aboutsummaryrefslogtreecommitdiffstats
path: root/presets
diff options
context:
space:
mode:
authorRémi Denis-Courmont <remi@remlab.net>2024-05-16 17:13:22 +0300
committerRémi Denis-Courmont <remi@remlab.net>2024-05-19 10:03:49 +0300
commit4ad5b9c8db737f3018360e179121fc0a9b7e7d1b (patch)
treed9c2f3ab5328bcc4283243d3c538efdd876ca49e /presets
parent8b8b555de008213894a29e453f27bf9de1c990dc (diff)
downloadffmpeg-4ad5b9c8db737f3018360e179121fc0a9b7e7d1b.tar.gz
lavc/startcode: add R-V Zbb startcode_find_candidate
The main loop processes 8 bytes in 5 instructions. For comparison, the optimal plain strnlen() requires 4 instructions per byte (6.4x worse): LBU; ADDI; BEQZ; BNE. The current libavcodec C code involves 5 instructions per byte (8x worse). Actual benchmarks may be slightly less favourable due to latency from ORC.B to BNE.
Diffstat (limited to 'presets')
0 files changed, 0 insertions, 0 deletions