1
0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2025-08-04 22:03:09 +02:00

swscale/swscale_unscaled: account for semi planar formats with data in the msb

Fixes fate failures introduced by recent tests that exercise the faulty code.

Signed-off-by: James Almer <jamrial@gmail.com>
This commit is contained in:
James Almer
2025-03-14 11:25:20 -03:00
parent 26c5d8cf5d
commit 819dec697a
11 changed files with 48 additions and 46 deletions

View File

@ -2146,17 +2146,17 @@ static int packedCopyWrapper(SwsInternal *c, const uint8_t *const src[],
if (c->opts.dither == SWS_DITHER_NONE) {\
for (i = 0; i < height; i++) {\
for (j = 0; j < length-7; j+=8) {\
tmp = (bswap(src[j+0]) + bias)>>shift; dst[j+0] = dbswap(tmp - (tmp>>dst_depth));\
tmp = (bswap(src[j+1]) + bias)>>shift; dst[j+1] = dbswap(tmp - (tmp>>dst_depth));\
tmp = (bswap(src[j+2]) + bias)>>shift; dst[j+2] = dbswap(tmp - (tmp>>dst_depth));\
tmp = (bswap(src[j+3]) + bias)>>shift; dst[j+3] = dbswap(tmp - (tmp>>dst_depth));\
tmp = (bswap(src[j+4]) + bias)>>shift; dst[j+4] = dbswap(tmp - (tmp>>dst_depth));\
tmp = (bswap(src[j+5]) + bias)>>shift; dst[j+5] = dbswap(tmp - (tmp>>dst_depth));\
tmp = (bswap(src[j+6]) + bias)>>shift; dst[j+6] = dbswap(tmp - (tmp>>dst_depth));\
tmp = (bswap(src[j+7]) + bias)>>shift; dst[j+7] = dbswap(tmp - (tmp>>dst_depth));\
tmp = ((bswap(src[j+0]) >> src_shift) + bias)>>shift; dst[j+0] = dbswap((tmp - (tmp>>dst_depth)) << dst_shift);\
tmp = ((bswap(src[j+1]) >> src_shift) + bias)>>shift; dst[j+1] = dbswap((tmp - (tmp>>dst_depth)) << dst_shift);\
tmp = ((bswap(src[j+2]) >> src_shift) + bias)>>shift; dst[j+2] = dbswap((tmp - (tmp>>dst_depth)) << dst_shift);\
tmp = ((bswap(src[j+3]) >> src_shift) + bias)>>shift; dst[j+3] = dbswap((tmp - (tmp>>dst_depth)) << dst_shift);\
tmp = ((bswap(src[j+4]) >> src_shift) + bias)>>shift; dst[j+4] = dbswap((tmp - (tmp>>dst_depth)) << dst_shift);\
tmp = ((bswap(src[j+5]) >> src_shift) + bias)>>shift; dst[j+5] = dbswap((tmp - (tmp>>dst_depth)) << dst_shift);\
tmp = ((bswap(src[j+6]) >> src_shift) + bias)>>shift; dst[j+6] = dbswap((tmp - (tmp>>dst_depth)) << dst_shift);\
tmp = ((bswap(src[j+7]) >> src_shift) + bias)>>shift; dst[j+7] = dbswap((tmp - (tmp>>dst_depth)) << dst_shift);\
}\
for (; j < length; j++) {\
tmp = (bswap(src[j]) + bias)>>shift; dst[j] = dbswap(tmp - (tmp>>dst_depth));\
tmp = (bswap(src[j]) + bias)>>shift; dst[j] = dbswap(tmp - (tmp>>dst_depth) << dst_shift);\
}\
dst += dstStride;\
src += srcStride;\
@ -2165,17 +2165,17 @@ static int packedCopyWrapper(SwsInternal *c, const uint8_t *const src[],
for (i = 0; i < height; i++) {\
const uint8_t *dither= dithers[shift-1][i&7];\
for (j = 0; j < length-7; j+=8) {\
tmp = (bswap(src[j+0]) + dither[0])>>shift; dst[j+0] = dbswap(tmp - (tmp>>dst_depth));\
tmp = (bswap(src[j+1]) + dither[1])>>shift; dst[j+1] = dbswap(tmp - (tmp>>dst_depth));\
tmp = (bswap(src[j+2]) + dither[2])>>shift; dst[j+2] = dbswap(tmp - (tmp>>dst_depth));\
tmp = (bswap(src[j+3]) + dither[3])>>shift; dst[j+3] = dbswap(tmp - (tmp>>dst_depth));\
tmp = (bswap(src[j+4]) + dither[4])>>shift; dst[j+4] = dbswap(tmp - (tmp>>dst_depth));\
tmp = (bswap(src[j+5]) + dither[5])>>shift; dst[j+5] = dbswap(tmp - (tmp>>dst_depth));\
tmp = (bswap(src[j+6]) + dither[6])>>shift; dst[j+6] = dbswap(tmp - (tmp>>dst_depth));\
tmp = (bswap(src[j+7]) + dither[7])>>shift; dst[j+7] = dbswap(tmp - (tmp>>dst_depth));\
tmp = ((bswap(src[j+0]) >> src_shift) + dither[0])>>shift; dst[j+0] = dbswap((tmp - (tmp>>dst_depth)) << dst_shift);\
tmp = ((bswap(src[j+1]) >> src_shift) + dither[1])>>shift; dst[j+1] = dbswap((tmp - (tmp>>dst_depth)) << dst_shift);\
tmp = ((bswap(src[j+2]) >> src_shift) + dither[2])>>shift; dst[j+2] = dbswap((tmp - (tmp>>dst_depth)) << dst_shift);\
tmp = ((bswap(src[j+3]) >> src_shift) + dither[3])>>shift; dst[j+3] = dbswap((tmp - (tmp>>dst_depth)) << dst_shift);\
tmp = ((bswap(src[j+4]) >> src_shift) + dither[4])>>shift; dst[j+4] = dbswap((tmp - (tmp>>dst_depth)) << dst_shift);\
tmp = ((bswap(src[j+5]) >> src_shift) + dither[5])>>shift; dst[j+5] = dbswap((tmp - (tmp>>dst_depth)) << dst_shift);\
tmp = ((bswap(src[j+6]) >> src_shift) + dither[6])>>shift; dst[j+6] = dbswap((tmp - (tmp>>dst_depth)) << dst_shift);\
tmp = ((bswap(src[j+7]) >> src_shift) + dither[7])>>shift; dst[j+7] = dbswap((tmp - (tmp>>dst_depth)) << dst_shift);\
}\
for (; j < length; j++) {\
tmp = (bswap(src[j]) + dither[j&7])>>shift; dst[j] = dbswap(tmp - (tmp>>dst_depth));\
tmp = (bswap(src[j]) + dither[j&7])>>shift; dst[j] = dbswap(tmp - (tmp>>dst_depth) << dst_shift);\
}\
dst += dstStride;\
src += srcStride;\
@ -2184,14 +2184,14 @@ static int packedCopyWrapper(SwsInternal *c, const uint8_t *const src[],
for (i = 0; i < height; i++) {\
const uint8_t *dither= dithers[shift-1][i&7];\
for (j = 0; j < length-7; j+=8) {\
tmp = bswap(src[j+0]); dst[j+0] = dbswap((tmp - (tmp>>dst_depth) + dither[0])>>shift);\
tmp = bswap(src[j+1]); dst[j+1] = dbswap((tmp - (tmp>>dst_depth) + dither[1])>>shift);\
tmp = bswap(src[j+2]); dst[j+2] = dbswap((tmp - (tmp>>dst_depth) + dither[2])>>shift);\
tmp = bswap(src[j+3]); dst[j+3] = dbswap((tmp - (tmp>>dst_depth) + dither[3])>>shift);\
tmp = bswap(src[j+4]); dst[j+4] = dbswap((tmp - (tmp>>dst_depth) + dither[4])>>shift);\
tmp = bswap(src[j+5]); dst[j+5] = dbswap((tmp - (tmp>>dst_depth) + dither[5])>>shift);\
tmp = bswap(src[j+6]); dst[j+6] = dbswap((tmp - (tmp>>dst_depth) + dither[6])>>shift);\
tmp = bswap(src[j+7]); dst[j+7] = dbswap((tmp - (tmp>>dst_depth) + dither[7])>>shift);\
tmp = bswap(src[j+0]) >> src_shift; dst[j+0] = dbswap(((tmp - (tmp>>dst_depth) + dither[0])>>shift) << dst_shift);\
tmp = bswap(src[j+1]) >> src_shift; dst[j+1] = dbswap(((tmp - (tmp>>dst_depth) + dither[1])>>shift) << dst_shift);\
tmp = bswap(src[j+2]) >> src_shift; dst[j+2] = dbswap(((tmp - (tmp>>dst_depth) + dither[2])>>shift) << dst_shift);\
tmp = bswap(src[j+3]) >> src_shift; dst[j+3] = dbswap(((tmp - (tmp>>dst_depth) + dither[3])>>shift) << dst_shift);\
tmp = bswap(src[j+4]) >> src_shift; dst[j+4] = dbswap(((tmp - (tmp>>dst_depth) + dither[4])>>shift) << dst_shift);\
tmp = bswap(src[j+5]) >> src_shift; dst[j+5] = dbswap(((tmp - (tmp>>dst_depth) + dither[5])>>shift) << dst_shift);\
tmp = bswap(src[j+6]) >> src_shift; dst[j+6] = dbswap(((tmp - (tmp>>dst_depth) + dither[6])>>shift) << dst_shift);\
tmp = bswap(src[j+7]) >> src_shift; dst[j+7] = dbswap(((tmp - (tmp>>dst_depth) + dither[7])>>shift) << dst_shift);\
}\
for (; j < length; j++) {\
tmp = bswap(src[j]); dst[j] = dbswap((tmp - (tmp>>dst_depth) + dither[j&7])>>shift);\
@ -2235,6 +2235,8 @@ static int planarCopyWrapper(SwsInternal *c, const uint8_t *const src[],
) {
const int src_depth = desc_src->comp[plane].depth;
const int dst_depth = desc_dst->comp[plane].depth;
const int src_shift = desc_src->comp[plane].shift;
const int dst_shift = desc_dst->comp[plane].shift;
const uint16_t *srcPtr2 = (const uint16_t *) srcPtr;
uint16_t *dstPtr2 = (uint16_t*)dstPtr;
@ -2250,11 +2252,11 @@ static int planarCopyWrapper(SwsInternal *c, const uint8_t *const src[],
#define COPY816(w)\
if (shiftonly) {\
for (j = 0; j < length; j++)\
w(&dstPtr2[j], srcPtr[j]<<(dst_depth-8));\
w(&dstPtr2[j], (srcPtr[j]<<(dst_depth-8)) << dst_shift);\
} else {\
for (j = 0; j < length; j++)\
w(&dstPtr2[j], (srcPtr[j]<<(dst_depth-8)) |\
(srcPtr[j]>>(2*8-dst_depth)));\
w(&dstPtr2[j], ((srcPtr[j]<<(dst_depth-8)) |\
(srcPtr[j]>>(2*8-dst_depth))) << dst_shift);\
}
if(isBE(c->opts.dst_format)){
COPY816(AV_WB16)
@ -2274,13 +2276,13 @@ static int planarCopyWrapper(SwsInternal *c, const uint8_t *const src[],
#if HAVE_FAST_64BIT
#define FAST_COPY_UP(shift) \
for (; j < length - 3; j += 4) { \
uint64_t v = AV_RN64A(srcPtr2 + j); \
uint64_t v = AV_RN64A(srcPtr2 + j) >> src_shift; \
AV_WN64A(dstPtr2 + j, v << shift); \
}
#else
#define FAST_COPY_UP(shift) \
for (; j < length - 1; j += 2) { \
uint32_t v = AV_RN32A(srcPtr2 + j); \
uint32_t v = AV_RN32A(srcPtr2 + j) >> src_shift; \
AV_WN32A(dstPtr2 + j, v << shift); \
}
#endif
@ -2293,14 +2295,14 @@ static int planarCopyWrapper(SwsInternal *c, const uint8_t *const src[],
#define COPY_UP(r,w) \
if(shiftonly){\
for (; j < length; j++){ \
unsigned int v= r(&srcPtr2[j]);\
w(&dstPtr2[j], v<<(dst_depth-src_depth));\
unsigned int v= r(&srcPtr2[j]) >> src_shift;\
w(&dstPtr2[j], (v << (dst_depth-src_depth)) << dst_shift);\
}\
}else{\
for (; j < length; j++){ \
unsigned int v= r(&srcPtr2[j]);\
w(&dstPtr2[j], (v<<(dst_depth-src_depth)) | \
(v>>(2*src_depth-dst_depth)));\
unsigned int v= r(&srcPtr2[j]) >> src_shift;\
w(&dstPtr2[j], ((v<<(dst_depth-src_depth)) | \
(v>>(2*src_depth-dst_depth))) << dst_shift);\
}\
}
if(isBE(c->opts.src_format)){

View File

@ -1,2 +1,2 @@
65d33af5a5828eb16b42573e9720b370 *tests/data/pixfmt/p410-nv24.yuv
bd4ba14d2828d10001aa6111adda31a8 *tests/data/pixfmt/p410-nv24.yuv
15206400 tests/data/pixfmt/p410-nv24.yuv

View File

@ -1,2 +1,2 @@
1b2996e48f9f2f9c08b47a7af2d1ad3a *tests/data/pixfmt/p410-p412be.yuv
60b454bfd6f48b64cc24de3041d0a01d *tests/data/pixfmt/p410-p412be.yuv
15206400 tests/data/pixfmt/p410-p412be.yuv

View File

@ -1,2 +1,2 @@
1b2996e48f9f2f9c08b47a7af2d1ad3a *tests/data/pixfmt/p410-p412le.yuv
60b454bfd6f48b64cc24de3041d0a01d *tests/data/pixfmt/p410-p412le.yuv
15206400 tests/data/pixfmt/p410-p412le.yuv

View File

@ -1,2 +1,2 @@
0a3bffd9200664676aa69684570a0053 *tests/data/pixfmt/p410-p416be.yuv
60b454bfd6f48b64cc24de3041d0a01d *tests/data/pixfmt/p410-p416be.yuv
15206400 tests/data/pixfmt/p410-p416be.yuv

View File

@ -1,2 +1,2 @@
cffbc4dedbfc209aec8bb4921c19fbf2 *tests/data/pixfmt/p410-p416le.yuv
60b454bfd6f48b64cc24de3041d0a01d *tests/data/pixfmt/p410-p416le.yuv
15206400 tests/data/pixfmt/p410-p416le.yuv

View File

@ -1,2 +1,2 @@
5400f2bfef1b3dc0fcea1fd8ef91e2c8 *tests/data/pixfmt/p412-nv24.yuv
bd4ba14d2828d10001aa6111adda31a8 *tests/data/pixfmt/p412-nv24.yuv
15206400 tests/data/pixfmt/p412-nv24.yuv

View File

@ -1,2 +1,2 @@
1e268eb904e8d173d1585474d12e6f10 *tests/data/pixfmt/p412-p410be.yuv
51e761c143c3fce8e9921d8cceb0a433 *tests/data/pixfmt/p412-p410be.yuv
15206400 tests/data/pixfmt/p412-p410be.yuv

View File

@ -1,2 +1,2 @@
1e268eb904e8d173d1585474d12e6f10 *tests/data/pixfmt/p412-p410le.yuv
51e761c143c3fce8e9921d8cceb0a433 *tests/data/pixfmt/p412-p410le.yuv
15206400 tests/data/pixfmt/p412-p410le.yuv

View File

@ -1,2 +1,2 @@
598b43d980ff2a5e24ee4003befcb875 *tests/data/pixfmt/p412-p416be.yuv
9b5fc804354cf299fcc4c3f7bd2781ad *tests/data/pixfmt/p412-p416be.yuv
15206400 tests/data/pixfmt/p412-p416be.yuv

View File

@ -1,2 +1,2 @@
598b43d980ff2a5e24ee4003befcb875 *tests/data/pixfmt/p412-p416le.yuv
9b5fc804354cf299fcc4c3f7bd2781ad *tests/data/pixfmt/p412-p416le.yuv
15206400 tests/data/pixfmt/p412-p416le.yuv