Date: Fri, 8 Feb 2002 10:23:51 -0800 (PST) From: David Wolfskill <david@catwhisker.org> To: current@freebsd.org Subject: Hang on flushing buffers w/today's -CURRENT, SMP system Message-ID: <200202081823.g18INp574522@bunrab.catwhisker.org>
next in thread | raw e-mail | index | archive | help
OK; I got today's -CURRENT built & running on each of my build machine (freebeast) & my laptop. (Got today's -STABLE built earlier; I mention this as a reference point/comparison. I similarly note that I've been tracking each daily on each machine for several months, and that today is the first time I've seen this problem.) Anyway, the laptop seems fairly normal: I got -CURRENT built, booted it up, ran a few things, used boot0cfg to switch to the slice with today's -STABLE, and it came right down (gracefully) and back up again: g1-7(4.5-S)[1] uname -a FreeBSD g1-7.catwhisker.org 4.5-STABLE FreeBSD 4.5-STABLE #74: Fri Feb 8 05:58:01 PST 2002 root@d144.catwhisker.org:/common/S1/obj/usr/src/sys/LAPTOP_30W i386 g1-7(4.5-S)[2] But on the build machine, I got it running -CURRENT, then did the same procedure (boot0cfg & reboot), but on the (serial) console, I see: Feb 8 08:45:32 freebeast mountd[181]: bad exports list line /cdrom -ro -alldirs apache cvsupd . Additional TCP options:. Starting background filesystem checks Fri Feb 8 08:45:34 PST 2002 FreeBSD/i386 (freebeast.catwhisker.org) (cuaa0) login: boot() called on cpu#0 Waiting (max 60 seconds) for system process `vnlru' to stop...stopped Waiting (max 60 seconds) for system process `bufdaemon' to stop...stopped Waiting (max 60 seconds) for system process `syncer' to stop...stopped syncing disks... 15 15 ...and it's been sitting like that for about half an hour. This is the second time I've tried this with today's -CURRENT on the box; the first time, it hung at "syncing disks... 7 7 " -- and I finally hit the reset button. (I don't have a keyboard on the machine, though I suppose I could try putting one on. Getting a (PC) monitor would be a little harder, though, since the only one we have is on my spouse's M$ machine, and the only spare monitor I have is a Sun-badged Sony for the SPARCstation. One other thing: the machine is pingable: bunrab(4.5-S)[1] ping freebeast PING freebeast.catwhisker.org (172.16.8.10): 56 data bytes 64 bytes from 172.16.8.10: icmp_seq=0 ttl=64 time=0.398 ms 64 bytes from 172.16.8.10: icmp_seq=1 ttl=64 time=0.189 ms ^C --- freebeast.catwhisker.org ping statistics --- 2 packets transmitted, 2 packets received, 0% packet loss round-trip min/avg/max/stddev = 0.189/0.293/0.398/0.105 ms bunrab(4.5-S)[2] So I'd guess that while things that require process creation are unlikely to succeed, the network is still capable of (primitive) functioning. I don't run much fancy stuff on the build machine -- Apache (so I can easily use cvsweb locally); cvsupd (for internal use); sshd (so I can login to the machine); that's about it. Is there a way to force the debugger from the serial console? About the only salient difference between the two machines that occurs to me is that the laptop has a single CPU (750 MHz PIII), while the build machine has a pair of them (2x866 MHz PIII). It's not a serious problem (in and of itself) for me at this time, but it certainly appears broken, so I'd like to help fix it. (I've reduced the list of likely candidate files that contributed to this down to 71 via a quick, casual inspection. I don't want to spam the list any worse than I already have, though....) Thanks, david -- David H. Wolfskill david@catwhisker.org I believe it would be irresponsible (and thus, unethical) for me to advise, recommend, or support the use of any product that is or depends on any Microsoft product for any purpose other than personal amusement. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200202081823.g18INp574522>