Just noticed metacity using 99% cpu. I investigated it a bit with mccann and it's because metacity is calling meta_finalize from a sigterm handler. This is a big no-no. There is a set of maybe 10 - 15 functions that are legal to call from a sigterm handler.
Ray provided a patched metacity for testing: http://koji.fedoraproject.org/koji/buildinfo?buildID=139997 a bug in that build has been discovered, and Ray is now working on another update: <Oxf13> halfline: and metacity is spiking the CPU <halfline> sweet <halfline> i can log in, but things definitely aren't right <halfline> investigating <Oxf13> halfline: ok, the second time I got logged in <Oxf13> but it's still spiking the CPU <halfline> Oxf13: yea, patch needs to be reworked <halfline> Oxf13: g_io_channel isn't firing <Oxf13> ok. glad I tested 9: <halfline> must be a leaked fd somewhere <halfline> rather than tackling that bug where ever it is, instead i'm going to just not rely on hup and do an explicit write call -- Fedora Bugzappers volunteer triage team https://fedoraproject.org/wiki/BugZappers
Improved fixed metacity is at: http://koji.fedoraproject.org/koji/buildinfo?buildID=140020 Jesse is testing. -- Fedora Bugzappers volunteer triage team https://fedoraproject.org/wiki/BugZappers
the fixed build was tagged. -- Fedora Bugzappers volunteer triage team https://fedoraproject.org/wiki/BugZappers