From: Clément Bœsch Date: Thu, 23 Mar 2017 11:03:25 +0000 (+0100) Subject: Merge commit 'e4128c08d786eb5513578e8c6063671ba03226ab' X-Git-Tag: n3.3~416 X-Git-Url: https://git.videolan.org/?p=ffmpeg.git;a=commitdiff_plain;h=733b13ad666c057c416a23bc694ae2e8391abd8b;hp=-c Merge commit 'e4128c08d786eb5513578e8c6063671ba03226ab' * commit 'e4128c08d786eb5513578e8c6063671ba03226ab': Revert "hevc: x86: Refactor IDCT macro declarations" So apparently this was technically correct be reverted due to authorship. Reverted as well in FFmpeg for now... See http://lists.libav.org/pipermail/libav-devel/2016-October/079560.html Merged-by: Clément Bœsch --- 733b13ad666c057c416a23bc694ae2e8391abd8b diff --combined libavcodec/x86/hevc_idct.asm index 26649d0d88,d662aa90ba..33b437c257 --- a/libavcodec/x86/hevc_idct.asm +++ b/libavcodec/x86/hevc_idct.asm @@@ -3,28 -3,28 +3,28 @@@ ;* Copyright (c) 2014 Pierre-Edouard LEPERE ;* Copyright (c) 2014 James Almer ;* -;* This file is part of Libav. +;* This file is part of FFmpeg. ;* -;* Libav is free software; you can redistribute it and/or +;* FFmpeg is free software; you can redistribute it and/or ;* modify it under the terms of the GNU Lesser General Public ;* License as published by the Free Software Foundation; either ;* version 2.1 of the License, or (at your option) any later version. ;* -;* Libav is distributed in the hope that it will be useful, +;* FFmpeg is distributed in the hope that it will be useful, ;* but WITHOUT ANY WARRANTY; without even the implied warranty of ;* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU ;* Lesser General Public License for more details. ;* ;* You should have received a copy of the GNU Lesser General Public -;* License along with Libav; if not, write to the Free Software +;* License along with FFmpeg; if not, write to the Free Software ;* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA ;****************************************************************************** %include "libavutil/x86/x86util.asm" -section .text +SECTION .text -; void ff_hevc_idctHxW_dc_{8,10}_(int16_t *coeffs) +; void ff_hevc_idct_HxW_dc_{8,10}_(int16_t *coeffs) ; %1 = HxW ; %2 = number of loops ; %3 = bitdepth @@@ -74,23 -74,34 +74,50 @@@ cglobal hevc_idct_%1x%1_dc_%2, 1, 2, 1 RET %endmacro - %macro INIT_IDCT_DC 1 + ; 8-bit INIT_MMX mmxext - IDCT_DC_NL 4, %1 - IDCT_DC 8, 2, %1 + IDCT_DC_NL 4, 8 + IDCT_DC 8, 2, 8 INIT_XMM sse2 - IDCT_DC_NL 8, %1 - IDCT_DC 16, 4, %1 - IDCT_DC 32, 16, %1 + IDCT_DC_NL 8, 8 + IDCT_DC 16, 4, 8 + IDCT_DC 32, 16, 8 %if HAVE_AVX2_EXTERNAL INIT_YMM avx2 - IDCT_DC 16, 2, %1 - IDCT_DC 32, 8, %1 + IDCT_DC 16, 2, 8 + IDCT_DC 32, 8, 8 %endif ;HAVE_AVX2_EXTERNAL - %endmacro - INIT_IDCT_DC 8 - INIT_IDCT_DC 10 - INIT_IDCT_DC 12 + ; 10-bit + INIT_MMX mmxext + IDCT_DC_NL 4, 10 + IDCT_DC 8, 2, 10 + + INIT_XMM sse2 + IDCT_DC_NL 8, 10 + IDCT_DC 16, 4, 10 + IDCT_DC 32, 16, 10 + + %if HAVE_AVX2_EXTERNAL + INIT_YMM avx2 + IDCT_DC 16, 2, 10 + IDCT_DC 32, 8, 10 + %endif ;HAVE_AVX2_EXTERNAL ++ ++; 12-bit ++INIT_MMX mmxext ++IDCT_DC_NL 4, 12 ++IDCT_DC 8, 2, 12 ++ ++INIT_XMM sse2 ++IDCT_DC_NL 8, 12 ++IDCT_DC 16, 4, 12 ++IDCT_DC 32, 16, 12 ++ ++%if HAVE_AVX2_EXTERNAL ++INIT_YMM avx2 ++IDCT_DC 16, 2, 12 ++IDCT_DC 32, 8, 12 ++%endif ;HAVE_AVX2_EXTERNAL