Description of Problem: After 3 days and 19 hours of running newburn (cerberus) version 9-19 on a Dell PowerEdge 7150 w/ seawolf gold ia64 installed, w/ bios A01, 2GB of ram, 512MB of swap. "Bad swap file entry 00000020" appeared repeatedly at the console and in dmesg. Newburn logs showed that newburn continued to run despite the error. How Reproducible: Haven't yet.
This defect considered SHOULD-FIX for Fairfax gold-release.
Created attachment 24188 [details] Errors in /var/log/messages
Errors messages now reproduced in fairfax beta2 ia64 on a pe7150 w/ 2GB ram, 4GB swap, booting from onboard qlogic scsi controller, running newburn version 9-20. Error messages begin within an hour of starting newburn. Errors no longer appear at console but are found in /var/log/messages. Attached is portion of /var/log/messages containing bootup, start of cerberus, and first appearance of errors.
Reproduced in beta3 on a 7150 w/ bios X15, 2GB ram, 2GB swap. Testing on 2.4.7-0.3 in progress...
Is this with eepro100 or e100 ?
We found a bug today that could explain weird things on ia64 under high load. Fix will be in kernels 2.4.7-0.11 and later.
Can you test this on 0.11 or later when it becomes available?
With Beta3 kernel 2.4.7-0.12.0 on IA64, 2GB ram, 2GB swap, onboard scsi controller running cerberus 9-20. "Bad swap file entry 00000020" errors now appearing in the DMESG cerberus log in /usr/bin/ctcs/.newburn*. File contains hundreds of them.
Is this an SMP or non-SMP kernel?
SMP
This should be fixed in 2.4.7-6.1 or later. Does that work for you?
Whoops, disregard that last comment.
"Bad swap file" error not appearing at console or in /var/log/messages running kernel 2.4.9-0.18smp w/ 2GB ram after 24 hours of newburn. Will continue to monitor.