diff mbox series

[8/8] drm/bridge: it66121: Allow link this driver as a lib

Message ID 20231114150130.497915-9-sui.jingfeng@linux.dev (mailing list archive)
State New, archived
Headers show
Series Allow link the it66121 display bridge driver as a lib | expand

Commit Message

Sui Jingfeng Nov. 14, 2023, 3:01 p.m. UTC
From: Sui Jingfeng <suijingfeng@loongson.cn>

The it66121_create_bridge() and it66121_destroy_bridge() are added to
export the core functionalities. Create a connector manually by using
bridge connector helpers when link as a lib.

Signed-off-by: Sui Jingfeng <suijingfeng@loongson.cn>
---
 drivers/gpu/drm/bridge/ite-it66121.c | 134 +++++++++++++++++++--------
 include/drm/bridge/ite-it66121.h     |  17 ++++
 2 files changed, 113 insertions(+), 38 deletions(-)
 create mode 100644 include/drm/bridge/ite-it66121.h

Comments

Dmitry Baryshkov Nov. 14, 2023, 4:30 p.m. UTC | #1
On Tue, 14 Nov 2023 at 17:09, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>
> From: Sui Jingfeng <suijingfeng@loongson.cn>
>
> The it66121_create_bridge() and it66121_destroy_bridge() are added to
> export the core functionalities. Create a connector manually by using
> bridge connector helpers when link as a lib.
>
> Signed-off-by: Sui Jingfeng <suijingfeng@loongson.cn>
> ---
>  drivers/gpu/drm/bridge/ite-it66121.c | 134 +++++++++++++++++++--------
>  include/drm/bridge/ite-it66121.h     |  17 ++++
>  2 files changed, 113 insertions(+), 38 deletions(-)
>  create mode 100644 include/drm/bridge/ite-it66121.h
>
> diff --git a/drivers/gpu/drm/bridge/ite-it66121.c b/drivers/gpu/drm/bridge/ite-it66121.c
> index 8971414a2a60..f5968b679c5d 100644
> --- a/drivers/gpu/drm/bridge/ite-it66121.c
> +++ b/drivers/gpu/drm/bridge/ite-it66121.c
> @@ -22,6 +22,7 @@
>
>  #include <drm/drm_atomic_helper.h>
>  #include <drm/drm_bridge.h>
> +#include <drm/drm_bridge_connector.h>
>  #include <drm/drm_edid.h>
>  #include <drm/drm_modes.h>
>  #include <drm/drm_print.h>
> @@ -703,14 +704,32 @@ static int it66121_bridge_attach(struct drm_bridge *bridge,
>                                  enum drm_bridge_attach_flags flags)
>  {
>         struct it66121_ctx *ctx = bridge_to_it66121(bridge);
> +       struct drm_bridge *next_bridge = ctx->next_bridge;
> +       struct drm_encoder *encoder = bridge->encoder;
>         int ret;
>
> -       if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR))
> -               return -EINVAL;
> +       if (next_bridge) {
> +               if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)) {
> +                       WARN_ON(1);

Why? At least use WARN() instead

> +                       flags |= DRM_BRIDGE_ATTACH_NO_CONNECTOR;
> +               }
> +               ret = drm_bridge_attach(encoder, next_bridge, bridge, flags);
> +               if (ret)
> +                       return ret;
> +       } else {
> +               struct drm_connector *connector;
>
> -       ret = drm_bridge_attach(bridge->encoder, ctx->next_bridge, bridge, flags);
> -       if (ret)
> -               return ret;
> +               if (flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)
> +                       WARN_ON(1);

No. It is perfectly fine to create attach a bridge with no next_bridge
and with the DRM_BRIDGE_ATTACH_NO_CONNECTOR flag.

> +
> +               connector = drm_bridge_connector_init(bridge->dev, encoder);
> +               if (IS_ERR(connector))
> +                       return PTR_ERR(connector);
> +
> +               drm_connector_attach_encoder(connector, encoder);

This goes into your device driver.

> +
> +               ctx->connector = connector;
> +       }
>
>         if (ctx->info->id == ID_IT66121) {
>                 ret = regmap_write_bits(ctx->regmap, IT66121_CLK_BANK_REG,
> @@ -1632,16 +1651,13 @@ static const char * const it66121_supplies[] = {
>         "vcn33", "vcn18", "vrf12"
>  };
>
> -static int it66121_probe(struct i2c_client *client)
> +int it66121_create_bridge(struct i2c_client *client, bool of_support,
> +                         bool hpd_support, bool audio_support,
> +                         struct drm_bridge **bridge)
>  {
> +       struct device *dev = &client->dev;
>         int ret;
>         struct it66121_ctx *ctx;
> -       struct device *dev = &client->dev;
> -
> -       if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) {
> -               dev_err(dev, "I2C check functionality failed.\n");
> -               return -ENXIO;
> -       }
>
>         ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
>         if (!ctx)
> @@ -1649,24 +1665,19 @@ static int it66121_probe(struct i2c_client *client)
>
>         ctx->dev = dev;
>         ctx->client = client;
> -       ctx->info = i2c_get_match_data(client);
> -
> -       ret = it66121_of_read_bus_width(dev, &ctx->bus_width);
> -       if (ret)
> -               return ret;
> -
> -       ret = it66121_of_get_next_bridge(dev, &ctx->next_bridge);
> -       if (ret)
> -               return ret;
> -
> -       i2c_set_clientdata(client, ctx);
>         mutex_init(&ctx->lock);
>
> -       ret = devm_regulator_bulk_get_enable(dev, ARRAY_SIZE(it66121_supplies),
> -                                            it66121_supplies);
> -       if (ret) {
> -               dev_err(dev, "Failed to enable power supplies\n");
> -               return ret;
> +       if (of_support) {
> +               ret = it66121_of_read_bus_width(dev, &ctx->bus_width);
> +               if (ret)
> +                       return ret;
> +
> +               ret = it66121_of_get_next_bridge(dev, &ctx->next_bridge);
> +               if (ret)
> +                       return ret;
> +       } else {
> +               ctx->bus_width = 24;
> +               ctx->next_bridge = NULL;
>         }

A better alternative would be to turn OF calls into fwnode calls and
to populate the fwnode properties. See
drivers/platform/x86/intel/chtwc_int33fe.c for example.

>
>         it66121_hw_reset(ctx);
> @@ -1679,33 +1690,80 @@ static int it66121_probe(struct i2c_client *client)
>         if (ret)
>                 return ret;
>
> -       if (ctx->vender_id != ctx->info->vid ||
> -           ctx->device_id != ctx->info->pid)
> +       ctx->info = it66121_get_match_data(ctx->vender_id, ctx->device_id);
> +       if (!ctx->info)
>                 return -ENODEV;
>
> -       ret = devm_request_threaded_irq(dev, client->irq, NULL, it66121_irq_threaded_handler,
> -                                       IRQF_ONESHOT, dev_name(dev), ctx);
> -       if (ret < 0) {
> -               dev_err(dev, "Failed to request irq %d:%d\n", client->irq, ret);
> -               return ret;
> +       if (hpd_support) {

Who handles HPD in your platform case?

> +               ret = devm_request_threaded_irq(dev, client->irq, NULL,
> +                                               it66121_irq_threaded_handler,
> +                                               IRQF_ONESHOT, dev_name(dev),
> +                                               ctx);
> +               if (ret < 0) {
> +                       dev_err(dev, "Failed to request irq: %d\n", ret);
> +                       return ret;
> +               }
>         }
>
>         it66121_bridge_init_base(&ctx->bridge, dev->of_node, true);
>
> -       it66121_audio_codec_init(ctx, dev);
> +       if (audio_support)
> +               it66121_audio_codec_init(ctx, dev);
> +
> +       *bridge = &ctx->bridge;
>
>         dev_info(dev, "IT66121 probed, chip id: 0x%x:0x%x, revision: %u\n",
>                  ctx->vender_id, ctx->device_id, ctx->revision);
>
>         return 0;
>  }
> +EXPORT_SYMBOL_GPL(it66121_create_bridge);
> +
> +static int it66121_probe(struct i2c_client *client)
> +{
> +       struct device *dev = &client->dev;
> +       struct it66121_ctx *ctx;
> +       struct drm_bridge *bridge;
> +       int ret;
> +
> +       if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) {
> +               dev_err(dev, "I2C check functionality failed.\n");
> +               return -ENXIO;
> +       }
> +
> +       ret = devm_regulator_bulk_get_enable(dev, ARRAY_SIZE(it66121_supplies),
> +                                            it66121_supplies);
> +       if (ret) {
> +               dev_err(dev, "Failed to enable power supplies\n");
> +               return ret;
> +       }
> +
> +       ret = it66121_create_bridge(client, true, true, true, &bridge);
> +       if (ret)
> +               return ret;
> +
> +       ctx = bridge_to_it66121(bridge);
> +
> +       i2c_set_clientdata(client, ctx);
> +
> +       return 0;
> +}
> +
> +void it66121_destroy_bridge(struct drm_bridge *bridge)
> +{
> +       struct it66121_ctx *ctx = bridge_to_it66121(bridge);
> +
> +       drm_bridge_remove(bridge);
> +
> +       mutex_destroy(&ctx->lock);
> +}
> +EXPORT_SYMBOL_GPL(it66121_destroy_bridge);
>
>  static void it66121_remove(struct i2c_client *client)
>  {
>         struct it66121_ctx *ctx = i2c_get_clientdata(client);
>
> -       drm_bridge_remove(&ctx->bridge);
> -       mutex_destroy(&ctx->lock);
> +       it66121_destroy_bridge(&ctx->bridge);
>  }
>
>  static const struct of_device_id it66121_dt_match[] = {
> diff --git a/include/drm/bridge/ite-it66121.h b/include/drm/bridge/ite-it66121.h
> new file mode 100644
> index 000000000000..e6753f695b7f
> --- /dev/null
> +++ b/include/drm/bridge/ite-it66121.h
> @@ -0,0 +1,17 @@
> +/* SPDX-License-Identifier: GPL-2.0-only */
> +
> +#ifndef __ITE_IT66121_H__
> +#define __ITE_IT66121_H__
> +
> +#include <linux/i2c.h>
> +
> +#include <drm/drm_bridge.h>
> +#include <drm/drm_device.h>
> +
> +int it66121_create_bridge(struct i2c_client *client, bool of_support,
> +                         bool hpd_support, bool audio_support,
> +                         struct drm_bridge **bridge);
> +
> +void it66121_destroy_bridge(struct drm_bridge *bridge);
> +
> +#endif
> --
> 2.34.1
>
kernel test robot Nov. 15, 2023, 1:02 p.m. UTC | #2
Hi Sui,

kernel test robot noticed the following build warnings:

[auto build test WARNING on drm-misc/drm-misc-next]
[also build test WARNING on linus/master v6.7-rc1 next-20231115]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Sui-Jingfeng/drm-bridge-it66121-Use-dev-replace-ctx-dev-in-the-it66121_probe/20231114-231203
base:   git://anongit.freedesktop.org/drm/drm-misc drm-misc-next
patch link:    https://lore.kernel.org/r/20231114150130.497915-9-sui.jingfeng%40linux.dev
patch subject: [PATCH 8/8] drm/bridge: it66121: Allow link this driver as a lib
config: x86_64-randconfig-005-20231115 (https://download.01.org/0day-ci/archive/20231115/202311152059.KuAZPGVE-lkp@intel.com/config)
compiler: gcc-12 (Debian 12.2.0-14) 12.2.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20231115/202311152059.KuAZPGVE-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202311152059.KuAZPGVE-lkp@intel.com/

All warnings (new ones prefixed by >>):

>> drivers/gpu/drm/bridge/ite-it66121.c:1654:5: warning: no previous prototype for 'it66121_create_bridge' [-Wmissing-prototypes]
    1654 | int it66121_create_bridge(struct i2c_client *client, bool of_support,
         |     ^~~~~~~~~~~~~~~~~~~~~
>> drivers/gpu/drm/bridge/ite-it66121.c:1752:6: warning: no previous prototype for 'it66121_destroy_bridge' [-Wmissing-prototypes]
    1752 | void it66121_destroy_bridge(struct drm_bridge *bridge)
         |      ^~~~~~~~~~~~~~~~~~~~~~


vim +/it66121_create_bridge +1654 drivers/gpu/drm/bridge/ite-it66121.c

  1653	
> 1654	int it66121_create_bridge(struct i2c_client *client, bool of_support,
  1655				  bool hpd_support, bool audio_support,
  1656				  struct drm_bridge **bridge)
  1657	{
  1658		struct device *dev = &client->dev;
  1659		int ret;
  1660		struct it66121_ctx *ctx;
  1661	
  1662		ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
  1663		if (!ctx)
  1664			return -ENOMEM;
  1665	
  1666		ctx->dev = dev;
  1667		ctx->client = client;
  1668		mutex_init(&ctx->lock);
  1669	
  1670		if (of_support) {
  1671			ret = it66121_of_read_bus_width(dev, &ctx->bus_width);
  1672			if (ret)
  1673				return ret;
  1674	
  1675			ret = it66121_of_get_next_bridge(dev, &ctx->next_bridge);
  1676			if (ret)
  1677				return ret;
  1678		} else {
  1679			ctx->bus_width = 24;
  1680			ctx->next_bridge = NULL;
  1681		}
  1682	
  1683		it66121_hw_reset(ctx);
  1684	
  1685		ctx->regmap = devm_regmap_init_i2c(client, &it66121_regmap_config);
  1686		if (IS_ERR(ctx->regmap))
  1687			return PTR_ERR(ctx->regmap);
  1688	
  1689		ret = it66121_read_chip_id(ctx, false);
  1690		if (ret)
  1691			return ret;
  1692	
  1693		ctx->info = it66121_get_match_data(ctx->vender_id, ctx->device_id);
  1694		if (!ctx->info)
  1695			return -ENODEV;
  1696	
  1697		if (hpd_support) {
  1698			ret = devm_request_threaded_irq(dev, client->irq, NULL,
  1699							it66121_irq_threaded_handler,
  1700							IRQF_ONESHOT, dev_name(dev),
  1701							ctx);
  1702			if (ret < 0) {
  1703				dev_err(dev, "Failed to request irq: %d\n", ret);
  1704				return ret;
  1705			}
  1706		}
  1707	
  1708		it66121_bridge_init_base(&ctx->bridge, dev->of_node, true);
  1709	
  1710		if (audio_support)
  1711			it66121_audio_codec_init(ctx, dev);
  1712	
  1713		*bridge = &ctx->bridge;
  1714	
  1715		dev_info(dev, "IT66121 probed, chip id: 0x%x:0x%x, revision: %u\n",
  1716			 ctx->vender_id, ctx->device_id, ctx->revision);
  1717	
  1718		return 0;
  1719	}
  1720	EXPORT_SYMBOL_GPL(it66121_create_bridge);
  1721	
  1722	static int it66121_probe(struct i2c_client *client)
  1723	{
  1724		struct device *dev = &client->dev;
  1725		struct it66121_ctx *ctx;
  1726		struct drm_bridge *bridge;
  1727		int ret;
  1728	
  1729		if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) {
  1730			dev_err(dev, "I2C check functionality failed.\n");
  1731			return -ENXIO;
  1732		}
  1733	
  1734		ret = devm_regulator_bulk_get_enable(dev, ARRAY_SIZE(it66121_supplies),
  1735						     it66121_supplies);
  1736		if (ret) {
  1737			dev_err(dev, "Failed to enable power supplies\n");
  1738			return ret;
  1739		}
  1740	
  1741		ret = it66121_create_bridge(client, true, true, true, &bridge);
  1742		if (ret)
  1743			return ret;
  1744	
  1745		ctx = bridge_to_it66121(bridge);
  1746	
  1747		i2c_set_clientdata(client, ctx);
  1748	
  1749		return 0;
  1750	}
  1751	
> 1752	void it66121_destroy_bridge(struct drm_bridge *bridge)
  1753	{
  1754		struct it66121_ctx *ctx = bridge_to_it66121(bridge);
  1755	
  1756		drm_bridge_remove(bridge);
  1757	
  1758		mutex_destroy(&ctx->lock);
  1759	}
  1760	EXPORT_SYMBOL_GPL(it66121_destroy_bridge);
  1761
Sui Jingfeng Nov. 16, 2023, 9:14 a.m. UTC | #3
Hi,

Thanks a lot for reviewing!


On 2023/11/15 00:30, Dmitry Baryshkov wrote:
> On Tue, 14 Nov 2023 at 17:09, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>> From: Sui Jingfeng <suijingfeng@loongson.cn>
>>
>> The it66121_create_bridge() and it66121_destroy_bridge() are added to
>> export the core functionalities. Create a connector manually by using
>> bridge connector helpers when link as a lib.
>>
>> Signed-off-by: Sui Jingfeng <suijingfeng@loongson.cn>
>> ---
>>   drivers/gpu/drm/bridge/ite-it66121.c | 134 +++++++++++++++++++--------
>>   include/drm/bridge/ite-it66121.h     |  17 ++++
>>   2 files changed, 113 insertions(+), 38 deletions(-)
>>   create mode 100644 include/drm/bridge/ite-it66121.h
>>
>> diff --git a/drivers/gpu/drm/bridge/ite-it66121.c b/drivers/gpu/drm/bridge/ite-it66121.c
>> index 8971414a2a60..f5968b679c5d 100644
>> --- a/drivers/gpu/drm/bridge/ite-it66121.c
>> +++ b/drivers/gpu/drm/bridge/ite-it66121.c
>> @@ -22,6 +22,7 @@
>>
>>   #include <drm/drm_atomic_helper.h>
>>   #include <drm/drm_bridge.h>
>> +#include <drm/drm_bridge_connector.h>
>>   #include <drm/drm_edid.h>
>>   #include <drm/drm_modes.h>
>>   #include <drm/drm_print.h>
>> @@ -703,14 +704,32 @@ static int it66121_bridge_attach(struct drm_bridge *bridge,
>>                                   enum drm_bridge_attach_flags flags)
>>   {
>>          struct it66121_ctx *ctx = bridge_to_it66121(bridge);
>> +       struct drm_bridge *next_bridge = ctx->next_bridge;
>> +       struct drm_encoder *encoder = bridge->encoder;
>>          int ret;
>>
>> -       if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR))
>> -               return -EINVAL;
>> +       if (next_bridge) {
>> +               if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)) {
>> +                       WARN_ON(1);
> Why? At least use WARN() instead

Originally I want to




>> +                       flags |= DRM_BRIDGE_ATTACH_NO_CONNECTOR;
>> +               }
>> +               ret = drm_bridge_attach(encoder, next_bridge, bridge, flags);
>> +               if (ret)
>> +                       return ret;
>> +       } else {
>> +               struct drm_connector *connector;
>>
>> -       ret = drm_bridge_attach(bridge->encoder, ctx->next_bridge, bridge, flags);
>> -       if (ret)
>> -               return ret;
>> +               if (flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)
>> +                       WARN_ON(1);
> No. It is perfectly fine to create attach a bridge with no next_bridge
> and with the DRM_BRIDGE_ATTACH_NO_CONNECTOR flag.
>

The document say when DRM_BRIDGE_ATTACH_NO_CONNECTOR flag is set
the bridge shall not create a drm_connector. So I think if a display
bridge driver don't have a next bridge attached (Currently, this is
told by the DT), it says that this is a non-DT environment. On such
a case, this display bridge driver(it66121.ko) should behavior like
a *agent*. Because the upstream port of it66121 is the DVO port of
the display controller, the downstream port of it66121 is the HDMI
connector. it66121 is on the middle. So I think the it66121.ko should
handle all of troubles on behalf of the display controller drivers.

Therefore (when in non-DT use case), the display controller drivers
side should not set DRM_BRIDGE_ATTACH_NO_CONNECTOR flag anymore.
Which is to hint that the it66121 should totally in charge of those
tasks (either by using bridge connector helper or create a connector
manually). I don't understand on such a case, why bother display
controller drivers anymore.


>> +
>> +               connector = drm_bridge_connector_init(bridge->dev, encoder);
>> +               if (IS_ERR(connector))
>> +                       return PTR_ERR(connector);
>> +
>> +               drm_connector_attach_encoder(connector, encoder);
> This goes into your device driver.
>
>> +
>> +               ctx->connector = connector;
>> +       }
>>
>>          if (ctx->info->id == ID_IT66121) {
>>                  ret = regmap_write_bits(ctx->regmap, IT66121_CLK_BANK_REG,
>> @@ -1632,16 +1651,13 @@ static const char * const it66121_supplies[] = {
>>          "vcn33", "vcn18", "vrf12"
>>   };
>
Dmitry Baryshkov Nov. 16, 2023, 9:30 a.m. UTC | #4
On Thu, 16 Nov 2023 at 11:14, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>
> Hi,
>
> Thanks a lot for reviewing!
>
>
> On 2023/11/15 00:30, Dmitry Baryshkov wrote:
> > On Tue, 14 Nov 2023 at 17:09, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
> >> From: Sui Jingfeng <suijingfeng@loongson.cn>
> >>
> >> The it66121_create_bridge() and it66121_destroy_bridge() are added to
> >> export the core functionalities. Create a connector manually by using
> >> bridge connector helpers when link as a lib.
> >>
> >> Signed-off-by: Sui Jingfeng <suijingfeng@loongson.cn>
> >> ---
> >>   drivers/gpu/drm/bridge/ite-it66121.c | 134 +++++++++++++++++++--------
> >>   include/drm/bridge/ite-it66121.h     |  17 ++++
> >>   2 files changed, 113 insertions(+), 38 deletions(-)
> >>   create mode 100644 include/drm/bridge/ite-it66121.h
> >>
> >> diff --git a/drivers/gpu/drm/bridge/ite-it66121.c b/drivers/gpu/drm/bridge/ite-it66121.c
> >> index 8971414a2a60..f5968b679c5d 100644
> >> --- a/drivers/gpu/drm/bridge/ite-it66121.c
> >> +++ b/drivers/gpu/drm/bridge/ite-it66121.c
> >> @@ -22,6 +22,7 @@
> >>
> >>   #include <drm/drm_atomic_helper.h>
> >>   #include <drm/drm_bridge.h>
> >> +#include <drm/drm_bridge_connector.h>
> >>   #include <drm/drm_edid.h>
> >>   #include <drm/drm_modes.h>
> >>   #include <drm/drm_print.h>
> >> @@ -703,14 +704,32 @@ static int it66121_bridge_attach(struct drm_bridge *bridge,
> >>                                   enum drm_bridge_attach_flags flags)
> >>   {
> >>          struct it66121_ctx *ctx = bridge_to_it66121(bridge);
> >> +       struct drm_bridge *next_bridge = ctx->next_bridge;
> >> +       struct drm_encoder *encoder = bridge->encoder;
> >>          int ret;
> >>
> >> -       if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR))
> >> -               return -EINVAL;
> >> +       if (next_bridge) {
> >> +               if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)) {
> >> +                       WARN_ON(1);
> > Why? At least use WARN() instead
>
> Originally I want to
>
>
>
>
> >> +                       flags |= DRM_BRIDGE_ATTACH_NO_CONNECTOR;
> >> +               }
> >> +               ret = drm_bridge_attach(encoder, next_bridge, bridge, flags);
> >> +               if (ret)
> >> +                       return ret;
> >> +       } else {
> >> +               struct drm_connector *connector;
> >>
> >> -       ret = drm_bridge_attach(bridge->encoder, ctx->next_bridge, bridge, flags);
> >> -       if (ret)
> >> -               return ret;
> >> +               if (flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)
> >> +                       WARN_ON(1);
> > No. It is perfectly fine to create attach a bridge with no next_bridge
> > and with the DRM_BRIDGE_ATTACH_NO_CONNECTOR flag.
> >
>
> The document say when DRM_BRIDGE_ATTACH_NO_CONNECTOR flag is set
> the bridge shall not create a drm_connector. So I think if a display
> bridge driver don't have a next bridge attached (Currently, this is
> told by the DT), it says that this is a non-DT environment. On such
> a case, this display bridge driver(it66121.ko) should behavior like
> a *agent*. Because the upstream port of it66121 is the DVO port of
> the display controller, the downstream port of it66121 is the HDMI
> connector. it66121 is on the middle. So I think the it66121.ko should
> handle all of troubles on behalf of the display controller drivers.

No. Don't make decisions for the other drivers. They might have different needs.

> Therefore (when in non-DT use case), the display controller drivers
> side should not set DRM_BRIDGE_ATTACH_NO_CONNECTOR flag anymore.
> Which is to hint that the it66121 should totally in charge of those
> tasks (either by using bridge connector helper or create a connector
> manually). I don't understand on such a case, why bother display
> controller drivers anymore.

This is the reason why we had introduced this flag. It allows the
driver to customise the connector. It even allows the driver to
implement a connector on its own, completely ignoring the
drm_bridge_connector.

>
>
> >> +
> >> +               connector = drm_bridge_connector_init(bridge->dev, encoder);
> >> +               if (IS_ERR(connector))
> >> +                       return PTR_ERR(connector);
> >> +
> >> +               drm_connector_attach_encoder(connector, encoder);
> > This goes into your device driver.
> >
> >> +
> >> +               ctx->connector = connector;
> >> +       }
> >>
> >>          if (ctx->info->id == ID_IT66121) {
> >>                  ret = regmap_write_bits(ctx->regmap, IT66121_CLK_BANK_REG,
> >> @@ -1632,16 +1651,13 @@ static const char * const it66121_supplies[] = {
> >>          "vcn33", "vcn18", "vrf12"
> >>   };
> >
Sui Jingfeng Nov. 16, 2023, 10:12 a.m. UTC | #5
Hi,


On 2023/11/16 17:30, Dmitry Baryshkov wrote:
> On Thu, 16 Nov 2023 at 11:14, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>> Hi,
>>
>> Thanks a lot for reviewing!
>>
>>
>> On 2023/11/15 00:30, Dmitry Baryshkov wrote:
>>> On Tue, 14 Nov 2023 at 17:09, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>>>> From: Sui Jingfeng <suijingfeng@loongson.cn>
>>>>
>>>> The it66121_create_bridge() and it66121_destroy_bridge() are added to
>>>> export the core functionalities. Create a connector manually by using
>>>> bridge connector helpers when link as a lib.
>>>>
>>>> Signed-off-by: Sui Jingfeng <suijingfeng@loongson.cn>
>>>> ---
>>>>    drivers/gpu/drm/bridge/ite-it66121.c | 134 +++++++++++++++++++--------
>>>>    include/drm/bridge/ite-it66121.h     |  17 ++++
>>>>    2 files changed, 113 insertions(+), 38 deletions(-)
>>>>    create mode 100644 include/drm/bridge/ite-it66121.h
>>>>
>>>> diff --git a/drivers/gpu/drm/bridge/ite-it66121.c b/drivers/gpu/drm/bridge/ite-it66121.c
>>>> index 8971414a2a60..f5968b679c5d 100644
>>>> --- a/drivers/gpu/drm/bridge/ite-it66121.c
>>>> +++ b/drivers/gpu/drm/bridge/ite-it66121.c
>>>> @@ -22,6 +22,7 @@
>>>>
>>>>    #include <drm/drm_atomic_helper.h>
>>>>    #include <drm/drm_bridge.h>
>>>> +#include <drm/drm_bridge_connector.h>
>>>>    #include <drm/drm_edid.h>
>>>>    #include <drm/drm_modes.h>
>>>>    #include <drm/drm_print.h>
>>>> @@ -703,14 +704,32 @@ static int it66121_bridge_attach(struct drm_bridge *bridge,
>>>>                                    enum drm_bridge_attach_flags flags)
>>>>    {
>>>>           struct it66121_ctx *ctx = bridge_to_it66121(bridge);
>>>> +       struct drm_bridge *next_bridge = ctx->next_bridge;
>>>> +       struct drm_encoder *encoder = bridge->encoder;
>>>>           int ret;
>>>>
>>>> -       if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR))
>>>> -               return -EINVAL;
>>>> +       if (next_bridge) {
>>>> +               if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)) {
>>>> +                       WARN_ON(1);
>>> Why? At least use WARN() instead
>> Originally I want to
>>
>>
>>
>>
>>>> +                       flags |= DRM_BRIDGE_ATTACH_NO_CONNECTOR;
>>>> +               }
>>>> +               ret = drm_bridge_attach(encoder, next_bridge, bridge, flags);
>>>> +               if (ret)
>>>> +                       return ret;
>>>> +       } else {
>>>> +               struct drm_connector *connector;
>>>>
>>>> -       ret = drm_bridge_attach(bridge->encoder, ctx->next_bridge, bridge, flags);
>>>> -       if (ret)
>>>> -               return ret;
>>>> +               if (flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)
>>>> +                       WARN_ON(1);
>>> No. It is perfectly fine to create attach a bridge with no next_bridge
>>> and with the DRM_BRIDGE_ATTACH_NO_CONNECTOR flag.
>>>
>> The document say when DRM_BRIDGE_ATTACH_NO_CONNECTOR flag is set
>> the bridge shall not create a drm_connector. So I think if a display
>> bridge driver don't have a next bridge attached (Currently, this is
>> told by the DT), it says that this is a non-DT environment. On such
>> a case, this display bridge driver(it66121.ko) should behavior like
>> a *agent*. Because the upstream port of it66121 is the DVO port of
>> the display controller, the downstream port of it66121 is the HDMI
>> connector. it66121 is on the middle. So I think the it66121.ko should
>> handle all of troubles on behalf of the display controller drivers.
> No. Don't make decisions for the other drivers. They might have different needs.

[...]


>
>> Therefore (when in non-DT use case), the display controller drivers
>> side should not set DRM_BRIDGE_ATTACH_NO_CONNECTOR flag anymore.
>> Which is to hint that the it66121 should totally in charge of those
>> tasks (either by using bridge connector helper or create a connector
>> manually). I don't understand on such a case, why bother display
>> controller drivers anymore.
> This is the reason why we had introduced this flag. It allows the
> driver to customise the connector. It even allows the driver to
> implement a connector on its own, completely ignoring the
> drm_bridge_connector.


I know what you said is right in the sense of the universe cases,
but I think the most frequent(majority) use case is that there is
only one display bridge on the middle. Therefore, I don't want to
movethe connector things into device driver if there is only one display 
bridge(say it66121) in the middle. After all, there is no *direct 
physical connection* from the perspective of the hardware. I means that 
there is no hardware wires connectthe HDMI connector and the DVO port. So display controller drivers 
should not interact with anything related with the connector on a 
perfect abstract on the software side. Especially for such a simple use 
case. It probably make senses to make a  decision for themost frequently use case, please also note
that this patch didn't introduce any-restriction for the more advance
uses cases(multiple bridges in the middle).


>>
>>>> +
>>>> +               connector = drm_bridge_connector_init(bridge->dev, encoder);
>>>> +               if (IS_ERR(connector))
>>>> +                       return PTR_ERR(connector);
>>>> +
>>>> +               drm_connector_attach_encoder(connector, encoder);
>>> This goes into your device driver.
>>>
>>>> +
>>>> +               ctx->connector = connector;
>>>> +       }
>>>>
>>>>           if (ctx->info->id == ID_IT66121) {
>>>>                   ret = regmap_write_bits(ctx->regmap, IT66121_CLK_BANK_REG,
>>>> @@ -1632,16 +1651,13 @@ static const char * const it66121_supplies[] = {
>>>>           "vcn33", "vcn18", "vrf12"
>>>>    };
>
>
Sui Jingfeng Nov. 16, 2023, 10:29 a.m. UTC | #6
Hi,


On 2023/11/15 00:30, Dmitry Baryshkov wrote:
>> @@ -703,14 +704,32 @@ static int it66121_bridge_attach(struct drm_bridge *bridge,
>>                                   enum drm_bridge_attach_flags flags)
>>   {
>>          struct it66121_ctx *ctx = bridge_to_it66121(bridge);
>> +       struct drm_bridge *next_bridge = ctx->next_bridge;
>> +       struct drm_encoder *encoder = bridge->encoder;
>>          int ret;
>>
>> -       if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR))
>> -               return -EINVAL;
>> +       if (next_bridge) {
>> +               if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)) {
>> +                       WARN_ON(1);
> Why? At least use WARN() instead

If (next_bridge) is true, it says that the driver *already* known that
it66121 have a next bridges attached. Then it66121 driver should certainly
attach it, no matter what it is. Either a connector or another display bridge.
It also says that this is a DT-based system on such a case. CallingWARN_ON(1) here helps to see(print) which DC driver is doing the wired 
things. Ok, I will remove the WARN_ON(1)  on the next version.
Dmitry Baryshkov Nov. 16, 2023, 11:11 a.m. UTC | #7
On Thu, 16 Nov 2023 at 12:29, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>
> Hi,
>
>
> On 2023/11/15 00:30, Dmitry Baryshkov wrote:
> >> @@ -703,14 +704,32 @@ static int it66121_bridge_attach(struct drm_bridge *bridge,
> >>                                   enum drm_bridge_attach_flags flags)
> >>   {
> >>          struct it66121_ctx *ctx = bridge_to_it66121(bridge);
> >> +       struct drm_bridge *next_bridge = ctx->next_bridge;
> >> +       struct drm_encoder *encoder = bridge->encoder;
> >>          int ret;
> >>
> >> -       if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR))
> >> -               return -EINVAL;
> >> +       if (next_bridge) {
> >> +               if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)) {
> >> +                       WARN_ON(1);
> > Why? At least use WARN() instead
>
> If (next_bridge) is true, it says that the driver *already* known that
> it66121 have a next bridges attached. Then it66121 driver should certainly
> attach it, no matter what it is. Either a connector or another display bridge.
> It also says that this is a DT-based system on such a case. CallingWARN_ON(1) here helps to see(print) which DC driver is doing the wired
> things. Ok, I will remove the WARN_ON(1)  on the next version.

That's why I pointed you to WARN(). WARN_ON(1) gives no information to
the user. WARN() allows you to add a message.
Jani Nikula Nov. 16, 2023, 11:15 a.m. UTC | #8
On Thu, 16 Nov 2023, Dmitry Baryshkov <dmitry.baryshkov@linaro.org> wrote:
> On Thu, 16 Nov 2023 at 12:29, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>>
>> Hi,
>>
>>
>> On 2023/11/15 00:30, Dmitry Baryshkov wrote:
>> >> @@ -703,14 +704,32 @@ static int it66121_bridge_attach(struct drm_bridge *bridge,
>> >>                                   enum drm_bridge_attach_flags flags)
>> >>   {
>> >>          struct it66121_ctx *ctx = bridge_to_it66121(bridge);
>> >> +       struct drm_bridge *next_bridge = ctx->next_bridge;
>> >> +       struct drm_encoder *encoder = bridge->encoder;
>> >>          int ret;
>> >>
>> >> -       if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR))
>> >> -               return -EINVAL;
>> >> +       if (next_bridge) {
>> >> +               if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)) {
>> >> +                       WARN_ON(1);
>> > Why? At least use WARN() instead
>>
>> If (next_bridge) is true, it says that the driver *already* known that
>> it66121 have a next bridges attached. Then it66121 driver should certainly
>> attach it, no matter what it is. Either a connector or another display bridge.
>> It also says that this is a DT-based system on such a case. CallingWARN_ON(1) here helps to see(print) which DC driver is doing the wired
>> things. Ok, I will remove the WARN_ON(1)  on the next version.
>
> That's why I pointed you to WARN(). WARN_ON(1) gives no information to
> the user. WARN() allows you to add a message.

Please use drm_WARN* while at it.

BR,
Jani.
Sui Jingfeng Nov. 16, 2023, 11:18 a.m. UTC | #9
Hi,


On 2023/11/15 00:30, Dmitry Baryshkov wrote:
>> +
>> +               ctx->connector = connector;
>> +       }
>>
>>          if (ctx->info->id == ID_IT66121) {
>>                  ret = regmap_write_bits(ctx->regmap, IT66121_CLK_BANK_REG,
>> @@ -1632,16 +1651,13 @@ static const char * const it66121_supplies[] = {
>>          "vcn33", "vcn18", "vrf12"
>>   };
>>
>> -static int it66121_probe(struct i2c_client *client)
>> +int it66121_create_bridge(struct i2c_client *client, bool of_support,
>> +                         bool hpd_support, bool audio_support,
>> +                         struct drm_bridge **bridge)
>>   {
>> +       struct device *dev = &client->dev;
>>          int ret;
>>          struct it66121_ctx *ctx;
>> -       struct device *dev = &client->dev;
>> -
>> -       if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) {
>> -               dev_err(dev, "I2C check functionality failed.\n");
>> -               return -ENXIO;
>> -       }
>>
>>          ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
>>          if (!ctx)
>> @@ -1649,24 +1665,19 @@ static int it66121_probe(struct i2c_client *client)
>>
>>          ctx->dev = dev;
>>          ctx->client = client;
>> -       ctx->info = i2c_get_match_data(client);
>> -
>> -       ret = it66121_of_read_bus_width(dev, &ctx->bus_width);
>> -       if (ret)
>> -               return ret;
>> -
>> -       ret = it66121_of_get_next_bridge(dev, &ctx->next_bridge);
>> -       if (ret)
>> -               return ret;
>> -
>> -       i2c_set_clientdata(client, ctx);
>>          mutex_init(&ctx->lock);
>>
>> -       ret = devm_regulator_bulk_get_enable(dev, ARRAY_SIZE(it66121_supplies),
>> -                                            it66121_supplies);
>> -       if (ret) {
>> -               dev_err(dev, "Failed to enable power supplies\n");
>> -               return ret;
>> +       if (of_support) {
>> +               ret = it66121_of_read_bus_width(dev, &ctx->bus_width);
>> +               if (ret)
>> +                       return ret;
>> +
>> +               ret = it66121_of_get_next_bridge(dev, &ctx->next_bridge);
>> +               if (ret)
>> +                       return ret;
>> +       } else {
>> +               ctx->bus_width = 24;
>> +               ctx->next_bridge = NULL;
>>          }
> A better alternative would be to turn OF calls into fwnode calls and
> to populate the fwnode properties. See
> drivers/platform/x86/intel/chtwc_int33fe.c for example.


Honestly, I don't want to leave any scratch(breadcrumbs).
I'm worries about that turn OF calls into fwnode calls will leave something unwanted.

Because I am not sure if fwnode calls will make sense in the DT world, while my patch
*still* be useful in the DT world. Because the newly introduced it66121_create_bridge()
function is a core. I think It's better leave this task to a more advance programmer.
if there have use case. It can be introduced at a latter time, probably parallel with
the DT.

I think DT and/or ACPI is best for integrated devices, but it66121 display bridges is
a i2c slave device. Personally, I think slave device shouldn't be standalone. I'm more
prefer to turn this driver to support hot-plug, even remove the device on the run time
freely when detach and allow reattach. Like the I2C EEPROM device in the monitor (which
contains the EDID, with I2C slave address 0x50). The I2C EEPROM device *also* don't has
a corresponding struct device representation in linux kernel.

so I still think It is best to make this drivers functional as a static lib, but I want
to hear you to say more. Why it would be a *better* alternative to turn OF calls into
fwnode calls? what are the potential benefits?
Dmitry Baryshkov Nov. 16, 2023, 11:19 a.m. UTC | #10
On Thu, 16 Nov 2023 at 12:13, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>
> Hi,
>
>
> On 2023/11/16 17:30, Dmitry Baryshkov wrote:
> > On Thu, 16 Nov 2023 at 11:14, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
> >> Hi,
> >>
> >> Thanks a lot for reviewing!
> >>
> >>
> >> On 2023/11/15 00:30, Dmitry Baryshkov wrote:
> >>> On Tue, 14 Nov 2023 at 17:09, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
> >>>> From: Sui Jingfeng <suijingfeng@loongson.cn>
> >>>>
> >>>> The it66121_create_bridge() and it66121_destroy_bridge() are added to
> >>>> export the core functionalities. Create a connector manually by using
> >>>> bridge connector helpers when link as a lib.
> >>>>
> >>>> Signed-off-by: Sui Jingfeng <suijingfeng@loongson.cn>
> >>>> ---
> >>>>    drivers/gpu/drm/bridge/ite-it66121.c | 134 +++++++++++++++++++--------
> >>>>    include/drm/bridge/ite-it66121.h     |  17 ++++
> >>>>    2 files changed, 113 insertions(+), 38 deletions(-)
> >>>>    create mode 100644 include/drm/bridge/ite-it66121.h
> >>>>
> >>>> diff --git a/drivers/gpu/drm/bridge/ite-it66121.c b/drivers/gpu/drm/bridge/ite-it66121.c
> >>>> index 8971414a2a60..f5968b679c5d 100644
> >>>> --- a/drivers/gpu/drm/bridge/ite-it66121.c
> >>>> +++ b/drivers/gpu/drm/bridge/ite-it66121.c
> >>>> @@ -22,6 +22,7 @@
> >>>>
> >>>>    #include <drm/drm_atomic_helper.h>
> >>>>    #include <drm/drm_bridge.h>
> >>>> +#include <drm/drm_bridge_connector.h>
> >>>>    #include <drm/drm_edid.h>
> >>>>    #include <drm/drm_modes.h>
> >>>>    #include <drm/drm_print.h>
> >>>> @@ -703,14 +704,32 @@ static int it66121_bridge_attach(struct drm_bridge *bridge,
> >>>>                                    enum drm_bridge_attach_flags flags)
> >>>>    {
> >>>>           struct it66121_ctx *ctx = bridge_to_it66121(bridge);
> >>>> +       struct drm_bridge *next_bridge = ctx->next_bridge;
> >>>> +       struct drm_encoder *encoder = bridge->encoder;
> >>>>           int ret;
> >>>>
> >>>> -       if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR))
> >>>> -               return -EINVAL;
> >>>> +       if (next_bridge) {
> >>>> +               if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)) {
> >>>> +                       WARN_ON(1);
> >>> Why? At least use WARN() instead
> >> Originally I want to
> >>
> >>
> >>
> >>
> >>>> +                       flags |= DRM_BRIDGE_ATTACH_NO_CONNECTOR;
> >>>> +               }
> >>>> +               ret = drm_bridge_attach(encoder, next_bridge, bridge, flags);
> >>>> +               if (ret)
> >>>> +                       return ret;
> >>>> +       } else {
> >>>> +               struct drm_connector *connector;
> >>>>
> >>>> -       ret = drm_bridge_attach(bridge->encoder, ctx->next_bridge, bridge, flags);
> >>>> -       if (ret)
> >>>> -               return ret;
> >>>> +               if (flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)
> >>>> +                       WARN_ON(1);
> >>> No. It is perfectly fine to create attach a bridge with no next_bridge
> >>> and with the DRM_BRIDGE_ATTACH_NO_CONNECTOR flag.
> >>>
> >> The document say when DRM_BRIDGE_ATTACH_NO_CONNECTOR flag is set
> >> the bridge shall not create a drm_connector. So I think if a display
> >> bridge driver don't have a next bridge attached (Currently, this is
> >> told by the DT), it says that this is a non-DT environment. On such
> >> a case, this display bridge driver(it66121.ko) should behavior like
> >> a *agent*. Because the upstream port of it66121 is the DVO port of
> >> the display controller, the downstream port of it66121 is the HDMI
> >> connector. it66121 is on the middle. So I think the it66121.ko should
> >> handle all of troubles on behalf of the display controller drivers.
> > No. Don't make decisions for the other drivers. They might have different needs.
>
> [...]
>
>
> >
> >> Therefore (when in non-DT use case), the display controller drivers
> >> side should not set DRM_BRIDGE_ATTACH_NO_CONNECTOR flag anymore.
> >> Which is to hint that the it66121 should totally in charge of those
> >> tasks (either by using bridge connector helper or create a connector
> >> manually). I don't understand on such a case, why bother display
> >> controller drivers anymore.
> > This is the reason why we had introduced this flag. It allows the
> > driver to customise the connector. It even allows the driver to
> > implement a connector on its own, completely ignoring the
> > drm_bridge_connector.
>
>
> I know what you said is right in the sense of the universe cases,
> but I think the most frequent(majority) use case is that there is
> only one display bridge on the middle. Therefore, I don't want to
> movethe connector things into device driver if there is only one display
> bridge(say it66121) in the middle. After all, there is no *direct
> physical connection* from the perspective of the hardware. I means that
> there is no hardware wires connectthe HDMI connector and the DVO port. So display controller drivers
> should not interact with anything related with the connector on a
> perfect abstract on the software side. Especially for such a simple use
> case. It probably make senses to make a  decision for themost frequently use case, please also note
> that this patch didn't introduce any-restriction for the more advance
> uses cases(multiple bridges in the middle).

So, for the sake of not having the connector in the display driver,
you want to add boilerplate code basically to each and every bridge
driver. In the end, they should all behave in the same way.

Moreover, there is no way this implementation can work without a
warning if there are two bridges in a chain and the it66121 is the
second (the last) one. The host can not specify the
DRM_BRIDGE_ATTACH_NO_CONNECTOR)
> >>>> +                       WARN_ON(1);
> >>> No. It is perfectly fine to create attach a bridge with no next_bridge
> >>> and with the DRM_BRIDGE_ATTACH_NO_CONNECTOR flag.
> >>>
> >> The document say when DRM_BRIDGE_ATTACH_NO_CONNECTOR flag is set
> >> the bridge shall not create a drm_connector. So I think if a display
> >> bridge driver don't have a next bridge attached (Currently, this is
> >> told by the DT), it says that this is a non-DT environment. On such
> >> a case, this display bridge driver(it66121.ko) should behavior like
> >> a *agent*. Because the upstream port of it66121 is the DVO port of
> >> the display controller, the downstream port of it66121 is the HDMI
> >> connector. it66121 is on the middle. So I think the it66121.ko should
> >> handle all of troubles on behalf of the display controller drivers.
> > No. Don't make decisions for the other drivers. They might have different needs.
>
> [...]
>
>
> >
> >> Therefore (when in non-DT use case), the display controller drivers
> >> side should not set DRM_BRIDGE_ATTACH_NO_CONNECTOR flag anymore.
> >> Which is to hint that the it66121 should totally in charge of those
> >> tasks (either by using bridge connector helper or create a connector
> >> manually). I don't understand on such a case, why bother display
> >> controller drivers anymore.
> > This is the reason why we had introduced this flag. It allows the
> > driver to customise the connector. It even allows the driver to
> > implement a connector on its own, completely ignoring the
> > drm_bridge_connector.
>
>
> I know what you said is right in the sense of the universe cases,
> but I think the most frequent(majority) use case is that there is
> only one display bridge on the middle. Therefore, I don't want to
> movethe connector things into device driver if there is only one display
> bridge(say it66121) in the middle. After all, there is no *direct
> physical connection* from the perspective of the hardware. I means that
> there is no hardware wires connectthe HDMI connector and the DVO port. So display controller drivers
> should not interact with anything related with the connector on a
> perfect abstract on the software side. Especially for such a simple use
> case. It probably make senses to make a  decision for themost frequently use case, please also note
> that this patch didn't introduce any-restriction for the more advance
> uses cases(multiple bridges in the middle).

So, for the sake of not having the connector in the display driver,
you want to add boilerplate code basically to each and every bridge
driver. In the end, they should all behave in the same way.

Moreover, there is no way this implementation can work without a
warning if there are two bridges in a chain and the it66121 is the
second (the last) one. The host can not specify the
DRM_BRIDGE_ATTACH_NO_CONNECTOR flag, it will cause a warning here. And
it can not omit the flag (as otherwise the first bridge will create a
connector, without consulting the second bridge).

>
>
> >>
> >>>> +
> >>>> +               connector = drm_bridge_connector_init(bridge->dev, encoder);
> >>>> +               if (IS_ERR(connector))
> >>>> +                       return PTR_ERR(connector);
> >>>> +
> >>>> +               drm_connector_attach_encoder(connector, encoder);
> >>> This goes into your device driver.
> >>>
> >>>> +
> >>>> +               ctx->connector = connector;
> >>>> +       }
> >>>>
> >>>>           if (ctx->info->id == ID_IT66121) {
> >>>>                   ret = regmap_write_bits(ctx->regmap, IT66121_CLK_BANK_REG,
> >>>> @@ -1632,16 +1651,13 @@ static const char * const it66121_supplies[] = {
> >>>>           "vcn33", "vcn18", "vrf12"
> >>>>    };
> >
> >
Dmitry Baryshkov Nov. 16, 2023, 11:29 a.m. UTC | #11
On Thu, 16 Nov 2023 at 13:18, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>
> Hi,
>
>
> On 2023/11/15 00:30, Dmitry Baryshkov wrote:
> >> +
> >> +               ctx->connector = connector;
> >> +       }
> >>
> >>          if (ctx->info->id == ID_IT66121) {
> >>                  ret = regmap_write_bits(ctx->regmap, IT66121_CLK_BANK_REG,
> >> @@ -1632,16 +1651,13 @@ static const char * const it66121_supplies[] = {
> >>          "vcn33", "vcn18", "vrf12"
> >>   };
> >>
> >> -static int it66121_probe(struct i2c_client *client)
> >> +int it66121_create_bridge(struct i2c_client *client, bool of_support,
> >> +                         bool hpd_support, bool audio_support,
> >> +                         struct drm_bridge **bridge)
> >>   {
> >> +       struct device *dev = &client->dev;
> >>          int ret;
> >>          struct it66121_ctx *ctx;
> >> -       struct device *dev = &client->dev;
> >> -
> >> -       if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) {
> >> -               dev_err(dev, "I2C check functionality failed.\n");
> >> -               return -ENXIO;
> >> -       }
> >>
> >>          ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
> >>          if (!ctx)
> >> @@ -1649,24 +1665,19 @@ static int it66121_probe(struct i2c_client *client)
> >>
> >>          ctx->dev = dev;
> >>          ctx->client = client;
> >> -       ctx->info = i2c_get_match_data(client);
> >> -
> >> -       ret = it66121_of_read_bus_width(dev, &ctx->bus_width);
> >> -       if (ret)
> >> -               return ret;
> >> -
> >> -       ret = it66121_of_get_next_bridge(dev, &ctx->next_bridge);
> >> -       if (ret)
> >> -               return ret;
> >> -
> >> -       i2c_set_clientdata(client, ctx);
> >>          mutex_init(&ctx->lock);
> >>
> >> -       ret = devm_regulator_bulk_get_enable(dev, ARRAY_SIZE(it66121_supplies),
> >> -                                            it66121_supplies);
> >> -       if (ret) {
> >> -               dev_err(dev, "Failed to enable power supplies\n");
> >> -               return ret;
> >> +       if (of_support) {
> >> +               ret = it66121_of_read_bus_width(dev, &ctx->bus_width);
> >> +               if (ret)
> >> +                       return ret;
> >> +
> >> +               ret = it66121_of_get_next_bridge(dev, &ctx->next_bridge);
> >> +               if (ret)
> >> +                       return ret;
> >> +       } else {
> >> +               ctx->bus_width = 24;
> >> +               ctx->next_bridge = NULL;
> >>          }
> > A better alternative would be to turn OF calls into fwnode calls and
> > to populate the fwnode properties. See
> > drivers/platform/x86/intel/chtwc_int33fe.c for example.
>
>
> Honestly, I don't want to leave any scratch(breadcrumbs).
> I'm worries about that turn OF calls into fwnode calls will leave something unwanted.
>
> Because I am not sure if fwnode calls will make sense in the DT world, while my patch
> *still* be useful in the DT world.

fwnode calls work for both DT and non-DT cases. In the DT case they
work with DT nodes and properties. In the non-DT case, they work with
manually populated properties.

> Because the newly introduced it66121_create_bridge()
> function is a core. I think It's better leave this task to a more advance programmer.
> if there have use case. It can be introduced at a latter time, probably parallel with
> the DT.
>
> I think DT and/or ACPI is best for integrated devices, but it66121 display bridges is
> a i2c slave device. Personally, I think slave device shouldn't be standalone. I'm more
> prefer to turn this driver to support hot-plug, even remove the device on the run time
> freely when detach and allow reattach. Like the I2C EEPROM device in the monitor (which
> contains the EDID, with I2C slave address 0x50). The I2C EEPROM device *also* don't has
> a corresponding struct device representation in linux kernel.

It has. See i2c_client::dev.

> so I still think It is best to make this drivers functional as a static lib, but I want
> to hear you to say more. Why it would be a *better* alternative to turn OF calls into
> fwnode calls? what are the potential benefits?

Because then you can populate device properties from your root device.
Because it allows the platform to specify the bus width instead of
hardcoding 24 bits (which might work in your case, but might not be
applicable to another user next week).

Anyway, even without fwnode, I'd strongly suggest you to drop the
it66121_create_bridge() as it is now and start by populating the i2c
bus from your root device. Then you will need some way (fwnode?) to
discover the bridge chain. And at the last point you will get into the
device data and/or properties business.
Sui Jingfeng Nov. 16, 2023, 11:53 a.m. UTC | #12
Hi,


On 2023/11/16 19:29, Dmitry Baryshkov wrote:
> On Thu, 16 Nov 2023 at 13:18, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>> Hi,
>>
>>
>> On 2023/11/15 00:30, Dmitry Baryshkov wrote:
>>>> +
>>>> +               ctx->connector = connector;
>>>> +       }
>>>>
>>>>           if (ctx->info->id == ID_IT66121) {
>>>>                   ret = regmap_write_bits(ctx->regmap, IT66121_CLK_BANK_REG,
>>>> @@ -1632,16 +1651,13 @@ static const char * const it66121_supplies[] = {
>>>>           "vcn33", "vcn18", "vrf12"
>>>>    };
>>>>
>>>> -static int it66121_probe(struct i2c_client *client)
>>>> +int it66121_create_bridge(struct i2c_client *client, bool of_support,
>>>> +                         bool hpd_support, bool audio_support,
>>>> +                         struct drm_bridge **bridge)
>>>>    {
>>>> +       struct device *dev = &client->dev;
>>>>           int ret;
>>>>           struct it66121_ctx *ctx;
>>>> -       struct device *dev = &client->dev;
>>>> -
>>>> -       if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) {
>>>> -               dev_err(dev, "I2C check functionality failed.\n");
>>>> -               return -ENXIO;
>>>> -       }
>>>>
>>>>           ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
>>>>           if (!ctx)
>>>> @@ -1649,24 +1665,19 @@ static int it66121_probe(struct i2c_client *client)
>>>>
>>>>           ctx->dev = dev;
>>>>           ctx->client = client;
>>>> -       ctx->info = i2c_get_match_data(client);
>>>> -
>>>> -       ret = it66121_of_read_bus_width(dev, &ctx->bus_width);
>>>> -       if (ret)
>>>> -               return ret;
>>>> -
>>>> -       ret = it66121_of_get_next_bridge(dev, &ctx->next_bridge);
>>>> -       if (ret)
>>>> -               return ret;
>>>> -
>>>> -       i2c_set_clientdata(client, ctx);
>>>>           mutex_init(&ctx->lock);
>>>>
>>>> -       ret = devm_regulator_bulk_get_enable(dev, ARRAY_SIZE(it66121_supplies),
>>>> -                                            it66121_supplies);
>>>> -       if (ret) {
>>>> -               dev_err(dev, "Failed to enable power supplies\n");
>>>> -               return ret;
>>>> +       if (of_support) {
>>>> +               ret = it66121_of_read_bus_width(dev, &ctx->bus_width);
>>>> +               if (ret)
>>>> +                       return ret;
>>>> +
>>>> +               ret = it66121_of_get_next_bridge(dev, &ctx->next_bridge);
>>>> +               if (ret)
>>>> +                       return ret;
>>>> +       } else {
>>>> +               ctx->bus_width = 24;
>>>> +               ctx->next_bridge = NULL;
>>>>           }
>>> A better alternative would be to turn OF calls into fwnode calls and
>>> to populate the fwnode properties. See
>>> drivers/platform/x86/intel/chtwc_int33fe.c for example.
>>
>> Honestly, I don't want to leave any scratch(breadcrumbs).
>> I'm worries about that turn OF calls into fwnode calls will leave something unwanted.
>>
>> Because I am not sure if fwnode calls will make sense in the DT world, while my patch
>> *still* be useful in the DT world.
> fwnode calls work for both DT and non-DT cases. In the DT case they
> work with DT nodes and properties. In the non-DT case, they work with
> manually populated properties.
>
>> Because the newly introduced it66121_create_bridge()
>> function is a core. I think It's better leave this task to a more advance programmer.
>> if there have use case. It can be introduced at a latter time, probably parallel with
>> the DT.
>>
>> I think DT and/or ACPI is best for integrated devices, but it66121 display bridges is
>> a i2c slave device. Personally, I think slave device shouldn't be standalone. I'm more
>> prefer to turn this driver to support hot-plug, even remove the device on the run time
>> freely when detach and allow reattach. Like the I2C EEPROM device in the monitor (which
>> contains the EDID, with I2C slave address 0x50). The I2C EEPROM device *also* don't has
>> a corresponding struct device representation in linux kernel.
> It has. See i2c_client::dev.

No, what I mean is that there don't have a device driver for monitor(display) hardware entity.
And the drm_do_probe_ddc_edid() is the static linked driver, which is similar with the idea
this series want to express.


>> so I still think It is best to make this drivers functional as a static lib, but I want
>> to hear you to say more. Why it would be a *better* alternative to turn OF calls into
>> fwnode calls? what are the potential benefits?
> Because then you can populate device properties from your root device.
> Because it allows the platform to specify the bus width instead of
> hardcoding 24 bits (which might work in your case, but might not be
> applicable to another user next week).


No, this problem can be easily solved. Simply add another argument.

```

int it66121_create_bridge(struct i2c_client *client, bool of_support,
                           bool hpd_support, bool audio_support, u32 bus_width,
                           struct drm_bridge **bridge);
```


> Anyway, even without fwnode, I'd strongly suggest you to drop the
> it66121_create_bridge() as it is now and start by populating the i2c
> bus from your root device.

This will force all non-DT users to add the similar code patter at the display controller side,
which is another kind of duplication. The monitor is also as I2C slave device, can be abstract
as a identify drm bridges in theory, I guess.


> Then you will need some way (fwnode?) to
> discover the bridge chain. And at the last point you will get into the
> device data and/or properties business.
>
No, leave that chance to a more better programmer and forgive me please,
too difficult, I'm afraid of not able to solve. Thanks a lot for the trust!
Sui Jingfeng Nov. 16, 2023, 12:07 p.m. UTC | #13
On 2023/11/16 19:53, Sui Jingfeng wrote:
> Hi,
>
>
> On 2023/11/16 19:29, Dmitry Baryshkov wrote:
>> On Thu, 16 Nov 2023 at 13:18, Sui Jingfeng <sui.jingfeng@linux.dev> 
>> wrote:
>>> Hi,
>>>
>>>
>>> On 2023/11/15 00:30, Dmitry Baryshkov wrote:
>>>>> +
>>>>> +               ctx->connector = connector;
>>>>> +       }
>>>>>
>>>>>           if (ctx->info->id == ID_IT66121) {
>>>>>                   ret = regmap_write_bits(ctx->regmap, 
>>>>> IT66121_CLK_BANK_REG,
>>>>> @@ -1632,16 +1651,13 @@ static const char * const 
>>>>> it66121_supplies[] = {
>>>>>           "vcn33", "vcn18", "vrf12"
>>>>>    };
>>>>>
>>>>> -static int it66121_probe(struct i2c_client *client)
>>>>> +int it66121_create_bridge(struct i2c_client *client, bool 
>>>>> of_support,
>>>>> +                         bool hpd_support, bool audio_support,
>>>>> +                         struct drm_bridge **bridge)
>>>>>    {
>>>>> +       struct device *dev = &client->dev;
>>>>>           int ret;
>>>>>           struct it66121_ctx *ctx;
>>>>> -       struct device *dev = &client->dev;
>>>>> -
>>>>> -       if (!i2c_check_functionality(client->adapter, 
>>>>> I2C_FUNC_I2C)) {
>>>>> -               dev_err(dev, "I2C check functionality failed.\n");
>>>>> -               return -ENXIO;
>>>>> -       }
>>>>>
>>>>>           ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
>>>>>           if (!ctx)
>>>>> @@ -1649,24 +1665,19 @@ static int it66121_probe(struct i2c_client 
>>>>> *client)
>>>>>
>>>>>           ctx->dev = dev;
>>>>>           ctx->client = client;
>>>>> -       ctx->info = i2c_get_match_data(client);
>>>>> -
>>>>> -       ret = it66121_of_read_bus_width(dev, &ctx->bus_width);
>>>>> -       if (ret)
>>>>> -               return ret;
>>>>> -
>>>>> -       ret = it66121_of_get_next_bridge(dev, &ctx->next_bridge);
>>>>> -       if (ret)
>>>>> -               return ret;
>>>>> -
>>>>> -       i2c_set_clientdata(client, ctx);
>>>>>           mutex_init(&ctx->lock);
>>>>>
>>>>> -       ret = devm_regulator_bulk_get_enable(dev, 
>>>>> ARRAY_SIZE(it66121_supplies),
>>>>> - it66121_supplies);
>>>>> -       if (ret) {
>>>>> -               dev_err(dev, "Failed to enable power supplies\n");
>>>>> -               return ret;
>>>>> +       if (of_support) {
>>>>> +               ret = it66121_of_read_bus_width(dev, 
>>>>> &ctx->bus_width);
>>>>> +               if (ret)
>>>>> +                       return ret;
>>>>> +
>>>>> +               ret = it66121_of_get_next_bridge(dev, 
>>>>> &ctx->next_bridge);
>>>>> +               if (ret)
>>>>> +                       return ret;
>>>>> +       } else {
>>>>> +               ctx->bus_width = 24;
>>>>> +               ctx->next_bridge = NULL;
>>>>>           }
>>>> A better alternative would be to turn OF calls into fwnode calls and
>>>> to populate the fwnode properties. See
>>>> drivers/platform/x86/intel/chtwc_int33fe.c for example.
>>>
>>> Honestly, I don't want to leave any scratch(breadcrumbs).
>>> I'm worries about that turn OF calls into fwnode calls will leave 
>>> something unwanted.
>>>
>>> Because I am not sure if fwnode calls will make sense in the DT 
>>> world, while my patch
>>> *still* be useful in the DT world.
>> fwnode calls work for both DT and non-DT cases. In the DT case they
>> work with DT nodes and properties. In the non-DT case, they work with
>> manually populated properties.
>>
>>> Because the newly introduced it66121_create_bridge()
>>> function is a core. I think It's better leave this task to a more 
>>> advance programmer.
>>> if there have use case. It can be introduced at a latter time, 
>>> probably parallel with
>>> the DT.
>>>
>>> I think DT and/or ACPI is best for integrated devices, but it66121 
>>> display bridges is
>>> a i2c slave device. Personally, I think slave device shouldn't be 
>>> standalone. I'm more
>>> prefer to turn this driver to support hot-plug, even remove the 
>>> device on the run time
>>> freely when detach and allow reattach. Like the I2C EEPROM device in 
>>> the monitor (which
>>> contains the EDID, with I2C slave address 0x50). The I2C EEPROM 
>>> device *also* don't has
>>> a corresponding struct device representation in linux kernel.
>> It has. See i2c_client::dev.
>
> No, what I mean is that there don't have a device driver for 
> monitor(display) hardware entity.
> And the drm_do_probe_ddc_edid() is the static linked driver, which is 
> similar with the idea
> this series want to express.
>
>
>>> so I still think It is best to make this drivers functional as a 
>>> static lib, but I want
>>> to hear you to say more. Why it would be a *better* alternative to 
>>> turn OF calls into
>>> fwnode calls? what are the potential benefits?
>> Because then you can populate device properties from your root device.
>> Because it allows the platform to specify the bus width instead of
>> hardcoding 24 bits (which might work in your case, but might not be
>> applicable to another user next week).
>
>
> No, this problem can be easily solved. Simply add another argument.
>
> ```
>
> int it66121_create_bridge(struct i2c_client *client, bool of_support,
>                           bool hpd_support, bool audio_support, u32 
> bus_width,
>                           struct drm_bridge **bridge);
> ```
>
>
>> Anyway, even without fwnode, I'd strongly suggest you to drop the
>> it66121_create_bridge() as it is now and start by populating the i2c
>> bus from your root device.
>
> This will force all non-DT users to add the similar code patter at the 
> display controller side,
> which is another kind of duplication. The monitor is also as I2C slave 
> device, can be abstract
> as a identify drm bridges in theory, I guess.
>

'identify' -> 'identity'


>
>> Then you will need some way (fwnode?) to
>> discover the bridge chain. And at the last point you will get into the
>> device data and/or properties business.
>>
> No, leave that chance to a more better programmer and forgive me please,
> too difficult, I'm afraid of not able to solve. Thanks a lot for the 
> trust!
>
>
Dmitry Baryshkov Nov. 16, 2023, 3:23 p.m. UTC | #14
On Thu, 16 Nov 2023 at 14:08, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>
>
> On 2023/11/16 19:53, Sui Jingfeng wrote:
> > Hi,
> >
> >
> > On 2023/11/16 19:29, Dmitry Baryshkov wrote:
> >> On Thu, 16 Nov 2023 at 13:18, Sui Jingfeng <sui.jingfeng@linux.dev>
> >> wrote:
> >>> Hi,
> >>>
> >>>
> >>> On 2023/11/15 00:30, Dmitry Baryshkov wrote:
> >>>>> +
> >>>>> +               ctx->connector = connector;
> >>>>> +       }
> >>>>>
> >>>>>           if (ctx->info->id == ID_IT66121) {
> >>>>>                   ret = regmap_write_bits(ctx->regmap,
> >>>>> IT66121_CLK_BANK_REG,
> >>>>> @@ -1632,16 +1651,13 @@ static const char * const
> >>>>> it66121_supplies[] = {
> >>>>>           "vcn33", "vcn18", "vrf12"
> >>>>>    };
> >>>>>
> >>>>> -static int it66121_probe(struct i2c_client *client)
> >>>>> +int it66121_create_bridge(struct i2c_client *client, bool
> >>>>> of_support,
> >>>>> +                         bool hpd_support, bool audio_support,
> >>>>> +                         struct drm_bridge **bridge)
> >>>>>    {
> >>>>> +       struct device *dev = &client->dev;
> >>>>>           int ret;
> >>>>>           struct it66121_ctx *ctx;
> >>>>> -       struct device *dev = &client->dev;
> >>>>> -
> >>>>> -       if (!i2c_check_functionality(client->adapter,
> >>>>> I2C_FUNC_I2C)) {
> >>>>> -               dev_err(dev, "I2C check functionality failed.\n");
> >>>>> -               return -ENXIO;
> >>>>> -       }
> >>>>>
> >>>>>           ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
> >>>>>           if (!ctx)
> >>>>> @@ -1649,24 +1665,19 @@ static int it66121_probe(struct i2c_client
> >>>>> *client)
> >>>>>
> >>>>>           ctx->dev = dev;
> >>>>>           ctx->client = client;
> >>>>> -       ctx->info = i2c_get_match_data(client);
> >>>>> -
> >>>>> -       ret = it66121_of_read_bus_width(dev, &ctx->bus_width);
> >>>>> -       if (ret)
> >>>>> -               return ret;
> >>>>> -
> >>>>> -       ret = it66121_of_get_next_bridge(dev, &ctx->next_bridge);
> >>>>> -       if (ret)
> >>>>> -               return ret;
> >>>>> -
> >>>>> -       i2c_set_clientdata(client, ctx);
> >>>>>           mutex_init(&ctx->lock);
> >>>>>
> >>>>> -       ret = devm_regulator_bulk_get_enable(dev,
> >>>>> ARRAY_SIZE(it66121_supplies),
> >>>>> - it66121_supplies);
> >>>>> -       if (ret) {
> >>>>> -               dev_err(dev, "Failed to enable power supplies\n");
> >>>>> -               return ret;
> >>>>> +       if (of_support) {
> >>>>> +               ret = it66121_of_read_bus_width(dev,
> >>>>> &ctx->bus_width);
> >>>>> +               if (ret)
> >>>>> +                       return ret;
> >>>>> +
> >>>>> +               ret = it66121_of_get_next_bridge(dev,
> >>>>> &ctx->next_bridge);
> >>>>> +               if (ret)
> >>>>> +                       return ret;
> >>>>> +       } else {
> >>>>> +               ctx->bus_width = 24;
> >>>>> +               ctx->next_bridge = NULL;
> >>>>>           }
> >>>> A better alternative would be to turn OF calls into fwnode calls and
> >>>> to populate the fwnode properties. See
> >>>> drivers/platform/x86/intel/chtwc_int33fe.c for example.
> >>>
> >>> Honestly, I don't want to leave any scratch(breadcrumbs).
> >>> I'm worries about that turn OF calls into fwnode calls will leave
> >>> something unwanted.
> >>>
> >>> Because I am not sure if fwnode calls will make sense in the DT
> >>> world, while my patch
> >>> *still* be useful in the DT world.
> >> fwnode calls work for both DT and non-DT cases. In the DT case they
> >> work with DT nodes and properties. In the non-DT case, they work with
> >> manually populated properties.
> >>
> >>> Because the newly introduced it66121_create_bridge()
> >>> function is a core. I think It's better leave this task to a more
> >>> advance programmer.
> >>> if there have use case. It can be introduced at a latter time,
> >>> probably parallel with
> >>> the DT.
> >>>
> >>> I think DT and/or ACPI is best for integrated devices, but it66121
> >>> display bridges is
> >>> a i2c slave device. Personally, I think slave device shouldn't be
> >>> standalone. I'm more
> >>> prefer to turn this driver to support hot-plug, even remove the
> >>> device on the run time
> >>> freely when detach and allow reattach. Like the I2C EEPROM device in
> >>> the monitor (which
> >>> contains the EDID, with I2C slave address 0x50). The I2C EEPROM
> >>> device *also* don't has
> >>> a corresponding struct device representation in linux kernel.
> >> It has. See i2c_client::dev.
> >
> > No, what I mean is that there don't have a device driver for
> > monitor(display) hardware entity.
> > And the drm_do_probe_ddc_edid() is the static linked driver, which is
> > similar with the idea
> > this series want to express.

Because the monitor is not a part of the display pipeline.

> >
> >
> >>> so I still think It is best to make this drivers functional as a
> >>> static lib, but I want
> >>> to hear you to say more. Why it would be a *better* alternative to
> >>> turn OF calls into
> >>> fwnode calls? what are the potential benefits?
> >> Because then you can populate device properties from your root device.
> >> Because it allows the platform to specify the bus width instead of
> >> hardcoding 24 bits (which might work in your case, but might not be
> >> applicable to another user next week).
> >
> >
> > No, this problem can be easily solved. Simply add another argument.
> >
> > ```
> >
> > int it66121_create_bridge(struct i2c_client *client, bool of_support,
> >                           bool hpd_support, bool audio_support, u32
> > bus_width,
> >                           struct drm_bridge **bridge);
> > ```
> >
> >
> >> Anyway, even without fwnode, I'd strongly suggest you to drop the
> >> it66121_create_bridge() as it is now and start by populating the i2c
> >> bus from your root device.
> >
> > This will force all non-DT users to add the similar code patter at the
> > display controller side,
> > which is another kind of duplication. The monitor is also as I2C slave
> > device, can be abstract
> > as a identify drm bridges in theory, I guess.
> >
>
> 'identify' -> 'identity'
>
>
> >
> >> Then you will need some way (fwnode?) to
> >> discover the bridge chain. And at the last point you will get into the
> >> device data and/or properties business.
> >>
> > No, leave that chance to a more better programmer and forgive me please,
> > too difficult, I'm afraid of not able to solve. Thanks a lot for the
> > trust!

From my point of view: no.
Sui Jingfeng Nov. 16, 2023, 5:18 p.m. UTC | #15
On 2023/11/16 23:23, Dmitry Baryshkov wrote:
> On Thu, 16 Nov 2023 at 14:08, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>>
>> On 2023/11/16 19:53, Sui Jingfeng wrote:
>>> Hi,
>>>
>>>
>>> On 2023/11/16 19:29, Dmitry Baryshkov wrote:
>>>> On Thu, 16 Nov 2023 at 13:18, Sui Jingfeng <sui.jingfeng@linux.dev>
>>>> wrote:
>>>>> Hi,
>>>>>
>>>>>
>>>>> On 2023/11/15 00:30, Dmitry Baryshkov wrote:
>>>>>>> +
>>>>>>> +               ctx->connector = connector;
>>>>>>> +       }
>>>>>>>
>>>>>>>            if (ctx->info->id == ID_IT66121) {
>>>>>>>                    ret = regmap_write_bits(ctx->regmap,
>>>>>>> IT66121_CLK_BANK_REG,
>>>>>>> @@ -1632,16 +1651,13 @@ static const char * const
>>>>>>> it66121_supplies[] = {
>>>>>>>            "vcn33", "vcn18", "vrf12"
>>>>>>>     };
>>>>>>>
>>>>>>> -static int it66121_probe(struct i2c_client *client)
>>>>>>> +int it66121_create_bridge(struct i2c_client *client, bool
>>>>>>> of_support,
>>>>>>> +                         bool hpd_support, bool audio_support,
>>>>>>> +                         struct drm_bridge **bridge)
>>>>>>>     {
>>>>>>> +       struct device *dev = &client->dev;
>>>>>>>            int ret;
>>>>>>>            struct it66121_ctx *ctx;
>>>>>>> -       struct device *dev = &client->dev;
>>>>>>> -
>>>>>>> -       if (!i2c_check_functionality(client->adapter,
>>>>>>> I2C_FUNC_I2C)) {
>>>>>>> -               dev_err(dev, "I2C check functionality failed.\n");
>>>>>>> -               return -ENXIO;
>>>>>>> -       }
>>>>>>>
>>>>>>>            ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
>>>>>>>            if (!ctx)
>>>>>>> @@ -1649,24 +1665,19 @@ static int it66121_probe(struct i2c_client
>>>>>>> *client)
>>>>>>>
>>>>>>>            ctx->dev = dev;
>>>>>>>            ctx->client = client;
>>>>>>> -       ctx->info = i2c_get_match_data(client);
>>>>>>> -
>>>>>>> -       ret = it66121_of_read_bus_width(dev, &ctx->bus_width);
>>>>>>> -       if (ret)
>>>>>>> -               return ret;
>>>>>>> -
>>>>>>> -       ret = it66121_of_get_next_bridge(dev, &ctx->next_bridge);
>>>>>>> -       if (ret)
>>>>>>> -               return ret;
>>>>>>> -
>>>>>>> -       i2c_set_clientdata(client, ctx);
>>>>>>>            mutex_init(&ctx->lock);
>>>>>>>
>>>>>>> -       ret = devm_regulator_bulk_get_enable(dev,
>>>>>>> ARRAY_SIZE(it66121_supplies),
>>>>>>> - it66121_supplies);
>>>>>>> -       if (ret) {
>>>>>>> -               dev_err(dev, "Failed to enable power supplies\n");
>>>>>>> -               return ret;
>>>>>>> +       if (of_support) {
>>>>>>> +               ret = it66121_of_read_bus_width(dev,
>>>>>>> &ctx->bus_width);
>>>>>>> +               if (ret)
>>>>>>> +                       return ret;
>>>>>>> +
>>>>>>> +               ret = it66121_of_get_next_bridge(dev,
>>>>>>> &ctx->next_bridge);
>>>>>>> +               if (ret)
>>>>>>> +                       return ret;
>>>>>>> +       } else {
>>>>>>> +               ctx->bus_width = 24;
>>>>>>> +               ctx->next_bridge = NULL;
>>>>>>>            }
>>>>>> A better alternative would be to turn OF calls into fwnode calls and
>>>>>> to populate the fwnode properties. See
>>>>>> drivers/platform/x86/intel/chtwc_int33fe.c for example.
>>>>> Honestly, I don't want to leave any scratch(breadcrumbs).
>>>>> I'm worries about that turn OF calls into fwnode calls will leave
>>>>> something unwanted.
>>>>>
>>>>> Because I am not sure if fwnode calls will make sense in the DT
>>>>> world, while my patch
>>>>> *still* be useful in the DT world.
>>>> fwnode calls work for both DT and non-DT cases. In the DT case they
>>>> work with DT nodes and properties. In the non-DT case, they work with
>>>> manually populated properties.
>>>>
>>>>> Because the newly introduced it66121_create_bridge()
>>>>> function is a core. I think It's better leave this task to a more
>>>>> advance programmer.
>>>>> if there have use case. It can be introduced at a latter time,
>>>>> probably parallel with
>>>>> the DT.
>>>>>
>>>>> I think DT and/or ACPI is best for integrated devices, but it66121
>>>>> display bridges is
>>>>> a i2c slave device. Personally, I think slave device shouldn't be
>>>>> standalone. I'm more
>>>>> prefer to turn this driver to support hot-plug, even remove the
>>>>> device on the run time
>>>>> freely when detach and allow reattach. Like the I2C EEPROM device in
>>>>> the monitor (which
>>>>> contains the EDID, with I2C slave address 0x50). The I2C EEPROM
>>>>> device *also* don't has
>>>>> a corresponding struct device representation in linux kernel.
>>>> It has. See i2c_client::dev.
>>> No, what I mean is that there don't have a device driver for
>>> monitor(display) hardware entity.
>>> And the drm_do_probe_ddc_edid() is the static linked driver, which is
>>> similar with the idea
>>> this series want to express.
> Because the monitor is not a part of the display pipeline.
>
I think the monitor *is definitely* part of the display pipeline, and it
is the most important part of the entire display pipeline.

1)

DPMS, self-refreshing, display timings, resolutions supported, HDR, DSC,
gsync and freesync etc can be part of whole mode-set. Please consider
what the various ->mode_valid() and -> the atomic_check() are for?

2)

If the monitor is not a part of the display pipeline, then the various
display panels hardware should also not be part of the display pipeline.
Because they are all belong to display category.
  
the monitor = panel + panel drive IC(such as RTD2281CL, HT1622, ssd130x).

There are panel bridges which abstract the panel + connector as a drm bridge,
why the bare panel can be part of the display pipeline, while the more complex
monitor can't be?
Sui Jingfeng Nov. 17, 2023, 4:24 a.m. UTC | #16
Hi,

On 2023/11/16 23:23, Dmitry Baryshkov wrote:
>>>> Then you will need some way (fwnode?) to
>>>> discover the bridge chain. And at the last point you will get into the
>>>> device data and/or properties business.
>>>>
>>> No, leave that chance to a more better programmer and forgive me please,
>>> too difficult, I'm afraid of not able to solve. Thanks a lot for the
>>> trust!
>  From my point of view: no.


I respect the fact that the community prefer generic mechanisms.
If our approach is not what the community want, can I switch back
to my previous solution? I can reduce the duplication of our
localized it66121 driver to a minimal, rewrite it until it meets
the community's requirement. I know our device looks weird and
our approach is not elegant. But at the very least, we could not
mess the community's design up by localize. Otherwise, I don't know
what is the better approach to solve such a problem.

Can I switch back or any other ideas?
Dmitry Baryshkov Nov. 17, 2023, 9:03 a.m. UTC | #17
On Fri, 17 Nov 2023 at 06:24, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>
> Hi,
>
> On 2023/11/16 23:23, Dmitry Baryshkov wrote:
> >>>> Then you will need some way (fwnode?) to
> >>>> discover the bridge chain. And at the last point you will get into the
> >>>> device data and/or properties business.
> >>>>
> >>> No, leave that chance to a more better programmer and forgive me please,
> >>> too difficult, I'm afraid of not able to solve. Thanks a lot for the
> >>> trust!
> >  From my point of view: no.
>
>
> I respect the fact that the community prefer generic mechanisms.
> If our approach is not what the community want, can I switch back
> to my previous solution? I can reduce the duplication of our
> localized it66121 driver to a minimal, rewrite it until it meets
> the community's requirement. I know our device looks weird and
> our approach is not elegant. But at the very least, we could not
> mess the community's design up by localize. Otherwise, I don't know
> what is the better approach to solve such a problem.
>
> Can I switch back or any other ideas?

I keep on repeating: create the i2c device from your root device
driver, which parses BIOS data.
Maxime Ripard Nov. 17, 2023, 9:52 a.m. UTC | #18
On Fri, Nov 17, 2023 at 01:18:49AM +0800, Sui Jingfeng wrote:
> 
> On 2023/11/16 23:23, Dmitry Baryshkov wrote:
> > On Thu, 16 Nov 2023 at 14:08, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
> > > 
> > > On 2023/11/16 19:53, Sui Jingfeng wrote:
> > > > Hi,
> > > > 
> > > > 
> > > > On 2023/11/16 19:29, Dmitry Baryshkov wrote:
> > > > > On Thu, 16 Nov 2023 at 13:18, Sui Jingfeng <sui.jingfeng@linux.dev>
> > > > > wrote:
> > > > > > Hi,
> > > > > > 
> > > > > > 
> > > > > > On 2023/11/15 00:30, Dmitry Baryshkov wrote:
> > > > > > > > +
> > > > > > > > +               ctx->connector = connector;
> > > > > > > > +       }
> > > > > > > > 
> > > > > > > >            if (ctx->info->id == ID_IT66121) {
> > > > > > > >                    ret = regmap_write_bits(ctx->regmap,
> > > > > > > > IT66121_CLK_BANK_REG,
> > > > > > > > @@ -1632,16 +1651,13 @@ static const char * const
> > > > > > > > it66121_supplies[] = {
> > > > > > > >            "vcn33", "vcn18", "vrf12"
> > > > > > > >     };
> > > > > > > > 
> > > > > > > > -static int it66121_probe(struct i2c_client *client)
> > > > > > > > +int it66121_create_bridge(struct i2c_client *client, bool
> > > > > > > > of_support,
> > > > > > > > +                         bool hpd_support, bool audio_support,
> > > > > > > > +                         struct drm_bridge **bridge)
> > > > > > > >     {
> > > > > > > > +       struct device *dev = &client->dev;
> > > > > > > >            int ret;
> > > > > > > >            struct it66121_ctx *ctx;
> > > > > > > > -       struct device *dev = &client->dev;
> > > > > > > > -
> > > > > > > > -       if (!i2c_check_functionality(client->adapter,
> > > > > > > > I2C_FUNC_I2C)) {
> > > > > > > > -               dev_err(dev, "I2C check functionality failed.\n");
> > > > > > > > -               return -ENXIO;
> > > > > > > > -       }
> > > > > > > > 
> > > > > > > >            ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
> > > > > > > >            if (!ctx)
> > > > > > > > @@ -1649,24 +1665,19 @@ static int it66121_probe(struct i2c_client
> > > > > > > > *client)
> > > > > > > > 
> > > > > > > >            ctx->dev = dev;
> > > > > > > >            ctx->client = client;
> > > > > > > > -       ctx->info = i2c_get_match_data(client);
> > > > > > > > -
> > > > > > > > -       ret = it66121_of_read_bus_width(dev, &ctx->bus_width);
> > > > > > > > -       if (ret)
> > > > > > > > -               return ret;
> > > > > > > > -
> > > > > > > > -       ret = it66121_of_get_next_bridge(dev, &ctx->next_bridge);
> > > > > > > > -       if (ret)
> > > > > > > > -               return ret;
> > > > > > > > -
> > > > > > > > -       i2c_set_clientdata(client, ctx);
> > > > > > > >            mutex_init(&ctx->lock);
> > > > > > > > 
> > > > > > > > -       ret = devm_regulator_bulk_get_enable(dev,
> > > > > > > > ARRAY_SIZE(it66121_supplies),
> > > > > > > > - it66121_supplies);
> > > > > > > > -       if (ret) {
> > > > > > > > -               dev_err(dev, "Failed to enable power supplies\n");
> > > > > > > > -               return ret;
> > > > > > > > +       if (of_support) {
> > > > > > > > +               ret = it66121_of_read_bus_width(dev,
> > > > > > > > &ctx->bus_width);
> > > > > > > > +               if (ret)
> > > > > > > > +                       return ret;
> > > > > > > > +
> > > > > > > > +               ret = it66121_of_get_next_bridge(dev,
> > > > > > > > &ctx->next_bridge);
> > > > > > > > +               if (ret)
> > > > > > > > +                       return ret;
> > > > > > > > +       } else {
> > > > > > > > +               ctx->bus_width = 24;
> > > > > > > > +               ctx->next_bridge = NULL;
> > > > > > > >            }
> > > > > > > A better alternative would be to turn OF calls into fwnode calls and
> > > > > > > to populate the fwnode properties. See
> > > > > > > drivers/platform/x86/intel/chtwc_int33fe.c for example.
> > > > > > Honestly, I don't want to leave any scratch(breadcrumbs).
> > > > > > I'm worries about that turn OF calls into fwnode calls will leave
> > > > > > something unwanted.
> > > > > > 
> > > > > > Because I am not sure if fwnode calls will make sense in the DT
> > > > > > world, while my patch
> > > > > > *still* be useful in the DT world.
> > > > > fwnode calls work for both DT and non-DT cases. In the DT case they
> > > > > work with DT nodes and properties. In the non-DT case, they work with
> > > > > manually populated properties.
> > > > > 
> > > > > > Because the newly introduced it66121_create_bridge()
> > > > > > function is a core. I think It's better leave this task to a more
> > > > > > advance programmer.
> > > > > > if there have use case. It can be introduced at a latter time,
> > > > > > probably parallel with
> > > > > > the DT.
> > > > > > 
> > > > > > I think DT and/or ACPI is best for integrated devices, but it66121
> > > > > > display bridges is
> > > > > > a i2c slave device. Personally, I think slave device shouldn't be
> > > > > > standalone. I'm more
> > > > > > prefer to turn this driver to support hot-plug, even remove the
> > > > > > device on the run time
> > > > > > freely when detach and allow reattach. Like the I2C EEPROM device in
> > > > > > the monitor (which
> > > > > > contains the EDID, with I2C slave address 0x50). The I2C EEPROM
> > > > > > device *also* don't has
> > > > > > a corresponding struct device representation in linux kernel.
> > > > > It has. See i2c_client::dev.
> > > > No, what I mean is that there don't have a device driver for
> > > > monitor(display) hardware entity.
> > > > And the drm_do_probe_ddc_edid() is the static linked driver, which is
> > > > similar with the idea
> > > > this series want to express.
> > Because the monitor is not a part of the display pipeline.
> > 
> I think the monitor *is definitely* part of the display pipeline, and it
> is the most important part of the entire display pipeline.
> 
> 1)
> 
> DPMS, self-refreshing, display timings, resolutions supported, HDR, DSC,
> gsync and freesync etc can be part of whole mode-set. Please consider
> what the various ->mode_valid() and -> the atomic_check() are for?
> 
> 2)
> 
> If the monitor is not a part of the display pipeline, then the various
> display panels hardware should also not be part of the display pipeline.
> Because they are all belong to display category.
> the monitor = panel + panel drive IC(such as RTD2281CL, HT1622, ssd130x).

To expand further on that, I guess one of the key difference is that you
don't really expect to interact with the EEPROM, you'll only read it,
which is fairly different from your bridge.

And if someone wanted to instatiate nvmem devices for the various
EEPROMs in the monitor, I would very much welcome that change.

Maxime
Maxime Ripard Nov. 17, 2023, 12:13 p.m. UTC | #19
On Fri, Nov 17, 2023 at 12:24:22PM +0800, Sui Jingfeng wrote:
> Hi,
> 
> On 2023/11/16 23:23, Dmitry Baryshkov wrote:
> > > > > Then you will need some way (fwnode?) to
> > > > > discover the bridge chain. And at the last point you will get into the
> > > > > device data and/or properties business.
> > > > > 
> > > > No, leave that chance to a more better programmer and forgive me please,
> > > > too difficult, I'm afraid of not able to solve. Thanks a lot for the
> > > > trust!
> >  From my point of view: no.
>
> I respect the fact that the community prefer generic mechanisms.
> If our approach is not what the community want, can I switch back
> to my previous solution?

By your previous solution, you mean rolling your own bridge driver? If
so, then no, it's not acceptable either.

> I can reduce the duplication of our localized it66121 driver to a
> minimal, rewrite it until it meets the community's requirement. I know
> our device looks weird and our approach is not elegant.

I'm glad we agree then :)

> But at the very least, we could not mess the community's design up by
> localize. Otherwise, I don't know what is the better approach to solve
> such a problem.

I think there's a gap between what we want from you and what you want
from us.

What we really care about is maintenance. In other words, it's mostly
about two things:

  - Once you and/or your company have moved on to other things, how easy
    it will be for us to keep that driver in good shape, and how much it
    will hold back any future development.

  - If we want to do a big rework, how much your driver will stand in
    the way.

That's pretty much all that we care about, and we will very much prefer
not to merge a driver in the first place than to have to maintain it for
10y while it stands in our way and we don't have any real documentation
or help.

So by making it "not weird" or "elegant" or whatever we can call it, you
effectively remove any concern we might have about merging your driver,
and there's only an upside (more hardware support and company
involvement is good!). So you're making it easy for you too.

Maxime
Sui Jingfeng Nov. 17, 2023, 5:14 p.m. UTC | #20
Hi,

On 2023/11/17 17:03, Dmitry Baryshkov wrote:
> On Fri, 17 Nov 2023 at 06:24, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>> Hi,
>>
>> On 2023/11/16 23:23, Dmitry Baryshkov wrote:
>>>>>> Then you will need some way (fwnode?) to
>>>>>> discover the bridge chain. And at the last point you will get into the
>>>>>> device data and/or properties business.
>>>>>>
>>>>> No, leave that chance to a more better programmer and forgive me please,
>>>>> too difficult, I'm afraid of not able to solve. Thanks a lot for the
>>>>> trust!
>>>   From my point of view: no.
>>
>> I respect the fact that the community prefer generic mechanisms.
>> If our approach is not what the community want, can I switch back
>> to my previous solution? I can reduce the duplication of our
>> localized it66121 driver to a minimal, rewrite it until it meets
>> the community's requirement. I know our device looks weird and
>> our approach is not elegant. But at the very least, we could not
>> mess the community's design up by localize. Otherwise, I don't know
>> what is the better approach to solve such a problem.
>>
>> Can I switch back or any other ideas?
> I keep on repeating: create the i2c device from your root device
> driver, which parses BIOS data.
>
This is not my own problems, currently it66121 (but not only) display bridge driver
don't works on X86 either. What we are trying to do is to provide a generic, non-platform
dependent solution. It is not only relevant to my driver. In fact, this series made
no assumption which hardware/display controller will be the user.

I have investigated before respin this patch, there are other hardwares which
ship the it66121 display bridge. For example, the Fresco Logic FL2000dx USB 3.0
to VGA display adapter[1][2]. Even the windows have a driver.

[1] https://github.com/FrescoLogic/FL2000
[2] https://oemdrivers.com/graphics-fresco-logic-fl2000
Sui Jingfeng Nov. 17, 2023, 5:35 p.m. UTC | #21
Hi,


On 2023/11/17 17:03, Dmitry Baryshkov wrote:
> On Fri, 17 Nov 2023 at 06:24, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>> Hi,
>>
>> On 2023/11/16 23:23, Dmitry Baryshkov wrote:
>>>>>> Then you will need some way (fwnode?) to
>>>>>> discover the bridge chain. And at the last point you will get into the
>>>>>> device data and/or properties business.
>>>>>>
>>>>> No, leave that chance to a more better programmer and forgive me please,
>>>>> too difficult, I'm afraid of not able to solve. Thanks a lot for the
>>>>> trust!
>>>   From my point of view: no.
>>
>> I respect the fact that the community prefer generic mechanisms.
>> If our approach is not what the community want, can I switch back
>> to my previous solution? I can reduce the duplication of our
>> localized it66121 driver to a minimal, rewrite it until it meets
>> the community's requirement. I know our device looks weird and
>> our approach is not elegant. But at the very least, we could not
>> mess the community's design up by localize. Otherwise, I don't know
>> what is the better approach to solve such a problem.
>>
>> Can I switch back or any other ideas?
> I keep on repeating: create the i2c device from your root device
> driver, which parses BIOS data.


You didn't focus on solve the problem, You are focus on solving me.
How does the method that parsing BIOS data can be generic and applied
universally?
Neil Armstrong Nov. 20, 2023, 8:23 a.m. UTC | #22
On 17/11/2023 18:14, Sui Jingfeng wrote:
> Hi,
> 
> On 2023/11/17 17:03, Dmitry Baryshkov wrote:
>> On Fri, 17 Nov 2023 at 06:24, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>>> Hi,
>>>
>>> On 2023/11/16 23:23, Dmitry Baryshkov wrote:
>>>>>>> Then you will need some way (fwnode?) to
>>>>>>> discover the bridge chain. And at the last point you will get into the
>>>>>>> device data and/or properties business.
>>>>>>>
>>>>>> No, leave that chance to a more better programmer and forgive me please,
>>>>>> too difficult, I'm afraid of not able to solve. Thanks a lot for the
>>>>>> trust!
>>>>   From my point of view: no.
>>>
>>> I respect the fact that the community prefer generic mechanisms.
>>> If our approach is not what the community want, can I switch back
>>> to my previous solution? I can reduce the duplication of our
>>> localized it66121 driver to a minimal, rewrite it until it meets
>>> the community's requirement. I know our device looks weird and
>>> our approach is not elegant. But at the very least, we could not
>>> mess the community's design up by localize. Otherwise, I don't know
>>> what is the better approach to solve such a problem.
>>>
>>> Can I switch back or any other ideas?
>> I keep on repeating: create the i2c device from your root device
>> driver, which parses BIOS data.
>>
> This is not my own problems, currently it66121 (but not only) display bridge driver
> don't works on X86 either. What we are trying to do is to provide a generic, non-platform
> dependent solution. It is not only relevant to my driver. In fact, this series made
> no assumption which hardware/display controller will be the user.
> 
> I have investigated before respin this patch, there are other hardwares which
> ship the it66121 display bridge. For example, the Fresco Logic FL2000dx USB 3.0
> to VGA display adapter[1][2]. Even the windows have a driver.
> 
> [1] https://github.com/FrescoLogic/FL2000
> [2] https://oemdrivers.com/graphics-fresco-logic-fl2000

Switching to fwnodes, registering an i2c bus and generating fwnode data matching the
interconnect architecture is the way.

DRM Bridge transition to fwnode only should be done first, this will open bridge
to any architecture and device description (DT or ACPI).

Neil

>
kernel test robot Nov. 20, 2023, 9:08 a.m. UTC | #23
Hi Sui,

kernel test robot noticed the following build warnings:

[auto build test WARNING on drm-misc/drm-misc-next]
[also build test WARNING on linus/master v6.7-rc2 next-20231120]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Sui-Jingfeng/drm-bridge-it66121-Use-dev-replace-ctx-dev-in-the-it66121_probe/20231114-231203
base:   git://anongit.freedesktop.org/drm/drm-misc drm-misc-next
patch link:    https://lore.kernel.org/r/20231114150130.497915-9-sui.jingfeng%40linux.dev
patch subject: [PATCH 8/8] drm/bridge: it66121: Allow link this driver as a lib
config: x86_64-buildonly-randconfig-004-20231120 (https://download.01.org/0day-ci/archive/20231120/202311201649.qjEbx5YL-lkp@intel.com/config)
compiler: clang version 16.0.4 (https://github.com/llvm/llvm-project.git ae42196bc493ffe877a7e3dff8be32035dea4d07)
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20231120/202311201649.qjEbx5YL-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202311201649.qjEbx5YL-lkp@intel.com/

All warnings (new ones prefixed by >>):

>> drivers/gpu/drm/bridge/ite-it66121.c:1654:5: warning: no previous prototype for function 'it66121_create_bridge' [-Wmissing-prototypes]
   int it66121_create_bridge(struct i2c_client *client, bool of_support,
       ^
   drivers/gpu/drm/bridge/ite-it66121.c:1654:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
   int it66121_create_bridge(struct i2c_client *client, bool of_support,
   ^
   static 
>> drivers/gpu/drm/bridge/ite-it66121.c:1752:6: warning: no previous prototype for function 'it66121_destroy_bridge' [-Wmissing-prototypes]
   void it66121_destroy_bridge(struct drm_bridge *bridge)
        ^
   drivers/gpu/drm/bridge/ite-it66121.c:1752:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
   void it66121_destroy_bridge(struct drm_bridge *bridge)
   ^
   static 
   2 warnings generated.


vim +/it66121_create_bridge +1654 drivers/gpu/drm/bridge/ite-it66121.c

  1653	
> 1654	int it66121_create_bridge(struct i2c_client *client, bool of_support,
  1655				  bool hpd_support, bool audio_support,
  1656				  struct drm_bridge **bridge)
  1657	{
  1658		struct device *dev = &client->dev;
  1659		int ret;
  1660		struct it66121_ctx *ctx;
  1661	
  1662		ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
  1663		if (!ctx)
  1664			return -ENOMEM;
  1665	
  1666		ctx->dev = dev;
  1667		ctx->client = client;
  1668		mutex_init(&ctx->lock);
  1669	
  1670		if (of_support) {
  1671			ret = it66121_of_read_bus_width(dev, &ctx->bus_width);
  1672			if (ret)
  1673				return ret;
  1674	
  1675			ret = it66121_of_get_next_bridge(dev, &ctx->next_bridge);
  1676			if (ret)
  1677				return ret;
  1678		} else {
  1679			ctx->bus_width = 24;
  1680			ctx->next_bridge = NULL;
  1681		}
  1682	
  1683		it66121_hw_reset(ctx);
  1684	
  1685		ctx->regmap = devm_regmap_init_i2c(client, &it66121_regmap_config);
  1686		if (IS_ERR(ctx->regmap))
  1687			return PTR_ERR(ctx->regmap);
  1688	
  1689		ret = it66121_read_chip_id(ctx, false);
  1690		if (ret)
  1691			return ret;
  1692	
  1693		ctx->info = it66121_get_match_data(ctx->vender_id, ctx->device_id);
  1694		if (!ctx->info)
  1695			return -ENODEV;
  1696	
  1697		if (hpd_support) {
  1698			ret = devm_request_threaded_irq(dev, client->irq, NULL,
  1699							it66121_irq_threaded_handler,
  1700							IRQF_ONESHOT, dev_name(dev),
  1701							ctx);
  1702			if (ret < 0) {
  1703				dev_err(dev, "Failed to request irq: %d\n", ret);
  1704				return ret;
  1705			}
  1706		}
  1707	
  1708		it66121_bridge_init_base(&ctx->bridge, dev->of_node, true);
  1709	
  1710		if (audio_support)
  1711			it66121_audio_codec_init(ctx, dev);
  1712	
  1713		*bridge = &ctx->bridge;
  1714	
  1715		dev_info(dev, "IT66121 probed, chip id: 0x%x:0x%x, revision: %u\n",
  1716			 ctx->vender_id, ctx->device_id, ctx->revision);
  1717	
  1718		return 0;
  1719	}
  1720	EXPORT_SYMBOL_GPL(it66121_create_bridge);
  1721	
  1722	static int it66121_probe(struct i2c_client *client)
  1723	{
  1724		struct device *dev = &client->dev;
  1725		struct it66121_ctx *ctx;
  1726		struct drm_bridge *bridge;
  1727		int ret;
  1728	
  1729		if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) {
  1730			dev_err(dev, "I2C check functionality failed.\n");
  1731			return -ENXIO;
  1732		}
  1733	
  1734		ret = devm_regulator_bulk_get_enable(dev, ARRAY_SIZE(it66121_supplies),
  1735						     it66121_supplies);
  1736		if (ret) {
  1737			dev_err(dev, "Failed to enable power supplies\n");
  1738			return ret;
  1739		}
  1740	
  1741		ret = it66121_create_bridge(client, true, true, true, &bridge);
  1742		if (ret)
  1743			return ret;
  1744	
  1745		ctx = bridge_to_it66121(bridge);
  1746	
  1747		i2c_set_clientdata(client, ctx);
  1748	
  1749		return 0;
  1750	}
  1751	
> 1752	void it66121_destroy_bridge(struct drm_bridge *bridge)
  1753	{
  1754		struct it66121_ctx *ctx = bridge_to_it66121(bridge);
  1755	
  1756		drm_bridge_remove(bridge);
  1757	
  1758		mutex_destroy(&ctx->lock);
  1759	}
  1760	EXPORT_SYMBOL_GPL(it66121_destroy_bridge);
  1761
Dmitry Baryshkov Nov. 20, 2023, 10:06 a.m. UTC | #24
On Fri, 17 Nov 2023 at 19:36, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>
> Hi,
>
>
> On 2023/11/17 17:03, Dmitry Baryshkov wrote:
> > On Fri, 17 Nov 2023 at 06:24, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
> >> Hi,
> >>
> >> On 2023/11/16 23:23, Dmitry Baryshkov wrote:
> >>>>>> Then you will need some way (fwnode?) to
> >>>>>> discover the bridge chain. And at the last point you will get into the
> >>>>>> device data and/or properties business.
> >>>>>>
> >>>>> No, leave that chance to a more better programmer and forgive me please,
> >>>>> too difficult, I'm afraid of not able to solve. Thanks a lot for the
> >>>>> trust!
> >>>   From my point of view: no.
> >>
> >> I respect the fact that the community prefer generic mechanisms.
> >> If our approach is not what the community want, can I switch back
> >> to my previous solution? I can reduce the duplication of our
> >> localized it66121 driver to a minimal, rewrite it until it meets
> >> the community's requirement. I know our device looks weird and
> >> our approach is not elegant. But at the very least, we could not
> >> mess the community's design up by localize. Otherwise, I don't know
> >> what is the better approach to solve such a problem.
> >>
> >> Can I switch back or any other ideas?
> > I keep on repeating: create the i2c device from your root device
> > driver, which parses BIOS data.
>
>
> You didn't focus on solve the problem, You are focus on solving me.
> How does the method that parsing BIOS data can be generic and applied
> universally?

Parsing BIOS data is unique to your platform (as well as your BIOS
tables). However using and extending (instead of replacing it just for
your platform) is a generic item.
Sui Jingfeng Nov. 23, 2023, 5:05 a.m. UTC | #25
Hi,


On 2023/11/16 19:19, Dmitry Baryshkov wrote:
> On Thu, 16 Nov 2023 at 12:13, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>> Hi,
>>
>>
>> On 2023/11/16 17:30, Dmitry Baryshkov wrote:
>>> On Thu, 16 Nov 2023 at 11:14, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>>>> Hi,
>>>>
>>>> Thanks a lot for reviewing!
>>>>
>>>>
>>>> On 2023/11/15 00:30, Dmitry Baryshkov wrote:
>>>>> On Tue, 14 Nov 2023 at 17:09, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>>>>>> From: Sui Jingfeng <suijingfeng@loongson.cn>
>>>>>>
>>>>>> The it66121_create_bridge() and it66121_destroy_bridge() are added to
>>>>>> export the core functionalities. Create a connector manually by using
>>>>>> bridge connector helpers when link as a lib.
>>>>>>
>>>>>> Signed-off-by: Sui Jingfeng <suijingfeng@loongson.cn>
>>>>>> ---
>>>>>>     drivers/gpu/drm/bridge/ite-it66121.c | 134 +++++++++++++++++++--------
>>>>>>     include/drm/bridge/ite-it66121.h     |  17 ++++
>>>>>>     2 files changed, 113 insertions(+), 38 deletions(-)
>>>>>>     create mode 100644 include/drm/bridge/ite-it66121.h
>>>>>>
>>>>>> diff --git a/drivers/gpu/drm/bridge/ite-it66121.c b/drivers/gpu/drm/bridge/ite-it66121.c
>>>>>> index 8971414a2a60..f5968b679c5d 100644
>>>>>> --- a/drivers/gpu/drm/bridge/ite-it66121.c
>>>>>> +++ b/drivers/gpu/drm/bridge/ite-it66121.c
>>>>>> @@ -22,6 +22,7 @@
>>>>>>
>>>>>>     #include <drm/drm_atomic_helper.h>
>>>>>>     #include <drm/drm_bridge.h>
>>>>>> +#include <drm/drm_bridge_connector.h>
>>>>>>     #include <drm/drm_edid.h>
>>>>>>     #include <drm/drm_modes.h>
>>>>>>     #include <drm/drm_print.h>
>>>>>> @@ -703,14 +704,32 @@ static int it66121_bridge_attach(struct drm_bridge *bridge,
>>>>>>                                     enum drm_bridge_attach_flags flags)
>>>>>>     {
>>>>>>            struct it66121_ctx *ctx = bridge_to_it66121(bridge);
>>>>>> +       struct drm_bridge *next_bridge = ctx->next_bridge;
>>>>>> +       struct drm_encoder *encoder = bridge->encoder;
>>>>>>            int ret;
>>>>>>
>>>>>> -       if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR))
>>>>>> -               return -EINVAL;
>>>>>> +       if (next_bridge) {
>>>>>> +               if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)) {
>>>>>> +                       WARN_ON(1);
>>>>> Why? At least use WARN() instead
>>>> Originally I want to
>>>>
>>>>
>>>>
>>>>
>>>>>> +                       flags |= DRM_BRIDGE_ATTACH_NO_CONNECTOR;
>>>>>> +               }
>>>>>> +               ret = drm_bridge_attach(encoder, next_bridge, bridge, flags);
>>>>>> +               if (ret)
>>>>>> +                       return ret;
>>>>>> +       } else {
>>>>>> +               struct drm_connector *connector;
>>>>>>
>>>>>> -       ret = drm_bridge_attach(bridge->encoder, ctx->next_bridge, bridge, flags);
>>>>>> -       if (ret)
>>>>>> -               return ret;
>>>>>> +               if (flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)
>>>>>> +                       WARN_ON(1);
>>>>> No. It is perfectly fine to create attach a bridge with no next_bridge
>>>>> and with the DRM_BRIDGE_ATTACH_NO_CONNECTOR flag.
>>>>>
>>>> The document say when DRM_BRIDGE_ATTACH_NO_CONNECTOR flag is set
>>>> the bridge shall not create a drm_connector. So I think if a display
>>>> bridge driver don't have a next bridge attached (Currently, this is
>>>> told by the DT), it says that this is a non-DT environment. On such
>>>> a case, this display bridge driver(it66121.ko) should behavior like
>>>> a *agent*. Because the upstream port of it66121 is the DVO port of
>>>> the display controller, the downstream port of it66121 is the HDMI
>>>> connector. it66121 is on the middle. So I think the it66121.ko should
>>>> handle all of troubles on behalf of the display controller drivers.
>>> No. Don't make decisions for the other drivers. They might have different needs.
>> [...]
>>
>>
>>>> Therefore (when in non-DT use case), the display controller drivers
>>>> side should not set DRM_BRIDGE_ATTACH_NO_CONNECTOR flag anymore.
>>>> Which is to hint that the it66121 should totally in charge of those
>>>> tasks (either by using bridge connector helper or create a connector
>>>> manually). I don't understand on such a case, why bother display
>>>> controller drivers anymore.
>>> This is the reason why we had introduced this flag. It allows the
>>> driver to customise the connector. It even allows the driver to
>>> implement a connector on its own, completely ignoring the
>>> drm_bridge_connector.
>>
>> I know what you said is right in the sense of the universe cases,
>> but I think the most frequent(majority) use case is that there is
>> only one display bridge on the middle. Therefore, I don't want to
>> movethe connector things into device driver if there is only one display
>> bridge(say it66121) in the middle. After all, there is no *direct
>> physical connection* from the perspective of the hardware. I means that
>> there is no hardware wires connectthe HDMI connector and the DVO port. So display controller drivers
>> should not interact with anything related with the connector on a
>> perfect abstract on the software side. Especially for such a simple use
>> case. It probably make senses to make a  decision for themost frequently use case, please also note
>> that this patch didn't introduce any-restriction for the more advance
>> uses cases(multiple bridges in the middle).
> So, for the sake of not having the connector in the display driver,
> you want to add boilerplate code basically to each and every bridge
> driver. In the end, they should all behave in the same way.
>
> Moreover, there is no way this implementation can work without a
> warning if there are two bridges in a chain and the it66121 is the
> second (the last) one. The host can not specify the
> DRM_BRIDGE_ATTACH_NO_CONNECTOR)
>>>>>> +                       WARN_ON(1);
>>>>> No. It is perfectly fine to create attach a bridge with no next_bridge
>>>>> and with the DRM_BRIDGE_ATTACH_NO_CONNECTOR flag.
>>>>>
>>>> The document say when DRM_BRIDGE_ATTACH_NO_CONNECTOR flag is set
>>>> the bridge shall not create a drm_connector. So I think if a display
>>>> bridge driver don't have a next bridge attached (Currently, this is
>>>> told by the DT), it says that this is a non-DT environment. On such
>>>> a case, this display bridge driver(it66121.ko) should behavior like
>>>> a *agent*. Because the upstream port of it66121 is the DVO port of
>>>> the display controller, the downstream port of it66121 is the HDMI
>>>> connector. it66121 is on the middle. So I think the it66121.ko should
>>>> handle all of troubles on behalf of the display controller drivers.
>>> No. Don't make decisions for the other drivers. They might have different needs.
>> [...]
>>
>>
>>>> Therefore (when in non-DT use case), the display controller drivers
>>>> side should not set DRM_BRIDGE_ATTACH_NO_CONNECTOR flag anymore.
>>>> Which is to hint that the it66121 should totally in charge of those
>>>> tasks (either by using bridge connector helper or create a connector
>>>> manually). I don't understand on such a case, why bother display
>>>> controller drivers anymore.
>>> This is the reason why we had introduced this flag. It allows the
>>> driver to customise the connector. It even allows the driver to
>>> implement a connector on its own, completely ignoring the
>>> drm_bridge_connector.
>>
>> I know what you said is right in the sense of the universe cases,
>> but I think the most frequent(majority) use case is that there is
>> only one display bridge on the middle. Therefore, I don't want to
>> movethe connector things into device driver if there is only one display
>> bridge(say it66121) in the middle. After all, there is no *direct
>> physical connection* from the perspective of the hardware. I means that
>> there is no hardware wires connectthe HDMI connector and the DVO port. So display controller drivers
>> should not interact with anything related with the connector on a
>> perfect abstract on the software side. Especially for such a simple use
>> case. It probably make senses to make a  decision for themost frequently use case, please also note
>> that this patch didn't introduce any-restriction for the more advance
>> uses cases(multiple bridges in the middle).
> So, for the sake of not having the connector in the display driver,
> you want to add boilerplate code basically to each and every bridge
> driver. In the end, they should all behave in the same way.

No, I'm only intend to modify the one when there has a user emerged.
If we have the connector related code in the KMS display driver side,
then I think we don't honor the meaning of the word *bridge*. I was
told drm_bridge is a modern design, if we still need the DC side
worry about something do not have a physical connection, then it will
not be modern anymore, it is not a good bridge.


> Moreover, there is no way this implementation can work without a
> warning if there are two bridges in a chain and the it66121 is the
> second (the last) one.

Yes and no!

If one of them are transparent, then thisimplementation still can works. It is just that this will not be a good 
abstract anymore.I do have seen such design on some notebook hardware.  But from my programming experiences,
using two bridges are typically a bad practice in engineering. As it tend
to increase the PCB board area and increase entire cost. As you need buy
one more TX encoder chip. Please also consider that the embedded world focus
on low cost and low power consume.

I think, multiple display bridges case should be avoided for middle/low end
application. Or allow us to handle the two and/or more bridges cases in the
future. When there has at least one user emerged, we will introduce new
approach to handle then.

Do you find any product level boards that using two external display bridge and
one of them is it66121? If we can not even find a user, we are not even have a
board to test if current design (state of art) works. Does it suffer from module
loading order problems? what if their i2c slave address is same? Does such a use
case will past the S3/S4 test? All of those concerns are imposed to every display
bridges programmer from the very beginning.

I'm agree with the idea that drm bridges drivers involved toward to a direction
that support more complex design, but I think we should also leave a way for the
most frequent use case. Make it straight-forward as a canonical design.

> The host can not specify the
> DRM_BRIDGE_ATTACH_NO_CONNECTOR flag, it will cause a warning here. And
> it can not omit the flag (as otherwise the first bridge will create a
> connector, without consulting the second bridge).

The semantics of DRM_BRIDGE_ATTACH_NO_CONNECTOR flagare implement-defined, for our case, I could just ignore it if their 
don't have a signal(DT or ACPI) tell me that there are multiple bridges 
in the chain. This depend on community's attitude.

For it66121 with a canonical design, the host should not need to specify this flag.
Because at the time of when the device driver(it66121.ko) get loaded, the it66121
driver could parse the DT by itself, and detect if there has a next bridge, is it a
connector or is it yet another display bridges. The DT speak everything about the
topology. The flag is there just for the KMS display controller driver to explicit
control, use it and make it more useful is the right way, is it?


>>
>>>>>> +
>>>>>> +               connector = drm_bridge_connector_init(bridge->dev, encoder);
>>>>>> +               if (IS_ERR(connector))
>>>>>> +                       return PTR_ERR(connector);
>>>>>> +
>>>>>> +               drm_connector_attach_encoder(connector, encoder);
>>>>> This goes into your device driver.
>>>>>
>>>>>> +
>>>>>> +               ctx->connector = connector;
>>>>>> +       }
>>>>>>
>>>>>>            if (ctx->info->id == ID_IT66121) {
>>>>>>                    ret = regmap_write_bits(ctx->regmap, IT66121_CLK_BANK_REG,
>>>>>> @@ -1632,16 +1651,13 @@ static const char * const it66121_supplies[] = {
>>>>>>            "vcn33", "vcn18", "vrf12"
>>>>>>     };
>>>
>
>
Dmitry Baryshkov Nov. 23, 2023, 8:08 a.m. UTC | #26
On Thu, 23 Nov 2023 at 07:05, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>
> Hi,
>
>
> On 2023/11/16 19:19, Dmitry Baryshkov wrote:
> > On Thu, 16 Nov 2023 at 12:13, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
> >> Hi,
> >>
> >>
> >> On 2023/11/16 17:30, Dmitry Baryshkov wrote:
> >>> On Thu, 16 Nov 2023 at 11:14, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
> >>>> Hi,
> >>>>
> >>>> Thanks a lot for reviewing!
> >>>>
> >>>>
> >>>> On 2023/11/15 00:30, Dmitry Baryshkov wrote:
> >>>>> On Tue, 14 Nov 2023 at 17:09, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
> >>>>>> From: Sui Jingfeng <suijingfeng@loongson.cn>
> >>>>>>
> >>>>>> The it66121_create_bridge() and it66121_destroy_bridge() are added to
> >>>>>> export the core functionalities. Create a connector manually by using
> >>>>>> bridge connector helpers when link as a lib.
> >>>>>>
> >>>>>> Signed-off-by: Sui Jingfeng <suijingfeng@loongson.cn>
> >>>>>> ---
> >>>>>>     drivers/gpu/drm/bridge/ite-it66121.c | 134 +++++++++++++++++++--------
> >>>>>>     include/drm/bridge/ite-it66121.h     |  17 ++++
> >>>>>>     2 files changed, 113 insertions(+), 38 deletions(-)
> >>>>>>     create mode 100644 include/drm/bridge/ite-it66121.h
> >>>>>>
> >>>>>> diff --git a/drivers/gpu/drm/bridge/ite-it66121.c b/drivers/gpu/drm/bridge/ite-it66121.c
> >>>>>> index 8971414a2a60..f5968b679c5d 100644
> >>>>>> --- a/drivers/gpu/drm/bridge/ite-it66121.c
> >>>>>> +++ b/drivers/gpu/drm/bridge/ite-it66121.c
> >>>>>> @@ -22,6 +22,7 @@
> >>>>>>
> >>>>>>     #include <drm/drm_atomic_helper.h>
> >>>>>>     #include <drm/drm_bridge.h>
> >>>>>> +#include <drm/drm_bridge_connector.h>
> >>>>>>     #include <drm/drm_edid.h>
> >>>>>>     #include <drm/drm_modes.h>
> >>>>>>     #include <drm/drm_print.h>
> >>>>>> @@ -703,14 +704,32 @@ static int it66121_bridge_attach(struct drm_bridge *bridge,
> >>>>>>                                     enum drm_bridge_attach_flags flags)
> >>>>>>     {
> >>>>>>            struct it66121_ctx *ctx = bridge_to_it66121(bridge);
> >>>>>> +       struct drm_bridge *next_bridge = ctx->next_bridge;
> >>>>>> +       struct drm_encoder *encoder = bridge->encoder;
> >>>>>>            int ret;
> >>>>>>
> >>>>>> -       if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR))
> >>>>>> -               return -EINVAL;
> >>>>>> +       if (next_bridge) {
> >>>>>> +               if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)) {
> >>>>>> +                       WARN_ON(1);
> >>>>> Why? At least use WARN() instead
> >>>> Originally I want to
> >>>>
> >>>>
> >>>>
> >>>>
> >>>>>> +                       flags |= DRM_BRIDGE_ATTACH_NO_CONNECTOR;
> >>>>>> +               }
> >>>>>> +               ret = drm_bridge_attach(encoder, next_bridge, bridge, flags);
> >>>>>> +               if (ret)
> >>>>>> +                       return ret;
> >>>>>> +       } else {
> >>>>>> +               struct drm_connector *connector;
> >>>>>>
> >>>>>> -       ret = drm_bridge_attach(bridge->encoder, ctx->next_bridge, bridge, flags);
> >>>>>> -       if (ret)
> >>>>>> -               return ret;
> >>>>>> +               if (flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)
> >>>>>> +                       WARN_ON(1);
> >>>>> No. It is perfectly fine to create attach a bridge with no next_bridge
> >>>>> and with the DRM_BRIDGE_ATTACH_NO_CONNECTOR flag.
> >>>>>
> >>>> The document say when DRM_BRIDGE_ATTACH_NO_CONNECTOR flag is set
> >>>> the bridge shall not create a drm_connector. So I think if a display
> >>>> bridge driver don't have a next bridge attached (Currently, this is
> >>>> told by the DT), it says that this is a non-DT environment. On such
> >>>> a case, this display bridge driver(it66121.ko) should behavior like
> >>>> a *agent*. Because the upstream port of it66121 is the DVO port of
> >>>> the display controller, the downstream port of it66121 is the HDMI
> >>>> connector. it66121 is on the middle. So I think the it66121.ko should
> >>>> handle all of troubles on behalf of the display controller drivers.
> >>> No. Don't make decisions for the other drivers. They might have different needs.
> >> [...]
> >>
> >>
> >>>> Therefore (when in non-DT use case), the display controller drivers
> >>>> side should not set DRM_BRIDGE_ATTACH_NO_CONNECTOR flag anymore.
> >>>> Which is to hint that the it66121 should totally in charge of those
> >>>> tasks (either by using bridge connector helper or create a connector
> >>>> manually). I don't understand on such a case, why bother display
> >>>> controller drivers anymore.
> >>> This is the reason why we had introduced this flag. It allows the
> >>> driver to customise the connector. It even allows the driver to
> >>> implement a connector on its own, completely ignoring the
> >>> drm_bridge_connector.
> >>
> >> I know what you said is right in the sense of the universe cases,
> >> but I think the most frequent(majority) use case is that there is
> >> only one display bridge on the middle. Therefore, I don't want to
> >> movethe connector things into device driver if there is only one display
> >> bridge(say it66121) in the middle. After all, there is no *direct
> >> physical connection* from the perspective of the hardware. I means that
> >> there is no hardware wires connectthe HDMI connector and the DVO port. So display controller drivers
> >> should not interact with anything related with the connector on a
> >> perfect abstract on the software side. Especially for such a simple use
> >> case. It probably make senses to make a  decision for themost frequently use case, please also note
> >> that this patch didn't introduce any-restriction for the more advance
> >> uses cases(multiple bridges in the middle).
> > So, for the sake of not having the connector in the display driver,
> > you want to add boilerplate code basically to each and every bridge
> > driver. In the end, they should all behave in the same way.
> >
> > Moreover, there is no way this implementation can work without a
> > warning if there are two bridges in a chain and the it66121 is the
> > second (the last) one. The host can not specify the
> > DRM_BRIDGE_ATTACH_NO_CONNECTOR)
> >>>>>> +                       WARN_ON(1);
> >>>>> No. It is perfectly fine to create attach a bridge with no next_bridge
> >>>>> and with the DRM_BRIDGE_ATTACH_NO_CONNECTOR flag.
> >>>>>
> >>>> The document say when DRM_BRIDGE_ATTACH_NO_CONNECTOR flag is set
> >>>> the bridge shall not create a drm_connector. So I think if a display
> >>>> bridge driver don't have a next bridge attached (Currently, this is
> >>>> told by the DT), it says that this is a non-DT environment. On such
> >>>> a case, this display bridge driver(it66121.ko) should behavior like
> >>>> a *agent*. Because the upstream port of it66121 is the DVO port of
> >>>> the display controller, the downstream port of it66121 is the HDMI
> >>>> connector. it66121 is on the middle. So I think the it66121.ko should
> >>>> handle all of troubles on behalf of the display controller drivers.
> >>> No. Don't make decisions for the other drivers. They might have different needs.
> >> [...]
> >>
> >>
> >>>> Therefore (when in non-DT use case), the display controller drivers
> >>>> side should not set DRM_BRIDGE_ATTACH_NO_CONNECTOR flag anymore.
> >>>> Which is to hint that the it66121 should totally in charge of those
> >>>> tasks (either by using bridge connector helper or create a connector
> >>>> manually). I don't understand on such a case, why bother display
> >>>> controller drivers anymore.
> >>> This is the reason why we had introduced this flag. It allows the
> >>> driver to customise the connector. It even allows the driver to
> >>> implement a connector on its own, completely ignoring the
> >>> drm_bridge_connector.
> >>
> >> I know what you said is right in the sense of the universe cases,
> >> but I think the most frequent(majority) use case is that there is
> >> only one display bridge on the middle. Therefore, I don't want to
> >> movethe connector things into device driver if there is only one display
> >> bridge(say it66121) in the middle. After all, there is no *direct
> >> physical connection* from the perspective of the hardware. I means that
> >> there is no hardware wires connectthe HDMI connector and the DVO port. So display controller drivers
> >> should not interact with anything related with the connector on a
> >> perfect abstract on the software side. Especially for such a simple use
> >> case. It probably make senses to make a  decision for themost frequently use case, please also note
> >> that this patch didn't introduce any-restriction for the more advance
> >> uses cases(multiple bridges in the middle).
> > So, for the sake of not having the connector in the display driver,
> > you want to add boilerplate code basically to each and every bridge
> > driver. In the end, they should all behave in the same way.
>
> No, I'm only intend to modify the one when there has a user emerged.
> If we have the connector related code in the KMS display driver side,
> then I think we don't honor the meaning of the word *bridge*. I was
> told drm_bridge is a modern design, if we still need the DC side
> worry about something do not have a physical connection, then it will
> not be modern anymore, it is not a good bridge.

Next time the user emerges for another bridge. And then for another.
This way the very similar code is copy-pasted over all bridge drivers.
So instead it was decided to keep it in the display driver code.

>
>
> > Moreover, there is no way this implementation can work without a
> > warning if there are two bridges in a chain and the it66121 is the
> > second (the last) one.
>
> Yes and no!
>
> If one of them are transparent, then thisimplementation still can works. It is just that this will not be a good
> abstract anymore.I do have seen such design on some notebook hardware.  But from my programming experiences,
> using two bridges are typically a bad practice in engineering. As it tend
> to increase the PCB board area and increase entire cost. As you need buy
> one more TX encoder chip. Please also consider that the embedded world focus
> on low cost and low power consume.

A typical pipeline for an embedded device can perfectly look like:
- DSI host (drm_encoder)
- DSI-HDMI or DSI-eDP bridge (drm_bridge)
- hdmi-connector or panel-bridge (drm_bridge)
- drm_bridge_connector.

Two drm_bridge instances.

>
> I think, multiple display bridges case should be avoided for middle/low end
> application. Or allow us to handle the two and/or more bridges cases in the
> future. When there has at least one user emerged, we will introduce new
> approach to handle then.
>
> Do you find any product level boards that using two external display bridge and
> one of them is it66121? If we can not even find a user, we are not even have a
> board to test if current design (state of art) works. Does it suffer from module
> loading order problems? what if their i2c slave address is same? Does such a use
> case will past the S3/S4 test? All of those concerns are imposed to every display
> bridges programmer from the very beginning.

Please add a hdmi-connector device to your testing model. You don't
have to use it, but it is a fully legit use case. And suddenly you
have to drm_bridge instances in your chain.

>
> I'm agree with the idea that drm bridges drivers involved toward to a direction
> that support more complex design, but I think we should also leave a way for the
> most frequent use case. Make it straight-forward as a canonical design.

Not having anything connector-related in the drm_bridge driver is a
canonical design.

>
> > The host can not specify the
> > DRM_BRIDGE_ATTACH_NO_CONNECTOR flag, it will cause a warning here. And
> > it can not omit the flag (as otherwise the first bridge will create a
> > connector, without consulting the second bridge).
>
> The semantics of DRM_BRIDGE_ATTACH_NO_CONNECTOR flagare implement-defined,

No, they are not. Semantics are pretty simple: do not create the
drm_connector instance. Pass the flag to the next bridge in the chain.

> for our case, I could just ignore it if their
> don't have a signal(DT or ACPI) tell me that there are multiple bridges
> in the chain. This depend on community's attitude.

Ignoring a flag is a bad idea.

>
> For it66121 with a canonical design, the host should not need to specify this flag.
> Because at the time of when the device driver(it66121.ko) get loaded, the it66121
> driver could parse the DT by itself, and detect if there has a next bridge, is it a
> connector or is it yet another display bridges. The DT speak everything about the
> topology. The flag is there just for the KMS display controller driver to explicit
> control, use it and make it more useful is the right way, is it?

No. We have been there (before the DRM_BRIDGE_ATTACH_NO_CONNECTOR was
introduced), we have gone away from it.

>
>
> >>
> >>>>>> +
> >>>>>> +               connector = drm_bridge_connector_init(bridge->dev, encoder);
> >>>>>> +               if (IS_ERR(connector))
> >>>>>> +                       return PTR_ERR(connector);
> >>>>>> +
> >>>>>> +               drm_connector_attach_encoder(connector, encoder);
> >>>>> This goes into your device driver.
> >>>>>
> >>>>>> +
> >>>>>> +               ctx->connector = connector;
> >>>>>> +       }
> >>>>>>
> >>>>>>            if (ctx->info->id == ID_IT66121) {
> >>>>>>                    ret = regmap_write_bits(ctx->regmap, IT66121_CLK_BANK_REG,
> >>>>>> @@ -1632,16 +1651,13 @@ static const char * const it66121_supplies[] = {
> >>>>>>            "vcn33", "vcn18", "vrf12"
> >>>>>>     };
> >>>
> >
> >
Neil Armstrong Nov. 23, 2023, 1:39 p.m. UTC | #27
On 23/11/2023 09:08, Dmitry Baryshkov wrote:
> On Thu, 23 Nov 2023 at 07:05, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>>
>> Hi,
>>
>>
>> On 2023/11/16 19:19, Dmitry Baryshkov wrote:
>>> On Thu, 16 Nov 2023 at 12:13, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>>>> Hi,
>>>>
>>>>
>>>> On 2023/11/16 17:30, Dmitry Baryshkov wrote:
>>>>> On Thu, 16 Nov 2023 at 11:14, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>>>>>> Hi,
>>>>>>
>>>>>> Thanks a lot for reviewing!
>>>>>>
>>>>>>
>>>>>> On 2023/11/15 00:30, Dmitry Baryshkov wrote:
>>>>>>> On Tue, 14 Nov 2023 at 17:09, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>>>>>>>> From: Sui Jingfeng <suijingfeng@loongson.cn>
>>>>>>>>
>>>>>>>> The it66121_create_bridge() and it66121_destroy_bridge() are added to
>>>>>>>> export the core functionalities. Create a connector manually by using
>>>>>>>> bridge connector helpers when link as a lib.
>>>>>>>>
>>>>>>>> Signed-off-by: Sui Jingfeng <suijingfeng@loongson.cn>
>>>>>>>> ---
>>>>>>>>      drivers/gpu/drm/bridge/ite-it66121.c | 134 +++++++++++++++++++--------
>>>>>>>>      include/drm/bridge/ite-it66121.h     |  17 ++++
>>>>>>>>      2 files changed, 113 insertions(+), 38 deletions(-)
>>>>>>>>      create mode 100644 include/drm/bridge/ite-it66121.h
>>>>>>>>
>>>>>>>> diff --git a/drivers/gpu/drm/bridge/ite-it66121.c b/drivers/gpu/drm/bridge/ite-it66121.c
>>>>>>>> index 8971414a2a60..f5968b679c5d 100644
>>>>>>>> --- a/drivers/gpu/drm/bridge/ite-it66121.c
>>>>>>>> +++ b/drivers/gpu/drm/bridge/ite-it66121.c
>>>>>>>> @@ -22,6 +22,7 @@
>>>>>>>>
>>>>>>>>      #include <drm/drm_atomic_helper.h>
>>>>>>>>      #include <drm/drm_bridge.h>
>>>>>>>> +#include <drm/drm_bridge_connector.h>
>>>>>>>>      #include <drm/drm_edid.h>
>>>>>>>>      #include <drm/drm_modes.h>
>>>>>>>>      #include <drm/drm_print.h>
>>>>>>>> @@ -703,14 +704,32 @@ static int it66121_bridge_attach(struct drm_bridge *bridge,
>>>>>>>>                                      enum drm_bridge_attach_flags flags)
>>>>>>>>      {
>>>>>>>>             struct it66121_ctx *ctx = bridge_to_it66121(bridge);
>>>>>>>> +       struct drm_bridge *next_bridge = ctx->next_bridge;
>>>>>>>> +       struct drm_encoder *encoder = bridge->encoder;
>>>>>>>>             int ret;
>>>>>>>>
>>>>>>>> -       if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR))
>>>>>>>> -               return -EINVAL;
>>>>>>>> +       if (next_bridge) {
>>>>>>>> +               if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)) {
>>>>>>>> +                       WARN_ON(1);
>>>>>>> Why? At least use WARN() instead
>>>>>> Originally I want to
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>>> +                       flags |= DRM_BRIDGE_ATTACH_NO_CONNECTOR;
>>>>>>>> +               }
>>>>>>>> +               ret = drm_bridge_attach(encoder, next_bridge, bridge, flags);
>>>>>>>> +               if (ret)
>>>>>>>> +                       return ret;
>>>>>>>> +       } else {
>>>>>>>> +               struct drm_connector *connector;
>>>>>>>>
>>>>>>>> -       ret = drm_bridge_attach(bridge->encoder, ctx->next_bridge, bridge, flags);
>>>>>>>> -       if (ret)
>>>>>>>> -               return ret;
>>>>>>>> +               if (flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)
>>>>>>>> +                       WARN_ON(1);
>>>>>>> No. It is perfectly fine to create attach a bridge with no next_bridge
>>>>>>> and with the DRM_BRIDGE_ATTACH_NO_CONNECTOR flag.
>>>>>>>
>>>>>> The document say when DRM_BRIDGE_ATTACH_NO_CONNECTOR flag is set
>>>>>> the bridge shall not create a drm_connector. So I think if a display
>>>>>> bridge driver don't have a next bridge attached (Currently, this is
>>>>>> told by the DT), it says that this is a non-DT environment. On such
>>>>>> a case, this display bridge driver(it66121.ko) should behavior like
>>>>>> a *agent*. Because the upstream port of it66121 is the DVO port of
>>>>>> the display controller, the downstream port of it66121 is the HDMI
>>>>>> connector. it66121 is on the middle. So I think the it66121.ko should
>>>>>> handle all of troubles on behalf of the display controller drivers.
>>>>> No. Don't make decisions for the other drivers. They might have different needs.
>>>> [...]
>>>>
>>>>
>>>>>> Therefore (when in non-DT use case), the display controller drivers
>>>>>> side should not set DRM_BRIDGE_ATTACH_NO_CONNECTOR flag anymore.
>>>>>> Which is to hint that the it66121 should totally in charge of those
>>>>>> tasks (either by using bridge connector helper or create a connector
>>>>>> manually). I don't understand on such a case, why bother display
>>>>>> controller drivers anymore.
>>>>> This is the reason why we had introduced this flag. It allows the
>>>>> driver to customise the connector. It even allows the driver to
>>>>> implement a connector on its own, completely ignoring the
>>>>> drm_bridge_connector.
>>>>
>>>> I know what you said is right in the sense of the universe cases,
>>>> but I think the most frequent(majority) use case is that there is
>>>> only one display bridge on the middle. Therefore, I don't want to
>>>> movethe connector things into device driver if there is only one display
>>>> bridge(say it66121) in the middle. After all, there is no *direct
>>>> physical connection* from the perspective of the hardware. I means that
>>>> there is no hardware wires connectthe HDMI connector and the DVO port. So display controller drivers
>>>> should not interact with anything related with the connector on a
>>>> perfect abstract on the software side. Especially for such a simple use
>>>> case. It probably make senses to make a  decision for themost frequently use case, please also note
>>>> that this patch didn't introduce any-restriction for the more advance
>>>> uses cases(multiple bridges in the middle).
>>> So, for the sake of not having the connector in the display driver,
>>> you want to add boilerplate code basically to each and every bridge
>>> driver. In the end, they should all behave in the same way.
>>>
>>> Moreover, there is no way this implementation can work without a
>>> warning if there are two bridges in a chain and the it66121 is the
>>> second (the last) one. The host can not specify the
>>> DRM_BRIDGE_ATTACH_NO_CONNECTOR)
>>>>>>>> +                       WARN_ON(1);
>>>>>>> No. It is perfectly fine to create attach a bridge with no next_bridge
>>>>>>> and with the DRM_BRIDGE_ATTACH_NO_CONNECTOR flag.
>>>>>>>
>>>>>> The document say when DRM_BRIDGE_ATTACH_NO_CONNECTOR flag is set
>>>>>> the bridge shall not create a drm_connector. So I think if a display
>>>>>> bridge driver don't have a next bridge attached (Currently, this is
>>>>>> told by the DT), it says that this is a non-DT environment. On such
>>>>>> a case, this display bridge driver(it66121.ko) should behavior like
>>>>>> a *agent*. Because the upstream port of it66121 is the DVO port of
>>>>>> the display controller, the downstream port of it66121 is the HDMI
>>>>>> connector. it66121 is on the middle. So I think the it66121.ko should
>>>>>> handle all of troubles on behalf of the display controller drivers.
>>>>> No. Don't make decisions for the other drivers. They might have different needs.
>>>> [...]
>>>>
>>>>
>>>>>> Therefore (when in non-DT use case), the display controller drivers
>>>>>> side should not set DRM_BRIDGE_ATTACH_NO_CONNECTOR flag anymore.
>>>>>> Which is to hint that the it66121 should totally in charge of those
>>>>>> tasks (either by using bridge connector helper or create a connector
>>>>>> manually). I don't understand on such a case, why bother display
>>>>>> controller drivers anymore.
>>>>> This is the reason why we had introduced this flag. It allows the
>>>>> driver to customise the connector. It even allows the driver to
>>>>> implement a connector on its own, completely ignoring the
>>>>> drm_bridge_connector.
>>>>
>>>> I know what you said is right in the sense of the universe cases,
>>>> but I think the most frequent(majority) use case is that there is
>>>> only one display bridge on the middle. Therefore, I don't want to
>>>> movethe connector things into device driver if there is only one display
>>>> bridge(say it66121) in the middle. After all, there is no *direct
>>>> physical connection* from the perspective of the hardware. I means that
>>>> there is no hardware wires connectthe HDMI connector and the DVO port. So display controller drivers
>>>> should not interact with anything related with the connector on a
>>>> perfect abstract on the software side. Especially for such a simple use
>>>> case. It probably make senses to make a  decision for themost frequently use case, please also note
>>>> that this patch didn't introduce any-restriction for the more advance
>>>> uses cases(multiple bridges in the middle).
>>> So, for the sake of not having the connector in the display driver,
>>> you want to add boilerplate code basically to each and every bridge
>>> driver. In the end, they should all behave in the same way.
>>
>> No, I'm only intend to modify the one when there has a user emerged.
>> If we have the connector related code in the KMS display driver side,
>> then I think we don't honor the meaning of the word *bridge*. I was
>> told drm_bridge is a modern design, if we still need the DC side
>> worry about something do not have a physical connection, then it will
>> not be modern anymore, it is not a good bridge.
> 
> Next time the user emerges for another bridge. And then for another.
> This way the very similar code is copy-pasted over all bridge drivers.
> So instead it was decided to keep it in the display driver code.
> 
>>
>>
>>> Moreover, there is no way this implementation can work without a
>>> warning if there are two bridges in a chain and the it66121 is the
>>> second (the last) one.
>>
>> Yes and no!
>>
>> If one of them are transparent, then thisimplementation still can works. It is just that this will not be a good
>> abstract anymore.I do have seen such design on some notebook hardware.  But from my programming experiences,
>> using two bridges are typically a bad practice in engineering. As it tend
>> to increase the PCB board area and increase entire cost. As you need buy
>> one more TX encoder chip. Please also consider that the embedded world focus
>> on low cost and low power consume.
> 
> A typical pipeline for an embedded device can perfectly look like:
> - DSI host (drm_encoder)
> - DSI-HDMI or DSI-eDP bridge (drm_bridge)
> - hdmi-connector or panel-bridge (drm_bridge)
> - drm_bridge_connector.
> 
> Two drm_bridge instances.

Nowadays, we are moving the encoder code into bridge so we can have way
more than 2 bridges, and connector code has been moved to a bridge.

On Amlogic SoCs, the HDMI chain has at least 3 bridges, it can have up to 4
on DSI usecase if you plug a DSI to eDP bridge.

On iMX8 platform, they have multiple internal SoC bridges even before the HDMI or DSI bridge.

The model works very well across extremely heterogeneous embedded platforms.

> 
>>
>> I think, multiple display bridges case should be avoided for middle/low end
>> application. Or allow us to handle the two and/or more bridges cases in the
>> future. When there has at least one user emerged, we will introduce new
>> approach to handle then.
>>
>> Do you find any product level boards that using two external display bridge and
>> one of them is it66121? If we can not even find a user, we are not even have a
>> board to test if current design (state of art) works. Does it suffer from module
>> loading order problems? what if their i2c slave address is same? Does such a use
>> case will past the S3/S4 test? All of those concerns are imposed to every display
>> bridges programmer from the very beginning.
> 
> Please add a hdmi-connector device to your testing model. You don't
> have to use it, but it is a fully legit use case. And suddenly you
> have to drm_bridge instances in your chain.
> 
>>
>> I'm agree with the idea that drm bridges drivers involved toward to a direction
>> that support more complex design, but I think we should also leave a way for the
>> most frequent use case. Make it straight-forward as a canonical design.
> 
> Not having anything connector-related in the drm_bridge driver is a
> canonical design.
> 
>>
>>> The host can not specify the
>>> DRM_BRIDGE_ATTACH_NO_CONNECTOR flag, it will cause a warning here. And
>>> it can not omit the flag (as otherwise the first bridge will create a
>>> connector, without consulting the second bridge).
>>
>> The semantics of DRM_BRIDGE_ATTACH_NO_CONNECTOR flagare implement-defined,
> 
> No, they are not. Semantics are pretty simple: do not create the
> drm_connector instance. Pass the flag to the next bridge in the chain.
> 
>> for our case, I could just ignore it if their
>> don't have a signal(DT or ACPI) tell me that there are multiple bridges
>> in the chain. This depend on community's attitude.
> 
> Ignoring a flag is a bad idea.
> 
>>
>> For it66121 with a canonical design, the host should not need to specify this flag.
>> Because at the time of when the device driver(it66121.ko) get loaded, the it66121
>> driver could parse the DT by itself, and detect if there has a next bridge, is it a
>> connector or is it yet another display bridges. The DT speak everything about the
>> topology. The flag is there just for the KMS display controller driver to explicit
>> control, use it and make it more useful is the right way, is it?
> 
> No. We have been there (before the DRM_BRIDGE_ATTACH_NO_CONNECTOR was
> introduced), we have gone away from it.
> 
>>
>>
>>>>
>>>>>>>> +
>>>>>>>> +               connector = drm_bridge_connector_init(bridge->dev, encoder);
>>>>>>>> +               if (IS_ERR(connector))
>>>>>>>> +                       return PTR_ERR(connector);
>>>>>>>> +
>>>>>>>> +               drm_connector_attach_encoder(connector, encoder);
>>>>>>> This goes into your device driver.
>>>>>>>
>>>>>>>> +
>>>>>>>> +               ctx->connector = connector;
>>>>>>>> +       }
>>>>>>>>
>>>>>>>>             if (ctx->info->id == ID_IT66121) {
>>>>>>>>                     ret = regmap_write_bits(ctx->regmap, IT66121_CLK_BANK_REG,
>>>>>>>> @@ -1632,16 +1651,13 @@ static const char * const it66121_supplies[] = {
>>>>>>>>             "vcn33", "vcn18", "vrf12"
>>>>>>>>      };
>>>>>
>>>
>>>
> 
> 
>
Sui Jingfeng Nov. 23, 2023, 3:41 p.m. UTC | #28
Hi,


On 2023/11/23 16:08, Dmitry Baryshkov wrote:
>> The semantics of DRM_BRIDGE_ATTACH_NO_CONNECTOR flag are implement-defined,
> No, they are not. Semantics are pretty simple: do not create the
> drm_connector instance. Pass the flag to the next bridge in the chain.


Then, my problem is that do we allow create a drm_connector instance in drm bridge
driver nowadays?
Dmitry Baryshkov Nov. 23, 2023, 4:06 p.m. UTC | #29
On Thu, 23 Nov 2023 at 17:41, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>
> Hi,
>
>
> On 2023/11/23 16:08, Dmitry Baryshkov wrote:
> >> The semantics of DRM_BRIDGE_ATTACH_NO_CONNECTOR flag are implement-defined,
> > No, they are not. Semantics are pretty simple: do not create the
> > drm_connector instance. Pass the flag to the next bridge in the chain.
>
>
> Then, my problem is that do we allow create a drm_connector instance in drm bridge
> driver nowadays?

Yes, if there is no DRM_BRIDGE_ATTACH_NO_CONNECTOR. But that's deprecated IMO.
Sui Jingfeng Nov. 23, 2023, 4:20 p.m. UTC | #30
Hi,


On 2023/11/23 21:39, Neil Armstrong wrote:
> On 23/11/2023 09:08, Dmitry Baryshkov wrote:
>> On Thu, 23 Nov 2023 at 07:05, Sui Jingfeng <sui.jingfeng@linux.dev> 
>> wrote:
>>>
>>> Hi,
>>>
>>>
>>> On 2023/11/16 19:19, Dmitry Baryshkov wrote:
>>>> On Thu, 16 Nov 2023 at 12:13, Sui Jingfeng <sui.jingfeng@linux.dev> 
>>>> wrote:
>>>>> Hi,
>>>>>
>>>>>
>>>>> On 2023/11/16 17:30, Dmitry Baryshkov wrote:
>>>>>> On Thu, 16 Nov 2023 at 11:14, Sui Jingfeng 
>>>>>> <sui.jingfeng@linux.dev> wrote:
>>>>>>> Hi,
>>>>>>>
>>>>>>> Thanks a lot for reviewing!
>>>>>>>
>>>>>>>
>>>>>>> On 2023/11/15 00:30, Dmitry Baryshkov wrote:
>>>>>>>> On Tue, 14 Nov 2023 at 17:09, Sui Jingfeng 
>>>>>>>> <sui.jingfeng@linux.dev> wrote:
>>>>>>>>> From: Sui Jingfeng <suijingfeng@loongson.cn>
>>>>>>>>>
>>>>>>>>> The it66121_create_bridge() and it66121_destroy_bridge() are 
>>>>>>>>> added to
>>>>>>>>> export the core functionalities. Create a connector manually 
>>>>>>>>> by using
>>>>>>>>> bridge connector helpers when link as a lib.
>>>>>>>>>
>>>>>>>>> Signed-off-by: Sui Jingfeng <suijingfeng@loongson.cn>
>>>>>>>>> ---
>>>>>>>>>      drivers/gpu/drm/bridge/ite-it66121.c | 134 
>>>>>>>>> +++++++++++++++++++--------
>>>>>>>>>      include/drm/bridge/ite-it66121.h     |  17 ++++
>>>>>>>>>      2 files changed, 113 insertions(+), 38 deletions(-)
>>>>>>>>>      create mode 100644 include/drm/bridge/ite-it66121.h
>>>>>>>>>
>>>>>>>>> diff --git a/drivers/gpu/drm/bridge/ite-it66121.c 
>>>>>>>>> b/drivers/gpu/drm/bridge/ite-it66121.c
>>>>>>>>> index 8971414a2a60..f5968b679c5d 100644
>>>>>>>>> --- a/drivers/gpu/drm/bridge/ite-it66121.c
>>>>>>>>> +++ b/drivers/gpu/drm/bridge/ite-it66121.c
>>>>>>>>> @@ -22,6 +22,7 @@
>>>>>>>>>
>>>>>>>>>      #include <drm/drm_atomic_helper.h>
>>>>>>>>>      #include <drm/drm_bridge.h>
>>>>>>>>> +#include <drm/drm_bridge_connector.h>
>>>>>>>>>      #include <drm/drm_edid.h>
>>>>>>>>>      #include <drm/drm_modes.h>
>>>>>>>>>      #include <drm/drm_print.h>
>>>>>>>>> @@ -703,14 +704,32 @@ static int it66121_bridge_attach(struct 
>>>>>>>>> drm_bridge *bridge,
>>>>>>>>>                                      enum 
>>>>>>>>> drm_bridge_attach_flags flags)
>>>>>>>>>      {
>>>>>>>>>             struct it66121_ctx *ctx = bridge_to_it66121(bridge);
>>>>>>>>> +       struct drm_bridge *next_bridge = ctx->next_bridge;
>>>>>>>>> +       struct drm_encoder *encoder = bridge->encoder;
>>>>>>>>>             int ret;
>>>>>>>>>
>>>>>>>>> -       if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR))
>>>>>>>>> -               return -EINVAL;
>>>>>>>>> +       if (next_bridge) {
>>>>>>>>> +               if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)) {
>>>>>>>>> +                       WARN_ON(1);
>>>>>>>> Why? At least use WARN() instead
>>>>>>> Originally I want to
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>>> + flags |= DRM_BRIDGE_ATTACH_NO_CONNECTOR;
>>>>>>>>> +               }
>>>>>>>>> +               ret = drm_bridge_attach(encoder, next_bridge, 
>>>>>>>>> bridge, flags);
>>>>>>>>> +               if (ret)
>>>>>>>>> +                       return ret;
>>>>>>>>> +       } else {
>>>>>>>>> +               struct drm_connector *connector;
>>>>>>>>>
>>>>>>>>> -       ret = drm_bridge_attach(bridge->encoder, 
>>>>>>>>> ctx->next_bridge, bridge, flags);
>>>>>>>>> -       if (ret)
>>>>>>>>> -               return ret;
>>>>>>>>> +               if (flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)
>>>>>>>>> +                       WARN_ON(1);
>>>>>>>> No. It is perfectly fine to create attach a bridge with no 
>>>>>>>> next_bridge
>>>>>>>> and with the DRM_BRIDGE_ATTACH_NO_CONNECTOR flag.
>>>>>>>>
>>>>>>> The document say when DRM_BRIDGE_ATTACH_NO_CONNECTOR flag is set
>>>>>>> the bridge shall not create a drm_connector. So I think if a 
>>>>>>> display
>>>>>>> bridge driver don't have a next bridge attached (Currently, this is
>>>>>>> told by the DT), it says that this is a non-DT environment. On such
>>>>>>> a case, this display bridge driver(it66121.ko) should behavior like
>>>>>>> a *agent*. Because the upstream port of it66121 is the DVO port of
>>>>>>> the display controller, the downstream port of it66121 is the HDMI
>>>>>>> connector. it66121 is on the middle. So I think the it66121.ko 
>>>>>>> should
>>>>>>> handle all of troubles on behalf of the display controller drivers.
>>>>>> No. Don't make decisions for the other drivers. They might have 
>>>>>> different needs.
>>>>> [...]
>>>>>
>>>>>
>>>>>>> Therefore (when in non-DT use case), the display controller drivers
>>>>>>> side should not set DRM_BRIDGE_ATTACH_NO_CONNECTOR flag anymore.
>>>>>>> Which is to hint that the it66121 should totally in charge of those
>>>>>>> tasks (either by using bridge connector helper or create a 
>>>>>>> connector
>>>>>>> manually). I don't understand on such a case, why bother display
>>>>>>> controller drivers anymore.
>>>>>> This is the reason why we had introduced this flag. It allows the
>>>>>> driver to customise the connector. It even allows the driver to
>>>>>> implement a connector on its own, completely ignoring the
>>>>>> drm_bridge_connector.
>>>>>
>>>>> I know what you said is right in the sense of the universe cases,
>>>>> but I think the most frequent(majority) use case is that there is
>>>>> only one display bridge on the middle. Therefore, I don't want to
>>>>> movethe connector things into device driver if there is only one 
>>>>> display
>>>>> bridge(say it66121) in the middle. After all, there is no *direct
>>>>> physical connection* from the perspective of the hardware. I means 
>>>>> that
>>>>> there is no hardware wires connectthe HDMI connector and the DVO 
>>>>> port. So display controller drivers
>>>>> should not interact with anything related with the connector on a
>>>>> perfect abstract on the software side. Especially for such a 
>>>>> simple use
>>>>> case. It probably make senses to make a  decision for themost 
>>>>> frequently use case, please also note
>>>>> that this patch didn't introduce any-restriction for the more advance
>>>>> uses cases(multiple bridges in the middle).
>>>> So, for the sake of not having the connector in the display driver,
>>>> you want to add boilerplate code basically to each and every bridge
>>>> driver. In the end, they should all behave in the same way.
>>>>
>>>> Moreover, there is no way this implementation can work without a
>>>> warning if there are two bridges in a chain and the it66121 is the
>>>> second (the last) one. The host can not specify the
>>>> DRM_BRIDGE_ATTACH_NO_CONNECTOR)
>>>>>>>>> + WARN_ON(1);
>>>>>>>> No. It is perfectly fine to create attach a bridge with no 
>>>>>>>> next_bridge
>>>>>>>> and with the DRM_BRIDGE_ATTACH_NO_CONNECTOR flag.
>>>>>>>>
>>>>>>> The document say when DRM_BRIDGE_ATTACH_NO_CONNECTOR flag is set
>>>>>>> the bridge shall not create a drm_connector. So I think if a 
>>>>>>> display
>>>>>>> bridge driver don't have a next bridge attached (Currently, this is
>>>>>>> told by the DT), it says that this is a non-DT environment. On such
>>>>>>> a case, this display bridge driver(it66121.ko) should behavior like
>>>>>>> a *agent*. Because the upstream port of it66121 is the DVO port of
>>>>>>> the display controller, the downstream port of it66121 is the HDMI
>>>>>>> connector. it66121 is on the middle. So I think the it66121.ko 
>>>>>>> should
>>>>>>> handle all of troubles on behalf of the display controller drivers.
>>>>>> No. Don't make decisions for the other drivers. They might have 
>>>>>> different needs.
>>>>> [...]
>>>>>
>>>>>
>>>>>>> Therefore (when in non-DT use case), the display controller drivers
>>>>>>> side should not set DRM_BRIDGE_ATTACH_NO_CONNECTOR flag anymore.
>>>>>>> Which is to hint that the it66121 should totally in charge of those
>>>>>>> tasks (either by using bridge connector helper or create a 
>>>>>>> connector
>>>>>>> manually). I don't understand on such a case, why bother display
>>>>>>> controller drivers anymore.
>>>>>> This is the reason why we had introduced this flag. It allows the
>>>>>> driver to customise the connector. It even allows the driver to
>>>>>> implement a connector on its own, completely ignoring the
>>>>>> drm_bridge_connector.
>>>>>
>>>>> I know what you said is right in the sense of the universe cases,
>>>>> but I think the most frequent(majority) use case is that there is
>>>>> only one display bridge on the middle. Therefore, I don't want to
>>>>> movethe connector things into device driver if there is only one 
>>>>> display
>>>>> bridge(say it66121) in the middle. After all, there is no *direct
>>>>> physical connection* from the perspective of the hardware. I means 
>>>>> that
>>>>> there is no hardware wires connectthe HDMI connector and the DVO 
>>>>> port. So display controller drivers
>>>>> should not interact with anything related with the connector on a
>>>>> perfect abstract on the software side. Especially for such a 
>>>>> simple use
>>>>> case. It probably make senses to make a  decision for themost 
>>>>> frequently use case, please also note
>>>>> that this patch didn't introduce any-restriction for the more advance
>>>>> uses cases(multiple bridges in the middle).
>>>> So, for the sake of not having the connector in the display driver,
>>>> you want to add boilerplate code basically to each and every bridge
>>>> driver. In the end, they should all behave in the same way.
>>>
>>> No, I'm only intend to modify the one when there has a user emerged.
>>> If we have the connector related code in the KMS display driver side,
>>> then I think we don't honor the meaning of the word *bridge*. I was
>>> told drm_bridge is a modern design, if we still need the DC side
>>> worry about something do not have a physical connection, then it will
>>> not be modern anymore, it is not a good bridge.
>>
>> Next time the user emerges for another bridge. And then for another.
>> This way the very similar code is copy-pasted over all bridge drivers.
>> So instead it was decided to keep it in the display driver code.
>>
>>>
>>>
>>>> Moreover, there is no way this implementation can work without a
>>>> warning if there are two bridges in a chain and the it66121 is the
>>>> second (the last) one.
>>>
>>> Yes and no!
>>>
>>> If one of them are transparent, then thisimplementation still can 
>>> works. It is just that this will not be a good
>>> abstract anymore.I do have seen such design on some notebook 
>>> hardware.  But from my programming experiences,
>>> using two bridges are typically a bad practice in engineering. As it 
>>> tend
>>> to increase the PCB board area and increase entire cost. As you need 
>>> buy
>>> one more TX encoder chip. Please also consider that the embedded 
>>> world focus
>>> on low cost and low power consume.
>>
>> A typical pipeline for an embedded device can perfectly look like:
>> - DSI host (drm_encoder)
>> - DSI-HDMI or DSI-eDP bridge (drm_bridge)
>> - hdmi-connector or panel-bridge (drm_bridge)
>> - drm_bridge_connector.
>>
>> Two drm_bridge instances.
>
> Nowadays, we are moving the encoder code into bridge so we can have way
> more than 2 bridges, and connector code has been moved to a bridge.
>
> On Amlogic SoCs, the HDMI chain has at least 3 bridges, it can have up 
> to 4
> on DSI usecase if you plug a DSI to eDP bridge.
>
> On iMX8 platform, they have multiple internal SoC bridges even before 
> the HDMI or DSI bridge.
>
> The model works very well across extremely heterogeneous embedded 
> platforms.
>
But you don't mention the prerequisite, there have difference between
display bridge devices *inside* of the SoC and display bridge devices
*outside* of the SoC. IT66121 is a onboard device. Suppose you are the
hardware vendor ITE, you definitely want your chip can be used on any
platform. The model probably works well on ARM, but on X86 platform
there no DT support, the abstract for the onboard device becomes pale.


>>
>>>
>>> I think, multiple display bridges case should be avoided for 
>>> middle/low end
>>> application. Or allow us to handle the two and/or more bridges cases 
>>> in the
>>> future. When there has at least one user emerged, we will introduce new
>>> approach to handle then.
>>>
>>> Do you find any product level boards that using two external display 
>>> bridge and
>>> one of them is it66121? If we can not even find a user, we are not 
>>> even have a
>>> board to test if current design (state of art) works. Does it suffer 
>>> from module
>>> loading order problems? what if their i2c slave address is same? 
>>> Does such a use
>>> case will past the S3/S4 test? All of those concerns are imposed to 
>>> every display
>>> bridges programmer from the very beginning.
>>
>> Please add a hdmi-connector device to your testing model. You don't
>> have to use it, but it is a fully legit use case. And suddenly you
>> have to drm_bridge instances in your chain.
>>
>>>
>>> I'm agree with the idea that drm bridges drivers involved toward to 
>>> a direction
>>> that support more complex design, but I think we should also leave a 
>>> way for the
>>> most frequent use case. Make it straight-forward as a canonical design.
>>
>> Not having anything connector-related in the drm_bridge driver is a
>> canonical design.
>>
>>>
>>>> The host can not specify the
>>>> DRM_BRIDGE_ATTACH_NO_CONNECTOR flag, it will cause a warning here. And
>>>> it can not omit the flag (as otherwise the first bridge will create a
>>>> connector, without consulting the second bridge).
>>>
>>> The semantics of DRM_BRIDGE_ATTACH_NO_CONNECTOR flagare 
>>> implement-defined,
>>
>> No, they are not. Semantics are pretty simple: do not create the
>> drm_connector instance. Pass the flag to the next bridge in the chain.
>>
>>> for our case, I could just ignore it if their
>>> don't have a signal(DT or ACPI) tell me that there are multiple bridges
>>> in the chain. This depend on community's attitude.
>>
>> Ignoring a flag is a bad idea.
>>
>>>
>>> For it66121 with a canonical design, the host should not need to 
>>> specify this flag.
>>> Because at the time of when the device driver(it66121.ko) get 
>>> loaded, the it66121
>>> driver could parse the DT by itself, and detect if there has a next 
>>> bridge, is it a
>>> connector or is it yet another display bridges. The DT speak 
>>> everything about the
>>> topology. The flag is there just for the KMS display controller 
>>> driver to explicit
>>> control, use it and make it more useful is the right way, is it?
>>
>> No. We have been there (before the DRM_BRIDGE_ATTACH_NO_CONNECTOR was
>> introduced), we have gone away from it.
>>
>>>
>>>
>>>>>
>>>>>>>>> +
>>>>>>>>> +               connector = 
>>>>>>>>> drm_bridge_connector_init(bridge->dev, encoder);
>>>>>>>>> +               if (IS_ERR(connector))
>>>>>>>>> +                       return PTR_ERR(connector);
>>>>>>>>> +
>>>>>>>>> + drm_connector_attach_encoder(connector, encoder);
>>>>>>>> This goes into your device driver.
>>>>>>>>
>>>>>>>>> +
>>>>>>>>> +               ctx->connector = connector;
>>>>>>>>> +       }
>>>>>>>>>
>>>>>>>>>             if (ctx->info->id == ID_IT66121) {
>>>>>>>>>                     ret = regmap_write_bits(ctx->regmap, 
>>>>>>>>> IT66121_CLK_BANK_REG,
>>>>>>>>> @@ -1632,16 +1651,13 @@ static const char * const 
>>>>>>>>> it66121_supplies[] = {
>>>>>>>>>             "vcn33", "vcn18", "vrf12"
>>>>>>>>>      };
>>>>>>
>>>>
>>>>
>>
>>
>>
>
Sui Jingfeng Nov. 23, 2023, 4:29 p.m. UTC | #31
Hi,

On 2023/11/24 00:06, Dmitry Baryshkov wrote:
> On Thu, 23 Nov 2023 at 17:41, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>> Hi,
>>
>>
>> On 2023/11/23 16:08, Dmitry Baryshkov wrote:
>>>> The semantics of DRM_BRIDGE_ATTACH_NO_CONNECTOR flag are implement-defined,
>>> No, they are not. Semantics are pretty simple: do not create the
>>> drm_connector instance. Pass the flag to the next bridge in the chain.
>>
>> Then, my problem is that do we allow create a drm_connector instance in drm bridge
>> driver nowadays?
> Yes, if there is no DRM_BRIDGE_ATTACH_NO_CONNECTOR. But that's deprecated IMO.

Then can you read the code in bridge/ti-tfp410.c please?
The tfp410_attach() function will create drm_connector instance manually
if the conditional (flags == 0) is true.
Sui Jingfeng Nov. 23, 2023, 5:04 p.m. UTC | #32
Hi,


On 2023/11/23 16:08, Dmitry Baryshkov wrote:
>>> The host can not specify the
>>> DRM_BRIDGE_ATTACH_NO_CONNECTOR flag, it will cause a warning here. And
>>> it can not omit the flag (as otherwise the first bridge will create a
>>> connector, without consulting the second bridge).
>> The semantics of DRM_BRIDGE_ATTACH_NO_CONNECTOR flagare implement-defined,
> No, they are not. Semantics are pretty simple: do not create the
> drm_connector instance. Pass the flag to the next bridge in the chain.
>
>> for our case, I could just ignore it if their
>> don't have a signal(DT or ACPI) tell me that there are multiple bridges
>> in the chain. This depend on community's attitude.
> Ignoring a flag is a bad idea.


Can you also read the code in the bridge/lontium-lt8912.c please?
when flags == 0 is true, the lt8912 driver will just create
a drm_connector instance in the drm bridge drivers. The behavior
is similar with this patch in the perspective of spirit.

And the most important thing is that no matter what the flag the upstream
port is passed, lt8912 just always pass the DRM_BRIDGE_ATTACH_NO_CONNECTOR
flags to the next bridge. Does this count as a kind of ignore? or

This is to say that both the lt8912 and the tfp410 drm bridge drivers are
allowing create a drm_connector manually in drm bridge drivers. They didn't
being asked to move the drm_connector related code to display controller
driver. I don't know why I can't follow this way?

Do you really read the code before do comments or I failed to understand something?
Dmitry Baryshkov Nov. 23, 2023, 5:39 p.m. UTC | #33
On Thu, 23 Nov 2023 at 19:04, Sui Jingfeng <sui.jingfeng@linux.dev> wrote:
>
> Hi,
>
>
> On 2023/11/23 16:08, Dmitry Baryshkov wrote:
> >>> The host can not specify the
> >>> DRM_BRIDGE_ATTACH_NO_CONNECTOR flag, it will cause a warning here. And
> >>> it can not omit the flag (as otherwise the first bridge will create a
> >>> connector, without consulting the second bridge).
> >> The semantics of DRM_BRIDGE_ATTACH_NO_CONNECTOR flagare implement-defined,
> > No, they are not. Semantics are pretty simple: do not create the
> > drm_connector instance. Pass the flag to the next bridge in the chain.
> >
> >> for our case, I could just ignore it if their
> >> don't have a signal(DT or ACPI) tell me that there are multiple bridges
> >> in the chain. This depend on community's attitude.
> > Ignoring a flag is a bad idea.
>
>
> Can you also read the code in the bridge/lontium-lt8912.c please?
> when flags == 0 is true, the lt8912 driver will just create
> a drm_connector instance in the drm bridge drivers. The behavior
> is similar with this patch in the perspective of spirit.
>
> And the most important thing is that no matter what the flag the upstream
> port is passed, lt8912 just always pass the DRM_BRIDGE_ATTACH_NO_CONNECTOR
> flags to the next bridge. Does this count as a kind of ignore? or
>
> This is to say that both the lt8912 and the tfp410 drm bridge drivers are
> allowing create a drm_connector manually in drm bridge drivers. They didn't
> being asked to move the drm_connector related code to display controller
> driver. I don't know why I can't follow this way?

This is called 'legacy'.

>
> Do you really read the code before do comments or I failed to understand something?
Sui Jingfeng Nov. 23, 2023, 5:52 p.m. UTC | #34
Hi,


On 2023/11/23 16:08, Dmitry Baryshkov wrote:
>> I'm agree with the idea that drm bridges drivers involved toward to a direction
>> that support more complex design, but I think we should also leave a way for the
>> most frequent use case. Make it straight-forward as a canonical design.
> Not having anything connector-related in the drm_bridge driver is a
> canonical design.

What you said is just for the more complex uses case. I can't agree, sorry.

By choosing the word "canonical design", I means that the most frequently used
cases in practice are the canonical design, 95+% motherboards I have seen has
only one *onboard* display bridges chip. For my driver, I abstract the internal
(inside of the chip) encoder as drm_encoder and abstract the external TX chip as
drm_bridge, this design still works very well.


Originally, I means that this is a concept of the hardware design.
You are wrong even through in the software design context, the
transparent simple drm bridge drivers(simple-bridge.c) also *allow*
to create drm connector manually. I don't think I need to emulate
more example, please read the code by youself.

Canonical or not canonical is not a question to argue, if other programmers
are allowed to do such kind of abstraction, I should also allowed too. Thanks.
Maxime Ripard Nov. 24, 2023, 7:38 a.m. UTC | #35
On Fri, Nov 24, 2023 at 01:52:26AM +0800, Sui Jingfeng wrote:
> On 2023/11/23 16:08, Dmitry Baryshkov wrote:
> > > I'm agree with the idea that drm bridges drivers involved toward to a direction
> > > that support more complex design, but I think we should also leave a way for the
> > > most frequent use case. Make it straight-forward as a canonical design.
> > Not having anything connector-related in the drm_bridge driver is a
> > canonical design.
> 
> What you said is just for the more complex uses case. I can't agree, sorry.
> 
> By choosing the word "canonical design", I means that the most frequently used
> cases in practice are the canonical design, 95+% motherboards I have seen has
> only one *onboard* display bridges chip. For my driver, I abstract the internal
> (inside of the chip) encoder as drm_encoder and abstract the external TX chip as
> drm_bridge, this design still works very well.
> 
> 
> Originally, I means that this is a concept of the hardware design.
> You are wrong even through in the software design context, the
> transparent simple drm bridge drivers(simple-bridge.c) also *allow*
> to create drm connector manually. I don't think I need to emulate
> more example, please read the code by youself.

Ok. That's it. We've been patient long enough. You have been given a
review and a list of things to fix for your driver to be merged. Whether
you follow them or not is your decision.

We won't tolerate insulting comments though.

Maxime
Sui Jingfeng Nov. 24, 2023, 7:51 a.m. UTC | #36
Hi,

On 2023/11/24 15:38, Maxime Ripard wrote:
> On Fri, Nov 24, 2023 at 01:52:26AM +0800, Sui Jingfeng wrote:
>> On 2023/11/23 16:08, Dmitry Baryshkov wrote:
>>>> I'm agree with the idea that drm bridges drivers involved toward to a direction
>>>> that support more complex design, but I think we should also leave a way for the
>>>> most frequent use case. Make it straight-forward as a canonical design.
>>> Not having anything connector-related in the drm_bridge driver is a
>>> canonical design.
>> What you said is just for the more complex uses case. I can't agree, sorry.
>>
>> By choosing the word "canonical design", I means that the most frequently used
>> cases in practice are the canonical design, 95+% motherboards I have seen has
>> only one *onboard* display bridges chip. For my driver, I abstract the internal
>> (inside of the chip) encoder as drm_encoder and abstract the external TX chip as
>> drm_bridge, this design still works very well.
>>
>>
>> Originally, I means that this is a concept of the hardware design.
>> You are wrong even through in the software design context, the
>> transparent simple drm bridge drivers(simple-bridge.c) also *allow*
>> to create drm connector manually. I don't think I need to emulate
>> more example, please read the code by youself.

'emulate' -> 'enumerate'


> Ok. That's it. We've been patient long enough. You have been given a
> review and a list of things to fix for your driver to be merged.


This series is not relevant to my driver, can we please *limit*
the discussion to this series?


>   Whether
> you follow them or not is your decision.


I'm not saying that I will not follow, just to make sure what's solution is you want.
I need discussion to figure out.


> We won't tolerate insulting comments though.


There is *no* insulting, please don't misunderstanding before *sufficient* communication, OK?
Originally, I thought Dmitry may ignore(or overlook) what is the current status.


> Maxime
Maxime Ripard Nov. 24, 2023, 8:13 a.m. UTC | #37
On Fri, Nov 24, 2023 at 03:51:00PM +0800, Sui Jingfeng wrote:
> Hi,
> 
> On 2023/11/24 15:38, Maxime Ripard wrote:
> > On Fri, Nov 24, 2023 at 01:52:26AM +0800, Sui Jingfeng wrote:
> > > On 2023/11/23 16:08, Dmitry Baryshkov wrote:
> > > > > I'm agree with the idea that drm bridges drivers involved toward to a direction
> > > > > that support more complex design, but I think we should also leave a way for the
> > > > > most frequent use case. Make it straight-forward as a canonical design.
> > > > Not having anything connector-related in the drm_bridge driver is a
> > > > canonical design.
> > > What you said is just for the more complex uses case. I can't agree, sorry.
> > > 
> > > By choosing the word "canonical design", I means that the most frequently used
> > > cases in practice are the canonical design, 95+% motherboards I have seen has
> > > only one *onboard* display bridges chip. For my driver, I abstract the internal
> > > (inside of the chip) encoder as drm_encoder and abstract the external TX chip as
> > > drm_bridge, this design still works very well.
> > > 
> > > 
> > > Originally, I means that this is a concept of the hardware design.
> > > You are wrong even through in the software design context, the
> > > transparent simple drm bridge drivers(simple-bridge.c) also *allow*
> > > to create drm connector manually. I don't think I need to emulate
> > > more example, please read the code by youself.
> 
> 'emulate' -> 'enumerate'
>
> > Ok. That's it. We've been patient long enough. You have been given a
> > review and a list of things to fix for your driver to be merged.
>
> This series is not relevant to my driver, can we please *limit* the
> discussion to this series?

Right, I conflated the two, I meant this series, or the general goal to
enable that bridge with your driver. The rest of the driver is of course
unaffected.

> > Whether you follow them or not is your decision.
>
> I'm not saying that I will not follow, just to make sure what's
> solution is you want. I need discussion to figure out.

You had direct, repeated, feedback on that already by a maintainer and
one of the most experienced dev and reviewer on bridges. If you need
more guidance, you can definitely ask questions, but asking questions
and telling them they are wrong is very different.

> > We won't tolerate insulting comments though.
>
> There is *no* insulting, please don't misunderstanding before
> *sufficient* communication, OK? Originally, I thought Dmitry may
> ignore(or overlook) what is the current status.

Saying to someone maintaining and/or reviewing that code for years now
that they are wrong and should go read the code is insulting.

Maxime
Sui Jingfeng Nov. 24, 2023, 8:48 a.m. UTC | #38
Hi,

On 2023/11/24 16:13, Maxime Ripard wrote:
> On Fri, Nov 24, 2023 at 03:51:00PM +0800, Sui Jingfeng wrote:
>> Hi,
>>
>> On 2023/11/24 15:38, Maxime Ripard wrote:
>>> On Fri, Nov 24, 2023 at 01:52:26AM +0800, Sui Jingfeng wrote:
>>>> On 2023/11/23 16:08, Dmitry Baryshkov wrote:
>>>>>> I'm agree with the idea that drm bridges drivers involved toward to a direction
>>>>>> that support more complex design, but I think we should also leave a way for the
>>>>>> most frequent use case. Make it straight-forward as a canonical design.
>>>>> Not having anything connector-related in the drm_bridge driver is a
>>>>> canonical design.
>>>> What you said is just for the more complex uses case. I can't agree, sorry.
>>>>
>>>> By choosing the word "canonical design", I means that the most frequently used
>>>> cases in practice are the canonical design, 95+% motherboards I have seen has
>>>> only one *onboard* display bridges chip. For my driver, I abstract the internal
>>>> (inside of the chip) encoder as drm_encoder and abstract the external TX chip as
>>>> drm_bridge, this design still works very well.
>>>>
>>>>
>>>> Originally, I means that this is a concept of the hardware design.
>>>> You are wrong even through in the software design context, the
>>>> transparent simple drm bridge drivers(simple-bridge.c) also *allow*
>>>> to create drm connector manually. I don't think I need to emulate
>>>> more example, please read the code by youself.
>> 'emulate' -> 'enumerate'
>>
>>> Ok. That's it. We've been patient long enough. You have been given a
>>> review and a list of things to fix for your driver to be merged.
>> This series is not relevant to my driver, can we please *limit* the
>> discussion to this series?
> Right, I conflated the two, I meant this series, or the general goal to
> enable that bridge with your driver. The rest of the driver is of course
> unaffected.
>
>>> Whether you follow them or not is your decision.
>> I'm not saying that I will not follow, just to make sure what's
>> solution is you want. I need discussion to figure out.
> You had direct, repeated, feedback on that already by a maintainer and
> one of the most experienced dev and reviewer on bridges. If you need
> more guidance, you can definitely ask questions, but asking questions
> and telling them they are wrong is very different.
>
>>> We won't tolerate insulting comments though.
>> There is *no* insulting, please don't misunderstanding before
>> *sufficient* communication, OK? Originally, I thought Dmitry may
>> ignore(or overlook) what is the current status.
> Saying to someone maintaining and/or reviewing that code for years now
> that they are wrong and should go read the code is insulting.

OK, I will remind my written words in the future.
I will back to investigate for a period of time.
Thanks a lot for reviewing.


> Maxime
Sui Jingfeng Nov. 25, 2023, 2:30 a.m. UTC | #39
On 2023/11/24 16:13, Maxime Ripard wrote:
> On Fri, Nov 24, 2023 at 03:51:00PM +0800, Sui Jingfeng wrote:
>> Hi,
>>
>> On 2023/11/24 15:38, Maxime Ripard wrote:
>>> On Fri, Nov 24, 2023 at 01:52:26AM +0800, Sui Jingfeng wrote:
>>>> On 2023/11/23 16:08, Dmitry Baryshkov wrote:
>>>>>> I'm agree with the idea that drm bridges drivers involved toward to a direction
>>>>>> that support more complex design, but I think we should also leave a way for the
>>>>>> most frequent use case. Make it straight-forward as a canonical design.
>>>>> Not having anything connector-related in the drm_bridge driver is a
>>>>> canonical design.
>>>> What you said is just for the more complex uses case. I can't agree, sorry.
>>>>
>>>> By choosing the word "canonical design", I means that the most frequently used
>>>> cases in practice are the canonical design, 95+% motherboards I have seen has
>>>> only one *onboard* display bridges chip. For my driver, I abstract the internal
>>>> (inside of the chip) encoder as drm_encoder and abstract the external TX chip as
>>>> drm_bridge, this design still works very well.
>>>>
>>>>
>>>> Originally, I means that this is a concept of the hardware design.
>>>> You are wrong even through in the software design context, the
>>>> transparent simple drm bridge drivers(simple-bridge.c) also *allow*
>>>> to create drm connector manually. I don't think I need to emulate
>>>> more example, please read the code by youself.
>> 'emulate' -> 'enumerate'
>>
>>> Ok. That's it. We've been patient long enough. You have been given a
>>> review and a list of things to fix for your driver to be merged.
>> This series is not relevant to my driver, can we please *limit* the
>> discussion to this series?
> Right, I conflated the two, I meant this series, or the general goal to
> enable that bridge with your driver. The rest of the driver is of course
> unaffected.
>
>>> Whether you follow them or not is your decision.
>> I'm not saying that I will not follow, just to make sure what's
>> solution is you want. I need discussion to figure out.
> You had direct, repeated, feedback on that already by a maintainer and
> one of the most experienced dev and reviewer on bridges. If you need
> more guidance, you can definitely ask questions, but asking questions
> and telling them they are wrong is very different.
>
>>> We won't tolerate insulting comments though.
>> There is *no* insulting, please don't misunderstanding before
>> *sufficient* communication, OK? Originally, I thought Dmitry may
>> ignore(or overlook) what is the current status.
> Saying to someone maintaining and/or reviewing that code for years now
> that they are wrong and should go read the code is insulting.


Back to that time, I'm focus on kindly technique debating, this is a kind
of defensive reply for the patch. This is a kind of remind in case of ignores.
Probably a bit of negative, but please don't misunderstanding as insult anyway.

Dmitry, really thanks a lot for the instructs, I have learned a lot while
talking with you. I will back to try mentioned method.
diff mbox series

Patch

diff --git a/drivers/gpu/drm/bridge/ite-it66121.c b/drivers/gpu/drm/bridge/ite-it66121.c
index 8971414a2a60..f5968b679c5d 100644
--- a/drivers/gpu/drm/bridge/ite-it66121.c
+++ b/drivers/gpu/drm/bridge/ite-it66121.c
@@ -22,6 +22,7 @@ 
 
 #include <drm/drm_atomic_helper.h>
 #include <drm/drm_bridge.h>
+#include <drm/drm_bridge_connector.h>
 #include <drm/drm_edid.h>
 #include <drm/drm_modes.h>
 #include <drm/drm_print.h>
@@ -703,14 +704,32 @@  static int it66121_bridge_attach(struct drm_bridge *bridge,
 				 enum drm_bridge_attach_flags flags)
 {
 	struct it66121_ctx *ctx = bridge_to_it66121(bridge);
+	struct drm_bridge *next_bridge = ctx->next_bridge;
+	struct drm_encoder *encoder = bridge->encoder;
 	int ret;
 
-	if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR))
-		return -EINVAL;
+	if (next_bridge) {
+		if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)) {
+			WARN_ON(1);
+			flags |= DRM_BRIDGE_ATTACH_NO_CONNECTOR;
+		}
+		ret = drm_bridge_attach(encoder, next_bridge, bridge, flags);
+		if (ret)
+			return ret;
+	} else {
+		struct drm_connector *connector;
 
-	ret = drm_bridge_attach(bridge->encoder, ctx->next_bridge, bridge, flags);
-	if (ret)
-		return ret;
+		if (flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)
+			WARN_ON(1);
+
+		connector = drm_bridge_connector_init(bridge->dev, encoder);
+		if (IS_ERR(connector))
+			return PTR_ERR(connector);
+
+		drm_connector_attach_encoder(connector, encoder);
+
+		ctx->connector = connector;
+	}
 
 	if (ctx->info->id == ID_IT66121) {
 		ret = regmap_write_bits(ctx->regmap, IT66121_CLK_BANK_REG,
@@ -1632,16 +1651,13 @@  static const char * const it66121_supplies[] = {
 	"vcn33", "vcn18", "vrf12"
 };
 
-static int it66121_probe(struct i2c_client *client)
+int it66121_create_bridge(struct i2c_client *client, bool of_support,
+			  bool hpd_support, bool audio_support,
+			  struct drm_bridge **bridge)
 {
+	struct device *dev = &client->dev;
 	int ret;
 	struct it66121_ctx *ctx;
-	struct device *dev = &client->dev;
-
-	if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) {
-		dev_err(dev, "I2C check functionality failed.\n");
-		return -ENXIO;
-	}
 
 	ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
 	if (!ctx)
@@ -1649,24 +1665,19 @@  static int it66121_probe(struct i2c_client *client)
 
 	ctx->dev = dev;
 	ctx->client = client;
-	ctx->info = i2c_get_match_data(client);
-
-	ret = it66121_of_read_bus_width(dev, &ctx->bus_width);
-	if (ret)
-		return ret;
-
-	ret = it66121_of_get_next_bridge(dev, &ctx->next_bridge);
-	if (ret)
-		return ret;
-
-	i2c_set_clientdata(client, ctx);
 	mutex_init(&ctx->lock);
 
-	ret = devm_regulator_bulk_get_enable(dev, ARRAY_SIZE(it66121_supplies),
-					     it66121_supplies);
-	if (ret) {
-		dev_err(dev, "Failed to enable power supplies\n");
-		return ret;
+	if (of_support) {
+		ret = it66121_of_read_bus_width(dev, &ctx->bus_width);
+		if (ret)
+			return ret;
+
+		ret = it66121_of_get_next_bridge(dev, &ctx->next_bridge);
+		if (ret)
+			return ret;
+	} else {
+		ctx->bus_width = 24;
+		ctx->next_bridge = NULL;
 	}
 
 	it66121_hw_reset(ctx);
@@ -1679,33 +1690,80 @@  static int it66121_probe(struct i2c_client *client)
 	if (ret)
 		return ret;
 
-	if (ctx->vender_id != ctx->info->vid ||
-	    ctx->device_id != ctx->info->pid)
+	ctx->info = it66121_get_match_data(ctx->vender_id, ctx->device_id);
+	if (!ctx->info)
 		return -ENODEV;
 
-	ret = devm_request_threaded_irq(dev, client->irq, NULL,	it66121_irq_threaded_handler,
-					IRQF_ONESHOT, dev_name(dev), ctx);
-	if (ret < 0) {
-		dev_err(dev, "Failed to request irq %d:%d\n", client->irq, ret);
-		return ret;
+	if (hpd_support) {
+		ret = devm_request_threaded_irq(dev, client->irq, NULL,
+						it66121_irq_threaded_handler,
+						IRQF_ONESHOT, dev_name(dev),
+						ctx);
+		if (ret < 0) {
+			dev_err(dev, "Failed to request irq: %d\n", ret);
+			return ret;
+		}
 	}
 
 	it66121_bridge_init_base(&ctx->bridge, dev->of_node, true);
 
-	it66121_audio_codec_init(ctx, dev);
+	if (audio_support)
+		it66121_audio_codec_init(ctx, dev);
+
+	*bridge = &ctx->bridge;
 
 	dev_info(dev, "IT66121 probed, chip id: 0x%x:0x%x, revision: %u\n",
 		 ctx->vender_id, ctx->device_id, ctx->revision);
 
 	return 0;
 }
+EXPORT_SYMBOL_GPL(it66121_create_bridge);
+
+static int it66121_probe(struct i2c_client *client)
+{
+	struct device *dev = &client->dev;
+	struct it66121_ctx *ctx;
+	struct drm_bridge *bridge;
+	int ret;
+
+	if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) {
+		dev_err(dev, "I2C check functionality failed.\n");
+		return -ENXIO;
+	}
+
+	ret = devm_regulator_bulk_get_enable(dev, ARRAY_SIZE(it66121_supplies),
+					     it66121_supplies);
+	if (ret) {
+		dev_err(dev, "Failed to enable power supplies\n");
+		return ret;
+	}
+
+	ret = it66121_create_bridge(client, true, true, true, &bridge);
+	if (ret)
+		return ret;
+
+	ctx = bridge_to_it66121(bridge);
+
+	i2c_set_clientdata(client, ctx);
+
+	return 0;
+}
+
+void it66121_destroy_bridge(struct drm_bridge *bridge)
+{
+	struct it66121_ctx *ctx = bridge_to_it66121(bridge);
+
+	drm_bridge_remove(bridge);
+
+	mutex_destroy(&ctx->lock);
+}
+EXPORT_SYMBOL_GPL(it66121_destroy_bridge);
 
 static void it66121_remove(struct i2c_client *client)
 {
 	struct it66121_ctx *ctx = i2c_get_clientdata(client);
 
-	drm_bridge_remove(&ctx->bridge);
-	mutex_destroy(&ctx->lock);
+	it66121_destroy_bridge(&ctx->bridge);
 }
 
 static const struct of_device_id it66121_dt_match[] = {
diff --git a/include/drm/bridge/ite-it66121.h b/include/drm/bridge/ite-it66121.h
new file mode 100644
index 000000000000..e6753f695b7f
--- /dev/null
+++ b/include/drm/bridge/ite-it66121.h
@@ -0,0 +1,17 @@ 
+/* SPDX-License-Identifier: GPL-2.0-only */
+
+#ifndef __ITE_IT66121_H__
+#define __ITE_IT66121_H__
+
+#include <linux/i2c.h>
+
+#include <drm/drm_bridge.h>
+#include <drm/drm_device.h>
+
+int it66121_create_bridge(struct i2c_client *client, bool of_support,
+			  bool hpd_support, bool audio_support,
+			  struct drm_bridge **bridge);
+
+void it66121_destroy_bridge(struct drm_bridge *bridge);
+
+#endif