Builder ffmpeg-solaris10-i386 Build #13920
Results:
Failed shell_2 shell_3 shell_4 shell_5
SourceStamp:
| Project | ffmpeg |
| Repository | https://git.ffmpeg.org/ffmpeg.git |
| Branch | master |
| Revision | d9e2239f3c5e17d908db0b79f68ff06ee1867828 |
| Got Revision | d9e2239f3c5e17d908db0b79f68ff06ee1867828 |
| Changes | 7 changes |
BuildSlave:
unstable10xReason:
The SingleBranchScheduler scheduler named 'schedule-ffmpeg-solaris10-i386' triggered this build
Steps and Logfiles:
-
git update ( 4 secs )
-
shell 'gsed -i ...' ( 0 secs )
-
shell_1 'gsed -i ...' ( 0 secs )
-
shell_2 'gsed -i ...' failed ( 0 secs )
-
shell_3 './configure --samples="../../../ffmpeg/fate-suite" ...' failed ( 6 secs )
-
shell_4 'gmake fate-rsync' failed ( 0 secs )
-
shell_5 '../../../ffmpeg/fate.sh ../../../ffmpeg/fate_config.sh' failed ( 0 secs )
Build Properties:
| Name | Value | Source |
|---|---|---|
| branch | master | Build |
| builddir | /export/home/buildbot/slave/ffmpeg-solaris10-i386 | slave |
| buildername | ffmpeg-solaris10-i386 | Builder |
| buildnumber | 13920 | Build |
| codebase | Build | |
| got_revision | d9e2239f3c5e17d908db0b79f68ff06ee1867828 | Git |
| project | ffmpeg | Build |
| repository | https://git.ffmpeg.org/ffmpeg.git | Build |
| revision | d9e2239f3c5e17d908db0b79f68ff06ee1867828 | Build |
| scheduler | schedule-ffmpeg-solaris10-i386 | Scheduler |
| slavename | unstable10x | BuildSlave |
| workdir | /export/home/buildbot/slave/ffmpeg-solaris10-i386 | slave (deprecated) |
Forced Build Properties:
| Name | Label | Value |
|---|
Responsible Users:
- DROOdotFOOdrew@axol.io
Timing:
| Start | Sat Jun 6 19:53:44 2026 |
| End | Sat Jun 6 19:53:57 2026 |
| Elapsed | 12 secs |
All Changes:
:
Change #269648
Category ffmpeg Changed by DROOdotFOO <drew@axol.io> Changed at Sat 06 Jun 2026 19:38:40 Repository https://git.ffmpeg.org/ffmpeg.git Project ffmpeg Branch master Revision e0fa6412408fd776776d9571e4ca79f630a13e14 Comments
swscale/aarch64/yuv2rgb_neon: chroma-preserve compute_rgb Macro writes per-luma sums into the destination registers, leaving v20-v25 (chroma -> RGB offsets) intact for the 2-line callers. Takes bare register names. compute_rgba and compute_rgba_alpha follow suit. Single-row callers reload v20-v25 each iteration via chroma_to_rgb_offsets, so the change is a no-op for them: Apple M1 width=1920 mean -0.54% across 55 paths, within bench noise. Co-authored-by: Ramiro Polla <ramiro.polla@gmail.com> Signed-off-by: DROOdotFOO <drew@axol.io>
Changed files
- libswscale/aarch64/yuv2rgb_neon.S
Change #269649
Category ffmpeg Changed by DROOdotFOO <drew@axol.io> Changed at Sat 06 Jun 2026 19:38:40 Repository https://git.ffmpeg.org/ffmpeg.git Project ffmpeg Branch master Revision 8dbc7299509f7e30530ded2ea868249e05ff54e8 Comments
swscale/aarch64/yuv2rgb_neon: name registers the loop body. .text byte-identical. Co-authored-by: Ramiro Polla <ramiro.polla@gmail.com> Signed-off-by: DROOdotFOO <drew@axol.io>
Changed files
- libswscale/aarch64/yuv2rgb_neon.S
Change #269650
Category ffmpeg Changed by DROOdotFOO <drew@axol.io> Changed at Sat 06 Jun 2026 19:38:40 Repository https://git.ffmpeg.org/ffmpeg.git Project ffmpeg Branch master Revision 11b1721b11becf9692d855802b990cb032de6f3c Comments
swscale/aarch64/yuv2rgb_neon: reorder params, unify signature Pass src[]/srcStride[] as arrays (x5/x6), move y_offset/y_coeff into register args (w2/w3). Only int-after-pointer stack args remain, so Apple and AAPCS64 lay them out identically; every __APPLE__ is gone. nv12/nv21/yuv420p/yuv422p/yuva420p share one signature. Co-authored-by: Ramiro Polla <ramiro.polla@gmail.com> Signed-off-by: DROOdotFOO <drew@axol.io>
Changed files
- libswscale/aarch64/swscale_unscaled.c
- libswscale/aarch64/yuv2rgb_neon.S
Change #269651
Category ffmpeg Changed by DROOdotFOO <drew@axol.io> Changed at Sat 06 Jun 2026 19:38:40 Repository https://git.ffmpeg.org/ffmpeg.git Project ffmpeg Branch master Revision 4bfe7efd0c3ffa35a39d34af25c4c64ac484fe95 Comments
swscale/aarch64/yuv2rgb_neon: 2 lines at a time, packed RGB Vertically-subsampled inputs (nv12, nv21, yuv420p) share a chroma row across two output rows; compute the chroma -> RGB offsets once and apply to both luma rows. Covers argb/rgba/abgr/bgra/rgb24/bgr24. Test Name A55-gcc M1-clang A76-gcc ---------------------------------------------------------------------------------------- nv12_to_argb_neon 21647.2 (1.16x) 40.1 (1.24x) 13813.3 (1.16x) nv12_to_rgba_neon 21653.7 (1.16x) 40.8 (1.32x) 14105.0 (1.13x) nv12_to_abgr_neon 22122.2 (1.15x) 40.3 (1.27x) 14100.2 (1.16x) nv12_to_bgra_neon 22121.6 (1.15x) 39.6 (1.24x) 14125.9 (1.16x) nv12_to_rgb24_neon 19842.0 (1.18x) 33.4 (1.28x) 12868.9 (1.17x) nv12_to_bgr24_neon 20318.0 (1.18x) 34.6 (1.23x) 12868.8 (1.17x) nv21_to_argb_neon 21648.5 (1.16x) 41.0 (1.29x) 13978.5 (1.14x) nv21_to_rgba_neon 21653.0 (1.16x) 41.3 (1.21x) 14173.5 (1.11x) nv21_to_abgr_neon 22120.6 (1.15x) 41.1 (1.20x) 14505.4 (1.14x) nv21_to_bgra_neon 22120.8 (1.15x) 41.0 (1.22x) 14520.1 (1.14x) nv21_to_rgb24_neon 19830.5 (1.19x) 35.1 (1.28x) 12832.4 (1.17x) nv21_to_bgr24_neon 20317.1 (1.18x) 34.6 (1.27x) 12833.1 (1.17x) yuv420p_to_argb_neon 21450.2 (1.15x) 39.2 (1.19x) 14118.3 (1.12x) yuv420p_to_rgba_neon 21447.2 (1.15x) 38.8 (1.24x) 14326.0 (1.14x) yuv420p_to_abgr_neon 21927.0 (1.15x) 38.9 (1.25x) 14826.6 (1.13x) yuv420p_to_bgra_neon 21930.8 (1.15x) 41.4 (1.18x) 14822.9 (1.13x) yuv420p_to_rgb24_neon 19365.5 (1.17x) 33.5 (1.25x) 13291.8 (1.16x) yuv420p_to_bgr24_neon 19848.8 (1.16x) 34.1 (1.35x) 13292.8 (1.16x) Co-authored-by: Ramiro Polla <ramiro.polla@gmail.com> Signed-off-by: DROOdotFOO <drew@axol.io>
Changed files
- libswscale/aarch64/yuv2rgb_neon.S
Change #269652
Category ffmpeg Changed by DROOdotFOO <drew@axol.io> Changed at Sat 06 Jun 2026 19:38:40 Repository https://git.ffmpeg.org/ffmpeg.git Project ffmpeg Branch master Revision dad212060c777ad21a8f785ba11f1add3a15e432 Comments
swscale/aarch64/yuv2rgb_neon: 2 lines at a time, gbrp Six dst pointers exhaust the caller-saved registers; spill x19/x20. yuva420p_to_gbrp_neon is routed through the yuv420p path by the dispatcher (gbrp has no alpha channel). Test Name A55-gcc M1-clang A76-gcc ---------------------------------------------------------------------------------------- nv12_to_gbrp_neon 20017.8 (1.15x) 32.8 (1.34x) 10658.0 (1.27x) nv21_to_gbrp_neon 20020.9 (1.15x) 32.5 (1.36x) 10691.1 (1.26x) yuv420p_to_gbrp_neon 19856.3 (1.14x) 31.4 (1.34x) 10348.0 (1.37x) yuva420p_to_gbrp_neon 19859.8 (1.14x) 30.9 (1.27x) 10350.9 (1.37x) Co-authored-by: Ramiro Polla <ramiro.polla@gmail.com> Signed-off-by: DROOdotFOO <drew@axol.io>
Changed files
- libswscale/aarch64/yuv2rgb_neon.S
Change #269653
Category ffmpeg Changed by DROOdotFOO <drew@axol.io> Changed at Sat 06 Jun 2026 19:38:40 Repository https://git.ffmpeg.org/ffmpeg.git Project ffmpeg Branch master Revision 4b6f7c2a05f456cc27241f6ed1e44e77aa7861d6 Comments
swscale/aarch64/yuv2rgb_neon: 2 lines at a time, rgb16 pack_rgb16_2l uses v26-v29 as scratch (luma temps, dead by then) instead of v20-v23, so v20-v25 chroma survives the pack step. A .error trips if yuva420p hits rgb16 (v28/v29 would clobber alpha); the dispatcher routes that combination through yuv420p anyway. Test Name A55-gcc M1-clang A76-gcc ---------------------------------------------------------------------------------------- nv12_to_rgb565le_neon 28531.9 (1.12x) 46.8 (1.28x) 19252.9 (1.09x) nv12_to_bgr565le_neon 29018.1 (1.12x) 48.1 (1.17x) 19252.0 (1.09x) nv12_to_rgb555le_neon 28531.3 (1.12x) 47.2 (1.24x) 19253.6 (1.09x) nv12_to_bgr555le_neon 29012.1 (1.12x) 45.8 (1.22x) 19252.5 (1.09x) nv21_to_rgb565le_neon 28532.3 (1.12x) 48.4 (1.15x) 19430.0 (1.09x) nv21_to_bgr565le_neon 29013.8 (1.12x) 47.2 (1.21x) 19428.8 (1.09x) nv21_to_rgb555le_neon 28533.3 (1.12x) 49.7 (1.16x) 19430.5 (1.09x) nv21_to_bgr555le_neon 29011.4 (1.12x) 48.5 (1.18x) 19428.7 (1.09x) yuv420p_to_rgb565le_neon 28351.9 (1.11x) 46.4 (1.18x) 19635.3 (1.08x) yuv420p_to_bgr565le_neon 28831.8 (1.11x) 50.8 (1.09x) 19634.5 (1.08x) yuv420p_to_rgb555le_neon 28351.3 (1.11x) 46.3 (1.23x) 19634.2 (1.08x) yuv420p_to_bgr555le_neon 28829.1 (1.11x) 46.5 (1.21x) 19634.3 (1.08x) yuva420p_to_rgb565le_neon 28349.5 (1.11x) 51.2 (1.06x) 19634.7 (1.08x) yuva420p_to_bgr565le_neon 28833.1 (1.11x) 48.6 (1.17x) 19633.9 (1.08x) yuva420p_to_rgb555le_neon 28351.6 (1.11x) 47.8 (1.16x) 19635.2 (1.08x) yuva420p_to_bgr555le_neon 28831.5 (1.11x) 46.4 (1.14x) 19634.8 (1.08x) Co-authored-by: Ramiro Polla <ramiro.polla@gmail.com> Signed-off-by: DROOdotFOO <drew@axol.io>
Changed files
- libswscale/aarch64/yuv2rgb_neon.S
Change #269654
Category ffmpeg Changed by DROOdotFOO <drew@axol.io> Changed at Sat 06 Jun 2026 19:38:40 Repository https://git.ffmpeg.org/ffmpeg.git Project ffmpeg Branch master Revision d9e2239f3c5e17d908db0b79f68ff06ee1867828 Comments
swscale/aarch64/yuv2rgb_neon: 2 lines at a time, yuva420p Alpha is full resolution, so each row loads its own 16 alpha bytes via process_row's \rsrcA arg. Test Name A55-gcc M1-clang A76-gcc ---------------------------------------------------------------------------------------- yuva420p_to_argb_neon 22607.6 (1.16x) 39.2 (1.24x) 13631.6 (1.12x) yuva420p_to_rgba_neon 22608.2 (1.16x) 38.3 (1.21x) 13912.8 (1.12x) yuva420p_to_abgr_neon 23074.6 (1.16x) 38.8 (1.22x) 14492.1 (1.08x) yuva420p_to_bgra_neon 23079.7 (1.16x) 39.9 (1.19x) 14472.6 (1.08x) Co-authored-by: Ramiro Polla <ramiro.polla@gmail.com> Signed-off-by: DROOdotFOO <drew@axol.io>
Changed files
- libswscale/aarch64/yuv2rgb_neon.S