Description of problem: When connecting network cable to NIC after machine is up, I get the message, "Spurious 8259A interrupt: IRQ7" and I don't see an IRQ 7 in /proc/interrupts (is this the right place to look?). Moments after this error appears in /var/log/messages, the machine hangs and needs to be restarted. Sometimes a thing or two can be done after this message appears, but it ALWAYS hangs, regardless of run level. The trick is to get it to appear, and it isn't always consistent. I have an HP OmniBook 900B with a Xircom REM58G-100 PCMCIA card....dont' know if the card is bad yet, as I don't have a replacement test case. The machine was upgraded from Red Hat 8.0, latest kernel version, to Fedora Core 1 from initial ISOs, then to current updates as of 1/8/04. Version-Release number of selected component (if applicable): 2.4.22-1.2140.n How reproducible: Somewhat Steps to Reproduce: 1.Start machine (any run level allowing network connectivity). 2.Plug in network cable. 3.Watch /var/log/messages for this message. Actual results: Complete freeze up. Expected results: Network connectivity. Additional info: This all seemed to have begun concurrent with reporting bug number 113229. Machine hangs frequently under seemingly different circumstances, where it was rock-solid before in Red Hat kernel. Also, messages in /var/log/messages such as: 1. localhost ntpd[775]: synchronisation lost and 2. localhost modprobe: modprobe: Can't locate module sound-slot-1 localhost modprobe: modprobe: can't locate module sound-service-1-0 localhost kernel: ttyS3 unloaded localhost cardmgr[520]: executing: './network stop eth0' localhost kernel: vfree(): sleeping in interrupt!! and even 3. localhost kernel: spurious 8259A interrupt: IRQ7. localhost ntpd[1388]: time reset 0.304193 s localhost ntpd[1388]: kernel time discipline status change 41 localhost ntpd[1388]: synchronisation lost appear shortly (i.e., seconds to about 2 minutes) before hanging (these are usually when the network cable is connected, however). Again, don't know how the different means of hanging might be related, and can't test a different PCMCIA card.
No sooner did I commit this bug, but it happened again, and I figured out how to avoid a lock-up: ifdown eth0; cardctl eject Then, when I reinserted the card (with the cable still connected), it seems to be working, but I'm still not completely convinced the system is not going to turn fragile on me again....I've seen too many different problems crop up since bug 113229. Thanks again, Kara
Right about the time I entered the above comment, I lost the box again--this time, I was in a man page and about to hit "k" to move up, and nada. Zilch. Zippo. No comment in /var/log/messages in single-user mode indicates any particular issues. In fact, it isn't even up-to-date with this current restart (I should have seen messages in there, even though I'm just in single-user mode, right?), let alone any last gasps. Is there anything I can turn on to give me more verbose and possibly helpful messages, like cranking up a log level somewhere?
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/