On Sun, Aug 30, 2015 at 12:39 PM, Richard W.M. Jones rjones@redhat.com wrote:
I have i686 hardware (ancient Dell laptop) running 4.2.0-0.rc8.git0.1.fc23.i686 for over 24 hours with some heavy btrfs + rsync stuff, and no oopses. What would I need to run to try to reproduce? And is there a meaningful difference this is an fc23 kernel, not fc24?
On Sun, Aug 30, 2015 at 01:07:36PM -0600, Chris Murphy wrote:
On Sun, Aug 30, 2015 at 12:39 PM, Richard W.M. Jones rjones@redhat.com wrote:
I have i686 hardware (ancient Dell laptop) running 4.2.0-0.rc8.git0.1.fc23.i686 for over 24 hours with some heavy btrfs + rsync stuff, and no oopses. What would I need to run to try to reproduce? And is there a meaningful difference this is an fc23 kernel, not fc24?
It happens during the first second of boot, and seems to happen inside the 8250 serial driver, so unless you're repeatedly booting the laptop and it is attached to a serial console you wouldn't see this.
To try reproducing it, this is probably the easiest way:
# yum install libguestfs
$ export LIBGUESTFS_BACKEND=direct $ while libguestfs-test-tool >/tmp/log 2>&1; do echo -n .; done
After (possibly many many) 100s of iterations, it may exit and you can examine the log file (/tmp/log) to see if the failure was this bug or some other thing.
It may (or may not) help to have some other source of load on the laptop at the same time.
Rich.
On Sun, Aug 30, 2015 at 01:07:36PM -0600, Chris Murphy wrote:
And is there a meaningful difference this is an fc23 kernel, not fc24?
I don't know. If fc23 has the same kernel, I see no reason why it wouldn't have the same bug, but I didn't try it on fc23.
Rich.