Date: Mon, 2 Jul 2012 15:42:30 -0700 From: Doug Hardie <bc979@lafn.org> To: Wojciech Puchar <wojtek@wojtek.tensor.gdynia.pl> Cc: "questions@FreeBSD.org Questions" <questions@freebsd.org> Subject: Re: FreeBSD 9.0 hang Message-ID: <144523C4-23BE-460C-9690-79E3B3EFB4C1@lafn.org> In-Reply-To: <alpine.BSF.2.00.1207021747140.4054@wojtek.tensor.gdynia.pl> References: <3BFFBC01-AF69-401F-AAB7-6F10781007B7@lafn.org> <alpine.BSF.2.00.1207021747140.4054@wojtek.tensor.gdynia.pl>
next in thread | previous in thread | raw e-mail | index | archive | help
On 2 July 2012, at 08:50, Wojciech Puchar wrote:
>>=20
>> link_eif symbol atm_event undefined
>> KLD if_en.ko: depends on utopia - not available or version mismatch.
Those were the last 2 lines on the console before the hang. There is =
nothing at all in messages about this. I suspect the system was not =
totally hung, just the etnernet interfaces (2 different ones) as I could =
still ping both interfaces successfully. However, no attempt to access =
any service worked.
>>=20
>> I haven't found anything relevant on those through Google. if_en.ko =
os present as is utopia.ko. I don't understand why the kernel would try =
to load if_en as I don't have any of those devices. There are em0 and =
dc0 ethernet interfaces. This is almost a generic kernel. The config =
file contains:
> Was that line printed just before hangup?
>=20
> Do you actually tried to load ATM interface driver. If no - check why =
it loads at all.
I have no ATM interfaces so it shouldn't be loading to my way of =
thinking.
>=20
> Check what is last in your logfile.
Nothing.
>=20
> Check if any addon drivers you use (fuse.ko, vboxdrv.ko) was compiled =
with the same kernel sources that you compiled kernel.
kldstat -v shows:
2 1 0xc5b36000 4000 fdescfs.ko (/boot/kernel/fdescfs.ko)
Contains modules:
Id Name
493 fdescfs
3 1 0xc5c8f000 3000 pflog.ko (/boot/kernel/pflog.ko)
Contains modules:
Id Name
495 pflog
4 1 0xc5c92000 34000 pf.ko (/boot/kernel/pf.ko)
Contains modules:
Id Name
494 pf
if_en is not listed as compiled into the kernel.
The kernel and userland were built shortly after an install from =
memstick image using the procedure in UPDATING:
To rebuild everything and install it on the current system.
-----------------------------------------------------------
# Note: sometimes if you are running current you gotta do more =
than
# is listed here if you are upgrading from a really old current.
<make sure you have good level 0 dumps>
make buildworld
make kernel KERNCONF=3DYOUR_KERNEL_HERE
[1]
<reboot in single user> [3]
mergemaster -p [5]
make installworld
mergemaster -i [4]
make delete-old [6]
<reboot>
After that the ports and application software were installed. Basically =
the only services that run on this system are nagios and mrtg. It is =
used only as a monitoring system for my production server and for =
testing new software. It has only been used for monitoring since the =
upgrade. I can't do any development work till I get the production =
servers upgraded from 8.2 to 9.0.
>=20
> If this doesn't help then recompile your kernel with
>=20
> makeoptions DEBUG=3D"-O0 -g"
> options INCLUDE_CONFIG_FILE
> options DEADLKRES
> options KDB
> options DDB
> options INVARIANTS
> options INVARIANT_SUPPORT
> options WITNESS
> options WITNESS_SKIPSPIN
> options DIAGNOSTIC
>=20
> make sure that dump device is active
>=20
> dumpon=3D"/dev/dumpdevicename"
>=20
> and reboot with that kernel.
>=20
> At next crash you will get full dump with all symbols and all data =
where it crashes.
When this repeats I will do that. Thanks for the help.
-- Doug
>=20
>=20
>=20
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?144523C4-23BE-460C-9690-79E3B3EFB4C1>