adam
|
eb56c8e56a
|
Changes 20170615:
* x86: Avoid self-relative expressions on macho64
* configure: Don't try to detect clang by $CC
* checkasm: Use the right variable in a loop condition
* x86: Fix linking with 8-bit depth shared libx264
* x86: Only enable AVX-512 in 8-bit mode
* x86: AVX-512 cabac_block_residual
* x86: AVX-512 pixel_sad_x3 and pixel_sad_x4
* x86: AVX-512 pixel_sad
* x86: AVX-512 decimate_score
* x86: AVX-512 pixel_var2_8x8 and 8x16
* Rework pixel_var2
* x86: AVX-512 pixel_var_8x8, 8x16, and 16x16
* x86: AVX-512 pixel_sa8d_8x8
* x86: AVX-512 pixel_satd
* x86: AVX-512 deblock_strength
* x86: AVX-512 plane_copy_deinterleave_v210
* x86: AVX-512 memzero_aligned
* x86: AVX and AVX-512 memcpy_aligned
* x86: AVX-512 dequant_8x8_flat16
* x86: AVX-512 dequant_8x8
* x86: AVX-512 dequant_4x4
* x86: AVX-512 mbtree_propagate_cost
* x86: AVX-512 coeff_last
* x86: AVX-512 zigzag_interleave_8x8_cavlc
* x86: AVX-512 zigzag_scan_8x8_field
* x86: AVX-512 zigzag_scan_4x4_field
* x86: AVX-512 zigzag_scan_8x8_frame
* x86: AVX-512 zigzag_scan_4x4_frame
* checkasm: x86: More accurate ymm/zmm measurements
* x86: AVX-512 support
* x86: Change assembler from yasm to nasm
* x86: Add some additional cpuflag relations
* x86: Faster SSE2 pixel_sad_16x16 and 16x8
* msvs/icl: Improve target host detection
* ppc: Optimize add8x8_idct_dc
|
2017-06-16 05:39:23 +00:00 |
|