ffmpeg | branch: master | Bin Peng <peng...@visionular.com> | Mon Dec 16 10:31:23 2024 +0800| [decc9e643cc3ac5537f42b465e2637fbefbf41cc] | committer: Martin Storsjö
lavc/aarch64: Fix ff_pred8x8_plane_neon_10 Fix test failure on aarch64: ./tests/checkasm/checkasm --test=h264pred 479612 The mismatch between neon and C functions can also be reproduced using the following bitstream and command line. wget https://streams.videolan.org/ffmpeg/incoming/intra8x8pred_10bit.264 ./ffmpeg -cpuflags 0 -threads 1 -i intra8x8pred_10bit.264 -f framemd5 -y md5_ref ./ffmpeg -threads 1 -i intra8x8pred_10bit.264 -f framemd5 -y md5_neon Signed-off-by: Bin Peng <peng...@visionular.com> Signed-off-by: Martin Storsjö <mar...@martin.st> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=decc9e643cc3ac5537f42b465e2637fbefbf41cc --- libavcodec/aarch64/h264pred_neon.S | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/libavcodec/aarch64/h264pred_neon.S b/libavcodec/aarch64/h264pred_neon.S index ea37689f34..168f8191ad 100644 --- a/libavcodec/aarch64/h264pred_neon.S +++ b/libavcodec/aarch64/h264pred_neon.S @@ -595,12 +595,11 @@ function ff_pred8x8_plane_neon_10, export=1 ssubl v2.4s, v2.4h, v3.4h ext v0.16b, v0.16b, v0.16b, #14 mov v0.h[0], wzr - mul v0.8h, v0.8h, v5.h[0] dup v1.4s, v2.s[0] dup v2.4s, v2.s[0] dup v3.8h, v5.h[1] - saddw v1.4s, v1.4s, v0.4h - saddw2 v2.4s, v2.4s, v0.8h + smlal v1.4s, v0.4h, v5.h[0] + smlal2 v2.4s, v0.8h, v5.h[0] mov w3, #8 mvni v4.8h, #0xFC, lsl #8 // 1023 for clipping 1: _______________________________________________ ffmpeg-cvslog mailing list ffmpeg-cvslog@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-cvslog To unsubscribe, visit link above, or email ffmpeg-cvslog-requ...@ffmpeg.org with subject "unsubscribe".
