Bug#551101: still exists
On Mon, 2010-06-21 at 11:33 +0200, martin f krafft wrote:
> also sprach Ben Hutchings <firstname.lastname@example.org> [2010.06.21.0120 +0200]:
> > > I am running 2.6.32 and am faced by exactly this problem on
> > > a machine I cannot just reboot:
> > >
> > > % ps aux | egrep -c '<D>.+vgdisplay'
> > > 37821
> > >
> > > lsof or strace on existing processes don't show anything, but strace
> > > on a new vgdisplay process shows that it blocks on opening
> > > /dev/cdrom, which seems to be defective.
> > Was this a version before or after the libata transition?
> I have not really followed the transition, but this is on a lenny
> system with a 2.6.32 backports.org kernel. How can I check?
We switched from IDE drivers to libata-based drivers in 2.6.32-10. My
suspicion is that this is a bug in ide-cd which will no longer apply.
Once a job is fouled up, anything done to improve it makes it worse.