[Top] [All Lists]

Re: 2.6.22-rc3 hibernate(?) fails totally - regression (xfs on raid6)

To: David Chinner <dgc@xxxxxxx>
Subject: Re: 2.6.22-rc3 hibernate(?) fails totally - regression (xfs on raid6)
From: David Greaves <david@xxxxxxxxxxxx>
Date: Thu, 07 Jun 2007 14:59:58 +0100
Cc: Tejun Heo <htejun@xxxxxxxxx>, Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>, "Rafael J. Wysocki" <rjw@xxxxxxx>, xfs@xxxxxxxxxxx, "'linux-kernel@xxxxxxxxxxxxxxx'" <linux-kernel@xxxxxxxxxxxxxxx>, linux-pm <linux-pm@xxxxxxxxxxxxxx>, Neil Brown <neilb@xxxxxxx>
In-reply-to: <20070607110708.GS86004887@sgi.com>
References: <200706012342.45657.rjw@sisk.pl> <46609FAD.7010203@dgreaves.com> <200706020122.49989.rjw@sisk.pl> <4661EFBB.5010406@dgreaves.com> <alpine.LFD.0.98.0706021538360.23741@woody.linux-foundation.org> <4662D852.4000005@dgreaves.com> <46667160.80905@gmail.com> <46668EE0.2030509@dgreaves.com> <46679D56.7040001@gmail.com> <4667DE2D.6050903@dgreaves.com> <20070607110708.GS86004887@sgi.com>
Sender: xfs-bounce@xxxxxxxxxxx
User-agent: Mozilla-Thunderbird (X11/20070601)
David Chinner wrote:
On Thu, Jun 07, 2007 at 11:30:05AM +0100, David Greaves wrote:
Tejun Heo wrote:

David Greaves wrote:
Just to be clear. This problem is where my system won't resume after s2d
unless I umount my xfs over raid6 filesystem.
This is really weird. I don't see how xfs mount can affect this at all.
It does :)

Ok, so lets determine if it really is XFS.
Seems like a good next step...

Does the lockup happen with a
different filesystem on the md device? Or if you can't test that, does
any other XFS filesystem you have show the same problem?
It's a rather full 1.2Tb raid6 array - can't reformat it - sorry :)
I only noticed the problem when I umounted the fs during tests to prevent corruption - and it worked. I'm doing a sync each time it hibernates (see below) and a couple of paranoia xfs_repairs haven't shown any problems.

I do have another xfs filesystem on /dev/hdb2 (mentioned when I noticed the md/XFS correlation). It doesn't seem to have/cause any problems.

If it is xfs that is causing the problem, what happens if you
remount read-only instead of unmounting before shutting down?
Yes, I'm happy to try these tests.
nb, the hibernate script is:
ethtool -s eth0 wol g
echo platform > /sys/power/disk
echo disk > /sys/power/state

So there has always been a sync before any hibernate.

cu:~# mount -oremount,ro /huge
cu:~# mount
/dev/hda2 on / type xfs (rw)
proc on /proc type proc (rw)
sysfs on /sys type sysfs (rw)
usbfs on /proc/bus/usb type usbfs (rw)
tmpfs on /dev/shm type tmpfs (rw)
devpts on /dev/pts type devpts (rw,gid=5,mode=620)
nfsd on /proc/fs/nfsd type nfsd (rw)
/dev/hda1 on /boot type ext3 (rw)
/dev/md0 on /huge type xfs (ro)
/dev/hdb2 on /scratch type xfs (rw)
tmpfs on /dev type tmpfs (rw,size=10M,mode=0755)
rpc_pipefs on /var/lib/nfs/rpc_pipefs type rpc_pipefs (rw)
cu:(pid2862,port1022) on /net type nfs (intr,rw,port=1022,toplvl,map=/usr/share/am-utils/amd.net,noac)
elm:/space on /amd/elm/root/space type nfs (rw,vers=3,proto=tcp)
elm:/space-backup on /amd/elm/root/space-backup type nfs (rw,vers=3,proto=tcp)
elm:/usr/src on /amd/elm/root/usr/src type nfs (rw,vers=3,proto=tcp)
cu:~# /usr/net/bin/hibernate
[this works and resumes]

cu:~# mount -oremount,rw /huge
cu:~# /usr/net/bin/hibernate
[this works and resumes too !]

cu:~# touch /huge/tst
cu:~# /usr/net/bin/hibernate
[but this doesn't even hibernate]

> What about freezing the filesystem? cu:~# xfs_freeze -f /huge cu:~# /usr/net/bin/hibernate [but this doesn't even hibernate - same as the 'touch']

Nb the screen looks like this:
whether it hangs on suspend or resume.

So I wouldn't say it *is* XFS at fault - but there certainly seems to be an interaction...
At least it's easily reproducible :) Shame about the sysrq

I can think of other permutations of freeze/ro/writing tests but I'm just thrashing really. Happy for you to tell me what to try next ...


<Prev in Thread] Current Thread [Next in Thread>