On Sun, 2 Mar 2025, Martin Storsjö wrote:
On Sat, 1 Mar 2025, Krzysztof Pyrkosz via ffmpeg-devel wrote:
Before/after:
A78
hscale_16_to_15__fs_4_dstW_8_neon: 86.8 ( 1.72x)
hscale_16_to_15__fs_4_dstW_24_neon: 147.5 ( 2.73x)
hscale_16_to_15__fs_4_dstW_128_neon: 614.0 ( 3.14x)
hscale_16_to_15__fs_4_dstW_144_neon: 680.5 ( 3.18x)
hscale_16_to_15__fs_4_dstW_256_neon: 1193.2 ( 3.19x)
hscale_16_to_15__fs_4_dstW_512_neon: 2305.0 ( 3.27x)
hscale_16_to_15__fs_4_dstW_8_neon: 86.0 ( 1.74x)
hscale_16_to_15__fs_4_dstW_24_neon: 106.8 ( 3.78x)
hscale_16_to_15__fs_4_dstW_128_neon: 404.0 ( 4.81x)
hscale_16_to_15__fs_4_dstW_144_neon: 451.8 ( 4.80x)
hscale_16_to_15__fs_4_dstW_256_neon: 760.5 ( 5.06x)
hscale_16_to_15__fs_4_dstW_512_neon: 1520.0 ( 5.01x)
A72
hscale_16_to_15__fs_4_dstW_8_neon: 156.8 ( 1.52x)
hscale_16_to_15__fs_4_dstW_24_neon: 217.8 ( 2.52x)
hscale_16_to_15__fs_4_dstW_128_neon: 906.8 ( 2.90x)
hscale_16_to_15__fs_4_dstW_144_neon: 1014.5 ( 2.91x)
hscale_16_to_15__fs_4_dstW_256_neon: 1751.5 ( 2.96x)
hscale_16_to_15__fs_4_dstW_512_neon: 3469.3 ( 2.97x)
hscale_16_to_15__fs_4_dstW_8_neon: 151.2 ( 1.54x)
hscale_16_to_15__fs_4_dstW_24_neon: 173.4 ( 3.15x)
hscale_16_to_15__fs_4_dstW_128_neon: 660.0 ( 3.98x)
hscale_16_to_15__fs_4_dstW_144_neon: 735.7 ( 4.00x)
hscale_16_to_15__fs_4_dstW_256_neon: 1273.5 ( 4.09x)
hscale_16_to_15__fs_4_dstW_512_neon: 2488.2 ( 4.16x)
---
This patch removes the use of stack for temporary state and replaces
interleaved ld4 loads with ld1.
I'm aware the component is being deprecated, however in my use case
(screen recording) the total time spent in this function is roughly 15%,
the improvement is significant and worth sharing.
The patch looks good. I didn't follow it in exact detail, but it overall
looks reasonable, and looks much better than the previous form. This
description of what the patch does and why also is worth keeping in the final
commit message, but as there's no need to repost the patch, I could just
adjust the message myself before pushing it.
I pushed this one, and the second ac3dsp patch now, with the commit
messages readjusted a little bit. The first ac3dsp patch should be good
too if someone verifies that it's ok to handle 16 elements at a time.
// Martin
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
