HP says that a high percentage of their hardware manifests this bug. They get a constant stream of errors from autofs about floppy disk failures (sorry, no specifics...more info coming on the exact error). They've looked into it and tracked it down to a constant string of disk change detections coming from floppy.c There are two possibilities; the hardware is telling floppy.c that a disk change has occurred or there is a bug in floppy.c so that it returns a false positive on this check. The overwhelming evidence of our install base points towards the former and not the latter, but HP has looked at the code and disagrees. So, if we could look into this code and see if there are any obvious problems that would be helpful. If we don't see anything, then we can test on HP hardware and see what we get. The final possibility is that HP has tracked this to the wrong place; if so, we should test on their hardware and trace the errors ourselves. I've glanced at their config, btw, and they look fine. Contact nthomas if you need more info.
Nathan, is this bug still an issue?
This is probably not an issue, at least with reguards to why it was opened. This turned out to be a SNMP issue. This is alans response: "Its correct behaviour. They are querying the floppy drive and other devices by snmp and it is trying to see if there is a disk in and getting errors back from the drive as it is empty" HP agrees with this and has closed the issue.