linux-kernel - Re: [PATCH v14 1/2] drm: add kms driver for loongson display controller

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <d8e7a1ee-317c-6b44-27eb-ea637f8813ec@189.cn>
Date:   Mon, 22 May 2023 16:29:29 +0800
From:   Sui Jingfeng <15330273260@....cn>
To:     WANG Xuerui <kernel@...0n.name>,
        Maarten Lankhorst <maarten.lankhorst@...ux.intel.com>,
        Maxime Ripard <mripard@...nel.org>,
        Thomas Zimmermann <tzimmermann@...e.de>,
        David Airlie <airlied@...il.com>,
        Daniel Vetter <daniel@...ll.ch>,
        Sui Jingfeng <suijingfeng@...ngson.cn>,
        Li Yi <liyi@...ngson.cn>,
        Sumit Semwal <sumit.semwal@...aro.org>,
        Christian Koenig <christian.koenig@....com>,
        Emil Velikov <emil.l.velikov@...il.com>
Cc:     linaro-mm-sig@...ts.linaro.org, loongson-kernel@...ts.loongnix.cn,
        Geert Uytterhoeven <geert+renesas@...der.be>,
        linux-kernel@...r.kernel.org, dri-devel@...ts.freedesktop.org,
        Javier Martinez Canillas <javierm@...hat.com>,
        Nathan Chancellor <nathan@...nel.org>,
        Liu Peibao <liupeibao@...ngson.cn>, linux-media@...r.kernel.org
Subject: Re: [PATCH v14 1/2] drm: add kms driver for loongson display
 controller


On 2023/5/22 16:09, WANG Xuerui wrote:
> On 2023/5/22 16:02, Sui Jingfeng wrote:
>> Hi,
>>
>> On 2023/5/21 20:21, WANG Xuerui wrote:
>>>> --- /dev/null
>>>> +++ b/drivers/gpu/drm/loongson/Kconfig
>>>> @@ -0,0 +1,17 @@
>>>> +# SPDX-License-Identifier: GPL-2.0
>>>> +
>>>> +config DRM_LOONGSON
>>>> +    tristate "DRM support for Loongson Graphics"
>>>> +    depends on DRM && PCI && MMU
>>>> +    select DRM_KMS_HELPER
>>>> +    select DRM_TTM
>>>> +    select I2C
>>>> +    select I2C_ALGOBIT
>>>> +    help
>>>> +      This is a DRM driver for Loongson Graphics, it may including
>>>
>>> Drop "it may"; "including" should be enough.
>>>
>> 'it may' is more *precise* here, because currently we don't ship with 
>> the support for loongson 2K series SoC.
>>
>> I'm try to be precise as far as I can, we avoid made this driver too 
>> large by ignore loongson 2K series SoC temporary.
>
> That's a good idea! For now the patch is so large that my review reply 
> is said to be dropped by the lists. Focusing on one bunch of similar 
> models first then adding support for the rest not-so-similar models is 
> very friendly towards the reviewing process and will help code quality 
> too.
>
>>
>>>> +      LS7A2000, LS7A1000, LS2K2000 and LS2K1000 etc. Loongson LS7A
>>>> +      series are bridge chipset, while Loongson LS2K series are SoC.
>>>> +
>>>> +      If "M" is selected, the module will be called loongson.
>>>
>>> Just "loongson"? 
>>
>> Yes,  when compile this driver as module,  loongson.ko will be 
>> generated.
>>
>>   drm radeon is also doing so, See drm/radeon/Kconfig.
>>
>>> I know it's like this for ages (at least dating back to the MIPS 
>>> days) but you really don't want to imply Loongson is mainly a GPU 
>>> company. Something like "loongson_drm" or "lsdc" or "gsgpu" could be 
>>> better. 
>>
>> No, these name may have backward compatibility problems.
>>
>> Downstream driver already taken those name.
>>
>> userspace driver need to differentiate them who is who.
>
> IMO this shouldn't be a problem. Let me try explaining this: 
> currently, upstream / the "new world" doesn't have any support for 
> this driver at all, so any name will work; just use whatever is 
> appropriate from an upstream's perspective, then make the userspace 
> bits recognize both variants, and you'll be fine. And the "existing" 
> userspace drivers can also carry the change, it'll just be a branch 
> never taken in that setup.
>
> So, I'm still in favor of keeping the upstream "clean" without dubious 
> names like this (bare "loongson"). What do you think about my 
> suggestion above?
>
No,

there is a 'arm' folder in the drivers/gpu/drm/,  It doesn't say that 
arm is a pure gpu company.

there is a 'ingenic' folder in the drivers/gpu/drm/, ingenic also have 
their own custom CPUs.

there is a 'amd' folder in the drivers/gpu/drm/, these doesn't imply amd 
is mainly a GPU company.

when a folder emerged in drm/, it stand for the GPU related part of this 
company.