python-pillow/Pillow
1
1
Fork 0
mirror of https://github.com/python-pillow/Pillow.git synced 2025-08-20 20:24:45 +03:00
Code Issues Packages Projects Releases Wiki Activity

SIMD Filter. Rearrange instruction for speedup

Browse Source
This commit is contained in:
Alexander 2017-09-13 03:22:03 +03:00 committed by Alexander Karpinsky
parent 74d95f5e9d
commit ab50dde9d2
1 changed files with 71 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

79
libImaging/FilterSIMD_3x3i_4u8.c
View File

@ -58,20 +58,83 @@ ImagingFilter3x3i_4u8(Imaging imOut, Imaging im, const INT16* kernel,
__m128i ss1 = _mm_set1_epi32(offset); __m128i ss1 = _mm_set1_epi32(offset);
__m128i ss2 = _mm_set1_epi32(offset); __m128i ss2 = _mm_set1_epi32(offset);
__m128i ss3 = _mm_set1_epi32(offset); __m128i ss3 = _mm_set1_epi32(offset);
__m128i tmp;
tmp = _mm_unpacklo_epi8(pix00, _mm_setzero_si128());
ss0 = _mm_add_epi32(ss0, _mm_madd_epi16(
tmp, _mm_unpacklo_epi32(kernel00, kernel00)));
ss1 = _mm_add_epi32(ss1, _mm_madd_epi16(
tmp, _mm_unpacklo_epi32(kernel01, kernel01)));
tmp = _mm_unpackhi_epi8(pix00, _mm_setzero_si128());
ss0 = _mm_add_epi32(ss0, _mm_madd_epi16(
tmp, _mm_unpackhi_epi32(kernel00, kernel00)));
ss1 = _mm_add_epi32(ss1, _mm_madd_epi16(
tmp, _mm_unpackhi_epi32(kernel01, kernel01)));
tmp = _mm_unpacklo_epi8(pix10, _mm_setzero_si128());
ss0 = _mm_add_epi32(ss0, _mm_madd_epi16(
tmp, _mm_unpacklo_epi32(kernel10, kernel10)));
ss1 = _mm_add_epi32(ss1, _mm_madd_epi16(
tmp, _mm_unpacklo_epi32(kernel11, kernel11)));
tmp = _mm_unpackhi_epi8(pix10, _mm_setzero_si128());
ss0 = _mm_add_epi32(ss0, _mm_madd_epi16(
tmp, _mm_unpackhi_epi32(kernel10, kernel10)));
ss1 = _mm_add_epi32(ss1, _mm_madd_epi16(
tmp, _mm_unpackhi_epi32(kernel11, kernel11)));
tmp = _mm_unpacklo_epi8(pix20, _mm_setzero_si128());
ss0 = _mm_add_epi32(ss0, _mm_madd_epi16(
tmp, _mm_unpacklo_epi32(kernel20, kernel20)));
ss1 = _mm_add_epi32(ss1, _mm_madd_epi16(
tmp, _mm_unpacklo_epi32(kernel21, kernel21)));
tmp = _mm_unpackhi_epi8(pix20, _mm_setzero_si128());
ss0 = _mm_add_epi32(ss0, _mm_madd_epi16(
tmp, _mm_unpackhi_epi32(kernel20, kernel20)));
ss1 = _mm_add_epi32(ss1, _mm_madd_epi16(
tmp, _mm_unpackhi_epi32(kernel21, kernel21)));
MM_KERNEL_SUM(ss0, 0, 0, _mm_unpacklo_epi8, _mm_unpacklo_epi32);
MM_KERNEL_SUM(ss0, 0, 0, _mm_unpackhi_epi8, _mm_unpackhi_epi32);
MM_KERNEL_SUM(ss1, 0, 1, _mm_unpacklo_epi8, _mm_unpacklo_epi32);
MM_KERNEL_SUM(ss1, 0, 1, _mm_unpackhi_epi8, _mm_unpackhi_epi32);
ss0 = _mm_packs_epi32( ss0 = _mm_packs_epi32(
_mm_srai_epi32(ss0, PRECISION_BITS), _mm_srai_epi32(ss0, PRECISION_BITS),
_mm_srai_epi32(ss1, PRECISION_BITS)); _mm_srai_epi32(ss1, PRECISION_BITS));
MM_KERNEL_SUM(ss2, 0, 0, _mm_unpackhi_epi8, _mm_unpacklo_epi32);
MM_KERNEL_SUM(ss3, 0, 1, _mm_unpackhi_epi8, _mm_unpacklo_epi32); tmp = _mm_unpackhi_epi8(pix00, _mm_setzero_si128());
ss2 = _mm_add_epi32(ss2, _mm_madd_epi16(
tmp, _mm_unpacklo_epi32(kernel00, kernel00)));
ss3 = _mm_add_epi32(ss3, _mm_madd_epi16(
tmp, _mm_unpacklo_epi32(kernel01, kernel01)));
tmp = _mm_unpackhi_epi8(pix10, _mm_setzero_si128());
ss2 = _mm_add_epi32(ss2, _mm_madd_epi16(
tmp, _mm_unpacklo_epi32(kernel10, kernel10)));
ss3 = _mm_add_epi32(ss3, _mm_madd_epi16(
tmp, _mm_unpacklo_epi32(kernel11, kernel11)));
tmp = _mm_unpackhi_epi8(pix20, _mm_setzero_si128());
ss2 = _mm_add_epi32(ss2, _mm_madd_epi16(
tmp, _mm_unpacklo_epi32(kernel20, kernel20)));
ss3 = _mm_add_epi32(ss3, _mm_madd_epi16(
tmp, _mm_unpacklo_epi32(kernel21, kernel21)));
MM_KERNEL_LOAD(0, x+3); MM_KERNEL_LOAD(0, x+3);
MM_KERNEL_SUM(ss2, 0, 0, _mm_unpacklo_epi8, _mm_unpackhi_epi32);
MM_KERNEL_SUM(ss3, 0, 1, _mm_unpacklo_epi8, _mm_unpackhi_epi32); tmp = _mm_unpacklo_epi8(pix00, _mm_setzero_si128());
ss2 = _mm_add_epi32(ss2, _mm_madd_epi16(
tmp, _mm_unpackhi_epi32(kernel00, kernel00)));
ss3 = _mm_add_epi32(ss3, _mm_madd_epi16(
tmp, _mm_unpackhi_epi32(kernel01, kernel01)));
tmp = _mm_unpacklo_epi8(pix10, _mm_setzero_si128());
ss2 = _mm_add_epi32(ss2, _mm_madd_epi16(
tmp, _mm_unpackhi_epi32(kernel10, kernel10)));
ss3 = _mm_add_epi32(ss3, _mm_madd_epi16(
tmp, _mm_unpackhi_epi32(kernel11, kernel11)));
tmp = _mm_unpacklo_epi8(pix20, _mm_setzero_si128());
ss2 = _mm_add_epi32(ss2, _mm_madd_epi16(
tmp, _mm_unpackhi_epi32(kernel20, kernel20)));
ss3 = _mm_add_epi32(ss3, _mm_madd_epi16(
tmp, _mm_unpackhi_epi32(kernel21, kernel21)));
ss2 = _mm_packs_epi32( ss2 = _mm_packs_epi32(
_mm_srai_epi32(ss2, PRECISION_BITS), _mm_srai_epi32(ss2, PRECISION_BITS),
_mm_srai_epi32(ss3, PRECISION_BITS)); _mm_srai_epi32(ss3, PRECISION_BITS));