[Mesa-dev] [PATCH 11/13] i965/miptree: Map with movntdqa for linear buffers only

[Scott D Phillips]

Kenneth Graunke <kenneth at whitecape.org> writes:

> On Monday, April 30, 2018 10:25:50 AM PDT Scott D Phillips wrote:
>> Removes a place where gtt mapping is used.
>> 
>> Reviewed-by: Nanley Chery <nanley.g.chery at intel.com>
>> Reviewed-by: Chris Wilson <chris at chris-wilson.co.uk>
>> ---
>>  src/mesa/drivers/dri/i965/intel_mipmap_tree.c | 3 ++-
>>  1 file changed, 2 insertions(+), 1 deletion(-)
>> 
>> diff --git a/src/mesa/drivers/dri/i965/intel_mipmap_tree.c b/src/mesa/drivers/dri/i965/intel_mipmap_tree.c
>> index 498eebd2f86..b3b27c50de3 100644
>> --- a/src/mesa/drivers/dri/i965/intel_mipmap_tree.c
>> +++ b/src/mesa/drivers/dri/i965/intel_mipmap_tree.c
>> @@ -3770,7 +3770,8 @@ intel_miptree_map(struct brw_context *brw,
>>  #if defined(USE_SSE41)
>>     } else if (!(mode & GL_MAP_WRITE_BIT) &&
>>                !mt->compressed && cpu_has_sse4_1 &&
>> -              (mt->surf.row_pitch % 16 == 0)) {
>> +              (mt->surf.row_pitch % 16 == 0) &&
>> +              (mt->surf.tiling == ISL_TILING_LINEAR)) {
>>        intel_miptree_map_movntdqa(brw, mt, map, level, slice);
>>  #endif
>>     } else if (mt->surf.tiling != ISL_TILING_LINEAR &&
>> 
>
> You added intel_miptree_map_tiled_memcpy near the end, but couldn't you
> just put it before intel_miptree_map_movntdqa?  Then it would try that
> first and fall back to this if it couldn't be handled.

Ya, I can definitely do that. I was putting tiled_memcpy right where
gtt_map was in the hopes of making a small change, but that could be
more sensible.

> --Ken