Red Hat Bugzilla – Bug 129097
With kudzu-1.1.75-1 a guaranteed hangup on reboot
Last modified: 2015-01-04 17:08:21 EST
Description of problem: On a cold boot kudzu does not create problems but I will try to reboot then this invariably hangs in kudzu. A machine is dead to that extent that even trying sysrq key, which I am turning on in a modified kudzu startup script, does not have any effects. Configuring it "SAFE" also has no effect. It trying a manual run after after a reboot with kudzu service turned off then one can see '/usr/sbin/kudzu -s -t 30' on a screen nothing happens after that. This was happening before often enough to be annoying but now it is certain. As a matter of fact the real culprit can be USB because re-booting with 'nousb' in kernel parameters makes this hang to go away. OTOH this makes USB clearly inaccessible. Hardware: Tyan Tiger MP with 2 Athlon MP 1700+ processors.
I'm guessing this is a warm kernel reboot problem; there are other reports of this.
I am inclined right now to put that to "USB handling on this mobo" bug. This motherboard is known to have problems with USB and, as a result of testing, some new OHCI Root Hub was added recently. It could be too that kernel is not handling properly some weird USB behaviour. I cannot be sure. If I will skip kudzu and will leave "nousb" out of the picture then I will boot fine into level 3 but then X will hang that machine without any traces anywhere. Things are usually fine from a cold boot even if kudzu will get stuck there too on some infrequent occasions. Hard for me to say anything about "other reports" Bill mentioned.
any better with the current updates ?
> any better with the current updates ? With 2.6.9-1.1020_FC4 turning on kudzu while 'nousb' was omitted from boot parameters still kills my test box with an absolute perfection. But there is some progress. Namely if I will leave kudzu out of the picture then I can start X, which was impossible before, and even a USB mouse does then work as well as a USB floppy (even if the later did not mount itself; probably because it showed as /dev/sda).
An update has been released for Fedora Core 3 (kernel-2.6.12-1.1372_FC3) which may contain a fix for your problem. Please update to this new kernel, and report whether or not it fixes your problem. If you have updated to Fedora Core 4 since this bug was opened, and the problem still occurs with the latest updates for that release, please change the version field of this bug to 'fc4'. Thank you.