Summary: | sci-geosciences/gpsd is unstable when built with -O3 | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | János Tóth F. <toth.f.janos> |
Component: | Current packages | Assignee: | Sci-geo Project <sci-geosciences> |
Status: | UNCONFIRMED --- | ||
Severity: | minor | CC: | esigra, gem, negril.nx+gentoo, toth.f.janos |
Priority: | Normal | ||
Version: | unspecified | ||
Hardware: | AMD64 | ||
OS: | Linux | ||
See Also: | https://bugs.gentoo.org/show_bug.cgi?id=932307 | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Bug Depends on: | |||
Bug Blocks: | 915000 |
Description
János Tóth F.
2024-06-27 14:22:39 UTC
I have hardware to test this, but obviously gonna take a few days to reproduce. Any way to find out what -O3 does differently than -O2? The SOCKS PPS is in a thread, just for PPS. It has to do some memory handshakes with Chrony. Do the gpsd regression tests still pass when you build with -O3? "man gcc" says: ```` -O3 Optimize yet more. -O3 turns on all optimizations specified by -O2 and also turns on the following optimization flags: -fgcse-after-reload -fipa-cp-clone -floop-interchange -floop-unroll-and-jam -fpeel-loops -fpredictive-commoning -fsplit-loops -fsplit-paths -ftree-loop-distribution -ftree-partial-pre -funswitch-loops -fvect-cost-model=dynamic -fversion-loops-for-strides ```` So one of those is the problem. Sorry for my false report. It's not an -O3 issue. It just took a lot longer this time to glitch out after I rebuilt and restarted gpsd. I will try to build both gpsd and chrony with -O2 and -fno-lto but I guess the problem might lays elsewhere. I also tried to use the older SHM interface and that also stopped working after a while. You should consider moving this issue to gpsd's big tracker: https://gitlab.com/gpsd/gpsd/ Or visit #gpsd on Librechat IRC. |