FS#24747 - [kernel26] 2.6.39.1-1 cannot reboot on fitpc2 ; regression (2.6.38.7-1 is fine)
Attached to Project:
Arch Linux
Opened by Dave Hooper (stripwax) - Thursday, 16 June 2011, 22:36 GMT
Last edited by Tobias Powalowski (tpowa) - Thursday, 16 February 2012, 17:59 GMT
Opened by Dave Hooper (stripwax) - Thursday, 16 June 2011, 22:36 GMT
Last edited by Tobias Powalowski (tpowa) - Thursday, 16 February 2012, 17:59 GMT
|
Details
Description:
On an otherwise up-to-date arch system (i686, fitpc2, Intel Atom Z510), running kernel 2.6.39.1-1, reboot (and "shutdown -r") do not work and simply hang. "shutdown -h" still works fine, and powers down as expected here. Downgrading just the kernel26 package to 2.6.38.7-1 (happened to be my prior version) and reboot works again. fitpc2 uses BIOS reboot. Initially I suspected this bug: https://bugzilla.kernel.org/show_bug.cgi?id=33302 however that is marked closed and should therefore already be in 2.6.39 (and presumably 2.6.39.1 also) Originally reported on forums here, before I found the archlinux flyspray: https://bbs.archlinux.org/viewtopic.php?pid=949245 |
This task depends upon
Closed by Tobias Powalowski (tpowa)
Thursday, 16 February 2012, 17:59 GMT
Reason for closing: Upstream
Thursday, 16 February 2012, 17:59 GMT
Reason for closing: Upstream
With kernel 2.6.38.7-1 , the system reboots as expected .
https://bugzilla.kernel.org/show_bug.cgi?id=37712
https://bbs.archlinux.org/viewtopic.php?pid=952292#p952292
I have separately confirmed that the problem still persists on 2.6.39.1-vanilla i.e. kernel built from AUR package without ARCH-specific patches. This suggests to me that this must just be an upstream kernel bug (assuming the AUR kernel26-vanilla package is indeed entirely vanilla).
I raised a bugzilla issue (before kernel.org went down...) - anyone know if the linux kernel folks still look at bug reports??
http://www.gossamer-threads.com/lists/linux/kernel/1404953