From 3a39195b1db5620901a049cd690752b1935f5e0f Mon Sep 17 00:00:00 2001 From: "Ronald S. Bultje" Date: Sat, 23 Jul 2011 18:36:35 -0700 Subject: Move x86inc.asm to libavutil/. This allows using it in libswscale/ also. --- libavcodec/x86/h264_idct.asm | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'libavcodec/x86/h264_idct.asm') diff --git a/libavcodec/x86/h264_idct.asm b/libavcodec/x86/h264_idct.asm index 37c2c90476..482881b332 100644 --- a/libavcodec/x86/h264_idct.asm +++ b/libavcodec/x86/h264_idct.asm @@ -26,7 +26,7 @@ ;* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA ;***************************************************************************** -%include "x86inc.asm" +%include "libavutil/x86/x86inc.asm" %include "x86util.asm" SECTION_RODATA -- cgit v1.2.3 From b2c087871dafc7d030b2d48457ddff597dfd4925 Mon Sep 17 00:00:00 2001 From: "Ronald S. Bultje" Date: Wed, 3 Aug 2011 09:48:08 -0700 Subject: Move x86util.asm from libavcodec/ to libavutil/. This allows using it in swscale also. --- libavcodec/x86/ac3dsp.asm | 2 +- libavcodec/x86/dct32_sse.asm | 2 +- libavcodec/x86/deinterlace.asm | 2 +- libavcodec/x86/dsputilenc_yasm.asm | 2 +- libavcodec/x86/fmtconvert.asm | 2 +- libavcodec/x86/h264_chromamc.asm | 2 +- libavcodec/x86/h264_chromamc_10bit.asm | 2 +- libavcodec/x86/h264_deblock.asm | 2 +- libavcodec/x86/h264_deblock_10bit.asm | 2 +- libavcodec/x86/h264_idct.asm | 2 +- libavcodec/x86/h264_idct_10bit.asm | 2 +- libavcodec/x86/h264_intrapred.asm | 2 +- libavcodec/x86/h264_intrapred_10bit.asm | 2 +- libavcodec/x86/h264_qpel_10bit.asm | 2 +- libavcodec/x86/h264_weight_10bit.asm | 2 +- libavcodec/x86/vc1dsp_yasm.asm | 2 +- libavcodec/x86/vp3dsp.asm | 2 +- libavcodec/x86/vp56dsp.asm | 2 +- libavcodec/x86/vp8dsp.asm | 2 +- libavcodec/x86/x86util.asm | 542 -------------------------------- libavutil/x86/x86util.asm | 542 ++++++++++++++++++++++++++++++++ 21 files changed, 561 insertions(+), 561 deletions(-) delete mode 100644 libavcodec/x86/x86util.asm create mode 100644 libavutil/x86/x86util.asm (limited to 'libavcodec/x86/h264_idct.asm') diff --git a/libavcodec/x86/ac3dsp.asm b/libavcodec/x86/ac3dsp.asm index 44124c5397..9e92678afe 100644 --- a/libavcodec/x86/ac3dsp.asm +++ b/libavcodec/x86/ac3dsp.asm @@ -20,7 +20,7 @@ ;****************************************************************************** %include "libavutil/x86/x86inc.asm" -%include "x86util.asm" +%include "libavutil/x86/x86util.asm" SECTION_RODATA diff --git a/libavcodec/x86/dct32_sse.asm b/libavcodec/x86/dct32_sse.asm index 4e938ad198..f3eaf3a956 100644 --- a/libavcodec/x86/dct32_sse.asm +++ b/libavcodec/x86/dct32_sse.asm @@ -20,7 +20,7 @@ ;****************************************************************************** %include "libavutil/x86/x86inc.asm" -%include "x86util.asm" +%include "libavutil/x86/x86util.asm" SECTION_RODATA 32 diff --git a/libavcodec/x86/deinterlace.asm b/libavcodec/x86/deinterlace.asm index a782b2aae1..9150f4578e 100644 --- a/libavcodec/x86/deinterlace.asm +++ b/libavcodec/x86/deinterlace.asm @@ -21,7 +21,7 @@ ;****************************************************************************** %include "libavutil/x86/x86inc.asm" -%include "x86util.asm" +%include "libavutil/x86/x86util.asm" SECTION_RODATA diff --git a/libavcodec/x86/dsputilenc_yasm.asm b/libavcodec/x86/dsputilenc_yasm.asm index 09450226e0..a0531b0b7e 100644 --- a/libavcodec/x86/dsputilenc_yasm.asm +++ b/libavcodec/x86/dsputilenc_yasm.asm @@ -22,7 +22,7 @@ ;***************************************************************************** %include "libavutil/x86/x86inc.asm" -%include "x86util.asm" +%include "libavutil/x86/x86util.asm" SECTION .text diff --git a/libavcodec/x86/fmtconvert.asm b/libavcodec/x86/fmtconvert.asm index a69b3ff357..e384e8f0b2 100644 --- a/libavcodec/x86/fmtconvert.asm +++ b/libavcodec/x86/fmtconvert.asm @@ -20,7 +20,7 @@ ;****************************************************************************** %include "libavutil/x86/x86inc.asm" -%include "x86util.asm" +%include "libavutil/x86/x86util.asm" SECTION_TEXT diff --git a/libavcodec/x86/h264_chromamc.asm b/libavcodec/x86/h264_chromamc.asm index 201dcd85a6..caef7dd7be 100644 --- a/libavcodec/x86/h264_chromamc.asm +++ b/libavcodec/x86/h264_chromamc.asm @@ -21,7 +21,7 @@ ;****************************************************************************** %include "libavutil/x86/x86inc.asm" -%include "x86util.asm" +%include "libavutil/x86/x86util.asm" SECTION_RODATA diff --git a/libavcodec/x86/h264_chromamc_10bit.asm b/libavcodec/x86/h264_chromamc_10bit.asm index af574844dc..56c0511857 100644 --- a/libavcodec/x86/h264_chromamc_10bit.asm +++ b/libavcodec/x86/h264_chromamc_10bit.asm @@ -23,7 +23,7 @@ ;****************************************************************************** %include "libavutil/x86/x86inc.asm" -%include "x86util.asm" +%include "libavutil/x86/x86util.asm" SECTION_RODATA diff --git a/libavcodec/x86/h264_deblock.asm b/libavcodec/x86/h264_deblock.asm index ac25dca92d..1304e40261 100644 --- a/libavcodec/x86/h264_deblock.asm +++ b/libavcodec/x86/h264_deblock.asm @@ -25,7 +25,7 @@ ;****************************************************************************** %include "libavutil/x86/x86inc.asm" -%include "x86util.asm" +%include "libavutil/x86/x86util.asm" SECTION .text diff --git a/libavcodec/x86/h264_deblock_10bit.asm b/libavcodec/x86/h264_deblock_10bit.asm index 0f525449e3..0a1809c188 100644 --- a/libavcodec/x86/h264_deblock_10bit.asm +++ b/libavcodec/x86/h264_deblock_10bit.asm @@ -25,7 +25,7 @@ ;****************************************************************************** %include "libavutil/x86/x86inc.asm" -%include "x86util.asm" +%include "libavutil/x86/x86util.asm" SECTION_RODATA diff --git a/libavcodec/x86/h264_idct.asm b/libavcodec/x86/h264_idct.asm index 482881b332..42c175a0d8 100644 --- a/libavcodec/x86/h264_idct.asm +++ b/libavcodec/x86/h264_idct.asm @@ -27,7 +27,7 @@ ;***************************************************************************** %include "libavutil/x86/x86inc.asm" -%include "x86util.asm" +%include "libavutil/x86/x86util.asm" SECTION_RODATA diff --git a/libavcodec/x86/h264_idct_10bit.asm b/libavcodec/x86/h264_idct_10bit.asm index 6a45724f3f..a9c12da5a2 100644 --- a/libavcodec/x86/h264_idct_10bit.asm +++ b/libavcodec/x86/h264_idct_10bit.asm @@ -23,7 +23,7 @@ ;****************************************************************************** %include "libavutil/x86/x86inc.asm" -%include "x86util.asm" +%include "libavutil/x86/x86util.asm" SECTION_RODATA diff --git a/libavcodec/x86/h264_intrapred.asm b/libavcodec/x86/h264_intrapred.asm index 258eef563f..71b5aa130c 100644 --- a/libavcodec/x86/h264_intrapred.asm +++ b/libavcodec/x86/h264_intrapred.asm @@ -23,7 +23,7 @@ ;****************************************************************************** %include "libavutil/x86/x86inc.asm" -%include "x86util.asm" +%include "libavutil/x86/x86util.asm" SECTION_RODATA diff --git a/libavcodec/x86/h264_intrapred_10bit.asm b/libavcodec/x86/h264_intrapred_10bit.asm index 7d8ce08690..5a06896afe 100644 --- a/libavcodec/x86/h264_intrapred_10bit.asm +++ b/libavcodec/x86/h264_intrapred_10bit.asm @@ -23,7 +23,7 @@ ;****************************************************************************** %include "libavutil/x86/x86inc.asm" -%include "x86util.asm" +%include "libavutil/x86/x86util.asm" SECTION_RODATA diff --git a/libavcodec/x86/h264_qpel_10bit.asm b/libavcodec/x86/h264_qpel_10bit.asm index 3df3048100..30d0913cd1 100644 --- a/libavcodec/x86/h264_qpel_10bit.asm +++ b/libavcodec/x86/h264_qpel_10bit.asm @@ -23,7 +23,7 @@ ;****************************************************************************** %include "libavutil/x86/x86inc.asm" -%include "x86util.asm" +%include "libavutil/x86/x86util.asm" SECTION_RODATA 32 diff --git a/libavcodec/x86/h264_weight_10bit.asm b/libavcodec/x86/h264_weight_10bit.asm index abf382389b..affc4ce5c4 100644 --- a/libavcodec/x86/h264_weight_10bit.asm +++ b/libavcodec/x86/h264_weight_10bit.asm @@ -23,7 +23,7 @@ ;****************************************************************************** %include "libavutil/x86/x86inc.asm" -%include "x86util.asm" +%include "libavutil/x86/x86util.asm" SECTION_RODATA 32 diff --git a/libavcodec/x86/vc1dsp_yasm.asm b/libavcodec/x86/vc1dsp_yasm.asm index 0bf7185a61..bc53239e9d 100644 --- a/libavcodec/x86/vc1dsp_yasm.asm +++ b/libavcodec/x86/vc1dsp_yasm.asm @@ -20,7 +20,7 @@ ;****************************************************************************** %include "libavutil/x86/x86inc.asm" -%include "x86util.asm" +%include "libavutil/x86/x86util.asm" cextern pw_4 cextern pw_5 diff --git a/libavcodec/x86/vp3dsp.asm b/libavcodec/x86/vp3dsp.asm index 23616a1863..096879bc47 100644 --- a/libavcodec/x86/vp3dsp.asm +++ b/libavcodec/x86/vp3dsp.asm @@ -20,7 +20,7 @@ ;****************************************************************************** %include "libavutil/x86/x86inc.asm" -%include "x86util.asm" +%include "libavutil/x86/x86util.asm" ; MMX-optimized functions cribbed from the original VP3 source code. diff --git a/libavcodec/x86/vp56dsp.asm b/libavcodec/x86/vp56dsp.asm index fc68cabef0..5c77bea55a 100644 --- a/libavcodec/x86/vp56dsp.asm +++ b/libavcodec/x86/vp56dsp.asm @@ -21,7 +21,7 @@ ;****************************************************************************** %include "libavutil/x86/x86inc.asm" -%include "x86util.asm" +%include "libavutil/x86/x86util.asm" cextern pw_64 diff --git a/libavcodec/x86/vp8dsp.asm b/libavcodec/x86/vp8dsp.asm index ae2055ce96..1434be629e 100644 --- a/libavcodec/x86/vp8dsp.asm +++ b/libavcodec/x86/vp8dsp.asm @@ -21,7 +21,7 @@ ;****************************************************************************** %include "libavutil/x86/x86inc.asm" -%include "x86util.asm" +%include "libavutil/x86/x86util.asm" SECTION_RODATA diff --git a/libavcodec/x86/x86util.asm b/libavcodec/x86/x86util.asm deleted file mode 100644 index 5f39107379..0000000000 --- a/libavcodec/x86/x86util.asm +++ /dev/null @@ -1,542 +0,0 @@ -;***************************************************************************** -;* x86util.asm -;***************************************************************************** -;* Copyright (C) 2008-2010 x264 project -;* -;* Authors: Loren Merritt -;* Holger Lubitz -;* -;* This file is part of Libav. -;* -;* Libav 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, -;* 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 -;* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA -;****************************************************************************** - -%macro SBUTTERFLY 4 -%if avx_enabled == 0 - mova m%4, m%2 - punpckl%1 m%2, m%3 - punpckh%1 m%4, m%3 -%else - punpckh%1 m%4, m%2, m%3 - punpckl%1 m%2, m%3 -%endif - SWAP %3, %4 -%endmacro - -%macro SBUTTERFLY2 4 - punpckl%1 m%4, m%2, m%3 - punpckh%1 m%2, m%2, m%3 - SWAP %2, %4, %3 -%endmacro - -%macro SBUTTERFLYPS 3 - movaps m%3, m%1 - unpcklps m%1, m%2 - unpckhps m%3, m%2 - SWAP %2, %3 -%endmacro - -%macro TRANSPOSE4x4B 5 - SBUTTERFLY bw, %1, %2, %5 - SBUTTERFLY bw, %3, %4, %5 - SBUTTERFLY wd, %1, %3, %5 - SBUTTERFLY wd, %2, %4, %5 - SWAP %2, %3 -%endmacro - -%macro TRANSPOSE4x4W 5 - SBUTTERFLY wd, %1, %2, %5 - SBUTTERFLY wd, %3, %4, %5 - SBUTTERFLY dq, %1, %3, %5 - SBUTTERFLY dq, %2, %4, %5 - SWAP %2, %3 -%endmacro - -%macro TRANSPOSE2x4x4W 5 - SBUTTERFLY wd, %1, %2, %5 - SBUTTERFLY wd, %3, %4, %5 - SBUTTERFLY dq, %1, %3, %5 - SBUTTERFLY dq, %2, %4, %5 - SBUTTERFLY qdq, %1, %2, %5 - SBUTTERFLY qdq, %3, %4, %5 -%endmacro - -%macro TRANSPOSE4x4D 5 - SBUTTERFLY dq, %1, %2, %5 - SBUTTERFLY dq, %3, %4, %5 - SBUTTERFLY qdq, %1, %3, %5 - SBUTTERFLY qdq, %2, %4, %5 - SWAP %2, %3 -%endmacro - -; identical behavior to TRANSPOSE4x4D, but using SSE1 float ops -%macro TRANSPOSE4x4PS 5 - SBUTTERFLYPS %1, %2, %5 - SBUTTERFLYPS %3, %4, %5 - movaps m%5, m%1 - movlhps m%1, m%3 - movhlps m%3, m%5 - movaps m%5, m%2 - movlhps m%2, m%4 - movhlps m%4, m%5 - SWAP %2, %3 -%endmacro - -%macro TRANSPOSE8x8W 9-11 -%ifdef ARCH_X86_64 - SBUTTERFLY wd, %1, %2, %9 - SBUTTERFLY wd, %3, %4, %9 - SBUTTERFLY wd, %5, %6, %9 - SBUTTERFLY wd, %7, %8, %9 - SBUTTERFLY dq, %1, %3, %9 - SBUTTERFLY dq, %2, %4, %9 - SBUTTERFLY dq, %5, %7, %9 - SBUTTERFLY dq, %6, %8, %9 - SBUTTERFLY qdq, %1, %5, %9 - SBUTTERFLY qdq, %2, %6, %9 - SBUTTERFLY qdq, %3, %7, %9 - SBUTTERFLY qdq, %4, %8, %9 - SWAP %2, %5 - SWAP %4, %7 -%else -; in: m0..m7, unless %11 in which case m6 is in %9 -; out: m0..m7, unless %11 in which case m4 is in %10 -; spills into %9 and %10 -%if %0<11 - movdqa %9, m%7 -%endif - SBUTTERFLY wd, %1, %2, %7 - movdqa %10, m%2 - movdqa m%7, %9 - SBUTTERFLY wd, %3, %4, %2 - SBUTTERFLY wd, %5, %6, %2 - SBUTTERFLY wd, %7, %8, %2 - SBUTTERFLY dq, %1, %3, %2 - movdqa %9, m%3 - movdqa m%2, %10 - SBUTTERFLY dq, %2, %4, %3 - SBUTTERFLY dq, %5, %7, %3 - SBUTTERFLY dq, %6, %8, %3 - SBUTTERFLY qdq, %1, %5, %3 - SBUTTERFLY qdq, %2, %6, %3 - movdqa %10, m%2 - movdqa m%3, %9 - SBUTTERFLY qdq, %3, %7, %2 - SBUTTERFLY qdq, %4, %8, %2 - SWAP %2, %5 - SWAP %4, %7 -%if %0<11 - movdqa m%5, %10 -%endif -%endif -%endmacro - -; PABSW macros assume %1 != %2, while ABS1/2 macros work in-place -%macro PABSW_MMX 2 - pxor %1, %1 - pcmpgtw %1, %2 - pxor %2, %1 - psubw %2, %1 - SWAP %1, %2 -%endmacro - -%macro PSIGNW_MMX 2 - pxor %1, %2 - psubw %1, %2 -%endmacro - -%macro PABSW_MMX2 2 - pxor %1, %1 - psubw %1, %2 - pmaxsw %1, %2 -%endmacro - -%macro PABSW_SSSE3 2 - pabsw %1, %2 -%endmacro - -%macro PSIGNW_SSSE3 2 - psignw %1, %2 -%endmacro - -%macro ABS1_MMX 2 ; a, tmp - pxor %2, %2 - pcmpgtw %2, %1 - pxor %1, %2 - psubw %1, %2 -%endmacro - -%macro ABS2_MMX 4 ; a, b, tmp0, tmp1 - pxor %3, %3 - pxor %4, %4 - pcmpgtw %3, %1 - pcmpgtw %4, %2 - pxor %1, %3 - pxor %2, %4 - psubw %1, %3 - psubw %2, %4 -%endmacro - -%macro ABS1_MMX2 2 ; a, tmp - pxor %2, %2 - psubw %2, %1 - pmaxsw %1, %2 -%endmacro - -%macro ABS2_MMX2 4 ; a, b, tmp0, tmp1 - pxor %3, %3 - pxor %4, %4 - psubw %3, %1 - psubw %4, %2 - pmaxsw %1, %3 - pmaxsw %2, %4 -%endmacro - -%macro ABS1_SSSE3 2 - pabsw %1, %1 -%endmacro - -%macro ABS2_SSSE3 4 - pabsw %1, %1 - pabsw %2, %2 -%endmacro - -%macro ABSB_MMX 2 - pxor %2, %2 - psubb %2, %1 - pminub %1, %2 -%endmacro - -%macro ABSB2_MMX 4 - pxor %3, %3 - pxor %4, %4 - psubb %3, %1 - psubb %4, %2 - pminub %1, %3 - pminub %2, %4 -%endmacro - -%macro ABSD2_MMX 4 - pxor %3, %3 - pxor %4, %4 - pcmpgtd %3, %1 - pcmpgtd %4, %2 - pxor %1, %3 - pxor %2, %4 - psubd %1, %3 - psubd %2, %4 -%endmacro - -%macro ABSB_SSSE3 2 - pabsb %1, %1 -%endmacro - -%macro ABSB2_SSSE3 4 - pabsb %1, %1 - pabsb %2, %2 -%endmacro - -%macro ABS4 6 - ABS2 %1, %2, %5, %6 - ABS2 %3, %4, %5, %6 -%endmacro - -%define ABS1 ABS1_MMX -%define ABS2 ABS2_MMX -%define ABSB ABSB_MMX -%define ABSB2 ABSB2_MMX - -%macro SPLATB_MMX 3 - movd %1, [%2-3] ;to avoid crossing a cacheline - punpcklbw %1, %1 - SPLATW %1, %1, 3 -%endmacro - -%macro SPLATB_SSSE3 3 - movd %1, [%2-3] - pshufb %1, %3 -%endmacro - -%macro PALIGNR_MMX 4-5 ; [dst,] src1, src2, imm, tmp - %define %%dst %1 -%if %0==5 -%ifnidn %1, %2 - mova %%dst, %2 -%endif - %rotate 1 -%endif -%ifnidn %4, %2 - mova %4, %2 -%endif -%if mmsize==8 - psllq %%dst, (8-%3)*8 - psrlq %4, %3*8 -%else - pslldq %%dst, 16-%3 - psrldq %4, %3 -%endif - por %%dst, %4 -%endmacro - -%macro PALIGNR_SSSE3 4-5 -%if %0==5 - palignr %1, %2, %3, %4 -%else - palignr %1, %2, %3 -%endif -%endmacro - -%macro DEINTB 5 ; mask, reg1, mask, reg2, optional src to fill masks from -%ifnum %5 - pand m%3, m%5, m%4 ; src .. y6 .. y4 - pand m%1, m%5, m%2 ; dst .. y6 .. y4 -%else - mova m%1, %5 - pand m%3, m%1, m%4 ; src .. y6 .. y4 - pand m%1, m%1, m%2 ; dst .. y6 .. y4 -%endif - psrlw m%2, 8 ; dst .. y7 .. y5 - psrlw m%4, 8 ; src .. y7 .. y5 -%endmacro - -%macro SUMSUB_BA 3-4 -%if %0==3 - padd%1 m%2, m%3 - padd%1 m%3, m%3 - psub%1 m%3, m%2 -%else -%if avx_enabled == 0 - mova m%4, m%2 - padd%1 m%2, m%3 - psub%1 m%3, m%4 -%else - padd%1 m%4, m%2, m%3 - psub%1 m%3, m%2 - SWAP %2, %4 -%endif -%endif -%endmacro - -%macro SUMSUB_BADC 5-6 -%if %0==6 - SUMSUB_BA %1, %2, %3, %6 - SUMSUB_BA %1, %4, %5, %6 -%else - padd%1 m%2, m%3 - padd%1 m%4, m%5 - padd%1 m%3, m%3 - padd%1 m%5, m%5 - psub%1 m%3, m%2 - psub%1 m%5, m%4 -%endif -%endmacro - -%macro SUMSUB2_AB 4 -%ifnum %3 - psub%1 m%4, m%2, m%3 - psub%1 m%4, m%3 - padd%1 m%2, m%2 - padd%1 m%2, m%3 -%else - mova m%4, m%2 - padd%1 m%2, m%2 - padd%1 m%2, %3 - psub%1 m%4, %3 - psub%1 m%4, %3 -%endif -%endmacro - -%macro SUMSUB2_BA 4 -%if avx_enabled == 0 - mova m%4, m%2 - padd%1 m%2, m%3 - padd%1 m%2, m%3 - psub%1 m%3, m%4 - psub%1 m%3, m%4 -%else - padd%1 m%4, m%2, m%3 - padd%1 m%4, m%3 - psub%1 m%3, m%2 - psub%1 m%3, m%2 - SWAP %2, %4 -%endif -%endmacro - -%macro SUMSUBD2_AB 5 -%ifnum %4 - psra%1 m%5, m%2, 1 ; %3: %3>>1 - psra%1 m%4, m%3, 1 ; %2: %2>>1 - padd%1 m%4, m%2 ; %3: %3>>1+%2 - psub%1 m%5, m%3 ; %2: %2>>1-%3 - SWAP %2, %5 - SWAP %3, %4 -%else - mova %5, m%2 - mova %4, m%3 - psra%1 m%3, 1 ; %3: %3>>1 - psra%1 m%2, 1 ; %2: %2>>1 - padd%1 m%3, %5 ; %3: %3>>1+%2 - psub%1 m%2, %4 ; %2: %2>>1-%3 -%endif -%endmacro - -%macro DCT4_1D 5 -%ifnum %5 - SUMSUB_BADC w, %4, %1, %3, %2, %5 - SUMSUB_BA w, %3, %4, %5 - SUMSUB2_AB w, %1, %2, %5 - SWAP %1, %3, %4, %5, %2 -%else - SUMSUB_BADC w, %4, %1, %3, %2 - SUMSUB_BA w, %3, %4 - mova [%5], m%2 - SUMSUB2_AB w, %1, [%5], %2 - SWAP %1, %3, %4, %2 -%endif -%endmacro - -%macro IDCT4_1D 6-7 -%ifnum %6 - SUMSUBD2_AB %1, %3, %5, %7, %6 - ; %3: %3>>1-%5 %5: %3+%5>>1 - SUMSUB_BA %1, %4, %2, %7 - ; %4: %2+%4 %2: %2-%4 - SUMSUB_BADC %1, %5, %4, %3, %2, %7 - ; %5: %2+%4 + (%3+%5>>1) - ; %4: %2+%4 - (%3+%5>>1) - ; %3: %2-%4 + (%3>>1-%5) - ; %2: %2-%4 - (%3>>1-%5) -%else -%ifidn %1, w - SUMSUBD2_AB %1, %3, %5, [%6], [%6+16] -%else - SUMSUBD2_AB %1, %3, %5, [%6], [%6+32] -%endif - SUMSUB_BA %1, %4, %2 - SUMSUB_BADC %1, %5, %4, %3, %2 -%endif - SWAP %2, %5, %4 - ; %2: %2+%4 + (%3+%5>>1) row0 - ; %3: %2-%4 + (%3>>1-%5) row1 - ; %4: %2-%4 - (%3>>1-%5) row2 - ; %5: %2+%4 - (%3+%5>>1) row3 -%endmacro - - -%macro LOAD_DIFF 5 -%ifidn %3, none - movh %1, %4 - movh %2, %5 - punpcklbw %1, %2 - punpcklbw %2, %2 - psubw %1, %2 -%else - movh %1, %4 - punpcklbw %1, %3 - movh %2, %5 - punpcklbw %2, %3 - psubw %1, %2 -%endif -%endmacro - -%macro STORE_DCT 6 - movq [%5+%6+ 0], m%1 - movq [%5+%6+ 8], m%2 - movq [%5+%6+16], m%3 - movq [%5+%6+24], m%4 - movhps [%5+%6+32], m%1 - movhps [%5+%6+40], m%2 - movhps [%5+%6+48], m%3 - movhps [%5+%6+56], m%4 -%endmacro - -%macro LOAD_DIFF_8x4P 7-10 r0,r2,0 ; 4x dest, 2x temp, 2x pointer, increment? - LOAD_DIFF m%1, m%5, m%7, [%8], [%9] - LOAD_DIFF m%2, m%6, m%7, [%8+r1], [%9+r3] - LOAD_DIFF m%3, m%5, m%7, [%8+2*r1], [%9+2*r3] - LOAD_DIFF m%4, m%6, m%7, [%8+r4], [%9+r5] -%if %10 - lea %8, [%8+4*r1] - lea %9, [%9+4*r3] -%endif -%endmacro - -%macro DIFFx2 6-7 - movh %3, %5 - punpcklbw %3, %4 - psraw %1, 6 - paddsw %1, %3 - movh %3, %6 - punpcklbw %3, %4 - psraw %2, 6 - paddsw %2, %3 - packuswb %2, %1 -%endmacro - -%macro STORE_DIFF 4 - movh %2, %4 - punpcklbw %2, %3 - psraw %1, 6 - paddsw %1, %2 - packuswb %1, %1 - movh %4, %1 -%endmacro - -%macro STORE_DIFFx2 8 ; add1, add2, reg1, reg2, zero, shift, source, stride - movh %3, [%7] - movh %4, [%7+%8] - psraw %1, %6 - psraw %2, %6 - punpcklbw %3, %5 - punpcklbw %4, %5 - paddw %3, %1 - paddw %4, %2 - packuswb %3, %5 - packuswb %4, %5 - movh [%7], %3 - movh [%7+%8], %4 -%endmacro - -%macro PMINUB_MMX 3 ; dst, src, tmp - mova %3, %1 - psubusb %3, %2 - psubb %1, %3 -%endmacro - -%macro PMINUB_MMXEXT 3 ; dst, src, ignored - pminub %1, %2 -%endmacro - -%macro SPLATW 2-3 0 -%if mmsize == 16 - pshuflw %1, %2, (%3)*0x55 - punpcklqdq %1, %1 -%else - pshufw %1, %2, (%3)*0x55 -%endif -%endmacro - -%macro SPLATD 2-3 0 -%if mmsize == 16 - pshufd %1, %2, (%3)*0x55 -%else - pshufw %1, %2, (%3)*0x11 + ((%3)+1)*0x44 -%endif -%endmacro - -%macro CLIPW 3 ;(dst, min, max) - pmaxsw %1, %2 - pminsw %1, %3 -%endmacro diff --git a/libavutil/x86/x86util.asm b/libavutil/x86/x86util.asm new file mode 100644 index 0000000000..5f39107379 --- /dev/null +++ b/libavutil/x86/x86util.asm @@ -0,0 +1,542 @@ +;***************************************************************************** +;* x86util.asm +;***************************************************************************** +;* Copyright (C) 2008-2010 x264 project +;* +;* Authors: Loren Merritt +;* Holger Lubitz +;* +;* This file is part of Libav. +;* +;* Libav 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, +;* 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 +;* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA +;****************************************************************************** + +%macro SBUTTERFLY 4 +%if avx_enabled == 0 + mova m%4, m%2 + punpckl%1 m%2, m%3 + punpckh%1 m%4, m%3 +%else + punpckh%1 m%4, m%2, m%3 + punpckl%1 m%2, m%3 +%endif + SWAP %3, %4 +%endmacro + +%macro SBUTTERFLY2 4 + punpckl%1 m%4, m%2, m%3 + punpckh%1 m%2, m%2, m%3 + SWAP %2, %4, %3 +%endmacro + +%macro SBUTTERFLYPS 3 + movaps m%3, m%1 + unpcklps m%1, m%2 + unpckhps m%3, m%2 + SWAP %2, %3 +%endmacro + +%macro TRANSPOSE4x4B 5 + SBUTTERFLY bw, %1, %2, %5 + SBUTTERFLY bw, %3, %4, %5 + SBUTTERFLY wd, %1, %3, %5 + SBUTTERFLY wd, %2, %4, %5 + SWAP %2, %3 +%endmacro + +%macro TRANSPOSE4x4W 5 + SBUTTERFLY wd, %1, %2, %5 + SBUTTERFLY wd, %3, %4, %5 + SBUTTERFLY dq, %1, %3, %5 + SBUTTERFLY dq, %2, %4, %5 + SWAP %2, %3 +%endmacro + +%macro TRANSPOSE2x4x4W 5 + SBUTTERFLY wd, %1, %2, %5 + SBUTTERFLY wd, %3, %4, %5 + SBUTTERFLY dq, %1, %3, %5 + SBUTTERFLY dq, %2, %4, %5 + SBUTTERFLY qdq, %1, %2, %5 + SBUTTERFLY qdq, %3, %4, %5 +%endmacro + +%macro TRANSPOSE4x4D 5 + SBUTTERFLY dq, %1, %2, %5 + SBUTTERFLY dq, %3, %4, %5 + SBUTTERFLY qdq, %1, %3, %5 + SBUTTERFLY qdq, %2, %4, %5 + SWAP %2, %3 +%endmacro + +; identical behavior to TRANSPOSE4x4D, but using SSE1 float ops +%macro TRANSPOSE4x4PS 5 + SBUTTERFLYPS %1, %2, %5 + SBUTTERFLYPS %3, %4, %5 + movaps m%5, m%1 + movlhps m%1, m%3 + movhlps m%3, m%5 + movaps m%5, m%2 + movlhps m%2, m%4 + movhlps m%4, m%5 + SWAP %2, %3 +%endmacro + +%macro TRANSPOSE8x8W 9-11 +%ifdef ARCH_X86_64 + SBUTTERFLY wd, %1, %2, %9 + SBUTTERFLY wd, %3, %4, %9 + SBUTTERFLY wd, %5, %6, %9 + SBUTTERFLY wd, %7, %8, %9 + SBUTTERFLY dq, %1, %3, %9 + SBUTTERFLY dq, %2, %4, %9 + SBUTTERFLY dq, %5, %7, %9 + SBUTTERFLY dq, %6, %8, %9 + SBUTTERFLY qdq, %1, %5, %9 + SBUTTERFLY qdq, %2, %6, %9 + SBUTTERFLY qdq, %3, %7, %9 + SBUTTERFLY qdq, %4, %8, %9 + SWAP %2, %5 + SWAP %4, %7 +%else +; in: m0..m7, unless %11 in which case m6 is in %9 +; out: m0..m7, unless %11 in which case m4 is in %10 +; spills into %9 and %10 +%if %0<11 + movdqa %9, m%7 +%endif + SBUTTERFLY wd, %1, %2, %7 + movdqa %10, m%2 + movdqa m%7, %9 + SBUTTERFLY wd, %3, %4, %2 + SBUTTERFLY wd, %5, %6, %2 + SBUTTERFLY wd, %7, %8, %2 + SBUTTERFLY dq, %1, %3, %2 + movdqa %9, m%3 + movdqa m%2, %10 + SBUTTERFLY dq, %2, %4, %3 + SBUTTERFLY dq, %5, %7, %3 + SBUTTERFLY dq, %6, %8, %3 + SBUTTERFLY qdq, %1, %5, %3 + SBUTTERFLY qdq, %2, %6, %3 + movdqa %10, m%2 + movdqa m%3, %9 + SBUTTERFLY qdq, %3, %7, %2 + SBUTTERFLY qdq, %4, %8, %2 + SWAP %2, %5 + SWAP %4, %7 +%if %0<11 + movdqa m%5, %10 +%endif +%endif +%endmacro + +; PABSW macros assume %1 != %2, while ABS1/2 macros work in-place +%macro PABSW_MMX 2 + pxor %1, %1 + pcmpgtw %1, %2 + pxor %2, %1 + psubw %2, %1 + SWAP %1, %2 +%endmacro + +%macro PSIGNW_MMX 2 + pxor %1, %2 + psubw %1, %2 +%endmacro + +%macro PABSW_MMX2 2 + pxor %1, %1 + psubw %1, %2 + pmaxsw %1, %2 +%endmacro + +%macro PABSW_SSSE3 2 + pabsw %1, %2 +%endmacro + +%macro PSIGNW_SSSE3 2 + psignw %1, %2 +%endmacro + +%macro ABS1_MMX 2 ; a, tmp + pxor %2, %2 + pcmpgtw %2, %1 + pxor %1, %2 + psubw %1, %2 +%endmacro + +%macro ABS2_MMX 4 ; a, b, tmp0, tmp1 + pxor %3, %3 + pxor %4, %4 + pcmpgtw %3, %1 + pcmpgtw %4, %2 + pxor %1, %3 + pxor %2, %4 + psubw %1, %3 + psubw %2, %4 +%endmacro + +%macro ABS1_MMX2 2 ; a, tmp + pxor %2, %2 + psubw %2, %1 + pmaxsw %1, %2 +%endmacro + +%macro ABS2_MMX2 4 ; a, b, tmp0, tmp1 + pxor %3, %3 + pxor %4, %4 + psubw %3, %1 + psubw %4, %2 + pmaxsw %1, %3 + pmaxsw %2, %4 +%endmacro + +%macro ABS1_SSSE3 2 + pabsw %1, %1 +%endmacro + +%macro ABS2_SSSE3 4 + pabsw %1, %1 + pabsw %2, %2 +%endmacro + +%macro ABSB_MMX 2 + pxor %2, %2 + psubb %2, %1 + pminub %1, %2 +%endmacro + +%macro ABSB2_MMX 4 + pxor %3, %3 + pxor %4, %4 + psubb %3, %1 + psubb %4, %2 + pminub %1, %3 + pminub %2, %4 +%endmacro + +%macro ABSD2_MMX 4 + pxor %3, %3 + pxor %4, %4 + pcmpgtd %3, %1 + pcmpgtd %4, %2 + pxor %1, %3 + pxor %2, %4 + psubd %1, %3 + psubd %2, %4 +%endmacro + +%macro ABSB_SSSE3 2 + pabsb %1, %1 +%endmacro + +%macro ABSB2_SSSE3 4 + pabsb %1, %1 + pabsb %2, %2 +%endmacro + +%macro ABS4 6 + ABS2 %1, %2, %5, %6 + ABS2 %3, %4, %5, %6 +%endmacro + +%define ABS1 ABS1_MMX +%define ABS2 ABS2_MMX +%define ABSB ABSB_MMX +%define ABSB2 ABSB2_MMX + +%macro SPLATB_MMX 3 + movd %1, [%2-3] ;to avoid crossing a cacheline + punpcklbw %1, %1 + SPLATW %1, %1, 3 +%endmacro + +%macro SPLATB_SSSE3 3 + movd %1, [%2-3] + pshufb %1, %3 +%endmacro + +%macro PALIGNR_MMX 4-5 ; [dst,] src1, src2, imm, tmp + %define %%dst %1 +%if %0==5 +%ifnidn %1, %2 + mova %%dst, %2 +%endif + %rotate 1 +%endif +%ifnidn %4, %2 + mova %4, %2 +%endif +%if mmsize==8 + psllq %%dst, (8-%3)*8 + psrlq %4, %3*8 +%else + pslldq %%dst, 16-%3 + psrldq %4, %3 +%endif + por %%dst, %4 +%endmacro + +%macro PALIGNR_SSSE3 4-5 +%if %0==5 + palignr %1, %2, %3, %4 +%else + palignr %1, %2, %3 +%endif +%endmacro + +%macro DEINTB 5 ; mask, reg1, mask, reg2, optional src to fill masks from +%ifnum %5 + pand m%3, m%5, m%4 ; src .. y6 .. y4 + pand m%1, m%5, m%2 ; dst .. y6 .. y4 +%else + mova m%1, %5 + pand m%3, m%1, m%4 ; src .. y6 .. y4 + pand m%1, m%1, m%2 ; dst .. y6 .. y4 +%endif + psrlw m%2, 8 ; dst .. y7 .. y5 + psrlw m%4, 8 ; src .. y7 .. y5 +%endmacro + +%macro SUMSUB_BA 3-4 +%if %0==3 + padd%1 m%2, m%3 + padd%1 m%3, m%3 + psub%1 m%3, m%2 +%else +%if avx_enabled == 0 + mova m%4, m%2 + padd%1 m%2, m%3 + psub%1 m%3, m%4 +%else + padd%1 m%4, m%2, m%3 + psub%1 m%3, m%2 + SWAP %2, %4 +%endif +%endif +%endmacro + +%macro SUMSUB_BADC 5-6 +%if %0==6 + SUMSUB_BA %1, %2, %3, %6 + SUMSUB_BA %1, %4, %5, %6 +%else + padd%1 m%2, m%3 + padd%1 m%4, m%5 + padd%1 m%3, m%3 + padd%1 m%5, m%5 + psub%1 m%3, m%2 + psub%1 m%5, m%4 +%endif +%endmacro + +%macro SUMSUB2_AB 4 +%ifnum %3 + psub%1 m%4, m%2, m%3 + psub%1 m%4, m%3 + padd%1 m%2, m%2 + padd%1 m%2, m%3 +%else + mova m%4, m%2 + padd%1 m%2, m%2 + padd%1 m%2, %3 + psub%1 m%4, %3 + psub%1 m%4, %3 +%endif +%endmacro + +%macro SUMSUB2_BA 4 +%if avx_enabled == 0 + mova m%4, m%2 + padd%1 m%2, m%3 + padd%1 m%2, m%3 + psub%1 m%3, m%4 + psub%1 m%3, m%4 +%else + padd%1 m%4, m%2, m%3 + padd%1 m%4, m%3 + psub%1 m%3, m%2 + psub%1 m%3, m%2 + SWAP %2, %4 +%endif +%endmacro + +%macro SUMSUBD2_AB 5 +%ifnum %4 + psra%1 m%5, m%2, 1 ; %3: %3>>1 + psra%1 m%4, m%3, 1 ; %2: %2>>1 + padd%1 m%4, m%2 ; %3: %3>>1+%2 + psub%1 m%5, m%3 ; %2: %2>>1-%3 + SWAP %2, %5 + SWAP %3, %4 +%else + mova %5, m%2 + mova %4, m%3 + psra%1 m%3, 1 ; %3: %3>>1 + psra%1 m%2, 1 ; %2: %2>>1 + padd%1 m%3, %5 ; %3: %3>>1+%2 + psub%1 m%2, %4 ; %2: %2>>1-%3 +%endif +%endmacro + +%macro DCT4_1D 5 +%ifnum %5 + SUMSUB_BADC w, %4, %1, %3, %2, %5 + SUMSUB_BA w, %3, %4, %5 + SUMSUB2_AB w, %1, %2, %5 + SWAP %1, %3, %4, %5, %2 +%else + SUMSUB_BADC w, %4, %1, %3, %2 + SUMSUB_BA w, %3, %4 + mova [%5], m%2 + SUMSUB2_AB w, %1, [%5], %2 + SWAP %1, %3, %4, %2 +%endif +%endmacro + +%macro IDCT4_1D 6-7 +%ifnum %6 + SUMSUBD2_AB %1, %3, %5, %7, %6 + ; %3: %3>>1-%5 %5: %3+%5>>1 + SUMSUB_BA %1, %4, %2, %7 + ; %4: %2+%4 %2: %2-%4 + SUMSUB_BADC %1, %5, %4, %3, %2, %7 + ; %5: %2+%4 + (%3+%5>>1) + ; %4: %2+%4 - (%3+%5>>1) + ; %3: %2-%4 + (%3>>1-%5) + ; %2: %2-%4 - (%3>>1-%5) +%else +%ifidn %1, w + SUMSUBD2_AB %1, %3, %5, [%6], [%6+16] +%else + SUMSUBD2_AB %1, %3, %5, [%6], [%6+32] +%endif + SUMSUB_BA %1, %4, %2 + SUMSUB_BADC %1, %5, %4, %3, %2 +%endif + SWAP %2, %5, %4 + ; %2: %2+%4 + (%3+%5>>1) row0 + ; %3: %2-%4 + (%3>>1-%5) row1 + ; %4: %2-%4 - (%3>>1-%5) row2 + ; %5: %2+%4 - (%3+%5>>1) row3 +%endmacro + + +%macro LOAD_DIFF 5 +%ifidn %3, none + movh %1, %4 + movh %2, %5 + punpcklbw %1, %2 + punpcklbw %2, %2 + psubw %1, %2 +%else + movh %1, %4 + punpcklbw %1, %3 + movh %2, %5 + punpcklbw %2, %3 + psubw %1, %2 +%endif +%endmacro + +%macro STORE_DCT 6 + movq [%5+%6+ 0], m%1 + movq [%5+%6+ 8], m%2 + movq [%5+%6+16], m%3 + movq [%5+%6+24], m%4 + movhps [%5+%6+32], m%1 + movhps [%5+%6+40], m%2 + movhps [%5+%6+48], m%3 + movhps [%5+%6+56], m%4 +%endmacro + +%macro LOAD_DIFF_8x4P 7-10 r0,r2,0 ; 4x dest, 2x temp, 2x pointer, increment? + LOAD_DIFF m%1, m%5, m%7, [%8], [%9] + LOAD_DIFF m%2, m%6, m%7, [%8+r1], [%9+r3] + LOAD_DIFF m%3, m%5, m%7, [%8+2*r1], [%9+2*r3] + LOAD_DIFF m%4, m%6, m%7, [%8+r4], [%9+r5] +%if %10 + lea %8, [%8+4*r1] + lea %9, [%9+4*r3] +%endif +%endmacro + +%macro DIFFx2 6-7 + movh %3, %5 + punpcklbw %3, %4 + psraw %1, 6 + paddsw %1, %3 + movh %3, %6 + punpcklbw %3, %4 + psraw %2, 6 + paddsw %2, %3 + packuswb %2, %1 +%endmacro + +%macro STORE_DIFF 4 + movh %2, %4 + punpcklbw %2, %3 + psraw %1, 6 + paddsw %1, %2 + packuswb %1, %1 + movh %4, %1 +%endmacro + +%macro STORE_DIFFx2 8 ; add1, add2, reg1, reg2, zero, shift, source, stride + movh %3, [%7] + movh %4, [%7+%8] + psraw %1, %6 + psraw %2, %6 + punpcklbw %3, %5 + punpcklbw %4, %5 + paddw %3, %1 + paddw %4, %2 + packuswb %3, %5 + packuswb %4, %5 + movh [%7], %3 + movh [%7+%8], %4 +%endmacro + +%macro PMINUB_MMX 3 ; dst, src, tmp + mova %3, %1 + psubusb %3, %2 + psubb %1, %3 +%endmacro + +%macro PMINUB_MMXEXT 3 ; dst, src, ignored + pminub %1, %2 +%endmacro + +%macro SPLATW 2-3 0 +%if mmsize == 16 + pshuflw %1, %2, (%3)*0x55 + punpcklqdq %1, %1 +%else + pshufw %1, %2, (%3)*0x55 +%endif +%endmacro + +%macro SPLATD 2-3 0 +%if mmsize == 16 + pshufd %1, %2, (%3)*0x55 +%else + pshufw %1, %2, (%3)*0x11 + ((%3)+1)*0x44 +%endif +%endmacro + +%macro CLIPW 3 ;(dst, min, max) + pmaxsw %1, %2 + pminsw %1, %3 +%endmacro -- cgit v1.2.3