# HG changeset patch # User gpoirier # Date 1155400651 0 # Node ID f7125bf10892830d72b0a4ac305c5444bb16beca # Parent 4baed6f6577ba6bfb7df9a7076283df2d153d9cc Support for MacIntel, last part: balign directives Determines whether .align's arg is power-of-two or not, then defines ASMALIGN appropriately in config.h. Changes all .baligns to ASMALIGNs. Patch by John Dalgliesh % johnd AH defyne P org % Original thread: Date: Aug 11, 2006 8:00 AM Subject: Re: [Ffmpeg-devel] Mac OS X Intel last part: balign directives diff -r 4baed6f6577b -r f7125bf10892 i386/dsputil_mmx.c --- a/i386/dsputil_mmx.c Sat Aug 12 16:31:17 2006 +0000 +++ b/i386/dsputil_mmx.c Sat Aug 12 16:37:31 2006 +0000 @@ -56,7 +56,7 @@ static const uint64_t ff_pb_3F attribute_used __attribute__ ((aligned(8))) = 0x3F3F3F3F3F3F3F3FULL; static const uint64_t ff_pb_FC attribute_used __attribute__ ((aligned(8))) = 0xFCFCFCFCFCFCFCFCULL; -#define JUMPALIGN() __asm __volatile (".balign 8"::) +#define JUMPALIGN() __asm __volatile (ASMALIGN(3)::) #define MOVQ_ZERO(regd) __asm __volatile ("pxor %%" #regd ", %%" #regd ::) #define MOVQ_WONE(regd) \ @@ -204,7 +204,7 @@ asm volatile( "mov $-128, %%"REG_a" \n\t" "pxor %%mm7, %%mm7 \n\t" - ".balign 16 \n\t" + ASMALIGN(4) "1: \n\t" "movq (%0), %%mm0 \n\t" "movq (%0, %2), %%mm2 \n\t" @@ -232,7 +232,7 @@ asm volatile( "pxor %%mm7, %%mm7 \n\t" "mov $-128, %%"REG_a" \n\t" - ".balign 16 \n\t" + ASMALIGN(4) "1: \n\t" "movq (%0), %%mm0 \n\t" "movq (%1), %%mm2 \n\t" @@ -375,7 +375,7 @@ { __asm __volatile( "lea (%3, %3), %%"REG_a" \n\t" - ".balign 8 \n\t" + ASMALIGN(3) "1: \n\t" "movd (%1), %%mm0 \n\t" "movd (%1, %3), %%mm1 \n\t" @@ -401,7 +401,7 @@ { __asm __volatile( "lea (%3, %3), %%"REG_a" \n\t" - ".balign 8 \n\t" + ASMALIGN(3) "1: \n\t" "movq (%1), %%mm0 \n\t" "movq (%1, %3), %%mm1 \n\t" @@ -427,7 +427,7 @@ { __asm __volatile( "lea (%3, %3), %%"REG_a" \n\t" - ".balign 8 \n\t" + ASMALIGN(3) "1: \n\t" "movq (%1), %%mm0 \n\t" "movq 8(%1), %%mm4 \n\t" diff -r 4baed6f6577b -r f7125bf10892 i386/dsputil_mmx_avg.h --- a/i386/dsputil_mmx_avg.h Sat Aug 12 16:31:17 2006 +0000 +++ b/i386/dsputil_mmx_avg.h Sat Aug 12 16:37:31 2006 +0000 @@ -754,7 +754,7 @@ "lea (%3, %3), %%"REG_a" \n\t" "movq (%1), %%mm0 \n\t" PAVGB" 1(%1), %%mm0 \n\t" - ".balign 8 \n\t" + ASMALIGN(3) "1: \n\t" "movq (%1, %%"REG_a"), %%mm2 \n\t" "movq (%1, %3), %%mm1 \n\t" diff -r 4baed6f6577b -r f7125bf10892 i386/dsputil_mmx_rnd.h --- a/i386/dsputil_mmx_rnd.h Sat Aug 12 16:31:17 2006 +0000 +++ b/i386/dsputil_mmx_rnd.h Sat Aug 12 16:37:31 2006 +0000 @@ -28,7 +28,7 @@ MOVQ_BFE(mm6); __asm __volatile( "lea (%3, %3), %%"REG_a" \n\t" - ".balign 8 \n\t" + ASMALIGN(3) "1: \n\t" "movq (%1), %%mm0 \n\t" "movq 1(%1), %%mm1 \n\t" @@ -69,7 +69,7 @@ "movq %%mm4, (%3) \n\t" "add %5, %3 \n\t" "decl %0 \n\t" - ".balign 8 \n\t" + ASMALIGN(3) "1: \n\t" "movq (%1), %%mm0 \n\t" "movq (%2), %%mm1 \n\t" @@ -110,7 +110,7 @@ MOVQ_BFE(mm6); __asm __volatile( "lea (%3, %3), %%"REG_a" \n\t" - ".balign 8 \n\t" + ASMALIGN(3) "1: \n\t" "movq (%1), %%mm0 \n\t" "movq 1(%1), %%mm1 \n\t" @@ -168,7 +168,7 @@ "movq %%mm5, 8(%3) \n\t" "add %5, %3 \n\t" "decl %0 \n\t" - ".balign 8 \n\t" + ASMALIGN(3) "1: \n\t" "movq (%1), %%mm0 \n\t" "movq (%2), %%mm1 \n\t" @@ -206,7 +206,7 @@ __asm __volatile( "lea (%3, %3), %%"REG_a" \n\t" "movq (%1), %%mm0 \n\t" - ".balign 8 \n\t" + ASMALIGN(3) "1: \n\t" "movq (%1, %3), %%mm1 \n\t" "movq (%1, %%"REG_a"),%%mm2 \n\t" @@ -246,7 +246,7 @@ "paddusw %%mm1, %%mm5 \n\t" "xor %%"REG_a", %%"REG_a" \n\t" "add %3, %1 \n\t" - ".balign 8 \n\t" + ASMALIGN(3) "1: \n\t" "movq (%1, %%"REG_a"), %%mm0 \n\t" "movq 1(%1, %%"REG_a"), %%mm2 \n\t" @@ -458,7 +458,7 @@ __asm __volatile( "lea (%3, %3), %%"REG_a" \n\t" "movq (%1), %%mm0 \n\t" - ".balign 8 \n\t" + ASMALIGN(3) "1: \n\t" "movq (%1, %3), %%mm1 \n\t" "movq (%1, %%"REG_a"), %%mm2 \n\t" @@ -509,7 +509,7 @@ "paddusw %%mm1, %%mm5 \n\t" "xor %%"REG_a", %%"REG_a" \n\t" "add %3, %1 \n\t" - ".balign 8 \n\t" + ASMALIGN(3) "1: \n\t" "movq (%1, %%"REG_a"), %%mm0 \n\t" "movq 1(%1, %%"REG_a"), %%mm2 \n\t" diff -r 4baed6f6577b -r f7125bf10892 i386/motion_est_mmx.c --- a/i386/motion_est_mmx.c Sat Aug 12 16:31:17 2006 +0000 +++ b/i386/motion_est_mmx.c Sat Aug 12 16:37:31 2006 +0000 @@ -34,7 +34,7 @@ { long len= -(stride*h); asm volatile( - ".balign 16 \n\t" + ASMALIGN(4) "1: \n\t" "movq (%1, %%"REG_a"), %%mm0 \n\t" "movq (%2, %%"REG_a"), %%mm2 \n\t" @@ -70,7 +70,7 @@ { long len= -(stride*h); asm volatile( - ".balign 16 \n\t" + ASMALIGN(4) "1: \n\t" "movq (%1, %%"REG_a"), %%mm0 \n\t" "movq (%2, %%"REG_a"), %%mm2 \n\t" @@ -92,7 +92,7 @@ { long len= -(stride*h); asm volatile( - ".balign 16 \n\t" + ASMALIGN(4) "1: \n\t" "movq (%1, %%"REG_a"), %%mm0 \n\t" "movq (%2, %%"REG_a"), %%mm2 \n\t" @@ -118,7 +118,7 @@ { //FIXME reuse src long len= -(stride*h); asm volatile( - ".balign 16 \n\t" + ASMALIGN(4) "movq "MANGLE(bone)", %%mm5 \n\t" "1: \n\t" "movq (%1, %%"REG_a"), %%mm0 \n\t" @@ -155,7 +155,7 @@ { long len= -(stride*h); asm volatile( - ".balign 16 \n\t" + ASMALIGN(4) "1: \n\t" "movq (%1, %%"REG_a"), %%mm0 \n\t" "movq (%2, %%"REG_a"), %%mm1 \n\t" @@ -193,7 +193,7 @@ { long len= -(stride*h); asm volatile( - ".balign 16 \n\t" + ASMALIGN(4) "1: \n\t" "movq (%1, %%"REG_a"), %%mm0 \n\t" "movq (%2, %%"REG_a"), %%mm1 \n\t" diff -r 4baed6f6577b -r f7125bf10892 i386/mpegvideo_mmx.c --- a/i386/mpegvideo_mmx.c Sat Aug 12 16:31:17 2006 +0000 +++ b/i386/mpegvideo_mmx.c Sat Aug 12 16:37:31 2006 +0000 @@ -66,7 +66,7 @@ "packssdw %%mm5, %%mm5 \n\t" "psubw %%mm5, %%mm7 \n\t" "pxor %%mm4, %%mm4 \n\t" - ".balign 16 \n\t" + ASMALIGN(4) "1: \n\t" "movq (%0, %3), %%mm0 \n\t" "movq 8(%0, %3), %%mm1 \n\t" @@ -129,7 +129,7 @@ "packssdw %%mm5, %%mm5 \n\t" "psubw %%mm5, %%mm7 \n\t" "pxor %%mm4, %%mm4 \n\t" - ".balign 16 \n\t" + ASMALIGN(4) "1: \n\t" "movq (%0, %3), %%mm0 \n\t" "movq 8(%0, %3), %%mm1 \n\t" @@ -222,7 +222,7 @@ "packssdw %%mm6, %%mm6 \n\t" "packssdw %%mm6, %%mm6 \n\t" "mov %3, %%"REG_a" \n\t" - ".balign 16 \n\t" + ASMALIGN(4) "1: \n\t" "movq (%0, %%"REG_a"), %%mm0 \n\t" "movq 8(%0, %%"REG_a"), %%mm1 \n\t" @@ -285,7 +285,7 @@ "packssdw %%mm6, %%mm6 \n\t" "packssdw %%mm6, %%mm6 \n\t" "mov %3, %%"REG_a" \n\t" - ".balign 16 \n\t" + ASMALIGN(4) "1: \n\t" "movq (%0, %%"REG_a"), %%mm0 \n\t" "movq 8(%0, %%"REG_a"), %%mm1 \n\t" @@ -357,7 +357,7 @@ "packssdw %%mm6, %%mm6 \n\t" "packssdw %%mm6, %%mm6 \n\t" "mov %3, %%"REG_a" \n\t" - ".balign 16 \n\t" + ASMALIGN(4) "1: \n\t" "movq (%0, %%"REG_a"), %%mm0 \n\t" "movq 8(%0, %%"REG_a"), %%mm1 \n\t" @@ -418,7 +418,7 @@ "packssdw %%mm6, %%mm6 \n\t" "packssdw %%mm6, %%mm6 \n\t" "mov %3, %%"REG_a" \n\t" - ".balign 16 \n\t" + ASMALIGN(4) "1: \n\t" "movq (%0, %%"REG_a"), %%mm0 \n\t" "movq 8(%0, %%"REG_a"), %%mm1 \n\t" diff -r 4baed6f6577b -r f7125bf10892 i386/mpegvideo_mmx_template.c --- a/i386/mpegvideo_mmx_template.c Sat Aug 12 16:31:17 2006 +0000 +++ b/i386/mpegvideo_mmx_template.c Sat Aug 12 16:37:31 2006 +0000 @@ -112,7 +112,7 @@ "pxor %%mm6, %%mm6 \n\t" "psubw (%3), %%mm6 \n\t" // -bias[0] "mov $-128, %%"REG_a" \n\t" - ".balign 16 \n\t" + ASMALIGN(4) "1: \n\t" "pxor %%mm1, %%mm1 \n\t" // 0 "movq (%1, %%"REG_a"), %%mm0 \n\t" // block[i] @@ -156,7 +156,7 @@ "pxor %%mm7, %%mm7 \n\t" // 0 "pxor %%mm4, %%mm4 \n\t" // 0 "mov $-128, %%"REG_a" \n\t" - ".balign 16 \n\t" + ASMALIGN(4) "1: \n\t" "pxor %%mm1, %%mm1 \n\t" // 0 "movq (%1, %%"REG_a"), %%mm0 \n\t" // block[i] diff -r 4baed6f6577b -r f7125bf10892 i386/simple_idct_mmx.c --- a/i386/simple_idct_mmx.c Sat Aug 12 16:31:17 2006 +0000 +++ b/i386/simple_idct_mmx.c Sat Aug 12 16:37:31 2006 +0000 @@ -785,7 +785,7 @@ IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0), 20) "jmp 9f \n\t" - "#.balign 16 \n\t"\ + "#" ASMALIGN(4) \ "4: \n\t" Z_COND_IDCT( 64(%0), 72(%0), 80(%0), 88(%0), 64(%1),paddd (%2), 11, 6f) Z_COND_IDCT( 96(%0),104(%0),112(%0),120(%0), 96(%1),paddd (%2), 11, 5f) @@ -860,7 +860,7 @@ IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0), 20) "jmp 9f \n\t" - "#.balign 16 \n\t"\ + "#" ASMALIGN(4) \ "6: \n\t" Z_COND_IDCT( 96(%0),104(%0),112(%0),120(%0), 96(%1),paddd (%2), 11, 7f) @@ -926,7 +926,7 @@ IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0), 20) "jmp 9f \n\t" - "#.balign 16 \n\t"\ + "#" ASMALIGN(4) \ "2: \n\t" Z_COND_IDCT( 96(%0),104(%0),112(%0),120(%0), 96(%1),paddd (%2), 11, 3f) @@ -1003,7 +1003,7 @@ IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0), 20) "jmp 9f \n\t" - "#.balign 16 \n\t"\ + "#" ASMALIGN(4) \ "3: \n\t" #undef IDCT #define IDCT(src0, src4, src1, src5, dst, shift) \ @@ -1067,7 +1067,7 @@ IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0), 20) "jmp 9f \n\t" - "#.balign 16 \n\t"\ + "#" ASMALIGN(4) \ "5: \n\t" #undef IDCT #define IDCT(src0, src4, src1, src5, dst, shift) \ @@ -1132,7 +1132,7 @@ "jmp 9f \n\t" - "#.balign 16 \n\t"\ + "#" ASMALIGN(4) \ "1: \n\t" #undef IDCT #define IDCT(src0, src4, src1, src5, dst, shift) \ @@ -1206,7 +1206,7 @@ "jmp 9f \n\t" - "#.balign 16 \n\t" + "#" ASMALIGN(4) "7: \n\t" #undef IDCT #define IDCT(src0, src4, src1, src5, dst, shift) \