[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <8737da50dj.fsf@eliezer.anholt.net>
Date: Fri, 14 Apr 2017 11:07:20 -0700
From: Eric Anholt <eric@...olt.net>
To: Boris Brezillon <boris.brezillon@...e-electrons.com>
Cc: dri-devel@...ts.freedesktop.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] drm/vc4: Allow using more than 256MB of CMA memory.
Boris Brezillon <boris.brezillon@...e-electrons.com> writes:
> On Mon, 27 Mar 2017 16:10:25 -0700
> Eric Anholt <eric@...olt.net> wrote:
>
>> Until now, we've had to limit Raspberry Pi to 256MB of CMA memory to
>> keep from triggering the hardware addressing bug between of the tile
>> binner of the tile alloc memory (where the top 4 bits come from the
>> tile state data array's address).
>>
>> To work around that and allow more memory to be reserved for graphics,
>> allocate a single BO to store tile state data arrays and tile
>> alloc/overflow memory while the GPU is active, and make sure that that
>> one BO doesn't happen to cross a 256MB boundary. With that in place,
>> we can allocate textures and shaders anywhere in system memory (still
>> contiguous, of course).
>
> It's hard to review something I don't quite understand, but I didn't
> spot any problem and the code seems to follow what the commit message
> says: make sure the memory used by the tile binner (still have to look
> at what a tile binner is :-)) does not cross a 256MB.
>
> So, not sure my review has a real value here, but
>
> Reviewed-by: Boris Brezillon <boris.brezillon@...e-electrons.com>
It's still more review than vc4 patches used to get, so I'm happy.
Thanks!
Download attachment "signature.asc" of type "application/pgp-signature" (833 bytes)
Powered by blists - more mailing lists