Date: Wed, 11 Mar 1998 08:38:20 -0800 (PST) From: Chris Timmons <skynyrd@opus.cts.cwu.edu> To: freebsd-stable@FreeBSD.ORG Subject: panic at getnewbuf+0x21f, new mount behavior Message-ID: <Pine.BSF.3.96.980311082432.21355C-100000@opus.cts.cwu.edu>
next in thread | raw e-mail | index | archive | help
I still haven't figured out what's going on, but I get reliable panics during CVSup and other high volume disk activity in this environment: -stable as of 0445 pst today, full build/installworld genuine intel p5 NO_F00F_HACK CLK_USE_I586_CALIBRATION aha-3940 de-500 100/mb ether /home with CVS repository on 2 disk ccd which is NFS exported to a -current machine that has it mounted As soon as I start CVSup updating the CVS repository on the ccd, bang, into ddb at getnewbuf+0x21f. I am slowly narrowing down my kernel configuration and looking at the source tree during the time when the problem appeared (between 2/27-1200 and 2/28-000 utc, and the diffs are very innocent looking.) However, as of the mount/slice changes, I have another problem. When the system goes down hard and comes up, fsck's, it refuses to mount / R/W - claiming it is _still_dirty. Ok, I run fsck again, but it still refuses and so I have to reboot again to finally mount /. Somehow the fact that the file system has been cleaned is not noticed and each successive attempt to 'mount /dev/sd0a /' or 'mount /dev/sd0s2a /' fails. I'm alternating running kernels from about 2/27-1200 which don't panic given the same config, so maybe there is some sort of mismatch with the new mount stuff when I come up after a crash booting /kernel.good. Those kernels like /dev/sd0a rather than /dev/sd0s2a for /, so I had to back out the recommended change that mount told me about as I booted the latest kernel. -Chris To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.3.96.980311082432.21355C-100000>