Messages in this thread |  | | | Date | Fri, 3 Jul 2009 16:54:22 -0600 | | From | dann frazier <> | | Subject | Re: ACPI issues on HP DL585 G2 |
On Tue, Jun 30, 2009 at 09:35:04PM -0400, Stephen Frost wrote: > * dann frazier (dannf@dannf.org) wrote: > > On Mon, Jun 29, 2009 at 10:06:44PM -0400, Stephen Frost wrote: > > > Don't suppose you can grab a P600 from somewhere and try? > > > > Got one, and I am able to reproduce a hang while initializing > > cciss. I'll try narrowing down the issue. > > Great! Glad to hear it. If there's anything I can do to help, please > let me know.
Here's what I've found.
Bisecting the hang lands on this commit, which went in right before 2.6.26 went final:
77b96bd: cciss: fix regression that no device nodes are created if no logical drives
This seems to have caused problems for 2.6.26 era kernels (including Debian's current stable release). I'm not sure if this problem still exists in 2.6.31-rc1, because I think we are hitting an earlier issue which also manifests as a hang.
In 2.6.31-rc1, I see a hang while trying to request the firmware version using sendcmd_withirq(). If I switch this to a sendcmd(), it proceeds until the next sendcmd_withirq(). /proc/interrupts shows a 0 count for all cpus on the chosen IRQ line (26). This makes me wonder if its related to the issue: http://www.spinics.net/lists/linux-scsi/msg37177.html
-- dann frazier
|  |