From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; Linux 2.4.2-2 i686; en-US; rv:0.9.1) Gecko/20010622 Description of problem: Fresh install of 7.1, with updates on FIC SD11 MB with Athalon 600 and extra promise ultra66 udma ide controller. New kernel 2.4.3-12 from redhat panics on boot with messages that /dev/hdg id now /dev/hdc, can't find root device, driver not present for /dev/hdg and lots of other stuff. Works fine with kernel 2.4.2-2. Version-Release number of selected component (if applicable): How reproducible: Always Steps to Reproduce: 1.Install RH7.1 (works fine) 2.Upgrade kernel to 2.4.3-12.rpm (or 2.4.7 from rawhide) 3.Try booting new kernel Actual Results: ide devices get whacked Expected Results: I expected my ide device names to stay the same. They are different, possibly missing? Additional info: Here's the setup: ide0(onboard) hda = WD172AA 17 Gig hdb = no device ide1(onboard) hdc = Mitsumi CDROM hdd = no device ide2(offboard promise ultra66) hde = LiteOn CDRW hdf = ZIP 100 ide3(offboard promise ultra66) hdg = IBM 30 Gig hdh = no device I suspect that some of the ide controllers are not being detected, hence the change in device names. The kernel produces tons of messages about /dev/md0, among others, that all scroll off the screen. I will gladly provide these if you can tell me how to capture them. I can't find them in dmesg or messages. Otherwise I will just type up what is left on the screen when the kernel panics. Right now I can't tell if the devices are detected, then dropped (udma problem?), or just not detected at all. The one thing I've tried so far is ide=nodma at the lilo prompt, and this didn't work.
Created attachment 29815 [details] boot messages from working kernel 2.4.2-2 and lilo.conf
*** See attachment *** Is the way the new kernel ide drivers handle disk geometry different? It seems with the lilo append line I'm using for the new kernel, with the geometry that is reported by the working old kernel, that everything "works" except for hdg. The new kernel reports that hdg has a bad superblock, has a read error, etc., and then panics. -If I probe hdg, the kernel panics with no device driver for hdg. -If I don't probe, the same. -If I don't probe AND specify geometry, the kernel finds the device and gives a read error. If someone could tell me how to capture dmesg from a bombed (at boot) kernel, I sure would be happpy :)
Thanks for the bug report. However, Red Hat no longer maintains this version of the product. Please upgrade to the latest version and open a new bug if the problem persists. The Fedora Legacy project (http://fedoralegacy.org/) maintains some older releases, and if you believe this bug is interesting to them, please report the problem in the bug tracker at: http://bugzilla.fedora.us/