Message-ID: <297d2c67-62ad-4cb7-8f30-d87c90c084ad@linux.dev>
Date:   Fri, 24 Nov 2023 00:20:10 +0800
From:   Sui Jingfeng <sui.jingfeng@...ux.dev>
To:     neil.armstrong@...aro.org,
        Dmitry Baryshkov <dmitry.baryshkov@...aro.org>
Cc:     Phong LE <ple@...libre.com>, Maxime Ripard <mripard@...nel.org>,
        Sui Jingfeng <suijingfeng@...ngson.cn>,
        linux-kernel@...r.kernel.org, dri-devel@...ts.freedesktop.org,
        Laurent Pinchart <Laurent.pinchart@...asonboard.com>,
        Thomas Zimmermann <tzimmermann@...e.de>
Subject: Re: [PATCH 8/8] drm/bridge: it66121: Allow link this driver as a lib

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@...ux.dev> 
>> wrote:
>>>
>>> Hi,
>>>
>>>
>>> On 2023/11/16 19:19, Dmitry Baryshkov wrote:
>>>> On Thu, 16 Nov 2023 at 12:13, Sui Jingfeng <sui.jingfeng@...ux.dev> 
>>>> wrote:
>>>>> Hi,
>>>>>
>>>>>
>>>>> On 2023/11/16 17:30, Dmitry Baryshkov wrote:
>>>>>> On Thu, 16 Nov 2023 at 11:14, Sui Jingfeng 
>>>>>> <sui.jingfeng@...ux.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@...ux.dev> wrote:
>>>>>>>>> From: Sui Jingfeng <suijingfeng@...ngson.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@...ngson.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"
>>>>>>>>>      };
>>>>>>
>>>>
>>>>
>>
>>
>>
>

Powered by blists - more mailing lists