[FFmpeg-devel] [PATCH 1/5] avcodec/ppc/hpeldsp_altivec: Fix left-shift of negative number

[FFmpeg-devel] [PATCH 1/5] avcodec/ppc/hpeldsp_altivec: Fix left-shift of negative number

[Andreas Rheinhardt]

It is UB and affected e.g. the vp5 and vp61 FATE tests:
https://fate.ffmpeg.org/report.cgi?time=20240327083327&slot=ppc-linux-gcc-13.2-ubsan-altivec-qemu

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
---
 libavcodec/ppc/hpeldsp_altivec.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/libavcodec/ppc/hpeldsp_altivec.c b/libavcodec/ppc/hpeldsp_altivec.c
index a531b6b6ec..4bf6b28ed6 100644
--- a/libavcodec/ppc/hpeldsp_altivec.c
+++ b/libavcodec/ppc/hpeldsp_altivec.c
@@ -41,9 +41,9 @@ void ff_put_pixels16_altivec(uint8_t *block, const uint8_t *pixels, ptrdiff_t li
     register vector unsigned char pixelsv1D;
 
     int i;
-    register ptrdiff_t line_size_2 = line_size << 1;
+    register ptrdiff_t line_size_2 = line_size * (1 << 1);
     register ptrdiff_t line_size_3 = line_size + line_size_2;
-    register ptrdiff_t line_size_4 = line_size << 2;
+    register ptrdiff_t line_size_4 = line_size * (1 << 2);
 
 // hand-unrolling the loop by 4 gains about 15%
 // mininum execution time goes from 74 to 60 cycles
-- 
2.40.1

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

[FFmpeg-devel] [PATCH 2/5] fate/ffmpeg: Explicitly set pix fmt for sub2video tests

[Andreas Rheinhardt]

Otherwise the test would use bgra on little endian and argb
on big endian.

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
---
 tests/fate/ffmpeg.mak | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/tests/fate/ffmpeg.mak b/tests/fate/ffmpeg.mak
index 3c549b265e..fda3a29239 100644
--- a/tests/fate/ffmpeg.mak
+++ b/tests/fate/ffmpeg.mak
@@ -64,7 +64,7 @@ fate-sub2video_basic: CMD = framecrc -auto_conversion_filters \
   -i $(TARGET_SAMPLES)/sub/vobsub.idx \
   -fps_mode passthrough -copyts \
   -filter_complex "sws_flags=+accurate_rnd+bitexact\;[0:s:0]scale" \
-  -c:v rawvideo -threads 1
+  -c:v rawvideo -pix_fmt bgra -threads 1
 
 # Time-limited run with a sample that doesn't require seeking and
 # contains samples within the initial period.
@@ -74,7 +74,7 @@ fate-sub2video_time_limited: CMD = framecrc -auto_conversion_filters \
   -fps_mode passthrough -copyts \
   -t 15 \
   -filter_complex "sws_flags=+accurate_rnd+bitexact\;[0:s:0]scale" \
-  -c:v rawvideo -threads 1
+  -c:v rawvideo -threads 1 -pix_fmt bgra
 
 FATE_FFMPEG-$(call ENCDEC, PCM_S16LE, PCM_S16LE) += fate-unknown_layout-pcm
 fate-unknown_layout-pcm: $(AREF)
-- 
2.40.1

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] [PATCH 2/5] fate/ffmpeg: Explicitly set pix fmt for sub2video tests

[Sean McGovern]

Hi,


On Thu, Mar 28, 2024, 23:24 Andreas Rheinhardt <
andreas.rheinhardt@outlook.com> wrote:

> Otherwise the test would use bgra on little endian and argb
> on big endian.
>
> Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
> ---
>  tests/fate/ffmpeg.mak | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/tests/fate/ffmpeg.mak b/tests/fate/ffmpeg.mak
> index 3c549b265e..fda3a29239 100644
> --- a/tests/fate/ffmpeg.mak
> +++ b/tests/fate/ffmpeg.mak
> @@ -64,7 +64,7 @@ fate-sub2video_basic: CMD = framecrc
> -auto_conversion_filters \
>    -i $(TARGET_SAMPLES)/sub/vobsub.idx \
>    -fps_mode passthrough -copyts \
>    -filter_complex "sws_flags=+accurate_rnd+bitexact\;[0:s:0]scale" \
> -  -c:v rawvideo -threads 1
> +  -c:v rawvideo -pix_fmt bgra -threads 1
>
>  # Time-limited run with a sample that doesn't require seeking and
>  # contains samples within the initial period.
> @@ -74,7 +74,7 @@ fate-sub2video_time_limited: CMD = framecrc
> -auto_conversion_filters \
>    -fps_mode passthrough -copyts \
>    -t 15 \
>    -filter_complex "sws_flags=+accurate_rnd+bitexact\;[0:s:0]scale" \
> -  -c:v rawvideo -threads 1
> +  -c:v rawvideo -threads 1 -pix_fmt bgra
>
>  FATE_FFMPEG-$(call ENCDEC, PCM_S16LE, PCM_S16LE) +=
> fate-unknown_layout-pcm
>  fate-unknown_layout-pcm: $(AREF)
> --
> 2.40.1
>
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel@ffmpeg.org
> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>
> To unsubscribe, visit link above, or email
> ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
>

I already submitted this change a few weeks ago (March 6th to be precise)
and it has been left unreviewed.

-- Sean McGovern

>
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] [PATCH 2/5] fate/ffmpeg: Explicitly set pix fmt for sub2video tests

[Andreas Rheinhardt]

Sean McGovern:
> Hi,
> 
> 
> On Thu, Mar 28, 2024, 23:24 Andreas Rheinhardt <
> andreas.rheinhardt@outlook.com> wrote:
> 
>> Otherwise the test would use bgra on little endian and argb
>> on big endian.
>>
>> Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
>> ---
>>  tests/fate/ffmpeg.mak | 4 ++--
>>  1 file changed, 2 insertions(+), 2 deletions(-)
>>
>> diff --git a/tests/fate/ffmpeg.mak b/tests/fate/ffmpeg.mak
>> index 3c549b265e..fda3a29239 100644
>> --- a/tests/fate/ffmpeg.mak
>> +++ b/tests/fate/ffmpeg.mak
>> @@ -64,7 +64,7 @@ fate-sub2video_basic: CMD = framecrc
>> -auto_conversion_filters \
>>    -i $(TARGET_SAMPLES)/sub/vobsub.idx \
>>    -fps_mode passthrough -copyts \
>>    -filter_complex "sws_flags=+accurate_rnd+bitexact\;[0:s:0]scale" \
>> -  -c:v rawvideo -threads 1
>> +  -c:v rawvideo -pix_fmt bgra -threads 1
>>
>>  # Time-limited run with a sample that doesn't require seeking and
>>  # contains samples within the initial period.
>> @@ -74,7 +74,7 @@ fate-sub2video_time_limited: CMD = framecrc
>> -auto_conversion_filters \
>>    -fps_mode passthrough -copyts \
>>    -t 15 \
>>    -filter_complex "sws_flags=+accurate_rnd+bitexact\;[0:s:0]scale" \
>> -  -c:v rawvideo -threads 1
>> +  -c:v rawvideo -threads 1 -pix_fmt bgra
>>
>>  FATE_FFMPEG-$(call ENCDEC, PCM_S16LE, PCM_S16LE) +=
>> fate-unknown_layout-pcm
>>  fate-unknown_layout-pcm: $(AREF)
>> --
>> 2.40.1
>>
>> _______________________________________________
>> ffmpeg-devel mailing list
>> ffmpeg-devel@ffmpeg.org
>> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>>
>> To unsubscribe, visit link above, or email
>> ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
>>
> 
> I already submitted this change a few weeks ago (March 6th to be precise)
> and it has been left unreviewed.
> 

Ok, will apply yours then.

- Andreas

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

[FFmpeg-devel] [PATCH 3/5] avcodec/pngdsp: Fix unaligned accesses, effective type violations

[Andreas Rheinhardt]

Affected the lscr fate-test (only visible on x86 if
the SSE2 is disabled).

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
---
 libavcodec/pngdsp.c | 25 +++++++++++++++++++------
 1 file changed, 19 insertions(+), 6 deletions(-)

diff --git a/libavcodec/pngdsp.c b/libavcodec/pngdsp.c
index 65916b1386..50ee96a684 100644
--- a/libavcodec/pngdsp.c
+++ b/libavcodec/pngdsp.c
@@ -21,20 +21,33 @@
 
 #include "config.h"
 #include "libavutil/attributes.h"
+#include "libavutil/intreadwrite.h"
+#include "libavutil/macros.h"
 #include "png.h"
 #include "pngdsp.h"
 
+#if HAVE_FAST_64BIT
+#define BITS 64
+typedef uint64_t uint_native;
+#else
+#define BITS 32
+typedef uint32_t uint_native;
+#endif
+#define RN  AV_JOIN(AV_RN, BITS)
+#define RNA AV_JOIN(AV_JOIN(AV_RN, BITS),  A)
+#define WN  AV_JOIN(AV_WN, BITS)
+
 // 0x7f7f7f7f or 0x7f7f7f7f7f7f7f7f or whatever, depending on the cpu's native arithmetic size
-#define pb_7f (~0UL / 255 * 0x7f)
-#define pb_80 (~0UL / 255 * 0x80)
+#define pb_7f (~(uint_native)0 / 255 * 0x7f)
+#define pb_80 (~(uint_native)0 / 255 * 0x80)
 
 static void add_bytes_l2_c(uint8_t *dst, uint8_t *src1, uint8_t *src2, int w)
 {
     long i;
-    for (i = 0; i <= w - (int) sizeof(long); i += sizeof(long)) {
-        long a = *(long *)(src1 + i);
-        long b = *(long *)(src2 + i);
-        *(long *)(dst + i) = ((a & pb_7f) + (b & pb_7f)) ^ ((a ^ b) & pb_80);
+    for (i = 0; i <= w - (int) sizeof(uint_native); i += sizeof(uint_native)) {
+        uint_native a = RNA(src1 + i);
+        uint_native b = RN (src2 + i);
+        WN(dst + i, ((a & pb_7f) + (b & pb_7f)) ^ ((a ^ b) & pb_80));
     }
     for (; i < w; i++)
         dst[i] = src1[i] + src2[i];
-- 
2.40.1

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

[FFmpeg-devel] [PATCH 4/5] avfilter/vf_spp: Fix left-shift of negative value

[Andreas Rheinhardt]

Affected the vf-spp FATE-test (on x86 only when MMX
is disabled).

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
---
 libavfilter/vf_spp.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/libavfilter/vf_spp.c b/libavfilter/vf_spp.c
index 5c6495612b..c8366ae319 100644
--- a/libavfilter/vf_spp.c
+++ b/libavfilter/vf_spp.c
@@ -171,7 +171,7 @@ static void store_slice_c(uint8_t *dst, const int16_t *src,
     int y, x;
 
 #define STORE(pos) do {                                                     \
-    temp = ((src[x + y*src_linesize + pos] << log2_scale) + d[pos]) >> 6;   \
+    temp = (src[x + y*src_linesize + pos] * (1 << log2_scale) + d[pos]) >> 6;\
     if (temp & 0x100)                                                       \
         temp = ~(temp >> 31);                                               \
     dst[x + y*dst_linesize + pos] = temp;                                   \
@@ -202,7 +202,7 @@ static void store_slice16_c(uint16_t *dst, const int16_t *src,
     unsigned int mask = -1<<depth;
 
 #define STORE16(pos) do {                                                   \
-    temp = ((src[x + y*src_linesize + pos] << log2_scale) + (d[pos]>>1)) >> 5;   \
+    temp = (src[x + y*src_linesize + pos] * (1 << log2_scale) + (d[pos]>>1)) >> 5; \
     if (temp & mask )                                                       \
         temp = ~(temp >> 31);                                               \
     dst[x + y*dst_linesize + pos] = temp;                                   \
-- 
2.40.1

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

[FFmpeg-devel] [PATCH 5/5] avcodec/huffyuvencdsp: Fix load of misaligned values

[Andreas Rheinhardt]

Affected many ffvhuff FATE tests.

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
---
 libavcodec/huffyuvencdsp.c | 32 ++++++++++++++++++++++++--------
 1 file changed, 24 insertions(+), 8 deletions(-)

diff --git a/libavcodec/huffyuvencdsp.c b/libavcodec/huffyuvencdsp.c
index 36e8f6130b..27428635af 100644
--- a/libavcodec/huffyuvencdsp.c
+++ b/libavcodec/huffyuvencdsp.c
@@ -18,16 +18,32 @@
 
 #include "config.h"
 #include "libavutil/attributes.h"
+#include "libavutil/intreadwrite.h"
 #include "huffyuvencdsp.h"
 #include "mathops.h"
 
+#if HAVE_FAST_64BIT
+#define BITS 64
+typedef uint64_t uint_native;
+#else
+#define BITS 32
+typedef uint32_t uint_native;
+#endif
+#define RN  AV_JOIN(AV_RN, BITS)
+#define RNA AV_JOIN(AV_JOIN(AV_RN, BITS), A)
+#define WNA AV_JOIN(AV_JOIN(AV_WN, BITS), A)
+
+// 0x7f7f7f7f or 0x7f7f7f7f7f7f7f7f or whatever, depending on the cpu's native arithmetic size
+#define pb_7f (~(uint_native)0 / 255 * 0x7f)
+#define pb_80 (~(uint_native)0 / 255 * 0x80)
+
 // 0x00010001 or 0x0001000100010001 or whatever, depending on the cpu's native arithmetic size
-#define pw_1 (ULONG_MAX / UINT16_MAX)
+#define pw_1 ((uint_native)-1 / UINT16_MAX)
 
 static void diff_int16_c(uint16_t *dst, const uint16_t *src1, const uint16_t *src2, unsigned mask, int w){
     long i;
 #if !HAVE_FAST_UNALIGNED
-    if((long)src2 & (sizeof(long)-1)){
+    if ((uintptr_t)src2 & (sizeof(uint_native) - 1)) {
         for(i=0; i+3<w; i+=4){
             dst[i+0] = (src1[i+0]-src2[i+0]) & mask;
             dst[i+1] = (src1[i+1]-src2[i+1]) & mask;
@@ -37,13 +53,13 @@ static void diff_int16_c(uint16_t *dst, const uint16_t *src1, const uint16_t *sr
     }else
 #endif
     {
-        unsigned long pw_lsb = (mask >> 1) * pw_1;
-        unsigned long pw_msb = pw_lsb +  pw_1;
+        uint_native pw_lsb = (mask >> 1) * pw_1;
+        uint_native pw_msb = pw_lsb +  pw_1;
 
-        for (i = 0; i <= w - (int)sizeof(long)/2; i += sizeof(long)/2) {
-            long a = *(long*)(src1+i);
-            long b = *(long*)(src2+i);
-            *(long*)(dst+i) = ((a|pw_msb) - (b&pw_lsb)) ^ ((a^b^pw_msb)&pw_msb);
+        for (i = 0; i <= w - (int)sizeof(uint_native)/2; i += sizeof(uint_native)/2) {
+            uint_native a = RNA(src1 + i);
+            uint_native b = RN (src2 + i);
+            WNA(dst + i, ((a | pw_msb) - (b & pw_lsb)) ^ ((a^b^pw_msb) & pw_msb));
         }
     }
     for (; i<w; i++)
-- 
2.40.1

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

[FFmpeg-devel] [PATCH 6/6] fate/image: Fix EXR tests on big endian

[Andreas Rheinhardt]

These tests need a scale filter to convert to the prescribed
pixel format (the native format is endian-dependent).

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
---
And now I will deduplicate this mess...

 tests/fate/image.mak | 150 +++++++++++++++++++++----------------------
 1 file changed, 75 insertions(+), 75 deletions(-)

diff --git a/tests/fate/image.mak b/tests/fate/image.mak
index 7c0e0fec08..753936ec20 100644
--- a/tests/fate/image.mak
+++ b/tests/fate/image.mak
@@ -104,229 +104,229 @@ FATE_IMAGE_PROBE-$(call DEMDEC, IMAGE2, DPX) += fate-dpx-probe
 fate-dpx-probe: CMD = probeframes -show_entries frame=color_transfer,color_range,color_space,color_primaries,sample_aspect_ratio $(TARGET_SAMPLES)/dpx/cyan.dpx
 
 FATE_EXR += fate-exr-slice-raw
-fate-exr-slice-raw: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_slice_raw.exr -pix_fmt gbrapf32le
+fate-exr-slice-raw: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_slice_raw.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-slice-rle
-fate-exr-slice-rle: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_slice_rle.exr -pix_fmt gbrapf32le
+fate-exr-slice-rle: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_slice_rle.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-slice-zip1
-fate-exr-slice-zip1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_slice_zip1.exr -pix_fmt gbrapf32le
+fate-exr-slice-zip1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_slice_zip1.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-slice-zip16
-fate-exr-slice-zip16: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_slice_zip16.exr -pix_fmt gbrapf32le
+fate-exr-slice-zip16: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_slice_zip16.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-slice-pxr24
-fate-exr-slice-pxr24: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_slice_pxr24.exr -pix_fmt gbrpf32le
+fate-exr-slice-pxr24: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_slice_pxr24.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-pxr24-float-12x8
-fate-exr-rgb-scanline-pxr24-float-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_float_12x8.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-pxr24-float-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_float_12x8.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgba-multiscanline-half-b44
-fate-exr-rgba-multiscanline-half-b44: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_multiscanline_half_b44.exr -pix_fmt gbrapf32le
+fate-exr-rgba-multiscanline-half-b44: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_multiscanline_half_b44.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgb-scanline-float-b44
-fate-exr-rgb-scanline-float-b44: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_float_b44.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-float-b44: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_float_b44.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-half-b44-12x8
-fate-exr-rgb-scanline-half-b44-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_b44_12x8.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-half-b44-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_b44_12x8.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-half-b44-13x9
-fate-exr-rgb-scanline-half-b44-13x9: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_b44_13x9.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-half-b44-13x9: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_b44_13x9.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-tile-float-raw-12x8
-fate-exr-rgb-tile-float-raw-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_float_raw_12x8.exr -pix_fmt gbrpf32le
+fate-exr-rgb-tile-float-raw-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_float_raw_12x8.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-tile-float-raw-150x130
-fate-exr-rgb-tile-float-raw-150x130: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_float_raw_150x130.exr -pix_fmt gbrpf32le
+fate-exr-rgb-tile-float-raw-150x130: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_float_raw_150x130.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-tile-half-raw-12x8
-fate-exr-rgb-tile-half-raw-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_raw_12x8.exr -pix_fmt gbrpf32le
+fate-exr-rgb-tile-half-raw-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_raw_12x8.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgba-scanline-float-half-b44-13x9-l1
-fate-exr-rgba-scanline-float-half-b44-13x9-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44_13x9.exr -pix_fmt gbrapf32le
+fate-exr-rgba-scanline-float-half-b44-13x9-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44_13x9.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgba-scanline-float-half-b44-13x9-l2
-fate-exr-rgba-scanline-float-half-b44-13x9-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44_13x9.exr -pix_fmt gbrapf32le
+fate-exr-rgba-scanline-float-half-b44-13x9-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44_13x9.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgba-scanline-float-half-b44-12x8-l1
-fate-exr-rgba-scanline-float-half-b44-12x8-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44_12x8.exr -pix_fmt gbrapf32le
+fate-exr-rgba-scanline-float-half-b44-12x8-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44_12x8.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgba-scanline-float-half-b44-12x8-l2
-fate-exr-rgba-scanline-float-half-b44-12x8-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44_12x8.exr -pix_fmt gbrapf32le
+fate-exr-rgba-scanline-float-half-b44-12x8-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44_12x8.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgba-scanline-float-half-b44a-12x8-l1
-fate-exr-rgba-scanline-float-half-b44a-12x8-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44a_12x8.exr -pix_fmt gbrapf32le
+fate-exr-rgba-scanline-float-half-b44a-12x8-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44a_12x8.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgba-scanline-float-half-b44a-12x8-l2
-fate-exr-rgba-scanline-float-half-b44a-12x8-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44a_12x8.exr -pix_fmt gbrapf32le
+fate-exr-rgba-scanline-float-half-b44a-12x8-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44a_12x8.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgba-scanline-float-half-b44a-13x9-l1
-fate-exr-rgba-scanline-float-half-b44a-13x9-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44a_13x9.exr -pix_fmt gbrapf32le
+fate-exr-rgba-scanline-float-half-b44a-13x9-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44a_13x9.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgba-scanline-float-half-b44a-13x9-l2
-fate-exr-rgba-scanline-float-half-b44a-13x9-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44a_13x9.exr -pix_fmt gbrapf32le
+fate-exr-rgba-scanline-float-half-b44a-13x9-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44a_13x9.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgb-tile-pxr24-float-half-l1
-fate-exr-rgb-tile-pxr24-float-half-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_pxr24_float_half.exr -pix_fmt gbrpf32le
+fate-exr-rgb-tile-pxr24-float-half-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_pxr24_float_half.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-tile-pxr24-float-half-l2
-fate-exr-rgb-tile-pxr24-float-half-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_pxr24_float_half.exr -pix_fmt gbrapf32le
+fate-exr-rgb-tile-pxr24-float-half-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_pxr24_float_half.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgb-tile-pxr24-half-float-l1
-fate-exr-rgb-tile-pxr24-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_pxr24_half_float.exr -pix_fmt gbrpf32le
+fate-exr-rgb-tile-pxr24-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_pxr24_half_float.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-tile-pxr24-half-float-l2
-fate-exr-rgb-tile-pxr24-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_pxr24_half_float.exr -pix_fmt gbrapf32le
+fate-exr-rgb-tile-pxr24-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_pxr24_half_float.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgb-tile-half-float-b44-12x8-l1
-fate-exr-rgb-tile-half-float-b44-12x8-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_float_b44_12x8.exr -pix_fmt gbrpf32le
+fate-exr-rgb-tile-half-float-b44-12x8-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_float_b44_12x8.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-tile-half-float-b44-12x8-l2
-fate-exr-rgb-tile-half-float-b44-12x8-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_half_float_b44_12x8.exr -pix_fmt gbrapf32le
+fate-exr-rgb-tile-half-float-b44-12x8-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_half_float_b44_12x8.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgb-tile-zip-half-float-l1
-fate-exr-rgb-tile-zip-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_zip_half_float.exr -pix_fmt gbrpf32le
+fate-exr-rgb-tile-zip-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_zip_half_float.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-tile-zip-half-float-l2
-fate-exr-rgb-tile-zip-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_zip_half_float.exr -pix_fmt gbrapf32le
+fate-exr-rgb-tile-zip-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_zip_half_float.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgb-tile-zip1-half-float-l1
-fate-exr-rgb-tile-zip1-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_zip1_half_float.exr -pix_fmt gbrpf32le
+fate-exr-rgb-tile-zip1-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_zip1_half_float.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-tile-zip1-half-float-l2
-fate-exr-rgb-tile-zip1-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_zip1_half_float.exr -pix_fmt gbrapf32le
+fate-exr-rgb-tile-zip1-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_zip1_half_float.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgb-tile-rle-half-float-l1
-fate-exr-rgb-tile-rle-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_rle_half_float.exr -pix_fmt gbrpf32le
+fate-exr-rgb-tile-rle-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_rle_half_float.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-tile-rle-half-float-l2
-fate-exr-rgb-tile-rle-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_rle_half_float.exr -pix_fmt gbrapf32le
+fate-exr-rgb-tile-rle-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_rle_half_float.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgb-tile-raw-half-float-l1
-fate-exr-rgb-tile-raw-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_raw_half_float.exr -pix_fmt gbrpf32le
+fate-exr-rgb-tile-raw-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_raw_half_float.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-tile-raw-half-float-l2
-fate-exr-rgb-tile-raw-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_raw_half_float.exr -pix_fmt gbrapf32le
+fate-exr-rgb-tile-raw-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_raw_half_float.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgb-scanline-b44-half-float-12x8-l1
-fate-exr-rgb-scanline-b44-half-float-12x8-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_b44_half_float_12x8.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-b44-half-float-12x8-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_b44_half_float_12x8.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-b44-half-float-12x8-l2
-fate-exr-rgb-scanline-b44-half-float-12x8-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_b44_half_float_12x8.exr -pix_fmt gbrapf32le
+fate-exr-rgb-scanline-b44-half-float-12x8-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_b44_half_float_12x8.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgb-scanline-pxr24-half-float-l1
-fate-exr-rgb-scanline-pxr24-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_half_float.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-pxr24-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_half_float.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-pxr24-half-float-l2
-fate-exr-rgb-scanline-pxr24-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_half_float.exr -pix_fmt gbrapf32le
+fate-exr-rgb-scanline-pxr24-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_half_float.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgb-scanline-pxr24-float-half-l1
-fate-exr-rgb-scanline-pxr24-float-half-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_float_half.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-pxr24-float-half-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_float_half.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-pxr24-float-half-l2
-fate-exr-rgb-scanline-pxr24-float-half-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_float_half.exr -pix_fmt gbrapf32le
+fate-exr-rgb-scanline-pxr24-float-half-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_float_half.exr -vf scale -pix_fmt gbrapf32le
 
-FATE_EXR-$(call DEMDEC, IMAGE2, EXR, SCALE_FILTER) += fate-exr-rgb-scanline-pxr24-half-uint32-13x9
+FATE_EXR += fate-exr-rgb-scanline-pxr24-half-uint32-13x9
 fate-exr-rgb-scanline-pxr24-half-uint32-13x9: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_half_uint32_13x9.exr -pix_fmt rgb48le -vf scale
 
 FATE_EXR += fate-exr-rgb-scanline-zip-half-float-l1
-fate-exr-rgb-scanline-zip-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip_half_float.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-zip-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip_half_float.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-zip-half-float-l2
-fate-exr-rgb-scanline-zip-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip_half_float.exr -pix_fmt gbrapf32le
+fate-exr-rgb-scanline-zip-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip_half_float.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgb-scanline-zip1-half-float-l1
-fate-exr-rgb-scanline-zip1-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip1_half_float.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-zip1-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip1_half_float.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-zip1-half-float-l2
-fate-exr-rgb-scanline-zip1-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip1_half_float.exr -pix_fmt gbrapf32le
+fate-exr-rgb-scanline-zip1-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip1_half_float.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgb-scanline-rle-half-float-l1
-fate-exr-rgb-scanline-rle-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_rle_half_float.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-rle-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_rle_half_float.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-rle-half-float-l2
-fate-exr-rgb-scanline-rle-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_rle_half_float.exr -pix_fmt gbrapf32le
+fate-exr-rgb-scanline-rle-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_rle_half_float.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgb-scanline-raw-half-float-l1
-fate-exr-rgb-scanline-raw-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_raw_half_float.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-raw-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_raw_half_float.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-raw-half-float-l2
-fate-exr-rgb-scanline-raw-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_raw_half_float.exr -pix_fmt gbrapf32le
+fate-exr-rgb-scanline-raw-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_raw_half_float.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgb-scanline-b44-uint32
-fate-exr-rgb-scanline-b44-uint32: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_b44_uint32.exr -pix_fmt rgb48le
+fate-exr-rgb-scanline-b44-uint32: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_b44_uint32.exr -vf scale -pix_fmt rgb48le
 
 FATE_EXR += fate-exr-rgb-scanline-pxr24-uint32
-fate-exr-rgb-scanline-pxr24-uint32: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_uint32.exr -pix_fmt rgb48le
+fate-exr-rgb-scanline-pxr24-uint32: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_uint32.exr -vf scale -pix_fmt rgb48le
 
 FATE_EXR += fate-exr-rgb-scanline-zip1-half-float-l1-zero-offsets
-fate-exr-rgb-scanline-zip1-half-float-l1-zero-offsets: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip1_half_float_zero_offsets.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-zip1-half-float-l1-zero-offsets: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip1_half_float_zero_offsets.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-half-piz-bw
-fate-exr-rgb-scanline-half-piz-bw: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_bw.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-half-piz-bw: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_bw.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-half-piz-color
-fate-exr-rgb-scanline-half-piz-color: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_color.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-half-piz-color: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_color.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-half-piz-dw-t01
-fate-exr-rgb-scanline-half-piz-dw-t01: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_dw_t01.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-half-piz-dw-t01: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_dw_t01.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-float-piz-48x32
-fate-exr-rgb-scanline-float-piz-48x32: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_float_piz_48x32.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-float-piz-48x32: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_float_piz_48x32.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-none-negative-red
-fate-exr-rgb-scanline-none-negative-red: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_none_negative_red.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-none-negative-red: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_none_negative_red.exr -vf scale -pix_fmt gbrpf32le
 
 
 FATE_EXR += fate-exr-rgb-b44a-half-negative-4x4
-fate-exr-rgb-b44a-half-negative-4x4: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_b44a_half_negative_4x4.exr -pix_fmt gbrpf32le
+fate-exr-rgb-b44a-half-negative-4x4: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_b44a_half_negative_4x4.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-y-tile-zip-half-12x8
-fate-exr-y-tile-zip-half-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/y_tile_zip_half_12x8.exr -pix_fmt grayf32le
+fate-exr-y-tile-zip-half-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/y_tile_zip_half_12x8.exr -vf scale -pix_fmt grayf32le
 
 FATE_EXR += fate-exr-y-scanline-zip-half-12x8
-fate-exr-y-scanline-zip-half-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/y_scanline_zip_half_12x8.exr -pix_fmt grayf32le
+fate-exr-y-scanline-zip-half-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/y_scanline_zip_half_12x8.exr -vf scale -pix_fmt grayf32le
 
 FATE_EXR += fate-exr-rgb-scanline-half-piz-dw-t08
-fate-exr-rgb-scanline-half-piz-dw-t08: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_dw_t08.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-half-piz-dw-t08: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_dw_t08.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgba-zip16-16x32-flag4
-fate-exr-rgba-zip16-16x32-flag4: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_zip16_16x32_flag4.exr -pix_fmt gbrapf32le
+fate-exr-rgba-zip16-16x32-flag4: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_zip16_16x32_flag4.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-ya-scanline-zip-half-12x8
-fate-exr-ya-scanline-zip-half-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/ya_scanline_zip_half_12x8.exr -pix_fmt gbrapf32le
+fate-exr-ya-scanline-zip-half-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/ya_scanline_zip_half_12x8.exr -vf scale -pix_fmt gbrapf32le
 
 FATE_EXR += fate-exr-rgb-tile-half-zip
-fate-exr-rgb-tile-half-zip: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_zip.exr -pix_fmt gbrpf32le
+fate-exr-rgb-tile-half-zip: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_zip.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-float-zip-dw-large
-fate-exr-rgb-scanline-float-zip-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_float_zip_dw_large.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-float-zip-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_float_zip_dw_large.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-half-piz-dw-large
-fate-exr-rgb-scanline-half-piz-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_dw_large.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-half-piz-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_dw_large.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-half-zip-dw-large
-fate-exr-rgb-scanline-half-zip-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_zip_dw_large.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-half-zip-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_zip_dw_large.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-uint32-piz-dw-large
-fate-exr-rgb-scanline-uint32-piz-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_uint32_piz_dw_large.exr -pix_fmt rgb48le
+fate-exr-rgb-scanline-uint32-piz-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_uint32_piz_dw_large.exr -vf scale -pix_fmt rgb48le
 
 FATE_EXR += fate-exr-rgb-tile-half-piz-dw-large
-fate-exr-rgb-tile-half-piz-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_piz_dw_large.exr -pix_fmt gbrpf32le
+fate-exr-rgb-tile-half-piz-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_piz_dw_large.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-tile-uint32-piz-dw-large
-fate-exr-rgb-tile-uint32-piz-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_uint32_piz_dw_large.exr -pix_fmt rgb48le
+fate-exr-rgb-tile-uint32-piz-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_uint32_piz_dw_large.exr -vf scale -pix_fmt rgb48le
 
 FATE_EXR += fate-exr-rgb-scanline-half-zip-dw-outside
-fate-exr-rgb-scanline-half-zip-dw-outside: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_zip_dw_outside.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-half-zip-dw-outside: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_zip_dw_outside.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-tile-half-zip-dw-outside
-fate-exr-rgb-tile-half-zip-dw-outside: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_zip_dw_outside.exr -pix_fmt gbrpf32le
+fate-exr-rgb-tile-half-zip-dw-outside: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_zip_dw_outside.exr -vf scale -pix_fmt gbrpf32le
 
 FATE_EXR += fate-exr-rgb-scanline-zip-half-0x0-0xFFFF
-fate-exr-rgb-scanline-zip-half-0x0-0xFFFF: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip_half_float_0x0_to_0xFFFF.exr -pix_fmt gbrpf32le
+fate-exr-rgb-scanline-zip-half-0x0-0xFFFF: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip_half_float_0x0_to_0xFFFF.exr -vf scale -pix_fmt gbrpf32le
 
-FATE_EXR-$(call DEMDEC, IMAGE2, EXR) += $(FATE_EXR)
+FATE_EXR-$(call DEMDEC, IMAGE2, EXR, SCALE_FILTER) += $(FATE_EXR)
 
 FATE_IMAGE_FRAMECRC += $(FATE_EXR-yes)
 fate-exr: $(FATE_EXR-yes)
-- 
2.40.1

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] [PATCH 6/6] fate/image: Fix EXR tests on big endian

[Sean McGovern]

Andreas,

On Fri, Mar 29, 2024 at 12:45 PM Andreas Rheinhardt
<andreas.rheinhardt@outlook.com> wrote:
>
> These tests need a scale filter to convert to the prescribed
> pixel format (the native format is endian-dependent).
>
> Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
> ---
> And now I will deduplicate this mess...
>
>  tests/fate/image.mak | 150 +++++++++++++++++++++----------------------
>  1 file changed, 75 insertions(+), 75 deletions(-)
>
> diff --git a/tests/fate/image.mak b/tests/fate/image.mak
> index 7c0e0fec08..753936ec20 100644
> --- a/tests/fate/image.mak
> +++ b/tests/fate/image.mak
> @@ -104,229 +104,229 @@ FATE_IMAGE_PROBE-$(call DEMDEC, IMAGE2, DPX) += fate-dpx-probe
>  fate-dpx-probe: CMD = probeframes -show_entries frame=color_transfer,color_range,color_space,color_primaries,sample_aspect_ratio $(TARGET_SAMPLES)/dpx/cyan.dpx
>
>  FATE_EXR += fate-exr-slice-raw
> -fate-exr-slice-raw: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_slice_raw.exr -pix_fmt gbrapf32le
> +fate-exr-slice-raw: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_slice_raw.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-slice-rle
> -fate-exr-slice-rle: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_slice_rle.exr -pix_fmt gbrapf32le
> +fate-exr-slice-rle: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_slice_rle.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-slice-zip1
> -fate-exr-slice-zip1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_slice_zip1.exr -pix_fmt gbrapf32le
> +fate-exr-slice-zip1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_slice_zip1.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-slice-zip16
> -fate-exr-slice-zip16: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_slice_zip16.exr -pix_fmt gbrapf32le
> +fate-exr-slice-zip16: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_slice_zip16.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-slice-pxr24
> -fate-exr-slice-pxr24: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_slice_pxr24.exr -pix_fmt gbrpf32le
> +fate-exr-slice-pxr24: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_slice_pxr24.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-pxr24-float-12x8
> -fate-exr-rgb-scanline-pxr24-float-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_float_12x8.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-pxr24-float-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_float_12x8.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgba-multiscanline-half-b44
> -fate-exr-rgba-multiscanline-half-b44: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_multiscanline_half_b44.exr -pix_fmt gbrapf32le
> +fate-exr-rgba-multiscanline-half-b44: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_multiscanline_half_b44.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-float-b44
> -fate-exr-rgb-scanline-float-b44: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_float_b44.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-float-b44: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_float_b44.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-half-b44-12x8
> -fate-exr-rgb-scanline-half-b44-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_b44_12x8.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-half-b44-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_b44_12x8.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-half-b44-13x9
> -fate-exr-rgb-scanline-half-b44-13x9: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_b44_13x9.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-half-b44-13x9: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_b44_13x9.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-tile-float-raw-12x8
> -fate-exr-rgb-tile-float-raw-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_float_raw_12x8.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-tile-float-raw-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_float_raw_12x8.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-tile-float-raw-150x130
> -fate-exr-rgb-tile-float-raw-150x130: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_float_raw_150x130.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-tile-float-raw-150x130: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_float_raw_150x130.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-tile-half-raw-12x8
> -fate-exr-rgb-tile-half-raw-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_raw_12x8.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-tile-half-raw-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_raw_12x8.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgba-scanline-float-half-b44-13x9-l1
> -fate-exr-rgba-scanline-float-half-b44-13x9-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44_13x9.exr -pix_fmt gbrapf32le
> +fate-exr-rgba-scanline-float-half-b44-13x9-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44_13x9.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgba-scanline-float-half-b44-13x9-l2
> -fate-exr-rgba-scanline-float-half-b44-13x9-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44_13x9.exr -pix_fmt gbrapf32le
> +fate-exr-rgba-scanline-float-half-b44-13x9-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44_13x9.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgba-scanline-float-half-b44-12x8-l1
> -fate-exr-rgba-scanline-float-half-b44-12x8-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44_12x8.exr -pix_fmt gbrapf32le
> +fate-exr-rgba-scanline-float-half-b44-12x8-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44_12x8.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgba-scanline-float-half-b44-12x8-l2
> -fate-exr-rgba-scanline-float-half-b44-12x8-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44_12x8.exr -pix_fmt gbrapf32le
> +fate-exr-rgba-scanline-float-half-b44-12x8-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44_12x8.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgba-scanline-float-half-b44a-12x8-l1
> -fate-exr-rgba-scanline-float-half-b44a-12x8-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44a_12x8.exr -pix_fmt gbrapf32le
> +fate-exr-rgba-scanline-float-half-b44a-12x8-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44a_12x8.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgba-scanline-float-half-b44a-12x8-l2
> -fate-exr-rgba-scanline-float-half-b44a-12x8-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44a_12x8.exr -pix_fmt gbrapf32le
> +fate-exr-rgba-scanline-float-half-b44a-12x8-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44a_12x8.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgba-scanline-float-half-b44a-13x9-l1
> -fate-exr-rgba-scanline-float-half-b44a-13x9-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44a_13x9.exr -pix_fmt gbrapf32le
> +fate-exr-rgba-scanline-float-half-b44a-13x9-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44a_13x9.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgba-scanline-float-half-b44a-13x9-l2
> -fate-exr-rgba-scanline-float-half-b44a-13x9-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44a_13x9.exr -pix_fmt gbrapf32le
> +fate-exr-rgba-scanline-float-half-b44a-13x9-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgba_scanline_float_half_b44a_13x9.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgb-tile-pxr24-float-half-l1
> -fate-exr-rgb-tile-pxr24-float-half-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_pxr24_float_half.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-tile-pxr24-float-half-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_pxr24_float_half.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-tile-pxr24-float-half-l2
> -fate-exr-rgb-tile-pxr24-float-half-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_pxr24_float_half.exr -pix_fmt gbrapf32le
> +fate-exr-rgb-tile-pxr24-float-half-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_pxr24_float_half.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgb-tile-pxr24-half-float-l1
> -fate-exr-rgb-tile-pxr24-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_pxr24_half_float.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-tile-pxr24-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_pxr24_half_float.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-tile-pxr24-half-float-l2
> -fate-exr-rgb-tile-pxr24-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_pxr24_half_float.exr -pix_fmt gbrapf32le
> +fate-exr-rgb-tile-pxr24-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_pxr24_half_float.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgb-tile-half-float-b44-12x8-l1
> -fate-exr-rgb-tile-half-float-b44-12x8-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_float_b44_12x8.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-tile-half-float-b44-12x8-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_float_b44_12x8.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-tile-half-float-b44-12x8-l2
> -fate-exr-rgb-tile-half-float-b44-12x8-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_half_float_b44_12x8.exr -pix_fmt gbrapf32le
> +fate-exr-rgb-tile-half-float-b44-12x8-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_half_float_b44_12x8.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgb-tile-zip-half-float-l1
> -fate-exr-rgb-tile-zip-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_zip_half_float.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-tile-zip-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_zip_half_float.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-tile-zip-half-float-l2
> -fate-exr-rgb-tile-zip-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_zip_half_float.exr -pix_fmt gbrapf32le
> +fate-exr-rgb-tile-zip-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_zip_half_float.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgb-tile-zip1-half-float-l1
> -fate-exr-rgb-tile-zip1-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_zip1_half_float.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-tile-zip1-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_zip1_half_float.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-tile-zip1-half-float-l2
> -fate-exr-rgb-tile-zip1-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_zip1_half_float.exr -pix_fmt gbrapf32le
> +fate-exr-rgb-tile-zip1-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_zip1_half_float.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgb-tile-rle-half-float-l1
> -fate-exr-rgb-tile-rle-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_rle_half_float.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-tile-rle-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_rle_half_float.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-tile-rle-half-float-l2
> -fate-exr-rgb-tile-rle-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_rle_half_float.exr -pix_fmt gbrapf32le
> +fate-exr-rgb-tile-rle-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_rle_half_float.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgb-tile-raw-half-float-l1
> -fate-exr-rgb-tile-raw-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_raw_half_float.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-tile-raw-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_raw_half_float.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-tile-raw-half-float-l2
> -fate-exr-rgb-tile-raw-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_raw_half_float.exr -pix_fmt gbrapf32le
> +fate-exr-rgb-tile-raw-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_tile_raw_half_float.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-b44-half-float-12x8-l1
> -fate-exr-rgb-scanline-b44-half-float-12x8-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_b44_half_float_12x8.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-b44-half-float-12x8-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_b44_half_float_12x8.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-b44-half-float-12x8-l2
> -fate-exr-rgb-scanline-b44-half-float-12x8-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_b44_half_float_12x8.exr -pix_fmt gbrapf32le
> +fate-exr-rgb-scanline-b44-half-float-12x8-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_b44_half_float_12x8.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-pxr24-half-float-l1
> -fate-exr-rgb-scanline-pxr24-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_half_float.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-pxr24-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_half_float.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-pxr24-half-float-l2
> -fate-exr-rgb-scanline-pxr24-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_half_float.exr -pix_fmt gbrapf32le
> +fate-exr-rgb-scanline-pxr24-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_half_float.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-pxr24-float-half-l1
> -fate-exr-rgb-scanline-pxr24-float-half-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_float_half.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-pxr24-float-half-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_float_half.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-pxr24-float-half-l2
> -fate-exr-rgb-scanline-pxr24-float-half-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_float_half.exr -pix_fmt gbrapf32le
> +fate-exr-rgb-scanline-pxr24-float-half-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_float_half.exr -vf scale -pix_fmt gbrapf32le
>
> -FATE_EXR-$(call DEMDEC, IMAGE2, EXR, SCALE_FILTER) += fate-exr-rgb-scanline-pxr24-half-uint32-13x9
> +FATE_EXR += fate-exr-rgb-scanline-pxr24-half-uint32-13x9
>  fate-exr-rgb-scanline-pxr24-half-uint32-13x9: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_half_uint32_13x9.exr -pix_fmt rgb48le -vf scale
>
>  FATE_EXR += fate-exr-rgb-scanline-zip-half-float-l1
> -fate-exr-rgb-scanline-zip-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip_half_float.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-zip-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip_half_float.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-zip-half-float-l2
> -fate-exr-rgb-scanline-zip-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip_half_float.exr -pix_fmt gbrapf32le
> +fate-exr-rgb-scanline-zip-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip_half_float.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-zip1-half-float-l1
> -fate-exr-rgb-scanline-zip1-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip1_half_float.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-zip1-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip1_half_float.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-zip1-half-float-l2
> -fate-exr-rgb-scanline-zip1-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip1_half_float.exr -pix_fmt gbrapf32le
> +fate-exr-rgb-scanline-zip1-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip1_half_float.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-rle-half-float-l1
> -fate-exr-rgb-scanline-rle-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_rle_half_float.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-rle-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_rle_half_float.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-rle-half-float-l2
> -fate-exr-rgb-scanline-rle-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_rle_half_float.exr -pix_fmt gbrapf32le
> +fate-exr-rgb-scanline-rle-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_rle_half_float.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-raw-half-float-l1
> -fate-exr-rgb-scanline-raw-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_raw_half_float.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-raw-half-float-l1: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_raw_half_float.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-raw-half-float-l2
> -fate-exr-rgb-scanline-raw-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_raw_half_float.exr -pix_fmt gbrapf32le
> +fate-exr-rgb-scanline-raw-half-float-l2: CMD = framecrc -layer "VRaySamplerInfo" -i $(TARGET_SAMPLES)/exr/rgb_scanline_raw_half_float.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-b44-uint32
> -fate-exr-rgb-scanline-b44-uint32: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_b44_uint32.exr -pix_fmt rgb48le
> +fate-exr-rgb-scanline-b44-uint32: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_b44_uint32.exr -vf scale -pix_fmt rgb48le
>
>  FATE_EXR += fate-exr-rgb-scanline-pxr24-uint32
> -fate-exr-rgb-scanline-pxr24-uint32: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_uint32.exr -pix_fmt rgb48le
> +fate-exr-rgb-scanline-pxr24-uint32: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_pxr24_uint32.exr -vf scale -pix_fmt rgb48le
>
>  FATE_EXR += fate-exr-rgb-scanline-zip1-half-float-l1-zero-offsets
> -fate-exr-rgb-scanline-zip1-half-float-l1-zero-offsets: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip1_half_float_zero_offsets.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-zip1-half-float-l1-zero-offsets: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip1_half_float_zero_offsets.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-half-piz-bw
> -fate-exr-rgb-scanline-half-piz-bw: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_bw.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-half-piz-bw: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_bw.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-half-piz-color
> -fate-exr-rgb-scanline-half-piz-color: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_color.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-half-piz-color: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_color.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-half-piz-dw-t01
> -fate-exr-rgb-scanline-half-piz-dw-t01: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_dw_t01.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-half-piz-dw-t01: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_dw_t01.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-float-piz-48x32
> -fate-exr-rgb-scanline-float-piz-48x32: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_float_piz_48x32.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-float-piz-48x32: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_float_piz_48x32.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-none-negative-red
> -fate-exr-rgb-scanline-none-negative-red: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_none_negative_red.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-none-negative-red: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_none_negative_red.exr -vf scale -pix_fmt gbrpf32le
>
>
>  FATE_EXR += fate-exr-rgb-b44a-half-negative-4x4
> -fate-exr-rgb-b44a-half-negative-4x4: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_b44a_half_negative_4x4.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-b44a-half-negative-4x4: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_b44a_half_negative_4x4.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-y-tile-zip-half-12x8
> -fate-exr-y-tile-zip-half-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/y_tile_zip_half_12x8.exr -pix_fmt grayf32le
> +fate-exr-y-tile-zip-half-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/y_tile_zip_half_12x8.exr -vf scale -pix_fmt grayf32le
>
>  FATE_EXR += fate-exr-y-scanline-zip-half-12x8
> -fate-exr-y-scanline-zip-half-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/y_scanline_zip_half_12x8.exr -pix_fmt grayf32le
> +fate-exr-y-scanline-zip-half-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/y_scanline_zip_half_12x8.exr -vf scale -pix_fmt grayf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-half-piz-dw-t08
> -fate-exr-rgb-scanline-half-piz-dw-t08: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_dw_t08.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-half-piz-dw-t08: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_dw_t08.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgba-zip16-16x32-flag4
> -fate-exr-rgba-zip16-16x32-flag4: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_zip16_16x32_flag4.exr -pix_fmt gbrapf32le
> +fate-exr-rgba-zip16-16x32-flag4: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgba_zip16_16x32_flag4.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-ya-scanline-zip-half-12x8
> -fate-exr-ya-scanline-zip-half-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/ya_scanline_zip_half_12x8.exr -pix_fmt gbrapf32le
> +fate-exr-ya-scanline-zip-half-12x8: CMD = framecrc -i $(TARGET_SAMPLES)/exr/ya_scanline_zip_half_12x8.exr -vf scale -pix_fmt gbrapf32le
>
>  FATE_EXR += fate-exr-rgb-tile-half-zip
> -fate-exr-rgb-tile-half-zip: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_zip.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-tile-half-zip: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_zip.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-float-zip-dw-large
> -fate-exr-rgb-scanline-float-zip-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_float_zip_dw_large.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-float-zip-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_float_zip_dw_large.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-half-piz-dw-large
> -fate-exr-rgb-scanline-half-piz-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_dw_large.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-half-piz-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_piz_dw_large.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-half-zip-dw-large
> -fate-exr-rgb-scanline-half-zip-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_zip_dw_large.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-half-zip-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_zip_dw_large.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-uint32-piz-dw-large
> -fate-exr-rgb-scanline-uint32-piz-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_uint32_piz_dw_large.exr -pix_fmt rgb48le
> +fate-exr-rgb-scanline-uint32-piz-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_uint32_piz_dw_large.exr -vf scale -pix_fmt rgb48le
>
>  FATE_EXR += fate-exr-rgb-tile-half-piz-dw-large
> -fate-exr-rgb-tile-half-piz-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_piz_dw_large.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-tile-half-piz-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_piz_dw_large.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-tile-uint32-piz-dw-large
> -fate-exr-rgb-tile-uint32-piz-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_uint32_piz_dw_large.exr -pix_fmt rgb48le
> +fate-exr-rgb-tile-uint32-piz-dw-large: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_uint32_piz_dw_large.exr -vf scale -pix_fmt rgb48le
>
>  FATE_EXR += fate-exr-rgb-scanline-half-zip-dw-outside
> -fate-exr-rgb-scanline-half-zip-dw-outside: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_zip_dw_outside.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-half-zip-dw-outside: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_half_zip_dw_outside.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-tile-half-zip-dw-outside
> -fate-exr-rgb-tile-half-zip-dw-outside: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_zip_dw_outside.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-tile-half-zip-dw-outside: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_tile_half_zip_dw_outside.exr -vf scale -pix_fmt gbrpf32le
>
>  FATE_EXR += fate-exr-rgb-scanline-zip-half-0x0-0xFFFF
> -fate-exr-rgb-scanline-zip-half-0x0-0xFFFF: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip_half_float_0x0_to_0xFFFF.exr -pix_fmt gbrpf32le
> +fate-exr-rgb-scanline-zip-half-0x0-0xFFFF: CMD = framecrc -i $(TARGET_SAMPLES)/exr/rgb_scanline_zip_half_float_0x0_to_0xFFFF.exr -vf scale -pix_fmt gbrpf32le
>
> -FATE_EXR-$(call DEMDEC, IMAGE2, EXR) += $(FATE_EXR)
> +FATE_EXR-$(call DEMDEC, IMAGE2, EXR, SCALE_FILTER) += $(FATE_EXR)
>
>  FATE_IMAGE_FRAMECRC += $(FATE_EXR-yes)
>  fate-exr: $(FATE_EXR-yes)
> --
> 2.40.1
>
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel@ffmpeg.org
> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>
> To unsubscribe, visit link above, or email
> ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Confirming this fixes fate-exr-* on a big-endian POWER7.

Thanks,
Sean McGovern.
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

[FFmpeg-devel] [PATCH 7/7] fate/video: Only use bitexact IDCT in avid meridian

[Andreas Rheinhardt]

Precludes the usage of the altivec IDCT which fixes
the avid-meridian FATE test on ppc64be here.

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
---
 tests/fate/video.mak | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tests/fate/video.mak b/tests/fate/video.mak
index 8f51a42077..9ad39e21ee 100644
--- a/tests/fate/video.mak
+++ b/tests/fate/video.mak
@@ -48,7 +48,7 @@ FATE_VIDEO-$(call FRAMECRC, AVI, AVRN) += fate-avid-interlaced
 fate-avid-interlaced: CMD = framecrc -i $(TARGET_SAMPLES)/avid/avid_ntsc_interlaced.avi
 
 FATE_VIDEO-$(call FRAMECRC, MOV, MJPEG) += fate-avid-meridian
-fate-avid-meridian: CMD = framecrc -i $(TARGET_SAMPLES)/avid/avidmeridianntsc.mov
+fate-avid-meridian: CMD = framecrc -bitexact -i $(TARGET_SAMPLES)/avid/avidmeridianntsc.mov
 
 FATE_VIDEO-$(call FRAMECRC, BETHSOFTVID, BETHSOFTVID, ARESAMPLE_FILTER SCALE_FILTER) += fate-bethsoft-vid
 fate-bethsoft-vid: CMD = framecrc -i $(TARGET_SAMPLES)/bethsoft-vid/ANIM0001.VID -t 5 -pix_fmt rgb24 -vf scale -af aresample
-- 
2.40.1

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

[FFmpeg-devel] [PATCH 8/8] fate/filter-video: Always use little endian pixel format

[Andreas Rheinhardt]

Fixes filter-metadata-signalstats-yuv420p10 on BE arches.

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
---
 tests/fate/filter-video.mak | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tests/fate/filter-video.mak b/tests/fate/filter-video.mak
index ee9f0f5e40..681cddc33e 100644
--- a/tests/fate/filter-video.mak
+++ b/tests/fate/filter-video.mak
@@ -692,7 +692,7 @@ fate-filter-metadata-freezedetect: CMD = run $(FILTER_METADATA_COMMAND) "sws_fla
 SIGNALSTATS_DEPS = LAVFI_INDEV COLOR_FILTER SCALE_FILTER SIGNALSTATS_FILTER
 FATE_METADATA_FILTER-$(call ALLYES, $(SIGNALSTATS_DEPS)) += fate-filter-metadata-signalstats-yuv420p fate-filter-metadata-signalstats-yuv420p10
 fate-filter-metadata-signalstats-yuv420p: CMD = run $(FILTER_METADATA_COMMAND) "sws_flags=+accurate_rnd+bitexact;color=white:duration=1:r=1,signalstats"
-fate-filter-metadata-signalstats-yuv420p10: CMD = run $(FILTER_METADATA_COMMAND) "sws_flags=+accurate_rnd+bitexact;color=white:duration=1:r=1,format=yuv420p10,signalstats"
+fate-filter-metadata-signalstats-yuv420p10: CMD = run $(FILTER_METADATA_COMMAND) "sws_flags=+accurate_rnd+bitexact;color=white:duration=1:r=1,format=yuv420p10le,signalstats"
 
 SILENCEDETECT_DEPS = LAVFI_INDEV FILE_PROTOCOL AMOVIE_FILTER TTA_DEMUXER TTA_DECODER SILENCEDETECT_FILTER
 FATE_METADATA_FILTER-$(call ALLYES, $(SILENCEDETECT_DEPS)) += fate-filter-metadata-silencedetect
-- 
2.40.1

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

[FFmpeg-devel] [PATCH 9/9] fate/filter-video: Insert scale, format filters in filter-yadif tests

[Andreas Rheinhardt]

The format and the first scale filter ensures that the filter
processing actually happens in high bit depth; the second
scale filter is only necessary for big endian arches.

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
---
 tests/fate/filter-video.mak | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/tests/fate/filter-video.mak b/tests/fate/filter-video.mak
index 681cddc33e..d8d29866b8 100644
--- a/tests/fate/filter-video.mak
+++ b/tests/fate/filter-video.mak
@@ -25,8 +25,8 @@ fate-filter-yadif-mode0: CMD = framecrc -ec 0 -flags bitexact -idct simple -i $(
 fate-filter-yadif-mode1: CMD = framecrc -ec 0 -flags bitexact -idct simple -i $(TARGET_SAMPLES)/mpeg2/mpeg2_field_encoding.ts -frames:v 59 -vf yadif=1
 
 FATE_YADIF-$(call FILTERDEMDEC, YADIF SCALE, MPEGTS, MPEG2VIDEO) += fate-filter-yadif10 fate-filter-yadif16
-fate-filter-yadif10: CMD = framecrc -ec 0 -flags bitexact -idct simple -i $(TARGET_SAMPLES)/mpeg2/mpeg2_field_encoding.ts -flags bitexact -pix_fmt yuv420p10le -frames:v 30 -vf scale,yadif=0
-fate-filter-yadif16: CMD = framecrc -ec 0 -flags bitexact -idct simple -i $(TARGET_SAMPLES)/mpeg2/mpeg2_field_encoding.ts -flags bitexact -pix_fmt yuv420p16le -frames:v 30 -vf scale,yadif=0
+fate-filter-yadif10: CMD = framecrc -ec 0 -flags bitexact -idct simple -i $(TARGET_SAMPLES)/mpeg2/mpeg2_field_encoding.ts -flags bitexact -pix_fmt yuv420p10le -frames:v 30 -vf scale,format=yuv420p10,yadif=0,scale
+fate-filter-yadif16: CMD = framecrc -ec 0 -flags bitexact -idct simple -i $(TARGET_SAMPLES)/mpeg2/mpeg2_field_encoding.ts -flags bitexact -pix_fmt yuv420p16le -frames:v 30 -vf scale,format=yuv420p16,yadif=0,scale
 
 FATE_FILTER_SAMPLES-yes += $(FATE_YADIF-yes)
 
-- 
2.40.1

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

[FFmpeg-devel] [PATCH v2 9/9] fate/filter-video: Insert scale, format filters in filter-yadif, bwdif10

[Andreas Rheinhardt]

The format and the first scale filter ensures that the filter
processing actually happens in high bit depth; the second
scale filter is only necessary for big endian arches.

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
---
 tests/fate/filter-video.mak | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/tests/fate/filter-video.mak b/tests/fate/filter-video.mak
index 681cddc33e..7f8dc3aa27 100644
--- a/tests/fate/filter-video.mak
+++ b/tests/fate/filter-video.mak
@@ -16,7 +16,7 @@ fate-filter-bwdif-mode0: CMD = framecrc -ec 0 -flags bitexact -idct simple -i $(
 fate-filter-bwdif-mode1: CMD = framecrc -ec 0 -flags bitexact -idct simple -i $(TARGET_SAMPLES)/mpeg2/mpeg2_field_encoding.ts -frames:v 59 -vf bwdif=send_field
 
 FATE_BWDIF-$(call FILTERDEMDEC, BWDIF SCALE, MPEGTS, MPEG2VIDEO) += fate-filter-bwdif10
-fate-filter-bwdif10: CMD = framecrc -ec 0 -flags bitexact -idct simple -i $(TARGET_SAMPLES)/mpeg2/mpeg2_field_encoding.ts -flags bitexact -pix_fmt yuv420p10le -frames:v 30 -vf scale,bwdif=0
+fate-filter-bwdif10: CMD = framecrc -ec 0 -flags bitexact -idct simple -i $(TARGET_SAMPLES)/mpeg2/mpeg2_field_encoding.ts -flags bitexact -pix_fmt yuv420p10le -frames:v 30 -vf scale,format=yuv420p10,bwdif=0,scale
 
 FATE_FILTER_SAMPLES-yes += $(FATE_BWDIF-yes)
 
@@ -25,8 +25,8 @@ fate-filter-yadif-mode0: CMD = framecrc -ec 0 -flags bitexact -idct simple -i $(
 fate-filter-yadif-mode1: CMD = framecrc -ec 0 -flags bitexact -idct simple -i $(TARGET_SAMPLES)/mpeg2/mpeg2_field_encoding.ts -frames:v 59 -vf yadif=1
 
 FATE_YADIF-$(call FILTERDEMDEC, YADIF SCALE, MPEGTS, MPEG2VIDEO) += fate-filter-yadif10 fate-filter-yadif16
-fate-filter-yadif10: CMD = framecrc -ec 0 -flags bitexact -idct simple -i $(TARGET_SAMPLES)/mpeg2/mpeg2_field_encoding.ts -flags bitexact -pix_fmt yuv420p10le -frames:v 30 -vf scale,yadif=0
-fate-filter-yadif16: CMD = framecrc -ec 0 -flags bitexact -idct simple -i $(TARGET_SAMPLES)/mpeg2/mpeg2_field_encoding.ts -flags bitexact -pix_fmt yuv420p16le -frames:v 30 -vf scale,yadif=0
+fate-filter-yadif10: CMD = framecrc -ec 0 -flags bitexact -idct simple -i $(TARGET_SAMPLES)/mpeg2/mpeg2_field_encoding.ts -flags bitexact -pix_fmt yuv420p10le -frames:v 30 -vf scale,format=yuv420p10,yadif=0,scale
+fate-filter-yadif16: CMD = framecrc -ec 0 -flags bitexact -idct simple -i $(TARGET_SAMPLES)/mpeg2/mpeg2_field_encoding.ts -flags bitexact -pix_fmt yuv420p16le -frames:v 30 -vf scale,format=yuv420p16,yadif=0,scale
 
 FATE_FILTER_SAMPLES-yes += $(FATE_YADIF-yes)
 
-- 
2.40.1

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

[FFmpeg-devel] [PATCH 10/12] fate/fits: Fix tests on BE

[Andreas Rheinhardt]

The fits decoder decodes to native pixel formats; so
the fitsdec-gbrap16be fate test failed on BE despite
its name because the reference file is LE.
This patch fixes this by forcing a pixel format;
the forced pixel format is BE, causing a change
in the reference file.
The fitsdec-gbrp16be test was not affected, because
its source file (lena-rgb48.png from tne FATE suite)
is actually biendian (as if someone had multiplied
8bit content by 257...).

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
---
 tests/fate/fits.mak              | 2 +-
 tests/ref/fate/fitsdec-gbrap16be | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/tests/fate/fits.mak b/tests/fate/fits.mak
index d85946bc1a..b83900aaee 100644
--- a/tests/fate/fits.mak
+++ b/tests/fate/fits.mak
@@ -28,7 +28,7 @@ fate-fitsdec-multi: tests/data/fits-multi.fits
 fate-fitsdec-multi: CMD = framecrc -i $(TARGET_PATH)/tests/data/fits-multi.fits -pix_fmt gbrap
 
 fate-fitsdec%: PIXFMT = $(word 3, $(subst -, ,$(@)))
-fate-fitsdec%: CMD = transcode image2 $(TARGET_SAMPLES)/png1/lena-$(fits-png-map-$(PIXFMT)).png fits "-vf scale -pix_fmt $(PIXFMT)"
+fate-fitsdec%: CMD = transcode image2 $(TARGET_SAMPLES)/png1/lena-$(fits-png-map-$(PIXFMT)).png fits "-vf scale -pix_fmt $(PIXFMT)" "-vf scale -pix_fmt $(PIXFMT)"
 
 FATE_FITS_DEC_PIXFMT = gray gbrp gbrp16be gbrap16be
 FATE_FITS_DEC-$(call TRANSCODE, FITS, FITS, IMAGE2_DEMUXER PNG_DECODER SCALE_FILTER) += $(FATE_FITS_DEC_PIXFMT:%=fate-fitsdec-%)
diff --git a/tests/ref/fate/fitsdec-gbrap16be b/tests/ref/fate/fitsdec-gbrap16be
index 1174a0f1d8..e57a878845 100644
--- a/tests/ref/fate/fitsdec-gbrap16be
+++ b/tests/ref/fate/fitsdec-gbrap16be
@@ -5,4 +5,4 @@
 #codec_id 0: rawvideo
 #dimensions 0: 128x128
 #sar 0: 0/1
-0,          0,          0,        1,   131072, 0x487894b2
+0,          0,          0,        1,   131072, 0xebb194b2
-- 
2.40.1

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

[FFmpeg-devel] [PATCH 11/12] avcodec/vlc, bitstream: Fix multi VLC with uint8_t syms on BE

[Andreas Rheinhardt]

VLC_MULTI_ELEM contains an uint8_t array that is supposed
to be treated as an array of uint16_t when the used symbols
have a size of two; otherwise it should be treated as just
an array of uint8_t, but it was not always treated that way:

vlc_multi_gen() initialized the first entry of the array
by writing the symbol via AV_WN16; on big endian systems,
the intended value was instead written into the second entry
of the array (where it would likely be overwritten lateron
during initialization).

read_vlc_multi() also treated this case incorrectly: In case
the code is so long that it needs a classical multi-stage lookup,
the symbol has been written to the destination as if via AV_WN16.
On little endian systems, this sets the correct first symbol and
clobbers (zeroes) the next one, but the next one will be overwritten
lateron anyway, so it won't be recognized. But on big-endian systems,
the first symbol will be set to zero and the actually read symbol
will be put into the slot for the next one (where it will be overwritten
lateron).

This commit fixes this; this fixes the magicyuv and utvideo FATE-tests
on big endian arches.

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
---
 libavcodec/bitstream_template.h | 8 ++++++--
 libavcodec/get_bits.h           | 3 ++-
 libavcodec/magicyuv.c           | 2 +-
 libavcodec/utvideodec.c         | 2 +-
 libavcodec/vlc.c                | 5 ++++-
 5 files changed, 14 insertions(+), 6 deletions(-)

diff --git a/libavcodec/bitstream_template.h b/libavcodec/bitstream_template.h
index 4f3d07275f..c8e4a5131e 100644
--- a/libavcodec/bitstream_template.h
+++ b/libavcodec/bitstream_template.h
@@ -536,7 +536,8 @@ static inline int BS_FUNC(read_vlc)(BSCTX *bc, const VLCElem *table,
 static inline int BS_FUNC(read_vlc_multi)(BSCTX *bc, uint8_t dst[8],
                                           const VLC_MULTI_ELEM *const Jtable,
                                           const VLCElem *const table,
-                                          const int bits, const int max_depth)
+                                          const int bits, const int max_depth,
+                                          const int symbols_size)
 {
     unsigned idx = BS_FUNC(peek)(bc, bits);
     int ret, nb_bits, code, n = Jtable[idx].len;
@@ -554,7 +555,10 @@ static inline int BS_FUNC(read_vlc_multi)(BSCTX *bc, uint8_t dst[8],
                 code = BS_FUNC(priv_set_idx)(bc, code, &n, &nb_bits, table);
             }
         }
-        AV_WN16(dst, code);
+        if (symbols_size == 1)
+            *dst = code;
+        else
+            AV_WN16(dst, code);
         ret = n > 0;
     }
     BS_FUNC(priv_skip_remaining)(bc, n);
diff --git a/libavcodec/get_bits.h b/libavcodec/get_bits.h
index cfcf97c021..fe2f6378b4 100644
--- a/libavcodec/get_bits.h
+++ b/libavcodec/get_bits.h
@@ -667,7 +667,8 @@ static av_always_inline int get_vlc2(GetBitContext *s, const VLCElem *table,
 static inline int get_vlc_multi(GetBitContext *s, uint8_t *dst,
                                 const VLC_MULTI_ELEM *const Jtable,
                                 const VLCElem *const table,
-                                const int bits, const int max_depth)
+                                const int bits, const int max_depth,
+                                const int symbols_size)
 {
     dst[0] = get_vlc2(s, table, bits, max_depth);
     return 1;
diff --git a/libavcodec/magicyuv.c b/libavcodec/magicyuv.c
index 3f6348b531..4f30493626 100644
--- a/libavcodec/magicyuv.c
+++ b/libavcodec/magicyuv.c
@@ -124,7 +124,7 @@ static void magicyuv_median_pred16(uint16_t *dst, const uint16_t *src1,
     x = 0; \
     for (; CACHED_BITSTREAM_READER && x < width-c && get_bits_left(&gb) > 0;) {\
         ret = get_vlc_multi(&gb, (uint8_t *)dst + x * b, multi, \
-                            vlc, vlc_bits, 3); \
+                            vlc, vlc_bits, 3, b); \
         if (ret <= 0) \
             return AVERROR_INVALIDDATA; \
         x += ret; \
diff --git a/libavcodec/utvideodec.c b/libavcodec/utvideodec.c
index ce5d00f7af..0c2e67e282 100644
--- a/libavcodec/utvideodec.c
+++ b/libavcodec/utvideodec.c
@@ -120,7 +120,7 @@ static int build_huff(UtvideoContext *c, const uint8_t *src, VLC *vlc,
     i = 0; \
     for (; CACHED_BITSTREAM_READER && i < width-end && get_bits_left(&gb) > 0;) {\
         ret = get_vlc_multi(&gb, (uint8_t *)buf + i * b, multi.table, \
-                            vlc.table, VLC_BITS, 3); \
+                            vlc.table, VLC_BITS, 3, b); \
         if (ret > 0) \
             i += ret; \
         if (ret <= 0) \
diff --git a/libavcodec/vlc.c b/libavcodec/vlc.c
index 78510e30d6..e01cc41689 100644
--- a/libavcodec/vlc.c
+++ b/libavcodec/vlc.c
@@ -499,7 +499,10 @@ static int vlc_multi_gen(VLC_MULTI_ELEM *table, const VLC *single,
     for (int j = 0; j < 1<<numbits; j++) {
         table[j].len = single->table[j].len;
         table[j].num = single->table[j].len > 0 ? 1 : 0;
-        AV_WN16(table[j].val, single->table[j].sym);
+        if (is16bit)
+            AV_WN16(table[j].val, single->table[j].sym);
+        else
+            table[j].val[0] = single->table[j].sym;
     }
 
     add_level(table, is16bit, nb_codes, numbits, buf,
-- 
2.40.1

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

[FFmpeg-devel] [PATCH 12/12] avcodec/vlc: Use union of uint8_t and uint16_t in VLC_MULTI_ELEM

[Andreas Rheinhardt]

It is more natural and simplifies writing these arrays.

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
---
 libavcodec/bitstream_template.h | 2 +-
 libavcodec/vlc.c                | 8 ++++----
 libavcodec/vlc.h                | 5 ++++-
 3 files changed, 9 insertions(+), 6 deletions(-)

diff --git a/libavcodec/bitstream_template.h b/libavcodec/bitstream_template.h
index c8e4a5131e..bbb8dfa555 100644
--- a/libavcodec/bitstream_template.h
+++ b/libavcodec/bitstream_template.h
@@ -542,7 +542,7 @@ static inline int BS_FUNC(read_vlc_multi)(BSCTX *bc, uint8_t dst[8],
     unsigned idx = BS_FUNC(peek)(bc, bits);
     int ret, nb_bits, code, n = Jtable[idx].len;
     if (Jtable[idx].num) {
-        AV_COPY64U(dst, Jtable[idx].val);
+        AV_COPY64U(dst, Jtable[idx].val8);
         ret = Jtable[idx].num;
     } else {
         code = table[idx].sym;
diff --git a/libavcodec/vlc.c b/libavcodec/vlc.c
index e01cc41689..e89866e869 100644
--- a/libavcodec/vlc.c
+++ b/libavcodec/vlc.c
@@ -440,8 +440,8 @@ static void add_level(VLC_MULTI_ELEM *table, const int is16bit,
             code = curcode + (buf[t].code >> curlen);
             newlimit = curlimit - l;
             l  += curlen;
-            if (is16bit) AV_WN16(info.val+2*curlevel, sym);
-            else info.val[curlevel] = sym&0xFF;
+            if (is16bit) info.val16[curlevel] = sym;
+            else info.val8[curlevel] = sym&0xFF;
 
             if (curlevel) { // let's not add single entries
                 uint32_t val = code >> (32 - numbits);
@@ -500,9 +500,9 @@ static int vlc_multi_gen(VLC_MULTI_ELEM *table, const VLC *single,
         table[j].len = single->table[j].len;
         table[j].num = single->table[j].len > 0 ? 1 : 0;
         if (is16bit)
-            AV_WN16(table[j].val, single->table[j].sym);
+            table[j].val16[0] = single->table[j].sym;
         else
-            table[j].val[0] = single->table[j].sym;
+            table[j].val8[0]  = single->table[j].sym;
     }
 
     add_level(table, is16bit, nb_codes, numbits, buf,
diff --git a/libavcodec/vlc.h b/libavcodec/vlc.h
index 0cc106c499..bf7b0e65b4 100644
--- a/libavcodec/vlc.h
+++ b/libavcodec/vlc.h
@@ -40,7 +40,10 @@ typedef struct VLC {
 } VLC;
 
 typedef struct VLC_MULTI_ELEM {
-    uint8_t val[VLC_MULTI_MAX_SYMBOLS];
+    union {
+        uint8_t   val8[VLC_MULTI_MAX_SYMBOLS];
+        uint16_t val16[VLC_MULTI_MAX_SYMBOLS / 2];
+    };
     int8_t len; // -31,32
     uint8_t num;
 } VLC_MULTI_ELEM;
-- 
2.40.1

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

[FFmpeg-devel] [PATCH v2 12/12] avcodec/vlc: Use union of uint8_t and uint16_t in VLC_MULTI_ELEM

[Andreas Rheinhardt]

It is more natural and simplifies writing these arrays.

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
---
 libavcodec/bitstream_template.h |  2 +-
 libavcodec/vlc.c                | 10 +++++-----
 libavcodec/vlc.h                |  5 ++++-
 3 files changed, 10 insertions(+), 7 deletions(-)

diff --git a/libavcodec/bitstream_template.h b/libavcodec/bitstream_template.h
index c8e4a5131e..bbb8dfa555 100644
--- a/libavcodec/bitstream_template.h
+++ b/libavcodec/bitstream_template.h
@@ -542,7 +542,7 @@ static inline int BS_FUNC(read_vlc_multi)(BSCTX *bc, uint8_t dst[8],
     unsigned idx = BS_FUNC(peek)(bc, bits);
     int ret, nb_bits, code, n = Jtable[idx].len;
     if (Jtable[idx].num) {
-        AV_COPY64U(dst, Jtable[idx].val);
+        AV_COPY64U(dst, Jtable[idx].val8);
         ret = Jtable[idx].num;
     } else {
         code = table[idx].sym;
diff --git a/libavcodec/vlc.c b/libavcodec/vlc.c
index e01cc41689..ee09d96fd6 100644
--- a/libavcodec/vlc.c
+++ b/libavcodec/vlc.c
@@ -440,8 +440,8 @@ static void add_level(VLC_MULTI_ELEM *table, const int is16bit,
             code = curcode + (buf[t].code >> curlen);
             newlimit = curlimit - l;
             l  += curlen;
-            if (is16bit) AV_WN16(info.val+2*curlevel, sym);
-            else info.val[curlevel] = sym&0xFF;
+            if (is16bit) info.val16[curlevel] = sym;
+            else info.val8[curlevel] = sym&0xFF;
 
             if (curlevel) { // let's not add single entries
                 uint32_t val = code >> (32 - numbits);
@@ -468,7 +468,7 @@ static int vlc_multi_gen(VLC_MULTI_ELEM *table, const VLC *single,
 {
     int minbits, maxbits, max;
     unsigned count[VLC_MULTI_MAX_SYMBOLS-1] = { 0, };
-    VLC_MULTI_ELEM info = { { 0, }, 0, 0, };
+    VLC_MULTI_ELEM info = { 0 };
     int count0 = 0;
 
     for (int j = 0; j < 1<<numbits; j++) {
@@ -500,9 +500,9 @@ static int vlc_multi_gen(VLC_MULTI_ELEM *table, const VLC *single,
         table[j].len = single->table[j].len;
         table[j].num = single->table[j].len > 0 ? 1 : 0;
         if (is16bit)
-            AV_WN16(table[j].val, single->table[j].sym);
+            table[j].val16[0] = single->table[j].sym;
         else
-            table[j].val[0] = single->table[j].sym;
+            table[j].val8[0]  = single->table[j].sym;
     }
 
     add_level(table, is16bit, nb_codes, numbits, buf,
diff --git a/libavcodec/vlc.h b/libavcodec/vlc.h
index 0cc106c499..bf7b0e65b4 100644
--- a/libavcodec/vlc.h
+++ b/libavcodec/vlc.h
@@ -40,7 +40,10 @@ typedef struct VLC {
 } VLC;
 
 typedef struct VLC_MULTI_ELEM {
-    uint8_t val[VLC_MULTI_MAX_SYMBOLS];
+    union {
+        uint8_t   val8[VLC_MULTI_MAX_SYMBOLS];
+        uint16_t val16[VLC_MULTI_MAX_SYMBOLS / 2];
+    };
     int8_t len; // -31,32
     uint8_t num;
 } VLC_MULTI_ELEM;
-- 
2.40.1

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".

Re: [FFmpeg-devel] [PATCH 1/5] avcodec/ppc/hpeldsp_altivec: Fix left-shift of negative number

[Andreas Rheinhardt]

Andreas Rheinhardt:
> It is UB and affected e.g. the vp5 and vp61 FATE tests:
> https://fate.ffmpeg.org/report.cgi?time=20240327083327&slot=ppc-linux-gcc-13.2-ubsan-altivec-qemu
> 
> Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
> ---
>  libavcodec/ppc/hpeldsp_altivec.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/libavcodec/ppc/hpeldsp_altivec.c b/libavcodec/ppc/hpeldsp_altivec.c
> index a531b6b6ec..4bf6b28ed6 100644
> --- a/libavcodec/ppc/hpeldsp_altivec.c
> +++ b/libavcodec/ppc/hpeldsp_altivec.c
> @@ -41,9 +41,9 @@ void ff_put_pixels16_altivec(uint8_t *block, const uint8_t *pixels, ptrdiff_t li
>      register vector unsigned char pixelsv1D;
>  
>      int i;
> -    register ptrdiff_t line_size_2 = line_size << 1;
> +    register ptrdiff_t line_size_2 = line_size * (1 << 1);
>      register ptrdiff_t line_size_3 = line_size + line_size_2;
> -    register ptrdiff_t line_size_4 = line_size << 2;
> +    register ptrdiff_t line_size_4 = line_size * (1 << 2);
>  
>  // hand-unrolling the loop by 4 gains about 15%
>  // mininum execution time goes from 74 to 60 cycles

Will apply the remainder of this patchset tomorrow unless there are
objections.

- Andreas

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".