Date: Tue, 29 Dec 2009 19:40:05 GMT From: Manuel Tobias Schiller <mala@hinterbergen.de> To: freebsd-sparc64@FreeBSD.org Subject: Re: sparc64/142102: FreeBSD 8.0 kernel panics on sparc64 when accessing NFS Message-ID: <200912291940.nBTJe5cf016937@freefall.freebsd.org>
next in thread | raw e-mail | index | archive | help
The following reply was made to PR sparc64/142102; it has been noted by GNATS. From: Manuel Tobias Schiller <mala@hinterbergen.de> To: Marius Strobl <marius@alchemy.franken.de> Cc: Mark Linimon <linimon@lonesome.com>, FreeBSD-gnats-submit@FreeBSD.org Subject: Re: sparc64/142102: FreeBSD 8.0 kernel panics on sparc64 when accessing NFS Date: Tue, 29 Dec 2009 20:29:55 +0100 On Tue, 29 Dec 2009 19:53:58 +0100 Marius Strobl <marius@alchemy.franken.de> wrote: > Oh, sorry, I had a bug in there, the svc code used a stale > pointer to the mbuf with this. Could you please re-fetch > and try again? I can't guarantee that this fixes the hang > you experienced but at least in theory just re-aligning > the data can't make the issue with the unaligned access > any worse. No problem, I know what it's like if you try to come up with a fix without being able to check things for yourself... I think that I can start a rebuild in about two hours or so, so we should have results by tomorrow afternoon (tomorrow morning is reserved for my grandparents...). At the moment, I'm compiling a "dumb" variant which just uses bcopy for all operations in sys/nfs/xdr_subs.h - if that compiles and works (I tend to make mistakes when I code, and a make buildkernel takes quite some time), we should have a safe version to fall back on, no matter what alignment constraints the architecture imposes... It's going to be ugly, though, so I'd rather use your solution, if possible. I guess we just have to wait and see. > Using the LOM or by breaking into the kernel > by sending a break you should be able to power-cycle/ > reboot the machine without removing the power cord should > it hang again though. I know about the sending-a-break-over-serial-trick, and of course I tried it, but there was no reaction whatsoever. > What NFS client and with which mount options are you > using to trigger is problem? > > Marius Concerning the NFS client machines: They are running Debian lenny, with a 2.6.30-bpo.2-amd64 kernel (one client is a powerpc G4, the other one is an amd64 machine). Mount options on the Linux clients are rw,nosuid,nodev,hard,intr. This combination used to work fine and rock-stable with FreeBSD 7.2 (during the month I spent at CERN just before Christmas, the machine was up the entire time and never showed any sign of instability - admittedly, there was no NFS access during that time). I hope we manage to get the FreeBSD 8.0 kernel there as well - apart from the NFS issue, what I've seen so far is impressive, as usual... Manuel -- Homepage: http://www.hinterbergen.de/mala OpenPGP: 0xA330353E (DSA) or 0xD87D188C (RSA)
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200912291940.nBTJe5cf016937>