5.6: tips on analyzing crash dumps

5.6: tips on analyzing crash dumps

Post by Brian C Hi » Thu, 23 Apr 1998 04:00:00



        I have a some system crash dumps on a dual Ultra-II that I want
to analyze at least enough to get a clue of where the problem is coming
from. I am going to take advantage of the warranty support we have on the
system, but it takes forever, and I want to at least get some clues now.
This is all that I know how to do:

$ adb -k unix.0 vmcore.0
physmem
1f541
$c
complete_panic(?) + 24
do_panic(0x1,0x30042fa4,0x1040c028,0x0,0x20,0x0)
vcmn_err(0x3,0x1040d048,0x3,0x30042fa4,0x30043e80,0x4) + 190
cmn_err(0x3,0x1040d048,0x9,0x0,0x1040d2b8,0x9) + 1c
die(0x31,0x30043188,0x78,0x0,0x1040d048,0x0) + a0
trap(0x30043188,0x1,0x0,0x1,0x0,0x5) + 81c

        Can someone point me to web pages or give me some other clues
to go about doing this?

Thanks

   _____________________________________________________________________

  | Systems Programmer          University of California, Davis         |
  | Unix Specialist             BCH Technical Services                  |

 
 
 

5.6: tips on analyzing crash dumps

Post by David Hart » Fri, 24 Apr 1998 04:00:00


Couple of Tips,

Go on a Course - there is Crash Dump Analysis course by Sun,
by a book,
http://sunsite.queensu.ca/sunmicro/sunpress/books/Drake/Drake.html

other than that there are several documents on Sunsolve,
there is also a simple initial analysis script called iscda available
from Sun for Preliminary Analysis.

But First of all download and install all the recommended patches, 90%
of problems have been seen and fixed before, with patches, and Sun will
request this to be done prior inspection.


>         I have a some system crash dumps on a dual Ultra-II that I want
> to analyze at least enough to get a clue of where the problem is coming
> from. I am going to take advantage of the warranty support we have on the
> system, but it takes forever, and I want to at least get some clues now.
> This is all that I know how to do:

> $ adb -k unix.0 vmcore.0
> physmem
> 1f541
> $c
> complete_panic(?) + 24
> do_panic(0x1,0x30042fa4,0x1040c028,0x0,0x20,0x0)
> vcmn_err(0x3,0x1040d048,0x3,0x30042fa4,0x30043e80,0x4) + 190
> cmn_err(0x3,0x1040d048,0x9,0x0,0x1040d2b8,0x9) + 1c
> die(0x31,0x30043188,0x78,0x0,0x1040d048,0x0) + a0
> trap(0x30043188,0x1,0x0,0x1,0x0,0x5) + 81c

>         Can someone point me to web pages or give me some other clues
> to go about doing this?

Dave.
*Sometimes I don't know what I'm talking about, the rest of the time I
know i don't*

 
 
 

5.6: tips on analyzing crash dumps

Post by D. J. Hagbe » Fri, 24 Apr 1998 04:00:00



:       I have a some system crash dumps on a dual Ultra-II that I want
: to analyze at least enough to get a clue of where the problem is coming
: from. I am going to take advantage of the warranty support we have on the
: system, but it takes forever, and I want to at least get some clues now.
: This is all that I know how to do:

You might want to take a look at:

http://www.sun.com/smi/ssoftpress/books/Drake/brown/macros.html

but I don't know if these will work with 2.6.  I would think so.

                        -=- D. J.

 
 
 

5.6: tips on analyzing crash dumps

Post by y1gk » Thu, 30 Apr 1998 04:00:00


cockroff's book "Crash" will guide you....


> I have a some system crash dumps on a dual Ultra-II that I want
>to analyze at least enough to get a clue of where the problem is coming
>from. I am going to take advantage of the warranty support we have on the
>system, but it takes forever, and I want to at least get some clues now.
>This is all that I know how to do:

>$ adb -k unix.0 vmcore.0
>physmem
>1f541
>$c
>complete_panic(?) + 24
>do_panic(0x1,0x30042fa4,0x1040c028,0x0,0x20,0x0)
>vcmn_err(0x3,0x1040d048,0x3,0x30042fa4,0x30043e80,0x4) + 190
>cmn_err(0x3,0x1040d048,0x9,0x0,0x1040d2b8,0x9) + 1c
>die(0x31,0x30043188,0x78,0x0,0x1040d048,0x0) + a0
>trap(0x30043188,0x1,0x0,0x1,0x0,0x5) + 81c

> Can someone point me to web pages or give me some other clues
>to go about doing this?

>Thanks

>   _____________________________________________________________________

>  | Systems Programmer University of California, Davis    |
>  | Unix Specialist BCH Technical Services |

 
 
 

5.6: tips on analyzing crash dumps

Post by Dom Mitchel » Fri, 01 May 1998 04:00:00



Quote:>    I have a some system crash dumps on a dual Ultra-II that I want
> to analyze at least enough to get a clue of where the problem is coming
> from. I am going to take advantage of the warranty support we have on the
> system, but it takes forever, and I want to at least get some clues now.

Try getting hold of the iscda script.  It's on your sunsolve cd's and
it's also available on the web as
<URL:http://sunsolve.sun.com/sunsolve/iscda.html>.
--
Dom Mitchell          -- Palmer & Harvey McLane

 
 
 

1. SunOS 5.6/SUNW,Ultra-2: fails to core dump after crash

        I have a SUNW,Ultra-2 that keeps crashing. Here is
the console transaction....

{1} ok go
Fast Data Access MMU Miss
{1} ok sync
panic[cpu1]/thread=0x30043e80: zero
syncing file systems...BAD TRAP: cpu=1 type=0x31 rp=0xfffb18d8 addr=0x10 mmu_fsr
=0x0
sched: trap type = 0x31
addr=0x10
pid=0, pc=0x10058208, sp=0xfffb1968, tstate=0x4480001e06, context=0x0
g1-g7: 300, 1042ff94, 850000000ea56636, 22, f004052c, fffeff68, 30043e80
panic[cpu1]/thread=0x30043e80: trap
10841 static and sysmap kernel pages
   65 dynamic kernel data pages
  716 kernel-pageable pages
    0 segkmap kernel pages
    0 segvn kernel pages
    0 current user process pages
11622 total pages (11622 chunks)

dumping to vp 601b690c, offset 863744
BAD TRAP: cpu=1 type=0x31 rp=0xfffb1110 addr=0x0 mmu_fsr=0x0
BAD TRAP occurred in module "sd" due to an illegal access to a user address.
sched: trap type = 0x31
pid=0, pc=0x601c3118, sp=0xfffb11a0, tstate=0x1e05, context=0x0
g1-g7: 602252d0, 6006e9c8, c3, 3, f004052c, fffeff68, 30043e80
panic[cpu1]/thread=0x30043e80: trap
Dump Aborted.
{1} boot
....
Sun Ultra 2 UPA/SBus (2 X UltraSPARC-II 296MHz), No Keyboard
OpenBoot 3.7, 768 MB memory installed, Serial #9430213.
Ethernet address 8:0:20:8f:e4:c5, Host ID: 808fe4c5.
....

successful boot, but now savecore has nothing to grab...

        Any clues appreciated...

   _____________________________________________________________________

  | Systems Programmer          University of California, Davis         |
  | Unix Specialist             BCH Technical Services                  |

2. Linux VCD player ?

3. analysing crash dumps

4. localhost login:? what the hell?

5. Can't analyse crash dump...

6. SPK and X-Based Login

7. Crash dump analyzing

8. Solaris 2.6 with no configured swap

9. 5.6: niscat core dump

10. Wanted: Crash Dump Analysis Tips

11. SunOS 5.6 crashes

12. SunOS 5.6 / x86: mysterious crashes

13. How can I dump the contents of a crash/dump part'n onto floppy?