From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.4) Gecko/20011019 Netscape6/6.2 Description of problem: Bordeaux w/ 2.4.9-13.1smp, 32GB ram, 2GB swap, bios A02. Boot disk on a perc 3/QC in slot 5, qlogic2200 copper attached to PV650F in slot 6, and Intel/XT in slot 8. External storage has 10 configured LUNS. During boot: Initializing USB controller (usb-uhci): Unable to handle kernel paging request at virtual address 0000000000000020 Unable to handle kernel paging request at virtual address 00000000000000c8 Unable to handle kernel paging request at virtual address 00000000000000c8 Unable to handle kernel paging request at virtual address 00000000000000c8 Checking for new hardware: /etc/rc3.d/S05kudzu: line 78: 290 Segmentation fault /usr/sbin/kudzu $KUDZU_ARGS -t 30 Then system hangs at: Bringing up interface lo: **Sometimes the system boots fine, sometimes the system just hangs at "Bringing up interface lo:", and sometimes the above errors occur. Version-Release number of selected component (if applicable): How reproducible: Sometimes Steps to Reproduce: 1.Install a bordeaux with rc3 and as late as 13.1 kernel, a02 bios, a megaraid card as boot device, and a qlogic 2200 attached to external storage. 2.Attempt to boot up 3. Actual Results: "Unable to handle kernel paging request" errors and a kudzu segmentation fault during boot. Expected Results: Normal boot. Additional info:
2.4.9-13.1? Um, that kernel is not-yet-ready-for-prime-time.
We're rewriting the swiommu as that was one of the reasons for severe performance problems on ia64 and this was a first draft ... to be continued.
Does not occur with RC2 updated to 2.4.9-0.18smp kernel.
Error did not appear during 5 boots with qa1108 (2.4.9-13.3smp).
2.4.9-13.4 is now available at ftp://ftp.beta.redhat.com/pub/testing/kernel/ Please test and make sure we have not regressed on this point from 2.4.9-13.3. I'm going to close this now on the assumption that we have not regressed; if I'm wrong, please re-open it.
After many more reboots during other testing, saw this with qa1108 (2.4.9-13.3smp).
Reproduced with qa1108 updated to 2.4.9-13.4.
With qa1129 (2.4.9-17.3smp), reproduced a system hang at: "Bringing up interface lo:" This is one of the possible hangs I've seen when the initrd is loading the qla2x00 driver on a system with a qlogic 2200 card and attached storage.
I wanted to clarify that this issue was reproduced (as described in my previous entry) with the new qla2200 driver in the initrd. I remade the initrd using the qla2200 entry made by kudzu, which as you said is updated for the new driver. It is only in bug 55836 that I manually loaded the old qla2x00 driver mistakenly. BTW, The reason I was manually loading the driver is because the installer is not updated with qla2200 entries.
Did you also get: "Unable to handle kernel paging request..." messages with qa1129? Did you get a segfault?
I have not seen the kernel paging error or segfault yet, just the hang. But, like the hang, these errors are intermittent. I am attempting to reproduce all of them.
And on my very next reboot...kernel paging error, kudzu segfault, and hang at "bringing up interface lo:" w/ qla2200 driver in qa1129 (2.4.9-17.3smp).
I wonder if this really is the qlogic fault or the usb controller ;(
A nasty stack overflow has been found and fixed in the qlogic driver; it was found due to an oops like this :).... Should be fixed as of 2.4.9-17.4 (which Dell should already have)
Putting this bug in NEEDINFO state since we think it is fixed but need a while before Dell is sure since it can't be reproduced reliably.
This seems fixed. Haven't reproduced with 2.4.9-17.4smp after rebooting all day yesterday and today.
Great, thanks!
*** Bug 562369 has been marked as a duplicate of this bug. ***