[Mesa-dev] [PATCH 1/4] i915: Fix streamling loads for intel_tiled_memcpy

[Kenneth Graunke]

On Friday, May 25, 2018 4:33:56 PM PDT Chris Wilson wrote:
> We stream from a tiled and aligned source into an unaligned user buffer,
> so we need to use _mm_storeu_si128.
> ---
>  src/mesa/drivers/dri/i965/intel_tiled_memcpy.c | 10 +++++-----
>  1 file changed, 5 insertions(+), 5 deletions(-)
> 
> diff --git a/src/mesa/drivers/dri/i965/intel_tiled_memcpy.c b/src/mesa/drivers/dri/i965/intel_tiled_memcpy.c
> index fac5427d2ed..6440dceac36 100644
> --- a/src/mesa/drivers/dri/i965/intel_tiled_memcpy.c
> +++ b/src/mesa/drivers/dri/i965/intel_tiled_memcpy.c
> @@ -223,17 +223,17 @@ _memcpy_streaming_load(void *dest, const void *src, size_t count)
>  {
>     if (count == 16) {
>        __m128i val = _mm_stream_load_si128((__m128i *)src);
> -      _mm_store_si128((__m128i *)dest, val);
> +      _mm_storeu_si128((__m128i *)dest, val);
>        return dest;
>     } else if (count == 64) {
>        __m128i val0 = _mm_stream_load_si128(((__m128i *)src) + 0);
>        __m128i val1 = _mm_stream_load_si128(((__m128i *)src) + 1);
>        __m128i val2 = _mm_stream_load_si128(((__m128i *)src) + 2);
>        __m128i val3 = _mm_stream_load_si128(((__m128i *)src) + 3);
> -      _mm_store_si128(((__m128i *)dest) + 0, val0);
> -      _mm_store_si128(((__m128i *)dest) + 1, val1);
> -      _mm_store_si128(((__m128i *)dest) + 2, val2);
> -      _mm_store_si128(((__m128i *)dest) + 3, val3);
> +      _mm_storeu_si128(((__m128i *)dest) + 0, val0);
> +      _mm_storeu_si128(((__m128i *)dest) + 1, val1);
> +      _mm_storeu_si128(((__m128i *)dest) + 2, val2);
> +      _mm_storeu_si128(((__m128i *)dest) + 3, val3);
>        return dest;
>     } else {
>        assert(count < 64); /* and (count < 16) for ytiled */
> 

Fixes: d21c086d819d78fb3f6abcbb14aa492970f442aa (i965/tiled_memcpy: inline movntdqa loads in tiled_to_linear)
Reviewed-by: Kenneth Graunke <kenneth at whitecape.org>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: This is a digitally signed message part.
URL: <https://lists.freedesktop.org/archives/mesa-dev/attachments/20180525/2439ac40/attachment.sig>