diff options
| author | Michael Niedermayer <michaelni@gmx.at> | 2013-04-11 11:53:19 +0200 |
|---|---|---|
| committer | Michael Niedermayer <michaelni@gmx.at> | 2013-04-11 11:53:19 +0200 |
| commit | 0724b4a16d6b88b311610e96c7ba341a4b33965d (patch) | |
| tree | 4e80fea21e5c1dd58bb706cecb95283f81e8846f | |
| parent | 944ad46182483ff52573ec3c827c86058b225d16 (diff) | |
| parent | 62844c3fd66940c7747e9b2bb7804e265319f43f (diff) | |
Merge commit '62844c3fd66940c7747e9b2bb7804e265319f43f'
* commit '62844c3fd66940c7747e9b2bb7804e265319f43f':
h264: Integrate clear_blocks calls with IDCT
Conflicts:
libavcodec/arm/h264idct_neon.S
libavcodec/h264idct_template.c
libavcodec/x86/h264_idct.asm
Merged-by: Michael Niedermayer <michaelni@gmx.at>
| -rw-r--r-- | libavcodec/arm/h264idct_neon.S | 18 | ||||
| -rw-r--r-- | libavcodec/h264idct_template.c | 4 | ||||
| -rw-r--r-- | libavcodec/x86/h264_idct.asm | 8 |
3 files changed, 15 insertions, 15 deletions
diff --git a/libavcodec/arm/h264idct_neon.S b/libavcodec/arm/h264idct_neon.S index fa5b90c23a..2edeca2cbc 100644 --- a/libavcodec/arm/h264idct_neon.S +++ b/libavcodec/arm/h264idct_neon.S @@ -187,8 +187,8 @@ endfunc vshr.s16 q2, q10, #1 vadd.i16 q0, q8, q12 vld1.16 {q14-q15},[r1,:128] - vst1.16 {q7}, [r1,:128]! - vst1.16 {q7}, [r1,:128]! + vst1.16 {q3}, [r1,:128]! + vst1.16 {q3}, [r1,:128]! vsub.i16 q1, q8, q12 vshr.s16 q3, q14, #1 vsub.i16 q2, q2, q14 @@ -267,16 +267,16 @@ endfunc .endm function ff_h264_idct8_add_neon, export=1 - vmov.i16 q7, #0 + vmov.i16 q3, #0 vld1.16 {q8-q9}, [r1,:128] - vst1.16 {q7}, [r1,:128]! - vst1.16 {q7}, [r1,:128]! + vst1.16 {q3}, [r1,:128]! + vst1.16 {q3}, [r1,:128]! vld1.16 {q10-q11},[r1,:128] - vst1.16 {q7}, [r1,:128]! - vst1.16 {q7}, [r1,:128]! + vst1.16 {q3}, [r1,:128]! + vst1.16 {q3}, [r1,:128]! vld1.16 {q12-q13},[r1,:128] - vst1.16 {q7}, [r1,:128]! - vst1.16 {q7}, [r1,:128]! + vst1.16 {q3}, [r1,:128]! + vst1.16 {q3}, [r1,:128]! idct8x8_cols 0 idct8x8_cols 1 diff --git a/libavcodec/h264idct_template.c b/libavcodec/h264idct_template.c index 9f16e1d2e0..42c32d15ed 100644 --- a/libavcodec/h264idct_template.c +++ b/libavcodec/h264idct_template.c @@ -145,7 +145,7 @@ void FUNCC(ff_h264_idct_dc_add)(uint8_t *_dst, int16_t *_block, int stride){ pixel *dst = (pixel*)_dst; dctcoef *block = (dctcoef*)_block; int dc = (block[0] + 32) >> 6; - stride >>= sizeof(pixel)-1; + stride /= sizeof(pixel); block[0] = 0; for( j = 0; j < 4; j++ ) { @@ -161,7 +161,7 @@ void FUNCC(ff_h264_idct8_dc_add)(uint8_t *_dst, int16_t *_block, int stride){ dctcoef *block = (dctcoef*)_block; int dc = (block[0] + 32) >> 6; block[0] = 0; - stride >>= sizeof(pixel)-1; + stride /= sizeof(pixel); for( j = 0; j < 8; j++ ) { for( i = 0; i < 8; i++ ) diff --git a/libavcodec/x86/h264_idct.asm b/libavcodec/x86/h264_idct.asm index 7bb1653428..17bf794bce 100644 --- a/libavcodec/x86/h264_idct.asm +++ b/libavcodec/x86/h264_idct.asm @@ -312,7 +312,7 @@ INIT_MMX mmxext %if ARCH_X86_64 cglobal h264_idct_dc_add_8, 3, 4, 0 movsx r3, word [r1] - mov word [r1], 0 + mov dword [r1], 0 DC_ADD_MMXEXT_INIT r3, r2 DC_ADD_MMXEXT_OP movh, r0, r2, r3 RET @@ -320,7 +320,7 @@ cglobal h264_idct_dc_add_8, 3, 4, 0 ; ff_h264_idct8_dc_add_mmxext(uint8_t *dst, int16_t *block, int stride) cglobal h264_idct8_dc_add_8, 3, 4, 0 movsx r3, word [r1] - mov word [r1], 0 + mov dword [r1], 0 DC_ADD_MMXEXT_INIT r3, r2 DC_ADD_MMXEXT_OP mova, r0, r2, r3 lea r0, [r0+r2*4] @@ -329,7 +329,7 @@ cglobal h264_idct8_dc_add_8, 3, 4, 0 %else cglobal h264_idct_dc_add_8, 2, 3, 0 movsx r2, word [r1] - mov word [r1], 0 + mov dword [r1], 0 mov r1, r2m DC_ADD_MMXEXT_INIT r2, r1 DC_ADD_MMXEXT_OP movh, r0, r1, r2 @@ -338,7 +338,7 @@ cglobal h264_idct_dc_add_8, 2, 3, 0 ; ff_h264_idct8_dc_add_mmxext(uint8_t *dst, int16_t *block, int stride) cglobal h264_idct8_dc_add_8, 2, 3, 0 movsx r2, word [r1] - mov word [r1], 0 + mov dword [r1], 0 mov r1, r2m DC_ADD_MMXEXT_INIT r2, r1 DC_ADD_MMXEXT_OP mova, r0, r1, r2 |