AuroraMiddleware/skia2
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

same 16->8 bit packing trick for SSE2/SSE4.1

Browse Source 
It's funny how now that I'm on a machine that doesn't support AVX2,
it's suddenly important for me that pack() is optimized for SSE!

This is basically the same as this morning, without any weird AVX2
pack ordering issues.  This replaces something like

    movdqa     2300(%rip), %xmm0
    pshufb     %xmm0, %xmm3
    pshufb     %xmm0, %xmm2
    punpcklqdq %xmm3, %xmm2
    (This is SSE4.1; the SSE2 version is worse.)

with

    psrlw    $8, %xmm3
    psrlw    $8, %xmm2
    packuswb %xmm3, %xmm2
    (SSE2 and SSE4.1 both.)

It's always nice to not need to load a shuffle mask out of memory.

Change-Id: I56fb30b31fcedc0ee84a4a71c483a597c8dc1622
Reviewed-on: https://skia-review.googlesource.com/30583
Reviewed-by: Florin Malita <fmalita@chromium.org>
Commit-Queue: Mike Klein <mtklein@chromium.org>
This commit is contained in:
Mike Klein 2017-08-03 12:18:24 -04:00 committed by Skia Commit-Bot
parent d0677bc44f
commit 29dbae9acf
3 changed files with 777 additions and 1026 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

899
src/jumper/SkJumper_generated.S
View File

File diff suppressed because it is too large Load Diff

899
src/jumper/SkJumper_generated_win.S
View File

File diff suppressed because it is too large Load Diff

5
src/jumper/SkJumper_stages_8bit.cpp
View File

@ -65,6 +65,11 @@ SI U8x4 pack(U16x4 v) {
auto _02 = _mm256_permute2x128_si256(lo,hi, 0x20),
_13 = _mm256_permute2x128_si256(lo,hi, 0x31);
return _mm256_packus_epi16(_02, _13);
#elif defined(__SSE2__)
static_assert(sizeof(v) == 32, "");
auto lo = unaligned_load<__m128i>((char*)&v + 0),
hi = unaligned_load<__m128i>((char*)&v + 16);
return _mm_packus_epi16(lo,hi);
#else
return __builtin_convertvector(v, U8x4);
#endif