Re: unable to handle paging request in worker_thread on apm resume

2005-03-23 Thread Andrew Morton
"J. Bruce Fields" <[EMAIL PROTECTED]> wrote: > > > Have you added any code which does schedule_work()? > > Hm, I don't think so. But of course I have some nfsv4 patches, and the > nfsv4 nfsd code does use schedule_delayed_work(). It's possible the > cleanup is wrong, and that bringing nfsd up

Re: unable to handle paging request in worker_thread on apm resume

2005-03-23 Thread J. Bruce Fields
On Wed, Mar 23, 2005 at 02:33:44AM -0800, Andrew Morton wrote: > Seems that there's a `struct work_struct' which is still registered but its > memory has been freed. It's likely that CONFIG_DEBUG_PAGEALLOC caught this. > > Either that, or some module got unloaded without flushing its workqueue. >

Re: unable to handle paging request in worker_thread on apm resume

2005-03-23 Thread Andrew Morton
"J. Bruce Fields" <[EMAIL PROTECTED]> wrote: > > I got the following after an apm resume on a thinkpad X31, with > 2.6.12-rc1 plus some (hopefully unrelated) NFS patches. Any ideas? > > --Bruce Fields > > > Mar 21 18:22:44 puzzle apmd[1815]: Suspending now > Mar 21 22:37:36 puzzle kernel: P

unable to handle paging request in worker_thread on apm resume

2005-03-21 Thread J. Bruce Fields
I got the following after an apm resume on a thinkpad X31, with 2.6.12-rc1 plus some (hopefully unrelated) NFS patches. Any ideas? --Bruce Fields Mar 21 18:22:44 puzzle apmd[1815]: Suspending now Mar 21 22:37:36 puzzle kernel: PCI: Setting latency timer of device :00:1d.0 to 64 Mar 21 22:3