diff options
author | Rich Felker <dalias@aerifal.cx> | 2014-08-25 15:58:19 -0400 |
---|---|---|
committer | Rich Felker <dalias@aerifal.cx> | 2014-08-25 15:58:19 -0400 |
commit | b8a9c90e4fe29821f2ef40796c07c48bd8fac8cc (patch) | |
tree | 30d89022cbd1511876d2e2993238bc9906eea96a /src/math/fmal.c | |
parent | ea818ea8340c13742a4f41e6077f732291aea4bc (diff) | |
download | musl-b8a9c90e4fe29821f2ef40796c07c48bd8fac8cc.tar.gz musl-b8a9c90e4fe29821f2ef40796c07c48bd8fac8cc.tar.xz musl-b8a9c90e4fe29821f2ef40796c07c48bd8fac8cc.zip |
sanitize number of spins in userspace before futex wait
the previous spin limit of 10000 was utterly unreasonable. empirically, it could consume up to 200000 cycles, whereas a failed futex wait (EAGAIN) typically takes 1000 cycles or less, and even a true wait/wake round seems much less expensive. the new counts (100 for general wait, 200 in barrier) were simply chosen to be in the range of what's reasonable without having adverse effects on casual micro-benchmark tests I have been running. they may still be too high, from a standpoint of not wasting cpu cycles, but at least they're a lot better than before. rigorous testing across different archs and cpu models should be performed at some point to determine whether further adjustments should be made.
Diffstat (limited to 'src/math/fmal.c')
0 files changed, 0 insertions, 0 deletions