diff options
Diffstat (limited to 'libavcodec/x86/audiodsp.asm')
| -rw-r--r-- | libavcodec/x86/audiodsp.asm | 100 |
1 files changed, 70 insertions, 30 deletions
diff --git a/libavcodec/x86/audiodsp.asm b/libavcodec/x86/audiodsp.asm index f2e831df17..273b9ef660 100644 --- a/libavcodec/x86/audiodsp.asm +++ b/libavcodec/x86/audiodsp.asm @@ -2,20 +2,20 @@ ;* optimized audio functions ;* Copyright (c) 2008 Loren Merritt ;* -;* 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 ;****************************************************************************** @@ -40,15 +40,11 @@ cglobal scalarproduct_int16, 3,3,3, v1, v2, order paddd m2, m1 add orderq, mmsize*2 jl .loop -%if mmsize == 16 - movhlps m0, m2 - paddd m2, m0 - pshuflw m0, m2, 0x4e -%else - pshufw m0, m2, 0x4e -%endif - paddd m2, m0 + HADDD m2, m0 movd eax, m2 +%if mmsize == 8 + emms +%endif RET %endmacro @@ -80,17 +76,17 @@ cglobal vector_clip_int32%5, 5,5,%1, dst, src, min, max, len SPLATD m4 SPLATD m5 .loop: -%assign %%i 1 +%assign %%i 0 %rep %2 - mova m0, [srcq+mmsize*0*%%i] - mova m1, [srcq+mmsize*1*%%i] - mova m2, [srcq+mmsize*2*%%i] - mova m3, [srcq+mmsize*3*%%i] + mova m0, [srcq+mmsize*(0+%%i)] + mova m1, [srcq+mmsize*(1+%%i)] + mova m2, [srcq+mmsize*(2+%%i)] + mova m3, [srcq+mmsize*(3+%%i)] %if %3 - mova m7, [srcq+mmsize*4*%%i] - mova m8, [srcq+mmsize*5*%%i] - mova m9, [srcq+mmsize*6*%%i] - mova m10, [srcq+mmsize*7*%%i] + mova m7, [srcq+mmsize*(4+%%i)] + mova m8, [srcq+mmsize*(5+%%i)] + mova m9, [srcq+mmsize*(6+%%i)] + mova m10, [srcq+mmsize*(7+%%i)] %endif CLIPD m0, m4, m5, m6 CLIPD m1, m4, m5, m6 @@ -102,17 +98,17 @@ cglobal vector_clip_int32%5, 5,5,%1, dst, src, min, max, len CLIPD m9, m4, m5, m6 CLIPD m10, m4, m5, m6 %endif - mova [dstq+mmsize*0*%%i], m0 - mova [dstq+mmsize*1*%%i], m1 - mova [dstq+mmsize*2*%%i], m2 - mova [dstq+mmsize*3*%%i], m3 + mova [dstq+mmsize*(0+%%i)], m0 + mova [dstq+mmsize*(1+%%i)], m1 + mova [dstq+mmsize*(2+%%i)], m2 + mova [dstq+mmsize*(3+%%i)], m3 %if %3 - mova [dstq+mmsize*4*%%i], m7 - mova [dstq+mmsize*5*%%i], m8 - mova [dstq+mmsize*6*%%i], m9 - mova [dstq+mmsize*7*%%i], m10 + mova [dstq+mmsize*(4+%%i)], m7 + mova [dstq+mmsize*(5+%%i)], m8 + mova [dstq+mmsize*(6+%%i)], m9 + mova [dstq+mmsize*(7+%%i)], m10 %endif -%assign %%i %%i+1 +%assign %%i %%i+4*(%3+1) %endrep add srcq, mmsize*4*(%2+%3) add dstq, mmsize*4*(%2+%3) @@ -135,3 +131,47 @@ VECTOR_CLIP_INT32 11, 1, 1, 0 %else VECTOR_CLIP_INT32 6, 1, 0, 0 %endif + +;----------------------------------------------------- +;void ff_vector_clipf(float *dst, const float *src, +; float min, float max, int len) +;----------------------------------------------------- +INIT_XMM sse +%if UNIX64 +cglobal vector_clipf, 3,3,6, dst, src, len +%else +cglobal vector_clipf, 5,5,6, dst, src, min, max, len +%endif +%if WIN64 + SWAP 0, 2 + SWAP 1, 3 +%elif ARCH_X86_32 + movss m0, minm + movss m1, maxm +%endif + SPLATD m0 + SPLATD m1 + shl lend, 2 + add srcq, lenq + add dstq, lenq + neg lenq +.loop: + mova m2, [srcq+lenq+mmsize*0] + mova m3, [srcq+lenq+mmsize*1] + mova m4, [srcq+lenq+mmsize*2] + mova m5, [srcq+lenq+mmsize*3] + maxps m2, m0 + maxps m3, m0 + maxps m4, m0 + maxps m5, m0 + minps m2, m1 + minps m3, m1 + minps m4, m1 + minps m5, m1 + mova [dstq+lenq+mmsize*0], m2 + mova [dstq+lenq+mmsize*1], m3 + mova [dstq+lenq+mmsize*2], m4 + mova [dstq+lenq+mmsize*3], m5 + add lenq, mmsize*4 + jl .loop + REP_RET |