1
0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-11-21 10:55:51 +02:00

swscale: Commits that could not be pulled earlier due to bugs #2

commit 5a5a0f1613
Author: Diego Biurrun <diego@biurrun.de>
Date:   Fri May 27 19:46:39 2011 +0200

    swscale: Remove unused variables in x86 code.

    libswscale/x86/swscale_template.c:2072: warning: unused variable ‘canMMX2BeUsed’
    libswscale/x86/swscale_template.c:2145: warning: unused variable ‘canMMX2BeUsed’
    libswscale/x86/swscale_template.c:2209: warning: unused variable ‘chrVPixBuf’
    libswscale/x86/swscale_template.c:2237: warning: unused variable ‘chrVSrcPtr’

commit 389e2000eb
Author: Ronald S. Bultje <rsbultje@gmail.com>
Date:   Fri May 27 12:23:32 2011 -0400

    swscale: delay allocation of formatConvBuffer().

    That means it won't be allocated when not needed. Alongside
    this, it fixes valgrind/fate-detected memory leaks.

commit f327bfa6dc
Author: Ronald S. Bultje <rsbultje@gmail.com>
Date:   Fri May 27 11:36:43 2011 -0400

    swscale: fix build with --disable-swscale-alpha.

commit 9f5d45025e
Author: Ronald S. Bultje <rsbultje@gmail.com>
Date:   Fri May 27 09:28:38 2011 -0400

    swscale: fix non-bitexact yuv2yuv[X2]() MMX/MMX2 functions.
This commit is contained in:
Michael Niedermayer 2011-05-27 23:50:06 +02:00
parent 986f0d86cb
commit 39d607e5bb
3 changed files with 11 additions and 13 deletions

View File

@ -795,7 +795,6 @@ int sws_init_context(SwsContext *c, SwsFilter *srcFilter, SwsFilter *dstFilter)
srcW, srcH, dstW, dstH); srcW, srcH, dstW, dstH);
return AVERROR(EINVAL); return AVERROR(EINVAL);
} }
FF_ALLOC_OR_GOTO(c, c->formatConvBuffer, FFALIGN(srcW*2+78, 16) * 2, fail);
if (!dstFilter) dstFilter= &dummyFilter; if (!dstFilter) dstFilter= &dummyFilter;
if (!srcFilter) srcFilter= &dummyFilter; if (!srcFilter) srcFilter= &dummyFilter;
@ -851,6 +850,7 @@ int sws_init_context(SwsContext *c, SwsFilter *srcFilter, SwsFilter *dstFilter)
} }
} }
FF_ALLOC_OR_GOTO(c, c->formatConvBuffer, FFALIGN(srcW*2+78, 16) * 2, fail);
if (HAVE_MMX2 && cpu_flags & AV_CPU_FLAG_MMX2) { if (HAVE_MMX2 && cpu_flags & AV_CPU_FLAG_MMX2) {
c->canMMX2BeUsed= (dstW >=srcW && (dstW&31)==0 && (srcW&15)==0) ? 1 : 0; c->canMMX2BeUsed= (dstW >=srcW && (dstW&31)==0 && (srcW&15)==0) ? 1 : 0;
if (!c->canMMX2BeUsed && dstW >=srcW && (srcW&15)==0 && (flags&SWS_FAST_BILINEAR)) { if (!c->canMMX2BeUsed && dstW >=srcW && (srcW&15)==0 && (flags&SWS_FAST_BILINEAR)) {

View File

@ -59,7 +59,7 @@
"psraw $3, %%mm3 \n\t"\ "psraw $3, %%mm3 \n\t"\
"psraw $3, %%mm4 \n\t"\ "psraw $3, %%mm4 \n\t"\
"packuswb %%mm4, %%mm3 \n\t"\ "packuswb %%mm4, %%mm3 \n\t"\
MOVNTQ(%%mm3, (%1, %%REGa))\ MOVNTQ(%%mm3, (%1, %3))\
"add $8, %3 \n\t"\ "add $8, %3 \n\t"\
"cmp %2, %3 \n\t"\ "cmp %2, %3 \n\t"\
"movq "VROUNDER_OFFSET"(%0), %%mm3 \n\t"\ "movq "VROUNDER_OFFSET"(%0), %%mm3 \n\t"\
@ -81,8 +81,9 @@ static inline void RENAME(yuv2yuvX)(SwsContext *c, const int16_t *lumFilter,
uint8_t *aDest, long dstW, long chrDstW) uint8_t *aDest, long dstW, long chrDstW)
{ {
if (uDest) { if (uDest) {
x86_reg uv_off = c->uv_off;
YSCALEYUV2YV12X(CHR_MMX_FILTER_OFFSET, uDest, chrDstW, 0) YSCALEYUV2YV12X(CHR_MMX_FILTER_OFFSET, uDest, chrDstW, 0)
YSCALEYUV2YV12X(CHR_MMX_FILTER_OFFSET, vDest, chrDstW + c->uv_off, c->uv_off) YSCALEYUV2YV12X(CHR_MMX_FILTER_OFFSET, vDest - uv_off, chrDstW + uv_off, uv_off)
} }
if (CONFIG_SWSCALE_ALPHA && aDest) { if (CONFIG_SWSCALE_ALPHA && aDest) {
YSCALEYUV2YV12X(ALP_MMX_FILTER_OFFSET, aDest, dstW, 0) YSCALEYUV2YV12X(ALP_MMX_FILTER_OFFSET, aDest, dstW, 0)
@ -137,7 +138,7 @@ static inline void RENAME(yuv2yuvX)(SwsContext *c, const int16_t *lumFilter,
"psraw $3, %%mm4 \n\t"\ "psraw $3, %%mm4 \n\t"\
"psraw $3, %%mm6 \n\t"\ "psraw $3, %%mm6 \n\t"\
"packuswb %%mm6, %%mm4 \n\t"\ "packuswb %%mm6, %%mm4 \n\t"\
MOVNTQ(%%mm4, (%1, %%REGa))\ MOVNTQ(%%mm4, (%1, %3))\
"add $8, %3 \n\t"\ "add $8, %3 \n\t"\
"cmp %2, %3 \n\t"\ "cmp %2, %3 \n\t"\
"lea " offset "(%0), %%"REG_d" \n\t"\ "lea " offset "(%0), %%"REG_d" \n\t"\
@ -161,8 +162,9 @@ static inline void RENAME(yuv2yuvX_ar)(SwsContext *c, const int16_t *lumFilter,
uint8_t *aDest, long dstW, long chrDstW) uint8_t *aDest, long dstW, long chrDstW)
{ {
if (uDest) { if (uDest) {
x86_reg uv_off = c->uv_off;
YSCALEYUV2YV12X_ACCURATE(CHR_MMX_FILTER_OFFSET, uDest, chrDstW, 0) YSCALEYUV2YV12X_ACCURATE(CHR_MMX_FILTER_OFFSET, uDest, chrDstW, 0)
YSCALEYUV2YV12X_ACCURATE(CHR_MMX_FILTER_OFFSET, vDest, chrDstW + c->uv_off, c->uv_off) YSCALEYUV2YV12X_ACCURATE(CHR_MMX_FILTER_OFFSET, vDest - uv_off, chrDstW + uv_off, uv_off)
} }
if (CONFIG_SWSCALE_ALPHA && aDest) { if (CONFIG_SWSCALE_ALPHA && aDest) {
YSCALEYUV2YV12X_ACCURATE(ALP_MMX_FILTER_OFFSET, aDest, dstW, 0) YSCALEYUV2YV12X_ACCURATE(ALP_MMX_FILTER_OFFSET, aDest, dstW, 0)
@ -2223,7 +2225,6 @@ static inline void RENAME(hyscale_fast)(SwsContext *c, int16_t *dst,
{ {
int32_t *filterPos = c->hLumFilterPos; int32_t *filterPos = c->hLumFilterPos;
int16_t *filter = c->hLumFilter; int16_t *filter = c->hLumFilter;
int canMMX2BeUsed = c->canMMX2BeUsed;
void *mmx2FilterCode= c->lumMmx2FilterCode; void *mmx2FilterCode= c->lumMmx2FilterCode;
int i; int i;
#if defined(PIC) #if defined(PIC)
@ -2296,7 +2297,6 @@ static inline void RENAME(hcscale_fast)(SwsContext *c, int16_t *dst1, int16_t *d
{ {
int32_t *filterPos = c->hChrFilterPos; int32_t *filterPos = c->hChrFilterPos;
int16_t *filter = c->hChrFilter; int16_t *filter = c->hChrFilter;
int canMMX2BeUsed = c->canMMX2BeUsed;
void *mmx2FilterCode= c->chrMmx2FilterCode; void *mmx2FilterCode= c->chrMmx2FilterCode;
int i; int i;
#if defined(PIC) #if defined(PIC)
@ -2362,7 +2362,6 @@ static void updateMMXDitherTables(SwsContext *c, int dstY, int lumBufIndex, int
const int flags= c->flags; const int flags= c->flags;
int16_t **lumPixBuf= c->lumPixBuf; int16_t **lumPixBuf= c->lumPixBuf;
int16_t **chrUPixBuf= c->chrUPixBuf; int16_t **chrUPixBuf= c->chrUPixBuf;
int16_t **chrVPixBuf= c->chrVPixBuf;
int16_t **alpPixBuf= c->alpPixBuf; int16_t **alpPixBuf= c->alpPixBuf;
const int vLumBufSize= c->vLumBufSize; const int vLumBufSize= c->vLumBufSize;
const int vChrBufSize= c->vChrBufSize; const int vChrBufSize= c->vChrBufSize;
@ -2388,7 +2387,6 @@ static void updateMMXDitherTables(SwsContext *c, int dstY, int lumBufIndex, int
if (dstY < dstH - 2) { if (dstY < dstH - 2) {
const int16_t **lumSrcPtr= (const int16_t **) lumPixBuf + lumBufIndex + firstLumSrcY - lastInLumBuf + vLumBufSize; const int16_t **lumSrcPtr= (const int16_t **) lumPixBuf + lumBufIndex + firstLumSrcY - lastInLumBuf + vLumBufSize;
const int16_t **chrUSrcPtr= (const int16_t **) chrUPixBuf + chrBufIndex + firstChrSrcY - lastInChrBuf + vChrBufSize; const int16_t **chrUSrcPtr= (const int16_t **) chrUPixBuf + chrBufIndex + firstChrSrcY - lastInChrBuf + vChrBufSize;
const int16_t **chrVSrcPtr= (const int16_t **) chrVPixBuf + chrBufIndex + firstChrSrcY - lastInChrBuf + vChrBufSize;
const int16_t **alpSrcPtr= (CONFIG_SWSCALE_ALPHA && alpPixBuf) ? (const int16_t **) alpPixBuf + lumBufIndex + firstLumSrcY - lastInLumBuf + vLumBufSize : NULL; const int16_t **alpSrcPtr= (CONFIG_SWSCALE_ALPHA && alpPixBuf) ? (const int16_t **) alpPixBuf + lumBufIndex + firstLumSrcY - lastInLumBuf + vLumBufSize : NULL;
int i; int i;
if (flags & SWS_ACCURATE_RND) { if (flags & SWS_ACCURATE_RND) {

View File

@ -82,15 +82,15 @@ SwsFunc ff_yuv2rgb_init_mmx(SwsContext *c)
if (HAVE_MMX && cpu_flags & AV_CPU_FLAG_MMX) { if (HAVE_MMX && cpu_flags & AV_CPU_FLAG_MMX) {
switch (c->dstFormat) { switch (c->dstFormat) {
case PIX_FMT_RGB32: case PIX_FMT_RGB32:
if (CONFIG_SWSCALE_ALPHA && c->srcFormat == PIX_FMT_YUVA420P) { if (c->srcFormat == PIX_FMT_YUVA420P) {
#if HAVE_7REGS #if HAVE_7REGS && CONFIG_SWSCALE_ALPHA
return yuva420_rgb32_MMX; return yuva420_rgb32_MMX;
#endif #endif
break; break;
} else return yuv420_rgb32_MMX; } else return yuv420_rgb32_MMX;
case PIX_FMT_BGR32: case PIX_FMT_BGR32:
if (CONFIG_SWSCALE_ALPHA && c->srcFormat == PIX_FMT_YUVA420P) { if (c->srcFormat == PIX_FMT_YUVA420P) {
#if HAVE_7REGS #if HAVE_7REGS && CONFIG_SWSCALE_ALPHA
return yuva420_bgr32_MMX; return yuva420_bgr32_MMX;
#endif #endif
break; break;