Date: Mon, 29 Oct 2007 11:37:30 +0100 From: Kris Kennaway <kris@FreeBSD.org> To: Gleb Kozyrev <gkozyrev@gmail.com> Cc: freebsd-current@freebsd.org Subject: Re: Deadlock, exclusive sx so_rcv_sx, amd64 Message-ID: <4725B7EA.5080807@FreeBSD.org> In-Reply-To: <ca661e8a0710281849g7147a3ao97d68fd2f12b3daf@mail.gmail.com> References: <ca661e8a0710211529y2a48b4a9h78da4211b3f09274@mail.gmail.com> <200710251435.58984.jhb@freebsd.org> <ca661e8a0710260252u4f5c1c3di22828143e81e3df2@mail.gmail.com> <200710261222.28656.jhb@freebsd.org> <20071026184536.GE37471@deviant.kiev.zoral.com.ua> <ca661e8a0710261302j188a05c4h63728eab67f6068@mail.gmail.com> <ca661e8a0710281849g7147a3ao97d68fd2f12b3daf@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Gleb Kozyrev wrote: > On 26/10/2007, Gleb Kozyrev <gkozyrev@gmail.com> wrote: >> I managed to find a kernel message that precedes the freeze: >> Approaching the limit on PV entries, consider increasing sysctl >> vm.pmap.shpgperproc or vm.pmap.pv_entry_max >> >> rtorrent likes to mmap large files. >> After increasing (one of) these sysctls everything seems ok. >> Probably, setting rtorrent's memory limit could have helped either. >> >> Is there a way to know if some process is becoming close to that limit? > > It now just takes longer until freeze with shpgperproc = 5000 (originally 200). > So how about vm.pmap.pv_entry_max? Have you obtained the usual debugging information (DDB backtraces, etc) when the system freezes? Kris
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4725B7EA.5080807>