?
>>>>>
>>>>> Will add ARM64.
>>>>>
>>>>>>>>> +	select MTK_INFRACFG
>>>>>>>>> +	help
>>>>>>>>> +	  Say yes here to add support for the MediaTek Command Queue (CMDQ)
>>>>>>>>> +	  driver. The CMDQ is used to help read/write registers with critical
>>>>>>>>> +	  time limitation, such as updating display configuration during the
>>>>>>>>> +	  vblank.
>>>>>>>>> +
>>>>>>>>>       config MTK_INFRACFG
>>>>>>>>>       	bool "MediaTek INFRACFG Support"
>>>>>>>>>       	depends on ARCH_MEDIATEK || COMPILE_TEST
>>>>>>>>> diff --git a/drivers/soc/mediatek/Makefile b/drivers/soc/mediatek/Makefile
>>>>>>>>> index 12998b0..f7397ef 100644
>>>>>>>>> --- a/drivers/soc/mediatek/Makefile
>>>>>>>>> +++ b/drivers/soc/mediatek/Makefile
>>>>>>>>> @@ -1,3 +1,4 @@
>>>>>>>>> +obj-$(CONFIG_MTK_CMDQ) += mtk-cmdq.o
>>>>>>>>>       obj-$(CONFIG_MTK_INFRACFG) += mtk-infracfg.o
>>>>>>>>>       obj-$(CONFIG_MTK_PMIC_WRAP) += mtk-pmic-wrap.o
>>>>>>>>>       obj-$(CONFIG_MTK_SCPSYS) += mtk-scpsys.o
>>>>>>>>> diff --git a/drivers/soc/mediatek/mtk-cmdq.c b/drivers/soc/mediatek/mtk-cmdq.c
>>>>>>>>> new file mode 100644
>>>>>>>>> index 0000000..e9d6e1c
>>>>>>>>> --- /dev/null
>>>>>>>>> +++ b/drivers/soc/mediatek/mtk-cmdq.c
>>>>>>>>> @@ -0,0 +1,943 @@
>>>>>>>>> +/*
>>>>>>>>> + * Copyright (c) 2015 MediaTek Inc.
>>>>>>>>> + *
>>>>>>>>> + * This program is free software; you can redistribute it and/or modify
>>>>>>>>> + * it under the terms of the GNU General Public License version 2 as
>>>>>>>>> + * published by the Free Software Foundation.
>>>>>>>>> + *
>>>>>>>>> + * This program is distributed in the hope that it will be useful,
>>>>>>>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>>>>>>>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
>>>>>>>>> + * GNU General Public License for more details.
>>>>>>>>> + */
>>>>>>>>> +
>>>>>>>>> +#include <linux/clk.h>
>>>>>>>>> +#include <linux/clk-provider.h>
>>>>>>>>> +#include <linux/completion.h>
>>>>>>>>> +#include <linux/dma-mapping.h>
>>>>>>>>> +#include <linux/errno.h>
>>>>>>>>> +#include <linux/interrupt.h>
>>>>>>>>> +#include <linux/iopoll.h>
>>>>>>>>> +#include <linux/kernel.h>
>>>>>>>>> +#include <linux/kthread.h>
>>>>>>>>> +#include <linux/module.h>
>>>>>>>>> +#include <linux/mutex.h>
>>>>>>>>> +#include <linux/of_address.h>
>>>>>>>>> +#include <linux/of_irq.h>
>>>>>>>>> +#include <linux/platform_device.h>
>>>>>>>>> +#include <linux/slab.h>
>>>>>>>>> +#include <linux/spinlock.h>
>>>>>>>>> +#include <linux/suspend.h>
>>>>>>>>> +#include <linux/workqueue.h>
>>>>>>>>> +#include <soc/mediatek/cmdq.h>
>>>>>>>>> +
>>>>>>>>> +#define CMDQ_INITIAL_CMD_BLOCK_SIZE	PAGE_SIZE
>>>>>>>>> +#define CMDQ_INST_SIZE			8 /* instruction is 64-bit */
>>>>>>>>> +#define CMDQ_TIMEOUT_MS			1000
>>>>>>>>> +#define CMDQ_IRQ_MASK			0xffff
>>>>>>>>> +#define CMDQ_DRIVER_DEVICE_NAME		"mtk_cmdq"
>>>>>>>>> +#define CMDQ_CLK_NAME			"gce"
>>>>>>>>
>>>>>>>> We can put the names in directly to un-bloat the defines.
>>>>>>>
>>>>>>> I will use the names directly and remove defines.
>>>>>>>
>>>>>>>>> +
>>>>>>>>> +#define CMDQ_CURR_IRQ_STATUS		0x010
>>>>>>>>> +#define CMDQ_CURR_LOADED_THR		0x018
>>>>>>>>> +#define CMDQ_THR_SLOT_CYCLES		0x030
>>>>>>>>> +
>>>>>>>>> +#define CMDQ_THR_BASE			0x100
>>>>>>>>> +#define CMDQ_THR_SHIFT			0x080
>>>>>>>>
>>>>>>>> Wouldn't be CMDQ_THR_SIZE more accurate?
>>>>>>>
>>>>>>> Will rename it.
>>>>>>>
>>>>>>>>> +#define CMDQ_THR_WARM_RESET		0x00
>>>>>>>>> +#define CMDQ_THR_ENABLE_TASK		0x04
>>>>>>>>> +#define CMDQ_THR_SUSPEND_TASK		0x08
>>>>>>>>> +#define CMDQ_THR_CURR_STATUS		0x0c
>>>>>>>>> +#define CMDQ_THR_IRQ_STATUS		0x10
>>>>>>>>> +#define CMDQ_THR_IRQ_ENABLE		0x14
>>>>>>>>> +#define CMDQ_THR_CURR_ADDR		0x20
>>>>>>>>> +#define CMDQ_THR_END_ADDR		0x24
>>>>>>>>> +#define CMDQ_THR_CFG			0x40
>>>>>>>>> +
>>>>>>>>> +#define CMDQ_THR_ENABLED		0x1
>>>>>>>>> +#define CMDQ_THR_DISABLED		0x0
>>>>>>>>> +#define CMDQ_THR_SUSPEND		0x1
>>>>>>>>> +#define CMDQ_THR_RESUME			0x0
>>>>>>>>> +#define CMDQ_THR_STATUS_SUSPENDED	BIT(1)
>>>>>>>>> +#define CMDQ_THR_DO_WARM_RESET		BIT(0)
>>>>>>>>> +#define CMDQ_THR_ACTIVE_SLOT_CYCLES	0x3200
>>>>>>>>> +#define CMDQ_THR_PRIORITY		3
>>>>>>>>> +#define CMDQ_THR_IRQ_DONE		0x1
>>>>>>>>> +#define CMDQ_THR_IRQ_ERROR		0x12
>>>>>>>>> +#define CMDQ_THR_IRQ_EN			0x13 /* done + error */
>>>>>>>>
>>>>>>>> #define CMDQ_THR_IRQ_EN (CMDQ_THR_IRQ_ERROR | CMDQ_THR_IRQ_DONE)
>>>>>>>
>>>>>>> Will do.
>>>>>>>
>>>>>>>>> +#define CMDQ_THR_IRQ_MASK		0x13
>>>>>>>>
>>>>>>>> never used.
>>>>>>>
>>>>>>> Will remove.
>>>>>>>
>>>>>>>>> +#define CMDQ_THR_EXECUTING		BIT(31)
>>>>>>>>> +
>>>>>>>>> +#define CMDQ_ARG_A_WRITE_MASK		0xffff
>>>>>>>>> +#define CMDQ_SUBSYS_MASK		0x1f
>>>>>>>>> +#define CMDQ_OP_CODE_MASK		0xff000000
>>>>>>>>> +
>>>>>>>>> +#define CMDQ_OP_CODE_SHIFT		24
>>>>>>>>
>>>>>>>> Couldn't we connect the mask with the shift, or aren't they related?
>>>>>>>>
>>>>>>>> #define CMDQ_OP_CODE_MASK (0xff << CMDQ_OP_CODE_SHIFT)
>>>>>>>
>>>>>>> Will do.
>>>>>>>
>>>>>>>>> +#define CMDQ_SUBSYS_SHIFT		16
>>>>>>>>> +
>>>>>>>>> +#define CMDQ_WRITE_ENABLE_MASK		BIT(0)
>>>>>>>>> +#define CMDQ_JUMP_BY_OFFSET		0x10000000
>>>>>>>>> +#define CMDQ_JUMP_BY_PA			0x10000001
>>>>>>>>> +#define CMDQ_JUMP_PASS			CMDQ_INST_SIZE
>>>>>>>>> +#define CMDQ_WFE_UPDATE			BIT(31)
>>>>>>>>> +#define CMDQ_WFE_WAIT			BIT(15)
>>>>>>>>> +#define CMDQ_WFE_WAIT_VALUE		0x1
>>>>>>>>> +#define CMDQ_EOC_IRQ_EN			BIT(0)
>>>>>>>>> +
>>>>>>>>> +enum cmdq_thread_index {
>>>>>>>>> +	CMDQ_THR_DISP_MAIN_IDX,	/* main */
>>>>>>>>> +	CMDQ_THR_DISP_SUB_IDX,	/* sub */
>>>>>>>>> +	CMDQ_THR_DISP_MISC_IDX,	/* misc */
>>>>>>>>> +	CMDQ_THR_MAX_COUNT,	/* max */
>>>>>>>>> +};
>>>>>>>>> +
>>>>>>>>> +/*
>>>>>>>>> + * CMDQ_CODE_MOVE:
>>>>>>>>> + *   move value into internal register as mask
>>>>>>>>> + *   format: op mask
>>>>>>>>> + * CMDQ_CODE_WRITE:
>>>>>>>>> + *   write value into target register
>>>>>>>>> + *   format: op subsys address value
>>>>>>>>> + * CMDQ_CODE_JUMP:
>>>>>>>>> + *   jump by offset
>>>>>>>>> + *   format: op offset
>>>>>>>>> + * CMDQ_CODE_WFE:
>>>>>>>>> + *   wait for event and clear
>>>>>>>>> + *   it is just clear if no wait
>>>>>>>>> + *   format: [wait]  op event update:1 to_wait:1 wait:1
>>>>>>>>> + *           [clear] op event update:1 to_wait:0 wait:0
>>>>>>>>> + * CMDQ_CODE_EOC:
>>>>>>>>> + *   end of command
>>>>>>>>> + *   format: op irq_flag
>>>>>>>>> + */
>>>>>>>>
>>>>>>>> I think we need more documentation of how this command queue engine is
>>>>>>>> working. If not, I think it will be really complicated to understand how
>>>>>>>> to use this.
>>>>>>>>
>>>>>>>>> +enum cmdq_code {
>>>>>>>>> +	CMDQ_CODE_MOVE = 0x02,
>>>>>>>>> +	CMDQ_CODE_WRITE = 0x04,
>>>>>>>>> +	CMDQ_CODE_JUMP = 0x10,
>>>>>>>>> +	CMDQ_CODE_WFE = 0x20,
>>>>>>>>> +	CMDQ_CODE_EOC = 0x40,
>>>>>>>>> +};
>>>>>>>>> +
>>>>>>>>> +enum cmdq_task_state {
>>>>>>>>> +	TASK_STATE_BUSY, /* running on a GCE thread */
>>>>>>>>> +	TASK_STATE_ERROR,
>>>>>>>>> +	TASK_STATE_DONE,
>>>>>>>>> +};
>>>>>>>>> +
>>>>>>>>> +struct cmdq_task_cb {
>>>>>>>>> +	cmdq_async_flush_cb	cb;
>>>>>>>>> +	void			*data;
>>>>>>>>> +};
>>>>>>>>> +
>>>>>>>>> +struct cmdq_thread {
>>>>>>>>> +	void __iomem		*base;
>>>>>>>>> +	struct list_head	task_busy_list;
>>>>>>>>> +	wait_queue_head_t	wait_task_done;
>>>>>>>>> +};
>>>>>>>>> +
>>>>>>>>> +struct cmdq_task {
>>>>>>>>> +	struct cmdq		*cmdq;
>>>>>>>>> +	struct list_head	list_entry;
>>>>>>>>> +	enum cmdq_task_state	task_state;
>>>>>>>>> +	void			*va_base;
>>>>>>>>> +	dma_addr_t		pa_base;
>>>>>>>>> +	u64			engine_flag;
>>>>>>>>> +	size_t			command_size;
>>>>>>>>> +	u32			num_cmd;
>>>>>>>>
>>>>>>>> num_cmd is directly connected to command_size. I prefer to just keep
>>>>>>>> num_cmd and calculate the command_size where necessary.
>>>>>>>
>>>>>>> After I trace code, I prefer to keep command_size and calculate num_cmd
>>>>>>> where necessary. What do you think?
>>>>>>>
>>>>>>
>>>>>> I suppose you prefer this, as you are writing to the GCE depending on
>>>>>> the command_size. I think it is worth to create a macro for the
>>>>>> calculation of the number of commands, to make the code more readable.
>>>>>> Would be nice if you would just pass cmdq_task to it and it would return
>>>>>> the number. Just as an idea.
>>>>>
>>>>> Will add macro.
>>>>>
>>>>>>>>> +	struct cmdq_thread	*thread;
>>>>>>>>> +	struct cmdq_task_cb	cb;
>>>>>>>>> +	struct work_struct	release_work;
>>>>>>>>> +};
>>>>>>>>> +
>>>>>>>>> +struct cmdq {
>>>>>>>>> +	struct device		*dev;
>>>>>>>>> +	void __iomem		*base;
>>>>>>>>> +	u32			irq;
>>>>>>>>> +	struct workqueue_struct	*task_release_wq;
>>>>>>>>> +	struct cmdq_thread	thread[CMDQ_THR_MAX_COUNT];
>>>>>>>>> +	struct mutex		task_mutex;	/* for task */
>>>>>>>>> +	spinlock_t		exec_lock;	/* for exec */
>>>>>>>>> +	struct clk		*clock;
>>>>>>>>> +	bool			suspended;
>>>>>>>>> +};
>>>>>>>>> +
>>>>>>>>> +struct cmdq_subsys {
>>>>>>>>> +	u32	base;
>>>>>>>>> +	int	id;
>>>>>>>>> +};
>>>>>>>>> +
>>>>>>>>> +static const struct cmdq_subsys gce_subsys[] = {
>>>>>>>>> +	{0x1400, 1},
>>>>>>>>> +	{0x1401, 2},
>>>>>>>>> +	{0x1402, 3},
>>>>>>>>> +};
>>>>>>>>> +
>>>>>>>>> +static int cmdq_subsys_base_to_id(u32 base)
>>>>>>>>> +{
>>>>>>>>> +	int i;
>>>>>>>>> +
>>>>>>>>> +	for (i = 0; i < ARRAY_SIZE(gce_subsys); i++)
>>>>>>>>> +		if (gce_subsys[i].base == base)
>>>>>>>>> +			return gce_subsys[i].id;
>>>>>>>>> +	return -EFAULT;
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static int cmdq_eng_get_thread(u64 flag)
>>>>>>>>> +{
>>>>>>>>> +	if (flag & BIT_ULL(CMDQ_ENG_DISP_DSI0))
>>>>>>>>> +		return CMDQ_THR_DISP_MAIN_IDX;
>>>>>>>>> +	else if (flag & BIT_ULL(CMDQ_ENG_DISP_DPI0))
>>>>>>>>> +		return CMDQ_THR_DISP_SUB_IDX;
>>>>>>>>> +	else
>>>>>>>>> +		return CMDQ_THR_DISP_MISC_IDX;
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static void cmdq_task_release(struct cmdq_task *task)
>>>>>>>>> +{
>>>>>>>>> +	struct cmdq *cmdq = task->cmdq;
>>>>>>>>> +
>>>>>>>>> +	dma_free_coherent(cmdq->dev, task->command_size, task->va_base,
>>>>>>>>> +			  task->pa_base);
>>>>>>>>> +	kfree(task);
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static struct cmdq_task *cmdq_task_acquire(struct cmdq_rec *rec,
>>>>>>>>> +					   struct cmdq_task_cb cb)
>>>>>>>>> +{
>>>>>>>>> +	struct cmdq *cmdq = rec->cmdq;
>>>>>>>>> +	struct device *dev = cmdq->dev;
>>>>>>>>> +	struct cmdq_task *task;
>>>>>>>>> +
>>>>>>>>> +	task = kzalloc(sizeof(*task), GFP_KERNEL);
>>>>>>>>> +	INIT_LIST_HEAD(&task->list_entry);
>>>>>>>>> +	task->va_base = dma_alloc_coherent(dev, rec->command_size,
>>>>>>>>> +					   &task->pa_base, GFP_KERNEL);
>>>>>>>>> +	if (!task->va_base) {
>>>>>>>>> +		dev_err(dev, "allocate command buffer failed\n");
>>>>>>>>> +		kfree(task);
>>>>>>>>> +		return NULL;
>>>>>>>>> +	}
>>>>>>>>> +
>>>>>>>>> +	task->cmdq = cmdq;
>>>>>>>>> +	task->command_size = rec->command_size;
>>>>>>>>> +	task->engine_flag = rec->engine_flag;
>>>>>>>>> +	task->task_state = TASK_STATE_BUSY;
>>>>>>>>> +	task->cb = cb;
>>>>>>>>> +	memcpy(task->va_base, rec->buf, rec->command_size);
>>>>>>>>> +	task->num_cmd = task->command_size / CMDQ_INST_SIZE;
>>>>>>>>> +	return task;
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static void cmdq_thread_writel(struct cmdq_thread *thread, u32 value,
>>>>>>>>> +			       u32 offset)
>>>>>>>>> +{
>>>>>>>>> +	writel(value, thread->base + offset);
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static u32 cmdq_thread_readl(struct cmdq_thread *thread, u32 offset)
>>>>>>>>> +{
>>>>>>>>> +	return readl(thread->base + offset);
>>>>>>>>> +}
>>>>>>>>
>>>>>>>> We can get rid of cmdq_thread_readl/writel.
>>>>>>>
>>>>>>> Will do.
>>>>>>>
>>>>>>>>> +
>>>>>>>>> +static int cmdq_thread_suspend(struct cmdq *cmdq, struct cmdq_thread *thread)
>>>>>>>>> +{
>>>>>>>>> +	u32 status;
>>>>>>>>> +
>>>>>>>>> +	cmdq_thread_writel(thread, CMDQ_THR_SUSPEND, CMDQ_THR_SUSPEND_TASK);
>>>>>>>>> +
>>>>>>>>> +	/* If already disabled, treat as suspended successful. */
>>>>>>>>> +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
>>>>>>>>> +	      CMDQ_THR_ENABLED))
>>>>>>>>> +		return 0;
>>>>>>>>> +
>>>>>>>>> +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_CURR_STATUS,
>>>>>>>>> +			status, status & CMDQ_THR_STATUS_SUSPENDED, 0, 10)) {
>>>>>>>>> +		dev_err(cmdq->dev, "suspend GCE thread 0x%x failed\n",
>>>>>>>>> +			(u32)(thread->base - cmdq->base));
>>>>>>>>> +		return -EFAULT;
>>>>>>>>> +	}
>>>>>>>>> +
>>>>>>>>> +	return 0;
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static void cmdq_thread_resume(struct cmdq_thread *thread)
>>>>>>>>> +{
>>>>>>>>> +	cmdq_thread_writel(thread, CMDQ_THR_RESUME, CMDQ_THR_SUSPEND_TASK);
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static int cmdq_thread_reset(struct cmdq *cmdq, struct cmdq_thread *thread)
>>>>>>>>> +{
>>>>>>>>> +	u32 warm_reset;
>>>>>>>>> +
>>>>>>>>> +	cmdq_thread_writel(thread, CMDQ_THR_DO_WARM_RESET, CMDQ_THR_WARM_RESET);
>>>>>>>>> +	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_WARM_RESET,
>>>>>>>>> +			warm_reset, !(warm_reset & CMDQ_THR_DO_WARM_RESET),
>>>>>>>>> +			0, 10)) {
>>>>>>>>> +		dev_err(cmdq->dev, "reset GCE thread 0x%x failed\n",
>>>>>>>>> +			(u32)(thread->base - cmdq->base));
>>>>>>>>> +		return -EFAULT;
>>>>>>>>> +	}
>>>>>>>>> +	writel(CMDQ_THR_ACTIVE_SLOT_CYCLES, cmdq->base + CMDQ_THR_SLOT_CYCLES);
>>>>>>>>> +	return 0;
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static void cmdq_thread_disable(struct cmdq *cmdq, struct cmdq_thread *thread)
>>>>>>>>> +{
>>>>>>>>> +	cmdq_thread_reset(cmdq, thread);
>>>>>>>>> +	cmdq_thread_writel(thread, CMDQ_THR_DISABLED, CMDQ_THR_ENABLE_TASK);
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +/* notify GCE to re-fetch commands by setting GCE thread PC */
>>>>>>>>> +static void cmdq_thread_invalidate_fetched_data(struct cmdq_thread *thread)
>>>>>>>>> +{
>>>>>>>>> +	cmdq_thread_writel(thread,
>>>>>>>>> +			   cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR),
>>>>>>>>> +			   CMDQ_THR_CURR_ADDR);
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static void cmdq_task_insert_into_thread(struct cmdq_task *task)
>>>>>>>>> +{
>>>>>>>>> +	struct cmdq_thread *thread = task->thread;
>>>>>>>>> +	struct cmdq_task *prev_task = list_last_entry(
>>>>>>>>> +			&thread->task_busy_list, typeof(*task), list_entry);
>>>>>>>>> +	u64 *prev_task_base = prev_task->va_base;
>>>>>>>>> +
>>>>>>>>> +	/* let previous task jump to this task */
>>>>>>>>> +	prev_task_base[prev_task->num_cmd - 1] = (u64)CMDQ_JUMP_BY_PA << 32 |
>>>>>>>>> +						 task->pa_base;
>>>>>>>>> +
>>>>>>>>> +	cmdq_thread_invalidate_fetched_data(thread);
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +/* we assume tasks in the same display GCE thread are waiting the same event. */
>>>>>>>>> +static void cmdq_task_remove_wfe(struct cmdq_task *task)
>>>>>>>>> +{
>>>>>>>>> +	u32 wfe_option = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
>>>>>>>>> +	u32 wfe_op = CMDQ_CODE_WFE << CMDQ_OP_CODE_SHIFT;
>>>>>>>>> +	u32 *base = task->va_base;
>>>>>>>>> +	u32 num_cmd = task->num_cmd << 1;
>>>>>>>>> +	int i;
>>>>>>>>> +
>>>>>>>>> +	for (i = 0; i < num_cmd; i += 2)
>>>>>>>>> +		if (base[i] == wfe_option &&
>>>>>>>>> +		    (base[i + 1] & CMDQ_OP_CODE_MASK) == wfe_op) {
>>>>>>>>> +			base[i] = CMDQ_JUMP_PASS;
>>>>>>>>> +			base[i + 1] = CMDQ_JUMP_BY_OFFSET;
>>>>>>>>> +		}
>>>>>>>>
>>>>>>>> After using the command buffer as a void pointer a u64 pointer, we now
>>>>>>>> reference to it as u32. I would prefer to explain here, how the command
>>>>>>>> looks like we are searching for and use a for loop passing task->num_cmd
>>>>>>>> instead.
>>>>>>>
>>>>>>> Will use u64* to rewrite the above code.
>>>>>>>
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static void cmdq_task_exec(struct cmdq_task *task, struct cmdq_thread *thread)
>>>>>>>>> +{
>>>>>>>>> +	struct cmdq *cmdq = task->cmdq;
>>>>>>>>> +	unsigned long flags;
>>>>>>>>> +	unsigned long curr_pa, end_pa;
>>>>>>>>> +
>>>>>>>>> +	WARN_ON(clk_prepare_enable(cmdq->clock) < 0);
>>>>>>>>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
>>>>>>>>
>>>>>>>> cmdq_task_exec is called with cmdq->task_mutex held, so why do we need
>>>>>>>> the spin_lock here? Can't we just use one of the two?
>>>>>>>
>>>>>>> We can drop task_mutex, but we will get some side effects.
>>>>>>> 1. exec_lock needs to include more code, but I think it is not good for
>>>>>>>        spinlock.
>>>>>>> 2. In cmdq_rec_flush_async(), task_mutex needs to protect
>>>>>>>        (1) cmdq->suspended, (2) cmdq_task_exec(), and
>>>>>>>        (3) cmdq_task_wait_release_schedule().
>>>>>>>        If we drop task_mutex, we have to put cmdq->suspended if condition
>>>>>>>        just before cmdq_task_exec() and inside exec_lock, and we have to
>>>>>>>        release task and its command buffer if error. This will let flow
>>>>>>>        become more complex and enlarge code size.
>>>>>>>
>>>>>>> What do you think?
>>>>>>
>>>>>> Why do you need to protect cmdq_task_wait_release_schedule? We don't
>>>>>> care about the order of the workqueue elements, do we?
>>>>>
>>>>> According to CK's comment, we have to ensure order to remove previous
>>>>> task.
>>>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-May/005547.html
>>>>>
>>>>>> As far as I understand you would need to protect cmdq_task_acquire as
>>>>>> well, to "ensure" continously growing pa_base. More on that below.
>>>>>
>>>>> We need to ensure continuous physical addresses in a task, but we don't
>>>>> need to ensure continuous physical addresses between tasks.
>>>>> So, I think it is unnecessary to protect by mutex or spinlock.
>>>>>
>>>>
>>>> True, I didn't get that.
>>>>
>>>>>>>
>>>>>>>>> +	task->thread = thread;
>>>>>>>>> +	task->task_state = TASK_STATE_BUSY;
>>>>>>>>
>>>>>>>> That was already set in cmdq_task_acquire, why do we need to set it here
>>>>>>>> again?
>>>>>>>
>>>>>>> Will drop it.
>>>>>>>
>>>>>>
>>>>>> Yeah, but I think it makes more sense to drop it in cmdq_task_acquire
>>>>>> instead.
>>>>>
>>>>> Will drop it in cmdq_task_acquire instead.
>>>>>
>>>>>>>>> +	if (list_empty(&thread->task_busy_list)) {
>>>>>>>>> +		WARN_ON(cmdq_thread_reset(cmdq, thread) < 0);
>>>>>>>>> +
>>>>>>>>> +		cmdq_thread_writel(thread, task->pa_base, CMDQ_THR_CURR_ADDR);
>>>>>>>>> +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
>>>>>>>>> +				   CMDQ_THR_END_ADDR);
>>>>>>>>> +		cmdq_thread_writel(thread, CMDQ_THR_PRIORITY, CMDQ_THR_CFG);
>>>>>>>>> +		cmdq_thread_writel(thread, CMDQ_THR_IRQ_EN,
>>>>>>>>> +				   CMDQ_THR_IRQ_ENABLE);
>>>>>>>>> +
>>>>>>>>> +		cmdq_thread_writel(thread, CMDQ_THR_ENABLED,
>>>>>>>>> +				   CMDQ_THR_ENABLE_TASK);
>>>>>>>>> +	} else {
>>>>>>>>> +		WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
>>>>>>>>> +
>>>>>>>>> +		/*
>>>>>>>>> +		 * check boundary condition
>>>>>>>>> +		 * PC = END - 8, EOC is executed
>>>>>>>>> +		 * PC = END, all CMDs are executed
>>>>>>>>> +		 */
>>>>>>>>> +		curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
>>>>>>>>> +		end_pa = cmdq_thread_readl(thread, CMDQ_THR_END_ADDR);
>>>>>>>>> +		if (curr_pa == end_pa - 8 || curr_pa == end_pa) {
>>>>>>>>
>>>>>>>> 8 refers to CMDQ_INST_SIZE, right?
>>>>>>>
>>>>>>> Yes, I will use CMDQ_INST_SIZE.
>>>>>>>
>>>>>>>>> +			/* set to this task directly */
>>>>>>>>> +			cmdq_thread_writel(thread, task->pa_base,
>>>>>>>>> +					   CMDQ_THR_CURR_ADDR);
>>>>>>>>> +		} else {
>>>>>>>>> +			cmdq_task_insert_into_thread(task);
>>>>>>>>> +
>>>>>>>>> +			if (thread == &cmdq->thread[CMDQ_THR_DISP_MAIN_IDX] ||
>>>>>>>>> +			    thread == &cmdq->thread[CMDQ_THR_DISP_SUB_IDX])
>>>>>>>>> +				cmdq_task_remove_wfe(task);
>>>>>>>>
>>>>>>>> We could do this check using the task->engine_flag, I suppose that's
>>>>>>>> easier to undestand then.
>>>>>>>
>>>>>>> Will use task->engine_flag.
>>>>>>>
>>>>>>>>> +
>>>>>>>>> +			smp_mb(); /* modify jump before enable thread */
>>>>>>>>> +		}
>>>>>>>>> +
>>>>>>>>> +		cmdq_thread_writel(thread, task->pa_base + task->command_size,
>>>>>>>>> +				   CMDQ_THR_END_ADDR);
>>>>>>>>> +		cmdq_thread_resume(thread);
>>>>>>>>> +	}
>>>>>>>>> +	list_move_tail(&task->list_entry, &thread->task_busy_list);
>>>>>>>>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
>>>>>>>>> +				   struct cmdq_thread *thread, u32 irq_flag)
>>>>>>>>> +{
>>>>>>>>> +	struct cmdq_task *task, *tmp, *curr_task = NULL;
>>>>>>>>> +	u32 curr_pa;
>>>>>>>>> +	struct cmdq_cb_data cmdq_cb_data;
>>>>>>>>> +	bool err;
>>>>>>>>> +
>>>>>>>>> +	if (irq_flag & CMDQ_THR_IRQ_ERROR)
>>>>>>>>> +		err = true;
>>>>>>>>> +	else if (irq_flag & CMDQ_THR_IRQ_DONE)
>>>>>>>>> +		err = false;
>>>>>>>>> +	else
>>>>>>>>> +		return;
>>>>>>>>> +
>>>>>>>>> +	curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
>>>>>>>>> +
>>>>>>>>> +	list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
>>>>>>>>> +				 list_entry) {
>>>>>>>>> +		if (curr_pa >= task->pa_base &&
>>>>>>>>> +		    curr_pa < (task->pa_base + task->command_size))
>>>>>>>>
>>>>>>>> What are you checking here? It seems as if you make some implcit
>>>>>>>> assumptions about pa_base and the order of execution of commands in the
>>>>>>>> thread. Is it save to do so? Does dma_alloc_coherent give any guarantees
>>>>>>>> about dma_handle?
>>>>>>>
>>>>>>> 1. Check what is the current running task in this GCE thread.
>>>>>>> 2. Yes.
>>>>>>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
>>>>>>>
>>>>>>
>>>>>> Yes, physical addresses might be continous, but AFAIK there is no
>>>>>> guarantee that the dma_handle address is steadily growing, when calling
>>>>>> dma_alloc_coherent. And if I understand the code correctly, you use this
>>>>>> assumption to decide if the task picked from task_busy_list is currently
>>>>>> executing. So I think this mecanism is not working.
>>>>>
>>>>> I don't use dma_handle address, and just use physical addresses.
>>>>>    From CPU's point of view, tasks are linked by the busy list.
>>>>>    From GCE's point of view, tasks are linked by the JUMP command.
>>>>>
>>>>>> In which cases does the HW thread raise an interrupt.
>>>>>> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
>>>>>
>>>>> GCE will raise interrupt if any task is done or error.
>>>>> However, GCE is fast, so CPU may get multiple done tasks
>>>>> when it is running ISR.
>>>>>
>>>>> In case of error, that GCE thread will pause and raise interrupt.
>>>>> So, CPU may get multiple done tasks and one error task.
>>>>>
>>>>
>>>> I think we should reimplement the ISR mechanism. Can't we just read
>>>> CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave
>>>> cmdq_handle_error_done to the thread_fn? You will need to pass
>>>> information from the handler to thread_fn, but that shouldn't be an
>>>> issue. AFAIK interrupts are disabled in the handler, so we should stay
>>>> there as short as possible. Traversing task_busy_list is expensive, so
>>>> we need to do it in a thread context.
>>>
>>> Actually, our initial implementation is similar to your suggestion,
>>> but display needs CMDQ to return callback function very precisely,
>>> else display will drop frame.
>>> For display, CMDQ interrupt will be raised every 16 ~ 17 ms,
>>> and CMDQ needs to call callback function in ISR.
>>> If we defer callback to workqueue, the time interval may be larger than
>>> 32 ms.sometimes.
>>>
>>
>> I think the problem is, that you implemented the workqueue as a ordered
>> workqueue, so there is no parallel processing. I'm still not sure why
>> you need the workqueue to be ordered. Can you please explain.
>
> The order should be kept.
> Let me use mouse cursor as an example.
> If task 1 means move mouse cursor to point A, task 2 means point B,
> and task 3 means point C, our expected result is A -> B -> C.
> If the order is not kept, the result could become A -> C -> B.
>

Got it, thanks for the clarification.

Matthias

>>>> I keep thinking about how to get rid of the two data structures,
>>>> task_busy_list and the task_release_wq. We need the latter for the only
>>>> sake of getting a timeout.
>>>>
>>>> Did you have a look on how the mailbox framework handles this?
>>>> By the way, what is the reason to not implement the whole driver as a
>>>> mailbox controller? For me, this driver looks like a good fit.
>>>
>>> CMDQ needs to encode commands for GCE hardware. We think this behavior
>>> should be put in CMDQ driver, and client just call CMDQ functions.
>>> Therefore, if we want to use mailbox framework, cmdq_rec must be
>>> mailbox client, and the others must be mailbox controller.
>>>
>>
>> You mean the functions to fill the cmdq_rec and execute it?
>> I think this should be part of the driver.
>
> Yes.
>
>> Jassi, can you have a look on the interface this driver exports [0].
>> They are needed to actually create the message which will be send.
>> Could something like this be part of a mailbox driver?
>>
>> [0] https://patchwork.kernel.org/patch/9140221/
>>
>>> However, if we use mailbox controller, CMDQ driver still needs to
>>> control busy list for each GCE thread, and use workqueue to handle
>>> timeout tasks.
>>>
>>
>> Let me summarize my ideas around this driver:
>> When we enter the ISR, we know that all task in task_busy_list before
>> the entry which represents curr_task can be set to TASK_STATE_DONE.
>> The curr_task could be TASK_STATE_ERROR if the corresponding bit in the
>> irq status registers is set.
>> Do we need to call the callback in the same order as the tasks got
>> dispatched to the HW thread? If not, we could try to call all this
>> callbacks in a multithreaded workqueue.
>
> Yes, we should keep order.
>
>> Regards,
>> Matthias
>
> Thanks,
> HS
>
>>> The only thing that we can borrow from mailbox framework is the send
>>> (CMDQ flush) and receive (CMDQ callback) interface, However, we don't
>>> think we can gain many benefits from it, and we have some overheads to
>>> conform to mailbox interface.
>>>
>>>
>>>>
>>>>>>>>> +			curr_task = task;
>>>>>>>>> +		if (task->cb.cb) {
>>>>>>>>> +			cmdq_cb_data.err = curr_task ? err : false;
>>>>>>>>> +			cmdq_cb_data.data = task->cb.data;
>>>>>>>>> +			task->cb.cb(cmdq_cb_data);
>>>>>>>>> +		}
>>>>>>>>> +		task->task_state = (curr_task && err) ? TASK_STATE_ERROR :
>>>>>>>>> +				TASK_STATE_DONE;
>>>>>>>>> +		list_del(&task->list_entry);
>>>>>>>>> +		if (curr_task)
>>>>>>>>> +			break;
>>>>>>>>> +	}
>>>>>>>>> +
>>>>>>>>> +	wake_up(&thread->wait_task_done);
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static void cmdq_thread_irq_handler(struct cmdq *cmdq, int tid)
>>>>>>>>> +{
>>>>>>>>> +	struct cmdq_thread *thread = &cmdq->thread[tid];
>>>>>>>>> +	unsigned long flags = 0L;
>>>>>>>>> +	u32 irq_flag;
>>>>>>>>> +
>>>>>>>>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
>>>>>>>>> +
>>>>>>>>> +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
>>>>>>>>> +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
>>>>>>>>> +
>>>>>>>>> +	/*
>>>>>>>>> +	 * Another CPU core could run "release task" right before we acquire
>>>>>>>>> +	 * the spin lock, and thus reset / disable this GCE thread, so we
>>>>>>>>> +	 * need to check the enable bit of this GCE thread.
>>>>>>>>> +	 */
>>>>>>>>> +	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
>>>>>>>>> +	      CMDQ_THR_ENABLED))
>>>>>>>>> +		irq_flag = 0;
>>>>>>>>
>>>>>>>> cmdq_handle_error_done just retuns in this case. Programming this way
>>>>>>>> just makes things confusing. What about:
>>>>>>>>
>>>>>>>> if (cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
>>>>>>>> 	      CMDQ_THR_ENABLED)
>>>>>>>> 	cmdq_handle_error_done(cmdq, thread, irq_flag);
>>>>>>>> else
>>>>>>>> 		irq_flag = 0;
>>>>>>>>
>>>>>>>> spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>>>>
>>>>>>> We still need to clear irq_flag if GCE thread is disabled.
>>>>>>> So, I think we can just return here.
>>>>>>>
>>>>>>> 	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
>>>>>>> 	      CMDQ_THR_ENABLED))
>>>>>>> 		return;
>>>>>>>
>>>>>>> What do you think?
>>>>>>>
>>>>>>
>>>>>> No, you can't just return, you need to unlock the spinlock.
>>>>>> Anyway I would prefer it the other way round, as I put it in my last
>>>>>> mail. Just delete the else branch, we don't need to set irq_flag to zero.
>>>>>
>>>>> Sorry for my previous wrong reply since I merge your comment
>>>>> and CK's comment.
>>>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-May/005547.html
>>>>> So, I will put this if condition into cmdq_task_handle_error_result()
>>>>> and then just return it if GCE thread is disabled.
>>>>>
>>>>
>>>> You mean in cmdq_task_handle_done
>>>> We should rename this functions to not create confusion.
>>>
>>> Sorry again. I mean in cmdq_handle_error_done().
>>> This function handles both done and error.
>>>
>>> I agree the function name looks confusion.
>>> I think it can be renamed to cmdq_thread_irq_handler()
>>> since it is used to handle irq for GCE thread.
>>>
>>>> Regards,
>>>> Matthias
>>>
>>> Thanks,
>>> HS
>>>
>>>>>>>>> +
>>>>>>>>> +	cmdq_handle_error_done(cmdq, thread, irq_flag);
>>>>>>>>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static irqreturn_t cmdq_irq_handler(int irq, void *dev)
>>>>>>>>> +{
>>>>>>>>> +	struct cmdq *cmdq = dev;
>>>>>>>>> +	u32 irq_status;
>>>>>>>>> +	int i;
>>>>>>>>> +
>>>>>>>>> +	irq_status = readl(cmdq->base + CMDQ_CURR_IRQ_STATUS);
>>>>>>>>> +	irq_status &= CMDQ_IRQ_MASK;
>>>>>>>>> +	irq_status ^= CMDQ_IRQ_MASK;
>>>>>>>>
>>>>>>>> irq_status can be much bigger then 3, which is the number of threads in
>>>>>>>> the system (CMDQ_THR_MAX_COUNT). So why we use this mask here isn't
>>>>>>>> clear to me.
>>>>>>>
>>>>>>> Our GCE hardware has 16 threads, but we only use 3 threads currently.
>>>>>>>
>>>>>>
>>>>>> Ok, but please use bitops here.
>>>>>
>>>>> Will use bitops.
>>>>>
>>>>>>>>> +
>>>>>>>>> +	if (!irq_status)
>>>>>>>>> +		return IRQ_NONE;
>>>>>>>>> +
>>>>>>>>> +	while (irq_status) {
>>>>>>>>> +		i = ffs(irq_status) - 1;
>>>>>>>>> +		irq_status &= ~BIT(i);
>>>>>>>>> +		cmdq_thread_irq_handler(cmdq, i);
>>>>>>>>> +	}
>>>>>>>>
>>>>>>>> Can you explain how the irq status register looks like, that would it
>>>>>>>> make much easier to understand what happens here.
>>>>>>>
>>>>>>> Bit 0 ~ 15 of irq status register represents GCE thread 0 ~ 15
>>>>>>> interrupt. 0 means asserting interrupt; 1 means no interrupt.
>>>>>>>
>>>>>>
>>>>>> Thanks, that helped. :)
>>>>>>
>>>>>>>>> +
>>>>>>>>> +	return IRQ_HANDLED;	
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static int cmdq_task_handle_error_result(struct cmdq_task *task)
>>>>>>>>
>>>>>>>> We never check the return values, why do we have them?
>>>>>>>
>>>>>>> Will drop return value.
>>>>>>>
>>>>>>>>> +{
>>>>>>>>> +	struct cmdq *cmdq = task->cmdq;
>>>>>>>>> +	struct device *dev = cmdq->dev;
>>>>>>>>> +	struct cmdq_thread *thread = task->thread;
>>>>>>>>> +	struct cmdq_task *next_task, *prev_task;
>>>>>>>>> +	u32 irq_flag;
>>>>>>>>> +
>>>>>>>>> +	/* suspend GCE thread to ensure consistency */
>>>>>>>>> +	WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
>>>>>>>>> +
>>>>>>>>> +	/* ISR has handled this error task */
>>>>>>>>> +	if (task->task_state == TASK_STATE_ERROR) {
>>>>>>>>> +		next_task = list_first_entry_or_null(&thread->task_busy_list,
>>>>>>>>> +				struct cmdq_task, list_entry);
>>>>>>>>> +		if (next_task) /* move to next task */
>>>>>>>>> +			cmdq_thread_writel(thread, next_task->pa_base,
>>>>>>>>> +					   CMDQ_THR_CURR_ADDR);
>>>>>>>>
>>>>>>>> We have to do this, as we suppose that the thread did not reach the jump
>>>>>>>> instruction we put into it's command queue, right?
>>>>>>>
>>>>>>> Yes.
>>>>>>>
>>>>>>
>>>>>> So this should then go into it's own function. In wait_release_work,
>>>>>> something like this:
>>>>>>
>>>>>> if(task->task_state == TASK_STATE_ERROR)
>>>>>>          cmdq_task_handle_error(task)
>>>>>
>>>>> OK.
>>>>> I will write new function cmdq_task_handle_error() to handle error case.
>>>>>
>>>>>>>>> +		cmdq_thread_resume(thread);
>>>>>>>>> +		return -ECANCELED;
>>>>>>>>> +	}
>>>>>>>>> +
>>>>>>>>
>>>>>>>> if task_state != ERROR and != DONE. This means that the timeout of
>>>>>>>> task_release_wq has timed out, right?
>>>>>>>
>>>>>>> Yes.
>>>>>>>
>>>>>>>>> +	/*
>>>>>>>>> +	 * Save next_task and prev_task in advance
>>>>>>>>> +	 * since cmdq_handle_error_done will remove list_entry.
>>>>>>>>> +	 */
>>>>>>>>> +	next_task = prev_task = NULL;
>>>>>>>>> +	if (task->list_entry.next != &thread->task_busy_list)
>>>>>>>>> +		next_task = list_next_entry(task, list_entry);
>>>>>>>>> +	if (task->list_entry.prev != &thread->task_busy_list)
>>>>>>>>> +		prev_task = list_prev_entry(task, list_entry);
>>>>>>>>> +
>>>>>>>>> +	/*
>>>>>>>>> +	 * Although IRQ is disabled, GCE continues to execute.
>>>>>>>>> +	 * It may have pending IRQ before GCE thread is suspended,
>>>>>>>>> +	 * so check this condition again.
>>>>>>>>> +	 */
>>>>>>>>
>>>>>>>> The first thing we did in this function was suspending the thread. Why
>>>>>>>> do we need this then?
>>>>>>>
>>>>>>> Because timeout is CPU timeout not GCE timeout, GCE could just finish
>>>>>>> this task before the GCE thread is suspended.
>>>>>>>
>>>>>>
>>>>>> What are the reasons for a timeout? An error has happend, or the task is
>>>>>> still executing.
>>>>>
>>>>>    From GCE's point of view, this task is still executing.
>>>>> But, it could be an error of client.
>>>>> For example, task may never get event if display turn off hardware
>>>>> before waiting for task to finish its work.
>>>>>
>>>>>>>> To be honest this whole functions looks really like a design error. We
>>>>>>>> have to sperate the states much clearer so that it is possible to
>>>>>>>> understand what is happening in the GCE. Isn't it for example posible to
>>>>>>>> have worker queues for timed out tasks and tasks with an error? I'm not
>>>>>>>> sure how to do this, actually I'm not sure if I really understood how
>>>>>>>> this is supposed to work.
>>>>>>>
>>>>>>> GCE doesn't have timeout. The timeout is decided and controlled by CPU,
>>>>>>> so we check timeout in release work.
>>>>>>> For error and done, they are easy to check by register, and we have
>>>>>>> already created release work for timeout. So, I don't think we need to
>>>>>>> create work queue for each case.
>>>>>>>
>>>>>>> What do you think?
>>>>>>>
>>>>>>
>>>>>> I think, if we find in here, that the irq_flag is set, then the the
>>>>>> interrupt handler was triggered and is spinning the spinlock. If this is
>>>>>> not the case, we have a timeout and we handle this.
>>>>>
>>>>> I will write another function to handle error, and handle timeout here.
>>>>>
>>>>>>>> How much do we win, when we patch the thread command queue for every
>>>>>>>> task we add, instead of just taking one task after another from the
>>>>>>>> task_busy_list?
>>>>>>>
>>>>>>> GCE is used to help read/write registers with critical time limitation.
>>>>>>> Sometimes, client may ask to process multiple tasks in a short period
>>>>>>> of time, e.g. display flush multiple tasks for next vblank. So, CMDQ
>>>>>>> shouldn't limit to process one task after another from the
>>>>>>> task_busy_list. Currently, when interrupt or timeout, we will check
>>>>>>> how many tasks are done, and which one is error or timeout.
>>>>>>>
>>>>>>
>>>>>> So I suppose the device driver who use this are interested in throughput
>>>>>> and not in latency. The callback of every
>>>>>>
>>>>>>>>> +	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
>>>>>>>>> +	cmdq_handle_error_done(cmdq, thread, irq_flag);
>>>>>>>>> +	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
>>>>>>>>> +
>>>>>>>>> +	if (task->task_state == TASK_STATE_DONE) {
>>>>>>>>> +		cmdq_thread_resume(thread);
>>>>>>>>> +		return 0;
>>>>>>>>> +	}
>>>>>>>>> +
>>>>>>>>> +	if (task->task_state == TASK_STATE_ERROR) {
>>>>>>>>> +		dev_err(dev, "task 0x%p error\n", task);
>>>>>>>>> +		if (next_task) /* move to next task */
>>>>>>>>> +			cmdq_thread_writel(thread, next_task->pa_base,
>>>>>>>>> +					   CMDQ_THR_CURR_ADDR);
>>>>>>>>> +		cmdq_thread_resume(thread);
>>>>>>>>> +		return -ECANCELED;
>>>>>>>>> +	}
>>>>>>>>> +
>>>>>>>>> +	/* Task is running, so we force to remove it. */
>>>>>>>>> +	dev_err(dev, "task 0x%p timeout or killed\n", task);
>>>>>>>>> +	task->task_state = TASK_STATE_ERROR;
>>>>>>>>> +
>>>>>>>>> +	if (prev_task) {
>>>>>>>>> +		u64 *prev_va = prev_task->va_base;
>>>>>>>>> +		u64 *curr_va = task->va_base;
>>>>>>>>> +
>>>>>>>>> +		/* copy JUMP instruction */
>>>>>>>>> +		prev_va[prev_task->num_cmd - 1] = curr_va[task->num_cmd - 1];
>>>>>>>>> +
>>>>>>>>> +		cmdq_thread_invalidate_fetched_data(thread);
>>>>>>>>> +	} else if (next_task) { /* move to next task */
>>>>>>>>> +		cmdq_thread_writel(thread, next_task->pa_base,
>>>>>>>>> +				   CMDQ_THR_CURR_ADDR);
>>>>>>>>> +	}
>>>>>>>>> +
>>>>>>>>> +	list_del(&task->list_entry);
>>>>>>>>> +	cmdq_thread_resume(thread);
>>>>>>>>> +
>>>>>>>>> +	/* call cb here to prevent lock */
>>>>>>>>> +	if (task->cb.cb) {
>>>>>>>>> +		struct cmdq_cb_data cmdq_cb_data;
>>>>>>>>> +
>>>>>>>>> +		cmdq_cb_data.err = true;
>>>>>>>>> +		cmdq_cb_data.data = task->cb.data;
>>>>>>>>> +		task->cb.cb(cmdq_cb_data);
>>>>>>>>> +	}
>>>>>>>>> +
>>>>>>>>> +	return -ECANCELED;
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static void cmdq_task_wait_release_work(struct work_struct *work_item)
>>>>>>>>> +{
>>>>>>>>> +	struct cmdq_task *task = container_of(work_item, struct cmdq_task,
>>>>>>>>> +					      release_work);
>>>>>>>>> +	struct cmdq *cmdq = task->cmdq;
>>>>>>>>> +	struct cmdq_thread *thread = task->thread;
>>>>>>>>> +	unsigned long flags;
>>>>>>>>> +
>>>>>>>>> +	wait_event_timeout(thread->wait_task_done,
>>>>>>>>> +			   task->task_state != TASK_STATE_BUSY,
>>>>>>>>> +			   msecs_to_jiffies(CMDQ_TIMEOUT_MS));
>>>>>>>>> +
>>>>>>>>> +	spin_lock_irqsave(&cmdq->exec_lock, flags);
>>>>>>>>> +	if (task->task_state != TASK_STATE_DONE)
>>>>>>>>> +		cmdq_task_handle_error_result(task);
>>>>>>>>> +	if (list_empty(&thread->task_busy_list))
>>>>>>>>> +		cmdq_thread_disable(cmdq, thread);
>>>>>>>>> +	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>>>>>> +
>>>>>>>>> +	/* release regardless of success or not */
>>>>>>>>> +	clk_disable_unprepare(cmdq->clock);
>>>>>>>>> +	cmdq_task_release(task);
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static void cmdq_task_wait_release_schedule(struct cmdq_task *task)
>>>>>>>>> +{
>>>>>>>>> +	struct cmdq *cmdq = task->cmdq;
>>>>>>>>> +
>>>>>>>>> +	INIT_WORK(&task->release_work, cmdq_task_wait_release_work);
>>>>>>>>> +	queue_work(cmdq->task_release_wq, &task->release_work);
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static int cmdq_rec_realloc_cmd_buffer(struct cmdq_rec *rec, size_t size)
>>>>>>>>> +{
>>>>>>>>> +	void *new_buf;
>>>>>>>>> +
>>>>>>>>> +	new_buf = krealloc(rec->buf, size, GFP_KERNEL | __GFP_ZERO);
>>>>>>>>> +	if (!new_buf)
>>>>>>>>> +		return -ENOMEM;
>>>>>>>>> +	rec->buf = new_buf;
>>>>>>>>> +	rec->buf_size = size;
>>>>>>>>> +	return 0;
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +struct cmdq_base *cmdq_register_device(struct device *dev)
>>>>>>>>> +{
>>>>>>>>> +	struct cmdq_base *cmdq_base;
>>>>>>>>> +	struct resource res;
>>>>>>>>> +	int subsys;
>>>>>>>>> +	u32 base;
>>>>>>>>> +
>>>>>>>>> +	if (of_address_to_resource(dev->of_node, 0, &res))
>>>>>>>>> +		return NULL;
>>>>>>>>> +	base = (u32)res.start;
>>>>>>>>> +
>>>>>>>>> +	subsys = cmdq_subsys_base_to_id(base >> 16);
>>>>>>>>> +	if (subsys < 0)
>>>>>>>>> +		return NULL;
>>>>>>>>> +
>>>>>>>>> +	cmdq_base = devm_kmalloc(dev, sizeof(*cmdq_base), GFP_KERNEL);
>>>>>>>>> +	if (!cmdq_base)
>>>>>>>>> +		return NULL;
>>>>>>>>> +	cmdq_base->subsys = subsys;
>>>>>>>>> +	cmdq_base->base = base;
>>>>>>>>> +
>>>>>>>>> +	return cmdq_base;
>>>>>>>>> +}
>>>>>>>>> +EXPORT_SYMBOL(cmdq_register_device);
>>>>>>>>> +
>>>>>>>>> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
>>>>>>>>> +		    struct cmdq_rec **rec_ptr)
>>>>>>>>> +{
>>>>>>>>> +	struct cmdq_rec *rec;
>>>>>>>>> +	int err;
>>>>>>>>> +
>>>>>>>>> +	rec = kzalloc(sizeof(*rec), GFP_KERNEL);
>>>>>>>>> +	if (!rec)
>>>>>>>>> +		return -ENOMEM;
>>>>>>>>> +	rec->cmdq = dev_get_drvdata(dev);
>>>>>>>>> +	rec->engine_flag = engine_flag;
>>>>>>>>> +	err = cmdq_rec_realloc_cmd_buffer(rec, CMDQ_INITIAL_CMD_BLOCK_SIZE);
>>>>>>>>
>>>>>>>> Just pass PAGE_SIZE here, no need for CMDQ_INITIAL_CMD_BLOCK_SIZE.
>>>>>>>
>>>>>>> Will do.
>>>>>>>
>>>>>>>>> +	if (err < 0) {
>>>>>>>>> +		kfree(rec);
>>>>>>>>> +		return err;
>>>>>>>>> +	}
>>>>>>>>> +	*rec_ptr = rec;
>>>>>>>>> +	return 0;
>>>>>>>>> +}
>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_create);
>>>>>>>>> +
>>>>>>>>> +static int cmdq_rec_append_command(struct cmdq_rec *rec, enum cmdq_code code,
>>>>>>>>> +				   u32 arg_a, u32 arg_b)
>>>>>>>>> +{
>>>>>>>>> +	u64 *cmd_ptr;
>>>>>>>>> +	int err;
>>>>>>>>> +
>>>>>>>>> +	if (WARN_ON(rec->finalized))
>>>>>>>>> +		return -EBUSY;
>>>>>>>>> +	if (code < CMDQ_CODE_MOVE || code > CMDQ_CODE_EOC)
>>>>>>>>> +		return -EINVAL;
>>>>>>>>
>>>>>>>> cmdq_rec_append_command is just called from inside this driver and code
>>>>>>>> is a enum. We can expect it to be correct, no need for this check.
>>>>>>>
>>>>>>> Will drop this check.
>>>>>>>
>>>>>>>>> +	if (unlikely(rec->command_size + CMDQ_INST_SIZE > rec->buf_size)) {
>>>>>>>>
>>>>>>>> command_size is the offset into the buffer to which a new command is
>>>>>>>> written, so this name is highly confusing. I wonder if this would be
>>>>>>>> easier to understand if we redefine command_size to something like the
>>>>>>>> number of commands and divide/multiply CMDQ_INST_SIZE where this is needed.
>>>>>>>
>>>>>>> I can rename command_size to cmd_buf_size and calculate num_cmd by
>>>>>>> dividing CMDQ_INST_SIZE.
>>>>>>> What do you think?
>>>>>>>
>>>>>>>>> +		err = cmdq_rec_realloc_cmd_buffer(rec, rec->buf_size * 2);
>>>>>>>>> +		if (err < 0)
>>>>>>>>> +			return err;
>>>>>>>>> +	}
>>>>>>>>> +	cmd_ptr = rec->buf + rec->command_size;
>>>>>>>>> +	(*cmd_ptr) = (u64)((code << CMDQ_OP_CODE_SHIFT) | arg_a) << 32 | arg_b;
>>>>>>>>> +	rec->command_size += CMDQ_INST_SIZE;
>>>>>>>>> +	return 0;
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value, struct cmdq_base *base,
>>>>>>>>> +		   u32 offset)
>>>>>>>>> +{
>>>>>>>>> +	u32 arg_a = ((base->base + offset) & CMDQ_ARG_A_WRITE_MASK) |
>>>>>>>>> +		    ((base->subsys & CMDQ_SUBSYS_MASK) << CMDQ_SUBSYS_SHIFT);
>>>>>>>>
>>>>>>>> base->subsys is the id from gce_sybsys, so we can expect it to be
>>>>>>>> correct, no need to mask with CMDQ_SUBSYS_MASK.
>>>>>>>
>>>>>>> Will drop it.
>>>>>>>
>>>>>>>>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WRITE, arg_a, value);
>>>>>>>>> +}
>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_write);
>>>>>>>>> +
>>>>>>>>> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
>>>>>>>>> +			struct cmdq_base *base, u32 offset, u32 mask)
>>>>>>>>> +{
>>>>>>>>> +	u32 offset_mask = offset;
>>>>>>>>> +	int err;
>>>>>>>>> +
>>>>>>>>> +	if (mask != 0xffffffff) {
>>>>>>>>> +		err = cmdq_rec_append_command(rec, CMDQ_CODE_MOVE, 0, ~mask);
>>>>>>>>> +		if (err < 0)
>>>>>>>>> +			return err;
>>>>>>>>> +		offset_mask |= CMDQ_WRITE_ENABLE_MASK;
>>>>>>>>> +	}
>>>>>>>>> +	return cmdq_rec_write(rec, value, base, offset_mask);
>>>>>>>>> +}
>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_write_mask);
>>>>>>>>> +
>>>>>>>>> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event)
>>>>>>>>> +{
>>>>>>>>> +	u32 arg_b;
>>>>>>>>> +
>>>>>>>>> +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
>>>>>>>>> +		return -EINVAL;
>>>>>>>>> +
>>>>>>>>> +	/*
>>>>>>>>> +	 * bit 0-11: wait value
>>>>>>>>> +	 * bit 15: 1 - wait, 0 - no wait
>>>>>>>>> +	 * bit 16-27: update value
>>>>>>>>> +	 * bit 31: 1 - update, 0 - no update
>>>>>>>>> +	 */
>>>>>>>>
>>>>>>>> I don't understand this comments. What are they for?
>>>>>>>
>>>>>>> This is for WFE command. I will comment it.
>>>>>>>
>>>>>>>>> +	arg_b = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
>>>>>>>>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event, arg_b);
>>>>>>>>> +}
>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_wfe);
>>>>>>>>> +
>>>>>>>>> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event)
>>>>>>>>> +{
>>>>>>>>> +	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
>>>>>>>>> +		return -EINVAL;
>>>>>>>>> +
>>>>>>>>> +	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event,
>>>>>>>>> +				       CMDQ_WFE_UPDATE);
>>>>>>>>> +}
>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_clear_event);
>>>>>>>>> +
>>>>>>>>> +static int cmdq_rec_finalize(struct cmdq_rec *rec)
>>>>>>>>> +{
>>>>>>>>> +	int err;
>>>>>>>>> +
>>>>>>>>> +	if (rec->finalized)
>>>>>>>>> +		return 0;
>>>>>>>>> +
>>>>>>>>> +	/* insert EOC and generate IRQ for each command iteration */
>>>>>>>>> +	err = cmdq_rec_append_command(rec, CMDQ_CODE_EOC, 0, CMDQ_EOC_IRQ_EN);
>>>>>>>>> +	if (err < 0)
>>>>>>>>> +		return err;
>>>>>>>>> +
>>>>>>>>> +	/* JUMP to end */
>>>>>>>>> +	err = cmdq_rec_append_command(rec, CMDQ_CODE_JUMP, 0, CMDQ_JUMP_PASS);
>>>>>>>>> +	if (err < 0)
>>>>>>>>> +		return err;
>>>>>>>>> +
>>>>>>>>
>>>>>>>> Does this need to be atomic?
>>>>>>>> What happens if after CODE_EOC and before CODE_JUMP some
>>>>>>>> write/read/event gets added?
>>>>>>>> What happens if more commands get added to the queue after CODE_JUMP,
>>>>>>>> but before finalized is set to true. Why don't you use atomic functions
>>>>>>>> to access finalized?
>>>>>>>
>>>>>>> Since cmdq_rec doesn't guarantee thread safe, mutex is needed when
>>>>>>> client uses cmdq_rec.
>>>>>>>
>>>>>>
>>>>>> Well I think that rec->finalized tries to implement this, but might
>>>>>> fail, if two kernel threads work on the same cmdq_rec.
>>>>>>
>>>>>>>>> +	rec->finalized = true;
>>>>>>>>> +	return 0;
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
>>>>>>>>> +			 void *data)
>>>>>>>>> +{
>>>>>>>>> +	struct cmdq *cmdq = rec->cmdq;
>>>>>>>>> +	struct cmdq_task *task;
>>>>>>>>> +	struct cmdq_task_cb task_cb;
>>>>>>>>> +	struct cmdq_thread *thread;
>>>>>>>>> +	int err;
>>>>>>>>> +
>>>>>>>>> +	mutex_lock(&cmdq->task_mutex);
>>>>>>>>> +	if (cmdq->suspended) {
>>>>>>>>> +		dev_err(cmdq->dev, "%s is called after suspended\n", __func__);
>>>>>>>>> +		mutex_unlock(&cmdq->task_mutex);
>>>>>>>>> +		return -EPERM;
>>>>>>>>> +	}
>>>>>>>>> +
>>>>>>>>> +	err = cmdq_rec_finalize(rec);
>>>>>>>>> +	if (err < 0) {
>>>>>>>>> +		mutex_unlock(&cmdq->task_mutex);
>>>>>>>>> +		return err;
>>>>>>>>> +	}
>>>>>>>>> +
>>>>>>>>> +	task_cb.cb = cb;
>>>>>>>>> +	task_cb.data = data;
>>>>>>>>> +	task = cmdq_task_acquire(rec, task_cb);
>>>>>>>>> +	if (!task) {
>>>>>>>>> +		mutex_unlock(&cmdq->task_mutex);
>>>>>>>>> +		return -EFAULT;
>>>>>>>>> +	}
>>>>>>>>> +
>>>>>>>>> +	thread = &cmdq->thread[cmdq_eng_get_thread(task->engine_flag)];
>>>>>>>>> +	cmdq_task_exec(task, thread);
>>>>>>>>> +	cmdq_task_wait_release_schedule(task);
>>>>>>>>> +	mutex_unlock(&cmdq->task_mutex);
>>>>>>>>> +	return 0;
>>>>>>>>> +}
>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_flush_async);
>>>>>>>>> +
>>>>>>>>> +struct cmdq_flush_completion {
>>>>>>>>> +	struct completion cmplt;
>>>>>>>>> +	bool err;
>>>>>>>>> +};
>>>>>>>>> +
>>>>>>>>> +static int cmdq_rec_flush_cb(struct cmdq_cb_data data)
>>>>>>>>> +{
>>>>>>>>> +	struct cmdq_flush_completion *cmplt = data.data;
>>>>>>>>> +
>>>>>>>>> +	cmplt->err = data.err;
>>>>>>>>> +	complete(&cmplt->cmplt);
>>>>>>>>> +	return 0;
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +int cmdq_rec_flush(struct cmdq_rec *rec)
>>>>>>>>> +{
>>>>>>>>> +	struct cmdq_flush_completion cmplt;
>>>>>>>>> +	int err;
>>>>>>>>> +
>>>>>>>>> +	init_completion(&cmplt.cmplt);
>>>>>>>>> +	err = cmdq_rec_flush_async(rec, cmdq_rec_flush_cb, &cmplt);
>>>>>>>>> +	if (err < 0)
>>>>>>>>> +		return err;
>>>>>>>>> +	wait_for_completion(&cmplt.cmplt);
>>>>>>>>> +	return cmplt.err ? -EFAULT : 0;
>>>>>>>>> +}
>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_flush);
>>>>>>>>> +
>>>>>>>>> +void cmdq_rec_destroy(struct cmdq_rec *rec)
>>>>>>>>> +{
>>>>>>>>> +	kfree(rec->buf);
>>>>>>>>> +	kfree(rec);
>>>>>>>>> +}
>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_destroy);
>>>>>>>>> +
>>>>>>>>> +static bool cmdq_task_is_empty(struct cmdq *cmdq)
>>>>>>>>> +{
>>>>>>>>> +	struct cmdq_thread *thread;
>>>>>>>>> +	int i;
>>>>>>>>> +
>>>>>>>>> +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
>>>>>>>>> +		thread = &cmdq->thread[i];
>>>>>>>>> +		if (!list_empty(&thread->task_busy_list))
>>>>>>>>> +			return false;
>>>>>>>>> +	}
>>>>>>>>> +	return true;
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static int cmdq_suspend(struct device *dev)
>>>>>>>>> +{
>>>>>>>>> +	struct cmdq *cmdq = dev_get_drvdata(dev);
>>>>>>>>> +	u32 exec_threads;
>>>>>>>>> +
>>>>>>>>> +	mutex_lock(&cmdq->task_mutex);
>>>>>>>>> +	cmdq->suspended = true;
>>>>>>>>> +	mutex_unlock(&cmdq->task_mutex);
>>>>>>>>> +
>>>>>>>>> +	exec_threads = readl(cmdq->base + CMDQ_CURR_LOADED_THR);
>>>>>>>>> +	if ((exec_threads & CMDQ_THR_EXECUTING) && !cmdq_task_is_empty(cmdq)) {
>>>>>>>>> +		dev_err(dev, "wait active tasks timeout.\n");
>>>>>>>>> +		flush_workqueue(cmdq->task_release_wq);
>>>>>>>>> +	}
>>>>>>>>> +	return 0;
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static int cmdq_resume(struct device *dev)
>>>>>>>>> +{
>>>>>>>>> +	struct cmdq *cmdq = dev_get_drvdata(dev);
>>>>>>>>> +
>>>>>>>>> +	cmdq->suspended = false;
>>>>>>>>> +	return 0;
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static int cmdq_remove(struct platform_device *pdev)
>>>>>>>>> +{
>>>>>>>>> +	struct cmdq *cmdq = platform_get_drvdata(pdev);
>>>>>>>>> +
>>>>>>>>> +	destroy_workqueue(cmdq->task_release_wq);
>>>>>>>>> +	cmdq->task_release_wq = NULL;
>>>>>>>>> +	return 0;
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static int cmdq_probe(struct platform_device *pdev)
>>>>>>>>> +{
>>>>>>>>> +	struct device *dev = &pdev->dev;
>>>>>>>>> +	struct device_node *node = dev->of_node;
>>>>>>>>> +	struct resource *res;
>>>>>>>>> +	struct cmdq *cmdq;
>>>>>>>>> +	int err, i;
>>>>>>>>> +
>>>>>>>>> +	cmdq = devm_kzalloc(dev, sizeof(*cmdq), GFP_KERNEL);
>>>>>>>>> +	if (!cmdq)
>>>>>>>>> +		return -ENOMEM;
>>>>>>>>> +	cmdq->dev = dev;
>>>>>>>>> +
>>>>>>>>> +	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
>>>>>>>>> +	cmdq->base = devm_ioremap_resource(dev, res);
>>>>>>>>> +	if (IS_ERR(cmdq->base)) {
>>>>>>>>> +		dev_err(dev, "failed to ioremap gce\n");
>>>>>>>>> +		return PTR_ERR(cmdq->base);
>>>>>>>>> +	}
>>>>>>>>> +
>>>>>>>>> +	cmdq->irq = irq_of_parse_and_map(node, 0);
>>>>>>>>> +	if (!cmdq->irq) {
>>>>>>>>> +		dev_err(dev, "failed to get irq\n");
>>>>>>>>> +		return -EINVAL;
>>>>>>>>> +	}
>>>>>>>>> +
>>>>>>>>> +	dev_dbg(dev, "cmdq device: addr:0x%p, va:0x%p, irq:%d\n",
>>>>>>>>> +		dev, cmdq->base, cmdq->irq);
>>>>>>>>> +
>>>>>>>>> +	mutex_init(&cmdq->task_mutex);
>>>>>>>>> +	spin_lock_init(&cmdq->exec_lock);
>>>>>>>>> +	cmdq->task_release_wq = alloc_ordered_workqueue(
>>>>>>>>> +			"%s", WQ_MEM_RECLAIM | WQ_HIGHPRI,
>>>>>>>>> +			"cmdq_task_wait_release");
>>>>>>>>> +
>>>>>>>>> +	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
>>>>>>>>> +		cmdq->thread[i].base = cmdq->base + CMDQ_THR_BASE +
>>>>>>>>> +				CMDQ_THR_SHIFT * i;
>>>>>>>>> +		init_waitqueue_head(&cmdq->thread[i].wait_task_done);
>>>>>>>>> +		INIT_LIST_HEAD(&cmdq->thread[i].task_busy_list);
>>>>>>>>> +	}
>>>>>>>>> +
>>>>>>>>> +	platform_set_drvdata(pdev, cmdq);
>>>>>>>>> +
>>>>>>>>> +	err = devm_request_irq(dev, cmdq->irq, cmdq_irq_handler, IRQF_SHARED,
>>>>>>>>> +			       CMDQ_DRIVER_DEVICE_NAME, cmdq);
>>>>>>>>> +	if (err < 0) {
>>>>>>>>> +		dev_err(dev, "failed to register ISR (%d)\n", err);
>>>>>>>>> +		goto fail;
>>>>>>>>> +	}
>>>>>>>>> +
>>>>>>>>> +	cmdq->clock = devm_clk_get(dev, CMDQ_CLK_NAME);
>>>>>>>>> +	if (IS_ERR(cmdq->clock)) {
>>>>>>>>> +		dev_err(dev, "failed to get clk:%s\n", CMDQ_CLK_NAME);
>>>>>>>>> +		err = PTR_ERR(cmdq->clock);
>>>>>>>>> +		goto fail;
>>>>>>>>> +	}
>>>>>>>>> +	return 0;
>>>>>>>>> +
>>>>>>>>> +fail:
>>>>>>>>> +	cmdq_remove(pdev);
>>>>>>>>> +	return err;
>>>>>>>>> +}
>>>>>>>>> +
>>>>>>>>> +static const struct dev_pm_ops cmdq_pm_ops = {
>>>>>>>>> +	.suspend = cmdq_suspend,
>>>>>>>>> +	.resume = cmdq_resume,
>>>>>>>>> +};
>>>>>>>>> +
>>>>>>>>> +static const struct of_device_id cmdq_of_ids[] = {
>>>>>>>>> +	{.compatible = "mediatek,mt8173-gce",},
>>>>>>>>> +	{}
>>>>>>>>> +};
>>>>>>>>> +
>>>>>>>>> +static struct platform_driver cmdq_drv = {
>>>>>>>>> +	.probe = cmdq_probe,
>>>>>>>>> +	.remove = cmdq_remove,
>>>>>>>>> +	.driver = {
>>>>>>>>> +		.name = CMDQ_DRIVER_DEVICE_NAME,
>>>>>>>>> +		.owner = THIS_MODULE,
>>>>>>>>> +		.pm = &cmdq_pm_ops,
>>>>>>>>> +		.of_match_table = cmdq_of_ids,
>>>>>>>>> +	}
>>>>>>>>> +};
>>>>>>>>> +
>>>>>>>>> +builtin_platform_driver(cmdq_drv);
>>>>>>>>> diff --git a/include/soc/mediatek/cmdq.h b/include/soc/mediatek/cmdq.h
>>>>>>>>> new file mode 100644
>>>>>>>>> index 0000000..60eef3d
>>>>>>>>> --- /dev/null
>>>>>>>>> +++ b/include/soc/mediatek/cmdq.h
>>>>>>>>> @@ -0,0 +1,197 @@
>>>>>>>>> +/*
>>>>>>>>> + * Copyright (c) 2015 MediaTek Inc.
>>>>>>>>> + *
>>>>>>>>> + * This program is free software; you can redistribute it and/or modify
>>>>>>>>> + * it under the terms of the GNU General Public License version 2 as
>>>>>>>>> + * published by the Free Software Foundation.
>>>>>>>>> + *
>>>>>>>>> + * This program is distributed in the hope that it will be useful,
>>>>>>>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>>>>>>>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
>>>>>>>>> + * GNU General Public License for more details.
>>>>>>>>> + */
>>>>>>>>> +
>>>>>>>>> +#ifndef __MTK_CMDQ_H__
>>>>>>>>> +#define __MTK_CMDQ_H__
>>>>>>>>> +
>>>>>>>>> +#include <linux/platform_device.h>
>>>>>>>>> +#include <linux/types.h>
>>>>>>>>> +
>>>>>>>>> +enum cmdq_eng {
>>>>>>>>> +	CMDQ_ENG_DISP_AAL,
>>>>>>>>> +	CMDQ_ENG_DISP_COLOR0,
>>>>>>>>> +	CMDQ_ENG_DISP_COLOR1,
>>>>>>>>> +	CMDQ_ENG_DISP_DPI0,
>>>>>>>>> +	CMDQ_ENG_DISP_DSI0,
>>>>>>>>> +	CMDQ_ENG_DISP_DSI1,
>>>>>>>>> +	CMDQ_ENG_DISP_GAMMA,
>>>>>>>>> +	CMDQ_ENG_DISP_OD,
>>>>>>>>> +	CMDQ_ENG_DISP_OVL0,
>>>>>>>>> +	CMDQ_ENG_DISP_OVL1,
>>>>>>>>> +	CMDQ_ENG_DISP_PWM0,
>>>>>>>>> +	CMDQ_ENG_DISP_PWM1,
>>>>>>>>> +	CMDQ_ENG_DISP_RDMA0,
>>>>>>>>> +	CMDQ_ENG_DISP_RDMA1,
>>>>>>>>> +	CMDQ_ENG_DISP_RDMA2,
>>>>>>>>> +	CMDQ_ENG_DISP_UFOE,
>>>>>>>>> +	CMDQ_ENG_DISP_WDMA0,
>>>>>>>>> +	CMDQ_ENG_DISP_WDMA1,
>>>>>>>>> +	CMDQ_ENG_MAX,
>>>>>>>>> +};
>>>>>>>>> +
>>>>>>>>> +/* events for CMDQ and display */
>>>>>>>>> +enum cmdq_event {
>>>>>>>>> +	/* Display start of frame(SOF) events */
>>>>>>>>> +	CMDQ_EVENT_DISP_OVL0_SOF = 11,
>>>>>>>>> +	CMDQ_EVENT_DISP_OVL1_SOF = 12,
>>>>>>>>> +	CMDQ_EVENT_DISP_RDMA0_SOF = 13,
>>>>>>>>> +	CMDQ_EVENT_DISP_RDMA1_SOF = 14,
>>>>>>>>> +	CMDQ_EVENT_DISP_RDMA2_SOF = 15,
>>>>>>>>> +	CMDQ_EVENT_DISP_WDMA0_SOF = 16,
>>>>>>>>> +	CMDQ_EVENT_DISP_WDMA1_SOF = 17,
>>>>>>>>> +	/* Display end of frame(EOF) events */
>>>>>>>>> +	CMDQ_EVENT_DISP_OVL0_EOF = 39,
>>>>>>>>> +	CMDQ_EVENT_DISP_OVL1_EOF = 40,
>>>>>>>>> +	CMDQ_EVENT_DISP_RDMA0_EOF = 41,
>>>>>>>>> +	CMDQ_EVENT_DISP_RDMA1_EOF = 42,
>>>>>>>>> +	CMDQ_EVENT_DISP_RDMA2_EOF = 43,
>>>>>>>>> +	CMDQ_EVENT_DISP_WDMA0_EOF = 44,
>>>>>>>>> +	CMDQ_EVENT_DISP_WDMA1_EOF = 45,
>>>>>>>>> +	/* Mutex end of frame(EOF) events */
>>>>>>>>> +	CMDQ_EVENT_MUTEX0_STREAM_EOF = 53,
>>>>>>>>> +	CMDQ_EVENT_MUTEX1_STREAM_EOF = 54,
>>>>>>>>> +	CMDQ_EVENT_MUTEX2_STREAM_EOF = 55,
>>>>>>>>> +	CMDQ_EVENT_MUTEX3_STREAM_EOF = 56,
>>>>>>>>> +	CMDQ_EVENT_MUTEX4_STREAM_EOF = 57,
>>>>>>>>> +	/* Display underrun events */
>>>>>>>>> +	CMDQ_EVENT_DISP_RDMA0_UNDERRUN = 63,
>>>>>>>>> +	CMDQ_EVENT_DISP_RDMA1_UNDERRUN = 64,
>>>>>>>>> +	CMDQ_EVENT_DISP_RDMA2_UNDERRUN = 65,
>>>>>>>>> +	/* Keep this at the end of HW events */
>>>>>>>>> +	CMDQ_MAX_HW_EVENT_COUNT = 260,
>>>>>>>>> +};
>>>>>>>>> +
>>>>>>>>> +struct cmdq_cb_data {
>>>>>>>>> +	bool	err;
>>>>>>>>> +	void	*data;
>>>>>>>>> +};
>>>>>>>>> +
>>>>>>>>> +typedef int (*cmdq_async_flush_cb)(struct cmdq_cb_data data);
>>>>>>>>> +
>>>>>>>>> +struct cmdq_task;
>>>>>>>>> +struct cmdq;
>>>>>>>>> +
>>>>>>>>> +struct cmdq_rec {
>>>>>>>>> +	struct cmdq		*cmdq;
>>>>>>>>> +	u64			engine_flag;
>>>>>>>>> +	size_t			command_size;
>>>>>>>>> +	void			*buf;
>>>>>>>>> +	size_t			buf_size;
>>>>>>>>> +	bool			finalized;
>>>>>>>>> +};
>>>>>>
>>>>>> Why do we need cmdq_rec at all? Can't we just use the cmdq_task for that
>>>>>> and this way make the driver less complex?
>>>>>
>>>>> There are two main reasons for cmdq_rec.
>>>>> 1. It is slow to access dma too frequently.
>>>>>       So, we append commands to cacheable memory, and then flush to dma.
>>>>> 2. cmdq_rec is not thread safe, but cmdq_task needs thread safe.
>>>>>       If we merge them, we need to take care of some synchronization
>>>>>       issues.
>>>>>
>>>>>>>>> +
>>>>>>>>> +struct cmdq_base {
>>>>>>>>> +	int	subsys;
>>>>>>>>> +	u32	base;
>>>>>>>>
>>>>>>>> subsys can always be calculated via cmdq_subsys_base_to_id(base >> 16)
>>>>>>>> so we can get rid of the struct, right?
>>>>>>>
>>>>>>> Current subsys method is based on previous comment from Daniel Kurtz.
>>>>>>> Please take a look of our previous discussion.
>>>>>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/004483.html
>>>>>>> Thanks.
>>>>>>>
>>>>>>
>>>>>> I have to look deeper into this, but from what I read, the proposal from
>>>>>> Daniel [1] seems good to me.
>>>>>>
>>>>>> [1] https://patchwork.kernel.org/patch/8068311/
>>>>>
>>>>> The initial proposal has some problem, so please see the follow-up
>>>>> discussions. Thanks.
>>>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/003972.html
>>>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/004483.html
>>>>>
>>>>>
>>>>>>>>> +};
>>>>>>>>> +
>>>>>>>>> +/**
>>>>>>>>> + * cmdq_register_device() - register device which needs CMDQ
>>>>>>>>> + * @dev:		device
>>>>>>>>> + *
>>>>>>>>> + * Return: cmdq_base pointer or NULL for failed
>>>>>>>>> + */
>>>>>>>>> +struct cmdq_base *cmdq_register_device(struct device *dev);
>>>>>>>>> +
>>>>>>>>> +/**
>>>>>>>>> + * cmdq_rec_create() - create command queue record
>>>>>>>>> + * @dev:		device
>>>>>>>>> + * @engine_flag:	command queue engine flag
>>>>>>>>> + * @rec_ptr:		command queue record pointer to retrieve cmdq_rec
>>>>>>>>> + *
>>>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>>>> + */
>>>>>>>>> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
>>>>>>>>> +		    struct cmdq_rec **rec_ptr);
>>>>>>>>> +
>>>>>>>>> +/**
>>>>>>>>> + * cmdq_rec_write() - append write command to the command queue record
>>>>>>>>> + * @rec:	the command queue record
>>>>>>>>> + * @value:	the specified target register value
>>>>>>>>> + * @base:	the command queue base
>>>>>>>>> + * @offset:	register offset from module base
>>>>>>>>> + *
>>>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>>>> + */
>>>>>>>>> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value,
>>>>>>>>> +		   struct cmdq_base *base, u32 offset);
>>>>>>>>> +
>>>>>>>>> +/**
>>>>>>>>> + * cmdq_rec_write_mask() - append write command with mask to the command
>>>>>>>>> + *			   queue record
>>>>>>>>> + * @rec:	the command queue record
>>>>>>>>> + * @value:	the specified target register value
>>>>>>>>> + * @base:	the command queue base
>>>>>>>>> + * @offset:	register offset from module base
>>>>>>>>> + * @mask:	the specified target register mask
>>>>>>>>> + *
>>>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>>>> + */
>>>>>>>>> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
>>>>>>>>> +			struct cmdq_base *base, u32 offset, u32 mask);
>>>>>>>>> +
>>>>>>>>> +/**
>>>>>>>>> + * cmdq_rec_wfe() - append wait for event command to the command queue reco	rd
>>>>>>>>
>>>>>>>> reco rd -> record
>>>>>>>
>>>>>>> Will fix it.
>>>>>>>
>>>>>>>> Regards,
>>>>>>>> Matthias
>>>>>>>>
>>>>>>>>> + * @rec:	the command queue record
>>>>>>>>> + * @event:	the desired event type to "wait and CLEAR"
>>>>>>>>> + *
>>>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>>>> + */
>>>>>>>>> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event);
>>>>>>>>> +
>>>>>>>>> +/**
>>>>>>>>> + * cmdq_rec_clear_event() - append clear event command to the command queue
>>>>>>>>> + *			    record
>>>>>>>>> + * @rec:	the command queue record
>>>>>>>>> + * @event:	the desired event to be cleared
>>>>>>>>> + *
>>>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>>>> + */
>>>>>>>>> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event);
>>>>>>>>> +
>>>>>>>>> +/**
>>>>>>>>> + * cmdq_rec_flush() - trigger CMDQ to execute the recorded commands
>>>>>>>>> + * @rec:	the command queue record
>>>>>>>>> + *
>>>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>>>> + *
>>>>>>>>> + * Trigger CMDQ to execute the recorded commands. Note that this is a
>>>>>>>>> + * synchronous flush function. When the function returned, the recorded
>>>>>>>>> + * commands have been done.
>>>>>>>>> + */
>>>>>>>>> +int cmdq_rec_flush(struct cmdq_rec *rec);
>>>>>>>>> +
>>>>>>>>> +/**
>>>>>>>>> + * cmdq_rec_flush_async() - trigger CMDQ to asynchronously execute the recorded
>>>>>>>>> + *			    commands and call back after ISR is finished
>>>>>>>>> + * @rec:	the command queue record
>>>>>>>>> + * @cb:		called in the end of CMDQ ISR
>>>>>>>>> + * @data:	this data will pass back to cb
>>>>>>>>> + *
>>>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>>>> + *
>>>>>>>>> + * Trigger CMDQ to asynchronously execute the recorded commands and call back
>>>>>>>>> + * after ISR is finished. Note that this is an ASYNC function. When the function
>>>>>>>>> + * returned, it may or may not be finished. The ISR callback function is called
>>>>>>>>> + * in the end of ISR.
>>>>>>
>>>>>> "The callback is called from the ISR."
>>>>>>
>>>>>> Regards,
>>>>>> Matthias
>>>>>>
>>>>>>>>> + */
>>>>>>>>> +int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
>>>>>>>>> +			 void *data);
>>>>>>>>> +
>>>>>>>>> +/**
>>>>>>>>> + * cmdq_rec_destroy() - destroy command queue record
>>>>>>>>> + * @rec:	the command queue record
>>>>>>>>> + */
>>>>>>>>> +void cmdq_rec_destroy(struct cmdq_rec *rec);
>>>>>>>>> +
>>>>>>>>> +#endif	/* __MTK_CMDQ_H__ */
>>>>>>>>>
>>>>>>>
>>>>>>> Thanks,
>>>>>>> HS
>>>>>>>
>>>>>
>>>>> Thanks,
>>>>> HS
>>>>>
>>>
>>>
>
>

Jassi Brar June 3, 2016, 1:11 p.m. UTC | #12

On Fri, Jun 3, 2016 at 4:48 PM, Matthias Brugger <[e-mail]> wrote:
> On 03/06/16 08:12, Horng-Shyang Liao wrote:
>> On Thu, 2016-06-02 at 10:46 +0200, Matthias Brugger wrote:

>>> I keep thinking about how to get rid of the two data structures,
>>> task_busy_list and the task_release_wq. We need the latter for the only
>>> sake of getting a timeout.
>>>
>>> Did you have a look on how the mailbox framework handles this?
>>> By the way, what is the reason to not implement the whole driver as a
>>> mailbox controller? For me, this driver looks like a good fit.
>>
>>
>> CMDQ needs to encode commands for GCE hardware. We think this behavior
>> should be put in CMDQ driver, and client just call CMDQ functions.
>> Therefore, if we want to use mailbox framework, cmdq_rec must be
>> mailbox client, and the others must be mailbox controller.
>>
>
> You mean the functions to fill the cmdq_rec and execute it?
> I think this should be part of the driver.
>
> Jassi, can you have a look on the interface this driver exports [0].
> They are needed to actually create the message which will be send.
> Could something like this be part of a mailbox driver?
>
> [0] https://patchwork.kernel.org/patch/9140221/
>
Packet creating/parsing should not be a part of controller driver. As
the log of this patch says, today it is used for only display but in
future it could work with other h/w as well, so it makes sense to have
mailbox api do the message queuing, the controller driver do the
send/receive and client drivers implement display and other h/w
specific packaging of data (protocol handling).

So yes, I think this could use mailbox api.

Cheers.

[e-mail] June 6, 2016, 9:33 a.m. UTC | #13

Hi Matthias, Jassi,

On Fri, 2016-06-03 at 18:41 +0530, Jassi Brar wrote:
> On Fri, Jun 3, 2016 at 4:48 PM, Matthias Brugger <[e-mail]> wrote:
> > On 03/06/16 08:12, Horng-Shyang Liao wrote:
> >> On Thu, 2016-06-02 at 10:46 +0200, Matthias Brugger wrote:
> 
> >>> I keep thinking about how to get rid of the two data structures,
> >>> task_busy_list and the task_release_wq. We need the latter for the only
> >>> sake of getting a timeout.
> >>>
> >>> Did you have a look on how the mailbox framework handles this?
> >>> By the way, what is the reason to not implement the whole driver as a
> >>> mailbox controller? For me, this driver looks like a good fit.
> >>
> >>
> >> CMDQ needs to encode commands for GCE hardware. We think this behavior
> >> should be put in CMDQ driver, and client just call CMDQ functions.
> >> Therefore, if we want to use mailbox framework, cmdq_rec must be
> >> mailbox client, and the others must be mailbox controller.
> >>
> >
> > You mean the functions to fill the cmdq_rec and execute it?
> > I think this should be part of the driver.
> >
> > Jassi, can you have a look on the interface this driver exports [0].
> > They are needed to actually create the message which will be send.
> > Could something like this be part of a mailbox driver?
> >
> > [0] https://patchwork.kernel.org/patch/9140221/
> >
> Packet creating/parsing should not be a part of controller driver. As
> the log of this patch says, today it is used for only display but in
> future it could work with other h/w as well, so it makes sense to have
> mailbox api do the message queuing, the controller driver do the
> send/receive and client drivers implement display and other h/w
> specific packaging of data (protocol handling).
> 
> So yes, I think this could use mailbox api.
> 
> Cheers.


Let me use display as an example to do some further explanation
about CMDQ in advance. You can think CMDQ is a shadow register
replacement. Therefore, we use cmdq_rec_write(_mask), cmdq_rec_wfe, and
cmdq_rec_clear_event instead of accessing registers, and use
cmdq_rec_flush(_async) instead of atomic swap.

If we use mailbox to do the message queue, we can use mailbox framework
to implement flush and callback. However, I don't think mailbox is
suitable for cmdq_rec_write(_mask), cmdq_rec_wfe, and
cmdq_rec_clear_event since they are just record some commands. Is this
the same as your comment "Packet creating/parsing should not be a part
of controller driver."?

Therefore, do you mean we use mailbox framework to implement flush and
callback and keep other interfaces? Just want to confirm that I get the
correct idea from you. Many thanks for your kindly reply.

Thanks,
HS

[e-mail] June 7, 2016, 2:45 a.m. UTC | #14

On Fri, 2016-06-03 at 18:41 +0530, Jassi Brar wrote:
> On Fri, Jun 3, 2016 at 4:48 PM, Matthias Brugger <[e-mail]> wrote:
> > On 03/06/16 08:12, Horng-Shyang Liao wrote:
> >> On Thu, 2016-06-02 at 10:46 +0200, Matthias Brugger wrote:
> 
> >>> I keep thinking about how to get rid of the two data structures,
> >>> task_busy_list and the task_release_wq. We need the latter for the only
> >>> sake of getting a timeout.
> >>>
> >>> Did you have a look on how the mailbox framework handles this?
> >>> By the way, what is the reason to not implement the whole driver as a
> >>> mailbox controller? For me, this driver looks like a good fit.
> >>
> >>
> >> CMDQ needs to encode commands for GCE hardware. We think this behavior
> >> should be put in CMDQ driver, and client just call CMDQ functions.
> >> Therefore, if we want to use mailbox framework, cmdq_rec must be
> >> mailbox client, and the others must be mailbox controller.
> >>
> >
> > You mean the functions to fill the cmdq_rec and execute it?
> > I think this should be part of the driver.
> >
> > Jassi, can you have a look on the interface this driver exports [0].
> > They are needed to actually create the message which will be send.
> > Could something like this be part of a mailbox driver?
> >
> > [0] https://patchwork.kernel.org/patch/9140221/
> >
> Packet creating/parsing should not be a part of controller driver. As
> the log of this patch says, today it is used for only display but in
> future it could work with other h/w as well, so it makes sense to have
> mailbox api do the message queuing, the controller driver do the
> send/receive and client drivers implement display and other h/w
> specific packaging of data (protocol handling).
> 
> So yes, I think this could use mailbox api.
> 
> Cheers.

I will try to rewrite CMDQ driver by mailbox framework and discuss with
you if I encounter other problems.

Thanks,
HS

Matthias Brugger June 7, 2016, 4:59 p.m. UTC | #15

On 03/06/16 15:11, Matthias Brugger wrote:
>
>
[...]

>>>>>>>>>> +
>>>>>>>>>> +            smp_mb(); /* modify jump before enable thread */
>>>>>>>>>> +        }
>>>>>>>>>> +
>>>>>>>>>> +        cmdq_thread_writel(thread, task->pa_base +
>>>>>>>>>> task->command_size,
>>>>>>>>>> +                   CMDQ_THR_END_ADDR);
>>>>>>>>>> +        cmdq_thread_resume(thread);
>>>>>>>>>> +    }
>>>>>>>>>> +    list_move_tail(&task->list_entry, &thread->task_busy_list);
>>>>>>>>>> +    spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>>>>>>> +}
>>>>>>>>>> +
>>>>>>>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
>>>>>>>>>> +                   struct cmdq_thread *thread, u32 irq_flag)
>>>>>>>>>> +{
>>>>>>>>>> +    struct cmdq_task *task, *tmp, *curr_task = NULL;
>>>>>>>>>> +    u32 curr_pa;
>>>>>>>>>> +    struct cmdq_cb_data cmdq_cb_data;
>>>>>>>>>> +    bool err;
>>>>>>>>>> +
>>>>>>>>>> +    if (irq_flag & CMDQ_THR_IRQ_ERROR)
>>>>>>>>>> +        err = true;
>>>>>>>>>> +    else if (irq_flag & CMDQ_THR_IRQ_DONE)
>>>>>>>>>> +        err = false;
>>>>>>>>>> +    else
>>>>>>>>>> +        return;
>>>>>>>>>> +
>>>>>>>>>> +    curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
>>>>>>>>>> +
>>>>>>>>>> +    list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
>>>>>>>>>> +                 list_entry) {
>>>>>>>>>> +        if (curr_pa >= task->pa_base &&
>>>>>>>>>> +            curr_pa < (task->pa_base + task->command_size))
>>>>>>>>>
>>>>>>>>> What are you checking here? It seems as if you make some implcit
>>>>>>>>> assumptions about pa_base and the order of execution of
>>>>>>>>> commands in the
>>>>>>>>> thread. Is it save to do so? Does dma_alloc_coherent give any
>>>>>>>>> guarantees
>>>>>>>>> about dma_handle?
>>>>>>>>
>>>>>>>> 1. Check what is the current running task in this GCE thread.
>>>>>>>> 2. Yes.
>>>>>>>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
>>>>>>>>
>>>>>>>
>>>>>>> Yes, physical addresses might be continous, but AFAIK there is no
>>>>>>> guarantee that the dma_handle address is steadily growing, when
>>>>>>> calling
>>>>>>> dma_alloc_coherent. And if I understand the code correctly, you
>>>>>>> use this
>>>>>>> assumption to decide if the task picked from task_busy_list is
>>>>>>> currently
>>>>>>> executing. So I think this mecanism is not working.
>>>>>>
>>>>>> I don't use dma_handle address, and just use physical addresses.
>>>>>>    From CPU's point of view, tasks are linked by the busy list.
>>>>>>    From GCE's point of view, tasks are linked by the JUMP command.
>>>>>>
>>>>>>> In which cases does the HW thread raise an interrupt.
>>>>>>> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
>>>>>>
>>>>>> GCE will raise interrupt if any task is done or error.
>>>>>> However, GCE is fast, so CPU may get multiple done tasks
>>>>>> when it is running ISR.
>>>>>>
>>>>>> In case of error, that GCE thread will pause and raise interrupt.
>>>>>> So, CPU may get multiple done tasks and one error task.
>>>>>>
>>>>>
>>>>> I think we should reimplement the ISR mechanism. Can't we just read
>>>>> CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave
>>>>> cmdq_handle_error_done to the thread_fn? You will need to pass
>>>>> information from the handler to thread_fn, but that shouldn't be an
>>>>> issue. AFAIK interrupts are disabled in the handler, so we should stay
>>>>> there as short as possible. Traversing task_busy_list is expensive, so
>>>>> we need to do it in a thread context.
>>>>
>>>> Actually, our initial implementation is similar to your suggestion,
>>>> but display needs CMDQ to return callback function very precisely,
>>>> else display will drop frame.
>>>> For display, CMDQ interrupt will be raised every 16 ~ 17 ms,
>>>> and CMDQ needs to call callback function in ISR.
>>>> If we defer callback to workqueue, the time interval may be larger than
>>>> 32 ms.sometimes.
>>>>
>>>
>>> I think the problem is, that you implemented the workqueue as a ordered
>>> workqueue, so there is no parallel processing. I'm still not sure why
>>> you need the workqueue to be ordered. Can you please explain.
>>
>> The order should be kept.
>> Let me use mouse cursor as an example.
>> If task 1 means move mouse cursor to point A, task 2 means point B,
>> and task 3 means point C, our expected result is A -> B -> C.
>> If the order is not kept, the result could become A -> C -> B.
>>
>
> Got it, thanks for the clarification.
>

I think a way to get rid of the workqueue is to use a timer, which gets 
programmed to the time a timeout in the first task in the busy list 
would happen. Everytime we update the busy list (e.g. because of task 
got finished by the thread), we update the timer. When the timer 
triggers, which hopefully won't happen too often, we return timeout on 
the busy list elements, until the time is lower then the actual time.

At least with this we can reduce the data structures in this driver and 
make it more lightweight.

Best regards,
Matthias

>>>>> I keep thinking about how to get rid of the two data structures,
>>>>> task_busy_list and the task_release_wq. We need the latter for the
>>>>> only
>>>>> sake of getting a timeout.
>>>>>
>>>>> Did you have a look on how the mailbox framework handles this?
>>>>> By the way, what is the reason to not implement the whole driver as a
>>>>> mailbox controller? For me, this driver looks like a good fit.
>>>>
>>>> CMDQ needs to encode commands for GCE hardware. We think this behavior
>>>> should be put in CMDQ driver, and client just call CMDQ functions.
>>>> Therefore, if we want to use mailbox framework, cmdq_rec must be
>>>> mailbox client, and the others must be mailbox controller.
>>>>
>>>
>>> You mean the functions to fill the cmdq_rec and execute it?
>>> I think this should be part of the driver.
>>
>> Yes.
>>
>>> Jassi, can you have a look on the interface this driver exports [0].
>>> They are needed to actually create the message which will be send.
>>> Could something like this be part of a mailbox driver?
>>>
>>> [0] https://patchwork.kernel.org/patch/9140221/
>>>
>>>> However, if we use mailbox controller, CMDQ driver still needs to
>>>> control busy list for each GCE thread, and use workqueue to handle
>>>> timeout tasks.
>>>>
>>>
>>> Let me summarize my ideas around this driver:
>>> When we enter the ISR, we know that all task in task_busy_list before
>>> the entry which represents curr_task can be set to TASK_STATE_DONE.
>>> The curr_task could be TASK_STATE_ERROR if the corresponding bit in the
>>> irq status registers is set.
>>> Do we need to call the callback in the same order as the tasks got
>>> dispatched to the HW thread? If not, we could try to call all this
>>> callbacks in a multithreaded workqueue.
>>
>> Yes, we should keep order.
>>
>>> Regards,
>>> Matthias
>>
>> Thanks,
>> HS
>>
>>>> The only thing that we can borrow from mailbox framework is the send
>>>> (CMDQ flush) and receive (CMDQ callback) interface, However, we don't
>>>> think we can gain many benefits from it, and we have some overheads to
>>>> conform to mailbox interface.
>>>>
>>>>
>>>>>
>>>>>>>>>> +            curr_task = task;
>>>>>>>>>> +        if (task->cb.cb) {
>>>>>>>>>> +            cmdq_cb_data.err = curr_task ? err : false;
>>>>>>>>>> +            cmdq_cb_data.data = task->cb.data;
>>>>>>>>>> +            task->cb.cb(cmdq_cb_data);
>>>>>>>>>> +        }
>>>>>>>>>> +        task->task_state = (curr_task && err) ?
>>>>>>>>>> TASK_STATE_ERROR :
>>>>>>>>>> +                TASK_STATE_DONE;
>>>>>>>>>> +        list_del(&task->list_entry);
>>>>>>>>>> +        if (curr_task)
>>>>>>>>>> +            break;
>>>>>>>>>> +    }
>>>>>>>>>> +
>>>>>>>>>> +    wake_up(&thread->wait_task_done);
>>>>>>>>>> +}
>>>>>>>>>> +
>>>>>>>>>> +static void cmdq_thread_irq_handler(struct cmdq *cmdq, int tid)
>>>>>>>>>> +{
>>>>>>>>>> +    struct cmdq_thread *thread = &cmdq->thread[tid];
>>>>>>>>>> +    unsigned long flags = 0L;
>>>>>>>>>> +    u32 irq_flag;
>>>>>>>>>> +
>>>>>>>>>> +    spin_lock_irqsave(&cmdq->exec_lock, flags);
>>>>>>>>>> +
>>>>>>>>>> +    irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
>>>>>>>>>> +    cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
>>>>>>>>>> +
>>>>>>>>>> +    /*
>>>>>>>>>> +     * Another CPU core could run "release task" right before
>>>>>>>>>> we acquire
>>>>>>>>>> +     * the spin lock, and thus reset / disable this GCE
>>>>>>>>>> thread, so we
>>>>>>>>>> +     * need to check the enable bit of this GCE thread.
>>>>>>>>>> +     */
>>>>>>>>>> +    if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
>>>>>>>>>> +          CMDQ_THR_ENABLED))
>>>>>>>>>> +        irq_flag = 0;
>>>>>>>>>
>>>>>>>>> cmdq_handle_error_done just retuns in this case. Programming
>>>>>>>>> this way
>>>>>>>>> just makes things confusing. What about:
>>>>>>>>>
>>>>>>>>> if (cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
>>>>>>>>>           CMDQ_THR_ENABLED)
>>>>>>>>>     cmdq_handle_error_done(cmdq, thread, irq_flag);
>>>>>>>>> else
>>>>>>>>>         irq_flag = 0;
>>>>>>>>>
>>>>>>>>> spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>>>>>
>>>>>>>> We still need to clear irq_flag if GCE thread is disabled.
>>>>>>>> So, I think we can just return here.
>>>>>>>>
>>>>>>>>     if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
>>>>>>>>           CMDQ_THR_ENABLED))
>>>>>>>>         return;
>>>>>>>>
>>>>>>>> What do you think?
>>>>>>>>
>>>>>>>
>>>>>>> No, you can't just return, you need to unlock the spinlock.
>>>>>>> Anyway I would prefer it the other way round, as I put it in my last
>>>>>>> mail. Just delete the else branch, we don't need to set irq_flag
>>>>>>> to zero.
>>>>>>
>>>>>> Sorry for my previous wrong reply since I merge your comment
>>>>>> and CK's comment.
>>>>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-May/005547.html
>>>>>>
>>>>>> So, I will put this if condition into cmdq_task_handle_error_result()
>>>>>> and then just return it if GCE thread is disabled.
>>>>>>
>>>>>
>>>>> You mean in cmdq_task_handle_done
>>>>> We should rename this functions to not create confusion.
>>>>
>>>> Sorry again. I mean in cmdq_handle_error_done().
>>>> This function handles both done and error.
>>>>
>>>> I agree the function name looks confusion.
>>>> I think it can be renamed to cmdq_thread_irq_handler()
>>>> since it is used to handle irq for GCE thread.
>>>>
>>>>> Regards,
>>>>> Matthias
>>>>
>>>> Thanks,
>>>> HS
>>>>
>>>>>>>>>> +
>>>>>>>>>> +    cmdq_handle_error_done(cmdq, thread, irq_flag);
>>>>>>>>>> +    spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>>>>>>> +}
>>>>>>>>>> +
>>>>>>>>>> +static irqreturn_t cmdq_irq_handler(int irq, void *dev)
>>>>>>>>>> +{
>>>>>>>>>> +    struct cmdq *cmdq = dev;
>>>>>>>>>> +    u32 irq_status;
>>>>>>>>>> +    int i;
>>>>>>>>>> +
>>>>>>>>>> +    irq_status = readl(cmdq->base + CMDQ_CURR_IRQ_STATUS);
>>>>>>>>>> +    irq_status &= CMDQ_IRQ_MASK;
>>>>>>>>>> +    irq_status ^= CMDQ_IRQ_MASK;
>>>>>>>>>
>>>>>>>>> irq_status can be much bigger then 3, which is the number of
>>>>>>>>> threads in
>>>>>>>>> the system (CMDQ_THR_MAX_COUNT). So why we use this mask here
>>>>>>>>> isn't
>>>>>>>>> clear to me.
>>>>>>>>
>>>>>>>> Our GCE hardware has 16 threads, but we only use 3 threads
>>>>>>>> currently.
>>>>>>>>
>>>>>>>
>>>>>>> Ok, but please use bitops here.
>>>>>>
>>>>>> Will use bitops.
>>>>>>
>>>>>>>>>> +
>>>>>>>>>> +    if (!irq_status)
>>>>>>>>>> +        return IRQ_NONE;
>>>>>>>>>> +
>>>>>>>>>> +    while (irq_status) {
>>>>>>>>>> +        i = ffs(irq_status) - 1;
>>>>>>>>>> +        irq_status &= ~BIT(i);
>>>>>>>>>> +        cmdq_thread_irq_handler(cmdq, i);
>>>>>>>>>> +    }
>>>>>>>>>
>>>>>>>>> Can you explain how the irq status register looks like, that
>>>>>>>>> would it
>>>>>>>>> make much easier to understand what happens here.
>>>>>>>>
>>>>>>>> Bit 0 ~ 15 of irq status register represents GCE thread 0 ~ 15
>>>>>>>> interrupt. 0 means asserting interrupt; 1 means no interrupt.
>>>>>>>>
>>>>>>>
>>>>>>> Thanks, that helped. :)
>>>>>>>
>>>>>>>>>> +
>>>>>>>>>> +    return IRQ_HANDLED;
>>>>>>>>>> +}
>>>>>>>>>> +
>>>>>>>>>> +static int cmdq_task_handle_error_result(struct cmdq_task *task)
>>>>>>>>>
>>>>>>>>> We never check the return values, why do we have them?
>>>>>>>>
>>>>>>>> Will drop return value.
>>>>>>>>
>>>>>>>>>> +{
>>>>>>>>>> +    struct cmdq *cmdq = task->cmdq;
>>>>>>>>>> +    struct device *dev = cmdq->dev;
>>>>>>>>>> +    struct cmdq_thread *thread = task->thread;
>>>>>>>>>> +    struct cmdq_task *next_task, *prev_task;
>>>>>>>>>> +    u32 irq_flag;
>>>>>>>>>> +
>>>>>>>>>> +    /* suspend GCE thread to ensure consistency */
>>>>>>>>>> +    WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
>>>>>>>>>> +
>>>>>>>>>> +    /* ISR has handled this error task */
>>>>>>>>>> +    if (task->task_state == TASK_STATE_ERROR) {
>>>>>>>>>> +        next_task =
>>>>>>>>>> list_first_entry_or_null(&thread->task_busy_list,
>>>>>>>>>> +                struct cmdq_task, list_entry);
>>>>>>>>>> +        if (next_task) /* move to next task */
>>>>>>>>>> +            cmdq_thread_writel(thread, next_task->pa_base,
>>>>>>>>>> +                       CMDQ_THR_CURR_ADDR);
>>>>>>>>>
>>>>>>>>> We have to do this, as we suppose that the thread did not reach
>>>>>>>>> the jump
>>>>>>>>> instruction we put into it's command queue, right?
>>>>>>>>
>>>>>>>> Yes.
>>>>>>>>
>>>>>>>
>>>>>>> So this should then go into it's own function. In wait_release_work,
>>>>>>> something like this:
>>>>>>>
>>>>>>> if(task->task_state == TASK_STATE_ERROR)
>>>>>>>          cmdq_task_handle_error(task)
>>>>>>
>>>>>> OK.
>>>>>> I will write new function cmdq_task_handle_error() to handle error
>>>>>> case.
>>>>>>
>>>>>>>>>> +        cmdq_thread_resume(thread);
>>>>>>>>>> +        return -ECANCELED;
>>>>>>>>>> +    }
>>>>>>>>>> +
>>>>>>>>>
>>>>>>>>> if task_state != ERROR and != DONE. This means that the timeout of
>>>>>>>>> task_release_wq has timed out, right?
>>>>>>>>
>>>>>>>> Yes.
>>>>>>>>
>>>>>>>>>> +    /*
>>>>>>>>>> +     * Save next_task and prev_task in advance
>>>>>>>>>> +     * since cmdq_handle_error_done will remove list_entry.
>>>>>>>>>> +     */
>>>>>>>>>> +    next_task = prev_task = NULL;
>>>>>>>>>> +    if (task->list_entry.next != &thread->task_busy_list)
>>>>>>>>>> +        next_task = list_next_entry(task, list_entry);
>>>>>>>>>> +    if (task->list_entry.prev != &thread->task_busy_list)
>>>>>>>>>> +        prev_task = list_prev_entry(task, list_entry);
>>>>>>>>>> +
>>>>>>>>>> +    /*
>>>>>>>>>> +     * Although IRQ is disabled, GCE continues to execute.
>>>>>>>>>> +     * It may have pending IRQ before GCE thread is suspended,
>>>>>>>>>> +     * so check this condition again.
>>>>>>>>>> +     */
>>>>>>>>>
>>>>>>>>> The first thing we did in this function was suspending the
>>>>>>>>> thread. Why
>>>>>>>>> do we need this then?
>>>>>>>>
>>>>>>>> Because timeout is CPU timeout not GCE timeout, GCE could just
>>>>>>>> finish
>>>>>>>> this task before the GCE thread is suspended.
>>>>>>>>
>>>>>>>
>>>>>>> What are the reasons for a timeout? An error has happend, or the
>>>>>>> task is
>>>>>>> still executing.
>>>>>>
>>>>>>    From GCE's point of view, this task is still executing.
>>>>>> But, it could be an error of client.
>>>>>> For example, task may never get event if display turn off hardware
>>>>>> before waiting for task to finish its work.
>>>>>>
>>>>>>>>> To be honest this whole functions looks really like a design
>>>>>>>>> error. We
>>>>>>>>> have to sperate the states much clearer so that it is possible to
>>>>>>>>> understand what is happening in the GCE. Isn't it for example
>>>>>>>>> posible to
>>>>>>>>> have worker queues for timed out tasks and tasks with an error?
>>>>>>>>> I'm not
>>>>>>>>> sure how to do this, actually I'm not sure if I really
>>>>>>>>> understood how
>>>>>>>>> this is supposed to work.
>>>>>>>>
>>>>>>>> GCE doesn't have timeout. The timeout is decided and controlled
>>>>>>>> by CPU,
>>>>>>>> so we check timeout in release work.
>>>>>>>> For error and done, they are easy to check by register, and we have
>>>>>>>> already created release work for timeout. So, I don't think we
>>>>>>>> need to
>>>>>>>> create work queue for each case.
>>>>>>>>
>>>>>>>> What do you think?
>>>>>>>>
>>>>>>>
>>>>>>> I think, if we find in here, that the irq_flag is set, then the the
>>>>>>> interrupt handler was triggered and is spinning the spinlock. If
>>>>>>> this is
>>>>>>> not the case, we have a timeout and we handle this.
>>>>>>
>>>>>> I will write another function to handle error, and handle timeout
>>>>>> here.
>>>>>>
>>>>>>>>> How much do we win, when we patch the thread command queue for
>>>>>>>>> every
>>>>>>>>> task we add, instead of just taking one task after another from
>>>>>>>>> the
>>>>>>>>> task_busy_list?
>>>>>>>>
>>>>>>>> GCE is used to help read/write registers with critical time
>>>>>>>> limitation.
>>>>>>>> Sometimes, client may ask to process multiple tasks in a short
>>>>>>>> period
>>>>>>>> of time, e.g. display flush multiple tasks for next vblank. So,
>>>>>>>> CMDQ
>>>>>>>> shouldn't limit to process one task after another from the
>>>>>>>> task_busy_list. Currently, when interrupt or timeout, we will check
>>>>>>>> how many tasks are done, and which one is error or timeout.
>>>>>>>>
>>>>>>>
>>>>>>> So I suppose the device driver who use this are interested in
>>>>>>> throughput
>>>>>>> and not in latency. The callback of every
>>>>>>>
>>>>>>>>>> +    irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
>>>>>>>>>> +    cmdq_handle_error_done(cmdq, thread, irq_flag);
>>>>>>>>>> +    cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
>>>>>>>>>> +
>>>>>>>>>> +    if (task->task_state == TASK_STATE_DONE) {
>>>>>>>>>> +        cmdq_thread_resume(thread);
>>>>>>>>>> +        return 0;
>>>>>>>>>> +    }
>>>>>>>>>> +
>>>>>>>>>> +    if (task->task_state == TASK_STATE_ERROR) {
>>>>>>>>>> +        dev_err(dev, "task 0x%p error\n", task);
>>>>>>>>>> +        if (next_task) /* move to next task */
>>>>>>>>>> +            cmdq_thread_writel(thread, next_task->pa_base,
>>>>>>>>>> +                       CMDQ_THR_CURR_ADDR);
>>>>>>>>>> +        cmdq_thread_resume(thread);
>>>>>>>>>> +        return -ECANCELED;
>>>>>>>>>> +    }
>>>>>>>>>> +
>>>>>>>>>> +    /* Task is running, so we force to remove it. */
>>>>>>>>>> +    dev_err(dev, "task 0x%p timeout or killed\n", task);
>>>>>>>>>> +    task->task_state = TASK_STATE_ERROR;
>>>>>>>>>> +
>>>>>>>>>> +    if (prev_task) {
>>>>>>>>>> +        u64 *prev_va = prev_task->va_base;
>>>>>>>>>> +        u64 *curr_va = task->va_base;
>>>>>>>>>> +
>>>>>>>>>> +        /* copy JUMP instruction */
>>>>>>>>>> +        prev_va[prev_task->num_cmd - 1] =
>>>>>>>>>> curr_va[task->num_cmd - 1];
>>>>>>>>>> +
>>>>>>>>>> +        cmdq_thread_invalidate_fetched_data(thread);
>>>>>>>>>> +    } else if (next_task) { /* move to next task */
>>>>>>>>>> +        cmdq_thread_writel(thread, next_task->pa_base,
>>>>>>>>>> +                   CMDQ_THR_CURR_ADDR);
>>>>>>>>>> +    }
>>>>>>>>>> +
>>>>>>>>>> +    list_del(&task->list_entry);
>>>>>>>>>> +    cmdq_thread_resume(thread);
>>>>>>>>>> +
>>>>>>>>>> +    /* call cb here to prevent lock */
>>>>>>>>>> +    if (task->cb.cb) {
>>>>>>>>>> +        struct cmdq_cb_data cmdq_cb_data;
>>>>>>>>>> +
>>>>>>>>>> +        cmdq_cb_data.err = true;
>>>>>>>>>> +        cmdq_cb_data.data = task->cb.data;
>>>>>>>>>> +        task->cb.cb(cmdq_cb_data);
>>>>>>>>>> +    }
>>>>>>>>>> +
>>>>>>>>>> +    return -ECANCELED;
>>>>>>>>>> +}
>>>>>>>>>> +
>>>>>>>>>> +static void cmdq_task_wait_release_work(struct work_struct
>>>>>>>>>> *work_item)
>>>>>>>>>> +{
>>>>>>>>>> +    struct cmdq_task *task = container_of(work_item, struct
>>>>>>>>>> cmdq_task,
>>>>>>>>>> +                          release_work);
>>>>>>>>>> +    struct cmdq *cmdq = task->cmdq;
>>>>>>>>>> +    struct cmdq_thread *thread = task->thread;
>>>>>>>>>> +    unsigned long flags;
>>>>>>>>>> +
>>>>>>>>>> +    wait_event_timeout(thread->wait_task_done,
>>>>>>>>>> +               task->task_state != TASK_STATE_BUSY,
>>>>>>>>>> +               msecs_to_jiffies(CMDQ_TIMEOUT_MS));
>>>>>>>>>> +
>>>>>>>>>> +    spin_lock_irqsave(&cmdq->exec_lock, flags);
>>>>>>>>>> +    if (task->task_state != TASK_STATE_DONE)
>>>>>>>>>> +        cmdq_task_handle_error_result(task);
>>>>>>>>>> +    if (list_empty(&thread->task_busy_list))
>>>>>>>>>> +        cmdq_thread_disable(cmdq, thread);
>>>>>>>>>> +    spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>>>>>>> +
>>>>>>>>>> +    /* release regardless of success or not */
>>>>>>>>>> +    clk_disable_unprepare(cmdq->clock);
>>>>>>>>>> +    cmdq_task_release(task);
>>>>>>>>>> +}
>>>>>>>>>> +
>>>>>>>>>> +static void cmdq_task_wait_release_schedule(struct cmdq_task
>>>>>>>>>> *task)
>>>>>>>>>> +{
>>>>>>>>>> +    struct cmdq *cmdq = task->cmdq;
>>>>>>>>>> +
>>>>>>>>>> +    INIT_WORK(&task->release_work, cmdq_task_wait_release_work);
>>>>>>>>>> +    queue_work(cmdq->task_release_wq, &task->release_work);
>>>>>>>>>> +}
>>>>>>>>>> +
>>>>>>>>>> +static int cmdq_rec_realloc_cmd_buffer(struct cmdq_rec *rec,
>>>>>>>>>> size_t size)
>>>>>>>>>> +{
>>>>>>>>>> +    void *new_buf;
>>>>>>>>>> +
>>>>>>>>>> +    new_buf = krealloc(rec->buf, size, GFP_KERNEL | __GFP_ZERO);
>>>>>>>>>> +    if (!new_buf)
>>>>>>>>>> +        return -ENOMEM;
>>>>>>>>>> +    rec->buf = new_buf;
>>>>>>>>>> +    rec->buf_size = size;
>>>>>>>>>> +    return 0;
>>>>>>>>>> +}
>>>>>>>>>> +
>>>>>>>>>> +struct cmdq_base *cmdq_register_device(struct device *dev)
>>>>>>>>>> +{
>>>>>>>>>> +    struct cmdq_base *cmdq_base;
>>>>>>>>>> +    struct resource res;
>>>>>>>>>> +    int subsys;
>>>>>>>>>> +    u32 base;
>>>>>>>>>> +
>>>>>>>>>> +    if (of_address_to_resource(dev->of_node, 0, &res))
>>>>>>>>>> +        return NULL;
>>>>>>>>>> +    base = (u32)res.start;
>>>>>>>>>> +
>>>>>>>>>> +    subsys = cmdq_subsys_base_to_id(base >> 16);
>>>>>>>>>> +    if (subsys < 0)
>>>>>>>>>> +        return NULL;
>>>>>>>>>> +
>>>>>>>>>> +    cmdq_base = devm_kmalloc(dev, sizeof(*cmdq_base),
>>>>>>>>>> GFP_KERNEL);
>>>>>>>>>> +    if (!cmdq_base)
>>>>>>>>>> +        return NULL;
>>>>>>>>>> +    cmdq_base->subsys = subsys;
>>>>>>>>>> +    cmdq_base->base = base;
>>>>>>>>>> +
>>>>>>>>>> +    return cmdq_base;
>>>>>>>>>> +}
>>>>>>>>>> +EXPORT_SYMBOL(cmdq_register_device);
>>>>>>>>>> +
>>>>>>>>>> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
>>>>>>>>>> +            struct cmdq_rec **rec_ptr)
>>>>>>>>>> +{
>>>>>>>>>> +    struct cmdq_rec *rec;
>>>>>>>>>> +    int err;
>>>>>>>>>> +
>>>>>>>>>> +    rec = kzalloc(sizeof(*rec), GFP_KERNEL);
>>>>>>>>>> +    if (!rec)
>>>>>>>>>> +        return -ENOMEM;
>>>>>>>>>> +    rec->cmdq = dev_get_drvdata(dev);
>>>>>>>>>> +    rec->engine_flag = engine_flag;
>>>>>>>>>> +    err = cmdq_rec_realloc_cmd_buffer(rec,
>>>>>>>>>> CMDQ_INITIAL_CMD_BLOCK_SIZE);
>>>>>>>>>
>>>>>>>>> Just pass PAGE_SIZE here, no need for CMDQ_INITIAL_CMD_BLOCK_SIZE.
>>>>>>>>
>>>>>>>> Will do.
>>>>>>>>
>>>>>>>>>> +    if (err < 0) {
>>>>>>>>>> +        kfree(rec);
>>>>>>>>>> +        return err;
>>>>>>>>>> +    }
>>>>>>>>>> +    *rec_ptr = rec;
>>>>>>>>>> +    return 0;
>>>>>>>>>> +}
>>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_create);
>>>>>>>>>> +
>>>>>>>>>> +static int cmdq_rec_append_command(struct cmdq_rec *rec, enum
>>>>>>>>>> cmdq_code code,
>>>>>>>>>> +                   u32 arg_a, u32 arg_b)
>>>>>>>>>> +{
>>>>>>>>>> +    u64 *cmd_ptr;
>>>>>>>>>> +    int err;
>>>>>>>>>> +
>>>>>>>>>> +    if (WARN_ON(rec->finalized))
>>>>>>>>>> +        return -EBUSY;
>>>>>>>>>> +    if (code < CMDQ_CODE_MOVE || code > CMDQ_CODE_EOC)
>>>>>>>>>> +        return -EINVAL;
>>>>>>>>>
>>>>>>>>> cmdq_rec_append_command is just called from inside this driver
>>>>>>>>> and code
>>>>>>>>> is a enum. We can expect it to be correct, no need for this check.
>>>>>>>>
>>>>>>>> Will drop this check.
>>>>>>>>
>>>>>>>>>> +    if (unlikely(rec->command_size + CMDQ_INST_SIZE >
>>>>>>>>>> rec->buf_size)) {
>>>>>>>>>
>>>>>>>>> command_size is the offset into the buffer to which a new
>>>>>>>>> command is
>>>>>>>>> written, so this name is highly confusing. I wonder if this
>>>>>>>>> would be
>>>>>>>>> easier to understand if we redefine command_size to something
>>>>>>>>> like the
>>>>>>>>> number of commands and divide/multiply CMDQ_INST_SIZE where
>>>>>>>>> this is needed.
>>>>>>>>
>>>>>>>> I can rename command_size to cmd_buf_size and calculate num_cmd by
>>>>>>>> dividing CMDQ_INST_SIZE.
>>>>>>>> What do you think?
>>>>>>>>
>>>>>>>>>> +        err = cmdq_rec_realloc_cmd_buffer(rec, rec->buf_size
>>>>>>>>>> * 2);
>>>>>>>>>> +        if (err < 0)
>>>>>>>>>> +            return err;
>>>>>>>>>> +    }
>>>>>>>>>> +    cmd_ptr = rec->buf + rec->command_size;
>>>>>>>>>> +    (*cmd_ptr) = (u64)((code << CMDQ_OP_CODE_SHIFT) | arg_a)
>>>>>>>>>> << 32 | arg_b;
>>>>>>>>>> +    rec->command_size += CMDQ_INST_SIZE;
>>>>>>>>>> +    return 0;
>>>>>>>>>> +}
>>>>>>>>>> +
>>>>>>>>>> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value, struct
>>>>>>>>>> cmdq_base *base,
>>>>>>>>>> +           u32 offset)
>>>>>>>>>> +{
>>>>>>>>>> +    u32 arg_a = ((base->base + offset) &
>>>>>>>>>> CMDQ_ARG_A_WRITE_MASK) |
>>>>>>>>>> +            ((base->subsys & CMDQ_SUBSYS_MASK) <<
>>>>>>>>>> CMDQ_SUBSYS_SHIFT);
>>>>>>>>>
>>>>>>>>> base->subsys is the id from gce_sybsys, so we can expect it to be
>>>>>>>>> correct, no need to mask with CMDQ_SUBSYS_MASK.
>>>>>>>>
>>>>>>>> Will drop it.
>>>>>>>>
>>>>>>>>>> +    return cmdq_rec_append_command(rec, CMDQ_CODE_WRITE,
>>>>>>>>>> arg_a, value);
>>>>>>>>>> +}
>>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_write);
>>>>>>>>>> +
>>>>>>>>>> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
>>>>>>>>>> +            struct cmdq_base *base, u32 offset, u32 mask)
>>>>>>>>>> +{
>>>>>>>>>> +    u32 offset_mask = offset;
>>>>>>>>>> +    int err;
>>>>>>>>>> +
>>>>>>>>>> +    if (mask != 0xffffffff) {
>>>>>>>>>> +        err = cmdq_rec_append_command(rec, CMDQ_CODE_MOVE, 0,
>>>>>>>>>> ~mask);
>>>>>>>>>> +        if (err < 0)
>>>>>>>>>> +            return err;
>>>>>>>>>> +        offset_mask |= CMDQ_WRITE_ENABLE_MASK;
>>>>>>>>>> +    }
>>>>>>>>>> +    return cmdq_rec_write(rec, value, base, offset_mask);
>>>>>>>>>> +}
>>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_write_mask);
>>>>>>>>>> +
>>>>>>>>>> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event)
>>>>>>>>>> +{
>>>>>>>>>> +    u32 arg_b;
>>>>>>>>>> +
>>>>>>>>>> +    if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
>>>>>>>>>> +        return -EINVAL;
>>>>>>>>>> +
>>>>>>>>>> +    /*
>>>>>>>>>> +     * bit 0-11: wait value
>>>>>>>>>> +     * bit 15: 1 - wait, 0 - no wait
>>>>>>>>>> +     * bit 16-27: update value
>>>>>>>>>> +     * bit 31: 1 - update, 0 - no update
>>>>>>>>>> +     */
>>>>>>>>>
>>>>>>>>> I don't understand this comments. What are they for?
>>>>>>>>
>>>>>>>> This is for WFE command. I will comment it.
>>>>>>>>
>>>>>>>>>> +    arg_b = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT |
>>>>>>>>>> CMDQ_WFE_WAIT_VALUE;
>>>>>>>>>> +    return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event,
>>>>>>>>>> arg_b);
>>>>>>>>>> +}
>>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_wfe);
>>>>>>>>>> +
>>>>>>>>>> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum
>>>>>>>>>> cmdq_event event)
>>>>>>>>>> +{
>>>>>>>>>> +    if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
>>>>>>>>>> +        return -EINVAL;
>>>>>>>>>> +
>>>>>>>>>> +    return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event,
>>>>>>>>>> +                       CMDQ_WFE_UPDATE);
>>>>>>>>>> +}
>>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_clear_event);
>>>>>>>>>> +
>>>>>>>>>> +static int cmdq_rec_finalize(struct cmdq_rec *rec)
>>>>>>>>>> +{
>>>>>>>>>> +    int err;
>>>>>>>>>> +
>>>>>>>>>> +    if (rec->finalized)
>>>>>>>>>> +        return 0;
>>>>>>>>>> +
>>>>>>>>>> +    /* insert EOC and generate IRQ for each command iteration */
>>>>>>>>>> +    err = cmdq_rec_append_command(rec, CMDQ_CODE_EOC, 0,
>>>>>>>>>> CMDQ_EOC_IRQ_EN);
>>>>>>>>>> +    if (err < 0)
>>>>>>>>>> +        return err;
>>>>>>>>>> +
>>>>>>>>>> +    /* JUMP to end */
>>>>>>>>>> +    err = cmdq_rec_append_command(rec, CMDQ_CODE_JUMP, 0,
>>>>>>>>>> CMDQ_JUMP_PASS);
>>>>>>>>>> +    if (err < 0)
>>>>>>>>>> +        return err;
>>>>>>>>>> +
>>>>>>>>>
>>>>>>>>> Does this need to be atomic?
>>>>>>>>> What happens if after CODE_EOC and before CODE_JUMP some
>>>>>>>>> write/read/event gets added?
>>>>>>>>> What happens if more commands get added to the queue after
>>>>>>>>> CODE_JUMP,
>>>>>>>>> but before finalized is set to true. Why don't you use atomic
>>>>>>>>> functions
>>>>>>>>> to access finalized?
>>>>>>>>
>>>>>>>> Since cmdq_rec doesn't guarantee thread safe, mutex is needed when
>>>>>>>> client uses cmdq_rec.
>>>>>>>>
>>>>>>>
>>>>>>> Well I think that rec->finalized tries to implement this, but might
>>>>>>> fail, if two kernel threads work on the same cmdq_rec.
>>>>>>>
>>>>>>>>>> +    rec->finalized = true;
>>>>>>>>>> +    return 0;
>>>>>>>>>> +}
>>>>>>>>>> +
>>>>>>>>>> +int cmdq_rec_flush_async(struct cmdq_rec *rec,
>>>>>>>>>> cmdq_async_flush_cb cb,
>>>>>>>>>> +             void *data)
>>>>>>>>>> +{
>>>>>>>>>> +    struct cmdq *cmdq = rec->cmdq;
>>>>>>>>>> +    struct cmdq_task *task;
>>>>>>>>>> +    struct cmdq_task_cb task_cb;
>>>>>>>>>> +    struct cmdq_thread *thread;
>>>>>>>>>> +    int err;
>>>>>>>>>> +
>>>>>>>>>> +    mutex_lock(&cmdq->task_mutex);
>>>>>>>>>> +    if (cmdq->suspended) {
>>>>>>>>>> +        dev_err(cmdq->dev, "%s is called after suspended\n",
>>>>>>>>>> __func__);
>>>>>>>>>> +        mutex_unlock(&cmdq->task_mutex);
>>>>>>>>>> +        return -EPERM;
>>>>>>>>>> +    }
>>>>>>>>>> +
>>>>>>>>>> +    err = cmdq_rec_finalize(rec);
>>>>>>>>>> +    if (err < 0) {
>>>>>>>>>> +        mutex_unlock(&cmdq->task_mutex);
>>>>>>>>>> +        return err;
>>>>>>>>>> +    }
>>>>>>>>>> +
>>>>>>>>>> +    task_cb.cb = cb;
>>>>>>>>>> +    task_cb.data = data;
>>>>>>>>>> +    task = cmdq_task_acquire(rec, task_cb);
>>>>>>>>>> +    if (!task) {
>>>>>>>>>> +        mutex_unlock(&cmdq->task_mutex);
>>>>>>>>>> +        return -EFAULT;
>>>>>>>>>> +    }
>>>>>>>>>> +
>>>>>>>>>> +    thread =
>>>>>>>>>> &cmdq->thread[cmdq_eng_get_thread(task->engine_flag)];
>>>>>>>>>> +    cmdq_task_exec(task, thread);
>>>>>>>>>> +    cmdq_task_wait_release_schedule(task);
>>>>>>>>>> +    mutex_unlock(&cmdq->task_mutex);
>>>>>>>>>> +    return 0;
>>>>>>>>>> +}
>>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_flush_async);
>>>>>>>>>> +
>>>>>>>>>> +struct cmdq_flush_completion {
>>>>>>>>>> +    struct completion cmplt;
>>>>>>>>>> +    bool err;
>>>>>>>>>> +};
>>>>>>>>>> +
>>>>>>>>>> +static int cmdq_rec_flush_cb(struct cmdq_cb_data data)
>>>>>>>>>> +{
>>>>>>>>>> +    struct cmdq_flush_completion *cmplt = data.data;
>>>>>>>>>> +
>>>>>>>>>> +    cmplt->err = data.err;
>>>>>>>>>> +    complete(&cmplt->cmplt);
>>>>>>>>>> +    return 0;
>>>>>>>>>> +}
>>>>>>>>>> +
>>>>>>>>>> +int cmdq_rec_flush(struct cmdq_rec *rec)
>>>>>>>>>> +{
>>>>>>>>>> +    struct cmdq_flush_completion cmplt;
>>>>>>>>>> +    int err;
>>>>>>>>>> +
>>>>>>>>>> +    init_completion(&cmplt.cmplt);
>>>>>>>>>> +    err = cmdq_rec_flush_async(rec, cmdq_rec_flush_cb, &cmplt);
>>>>>>>>>> +    if (err < 0)
>>>>>>>>>> +        return err;
>>>>>>>>>> +    wait_for_completion(&cmplt.cmplt);
>>>>>>>>>> +    return cmplt.err ? -EFAULT : 0;
>>>>>>>>>> +}
>>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_flush);
>>>>>>>>>> +
>>>>>>>>>> +void cmdq_rec_destroy(struct cmdq_rec *rec)
>>>>>>>>>> +{
>>>>>>>>>> +    kfree(rec->buf);
>>>>>>>>>> +    kfree(rec);
>>>>>>>>>> +}
>>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_destroy);
>>>>>>>>>> +
>>>>>>>>>> +static bool cmdq_task_is_empty(struct cmdq *cmdq)
>>>>>>>>>> +{
>>>>>>>>>> +    struct cmdq_thread *thread;
>>>>>>>>>> +    int i;
>>>>>>>>>> +
>>>>>>>>>> +    for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
>>>>>>>>>> +        thread = &cmdq->thread[i];
>>>>>>>>>> +        if (!list_empty(&thread->task_busy_list))
>>>>>>>>>> +            return false;
>>>>>>>>>> +    }
>>>>>>>>>> +    return true;
>>>>>>>>>> +}
>>>>>>>>>> +
>>>>>>>>>> +static int cmdq_suspend(struct device *dev)
>>>>>>>>>> +{
>>>>>>>>>> +    struct cmdq *cmdq = dev_get_drvdata(dev);
>>>>>>>>>> +    u32 exec_threads;
>>>>>>>>>> +
>>>>>>>>>> +    mutex_lock(&cmdq->task_mutex);
>>>>>>>>>> +    cmdq->suspended = true;
>>>>>>>>>> +    mutex_unlock(&cmdq->task_mutex);
>>>>>>>>>> +
>>>>>>>>>> +    exec_threads = readl(cmdq->base + CMDQ_CURR_LOADED_THR);
>>>>>>>>>> +    if ((exec_threads & CMDQ_THR_EXECUTING) &&
>>>>>>>>>> !cmdq_task_is_empty(cmdq)) {
>>>>>>>>>> +        dev_err(dev, "wait active tasks timeout.\n");
>>>>>>>>>> +        flush_workqueue(cmdq->task_release_wq);
>>>>>>>>>> +    }
>>>>>>>>>> +    return 0;
>>>>>>>>>> +}
>>>>>>>>>> +
>>>>>>>>>> +static int cmdq_resume(struct device *dev)
>>>>>>>>>> +{
>>>>>>>>>> +    struct cmdq *cmdq = dev_get_drvdata(dev);
>>>>>>>>>> +
>>>>>>>>>> +    cmdq->suspended = false;
>>>>>>>>>> +    return 0;
>>>>>>>>>> +}
>>>>>>>>>> +
>>>>>>>>>> +static int cmdq_remove(struct platform_device *pdev)
>>>>>>>>>> +{
>>>>>>>>>> +    struct cmdq *cmdq = platform_get_drvdata(pdev);
>>>>>>>>>> +
>>>>>>>>>> +    destroy_workqueue(cmdq->task_release_wq);
>>>>>>>>>> +    cmdq->task_release_wq = NULL;
>>>>>>>>>> +    return 0;
>>>>>>>>>> +}
>>>>>>>>>> +
>>>>>>>>>> +static int cmdq_probe(struct platform_device *pdev)
>>>>>>>>>> +{
>>>>>>>>>> +    struct device *dev = &pdev->dev;
>>>>>>>>>> +    struct device_node *node = dev->of_node;
>>>>>>>>>> +    struct resource *res;
>>>>>>>>>> +    struct cmdq *cmdq;
>>>>>>>>>> +    int err, i;
>>>>>>>>>> +
>>>>>>>>>> +    cmdq = devm_kzalloc(dev, sizeof(*cmdq), GFP_KERNEL);
>>>>>>>>>> +    if (!cmdq)
>>>>>>>>>> +        return -ENOMEM;
>>>>>>>>>> +    cmdq->dev = dev;
>>>>>>>>>> +
>>>>>>>>>> +    res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
>>>>>>>>>> +    cmdq->base = devm_ioremap_resource(dev, res);
>>>>>>>>>> +    if (IS_ERR(cmdq->base)) {
>>>>>>>>>> +        dev_err(dev, "failed to ioremap gce\n");
>>>>>>>>>> +        return PTR_ERR(cmdq->base);
>>>>>>>>>> +    }
>>>>>>>>>> +
>>>>>>>>>> +    cmdq->irq = irq_of_parse_and_map(node, 0);
>>>>>>>>>> +    if (!cmdq->irq) {
>>>>>>>>>> +        dev_err(dev, "failed to get irq\n");
>>>>>>>>>> +        return -EINVAL;
>>>>>>>>>> +    }
>>>>>>>>>> +
>>>>>>>>>> +    dev_dbg(dev, "cmdq device: addr:0x%p, va:0x%p, irq:%d\n",
>>>>>>>>>> +        dev, cmdq->base, cmdq->irq);
>>>>>>>>>> +
>>>>>>>>>> +    mutex_init(&cmdq->task_mutex);
>>>>>>>>>> +    spin_lock_init(&cmdq->exec_lock);
>>>>>>>>>> +    cmdq->task_release_wq = alloc_ordered_workqueue(
>>>>>>>>>> +            "%s", WQ_MEM_RECLAIM | WQ_HIGHPRI,
>>>>>>>>>> +            "cmdq_task_wait_release");
>>>>>>>>>> +
>>>>>>>>>> +    for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
>>>>>>>>>> +        cmdq->thread[i].base = cmdq->base + CMDQ_THR_BASE +
>>>>>>>>>> +                CMDQ_THR_SHIFT * i;
>>>>>>>>>> +        init_waitqueue_head(&cmdq->thread[i].wait_task_done);
>>>>>>>>>> +        INIT_LIST_HEAD(&cmdq->thread[i].task_busy_list);
>>>>>>>>>> +    }
>>>>>>>>>> +
>>>>>>>>>> +    platform_set_drvdata(pdev, cmdq);
>>>>>>>>>> +
>>>>>>>>>> +    err = devm_request_irq(dev, cmdq->irq, cmdq_irq_handler,
>>>>>>>>>> IRQF_SHARED,
>>>>>>>>>> +                   CMDQ_DRIVER_DEVICE_NAME, cmdq);
>>>>>>>>>> +    if (err < 0) {
>>>>>>>>>> +        dev_err(dev, "failed to register ISR (%d)\n", err);
>>>>>>>>>> +        goto fail;
>>>>>>>>>> +    }
>>>>>>>>>> +
>>>>>>>>>> +    cmdq->clock = devm_clk_get(dev, CMDQ_CLK_NAME);
>>>>>>>>>> +    if (IS_ERR(cmdq->clock)) {
>>>>>>>>>> +        dev_err(dev, "failed to get clk:%s\n", CMDQ_CLK_NAME);
>>>>>>>>>> +        err = PTR_ERR(cmdq->clock);
>>>>>>>>>> +        goto fail;
>>>>>>>>>> +    }
>>>>>>>>>> +    return 0;
>>>>>>>>>> +
>>>>>>>>>> +fail:
>>>>>>>>>> +    cmdq_remove(pdev);
>>>>>>>>>> +    return err;
>>>>>>>>>> +}
>>>>>>>>>> +
>>>>>>>>>> +static const struct dev_pm_ops cmdq_pm_ops = {
>>>>>>>>>> +    .suspend = cmdq_suspend,
>>>>>>>>>> +    .resume = cmdq_resume,
>>>>>>>>>> +};
>>>>>>>>>> +
>>>>>>>>>> +static const struct of_device_id cmdq_of_ids[] = {
>>>>>>>>>> +    {.compatible = "mediatek,mt8173-gce",},
>>>>>>>>>> +    {}
>>>>>>>>>> +};
>>>>>>>>>> +
>>>>>>>>>> +static struct platform_driver cmdq_drv = {
>>>>>>>>>> +    .probe = cmdq_probe,
>>>>>>>>>> +    .remove = cmdq_remove,
>>>>>>>>>> +    .driver = {
>>>>>>>>>> +        .name = CMDQ_DRIVER_DEVICE_NAME,
>>>>>>>>>> +        .owner = THIS_MODULE,
>>>>>>>>>> +        .pm = &cmdq_pm_ops,
>>>>>>>>>> +        .of_match_table = cmdq_of_ids,
>>>>>>>>>> +    }
>>>>>>>>>> +};
>>>>>>>>>> +
>>>>>>>>>> +builtin_platform_driver(cmdq_drv);
>>>>>>>>>> diff --git a/include/soc/mediatek/cmdq.h
>>>>>>>>>> b/include/soc/mediatek/cmdq.h
>>>>>>>>>> new file mode 100644
>>>>>>>>>> index 0000000..60eef3d
>>>>>>>>>> --- /dev/null
>>>>>>>>>> +++ b/include/soc/mediatek/cmdq.h
>>>>>>>>>> @@ -0,0 +1,197 @@
>>>>>>>>>> +/*
>>>>>>>>>> + * Copyright (c) 2015 MediaTek Inc.
>>>>>>>>>> + *
>>>>>>>>>> + * This program is free software; you can redistribute it
>>>>>>>>>> and/or modify
>>>>>>>>>> + * it under the terms of the GNU General Public License
>>>>>>>>>> version 2 as
>>>>>>>>>> + * published by the Free Software Foundation.
>>>>>>>>>> + *
>>>>>>>>>> + * This program is distributed in the hope that it will be
>>>>>>>>>> useful,
>>>>>>>>>> + * but WITHOUT ANY WARRANTY; without even the implied
>>>>>>>>>> warranty of
>>>>>>>>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
>>>>>>>>>> + * GNU General Public License for more details.
>>>>>>>>>> + */
>>>>>>>>>> +
>>>>>>>>>> +#ifndef __MTK_CMDQ_H__
>>>>>>>>>> +#define __MTK_CMDQ_H__
>>>>>>>>>> +
>>>>>>>>>> +#include <linux/platform_device.h>
>>>>>>>>>> +#include <linux/types.h>
>>>>>>>>>> +
>>>>>>>>>> +enum cmdq_eng {
>>>>>>>>>> +    CMDQ_ENG_DISP_AAL,
>>>>>>>>>> +    CMDQ_ENG_DISP_COLOR0,
>>>>>>>>>> +    CMDQ_ENG_DISP_COLOR1,
>>>>>>>>>> +    CMDQ_ENG_DISP_DPI0,
>>>>>>>>>> +    CMDQ_ENG_DISP_DSI0,
>>>>>>>>>> +    CMDQ_ENG_DISP_DSI1,
>>>>>>>>>> +    CMDQ_ENG_DISP_GAMMA,
>>>>>>>>>> +    CMDQ_ENG_DISP_OD,
>>>>>>>>>> +    CMDQ_ENG_DISP_OVL0,
>>>>>>>>>> +    CMDQ_ENG_DISP_OVL1,
>>>>>>>>>> +    CMDQ_ENG_DISP_PWM0,
>>>>>>>>>> +    CMDQ_ENG_DISP_PWM1,
>>>>>>>>>> +    CMDQ_ENG_DISP_RDMA0,
>>>>>>>>>> +    CMDQ_ENG_DISP_RDMA1,
>>>>>>>>>> +    CMDQ_ENG_DISP_RDMA2,
>>>>>>>>>> +    CMDQ_ENG_DISP_UFOE,
>>>>>>>>>> +    CMDQ_ENG_DISP_WDMA0,
>>>>>>>>>> +    CMDQ_ENG_DISP_WDMA1,
>>>>>>>>>> +    CMDQ_ENG_MAX,
>>>>>>>>>> +};
>>>>>>>>>> +
>>>>>>>>>> +/* events for CMDQ and display */
>>>>>>>>>> +enum cmdq_event {
>>>>>>>>>> +    /* Display start of frame(SOF) events */
>>>>>>>>>> +    CMDQ_EVENT_DISP_OVL0_SOF = 11,
>>>>>>>>>> +    CMDQ_EVENT_DISP_OVL1_SOF = 12,
>>>>>>>>>> +    CMDQ_EVENT_DISP_RDMA0_SOF = 13,
>>>>>>>>>> +    CMDQ_EVENT_DISP_RDMA1_SOF = 14,
>>>>>>>>>> +    CMDQ_EVENT_DISP_RDMA2_SOF = 15,
>>>>>>>>>> +    CMDQ_EVENT_DISP_WDMA0_SOF = 16,
>>>>>>>>>> +    CMDQ_EVENT_DISP_WDMA1_SOF = 17,
>>>>>>>>>> +    /* Display end of frame(EOF) events */
>>>>>>>>>> +    CMDQ_EVENT_DISP_OVL0_EOF = 39,
>>>>>>>>>> +    CMDQ_EVENT_DISP_OVL1_EOF = 40,
>>>>>>>>>> +    CMDQ_EVENT_DISP_RDMA0_EOF = 41,
>>>>>>>>>> +    CMDQ_EVENT_DISP_RDMA1_EOF = 42,
>>>>>>>>>> +    CMDQ_EVENT_DISP_RDMA2_EOF = 43,
>>>>>>>>>> +    CMDQ_EVENT_DISP_WDMA0_EOF = 44,
>>>>>>>>>> +    CMDQ_EVENT_DISP_WDMA1_EOF = 45,
>>>>>>>>>> +    /* Mutex end of frame(EOF) events */
>>>>>>>>>> +    CMDQ_EVENT_MUTEX0_STREAM_EOF = 53,
>>>>>>>>>> +    CMDQ_EVENT_MUTEX1_STREAM_EOF = 54,
>>>>>>>>>> +    CMDQ_EVENT_MUTEX2_STREAM_EOF = 55,
>>>>>>>>>> +    CMDQ_EVENT_MUTEX3_STREAM_EOF = 56,
>>>>>>>>>> +    CMDQ_EVENT_MUTEX4_STREAM_EOF = 57,
>>>>>>>>>> +    /* Display underrun events */
>>>>>>>>>> +    CMDQ_EVENT_DISP_RDMA0_UNDERRUN = 63,
>>>>>>>>>> +    CMDQ_EVENT_DISP_RDMA1_UNDERRUN = 64,
>>>>>>>>>> +    CMDQ_EVENT_DISP_RDMA2_UNDERRUN = 65,
>>>>>>>>>> +    /* Keep this at the end of HW events */
>>>>>>>>>> +    CMDQ_MAX_HW_EVENT_COUNT = 260,
>>>>>>>>>> +};
>>>>>>>>>> +
>>>>>>>>>> +struct cmdq_cb_data {
>>>>>>>>>> +    bool    err;
>>>>>>>>>> +    void    *data;
>>>>>>>>>> +};
>>>>>>>>>> +
>>>>>>>>>> +typedef int (*cmdq_async_flush_cb)(struct cmdq_cb_data data);
>>>>>>>>>> +
>>>>>>>>>> +struct cmdq_task;
>>>>>>>>>> +struct cmdq;
>>>>>>>>>> +
>>>>>>>>>> +struct cmdq_rec {
>>>>>>>>>> +    struct cmdq        *cmdq;
>>>>>>>>>> +    u64            engine_flag;
>>>>>>>>>> +    size_t            command_size;
>>>>>>>>>> +    void            *buf;
>>>>>>>>>> +    size_t            buf_size;
>>>>>>>>>> +    bool            finalized;
>>>>>>>>>> +};
>>>>>>>
>>>>>>> Why do we need cmdq_rec at all? Can't we just use the cmdq_task
>>>>>>> for that
>>>>>>> and this way make the driver less complex?
>>>>>>
>>>>>> There are two main reasons for cmdq_rec.
>>>>>> 1. It is slow to access dma too frequently.
>>>>>>       So, we append commands to cacheable memory, and then flush
>>>>>> to dma.
>>>>>> 2. cmdq_rec is not thread safe, but cmdq_task needs thread safe.
>>>>>>       If we merge them, we need to take care of some synchronization
>>>>>>       issues.
>>>>>>
>>>>>>>>>> +
>>>>>>>>>> +struct cmdq_base {
>>>>>>>>>> +    int    subsys;
>>>>>>>>>> +    u32    base;
>>>>>>>>>
>>>>>>>>> subsys can always be calculated via cmdq_subsys_base_to_id(base
>>>>>>>>> >> 16)
>>>>>>>>> so we can get rid of the struct, right?
>>>>>>>>
>>>>>>>> Current subsys method is based on previous comment from Daniel
>>>>>>>> Kurtz.
>>>>>>>> Please take a look of our previous discussion.
>>>>>>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/004483.html
>>>>>>>>
>>>>>>>> Thanks.
>>>>>>>>
>>>>>>>
>>>>>>> I have to look deeper into this, but from what I read, the
>>>>>>> proposal from
>>>>>>> Daniel [1] seems good to me.
>>>>>>>
>>>>>>> [1] https://patchwork.kernel.org/patch/8068311/
>>>>>>
>>>>>> The initial proposal has some problem, so please see the follow-up
>>>>>> discussions. Thanks.
>>>>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/003972.html
>>>>>>
>>>>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/004483.html
>>>>>>
>>>>>>
>>>>>>
>>>>>>>>>> +};
>>>>>>>>>> +
>>>>>>>>>> +/**
>>>>>>>>>> + * cmdq_register_device() - register device which needs CMDQ
>>>>>>>>>> + * @dev:        device
>>>>>>>>>> + *
>>>>>>>>>> + * Return: cmdq_base pointer or NULL for failed
>>>>>>>>>> + */
>>>>>>>>>> +struct cmdq_base *cmdq_register_device(struct device *dev);
>>>>>>>>>> +
>>>>>>>>>> +/**
>>>>>>>>>> + * cmdq_rec_create() - create command queue record
>>>>>>>>>> + * @dev:        device
>>>>>>>>>> + * @engine_flag:    command queue engine flag
>>>>>>>>>> + * @rec_ptr:        command queue record pointer to retrieve
>>>>>>>>>> cmdq_rec
>>>>>>>>>> + *
>>>>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>>>>> + */
>>>>>>>>>> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
>>>>>>>>>> +            struct cmdq_rec **rec_ptr);
>>>>>>>>>> +
>>>>>>>>>> +/**
>>>>>>>>>> + * cmdq_rec_write() - append write command to the command
>>>>>>>>>> queue record
>>>>>>>>>> + * @rec:    the command queue record
>>>>>>>>>> + * @value:    the specified target register value
>>>>>>>>>> + * @base:    the command queue base
>>>>>>>>>> + * @offset:    register offset from module base
>>>>>>>>>> + *
>>>>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>>>>> + */
>>>>>>>>>> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value,
>>>>>>>>>> +           struct cmdq_base *base, u32 offset);
>>>>>>>>>> +
>>>>>>>>>> +/**
>>>>>>>>>> + * cmdq_rec_write_mask() - append write command with mask to
>>>>>>>>>> the command
>>>>>>>>>> + *               queue record
>>>>>>>>>> + * @rec:    the command queue record
>>>>>>>>>> + * @value:    the specified target register value
>>>>>>>>>> + * @base:    the command queue base
>>>>>>>>>> + * @offset:    register offset from module base
>>>>>>>>>> + * @mask:    the specified target register mask
>>>>>>>>>> + *
>>>>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>>>>> + */
>>>>>>>>>> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
>>>>>>>>>> +            struct cmdq_base *base, u32 offset, u32 mask);
>>>>>>>>>> +
>>>>>>>>>> +/**
>>>>>>>>>> + * cmdq_rec_wfe() - append wait for event command to the
>>>>>>>>>> command queue reco    rd
>>>>>>>>>
>>>>>>>>> reco rd -> record
>>>>>>>>
>>>>>>>> Will fix it.
>>>>>>>>
>>>>>>>>> Regards,
>>>>>>>>> Matthias
>>>>>>>>>
>>>>>>>>>> + * @rec:    the command queue record
>>>>>>>>>> + * @event:    the desired event type to "wait and CLEAR"
>>>>>>>>>> + *
>>>>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>>>>> + */
>>>>>>>>>> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event);
>>>>>>>>>> +
>>>>>>>>>> +/**
>>>>>>>>>> + * cmdq_rec_clear_event() - append clear event command to the
>>>>>>>>>> command queue
>>>>>>>>>> + *                record
>>>>>>>>>> + * @rec:    the command queue record
>>>>>>>>>> + * @event:    the desired event to be cleared
>>>>>>>>>> + *
>>>>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>>>>> + */
>>>>>>>>>> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum
>>>>>>>>>> cmdq_event event);
>>>>>>>>>> +
>>>>>>>>>> +/**
>>>>>>>>>> + * cmdq_rec_flush() - trigger CMDQ to execute the recorded
>>>>>>>>>> commands
>>>>>>>>>> + * @rec:    the command queue record
>>>>>>>>>> + *
>>>>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>>>>> + *
>>>>>>>>>> + * Trigger CMDQ to execute the recorded commands. Note that
>>>>>>>>>> this is a
>>>>>>>>>> + * synchronous flush function. When the function returned,
>>>>>>>>>> the recorded
>>>>>>>>>> + * commands have been done.
>>>>>>>>>> + */
>>>>>>>>>> +int cmdq_rec_flush(struct cmdq_rec *rec);
>>>>>>>>>> +
>>>>>>>>>> +/**
>>>>>>>>>> + * cmdq_rec_flush_async() - trigger CMDQ to asynchronously
>>>>>>>>>> execute the recorded
>>>>>>>>>> + *                commands and call back after ISR is finished
>>>>>>>>>> + * @rec:    the command queue record
>>>>>>>>>> + * @cb:        called in the end of CMDQ ISR
>>>>>>>>>> + * @data:    this data will pass back to cb
>>>>>>>>>> + *
>>>>>>>>>> + * Return: 0 for success; else the error code is returned
>>>>>>>>>> + *
>>>>>>>>>> + * Trigger CMDQ to asynchronously execute the recorded
>>>>>>>>>> commands and call back
>>>>>>>>>> + * after ISR is finished. Note that this is an ASYNC
>>>>>>>>>> function. When the function
>>>>>>>>>> + * returned, it may or may not be finished. The ISR callback
>>>>>>>>>> function is called
>>>>>>>>>> + * in the end of ISR.
>>>>>>>
>>>>>>> "The callback is called from the ISR."
>>>>>>>
>>>>>>> Regards,
>>>>>>> Matthias
>>>>>>>
>>>>>>>>>> + */
>>>>>>>>>> +int cmdq_rec_flush_async(struct cmdq_rec *rec,
>>>>>>>>>> cmdq_async_flush_cb cb,
>>>>>>>>>> +             void *data);
>>>>>>>>>> +
>>>>>>>>>> +/**
>>>>>>>>>> + * cmdq_rec_destroy() - destroy command queue record
>>>>>>>>>> + * @rec:    the command queue record
>>>>>>>>>> + */
>>>>>>>>>> +void cmdq_rec_destroy(struct cmdq_rec *rec);
>>>>>>>>>> +
>>>>>>>>>> +#endif    /* __MTK_CMDQ_H__ */
>>>>>>>>>>
>>>>>>>>
>>>>>>>> Thanks,
>>>>>>>> HS
>>>>>>>>
>>>>>>
>>>>>> Thanks,
>>>>>> HS
>>>>>>
>>>>
>>>>
>>
>>

Matthias Brugger June 7, 2016, 5:04 p.m. UTC | #16

On 30/05/16 05:19, HS Liao wrote:
> This patch is first version of Mediatek Command Queue(CMDQ) driver. The
> CMDQ is used to help read/write registers with critical time limitation,
> such as updating display configuration during the vblank. It controls
> Global Command Engine (GCE) hardware to achieve this requirement.
> Currently, CMDQ only supports display related hardwares, but we expect
> it can be extended to other hardwares for future requirements.
>
> Signed-off-by: HS Liao <[e-mail]>
> Signed-off-by: CK Hu <[e-mail]>
> ---

[...]

> +static void cmdq_handle_error_done(struct cmdq *cmdq,
> +				   struct cmdq_thread *thread, u32 irq_flag)
> +{
> +	struct cmdq_task *task, *tmp, *curr_task = NULL;
> +	u32 curr_pa;
> +	struct cmdq_cb_data cmdq_cb_data;
> +	bool err;
> +
> +	if (irq_flag & CMDQ_THR_IRQ_ERROR)
> +		err = true;
> +	else if (irq_flag & CMDQ_THR_IRQ_DONE)
> +		err = false;
> +	else
> +		return;
> +
> +	curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> +
> +	list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
> +				 list_entry) {
> +		if (curr_pa >= task->pa_base &&
> +		    curr_pa < (task->pa_base + task->command_size))
> +			curr_task = task;
> +		if (task->cb.cb) {
> +			cmdq_cb_data.err = curr_task ? err : false;
> +			cmdq_cb_data.data = task->cb.data;
> +			task->cb.cb(cmdq_cb_data);
> +		}

I think this is not right. If we got an IRQ_DONE, then the current task 
is in execution, we should not call the callback until it has finished.

Regards,
Matthias

[e-mail] June 8, 2016, 5:09 a.m. UTC | #17

Hi Matthias,

On Tue, 2016-06-07 at 19:04 +0200, Matthias Brugger wrote:
> 
> On 30/05/16 05:19, HS Liao wrote:
> > This patch is first version of Mediatek Command Queue(CMDQ) driver. The
> > CMDQ is used to help read/write registers with critical time limitation,
> > such as updating display configuration during the vblank. It controls
> > Global Command Engine (GCE) hardware to achieve this requirement.
> > Currently, CMDQ only supports display related hardwares, but we expect
> > it can be extended to other hardwares for future requirements.
> >
> > Signed-off-by: HS Liao <[e-mail]>
> > Signed-off-by: CK Hu <[e-mail]>
> > ---
> 
> [...]
> 
> > +static void cmdq_handle_error_done(struct cmdq *cmdq,
> > +				   struct cmdq_thread *thread, u32 irq_flag)
> > +{
> > +	struct cmdq_task *task, *tmp, *curr_task = NULL;
> > +	u32 curr_pa;
> > +	struct cmdq_cb_data cmdq_cb_data;
> > +	bool err;
> > +
> > +	if (irq_flag & CMDQ_THR_IRQ_ERROR)
> > +		err = true;
> > +	else if (irq_flag & CMDQ_THR_IRQ_DONE)
> > +		err = false;
> > +	else
> > +		return;
> > +
> > +	curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> > +
> > +	list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
> > +				 list_entry) {
> > +		if (curr_pa >= task->pa_base &&
> > +		    curr_pa < (task->pa_base + task->command_size))
> > +			curr_task = task;
> > +		if (task->cb.cb) {
> > +			cmdq_cb_data.err = curr_task ? err : false;
> > +			cmdq_cb_data.data = task->cb.data;
> > +			task->cb.cb(cmdq_cb_data);
> > +		}
> 
> I think this is not right. If we got an IRQ_DONE, then the current task 
> is in execution, we should not call the callback until it has finished.

Thanks for your finding. This is a bug from CMDQ v6.
I will fix it in next version (CMDQ v9).

> 
> Regards,
> Matthias

Thanks,
HS

[e-mail] June 8, 2016, 5:40 a.m. UTC | #18

Hi Matthias,

On Tue, 2016-06-07 at 18:59 +0200, Matthias Brugger wrote:
> 
> On 03/06/16 15:11, Matthias Brugger wrote:
> >
> >
> [...]
> 
> >>>>>>>>>> +
> >>>>>>>>>> +            smp_mb(); /* modify jump before enable thread */
> >>>>>>>>>> +        }
> >>>>>>>>>> +
> >>>>>>>>>> +        cmdq_thread_writel(thread, task->pa_base +
> >>>>>>>>>> task->command_size,
> >>>>>>>>>> +                   CMDQ_THR_END_ADDR);
> >>>>>>>>>> +        cmdq_thread_resume(thread);
> >>>>>>>>>> +    }
> >>>>>>>>>> +    list_move_tail(&task->list_entry, &thread->task_busy_list);
> >>>>>>>>>> +    spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>>>>>>>>> +}
> >>>>>>>>>> +
> >>>>>>>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
> >>>>>>>>>> +                   struct cmdq_thread *thread, u32 irq_flag)
> >>>>>>>>>> +{
> >>>>>>>>>> +    struct cmdq_task *task, *tmp, *curr_task = NULL;
> >>>>>>>>>> +    u32 curr_pa;
> >>>>>>>>>> +    struct cmdq_cb_data cmdq_cb_data;
> >>>>>>>>>> +    bool err;
> >>>>>>>>>> +
> >>>>>>>>>> +    if (irq_flag & CMDQ_THR_IRQ_ERROR)
> >>>>>>>>>> +        err = true;
> >>>>>>>>>> +    else if (irq_flag & CMDQ_THR_IRQ_DONE)
> >>>>>>>>>> +        err = false;
> >>>>>>>>>> +    else
> >>>>>>>>>> +        return;
> >>>>>>>>>> +
> >>>>>>>>>> +    curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> >>>>>>>>>> +
> >>>>>>>>>> +    list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
> >>>>>>>>>> +                 list_entry) {
> >>>>>>>>>> +        if (curr_pa >= task->pa_base &&
> >>>>>>>>>> +            curr_pa < (task->pa_base + task->command_size))
> >>>>>>>>>
> >>>>>>>>> What are you checking here? It seems as if you make some implcit
> >>>>>>>>> assumptions about pa_base and the order of execution of
> >>>>>>>>> commands in the
> >>>>>>>>> thread. Is it save to do so? Does dma_alloc_coherent give any
> >>>>>>>>> guarantees
> >>>>>>>>> about dma_handle?
> >>>>>>>>
> >>>>>>>> 1. Check what is the current running task in this GCE thread.
> >>>>>>>> 2. Yes.
> >>>>>>>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
> >>>>>>>>
> >>>>>>>
> >>>>>>> Yes, physical addresses might be continous, but AFAIK there is no
> >>>>>>> guarantee that the dma_handle address is steadily growing, when
> >>>>>>> calling
> >>>>>>> dma_alloc_coherent. And if I understand the code correctly, you
> >>>>>>> use this
> >>>>>>> assumption to decide if the task picked from task_busy_list is
> >>>>>>> currently
> >>>>>>> executing. So I think this mecanism is not working.
> >>>>>>
> >>>>>> I don't use dma_handle address, and just use physical addresses.
> >>>>>>    From CPU's point of view, tasks are linked by the busy list.
> >>>>>>    From GCE's point of view, tasks are linked by the JUMP command.
> >>>>>>
> >>>>>>> In which cases does the HW thread raise an interrupt.
> >>>>>>> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
> >>>>>>
> >>>>>> GCE will raise interrupt if any task is done or error.
> >>>>>> However, GCE is fast, so CPU may get multiple done tasks
> >>>>>> when it is running ISR.
> >>>>>>
> >>>>>> In case of error, that GCE thread will pause and raise interrupt.
> >>>>>> So, CPU may get multiple done tasks and one error task.
> >>>>>>
> >>>>>
> >>>>> I think we should reimplement the ISR mechanism. Can't we just read
> >>>>> CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave
> >>>>> cmdq_handle_error_done to the thread_fn? You will need to pass
> >>>>> information from the handler to thread_fn, but that shouldn't be an
> >>>>> issue. AFAIK interrupts are disabled in the handler, so we should stay
> >>>>> there as short as possible. Traversing task_busy_list is expensive, so
> >>>>> we need to do it in a thread context.
> >>>>
> >>>> Actually, our initial implementation is similar to your suggestion,
> >>>> but display needs CMDQ to return callback function very precisely,
> >>>> else display will drop frame.
> >>>> For display, CMDQ interrupt will be raised every 16 ~ 17 ms,
> >>>> and CMDQ needs to call callback function in ISR.
> >>>> If we defer callback to workqueue, the time interval may be larger than
> >>>> 32 ms.sometimes.
> >>>>
> >>>
> >>> I think the problem is, that you implemented the workqueue as a ordered
> >>> workqueue, so there is no parallel processing. I'm still not sure why
> >>> you need the workqueue to be ordered. Can you please explain.
> >>
> >> The order should be kept.
> >> Let me use mouse cursor as an example.
> >> If task 1 means move mouse cursor to point A, task 2 means point B,
> >> and task 3 means point C, our expected result is A -> B -> C.
> >> If the order is not kept, the result could become A -> C -> B.
> >>
> >
> > Got it, thanks for the clarification.
> >
> 
> I think a way to get rid of the workqueue is to use a timer, which gets 
> programmed to the time a timeout in the first task in the busy list 
> would happen. Everytime we update the busy list (e.g. because of task 
> got finished by the thread), we update the timer. When the timer 
> triggers, which hopefully won't happen too often, we return timeout on 
> the busy list elements, until the time is lower then the actual time.
> 
> At least with this we can reduce the data structures in this driver and 
> make it more lightweight.

From my understanding, your proposed method can handle timeout case.

However, the workqueue is also in charge of releasing tasks.
Do you take releasing tasks into consideration by using the proposed
timer method?
Furthermore, I think the code will become more complex if we also use
timer to implement releasing tasks.

> Best regards,
> Matthias

Thanks,
HS

> >>>>> I keep thinking about how to get rid of the two data structures,
> >>>>> task_busy_list and the task_release_wq. We need the latter for the
> >>>>> only
> >>>>> sake of getting a timeout.
> >>>>>
> >>>>> Did you have a look on how the mailbox framework handles this?
> >>>>> By the way, what is the reason to not implement the whole driver as a
> >>>>> mailbox controller? For me, this driver looks like a good fit.
> >>>>
> >>>> CMDQ needs to encode commands for GCE hardware. We think this behavior
> >>>> should be put in CMDQ driver, and client just call CMDQ functions.
> >>>> Therefore, if we want to use mailbox framework, cmdq_rec must be
> >>>> mailbox client, and the others must be mailbox controller.
> >>>>
> >>>
> >>> You mean the functions to fill the cmdq_rec and execute it?
> >>> I think this should be part of the driver.
> >>
> >> Yes.
> >>
> >>> Jassi, can you have a look on the interface this driver exports [0].
> >>> They are needed to actually create the message which will be send.
> >>> Could something like this be part of a mailbox driver?
> >>>
> >>> [0] https://patchwork.kernel.org/patch/9140221/
> >>>
> >>>> However, if we use mailbox controller, CMDQ driver still needs to
> >>>> control busy list for each GCE thread, and use workqueue to handle
> >>>> timeout tasks.
> >>>>
> >>>
> >>> Let me summarize my ideas around this driver:
> >>> When we enter the ISR, we know that all task in task_busy_list before
> >>> the entry which represents curr_task can be set to TASK_STATE_DONE.
> >>> The curr_task could be TASK_STATE_ERROR if the corresponding bit in the
> >>> irq status registers is set.
> >>> Do we need to call the callback in the same order as the tasks got
> >>> dispatched to the HW thread? If not, we could try to call all this
> >>> callbacks in a multithreaded workqueue.
> >>
> >> Yes, we should keep order.
> >>
> >>> Regards,
> >>> Matthias
> >>
> >> Thanks,
> >> HS
> >>
> >>>> The only thing that we can borrow from mailbox framework is the send
> >>>> (CMDQ flush) and receive (CMDQ callback) interface, However, we don't
> >>>> think we can gain many benefits from it, and we have some overheads to
> >>>> conform to mailbox interface.
> >>>>
> >>>>
> >>>>>
> >>>>>>>>>> +            curr_task = task;
> >>>>>>>>>> +        if (task->cb.cb) {
> >>>>>>>>>> +            cmdq_cb_data.err = curr_task ? err : false;
> >>>>>>>>>> +            cmdq_cb_data.data = task->cb.data;
> >>>>>>>>>> +            task->cb.cb(cmdq_cb_data);
> >>>>>>>>>> +        }
> >>>>>>>>>> +        task->task_state = (curr_task && err) ?
> >>>>>>>>>> TASK_STATE_ERROR :
> >>>>>>>>>> +                TASK_STATE_DONE;
> >>>>>>>>>> +        list_del(&task->list_entry);
> >>>>>>>>>> +        if (curr_task)
> >>>>>>>>>> +            break;
> >>>>>>>>>> +    }
> >>>>>>>>>> +
> >>>>>>>>>> +    wake_up(&thread->wait_task_done);
> >>>>>>>>>> +}
> >>>>>>>>>> +
> >>>>>>>>>> +static void cmdq_thread_irq_handler(struct cmdq *cmdq, int tid)
> >>>>>>>>>> +{
> >>>>>>>>>> +    struct cmdq_thread *thread = &cmdq->thread[tid];
> >>>>>>>>>> +    unsigned long flags = 0L;
> >>>>>>>>>> +    u32 irq_flag;
> >>>>>>>>>> +
> >>>>>>>>>> +    spin_lock_irqsave(&cmdq->exec_lock, flags);
> >>>>>>>>>> +
> >>>>>>>>>> +    irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
> >>>>>>>>>> +    cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
> >>>>>>>>>> +
> >>>>>>>>>> +    /*
> >>>>>>>>>> +     * Another CPU core could run "release task" right before
> >>>>>>>>>> we acquire
> >>>>>>>>>> +     * the spin lock, and thus reset / disable this GCE
> >>>>>>>>>> thread, so we
> >>>>>>>>>> +     * need to check the enable bit of this GCE thread.
> >>>>>>>>>> +     */
> >>>>>>>>>> +    if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> >>>>>>>>>> +          CMDQ_THR_ENABLED))
> >>>>>>>>>> +        irq_flag = 0;
> >>>>>>>>>
> >>>>>>>>> cmdq_handle_error_done just retuns in this case. Programming
> >>>>>>>>> this way
> >>>>>>>>> just makes things confusing. What about:
> >>>>>>>>>
> >>>>>>>>> if (cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> >>>>>>>>>           CMDQ_THR_ENABLED)
> >>>>>>>>>     cmdq_handle_error_done(cmdq, thread, irq_flag);
> >>>>>>>>> else
> >>>>>>>>>         irq_flag = 0;
> >>>>>>>>>
> >>>>>>>>> spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>>>>>>>
> >>>>>>>> We still need to clear irq_flag if GCE thread is disabled.
> >>>>>>>> So, I think we can just return here.
> >>>>>>>>
> >>>>>>>>     if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
> >>>>>>>>           CMDQ_THR_ENABLED))
> >>>>>>>>         return;
> >>>>>>>>
> >>>>>>>> What do you think?
> >>>>>>>>
> >>>>>>>
> >>>>>>> No, you can't just return, you need to unlock the spinlock.
> >>>>>>> Anyway I would prefer it the other way round, as I put it in my last
> >>>>>>> mail. Just delete the else branch, we don't need to set irq_flag
> >>>>>>> to zero.
> >>>>>>
> >>>>>> Sorry for my previous wrong reply since I merge your comment
> >>>>>> and CK's comment.
> >>>>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-May/005547.html
> >>>>>>
> >>>>>> So, I will put this if condition into cmdq_task_handle_error_result()
> >>>>>> and then just return it if GCE thread is disabled.
> >>>>>>
> >>>>>
> >>>>> You mean in cmdq_task_handle_done
> >>>>> We should rename this functions to not create confusion.
> >>>>
> >>>> Sorry again. I mean in cmdq_handle_error_done().
> >>>> This function handles both done and error.
> >>>>
> >>>> I agree the function name looks confusion.
> >>>> I think it can be renamed to cmdq_thread_irq_handler()
> >>>> since it is used to handle irq for GCE thread.
> >>>>
> >>>>> Regards,
> >>>>> Matthias
> >>>>
> >>>> Thanks,
> >>>> HS
> >>>>
> >>>>>>>>>> +
> >>>>>>>>>> +    cmdq_handle_error_done(cmdq, thread, irq_flag);
> >>>>>>>>>> +    spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>>>>>>>>> +}
> >>>>>>>>>> +
> >>>>>>>>>> +static irqreturn_t cmdq_irq_handler(int irq, void *dev)
> >>>>>>>>>> +{
> >>>>>>>>>> +    struct cmdq *cmdq = dev;
> >>>>>>>>>> +    u32 irq_status;
> >>>>>>>>>> +    int i;
> >>>>>>>>>> +
> >>>>>>>>>> +    irq_status = readl(cmdq->base + CMDQ_CURR_IRQ_STATUS);
> >>>>>>>>>> +    irq_status &= CMDQ_IRQ_MASK;
> >>>>>>>>>> +    irq_status ^= CMDQ_IRQ_MASK;
> >>>>>>>>>
> >>>>>>>>> irq_status can be much bigger then 3, which is the number of
> >>>>>>>>> threads in
> >>>>>>>>> the system (CMDQ_THR_MAX_COUNT). So why we use this mask here
> >>>>>>>>> isn't
> >>>>>>>>> clear to me.
> >>>>>>>>
> >>>>>>>> Our GCE hardware has 16 threads, but we only use 3 threads
> >>>>>>>> currently.
> >>>>>>>>
> >>>>>>>
> >>>>>>> Ok, but please use bitops here.
> >>>>>>
> >>>>>> Will use bitops.
> >>>>>>
> >>>>>>>>>> +
> >>>>>>>>>> +    if (!irq_status)
> >>>>>>>>>> +        return IRQ_NONE;
> >>>>>>>>>> +
> >>>>>>>>>> +    while (irq_status) {
> >>>>>>>>>> +        i = ffs(irq_status) - 1;
> >>>>>>>>>> +        irq_status &= ~BIT(i);
> >>>>>>>>>> +        cmdq_thread_irq_handler(cmdq, i);
> >>>>>>>>>> +    }
> >>>>>>>>>
> >>>>>>>>> Can you explain how the irq status register looks like, that
> >>>>>>>>> would it
> >>>>>>>>> make much easier to understand what happens here.
> >>>>>>>>
> >>>>>>>> Bit 0 ~ 15 of irq status register represents GCE thread 0 ~ 15
> >>>>>>>> interrupt. 0 means asserting interrupt; 1 means no interrupt.
> >>>>>>>>
> >>>>>>>
> >>>>>>> Thanks, that helped. :)
> >>>>>>>
> >>>>>>>>>> +
> >>>>>>>>>> +    return IRQ_HANDLED;
> >>>>>>>>>> +}
> >>>>>>>>>> +
> >>>>>>>>>> +static int cmdq_task_handle_error_result(struct cmdq_task *task)
> >>>>>>>>>
> >>>>>>>>> We never check the return values, why do we have them?
> >>>>>>>>
> >>>>>>>> Will drop return value.
> >>>>>>>>
> >>>>>>>>>> +{
> >>>>>>>>>> +    struct cmdq *cmdq = task->cmdq;
> >>>>>>>>>> +    struct device *dev = cmdq->dev;
> >>>>>>>>>> +    struct cmdq_thread *thread = task->thread;
> >>>>>>>>>> +    struct cmdq_task *next_task, *prev_task;
> >>>>>>>>>> +    u32 irq_flag;
> >>>>>>>>>> +
> >>>>>>>>>> +    /* suspend GCE thread to ensure consistency */
> >>>>>>>>>> +    WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
> >>>>>>>>>> +
> >>>>>>>>>> +    /* ISR has handled this error task */
> >>>>>>>>>> +    if (task->task_state == TASK_STATE_ERROR) {
> >>>>>>>>>> +        next_task =
> >>>>>>>>>> list_first_entry_or_null(&thread->task_busy_list,
> >>>>>>>>>> +                struct cmdq_task, list_entry);
> >>>>>>>>>> +        if (next_task) /* move to next task */
> >>>>>>>>>> +            cmdq_thread_writel(thread, next_task->pa_base,
> >>>>>>>>>> +                       CMDQ_THR_CURR_ADDR);
> >>>>>>>>>
> >>>>>>>>> We have to do this, as we suppose that the thread did not reach
> >>>>>>>>> the jump
> >>>>>>>>> instruction we put into it's command queue, right?
> >>>>>>>>
> >>>>>>>> Yes.
> >>>>>>>>
> >>>>>>>
> >>>>>>> So this should then go into it's own function. In wait_release_work,
> >>>>>>> something like this:
> >>>>>>>
> >>>>>>> if(task->task_state == TASK_STATE_ERROR)
> >>>>>>>          cmdq_task_handle_error(task)
> >>>>>>
> >>>>>> OK.
> >>>>>> I will write new function cmdq_task_handle_error() to handle error
> >>>>>> case.
> >>>>>>
> >>>>>>>>>> +        cmdq_thread_resume(thread);
> >>>>>>>>>> +        return -ECANCELED;
> >>>>>>>>>> +    }
> >>>>>>>>>> +
> >>>>>>>>>
> >>>>>>>>> if task_state != ERROR and != DONE. This means that the timeout of
> >>>>>>>>> task_release_wq has timed out, right?
> >>>>>>>>
> >>>>>>>> Yes.
> >>>>>>>>
> >>>>>>>>>> +    /*
> >>>>>>>>>> +     * Save next_task and prev_task in advance
> >>>>>>>>>> +     * since cmdq_handle_error_done will remove list_entry.
> >>>>>>>>>> +     */
> >>>>>>>>>> +    next_task = prev_task = NULL;
> >>>>>>>>>> +    if (task->list_entry.next != &thread->task_busy_list)
> >>>>>>>>>> +        next_task = list_next_entry(task, list_entry);
> >>>>>>>>>> +    if (task->list_entry.prev != &thread->task_busy_list)
> >>>>>>>>>> +        prev_task = list_prev_entry(task, list_entry);
> >>>>>>>>>> +
> >>>>>>>>>> +    /*
> >>>>>>>>>> +     * Although IRQ is disabled, GCE continues to execute.
> >>>>>>>>>> +     * It may have pending IRQ before GCE thread is suspended,
> >>>>>>>>>> +     * so check this condition again.
> >>>>>>>>>> +     */
> >>>>>>>>>
> >>>>>>>>> The first thing we did in this function was suspending the
> >>>>>>>>> thread. Why
> >>>>>>>>> do we need this then?
> >>>>>>>>
> >>>>>>>> Because timeout is CPU timeout not GCE timeout, GCE could just
> >>>>>>>> finish
> >>>>>>>> this task before the GCE thread is suspended.
> >>>>>>>>
> >>>>>>>
> >>>>>>> What are the reasons for a timeout? An error has happend, or the
> >>>>>>> task is
> >>>>>>> still executing.
> >>>>>>
> >>>>>>    From GCE's point of view, this task is still executing.
> >>>>>> But, it could be an error of client.
> >>>>>> For example, task may never get event if display turn off hardware
> >>>>>> before waiting for task to finish its work.
> >>>>>>
> >>>>>>>>> To be honest this whole functions looks really like a design
> >>>>>>>>> error. We
> >>>>>>>>> have to sperate the states much clearer so that it is possible to
> >>>>>>>>> understand what is happening in the GCE. Isn't it for example
> >>>>>>>>> posible to
> >>>>>>>>> have worker queues for timed out tasks and tasks with an error?
> >>>>>>>>> I'm not
> >>>>>>>>> sure how to do this, actually I'm not sure if I really
> >>>>>>>>> understood how
> >>>>>>>>> this is supposed to work.
> >>>>>>>>
> >>>>>>>> GCE doesn't have timeout. The timeout is decided and controlled
> >>>>>>>> by CPU,
> >>>>>>>> so we check timeout in release work.
> >>>>>>>> For error and done, they are easy to check by register, and we have
> >>>>>>>> already created release work for timeout. So, I don't think we
> >>>>>>>> need to
> >>>>>>>> create work queue for each case.
> >>>>>>>>
> >>>>>>>> What do you think?
> >>>>>>>>
> >>>>>>>
> >>>>>>> I think, if we find in here, that the irq_flag is set, then the the
> >>>>>>> interrupt handler was triggered and is spinning the spinlock. If
> >>>>>>> this is
> >>>>>>> not the case, we have a timeout and we handle this.
> >>>>>>
> >>>>>> I will write another function to handle error, and handle timeout
> >>>>>> here.
> >>>>>>
> >>>>>>>>> How much do we win, when we patch the thread command queue for
> >>>>>>>>> every
> >>>>>>>>> task we add, instead of just taking one task after another from
> >>>>>>>>> the
> >>>>>>>>> task_busy_list?
> >>>>>>>>
> >>>>>>>> GCE is used to help read/write registers with critical time
> >>>>>>>> limitation.
> >>>>>>>> Sometimes, client may ask to process multiple tasks in a short
> >>>>>>>> period
> >>>>>>>> of time, e.g. display flush multiple tasks for next vblank. So,
> >>>>>>>> CMDQ
> >>>>>>>> shouldn't limit to process one task after another from the
> >>>>>>>> task_busy_list. Currently, when interrupt or timeout, we will check
> >>>>>>>> how many tasks are done, and which one is error or timeout.
> >>>>>>>>
> >>>>>>>
> >>>>>>> So I suppose the device driver who use this are interested in
> >>>>>>> throughput
> >>>>>>> and not in latency. The callback of every
> >>>>>>>
> >>>>>>>>>> +    irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
> >>>>>>>>>> +    cmdq_handle_error_done(cmdq, thread, irq_flag);
> >>>>>>>>>> +    cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
> >>>>>>>>>> +
> >>>>>>>>>> +    if (task->task_state == TASK_STATE_DONE) {
> >>>>>>>>>> +        cmdq_thread_resume(thread);
> >>>>>>>>>> +        return 0;
> >>>>>>>>>> +    }
> >>>>>>>>>> +
> >>>>>>>>>> +    if (task->task_state == TASK_STATE_ERROR) {
> >>>>>>>>>> +        dev_err(dev, "task 0x%p error\n", task);
> >>>>>>>>>> +        if (next_task) /* move to next task */
> >>>>>>>>>> +            cmdq_thread_writel(thread, next_task->pa_base,
> >>>>>>>>>> +                       CMDQ_THR_CURR_ADDR);
> >>>>>>>>>> +        cmdq_thread_resume(thread);
> >>>>>>>>>> +        return -ECANCELED;
> >>>>>>>>>> +    }
> >>>>>>>>>> +
> >>>>>>>>>> +    /* Task is running, so we force to remove it. */
> >>>>>>>>>> +    dev_err(dev, "task 0x%p timeout or killed\n", task);
> >>>>>>>>>> +    task->task_state = TASK_STATE_ERROR;
> >>>>>>>>>> +
> >>>>>>>>>> +    if (prev_task) {
> >>>>>>>>>> +        u64 *prev_va = prev_task->va_base;
> >>>>>>>>>> +        u64 *curr_va = task->va_base;
> >>>>>>>>>> +
> >>>>>>>>>> +        /* copy JUMP instruction */
> >>>>>>>>>> +        prev_va[prev_task->num_cmd - 1] =
> >>>>>>>>>> curr_va[task->num_cmd - 1];
> >>>>>>>>>> +
> >>>>>>>>>> +        cmdq_thread_invalidate_fetched_data(thread);
> >>>>>>>>>> +    } else if (next_task) { /* move to next task */
> >>>>>>>>>> +        cmdq_thread_writel(thread, next_task->pa_base,
> >>>>>>>>>> +                   CMDQ_THR_CURR_ADDR);
> >>>>>>>>>> +    }
> >>>>>>>>>> +
> >>>>>>>>>> +    list_del(&task->list_entry);
> >>>>>>>>>> +    cmdq_thread_resume(thread);
> >>>>>>>>>> +
> >>>>>>>>>> +    /* call cb here to prevent lock */
> >>>>>>>>>> +    if (task->cb.cb) {
> >>>>>>>>>> +        struct cmdq_cb_data cmdq_cb_data;
> >>>>>>>>>> +
> >>>>>>>>>> +        cmdq_cb_data.err = true;
> >>>>>>>>>> +        cmdq_cb_data.data = task->cb.data;
> >>>>>>>>>> +        task->cb.cb(cmdq_cb_data);
> >>>>>>>>>> +    }
> >>>>>>>>>> +
> >>>>>>>>>> +    return -ECANCELED;
> >>>>>>>>>> +}
> >>>>>>>>>> +
> >>>>>>>>>> +static void cmdq_task_wait_release_work(struct work_struct
> >>>>>>>>>> *work_item)
> >>>>>>>>>> +{
> >>>>>>>>>> +    struct cmdq_task *task = container_of(work_item, struct
> >>>>>>>>>> cmdq_task,
> >>>>>>>>>> +                          release_work);
> >>>>>>>>>> +    struct cmdq *cmdq = task->cmdq;
> >>>>>>>>>> +    struct cmdq_thread *thread = task->thread;
> >>>>>>>>>> +    unsigned long flags;
> >>>>>>>>>> +
> >>>>>>>>>> +    wait_event_timeout(thread->wait_task_done,
> >>>>>>>>>> +               task->task_state != TASK_STATE_BUSY,
> >>>>>>>>>> +               msecs_to_jiffies(CMDQ_TIMEOUT_MS));
> >>>>>>>>>> +
> >>>>>>>>>> +    spin_lock_irqsave(&cmdq->exec_lock, flags);
> >>>>>>>>>> +    if (task->task_state != TASK_STATE_DONE)
> >>>>>>>>>> +        cmdq_task_handle_error_result(task);
> >>>>>>>>>> +    if (list_empty(&thread->task_busy_list))
> >>>>>>>>>> +        cmdq_thread_disable(cmdq, thread);
> >>>>>>>>>> +    spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>>>>>>>>> +
> >>>>>>>>>> +    /* release regardless of success or not */
> >>>>>>>>>> +    clk_disable_unprepare(cmdq->clock);
> >>>>>>>>>> +    cmdq_task_release(task);
> >>>>>>>>>> +}
> >>>>>>>>>> +
> >>>>>>>>>> +static void cmdq_task_wait_release_schedule(struct cmdq_task
> >>>>>>>>>> *task)
> >>>>>>>>>> +{
> >>>>>>>>>> +    struct cmdq *cmdq = task->cmdq;
> >>>>>>>>>> +
> >>>>>>>>>> +    INIT_WORK(&task->release_work, cmdq_task_wait_release_work);
> >>>>>>>>>> +    queue_work(cmdq->task_release_wq, &task->release_work);
> >>>>>>>>>> +}
> >>>>>>>>>> +
> >>>>>>>>>> +static int cmdq_rec_realloc_cmd_buffer(struct cmdq_rec *rec,
> >>>>>>>>>> size_t size)
> >>>>>>>>>> +{
> >>>>>>>>>> +    void *new_buf;
> >>>>>>>>>> +
> >>>>>>>>>> +    new_buf = krealloc(rec->buf, size, GFP_KERNEL | __GFP_ZERO);
> >>>>>>>>>> +    if (!new_buf)
> >>>>>>>>>> +        return -ENOMEM;
> >>>>>>>>>> +    rec->buf = new_buf;
> >>>>>>>>>> +    rec->buf_size = size;
> >>>>>>>>>> +    return 0;
> >>>>>>>>>> +}
> >>>>>>>>>> +
> >>>>>>>>>> +struct cmdq_base *cmdq_register_device(struct device *dev)
> >>>>>>>>>> +{
> >>>>>>>>>> +    struct cmdq_base *cmdq_base;
> >>>>>>>>>> +    struct resource res;
> >>>>>>>>>> +    int subsys;
> >>>>>>>>>> +    u32 base;
> >>>>>>>>>> +
> >>>>>>>>>> +    if (of_address_to_resource(dev->of_node, 0, &res))
> >>>>>>>>>> +        return NULL;
> >>>>>>>>>> +    base = (u32)res.start;
> >>>>>>>>>> +
> >>>>>>>>>> +    subsys = cmdq_subsys_base_to_id(base >> 16);
> >>>>>>>>>> +    if (subsys < 0)
> >>>>>>>>>> +        return NULL;
> >>>>>>>>>> +
> >>>>>>>>>> +    cmdq_base = devm_kmalloc(dev, sizeof(*cmdq_base),
> >>>>>>>>>> GFP_KERNEL);
> >>>>>>>>>> +    if (!cmdq_base)
> >>>>>>>>>> +        return NULL;
> >>>>>>>>>> +    cmdq_base->subsys = subsys;
> >>>>>>>>>> +    cmdq_base->base = base;
> >>>>>>>>>> +
> >>>>>>>>>> +    return cmdq_base;
> >>>>>>>>>> +}
> >>>>>>>>>> +EXPORT_SYMBOL(cmdq_register_device);
> >>>>>>>>>> +
> >>>>>>>>>> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
> >>>>>>>>>> +            struct cmdq_rec **rec_ptr)
> >>>>>>>>>> +{
> >>>>>>>>>> +    struct cmdq_rec *rec;
> >>>>>>>>>> +    int err;
> >>>>>>>>>> +
> >>>>>>>>>> +    rec = kzalloc(sizeof(*rec), GFP_KERNEL);
> >>>>>>>>>> +    if (!rec)
> >>>>>>>>>> +        return -ENOMEM;
> >>>>>>>>>> +    rec->cmdq = dev_get_drvdata(dev);
> >>>>>>>>>> +    rec->engine_flag = engine_flag;
> >>>>>>>>>> +    err = cmdq_rec_realloc_cmd_buffer(rec,
> >>>>>>>>>> CMDQ_INITIAL_CMD_BLOCK_SIZE);
> >>>>>>>>>
> >>>>>>>>> Just pass PAGE_SIZE here, no need for CMDQ_INITIAL_CMD_BLOCK_SIZE.
> >>>>>>>>
> >>>>>>>> Will do.
> >>>>>>>>
> >>>>>>>>>> +    if (err < 0) {
> >>>>>>>>>> +        kfree(rec);
> >>>>>>>>>> +        return err;
> >>>>>>>>>> +    }
> >>>>>>>>>> +    *rec_ptr = rec;
> >>>>>>>>>> +    return 0;
> >>>>>>>>>> +}
> >>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_create);
> >>>>>>>>>> +
> >>>>>>>>>> +static int cmdq_rec_append_command(struct cmdq_rec *rec, enum
> >>>>>>>>>> cmdq_code code,
> >>>>>>>>>> +                   u32 arg_a, u32 arg_b)
> >>>>>>>>>> +{
> >>>>>>>>>> +    u64 *cmd_ptr;
> >>>>>>>>>> +    int err;
> >>>>>>>>>> +
> >>>>>>>>>> +    if (WARN_ON(rec->finalized))
> >>>>>>>>>> +        return -EBUSY;
> >>>>>>>>>> +    if (code < CMDQ_CODE_MOVE || code > CMDQ_CODE_EOC)
> >>>>>>>>>> +        return -EINVAL;
> >>>>>>>>>
> >>>>>>>>> cmdq_rec_append_command is just called from inside this driver
> >>>>>>>>> and code
> >>>>>>>>> is a enum. We can expect it to be correct, no need for this check.
> >>>>>>>>
> >>>>>>>> Will drop this check.
> >>>>>>>>
> >>>>>>>>>> +    if (unlikely(rec->command_size + CMDQ_INST_SIZE >
> >>>>>>>>>> rec->buf_size)) {
> >>>>>>>>>
> >>>>>>>>> command_size is the offset into the buffer to which a new
> >>>>>>>>> command is
> >>>>>>>>> written, so this name is highly confusing. I wonder if this
> >>>>>>>>> would be
> >>>>>>>>> easier to understand if we redefine command_size to something
> >>>>>>>>> like the
> >>>>>>>>> number of commands and divide/multiply CMDQ_INST_SIZE where
> >>>>>>>>> this is needed.
> >>>>>>>>
> >>>>>>>> I can rename command_size to cmd_buf_size and calculate num_cmd by
> >>>>>>>> dividing CMDQ_INST_SIZE.
> >>>>>>>> What do you think?
> >>>>>>>>
> >>>>>>>>>> +        err = cmdq_rec_realloc_cmd_buffer(rec, rec->buf_size
> >>>>>>>>>> * 2);
> >>>>>>>>>> +        if (err < 0)
> >>>>>>>>>> +            return err;
> >>>>>>>>>> +    }
> >>>>>>>>>> +    cmd_ptr = rec->buf + rec->command_size;
> >>>>>>>>>> +    (*cmd_ptr) = (u64)((code << CMDQ_OP_CODE_SHIFT) | arg_a)
> >>>>>>>>>> << 32 | arg_b;
> >>>>>>>>>> +    rec->command_size += CMDQ_INST_SIZE;
> >>>>>>>>>> +    return 0;
> >>>>>>>>>> +}
> >>>>>>>>>> +
> >>>>>>>>>> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value, struct
> >>>>>>>>>> cmdq_base *base,
> >>>>>>>>>> +           u32 offset)
> >>>>>>>>>> +{
> >>>>>>>>>> +    u32 arg_a = ((base->base + offset) &
> >>>>>>>>>> CMDQ_ARG_A_WRITE_MASK) |
> >>>>>>>>>> +            ((base->subsys & CMDQ_SUBSYS_MASK) <<
> >>>>>>>>>> CMDQ_SUBSYS_SHIFT);
> >>>>>>>>>
> >>>>>>>>> base->subsys is the id from gce_sybsys, so we can expect it to be
> >>>>>>>>> correct, no need to mask with CMDQ_SUBSYS_MASK.
> >>>>>>>>
> >>>>>>>> Will drop it.
> >>>>>>>>
> >>>>>>>>>> +    return cmdq_rec_append_command(rec, CMDQ_CODE_WRITE,
> >>>>>>>>>> arg_a, value);
> >>>>>>>>>> +}
> >>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_write);
> >>>>>>>>>> +
> >>>>>>>>>> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
> >>>>>>>>>> +            struct cmdq_base *base, u32 offset, u32 mask)
> >>>>>>>>>> +{
> >>>>>>>>>> +    u32 offset_mask = offset;
> >>>>>>>>>> +    int err;
> >>>>>>>>>> +
> >>>>>>>>>> +    if (mask != 0xffffffff) {
> >>>>>>>>>> +        err = cmdq_rec_append_command(rec, CMDQ_CODE_MOVE, 0,
> >>>>>>>>>> ~mask);
> >>>>>>>>>> +        if (err < 0)
> >>>>>>>>>> +            return err;
> >>>>>>>>>> +        offset_mask |= CMDQ_WRITE_ENABLE_MASK;
> >>>>>>>>>> +    }
> >>>>>>>>>> +    return cmdq_rec_write(rec, value, base, offset_mask);
> >>>>>>>>>> +}
> >>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_write_mask);
> >>>>>>>>>> +
> >>>>>>>>>> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event)
> >>>>>>>>>> +{
> >>>>>>>>>> +    u32 arg_b;
> >>>>>>>>>> +
> >>>>>>>>>> +    if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
> >>>>>>>>>> +        return -EINVAL;
> >>>>>>>>>> +
> >>>>>>>>>> +    /*
> >>>>>>>>>> +     * bit 0-11: wait value
> >>>>>>>>>> +     * bit 15: 1 - wait, 0 - no wait
> >>>>>>>>>> +     * bit 16-27: update value
> >>>>>>>>>> +     * bit 31: 1 - update, 0 - no update
> >>>>>>>>>> +     */
> >>>>>>>>>
> >>>>>>>>> I don't understand this comments. What are they for?
> >>>>>>>>
> >>>>>>>> This is for WFE command. I will comment it.
> >>>>>>>>
> >>>>>>>>>> +    arg_b = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT |
> >>>>>>>>>> CMDQ_WFE_WAIT_VALUE;
> >>>>>>>>>> +    return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event,
> >>>>>>>>>> arg_b);
> >>>>>>>>>> +}
> >>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_wfe);
> >>>>>>>>>> +
> >>>>>>>>>> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum
> >>>>>>>>>> cmdq_event event)
> >>>>>>>>>> +{
> >>>>>>>>>> +    if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
> >>>>>>>>>> +        return -EINVAL;
> >>>>>>>>>> +
> >>>>>>>>>> +    return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event,
> >>>>>>>>>> +                       CMDQ_WFE_UPDATE);
> >>>>>>>>>> +}
> >>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_clear_event);
> >>>>>>>>>> +
> >>>>>>>>>> +static int cmdq_rec_finalize(struct cmdq_rec *rec)
> >>>>>>>>>> +{
> >>>>>>>>>> +    int err;
> >>>>>>>>>> +
> >>>>>>>>>> +    if (rec->finalized)
> >>>>>>>>>> +        return 0;
> >>>>>>>>>> +
> >>>>>>>>>> +    /* insert EOC and generate IRQ for each command iteration */
> >>>>>>>>>> +    err = cmdq_rec_append_command(rec, CMDQ_CODE_EOC, 0,
> >>>>>>>>>> CMDQ_EOC_IRQ_EN);
> >>>>>>>>>> +    if (err < 0)
> >>>>>>>>>> +        return err;
> >>>>>>>>>> +
> >>>>>>>>>> +    /* JUMP to end */
> >>>>>>>>>> +    err = cmdq_rec_append_command(rec, CMDQ_CODE_JUMP, 0,
> >>>>>>>>>> CMDQ_JUMP_PASS);
> >>>>>>>>>> +    if (err < 0)
> >>>>>>>>>> +        return err;
> >>>>>>>>>> +
> >>>>>>>>>
> >>>>>>>>> Does this need to be atomic?
> >>>>>>>>> What happens if after CODE_EOC and before CODE_JUMP some
> >>>>>>>>> write/read/event gets added?
> >>>>>>>>> What happens if more commands get added to the queue after
> >>>>>>>>> CODE_JUMP,
> >>>>>>>>> but before finalized is set to true. Why don't you use atomic
> >>>>>>>>> functions
> >>>>>>>>> to access finalized?
> >>>>>>>>
> >>>>>>>> Since cmdq_rec doesn't guarantee thread safe, mutex is needed when
> >>>>>>>> client uses cmdq_rec.
> >>>>>>>>
> >>>>>>>
> >>>>>>> Well I think that rec->finalized tries to implement this, but might
> >>>>>>> fail, if two kernel threads work on the same cmdq_rec.
> >>>>>>>
> >>>>>>>>>> +    rec->finalized = true;
> >>>>>>>>>> +    return 0;
> >>>>>>>>>> +}
> >>>>>>>>>> +
> >>>>>>>>>> +int cmdq_rec_flush_async(struct cmdq_rec *rec,
> >>>>>>>>>> cmdq_async_flush_cb cb,
> >>>>>>>>>> +             void *data)
> >>>>>>>>>> +{
> >>>>>>>>>> +    struct cmdq *cmdq = rec->cmdq;
> >>>>>>>>>> +    struct cmdq_task *task;
> >>>>>>>>>> +    struct cmdq_task_cb task_cb;
> >>>>>>>>>> +    struct cmdq_thread *thread;
> >>>>>>>>>> +    int err;
> >>>>>>>>>> +
> >>>>>>>>>> +    mutex_lock(&cmdq->task_mutex);
> >>>>>>>>>> +    if (cmdq->suspended) {
> >>>>>>>>>> +        dev_err(cmdq->dev, "%s is called after suspended\n",
> >>>>>>>>>> __func__);
> >>>>>>>>>> +        mutex_unlock(&cmdq->task_mutex);
> >>>>>>>>>> +        return -EPERM;
> >>>>>>>>>> +    }
> >>>>>>>>>> +
> >>>>>>>>>> +    err = cmdq_rec_finalize(rec);
> >>>>>>>>>> +    if (err < 0) {
> >>>>>>>>>> +        mutex_unlock(&cmdq->task_mutex);
> >>>>>>>>>> +        return err;
> >>>>>>>>>> +    }
> >>>>>>>>>> +
> >>>>>>>>>> +    task_cb.cb = cb;
> >>>>>>>>>> +    task_cb.data = data;
> >>>>>>>>>> +    task = cmdq_task_acquire(rec, task_cb);
> >>>>>>>>>> +    if (!task) {
> >>>>>>>>>> +        mutex_unlock(&cmdq->task_mutex);
> >>>>>>>>>> +        return -EFAULT;
> >>>>>>>>>> +    }
> >>>>>>>>>> +
> >>>>>>>>>> +    thread =
> >>>>>>>>>> &cmdq->thread[cmdq_eng_get_thread(task->engine_flag)];
> >>>>>>>>>> +    cmdq_task_exec(task, thread);
> >>>>>>>>>> +    cmdq_task_wait_release_schedule(task);
> >>>>>>>>>> +    mutex_unlock(&cmdq->task_mutex);
> >>>>>>>>>> +    return 0;
> >>>>>>>>>> +}
> >>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_flush_async);
> >>>>>>>>>> +
> >>>>>>>>>> +struct cmdq_flush_completion {
> >>>>>>>>>> +    struct completion cmplt;
> >>>>>>>>>> +    bool err;
> >>>>>>>>>> +};
> >>>>>>>>>> +
> >>>>>>>>>> +static int cmdq_rec_flush_cb(struct cmdq_cb_data data)
> >>>>>>>>>> +{
> >>>>>>>>>> +    struct cmdq_flush_completion *cmplt = data.data;
> >>>>>>>>>> +
> >>>>>>>>>> +    cmplt->err = data.err;
> >>>>>>>>>> +    complete(&cmplt->cmplt);
> >>>>>>>>>> +    return 0;
> >>>>>>>>>> +}
> >>>>>>>>>> +
> >>>>>>>>>> +int cmdq_rec_flush(struct cmdq_rec *rec)
> >>>>>>>>>> +{
> >>>>>>>>>> +    struct cmdq_flush_completion cmplt;
> >>>>>>>>>> +    int err;
> >>>>>>>>>> +
> >>>>>>>>>> +    init_completion(&cmplt.cmplt);
> >>>>>>>>>> +    err = cmdq_rec_flush_async(rec, cmdq_rec_flush_cb, &cmplt);
> >>>>>>>>>> +    if (err < 0)
> >>>>>>>>>> +        return err;
> >>>>>>>>>> +    wait_for_completion(&cmplt.cmplt);
> >>>>>>>>>> +    return cmplt.err ? -EFAULT : 0;
> >>>>>>>>>> +}
> >>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_flush);
> >>>>>>>>>> +
> >>>>>>>>>> +void cmdq_rec_destroy(struct cmdq_rec *rec)
> >>>>>>>>>> +{
> >>>>>>>>>> +    kfree(rec->buf);
> >>>>>>>>>> +    kfree(rec);
> >>>>>>>>>> +}
> >>>>>>>>>> +EXPORT_SYMBOL(cmdq_rec_destroy);
> >>>>>>>>>> +
> >>>>>>>>>> +static bool cmdq_task_is_empty(struct cmdq *cmdq)
> >>>>>>>>>> +{
> >>>>>>>>>> +    struct cmdq_thread *thread;
> >>>>>>>>>> +    int i;
> >>>>>>>>>> +
> >>>>>>>>>> +    for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
> >>>>>>>>>> +        thread = &cmdq->thread[i];
> >>>>>>>>>> +        if (!list_empty(&thread->task_busy_list))
> >>>>>>>>>> +            return false;
> >>>>>>>>>> +    }
> >>>>>>>>>> +    return true;
> >>>>>>>>>> +}
> >>>>>>>>>> +
> >>>>>>>>>> +static int cmdq_suspend(struct device *dev)
> >>>>>>>>>> +{
> >>>>>>>>>> +    struct cmdq *cmdq = dev_get_drvdata(dev);
> >>>>>>>>>> +    u32 exec_threads;
> >>>>>>>>>> +
> >>>>>>>>>> +    mutex_lock(&cmdq->task_mutex);
> >>>>>>>>>> +    cmdq->suspended = true;
> >>>>>>>>>> +    mutex_unlock(&cmdq->task_mutex);
> >>>>>>>>>> +
> >>>>>>>>>> +    exec_threads = readl(cmdq->base + CMDQ_CURR_LOADED_THR);
> >>>>>>>>>> +    if ((exec_threads & CMDQ_THR_EXECUTING) &&
> >>>>>>>>>> !cmdq_task_is_empty(cmdq)) {
> >>>>>>>>>> +        dev_err(dev, "wait active tasks timeout.\n");
> >>>>>>>>>> +        flush_workqueue(cmdq->task_release_wq);
> >>>>>>>>>> +    }
> >>>>>>>>>> +    return 0;
> >>>>>>>>>> +}
> >>>>>>>>>> +
> >>>>>>>>>> +static int cmdq_resume(struct device *dev)
> >>>>>>>>>> +{
> >>>>>>>>>> +    struct cmdq *cmdq = dev_get_drvdata(dev);
> >>>>>>>>>> +
> >>>>>>>>>> +    cmdq->suspended = false;
> >>>>>>>>>> +    return 0;
> >>>>>>>>>> +}
> >>>>>>>>>> +
> >>>>>>>>>> +static int cmdq_remove(struct platform_device *pdev)
> >>>>>>>>>> +{
> >>>>>>>>>> +    struct cmdq *cmdq = platform_get_drvdata(pdev);
> >>>>>>>>>> +
> >>>>>>>>>> +    destroy_workqueue(cmdq->task_release_wq);
> >>>>>>>>>> +    cmdq->task_release_wq = NULL;
> >>>>>>>>>> +    return 0;
> >>>>>>>>>> +}
> >>>>>>>>>> +
> >>>>>>>>>> +static int cmdq_probe(struct platform_device *pdev)
> >>>>>>>>>> +{
> >>>>>>>>>> +    struct device *dev = &pdev->dev;
> >>>>>>>>>> +    struct device_node *node = dev->of_node;
> >>>>>>>>>> +    struct resource *res;
> >>>>>>>>>> +    struct cmdq *cmdq;
> >>>>>>>>>> +    int err, i;
> >>>>>>>>>> +
> >>>>>>>>>> +    cmdq = devm_kzalloc(dev, sizeof(*cmdq), GFP_KERNEL);
> >>>>>>>>>> +    if (!cmdq)
> >>>>>>>>>> +        return -ENOMEM;
> >>>>>>>>>> +    cmdq->dev = dev;
> >>>>>>>>>> +
> >>>>>>>>>> +    res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> >>>>>>>>>> +    cmdq->base = devm_ioremap_resource(dev, res);
> >>>>>>>>>> +    if (IS_ERR(cmdq->base)) {
> >>>>>>>>>> +        dev_err(dev, "failed to ioremap gce\n");
> >>>>>>>>>> +        return PTR_ERR(cmdq->base);
> >>>>>>>>>> +    }
> >>>>>>>>>> +
> >>>>>>>>>> +    cmdq->irq = irq_of_parse_and_map(node, 0);
> >>>>>>>>>> +    if (!cmdq->irq) {
> >>>>>>>>>> +        dev_err(dev, "failed to get irq\n");
> >>>>>>>>>> +        return -EINVAL;
> >>>>>>>>>> +    }
> >>>>>>>>>> +
> >>>>>>>>>> +    dev_dbg(dev, "cmdq device: addr:0x%p, va:0x%p, irq:%d\n",
> >>>>>>>>>> +        dev, cmdq->base, cmdq->irq);
> >>>>>>>>>> +
> >>>>>>>>>> +    mutex_init(&cmdq->task_mutex);
> >>>>>>>>>> +    spin_lock_init(&cmdq->exec_lock);
> >>>>>>>>>> +    cmdq->task_release_wq = alloc_ordered_workqueue(
> >>>>>>>>>> +            "%s", WQ_MEM_RECLAIM | WQ_HIGHPRI,
> >>>>>>>>>> +            "cmdq_task_wait_release");
> >>>>>>>>>> +
> >>>>>>>>>> +    for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
> >>>>>>>>>> +        cmdq->thread[i].base = cmdq->base + CMDQ_THR_BASE +
> >>>>>>>>>> +                CMDQ_THR_SHIFT * i;
> >>>>>>>>>> +        init_waitqueue_head(&cmdq->thread[i].wait_task_done);
> >>>>>>>>>> +        INIT_LIST_HEAD(&cmdq->thread[i].task_busy_list);
> >>>>>>>>>> +    }
> >>>>>>>>>> +
> >>>>>>>>>> +    platform_set_drvdata(pdev, cmdq);
> >>>>>>>>>> +
> >>>>>>>>>> +    err = devm_request_irq(dev, cmdq->irq, cmdq_irq_handler,
> >>>>>>>>>> IRQF_SHARED,
> >>>>>>>>>> +                   CMDQ_DRIVER_DEVICE_NAME, cmdq);
> >>>>>>>>>> +    if (err < 0) {
> >>>>>>>>>> +        dev_err(dev, "failed to register ISR (%d)\n", err);
> >>>>>>>>>> +        goto fail;
> >>>>>>>>>> +    }
> >>>>>>>>>> +
> >>>>>>>>>> +    cmdq->clock = devm_clk_get(dev, CMDQ_CLK_NAME);
> >>>>>>>>>> +    if (IS_ERR(cmdq->clock)) {
> >>>>>>>>>> +        dev_err(dev, "failed to get clk:%s\n", CMDQ_CLK_NAME);
> >>>>>>>>>> +        err = PTR_ERR(cmdq->clock);
> >>>>>>>>>> +        goto fail;
> >>>>>>>>>> +    }
> >>>>>>>>>> +    return 0;
> >>>>>>>>>> +
> >>>>>>>>>> +fail:
> >>>>>>>>>> +    cmdq_remove(pdev);
> >>>>>>>>>> +    return err;
> >>>>>>>>>> +}
> >>>>>>>>>> +
> >>>>>>>>>> +static const struct dev_pm_ops cmdq_pm_ops = {
> >>>>>>>>>> +    .suspend = cmdq_suspend,
> >>>>>>>>>> +    .resume = cmdq_resume,
> >>>>>>>>>> +};
> >>>>>>>>>> +
> >>>>>>>>>> +static const struct of_device_id cmdq_of_ids[] = {
> >>>>>>>>>> +    {.compatible = "mediatek,mt8173-gce",},
> >>>>>>>>>> +    {}
> >>>>>>>>>> +};
> >>>>>>>>>> +
> >>>>>>>>>> +static struct platform_driver cmdq_drv = {
> >>>>>>>>>> +    .probe = cmdq_probe,
> >>>>>>>>>> +    .remove = cmdq_remove,
> >>>>>>>>>> +    .driver = {
> >>>>>>>>>> +        .name = CMDQ_DRIVER_DEVICE_NAME,
> >>>>>>>>>> +        .owner = THIS_MODULE,
> >>>>>>>>>> +        .pm = &cmdq_pm_ops,
> >>>>>>>>>> +        .of_match_table = cmdq_of_ids,
> >>>>>>>>>> +    }
> >>>>>>>>>> +};
> >>>>>>>>>> +
> >>>>>>>>>> +builtin_platform_driver(cmdq_drv);
> >>>>>>>>>> diff --git a/include/soc/mediatek/cmdq.h
> >>>>>>>>>> b/include/soc/mediatek/cmdq.h
> >>>>>>>>>> new file mode 100644
> >>>>>>>>>> index 0000000..60eef3d
> >>>>>>>>>> --- /dev/null
> >>>>>>>>>> +++ b/include/soc/mediatek/cmdq.h
> >>>>>>>>>> @@ -0,0 +1,197 @@
> >>>>>>>>>> +/*
> >>>>>>>>>> + * Copyright (c) 2015 MediaTek Inc.
> >>>>>>>>>> + *
> >>>>>>>>>> + * This program is free software; you can redistribute it
> >>>>>>>>>> and/or modify
> >>>>>>>>>> + * it under the terms of the GNU General Public License
> >>>>>>>>>> version 2 as
> >>>>>>>>>> + * published by the Free Software Foundation.
> >>>>>>>>>> + *
> >>>>>>>>>> + * This program is distributed in the hope that it will be
> >>>>>>>>>> useful,
> >>>>>>>>>> + * but WITHOUT ANY WARRANTY; without even the implied
> >>>>>>>>>> warranty of
> >>>>>>>>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> >>>>>>>>>> + * GNU General Public License for more details.
> >>>>>>>>>> + */
> >>>>>>>>>> +
> >>>>>>>>>> +#ifndef __MTK_CMDQ_H__
> >>>>>>>>>> +#define __MTK_CMDQ_H__
> >>>>>>>>>> +
> >>>>>>>>>> +#include <linux/platform_device.h>
> >>>>>>>>>> +#include <linux/types.h>
> >>>>>>>>>> +
> >>>>>>>>>> +enum cmdq_eng {
> >>>>>>>>>> +    CMDQ_ENG_DISP_AAL,
> >>>>>>>>>> +    CMDQ_ENG_DISP_COLOR0,
> >>>>>>>>>> +    CMDQ_ENG_DISP_COLOR1,
> >>>>>>>>>> +    CMDQ_ENG_DISP_DPI0,
> >>>>>>>>>> +    CMDQ_ENG_DISP_DSI0,
> >>>>>>>>>> +    CMDQ_ENG_DISP_DSI1,
> >>>>>>>>>> +    CMDQ_ENG_DISP_GAMMA,
> >>>>>>>>>> +    CMDQ_ENG_DISP_OD,
> >>>>>>>>>> +    CMDQ_ENG_DISP_OVL0,
> >>>>>>>>>> +    CMDQ_ENG_DISP_OVL1,
> >>>>>>>>>> +    CMDQ_ENG_DISP_PWM0,
> >>>>>>>>>> +    CMDQ_ENG_DISP_PWM1,
> >>>>>>>>>> +    CMDQ_ENG_DISP_RDMA0,
> >>>>>>>>>> +    CMDQ_ENG_DISP_RDMA1,
> >>>>>>>>>> +    CMDQ_ENG_DISP_RDMA2,
> >>>>>>>>>> +    CMDQ_ENG_DISP_UFOE,
> >>>>>>>>>> +    CMDQ_ENG_DISP_WDMA0,
> >>>>>>>>>> +    CMDQ_ENG_DISP_WDMA1,
> >>>>>>>>>> +    CMDQ_ENG_MAX,
> >>>>>>>>>> +};
> >>>>>>>>>> +
> >>>>>>>>>> +/* events for CMDQ and display */
> >>>>>>>>>> +enum cmdq_event {
> >>>>>>>>>> +    /* Display start of frame(SOF) events */
> >>>>>>>>>> +    CMDQ_EVENT_DISP_OVL0_SOF = 11,
> >>>>>>>>>> +    CMDQ_EVENT_DISP_OVL1_SOF = 12,
> >>>>>>>>>> +    CMDQ_EVENT_DISP_RDMA0_SOF = 13,
> >>>>>>>>>> +    CMDQ_EVENT_DISP_RDMA1_SOF = 14,
> >>>>>>>>>> +    CMDQ_EVENT_DISP_RDMA2_SOF = 15,
> >>>>>>>>>> +    CMDQ_EVENT_DISP_WDMA0_SOF = 16,
> >>>>>>>>>> +    CMDQ_EVENT_DISP_WDMA1_SOF = 17,
> >>>>>>>>>> +    /* Display end of frame(EOF) events */
> >>>>>>>>>> +    CMDQ_EVENT_DISP_OVL0_EOF = 39,
> >>>>>>>>>> +    CMDQ_EVENT_DISP_OVL1_EOF = 40,
> >>>>>>>>>> +    CMDQ_EVENT_DISP_RDMA0_EOF = 41,
> >>>>>>>>>> +    CMDQ_EVENT_DISP_RDMA1_EOF = 42,
> >>>>>>>>>> +    CMDQ_EVENT_DISP_RDMA2_EOF = 43,
> >>>>>>>>>> +    CMDQ_EVENT_DISP_WDMA0_EOF = 44,
> >>>>>>>>>> +    CMDQ_EVENT_DISP_WDMA1_EOF = 45,
> >>>>>>>>>> +    /* Mutex end of frame(EOF) events */
> >>>>>>>>>> +    CMDQ_EVENT_MUTEX0_STREAM_EOF = 53,
> >>>>>>>>>> +    CMDQ_EVENT_MUTEX1_STREAM_EOF = 54,
> >>>>>>>>>> +    CMDQ_EVENT_MUTEX2_STREAM_EOF = 55,
> >>>>>>>>>> +    CMDQ_EVENT_MUTEX3_STREAM_EOF = 56,
> >>>>>>>>>> +    CMDQ_EVENT_MUTEX4_STREAM_EOF = 57,
> >>>>>>>>>> +    /* Display underrun events */
> >>>>>>>>>> +    CMDQ_EVENT_DISP_RDMA0_UNDERRUN = 63,
> >>>>>>>>>> +    CMDQ_EVENT_DISP_RDMA1_UNDERRUN = 64,
> >>>>>>>>>> +    CMDQ_EVENT_DISP_RDMA2_UNDERRUN = 65,
> >>>>>>>>>> +    /* Keep this at the end of HW events */
> >>>>>>>>>> +    CMDQ_MAX_HW_EVENT_COUNT = 260,
> >>>>>>>>>> +};
> >>>>>>>>>> +
> >>>>>>>>>> +struct cmdq_cb_data {
> >>>>>>>>>> +    bool    err;
> >>>>>>>>>> +    void    *data;
> >>>>>>>>>> +};
> >>>>>>>>>> +
> >>>>>>>>>> +typedef int (*cmdq_async_flush_cb)(struct cmdq_cb_data data);
> >>>>>>>>>> +
> >>>>>>>>>> +struct cmdq_task;
> >>>>>>>>>> +struct cmdq;
> >>>>>>>>>> +
> >>>>>>>>>> +struct cmdq_rec {
> >>>>>>>>>> +    struct cmdq        *cmdq;
> >>>>>>>>>> +    u64            engine_flag;
> >>>>>>>>>> +    size_t            command_size;
> >>>>>>>>>> +    void            *buf;
> >>>>>>>>>> +    size_t            buf_size;
> >>>>>>>>>> +    bool            finalized;
> >>>>>>>>>> +};
> >>>>>>>
> >>>>>>> Why do we need cmdq_rec at all? Can't we just use the cmdq_task
> >>>>>>> for that
> >>>>>>> and this way make the driver less complex?
> >>>>>>
> >>>>>> There are two main reasons for cmdq_rec.
> >>>>>> 1. It is slow to access dma too frequently.
> >>>>>>       So, we append commands to cacheable memory, and then flush
> >>>>>> to dma.
> >>>>>> 2. cmdq_rec is not thread safe, but cmdq_task needs thread safe.
> >>>>>>       If we merge them, we need to take care of some synchronization
> >>>>>>       issues.
> >>>>>>
> >>>>>>>>>> +
> >>>>>>>>>> +struct cmdq_base {
> >>>>>>>>>> +    int    subsys;
> >>>>>>>>>> +    u32    base;
> >>>>>>>>>
> >>>>>>>>> subsys can always be calculated via cmdq_subsys_base_to_id(base
> >>>>>>>>> >> 16)
> >>>>>>>>> so we can get rid of the struct, right?
> >>>>>>>>
> >>>>>>>> Current subsys method is based on previous comment from Daniel
> >>>>>>>> Kurtz.
> >>>>>>>> Please take a look of our previous discussion.
> >>>>>>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/004483.html
> >>>>>>>>
> >>>>>>>> Thanks.
> >>>>>>>>
> >>>>>>>
> >>>>>>> I have to look deeper into this, but from what I read, the
> >>>>>>> proposal from
> >>>>>>> Daniel [1] seems good to me.
> >>>>>>>
> >>>>>>> [1] https://patchwork.kernel.org/patch/8068311/
> >>>>>>
> >>>>>> The initial proposal has some problem, so please see the follow-up
> >>>>>> discussions. Thanks.
> >>>>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/003972.html
> >>>>>>
> >>>>>> http://lists.infradead.org/pipermail/linux-mediatek/2016-February/004483.html
> >>>>>>
> >>>>>>
> >>>>>>
> >>>>>>>>>> +};
> >>>>>>>>>> +
> >>>>>>>>>> +/**
> >>>>>>>>>> + * cmdq_register_device() - register device which needs CMDQ
> >>>>>>>>>> + * @dev:        device
> >>>>>>>>>> + *
> >>>>>>>>>> + * Return: cmdq_base pointer or NULL for failed
> >>>>>>>>>> + */
> >>>>>>>>>> +struct cmdq_base *cmdq_register_device(struct device *dev);
> >>>>>>>>>> +
> >>>>>>>>>> +/**
> >>>>>>>>>> + * cmdq_rec_create() - create command queue record
> >>>>>>>>>> + * @dev:        device
> >>>>>>>>>> + * @engine_flag:    command queue engine flag
> >>>>>>>>>> + * @rec_ptr:        command queue record pointer to retrieve
> >>>>>>>>>> cmdq_rec
> >>>>>>>>>> + *
> >>>>>>>>>> + * Return: 0 for success; else the error code is returned
> >>>>>>>>>> + */
> >>>>>>>>>> +int cmdq_rec_create(struct device *dev, u64 engine_flag,
> >>>>>>>>>> +            struct cmdq_rec **rec_ptr);
> >>>>>>>>>> +
> >>>>>>>>>> +/**
> >>>>>>>>>> + * cmdq_rec_write() - append write command to the command
> >>>>>>>>>> queue record
> >>>>>>>>>> + * @rec:    the command queue record
> >>>>>>>>>> + * @value:    the specified target register value
> >>>>>>>>>> + * @base:    the command queue base
> >>>>>>>>>> + * @offset:    register offset from module base
> >>>>>>>>>> + *
> >>>>>>>>>> + * Return: 0 for success; else the error code is returned
> >>>>>>>>>> + */
> >>>>>>>>>> +int cmdq_rec_write(struct cmdq_rec *rec, u32 value,
> >>>>>>>>>> +           struct cmdq_base *base, u32 offset);
> >>>>>>>>>> +
> >>>>>>>>>> +/**
> >>>>>>>>>> + * cmdq_rec_write_mask() - append write command with mask to
> >>>>>>>>>> the command
> >>>>>>>>>> + *               queue record
> >>>>>>>>>> + * @rec:    the command queue record
> >>>>>>>>>> + * @value:    the specified target register value
> >>>>>>>>>> + * @base:    the command queue base
> >>>>>>>>>> + * @offset:    register offset from module base
> >>>>>>>>>> + * @mask:    the specified target register mask
> >>>>>>>>>> + *
> >>>>>>>>>> + * Return: 0 for success; else the error code is returned
> >>>>>>>>>> + */
> >>>>>>>>>> +int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
> >>>>>>>>>> +            struct cmdq_base *base, u32 offset, u32 mask);
> >>>>>>>>>> +
> >>>>>>>>>> +/**
> >>>>>>>>>> + * cmdq_rec_wfe() - append wait for event command to the
> >>>>>>>>>> command queue reco    rd
> >>>>>>>>>
> >>>>>>>>> reco rd -> record
> >>>>>>>>
> >>>>>>>> Will fix it.
> >>>>>>>>
> >>>>>>>>> Regards,
> >>>>>>>>> Matthias
> >>>>>>>>>
> >>>>>>>>>> + * @rec:    the command queue record
> >>>>>>>>>> + * @event:    the desired event type to "wait and CLEAR"
> >>>>>>>>>> + *
> >>>>>>>>>> + * Return: 0 for success; else the error code is returned
> >>>>>>>>>> + */
> >>>>>>>>>> +int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event);
> >>>>>>>>>> +
> >>>>>>>>>> +/**
> >>>>>>>>>> + * cmdq_rec_clear_event() - append clear event command to the
> >>>>>>>>>> command queue
> >>>>>>>>>> + *                record
> >>>>>>>>>> + * @rec:    the command queue record
> >>>>>>>>>> + * @event:    the desired event to be cleared
> >>>>>>>>>> + *
> >>>>>>>>>> + * Return: 0 for success; else the error code is returned
> >>>>>>>>>> + */
> >>>>>>>>>> +int cmdq_rec_clear_event(struct cmdq_rec *rec, enum
> >>>>>>>>>> cmdq_event event);
> >>>>>>>>>> +
> >>>>>>>>>> +/**
> >>>>>>>>>> + * cmdq_rec_flush() - trigger CMDQ to execute the recorded
> >>>>>>>>>> commands
> >>>>>>>>>> + * @rec:    the command queue record
> >>>>>>>>>> + *
> >>>>>>>>>> + * Return: 0 for success; else the error code is returned
> >>>>>>>>>> + *
> >>>>>>>>>> + * Trigger CMDQ to execute the recorded commands. Note that
> >>>>>>>>>> this is a
> >>>>>>>>>> + * synchronous flush function. When the function returned,
> >>>>>>>>>> the recorded
> >>>>>>>>>> + * commands have been done.
> >>>>>>>>>> + */
> >>>>>>>>>> +int cmdq_rec_flush(struct cmdq_rec *rec);
> >>>>>>>>>> +
> >>>>>>>>>> +/**
> >>>>>>>>>> + * cmdq_rec_flush_async() - trigger CMDQ to asynchronously
> >>>>>>>>>> execute the recorded
> >>>>>>>>>> + *                commands and call back after ISR is finished
> >>>>>>>>>> + * @rec:    the command queue record
> >>>>>>>>>> + * @cb:        called in the end of CMDQ ISR
> >>>>>>>>>> + * @data:    this data will pass back to cb
> >>>>>>>>>> + *
> >>>>>>>>>> + * Return: 0 for success; else the error code is returned
> >>>>>>>>>> + *
> >>>>>>>>>> + * Trigger CMDQ to asynchronously execute the recorded
> >>>>>>>>>> commands and call back
> >>>>>>>>>> + * after ISR is finished. Note that this is an ASYNC
> >>>>>>>>>> function. When the function
> >>>>>>>>>> + * returned, it may or may not be finished. The ISR callback
> >>>>>>>>>> function is called
> >>>>>>>>>> + * in the end of ISR.
> >>>>>>>
> >>>>>>> "The callback is called from the ISR."
> >>>>>>>
> >>>>>>> Regards,
> >>>>>>> Matthias
> >>>>>>>
> >>>>>>>>>> + */
> >>>>>>>>>> +int cmdq_rec_flush_async(struct cmdq_rec *rec,
> >>>>>>>>>> cmdq_async_flush_cb cb,
> >>>>>>>>>> +             void *data);
> >>>>>>>>>> +
> >>>>>>>>>> +/**
> >>>>>>>>>> + * cmdq_rec_destroy() - destroy command queue record
> >>>>>>>>>> + * @rec:    the command queue record
> >>>>>>>>>> + */
> >>>>>>>>>> +void cmdq_rec_destroy(struct cmdq_rec *rec);
> >>>>>>>>>> +
> >>>>>>>>>> +#endif    /* __MTK_CMDQ_H__ */
> >>>>>>>>>>
> >>>>>>>>
> >>>>>>>> Thanks,
> >>>>>>>> HS
> >>>>>>>>
> >>>>>>
> >>>>>> Thanks,
> >>>>>> HS
> >>>>>>
> >>>>
> >>>>
> >>
> >>

Matthias Brugger June 8, 2016, 10:45 a.m. UTC | #19

On 08/06/16 07:40, Horng-Shyang Liao wrote:
> Hi Matthias,
>
> On Tue, 2016-06-07 at 18:59 +0200, Matthias Brugger wrote:
>>
>> On 03/06/16 15:11, Matthias Brugger wrote:
>>>
>>>
>> [...]
>>
>>>>>>>>>>>> +
>>>>>>>>>>>> +            smp_mb(); /* modify jump before enable thread */
>>>>>>>>>>>> +        }
>>>>>>>>>>>> +
>>>>>>>>>>>> +        cmdq_thread_writel(thread, task->pa_base +
>>>>>>>>>>>> task->command_size,
>>>>>>>>>>>> +                   CMDQ_THR_END_ADDR);
>>>>>>>>>>>> +        cmdq_thread_resume(thread);
>>>>>>>>>>>> +    }
>>>>>>>>>>>> +    list_move_tail(&task->list_entry, &thread->task_busy_list);
>>>>>>>>>>>> +    spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>>>>>>>>> +}
>>>>>>>>>>>> +
>>>>>>>>>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
>>>>>>>>>>>> +                   struct cmdq_thread *thread, u32 irq_flag)
>>>>>>>>>>>> +{
>>>>>>>>>>>> +    struct cmdq_task *task, *tmp, *curr_task = NULL;
>>>>>>>>>>>> +    u32 curr_pa;
>>>>>>>>>>>> +    struct cmdq_cb_data cmdq_cb_data;
>>>>>>>>>>>> +    bool err;
>>>>>>>>>>>> +
>>>>>>>>>>>> +    if (irq_flag & CMDQ_THR_IRQ_ERROR)
>>>>>>>>>>>> +        err = true;
>>>>>>>>>>>> +    else if (irq_flag & CMDQ_THR_IRQ_DONE)
>>>>>>>>>>>> +        err = false;
>>>>>>>>>>>> +    else
>>>>>>>>>>>> +        return;
>>>>>>>>>>>> +
>>>>>>>>>>>> +    curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
>>>>>>>>>>>> +
>>>>>>>>>>>> +    list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
>>>>>>>>>>>> +                 list_entry) {
>>>>>>>>>>>> +        if (curr_pa >= task->pa_base &&
>>>>>>>>>>>> +            curr_pa < (task->pa_base + task->command_size))
>>>>>>>>>>>
>>>>>>>>>>> What are you checking here? It seems as if you make some implcit
>>>>>>>>>>> assumptions about pa_base and the order of execution of
>>>>>>>>>>> commands in the
>>>>>>>>>>> thread. Is it save to do so? Does dma_alloc_coherent give any
>>>>>>>>>>> guarantees
>>>>>>>>>>> about dma_handle?
>>>>>>>>>>
>>>>>>>>>> 1. Check what is the current running task in this GCE thread.
>>>>>>>>>> 2. Yes.
>>>>>>>>>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Yes, physical addresses might be continous, but AFAIK there is no
>>>>>>>>> guarantee that the dma_handle address is steadily growing, when
>>>>>>>>> calling
>>>>>>>>> dma_alloc_coherent. And if I understand the code correctly, you
>>>>>>>>> use this
>>>>>>>>> assumption to decide if the task picked from task_busy_list is
>>>>>>>>> currently
>>>>>>>>> executing. So I think this mecanism is not working.
>>>>>>>>
>>>>>>>> I don't use dma_handle address, and just use physical addresses.
>>>>>>>>     From CPU's point of view, tasks are linked by the busy list.
>>>>>>>>     From GCE's point of view, tasks are linked by the JUMP command.
>>>>>>>>
>>>>>>>>> In which cases does the HW thread raise an interrupt.
>>>>>>>>> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
>>>>>>>>
>>>>>>>> GCE will raise interrupt if any task is done or error.
>>>>>>>> However, GCE is fast, so CPU may get multiple done tasks
>>>>>>>> when it is running ISR.
>>>>>>>>
>>>>>>>> In case of error, that GCE thread will pause and raise interrupt.
>>>>>>>> So, CPU may get multiple done tasks and one error task.
>>>>>>>>
>>>>>>>
>>>>>>> I think we should reimplement the ISR mechanism. Can't we just read
>>>>>>> CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave
>>>>>>> cmdq_handle_error_done to the thread_fn? You will need to pass
>>>>>>> information from the handler to thread_fn, but that shouldn't be an
>>>>>>> issue. AFAIK interrupts are disabled in the handler, so we should stay
>>>>>>> there as short as possible. Traversing task_busy_list is expensive, so
>>>>>>> we need to do it in a thread context.
>>>>>>
>>>>>> Actually, our initial implementation is similar to your suggestion,
>>>>>> but display needs CMDQ to return callback function very precisely,
>>>>>> else display will drop frame.
>>>>>> For display, CMDQ interrupt will be raised every 16 ~ 17 ms,
>>>>>> and CMDQ needs to call callback function in ISR.
>>>>>> If we defer callback to workqueue, the time interval may be larger than
>>>>>> 32 ms.sometimes.
>>>>>>
>>>>>
>>>>> I think the problem is, that you implemented the workqueue as a ordered
>>>>> workqueue, so there is no parallel processing. I'm still not sure why
>>>>> you need the workqueue to be ordered. Can you please explain.
>>>>
>>>> The order should be kept.
>>>> Let me use mouse cursor as an example.
>>>> If task 1 means move mouse cursor to point A, task 2 means point B,
>>>> and task 3 means point C, our expected result is A -> B -> C.
>>>> If the order is not kept, the result could become A -> C -> B.
>>>>
>>>
>>> Got it, thanks for the clarification.
>>>
>>
>> I think a way to get rid of the workqueue is to use a timer, which gets
>> programmed to the time a timeout in the first task in the busy list
>> would happen. Everytime we update the busy list (e.g. because of task
>> got finished by the thread), we update the timer. When the timer
>> triggers, which hopefully won't happen too often, we return timeout on
>> the busy list elements, until the time is lower then the actual time.
>>
>> At least with this we can reduce the data structures in this driver and
>> make it more lightweight.
>
>  From my understanding, your proposed method can handle timeout case.
>
> However, the workqueue is also in charge of releasing tasks.
> Do you take releasing tasks into consideration by using the proposed
> timer method?
> Furthermore, I think the code will become more complex if we also use
> timer to implement releasing tasks.
>

Can't we call
         clk_disable_unprepare(cmdq->clock);
         cmdq_task_release(task);
after invoking the callback?
Regrading the clock, wouldn't it be easier to handle the clock 
enable/disable depending on the state of task_busy_list? I suppose we 
can't as we would need to check the task_busy_list of all threads, right?

Regards,
Matthias

[e-mail] June 8, 2016, 12:25 p.m. UTC | #20

Hi Matthias,

On Wed, 2016-06-08 at 12:45 +0200, Matthias Brugger wrote:
> 
> On 08/06/16 07:40, Horng-Shyang Liao wrote:
> > Hi Matthias,
> >
> > On Tue, 2016-06-07 at 18:59 +0200, Matthias Brugger wrote:
> >>
> >> On 03/06/16 15:11, Matthias Brugger wrote:
> >>>
> >>>
> >> [...]
> >>
> >>>>>>>>>>>> +
> >>>>>>>>>>>> +            smp_mb(); /* modify jump before enable thread */
> >>>>>>>>>>>> +        }
> >>>>>>>>>>>> +
> >>>>>>>>>>>> +        cmdq_thread_writel(thread, task->pa_base +
> >>>>>>>>>>>> task->command_size,
> >>>>>>>>>>>> +                   CMDQ_THR_END_ADDR);
> >>>>>>>>>>>> +        cmdq_thread_resume(thread);
> >>>>>>>>>>>> +    }
> >>>>>>>>>>>> +    list_move_tail(&task->list_entry, &thread->task_busy_list);
> >>>>>>>>>>>> +    spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>>>>>>>>>>> +}
> >>>>>>>>>>>> +
> >>>>>>>>>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
> >>>>>>>>>>>> +                   struct cmdq_thread *thread, u32 irq_flag)
> >>>>>>>>>>>> +{
> >>>>>>>>>>>> +    struct cmdq_task *task, *tmp, *curr_task = NULL;
> >>>>>>>>>>>> +    u32 curr_pa;
> >>>>>>>>>>>> +    struct cmdq_cb_data cmdq_cb_data;
> >>>>>>>>>>>> +    bool err;
> >>>>>>>>>>>> +
> >>>>>>>>>>>> +    if (irq_flag & CMDQ_THR_IRQ_ERROR)
> >>>>>>>>>>>> +        err = true;
> >>>>>>>>>>>> +    else if (irq_flag & CMDQ_THR_IRQ_DONE)
> >>>>>>>>>>>> +        err = false;
> >>>>>>>>>>>> +    else
> >>>>>>>>>>>> +        return;
> >>>>>>>>>>>> +
> >>>>>>>>>>>> +    curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> >>>>>>>>>>>> +
> >>>>>>>>>>>> +    list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
> >>>>>>>>>>>> +                 list_entry) {
> >>>>>>>>>>>> +        if (curr_pa >= task->pa_base &&
> >>>>>>>>>>>> +            curr_pa < (task->pa_base + task->command_size))
> >>>>>>>>>>>
> >>>>>>>>>>> What are you checking here? It seems as if you make some implcit
> >>>>>>>>>>> assumptions about pa_base and the order of execution of
> >>>>>>>>>>> commands in the
> >>>>>>>>>>> thread. Is it save to do so? Does dma_alloc_coherent give any
> >>>>>>>>>>> guarantees
> >>>>>>>>>>> about dma_handle?
> >>>>>>>>>>
> >>>>>>>>>> 1. Check what is the current running task in this GCE thread.
> >>>>>>>>>> 2. Yes.
> >>>>>>>>>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
> >>>>>>>>>>
> >>>>>>>>>
> >>>>>>>>> Yes, physical addresses might be continous, but AFAIK there is no
> >>>>>>>>> guarantee that the dma_handle address is steadily growing, when
> >>>>>>>>> calling
> >>>>>>>>> dma_alloc_coherent. And if I understand the code correctly, you
> >>>>>>>>> use this
> >>>>>>>>> assumption to decide if the task picked from task_busy_list is
> >>>>>>>>> currently
> >>>>>>>>> executing. So I think this mecanism is not working.
> >>>>>>>>
> >>>>>>>> I don't use dma_handle address, and just use physical addresses.
> >>>>>>>>     From CPU's point of view, tasks are linked by the busy list.
> >>>>>>>>     From GCE's point of view, tasks are linked by the JUMP command.
> >>>>>>>>
> >>>>>>>>> In which cases does the HW thread raise an interrupt.
> >>>>>>>>> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
> >>>>>>>>
> >>>>>>>> GCE will raise interrupt if any task is done or error.
> >>>>>>>> However, GCE is fast, so CPU may get multiple done tasks
> >>>>>>>> when it is running ISR.
> >>>>>>>>
> >>>>>>>> In case of error, that GCE thread will pause and raise interrupt.
> >>>>>>>> So, CPU may get multiple done tasks and one error task.
> >>>>>>>>
> >>>>>>>
> >>>>>>> I think we should reimplement the ISR mechanism. Can't we just read
> >>>>>>> CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave
> >>>>>>> cmdq_handle_error_done to the thread_fn? You will need to pass
> >>>>>>> information from the handler to thread_fn, but that shouldn't be an
> >>>>>>> issue. AFAIK interrupts are disabled in the handler, so we should stay
> >>>>>>> there as short as possible. Traversing task_busy_list is expensive, so
> >>>>>>> we need to do it in a thread context.
> >>>>>>
> >>>>>> Actually, our initial implementation is similar to your suggestion,
> >>>>>> but display needs CMDQ to return callback function very precisely,
> >>>>>> else display will drop frame.
> >>>>>> For display, CMDQ interrupt will be raised every 16 ~ 17 ms,
> >>>>>> and CMDQ needs to call callback function in ISR.
> >>>>>> If we defer callback to workqueue, the time interval may be larger than
> >>>>>> 32 ms.sometimes.
> >>>>>>
> >>>>>
> >>>>> I think the problem is, that you implemented the workqueue as a ordered
> >>>>> workqueue, so there is no parallel processing. I'm still not sure why
> >>>>> you need the workqueue to be ordered. Can you please explain.
> >>>>
> >>>> The order should be kept.
> >>>> Let me use mouse cursor as an example.
> >>>> If task 1 means move mouse cursor to point A, task 2 means point B,
> >>>> and task 3 means point C, our expected result is A -> B -> C.
> >>>> If the order is not kept, the result could become A -> C -> B.
> >>>>
> >>>
> >>> Got it, thanks for the clarification.
> >>>
> >>
> >> I think a way to get rid of the workqueue is to use a timer, which gets
> >> programmed to the time a timeout in the first task in the busy list
> >> would happen. Everytime we update the busy list (e.g. because of task
> >> got finished by the thread), we update the timer. When the timer
> >> triggers, which hopefully won't happen too often, we return timeout on
> >> the busy list elements, until the time is lower then the actual time.
> >>
> >> At least with this we can reduce the data structures in this driver and
> >> make it more lightweight.
> >
> >  From my understanding, your proposed method can handle timeout case.
> >
> > However, the workqueue is also in charge of releasing tasks.
> > Do you take releasing tasks into consideration by using the proposed
> > timer method?
> > Furthermore, I think the code will become more complex if we also use
> > timer to implement releasing tasks.
> >
> 
> Can't we call
>          clk_disable_unprepare(cmdq->clock);
>          cmdq_task_release(task);
> after invoking the callback?

Do you mean just call these two functions in ISR?
My major concern is dma_free_coherent() and kfree() in
cmdq_task_release(task).
Therefore, your suggestion is to use GFP_ATOMIC for both
dma_alloc_coherent() and kzalloc(). Right?
If so, I can try to implement timeout by timer, and discuss with you
if I have further questions.

> Regrading the clock, wouldn't it be easier to handle the clock 
> enable/disable depending on the state of task_busy_list? I suppose we 
> can't as we would need to check the task_busy_list of all threads, right?
> 
> Regards,
> Matthias

Thanks,
HS

Matthias Brugger June 8, 2016, 3:35 p.m. UTC | #21

On 08/06/16 14:25, Horng-Shyang Liao wrote:
> Hi Matthias,
>
> On Wed, 2016-06-08 at 12:45 +0200, Matthias Brugger wrote:
>>
>> On 08/06/16 07:40, Horng-Shyang Liao wrote:
>>> Hi Matthias,
>>>
>>> On Tue, 2016-06-07 at 18:59 +0200, Matthias Brugger wrote:
>>>>
>>>> On 03/06/16 15:11, Matthias Brugger wrote:
>>>>>
>>>>>
>>>> [...]
>>>>
>>>>>>>>>>>>>> +
>>>>>>>>>>>>>> +            smp_mb(); /* modify jump before enable thread */
>>>>>>>>>>>>>> +        }
>>>>>>>>>>>>>> +
>>>>>>>>>>>>>> +        cmdq_thread_writel(thread, task->pa_base +
>>>>>>>>>>>>>> task->command_size,
>>>>>>>>>>>>>> +                   CMDQ_THR_END_ADDR);
>>>>>>>>>>>>>> +        cmdq_thread_resume(thread);
>>>>>>>>>>>>>> +    }
>>>>>>>>>>>>>> +    list_move_tail(&task->list_entry, &thread->task_busy_list);
>>>>>>>>>>>>>> +    spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>>>>>>>>>>> +}
>>>>>>>>>>>>>> +
>>>>>>>>>>>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
>>>>>>>>>>>>>> +                   struct cmdq_thread *thread, u32 irq_flag)
>>>>>>>>>>>>>> +{
>>>>>>>>>>>>>> +    struct cmdq_task *task, *tmp, *curr_task = NULL;
>>>>>>>>>>>>>> +    u32 curr_pa;
>>>>>>>>>>>>>> +    struct cmdq_cb_data cmdq_cb_data;
>>>>>>>>>>>>>> +    bool err;
>>>>>>>>>>>>>> +
>>>>>>>>>>>>>> +    if (irq_flag & CMDQ_THR_IRQ_ERROR)
>>>>>>>>>>>>>> +        err = true;
>>>>>>>>>>>>>> +    else if (irq_flag & CMDQ_THR_IRQ_DONE)
>>>>>>>>>>>>>> +        err = false;
>>>>>>>>>>>>>> +    else
>>>>>>>>>>>>>> +        return;
>>>>>>>>>>>>>> +
>>>>>>>>>>>>>> +    curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
>>>>>>>>>>>>>> +
>>>>>>>>>>>>>> +    list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
>>>>>>>>>>>>>> +                 list_entry) {
>>>>>>>>>>>>>> +        if (curr_pa >= task->pa_base &&
>>>>>>>>>>>>>> +            curr_pa < (task->pa_base + task->command_size))
>>>>>>>>>>>>>
>>>>>>>>>>>>> What are you checking here? It seems as if you make some implcit
>>>>>>>>>>>>> assumptions about pa_base and the order of execution of
>>>>>>>>>>>>> commands in the
>>>>>>>>>>>>> thread. Is it save to do so? Does dma_alloc_coherent give any
>>>>>>>>>>>>> guarantees
>>>>>>>>>>>>> about dma_handle?
>>>>>>>>>>>>
>>>>>>>>>>>> 1. Check what is the current running task in this GCE thread.
>>>>>>>>>>>> 2. Yes.
>>>>>>>>>>>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
>>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> Yes, physical addresses might be continous, but AFAIK there is no
>>>>>>>>>>> guarantee that the dma_handle address is steadily growing, when
>>>>>>>>>>> calling
>>>>>>>>>>> dma_alloc_coherent. And if I understand the code correctly, you
>>>>>>>>>>> use this
>>>>>>>>>>> assumption to decide if the task picked from task_busy_list is
>>>>>>>>>>> currently
>>>>>>>>>>> executing. So I think this mecanism is not working.
>>>>>>>>>>
>>>>>>>>>> I don't use dma_handle address, and just use physical addresses.
>>>>>>>>>>      From CPU's point of view, tasks are linked by the busy list.
>>>>>>>>>>      From GCE's point of view, tasks are linked by the JUMP command.
>>>>>>>>>>
>>>>>>>>>>> In which cases does the HW thread raise an interrupt.
>>>>>>>>>>> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
>>>>>>>>>>
>>>>>>>>>> GCE will raise interrupt if any task is done or error.
>>>>>>>>>> However, GCE is fast, so CPU may get multiple done tasks
>>>>>>>>>> when it is running ISR.
>>>>>>>>>>
>>>>>>>>>> In case of error, that GCE thread will pause and raise interrupt.
>>>>>>>>>> So, CPU may get multiple done tasks and one error task.
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> I think we should reimplement the ISR mechanism. Can't we just read
>>>>>>>>> CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave
>>>>>>>>> cmdq_handle_error_done to the thread_fn? You will need to pass
>>>>>>>>> information from the handler to thread_fn, but that shouldn't be an
>>>>>>>>> issue. AFAIK interrupts are disabled in the handler, so we should stay
>>>>>>>>> there as short as possible. Traversing task_busy_list is expensive, so
>>>>>>>>> we need to do it in a thread context.
>>>>>>>>
>>>>>>>> Actually, our initial implementation is similar to your suggestion,
>>>>>>>> but display needs CMDQ to return callback function very precisely,
>>>>>>>> else display will drop frame.
>>>>>>>> For display, CMDQ interrupt will be raised every 16 ~ 17 ms,
>>>>>>>> and CMDQ needs to call callback function in ISR.
>>>>>>>> If we defer callback to workqueue, the time interval may be larger than
>>>>>>>> 32 ms.sometimes.
>>>>>>>>
>>>>>>>
>>>>>>> I think the problem is, that you implemented the workqueue as a ordered
>>>>>>> workqueue, so there is no parallel processing. I'm still not sure why
>>>>>>> you need the workqueue to be ordered. Can you please explain.
>>>>>>
>>>>>> The order should be kept.
>>>>>> Let me use mouse cursor as an example.
>>>>>> If task 1 means move mouse cursor to point A, task 2 means point B,
>>>>>> and task 3 means point C, our expected result is A -> B -> C.
>>>>>> If the order is not kept, the result could become A -> C -> B.
>>>>>>
>>>>>
>>>>> Got it, thanks for the clarification.
>>>>>
>>>>
>>>> I think a way to get rid of the workqueue is to use a timer, which gets
>>>> programmed to the time a timeout in the first task in the busy list
>>>> would happen. Everytime we update the busy list (e.g. because of task
>>>> got finished by the thread), we update the timer. When the timer
>>>> triggers, which hopefully won't happen too often, we return timeout on
>>>> the busy list elements, until the time is lower then the actual time.
>>>>
>>>> At least with this we can reduce the data structures in this driver and
>>>> make it more lightweight.
>>>
>>>   From my understanding, your proposed method can handle timeout case.
>>>
>>> However, the workqueue is also in charge of releasing tasks.
>>> Do you take releasing tasks into consideration by using the proposed
>>> timer method?
>>> Furthermore, I think the code will become more complex if we also use
>>> timer to implement releasing tasks.
>>>
>>
>> Can't we call
>>           clk_disable_unprepare(cmdq->clock);
>>           cmdq_task_release(task);
>> after invoking the callback?
>
> Do you mean just call these two functions in ISR?
> My major concern is dma_free_coherent() and kfree() in
> cmdq_task_release(task).

Why do we need the dma calls at all? Can't we just calculate the 
physical address using __pa(x)?

> Therefore, your suggestion is to use GFP_ATOMIC for both
> dma_alloc_coherent() and kzalloc(). Right?

I don't think we need GFP_ATOMIC, the critical path will just free the 
memory.

> If so, I can try to implement timeout by timer, and discuss with you
> if I have further questions.
>

Sounds good :)

Thanks,
Matthias

>> Regrading the clock, wouldn't it be easier to handle the clock
>> enable/disable depending on the state of task_busy_list? I suppose we
>> can't as we would need to check the task_busy_list of all threads, right?
>>
>> Regards,
>> Matthias
>
> Thanks,
> HS
>

[e-mail] June 14, 2016, 7:44 a.m. UTC | #22

Hi Matthias,

On Wed, 2016-06-08 at 17:35 +0200, Matthias Brugger wrote:
> 
> On 08/06/16 14:25, Horng-Shyang Liao wrote:
> > Hi Matthias,
> >
> > On Wed, 2016-06-08 at 12:45 +0200, Matthias Brugger wrote:
> >>
> >> On 08/06/16 07:40, Horng-Shyang Liao wrote:
> >>> Hi Matthias,
> >>>
> >>> On Tue, 2016-06-07 at 18:59 +0200, Matthias Brugger wrote:
> >>>>
> >>>> On 03/06/16 15:11, Matthias Brugger wrote:
> >>>>>
> >>>>>
> >>>> [...]
> >>>>
> >>>>>>>>>>>>>> +
> >>>>>>>>>>>>>> +            smp_mb(); /* modify jump before enable thread */
> >>>>>>>>>>>>>> +        }
> >>>>>>>>>>>>>> +
> >>>>>>>>>>>>>> +        cmdq_thread_writel(thread, task->pa_base +
> >>>>>>>>>>>>>> task->command_size,
> >>>>>>>>>>>>>> +                   CMDQ_THR_END_ADDR);
> >>>>>>>>>>>>>> +        cmdq_thread_resume(thread);
> >>>>>>>>>>>>>> +    }
> >>>>>>>>>>>>>> +    list_move_tail(&task->list_entry, &thread->task_busy_list);
> >>>>>>>>>>>>>> +    spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>>>>>>>>>>>>> +}
> >>>>>>>>>>>>>> +
> >>>>>>>>>>>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
> >>>>>>>>>>>>>> +                   struct cmdq_thread *thread, u32 irq_flag)
> >>>>>>>>>>>>>> +{
> >>>>>>>>>>>>>> +    struct cmdq_task *task, *tmp, *curr_task = NULL;
> >>>>>>>>>>>>>> +    u32 curr_pa;
> >>>>>>>>>>>>>> +    struct cmdq_cb_data cmdq_cb_data;
> >>>>>>>>>>>>>> +    bool err;
> >>>>>>>>>>>>>> +
> >>>>>>>>>>>>>> +    if (irq_flag & CMDQ_THR_IRQ_ERROR)
> >>>>>>>>>>>>>> +        err = true;
> >>>>>>>>>>>>>> +    else if (irq_flag & CMDQ_THR_IRQ_DONE)
> >>>>>>>>>>>>>> +        err = false;
> >>>>>>>>>>>>>> +    else
> >>>>>>>>>>>>>> +        return;
> >>>>>>>>>>>>>> +
> >>>>>>>>>>>>>> +    curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> >>>>>>>>>>>>>> +
> >>>>>>>>>>>>>> +    list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
> >>>>>>>>>>>>>> +                 list_entry) {
> >>>>>>>>>>>>>> +        if (curr_pa >= task->pa_base &&
> >>>>>>>>>>>>>> +            curr_pa < (task->pa_base + task->command_size))
> >>>>>>>>>>>>>
> >>>>>>>>>>>>> What are you checking here? It seems as if you make some implcit
> >>>>>>>>>>>>> assumptions about pa_base and the order of execution of
> >>>>>>>>>>>>> commands in the
> >>>>>>>>>>>>> thread. Is it save to do so? Does dma_alloc_coherent give any
> >>>>>>>>>>>>> guarantees
> >>>>>>>>>>>>> about dma_handle?
> >>>>>>>>>>>>
> >>>>>>>>>>>> 1. Check what is the current running task in this GCE thread.
> >>>>>>>>>>>> 2. Yes.
> >>>>>>>>>>>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
> >>>>>>>>>>>>
> >>>>>>>>>>>
> >>>>>>>>>>> Yes, physical addresses might be continous, but AFAIK there is no
> >>>>>>>>>>> guarantee that the dma_handle address is steadily growing, when
> >>>>>>>>>>> calling
> >>>>>>>>>>> dma_alloc_coherent. And if I understand the code correctly, you
> >>>>>>>>>>> use this
> >>>>>>>>>>> assumption to decide if the task picked from task_busy_list is
> >>>>>>>>>>> currently
> >>>>>>>>>>> executing. So I think this mecanism is not working.
> >>>>>>>>>>
> >>>>>>>>>> I don't use dma_handle address, and just use physical addresses.
> >>>>>>>>>>      From CPU's point of view, tasks are linked by the busy list.
> >>>>>>>>>>      From GCE's point of view, tasks are linked by the JUMP command.
> >>>>>>>>>>
> >>>>>>>>>>> In which cases does the HW thread raise an interrupt.
> >>>>>>>>>>> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
> >>>>>>>>>>
> >>>>>>>>>> GCE will raise interrupt if any task is done or error.
> >>>>>>>>>> However, GCE is fast, so CPU may get multiple done tasks
> >>>>>>>>>> when it is running ISR.
> >>>>>>>>>>
> >>>>>>>>>> In case of error, that GCE thread will pause and raise interrupt.
> >>>>>>>>>> So, CPU may get multiple done tasks and one error task.
> >>>>>>>>>>
> >>>>>>>>>
> >>>>>>>>> I think we should reimplement the ISR mechanism. Can't we just read
> >>>>>>>>> CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave
> >>>>>>>>> cmdq_handle_error_done to the thread_fn? You will need to pass
> >>>>>>>>> information from the handler to thread_fn, but that shouldn't be an
> >>>>>>>>> issue. AFAIK interrupts are disabled in the handler, so we should stay
> >>>>>>>>> there as short as possible. Traversing task_busy_list is expensive, so
> >>>>>>>>> we need to do it in a thread context.
> >>>>>>>>
> >>>>>>>> Actually, our initial implementation is similar to your suggestion,
> >>>>>>>> but display needs CMDQ to return callback function very precisely,
> >>>>>>>> else display will drop frame.
> >>>>>>>> For display, CMDQ interrupt will be raised every 16 ~ 17 ms,
> >>>>>>>> and CMDQ needs to call callback function in ISR.
> >>>>>>>> If we defer callback to workqueue, the time interval may be larger than
> >>>>>>>> 32 ms.sometimes.
> >>>>>>>>
> >>>>>>>
> >>>>>>> I think the problem is, that you implemented the workqueue as a ordered
> >>>>>>> workqueue, so there is no parallel processing. I'm still not sure why
> >>>>>>> you need the workqueue to be ordered. Can you please explain.
> >>>>>>
> >>>>>> The order should be kept.
> >>>>>> Let me use mouse cursor as an example.
> >>>>>> If task 1 means move mouse cursor to point A, task 2 means point B,
> >>>>>> and task 3 means point C, our expected result is A -> B -> C.
> >>>>>> If the order is not kept, the result could become A -> C -> B.
> >>>>>>
> >>>>>
> >>>>> Got it, thanks for the clarification.
> >>>>>
> >>>>
> >>>> I think a way to get rid of the workqueue is to use a timer, which gets
> >>>> programmed to the time a timeout in the first task in the busy list
> >>>> would happen. Everytime we update the busy list (e.g. because of task
> >>>> got finished by the thread), we update the timer. When the timer
> >>>> triggers, which hopefully won't happen too often, we return timeout on
> >>>> the busy list elements, until the time is lower then the actual time.
> >>>>
> >>>> At least with this we can reduce the data structures in this driver and
> >>>> make it more lightweight.
> >>>
> >>>   From my understanding, your proposed method can handle timeout case.
> >>>
> >>> However, the workqueue is also in charge of releasing tasks.
> >>> Do you take releasing tasks into consideration by using the proposed
> >>> timer method?
> >>> Furthermore, I think the code will become more complex if we also use
> >>> timer to implement releasing tasks.
> >>>
> >>
> >> Can't we call
> >>           clk_disable_unprepare(cmdq->clock);
> >>           cmdq_task_release(task);
> >> after invoking the callback?
> >
> > Do you mean just call these two functions in ISR?
> > My major concern is dma_free_coherent() and kfree() in
> > cmdq_task_release(task).
> 
> Why do we need the dma calls at all? Can't we just calculate the 
> physical address using __pa(x)?

I prefer to use dma_map_single/dma_unmap_single.

> > Therefore, your suggestion is to use GFP_ATOMIC for both
> > dma_alloc_coherent() and kzalloc(). Right?
> 
> I don't think we need GFP_ATOMIC, the critical path will just free the 
> memory.

I tested these two functions, and kfree was safe.
However, dma_free_coherent raised BUG.
BUG: failure at
/mnt/host/source/src/third_party/kernel/v3.18/mm/vmalloc.c:1514/vunmap()!
1512 void vunmap(const void *addr)
1513 {
1514         BUG_ON(in_interrupt());		// <-- here
1515         might_sleep();
1516         if (addr)
1517                 __vunmap(addr, 0);
1518 }
1519 EXPORT_SYMBOL(vunmap);

Therefore, I plan to use kmalloc + dma_map_single instead of
dma_alloc_coherent, and dma_unmap_single + kfree instead of
dma_free_coherent.

What do you think about the function replacement?

> > If so, I can try to implement timeout by timer, and discuss with you
> > if I have further questions.
> >
> 
> Sounds good :)
> 
> Thanks,
> Matthias

Thanks,
HS

> >> Regrading the clock, wouldn't it be easier to handle the clock
> >> enable/disable depending on the state of task_busy_list? I suppose we
> >> can't as we would need to check the task_busy_list of all threads, right?
> >>
> >> Regards,
> >> Matthias
> >
> > Thanks,
> > HS
> >

Matthias Brugger June 14, 2016, 10:17 a.m. UTC | #23

On 14/06/16 09:44, Horng-Shyang Liao wrote:
> Hi Matthias,
>
> On Wed, 2016-06-08 at 17:35 +0200, Matthias Brugger wrote:
>>
>> On 08/06/16 14:25, Horng-Shyang Liao wrote:
>>> Hi Matthias,
>>>
>>> On Wed, 2016-06-08 at 12:45 +0200, Matthias Brugger wrote:
>>>>
>>>> On 08/06/16 07:40, Horng-Shyang Liao wrote:
>>>>> Hi Matthias,
>>>>>
>>>>> On Tue, 2016-06-07 at 18:59 +0200, Matthias Brugger wrote:
>>>>>>
>>>>>> On 03/06/16 15:11, Matthias Brugger wrote:
>>>>>>>
>>>>>>>
>>>>>> [...]
>>>>>>
>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>> +            smp_mb(); /* modify jump before enable thread */
>>>>>>>>>>>>>>>> +        }
>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>> +        cmdq_thread_writel(thread, task->pa_base +
>>>>>>>>>>>>>>>> task->command_size,
>>>>>>>>>>>>>>>> +                   CMDQ_THR_END_ADDR);
>>>>>>>>>>>>>>>> +        cmdq_thread_resume(thread);
>>>>>>>>>>>>>>>> +    }
>>>>>>>>>>>>>>>> +    list_move_tail(&task->list_entry, &thread->task_busy_list);
>>>>>>>>>>>>>>>> +    spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>>>>>>>>>>>>> +}
>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
>>>>>>>>>>>>>>>> +                   struct cmdq_thread *thread, u32 irq_flag)
>>>>>>>>>>>>>>>> +{
>>>>>>>>>>>>>>>> +    struct cmdq_task *task, *tmp, *curr_task = NULL;
>>>>>>>>>>>>>>>> +    u32 curr_pa;
>>>>>>>>>>>>>>>> +    struct cmdq_cb_data cmdq_cb_data;
>>>>>>>>>>>>>>>> +    bool err;
>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>> +    if (irq_flag & CMDQ_THR_IRQ_ERROR)
>>>>>>>>>>>>>>>> +        err = true;
>>>>>>>>>>>>>>>> +    else if (irq_flag & CMDQ_THR_IRQ_DONE)
>>>>>>>>>>>>>>>> +        err = false;
>>>>>>>>>>>>>>>> +    else
>>>>>>>>>>>>>>>> +        return;
>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>> +    curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>> +    list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
>>>>>>>>>>>>>>>> +                 list_entry) {
>>>>>>>>>>>>>>>> +        if (curr_pa >= task->pa_base &&
>>>>>>>>>>>>>>>> +            curr_pa < (task->pa_base + task->command_size))
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> What are you checking here? It seems as if you make some implcit
>>>>>>>>>>>>>>> assumptions about pa_base and the order of execution of
>>>>>>>>>>>>>>> commands in the
>>>>>>>>>>>>>>> thread. Is it save to do so? Does dma_alloc_coherent give any
>>>>>>>>>>>>>>> guarantees
>>>>>>>>>>>>>>> about dma_handle?
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> 1. Check what is the current running task in this GCE thread.
>>>>>>>>>>>>>> 2. Yes.
>>>>>>>>>>>>>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
>>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>> Yes, physical addresses might be continous, but AFAIK there is no
>>>>>>>>>>>>> guarantee that the dma_handle address is steadily growing, when
>>>>>>>>>>>>> calling
>>>>>>>>>>>>> dma_alloc_coherent. And if I understand the code correctly, you
>>>>>>>>>>>>> use this
>>>>>>>>>>>>> assumption to decide if the task picked from task_busy_list is
>>>>>>>>>>>>> currently
>>>>>>>>>>>>> executing. So I think this mecanism is not working.
>>>>>>>>>>>>
>>>>>>>>>>>> I don't use dma_handle address, and just use physical addresses.
>>>>>>>>>>>>       From CPU's point of view, tasks are linked by the busy list.
>>>>>>>>>>>>       From GCE's point of view, tasks are linked by the JUMP command.
>>>>>>>>>>>>
>>>>>>>>>>>>> In which cases does the HW thread raise an interrupt.
>>>>>>>>>>>>> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
>>>>>>>>>>>>
>>>>>>>>>>>> GCE will raise interrupt if any task is done or error.
>>>>>>>>>>>> However, GCE is fast, so CPU may get multiple done tasks
>>>>>>>>>>>> when it is running ISR.
>>>>>>>>>>>>
>>>>>>>>>>>> In case of error, that GCE thread will pause and raise interrupt.
>>>>>>>>>>>> So, CPU may get multiple done tasks and one error task.
>>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> I think we should reimplement the ISR mechanism. Can't we just read
>>>>>>>>>>> CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave
>>>>>>>>>>> cmdq_handle_error_done to the thread_fn? You will need to pass
>>>>>>>>>>> information from the handler to thread_fn, but that shouldn't be an
>>>>>>>>>>> issue. AFAIK interrupts are disabled in the handler, so we should stay
>>>>>>>>>>> there as short as possible. Traversing task_busy_list is expensive, so
>>>>>>>>>>> we need to do it in a thread context.
>>>>>>>>>>
>>>>>>>>>> Actually, our initial implementation is similar to your suggestion,
>>>>>>>>>> but display needs CMDQ to return callback function very precisely,
>>>>>>>>>> else display will drop frame.
>>>>>>>>>> For display, CMDQ interrupt will be raised every 16 ~ 17 ms,
>>>>>>>>>> and CMDQ needs to call callback function in ISR.
>>>>>>>>>> If we defer callback to workqueue, the time interval may be larger than
>>>>>>>>>> 32 ms.sometimes.
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> I think the problem is, that you implemented the workqueue as a ordered
>>>>>>>>> workqueue, so there is no parallel processing. I'm still not sure why
>>>>>>>>> you need the workqueue to be ordered. Can you please explain.
>>>>>>>>
>>>>>>>> The order should be kept.
>>>>>>>> Let me use mouse cursor as an example.
>>>>>>>> If task 1 means move mouse cursor to point A, task 2 means point B,
>>>>>>>> and task 3 means point C, our expected result is A -> B -> C.
>>>>>>>> If the order is not kept, the result could become A -> C -> B.
>>>>>>>>
>>>>>>>
>>>>>>> Got it, thanks for the clarification.
>>>>>>>
>>>>>>
>>>>>> I think a way to get rid of the workqueue is to use a timer, which gets
>>>>>> programmed to the time a timeout in the first task in the busy list
>>>>>> would happen. Everytime we update the busy list (e.g. because of task
>>>>>> got finished by the thread), we update the timer. When the timer
>>>>>> triggers, which hopefully won't happen too often, we return timeout on
>>>>>> the busy list elements, until the time is lower then the actual time.
>>>>>>
>>>>>> At least with this we can reduce the data structures in this driver and
>>>>>> make it more lightweight.
>>>>>
>>>>>    From my understanding, your proposed method can handle timeout case.
>>>>>
>>>>> However, the workqueue is also in charge of releasing tasks.
>>>>> Do you take releasing tasks into consideration by using the proposed
>>>>> timer method?
>>>>> Furthermore, I think the code will become more complex if we also use
>>>>> timer to implement releasing tasks.
>>>>>
>>>>
>>>> Can't we call
>>>>            clk_disable_unprepare(cmdq->clock);
>>>>            cmdq_task_release(task);
>>>> after invoking the callback?
>>>
>>> Do you mean just call these two functions in ISR?
>>> My major concern is dma_free_coherent() and kfree() in
>>> cmdq_task_release(task).
>>
>> Why do we need the dma calls at all? Can't we just calculate the
>> physical address using __pa(x)?
>
> I prefer to use dma_map_single/dma_unmap_single.
>

Can you please elaborate why you need this. We don't do dma, so we 
should not use dma memory for this.

>>> Therefore, your suggestion is to use GFP_ATOMIC for both
>>> dma_alloc_coherent() and kzalloc(). Right?
>>
>> I don't think we need GFP_ATOMIC, the critical path will just free the
>> memory.
>
> I tested these two functions, and kfree was safe.
> However, dma_free_coherent raised BUG.
> BUG: failure at
> /mnt/host/source/src/third_party/kernel/v3.18/mm/vmalloc.c:1514/vunmap()!

Just a general hint. Please try to evaluate on a recent kernel. It looks 
like as if you tried this on a v3.18 based one.

Best regards,
Matthias

> 1512 void vunmap(const void *addr)
> 1513 {
> 1514         BUG_ON(in_interrupt());		// <-- here
> 1515         might_sleep();
> 1516         if (addr)
> 1517                 __vunmap(addr, 0);
> 1518 }
> 1519 EXPORT_SYMBOL(vunmap);
>
> Therefore, I plan to use kmalloc + dma_map_single instead of
> dma_alloc_coherent, and dma_unmap_single + kfree instead of
> dma_free_coherent.
>
> What do you think about the function replacement?
>
>>> If so, I can try to implement timeout by timer, and discuss with you
>>> if I have further questions.
>>>
>>
>> Sounds good :)
>>
>> Thanks,
>> Matthias
>
> Thanks,
> HS
>
>>>> Regrading the clock, wouldn't it be easier to handle the clock
>>>> enable/disable depending on the state of task_busy_list? I suppose we
>>>> can't as we would need to check the task_busy_list of all threads, right?
>>>>
>>>> Regards,
>>>> Matthias
>>>
>>> Thanks,
>>> HS
>>>
>
>

[e-mail] June 14, 2016, 12:07 p.m. UTC | #24

Hi Matthias,

On Tue, 2016-06-14 at 12:17 +0200, Matthias Brugger wrote:
> 
> On 14/06/16 09:44, Horng-Shyang Liao wrote:
> > Hi Matthias,
> >
> > On Wed, 2016-06-08 at 17:35 +0200, Matthias Brugger wrote:
> >>
> >> On 08/06/16 14:25, Horng-Shyang Liao wrote:
> >>> Hi Matthias,
> >>>
> >>> On Wed, 2016-06-08 at 12:45 +0200, Matthias Brugger wrote:
> >>>>
> >>>> On 08/06/16 07:40, Horng-Shyang Liao wrote:
> >>>>> Hi Matthias,
> >>>>>
> >>>>> On Tue, 2016-06-07 at 18:59 +0200, Matthias Brugger wrote:
> >>>>>>
> >>>>>> On 03/06/16 15:11, Matthias Brugger wrote:
> >>>>>>>
> >>>>>>>
> >>>>>> [...]
> >>>>>>
> >>>>>>>>>>>>>>>> +
> >>>>>>>>>>>>>>>> +            smp_mb(); /* modify jump before enable thread */
> >>>>>>>>>>>>>>>> +        }
> >>>>>>>>>>>>>>>> +
> >>>>>>>>>>>>>>>> +        cmdq_thread_writel(thread, task->pa_base +
> >>>>>>>>>>>>>>>> task->command_size,
> >>>>>>>>>>>>>>>> +                   CMDQ_THR_END_ADDR);
> >>>>>>>>>>>>>>>> +        cmdq_thread_resume(thread);
> >>>>>>>>>>>>>>>> +    }
> >>>>>>>>>>>>>>>> +    list_move_tail(&task->list_entry, &thread->task_busy_list);
> >>>>>>>>>>>>>>>> +    spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>>>>>>>>>>>>>>> +}
> >>>>>>>>>>>>>>>> +
> >>>>>>>>>>>>>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
> >>>>>>>>>>>>>>>> +                   struct cmdq_thread *thread, u32 irq_flag)
> >>>>>>>>>>>>>>>> +{
> >>>>>>>>>>>>>>>> +    struct cmdq_task *task, *tmp, *curr_task = NULL;
> >>>>>>>>>>>>>>>> +    u32 curr_pa;
> >>>>>>>>>>>>>>>> +    struct cmdq_cb_data cmdq_cb_data;
> >>>>>>>>>>>>>>>> +    bool err;
> >>>>>>>>>>>>>>>> +
> >>>>>>>>>>>>>>>> +    if (irq_flag & CMDQ_THR_IRQ_ERROR)
> >>>>>>>>>>>>>>>> +        err = true;
> >>>>>>>>>>>>>>>> +    else if (irq_flag & CMDQ_THR_IRQ_DONE)
> >>>>>>>>>>>>>>>> +        err = false;
> >>>>>>>>>>>>>>>> +    else
> >>>>>>>>>>>>>>>> +        return;
> >>>>>>>>>>>>>>>> +
> >>>>>>>>>>>>>>>> +    curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> >>>>>>>>>>>>>>>> +
> >>>>>>>>>>>>>>>> +    list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
> >>>>>>>>>>>>>>>> +                 list_entry) {
> >>>>>>>>>>>>>>>> +        if (curr_pa >= task->pa_base &&
> >>>>>>>>>>>>>>>> +            curr_pa < (task->pa_base + task->command_size))
> >>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>> What are you checking here? It seems as if you make some implcit
> >>>>>>>>>>>>>>> assumptions about pa_base and the order of execution of
> >>>>>>>>>>>>>>> commands in the
> >>>>>>>>>>>>>>> thread. Is it save to do so? Does dma_alloc_coherent give any
> >>>>>>>>>>>>>>> guarantees
> >>>>>>>>>>>>>>> about dma_handle?
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>> 1. Check what is the current running task in this GCE thread.
> >>>>>>>>>>>>>> 2. Yes.
> >>>>>>>>>>>>>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>
> >>>>>>>>>>>>> Yes, physical addresses might be continous, but AFAIK there is no
> >>>>>>>>>>>>> guarantee that the dma_handle address is steadily growing, when
> >>>>>>>>>>>>> calling
> >>>>>>>>>>>>> dma_alloc_coherent. And if I understand the code correctly, you
> >>>>>>>>>>>>> use this
> >>>>>>>>>>>>> assumption to decide if the task picked from task_busy_list is
> >>>>>>>>>>>>> currently
> >>>>>>>>>>>>> executing. So I think this mecanism is not working.
> >>>>>>>>>>>>
> >>>>>>>>>>>> I don't use dma_handle address, and just use physical addresses.
> >>>>>>>>>>>>       From CPU's point of view, tasks are linked by the busy list.
> >>>>>>>>>>>>       From GCE's point of view, tasks are linked by the JUMP command.
> >>>>>>>>>>>>
> >>>>>>>>>>>>> In which cases does the HW thread raise an interrupt.
> >>>>>>>>>>>>> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
> >>>>>>>>>>>>
> >>>>>>>>>>>> GCE will raise interrupt if any task is done or error.
> >>>>>>>>>>>> However, GCE is fast, so CPU may get multiple done tasks
> >>>>>>>>>>>> when it is running ISR.
> >>>>>>>>>>>>
> >>>>>>>>>>>> In case of error, that GCE thread will pause and raise interrupt.
> >>>>>>>>>>>> So, CPU may get multiple done tasks and one error task.
> >>>>>>>>>>>>
> >>>>>>>>>>>
> >>>>>>>>>>> I think we should reimplement the ISR mechanism. Can't we just read
> >>>>>>>>>>> CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave
> >>>>>>>>>>> cmdq_handle_error_done to the thread_fn? You will need to pass
> >>>>>>>>>>> information from the handler to thread_fn, but that shouldn't be an
> >>>>>>>>>>> issue. AFAIK interrupts are disabled in the handler, so we should stay
> >>>>>>>>>>> there as short as possible. Traversing task_busy_list is expensive, so
> >>>>>>>>>>> we need to do it in a thread context.
> >>>>>>>>>>
> >>>>>>>>>> Actually, our initial implementation is similar to your suggestion,
> >>>>>>>>>> but display needs CMDQ to return callback function very precisely,
> >>>>>>>>>> else display will drop frame.
> >>>>>>>>>> For display, CMDQ interrupt will be raised every 16 ~ 17 ms,
> >>>>>>>>>> and CMDQ needs to call callback function in ISR.
> >>>>>>>>>> If we defer callback to workqueue, the time interval may be larger than
> >>>>>>>>>> 32 ms.sometimes.
> >>>>>>>>>>
> >>>>>>>>>
> >>>>>>>>> I think the problem is, that you implemented the workqueue as a ordered
> >>>>>>>>> workqueue, so there is no parallel processing. I'm still not sure why
> >>>>>>>>> you need the workqueue to be ordered. Can you please explain.
> >>>>>>>>
> >>>>>>>> The order should be kept.
> >>>>>>>> Let me use mouse cursor as an example.
> >>>>>>>> If task 1 means move mouse cursor to point A, task 2 means point B,
> >>>>>>>> and task 3 means point C, our expected result is A -> B -> C.
> >>>>>>>> If the order is not kept, the result could become A -> C -> B.
> >>>>>>>>
> >>>>>>>
> >>>>>>> Got it, thanks for the clarification.
> >>>>>>>
> >>>>>>
> >>>>>> I think a way to get rid of the workqueue is to use a timer, which gets
> >>>>>> programmed to the time a timeout in the first task in the busy list
> >>>>>> would happen. Everytime we update the busy list (e.g. because of task
> >>>>>> got finished by the thread), we update the timer. When the timer
> >>>>>> triggers, which hopefully won't happen too often, we return timeout on
> >>>>>> the busy list elements, until the time is lower then the actual time.
> >>>>>>
> >>>>>> At least with this we can reduce the data structures in this driver and
> >>>>>> make it more lightweight.
> >>>>>
> >>>>>    From my understanding, your proposed method can handle timeout case.
> >>>>>
> >>>>> However, the workqueue is also in charge of releasing tasks.
> >>>>> Do you take releasing tasks into consideration by using the proposed
> >>>>> timer method?
> >>>>> Furthermore, I think the code will become more complex if we also use
> >>>>> timer to implement releasing tasks.
> >>>>>
> >>>>
> >>>> Can't we call
> >>>>            clk_disable_unprepare(cmdq->clock);
> >>>>            cmdq_task_release(task);
> >>>> after invoking the callback?
> >>>
> >>> Do you mean just call these two functions in ISR?
> >>> My major concern is dma_free_coherent() and kfree() in
> >>> cmdq_task_release(task).
> >>
> >> Why do we need the dma calls at all? Can't we just calculate the
> >> physical address using __pa(x)?
> >
> > I prefer to use dma_map_single/dma_unmap_single.
> >
> 
> Can you please elaborate why you need this. We don't do dma, so we 
> should not use dma memory for this.

We need a buffer to share between CPU and GCE, so we do need DMA.
CPU is in charge of writing GCE commands into this buffer.
GCE is in charge of reading and running GCE commands from this buffer.
When we chain CMDQ tasks, we also need to modify GCE JUMP command.
Therefore, I prefer to use dma_alloc_coherent and dma_free_coherent.

However, if we want to use timer to handle timeout, we need to release
memory in ISR.
In this case, using kmalloc/kfree + dma_map_single/dma_unmap_single
instead of dma_alloc_coherent/dma_free_coherent is an alternative
solution, but taking care the synchronization between cache and memory
is the expected overhead.

> >>> Therefore, your suggestion is to use GFP_ATOMIC for both
> >>> dma_alloc_coherent() and kzalloc(). Right?
> >>
> >> I don't think we need GFP_ATOMIC, the critical path will just free the
> >> memory.
> >
> > I tested these two functions, and kfree was safe.
> > However, dma_free_coherent raised BUG.
> > BUG: failure at
> > /mnt/host/source/src/third_party/kernel/v3.18/mm/vmalloc.c:1514/vunmap()!
> 
> Just a general hint. Please try to evaluate on a recent kernel. It looks 
> like as if you tried this on a v3.18 based one.

This driver should be backward compatible to v3.18 for a MTK project.

> Best regards,
> Matthias

Thanks,
HS

> > 1512 void vunmap(const void *addr)
> > 1513 {
> > 1514         BUG_ON(in_interrupt());		// <-- here
> > 1515         might_sleep();
> > 1516         if (addr)
> > 1517                 __vunmap(addr, 0);
> > 1518 }
> > 1519 EXPORT_SYMBOL(vunmap);
> >
> > Therefore, I plan to use kmalloc + dma_map_single instead of
> > dma_alloc_coherent, and dma_unmap_single + kfree instead of
> > dma_free_coherent.
> >
> > What do you think about the function replacement?
> >
> >>> If so, I can try to implement timeout by timer, and discuss with you
> >>> if I have further questions.
> >>>
> >>
> >> Sounds good :)
> >>
> >> Thanks,
> >> Matthias
> >
> > Thanks,
> > HS
> >
> >>>> Regrading the clock, wouldn't it be easier to handle the clock
> >>>> enable/disable depending on the state of task_busy_list? I suppose we
> >>>> can't as we would need to check the task_busy_list of all threads, right?
> >>>>
> >>>> Regards,
> >>>> Matthias
> >>>
> >>> Thanks,
> >>> HS
> >>>
> >
> >

[e-mail] June 17, 2016, 8:28 a.m. UTC | #25

Hi Matthias,

On Tue, 2016-06-14 at 20:07 +0800, Horng-Shyang Liao wrote:
> Hi Matthias,
> 
> On Tue, 2016-06-14 at 12:17 +0200, Matthias Brugger wrote:
> > 
> > On 14/06/16 09:44, Horng-Shyang Liao wrote:
> > > Hi Matthias,
> > >
> > > On Wed, 2016-06-08 at 17:35 +0200, Matthias Brugger wrote:
> > >>
> > >> On 08/06/16 14:25, Horng-Shyang Liao wrote:
> > >>> Hi Matthias,
> > >>>
> > >>> On Wed, 2016-06-08 at 12:45 +0200, Matthias Brugger wrote:
> > >>>>
> > >>>> On 08/06/16 07:40, Horng-Shyang Liao wrote:
> > >>>>> Hi Matthias,
> > >>>>>
> > >>>>> On Tue, 2016-06-07 at 18:59 +0200, Matthias Brugger wrote:
> > >>>>>>
> > >>>>>> On 03/06/16 15:11, Matthias Brugger wrote:
> > >>>>>>>
> > >>>>>>>
> > >>>>>> [...]
> > >>>>>>
> > >>>>>>>>>>>>>>>> +
> > >>>>>>>>>>>>>>>> +            smp_mb(); /* modify jump before enable thread */
> > >>>>>>>>>>>>>>>> +        }
> > >>>>>>>>>>>>>>>> +
> > >>>>>>>>>>>>>>>> +        cmdq_thread_writel(thread, task->pa_base +
> > >>>>>>>>>>>>>>>> task->command_size,
> > >>>>>>>>>>>>>>>> +                   CMDQ_THR_END_ADDR);
> > >>>>>>>>>>>>>>>> +        cmdq_thread_resume(thread);
> > >>>>>>>>>>>>>>>> +    }
> > >>>>>>>>>>>>>>>> +    list_move_tail(&task->list_entry, &thread->task_busy_list);
> > >>>>>>>>>>>>>>>> +    spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> > >>>>>>>>>>>>>>>> +}
> > >>>>>>>>>>>>>>>> +
> > >>>>>>>>>>>>>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
> > >>>>>>>>>>>>>>>> +                   struct cmdq_thread *thread, u32 irq_flag)
> > >>>>>>>>>>>>>>>> +{
> > >>>>>>>>>>>>>>>> +    struct cmdq_task *task, *tmp, *curr_task = NULL;
> > >>>>>>>>>>>>>>>> +    u32 curr_pa;
> > >>>>>>>>>>>>>>>> +    struct cmdq_cb_data cmdq_cb_data;
> > >>>>>>>>>>>>>>>> +    bool err;
> > >>>>>>>>>>>>>>>> +
> > >>>>>>>>>>>>>>>> +    if (irq_flag & CMDQ_THR_IRQ_ERROR)
> > >>>>>>>>>>>>>>>> +        err = true;
> > >>>>>>>>>>>>>>>> +    else if (irq_flag & CMDQ_THR_IRQ_DONE)
> > >>>>>>>>>>>>>>>> +        err = false;
> > >>>>>>>>>>>>>>>> +    else
> > >>>>>>>>>>>>>>>> +        return;
> > >>>>>>>>>>>>>>>> +
> > >>>>>>>>>>>>>>>> +    curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> > >>>>>>>>>>>>>>>> +
> > >>>>>>>>>>>>>>>> +    list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
> > >>>>>>>>>>>>>>>> +                 list_entry) {
> > >>>>>>>>>>>>>>>> +        if (curr_pa >= task->pa_base &&
> > >>>>>>>>>>>>>>>> +            curr_pa < (task->pa_base + task->command_size))
> > >>>>>>>>>>>>>>>
> > >>>>>>>>>>>>>>> What are you checking here? It seems as if you make some implcit
> > >>>>>>>>>>>>>>> assumptions about pa_base and the order of execution of
> > >>>>>>>>>>>>>>> commands in the
> > >>>>>>>>>>>>>>> thread. Is it save to do so? Does dma_alloc_coherent give any
> > >>>>>>>>>>>>>>> guarantees
> > >>>>>>>>>>>>>>> about dma_handle?
> > >>>>>>>>>>>>>>
> > >>>>>>>>>>>>>> 1. Check what is the current running task in this GCE thread.
> > >>>>>>>>>>>>>> 2. Yes.
> > >>>>>>>>>>>>>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
> > >>>>>>>>>>>>>>
> > >>>>>>>>>>>>>
> > >>>>>>>>>>>>> Yes, physical addresses might be continous, but AFAIK there is no
> > >>>>>>>>>>>>> guarantee that the dma_handle address is steadily growing, when
> > >>>>>>>>>>>>> calling
> > >>>>>>>>>>>>> dma_alloc_coherent. And if I understand the code correctly, you
> > >>>>>>>>>>>>> use this
> > >>>>>>>>>>>>> assumption to decide if the task picked from task_busy_list is
> > >>>>>>>>>>>>> currently
> > >>>>>>>>>>>>> executing. So I think this mecanism is not working.
> > >>>>>>>>>>>>
> > >>>>>>>>>>>> I don't use dma_handle address, and just use physical addresses.
> > >>>>>>>>>>>>       From CPU's point of view, tasks are linked by the busy list.
> > >>>>>>>>>>>>       From GCE's point of view, tasks are linked by the JUMP command.
> > >>>>>>>>>>>>
> > >>>>>>>>>>>>> In which cases does the HW thread raise an interrupt.
> > >>>>>>>>>>>>> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
> > >>>>>>>>>>>>
> > >>>>>>>>>>>> GCE will raise interrupt if any task is done or error.
> > >>>>>>>>>>>> However, GCE is fast, so CPU may get multiple done tasks
> > >>>>>>>>>>>> when it is running ISR.
> > >>>>>>>>>>>>
> > >>>>>>>>>>>> In case of error, that GCE thread will pause and raise interrupt.
> > >>>>>>>>>>>> So, CPU may get multiple done tasks and one error task.
> > >>>>>>>>>>>>
> > >>>>>>>>>>>
> > >>>>>>>>>>> I think we should reimplement the ISR mechanism. Can't we just read
> > >>>>>>>>>>> CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave
> > >>>>>>>>>>> cmdq_handle_error_done to the thread_fn? You will need to pass
> > >>>>>>>>>>> information from the handler to thread_fn, but that shouldn't be an
> > >>>>>>>>>>> issue. AFAIK interrupts are disabled in the handler, so we should stay
> > >>>>>>>>>>> there as short as possible. Traversing task_busy_list is expensive, so
> > >>>>>>>>>>> we need to do it in a thread context.
> > >>>>>>>>>>
> > >>>>>>>>>> Actually, our initial implementation is similar to your suggestion,
> > >>>>>>>>>> but display needs CMDQ to return callback function very precisely,
> > >>>>>>>>>> else display will drop frame.
> > >>>>>>>>>> For display, CMDQ interrupt will be raised every 16 ~ 17 ms,
> > >>>>>>>>>> and CMDQ needs to call callback function in ISR.
> > >>>>>>>>>> If we defer callback to workqueue, the time interval may be larger than
> > >>>>>>>>>> 32 ms.sometimes.
> > >>>>>>>>>>
> > >>>>>>>>>
> > >>>>>>>>> I think the problem is, that you implemented the workqueue as a ordered
> > >>>>>>>>> workqueue, so there is no parallel processing. I'm still not sure why
> > >>>>>>>>> you need the workqueue to be ordered. Can you please explain.
> > >>>>>>>>
> > >>>>>>>> The order should be kept.
> > >>>>>>>> Let me use mouse cursor as an example.
> > >>>>>>>> If task 1 means move mouse cursor to point A, task 2 means point B,
> > >>>>>>>> and task 3 means point C, our expected result is A -> B -> C.
> > >>>>>>>> If the order is not kept, the result could become A -> C -> B.
> > >>>>>>>>
> > >>>>>>>
> > >>>>>>> Got it, thanks for the clarification.
> > >>>>>>>
> > >>>>>>
> > >>>>>> I think a way to get rid of the workqueue is to use a timer, which gets
> > >>>>>> programmed to the time a timeout in the first task in the busy list
> > >>>>>> would happen. Everytime we update the busy list (e.g. because of task
> > >>>>>> got finished by the thread), we update the timer. When the timer
> > >>>>>> triggers, which hopefully won't happen too often, we return timeout on
> > >>>>>> the busy list elements, until the time is lower then the actual time.
> > >>>>>>
> > >>>>>> At least with this we can reduce the data structures in this driver and
> > >>>>>> make it more lightweight.
> > >>>>>
> > >>>>>    From my understanding, your proposed method can handle timeout case.
> > >>>>>
> > >>>>> However, the workqueue is also in charge of releasing tasks.
> > >>>>> Do you take releasing tasks into consideration by using the proposed
> > >>>>> timer method?
> > >>>>> Furthermore, I think the code will become more complex if we also use
> > >>>>> timer to implement releasing tasks.
> > >>>>>
> > >>>>
> > >>>> Can't we call
> > >>>>            clk_disable_unprepare(cmdq->clock);
> > >>>>            cmdq_task_release(task);
> > >>>> after invoking the callback?

After I put clk_disable_unprepare(cmdq->clock) into ISR, I encounter
another BUG.

(Quote some Linux 4.7 source code.)

 605 void clk_unprepare(struct clk *clk)
 606 {
 607         if (IS_ERR_OR_NULL(clk))
 608                 return;
 609
 610         clk_prepare_lock();                      // <-- Here
 611         clk_core_unprepare(clk->core);
 612         clk_prepare_unlock();
 613 }
 614 EXPORT_SYMBOL_GPL(clk_unprepare);

  91 static void clk_prepare_lock(void)
  92 {
  93         if (!mutex_trylock(&prepare_lock)) {     // <-- Here
  94                 if (prepare_owner == current) {
  95                         prepare_refcnt++;
  96                         return;
  97                 }
  98                 mutex_lock(&prepare_lock);
  99         }
 100         WARN_ON_ONCE(prepare_owner != NULL);
 101         WARN_ON_ONCE(prepare_refcnt != 0);
 102         prepare_owner = current;
 103         prepare_refcnt = 1;
 104 }

So, 'unprepare' can sleep and cannot be put into ISR.
I also try to put it into a timer, but the error is the same
since timer callback is executed by softirq.

We need clk_disable_unprepare() since it can save power consumption
in idle.
Therefore, I plan to
(1) move releasing buffer and task into ISR,
(2) move timeout into timer, and
(3) keep workqueue for clk_disable_unprepare().

What do you think?

Thanks,
HS

> > >>>
> > >>> Do you mean just call these two functions in ISR?
> > >>> My major concern is dma_free_coherent() and kfree() in
> > >>> cmdq_task_release(task).
> > >>
> > >> Why do we need the dma calls at all? Can't we just calculate the
> > >> physical address using __pa(x)?
> > >
> > > I prefer to use dma_map_single/dma_unmap_single.
> > >
> > 
> > Can you please elaborate why you need this. We don't do dma, so we 
> > should not use dma memory for this.
> 
> We need a buffer to share between CPU and GCE, so we do need DMA.
> CPU is in charge of writing GCE commands into this buffer.
> GCE is in charge of reading and running GCE commands from this buffer.
> When we chain CMDQ tasks, we also need to modify GCE JUMP command.
> Therefore, I prefer to use dma_alloc_coherent and dma_free_coherent.
> 
> However, if we want to use timer to handle timeout, we need to release
> memory in ISR.
> In this case, using kmalloc/kfree + dma_map_single/dma_unmap_single
> instead of dma_alloc_coherent/dma_free_coherent is an alternative
> solution, but taking care the synchronization between cache and memory
> is the expected overhead.
> 
> > >>> Therefore, your suggestion is to use GFP_ATOMIC for both
> > >>> dma_alloc_coherent() and kzalloc(). Right?
> > >>
> > >> I don't think we need GFP_ATOMIC, the critical path will just free the
> > >> memory.
> > >
> > > I tested these two functions, and kfree was safe.
> > > However, dma_free_coherent raised BUG.
> > > BUG: failure at
> > > /mnt/host/source/src/third_party/kernel/v3.18/mm/vmalloc.c:1514/vunmap()!
> > 
> > Just a general hint. Please try to evaluate on a recent kernel. It looks 
> > like as if you tried this on a v3.18 based one.
> 
> This driver should be backward compatible to v3.18 for a MTK project.
> 
> > Best regards,
> > Matthias
> 
> Thanks,
> HS
> 
> > > 1512 void vunmap(const void *addr)
> > > 1513 {
> > > 1514         BUG_ON(in_interrupt());		// <-- here
> > > 1515         might_sleep();
> > > 1516         if (addr)
> > > 1517                 __vunmap(addr, 0);
> > > 1518 }
> > > 1519 EXPORT_SYMBOL(vunmap);
> > >
> > > Therefore, I plan to use kmalloc + dma_map_single instead of
> > > dma_alloc_coherent, and dma_unmap_single + kfree instead of
> > > dma_free_coherent.
> > >
> > > What do you think about the function replacement?
> > >
> > >>> If so, I can try to implement timeout by timer, and discuss with you
> > >>> if I have further questions.
> > >>>
> > >>
> > >> Sounds good :)
> > >>
> > >> Thanks,
> > >> Matthias
> > >
> > > Thanks,
> > > HS
> > >
> > >>>> Regrading the clock, wouldn't it be easier to handle the clock
> > >>>> enable/disable depending on the state of task_busy_list? I suppose we
> > >>>> can't as we would need to check the task_busy_list of all threads, right?
> > >>>>
> > >>>> Regards,
> > >>>> Matthias
> > >>>
> > >>> Thanks,
> > >>> HS
> > >>>
> > >
> > >
> 
>

Matthias Brugger June 17, 2016, 3:57 p.m. UTC | #26

On 17/06/16 10:28, Horng-Shyang Liao wrote:
> Hi Matthias,
>
> On Tue, 2016-06-14 at 20:07 +0800, Horng-Shyang Liao wrote:
>> Hi Matthias,
>>
>> On Tue, 2016-06-14 at 12:17 +0200, Matthias Brugger wrote:
>>>
>>> On 14/06/16 09:44, Horng-Shyang Liao wrote:
>>>> Hi Matthias,
>>>>
>>>> On Wed, 2016-06-08 at 17:35 +0200, Matthias Brugger wrote:
>>>>>
>>>>> On 08/06/16 14:25, Horng-Shyang Liao wrote:
>>>>>> Hi Matthias,
>>>>>>
>>>>>> On Wed, 2016-06-08 at 12:45 +0200, Matthias Brugger wrote:
>>>>>>>
>>>>>>> On 08/06/16 07:40, Horng-Shyang Liao wrote:
>>>>>>>> Hi Matthias,
>>>>>>>>
>>>>>>>> On Tue, 2016-06-07 at 18:59 +0200, Matthias Brugger wrote:
>>>>>>>>>
>>>>>>>>> On 03/06/16 15:11, Matthias Brugger wrote:
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>> [...]
>>>>>>>>>
>>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>>> +            smp_mb(); /* modify jump before enable thread */
>>>>>>>>>>>>>>>>>>> +        }
>>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>>> +        cmdq_thread_writel(thread, task->pa_base +
>>>>>>>>>>>>>>>>>>> task->command_size,
>>>>>>>>>>>>>>>>>>> +                   CMDQ_THR_END_ADDR);
>>>>>>>>>>>>>>>>>>> +        cmdq_thread_resume(thread);
>>>>>>>>>>>>>>>>>>> +    }
>>>>>>>>>>>>>>>>>>> +    list_move_tail(&task->list_entry, &thread->task_busy_list);
>>>>>>>>>>>>>>>>>>> +    spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>>>>>>>>>>>>>>>> +}
>>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
>>>>>>>>>>>>>>>>>>> +                   struct cmdq_thread *thread, u32 irq_flag)
>>>>>>>>>>>>>>>>>>> +{
>>>>>>>>>>>>>>>>>>> +    struct cmdq_task *task, *tmp, *curr_task = NULL;
>>>>>>>>>>>>>>>>>>> +    u32 curr_pa;
>>>>>>>>>>>>>>>>>>> +    struct cmdq_cb_data cmdq_cb_data;
>>>>>>>>>>>>>>>>>>> +    bool err;
>>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>>> +    if (irq_flag & CMDQ_THR_IRQ_ERROR)
>>>>>>>>>>>>>>>>>>> +        err = true;
>>>>>>>>>>>>>>>>>>> +    else if (irq_flag & CMDQ_THR_IRQ_DONE)
>>>>>>>>>>>>>>>>>>> +        err = false;
>>>>>>>>>>>>>>>>>>> +    else
>>>>>>>>>>>>>>>>>>> +        return;
>>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>>> +    curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
>>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>>> +    list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
>>>>>>>>>>>>>>>>>>> +                 list_entry) {
>>>>>>>>>>>>>>>>>>> +        if (curr_pa >= task->pa_base &&
>>>>>>>>>>>>>>>>>>> +            curr_pa < (task->pa_base + task->command_size))
>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>> What are you checking here? It seems as if you make some implcit
>>>>>>>>>>>>>>>>>> assumptions about pa_base and the order of execution of
>>>>>>>>>>>>>>>>>> commands in the
>>>>>>>>>>>>>>>>>> thread. Is it save to do so? Does dma_alloc_coherent give any
>>>>>>>>>>>>>>>>>> guarantees
>>>>>>>>>>>>>>>>>> about dma_handle?
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>> 1. Check what is the current running task in this GCE thread.
>>>>>>>>>>>>>>>>> 2. Yes.
>>>>>>>>>>>>>>>>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> Yes, physical addresses might be continous, but AFAIK there is no
>>>>>>>>>>>>>>>> guarantee that the dma_handle address is steadily growing, when
>>>>>>>>>>>>>>>> calling
>>>>>>>>>>>>>>>> dma_alloc_coherent. And if I understand the code correctly, you
>>>>>>>>>>>>>>>> use this
>>>>>>>>>>>>>>>> assumption to decide if the task picked from task_busy_list is
>>>>>>>>>>>>>>>> currently
>>>>>>>>>>>>>>>> executing. So I think this mecanism is not working.
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> I don't use dma_handle address, and just use physical addresses.
>>>>>>>>>>>>>>>        From CPU's point of view, tasks are linked by the busy list.
>>>>>>>>>>>>>>>        From GCE's point of view, tasks are linked by the JUMP command.
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> In which cases does the HW thread raise an interrupt.
>>>>>>>>>>>>>>>> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> GCE will raise interrupt if any task is done or error.
>>>>>>>>>>>>>>> However, GCE is fast, so CPU may get multiple done tasks
>>>>>>>>>>>>>>> when it is running ISR.
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> In case of error, that GCE thread will pause and raise interrupt.
>>>>>>>>>>>>>>> So, CPU may get multiple done tasks and one error task.
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> I think we should reimplement the ISR mechanism. Can't we just read
>>>>>>>>>>>>>> CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave
>>>>>>>>>>>>>> cmdq_handle_error_done to the thread_fn? You will need to pass
>>>>>>>>>>>>>> information from the handler to thread_fn, but that shouldn't be an
>>>>>>>>>>>>>> issue. AFAIK interrupts are disabled in the handler, so we should stay
>>>>>>>>>>>>>> there as short as possible. Traversing task_busy_list is expensive, so
>>>>>>>>>>>>>> we need to do it in a thread context.
>>>>>>>>>>>>>
>>>>>>>>>>>>> Actually, our initial implementation is similar to your suggestion,
>>>>>>>>>>>>> but display needs CMDQ to return callback function very precisely,
>>>>>>>>>>>>> else display will drop frame.
>>>>>>>>>>>>> For display, CMDQ interrupt will be raised every 16 ~ 17 ms,
>>>>>>>>>>>>> and CMDQ needs to call callback function in ISR.
>>>>>>>>>>>>> If we defer callback to workqueue, the time interval may be larger than
>>>>>>>>>>>>> 32 ms.sometimes.
>>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> I think the problem is, that you implemented the workqueue as a ordered
>>>>>>>>>>>> workqueue, so there is no parallel processing. I'm still not sure why
>>>>>>>>>>>> you need the workqueue to be ordered. Can you please explain.
>>>>>>>>>>>
>>>>>>>>>>> The order should be kept.
>>>>>>>>>>> Let me use mouse cursor as an example.
>>>>>>>>>>> If task 1 means move mouse cursor to point A, task 2 means point B,
>>>>>>>>>>> and task 3 means point C, our expected result is A -> B -> C.
>>>>>>>>>>> If the order is not kept, the result could become A -> C -> B.
>>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> Got it, thanks for the clarification.
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> I think a way to get rid of the workqueue is to use a timer, which gets
>>>>>>>>> programmed to the time a timeout in the first task in the busy list
>>>>>>>>> would happen. Everytime we update the busy list (e.g. because of task
>>>>>>>>> got finished by the thread), we update the timer. When the timer
>>>>>>>>> triggers, which hopefully won't happen too often, we return timeout on
>>>>>>>>> the busy list elements, until the time is lower then the actual time.
>>>>>>>>>
>>>>>>>>> At least with this we can reduce the data structures in this driver and
>>>>>>>>> make it more lightweight.
>>>>>>>>
>>>>>>>>     From my understanding, your proposed method can handle timeout case.
>>>>>>>>
>>>>>>>> However, the workqueue is also in charge of releasing tasks.
>>>>>>>> Do you take releasing tasks into consideration by using the proposed
>>>>>>>> timer method?
>>>>>>>> Furthermore, I think the code will become more complex if we also use
>>>>>>>> timer to implement releasing tasks.
>>>>>>>>
>>>>>>>
>>>>>>> Can't we call
>>>>>>>             clk_disable_unprepare(cmdq->clock);
>>>>>>>             cmdq_task_release(task);
>>>>>>> after invoking the callback?
>
> After I put clk_disable_unprepare(cmdq->clock) into ISR, I encounter
> another BUG.
>
> (Quote some Linux 4.7 source code.)
>
>   605 void clk_unprepare(struct clk *clk)
>   606 {
>   607         if (IS_ERR_OR_NULL(clk))
>   608                 return;
>   609
>   610         clk_prepare_lock();                      // <-- Here
>   611         clk_core_unprepare(clk->core);
>   612         clk_prepare_unlock();
>   613 }
>   614 EXPORT_SYMBOL_GPL(clk_unprepare);
>
>    91 static void clk_prepare_lock(void)
>    92 {
>    93         if (!mutex_trylock(&prepare_lock)) {     // <-- Here
>    94                 if (prepare_owner == current) {
>    95                         prepare_refcnt++;
>    96                         return;
>    97                 }
>    98                 mutex_lock(&prepare_lock);
>    99         }
>   100         WARN_ON_ONCE(prepare_owner != NULL);
>   101         WARN_ON_ONCE(prepare_refcnt != 0);
>   102         prepare_owner = current;
>   103         prepare_refcnt = 1;
>   104 }
>
> So, 'unprepare' can sleep and cannot be put into ISR.
> I also try to put it into a timer, but the error is the same
> since timer callback is executed by softirq.
>
> We need clk_disable_unprepare() since it can save power consumption
> in idle.

We can call clk_prepare in probe and then use clk_enable/clk_disable, 
which don't sleep.

Regards,
Matthias

> Therefore, I plan to
> (1) move releasing buffer and task into ISR,
> (2) move timeout into timer, and
> (3) keep workqueue for clk_disable_unprepare().
>
> What do you think?
>
> Thanks,
> HS
>
>>>>>>
>>>>>> Do you mean just call these two functions in ISR?
>>>>>> My major concern is dma_free_coherent() and kfree() in
>>>>>> cmdq_task_release(task).
>>>>>
>>>>> Why do we need the dma calls at all? Can't we just calculate the
>>>>> physical address using __pa(x)?
>>>>
>>>> I prefer to use dma_map_single/dma_unmap_single.
>>>>
>>>
>>> Can you please elaborate why you need this. We don't do dma, so we
>>> should not use dma memory for this.
>>
>> We need a buffer to share between CPU and GCE, so we do need DMA.
>> CPU is in charge of writing GCE commands into this buffer.
>> GCE is in charge of reading and running GCE commands from this buffer.
>> When we chain CMDQ tasks, we also need to modify GCE JUMP command.
>> Therefore, I prefer to use dma_alloc_coherent and dma_free_coherent.
>>
>> However, if we want to use timer to handle timeout, we need to release
>> memory in ISR.
>> In this case, using kmalloc/kfree + dma_map_single/dma_unmap_single
>> instead of dma_alloc_coherent/dma_free_coherent is an alternative
>> solution, but taking care the synchronization between cache and memory
>> is the expected overhead.
>>
>>>>>> Therefore, your suggestion is to use GFP_ATOMIC for both
>>>>>> dma_alloc_coherent() and kzalloc(). Right?
>>>>>
>>>>> I don't think we need GFP_ATOMIC, the critical path will just free the
>>>>> memory.
>>>>
>>>> I tested these two functions, and kfree was safe.
>>>> However, dma_free_coherent raised BUG.
>>>> BUG: failure at
>>>> /mnt/host/source/src/third_party/kernel/v3.18/mm/vmalloc.c:1514/vunmap()!
>>>
>>> Just a general hint. Please try to evaluate on a recent kernel. It looks
>>> like as if you tried this on a v3.18 based one.
>>
>> This driver should be backward compatible to v3.18 for a MTK project.
>>
>>> Best regards,
>>> Matthias
>>
>> Thanks,
>> HS
>>
>>>> 1512 void vunmap(const void *addr)
>>>> 1513 {
>>>> 1514         BUG_ON(in_interrupt());		// <-- here
>>>> 1515         might_sleep();
>>>> 1516         if (addr)
>>>> 1517                 __vunmap(addr, 0);
>>>> 1518 }
>>>> 1519 EXPORT_SYMBOL(vunmap);
>>>>
>>>> Therefore, I plan to use kmalloc + dma_map_single instead of
>>>> dma_alloc_coherent, and dma_unmap_single + kfree instead of
>>>> dma_free_coherent.
>>>>
>>>> What do you think about the function replacement?
>>>>
>>>>>> If so, I can try to implement timeout by timer, and discuss with you
>>>>>> if I have further questions.
>>>>>>
>>>>>
>>>>> Sounds good :)
>>>>>
>>>>> Thanks,
>>>>> Matthias
>>>>
>>>> Thanks,
>>>> HS
>>>>
>>>>>>> Regrading the clock, wouldn't it be easier to handle the clock
>>>>>>> enable/disable depending on the state of task_busy_list? I suppose we
>>>>>>> can't as we would need to check the task_busy_list of all threads, right?
>>>>>>>
>>>>>>> Regards,
>>>>>>> Matthias
>>>>>>
>>>>>> Thanks,
>>>>>> HS
>>>>>>
>>>>
>>>>
>>
>>
>
>

Matthias Brugger June 17, 2016, 4:14 p.m. UTC | #27

On 14/06/16 14:07, Horng-Shyang Liao wrote:
> Hi Matthias,
>
> On Tue, 2016-06-14 at 12:17 +0200, Matthias Brugger wrote:
>>
>> On 14/06/16 09:44, Horng-Shyang Liao wrote:
>>> Hi Matthias,
>>>
>>> On Wed, 2016-06-08 at 17:35 +0200, Matthias Brugger wrote:
>>>>
>>>> On 08/06/16 14:25, Horng-Shyang Liao wrote:
>>>>> Hi Matthias,
>>>>>
>>>>> On Wed, 2016-06-08 at 12:45 +0200, Matthias Brugger wrote:
>>>>>>
>>>>>> On 08/06/16 07:40, Horng-Shyang Liao wrote:
>>>>>>> Hi Matthias,
>>>>>>>
>>>>>>> On Tue, 2016-06-07 at 18:59 +0200, Matthias Brugger wrote:
>>>>>>>>
>>>>>>>> On 03/06/16 15:11, Matthias Brugger wrote:
>>>>>>>>>
>>>>>>>>>
>>>>>>>> [...]
>>>>>>>>
>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>> +            smp_mb(); /* modify jump before enable thread */
>>>>>>>>>>>>>>>>>> +        }
>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>> +        cmdq_thread_writel(thread, task->pa_base +
>>>>>>>>>>>>>>>>>> task->command_size,
>>>>>>>>>>>>>>>>>> +                   CMDQ_THR_END_ADDR);
>>>>>>>>>>>>>>>>>> +        cmdq_thread_resume(thread);
>>>>>>>>>>>>>>>>>> +    }
>>>>>>>>>>>>>>>>>> +    list_move_tail(&task->list_entry, &thread->task_busy_list);
>>>>>>>>>>>>>>>>>> +    spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>>>>>>>>>>>>>>> +}
>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
>>>>>>>>>>>>>>>>>> +                   struct cmdq_thread *thread, u32 irq_flag)
>>>>>>>>>>>>>>>>>> +{
>>>>>>>>>>>>>>>>>> +    struct cmdq_task *task, *tmp, *curr_task = NULL;
>>>>>>>>>>>>>>>>>> +    u32 curr_pa;
>>>>>>>>>>>>>>>>>> +    struct cmdq_cb_data cmdq_cb_data;
>>>>>>>>>>>>>>>>>> +    bool err;
>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>> +    if (irq_flag & CMDQ_THR_IRQ_ERROR)
>>>>>>>>>>>>>>>>>> +        err = true;
>>>>>>>>>>>>>>>>>> +    else if (irq_flag & CMDQ_THR_IRQ_DONE)
>>>>>>>>>>>>>>>>>> +        err = false;
>>>>>>>>>>>>>>>>>> +    else
>>>>>>>>>>>>>>>>>> +        return;
>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>> +    curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>> +    list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
>>>>>>>>>>>>>>>>>> +                 list_entry) {
>>>>>>>>>>>>>>>>>> +        if (curr_pa >= task->pa_base &&
>>>>>>>>>>>>>>>>>> +            curr_pa < (task->pa_base + task->command_size))
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>> What are you checking here? It seems as if you make some implcit
>>>>>>>>>>>>>>>>> assumptions about pa_base and the order of execution of
>>>>>>>>>>>>>>>>> commands in the
>>>>>>>>>>>>>>>>> thread. Is it save to do so? Does dma_alloc_coherent give any
>>>>>>>>>>>>>>>>> guarantees
>>>>>>>>>>>>>>>>> about dma_handle?
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> 1. Check what is the current running task in this GCE thread.
>>>>>>>>>>>>>>>> 2. Yes.
>>>>>>>>>>>>>>>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> Yes, physical addresses might be continous, but AFAIK there is no
>>>>>>>>>>>>>>> guarantee that the dma_handle address is steadily growing, when
>>>>>>>>>>>>>>> calling
>>>>>>>>>>>>>>> dma_alloc_coherent. And if I understand the code correctly, you
>>>>>>>>>>>>>>> use this
>>>>>>>>>>>>>>> assumption to decide if the task picked from task_busy_list is
>>>>>>>>>>>>>>> currently
>>>>>>>>>>>>>>> executing. So I think this mecanism is not working.
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> I don't use dma_handle address, and just use physical addresses.
>>>>>>>>>>>>>>        From CPU's point of view, tasks are linked by the busy list.
>>>>>>>>>>>>>>        From GCE's point of view, tasks are linked by the JUMP command.
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> In which cases does the HW thread raise an interrupt.
>>>>>>>>>>>>>>> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> GCE will raise interrupt if any task is done or error.
>>>>>>>>>>>>>> However, GCE is fast, so CPU may get multiple done tasks
>>>>>>>>>>>>>> when it is running ISR.
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> In case of error, that GCE thread will pause and raise interrupt.
>>>>>>>>>>>>>> So, CPU may get multiple done tasks and one error task.
>>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>> I think we should reimplement the ISR mechanism. Can't we just read
>>>>>>>>>>>>> CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave
>>>>>>>>>>>>> cmdq_handle_error_done to the thread_fn? You will need to pass
>>>>>>>>>>>>> information from the handler to thread_fn, but that shouldn't be an
>>>>>>>>>>>>> issue. AFAIK interrupts are disabled in the handler, so we should stay
>>>>>>>>>>>>> there as short as possible. Traversing task_busy_list is expensive, so
>>>>>>>>>>>>> we need to do it in a thread context.
>>>>>>>>>>>>
>>>>>>>>>>>> Actually, our initial implementation is similar to your suggestion,
>>>>>>>>>>>> but display needs CMDQ to return callback function very precisely,
>>>>>>>>>>>> else display will drop frame.
>>>>>>>>>>>> For display, CMDQ interrupt will be raised every 16 ~ 17 ms,
>>>>>>>>>>>> and CMDQ needs to call callback function in ISR.
>>>>>>>>>>>> If we defer callback to workqueue, the time interval may be larger than
>>>>>>>>>>>> 32 ms.sometimes.
>>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> I think the problem is, that you implemented the workqueue as a ordered
>>>>>>>>>>> workqueue, so there is no parallel processing. I'm still not sure why
>>>>>>>>>>> you need the workqueue to be ordered. Can you please explain.
>>>>>>>>>>
>>>>>>>>>> The order should be kept.
>>>>>>>>>> Let me use mouse cursor as an example.
>>>>>>>>>> If task 1 means move mouse cursor to point A, task 2 means point B,
>>>>>>>>>> and task 3 means point C, our expected result is A -> B -> C.
>>>>>>>>>> If the order is not kept, the result could become A -> C -> B.
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Got it, thanks for the clarification.
>>>>>>>>>
>>>>>>>>
>>>>>>>> I think a way to get rid of the workqueue is to use a timer, which gets
>>>>>>>> programmed to the time a timeout in the first task in the busy list
>>>>>>>> would happen. Everytime we update the busy list (e.g. because of task
>>>>>>>> got finished by the thread), we update the timer. When the timer
>>>>>>>> triggers, which hopefully won't happen too often, we return timeout on
>>>>>>>> the busy list elements, until the time is lower then the actual time.
>>>>>>>>
>>>>>>>> At least with this we can reduce the data structures in this driver and
>>>>>>>> make it more lightweight.
>>>>>>>
>>>>>>>     From my understanding, your proposed method can handle timeout case.
>>>>>>>
>>>>>>> However, the workqueue is also in charge of releasing tasks.
>>>>>>> Do you take releasing tasks into consideration by using the proposed
>>>>>>> timer method?
>>>>>>> Furthermore, I think the code will become more complex if we also use
>>>>>>> timer to implement releasing tasks.
>>>>>>>
>>>>>>
>>>>>> Can't we call
>>>>>>             clk_disable_unprepare(cmdq->clock);
>>>>>>             cmdq_task_release(task);
>>>>>> after invoking the callback?
>>>>>
>>>>> Do you mean just call these two functions in ISR?
>>>>> My major concern is dma_free_coherent() and kfree() in
>>>>> cmdq_task_release(task).
>>>>
>>>> Why do we need the dma calls at all? Can't we just calculate the
>>>> physical address using __pa(x)?
>>>
>>> I prefer to use dma_map_single/dma_unmap_single.
>>>
>>
>> Can you please elaborate why you need this. We don't do dma, so we
>> should not use dma memory for this.
>
> We need a buffer to share between CPU and GCE, so we do need DMA.
> CPU is in charge of writing GCE commands into this buffer.
> GCE is in charge of reading and running GCE commands from this buffer.
> When we chain CMDQ tasks, we also need to modify GCE JUMP command.
> Therefore, I prefer to use dma_alloc_coherent and dma_free_coherent.
>
> However, if we want to use timer to handle timeout, we need to release
> memory in ISR.
> In this case, using kmalloc/kfree + dma_map_single/dma_unmap_single
> instead of dma_alloc_coherent/dma_free_coherent is an alternative
> solution, but taking care the synchronization between cache and memory
> is the expected overhead.
>
>>>>> Therefore, your suggestion is to use GFP_ATOMIC for both
>>>>> dma_alloc_coherent() and kzalloc(). Right?
>>>>
>>>> I don't think we need GFP_ATOMIC, the critical path will just free the
>>>> memory.
>>>
>>> I tested these two functions, and kfree was safe.
>>> However, dma_free_coherent raised BUG.
>>> BUG: failure at
>>> /mnt/host/source/src/third_party/kernel/v3.18/mm/vmalloc.c:1514/vunmap()!
>>
>> Just a general hint. Please try to evaluate on a recent kernel. It looks
>> like as if you tried this on a v3.18 based one.
>
> This driver should be backward compatible to v3.18 for a MTK project.
>

That is something the Linux community can't use as argument for design 
decisions. If the backporting get's cumbersome, I propose to tell your 
boss, giving him the hint, that if the driver would have been in 
mainline earlier, this would all be easier ;)
No, seriously, that's why it makes a lot of sense for companies like 
Mediatek to have their driver mainlined. A switch to a new kernel 
version comes for free for all the mainlined drivers. Knowing the source 
code of the Mediatek kernel, I suppose doing a kernel version switch 
takes a good bunch of time and nerves. :)

Best regards,
Matthias

CK Hu (胡俊光) June 20, 2016, 10:41 a.m. UTC | #28

Hi, HS:

One comment inline.

On Mon, 2016-05-30 at 11:19 +0800, HS Liao wrote:
> This patch is first version of Mediatek Command Queue(CMDQ) driver. The
> CMDQ is used to help read/write registers with critical time limitation,
> such as updating display configuration during the vblank. It controls
> Global Command Engine (GCE) hardware to achieve this requirement.
> Currently, CMDQ only supports display related hardwares, but we expect
> it can be extended to other hardwares for future requirements.
> 
> Signed-off-by: HS Liao <[e-mail]>
> Signed-off-by: CK Hu <[e-mail]>
> ---
>  drivers/soc/mediatek/Kconfig    |  10 +
>  drivers/soc/mediatek/Makefile   |   1 +
>  drivers/soc/mediatek/mtk-cmdq.c | 943 ++++++++++++++++++++++++++++++++++++++++
>  include/soc/mediatek/cmdq.h     | 197 +++++++++
>  4 files changed, 1151 insertions(+)
>  create mode 100644 drivers/soc/mediatek/mtk-cmdq.c
>  create mode 100644 include/soc/mediatek/cmdq.h
> 
> diff --git a/drivers/soc/mediatek/Kconfig b/drivers/soc/mediatek/Kconfig
> index 0a4ea80..c4ad75c 100644
> --- a/drivers/soc/mediatek/Kconfig
> +++ b/drivers/soc/mediatek/Kconfig
> @@ -1,6 +1,16 @@
>  #
>  # MediaTek SoC drivers
>  #
> +config MTK_CMDQ
> +	bool "MediaTek CMDQ Support"
> +	depends on ARCH_MEDIATEK || COMPILE_TEST
> +	select MTK_INFRACFG
> +	help
> +	  Say yes here to add support for the MediaTek Command Queue (CMDQ)
> +	  driver. The CMDQ is used to help read/write registers with critical
> +	  time limitation, such as updating display configuration during the
> +	  vblank.
> +
>  config MTK_INFRACFG
>  	bool "MediaTek INFRACFG Support"
>  	depends on ARCH_MEDIATEK || COMPILE_TEST
> diff --git a/drivers/soc/mediatek/Makefile b/drivers/soc/mediatek/Makefile
> index 12998b0..f7397ef 100644
> --- a/drivers/soc/mediatek/Makefile
> +++ b/drivers/soc/mediatek/Makefile

[Snip...]

> +
> +static int cmdq_eng_get_thread(u64 flag)
> +{
> +	if (flag & BIT_ULL(CMDQ_ENG_DISP_DSI0))
> +		return CMDQ_THR_DISP_MAIN_IDX;
> +	else if (flag & BIT_ULL(CMDQ_ENG_DISP_DPI0))
> +		return CMDQ_THR_DISP_SUB_IDX;
> +	else
> +		return CMDQ_THR_DISP_MISC_IDX;
> +}

I think cmdq should not have knowledge of client. These statement show
that cmdq know that DSI0-tasks is different with DPI0-tasks. I propose
the 'session' to replace engine flag. For example, main display create
one cmdq_session and external display create another cmdq_session. For
client driver, every tasks created by main display is bound to main
display session. For cmdq driver, it should dynamically bind a session
to a HW thread, and then dispatch tasks of this session to this HW
thread. After HW thread run out of tasks of this session, detach this
session with this thread.
For the problem of remove wfe cmd, I think a session can have a property
of merge_wfe_cmd. For display session, session->merge_wfe_cmd = true.
For other client, it's false.

Here is the sample code to create cmdq_rec with session.

merge_wfe_cmd = true;
cmdq_session_create(merge_wfe_cmd, &primary_display_session);
cmdq_rec_create(dev, primary_display_session, &rec);


Therefore, the below definition can be removed.

> +
> +enum cmdq_eng {
> +	CMDQ_ENG_DISP_AAL,
> +	CMDQ_ENG_DISP_COLOR0,
> +	CMDQ_ENG_DISP_COLOR1,
> +	CMDQ_ENG_DISP_DPI0,
> +	CMDQ_ENG_DISP_DSI0,
> +	CMDQ_ENG_DISP_DSI1,
> +	CMDQ_ENG_DISP_GAMMA,
> +	CMDQ_ENG_DISP_OD,
> +	CMDQ_ENG_DISP_OVL0,
> +	CMDQ_ENG_DISP_OVL1,
> +	CMDQ_ENG_DISP_PWM0,
> +	CMDQ_ENG_DISP_PWM1,
> +	CMDQ_ENG_DISP_RDMA0,
> +	CMDQ_ENG_DISP_RDMA1,
> +	CMDQ_ENG_DISP_RDMA2,
> +	CMDQ_ENG_DISP_UFOE,
> +	CMDQ_ENG_DISP_WDMA0,
> +	CMDQ_ENG_DISP_WDMA1,
> +	CMDQ_ENG_MAX,
> +};
> +


Regards,
CK

[e-mail] June 20, 2016, 11:22 a.m. UTC | #29

On Mon, 2016-06-20 at 18:41 +0800, CK Hu wrote:
> Hi, HS:
> 
> One comment inline.
> 
> On Mon, 2016-05-30 at 11:19 +0800, HS Liao wrote:
> > This patch is first version of Mediatek Command Queue(CMDQ) driver. The
> > CMDQ is used to help read/write registers with critical time limitation,
> > such as updating display configuration during the vblank. It controls
> > Global Command Engine (GCE) hardware to achieve this requirement.
> > Currently, CMDQ only supports display related hardwares, but we expect
> > it can be extended to other hardwares for future requirements.
> > 
> > Signed-off-by: HS Liao <[e-mail]>
> > Signed-off-by: CK Hu <[e-mail]>
> > ---
> >  drivers/soc/mediatek/Kconfig    |  10 +
> >  drivers/soc/mediatek/Makefile   |   1 +
> >  drivers/soc/mediatek/mtk-cmdq.c | 943 ++++++++++++++++++++++++++++++++++++++++
> >  include/soc/mediatek/cmdq.h     | 197 +++++++++
> >  4 files changed, 1151 insertions(+)
> >  create mode 100644 drivers/soc/mediatek/mtk-cmdq.c
> >  create mode 100644 include/soc/mediatek/cmdq.h
> > 
> > diff --git a/drivers/soc/mediatek/Kconfig b/drivers/soc/mediatek/Kconfig
> > index 0a4ea80..c4ad75c 100644
> > --- a/drivers/soc/mediatek/Kconfig
> > +++ b/drivers/soc/mediatek/Kconfig
> > @@ -1,6 +1,16 @@
> >  #
> >  # MediaTek SoC drivers
> >  #
> > +config MTK_CMDQ
> > +	bool "MediaTek CMDQ Support"
> > +	depends on ARCH_MEDIATEK || COMPILE_TEST
> > +	select MTK_INFRACFG
> > +	help
> > +	  Say yes here to add support for the MediaTek Command Queue (CMDQ)
> > +	  driver. The CMDQ is used to help read/write registers with critical
> > +	  time limitation, such as updating display configuration during the
> > +	  vblank.
> > +
> >  config MTK_INFRACFG
> >  	bool "MediaTek INFRACFG Support"
> >  	depends on ARCH_MEDIATEK || COMPILE_TEST
> > diff --git a/drivers/soc/mediatek/Makefile b/drivers/soc/mediatek/Makefile
> > index 12998b0..f7397ef 100644
> > --- a/drivers/soc/mediatek/Makefile
> > +++ b/drivers/soc/mediatek/Makefile
> 
> [Snip...]
> 
> > +
> > +static int cmdq_eng_get_thread(u64 flag)
> > +{
> > +	if (flag & BIT_ULL(CMDQ_ENG_DISP_DSI0))
> > +		return CMDQ_THR_DISP_MAIN_IDX;
> > +	else if (flag & BIT_ULL(CMDQ_ENG_DISP_DPI0))
> > +		return CMDQ_THR_DISP_SUB_IDX;
> > +	else
> > +		return CMDQ_THR_DISP_MISC_IDX;
> > +}
> 
> I think cmdq should not have knowledge of client. These statement show
> that cmdq know that DSI0-tasks is different with DPI0-tasks. I propose
> the 'session' to replace engine flag. For example, main display create
> one cmdq_session and external display create another cmdq_session. For
> client driver, every tasks created by main display is bound to main
> display session. For cmdq driver, it should dynamically bind a session
> to a HW thread, and then dispatch tasks of this session to this HW
> thread. After HW thread run out of tasks of this session, detach this
> session with this thread.
> For the problem of remove wfe cmd, I think a session can have a property
> of merge_wfe_cmd. For display session, session->merge_wfe_cmd = true.
> For other client, it's false.

Hi CK,

I think your suggestion is similar to CMDQ 'scenarios',
which was removed from CMDQ v2.

Daniel suggests to use engine flags instead of scenarios.
Quote from https://patchwork.kernel.org/patch/8068311/ .
'Instead of encoding policy (these arbitrary "scenarios"), perhaps the
 cmdq driver should just provide these flag bits, and the display
 subsystem can use them to build the "flag" parameter itself.

 After all, the exact configuration of mmsys components is somewhat
 flexible.'

Therefore, it would be better to discuss with Daniel before we change
it.


Hi Daniel,

Do you think we should use scenarios or sessions instead of flags?

Thanks,
HS

> Here is the sample code to create cmdq_rec with session.
> 
> merge_wfe_cmd = true;
> cmdq_session_create(merge_wfe_cmd, &primary_display_session);
> cmdq_rec_create(dev, primary_display_session, &rec);
> 
> 
> Therefore, the below definition can be removed.
> 
> > +
> > +enum cmdq_eng {
> > +	CMDQ_ENG_DISP_AAL,
> > +	CMDQ_ENG_DISP_COLOR0,
> > +	CMDQ_ENG_DISP_COLOR1,
> > +	CMDQ_ENG_DISP_DPI0,
> > +	CMDQ_ENG_DISP_DSI0,
> > +	CMDQ_ENG_DISP_DSI1,
> > +	CMDQ_ENG_DISP_GAMMA,
> > +	CMDQ_ENG_DISP_OD,
> > +	CMDQ_ENG_DISP_OVL0,
> > +	CMDQ_ENG_DISP_OVL1,
> > +	CMDQ_ENG_DISP_PWM0,
> > +	CMDQ_ENG_DISP_PWM1,
> > +	CMDQ_ENG_DISP_RDMA0,
> > +	CMDQ_ENG_DISP_RDMA1,
> > +	CMDQ_ENG_DISP_RDMA2,
> > +	CMDQ_ENG_DISP_UFOE,
> > +	CMDQ_ENG_DISP_WDMA0,
> > +	CMDQ_ENG_DISP_WDMA1,
> > +	CMDQ_ENG_MAX,
> > +};
> > +
> 
> 
> Regards,
> CK
>

CK Hu (胡俊光) June 21, 2016, 2:03 a.m. UTC | #30

On Mon, 2016-06-20 at 19:22 +0800, Horng-Shyang Liao wrote:
> On Mon, 2016-06-20 at 18:41 +0800, CK Hu wrote:
> > Hi, HS:
> > 
> > One comment inline.
> > 
> > On Mon, 2016-05-30 at 11:19 +0800, HS Liao wrote:
> > > This patch is first version of Mediatek Command Queue(CMDQ) driver. The
> > > CMDQ is used to help read/write registers with critical time limitation,
> > > such as updating display configuration during the vblank. It controls
> > > Global Command Engine (GCE) hardware to achieve this requirement.
> > > Currently, CMDQ only supports display related hardwares, but we expect
> > > it can be extended to other hardwares for future requirements.
> > > 
> > > Signed-off-by: HS Liao <[e-mail]>
> > > Signed-off-by: CK Hu <[e-mail]>
> > > ---
> > >  drivers/soc/mediatek/Kconfig    |  10 +
> > >  drivers/soc/mediatek/Makefile   |   1 +
> > >  drivers/soc/mediatek/mtk-cmdq.c | 943 ++++++++++++++++++++++++++++++++++++++++
> > >  include/soc/mediatek/cmdq.h     | 197 +++++++++
> > >  4 files changed, 1151 insertions(+)
> > >  create mode 100644 drivers/soc/mediatek/mtk-cmdq.c
> > >  create mode 100644 include/soc/mediatek/cmdq.h
> > > 
> > > diff --git a/drivers/soc/mediatek/Kconfig b/drivers/soc/mediatek/Kconfig
> > > index 0a4ea80..c4ad75c 100644
> > > --- a/drivers/soc/mediatek/Kconfig
> > > +++ b/drivers/soc/mediatek/Kconfig
> > > @@ -1,6 +1,16 @@
> > >  #
> > >  # MediaTek SoC drivers
> > >  #
> > > +config MTK_CMDQ
> > > +	bool "MediaTek CMDQ Support"
> > > +	depends on ARCH_MEDIATEK || COMPILE_TEST
> > > +	select MTK_INFRACFG
> > > +	help
> > > +	  Say yes here to add support for the MediaTek Command Queue (CMDQ)
> > > +	  driver. The CMDQ is used to help read/write registers with critical
> > > +	  time limitation, such as updating display configuration during the
> > > +	  vblank.
> > > +
> > >  config MTK_INFRACFG
> > >  	bool "MediaTek INFRACFG Support"
> > >  	depends on ARCH_MEDIATEK || COMPILE_TEST
> > > diff --git a/drivers/soc/mediatek/Makefile b/drivers/soc/mediatek/Makefile
> > > index 12998b0..f7397ef 100644
> > > --- a/drivers/soc/mediatek/Makefile
> > > +++ b/drivers/soc/mediatek/Makefile
> > 
> > [Snip...]
> > 
> > > +
> > > +static int cmdq_eng_get_thread(u64 flag)
> > > +{
> > > +	if (flag & BIT_ULL(CMDQ_ENG_DISP_DSI0))
> > > +		return CMDQ_THR_DISP_MAIN_IDX;
> > > +	else if (flag & BIT_ULL(CMDQ_ENG_DISP_DPI0))
> > > +		return CMDQ_THR_DISP_SUB_IDX;
> > > +	else
> > > +		return CMDQ_THR_DISP_MISC_IDX;
> > > +}
> > 
> > I think cmdq should not have knowledge of client. These statement show
> > that cmdq know that DSI0-tasks is different with DPI0-tasks. I propose
> > the 'session' to replace engine flag. For example, main display create
> > one cmdq_session and external display create another cmdq_session. For
> > client driver, every tasks created by main display is bound to main
> > display session. For cmdq driver, it should dynamically bind a session
> > to a HW thread, and then dispatch tasks of this session to this HW
> > thread. After HW thread run out of tasks of this session, detach this
> > session with this thread.
> > For the problem of remove wfe cmd, I think a session can have a property
> > of merge_wfe_cmd. For display session, session->merge_wfe_cmd = true.
> > For other client, it's false.
> 
> Hi CK,
> 
> I think your suggestion is similar to CMDQ 'scenarios',
> which was removed from CMDQ v2.
> 
> Daniel suggests to use engine flags instead of scenarios.
> Quote from https://patchwork.kernel.org/patch/8068311/ .
> 'Instead of encoding policy (these arbitrary "scenarios"), perhaps the
>  cmdq driver should just provide these flag bits, and the display
>  subsystem can use them to build the "flag" parameter itself.
> 
>  After all, the exact configuration of mmsys components is somewhat
>  flexible.'
> 
> Therefore, it would be better to discuss with Daniel before we change
> it.
> 
> 
> Hi Daniel,
> 
> Do you think we should use scenarios or sessions instead of flags?
> 
> Thanks,
> HS
> 

Hi, HS:

'session' is not similar to 'scenarios'.

In 'scenarios' mechanism, client bind a task with scenarios and send to
cmdq. Cmdq transfer scenarios to engine flag, and use engine flag to
dispatch task to HW thread.


In 'engine flag' mechanism, proposed by Daniel, client bind a task
directly with engine flag. Cmdq directly use engine flag to dispatch
task to HW thread without any translation.

But neither 'scenarios' mechanism nor 'engine flag' mechanism get rid of
engine flag, which make cmdq have knowledge of client.

In 'session' mechanism, there is no engine flag any more. Client bind
time-sequential tasks to the same session and tasks in different session
can execute parallelly. One thing cmdq need to know is to dispatch tasks
with the same session to the same HW thread, so cmdq does not have any
knowledge of client.

Daniel focus on reduce translating scenarios to engine flag. I think
'session' mechanism does not conflict with his opinion because we does
not translate 'session' to engine flag. Therefore, I think 'session' is
the best of these three mechanism.

Regards,
CK

> > Here is the sample code to create cmdq_rec with session.
> > 
> > merge_wfe_cmd = true;
> > cmdq_session_create(merge_wfe_cmd, &primary_display_session);
> > cmdq_rec_create(dev, primary_display_session, &rec);
> > 
> > 
> > Therefore, the below definition can be removed.
> > 
> > > +
> > > +enum cmdq_eng {
> > > +	CMDQ_ENG_DISP_AAL,
> > > +	CMDQ_ENG_DISP_COLOR0,
> > > +	CMDQ_ENG_DISP_COLOR1,
> > > +	CMDQ_ENG_DISP_DPI0,
> > > +	CMDQ_ENG_DISP_DSI0,
> > > +	CMDQ_ENG_DISP_DSI1,
> > > +	CMDQ_ENG_DISP_GAMMA,
> > > +	CMDQ_ENG_DISP_OD,
> > > +	CMDQ_ENG_DISP_OVL0,
> > > +	CMDQ_ENG_DISP_OVL1,
> > > +	CMDQ_ENG_DISP_PWM0,
> > > +	CMDQ_ENG_DISP_PWM1,
> > > +	CMDQ_ENG_DISP_RDMA0,
> > > +	CMDQ_ENG_DISP_RDMA1,
> > > +	CMDQ_ENG_DISP_RDMA2,
> > > +	CMDQ_ENG_DISP_UFOE,
> > > +	CMDQ_ENG_DISP_WDMA0,
> > > +	CMDQ_ENG_DISP_WDMA1,
> > > +	CMDQ_ENG_MAX,
> > > +};
> > > +
> > 
> > 
> > Regards,
> > CK
> > 
> 
>

[e-mail] June 21, 2016, 5:52 a.m. UTC | #31

On Fri, 2016-06-17 at 17:57 +0200, Matthias Brugger wrote:
> 
> On 17/06/16 10:28, Horng-Shyang Liao wrote:
> > Hi Matthias,
> >
> > On Tue, 2016-06-14 at 20:07 +0800, Horng-Shyang Liao wrote:
> >> Hi Matthias,
> >>
> >> On Tue, 2016-06-14 at 12:17 +0200, Matthias Brugger wrote:
> >>>
> >>> On 14/06/16 09:44, Horng-Shyang Liao wrote:
> >>>> Hi Matthias,
> >>>>
> >>>> On Wed, 2016-06-08 at 17:35 +0200, Matthias Brugger wrote:
> >>>>>
> >>>>> On 08/06/16 14:25, Horng-Shyang Liao wrote:
> >>>>>> Hi Matthias,
> >>>>>>
> >>>>>> On Wed, 2016-06-08 at 12:45 +0200, Matthias Brugger wrote:
> >>>>>>>
> >>>>>>> On 08/06/16 07:40, Horng-Shyang Liao wrote:
> >>>>>>>> Hi Matthias,
> >>>>>>>>
> >>>>>>>> On Tue, 2016-06-07 at 18:59 +0200, Matthias Brugger wrote:
> >>>>>>>>>
> >>>>>>>>> On 03/06/16 15:11, Matthias Brugger wrote:
> >>>>>>>>>>
> >>>>>>>>>>
> >>>>>>>>> [...]
> >>>>>>>>>
> >>>>>>>>>>>>>>>>>>> +
> >>>>>>>>>>>>>>>>>>> +            smp_mb(); /* modify jump before enable thread */
> >>>>>>>>>>>>>>>>>>> +        }
> >>>>>>>>>>>>>>>>>>> +
> >>>>>>>>>>>>>>>>>>> +        cmdq_thread_writel(thread, task->pa_base +
> >>>>>>>>>>>>>>>>>>> task->command_size,
> >>>>>>>>>>>>>>>>>>> +                   CMDQ_THR_END_ADDR);
> >>>>>>>>>>>>>>>>>>> +        cmdq_thread_resume(thread);
> >>>>>>>>>>>>>>>>>>> +    }
> >>>>>>>>>>>>>>>>>>> +    list_move_tail(&task->list_entry, &thread->task_busy_list);
> >>>>>>>>>>>>>>>>>>> +    spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>>>>>>>>>>>>>>>>>> +}
> >>>>>>>>>>>>>>>>>>> +
> >>>>>>>>>>>>>>>>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
> >>>>>>>>>>>>>>>>>>> +                   struct cmdq_thread *thread, u32 irq_flag)
> >>>>>>>>>>>>>>>>>>> +{
> >>>>>>>>>>>>>>>>>>> +    struct cmdq_task *task, *tmp, *curr_task = NULL;
> >>>>>>>>>>>>>>>>>>> +    u32 curr_pa;
> >>>>>>>>>>>>>>>>>>> +    struct cmdq_cb_data cmdq_cb_data;
> >>>>>>>>>>>>>>>>>>> +    bool err;
> >>>>>>>>>>>>>>>>>>> +
> >>>>>>>>>>>>>>>>>>> +    if (irq_flag & CMDQ_THR_IRQ_ERROR)
> >>>>>>>>>>>>>>>>>>> +        err = true;
> >>>>>>>>>>>>>>>>>>> +    else if (irq_flag & CMDQ_THR_IRQ_DONE)
> >>>>>>>>>>>>>>>>>>> +        err = false;
> >>>>>>>>>>>>>>>>>>> +    else
> >>>>>>>>>>>>>>>>>>> +        return;
> >>>>>>>>>>>>>>>>>>> +
> >>>>>>>>>>>>>>>>>>> +    curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> >>>>>>>>>>>>>>>>>>> +
> >>>>>>>>>>>>>>>>>>> +    list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
> >>>>>>>>>>>>>>>>>>> +                 list_entry) {
> >>>>>>>>>>>>>>>>>>> +        if (curr_pa >= task->pa_base &&
> >>>>>>>>>>>>>>>>>>> +            curr_pa < (task->pa_base + task->command_size))
> >>>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>>> What are you checking here? It seems as if you make some implcit
> >>>>>>>>>>>>>>>>>> assumptions about pa_base and the order of execution of
> >>>>>>>>>>>>>>>>>> commands in the
> >>>>>>>>>>>>>>>>>> thread. Is it save to do so? Does dma_alloc_coherent give any
> >>>>>>>>>>>>>>>>>> guarantees
> >>>>>>>>>>>>>>>>>> about dma_handle?
> >>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>> 1. Check what is the current running task in this GCE thread.
> >>>>>>>>>>>>>>>>> 2. Yes.
> >>>>>>>>>>>>>>>>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
> >>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>> Yes, physical addresses might be continous, but AFAIK there is no
> >>>>>>>>>>>>>>>> guarantee that the dma_handle address is steadily growing, when
> >>>>>>>>>>>>>>>> calling
> >>>>>>>>>>>>>>>> dma_alloc_coherent. And if I understand the code correctly, you
> >>>>>>>>>>>>>>>> use this
> >>>>>>>>>>>>>>>> assumption to decide if the task picked from task_busy_list is
> >>>>>>>>>>>>>>>> currently
> >>>>>>>>>>>>>>>> executing. So I think this mecanism is not working.
> >>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>> I don't use dma_handle address, and just use physical addresses.
> >>>>>>>>>>>>>>>        From CPU's point of view, tasks are linked by the busy list.
> >>>>>>>>>>>>>>>        From GCE's point of view, tasks are linked by the JUMP command.
> >>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>> In which cases does the HW thread raise an interrupt.
> >>>>>>>>>>>>>>>> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
> >>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>> GCE will raise interrupt if any task is done or error.
> >>>>>>>>>>>>>>> However, GCE is fast, so CPU may get multiple done tasks
> >>>>>>>>>>>>>>> when it is running ISR.
> >>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>> In case of error, that GCE thread will pause and raise interrupt.
> >>>>>>>>>>>>>>> So, CPU may get multiple done tasks and one error task.
> >>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>> I think we should reimplement the ISR mechanism. Can't we just read
> >>>>>>>>>>>>>> CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave
> >>>>>>>>>>>>>> cmdq_handle_error_done to the thread_fn? You will need to pass
> >>>>>>>>>>>>>> information from the handler to thread_fn, but that shouldn't be an
> >>>>>>>>>>>>>> issue. AFAIK interrupts are disabled in the handler, so we should stay
> >>>>>>>>>>>>>> there as short as possible. Traversing task_busy_list is expensive, so
> >>>>>>>>>>>>>> we need to do it in a thread context.
> >>>>>>>>>>>>>
> >>>>>>>>>>>>> Actually, our initial implementation is similar to your suggestion,
> >>>>>>>>>>>>> but display needs CMDQ to return callback function very precisely,
> >>>>>>>>>>>>> else display will drop frame.
> >>>>>>>>>>>>> For display, CMDQ interrupt will be raised every 16 ~ 17 ms,
> >>>>>>>>>>>>> and CMDQ needs to call callback function in ISR.
> >>>>>>>>>>>>> If we defer callback to workqueue, the time interval may be larger than
> >>>>>>>>>>>>> 32 ms.sometimes.
> >>>>>>>>>>>>>
> >>>>>>>>>>>>
> >>>>>>>>>>>> I think the problem is, that you implemented the workqueue as a ordered
> >>>>>>>>>>>> workqueue, so there is no parallel processing. I'm still not sure why
> >>>>>>>>>>>> you need the workqueue to be ordered. Can you please explain.
> >>>>>>>>>>>
> >>>>>>>>>>> The order should be kept.
> >>>>>>>>>>> Let me use mouse cursor as an example.
> >>>>>>>>>>> If task 1 means move mouse cursor to point A, task 2 means point B,
> >>>>>>>>>>> and task 3 means point C, our expected result is A -> B -> C.
> >>>>>>>>>>> If the order is not kept, the result could become A -> C -> B.
> >>>>>>>>>>>
> >>>>>>>>>>
> >>>>>>>>>> Got it, thanks for the clarification.
> >>>>>>>>>>
> >>>>>>>>>
> >>>>>>>>> I think a way to get rid of the workqueue is to use a timer, which gets
> >>>>>>>>> programmed to the time a timeout in the first task in the busy list
> >>>>>>>>> would happen. Everytime we update the busy list (e.g. because of task
> >>>>>>>>> got finished by the thread), we update the timer. When the timer
> >>>>>>>>> triggers, which hopefully won't happen too often, we return timeout on
> >>>>>>>>> the busy list elements, until the time is lower then the actual time.
> >>>>>>>>>
> >>>>>>>>> At least with this we can reduce the data structures in this driver and
> >>>>>>>>> make it more lightweight.
> >>>>>>>>
> >>>>>>>>     From my understanding, your proposed method can handle timeout case.
> >>>>>>>>
> >>>>>>>> However, the workqueue is also in charge of releasing tasks.
> >>>>>>>> Do you take releasing tasks into consideration by using the proposed
> >>>>>>>> timer method?
> >>>>>>>> Furthermore, I think the code will become more complex if we also use
> >>>>>>>> timer to implement releasing tasks.
> >>>>>>>>
> >>>>>>>
> >>>>>>> Can't we call
> >>>>>>>             clk_disable_unprepare(cmdq->clock);
> >>>>>>>             cmdq_task_release(task);
> >>>>>>> after invoking the callback?
> >
> > After I put clk_disable_unprepare(cmdq->clock) into ISR, I encounter
> > another BUG.
> >
> > (Quote some Linux 4.7 source code.)
> >
> >   605 void clk_unprepare(struct clk *clk)
> >   606 {
> >   607         if (IS_ERR_OR_NULL(clk))
> >   608                 return;
> >   609
> >   610         clk_prepare_lock();                      // <-- Here
> >   611         clk_core_unprepare(clk->core);
> >   612         clk_prepare_unlock();
> >   613 }
> >   614 EXPORT_SYMBOL_GPL(clk_unprepare);
> >
> >    91 static void clk_prepare_lock(void)
> >    92 {
> >    93         if (!mutex_trylock(&prepare_lock)) {     // <-- Here
> >    94                 if (prepare_owner == current) {
> >    95                         prepare_refcnt++;
> >    96                         return;
> >    97                 }
> >    98                 mutex_lock(&prepare_lock);
> >    99         }
> >   100         WARN_ON_ONCE(prepare_owner != NULL);
> >   101         WARN_ON_ONCE(prepare_refcnt != 0);
> >   102         prepare_owner = current;
> >   103         prepare_refcnt = 1;
> >   104 }
> >
> > So, 'unprepare' can sleep and cannot be put into ISR.
> > I also try to put it into a timer, but the error is the same
> > since timer callback is executed by softirq.
> >
> > We need clk_disable_unprepare() since it can save power consumption
> > in idle.
> 
> We can call clk_prepare in probe and then use clk_enable/clk_disable, 
> which don't sleep.
> 
> Regards,
> Matthias

Hi Matthias,

Because clock gate and MUX are controlled by clk_enable/clk_disable,
and PLL is controlled by clk_prepare/clk_unprepare,
I still need to call clk_unprepare.

After I remove releasing buffer, releasing task, and timeout task from
work, the work can be detached from task.

Therefore, I can use the following flow to reduce the number of works.

if task_busy_list from empty to non-empty
	clk_prepare_enable
if task_busy_list from non-empty to empty
	in ISR, add work for clk_disable_unprepare

What do you think of this solution?

Thanks,
HS

> > Therefore, I plan to
> > (1) move releasing buffer and task into ISR,
> > (2) move timeout into timer, and
> > (3) keep workqueue for clk_disable_unprepare().
> >
> > What do you think?
> >
> > Thanks,
> > HS
> >
> >>>>>>
> >>>>>> Do you mean just call these two functions in ISR?
> >>>>>> My major concern is dma_free_coherent() and kfree() in
> >>>>>> cmdq_task_release(task).
> >>>>>
> >>>>> Why do we need the dma calls at all? Can't we just calculate the
> >>>>> physical address using __pa(x)?
> >>>>
> >>>> I prefer to use dma_map_single/dma_unmap_single.
> >>>>
> >>>
> >>> Can you please elaborate why you need this. We don't do dma, so we
> >>> should not use dma memory for this.
> >>
> >> We need a buffer to share between CPU and GCE, so we do need DMA.
> >> CPU is in charge of writing GCE commands into this buffer.
> >> GCE is in charge of reading and running GCE commands from this buffer.
> >> When we chain CMDQ tasks, we also need to modify GCE JUMP command.
> >> Therefore, I prefer to use dma_alloc_coherent and dma_free_coherent.
> >>
> >> However, if we want to use timer to handle timeout, we need to release
> >> memory in ISR.
> >> In this case, using kmalloc/kfree + dma_map_single/dma_unmap_single
> >> instead of dma_alloc_coherent/dma_free_coherent is an alternative
> >> solution, but taking care the synchronization between cache and memory
> >> is the expected overhead.
> >>
> >>>>>> Therefore, your suggestion is to use GFP_ATOMIC for both
> >>>>>> dma_alloc_coherent() and kzalloc(). Right?
> >>>>>
> >>>>> I don't think we need GFP_ATOMIC, the critical path will just free the
> >>>>> memory.
> >>>>
> >>>> I tested these two functions, and kfree was safe.
> >>>> However, dma_free_coherent raised BUG.
> >>>> BUG: failure at
> >>>> /mnt/host/source/src/third_party/kernel/v3.18/mm/vmalloc.c:1514/vunmap()!
> >>>
> >>> Just a general hint. Please try to evaluate on a recent kernel. It looks
> >>> like as if you tried this on a v3.18 based one.
> >>
> >> This driver should be backward compatible to v3.18 for a MTK project.
> >>
> >>> Best regards,
> >>> Matthias
> >>
> >> Thanks,
> >> HS
> >>
> >>>> 1512 void vunmap(const void *addr)
> >>>> 1513 {
> >>>> 1514         BUG_ON(in_interrupt());		// <-- here
> >>>> 1515         might_sleep();
> >>>> 1516         if (addr)
> >>>> 1517                 __vunmap(addr, 0);
> >>>> 1518 }
> >>>> 1519 EXPORT_SYMBOL(vunmap);
> >>>>
> >>>> Therefore, I plan to use kmalloc + dma_map_single instead of
> >>>> dma_alloc_coherent, and dma_unmap_single + kfree instead of
> >>>> dma_free_coherent.
> >>>>
> >>>> What do you think about the function replacement?
> >>>>
> >>>>>> If so, I can try to implement timeout by timer, and discuss with you
> >>>>>> if I have further questions.
> >>>>>>
> >>>>>
> >>>>> Sounds good :)
> >>>>>
> >>>>> Thanks,
> >>>>> Matthias
> >>>>
> >>>> Thanks,
> >>>> HS
> >>>>
> >>>>>>> Regrading the clock, wouldn't it be easier to handle the clock
> >>>>>>> enable/disable depending on the state of task_busy_list? I suppose we
> >>>>>>> can't as we would need to check the task_busy_list of all threads, right?
> >>>>>>>
> >>>>>>> Regards,
> >>>>>>> Matthias
> >>>>>>
> >>>>>> Thanks,
> >>>>>> HS

[e-mail] June 21, 2016, 7:46 a.m. UTC | #32

On Tue, 2016-06-21 at 10:03 +0800, CK Hu wrote:
> On Mon, 2016-06-20 at 19:22 +0800, Horng-Shyang Liao wrote:
> > On Mon, 2016-06-20 at 18:41 +0800, CK Hu wrote:
> > > Hi, HS:
> > > 
> > > One comment inline.
> > > 
> > > On Mon, 2016-05-30 at 11:19 +0800, HS Liao wrote:
> > > > This patch is first version of Mediatek Command Queue(CMDQ) driver. The
> > > > CMDQ is used to help read/write registers with critical time limitation,
> > > > such as updating display configuration during the vblank. It controls
> > > > Global Command Engine (GCE) hardware to achieve this requirement.
> > > > Currently, CMDQ only supports display related hardwares, but we expect
> > > > it can be extended to other hardwares for future requirements.
> > > > 
> > > > Signed-off-by: HS Liao <[e-mail]>
> > > > Signed-off-by: CK Hu <[e-mail]>
> > > > ---
> > > >  drivers/soc/mediatek/Kconfig    |  10 +
> > > >  drivers/soc/mediatek/Makefile   |   1 +
> > > >  drivers/soc/mediatek/mtk-cmdq.c | 943 ++++++++++++++++++++++++++++++++++++++++
> > > >  include/soc/mediatek/cmdq.h     | 197 +++++++++
> > > >  4 files changed, 1151 insertions(+)
> > > >  create mode 100644 drivers/soc/mediatek/mtk-cmdq.c
> > > >  create mode 100644 include/soc/mediatek/cmdq.h
> > > > 
> > > > diff --git a/drivers/soc/mediatek/Kconfig b/drivers/soc/mediatek/Kconfig
> > > > index 0a4ea80..c4ad75c 100644
> > > > --- a/drivers/soc/mediatek/Kconfig
> > > > +++ b/drivers/soc/mediatek/Kconfig
> > > > @@ -1,6 +1,16 @@
> > > >  #
> > > >  # MediaTek SoC drivers
> > > >  #
> > > > +config MTK_CMDQ
> > > > +	bool "MediaTek CMDQ Support"
> > > > +	depends on ARCH_MEDIATEK || COMPILE_TEST
> > > > +	select MTK_INFRACFG
> > > > +	help
> > > > +	  Say yes here to add support for the MediaTek Command Queue (CMDQ)
> > > > +	  driver. The CMDQ is used to help read/write registers with critical
> > > > +	  time limitation, such as updating display configuration during the
> > > > +	  vblank.
> > > > +
> > > >  config MTK_INFRACFG
> > > >  	bool "MediaTek INFRACFG Support"
> > > >  	depends on ARCH_MEDIATEK || COMPILE_TEST
> > > > diff --git a/drivers/soc/mediatek/Makefile b/drivers/soc/mediatek/Makefile
> > > > index 12998b0..f7397ef 100644
> > > > --- a/drivers/soc/mediatek/Makefile
> > > > +++ b/drivers/soc/mediatek/Makefile
> > > 
> > > [Snip...]
> > > 
> > > > +
> > > > +static int cmdq_eng_get_thread(u64 flag)
> > > > +{
> > > > +	if (flag & BIT_ULL(CMDQ_ENG_DISP_DSI0))
> > > > +		return CMDQ_THR_DISP_MAIN_IDX;
> > > > +	else if (flag & BIT_ULL(CMDQ_ENG_DISP_DPI0))
> > > > +		return CMDQ_THR_DISP_SUB_IDX;
> > > > +	else
> > > > +		return CMDQ_THR_DISP_MISC_IDX;
> > > > +}
> > > 
> > > I think cmdq should not have knowledge of client. These statement show
> > > that cmdq know that DSI0-tasks is different with DPI0-tasks. I propose
> > > the 'session' to replace engine flag. For example, main display create
> > > one cmdq_session and external display create another cmdq_session. For
> > > client driver, every tasks created by main display is bound to main
> > > display session. For cmdq driver, it should dynamically bind a session
> > > to a HW thread, and then dispatch tasks of this session to this HW
> > > thread. After HW thread run out of tasks of this session, detach this
> > > session with this thread.
> > > For the problem of remove wfe cmd, I think a session can have a property
> > > of merge_wfe_cmd. For display session, session->merge_wfe_cmd = true.
> > > For other client, it's false.
> > 
> > Hi CK,
> > 
> > I think your suggestion is similar to CMDQ 'scenarios',
> > which was removed from CMDQ v2.
> > 
> > Daniel suggests to use engine flags instead of scenarios.
> > Quote from https://patchwork.kernel.org/patch/8068311/ .
> > 'Instead of encoding policy (these arbitrary "scenarios"), perhaps the
> >  cmdq driver should just provide these flag bits, and the display
> >  subsystem can use them to build the "flag" parameter itself.
> > 
> >  After all, the exact configuration of mmsys components is somewhat
> >  flexible.'
> > 
> > Therefore, it would be better to discuss with Daniel before we change
> > it.
> > 
> > 
> > Hi Daniel,
> > 
> > Do you think we should use scenarios or sessions instead of flags?
> > 
> > Thanks,
> > HS
> > 
> 
> Hi, HS:
> 
> 'session' is not similar to 'scenarios'.
> 
> In 'scenarios' mechanism, client bind a task with scenarios and send to
> cmdq. Cmdq transfer scenarios to engine flag, and use engine flag to
> dispatch task to HW thread.
> 
> 
> In 'engine flag' mechanism, proposed by Daniel, client bind a task
> directly with engine flag. Cmdq directly use engine flag to dispatch
> task to HW thread without any translation.
> 
> But neither 'scenarios' mechanism nor 'engine flag' mechanism get rid of
> engine flag, which make cmdq have knowledge of client.
> 
> In 'session' mechanism, there is no engine flag any more. Client bind
> time-sequential tasks to the same session and tasks in different session
> can execute parallelly. One thing cmdq need to know is to dispatch tasks
> with the same session to the same HW thread, so cmdq does not have any
> knowledge of client.
> 
> Daniel focus on reduce translating scenarios to engine flag. I think
> 'session' mechanism does not conflict with his opinion because we does
> not translate 'session' to engine flag. Therefore, I think 'session' is
> the best of these three mechanism.
> 
> Regards,
> CK

Hi CK,

'Session' looks like a group of options for CMDQ.
CMDQ driver can just follow the options to run its flow,
and doesn't need to know its client(s).

We don't have many options now, but it has good flexibility to extend
for future requirements.

I will add it in next version.

Thanks,
HS

> > > Here is the sample code to create cmdq_rec with session.
> > > 
> > > merge_wfe_cmd = true;
> > > cmdq_session_create(merge_wfe_cmd, &primary_display_session);
> > > cmdq_rec_create(dev, primary_display_session, &rec);
> > > 
> > > 
> > > Therefore, the below definition can be removed.
> > > 
> > > > +
> > > > +enum cmdq_eng {
> > > > +	CMDQ_ENG_DISP_AAL,
> > > > +	CMDQ_ENG_DISP_COLOR0,
> > > > +	CMDQ_ENG_DISP_COLOR1,
> > > > +	CMDQ_ENG_DISP_DPI0,
> > > > +	CMDQ_ENG_DISP_DSI0,
> > > > +	CMDQ_ENG_DISP_DSI1,
> > > > +	CMDQ_ENG_DISP_GAMMA,
> > > > +	CMDQ_ENG_DISP_OD,
> > > > +	CMDQ_ENG_DISP_OVL0,
> > > > +	CMDQ_ENG_DISP_OVL1,
> > > > +	CMDQ_ENG_DISP_PWM0,
> > > > +	CMDQ_ENG_DISP_PWM1,
> > > > +	CMDQ_ENG_DISP_RDMA0,
> > > > +	CMDQ_ENG_DISP_RDMA1,
> > > > +	CMDQ_ENG_DISP_RDMA2,
> > > > +	CMDQ_ENG_DISP_UFOE,
> > > > +	CMDQ_ENG_DISP_WDMA0,
> > > > +	CMDQ_ENG_DISP_WDMA1,
> > > > +	CMDQ_ENG_MAX,
> > > > +};
> > > > +
> > > 
> > > 
> > > Regards,
> > > CK
> > > 
> > 
> > 
> 
>

Matthias Brugger June 21, 2016, 1:41 p.m. UTC | #33

On 21/06/16 07:52, Horng-Shyang Liao wrote:
> On Fri, 2016-06-17 at 17:57 +0200, Matthias Brugger wrote:
>>
>> On 17/06/16 10:28, Horng-Shyang Liao wrote:
>>> Hi Matthias,
>>>
>>> On Tue, 2016-06-14 at 20:07 +0800, Horng-Shyang Liao wrote:
>>>> Hi Matthias,
>>>>
>>>> On Tue, 2016-06-14 at 12:17 +0200, Matthias Brugger wrote:
>>>>>
>>>>> On 14/06/16 09:44, Horng-Shyang Liao wrote:
>>>>>> Hi Matthias,
>>>>>>
>>>>>> On Wed, 2016-06-08 at 17:35 +0200, Matthias Brugger wrote:
>>>>>>>
>>>>>>> On 08/06/16 14:25, Horng-Shyang Liao wrote:
>>>>>>>> Hi Matthias,
>>>>>>>>
>>>>>>>> On Wed, 2016-06-08 at 12:45 +0200, Matthias Brugger wrote:
>>>>>>>>>
>>>>>>>>> On 08/06/16 07:40, Horng-Shyang Liao wrote:
>>>>>>>>>> Hi Matthias,
>>>>>>>>>>
>>>>>>>>>> On Tue, 2016-06-07 at 18:59 +0200, Matthias Brugger wrote:
>>>>>>>>>>>
>>>>>>>>>>> On 03/06/16 15:11, Matthias Brugger wrote:
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>> [...]
>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>>>>> +            smp_mb(); /* modify jump before enable thread */
>>>>>>>>>>>>>>>>>>>>> +        }
>>>>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>>>>> +        cmdq_thread_writel(thread, task->pa_base +
>>>>>>>>>>>>>>>>>>>>> task->command_size,
>>>>>>>>>>>>>>>>>>>>> +                   CMDQ_THR_END_ADDR);
>>>>>>>>>>>>>>>>>>>>> +        cmdq_thread_resume(thread);
>>>>>>>>>>>>>>>>>>>>> +    }
>>>>>>>>>>>>>>>>>>>>> +    list_move_tail(&task->list_entry, &thread->task_busy_list);
>>>>>>>>>>>>>>>>>>>>> +    spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>>>>>>>>>>>>>>>>>> +}
>>>>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
>>>>>>>>>>>>>>>>>>>>> +                   struct cmdq_thread *thread, u32 irq_flag)
>>>>>>>>>>>>>>>>>>>>> +{
>>>>>>>>>>>>>>>>>>>>> +    struct cmdq_task *task, *tmp, *curr_task = NULL;
>>>>>>>>>>>>>>>>>>>>> +    u32 curr_pa;
>>>>>>>>>>>>>>>>>>>>> +    struct cmdq_cb_data cmdq_cb_data;
>>>>>>>>>>>>>>>>>>>>> +    bool err;
>>>>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>>>>> +    if (irq_flag & CMDQ_THR_IRQ_ERROR)
>>>>>>>>>>>>>>>>>>>>> +        err = true;
>>>>>>>>>>>>>>>>>>>>> +    else if (irq_flag & CMDQ_THR_IRQ_DONE)
>>>>>>>>>>>>>>>>>>>>> +        err = false;
>>>>>>>>>>>>>>>>>>>>> +    else
>>>>>>>>>>>>>>>>>>>>> +        return;
>>>>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>>>>> +    curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
>>>>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>>>>> +    list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
>>>>>>>>>>>>>>>>>>>>> +                 list_entry) {
>>>>>>>>>>>>>>>>>>>>> +        if (curr_pa >= task->pa_base &&
>>>>>>>>>>>>>>>>>>>>> +            curr_pa < (task->pa_base + task->command_size))
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>> What are you checking here? It seems as if you make some implcit
>>>>>>>>>>>>>>>>>>>> assumptions about pa_base and the order of execution of
>>>>>>>>>>>>>>>>>>>> commands in the
>>>>>>>>>>>>>>>>>>>> thread. Is it save to do so? Does dma_alloc_coherent give any
>>>>>>>>>>>>>>>>>>>> guarantees
>>>>>>>>>>>>>>>>>>>> about dma_handle?
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>> 1. Check what is the current running task in this GCE thread.
>>>>>>>>>>>>>>>>>>> 2. Yes.
>>>>>>>>>>>>>>>>>>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>> Yes, physical addresses might be continous, but AFAIK there is no
>>>>>>>>>>>>>>>>>> guarantee that the dma_handle address is steadily growing, when
>>>>>>>>>>>>>>>>>> calling
>>>>>>>>>>>>>>>>>> dma_alloc_coherent. And if I understand the code correctly, you
>>>>>>>>>>>>>>>>>> use this
>>>>>>>>>>>>>>>>>> assumption to decide if the task picked from task_busy_list is
>>>>>>>>>>>>>>>>>> currently
>>>>>>>>>>>>>>>>>> executing. So I think this mecanism is not working.
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>> I don't use dma_handle address, and just use physical addresses.
>>>>>>>>>>>>>>>>>         From CPU's point of view, tasks are linked by the busy list.
>>>>>>>>>>>>>>>>>         From GCE's point of view, tasks are linked by the JUMP command.
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>> In which cases does the HW thread raise an interrupt.
>>>>>>>>>>>>>>>>>> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>> GCE will raise interrupt if any task is done or error.
>>>>>>>>>>>>>>>>> However, GCE is fast, so CPU may get multiple done tasks
>>>>>>>>>>>>>>>>> when it is running ISR.
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>> In case of error, that GCE thread will pause and raise interrupt.
>>>>>>>>>>>>>>>>> So, CPU may get multiple done tasks and one error task.
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> I think we should reimplement the ISR mechanism. Can't we just read
>>>>>>>>>>>>>>>> CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave
>>>>>>>>>>>>>>>> cmdq_handle_error_done to the thread_fn? You will need to pass
>>>>>>>>>>>>>>>> information from the handler to thread_fn, but that shouldn't be an
>>>>>>>>>>>>>>>> issue. AFAIK interrupts are disabled in the handler, so we should stay
>>>>>>>>>>>>>>>> there as short as possible. Traversing task_busy_list is expensive, so
>>>>>>>>>>>>>>>> we need to do it in a thread context.
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> Actually, our initial implementation is similar to your suggestion,
>>>>>>>>>>>>>>> but display needs CMDQ to return callback function very precisely,
>>>>>>>>>>>>>>> else display will drop frame.
>>>>>>>>>>>>>>> For display, CMDQ interrupt will be raised every 16 ~ 17 ms,
>>>>>>>>>>>>>>> and CMDQ needs to call callback function in ISR.
>>>>>>>>>>>>>>> If we defer callback to workqueue, the time interval may be larger than
>>>>>>>>>>>>>>> 32 ms.sometimes.
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> I think the problem is, that you implemented the workqueue as a ordered
>>>>>>>>>>>>>> workqueue, so there is no parallel processing. I'm still not sure why
>>>>>>>>>>>>>> you need the workqueue to be ordered. Can you please explain.
>>>>>>>>>>>>>
>>>>>>>>>>>>> The order should be kept.
>>>>>>>>>>>>> Let me use mouse cursor as an example.
>>>>>>>>>>>>> If task 1 means move mouse cursor to point A, task 2 means point B,
>>>>>>>>>>>>> and task 3 means point C, our expected result is A -> B -> C.
>>>>>>>>>>>>> If the order is not kept, the result could become A -> C -> B.
>>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> Got it, thanks for the clarification.
>>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> I think a way to get rid of the workqueue is to use a timer, which gets
>>>>>>>>>>> programmed to the time a timeout in the first task in the busy list
>>>>>>>>>>> would happen. Everytime we update the busy list (e.g. because of task
>>>>>>>>>>> got finished by the thread), we update the timer. When the timer
>>>>>>>>>>> triggers, which hopefully won't happen too often, we return timeout on
>>>>>>>>>>> the busy list elements, until the time is lower then the actual time.
>>>>>>>>>>>
>>>>>>>>>>> At least with this we can reduce the data structures in this driver and
>>>>>>>>>>> make it more lightweight.
>>>>>>>>>>
>>>>>>>>>>      From my understanding, your proposed method can handle timeout case.
>>>>>>>>>>
>>>>>>>>>> However, the workqueue is also in charge of releasing tasks.
>>>>>>>>>> Do you take releasing tasks into consideration by using the proposed
>>>>>>>>>> timer method?
>>>>>>>>>> Furthermore, I think the code will become more complex if we also use
>>>>>>>>>> timer to implement releasing tasks.
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Can't we call
>>>>>>>>>              clk_disable_unprepare(cmdq->clock);
>>>>>>>>>              cmdq_task_release(task);
>>>>>>>>> after invoking the callback?
>>>
>>> After I put clk_disable_unprepare(cmdq->clock) into ISR, I encounter
>>> another BUG.
>>>
>>> (Quote some Linux 4.7 source code.)
>>>
>>>    605 void clk_unprepare(struct clk *clk)
>>>    606 {
>>>    607         if (IS_ERR_OR_NULL(clk))
>>>    608                 return;
>>>    609
>>>    610         clk_prepare_lock();                      // <-- Here
>>>    611         clk_core_unprepare(clk->core);
>>>    612         clk_prepare_unlock();
>>>    613 }
>>>    614 EXPORT_SYMBOL_GPL(clk_unprepare);
>>>
>>>     91 static void clk_prepare_lock(void)
>>>     92 {
>>>     93         if (!mutex_trylock(&prepare_lock)) {     // <-- Here
>>>     94                 if (prepare_owner == current) {
>>>     95                         prepare_refcnt++;
>>>     96                         return;
>>>     97                 }
>>>     98                 mutex_lock(&prepare_lock);
>>>     99         }
>>>    100         WARN_ON_ONCE(prepare_owner != NULL);
>>>    101         WARN_ON_ONCE(prepare_refcnt != 0);
>>>    102         prepare_owner = current;
>>>    103         prepare_refcnt = 1;
>>>    104 }
>>>
>>> So, 'unprepare' can sleep and cannot be put into ISR.
>>> I also try to put it into a timer, but the error is the same
>>> since timer callback is executed by softirq.
>>>
>>> We need clk_disable_unprepare() since it can save power consumption
>>> in idle.
>>
>> We can call clk_prepare in probe and then use clk_enable/clk_disable,
>> which don't sleep.
>>
>> Regards,
>> Matthias
>
> Hi Matthias,
>
> Because clock gate and MUX are controlled by clk_enable/clk_disable,
> and PLL is controlled by clk_prepare/clk_unprepare,
> I still need to call clk_unprepare.
>
> After I remove releasing buffer, releasing task, and timeout task from
> work, the work can be detached from task.
>
> Therefore, I can use the following flow to reduce the number of works.
>
> if task_busy_list from empty to non-empty
> 	clk_prepare_enable
> if task_busy_list from non-empty to empty
> 	in ISR, add work for clk_disable_unprepare
>
> What do you think of this solution?

Can't we just call clk_prepare in probe and clk_unprepare in remove? I 
think this could be a good starting point, and if we see, that we need 
to save more energy in the future, we can think of some other mechanism. 
What do you think?

Regards,
Matthias

>
> Thanks,
> HS
>
>>> Therefore, I plan to
>>> (1) move releasing buffer and task into ISR,
>>> (2) move timeout into timer, and
>>> (3) keep workqueue for clk_disable_unprepare().
>>>
>>> What do you think?
>>>
>>> Thanks,
>>> HS
>>>
>>>>>>>>
>>>>>>>> Do you mean just call these two functions in ISR?
>>>>>>>> My major concern is dma_free_coherent() and kfree() in
>>>>>>>> cmdq_task_release(task).
>>>>>>>
>>>>>>> Why do we need the dma calls at all? Can't we just calculate the
>>>>>>> physical address using __pa(x)?
>>>>>>
>>>>>> I prefer to use dma_map_single/dma_unmap_single.
>>>>>>
>>>>>
>>>>> Can you please elaborate why you need this. We don't do dma, so we
>>>>> should not use dma memory for this.
>>>>
>>>> We need a buffer to share between CPU and GCE, so we do need DMA.
>>>> CPU is in charge of writing GCE commands into this buffer.
>>>> GCE is in charge of reading and running GCE commands from this buffer.
>>>> When we chain CMDQ tasks, we also need to modify GCE JUMP command.
>>>> Therefore, I prefer to use dma_alloc_coherent and dma_free_coherent.
>>>>
>>>> However, if we want to use timer to handle timeout, we need to release
>>>> memory in ISR.
>>>> In this case, using kmalloc/kfree + dma_map_single/dma_unmap_single
>>>> instead of dma_alloc_coherent/dma_free_coherent is an alternative
>>>> solution, but taking care the synchronization between cache and memory
>>>> is the expected overhead.
>>>>
>>>>>>>> Therefore, your suggestion is to use GFP_ATOMIC for both
>>>>>>>> dma_alloc_coherent() and kzalloc(). Right?
>>>>>>>
>>>>>>> I don't think we need GFP_ATOMIC, the critical path will just free the
>>>>>>> memory.
>>>>>>
>>>>>> I tested these two functions, and kfree was safe.
>>>>>> However, dma_free_coherent raised BUG.
>>>>>> BUG: failure at
>>>>>> /mnt/host/source/src/third_party/kernel/v3.18/mm/vmalloc.c:1514/vunmap()!
>>>>>
>>>>> Just a general hint. Please try to evaluate on a recent kernel. It looks
>>>>> like as if you tried this on a v3.18 based one.
>>>>
>>>> This driver should be backward compatible to v3.18 for a MTK project.
>>>>
>>>>> Best regards,
>>>>> Matthias
>>>>
>>>> Thanks,
>>>> HS
>>>>
>>>>>> 1512 void vunmap(const void *addr)
>>>>>> 1513 {
>>>>>> 1514         BUG_ON(in_interrupt());		// <-- here
>>>>>> 1515         might_sleep();
>>>>>> 1516         if (addr)
>>>>>> 1517                 __vunmap(addr, 0);
>>>>>> 1518 }
>>>>>> 1519 EXPORT_SYMBOL(vunmap);
>>>>>>
>>>>>> Therefore, I plan to use kmalloc + dma_map_single instead of
>>>>>> dma_alloc_coherent, and dma_unmap_single + kfree instead of
>>>>>> dma_free_coherent.
>>>>>>
>>>>>> What do you think about the function replacement?
>>>>>>
>>>>>>>> If so, I can try to implement timeout by timer, and discuss with you
>>>>>>>> if I have further questions.
>>>>>>>>
>>>>>>>
>>>>>>> Sounds good :)
>>>>>>>
>>>>>>> Thanks,
>>>>>>> Matthias
>>>>>>
>>>>>> Thanks,
>>>>>> HS
>>>>>>
>>>>>>>>> Regrading the clock, wouldn't it be easier to handle the clock
>>>>>>>>> enable/disable depending on the state of task_busy_list? I suppose we
>>>>>>>>> can't as we would need to check the task_busy_list of all threads, right?
>>>>>>>>>
>>>>>>>>> Regards,
>>>>>>>>> Matthias
>>>>>>>>
>>>>>>>> Thanks,
>>>>>>>> HS
>
>

[e-mail] June 22, 2016, 5:43 a.m. UTC | #34

On Tue, 2016-06-21 at 15:41 +0200, Matthias Brugger wrote:
> 
> On 21/06/16 07:52, Horng-Shyang Liao wrote:
> > On Fri, 2016-06-17 at 17:57 +0200, Matthias Brugger wrote:
> >>
> >> On 17/06/16 10:28, Horng-Shyang Liao wrote:
> >>> Hi Matthias,
> >>>
> >>> On Tue, 2016-06-14 at 20:07 +0800, Horng-Shyang Liao wrote:
> >>>> Hi Matthias,
> >>>>
> >>>> On Tue, 2016-06-14 at 12:17 +0200, Matthias Brugger wrote:
> >>>>>
> >>>>> On 14/06/16 09:44, Horng-Shyang Liao wrote:
> >>>>>> Hi Matthias,
> >>>>>>
> >>>>>> On Wed, 2016-06-08 at 17:35 +0200, Matthias Brugger wrote:
> >>>>>>>
> >>>>>>> On 08/06/16 14:25, Horng-Shyang Liao wrote:
> >>>>>>>> Hi Matthias,
> >>>>>>>>
> >>>>>>>> On Wed, 2016-06-08 at 12:45 +0200, Matthias Brugger wrote:
> >>>>>>>>>
> >>>>>>>>> On 08/06/16 07:40, Horng-Shyang Liao wrote:
> >>>>>>>>>> Hi Matthias,
> >>>>>>>>>>
> >>>>>>>>>> On Tue, 2016-06-07 at 18:59 +0200, Matthias Brugger wrote:
> >>>>>>>>>>>
> >>>>>>>>>>> On 03/06/16 15:11, Matthias Brugger wrote:
> >>>>>>>>>>>>
> >>>>>>>>>>>>
> >>>>>>>>>>> [...]
> >>>>>>>>>>>
> >>>>>>>>>>>>>>>>>>>>> +
> >>>>>>>>>>>>>>>>>>>>> +            smp_mb(); /* modify jump before enable thread */
> >>>>>>>>>>>>>>>>>>>>> +        }
> >>>>>>>>>>>>>>>>>>>>> +
> >>>>>>>>>>>>>>>>>>>>> +        cmdq_thread_writel(thread, task->pa_base +
> >>>>>>>>>>>>>>>>>>>>> task->command_size,
> >>>>>>>>>>>>>>>>>>>>> +                   CMDQ_THR_END_ADDR);
> >>>>>>>>>>>>>>>>>>>>> +        cmdq_thread_resume(thread);
> >>>>>>>>>>>>>>>>>>>>> +    }
> >>>>>>>>>>>>>>>>>>>>> +    list_move_tail(&task->list_entry, &thread->task_busy_list);
> >>>>>>>>>>>>>>>>>>>>> +    spin_unlock_irqrestore(&cmdq->exec_lock, flags);
> >>>>>>>>>>>>>>>>>>>>> +}
> >>>>>>>>>>>>>>>>>>>>> +
> >>>>>>>>>>>>>>>>>>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
> >>>>>>>>>>>>>>>>>>>>> +                   struct cmdq_thread *thread, u32 irq_flag)
> >>>>>>>>>>>>>>>>>>>>> +{
> >>>>>>>>>>>>>>>>>>>>> +    struct cmdq_task *task, *tmp, *curr_task = NULL;
> >>>>>>>>>>>>>>>>>>>>> +    u32 curr_pa;
> >>>>>>>>>>>>>>>>>>>>> +    struct cmdq_cb_data cmdq_cb_data;
> >>>>>>>>>>>>>>>>>>>>> +    bool err;
> >>>>>>>>>>>>>>>>>>>>> +
> >>>>>>>>>>>>>>>>>>>>> +    if (irq_flag & CMDQ_THR_IRQ_ERROR)
> >>>>>>>>>>>>>>>>>>>>> +        err = true;
> >>>>>>>>>>>>>>>>>>>>> +    else if (irq_flag & CMDQ_THR_IRQ_DONE)
> >>>>>>>>>>>>>>>>>>>>> +        err = false;
> >>>>>>>>>>>>>>>>>>>>> +    else
> >>>>>>>>>>>>>>>>>>>>> +        return;
> >>>>>>>>>>>>>>>>>>>>> +
> >>>>>>>>>>>>>>>>>>>>> +    curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
> >>>>>>>>>>>>>>>>>>>>> +
> >>>>>>>>>>>>>>>>>>>>> +    list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
> >>>>>>>>>>>>>>>>>>>>> +                 list_entry) {
> >>>>>>>>>>>>>>>>>>>>> +        if (curr_pa >= task->pa_base &&
> >>>>>>>>>>>>>>>>>>>>> +            curr_pa < (task->pa_base + task->command_size))
> >>>>>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>>>>> What are you checking here? It seems as if you make some implcit
> >>>>>>>>>>>>>>>>>>>> assumptions about pa_base and the order of execution of
> >>>>>>>>>>>>>>>>>>>> commands in the
> >>>>>>>>>>>>>>>>>>>> thread. Is it save to do so? Does dma_alloc_coherent give any
> >>>>>>>>>>>>>>>>>>>> guarantees
> >>>>>>>>>>>>>>>>>>>> about dma_handle?
> >>>>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>>>> 1. Check what is the current running task in this GCE thread.
> >>>>>>>>>>>>>>>>>>> 2. Yes.
> >>>>>>>>>>>>>>>>>>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
> >>>>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>>> Yes, physical addresses might be continous, but AFAIK there is no
> >>>>>>>>>>>>>>>>>> guarantee that the dma_handle address is steadily growing, when
> >>>>>>>>>>>>>>>>>> calling
> >>>>>>>>>>>>>>>>>> dma_alloc_coherent. And if I understand the code correctly, you
> >>>>>>>>>>>>>>>>>> use this
> >>>>>>>>>>>>>>>>>> assumption to decide if the task picked from task_busy_list is
> >>>>>>>>>>>>>>>>>> currently
> >>>>>>>>>>>>>>>>>> executing. So I think this mecanism is not working.
> >>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>> I don't use dma_handle address, and just use physical addresses.
> >>>>>>>>>>>>>>>>>         From CPU's point of view, tasks are linked by the busy list.
> >>>>>>>>>>>>>>>>>         From GCE's point of view, tasks are linked by the JUMP command.
> >>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>>> In which cases does the HW thread raise an interrupt.
> >>>>>>>>>>>>>>>>>> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
> >>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>> GCE will raise interrupt if any task is done or error.
> >>>>>>>>>>>>>>>>> However, GCE is fast, so CPU may get multiple done tasks
> >>>>>>>>>>>>>>>>> when it is running ISR.
> >>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>> In case of error, that GCE thread will pause and raise interrupt.
> >>>>>>>>>>>>>>>>> So, CPU may get multiple done tasks and one error task.
> >>>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>>> I think we should reimplement the ISR mechanism. Can't we just read
> >>>>>>>>>>>>>>>> CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave
> >>>>>>>>>>>>>>>> cmdq_handle_error_done to the thread_fn? You will need to pass
> >>>>>>>>>>>>>>>> information from the handler to thread_fn, but that shouldn't be an
> >>>>>>>>>>>>>>>> issue. AFAIK interrupts are disabled in the handler, so we should stay
> >>>>>>>>>>>>>>>> there as short as possible. Traversing task_busy_list is expensive, so
> >>>>>>>>>>>>>>>> we need to do it in a thread context.
> >>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>> Actually, our initial implementation is similar to your suggestion,
> >>>>>>>>>>>>>>> but display needs CMDQ to return callback function very precisely,
> >>>>>>>>>>>>>>> else display will drop frame.
> >>>>>>>>>>>>>>> For display, CMDQ interrupt will be raised every 16 ~ 17 ms,
> >>>>>>>>>>>>>>> and CMDQ needs to call callback function in ISR.
> >>>>>>>>>>>>>>> If we defer callback to workqueue, the time interval may be larger than
> >>>>>>>>>>>>>>> 32 ms.sometimes.
> >>>>>>>>>>>>>>>
> >>>>>>>>>>>>>>
> >>>>>>>>>>>>>> I think the problem is, that you implemented the workqueue as a ordered
> >>>>>>>>>>>>>> workqueue, so there is no parallel processing. I'm still not sure why
> >>>>>>>>>>>>>> you need the workqueue to be ordered. Can you please explain.
> >>>>>>>>>>>>>
> >>>>>>>>>>>>> The order should be kept.
> >>>>>>>>>>>>> Let me use mouse cursor as an example.
> >>>>>>>>>>>>> If task 1 means move mouse cursor to point A, task 2 means point B,
> >>>>>>>>>>>>> and task 3 means point C, our expected result is A -> B -> C.
> >>>>>>>>>>>>> If the order is not kept, the result could become A -> C -> B.
> >>>>>>>>>>>>>
> >>>>>>>>>>>>
> >>>>>>>>>>>> Got it, thanks for the clarification.
> >>>>>>>>>>>>
> >>>>>>>>>>>
> >>>>>>>>>>> I think a way to get rid of the workqueue is to use a timer, which gets
> >>>>>>>>>>> programmed to the time a timeout in the first task in the busy list
> >>>>>>>>>>> would happen. Everytime we update the busy list (e.g. because of task
> >>>>>>>>>>> got finished by the thread), we update the timer. When the timer
> >>>>>>>>>>> triggers, which hopefully won't happen too often, we return timeout on
> >>>>>>>>>>> the busy list elements, until the time is lower then the actual time.
> >>>>>>>>>>>
> >>>>>>>>>>> At least with this we can reduce the data structures in this driver and
> >>>>>>>>>>> make it more lightweight.
> >>>>>>>>>>
> >>>>>>>>>>      From my understanding, your proposed method can handle timeout case.
> >>>>>>>>>>
> >>>>>>>>>> However, the workqueue is also in charge of releasing tasks.
> >>>>>>>>>> Do you take releasing tasks into consideration by using the proposed
> >>>>>>>>>> timer method?
> >>>>>>>>>> Furthermore, I think the code will become more complex if we also use
> >>>>>>>>>> timer to implement releasing tasks.
> >>>>>>>>>>
> >>>>>>>>>
> >>>>>>>>> Can't we call
> >>>>>>>>>              clk_disable_unprepare(cmdq->clock);
> >>>>>>>>>              cmdq_task_release(task);
> >>>>>>>>> after invoking the callback?
> >>>
> >>> After I put clk_disable_unprepare(cmdq->clock) into ISR, I encounter
> >>> another BUG.
> >>>
> >>> (Quote some Linux 4.7 source code.)
> >>>
> >>>    605 void clk_unprepare(struct clk *clk)
> >>>    606 {
> >>>    607         if (IS_ERR_OR_NULL(clk))
> >>>    608                 return;
> >>>    609
> >>>    610         clk_prepare_lock();                      // <-- Here
> >>>    611         clk_core_unprepare(clk->core);
> >>>    612         clk_prepare_unlock();
> >>>    613 }
> >>>    614 EXPORT_SYMBOL_GPL(clk_unprepare);
> >>>
> >>>     91 static void clk_prepare_lock(void)
> >>>     92 {
> >>>     93         if (!mutex_trylock(&prepare_lock)) {     // <-- Here
> >>>     94                 if (prepare_owner == current) {
> >>>     95                         prepare_refcnt++;
> >>>     96                         return;
> >>>     97                 }
> >>>     98                 mutex_lock(&prepare_lock);
> >>>     99         }
> >>>    100         WARN_ON_ONCE(prepare_owner != NULL);
> >>>    101         WARN_ON_ONCE(prepare_refcnt != 0);
> >>>    102         prepare_owner = current;
> >>>    103         prepare_refcnt = 1;
> >>>    104 }
> >>>
> >>> So, 'unprepare' can sleep and cannot be put into ISR.
> >>> I also try to put it into a timer, but the error is the same
> >>> since timer callback is executed by softirq.
> >>>
> >>> We need clk_disable_unprepare() since it can save power consumption
> >>> in idle.
> >>
> >> We can call clk_prepare in probe and then use clk_enable/clk_disable,
> >> which don't sleep.
> >>
> >> Regards,
> >> Matthias
> >
> > Hi Matthias,
> >
> > Because clock gate and MUX are controlled by clk_enable/clk_disable,
> > and PLL is controlled by clk_prepare/clk_unprepare,
> > I still need to call clk_unprepare.
> >
> > After I remove releasing buffer, releasing task, and timeout task from
> > work, the work can be detached from task.
> >
> > Therefore, I can use the following flow to reduce the number of works.
> >
> > if task_busy_list from empty to non-empty
> > 	clk_prepare_enable
> > if task_busy_list from non-empty to empty
> > 	in ISR, add work for clk_disable_unprepare
> >
> > What do you think of this solution?
> 
> Can't we just call clk_prepare in probe and clk_unprepare in remove? I 
> think this could be a good starting point, and if we see, that we need 
> to save more energy in the future, we can think of some other mechanism. 
> What do you think?
> 
> Regards,
> Matthias

Hi Matthias,

As far as I know, we should call clk_unprepare to save more energy.

May I call clk_prepare in probe/resume and clk_unprepare in
remove/suspend in this patch, and then prepare another patch to call
clk_unprepare in idle to save more energy?

Thanks,
HS

> >
> > Thanks,
> > HS
> >
> >>> Therefore, I plan to
> >>> (1) move releasing buffer and task into ISR,
> >>> (2) move timeout into timer, and
> >>> (3) keep workqueue for clk_disable_unprepare().
> >>>
> >>> What do you think?
> >>>
> >>> Thanks,
> >>> HS
> >>>
> >>>>>>>>
> >>>>>>>> Do you mean just call these two functions in ISR?
> >>>>>>>> My major concern is dma_free_coherent() and kfree() in
> >>>>>>>> cmdq_task_release(task).
> >>>>>>>
> >>>>>>> Why do we need the dma calls at all? Can't we just calculate the
> >>>>>>> physical address using __pa(x)?
> >>>>>>
> >>>>>> I prefer to use dma_map_single/dma_unmap_single.
> >>>>>>
> >>>>>
> >>>>> Can you please elaborate why you need this. We don't do dma, so we
> >>>>> should not use dma memory for this.
> >>>>
> >>>> We need a buffer to share between CPU and GCE, so we do need DMA.
> >>>> CPU is in charge of writing GCE commands into this buffer.
> >>>> GCE is in charge of reading and running GCE commands from this buffer.
> >>>> When we chain CMDQ tasks, we also need to modify GCE JUMP command.
> >>>> Therefore, I prefer to use dma_alloc_coherent and dma_free_coherent.
> >>>>
> >>>> However, if we want to use timer to handle timeout, we need to release
> >>>> memory in ISR.
> >>>> In this case, using kmalloc/kfree + dma_map_single/dma_unmap_single
> >>>> instead of dma_alloc_coherent/dma_free_coherent is an alternative
> >>>> solution, but taking care the synchronization between cache and memory
> >>>> is the expected overhead.
> >>>>
> >>>>>>>> Therefore, your suggestion is to use GFP_ATOMIC for both
> >>>>>>>> dma_alloc_coherent() and kzalloc(). Right?
> >>>>>>>
> >>>>>>> I don't think we need GFP_ATOMIC, the critical path will just free the
> >>>>>>> memory.
> >>>>>>
> >>>>>> I tested these two functions, and kfree was safe.
> >>>>>> However, dma_free_coherent raised BUG.
> >>>>>> BUG: failure at
> >>>>>> /mnt/host/source/src/third_party/kernel/v3.18/mm/vmalloc.c:1514/vunmap()!
> >>>>>
> >>>>> Just a general hint. Please try to evaluate on a recent kernel. It looks
> >>>>> like as if you tried this on a v3.18 based one.
> >>>>
> >>>> This driver should be backward compatible to v3.18 for a MTK project.
> >>>>
> >>>>> Best regards,
> >>>>> Matthias
> >>>>
> >>>> Thanks,
> >>>> HS
> >>>>
> >>>>>> 1512 void vunmap(const void *addr)
> >>>>>> 1513 {
> >>>>>> 1514         BUG_ON(in_interrupt());		// <-- here
> >>>>>> 1515         might_sleep();
> >>>>>> 1516         if (addr)
> >>>>>> 1517                 __vunmap(addr, 0);
> >>>>>> 1518 }
> >>>>>> 1519 EXPORT_SYMBOL(vunmap);
> >>>>>>
> >>>>>> Therefore, I plan to use kmalloc + dma_map_single instead of
> >>>>>> dma_alloc_coherent, and dma_unmap_single + kfree instead of
> >>>>>> dma_free_coherent.
> >>>>>>
> >>>>>> What do you think about the function replacement?
> >>>>>>
> >>>>>>>> If so, I can try to implement timeout by timer, and discuss with you
> >>>>>>>> if I have further questions.
> >>>>>>>>
> >>>>>>>
> >>>>>>> Sounds good :)
> >>>>>>>
> >>>>>>> Thanks,
> >>>>>>> Matthias
> >>>>>>
> >>>>>> Thanks,
> >>>>>> HS
> >>>>>>
> >>>>>>>>> Regrading the clock, wouldn't it be easier to handle the clock
> >>>>>>>>> enable/disable depending on the state of task_busy_list? I suppose we
> >>>>>>>>> can't as we would need to check the task_busy_list of all threads, right?
> >>>>>>>>>
> >>>>>>>>> Regards,
> >>>>>>>>> Matthias
> >>>>>>>>
> >>>>>>>> Thanks,
> >>>>>>>> HS
> >
> >

Matthias Brugger June 22, 2016, 9:58 a.m. UTC | #35

On 06/22/2016 07:43 AM, Horng-Shyang Liao wrote:
> On Tue, 2016-06-21 at 15:41 +0200, Matthias Brugger wrote:
>>
>> On 21/06/16 07:52, Horng-Shyang Liao wrote:
>>> On Fri, 2016-06-17 at 17:57 +0200, Matthias Brugger wrote:
>>>>
>>>> On 17/06/16 10:28, Horng-Shyang Liao wrote:
>>>>> Hi Matthias,
>>>>>
>>>>> On Tue, 2016-06-14 at 20:07 +0800, Horng-Shyang Liao wrote:
>>>>>> Hi Matthias,
>>>>>>
>>>>>> On Tue, 2016-06-14 at 12:17 +0200, Matthias Brugger wrote:
>>>>>>>
>>>>>>> On 14/06/16 09:44, Horng-Shyang Liao wrote:
>>>>>>>> Hi Matthias,
>>>>>>>>
>>>>>>>> On Wed, 2016-06-08 at 17:35 +0200, Matthias Brugger wrote:
>>>>>>>>>
>>>>>>>>> On 08/06/16 14:25, Horng-Shyang Liao wrote:
>>>>>>>>>> Hi Matthias,
>>>>>>>>>>
>>>>>>>>>> On Wed, 2016-06-08 at 12:45 +0200, Matthias Brugger wrote:
>>>>>>>>>>>
>>>>>>>>>>> On 08/06/16 07:40, Horng-Shyang Liao wrote:
>>>>>>>>>>>> Hi Matthias,
>>>>>>>>>>>>
>>>>>>>>>>>> On Tue, 2016-06-07 at 18:59 +0200, Matthias Brugger wrote:
>>>>>>>>>>>>>
>>>>>>>>>>>>> On 03/06/16 15:11, Matthias Brugger wrote:
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>
>>>>>>>>>>>>> [...]
>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>>>>>>> +            smp_mb(); /* modify jump before enable thread */
>>>>>>>>>>>>>>>>>>>>>>> +        }
>>>>>>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>>>>>>> +        cmdq_thread_writel(thread, task->pa_base +
>>>>>>>>>>>>>>>>>>>>>>> task->command_size,
>>>>>>>>>>>>>>>>>>>>>>> +                   CMDQ_THR_END_ADDR);
>>>>>>>>>>>>>>>>>>>>>>> +        cmdq_thread_resume(thread);
>>>>>>>>>>>>>>>>>>>>>>> +    }
>>>>>>>>>>>>>>>>>>>>>>> +    list_move_tail(&task->list_entry, &thread->task_busy_list);
>>>>>>>>>>>>>>>>>>>>>>> +    spin_unlock_irqrestore(&cmdq->exec_lock, flags);
>>>>>>>>>>>>>>>>>>>>>>> +}
>>>>>>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>>>>>>> +static void cmdq_handle_error_done(struct cmdq *cmdq,
>>>>>>>>>>>>>>>>>>>>>>> +                   struct cmdq_thread *thread, u32 irq_flag)
>>>>>>>>>>>>>>>>>>>>>>> +{
>>>>>>>>>>>>>>>>>>>>>>> +    struct cmdq_task *task, *tmp, *curr_task = NULL;
>>>>>>>>>>>>>>>>>>>>>>> +    u32 curr_pa;
>>>>>>>>>>>>>>>>>>>>>>> +    struct cmdq_cb_data cmdq_cb_data;
>>>>>>>>>>>>>>>>>>>>>>> +    bool err;
>>>>>>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>>>>>>> +    if (irq_flag & CMDQ_THR_IRQ_ERROR)
>>>>>>>>>>>>>>>>>>>>>>> +        err = true;
>>>>>>>>>>>>>>>>>>>>>>> +    else if (irq_flag & CMDQ_THR_IRQ_DONE)
>>>>>>>>>>>>>>>>>>>>>>> +        err = false;
>>>>>>>>>>>>>>>>>>>>>>> +    else
>>>>>>>>>>>>>>>>>>>>>>> +        return;
>>>>>>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>>>>>>> +    curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
>>>>>>>>>>>>>>>>>>>>>>> +
>>>>>>>>>>>>>>>>>>>>>>> +    list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
>>>>>>>>>>>>>>>>>>>>>>> +                 list_entry) {
>>>>>>>>>>>>>>>>>>>>>>> +        if (curr_pa >= task->pa_base &&
>>>>>>>>>>>>>>>>>>>>>>> +            curr_pa < (task->pa_base + task->command_size))
>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>>> What are you checking here? It seems as if you make some implcit
>>>>>>>>>>>>>>>>>>>>>> assumptions about pa_base and the order of execution of
>>>>>>>>>>>>>>>>>>>>>> commands in the
>>>>>>>>>>>>>>>>>>>>>> thread. Is it save to do so? Does dma_alloc_coherent give any
>>>>>>>>>>>>>>>>>>>>>> guarantees
>>>>>>>>>>>>>>>>>>>>>> about dma_handle?
>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>> 1. Check what is the current running task in this GCE thread.
>>>>>>>>>>>>>>>>>>>>> 2. Yes.
>>>>>>>>>>>>>>>>>>>>> 3. Yes, CMDQ doesn't use iommu, so physical address is continuous.
>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>> Yes, physical addresses might be continous, but AFAIK there is no
>>>>>>>>>>>>>>>>>>>> guarantee that the dma_handle address is steadily growing, when
>>>>>>>>>>>>>>>>>>>> calling
>>>>>>>>>>>>>>>>>>>> dma_alloc_coherent. And if I understand the code correctly, you
>>>>>>>>>>>>>>>>>>>> use this
>>>>>>>>>>>>>>>>>>>> assumption to decide if the task picked from task_busy_list is
>>>>>>>>>>>>>>>>>>>> currently
>>>>>>>>>>>>>>>>>>>> executing. So I think this mecanism is not working.
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>> I don't use dma_handle address, and just use physical addresses.
>>>>>>>>>>>>>>>>>>>         From CPU's point of view, tasks are linked by the busy list.
>>>>>>>>>>>>>>>>>>>         From GCE's point of view, tasks are linked by the JUMP command.
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>>> In which cases does the HW thread raise an interrupt.
>>>>>>>>>>>>>>>>>>>> In case of error. When does CMDQ_THR_IRQ_DONE get raised?
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>> GCE will raise interrupt if any task is done or error.
>>>>>>>>>>>>>>>>>>> However, GCE is fast, so CPU may get multiple done tasks
>>>>>>>>>>>>>>>>>>> when it is running ISR.
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>> In case of error, that GCE thread will pause and raise interrupt.
>>>>>>>>>>>>>>>>>>> So, CPU may get multiple done tasks and one error task.
>>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>>> I think we should reimplement the ISR mechanism. Can't we just read
>>>>>>>>>>>>>>>>>> CURR_IRQ_STATUS and THR_IRQ_STATUS in the handler and leave
>>>>>>>>>>>>>>>>>> cmdq_handle_error_done to the thread_fn? You will need to pass
>>>>>>>>>>>>>>>>>> information from the handler to thread_fn, but that shouldn't be an
>>>>>>>>>>>>>>>>>> issue. AFAIK interrupts are disabled in the handler, so we should stay
>>>>>>>>>>>>>>>>>> there as short as possible. Traversing task_busy_list is expensive, so
>>>>>>>>>>>>>>>>>> we need to do it in a thread context.
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>> Actually, our initial implementation is similar to your suggestion,
>>>>>>>>>>>>>>>>> but display needs CMDQ to return callback function very precisely,
>>>>>>>>>>>>>>>>> else display will drop frame.
>>>>>>>>>>>>>>>>> For display, CMDQ interrupt will be raised every 16 ~ 17 ms,
>>>>>>>>>>>>>>>>> and CMDQ needs to call callback function in ISR.
>>>>>>>>>>>>>>>>> If we defer callback to workqueue, the time interval may be larger than
>>>>>>>>>>>>>>>>> 32 ms.sometimes.
>>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> I think the problem is, that you implemented the workqueue as a ordered
>>>>>>>>>>>>>>>> workqueue, so there is no parallel processing. I'm still not sure why
>>>>>>>>>>>>>>>> you need the workqueue to be ordered. Can you please explain.
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> The order should be kept.
>>>>>>>>>>>>>>> Let me use mouse cursor as an example.
>>>>>>>>>>>>>>> If task 1 means move mouse cursor to point A, task 2 means point B,
>>>>>>>>>>>>>>> and task 3 means point C, our expected result is A -> B -> C.
>>>>>>>>>>>>>>> If the order is not kept, the result could become A -> C -> B.
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> Got it, thanks for the clarification.
>>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>> I think a way to get rid of the workqueue is to use a timer, which gets
>>>>>>>>>>>>> programmed to the time a timeout in the first task in the busy list
>>>>>>>>>>>>> would happen. Everytime we update the busy list (e.g. because of task
>>>>>>>>>>>>> got finished by the thread), we update the timer. When the timer
>>>>>>>>>>>>> triggers, which hopefully won't happen too often, we return timeout on
>>>>>>>>>>>>> the busy list elements, until the time is lower then the actual time.
>>>>>>>>>>>>>
>>>>>>>>>>>>> At least with this we can reduce the data structures in this driver and
>>>>>>>>>>>>> make it more lightweight.
>>>>>>>>>>>>
>>>>>>>>>>>>      From my understanding, your proposed method can handle timeout case.
>>>>>>>>>>>>
>>>>>>>>>>>> However, the workqueue is also in charge of releasing tasks.
>>>>>>>>>>>> Do you take releasing tasks into consideration by using the proposed
>>>>>>>>>>>> timer method?
>>>>>>>>>>>> Furthermore, I think the code will become more complex if we also use
>>>>>>>>>>>> timer to implement releasing tasks.
>>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> Can't we call
>>>>>>>>>>>              clk_disable_unprepare(cmdq->clock);
>>>>>>>>>>>              cmdq_task_release(task);
>>>>>>>>>>> after invoking the callback?
>>>>>
>>>>> After I put clk_disable_unprepare(cmdq->clock) into ISR, I encounter
>>>>> another BUG.
>>>>>
>>>>> (Quote some Linux 4.7 source code.)
>>>>>
>>>>>    605 void clk_unprepare(struct clk *clk)
>>>>>    606 {
>>>>>    607         if (IS_ERR_OR_NULL(clk))
>>>>>    608                 return;
>>>>>    609
>>>>>    610         clk_prepare_lock();                      // <-- Here
>>>>>    611         clk_core_unprepare(clk->core);
>>>>>    612         clk_prepare_unlock();
>>>>>    613 }
>>>>>    614 EXPORT_SYMBOL_GPL(clk_unprepare);
>>>>>
>>>>>     91 static void clk_prepare_lock(void)
>>>>>     92 {
>>>>>     93         if (!mutex_trylock(&prepare_lock)) {     // <-- Here
>>>>>     94                 if (prepare_owner == current) {
>>>>>     95                         prepare_refcnt++;
>>>>>     96                         return;
>>>>>     97                 }
>>>>>     98                 mutex_lock(&prepare_lock);
>>>>>     99         }
>>>>>    100         WARN_ON_ONCE(prepare_owner != NULL);
>>>>>    101         WARN_ON_ONCE(prepare_refcnt != 0);
>>>>>    102         prepare_owner = current;
>>>>>    103         prepare_refcnt = 1;
>>>>>    104 }
>>>>>
>>>>> So, 'unprepare' can sleep and cannot be put into ISR.
>>>>> I also try to put it into a timer, but the error is the same
>>>>> since timer callback is executed by softirq.
>>>>>
>>>>> We need clk_disable_unprepare() since it can save power consumption
>>>>> in idle.
>>>>
>>>> We can call clk_prepare in probe and then use clk_enable/clk_disable,
>>>> which don't sleep.
>>>>
>>>> Regards,
>>>> Matthias
>>>
>>> Hi Matthias,
>>>
>>> Because clock gate and MUX are controlled by clk_enable/clk_disable,
>>> and PLL is controlled by clk_prepare/clk_unprepare,
>>> I still need to call clk_unprepare.
>>>
>>> After I remove releasing buffer, releasing task, and timeout task from
>>> work, the work can be detached from task.
>>>
>>> Therefore, I can use the following flow to reduce the number of works.
>>>
>>> if task_busy_list from empty to non-empty
>>> 	clk_prepare_enable
>>> if task_busy_list from non-empty to empty
>>> 	in ISR, add work for clk_disable_unprepare
>>>
>>> What do you think of this solution?
>>
>> Can't we just call clk_prepare in probe and clk_unprepare in remove? I
>> think this could be a good starting point, and if we see, that we need
>> to save more energy in the future, we can think of some other mechanism.
>> What do you think?
>>
>> Regards,
>> Matthias
>
> Hi Matthias,
>
> As far as I know, we should call clk_unprepare to save more energy.
>
> May I call clk_prepare in probe/resume and clk_unprepare in
> remove/suspend in this patch, and then prepare another patch to call
> clk_unprepare in idle to save more energy?
>

Sure. This was just a suggestion to a first working version of the 
driver to which we can add step-by-step new functionality.

Regards,
Matthias

> Thanks,
> HS
>
>>>
>>> Thanks,
>>> HS
>>>
>>>>> Therefore, I plan to
>>>>> (1) move releasing buffer and task into ISR,
>>>>> (2) move timeout into timer, and
>>>>> (3) keep workqueue for clk_disable_unprepare().
>>>>>
>>>>> What do you think?
>>>>>
>>>>> Thanks,
>>>>> HS
>>>>>
>>>>>>>>>>
>>>>>>>>>> Do you mean just call these two functions in ISR?
>>>>>>>>>> My major concern is dma_free_coherent() and kfree() in
>>>>>>>>>> cmdq_task_release(task).
>>>>>>>>>
>>>>>>>>> Why do we need the dma calls at all? Can't we just calculate the
>>>>>>>>> physical address using __pa(x)?
>>>>>>>>
>>>>>>>> I prefer to use dma_map_single/dma_unmap_single.
>>>>>>>>
>>>>>>>
>>>>>>> Can you please elaborate why you need this. We don't do dma, so we
>>>>>>> should not use dma memory for this.
>>>>>>
>>>>>> We need a buffer to share between CPU and GCE, so we do need DMA.
>>>>>> CPU is in charge of writing GCE commands into this buffer.
>>>>>> GCE is in charge of reading and running GCE commands from this buffer.
>>>>>> When we chain CMDQ tasks, we also need to modify GCE JUMP command.
>>>>>> Therefore, I prefer to use dma_alloc_coherent and dma_free_coherent.
>>>>>>
>>>>>> However, if we want to use timer to handle timeout, we need to release
>>>>>> memory in ISR.
>>>>>> In this case, using kmalloc/kfree + dma_map_single/dma_unmap_single
>>>>>> instead of dma_alloc_coherent/dma_free_coherent is an alternative
>>>>>> solution, but taking care the synchronization between cache and memory
>>>>>> is the expected overhead.
>>>>>>
>>>>>>>>>> Therefore, your suggestion is to use GFP_ATOMIC for both
>>>>>>>>>> dma_alloc_coherent() and kzalloc(). Right?
>>>>>>>>>
>>>>>>>>> I don't think we need GFP_ATOMIC, the critical path will just free the
>>>>>>>>> memory.
>>>>>>>>
>>>>>>>> I tested these two functions, and kfree was safe.
>>>>>>>> However, dma_free_coherent raised BUG.
>>>>>>>> BUG: failure at
>>>>>>>> /mnt/host/source/src/third_party/kernel/v3.18/mm/vmalloc.c:1514/vunmap()!
>>>>>>>
>>>>>>> Just a general hint. Please try to evaluate on a recent kernel. It looks
>>>>>>> like as if you tried this on a v3.18 based one.
>>>>>>
>>>>>> This driver should be backward compatible to v3.18 for a MTK project.
>>>>>>
>>>>>>> Best regards,
>>>>>>> Matthias
>>>>>>
>>>>>> Thanks,
>>>>>> HS
>>>>>>
>>>>>>>> 1512 void vunmap(const void *addr)
>>>>>>>> 1513 {
>>>>>>>> 1514         BUG_ON(in_interrupt());		// <-- here
>>>>>>>> 1515         might_sleep();
>>>>>>>> 1516         if (addr)
>>>>>>>> 1517                 __vunmap(addr, 0);
>>>>>>>> 1518 }
>>>>>>>> 1519 EXPORT_SYMBOL(vunmap);
>>>>>>>>
>>>>>>>> Therefore, I plan to use kmalloc + dma_map_single instead of
>>>>>>>> dma_alloc_coherent, and dma_unmap_single + kfree instead of
>>>>>>>> dma_free_coherent.
>>>>>>>>
>>>>>>>> What do you think about the function replacement?
>>>>>>>>
>>>>>>>>>> If so, I can try to implement timeout by timer, and discuss with you
>>>>>>>>>> if I have further questions.
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Sounds good :)
>>>>>>>>>
>>>>>>>>> Thanks,
>>>>>>>>> Matthias
>>>>>>>>
>>>>>>>> Thanks,
>>>>>>>> HS
>>>>>>>>
>>>>>>>>>>> Regrading the clock, wouldn't it be easier to handle the clock
>>>>>>>>>>> enable/disable depending on the state of task_busy_list? I suppose we
>>>>>>>>>>> can't as we would need to check the task_busy_list of all threads, right?
>>>>>>>>>>>
>>>>>>>>>>> Regards,
>>>>>>>>>>> Matthias
>>>>>>>>>>
>>>>>>>>>> Thanks,
>>>>>>>>>> HS
>>>
>>>
>
>

CK Hu (胡俊光) June 23, 2016, 6:03 a.m. UTC | #36

Hi, HS:

On Mon, 2016-05-30 at 11:19 +0800, HS Liao wrote:
> This patch is first version of Mediatek Command Queue(CMDQ) driver. The
> CMDQ is used to help read/write registers with critical time limitation,
> such as updating display configuration during the vblank. It controls
> Global Command Engine (GCE) hardware to achieve this requirement.
> Currently, CMDQ only supports display related hardwares, but we expect
> it can be extended to other hardwares for future requirements.
> 
> Signed-off-by: HS Liao <[e-mail]>
> Signed-off-by: CK Hu <[e-mail]>
> ---
>  drivers/soc/mediatek/Kconfig    |  10 +
>  drivers/soc/mediatek/Makefile   |   1 +
>  drivers/soc/mediatek/mtk-cmdq.c | 943 ++++++++++++++++++++++++++++++++++++++++
>  include/soc/mediatek/cmdq.h     | 197 +++++++++
>  4 files changed, 1151 insertions(+)
>  create mode 100644 drivers/soc/mediatek/mtk-cmdq.c
>  create mode 100644 include/soc/mediatek/cmdq.h

[...]

> +
> +/* events for CMDQ and display */
> +enum cmdq_event {
> +	/* Display start of frame(SOF) events */
> +	CMDQ_EVENT_DISP_OVL0_SOF = 11,
> +	CMDQ_EVENT_DISP_OVL1_SOF = 12,
> +	CMDQ_EVENT_DISP_RDMA0_SOF = 13,
> +	CMDQ_EVENT_DISP_RDMA1_SOF = 14,
> +	CMDQ_EVENT_DISP_RDMA2_SOF = 15,
> +	CMDQ_EVENT_DISP_WDMA0_SOF = 16,
> +	CMDQ_EVENT_DISP_WDMA1_SOF = 17,
> +	/* Display end of frame(EOF) events */
> +	CMDQ_EVENT_DISP_OVL0_EOF = 39,
> +	CMDQ_EVENT_DISP_OVL1_EOF = 40,
> +	CMDQ_EVENT_DISP_RDMA0_EOF = 41,
> +	CMDQ_EVENT_DISP_RDMA1_EOF = 42,
> +	CMDQ_EVENT_DISP_RDMA2_EOF = 43,
> +	CMDQ_EVENT_DISP_WDMA0_EOF = 44,
> +	CMDQ_EVENT_DISP_WDMA1_EOF = 45,
> +	/* Mutex end of frame(EOF) events */
> +	CMDQ_EVENT_MUTEX0_STREAM_EOF = 53,
> +	CMDQ_EVENT_MUTEX1_STREAM_EOF = 54,
> +	CMDQ_EVENT_MUTEX2_STREAM_EOF = 55,
> +	CMDQ_EVENT_MUTEX3_STREAM_EOF = 56,
> +	CMDQ_EVENT_MUTEX4_STREAM_EOF = 57,
> +	/* Display underrun events */
> +	CMDQ_EVENT_DISP_RDMA0_UNDERRUN = 63,
> +	CMDQ_EVENT_DISP_RDMA1_UNDERRUN = 64,
> +	CMDQ_EVENT_DISP_RDMA2_UNDERRUN = 65,
> +	/* Keep this at the end of HW events */
> +	CMDQ_MAX_HW_EVENT_COUNT = 260,
> +};

The value of these symbol is just the GCE-HW-defined value. I think it's
not appropriate to expose HW-defined value to client. For another SoC
GCE HW, these definition may change.

One way to solve this problem is to translate symbol to value
internally. But these events looks like interrupt and the value may vary
by each SoC, to prevent driver modified frequently, it's better to place
the value definition in device tree. It may looks like:

mmsys: clock-controller@14000000 {
	compatible = "mediatek,mt8173-mmsys";
	mediatek,gce = <&gce>;
	gce-events = <53 54>;
	gce-event-names = "MUTEX0_EOF","MUTEX1_EOF";
}

For cmdq driver, you just need modify interface from

int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event)
int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event)

to

int cmdq_rec_wfe(struct cmdq_rec *rec, u32 event)
int cmdq_rec_clear_event(struct cmdq_rec *rec, u32 event)

Regards,
CK

[e-mail] June 23, 2016, 7:54 a.m. UTC | #37

Hi CK,

On Thu, 2016-06-23 at 14:03 +0800, CK Hu wrote:
> Hi, HS:
> 
> On Mon, 2016-05-30 at 11:19 +0800, HS Liao wrote:
> [...]
> 
> > +
> > +/* events for CMDQ and display */
> > +enum cmdq_event {
> > +	/* Display start of frame(SOF) events */
> > +	CMDQ_EVENT_DISP_OVL0_SOF = 11,
> > +	CMDQ_EVENT_DISP_OVL1_SOF = 12,
> > +	CMDQ_EVENT_DISP_RDMA0_SOF = 13,
> > +	CMDQ_EVENT_DISP_RDMA1_SOF = 14,
> > +	CMDQ_EVENT_DISP_RDMA2_SOF = 15,
> > +	CMDQ_EVENT_DISP_WDMA0_SOF = 16,
> > +	CMDQ_EVENT_DISP_WDMA1_SOF = 17,
> > +	/* Display end of frame(EOF) events */
> > +	CMDQ_EVENT_DISP_OVL0_EOF = 39,
> > +	CMDQ_EVENT_DISP_OVL1_EOF = 40,
> > +	CMDQ_EVENT_DISP_RDMA0_EOF = 41,
> > +	CMDQ_EVENT_DISP_RDMA1_EOF = 42,
> > +	CMDQ_EVENT_DISP_RDMA2_EOF = 43,
> > +	CMDQ_EVENT_DISP_WDMA0_EOF = 44,
> > +	CMDQ_EVENT_DISP_WDMA1_EOF = 45,
> > +	/* Mutex end of frame(EOF) events */
> > +	CMDQ_EVENT_MUTEX0_STREAM_EOF = 53,
> > +	CMDQ_EVENT_MUTEX1_STREAM_EOF = 54,
> > +	CMDQ_EVENT_MUTEX2_STREAM_EOF = 55,
> > +	CMDQ_EVENT_MUTEX3_STREAM_EOF = 56,
> > +	CMDQ_EVENT_MUTEX4_STREAM_EOF = 57,
> > +	/* Display underrun events */
> > +	CMDQ_EVENT_DISP_RDMA0_UNDERRUN = 63,
> > +	CMDQ_EVENT_DISP_RDMA1_UNDERRUN = 64,
> > +	CMDQ_EVENT_DISP_RDMA2_UNDERRUN = 65,
> > +	/* Keep this at the end of HW events */
> > +	CMDQ_MAX_HW_EVENT_COUNT = 260,
> > +};
> 
> The value of these symbol is just the GCE-HW-defined value. I think it's
> not appropriate to expose HW-defined value to client. For another SoC
> GCE HW, these definition may change.

Agree.

> One way to solve this problem is to translate symbol to value
> internally. But these events looks like interrupt and the value may vary
> by each SoC, to prevent driver modified frequently, it's better to place
> the value definition in device tree. It may looks like:
> 
> mmsys: clock-controller@14000000 {
> 	compatible = "mediatek,mt8173-mmsys";
> 	mediatek,gce = <&gce>;
> 	gce-events = <53 54>;

However, client still needs to know the HW-defined value by using
this solution.

> 	gce-event-names = "MUTEX0_EOF","MUTEX1_EOF";
> }
> 
> For cmdq driver, you just need modify interface from
> 
> int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event)
> int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event)
> 
> to
> 
> int cmdq_rec_wfe(struct cmdq_rec *rec, u32 event)
> int cmdq_rec_clear_event(struct cmdq_rec *rec, u32 event)

I think an easy way for this issue is just removing HW-defined values
from include/soc/mediatek/cmdq.h (but keeping enum), and then mapping
abstract values into real values in driver.
The benefit of this solution is that we can keep interface and leave SoC
specific code into driver.
What do you think?

> Regards,
> CK

Thanks,
HS

CK Hu (胡俊光) June 23, 2016, 11:44 a.m. UTC | #38

On Thu, 2016-06-23 at 15:54 +0800, Horng-Shyang Liao wrote:
> Hi CK,
> 
> On Thu, 2016-06-23 at 14:03 +0800, CK Hu wrote:
> > Hi, HS:
> > 
> > On Mon, 2016-05-30 at 11:19 +0800, HS Liao wrote:
> > [...]
> > 
> > > +
> > > +/* events for CMDQ and display */
> > > +enum cmdq_event {
> > > +	/* Display start of frame(SOF) events */
> > > +	CMDQ_EVENT_DISP_OVL0_SOF = 11,
> > > +	CMDQ_EVENT_DISP_OVL1_SOF = 12,
> > > +	CMDQ_EVENT_DISP_RDMA0_SOF = 13,
> > > +	CMDQ_EVENT_DISP_RDMA1_SOF = 14,
> > > +	CMDQ_EVENT_DISP_RDMA2_SOF = 15,
> > > +	CMDQ_EVENT_DISP_WDMA0_SOF = 16,
> > > +	CMDQ_EVENT_DISP_WDMA1_SOF = 17,
> > > +	/* Display end of frame(EOF) events */
> > > +	CMDQ_EVENT_DISP_OVL0_EOF = 39,
> > > +	CMDQ_EVENT_DISP_OVL1_EOF = 40,
> > > +	CMDQ_EVENT_DISP_RDMA0_EOF = 41,
> > > +	CMDQ_EVENT_DISP_RDMA1_EOF = 42,
> > > +	CMDQ_EVENT_DISP_RDMA2_EOF = 43,
> > > +	CMDQ_EVENT_DISP_WDMA0_EOF = 44,
> > > +	CMDQ_EVENT_DISP_WDMA1_EOF = 45,
> > > +	/* Mutex end of frame(EOF) events */
> > > +	CMDQ_EVENT_MUTEX0_STREAM_EOF = 53,
> > > +	CMDQ_EVENT_MUTEX1_STREAM_EOF = 54,
> > > +	CMDQ_EVENT_MUTEX2_STREAM_EOF = 55,
> > > +	CMDQ_EVENT_MUTEX3_STREAM_EOF = 56,
> > > +	CMDQ_EVENT_MUTEX4_STREAM_EOF = 57,
> > > +	/* Display underrun events */
> > > +	CMDQ_EVENT_DISP_RDMA0_UNDERRUN = 63,
> > > +	CMDQ_EVENT_DISP_RDMA1_UNDERRUN = 64,
> > > +	CMDQ_EVENT_DISP_RDMA2_UNDERRUN = 65,
> > > +	/* Keep this at the end of HW events */
> > > +	CMDQ_MAX_HW_EVENT_COUNT = 260,
> > > +};
> > 
> > The value of these symbol is just the GCE-HW-defined value. I think it's
> > not appropriate to expose HW-defined value to client. For another SoC
> > GCE HW, these definition may change.
> 
> Agree.
> 
> > One way to solve this problem is to translate symbol to value
> > internally. But these events looks like interrupt and the value may vary
> > by each SoC, to prevent driver modified frequently, it's better to place
> > the value definition in device tree. It may looks like:
> > 
> > mmsys: clock-controller@14000000 {
> > 	compatible = "mediatek,mt8173-mmsys";
> > 	mediatek,gce = <&gce>;
> > 	gce-events = <53 54>;
> 
> However, client still needs to know the HW-defined value by using
> this solution.
> 
> > 	gce-event-names = "MUTEX0_EOF","MUTEX1_EOF";
> > }
> > 
> > For cmdq driver, you just need modify interface from
> > 
> > int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event)
> > int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event)
> > 
> > to
> > 
> > int cmdq_rec_wfe(struct cmdq_rec *rec, u32 event)
> > int cmdq_rec_clear_event(struct cmdq_rec *rec, u32 event)
> 
> I think an easy way for this issue is just removing HW-defined values
> from include/soc/mediatek/cmdq.h (but keeping enum), and then mapping
> abstract values into real values in driver.
> The benefit of this solution is that we can keep interface and leave SoC
> specific code into driver.
> What do you think?

Hi, HS:

OK, that's fine.

Regards,
CK

> 
> > Regards,
> > CK
> 
> Thanks,
> HS
> 
>

[e-mail] June 24, 2016, 11:39 a.m. UTC | #39

On Tue, 2016-06-21 at 15:46 +0800, Horng-Shyang Liao wrote:
> On Tue, 2016-06-21 at 10:03 +0800, CK Hu wrote:
> > On Mon, 2016-06-20 at 19:22 +0800, Horng-Shyang Liao wrote:
> > > On Mon, 2016-06-20 at 18:41 +0800, CK Hu wrote:
> > > > [Snip...]
> > > > 
> > > > > +
> > > > > +static int cmdq_eng_get_thread(u64 flag)
> > > > > +{
> > > > > +	if (flag & BIT_ULL(CMDQ_ENG_DISP_DSI0))
> > > > > +		return CMDQ_THR_DISP_MAIN_IDX;
> > > > > +	else if (flag & BIT_ULL(CMDQ_ENG_DISP_DPI0))
> > > > > +		return CMDQ_THR_DISP_SUB_IDX;
> > > > > +	else
> > > > > +		return CMDQ_THR_DISP_MISC_IDX;
> > > > > +}
> > > > 
> > > > I think cmdq should not have knowledge of client. These statement show
> > > > that cmdq know that DSI0-tasks is different with DPI0-tasks. I propose
> > > > the 'session' to replace engine flag. For example, main display create
> > > > one cmdq_session and external display create another cmdq_session. For
> > > > client driver, every tasks created by main display is bound to main
> > > > display session. For cmdq driver, it should dynamically bind a session
> > > > to a HW thread, and then dispatch tasks of this session to this HW
> > > > thread. After HW thread run out of tasks of this session, detach this
> > > > session with this thread.
> > > > For the problem of remove wfe cmd, I think a session can have a property
> > > > of merge_wfe_cmd. For display session, session->merge_wfe_cmd = true.
> > > > For other client, it's false.
> > > 
> > > Hi CK,
> > > 
> > > I think your suggestion is similar to CMDQ 'scenarios',
> > > which was removed from CMDQ v2.
> > > 
> > > Daniel suggests to use engine flags instead of scenarios.
> > > Quote from https://patchwork.kernel.org/patch/8068311/ .
> > > 'Instead of encoding policy (these arbitrary "scenarios"), perhaps the
> > >  cmdq driver should just provide these flag bits, and the display
> > >  subsystem can use them to build the "flag" parameter itself.
> > > 
> > >  After all, the exact configuration of mmsys components is somewhat
> > >  flexible.'
> > > 
> > > Therefore, it would be better to discuss with Daniel before we change
> > > it.
> > > 
> > > 
> > > Hi Daniel,
> > > 
> > > Do you think we should use scenarios or sessions instead of flags?
> > > 
> > > Thanks,
> > > HS
> > > 
> > 
> > Hi, HS:
> > 
> > 'session' is not similar to 'scenarios'.
> > 
> > In 'scenarios' mechanism, client bind a task with scenarios and send to
> > cmdq. Cmdq transfer scenarios to engine flag, and use engine flag to
> > dispatch task to HW thread.
> > 
> > 
> > In 'engine flag' mechanism, proposed by Daniel, client bind a task
> > directly with engine flag. Cmdq directly use engine flag to dispatch
> > task to HW thread without any translation.
> > 
> > But neither 'scenarios' mechanism nor 'engine flag' mechanism get rid of
> > engine flag, which make cmdq have knowledge of client.
> > 
> > In 'session' mechanism, there is no engine flag any more. Client bind
> > time-sequential tasks to the same session and tasks in different session
> > can execute parallelly. One thing cmdq need to know is to dispatch tasks
> > with the same session to the same HW thread, so cmdq does not have any
> > knowledge of client.
> > 
> > Daniel focus on reduce translating scenarios to engine flag. I think
> > 'session' mechanism does not conflict with his opinion because we does
> > not translate 'session' to engine flag. Therefore, I think 'session' is
> > the best of these three mechanism.
> > 
> > Regards,
> > CK
> 
> Hi CK,
> 
> 'Session' looks like a group of options for CMDQ.
> CMDQ driver can just follow the options to run its flow,
> and doesn't need to know its client(s).
> 
> We don't have many options now, but it has good flexibility to extend
> for future requirements.
> 
> I will add it in next version.
> 
> Thanks,
> HS

Hi CK,

I think session is very similar to mailbox channel.
We can treat session's parameters as channel's arguments in device tree.
Linking session to GCE thread is also just like linking channel to GCE
thread.
Because I will use mailbox framework in CMDQ v9, we can use mailbox
channel instead of session.
What do you think?

Thanks,
HS

> > > > Here is the sample code to create cmdq_rec with session.
> > > > 
> > > > merge_wfe_cmd = true;
> > > > cmdq_session_create(merge_wfe_cmd, &primary_display_session);
> > > > cmdq_rec_create(dev, primary_display_session, &rec);
> > > > 
> > > > 
> > > > Therefore, the below definition can be removed.
> > > > 
> > > > > +
> > > > > +enum cmdq_eng {
> > > > > +	CMDQ_ENG_DISP_AAL,
> > > > > +	CMDQ_ENG_DISP_COLOR0,
> > > > > +	CMDQ_ENG_DISP_COLOR1,
> > > > > +	CMDQ_ENG_DISP_DPI0,
> > > > > +	CMDQ_ENG_DISP_DSI0,
> > > > > +	CMDQ_ENG_DISP_DSI1,
> > > > > +	CMDQ_ENG_DISP_GAMMA,
> > > > > +	CMDQ_ENG_DISP_OD,
> > > > > +	CMDQ_ENG_DISP_OVL0,
> > > > > +	CMDQ_ENG_DISP_OVL1,
> > > > > +	CMDQ_ENG_DISP_PWM0,
> > > > > +	CMDQ_ENG_DISP_PWM1,
> > > > > +	CMDQ_ENG_DISP_RDMA0,
> > > > > +	CMDQ_ENG_DISP_RDMA1,
> > > > > +	CMDQ_ENG_DISP_RDMA2,
> > > > > +	CMDQ_ENG_DISP_UFOE,
> > > > > +	CMDQ_ENG_DISP_WDMA0,
> > > > > +	CMDQ_ENG_DISP_WDMA1,
> > > > > +	CMDQ_ENG_MAX,
> > > > > +};
> > > > > +
> > > > 
> > > > 
> > > > Regards,
> > > > CK

CK Hu (胡俊光) June 27, 2016, 2 a.m. UTC | #40

On Fri, 2016-06-24 at 19:39 +0800, Horng-Shyang Liao wrote:
> On Tue, 2016-06-21 at 15:46 +0800, Horng-Shyang Liao wrote:
> > On Tue, 2016-06-21 at 10:03 +0800, CK Hu wrote:
> > > On Mon, 2016-06-20 at 19:22 +0800, Horng-Shyang Liao wrote:
> > > > On Mon, 2016-06-20 at 18:41 +0800, CK Hu wrote:
> > > > > [Snip...]
> > > > > 
> > > > > > +
> > > > > > +static int cmdq_eng_get_thread(u64 flag)
> > > > > > +{
> > > > > > +	if (flag & BIT_ULL(CMDQ_ENG_DISP_DSI0))
> > > > > > +		return CMDQ_THR_DISP_MAIN_IDX;
> > > > > > +	else if (flag & BIT_ULL(CMDQ_ENG_DISP_DPI0))
> > > > > > +		return CMDQ_THR_DISP_SUB_IDX;
> > > > > > +	else
> > > > > > +		return CMDQ_THR_DISP_MISC_IDX;
> > > > > > +}
> > > > > 
> > > > > I think cmdq should not have knowledge of client. These statement show
> > > > > that cmdq know that DSI0-tasks is different with DPI0-tasks. I propose
> > > > > the 'session' to replace engine flag. For example, main display create
> > > > > one cmdq_session and external display create another cmdq_session. For
> > > > > client driver, every tasks created by main display is bound to main
> > > > > display session. For cmdq driver, it should dynamically bind a session
> > > > > to a HW thread, and then dispatch tasks of this session to this HW
> > > > > thread. After HW thread run out of tasks of this session, detach this
> > > > > session with this thread.
> > > > > For the problem of remove wfe cmd, I think a session can have a property
> > > > > of merge_wfe_cmd. For display session, session->merge_wfe_cmd = true.
> > > > > For other client, it's false.
> > > > 
> > > > Hi CK,
> > > > 
> > > > I think your suggestion is similar to CMDQ 'scenarios',
> > > > which was removed from CMDQ v2.
> > > > 
> > > > Daniel suggests to use engine flags instead of scenarios.
> > > > Quote from https://patchwork.kernel.org/patch/8068311/ .
> > > > 'Instead of encoding policy (these arbitrary "scenarios"), perhaps the
> > > >  cmdq driver should just provide these flag bits, and the display
> > > >  subsystem can use them to build the "flag" parameter itself.
> > > > 
> > > >  After all, the exact configuration of mmsys components is somewhat
> > > >  flexible.'
> > > > 
> > > > Therefore, it would be better to discuss with Daniel before we change
> > > > it.
> > > > 
> > > > 
> > > > Hi Daniel,
> > > > 
> > > > Do you think we should use scenarios or sessions instead of flags?
> > > > 
> > > > Thanks,
> > > > HS
> > > > 
> > > 
> > > Hi, HS:
> > > 
> > > 'session' is not similar to 'scenarios'.
> > > 
> > > In 'scenarios' mechanism, client bind a task with scenarios and send to
> > > cmdq. Cmdq transfer scenarios to engine flag, and use engine flag to
> > > dispatch task to HW thread.
> > > 
> > > 
> > > In 'engine flag' mechanism, proposed by Daniel, client bind a task
> > > directly with engine flag. Cmdq directly use engine flag to dispatch
> > > task to HW thread without any translation.
> > > 
> > > But neither 'scenarios' mechanism nor 'engine flag' mechanism get rid of
> > > engine flag, which make cmdq have knowledge of client.
> > > 
> > > In 'session' mechanism, there is no engine flag any more. Client bind
> > > time-sequential tasks to the same session and tasks in different session
> > > can execute parallelly. One thing cmdq need to know is to dispatch tasks
> > > with the same session to the same HW thread, so cmdq does not have any
> > > knowledge of client.
> > > 
> > > Daniel focus on reduce translating scenarios to engine flag. I think
> > > 'session' mechanism does not conflict with his opinion because we does
> > > not translate 'session' to engine flag. Therefore, I think 'session' is
> > > the best of these three mechanism.
> > > 
> > > Regards,
> > > CK
> > 
> > Hi CK,
> > 
> > 'Session' looks like a group of options for CMDQ.
> > CMDQ driver can just follow the options to run its flow,
> > and doesn't need to know its client(s).
> > 
> > We don't have many options now, but it has good flexibility to extend
> > for future requirements.
> > 
> > I will add it in next version.
> > 
> > Thanks,
> > HS
> 
> Hi CK,
> 
> I think session is very similar to mailbox channel.
> We can treat session's parameters as channel's arguments in device tree.
> Linking session to GCE thread is also just like linking channel to GCE
> thread.
> Because I will use mailbox framework in CMDQ v9, we can use mailbox
> channel instead of session.
> What do you think?
> 
> Thanks,
> HS
> 

Hi, HS:

I'm not familiar with mailbox, but this sounds good to me.

Regards,
CK

> > > > > Here is the sample code to create cmdq_rec with session.
> > > > > 
> > > > > merge_wfe_cmd = true;
> > > > > cmdq_session_create(merge_wfe_cmd, &primary_display_session);
> > > > > cmdq_rec_create(dev, primary_display_session, &rec);
> > > > > 
> > > > > 
> > > > > Therefore, the below definition can be removed.
> > > > > 
> > > > > > +
> > > > > > +enum cmdq_eng {
> > > > > > +	CMDQ_ENG_DISP_AAL,
> > > > > > +	CMDQ_ENG_DISP_COLOR0,
> > > > > > +	CMDQ_ENG_DISP_COLOR1,
> > > > > > +	CMDQ_ENG_DISP_DPI0,
> > > > > > +	CMDQ_ENG_DISP_DSI0,
> > > > > > +	CMDQ_ENG_DISP_DSI1,
> > > > > > +	CMDQ_ENG_DISP_GAMMA,
> > > > > > +	CMDQ_ENG_DISP_OD,
> > > > > > +	CMDQ_ENG_DISP_OVL0,
> > > > > > +	CMDQ_ENG_DISP_OVL1,
> > > > > > +	CMDQ_ENG_DISP_PWM0,
> > > > > > +	CMDQ_ENG_DISP_PWM1,
> > > > > > +	CMDQ_ENG_DISP_RDMA0,
> > > > > > +	CMDQ_ENG_DISP_RDMA1,
> > > > > > +	CMDQ_ENG_DISP_RDMA2,
> > > > > > +	CMDQ_ENG_DISP_UFOE,
> > > > > > +	CMDQ_ENG_DISP_WDMA0,
> > > > > > +	CMDQ_ENG_DISP_WDMA1,
> > > > > > +	CMDQ_ENG_MAX,
> > > > > > +};
> > > > > > +
> > > > > 
> > > > > 
> > > > > Regards,
> > > > > CK
> 
>

Patch

diff --git a/drivers/soc/mediatek/Kconfig b/drivers/soc/mediatek/Kconfig
index 0a4ea80..c4ad75c 100644
--- a/drivers/soc/mediatek/Kconfig
+++ b/drivers/soc/mediatek/Kconfig
@@ -1,6 +1,16 @@ 
 #
 # MediaTek SoC drivers
 #
+config MTK_CMDQ
+	bool "MediaTek CMDQ Support"
+	depends on ARCH_MEDIATEK || COMPILE_TEST
+	select MTK_INFRACFG
+	help
+	  Say yes here to add support for the MediaTek Command Queue (CMDQ)
+	  driver. The CMDQ is used to help read/write registers with critical
+	  time limitation, such as updating display configuration during the
+	  vblank.
+
 config MTK_INFRACFG
 	bool "MediaTek INFRACFG Support"
 	depends on ARCH_MEDIATEK || COMPILE_TEST
diff --git a/drivers/soc/mediatek/Makefile b/drivers/soc/mediatek/Makefile
index 12998b0..f7397ef 100644
--- a/drivers/soc/mediatek/Makefile
+++ b/drivers/soc/mediatek/Makefile
@@ -1,3 +1,4 @@ 
+obj-$(CONFIG_MTK_CMDQ) += mtk-cmdq.o
 obj-$(CONFIG_MTK_INFRACFG) += mtk-infracfg.o
 obj-$(CONFIG_MTK_PMIC_WRAP) += mtk-pmic-wrap.o
 obj-$(CONFIG_MTK_SCPSYS) += mtk-scpsys.o
diff --git a/drivers/soc/mediatek/mtk-cmdq.c b/drivers/soc/mediatek/mtk-cmdq.c
new file mode 100644
index 0000000..e9d6e1c
--- /dev/null
+++ b/drivers/soc/mediatek/mtk-cmdq.c
@@ -0,0 +1,943 @@ 
+/*
+ * Copyright (c) 2015 MediaTek Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/clk.h>
+#include <linux/clk-provider.h>
+#include <linux/completion.h>
+#include <linux/dma-mapping.h>
+#include <linux/errno.h>
+#include <linux/interrupt.h>
+#include <linux/iopoll.h>
+#include <linux/kernel.h>
+#include <linux/kthread.h>
+#include <linux/module.h>
+#include <linux/mutex.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+#include <linux/suspend.h>
+#include <linux/workqueue.h>
+#include <soc/mediatek/cmdq.h>
+
+#define CMDQ_INITIAL_CMD_BLOCK_SIZE	PAGE_SIZE
+#define CMDQ_INST_SIZE			8 /* instruction is 64-bit */
+#define CMDQ_TIMEOUT_MS			1000
+#define CMDQ_IRQ_MASK			0xffff
+#define CMDQ_DRIVER_DEVICE_NAME		"mtk_cmdq"
+#define CMDQ_CLK_NAME			"gce"
+
+#define CMDQ_CURR_IRQ_STATUS		0x010
+#define CMDQ_CURR_LOADED_THR		0x018
+#define CMDQ_THR_SLOT_CYCLES		0x030
+
+#define CMDQ_THR_BASE			0x100
+#define CMDQ_THR_SHIFT			0x080
+#define CMDQ_THR_WARM_RESET		0x00
+#define CMDQ_THR_ENABLE_TASK		0x04
+#define CMDQ_THR_SUSPEND_TASK		0x08
+#define CMDQ_THR_CURR_STATUS		0x0c
+#define CMDQ_THR_IRQ_STATUS		0x10
+#define CMDQ_THR_IRQ_ENABLE		0x14
+#define CMDQ_THR_CURR_ADDR		0x20
+#define CMDQ_THR_END_ADDR		0x24
+#define CMDQ_THR_CFG			0x40
+
+#define CMDQ_THR_ENABLED		0x1
+#define CMDQ_THR_DISABLED		0x0
+#define CMDQ_THR_SUSPEND		0x1
+#define CMDQ_THR_RESUME			0x0
+#define CMDQ_THR_STATUS_SUSPENDED	BIT(1)
+#define CMDQ_THR_DO_WARM_RESET		BIT(0)
+#define CMDQ_THR_ACTIVE_SLOT_CYCLES	0x3200
+#define CMDQ_THR_PRIORITY		3
+#define CMDQ_THR_IRQ_DONE		0x1
+#define CMDQ_THR_IRQ_ERROR		0x12
+#define CMDQ_THR_IRQ_EN			0x13 /* done + error */
+#define CMDQ_THR_IRQ_MASK		0x13
+#define CMDQ_THR_EXECUTING		BIT(31)
+
+#define CMDQ_ARG_A_WRITE_MASK		0xffff
+#define CMDQ_SUBSYS_MASK		0x1f
+#define CMDQ_OP_CODE_MASK		0xff000000
+
+#define CMDQ_OP_CODE_SHIFT		24
+#define CMDQ_SUBSYS_SHIFT		16
+
+#define CMDQ_WRITE_ENABLE_MASK		BIT(0)
+#define CMDQ_JUMP_BY_OFFSET		0x10000000
+#define CMDQ_JUMP_BY_PA			0x10000001
+#define CMDQ_JUMP_PASS			CMDQ_INST_SIZE
+#define CMDQ_WFE_UPDATE			BIT(31)
+#define CMDQ_WFE_WAIT			BIT(15)
+#define CMDQ_WFE_WAIT_VALUE		0x1
+#define CMDQ_EOC_IRQ_EN			BIT(0)
+
+enum cmdq_thread_index {
+	CMDQ_THR_DISP_MAIN_IDX,	/* main */
+	CMDQ_THR_DISP_SUB_IDX,	/* sub */
+	CMDQ_THR_DISP_MISC_IDX,	/* misc */
+	CMDQ_THR_MAX_COUNT,	/* max */
+};
+
+/*
+ * CMDQ_CODE_MOVE:
+ *   move value into internal register as mask
+ *   format: op mask
+ * CMDQ_CODE_WRITE:
+ *   write value into target register
+ *   format: op subsys address value
+ * CMDQ_CODE_JUMP:
+ *   jump by offset
+ *   format: op offset
+ * CMDQ_CODE_WFE:
+ *   wait for event and clear
+ *   it is just clear if no wait
+ *   format: [wait]  op event update:1 to_wait:1 wait:1
+ *           [clear] op event update:1 to_wait:0 wait:0
+ * CMDQ_CODE_EOC:
+ *   end of command
+ *   format: op irq_flag
+ */
+enum cmdq_code {
+	CMDQ_CODE_MOVE = 0x02,
+	CMDQ_CODE_WRITE = 0x04,
+	CMDQ_CODE_JUMP = 0x10,
+	CMDQ_CODE_WFE = 0x20,
+	CMDQ_CODE_EOC = 0x40,
+};
+
+enum cmdq_task_state {
+	TASK_STATE_BUSY, /* running on a GCE thread */
+	TASK_STATE_ERROR,
+	TASK_STATE_DONE,
+};
+
+struct cmdq_task_cb {
+	cmdq_async_flush_cb	cb;
+	void			*data;
+};
+
+struct cmdq_thread {
+	void __iomem		*base;
+	struct list_head	task_busy_list;
+	wait_queue_head_t	wait_task_done;
+};
+
+struct cmdq_task {
+	struct cmdq		*cmdq;
+	struct list_head	list_entry;
+	enum cmdq_task_state	task_state;
+	void			*va_base;
+	dma_addr_t		pa_base;
+	u64			engine_flag;
+	size_t			command_size;
+	u32			num_cmd;
+	struct cmdq_thread	*thread;
+	struct cmdq_task_cb	cb;
+	struct work_struct	release_work;
+};
+
+struct cmdq {
+	struct device		*dev;
+	void __iomem		*base;
+	u32			irq;
+	struct workqueue_struct	*task_release_wq;
+	struct cmdq_thread	thread[CMDQ_THR_MAX_COUNT];
+	struct mutex		task_mutex;	/* for task */
+	spinlock_t		exec_lock;	/* for exec */
+	struct clk		*clock;
+	bool			suspended;
+};
+
+struct cmdq_subsys {
+	u32	base;
+	int	id;
+};
+
+static const struct cmdq_subsys gce_subsys[] = {
+	{0x1400, 1},
+	{0x1401, 2},
+	{0x1402, 3},
+};
+
+static int cmdq_subsys_base_to_id(u32 base)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(gce_subsys); i++)
+		if (gce_subsys[i].base == base)
+			return gce_subsys[i].id;
+	return -EFAULT;
+}
+
+static int cmdq_eng_get_thread(u64 flag)
+{
+	if (flag & BIT_ULL(CMDQ_ENG_DISP_DSI0))
+		return CMDQ_THR_DISP_MAIN_IDX;
+	else if (flag & BIT_ULL(CMDQ_ENG_DISP_DPI0))
+		return CMDQ_THR_DISP_SUB_IDX;
+	else
+		return CMDQ_THR_DISP_MISC_IDX;
+}
+
+static void cmdq_task_release(struct cmdq_task *task)
+{
+	struct cmdq *cmdq = task->cmdq;
+
+	dma_free_coherent(cmdq->dev, task->command_size, task->va_base,
+			  task->pa_base);
+	kfree(task);
+}
+
+static struct cmdq_task *cmdq_task_acquire(struct cmdq_rec *rec,
+					   struct cmdq_task_cb cb)
+{
+	struct cmdq *cmdq = rec->cmdq;
+	struct device *dev = cmdq->dev;
+	struct cmdq_task *task;
+
+	task = kzalloc(sizeof(*task), GFP_KERNEL);
+	INIT_LIST_HEAD(&task->list_entry);
+	task->va_base = dma_alloc_coherent(dev, rec->command_size,
+					   &task->pa_base, GFP_KERNEL);
+	if (!task->va_base) {
+		dev_err(dev, "allocate command buffer failed\n");
+		kfree(task);
+		return NULL;
+	}
+
+	task->cmdq = cmdq;
+	task->command_size = rec->command_size;
+	task->engine_flag = rec->engine_flag;
+	task->task_state = TASK_STATE_BUSY;
+	task->cb = cb;
+	memcpy(task->va_base, rec->buf, rec->command_size);
+	task->num_cmd = task->command_size / CMDQ_INST_SIZE;
+	return task;
+}
+
+static void cmdq_thread_writel(struct cmdq_thread *thread, u32 value,
+			       u32 offset)
+{
+	writel(value, thread->base + offset);
+}
+
+static u32 cmdq_thread_readl(struct cmdq_thread *thread, u32 offset)
+{
+	return readl(thread->base + offset);
+}
+
+static int cmdq_thread_suspend(struct cmdq *cmdq, struct cmdq_thread *thread)
+{
+	u32 status;
+
+	cmdq_thread_writel(thread, CMDQ_THR_SUSPEND, CMDQ_THR_SUSPEND_TASK);
+
+	/* If already disabled, treat as suspended successful. */
+	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
+	      CMDQ_THR_ENABLED))
+		return 0;
+
+	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_CURR_STATUS,
+			status, status & CMDQ_THR_STATUS_SUSPENDED, 0, 10)) {
+		dev_err(cmdq->dev, "suspend GCE thread 0x%x failed\n",
+			(u32)(thread->base - cmdq->base));
+		return -EFAULT;
+	}
+
+	return 0;
+}
+
+static void cmdq_thread_resume(struct cmdq_thread *thread)
+{
+	cmdq_thread_writel(thread, CMDQ_THR_RESUME, CMDQ_THR_SUSPEND_TASK);
+}
+
+static int cmdq_thread_reset(struct cmdq *cmdq, struct cmdq_thread *thread)
+{
+	u32 warm_reset;
+
+	cmdq_thread_writel(thread, CMDQ_THR_DO_WARM_RESET, CMDQ_THR_WARM_RESET);
+	if (readl_poll_timeout_atomic(thread->base + CMDQ_THR_WARM_RESET,
+			warm_reset, !(warm_reset & CMDQ_THR_DO_WARM_RESET),
+			0, 10)) {
+		dev_err(cmdq->dev, "reset GCE thread 0x%x failed\n",
+			(u32)(thread->base - cmdq->base));
+		return -EFAULT;
+	}
+	writel(CMDQ_THR_ACTIVE_SLOT_CYCLES, cmdq->base + CMDQ_THR_SLOT_CYCLES);
+	return 0;
+}
+
+static void cmdq_thread_disable(struct cmdq *cmdq, struct cmdq_thread *thread)
+{
+	cmdq_thread_reset(cmdq, thread);
+	cmdq_thread_writel(thread, CMDQ_THR_DISABLED, CMDQ_THR_ENABLE_TASK);
+}
+
+/* notify GCE to re-fetch commands by setting GCE thread PC */
+static void cmdq_thread_invalidate_fetched_data(struct cmdq_thread *thread)
+{
+	cmdq_thread_writel(thread,
+			   cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR),
+			   CMDQ_THR_CURR_ADDR);
+}
+
+static void cmdq_task_insert_into_thread(struct cmdq_task *task)
+{
+	struct cmdq_thread *thread = task->thread;
+	struct cmdq_task *prev_task = list_last_entry(
+			&thread->task_busy_list, typeof(*task), list_entry);
+	u64 *prev_task_base = prev_task->va_base;
+
+	/* let previous task jump to this task */
+	prev_task_base[prev_task->num_cmd - 1] = (u64)CMDQ_JUMP_BY_PA << 32 |
+						 task->pa_base;
+
+	cmdq_thread_invalidate_fetched_data(thread);
+}
+
+/* we assume tasks in the same display GCE thread are waiting the same event. */
+static void cmdq_task_remove_wfe(struct cmdq_task *task)
+{
+	u32 wfe_option = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
+	u32 wfe_op = CMDQ_CODE_WFE << CMDQ_OP_CODE_SHIFT;
+	u32 *base = task->va_base;
+	u32 num_cmd = task->num_cmd << 1;
+	int i;
+
+	for (i = 0; i < num_cmd; i += 2)
+		if (base[i] == wfe_option &&
+		    (base[i + 1] & CMDQ_OP_CODE_MASK) == wfe_op) {
+			base[i] = CMDQ_JUMP_PASS;
+			base[i + 1] = CMDQ_JUMP_BY_OFFSET;
+		}
+}
+
+static void cmdq_task_exec(struct cmdq_task *task, struct cmdq_thread *thread)
+{
+	struct cmdq *cmdq = task->cmdq;
+	unsigned long flags;
+	unsigned long curr_pa, end_pa;
+
+	WARN_ON(clk_prepare_enable(cmdq->clock) < 0);
+	spin_lock_irqsave(&cmdq->exec_lock, flags);
+	task->thread = thread;
+	task->task_state = TASK_STATE_BUSY;
+	if (list_empty(&thread->task_busy_list)) {
+		WARN_ON(cmdq_thread_reset(cmdq, thread) < 0);
+
+		cmdq_thread_writel(thread, task->pa_base, CMDQ_THR_CURR_ADDR);
+		cmdq_thread_writel(thread, task->pa_base + task->command_size,
+				   CMDQ_THR_END_ADDR);
+		cmdq_thread_writel(thread, CMDQ_THR_PRIORITY, CMDQ_THR_CFG);
+		cmdq_thread_writel(thread, CMDQ_THR_IRQ_EN,
+				   CMDQ_THR_IRQ_ENABLE);
+
+		cmdq_thread_writel(thread, CMDQ_THR_ENABLED,
+				   CMDQ_THR_ENABLE_TASK);
+	} else {
+		WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
+
+		/*
+		 * check boundary condition
+		 * PC = END - 8, EOC is executed
+		 * PC = END, all CMDs are executed
+		 */
+		curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
+		end_pa = cmdq_thread_readl(thread, CMDQ_THR_END_ADDR);
+		if (curr_pa == end_pa - 8 || curr_pa == end_pa) {
+			/* set to this task directly */
+			cmdq_thread_writel(thread, task->pa_base,
+					   CMDQ_THR_CURR_ADDR);
+		} else {
+			cmdq_task_insert_into_thread(task);
+
+			if (thread == &cmdq->thread[CMDQ_THR_DISP_MAIN_IDX] ||
+			    thread == &cmdq->thread[CMDQ_THR_DISP_SUB_IDX])
+				cmdq_task_remove_wfe(task);
+
+			smp_mb(); /* modify jump before enable thread */
+		}
+
+		cmdq_thread_writel(thread, task->pa_base + task->command_size,
+				   CMDQ_THR_END_ADDR);
+		cmdq_thread_resume(thread);
+	}
+	list_move_tail(&task->list_entry, &thread->task_busy_list);
+	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
+}
+
+static void cmdq_handle_error_done(struct cmdq *cmdq,
+				   struct cmdq_thread *thread, u32 irq_flag)
+{
+	struct cmdq_task *task, *tmp, *curr_task = NULL;
+	u32 curr_pa;
+	struct cmdq_cb_data cmdq_cb_data;
+	bool err;
+
+	if (irq_flag & CMDQ_THR_IRQ_ERROR)
+		err = true;
+	else if (irq_flag & CMDQ_THR_IRQ_DONE)
+		err = false;
+	else
+		return;
+
+	curr_pa = cmdq_thread_readl(thread, CMDQ_THR_CURR_ADDR);
+
+	list_for_each_entry_safe(task, tmp, &thread->task_busy_list,
+				 list_entry) {
+		if (curr_pa >= task->pa_base &&
+		    curr_pa < (task->pa_base + task->command_size))
+			curr_task = task;
+		if (task->cb.cb) {
+			cmdq_cb_data.err = curr_task ? err : false;
+			cmdq_cb_data.data = task->cb.data;
+			task->cb.cb(cmdq_cb_data);
+		}
+		task->task_state = (curr_task && err) ? TASK_STATE_ERROR :
+				TASK_STATE_DONE;
+		list_del(&task->list_entry);
+		if (curr_task)
+			break;
+	}
+
+	wake_up(&thread->wait_task_done);
+}
+
+static void cmdq_thread_irq_handler(struct cmdq *cmdq, int tid)
+{
+	struct cmdq_thread *thread = &cmdq->thread[tid];
+	unsigned long flags = 0L;
+	u32 irq_flag;
+
+	spin_lock_irqsave(&cmdq->exec_lock, flags);
+
+	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
+	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
+
+	/*
+	 * Another CPU core could run "release task" right before we acquire
+	 * the spin lock, and thus reset / disable this GCE thread, so we
+	 * need to check the enable bit of this GCE thread.
+	 */
+	if (!(cmdq_thread_readl(thread, CMDQ_THR_ENABLE_TASK) &
+	      CMDQ_THR_ENABLED))
+		irq_flag = 0;
+
+	cmdq_handle_error_done(cmdq, thread, irq_flag);
+	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
+}
+
+static irqreturn_t cmdq_irq_handler(int irq, void *dev)
+{
+	struct cmdq *cmdq = dev;
+	u32 irq_status;
+	int i;
+
+	irq_status = readl(cmdq->base + CMDQ_CURR_IRQ_STATUS);
+	irq_status &= CMDQ_IRQ_MASK;
+	irq_status ^= CMDQ_IRQ_MASK;
+
+	if (!irq_status)
+		return IRQ_NONE;
+
+	while (irq_status) {
+		i = ffs(irq_status) - 1;
+		irq_status &= ~BIT(i);
+		cmdq_thread_irq_handler(cmdq, i);
+	}
+
+	return IRQ_HANDLED;
+}
+
+static int cmdq_task_handle_error_result(struct cmdq_task *task)
+{
+	struct cmdq *cmdq = task->cmdq;
+	struct device *dev = cmdq->dev;
+	struct cmdq_thread *thread = task->thread;
+	struct cmdq_task *next_task, *prev_task;
+	u32 irq_flag;
+
+	/* suspend GCE thread to ensure consistency */
+	WARN_ON(cmdq_thread_suspend(cmdq, thread) < 0);
+
+	/* ISR has handled this error task */
+	if (task->task_state == TASK_STATE_ERROR) {
+		next_task = list_first_entry_or_null(&thread->task_busy_list,
+				struct cmdq_task, list_entry);
+		if (next_task) /* move to next task */
+			cmdq_thread_writel(thread, next_task->pa_base,
+					   CMDQ_THR_CURR_ADDR);
+		cmdq_thread_resume(thread);
+		return -ECANCELED;
+	}
+
+	/*
+	 * Save next_task and prev_task in advance
+	 * since cmdq_handle_error_done will remove list_entry.
+	 */
+	next_task = prev_task = NULL;
+	if (task->list_entry.next != &thread->task_busy_list)
+		next_task = list_next_entry(task, list_entry);
+	if (task->list_entry.prev != &thread->task_busy_list)
+		prev_task = list_prev_entry(task, list_entry);
+
+	/*
+	 * Although IRQ is disabled, GCE continues to execute.
+	 * It may have pending IRQ before GCE thread is suspended,
+	 * so check this condition again.
+	 */
+	irq_flag = cmdq_thread_readl(thread, CMDQ_THR_IRQ_STATUS);
+	cmdq_handle_error_done(cmdq, thread, irq_flag);
+	cmdq_thread_writel(thread, ~irq_flag, CMDQ_THR_IRQ_STATUS);
+
+	if (task->task_state == TASK_STATE_DONE) {
+		cmdq_thread_resume(thread);
+		return 0;
+	}
+
+	if (task->task_state == TASK_STATE_ERROR) {
+		dev_err(dev, "task 0x%p error\n", task);
+		if (next_task) /* move to next task */
+			cmdq_thread_writel(thread, next_task->pa_base,
+					   CMDQ_THR_CURR_ADDR);
+		cmdq_thread_resume(thread);
+		return -ECANCELED;
+	}
+
+	/* Task is running, so we force to remove it. */
+	dev_err(dev, "task 0x%p timeout or killed\n", task);
+	task->task_state = TASK_STATE_ERROR;
+
+	if (prev_task) {
+		u64 *prev_va = prev_task->va_base;
+		u64 *curr_va = task->va_base;
+
+		/* copy JUMP instruction */
+		prev_va[prev_task->num_cmd - 1] = curr_va[task->num_cmd - 1];
+
+		cmdq_thread_invalidate_fetched_data(thread);
+	} else if (next_task) { /* move to next task */
+		cmdq_thread_writel(thread, next_task->pa_base,
+				   CMDQ_THR_CURR_ADDR);
+	}
+
+	list_del(&task->list_entry);
+	cmdq_thread_resume(thread);
+
+	/* call cb here to prevent lock */
+	if (task->cb.cb) {
+		struct cmdq_cb_data cmdq_cb_data;
+
+		cmdq_cb_data.err = true;
+		cmdq_cb_data.data = task->cb.data;
+		task->cb.cb(cmdq_cb_data);
+	}
+
+	return -ECANCELED;
+}
+
+static void cmdq_task_wait_release_work(struct work_struct *work_item)
+{
+	struct cmdq_task *task = container_of(work_item, struct cmdq_task,
+					      release_work);
+	struct cmdq *cmdq = task->cmdq;
+	struct cmdq_thread *thread = task->thread;
+	unsigned long flags;
+
+	wait_event_timeout(thread->wait_task_done,
+			   task->task_state != TASK_STATE_BUSY,
+			   msecs_to_jiffies(CMDQ_TIMEOUT_MS));
+
+	spin_lock_irqsave(&cmdq->exec_lock, flags);
+	if (task->task_state != TASK_STATE_DONE)
+		cmdq_task_handle_error_result(task);
+	if (list_empty(&thread->task_busy_list))
+		cmdq_thread_disable(cmdq, thread);
+	spin_unlock_irqrestore(&cmdq->exec_lock, flags);
+
+	/* release regardless of success or not */
+	clk_disable_unprepare(cmdq->clock);
+	cmdq_task_release(task);
+}
+
+static void cmdq_task_wait_release_schedule(struct cmdq_task *task)
+{
+	struct cmdq *cmdq = task->cmdq;
+
+	INIT_WORK(&task->release_work, cmdq_task_wait_release_work);
+	queue_work(cmdq->task_release_wq, &task->release_work);
+}
+
+static int cmdq_rec_realloc_cmd_buffer(struct cmdq_rec *rec, size_t size)
+{
+	void *new_buf;
+
+	new_buf = krealloc(rec->buf, size, GFP_KERNEL | __GFP_ZERO);
+	if (!new_buf)
+		return -ENOMEM;
+	rec->buf = new_buf;
+	rec->buf_size = size;
+	return 0;
+}
+
+struct cmdq_base *cmdq_register_device(struct device *dev)
+{
+	struct cmdq_base *cmdq_base;
+	struct resource res;
+	int subsys;
+	u32 base;
+
+	if (of_address_to_resource(dev->of_node, 0, &res))
+		return NULL;
+	base = (u32)res.start;
+
+	subsys = cmdq_subsys_base_to_id(base >> 16);
+	if (subsys < 0)
+		return NULL;
+
+	cmdq_base = devm_kmalloc(dev, sizeof(*cmdq_base), GFP_KERNEL);
+	if (!cmdq_base)
+		return NULL;
+	cmdq_base->subsys = subsys;
+	cmdq_base->base = base;
+
+	return cmdq_base;
+}
+EXPORT_SYMBOL(cmdq_register_device);
+
+int cmdq_rec_create(struct device *dev, u64 engine_flag,
+		    struct cmdq_rec **rec_ptr)
+{
+	struct cmdq_rec *rec;
+	int err;
+
+	rec = kzalloc(sizeof(*rec), GFP_KERNEL);
+	if (!rec)
+		return -ENOMEM;
+	rec->cmdq = dev_get_drvdata(dev);
+	rec->engine_flag = engine_flag;
+	err = cmdq_rec_realloc_cmd_buffer(rec, CMDQ_INITIAL_CMD_BLOCK_SIZE);
+	if (err < 0) {
+		kfree(rec);
+		return err;
+	}
+	*rec_ptr = rec;
+	return 0;
+}
+EXPORT_SYMBOL(cmdq_rec_create);
+
+static int cmdq_rec_append_command(struct cmdq_rec *rec, enum cmdq_code code,
+				   u32 arg_a, u32 arg_b)
+{
+	u64 *cmd_ptr;
+	int err;
+
+	if (WARN_ON(rec->finalized))
+		return -EBUSY;
+	if (code < CMDQ_CODE_MOVE || code > CMDQ_CODE_EOC)
+		return -EINVAL;
+	if (unlikely(rec->command_size + CMDQ_INST_SIZE > rec->buf_size)) {
+		err = cmdq_rec_realloc_cmd_buffer(rec, rec->buf_size * 2);
+		if (err < 0)
+			return err;
+	}
+	cmd_ptr = rec->buf + rec->command_size;
+	(*cmd_ptr) = (u64)((code << CMDQ_OP_CODE_SHIFT) | arg_a) << 32 | arg_b;
+	rec->command_size += CMDQ_INST_SIZE;
+	return 0;
+}
+
+int cmdq_rec_write(struct cmdq_rec *rec, u32 value, struct cmdq_base *base,
+		   u32 offset)
+{
+	u32 arg_a = ((base->base + offset) & CMDQ_ARG_A_WRITE_MASK) |
+		    ((base->subsys & CMDQ_SUBSYS_MASK) << CMDQ_SUBSYS_SHIFT);
+	return cmdq_rec_append_command(rec, CMDQ_CODE_WRITE, arg_a, value);
+}
+EXPORT_SYMBOL(cmdq_rec_write);
+
+int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
+			struct cmdq_base *base, u32 offset, u32 mask)
+{
+	u32 offset_mask = offset;
+	int err;
+
+	if (mask != 0xffffffff) {
+		err = cmdq_rec_append_command(rec, CMDQ_CODE_MOVE, 0, ~mask);
+		if (err < 0)
+			return err;
+		offset_mask |= CMDQ_WRITE_ENABLE_MASK;
+	}
+	return cmdq_rec_write(rec, value, base, offset_mask);
+}
+EXPORT_SYMBOL(cmdq_rec_write_mask);
+
+int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event)
+{
+	u32 arg_b;
+
+	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
+		return -EINVAL;
+
+	/*
+	 * bit 0-11: wait value
+	 * bit 15: 1 - wait, 0 - no wait
+	 * bit 16-27: update value
+	 * bit 31: 1 - update, 0 - no update
+	 */
+	arg_b = CMDQ_WFE_UPDATE | CMDQ_WFE_WAIT | CMDQ_WFE_WAIT_VALUE;
+	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event, arg_b);
+}
+EXPORT_SYMBOL(cmdq_rec_wfe);
+
+int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event)
+{
+	if (event >= CMDQ_MAX_HW_EVENT_COUNT || event < 0)
+		return -EINVAL;
+
+	return cmdq_rec_append_command(rec, CMDQ_CODE_WFE, event,
+				       CMDQ_WFE_UPDATE);
+}
+EXPORT_SYMBOL(cmdq_rec_clear_event);
+
+static int cmdq_rec_finalize(struct cmdq_rec *rec)
+{
+	int err;
+
+	if (rec->finalized)
+		return 0;
+
+	/* insert EOC and generate IRQ for each command iteration */
+	err = cmdq_rec_append_command(rec, CMDQ_CODE_EOC, 0, CMDQ_EOC_IRQ_EN);
+	if (err < 0)
+		return err;
+
+	/* JUMP to end */
+	err = cmdq_rec_append_command(rec, CMDQ_CODE_JUMP, 0, CMDQ_JUMP_PASS);
+	if (err < 0)
+		return err;
+
+	rec->finalized = true;
+	return 0;
+}
+
+int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
+			 void *data)
+{
+	struct cmdq *cmdq = rec->cmdq;
+	struct cmdq_task *task;
+	struct cmdq_task_cb task_cb;
+	struct cmdq_thread *thread;
+	int err;
+
+	mutex_lock(&cmdq->task_mutex);
+	if (cmdq->suspended) {
+		dev_err(cmdq->dev, "%s is called after suspended\n", __func__);
+		mutex_unlock(&cmdq->task_mutex);
+		return -EPERM;
+	}
+
+	err = cmdq_rec_finalize(rec);
+	if (err < 0) {
+		mutex_unlock(&cmdq->task_mutex);
+		return err;
+	}
+
+	task_cb.cb = cb;
+	task_cb.data = data;
+	task = cmdq_task_acquire(rec, task_cb);
+	if (!task) {
+		mutex_unlock(&cmdq->task_mutex);
+		return -EFAULT;
+	}
+
+	thread = &cmdq->thread[cmdq_eng_get_thread(task->engine_flag)];
+	cmdq_task_exec(task, thread);
+	cmdq_task_wait_release_schedule(task);
+	mutex_unlock(&cmdq->task_mutex);
+	return 0;
+}
+EXPORT_SYMBOL(cmdq_rec_flush_async);
+
+struct cmdq_flush_completion {
+	struct completion cmplt;
+	bool err;
+};
+
+static int cmdq_rec_flush_cb(struct cmdq_cb_data data)
+{
+	struct cmdq_flush_completion *cmplt = data.data;
+
+	cmplt->err = data.err;
+	complete(&cmplt->cmplt);
+	return 0;
+}
+
+int cmdq_rec_flush(struct cmdq_rec *rec)
+{
+	struct cmdq_flush_completion cmplt;
+	int err;
+
+	init_completion(&cmplt.cmplt);
+	err = cmdq_rec_flush_async(rec, cmdq_rec_flush_cb, &cmplt);
+	if (err < 0)
+		return err;
+	wait_for_completion(&cmplt.cmplt);
+	return cmplt.err ? -EFAULT : 0;
+}
+EXPORT_SYMBOL(cmdq_rec_flush);
+
+void cmdq_rec_destroy(struct cmdq_rec *rec)
+{
+	kfree(rec->buf);
+	kfree(rec);
+}
+EXPORT_SYMBOL(cmdq_rec_destroy);
+
+static bool cmdq_task_is_empty(struct cmdq *cmdq)
+{
+	struct cmdq_thread *thread;
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
+		thread = &cmdq->thread[i];
+		if (!list_empty(&thread->task_busy_list))
+			return false;
+	}
+	return true;
+}
+
+static int cmdq_suspend(struct device *dev)
+{
+	struct cmdq *cmdq = dev_get_drvdata(dev);
+	u32 exec_threads;
+
+	mutex_lock(&cmdq->task_mutex);
+	cmdq->suspended = true;
+	mutex_unlock(&cmdq->task_mutex);
+
+	exec_threads = readl(cmdq->base + CMDQ_CURR_LOADED_THR);
+	if ((exec_threads & CMDQ_THR_EXECUTING) && !cmdq_task_is_empty(cmdq)) {
+		dev_err(dev, "wait active tasks timeout.\n");
+		flush_workqueue(cmdq->task_release_wq);
+	}
+	return 0;
+}
+
+static int cmdq_resume(struct device *dev)
+{
+	struct cmdq *cmdq = dev_get_drvdata(dev);
+
+	cmdq->suspended = false;
+	return 0;
+}
+
+static int cmdq_remove(struct platform_device *pdev)
+{
+	struct cmdq *cmdq = platform_get_drvdata(pdev);
+
+	destroy_workqueue(cmdq->task_release_wq);
+	cmdq->task_release_wq = NULL;
+	return 0;
+}
+
+static int cmdq_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct device_node *node = dev->of_node;
+	struct resource *res;
+	struct cmdq *cmdq;
+	int err, i;
+
+	cmdq = devm_kzalloc(dev, sizeof(*cmdq), GFP_KERNEL);
+	if (!cmdq)
+		return -ENOMEM;
+	cmdq->dev = dev;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	cmdq->base = devm_ioremap_resource(dev, res);
+	if (IS_ERR(cmdq->base)) {
+		dev_err(dev, "failed to ioremap gce\n");
+		return PTR_ERR(cmdq->base);
+	}
+
+	cmdq->irq = irq_of_parse_and_map(node, 0);
+	if (!cmdq->irq) {
+		dev_err(dev, "failed to get irq\n");
+		return -EINVAL;
+	}
+
+	dev_dbg(dev, "cmdq device: addr:0x%p, va:0x%p, irq:%d\n",
+		dev, cmdq->base, cmdq->irq);
+
+	mutex_init(&cmdq->task_mutex);
+	spin_lock_init(&cmdq->exec_lock);
+	cmdq->task_release_wq = alloc_ordered_workqueue(
+			"%s", WQ_MEM_RECLAIM | WQ_HIGHPRI,
+			"cmdq_task_wait_release");
+
+	for (i = 0; i < ARRAY_SIZE(cmdq->thread); i++) {
+		cmdq->thread[i].base = cmdq->base + CMDQ_THR_BASE +
+				CMDQ_THR_SHIFT * i;
+		init_waitqueue_head(&cmdq->thread[i].wait_task_done);
+		INIT_LIST_HEAD(&cmdq->thread[i].task_busy_list);
+	}
+
+	platform_set_drvdata(pdev, cmdq);
+
+	err = devm_request_irq(dev, cmdq->irq, cmdq_irq_handler, IRQF_SHARED,
+			       CMDQ_DRIVER_DEVICE_NAME, cmdq);
+	if (err < 0) {
+		dev_err(dev, "failed to register ISR (%d)\n", err);
+		goto fail;
+	}
+
+	cmdq->clock = devm_clk_get(dev, CMDQ_CLK_NAME);
+	if (IS_ERR(cmdq->clock)) {
+		dev_err(dev, "failed to get clk:%s\n", CMDQ_CLK_NAME);
+		err = PTR_ERR(cmdq->clock);
+		goto fail;
+	}
+	return 0;
+
+fail:
+	cmdq_remove(pdev);
+	return err;
+}
+
+static const struct dev_pm_ops cmdq_pm_ops = {
+	.suspend = cmdq_suspend,
+	.resume = cmdq_resume,
+};
+
+static const struct of_device_id cmdq_of_ids[] = {
+	{.compatible = "mediatek,mt8173-gce",},
+	{}
+};
+
+static struct platform_driver cmdq_drv = {
+	.probe = cmdq_probe,
+	.remove = cmdq_remove,
+	.driver = {
+		.name = CMDQ_DRIVER_DEVICE_NAME,
+		.owner = THIS_MODULE,
+		.pm = &cmdq_pm_ops,
+		.of_match_table = cmdq_of_ids,
+	}
+};
+
+builtin_platform_driver(cmdq_drv);
diff --git a/include/soc/mediatek/cmdq.h b/include/soc/mediatek/cmdq.h
new file mode 100644
index 0000000..60eef3d
--- /dev/null
+++ b/include/soc/mediatek/cmdq.h
@@ -0,0 +1,197 @@ 
+/*
+ * Copyright (c) 2015 MediaTek Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#ifndef __MTK_CMDQ_H__
+#define __MTK_CMDQ_H__
+
+#include <linux/platform_device.h>
+#include <linux/types.h>
+
+enum cmdq_eng {
+	CMDQ_ENG_DISP_AAL,
+	CMDQ_ENG_DISP_COLOR0,
+	CMDQ_ENG_DISP_COLOR1,
+	CMDQ_ENG_DISP_DPI0,
+	CMDQ_ENG_DISP_DSI0,
+	CMDQ_ENG_DISP_DSI1,
+	CMDQ_ENG_DISP_GAMMA,
+	CMDQ_ENG_DISP_OD,
+	CMDQ_ENG_DISP_OVL0,
+	CMDQ_ENG_DISP_OVL1,
+	CMDQ_ENG_DISP_PWM0,
+	CMDQ_ENG_DISP_PWM1,
+	CMDQ_ENG_DISP_RDMA0,
+	CMDQ_ENG_DISP_RDMA1,
+	CMDQ_ENG_DISP_RDMA2,
+	CMDQ_ENG_DISP_UFOE,
+	CMDQ_ENG_DISP_WDMA0,
+	CMDQ_ENG_DISP_WDMA1,
+	CMDQ_ENG_MAX,
+};
+
+/* events for CMDQ and display */
+enum cmdq_event {
+	/* Display start of frame(SOF) events */
+	CMDQ_EVENT_DISP_OVL0_SOF = 11,
+	CMDQ_EVENT_DISP_OVL1_SOF = 12,
+	CMDQ_EVENT_DISP_RDMA0_SOF = 13,
+	CMDQ_EVENT_DISP_RDMA1_SOF = 14,
+	CMDQ_EVENT_DISP_RDMA2_SOF = 15,
+	CMDQ_EVENT_DISP_WDMA0_SOF = 16,
+	CMDQ_EVENT_DISP_WDMA1_SOF = 17,
+	/* Display end of frame(EOF) events */
+	CMDQ_EVENT_DISP_OVL0_EOF = 39,
+	CMDQ_EVENT_DISP_OVL1_EOF = 40,
+	CMDQ_EVENT_DISP_RDMA0_EOF = 41,
+	CMDQ_EVENT_DISP_RDMA1_EOF = 42,
+	CMDQ_EVENT_DISP_RDMA2_EOF = 43,
+	CMDQ_EVENT_DISP_WDMA0_EOF = 44,
+	CMDQ_EVENT_DISP_WDMA1_EOF = 45,
+	/* Mutex end of frame(EOF) events */
+	CMDQ_EVENT_MUTEX0_STREAM_EOF = 53,
+	CMDQ_EVENT_MUTEX1_STREAM_EOF = 54,
+	CMDQ_EVENT_MUTEX2_STREAM_EOF = 55,
+	CMDQ_EVENT_MUTEX3_STREAM_EOF = 56,
+	CMDQ_EVENT_MUTEX4_STREAM_EOF = 57,
+	/* Display underrun events */
+	CMDQ_EVENT_DISP_RDMA0_UNDERRUN = 63,
+	CMDQ_EVENT_DISP_RDMA1_UNDERRUN = 64,
+	CMDQ_EVENT_DISP_RDMA2_UNDERRUN = 65,
+	/* Keep this at the end of HW events */
+	CMDQ_MAX_HW_EVENT_COUNT = 260,
+};
+
+struct cmdq_cb_data {
+	bool	err;
+	void	*data;
+};
+
+typedef int (*cmdq_async_flush_cb)(struct cmdq_cb_data data);
+
+struct cmdq_task;
+struct cmdq;
+
+struct cmdq_rec {
+	struct cmdq		*cmdq;
+	u64			engine_flag;
+	size_t			command_size;
+	void			*buf;
+	size_t			buf_size;
+	bool			finalized;
+};
+
+struct cmdq_base {
+	int	subsys;
+	u32	base;
+};
+
+/**
+ * cmdq_register_device() - register device which needs CMDQ
+ * @dev:		device
+ *
+ * Return: cmdq_base pointer or NULL for failed
+ */
+struct cmdq_base *cmdq_register_device(struct device *dev);
+
+/**
+ * cmdq_rec_create() - create command queue record
+ * @dev:		device
+ * @engine_flag:	command queue engine flag
+ * @rec_ptr:		command queue record pointer to retrieve cmdq_rec
+ *
+ * Return: 0 for success; else the error code is returned
+ */
+int cmdq_rec_create(struct device *dev, u64 engine_flag,
+		    struct cmdq_rec **rec_ptr);
+
+/**
+ * cmdq_rec_write() - append write command to the command queue record
+ * @rec:	the command queue record
+ * @value:	the specified target register value
+ * @base:	the command queue base
+ * @offset:	register offset from module base
+ *
+ * Return: 0 for success; else the error code is returned
+ */
+int cmdq_rec_write(struct cmdq_rec *rec, u32 value,
+		   struct cmdq_base *base, u32 offset);
+
+/**
+ * cmdq_rec_write_mask() - append write command with mask to the command
+ *			   queue record
+ * @rec:	the command queue record
+ * @value:	the specified target register value
+ * @base:	the command queue base
+ * @offset:	register offset from module base
+ * @mask:	the specified target register mask
+ *
+ * Return: 0 for success; else the error code is returned
+ */
+int cmdq_rec_write_mask(struct cmdq_rec *rec, u32 value,
+			struct cmdq_base *base, u32 offset, u32 mask);
+
+/**
+ * cmdq_rec_wfe() - append wait for event command to the command queue record
+ * @rec:	the command queue record
+ * @event:	the desired event type to "wait and CLEAR"
+ *
+ * Return: 0 for success; else the error code is returned
+ */
+int cmdq_rec_wfe(struct cmdq_rec *rec, enum cmdq_event event);
+
+/**
+ * cmdq_rec_clear_event() - append clear event command to the command queue
+ *			    record
+ * @rec:	the command queue record
+ * @event:	the desired event to be cleared
+ *
+ * Return: 0 for success; else the error code is returned
+ */
+int cmdq_rec_clear_event(struct cmdq_rec *rec, enum cmdq_event event);
+
+/**
+ * cmdq_rec_flush() - trigger CMDQ to execute the recorded commands
+ * @rec:	the command queue record
+ *
+ * Return: 0 for success; else the error code is returned
+ *
+ * Trigger CMDQ to execute the recorded commands. Note that this is a
+ * synchronous flush function. When the function returned, the recorded
+ * commands have been done.
+ */
+int cmdq_rec_flush(struct cmdq_rec *rec);
+
+/**
+ * cmdq_rec_flush_async() - trigger CMDQ to asynchronously execute the recorded
+ *			    commands and call back after ISR is finished
+ * @rec:	the command queue record
+ * @cb:		called in the end of CMDQ ISR
+ * @data:	this data will pass back to cb
+ *
+ * Return: 0 for success; else the error code is returned
+ *
+ * Trigger CMDQ to asynchronously execute the recorded commands and call back
+ * after ISR is finished. Note that this is an ASYNC function. When the function
+ * returned, it may or may not be finished. The ISR callback function is called
+ * in the end of ISR.
+ */
+int cmdq_rec_flush_async(struct cmdq_rec *rec, cmdq_async_flush_cb cb,
+			 void *data);
+
+/**
+ * cmdq_rec_destroy() - destroy command queue record
+ * @rec:	the command queue record
+ */
+void cmdq_rec_destroy(struct cmdq_rec *rec);
+
+#endif	/* __MTK_CMDQ_H__ */