Re: Crash after world/kernel upgrade
> Jun 14 10:36:08 calvin syslogd: kernel boot file is /boot/kernel/kernel > Jun 14 10:36:08 calvin kernel: panic: lock (sleep mutex) inp not locked @ /usr/src/sys/netinet/tcp_subr.c:1069 Thanks for the bug report, Sid. This panic should be addressed by the following commit: hsu 2002/06/14 01:35:21 PDT Modified files: sys/netinet in_pcb.c in_pcb.h tcp_subr.c tcp_var.h udp_usrreq.c udp_var.h sys/netinet6 in6_pcb.c in6_pcb.h raw_ip6.c udp6_usrreq.c Log: Notify functions can destroy the pcb, so they have to return an indication of whether this happenned so the calling function knows whether or not to unlock the pcb. Submitted by: Jennifer Yang ([EMAIL PROTECTED]) Bug reported by: Sid Carter ([EMAIL PROTECTED]) Revision ChangesPath 1.108 +3 -2 src/sys/netinet/in_pcb.c 1.50 +3 -2 src/sys/netinet/in_pcb.h 1.132 +24 -15src/sys/netinet/tcp_subr.c 1.81 +6 -3 src/sys/netinet/tcp_var.h 1.113 +3 -2 src/sys/netinet/udp_usrreq.c 1.26 +2 -1 src/sys/netinet/udp_var.h 1.31 +3 -2 src/sys/netinet6/in6_pcb.c 1.8 +3 -2 src/sys/netinet6/in6_pcb.h 1.21 +1 -1 src/sys/netinet6/raw_ip6.c 1.28 +1 -1 src/sys/netinet6/udp6_usrreq.c To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Crash after world/kernel upgrade
That looks like a different problem. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Crash after world/kernel upgrade
An Fri, Jun 14, 2002 at 10:44:59AM +0530, Sid Carter schreib : > An Thu, Jun 13, 2002 at 02:19:52PM -0700, Jeffrey Hsu schreib : > > Duplicate inp warnings occur whenever 2 inp locks are held, even though they're > > on different inps, because all the inp locks are named "inp". I've > > investigated a couple of these already and they were all places where > > we intended to grab two inp locks. So, for the most part, these warning are > > harmless. As John said though, panics are serious, so if you get any of > > those, please report them to me directly, as I only monitor this list > > periodically. > Just came today and my computer has rebooted, asking me for the power-on > password. These are the warnings or panic messages I see. > > Jun 14 10:36:08 calvin kernel: syncing disks... panic: bremfree: bp > 0xc764e1a0 not locked > Jun 14 10:36:08 calvin kernel: Uptime: 9h53m57s > Jun 14 10:36:08 calvin kernel: pfs_vncache_unload(): 1 entries > remaining > Jun 14 10:36:08 calvin kernel: Terminate ACPI Hi, Similar Crash happened again. Jun 14 11:30:32 calvin syslogd: kernel boot file is /boot/kernel/kernel Jun 14 11:30:32 calvin kernel: Memory modified after free 0xd3fc5500(252) Jun 14 11:30:32 calvin kernel: panic: Most recently used by kqueue Jun 14 11:30:32 calvin kernel: Jun 14 11:30:32 calvin kernel: Jun 14 11:30:32 calvin kernel: syncing disks... panic: bremfree: bp0xc767f070 n ot locked Jun 14 11:30:32 calvin kernel: Uptime: 53m5s Jun 14 11:30:32 calvin kernel: Terminate ACPI Jun 14 11:30:32 calvin kernel: Automatic reboot in 15 seconds - press a key on the console to abort Regards Sid -- I've known him as a man, as an adolescent and as a child -- sometimes on the same day. Sid Carter FreeBSD oder Debian GNU/Linux. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Crash after world/kernel upgrade
An Thu, Jun 13, 2002 at 02:19:52PM -0700, Jeffrey Hsu schreib : > Duplicate inp warnings occur whenever 2 inp locks are held, even though they're > on different inps, because all the inp locks are named "inp". I've > investigated a couple of these already and they were all places where > we intended to grab two inp locks. So, for the most part, these warning are > harmless. As John said though, panics are serious, so if you get any of > those, please report them to me directly, as I only monitor this list > periodically. Hi, Just came today and my computer has rebooted, asking me for the power-on password. These are the warnings or panic messages I see. --- Jun 14 10:25:00 calvin kernel: /usr/src/sys/vm/uma_core.c:1327: could sleep with "process lock" locked from /usr/src/sys/kern/kern_prot.c:613 Jun 14 10:30:00 calvin kernel: /usr/src/sys/vm/uma_core.c:1327: could sleep with "process lock" locked from /usr/src/sys/kern/kern_prot.c:613 Jun 14 10:36:08 calvin syslogd: kernel boot file is /boot/kernel/kernel Jun 14 10:36:08 calvin kernel: panic: lock (sleep mutex) inp not locked @ /usr/src/sys/netinet/tcp_subr.c:1069 Jun 14 10:36:08 calvin kernel: Jun 14 10:36:08 calvin kernel: syncing disks... panic: bremfree: bp 0xc764e1a0 not locked Jun 14 10:36:08 calvin kernel: Uptime: 9h53m57s Jun 14 10:36:08 calvin kernel: pfs_vncache_unload(): 1 entries remaining Jun 14 10:36:08 calvin kernel: Terminate ACPI Jun 14 10:36:08 calvin kernel: Automatic reboot in 15 seconds - press a key on the console to abort Jun 14 10:36:08 calvin kernel: Rebooting... --- Regards Sid -- I've known him as a man, as an adolescent and as a child -- sometimes on the same day. Sid Carter FreeBSD oder Debian GNU/Linux. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Crash after world/kernel upgrade
On 13-Jun-2002 Jeffrey Hsu wrote: > Duplicate inp warnings occur whenever 2 inp locks are held, even though they're > on different inps, because all the inp locks are named "inp". I've > investigated a couple of these already and they were all places where > we intended to grab two inp locks. So, for the most part, these warning are > harmless. As John said though, panics are serious, so if you get any of > those, please report them to me directly, as I only monitor this list > periodically. How do you define the order between teh two inp locks to ensure you don't end up deadlocking? That is why the warning is there. Unless you have a well-defined lock order between inp locks you can easily get into deadlocks. -- John Baldwin <[EMAIL PROTECTED]> <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Crash after world/kernel upgrade
Duplicate inp warnings occur whenever 2 inp locks are held, even though they're on different inps, because all the inp locks are named "inp". I've investigated a couple of these already and they were all places where we intended to grab two inp locks. So, for the most part, these warning are harmless. As John said though, panics are serious, so if you get any of those, please report them to me directly, as I only monitor this list periodically. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Crash after world/kernel upgrade
> Jun 11 22:34:47 calvin kernel: lock order reversal > Jun 11 22:34:47 calvin kernel: 1st 0xd311c2f8 inp (inp) @ > /usr/src/sys/netinet/tcp_usrreq.c:581 > Jun 11 22:34:47 calvin kernel: 2nd 0xc04a5340 Giant (Giant) @ > /usr/src/sys/kern/subr_trap.c:76 > Jun 11 22:34:47 calvin kernel: exclusive sleep mutex inp r = 0 > (0xd311c2f8) locked @ /usr/src/sys/netinet/tcp_usrreq.c:581 > Jun 11 22:34:47 calvin kernel: panic: system call shutdown returning with mutex(s) held Some of the networking lock debug messages are supurious, but this one is real and was fixed late Tuesday night. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Crash after world/kernel upgrade
On 13-Jun-2002 Sid Carter wrote: > An Tue, Jun 11, 2002 at 11:06:44AM -0700, Edwin Culp schreib : >> Quoting Steve Ames <[EMAIL PROTECTED]>: >> | On Tue, Jun 11, 2002 at 10:51:25PM +0530, Sid Carter wrote: >> | > An Tue, Jun 11, 2002 at 04:58:39PM +0200, Thomas Ugland schreib : >> I get blown away with network setup with a Fatal trap 12, too. I've >> cvsuped and rebuilt twice with no change. > > Hi, > > Tell me about it. Luckily, I am able to get to FreeBSD after a fresh > cvsup and "make kernel". But I keep getting these errors from dmesg. s/errors/warnings/. If they were errors your kernel would panic. They are warnings of problems but your machine should still run fine. The duplicate lock of "inp" is a fairly big problem you should send to hsu@. -- John Baldwin <[EMAIL PROTECTED]> <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Crash after world/kernel upgrade
An Fri, Jun 14, 2002 at 12:43:51AM +0530, Sid Carter schreib : > An Tue, Jun 11, 2002 at 11:06:44AM -0700, Edwin Culp schreib : > > Quoting Steve Ames <[EMAIL PROTECTED]>: > > | On Tue, Jun 11, 2002 at 10:51:25PM +0530, Sid Carter wrote: > > | > An Tue, Jun 11, 2002 at 04:58:39PM +0200, Thomas Ugland schreib : > > I get blown away with network setup with a Fatal trap 12, too. I've > > cvsuped and rebuilt twice with no change. > I have to yet to start any further apps, cause I just rebooted. Hi, And this are some messages I get when I started "fetchmail". - Jun 14 00:44:00 calvin kernel: /usr/src/sys/vm/uma_core.c:1327: could sleep with "process lock" locked from /usr/src/sys/kern/kern_prot.c:511 Jun 14 00:44:04 calvin kernel: acquiring duplicate lock of same type: "inp" Jun 14 00:44:04 calvin kernel: 1st inp @ /usr/src/sys/netinet/tcp_input.c:631 Jun 14 00:44:04 calvin kernel: 2nd inp @ /usr/src/sys/netinet/tcp_usrreq.c:144 Jun 14 00:45:00 calvin kernel: /usr/src/sys/vm/uma_core.c:1327: could sleep with "process lock" locked from /usr/src/sys/kern/kern_prot.c:613 - Of course, previously my machine used to crash and reboot, whenever I started fetchmail with similar errors as pasted above. uname -a FreeBSD calvin 5.0-CURRENT FreeBSD 5.0-CURRENT #0: Fri Jun 14 05:46:24 IST 2002 root@:/usr/obj/usr/src/sys/GENERIC i386 Regards Sid -- I've known him as a man, as an adolescent and as a child -- sometimes on the same day. Sid Carter FreeBSD oder Debian GNU/Linux. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Crash after world/kernel upgrade
An Tue, Jun 11, 2002 at 11:06:44AM -0700, Edwin Culp schreib : > Quoting Steve Ames <[EMAIL PROTECTED]>: > | On Tue, Jun 11, 2002 at 10:51:25PM +0530, Sid Carter wrote: > | > An Tue, Jun 11, 2002 at 04:58:39PM +0200, Thomas Ugland schreib : > I get blown away with network setup with a Fatal trap 12, too. I've > cvsuped and rebuilt twice with no change. Hi, Tell me about it. Luckily, I am able to get to FreeBSD after a fresh cvsup and "make kernel". But I keep getting these errors from dmesg. -- Jun 14 00:36:27 calvin kernel: /usr/src/sys/vm/uma_core.c:1327: could sleep with "process lock" locked from /usr/src/sys/kern/kern_prot.c:511 Jun 14 00:36:30 calvin kernel: /usr/src/sys/vm/uma_core.c:1327: could sleep with "process lock" locked from /usr/src/sys/kern/kern_prot.c:511 Jun 14 00:36:31 calvin kernel: /usr/src/sys/vm/uma_core.c:1327: could sleep with "process lock" locked from /usr/src/sys/kern/kern_prot.c:613 Jun 14 00:36:33 calvin kernel: /usr/src/sys/vm/uma_core.c:1327: could sleep with "process lock" locked from /usr/src/sys/kern/kern_prot.c:867 Jun 14 00:36:38 calvin kernel: /usr/src/sys/vm/uma_core.c:1327: could sleep with "process lock" locked from /usr/src/sys/kern/kern_prot.c:867 Jun 14 00:36:55 calvin last message repeated 5 times Jun 14 00:37:21 calvin kernel: /usr/src/sys/vm/uma_core.c:1327: could sleep with "inp" locked from /usr/src/sys/netinet/tcp_input.c:631 Jun 14 00:40:00 calvin kernel: /usr/src/sys/vm/uma_core.c:1327: could sleep with "process lock" locked from /usr/src/sys/kern/kern_prot.c:613 -- I have to yet to start any further apps, cause I just rebooted. Let me see how long I can stay without a reboot. Regards Sid -- I've known him as a man, as an adolescent and as a child -- sometimes on the same day. Sid Carter FreeBSD oder Debian GNU/Linux. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Crash after world/kernel upgrade
> typo in unlocking (causing recursive lock instead) Fixed. Thanks. > lack of inet6 support for inpcb locking, e.g. no > handling of locks in in6_pcbdetach. > attempts to destroy held lock in in_pcbdetach Gurg, IPv6 isn't locked up yet! These must be the result of inadvertent interactions with the shared IPv4 code. I'll see what can be done to either mitigate the effects for IPv6 or quickly lock up v6. Jeffrey To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Crash after world/kernel upgrade
On Tue, Jun 11, 2002 at 12:28:48PM -0500, Steve Ames wrote: > > You got way more info than I did. With this mornings -CURRENT (around 9AM EDT): > > Jun 11 09:22:10 : Doing initial network setup: > Jun 11 09:22:10 : host.conf > Jun 11 09:22:10 : hostname > Jun 11 09:22:10 : . > Jun 11 09:22:10 : > Jun 11 09:22:10 : > Jun 11 09:22:10 : Fatal trap 12: page fault while in kernel mode > Jun 11 09:22:10 : fault virtual address = 0x28 > Jun 11 09:22:10 : fault code = supervisor read, page not present > Jun 11 09:22:10 : instruction pointer = 0x8:0xc01851c1 > Jun 11 09:22:10 : stack pointer = 0x10:0xd7865bdc > Jun 11 09:22:10 : frame pointer = 0x10:0xd7865bec > Jun 11 09:22:10 : code segment= base 0x0, limit 0xf, type 0x1b > Jun 11 09:22:10 : = DPL 0, pres 1, def32 1, gran 1 > Jun 11 09:22:10 : processor eflags= interrupt enabled, resume, IOPL = 0 > Jun 11 09:22:10 : current process = 11 (swi1: net) > Jun 11 09:22:10 : trap number = 12 > Jun 11 09:22:10 : panic: page fault > Jun 11 09:22:10 : > Jun 11 09:22:10 : syncing disks... panic: bdwrite: buffer is not busy > Jun 11 09:22:10 : Uptime: 3s > Jun 11 09:22:10 : pfs_vncache_unload(): 1 entries remaining > Jun 11 09:22:10 : Terminate ACPI > Jun 11 09:22:10 : Automatic reboot in 15 seconds - press a key on the console to >abort > My bad, this is exactly what I was referring to in the other thread. -- Munish Chopra The FreeBSD NVIDIA Driver Initiative http://nvidia.netexplorer.org To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Crash after world/kernel upgrade
Quoting Steve Ames <[EMAIL PROTECTED]>: | On Tue, Jun 11, 2002 at 10:51:25PM +0530, Sid Carter wrote: | > An Tue, Jun 11, 2002 at 04:58:39PM +0200, Thomas Ugland schreib : | > > System crashed after updating today. | > > During the start of system services, in specific | > > at the start of sendmail the system crashes with | > > the new kernel. :/ | > > uname -a: | > > - | > > FreeBSD vampire.lothlorien.no 5.0-CURRENT FreeBSD 5.0-CURRENT #0: Tue | > > Jun 11 15:41:09 CEST 2002 root@:/usr/src/sys/i386/compile/VAMPIRE | > > i386 | > | > Hi, | > | > Just rebooted after a crash and am presently doing a cvsup for a | > rebuild. I got similar error messages and this is what my | > /var/log/messages shows. | > | | You got way more info than I did. With this mornings -CURRENT (around 9AM | EDT): | I get blown away with network setup with a Fatal trap 12, too. I've cvsuped and rebuilt twice with no change. ed | Jun 11 09:22:10 : Doing initial network setup: | Jun 11 09:22:10 : host.conf | Jun 11 09:22:10 : hostname | Jun 11 09:22:10 : . | Jun 11 09:22:10 : | Jun 11 09:22:10 : | Jun 11 09:22:10 : Fatal trap 12: page fault while in kernel mode | Jun 11 09:22:10 : fault virtual address = 0x28 | Jun 11 09:22:10 : fault code = supervisor read, page not present | Jun 11 09:22:10 : instruction pointer = 0x8:0xc01851c1 | Jun 11 09:22:10 : stack pointer = 0x10:0xd7865bdc | Jun 11 09:22:10 : frame pointer = 0x10:0xd7865bec | Jun 11 09:22:10 : code segment= base 0x0, limit 0xf, type 0x1b | Jun 11 09:22:10 : = DPL 0, pres 1, def32 1, gran 1 | Jun 11 09:22:10 : processor eflags= interrupt enabled, resume, IOPL = 0 | Jun 11 09:22:10 : current process = 11 (swi1: net) | Jun 11 09:22:10 : trap number = 12 | Jun 11 09:22:10 : panic: page fault | Jun 11 09:22:10 : | Jun 11 09:22:10 : syncing disks... panic: bdwrite: buffer is not busy | Jun 11 09:22:10 : Uptime: 3s | Jun 11 09:22:10 : pfs_vncache_unload(): 1 entries remaining | Jun 11 09:22:10 : Terminate ACPI | Jun 11 09:22:10 : Automatic reboot in 15 seconds - press a key on the console | to abort | | To Unsubscribe: send mail to [EMAIL PROTECTED] | with "unsubscribe freebsd-current" in the body of the message -- To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Crash after world/kernel upgrade
On Tue, Jun 11, 2002 at 10:51:25PM +0530, Sid Carter wrote: > An Tue, Jun 11, 2002 at 04:58:39PM +0200, Thomas Ugland schreib : > > System crashed after updating today. > > During the start of system services, in specific > > at the start of sendmail the system crashes with > > the new kernel. :/ > > uname -a: > > - > > FreeBSD vampire.lothlorien.no 5.0-CURRENT FreeBSD 5.0-CURRENT #0: Tue > > Jun 11 15:41:09 CEST 2002 root@:/usr/src/sys/i386/compile/VAMPIRE > > i386 > > Hi, > > Just rebooted after a crash and am presently doing a cvsup for a > rebuild. I got similar error messages and this is what my > /var/log/messages shows. > You got way more info than I did. With this mornings -CURRENT (around 9AM EDT): Jun 11 09:22:10 : Doing initial network setup: Jun 11 09:22:10 : host.conf Jun 11 09:22:10 : hostname Jun 11 09:22:10 : . Jun 11 09:22:10 : Jun 11 09:22:10 : Jun 11 09:22:10 : Fatal trap 12: page fault while in kernel mode Jun 11 09:22:10 : fault virtual address = 0x28 Jun 11 09:22:10 : fault code= supervisor read, page not present Jun 11 09:22:10 : instruction pointer = 0x8:0xc01851c1 Jun 11 09:22:10 : stack pointer = 0x10:0xd7865bdc Jun 11 09:22:10 : frame pointer = 0x10:0xd7865bec Jun 11 09:22:10 : code segment = base 0x0, limit 0xf, type 0x1b Jun 11 09:22:10 : = DPL 0, pres 1, def32 1, gran 1 Jun 11 09:22:10 : processor eflags = interrupt enabled, resume, IOPL = 0 Jun 11 09:22:10 : current process = 11 (swi1: net) Jun 11 09:22:10 : trap number = 12 Jun 11 09:22:10 : panic: page fault Jun 11 09:22:10 : Jun 11 09:22:10 : syncing disks... panic: bdwrite: buffer is not busy Jun 11 09:22:10 : Uptime: 3s Jun 11 09:22:10 : pfs_vncache_unload(): 1 entries remaining Jun 11 09:22:10 : Terminate ACPI Jun 11 09:22:10 : Automatic reboot in 15 seconds - press a key on the console to abort To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Crash after world/kernel upgrade
An Tue, Jun 11, 2002 at 04:58:39PM +0200, Thomas Ugland schreib : > System crashed after updating today. > During the start of system services, in specific > at the start of sendmail the system crashes with > the new kernel. :/ > uname -a: > - > FreeBSD vampire.lothlorien.no 5.0-CURRENT FreeBSD 5.0-CURRENT #0: Tue > Jun 11 15:41:09 CEST 2002 root@:/usr/src/sys/i386/compile/VAMPIRE > i386 Hi, Just rebooted after a crash and am presently doing a cvsup for a rebuild. I got similar error messages and this is what my /var/log/messages shows. -- Jun 11 22:29:29 calvin kernel: /usr/src/sys/vm/uma_core.c:1327: could sleep with "process lock" locked from /usr/src/sys/kern/kern_prot.c:511 Jun 11 22:29:38 calvin kernel: /usr/src/sys/vm/uma_core.c:1327: could sleep with "inp" locked from /usr/src/sys/netinet/tcp_usrreq.c:650 Jun 11 22:29:38 calvin kernel: /usr/src/sys/vm/uma_core.c:1327: could sleep with "tcp" locked from /usr/src/sys/netinet/tcp_usrreq.c:633 Jun 11 22:29:38 calvin kernel: acquiring duplicate lock of same type: "inp" Jun 11 22:29:38 calvin kernel: 1st inp @ /usr/src/sys/netinet/tcp_input.c:631 Jun 11 22:29:38 calvin kernel: 2nd inp @ /usr/src/sys/netinet/tcp_usrreq.c:144 Jun 11 22:34:47 calvin syslogd: kernel boot file is /boot/kernel/kernel Jun 11 22:34:47 calvin kernel: lock order reversal Jun 11 22:34:47 calvin kernel: 1st 0xd311c2f8 inp (inp) @ /usr/src/sys/netinet/tcp_usrreq.c:581 Jun 11 22:34:47 calvin kernel: 2nd 0xc04a5340 Giant (Giant) @ /usr/src/sys/kern/subr_trap.c:76 Jun 11 22:34:47 calvin kernel: exclusive sleep mutex inp r = 0 (0xd311c2f8) locked @ /usr/src/sys/netinet/tcp_usrreq.c:581 Jun 11 22:34:47 calvin kernel: panic: system call shutdown returning with mutex(s) held -- uname -a FreeBSD calvin 5.0-CURRENT FreeBSD 5.0-CURRENT #0: Tue Jun 11 21:20:29 IST 2002 root@calvin:/usr/obj/usr/src/sys/GENERIC i386 Regards Sid -- I've known him as a man, as an adolescent and as a child -- sometimes on the same day. Sid Carter FreeBSD oder Debian GNU/Linux. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Crash after world/kernel upgrade
> System crashed after updating today. > During the start of system services, in specific > at the start of sendmail the system crashes with > the new kernel. :/ There are some problems with the inpcb locking: - attempts to destroy held lock in in_pcbdetach. - typo in unlocking (causing recursive lock instead) - lack of inet6 support for inpcb locking, e.g. no handling of locks in in6_pcbdetach. I had to comment out INET6 from my kernel config file and apply the enclosed patch to get my machine to boot today. - Tor Egge Index: sys/netinet/in_pcb.c === RCS file: /home/ncvs/src/sys/netinet/in_pcb.c,v retrieving revision 1.106 diff -u -r1.106 in_pcb.c --- sys/netinet/in_pcb.c10 Jun 2002 20:05:36 - 1.106 +++ sys/netinet/in_pcb.c11 Jun 2002 16:13:29 - @@ -573,6 +573,11 @@ rtfree(inp->inp_route.ro_rt); ip_freemoptions(inp->inp_moptions); inp->inp_vflag = 0; + /* XXX: Kludge: Unlock inp before crashing */ + if (mtx_owned(&inp->inp_mtx)) { + printf("Warning: INP_LOCK held in in_pcbdetach\n"); + INP_UNLOCK(inp); + } INP_LOCK_DESTROY(inp); uma_zfree(ipi->ipi_zone, inp); } @@ -741,7 +746,7 @@ } INP_UNLOCK(inp); } - INP_INFO_RLOCK(pcbinfo); + INP_INFO_RUNLOCK(pcbinfo); } /*
Crash after world/kernel upgrade
System crashed after updating today. During the start of system services, in specific at the start of sendmail the system crashes with the new kernel. :/ uname -a: - FreeBSD vampire.lothlorien.no 5.0-CURRENT FreeBSD 5.0-CURRENT #0: Tue Jun 11 15:41:09 CEST 2002 root@:/usr/src/sys/i386/compile/VAMPIRE i386 gdb/bt: --- GNU gdb 5.2 (FreeBSD) This GDB was configured as "i386-portbld-freebsd5.0"... IdlePTD at phsyical address 0x0050d000 initial pcb at physical address 0x003d9b80 panicstr: from debugger panic messages: --- panic: system call close returning with mutex(s) held panic: from debugger Uptime: 1m15s Dumping 255 MB ata0: resetting devices .. ad0: DMA limited to UDMA33, non-ATA66 cable or device done 16 32 48 64 80 96 112 128 144 160 176 192 208 224 240 --- #0 doadump () at /usr/src/sys/kern/kern_shutdown.c:213 213 dumping++; (kgdb) bt #0 doadump () at /usr/src/sys/kern/kern_shutdown.c:213 #1 0xc0202024 in boot (howto=260) at /usr/src/sys/kern/kern_shutdown.c:346 #2 0xc0202201 in panic () at /usr/src/sys/kern/kern_shutdown.c:490 #3 0xc0143dd4 in db_command_loop () at /usr/src/sys/ddb/db_command.c:449 #4 0xc0143d69 in db_command (last_cmdp=0xc03810a0, cmd_table=0xcdffeb10, aux_cmd_tablep=0xc0143dd4, aux_cmd_tablep_end=0xc0339bc8) at /usr/src/sys/ddb/db_command.c:345 #5 0xc0143e4c in db_command_loop () at /usr/src/sys/ddb/db_command.c:471 #6 0xc014656c in db_trap (type=3, code=0) at /usr/src/sys/ddb/db_trap.c:72 #7 0xc030efeb in kdb_trap (type=3, code=0, regs=0xcdffec78) at /usr/src/sys/i386/i386/db_interface.c:161 #8 0xc031e231 in trap (frame= {tf_fs = 24, tf_es = 16, tf_ds = 16, tf_edi = -838889472, tf_esi = 256, tf_ebp = -838865724, tf_isp = -838865756, tf_ebx = 0, tf_edx = 0, tf_ecx = 32, tf_eax = 18, tf_trapno = 3, tf_err = 0, tf_eip = -1070534051, tf_cs = 8, tf_eflags = 662, tf_esp = -1070127393, tf_ss = -1070254436}) at /usr/src/sys/i386/i386/trap.c:585 #9 0xc030f25d in Debugger (msg=0x0) at cpufunc.h:68 #10 0xc02021ea in panic (fmt=0x0) at /usr/src/sys/kern/kern_shutdown.c:477 #11 0xc031ec5e in syscall (frame= {tf_fs = 47, tf_es = 47, tf_ds = 47, tf_edi = 135235593, tf_esi = 420, tf_ebp = -1077943416, tf_isp = -838865548, tf_ebx = 673845760, tf_edx = 135217248, tf_ecx = 0, tf_eax = 0, tf_trapno = 12, tf_err = 2, tf_eip = 673441123, tf_cs = 31, tf_eflags = 518, tf_esp = -1077943460, tf_ss = 47}) at /usr/src/sys/i386/i386/trap.c:1081 #12 0xc031060d in syscall_with_err_pushed () at {standard input}:128 #13 0x0806c149 in ?? () #14 0x0804c08d in ?? () (kgdb) -- Thomas Ugland -> If everything is coming your way then you're in the wrong lane. <- signature.asc Description: This is a digitally signed message part