[FFmpeg-cvslog] swscale: remove if(canMMX2BeUsed) conditional.
Ronald S. Bultje
git at videolan.org
Wed May 25 06:37:13 CEST 2011
ffmpeg | branch: master | Ronald S. Bultje <rsbultje at gmail.com> | Tue May 24 12:15:14 2011 -0400| [566b5fbbb3753f71b60c4da491236a62232a8cf9] | committer: Ronald S. Bultje
swscale: remove if(canMMX2BeUsed) conditional.
Instead, set function pointers conditionally during init. This
patch also reveals a whole branch of dead assembly code that is
therefore also removed.
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=566b5fbbb3753f71b60c4da491236a62232a8cf9
---
libswscale/x86/swscale_template.c | 99 +++----------------------------------
1 files changed, 8 insertions(+), 91 deletions(-)
diff --git a/libswscale/x86/swscale_template.c b/libswscale/x86/swscale_template.c
index 37a6bb5..0bee76e 100644
--- a/libswscale/x86/swscale_template.c
+++ b/libswscale/x86/swscale_template.c
@@ -1918,19 +1918,11 @@ static inline void RENAME(hScale)(int16_t *dst, int dstW, const uint8_t *src, in
}
}
-#define FAST_BILINEAR_X86 \
- "subl %%edi, %%esi \n\t" /* src[xx+1] - src[xx] */ \
- "imull %%ecx, %%esi \n\t" /* (src[xx+1] - src[xx])*xalpha */ \
- "shll $16, %%edi \n\t" \
- "addl %%edi, %%esi \n\t" /* src[xx+1]*xalpha + src[xx]*(1-xalpha) */ \
- "mov %1, %%"REG_D"\n\t" \
- "shrl $9, %%esi \n\t" \
-
+#if COMPILE_TEMPLATE_MMX2
static inline void RENAME(hyscale_fast)(SwsContext *c, int16_t *dst,
long dstWidth, const uint8_t *src, int srcW,
int xInc)
{
-#if COMPILE_TEMPLATE_MMX2
int32_t *filterPos = c->hLumFilterPos;
int16_t *filter = c->hLumFilter;
int canMMX2BeUsed = c->canMMX2BeUsed;
@@ -1939,7 +1931,7 @@ static inline void RENAME(hyscale_fast)(SwsContext *c, int16_t *dst,
#if defined(PIC)
DECLARE_ALIGNED(8, uint64_t, ebxsave);
#endif
- if (canMMX2BeUsed) {
+
__asm__ volatile(
#if defined(PIC)
"mov %%"REG_b", %5 \n\t"
@@ -1998,51 +1990,12 @@ static inline void RENAME(hyscale_fast)(SwsContext *c, int16_t *dst,
#endif
);
for (i=dstWidth-1; (i*xInc)>>16 >=srcW-1; i--) dst[i] = src[srcW-1]*128;
- } else {
-#endif /* COMPILE_TEMPLATE_MMX2 */
- x86_reg xInc_shr16 = xInc >> 16;
- uint16_t xInc_mask = xInc & 0xffff;
- x86_reg dstWidth_reg = dstWidth;
- //NO MMX just normal asm ...
- __asm__ volatile(
- "xor %%"REG_a", %%"REG_a" \n\t" // i
- "xor %%"REG_d", %%"REG_d" \n\t" // xx
- "xorl %%ecx, %%ecx \n\t" // xalpha
- ".p2align 4 \n\t"
- "1: \n\t"
- "movzbl (%0, %%"REG_d"), %%edi \n\t" //src[xx]
- "movzbl 1(%0, %%"REG_d"), %%esi \n\t" //src[xx+1]
- FAST_BILINEAR_X86
- "movw %%si, (%%"REG_D", %%"REG_a", 2) \n\t"
- "addw %4, %%cx \n\t" //xalpha += xInc&0xFFFF
- "adc %3, %%"REG_d" \n\t" //xx+= xInc>>16 + carry
-
- "movzbl (%0, %%"REG_d"), %%edi \n\t" //src[xx]
- "movzbl 1(%0, %%"REG_d"), %%esi \n\t" //src[xx+1]
- FAST_BILINEAR_X86
- "movw %%si, 2(%%"REG_D", %%"REG_a", 2) \n\t"
- "addw %4, %%cx \n\t" //xalpha += xInc&0xFFFF
- "adc %3, %%"REG_d" \n\t" //xx+= xInc>>16 + carry
-
-
- "add $2, %%"REG_a" \n\t"
- "cmp %2, %%"REG_a" \n\t"
- " jb 1b \n\t"
-
-
- :: "r" (src), "m" (dst), "m" (dstWidth_reg), "m" (xInc_shr16), "m" (xInc_mask)
- : "%"REG_a, "%"REG_d, "%ecx", "%"REG_D, "%esi"
- );
-#if COMPILE_TEMPLATE_MMX2
- } //if MMX2 can't be used
-#endif
}
static inline void RENAME(hcscale_fast)(SwsContext *c, int16_t *dst,
long dstWidth, const uint8_t *src1,
const uint8_t *src2, int srcW, int xInc)
{
-#if COMPILE_TEMPLATE_MMX2
int32_t *filterPos = c->hChrFilterPos;
int16_t *filter = c->hChrFilter;
int canMMX2BeUsed = c->canMMX2BeUsed;
@@ -2051,7 +2004,7 @@ static inline void RENAME(hcscale_fast)(SwsContext *c, int16_t *dst,
#if defined(PIC)
DECLARE_ALIGNED(8, uint64_t, ebxsave);
#endif
- if (canMMX2BeUsed) {
+
__asm__ volatile(
#if defined(PIC)
"mov %%"REG_b", %6 \n\t"
@@ -2101,48 +2054,8 @@ static inline void RENAME(hcscale_fast)(SwsContext *c, int16_t *dst,
dst[i] = src1[srcW-1]*128;
dst[i+VOFW] = src2[srcW-1]*128;
}
- } else {
-#endif /* COMPILE_TEMPLATE_MMX2 */
- x86_reg xInc_shr16 = (x86_reg) (xInc >> 16);
- uint16_t xInc_mask = xInc & 0xffff;
- x86_reg dstWidth_reg = dstWidth;
- __asm__ volatile(
- "xor %%"REG_a", %%"REG_a" \n\t" // i
- "xor %%"REG_d", %%"REG_d" \n\t" // xx
- "xorl %%ecx, %%ecx \n\t" // xalpha
- ".p2align 4 \n\t"
- "1: \n\t"
- "mov %0, %%"REG_S" \n\t"
- "movzbl (%%"REG_S", %%"REG_d"), %%edi \n\t" //src[xx]
- "movzbl 1(%%"REG_S", %%"REG_d"), %%esi \n\t" //src[xx+1]
- FAST_BILINEAR_X86
- "movw %%si, (%%"REG_D", %%"REG_a", 2) \n\t"
-
- "movzbl (%5, %%"REG_d"), %%edi \n\t" //src[xx]
- "movzbl 1(%5, %%"REG_d"), %%esi \n\t" //src[xx+1]
- FAST_BILINEAR_X86
- "movw %%si, "AV_STRINGIFY(VOF)"(%%"REG_D", %%"REG_a", 2) \n\t"
-
- "addw %4, %%cx \n\t" //xalpha += xInc&0xFFFF
- "adc %3, %%"REG_d" \n\t" //xx+= xInc>>16 + carry
- "add $1, %%"REG_a" \n\t"
- "cmp %2, %%"REG_a" \n\t"
- " jb 1b \n\t"
-
-/* GCC 3.3 makes MPlayer crash on IA-32 machines when using "g" operand here,
-which is needed to support GCC 4.0. */
-#if ARCH_X86_64 && AV_GCC_VERSION_AT_LEAST(3,4)
- :: "m" (src1), "m" (dst), "g" (dstWidth_reg), "m" (xInc_shr16), "m" (xInc_mask),
-#else
- :: "m" (src1), "m" (dst), "m" (dstWidth_reg), "m" (xInc_shr16), "m" (xInc_mask),
-#endif
- "r" (src2)
- : "%"REG_a, "%"REG_d, "%ecx", "%"REG_D, "%esi"
- );
-#if COMPILE_TEMPLATE_MMX2
- } //if MMX2 can't be used
-#endif
}
+#endif /* COMPILE_TEMPLATE_MMX2 */
#if !COMPILE_TEMPLATE_MMX2
static void updateMMXDitherTables(SwsContext *c, int dstY, int lumBufIndex, int chrBufIndex,
@@ -2240,14 +2153,18 @@ static void RENAME(sws_init_swScale)(SwsContext *c)
c->hScale = RENAME(hScale );
// Use the new MMX scaler if the MMX2 one can't be used (it is faster than the x86 ASM one).
+#if COMPILE_TEMPLATE_MMX2
if (c->flags & SWS_FAST_BILINEAR && c->canMMX2BeUsed)
{
c->hyscale_fast = RENAME(hyscale_fast);
c->hcscale_fast = RENAME(hcscale_fast);
} else {
+#endif /* COMPILE_TEMPLATE_MMX2 */
c->hyscale_fast = NULL;
c->hcscale_fast = NULL;
+#if COMPILE_TEMPLATE_MMX2
}
+#endif /* COMPILE_TEMPLATE_MMX2 */
switch(srcFormat) {
case PIX_FMT_YUYV422 : c->chrToYV12 = RENAME(yuy2ToUV); break;
More information about the ffmpeg-cvslog
mailing list