[Mesa-dev] [PATCH 1/4] radeonsi: disable DCC MSAA for 128bpp formats on Stoney
Marek Olšák
maraeo at gmail.com
Fri Jun 22 21:16:36 UTC 2018
On Fri, Jun 22, 2018 at 12:48 PM, Dylan Baker <dylan at pnwbakers.com> wrote:
> Quoting Marek Olšák (2018-06-20 18:01:55)
>> From: Marek Olšák <marek.olsak at amd.com>
>>
>> Cc: 18.1 <mesa-stable at lists.freedesktop.org>
>> ---
>> src/gallium/drivers/radeonsi/si_texture.c | 5 +++++
>> 1 file changed, 5 insertions(+)
>>
>> diff --git a/src/gallium/drivers/radeonsi/si_texture.c b/src/gallium/drivers/radeonsi/si_texture.c
>> index b46208be252..05d5d1b8a6d 100644
>> --- a/src/gallium/drivers/radeonsi/si_texture.c
>> +++ b/src/gallium/drivers/radeonsi/si_texture.c
>> @@ -266,20 +266,25 @@ static int si_init_surface(struct si_screen *sscreen,
>> if (is_stencil)
>> flags |= RADEON_SURF_SBUFFER;
>> }
>>
>> if (sscreen->info.chip_class >= VI &&
>> (ptex->flags & SI_RESOURCE_FLAG_DISABLE_DCC ||
>> ptex->format == PIPE_FORMAT_R9G9B9E5_FLOAT ||
>> (ptex->nr_samples >= 2 && !sscreen->dcc_msaa_allowed)))
>> flags |= RADEON_SURF_DISABLE_DCC;
>>
>> + /* Stoney: 128bpp MSAA textures randomly fail piglit tests with DCC. */
>> + if (sscreen->info.family == CHIP_STONEY &&
>> + bpe == 16 && ptex->nr_samples >= 2)
>> + flags |= RADEON_SURF_DISABLE_DCC;
>> +
>> /* VI: DCC clear for 4x and 8x MSAA array textures unimplemented. */
>> if (sscreen->info.chip_class == VI &&
>> num_color_samples >= 4 &&
>> ptex->array_size > 1)
>> flags |= RADEON_SURF_DISABLE_DCC;
>>
>> /* GFX9: DCC clear for 4x and 8x MSAA textures unimplemented. */
>> if (sscreen->info.chip_class >= GFX9 &&
>> num_color_samples >= 4)
>> flags |= RADEON_SURF_DISABLE_DCC;
>> --
>> 2.17.1
>>
>> _______________________________________________
>> mesa-dev mailing list
>> mesa-dev at lists.freedesktop.org
>> https://lists.freedesktop.org/mailman/listinfo/mesa-dev
>
> Hi Marek,
>
> For this to apply to 18.1 I needed to pull
> 1cc2e0cc6b47bd5efbf2af266405060785085e6b
> "radeonsi: fully enable 2x DCC MSAA for array and non-array textures",
> is that reasonable, or would you like me to do something else?
OK, then this fix is not necessary.
Thanks,
Marek
More information about the mesa-dev
mailing list