This code I reviewed before he post I think he like split function and integrate into two patches.
At 2013-11-23 02:11:16,"Steve Borho" <[email protected]> wrote: >Did you forget to include the asm-primitive.cpp change, or this this meant to >be for review only? > >On Nov 22, 2013, at 4:49 AM, [email protected] wrote: > >> # HG changeset patch >> # User Nabajit Deka >> # Date 1385116295 -19800 >> # Fri Nov 22 16:01:35 2013 +0530 >> # Node ID 31f6c6e8e965f06825a8b72e5dc42bfb5ce981ff >> # Parent 5009254d3d3ac92e90b1551444c5eb32ba2f8d31 >> asm : routine for weightUnidirPixel(), for input width in multiples of 16. >> >> diff -r 5009254d3d3a -r 31f6c6e8e965 source/common/x86/pixel-util.asm >> --- a/source/common/x86/pixel-util.asm Fri Nov 22 00:17:46 2013 -0600 >> +++ b/source/common/x86/pixel-util.asm Fri Nov 22 16:01:35 2013 +0530 >> @@ -29,6 +29,7 @@ >> c_d_4: dd 4, 4, 4, 4 >> c_d_1234: dd 1, 2, 3, 4 >> >> +tab_c_1: times 8 dw 1 >> >> SECTION .text >> >> @@ -670,3 +671,81 @@ >> movd eax, m7 >> >> RET >> + >> +;------------------------------------------------------------------------------------------------------------------------------------------------------- >> +;void weightUnidirPixel(pixel *src, pixel *dst, intptr_t srcStride, >> intptr_t dstStride, int width, int height, int w0, int round, int shift, int >> offset) >> +;------------------------------------------------------------------------------------------------------------------------------------------------------- >> +INIT_XMM sse4 >> +cglobal weightUnidirPixel, 6, 7, 6 >> + >> + mov r6d, r6m >> + shl r6d, 6 >> + movd m0, r6d ; m0 = [w0<<6] >> + >> + movd m1, r7m ; m1 = [round] >> + punpcklwd m0, m1 ; assuming both (w0<<6) and round are using >> maximum of 16 bits each. >> + pshufd m0, m0, 0 ; m0 = [w0<<6 round] >> + >> + movd m1, r8m >> + >> + movd m2, r9m >> + pshufd m2, m2, 0 >> + >> + mova m5, [tab_c_1] >> + >> + sub r2d, r4d >> + sub r3d, r4d >> + >> +.loopH >> + mov r6d, r4d >> + shr r6d, 4 >> +.loopW: >> + movh m4, [r0] >> + pmovzxbw m4, m4 >> + >> + punpcklwd m3, m4, m5 >> + pmaddwd m3, m0 >> + psrad m3, m1 >> + paddd m3, m2 >> + >> + punpckhwd m4, m5 >> + pmaddwd m4, m0 >> + psrad m4, m1 >> + paddd m4, m2 >> + >> + packssdw m3, m4 >> + packuswb m3, m3 >> + >> + movh [r1], m3 >> + >> + movh m4, [r0 + 8] >> + pmovzxbw m4, m4 >> + >> + punpcklwd m3, m4, m5 >> + pmaddwd m3, m0 >> + psrad m3, m1 >> + paddd m3, m2 >> + >> + punpckhwd m4, m5 >> + pmaddwd m4, m0 >> + psrad m4, m1 >> + paddd m4, m2 >> + >> + packssdw m3, m4 >> + packuswb m3, m3 >> + >> + movh [r1 + 8], m3 >> + >> + add r0, 16 >> + add r1, 16 >> + >> + dec r6d >> + jnz .loopW >> + >> + lea r0, [r0 + r2] >> + lea r1, [r1 + r3] >> + >> + dec r5d >> + jnz .loopH >> + >> + RET >> _______________________________________________ >> x265-devel mailing list >> [email protected] >> https://mailman.videolan.org/listinfo/x265-devel >
_______________________________________________ x265-devel mailing list [email protected] https://mailman.videolan.org/listinfo/x265-devel
