diff options
author | Dr.Smile <vabnick@gmail.com> | 2021-03-09 04:24:53 +0300 |
---|---|---|
committer | Dr.Smile <vabnick@gmail.com> | 2021-04-21 20:58:01 +0300 |
commit | 2c9b36d4800a18de7bfad1d644fd2934b9f95799 (patch) | |
tree | 6af9402febf3aefe2b1e16fb5c24dd2191e538e1 /libass/ass_func_template.h | |
parent | 904d9219395d8f471e949506feaa05fa18183f4e (diff) | |
download | libass-2c9b36d4800a18de7bfad1d644fd2934b9f95799.tar.bz2 libass-2c9b36d4800a18de7bfad1d644fd2934b9f95799.tar.xz |
Rewrite mul_bitmaps() assembly
Change list:
- No special handling of unaligned case.
- Common macro for SSE2 and AVX2 versions,
AVX2 has got significantly faster.
- Reduced register usage and efficient 32-bit version.
- Full width memory operations instead of half-register.
- Vectorized handling of width tails instead of byte loops.
Diffstat (limited to 'libass/ass_func_template.h')
-rw-r--r-- | libass/ass_func_template.h | 4 |
1 files changed, 0 insertions, 4 deletions
diff --git a/libass/ass_func_template.h b/libass/ass_func_template.h index 4737b87..b6905ad 100644 --- a/libass/ass_func_template.h +++ b/libass/ass_func_template.h @@ -106,11 +106,7 @@ const BitmapEngine DECORATE(bitmap_engine) = { .add_bitmaps = DECORATE(add_bitmaps), .sub_bitmaps = DECORATE(sub_bitmaps), -#ifdef __x86_64__ .mul_bitmaps = DECORATE(mul_bitmaps), -#else - .mul_bitmaps = ass_mul_bitmaps_c, -#endif #ifdef __x86_64__ .be_blur = DECORATE(be_blur), |