system panics with bad mutex

system panics with bad mutex

Post by Thomas Wol » Thu, 20 Jan 2000 04:00:00



Hi there,

We have a intel box running Sol 2.6 for years without any problems.
Recently,
we did the following upgrades:
Memory from 64->80
SCSI from AHA2940UW -> AHA2940U2W
additional IBM 9,1 U2WLVD Disk
Sol 2.6 -> Sol 7

Now the machine reboots from time to time, most likely when there is load
caused by a batch creating large database reports over the weekend.

The error found in /var/adm/messages is:

Jan 15 18:20:36 solbox unix: panic:
Jan 15 18:20:36 solbox: mutex_destroy: bad mutex, lp=fecce5dc owner=e0b777
a0 thread=ef577e60
Jan 15 18:20:36 solbox unix:
Jan 15 18:20:36 solbox: syncing file systems...

'lp=' is alwas the same
'owner' changes
'thread' is alwas the same

After rebooting, the batch restarts automatically, and succeds the next time
or
the next ...

Anybody has any hints where to start solving this problem ?

TIA
Thomas Wolf

 
 
 

system panics with bad mutex

Post by Russell A. Johnso » Thu, 20 Jan 2000 04:00:00


Is your system setup to save crash dumps?  Run dumpadm on Solaris 7 to
check or configure it.  After the system reboots from the next crash, a
system core dump will exist whereever dumpadm was told to put it.  Run
adb or crash on the system core dump and figure out what the stack was
like when the crash occurred.  That might give a clue as to what
application or kernel driver caused the mutex failure.

Mutexes are used within device drivers so because you changed your SCSI
card, you might need an updated SCSI driver.  You could check to see if
one was available and update the driver to see if that helps...

Russ


> SCSI from AHA2940UW -> AHA2940U2W
> Sol 2.6 -> Sol 7

> Jan 15 18:20:36 solbox: mutex_destroy: bad mutex, lp=fecce5dc owner=e0b777
> a0 thread=ef577e60
> Anybody has any hints where to start solving this problem ?


 
 
 

system panics with bad mutex

Post by Roland Main » Thu, 20 Jan 2000 04:00:00


Hi !

----


> We have a intel box running Sol 2.6 for years without any problems.
> Recently,
> we did the following upgrades:
> Memory from 64->80
> SCSI from AHA2940UW -> AHA2940U2W
> additional IBM 9,1 U2WLVD Disk
> Sol 2.6 -> Sol 7

> Now the machine reboots from time to time, most likely when there is load
> caused by a batch creating large database reports over the weekend.

> Anybody has any hints where to start solving this problem ?

Install the recent kernel patch for Solaris 2.7 Intel, see
http://sunsolve.sun.de/pub-cgi/show.pl?target=patches/xos-7_x86&nav=p...

This fixes serveral "recursive mutex enter"-related panics...

If the problem still occurs after patching, see "Russell A. Johnson"

Other solution: Buy a SPARC =:-)

----

Bye,
Roland

--
  __ .  . __


  /O /==\ O\  MPEG specialist, C&&JAVA&&Sun&&Unix programmer
 (;O/ \/ \O;) TEL +49 (0) 641/99-13193 FAX +49 (0) 641/99-41359

 
 
 

1. PANIC PANIC BAD DIR INO ............

Hello
i v got a problem with my hp-ux station during the process of boot.I ve
always got this type of message;

BAD DIR INO 79948 AT OFFSET 0
MANGLED ENTRY..:BAD DIR

I CAN T remove this directory wich had corrupted my filesystem.WHAT
SHOULD I DO?
BEST REGARDS

2. rexec

3. SYSTEM CRASH: CPU PANIC : BAD TRAP !!!!!

4. QT Designer Crash

5. recursive mutex panic doing ftp

6. Boot Problem

7. mutex_enter: bad mutex

8. Driver for digitizer

9. This is disconcerting: System panic: panic (cpu 0): Processor Machine Check

10. System Tuning & Mutex

11. Mutex failure on MP systems?!

12. 014 Bad Bad Bad !!! for Linux

13. Bad, bad, bad VM behaviour in 2.4.10