diff mbox series

[v2,6/7] fpga: add PolarFire SoC Auto Update support

Message ID 20230331071823.956087-7-conor.dooley@microchip.com (mailing list archive)
State Handled Elsewhere
Headers show
Series PolarFire SoC Auto Update Support | expand

Checks

Context Check Description
conchuod/tree_selection fail Failed to apply to next/pending-fixes or riscv/for-next

Commit Message

Conor Dooley March 31, 2023, 7:18 a.m. UTC 
Add support for Auto Update reprogramming of the FPGA fabric on
PolarFire SoC, using the fw_upload mechanism a la the
intel-m10-bmc-sec-update driver.

This driver only writes the image to the spi flash & performs
validation on it, as the entire FPGA becomes unusable during the actual
reprogramming of a bitstream. To initiate the reprogramming itself, a
device reset is required.
The SBI SRST extension's "cold reboot" can trigger such a device reset,
by means of a "Tamper Macro" in the FPGA fabric, provided corresponding
support has been enabled in the HSS (Hart Software Services).

Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
---
 drivers/fpga/Kconfig                 |  11 +
 drivers/fpga/Makefile                |   3 +-
 drivers/fpga/microchip-auto-update.c | 494 +++++++++++++++++++++++++++
 3 files changed, 507 insertions(+), 1 deletion(-)
 create mode 100644 drivers/fpga/microchip-auto-update.c

Comments

Xu Yilun April 2, 2023, 10:57 a.m. UTC | #1 
On 2023-03-31 at 08:18:22 +0100, Conor Dooley wrote:
> Add support for Auto Update reprogramming of the FPGA fabric on
> PolarFire SoC, using the fw_upload mechanism a la the
> intel-m10-bmc-sec-update driver.

I'm not an expert of MTD and have several concerns about the mtd
writing part:

- This seems a generic case that writing a file content into the
  driver restricted offset on mtd.

- Seems the firmware update interface could be created for mtd device,
  like the nvmem interface?

- I assume the purpose of the fw upload interface is to protect
  unexpected writing from user. So how to protect the mtd be
  written by other user interfaces like mtdchar/block or nvmem?


For other parts:

- directory block check, fw upload don't touch this block, so why check
  it in fw upload process.

- image verification, no matter pass or fail, fw upload is actually done,
  the flash is written and no way to rollback. So may remove from fw
  upload process.

Thanks,
Yilun

> 
> This driver only writes the image to the spi flash & performs
> validation on it, as the entire FPGA becomes unusable during the actual
> reprogramming of a bitstream. To initiate the reprogramming itself, a
> device reset is required.
> The SBI SRST extension's "cold reboot" can trigger such a device reset,
> by means of a "Tamper Macro" in the FPGA fabric, provided corresponding
> support has been enabled in the HSS (Hart Software Services).
> 
> Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
> ---
>  drivers/fpga/Kconfig                 |  11 +
>  drivers/fpga/Makefile                |   3 +-
>  drivers/fpga/microchip-auto-update.c | 494 +++++++++++++++++++++++++++
>  3 files changed, 507 insertions(+), 1 deletion(-)
>  create mode 100644 drivers/fpga/microchip-auto-update.c
> 
> diff --git a/drivers/fpga/Kconfig b/drivers/fpga/Kconfig
> index 0a00763b9f28..2838ce7ddafe 100644
> --- a/drivers/fpga/Kconfig
> +++ b/drivers/fpga/Kconfig
> @@ -257,6 +257,17 @@ config FPGA_M10_BMC_SEC_UPDATE
>  	  (BMC) and provides support for secure updates for the BMC image,
>  	  the FPGA image, the Root Entry Hashes, etc.
>  
> +config FPGA_MICROCHIP_AUTO_UPDATE
> +	tristate "Microchip PolarFire SoC AUTO UPDATE"
> +	depends on POLARFIRE_SOC_SYS_CTRL
> +	select FW_LOADER
> +	select FW_UPLOAD
> +	help
> +	  FPGA manager driver support for reprogramming PolarFire SoC from
> +	  within Linux, using the Auto Upgrade feature of the system controller.
> +
> +	  If built as a module, it will be called microchip-auto-update.
> +
>  config FPGA_MGR_MICROCHIP_SPI
>  	tristate "Microchip Polarfire SPI FPGA manager"
>  	depends on SPI
> diff --git a/drivers/fpga/Makefile b/drivers/fpga/Makefile
> index 72e554b4d2f7..0b1ba6db860d 100644
> --- a/drivers/fpga/Makefile
> +++ b/drivers/fpga/Makefile
> @@ -25,8 +25,9 @@ obj-$(CONFIG_FPGA_MGR_LATTICE_SYSCONFIG_SPI)	+= lattice-sysconfig-spi.o
>  obj-$(CONFIG_ALTERA_PR_IP_CORE)		+= altera-pr-ip-core.o
>  obj-$(CONFIG_ALTERA_PR_IP_CORE_PLAT)	+= altera-pr-ip-core-plat.o
>  
> -# FPGA Secure Update Drivers
> +# FPGA Firmware Upload Drivers
>  obj-$(CONFIG_FPGA_M10_BMC_SEC_UPDATE)	+= intel-m10-bmc-sec-update.o
> +obj-$(CONFIG_FPGA_MICROCHIP_AUTO_UPDATE)	+= microchip-auto-update.o
>  
>  # FPGA Bridge Drivers
>  obj-$(CONFIG_FPGA_BRIDGE)		+= fpga-bridge.o
> diff --git a/drivers/fpga/microchip-auto-update.c b/drivers/fpga/microchip-auto-update.c
> new file mode 100644
> index 000000000000..5fa920b659c6
> --- /dev/null
> +++ b/drivers/fpga/microchip-auto-update.c
> @@ -0,0 +1,494 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * Microchip Polarfire SoC "Auto Update" FPGA reprogramming.
> + *
> + * Documentation of this functionality is available in the "PolarFire® FPGA and
> + * PolarFire SoC FPGA Programming" User Guide.
> + *
> + * Copyright (c) 2022-2023 Microchip Corporation. All rights reserved.
> + *
> + * Author: Conor Dooley <conor.dooley@microchip.com>
> + */
> +#include <linux/debugfs.h>
> +#include <linux/firmware.h>
> +#include <linux/math.h>
> +#include <linux/module.h>
> +#include <linux/mtd/mtd.h>
> +#include <linux/of_device.h>
> +#include <linux/sizes.h>
> +
> +#include <soc/microchip/mpfs.h>
> +
> +#define AUTO_UPDATE_DEFAULT_MBOX_OFFSET		0u
> +#define AUTO_UPDATE_DEFAULT_RESP_OFFSET		0u
> +
> +#define AUTO_UPDATE_FEATURE_CMD_OPCODE		0x05u
> +#define AUTO_UPDATE_FEATURE_CMD_DATA_SIZE	0u
> +#define AUTO_UPDATE_FEATURE_RESP_SIZE		33u
> +#define AUTO_UPDATE_FEATURE_CMD_DATA		NULL
> +#define AUTO_UPDATE_FEATURE_ENABLED		BIT(5)
> +
> +#define AUTO_UPDATE_AUTHENTICATE_CMD_OPCODE	0x22u
> +#define AUTO_UPDATE_AUTHENTICATE_CMD_DATA_SIZE	0u
> +#define AUTO_UPDATE_AUTHENTICATE_RESP_SIZE	1u
> +#define AUTO_UPDATE_AUTHENTICATE_CMD_DATA	NULL
> +
> +#define AUTO_UPDATE_PROGRAM_CMD_OPCODE		0x46u
> +#define AUTO_UPDATE_PROGRAM_CMD_DATA_SIZE	0u
> +#define AUTO_UPDATE_PROGRAM_RESP_SIZE		1u
> +#define AUTO_UPDATE_PROGRAM_CMD_DATA		NULL
> +
> +/*
> + * SPI Flash layout example:
> + * |------------------------------| 0x0000000
> + * | 1 KiB                        |
> + * | SPI "directories"            |
> + * |------------------------------| 0x0000400
> + * | 1 MiB                        |
> + * | Reserved area                |
> + * | Used for bitstream info      |
> + * |------------------------------| 0x0100400
> + * | 20 MiB                       |
> + * | Golden Image                 |
> + * |------------------------------| 0x1500400
> + * | 20 MiB                       |
> + * | Auto Upgrade Image           |
> + * |------------------------------| 0x2900400
> + * | 20 MiB                       |
> + * | Reserved for multi-image IAP |
> + * | Unused for Auto Upgrade      |
> + * |------------------------------| 0x3D00400
> + * | ? B                          |
> + * | Unused                       |
> + * |------------------------------| 0x?
> + */
> +#define AUTO_UPDATE_DIRECTORY_BASE	0u
> +#define AUTO_UPDATE_DIRECTORY_WIDTH	4u
> +#define AUTO_UPDATE_GOLDEN_INDEX	0u
> +#define AUTO_UPDATE_UPGRADE_INDEX	1u
> +#define AUTO_UPDATE_BLANK_INDEX		2u
> +#define AUTO_UPDATE_GOLDEN_DIRECTORY	(AUTO_UPDATE_DIRECTORY_WIDTH * AUTO_UPDATE_GOLDEN_INDEX)
> +#define AUTO_UPDATE_UPGRADE_DIRECTORY	(AUTO_UPDATE_DIRECTORY_WIDTH * AUTO_UPDATE_UPGRADE_INDEX)
> +#define AUTO_UPDATE_BLANK_DIRECTORY	(AUTO_UPDATE_DIRECTORY_WIDTH * AUTO_UPDATE_BLANK_INDEX)
> +#define AUTO_UPDATE_DIRECTORY_SIZE	SZ_1K
> +#define AUTO_UPDATE_RESERVED_SIZE	SZ_1M
> +#define AUTO_UPDATE_BITSTREAM_BASE	(AUTO_UPDATE_DIRECTORY_SIZE + AUTO_UPDATE_RESERVED_SIZE)
> +
> +#define AUTO_UPDATE_TIMEOUT_MS		60000
> +
> +struct mpfs_auto_update_priv {
> +	struct mpfs_sys_controller *sys_controller;
> +	struct device *dev;
> +	struct mtd_info *flash;
> +	struct fw_upload *fw_uploader;
> +	struct completion programming_complete;
> +	size_t size_per_bitstream;
> +	bool cancel_request;
> +};
> +
> +static enum fw_upload_err mpfs_auto_update_prepare(struct fw_upload *fw_uploader, const u8 *data,
> +						   u32 size)
> +{
> +	struct mpfs_auto_update_priv *priv = fw_uploader->dd_handle;
> +	size_t erase_size = AUTO_UPDATE_DIRECTORY_SIZE;
> +
> +	/*
> +	 * Verifying the Golden Image is idealistic. It will be evaluated
> +	 * against the currently programmed image and thus may fail - due to
> +	 * either rollback protection (if its an older version than that in use)
> +	 * or if the version is the same as that of the in-use image.
> +	 * Extracting the information as to why a failure occurred is not
> +	 * currently possible due to limitations of the system controller
> +	 * driver. If those are fixed, verification of the Golden Image should
> +	 * be added here.
> +	 */
> +
> +	priv->flash = mpfs_sys_controller_get_flash(priv->sys_controller);
> +	if (!priv->flash)
> +		return FW_UPLOAD_ERR_HW_ERROR;
> +
> +	erase_size = round_up(erase_size, (u64)priv->flash->erasesize);
> +
> +	/*
> +	 * We need to calculate if we have enough space in the flash for the
> +	 * new image.
> +	 * First, chop off the first 1 KiB as it's reserved for the directory.
> +	 * The 1 MiB reserved for design info needs to be ignored also.
> +	 * All that remains is carved into 3 & rounded down to the erasesize.
> +	 * If this is smaller than the image size, we abort.
> +	 * There's also no need to consume more than 20 MiB per image.
> +	 */
> +	priv->size_per_bitstream = priv->flash->size - SZ_1K - SZ_1M;
> +	priv->size_per_bitstream = round_down(priv->size_per_bitstream / 3, erase_size);
> +	if (priv->size_per_bitstream > 20 * SZ_1M)
> +		priv->size_per_bitstream = 20 * SZ_1M;
> +
> +	if (priv->size_per_bitstream < size) {
> +		dev_err(priv->dev,
> +			"flash device has insufficient capacity to store this bitstream\n");
> +		return FW_UPLOAD_ERR_INVALID_SIZE;
> +	}
> +
> +	priv->cancel_request = false;
> +
> +	return FW_UPLOAD_ERR_NONE;
> +}
> +
> +static void mpfs_auto_update_cancel(struct fw_upload *fw_uploader)
> +{
> +	struct mpfs_auto_update_priv *priv = fw_uploader->dd_handle;
> +
> +	priv->cancel_request = true;
> +}
> +
> +static enum fw_upload_err mpfs_auto_update_poll_complete(struct fw_upload *fw_uploader)
> +{
> +	struct mpfs_auto_update_priv *priv = fw_uploader->dd_handle;
> +	int ret;
> +
> +	/*
> +	 * There is no meaningful way to get the status of the programming while
> +	 * it is in progress, so attempting anything other than waiting for it
> +	 * to complete would be misplaced.
> +	 */
> +	ret = wait_for_completion_timeout(&priv->programming_complete,
> +					  msecs_to_jiffies(AUTO_UPDATE_TIMEOUT_MS));
> +	if (ret)
> +		return FW_UPLOAD_ERR_TIMEOUT;
> +
> +	return FW_UPLOAD_ERR_NONE;
> +}
> +
> +static int mpfs_auto_update_verify_image(struct fw_upload *fw_uploader)
> +{
> +	struct mpfs_auto_update_priv *priv = fw_uploader->dd_handle;
> +	struct mpfs_mss_response *response;
> +	struct mpfs_mss_msg *message;
> +	u32 *response_msg;
> +	int ret;
> +
> +	response_msg = devm_kzalloc(priv->dev, AUTO_UPDATE_FEATURE_RESP_SIZE * sizeof(response_msg),
> +				    GFP_KERNEL);
> +	if (!response_msg)
> +		return -ENOMEM;
> +
> +	response = devm_kzalloc(priv->dev, sizeof(struct mpfs_mss_response), GFP_KERNEL);
> +	if (!response) {
> +		ret = -ENOMEM;
> +		goto free_response_msg;
> +	}
> +
> +	message = devm_kzalloc(priv->dev, sizeof(struct mpfs_mss_msg), GFP_KERNEL);
> +	if (!message) {
> +		ret = -ENOMEM;
> +		goto free_response;
> +	}
> +
> +	/*
> +	 * The system controller can verify that an image in the flash is valid.
> +	 * Rather than duplicate the check in this driver, call the relevant
> +	 * service from the system controller instead.
> +	 * This service has no command data and no response data. It overloads
> +	 * mbox_offset with the image index in the flash's SPI directory where
> +	 * the bitstream is located.
> +	 */
> +	response->resp_msg = response_msg;
> +	response->resp_size = AUTO_UPDATE_AUTHENTICATE_RESP_SIZE;
> +	message->cmd_opcode = AUTO_UPDATE_AUTHENTICATE_CMD_OPCODE;
> +	message->cmd_data_size = AUTO_UPDATE_AUTHENTICATE_CMD_DATA_SIZE;
> +	message->response = response;
> +	message->cmd_data = AUTO_UPDATE_AUTHENTICATE_CMD_DATA;
> +	message->mbox_offset = AUTO_UPDATE_UPGRADE_INDEX;
> +	message->resp_offset = AUTO_UPDATE_DEFAULT_RESP_OFFSET;
> +
> +	dev_info(priv->dev, "Running verification of Upgrade Image\n");
> +	ret = mpfs_blocking_transaction(priv->sys_controller, message);
> +	if (ret | response->resp_status) {
> +		dev_warn(priv->dev, "Verification of Upgrade Image failed!\n");
> +		ret = ret ? ret : -EBADMSG;
> +	}
> +
> +	dev_info(priv->dev, "Verification of Upgrade Image passed!\n");
> +
> +	devm_kfree(priv->dev, message);
> +free_response:
> +	devm_kfree(priv->dev, response);
> +free_response_msg:
> +	devm_kfree(priv->dev, response_msg);
> +
> +	return ret;
> +}
> +
> +static int mpfs_auto_update_set_image_address(struct mpfs_auto_update_priv *priv, char *buffer,
> +					      u32 image_address, loff_t directory_address)
> +{
> +	struct erase_info erase;
> +	size_t erase_size = AUTO_UPDATE_DIRECTORY_SIZE;
> +	size_t bytes_written = 0, bytes_read = 0;
> +	int ret;
> +
> +	erase_size = round_up(erase_size, (u64)priv->flash->erasesize);
> +
> +	erase.addr = AUTO_UPDATE_DIRECTORY_BASE;
> +	erase.len = erase_size;
> +
> +	/*
> +	 * We need to write the "SPI DIRECTORY" to the first 1 KiB, telling
> +	 * the system controller where to find the actual bitstream. Since
> +	 * this is spi-nor, we have to read the first eraseblock, erase that
> +	 * portion of the flash, modify the data and then write it back.
> +	 * There's no need to do this though if things are already the way they
> +	 * should be, so check and save the write in that case.
> +	 */
> +	ret = mtd_read(priv->flash, AUTO_UPDATE_DIRECTORY_BASE, erase_size, &bytes_read,
> +		       (u_char *)buffer);
> +	if (ret)
> +		return ret;
> +
> +	if (bytes_read != erase_size)
> +		return -EIO;
> +
> +	if ((*(u32 *)(buffer + AUTO_UPDATE_UPGRADE_DIRECTORY) == image_address) &&
> +	    !(*(u32 *)(buffer + AUTO_UPDATE_BLANK_DIRECTORY)))
> +		return 0;
> +
> +	ret = mtd_erase(priv->flash, &erase);
> +	if (ret)
> +		return ret;
> +
> +	/*
> +	 * Populate the image address and then zero out the next directory so
> +	 * that the system controller doesn't complain if in "Single Image"
> +	 * mode.
> +	 */
> +	memcpy(buffer + AUTO_UPDATE_UPGRADE_DIRECTORY, &image_address,
> +	       AUTO_UPDATE_DIRECTORY_WIDTH);
> +	memset(buffer + AUTO_UPDATE_BLANK_DIRECTORY, 0x0, AUTO_UPDATE_DIRECTORY_WIDTH);
> +
> +	dev_info(priv->dev, "Writing the image address (%x) to the flash directory (%llx)\n",
> +		 image_address, directory_address);
> +
> +	ret = mtd_write(priv->flash, 0x0, erase_size, &bytes_written, (u_char *)buffer);
> +	if (ret)
> +		return ret;
> +
> +	if (bytes_written != erase_size)
> +		return ret;
> +
> +	return 0;
> +}
> +
> +static int mpfs_auto_update_write_bitstream(struct fw_upload *fw_uploader, const u8 *data,
> +					    u32 offset, u32 size, u32 *written)
> +{
> +	struct mpfs_auto_update_priv *priv = fw_uploader->dd_handle;
> +	struct erase_info erase;
> +	char *buffer;
> +	loff_t directory_address = AUTO_UPDATE_UPGRADE_DIRECTORY;
> +	size_t erase_size = AUTO_UPDATE_DIRECTORY_SIZE;
> +	size_t bytes_written = 0;
> +	u32 image_address;
> +	int ret;
> +
> +	erase_size = round_up(erase_size, (u64)priv->flash->erasesize);
> +
> +	image_address = AUTO_UPDATE_BITSTREAM_BASE +
> +		AUTO_UPDATE_UPGRADE_INDEX * priv->size_per_bitstream;
> +
> +	buffer = devm_kzalloc(priv->dev, erase_size, GFP_KERNEL);
> +	if (!buffer)
> +		return -ENOMEM;
> +
> +	ret = mpfs_auto_update_set_image_address(priv, buffer, image_address, directory_address);
> +	if (ret) {
> +		dev_err(priv->dev, "failed to set image address in the SPI directory: %d\n", ret);
> +		goto out;
> +	}
> +
> +	/*
> +	 * Now the .spi image itself can be written to the flash. Preservation
> +	 * of contents here is not important here, unlike the spi "directory"
> +	 * which must be RMWed.
> +	 */
> +	erase.len = round_up(size, (size_t)priv->flash->erasesize);
> +	erase.addr = image_address;
> +
> +	dev_info(priv->dev, "Erasing the flash at address (%x)\n", image_address);
> +	ret = mtd_erase(priv->flash, &erase);
> +	if (ret)
> +		goto out;
> +
> +	/*
> +	 * No parsing etc of the bitstream is required. The system controller
> +	 * will do all of that itself - including verifying that the bitstream
> +	 * is valid.
> +	 */
> +	dev_info(priv->dev, "Writing the image to the flash at address (%x)\n", image_address);
> +	ret = mtd_write(priv->flash, (loff_t)image_address, size, &bytes_written, data);
> +	if (ret)
> +		goto out;
> +
> +	if (bytes_written != size) {
> +		ret = -EIO;
> +		goto out;
> +	}
> +
> +	*written = bytes_written;
> +
> +out:
> +	devm_kfree(priv->dev, buffer);
> +	return ret;
> +}
> +
> +static enum fw_upload_err mpfs_auto_update_write(struct fw_upload *fw_uploader, const u8 *data,
> +						 u32 offset, u32 size, u32 *written)
> +{
> +	struct mpfs_auto_update_priv *priv = fw_uploader->dd_handle;
> +	enum fw_upload_err err = FW_UPLOAD_ERR_NONE;
> +	int ret;
> +
> +	reinit_completion(&priv->programming_complete);
> +
> +	ret = mpfs_auto_update_write_bitstream(fw_uploader, data, offset, size, written);
> +	if (ret) {
> +		err = FW_UPLOAD_ERR_RW_ERROR;
> +		goto out;
> +	}
> +
> +	if (priv->cancel_request) {
> +		err = FW_UPLOAD_ERR_CANCELED;
> +		goto out;
> +	}
> +
> +	ret = mpfs_auto_update_verify_image(fw_uploader);
> +	if (ret)
> +		err = FW_UPLOAD_ERR_HW_ERROR;
> +
> +out:
> +	complete(&priv->programming_complete);
> +
> +	return err;
> +}
> +
> +static const struct fw_upload_ops mpfs_auto_update_ops = {
> +	.prepare = mpfs_auto_update_prepare,
> +	.write = mpfs_auto_update_write,
> +	.poll_complete = mpfs_auto_update_poll_complete,
> +	.cancel = mpfs_auto_update_cancel,
> +};
> +
> +static int mpfs_auto_update_available(struct mpfs_auto_update_priv *priv)
> +{
> +	struct mpfs_mss_response *response;
> +	struct mpfs_mss_msg *message;
> +	u32 *response_msg;
> +	int ret;
> +
> +	response_msg = devm_kzalloc(priv->dev, AUTO_UPDATE_FEATURE_RESP_SIZE * sizeof(response_msg),
> +				    GFP_KERNEL);
> +	if (!response_msg)
> +		return -ENOMEM;
> +
> +	response = devm_kzalloc(priv->dev, sizeof(struct mpfs_mss_response), GFP_KERNEL);
> +	if (!response)
> +		return -ENOMEM;
> +
> +	message = devm_kzalloc(priv->dev, sizeof(struct mpfs_mss_msg), GFP_KERNEL);
> +	if (!message)
> +		return -ENOMEM;
> +
> +	/*
> +	 * To verify that Auto Update is possible, the "Query Security Service
> +	 * Request" is performed.
> +	 * This service has no command data & does not overload mbox_offset.
> +	 */
> +	response->resp_msg = response_msg;
> +	response->resp_size = AUTO_UPDATE_FEATURE_RESP_SIZE;
> +	message->cmd_opcode = AUTO_UPDATE_FEATURE_CMD_OPCODE;
> +	message->cmd_data_size = AUTO_UPDATE_FEATURE_CMD_DATA_SIZE;
> +	message->response = response;
> +	message->cmd_data = AUTO_UPDATE_FEATURE_CMD_DATA;
> +	message->mbox_offset = AUTO_UPDATE_DEFAULT_MBOX_OFFSET;
> +	message->resp_offset = AUTO_UPDATE_DEFAULT_RESP_OFFSET;
> +
> +	ret = mpfs_blocking_transaction(priv->sys_controller, message);
> +	if (ret)
> +		return ret;
> +
> +	/*
> +	 * Currently, the system controller's firmware does not generate any
> +	 * interrupts for failed services, so mpfs_blocking_transaction() should
> +	 * time out & therefore return an error.
> +	 * Hitting this check is highly unlikely at present, but if the system
> +	 * controller's behaviour changes so that it does generate interrupts
> +	 * for failed services, it will be required.
> +	 */
> +	if (response->resp_status)
> +		return -EIO;
> +
> +	/*
> +	 * Bit 5 of byte 1 is "UL_Auto Update" & if it is set, Auto Update is
> +	 * not possible.
> +	 */
> +	if (response_msg[1] & AUTO_UPDATE_FEATURE_ENABLED)
> +		return -EPERM;
> +
> +	return 0;
> +}
> +
> +static int mpfs_auto_update_probe(struct platform_device *pdev)
> +{
> +	struct device *dev = &pdev->dev;
> +	struct mpfs_auto_update_priv *priv;
> +	struct fw_upload *fw_uploader;
> +	int ret;
> +
> +	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
> +	if (!priv)
> +		return -ENOMEM;
> +
> +	priv->sys_controller = mpfs_sys_controller_get(dev);
> +	if (IS_ERR(priv->sys_controller))
> +		return dev_err_probe(dev, PTR_ERR(priv->sys_controller),
> +				     "Could not register as a sub device of the system controller\n");
> +
> +	priv->dev = dev;
> +	platform_set_drvdata(pdev, priv);
> +
> +	ret = mpfs_auto_update_available(priv);
> +	if (ret)
> +		return dev_err_probe(dev, ret,
> +				     "The current bitstream does not support auto-update\n");
> +
> +	init_completion(&priv->programming_complete);
> +
> +	fw_uploader = firmware_upload_register(THIS_MODULE, dev, "mpfs_bitstream",
> +					       &mpfs_auto_update_ops, priv);
> +	if (IS_ERR(fw_uploader))
> +		return dev_err_probe(dev, PTR_ERR(fw_uploader),
> +				     "Failed to register the bitstream uploader\n");
> +
> +	priv->fw_uploader = fw_uploader;
> +
> +	return 0;
> +}
> +
> +static void mpfs_auto_update_remove(struct platform_device *pdev)
> +{
> +	struct mpfs_auto_update_priv *priv = platform_get_drvdata(pdev);
> +
> +	firmware_upload_unregister(priv->fw_uploader);
> +}
> +
> +static struct platform_driver mpfs_auto_update_driver = {
> +	.driver = {
> +		.name = "mpfs-auto-update",
> +	},
> +	.probe = mpfs_auto_update_probe,
> +	.remove_new = mpfs_auto_update_remove,
> +};
> +module_platform_driver(mpfs_auto_update_driver);
> +
> +MODULE_LICENSE("GPL");
> +MODULE_AUTHOR("Conor Dooley <conor.dooley@microchip.com>");
> +MODULE_DESCRIPTION("PolarFire SoC Auto Update FPGA reprogramming");
> -- 
> 2.39.2
>
Conor Dooley April 11, 2023, 11:51 a.m. UTC | #2 
On Sun, Apr 02, 2023 at 06:57:18PM +0800, Xu Yilun wrote:
> On 2023-03-31 at 08:18:22 +0100, Conor Dooley wrote:
> > Add support for Auto Update reprogramming of the FPGA fabric on
> > PolarFire SoC, using the fw_upload mechanism a la the
> > intel-m10-bmc-sec-update driver.
> 
> I'm not an expert of MTD and have several concerns about the mtd
> writing part:
> 
> - This seems a generic case that writing a file content into the
>   driver restricted offset on mtd.

Right, writing at an offset is pretty generic.

> - Seems the firmware update interface could be created for mtd device,
>   like the nvmem interface?

I'm not entirely sure what you mean by this one.
Are you saying that there should be a generic way to do "firmware"
uploads to mtd devices, based on some sort of structure defined in
devicetree etc?
And perhaps then, this driver would register as a consumer of that
interface, pushing all of the direct mtd stuff out to that interface?

> - I assume the purpose of the fw upload interface is to protect
>   unexpected writing from user. So how to protect the mtd be
>   written by other user interfaces like mtdchar/block or nvmem?

I'm not very good with the security knobs so I don't have an answer for
you here (yet). But your assumption about the rationale here is correct
and I would like to block access.

> For other parts:
> 
> - directory block check, fw upload don't touch this block, so why check
>   it in fw upload process.

I don't know what you mean. The mtd device could be empty and the
directory would need to be populated with the image address etc.
Or, it could previously have been used for something else, and then it
would be required to reconfigure it correctly for Auto Upgrade.

Maybe I am misunderstanding you?

> 
> - image verification, no matter pass or fail, fw upload is actually done,
>   the flash is written and no way to rollback. So may remove from fw
>   upload process.

The FPGA is flash based, so writing to the mtd device is not going to
actually program the FPGA with this content. The original image/bitstream
in the mtd device is either going to be a) equivalent to what is already
programmed in the FPGA, b) not an upgrade over what is current in the
FPGA, c) something that was uploaded since the last time the device was
restarted.
I don't think that overwriting any of those is problematic, as if the
verification fails, the original bitstream will continue being used
until an image that does actually pass is provided.

Without doing the verification though, there's no way to tell for certain
if the FPGA is actually going to perform an upgrade to this new bitstream
once it is restarted, so I think there is value in keeping this.

If you're still very sceptical about whether this belongs in
drivers/fpga, I can always dump it in drivers/soc instead I guess.

Cheers,
Conor.
Xu Yilun April 14, 2023, 12:40 p.m. UTC | #3 
On 2023-04-11 at 12:51:34 +0100, Conor Dooley wrote:
> On Sun, Apr 02, 2023 at 06:57:18PM +0800, Xu Yilun wrote:
> > On 2023-03-31 at 08:18:22 +0100, Conor Dooley wrote:
> > > Add support for Auto Update reprogramming of the FPGA fabric on
> > > PolarFire SoC, using the fw_upload mechanism a la the
> > > intel-m10-bmc-sec-update driver.
> > 
> > I'm not an expert of MTD and have several concerns about the mtd
> > writing part:
> > 
> > - This seems a generic case that writing a file content into the
> >   driver restricted offset on mtd.
> 
> Right, writing at an offset is pretty generic.
> 
> > - Seems the firmware update interface could be created for mtd device,
> >   like the nvmem interface?
> 
> I'm not entirely sure what you mean by this one.
> Are you saying that there should be a generic way to do "firmware"
> uploads to mtd devices, based on some sort of structure defined in

Yes, I think if the firmware upload interface could be generic, then add
it to the mtd devices, rather than creating a "virtual mtd management"
device as this patch does.

> devicetree etc?
> And perhaps then, this driver would register as a consumer of that
> interface, pushing all of the direct mtd stuff out to that interface?
> 
> > - I assume the purpose of the fw upload interface is to protect
> >   unexpected writing from user. So how to protect the mtd be
> >   written by other user interfaces like mtdchar/block or nvmem?
> 
> I'm not very good with the security knobs so I don't have an answer for
> you here (yet). But your assumption about the rationale here is correct
> and I would like to block access.
> 
> > For other parts:
> > 
> > - directory block check, fw upload don't touch this block, so why check
> >   it in fw upload process.
> 
> I don't know what you mean. The mtd device could be empty and the
> directory would need to be populated with the image address etc.
> Or, it could previously have been used for something else, and then it
> would be required to reconfigure it correctly for Auto Upgrade.
> 
> Maybe I am misunderstanding you?

This directory block content is unrelated to the firmware image, so
could be written before firmware upload interface is created. The
firmware image uploading could be done without touching directory
blocks. That means directory blocks could be set by driver as
no-touch area for firmware uploading interface.

> 
> > 
> > - image verification, no matter pass or fail, fw upload is actually done,
> >   the flash is written and no way to rollback. So may remove from fw
> >   upload process.
> 
> The FPGA is flash based, so writing to the mtd device is not going to
> actually program the FPGA with this content. The original image/bitstream
> in the mtd device is either going to be a) equivalent to what is already
> programmed in the FPGA, b) not an upgrade over what is current in the
> FPGA, c) something that was uploaded since the last time the device was
> restarted.
> I don't think that overwriting any of those is problematic, as if the
> verification fails, the original bitstream will continue being used
> until an image that does actually pass is provided.
> 
> Without doing the verification though, there's no way to tell for certain
> if the FPGA is actually going to perform an upgrade to this new bitstream
> once it is restarted, so I think there is value in keeping this.

I don't mean to remove the support of verification from driver. I say
remove the verification from MTD firmware upload process. You can
use another interface in your driver to verify if FPGA could load the
content in user block at any time.

In my opinion, generic firmware upload interface for MTD could be used
out of FPGA cases, it just deals with writing the image file to the MTD.
Once you have nothing more to write for the image file, the job for the
interface could be successfully done.

> 
> If you're still very sceptical about whether this belongs in
> drivers/fpga, I can always dump it in drivers/soc instead I guess.

I actually don't think this belongs to FPGA domain. My pereference is a
generic MTD firmware upload interface for MTD domain.

But this is just my opinion. You may talk to MTD or SoC maintainers for
final solution.

Thanks,
Yilun

> 
> Cheers,
> Conor.
>
diff mbox series

Patch

diff --git a/drivers/fpga/Kconfig b/drivers/fpga/Kconfig
index 0a00763b9f28..2838ce7ddafe 100644
--- a/drivers/fpga/Kconfig
+++ b/drivers/fpga/Kconfig
@@ -257,6 +257,17 @@  config FPGA_M10_BMC_SEC_UPDATE
 	  (BMC) and provides support for secure updates for the BMC image,
 	  the FPGA image, the Root Entry Hashes, etc.
 
+config FPGA_MICROCHIP_AUTO_UPDATE
+	tristate "Microchip PolarFire SoC AUTO UPDATE"
+	depends on POLARFIRE_SOC_SYS_CTRL
+	select FW_LOADER
+	select FW_UPLOAD
+	help
+	  FPGA manager driver support for reprogramming PolarFire SoC from
+	  within Linux, using the Auto Upgrade feature of the system controller.
+
+	  If built as a module, it will be called microchip-auto-update.
+
 config FPGA_MGR_MICROCHIP_SPI
 	tristate "Microchip Polarfire SPI FPGA manager"
 	depends on SPI
diff --git a/drivers/fpga/Makefile b/drivers/fpga/Makefile
index 72e554b4d2f7..0b1ba6db860d 100644
--- a/drivers/fpga/Makefile
+++ b/drivers/fpga/Makefile
@@ -25,8 +25,9 @@  obj-$(CONFIG_FPGA_MGR_LATTICE_SYSCONFIG_SPI)	+= lattice-sysconfig-spi.o
 obj-$(CONFIG_ALTERA_PR_IP_CORE)		+= altera-pr-ip-core.o
 obj-$(CONFIG_ALTERA_PR_IP_CORE_PLAT)	+= altera-pr-ip-core-plat.o
 
-# FPGA Secure Update Drivers
+# FPGA Firmware Upload Drivers
 obj-$(CONFIG_FPGA_M10_BMC_SEC_UPDATE)	+= intel-m10-bmc-sec-update.o
+obj-$(CONFIG_FPGA_MICROCHIP_AUTO_UPDATE)	+= microchip-auto-update.o
 
 # FPGA Bridge Drivers
 obj-$(CONFIG_FPGA_BRIDGE)		+= fpga-bridge.o
diff --git a/drivers/fpga/microchip-auto-update.c b/drivers/fpga/microchip-auto-update.c
new file mode 100644
index 000000000000..5fa920b659c6
--- /dev/null
+++ b/drivers/fpga/microchip-auto-update.c
@@ -0,0 +1,494 @@ 
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Microchip Polarfire SoC "Auto Update" FPGA reprogramming.
+ *
+ * Documentation of this functionality is available in the "PolarFire® FPGA and
+ * PolarFire SoC FPGA Programming" User Guide.
+ *
+ * Copyright (c) 2022-2023 Microchip Corporation. All rights reserved.
+ *
+ * Author: Conor Dooley <conor.dooley@microchip.com>
+ */
+#include <linux/debugfs.h>
+#include <linux/firmware.h>
+#include <linux/math.h>
+#include <linux/module.h>
+#include <linux/mtd/mtd.h>
+#include <linux/of_device.h>
+#include <linux/sizes.h>
+
+#include <soc/microchip/mpfs.h>
+
+#define AUTO_UPDATE_DEFAULT_MBOX_OFFSET		0u
+#define AUTO_UPDATE_DEFAULT_RESP_OFFSET		0u
+
+#define AUTO_UPDATE_FEATURE_CMD_OPCODE		0x05u
+#define AUTO_UPDATE_FEATURE_CMD_DATA_SIZE	0u
+#define AUTO_UPDATE_FEATURE_RESP_SIZE		33u
+#define AUTO_UPDATE_FEATURE_CMD_DATA		NULL
+#define AUTO_UPDATE_FEATURE_ENABLED		BIT(5)
+
+#define AUTO_UPDATE_AUTHENTICATE_CMD_OPCODE	0x22u
+#define AUTO_UPDATE_AUTHENTICATE_CMD_DATA_SIZE	0u
+#define AUTO_UPDATE_AUTHENTICATE_RESP_SIZE	1u
+#define AUTO_UPDATE_AUTHENTICATE_CMD_DATA	NULL
+
+#define AUTO_UPDATE_PROGRAM_CMD_OPCODE		0x46u
+#define AUTO_UPDATE_PROGRAM_CMD_DATA_SIZE	0u
+#define AUTO_UPDATE_PROGRAM_RESP_SIZE		1u
+#define AUTO_UPDATE_PROGRAM_CMD_DATA		NULL
+
+/*
+ * SPI Flash layout example:
+ * |------------------------------| 0x0000000
+ * | 1 KiB                        |
+ * | SPI "directories"            |
+ * |------------------------------| 0x0000400
+ * | 1 MiB                        |
+ * | Reserved area                |
+ * | Used for bitstream info      |
+ * |------------------------------| 0x0100400
+ * | 20 MiB                       |
+ * | Golden Image                 |
+ * |------------------------------| 0x1500400
+ * | 20 MiB                       |
+ * | Auto Upgrade Image           |
+ * |------------------------------| 0x2900400
+ * | 20 MiB                       |
+ * | Reserved for multi-image IAP |
+ * | Unused for Auto Upgrade      |
+ * |------------------------------| 0x3D00400
+ * | ? B                          |
+ * | Unused                       |
+ * |------------------------------| 0x?
+ */
+#define AUTO_UPDATE_DIRECTORY_BASE	0u
+#define AUTO_UPDATE_DIRECTORY_WIDTH	4u
+#define AUTO_UPDATE_GOLDEN_INDEX	0u
+#define AUTO_UPDATE_UPGRADE_INDEX	1u
+#define AUTO_UPDATE_BLANK_INDEX		2u
+#define AUTO_UPDATE_GOLDEN_DIRECTORY	(AUTO_UPDATE_DIRECTORY_WIDTH * AUTO_UPDATE_GOLDEN_INDEX)
+#define AUTO_UPDATE_UPGRADE_DIRECTORY	(AUTO_UPDATE_DIRECTORY_WIDTH * AUTO_UPDATE_UPGRADE_INDEX)
+#define AUTO_UPDATE_BLANK_DIRECTORY	(AUTO_UPDATE_DIRECTORY_WIDTH * AUTO_UPDATE_BLANK_INDEX)
+#define AUTO_UPDATE_DIRECTORY_SIZE	SZ_1K
+#define AUTO_UPDATE_RESERVED_SIZE	SZ_1M
+#define AUTO_UPDATE_BITSTREAM_BASE	(AUTO_UPDATE_DIRECTORY_SIZE + AUTO_UPDATE_RESERVED_SIZE)
+
+#define AUTO_UPDATE_TIMEOUT_MS		60000
+
+struct mpfs_auto_update_priv {
+	struct mpfs_sys_controller *sys_controller;
+	struct device *dev;
+	struct mtd_info *flash;
+	struct fw_upload *fw_uploader;
+	struct completion programming_complete;
+	size_t size_per_bitstream;
+	bool cancel_request;
+};
+
+static enum fw_upload_err mpfs_auto_update_prepare(struct fw_upload *fw_uploader, const u8 *data,
+						   u32 size)
+{
+	struct mpfs_auto_update_priv *priv = fw_uploader->dd_handle;
+	size_t erase_size = AUTO_UPDATE_DIRECTORY_SIZE;
+
+	/*
+	 * Verifying the Golden Image is idealistic. It will be evaluated
+	 * against the currently programmed image and thus may fail - due to
+	 * either rollback protection (if its an older version than that in use)
+	 * or if the version is the same as that of the in-use image.
+	 * Extracting the information as to why a failure occurred is not
+	 * currently possible due to limitations of the system controller
+	 * driver. If those are fixed, verification of the Golden Image should
+	 * be added here.
+	 */
+
+	priv->flash = mpfs_sys_controller_get_flash(priv->sys_controller);
+	if (!priv->flash)
+		return FW_UPLOAD_ERR_HW_ERROR;
+
+	erase_size = round_up(erase_size, (u64)priv->flash->erasesize);
+
+	/*
+	 * We need to calculate if we have enough space in the flash for the
+	 * new image.
+	 * First, chop off the first 1 KiB as it's reserved for the directory.
+	 * The 1 MiB reserved for design info needs to be ignored also.
+	 * All that remains is carved into 3 & rounded down to the erasesize.
+	 * If this is smaller than the image size, we abort.
+	 * There's also no need to consume more than 20 MiB per image.
+	 */
+	priv->size_per_bitstream = priv->flash->size - SZ_1K - SZ_1M;
+	priv->size_per_bitstream = round_down(priv->size_per_bitstream / 3, erase_size);
+	if (priv->size_per_bitstream > 20 * SZ_1M)
+		priv->size_per_bitstream = 20 * SZ_1M;
+
+	if (priv->size_per_bitstream < size) {
+		dev_err(priv->dev,
+			"flash device has insufficient capacity to store this bitstream\n");
+		return FW_UPLOAD_ERR_INVALID_SIZE;
+	}
+
+	priv->cancel_request = false;
+
+	return FW_UPLOAD_ERR_NONE;
+}
+
+static void mpfs_auto_update_cancel(struct fw_upload *fw_uploader)
+{
+	struct mpfs_auto_update_priv *priv = fw_uploader->dd_handle;
+
+	priv->cancel_request = true;
+}
+
+static enum fw_upload_err mpfs_auto_update_poll_complete(struct fw_upload *fw_uploader)
+{
+	struct mpfs_auto_update_priv *priv = fw_uploader->dd_handle;
+	int ret;
+
+	/*
+	 * There is no meaningful way to get the status of the programming while
+	 * it is in progress, so attempting anything other than waiting for it
+	 * to complete would be misplaced.
+	 */
+	ret = wait_for_completion_timeout(&priv->programming_complete,
+					  msecs_to_jiffies(AUTO_UPDATE_TIMEOUT_MS));
+	if (ret)
+		return FW_UPLOAD_ERR_TIMEOUT;
+
+	return FW_UPLOAD_ERR_NONE;
+}
+
+static int mpfs_auto_update_verify_image(struct fw_upload *fw_uploader)
+{
+	struct mpfs_auto_update_priv *priv = fw_uploader->dd_handle;
+	struct mpfs_mss_response *response;
+	struct mpfs_mss_msg *message;
+	u32 *response_msg;
+	int ret;
+
+	response_msg = devm_kzalloc(priv->dev, AUTO_UPDATE_FEATURE_RESP_SIZE * sizeof(response_msg),
+				    GFP_KERNEL);
+	if (!response_msg)
+		return -ENOMEM;
+
+	response = devm_kzalloc(priv->dev, sizeof(struct mpfs_mss_response), GFP_KERNEL);
+	if (!response) {
+		ret = -ENOMEM;
+		goto free_response_msg;
+	}
+
+	message = devm_kzalloc(priv->dev, sizeof(struct mpfs_mss_msg), GFP_KERNEL);
+	if (!message) {
+		ret = -ENOMEM;
+		goto free_response;
+	}
+
+	/*
+	 * The system controller can verify that an image in the flash is valid.
+	 * Rather than duplicate the check in this driver, call the relevant
+	 * service from the system controller instead.
+	 * This service has no command data and no response data. It overloads
+	 * mbox_offset with the image index in the flash's SPI directory where
+	 * the bitstream is located.
+	 */
+	response->resp_msg = response_msg;
+	response->resp_size = AUTO_UPDATE_AUTHENTICATE_RESP_SIZE;
+	message->cmd_opcode = AUTO_UPDATE_AUTHENTICATE_CMD_OPCODE;
+	message->cmd_data_size = AUTO_UPDATE_AUTHENTICATE_CMD_DATA_SIZE;
+	message->response = response;
+	message->cmd_data = AUTO_UPDATE_AUTHENTICATE_CMD_DATA;
+	message->mbox_offset = AUTO_UPDATE_UPGRADE_INDEX;
+	message->resp_offset = AUTO_UPDATE_DEFAULT_RESP_OFFSET;
+
+	dev_info(priv->dev, "Running verification of Upgrade Image\n");
+	ret = mpfs_blocking_transaction(priv->sys_controller, message);
+	if (ret | response->resp_status) {
+		dev_warn(priv->dev, "Verification of Upgrade Image failed!\n");
+		ret = ret ? ret : -EBADMSG;
+	}
+
+	dev_info(priv->dev, "Verification of Upgrade Image passed!\n");
+
+	devm_kfree(priv->dev, message);
+free_response:
+	devm_kfree(priv->dev, response);
+free_response_msg:
+	devm_kfree(priv->dev, response_msg);
+
+	return ret;
+}
+
+static int mpfs_auto_update_set_image_address(struct mpfs_auto_update_priv *priv, char *buffer,
+					      u32 image_address, loff_t directory_address)
+{
+	struct erase_info erase;
+	size_t erase_size = AUTO_UPDATE_DIRECTORY_SIZE;
+	size_t bytes_written = 0, bytes_read = 0;
+	int ret;
+
+	erase_size = round_up(erase_size, (u64)priv->flash->erasesize);
+
+	erase.addr = AUTO_UPDATE_DIRECTORY_BASE;
+	erase.len = erase_size;
+
+	/*
+	 * We need to write the "SPI DIRECTORY" to the first 1 KiB, telling
+	 * the system controller where to find the actual bitstream. Since
+	 * this is spi-nor, we have to read the first eraseblock, erase that
+	 * portion of the flash, modify the data and then write it back.
+	 * There's no need to do this though if things are already the way they
+	 * should be, so check and save the write in that case.
+	 */
+	ret = mtd_read(priv->flash, AUTO_UPDATE_DIRECTORY_BASE, erase_size, &bytes_read,
+		       (u_char *)buffer);
+	if (ret)
+		return ret;
+
+	if (bytes_read != erase_size)
+		return -EIO;
+
+	if ((*(u32 *)(buffer + AUTO_UPDATE_UPGRADE_DIRECTORY) == image_address) &&
+	    !(*(u32 *)(buffer + AUTO_UPDATE_BLANK_DIRECTORY)))
+		return 0;
+
+	ret = mtd_erase(priv->flash, &erase);
+	if (ret)
+		return ret;
+
+	/*
+	 * Populate the image address and then zero out the next directory so
+	 * that the system controller doesn't complain if in "Single Image"
+	 * mode.
+	 */
+	memcpy(buffer + AUTO_UPDATE_UPGRADE_DIRECTORY, &image_address,
+	       AUTO_UPDATE_DIRECTORY_WIDTH);
+	memset(buffer + AUTO_UPDATE_BLANK_DIRECTORY, 0x0, AUTO_UPDATE_DIRECTORY_WIDTH);
+
+	dev_info(priv->dev, "Writing the image address (%x) to the flash directory (%llx)\n",
+		 image_address, directory_address);
+
+	ret = mtd_write(priv->flash, 0x0, erase_size, &bytes_written, (u_char *)buffer);
+	if (ret)
+		return ret;
+
+	if (bytes_written != erase_size)
+		return ret;
+
+	return 0;
+}
+
+static int mpfs_auto_update_write_bitstream(struct fw_upload *fw_uploader, const u8 *data,
+					    u32 offset, u32 size, u32 *written)
+{
+	struct mpfs_auto_update_priv *priv = fw_uploader->dd_handle;
+	struct erase_info erase;
+	char *buffer;
+	loff_t directory_address = AUTO_UPDATE_UPGRADE_DIRECTORY;
+	size_t erase_size = AUTO_UPDATE_DIRECTORY_SIZE;
+	size_t bytes_written = 0;
+	u32 image_address;
+	int ret;
+
+	erase_size = round_up(erase_size, (u64)priv->flash->erasesize);
+
+	image_address = AUTO_UPDATE_BITSTREAM_BASE +
+		AUTO_UPDATE_UPGRADE_INDEX * priv->size_per_bitstream;
+
+	buffer = devm_kzalloc(priv->dev, erase_size, GFP_KERNEL);
+	if (!buffer)
+		return -ENOMEM;
+
+	ret = mpfs_auto_update_set_image_address(priv, buffer, image_address, directory_address);
+	if (ret) {
+		dev_err(priv->dev, "failed to set image address in the SPI directory: %d\n", ret);
+		goto out;
+	}
+
+	/*
+	 * Now the .spi image itself can be written to the flash. Preservation
+	 * of contents here is not important here, unlike the spi "directory"
+	 * which must be RMWed.
+	 */
+	erase.len = round_up(size, (size_t)priv->flash->erasesize);
+	erase.addr = image_address;
+
+	dev_info(priv->dev, "Erasing the flash at address (%x)\n", image_address);
+	ret = mtd_erase(priv->flash, &erase);
+	if (ret)
+		goto out;
+
+	/*
+	 * No parsing etc of the bitstream is required. The system controller
+	 * will do all of that itself - including verifying that the bitstream
+	 * is valid.
+	 */
+	dev_info(priv->dev, "Writing the image to the flash at address (%x)\n", image_address);
+	ret = mtd_write(priv->flash, (loff_t)image_address, size, &bytes_written, data);
+	if (ret)
+		goto out;
+
+	if (bytes_written != size) {
+		ret = -EIO;
+		goto out;
+	}
+
+	*written = bytes_written;
+
+out:
+	devm_kfree(priv->dev, buffer);
+	return ret;
+}
+
+static enum fw_upload_err mpfs_auto_update_write(struct fw_upload *fw_uploader, const u8 *data,
+						 u32 offset, u32 size, u32 *written)
+{
+	struct mpfs_auto_update_priv *priv = fw_uploader->dd_handle;
+	enum fw_upload_err err = FW_UPLOAD_ERR_NONE;
+	int ret;
+
+	reinit_completion(&priv->programming_complete);
+
+	ret = mpfs_auto_update_write_bitstream(fw_uploader, data, offset, size, written);
+	if (ret) {
+		err = FW_UPLOAD_ERR_RW_ERROR;
+		goto out;
+	}
+
+	if (priv->cancel_request) {
+		err = FW_UPLOAD_ERR_CANCELED;
+		goto out;
+	}
+
+	ret = mpfs_auto_update_verify_image(fw_uploader);
+	if (ret)
+		err = FW_UPLOAD_ERR_HW_ERROR;
+
+out:
+	complete(&priv->programming_complete);
+
+	return err;
+}
+
+static const struct fw_upload_ops mpfs_auto_update_ops = {
+	.prepare = mpfs_auto_update_prepare,
+	.write = mpfs_auto_update_write,
+	.poll_complete = mpfs_auto_update_poll_complete,
+	.cancel = mpfs_auto_update_cancel,
+};
+
+static int mpfs_auto_update_available(struct mpfs_auto_update_priv *priv)
+{
+	struct mpfs_mss_response *response;
+	struct mpfs_mss_msg *message;
+	u32 *response_msg;
+	int ret;
+
+	response_msg = devm_kzalloc(priv->dev, AUTO_UPDATE_FEATURE_RESP_SIZE * sizeof(response_msg),
+				    GFP_KERNEL);
+	if (!response_msg)
+		return -ENOMEM;
+
+	response = devm_kzalloc(priv->dev, sizeof(struct mpfs_mss_response), GFP_KERNEL);
+	if (!response)
+		return -ENOMEM;
+
+	message = devm_kzalloc(priv->dev, sizeof(struct mpfs_mss_msg), GFP_KERNEL);
+	if (!message)
+		return -ENOMEM;
+
+	/*
+	 * To verify that Auto Update is possible, the "Query Security Service
+	 * Request" is performed.
+	 * This service has no command data & does not overload mbox_offset.
+	 */
+	response->resp_msg = response_msg;
+	response->resp_size = AUTO_UPDATE_FEATURE_RESP_SIZE;
+	message->cmd_opcode = AUTO_UPDATE_FEATURE_CMD_OPCODE;
+	message->cmd_data_size = AUTO_UPDATE_FEATURE_CMD_DATA_SIZE;
+	message->response = response;
+	message->cmd_data = AUTO_UPDATE_FEATURE_CMD_DATA;
+	message->mbox_offset = AUTO_UPDATE_DEFAULT_MBOX_OFFSET;
+	message->resp_offset = AUTO_UPDATE_DEFAULT_RESP_OFFSET;
+
+	ret = mpfs_blocking_transaction(priv->sys_controller, message);
+	if (ret)
+		return ret;
+
+	/*
+	 * Currently, the system controller's firmware does not generate any
+	 * interrupts for failed services, so mpfs_blocking_transaction() should
+	 * time out & therefore return an error.
+	 * Hitting this check is highly unlikely at present, but if the system
+	 * controller's behaviour changes so that it does generate interrupts
+	 * for failed services, it will be required.
+	 */
+	if (response->resp_status)
+		return -EIO;
+
+	/*
+	 * Bit 5 of byte 1 is "UL_Auto Update" & if it is set, Auto Update is
+	 * not possible.
+	 */
+	if (response_msg[1] & AUTO_UPDATE_FEATURE_ENABLED)
+		return -EPERM;
+
+	return 0;
+}
+
+static int mpfs_auto_update_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct mpfs_auto_update_priv *priv;
+	struct fw_upload *fw_uploader;
+	int ret;
+
+	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	priv->sys_controller = mpfs_sys_controller_get(dev);
+	if (IS_ERR(priv->sys_controller))
+		return dev_err_probe(dev, PTR_ERR(priv->sys_controller),
+				     "Could not register as a sub device of the system controller\n");
+
+	priv->dev = dev;
+	platform_set_drvdata(pdev, priv);
+
+	ret = mpfs_auto_update_available(priv);
+	if (ret)
+		return dev_err_probe(dev, ret,
+				     "The current bitstream does not support auto-update\n");
+
+	init_completion(&priv->programming_complete);
+
+	fw_uploader = firmware_upload_register(THIS_MODULE, dev, "mpfs_bitstream",
+					       &mpfs_auto_update_ops, priv);
+	if (IS_ERR(fw_uploader))
+		return dev_err_probe(dev, PTR_ERR(fw_uploader),
+				     "Failed to register the bitstream uploader\n");
+
+	priv->fw_uploader = fw_uploader;
+
+	return 0;
+}
+
+static void mpfs_auto_update_remove(struct platform_device *pdev)
+{
+	struct mpfs_auto_update_priv *priv = platform_get_drvdata(pdev);
+
+	firmware_upload_unregister(priv->fw_uploader);
+}
+
+static struct platform_driver mpfs_auto_update_driver = {
+	.driver = {
+		.name = "mpfs-auto-update",
+	},
+	.probe = mpfs_auto_update_probe,
+	.remove_new = mpfs_auto_update_remove,
+};
+module_platform_driver(mpfs_auto_update_driver);
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Conor Dooley <conor.dooley@microchip.com>");
+MODULE_DESCRIPTION("PolarFire SoC Auto Update FPGA reprogramming");