[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#476890: linux-image-2.6.24-1-686: nfs4 crash after resuming from suspend2disk

> tags 476890 moreinfo
> thanks
> On Sun, Apr 20, 2008 at 01:00:26AM +0200, Dunsens wrote:
> > After resuming from a suspend2disk, the system comes back up (the whole disk is luks+lvm including swap space), but then fails to reinitialize the wireless lan connection (iwl3945), configured within /etc/network/interfaces with wpa_supplicant. 
> > I get a backtrace like with dmesg (see below), showing that sunrpc/nfs4 crashes occur inside the kernel before I am able to try to reconfigure the iwl3945 driver with ifup/ifdown. I cannot get any wlan information, like iwlist scanning, afterwards, so it might be a problem with the iwl3945 driver, since I have used ipw3945 before. On my iBook I have to reconfigure the b43 wlan driver with ifup/ifdown after suspending, too.
> I use a similar configuration. iwl3945, some nfs4 mounts, suspend to ram
> and suspend to disk. I've never seen such bugs with .23, .24 and later.
S2Ram hangs after switching on (hd-led switched on) on since 2.6.24
here. Even alt+print+reisub does not show any effect. This might be
another bug report.
Can you connect to your NFS4 mounts without using the portmapper? I have
had to open it in the firewall to reenable my clients again, which
should not be necessary.
> > I am using /home on nfs4, which renders the whole system for all users with /home unusable (in my case only me) and causes data loss. 
> You use iwl3945, which is usualy a mini-PCIe card, in a multiuser
> machine? Anyway, any kernel crash can cause data loss.
Several people can be logged in a graphical session at the same time,
even on a laptop.
> Also I see loaded vboxdrv as tainting because it modifies core kernel
> status. So, please recheck without vboxdrv and/or with 2.6.25[1].
I have removed the module and have tried 2.6.25-trunk. Then I have tried
2.6.24-1 again. It always worked. It even worked with going back to my
old setup, so I guess that the error is not reproducable with a freshly
booted system and this information is useless. I will try to recheck it
during the next days with longer sessions.
> Bastian
> [1]: http://wiki.debian.org/DebianKernel


Reply to: