Re: 5.8-stable: panic: mtx_enter locking against myself

2015-09-24 Thread mxb
Looks like I found the root cause. At least it is stable as it suppose to be. In need to reproduce this in lab before making next move. //mxb > On 17 sep. 2015, at 10:35, mxb wrote: > > > Hey, > getting panics with 5.8-STABLE kernel. > > panic: mix_enter: locking

Re: 5.8-stable: panic: mtx_enter locking against myself

2015-09-17 Thread mxb
It is from CVS. dmesg below. OpenBSD 5.8-stable (GENERIC.MP) #0: Tue Sep 15 22:38:07 CEST 2015 r...@fw1.home.unixconn.com:/usr/src/sys/arch/amd64/compile/GENERIC.MP real mem = 17051418624 (16261MB) avail mem = 16530747392 (15764MB) mpath0 at root scsibus0 at mpath0: 256 targets mainbus0 at

5.8-stable: panic: mtx_enter locking against myself

2015-09-17 Thread mxb
Hey, getting panics with 5.8-STABLE kernel. panic: mix_enter: locking against myself Starting stack trace… panic() at panic+0x10b mtx_enter() at mtx_enter+0x60 sofree() at sofree+0xa0 in_pcbdetach() at in_pcbdetach+0x40 tcp_close() at tcp_close+0xad tcp_timer_2msl() at tcp_timer_2msl+0x90

Re: 5.8-stable: panic: mtx_enter locking against myself

2015-09-17 Thread mxb
and this is for a crash just 10min ago. (gdb) file /var/crash/bsd.0 Reading symbols from /var/crash/bsd.0...(no debugging symbols found)...done. (gdb) target kvm /var/crash/bsd.0.core #0 0x8131cae4 in dumpsys () (gdb) where #0 0x8131cae4 in dumpsys () #1 0x00030272 in

Re: 5.8-stable: panic: mtx_enter locking against myself

2015-09-17 Thread koko
On Thu, 17 Sep 2015 10:35:46 +0200 mxb wrote: > getting panics with 5.8-STABLE kernel. > 5.8-STABLE not released yet. you mean 5.8-CURRENT?

Re: 5.8-stable: panic: mtx_enter locking against myself

2015-09-17 Thread Stuart Henderson
On 2015-09-17, mxb wrote: > Hey, > getting panics with 5.8-STABLE kernel. > > panic: mix_enter: locking against myself > Starting stack trace… > panic() at panic+0x10b > mtx_enter() at mtx_enter+0x60 > sofree() at sofree+0xa0 > in_pcbdetach() at in_pcbdetach+0x40 >

Re: 5.8-stable: panic: mtx_enter locking against myself

2015-09-17 Thread mxb
Seems to happen very random. I have two systems which do this. Both are 2-node CARP. It is only kernel which is -STABLE on those two. Userland is from a snap approx. 14 days old. Update to stable kernel is done recently, because of those crashes. Kernel before is running OK on more heavily loaded