Hello Geert,
On 2/9/22 15:27, Geert Uytterhoeven wrote:
> Hi Andy,
>
> On Wed, Feb 9, 2022 at 2:48 PM Andy Shevchenko
> <andriy.shevchenko@xxxxxxxxxxxxxxx> wrote:
>> On Tue, Feb 08, 2022 at 04:10:49PM +0100, Javier Martinez Canillas wrote:
>>> On 2/8/22 15:19, Geert Uytterhoeven wrote:
>>>> On Fri, Feb 4, 2022 at 2:43 PM Javier Martinez Canillas
>>>> <javierm@xxxxxxxxxx> wrote:
>>>> - Kernel size increased by 349 KiB,
>>>> - The "Memory:" line reports 412 KiB less memory,
>>>> - On top of that, "free" shows ca. 92 KiB more memory in use after
>>>> bootup.
>>
>> The memory consumption should really be taken seriously, because these kind of
>> displays are for embedded platforms with limited amount of resources.
>
> Thanks for your concern!
>
> Looking at the options that are auto-enabled, a few stand out that
> look like they're not needed on systems witch such small displays,
> or on legacy systems predating DDC:
Thanks for your analysis.
Since drivers are replacing the {simple,efi}fb drivers and others with the
simpledrm driver, the DRM subsystem is now built into the kernel and no
longer a loadable module.
So there has been some effort to make it more modular and smaller, as an
example the following patch-set from Thomas:
https://www.spinics.net/lists/dri-devel/msg329120.html
But there are still a lot of room to reduce this and certainly enabling
CONFIG_DRM will be noticeable for such memory constrainted systems.
This is outside the scope of this patch series though, that is only about
adding a new DRM driver :)
Now, this is a reason why I mentioned that the old fbdev driver shouldn't
be removed yet.
Best regards,
--
Javier Martinez Canillas
Linux Engineering
Red Hat
