I was trying to do 'portupgrade portupgrade' today and my -CURRENT system rebooted. Strange - I haven't seen such behaviour before. I believe it wouldn't happen before my last upgrade (it was 30-10-2002 -CURRENT system, and I cvsup'ed it on 6-11-2002)
After reboot I retried portupgrade - samre result. During next few reboot cycles I managed to narrow the problem a little - I am getting 100% reproducible panics when running this script as root: #!/bin/sh export FTP_PASSIVE_MODE=no fetch -v ftp://ftp.iDaemons.org/pub/distfiles/pkgtools-20021103-20021106.diff.bz2 So I've compiled in DDB, INVARIANTS and WITNESS stuff I have previously removed from kernel. Here is the information I gathered: 1. First - I think this might be important (kernel message after samba has started): Local package initialization: apache2 proftpd rtc Sambaacquiring duplicate lock of same type: "inp" 1st inp @ /usr/src/sys/netinet/udp_usrreq.c:290 2nd inp @ /usr/src/sys/netinet/udp_usrreq.c:290 2. The crash (result of running the script). This is hand-copied, I don't even have swap-space for crash dumps... panic: process 657(fetch):32 holds inp but isn't blocked on a mutex Debugger("panic") Stopped at Debugger+0x4d: xchgl %ebx,in_Debugger.0 db> trace Debugger(c03906fb,c0405100,c038f900,d68bdc30,1) at ... panic ... propagate_priority(c152ea90,2,c038f82d23b,c4522c7c) at propagate_priority+0x1bf _mtx_lock_sleep(c4522c7c,0,c03968a9,182,c4524ee0) at _mtx_lock_sleep+0x217 _mtx_lock_flags(c4522c72,0,c039b8a9,182,c03913eb) at _mtx_lock_flags+0x97 syncache_timer(2,0,c03913eb,bf,39a5) at syncache_timer+0xa4 softclock(0,0,c038e13c,230,c152da80) at softclock+0x18c ithread_loop(c151c600,d68bdd48,c038de9b,354,c152ea90) at ithread_loop+0x16c fork_exit(c01efef0,c151c600,a68bdd48) at fork_exit+0xa5 fork_trampoline() at fork_trampoline+0x1a ---trap 0x1, eip=0, esp=0xd68bdd7c, ebp=0 --- -- Krzysztof Jędruczyk To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
