]> git.sesse.net Git - ffmpeg/commit
swscale: aarch64: Add a NEON implementation of interleaveBytes
authorMartin Storsjö <martin@martin.st>
Fri, 15 May 2020 07:54:10 +0000 (10:54 +0300)
committerMartin Storsjö <martin@martin.st>
Fri, 15 May 2020 20:38:17 +0000 (23:38 +0300)
commite0604d508e97322b88c8e4318516ead0cbc55170
tree2157f7cc3ce3f7cbeae0bcb60cd00c907b96129f
parenteba1ebd9bfd667619add3e5d8042bf0e5cf928e5
swscale: aarch64: Add a NEON implementation of interleaveBytes

This allows speeding up format conversions from yuv420 to nv12.

                             Cortex A53      A72      A73
interleave_bytes_c:             86077.5  51433.0  66972.0
interleave_bytes_neon:          19701.7  23019.2  15859.2
interleave_bytes_aligned_c:     86603.0  52017.2  67484.2
interleave_bytes_aligned_neon:   9061.0   7623.0   6309.0

Signed-off-by: Martin Storsjö <martin@martin.st>
libswscale/aarch64/Makefile
libswscale/aarch64/rgb2rgb.c [new file with mode: 0644]
libswscale/aarch64/rgb2rgb_neon.S [new file with mode: 0644]
libswscale/rgb2rgb.c
libswscale/rgb2rgb.h