[FFmpeg-devel] [PATCH] Replace ASMALIGN() with .p2align
David Conrad
lessen42
Wed Jun 2 16:05:31 CEST 2010
On Jun 2, 2010, at 9:50 AM, Mans Rullgard wrote:
> The test for .align syntax was apparently unreliable with some compilers.
> Using the .p2align directive instead avoids the need for this test.
Don't forget swscale.
> ---
> configure | 7 -------
> libavcodec/x86/dsputil_mmx.c | 8 ++++----
> libavcodec/x86/dsputil_mmx_avg_template.c | 2 +-
> libavcodec/x86/dsputil_mmx_qns_template.c | 4 ++--
> libavcodec/x86/dsputil_mmx_rnd_template.c | 16 ++++++++--------
> libavcodec/x86/dsputilenc_mmx.c | 4 ++--
> libavcodec/x86/idct_sse2_xvid.c | 2 +-
> libavcodec/x86/motion_est_mmx.c | 16 ++++++++--------
> libavcodec/x86/mpegvideo_mmx.c | 12 ++++++------
> libavcodec/x86/mpegvideo_mmx_template.c | 4 ++--
> libavcodec/x86/simple_idct_mmx.c | 14 +++++++-------
> libavcodec/x86/vc1dsp_mmx.c | 6 +++---
> 12 files changed, 44 insertions(+), 51 deletions(-)
>
> diff --git a/configure b/configure
> index 4f65143..d2635ad 100755
> --- a/configure
> +++ b/configure
> @@ -2794,9 +2794,6 @@ if enabled gprof; then
> add_ldflags -p
> fi
>
> -# Find out if the .align argument is a power of two or not.
> -check_asm asmalign_pot '".align 3"'
> -
> enabled_any $THREADS_LIST && enable threads
>
> check_deps $CONFIG_LIST \
> @@ -2819,7 +2816,6 @@ enabled asm || { arch=c; disable $ARCH_LIST $ARCH_EXT_LIST; }
> echo "install prefix $prefix"
> echo "source path $source_path"
> echo "C compiler $cc"
> -echo ".align is power-of-two $asmalign_pot"
> echo "ARCH $arch ($cpu)"
> if test "$build_suffix" != ""; then
> echo "build suffix $build_suffix"
> @@ -3051,8 +3047,6 @@ get_version LIBAVFORMAT libavformat/avformat.h
> get_version LIBAVUTIL libavutil/avutil.h
> get_version LIBAVFILTER libavfilter/avfilter.h
>
> -enabled asmalign_pot || align_shift="1 <<"
> -
> cat > $TMPH <<EOF
> /* Automatically generated by configure - do not modify! */
> #ifndef FFMPEG_CONFIG_H
> @@ -3063,7 +3057,6 @@ cat > $TMPH <<EOF
> #define CC_TYPE "$cc_type"
> #define CC_VERSION $cc_version
> #define restrict $_restrict
> -#define ASMALIGN(ZEROBITS) ".align $align_shift " #ZEROBITS "\\n\\t"
> #define EXTERN_PREFIX "${extern_prefix}"
> #define EXTERN_ASM ${extern_prefix}
> EOF
> diff --git a/libavcodec/x86/dsputil_mmx.c b/libavcodec/x86/dsputil_mmx.c
> index cc2f881..7bb3296 100644
> --- a/libavcodec/x86/dsputil_mmx.c
> +++ b/libavcodec/x86/dsputil_mmx.c
> @@ -73,7 +73,7 @@ DECLARE_ALIGNED(8, const uint64_t, ff_pb_FC ) = 0xFCFCFCFCFCFCFCFCULL;
> DECLARE_ALIGNED(16, const double, ff_pd_1)[2] = { 1.0, 1.0 };
> DECLARE_ALIGNED(16, const double, ff_pd_2)[2] = { 2.0, 2.0 };
>
> -#define JUMPALIGN() __asm__ volatile (ASMALIGN(3)::)
> +#define JUMPALIGN() __asm__ volatile (".p2align 3"::)
> #define MOVQ_ZERO(regd) __asm__ volatile ("pxor %%" #regd ", %%" #regd ::)
>
> #define MOVQ_BFE(regd) \
> @@ -360,7 +360,7 @@ static void put_pixels4_mmx(uint8_t *block, const uint8_t *pixels, int line_size
> {
> __asm__ volatile(
> "lea (%3, %3), %%"REG_a" \n\t"
> - ASMALIGN(3)
> + ".p2align 3 \n\t"
> "1: \n\t"
> "movd (%1), %%mm0 \n\t"
> "movd (%1, %3), %%mm1 \n\t"
> @@ -386,7 +386,7 @@ static void put_pixels8_mmx(uint8_t *block, const uint8_t *pixels, int line_size
> {
> __asm__ volatile(
> "lea (%3, %3), %%"REG_a" \n\t"
> - ASMALIGN(3)
> + ".p2align 3 \n\t"
> "1: \n\t"
> "movq (%1), %%mm0 \n\t"
> "movq (%1, %3), %%mm1 \n\t"
> @@ -412,7 +412,7 @@ static void put_pixels16_mmx(uint8_t *block, const uint8_t *pixels, int line_siz
> {
> __asm__ volatile(
> "lea (%3, %3), %%"REG_a" \n\t"
> - ASMALIGN(3)
> + ".p2align 3 \n\t"
> "1: \n\t"
> "movq (%1), %%mm0 \n\t"
> "movq 8(%1), %%mm4 \n\t"
> diff --git a/libavcodec/x86/dsputil_mmx_avg_template.c b/libavcodec/x86/dsputil_mmx_avg_template.c
> index 8220867..d0e6e01 100644
> --- a/libavcodec/x86/dsputil_mmx_avg_template.c
> +++ b/libavcodec/x86/dsputil_mmx_avg_template.c
> @@ -757,7 +757,7 @@ static void DEF(avg_pixels8_xy2)(uint8_t *block, const uint8_t *pixels, int line
> "lea (%3, %3), %%"REG_a" \n\t"
> "movq (%1), %%mm0 \n\t"
> PAVGB" 1(%1), %%mm0 \n\t"
> - ASMALIGN(3)
> + ".p2align 3 \n\t"
> "1: \n\t"
> "movq (%1, %%"REG_a"), %%mm2 \n\t"
> "movq (%1, %3), %%mm1 \n\t"
> diff --git a/libavcodec/x86/dsputil_mmx_qns_template.c b/libavcodec/x86/dsputil_mmx_qns_template.c
> index d2dbfc5..77a41b9 100644
> --- a/libavcodec/x86/dsputil_mmx_qns_template.c
> +++ b/libavcodec/x86/dsputil_mmx_qns_template.c
> @@ -37,7 +37,7 @@ static int DEF(try_8x8basis)(int16_t rem[64], int16_t weight[64], int16_t basis[
> "movd %4, %%mm5 \n\t"
> "punpcklwd %%mm5, %%mm5 \n\t"
> "punpcklwd %%mm5, %%mm5 \n\t"
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> "movq (%1, %0), %%mm0 \n\t"
> "movq 8(%1, %0), %%mm1 \n\t"
> @@ -77,7 +77,7 @@ static void DEF(add_8x8basis)(int16_t rem[64], int16_t basis[64], int scale)
> "movd %3, %%mm5 \n\t"
> "punpcklwd %%mm5, %%mm5 \n\t"
> "punpcklwd %%mm5, %%mm5 \n\t"
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> "movq (%1, %0), %%mm0 \n\t"
> "movq 8(%1, %0), %%mm1 \n\t"
> diff --git a/libavcodec/x86/dsputil_mmx_rnd_template.c b/libavcodec/x86/dsputil_mmx_rnd_template.c
> index 2fc1756..e4c9138 100644
> --- a/libavcodec/x86/dsputil_mmx_rnd_template.c
> +++ b/libavcodec/x86/dsputil_mmx_rnd_template.c
> @@ -30,7 +30,7 @@ static void DEF(put, pixels8_x2)(uint8_t *block, const uint8_t *pixels, int line
> MOVQ_BFE(mm6);
> __asm__ volatile(
> "lea (%3, %3), %%"REG_a" \n\t"
> - ASMALIGN(3)
> + ".p2align 3 \n\t"
> "1: \n\t"
> "movq (%1), %%mm0 \n\t"
> "movq 1(%1), %%mm1 \n\t"
> @@ -71,7 +71,7 @@ static void av_unused DEF(put, pixels8_l2)(uint8_t *dst, uint8_t *src1, uint8_t
> "movq %%mm4, (%3) \n\t"
> "add %5, %3 \n\t"
> "decl %0 \n\t"
> - ASMALIGN(3)
> + ".p2align 3 \n\t"
> "1: \n\t"
> "movq (%1), %%mm0 \n\t"
> "movq (%2), %%mm1 \n\t"
> @@ -112,7 +112,7 @@ static void DEF(put, pixels16_x2)(uint8_t *block, const uint8_t *pixels, int lin
> MOVQ_BFE(mm6);
> __asm__ volatile(
> "lea (%3, %3), %%"REG_a" \n\t"
> - ASMALIGN(3)
> + ".p2align 3 \n\t"
> "1: \n\t"
> "movq (%1), %%mm0 \n\t"
> "movq 1(%1), %%mm1 \n\t"
> @@ -170,7 +170,7 @@ static void av_unused DEF(put, pixels16_l2)(uint8_t *dst, uint8_t *src1, uint8_t
> "movq %%mm5, 8(%3) \n\t"
> "add %5, %3 \n\t"
> "decl %0 \n\t"
> - ASMALIGN(3)
> + ".p2align 3 \n\t"
> "1: \n\t"
> "movq (%1), %%mm0 \n\t"
> "movq (%2), %%mm1 \n\t"
> @@ -208,7 +208,7 @@ static void DEF(put, pixels8_y2)(uint8_t *block, const uint8_t *pixels, int line
> __asm__ volatile(
> "lea (%3, %3), %%"REG_a" \n\t"
> "movq (%1), %%mm0 \n\t"
> - ASMALIGN(3)
> + ".p2align 3 \n\t"
> "1: \n\t"
> "movq (%1, %3), %%mm1 \n\t"
> "movq (%1, %%"REG_a"),%%mm2 \n\t"
> @@ -248,7 +248,7 @@ static void DEF(put, pixels8_xy2)(uint8_t *block, const uint8_t *pixels, int lin
> "paddusw %%mm1, %%mm5 \n\t"
> "xor %%"REG_a", %%"REG_a" \n\t"
> "add %3, %1 \n\t"
> - ASMALIGN(3)
> + ".p2align 3 \n\t"
> "1: \n\t"
> "movq (%1, %%"REG_a"), %%mm0 \n\t"
> "movq 1(%1, %%"REG_a"), %%mm2 \n\t"
> @@ -460,7 +460,7 @@ static void DEF(avg, pixels8_y2)(uint8_t *block, const uint8_t *pixels, int line
> __asm__ volatile(
> "lea (%3, %3), %%"REG_a" \n\t"
> "movq (%1), %%mm0 \n\t"
> - ASMALIGN(3)
> + ".p2align 3 \n\t"
> "1: \n\t"
> "movq (%1, %3), %%mm1 \n\t"
> "movq (%1, %%"REG_a"), %%mm2 \n\t"
> @@ -511,7 +511,7 @@ static void DEF(avg, pixels8_xy2)(uint8_t *block, const uint8_t *pixels, int lin
> "paddusw %%mm1, %%mm5 \n\t"
> "xor %%"REG_a", %%"REG_a" \n\t"
> "add %3, %1 \n\t"
> - ASMALIGN(3)
> + ".p2align 3 \n\t"
> "1: \n\t"
> "movq (%1, %%"REG_a"), %%mm0 \n\t"
> "movq 1(%1, %%"REG_a"), %%mm2 \n\t"
> diff --git a/libavcodec/x86/dsputilenc_mmx.c b/libavcodec/x86/dsputilenc_mmx.c
> index f491111..886b1e9 100644
> --- a/libavcodec/x86/dsputilenc_mmx.c
> +++ b/libavcodec/x86/dsputilenc_mmx.c
> @@ -34,7 +34,7 @@ static void get_pixels_mmx(DCTELEM *block, const uint8_t *pixels, int line_size)
> __asm__ volatile(
> "mov $-128, %%"REG_a" \n\t"
> "pxor %%mm7, %%mm7 \n\t"
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> "movq (%0), %%mm0 \n\t"
> "movq (%0, %2), %%mm2 \n\t"
> @@ -96,7 +96,7 @@ static inline void diff_pixels_mmx(DCTELEM *block, const uint8_t *s1, const uint
> __asm__ volatile(
> "pxor %%mm7, %%mm7 \n\t"
> "mov $-128, %%"REG_a" \n\t"
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> "movq (%0), %%mm0 \n\t"
> "movq (%1), %%mm2 \n\t"
> diff --git a/libavcodec/x86/idct_sse2_xvid.c b/libavcodec/x86/idct_sse2_xvid.c
> index fc670e2..f64261d 100644
> --- a/libavcodec/x86/idct_sse2_xvid.c
> +++ b/libavcodec/x86/idct_sse2_xvid.c
> @@ -355,7 +355,7 @@ inline void ff_idct_xvid_sse2(short *block)
> TEST_TWO_ROWS("5*16(%0)", "6*16(%0)", "%%eax", "%%edx", CLEAR_ODD(ROW5), CLEAR_EVEN(ROW6))
> TEST_ONE_ROW("7*16(%0)", "%%esi", CLEAR_ODD(ROW7))
> iLLM_HEAD
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> JNZ("%%ecx", "2f")
> JNZ("%%eax", "3f")
> JNZ("%%edx", "4f")
> diff --git a/libavcodec/x86/motion_est_mmx.c b/libavcodec/x86/motion_est_mmx.c
> index 0272410..91c7582 100644
> --- a/libavcodec/x86/motion_est_mmx.c
> +++ b/libavcodec/x86/motion_est_mmx.c
> @@ -38,7 +38,7 @@ static inline void sad8_1_mmx(uint8_t *blk1, uint8_t *blk2, int stride, int h)
> {
> x86_reg len= -(stride*h);
> __asm__ volatile(
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> "movq (%1, %%"REG_a"), %%mm0 \n\t"
> "movq (%2, %%"REG_a"), %%mm2 \n\t"
> @@ -73,7 +73,7 @@ static inline void sad8_1_mmx(uint8_t *blk1, uint8_t *blk2, int stride, int h)
> static inline void sad8_1_mmx2(uint8_t *blk1, uint8_t *blk2, int stride, int h)
> {
> __asm__ volatile(
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> "movq (%1), %%mm0 \n\t"
> "movq (%1, %3), %%mm1 \n\t"
> @@ -95,7 +95,7 @@ static int sad16_sse2(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)
> int ret;
> __asm__ volatile(
> "pxor %%xmm6, %%xmm6 \n\t"
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> "movdqu (%1), %%xmm0 \n\t"
> "movdqu (%1, %3), %%xmm1 \n\t"
> @@ -122,7 +122,7 @@ static int sad16_sse2(void *v, uint8_t *blk2, uint8_t *blk1, int stride, int h)
> static inline void sad8_x2a_mmx2(uint8_t *blk1, uint8_t *blk2, int stride, int h)
> {
> __asm__ volatile(
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> "movq (%1), %%mm0 \n\t"
> "movq (%1, %3), %%mm1 \n\t"
> @@ -146,7 +146,7 @@ static inline void sad8_y2a_mmx2(uint8_t *blk1, uint8_t *blk2, int stride, int h
> __asm__ volatile(
> "movq (%1), %%mm0 \n\t"
> "add %3, %1 \n\t"
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> "movq (%1), %%mm1 \n\t"
> "movq (%1, %3), %%mm2 \n\t"
> @@ -173,7 +173,7 @@ static inline void sad8_4_mmx2(uint8_t *blk1, uint8_t *blk2, int stride, int h)
> "movq (%1), %%mm0 \n\t"
> "pavgb 1(%1), %%mm0 \n\t"
> "add %3, %1 \n\t"
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> "movq (%1), %%mm1 \n\t"
> "movq (%1,%3), %%mm2 \n\t"
> @@ -200,7 +200,7 @@ static inline void sad8_2_mmx(uint8_t *blk1a, uint8_t *blk1b, uint8_t *blk2, int
> {
> x86_reg len= -(stride*h);
> __asm__ volatile(
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> "movq (%1, %%"REG_a"), %%mm0 \n\t"
> "movq (%2, %%"REG_a"), %%mm1 \n\t"
> @@ -248,7 +248,7 @@ static inline void sad8_4_mmx(uint8_t *blk1, uint8_t *blk2, int stride, int h)
> "punpckhbw %%mm7, %%mm3 \n\t"
> "paddw %%mm2, %%mm0 \n\t"
> "paddw %%mm3, %%mm1 \n\t"
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> "movq (%2, %%"REG_a"), %%mm2 \n\t"
> "movq 1(%2, %%"REG_a"), %%mm4 \n\t"
> diff --git a/libavcodec/x86/mpegvideo_mmx.c b/libavcodec/x86/mpegvideo_mmx.c
> index 5deb68d..f89100e 100644
> --- a/libavcodec/x86/mpegvideo_mmx.c
> +++ b/libavcodec/x86/mpegvideo_mmx.c
> @@ -65,7 +65,7 @@ __asm__ volatile(
> "packssdw %%mm5, %%mm5 \n\t"
> "psubw %%mm5, %%mm7 \n\t"
> "pxor %%mm4, %%mm4 \n\t"
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> "movq (%0, %3), %%mm0 \n\t"
> "movq 8(%0, %3), %%mm1 \n\t"
> @@ -128,7 +128,7 @@ __asm__ volatile(
> "packssdw %%mm5, %%mm5 \n\t"
> "psubw %%mm5, %%mm7 \n\t"
> "pxor %%mm4, %%mm4 \n\t"
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> "movq (%0, %3), %%mm0 \n\t"
> "movq 8(%0, %3), %%mm1 \n\t"
> @@ -221,7 +221,7 @@ __asm__ volatile(
> "packssdw %%mm6, %%mm6 \n\t"
> "packssdw %%mm6, %%mm6 \n\t"
> "mov %3, %%"REG_a" \n\t"
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> "movq (%0, %%"REG_a"), %%mm0 \n\t"
> "movq 8(%0, %%"REG_a"), %%mm1 \n\t"
> @@ -284,7 +284,7 @@ __asm__ volatile(
> "packssdw %%mm6, %%mm6 \n\t"
> "packssdw %%mm6, %%mm6 \n\t"
> "mov %3, %%"REG_a" \n\t"
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> "movq (%0, %%"REG_a"), %%mm0 \n\t"
> "movq 8(%0, %%"REG_a"), %%mm1 \n\t"
> @@ -356,7 +356,7 @@ __asm__ volatile(
> "packssdw %%mm6, %%mm6 \n\t"
> "packssdw %%mm6, %%mm6 \n\t"
> "mov %3, %%"REG_a" \n\t"
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> "movq (%0, %%"REG_a"), %%mm0 \n\t"
> "movq 8(%0, %%"REG_a"), %%mm1 \n\t"
> @@ -417,7 +417,7 @@ __asm__ volatile(
> "packssdw %%mm6, %%mm6 \n\t"
> "packssdw %%mm6, %%mm6 \n\t"
> "mov %3, %%"REG_a" \n\t"
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> "movq (%0, %%"REG_a"), %%mm0 \n\t"
> "movq 8(%0, %%"REG_a"), %%mm1 \n\t"
> diff --git a/libavcodec/x86/mpegvideo_mmx_template.c b/libavcodec/x86/mpegvideo_mmx_template.c
> index 0d92792..a7e999f 100644
> --- a/libavcodec/x86/mpegvideo_mmx_template.c
> +++ b/libavcodec/x86/mpegvideo_mmx_template.c
> @@ -158,7 +158,7 @@ static int RENAME(dct_quantize)(MpegEncContext *s,
> "pxor "MM"6, "MM"6 \n\t"
> "psubw (%3), "MM"6 \n\t" // -bias[0]
> "mov $-128, %%"REG_a" \n\t"
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> MOVQ" (%1, %%"REG_a"), "MM"0 \n\t" // block[i]
> SAVE_SIGN(MM"1", MM"0") // ABS(block[i])
> @@ -188,7 +188,7 @@ static int RENAME(dct_quantize)(MpegEncContext *s,
> "pxor "MM"7, "MM"7 \n\t" // 0
> "pxor "MM"4, "MM"4 \n\t" // 0
> "mov $-128, %%"REG_a" \n\t"
> - ASMALIGN(4)
> + ".p2align 4 \n\t"
> "1: \n\t"
> MOVQ" (%1, %%"REG_a"), "MM"0 \n\t" // block[i]
> SAVE_SIGN(MM"1", MM"0") // ABS(block[i])
> diff --git a/libavcodec/x86/simple_idct_mmx.c b/libavcodec/x86/simple_idct_mmx.c
> index 5ea4c84..abbc0f2 100644
> --- a/libavcodec/x86/simple_idct_mmx.c
> +++ b/libavcodec/x86/simple_idct_mmx.c
> @@ -789,7 +789,7 @@ IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0), 20)
> IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0), 20)
> "jmp 9f \n\t"
>
> - "#" ASMALIGN(4) \
> + "# .p2align 4 \n\t"\
> "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)
> @@ -864,7 +864,7 @@ IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0), 20)
> IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0), 20)
> "jmp 9f \n\t"
>
> - "#" ASMALIGN(4) \
> + "# .p2align 4 \n\t"\
> "6: \n\t"
> Z_COND_IDCT( 96(%0),104(%0),112(%0),120(%0), 96(%1),paddd (%2), 11, 7f)
>
> @@ -930,7 +930,7 @@ IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0), 20)
> IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0), 20)
> "jmp 9f \n\t"
>
> - "#" ASMALIGN(4) \
> + "# .p2align 4 \n\t"\
> "2: \n\t"
> Z_COND_IDCT( 96(%0),104(%0),112(%0),120(%0), 96(%1),paddd (%2), 11, 3f)
>
> @@ -1007,7 +1007,7 @@ IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0), 20)
> IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0), 20)
> "jmp 9f \n\t"
>
> - "#" ASMALIGN(4) \
> + "# .p2align 4 \n\t"\
> "3: \n\t"
> #undef IDCT
> #define IDCT(src0, src4, src1, src5, dst, shift) \
> @@ -1071,7 +1071,7 @@ IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0), 20)
> IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0), 20)
> "jmp 9f \n\t"
>
> - "#" ASMALIGN(4) \
> + "# .p2align 4 \n\t"\
> "5: \n\t"
> #undef IDCT
> #define IDCT(src0, src4, src1, src5, dst, shift) \
> @@ -1136,7 +1136,7 @@ IDCT( 16(%1), 80(%1), 48(%1), 112(%1), 8(%0), 20)
> "jmp 9f \n\t"
>
>
> - "#" ASMALIGN(4) \
> + "# .p2align 4 \n\t"\
> "1: \n\t"
> #undef IDCT
> #define IDCT(src0, src4, src1, src5, dst, shift) \
> @@ -1210,7 +1210,7 @@ IDCT( 24(%1), 88(%1), 56(%1), 120(%1), 12(%0), 20)
> "jmp 9f \n\t"
>
>
> - "#" ASMALIGN(4)
> + "# .p2align 4 \n\t"
> "7: \n\t"
> #undef IDCT
> #define IDCT(src0, src4, src1, src5, dst, shift) \
> diff --git a/libavcodec/x86/vc1dsp_mmx.c b/libavcodec/x86/vc1dsp_mmx.c
> index e0b1f5b..092976f 100644
> --- a/libavcodec/x86/vc1dsp_mmx.c
> +++ b/libavcodec/x86/vc1dsp_mmx.c
> @@ -283,7 +283,7 @@ vc1_put_ver_16b_ ## NAME ## _mmx(int16_t *dst, const uint8_t *src, \
> LOAD_ROUNDER_MMX("%5") \
> "movq "MANGLE(ff_pw_53)", %%mm5\n\t" \
> "movq "MANGLE(ff_pw_18)", %%mm6\n\t" \
> - ASMALIGN(3) \
> + ".p2align 3 \n\t" \
> "1: \n\t" \
> MSPEL_FILTER13_CORE(DO_UNPACK, "movd 1", A1, A2, A3, A4) \
> NORMALIZE_MMX("%6") \
> @@ -339,7 +339,7 @@ OPNAME ## vc1_hor_16b_ ## NAME ## _mmx(uint8_t *dst, x86_reg stride, \
> LOAD_ROUNDER_MMX("%4") \
> "movq "MANGLE(ff_pw_18)", %%mm6 \n\t" \
> "movq "MANGLE(ff_pw_53)", %%mm5 \n\t" \
> - ASMALIGN(3) \
> + ".p2align 3 \n\t" \
> "1: \n\t" \
> MSPEL_FILTER13_CORE(DONT_UNPACK, "movq 2", A1, A2, A3, A4) \
> NORMALIZE_MMX("$7") \
> @@ -377,7 +377,7 @@ OPNAME ## vc1_## NAME ## _mmx(uint8_t *dst, const uint8_t *src, \
> LOAD_ROUNDER_MMX("%6") \
> "movq "MANGLE(ff_pw_53)", %%mm5 \n\t" \
> "movq "MANGLE(ff_pw_18)", %%mm6 \n\t" \
> - ASMALIGN(3) \
> + ".p2align 3 \n\t" \
> "1: \n\t" \
> MSPEL_FILTER13_CORE(DO_UNPACK, "movd 1", A1, A2, A3, A4) \
> NORMALIZE_MMX("$6") \
> --
> 1.7.1
>
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel at mplayerhq.hu
> https://lists.mplayerhq.hu/mailman/listinfo/ffmpeg-devel
More information about the ffmpeg-devel
mailing list