1. compile the source with the debug switch on gcc -g -o myprog myprog.c
2. To force a crash run your program, switch to another process , then kill -s SEGV [pid]
where pid == the process id of the running program
3. get into gdb
ba is backtrace which shows you the contents of the stack at the time the program crashed.
hi friends,
i know that when there is a crash then that memory image is
put into /var/adm/crash
but if the system hangs up and if i have access to console of
that machine then how can i take the crash dump manully.
thanks (2 Replies)
Is it true that you can't have the crash dump server/client on the same server?
I know I've installed Nagios open source before, I though it's only for that kind of thing. I never though that Red hat ent 4 would be like client/server on the crash dump.
if someone is having problem with high... (0 Replies)
I am initiating the discussion.Hope everyone will add on to this.
Whenever any application crashes the system's crash reporter(/System/Library/CoreServices/Crash Reporter.app) creates a crash dump.A crash dump is the image of the state of the kernel that was in physical memory when the system... (0 Replies)
My sparc solaris 8 server crashed/rebooted yesterday and I have the vmcore files. I have some initial output from SCAT and ACT. I have not included all but any info would be helpful. I can supply more output if necessary.
Thanks you.
Can I have any information about this ACT and SCAT... (2 Replies)
Hi,
I'm running Solaris 10 with a zone called "testzone"
If I do a "reboot -d" on the host, as expected the kernet panics, reboots and creates a crash dump file in /var/crash
However no crash file is created in testzone.
My question is how can I tell if a zone crashs or shuts down... (4 Replies)
Hi gurus,
I will be glad if anyone can help me with this:
How do you copy a crash dump file to send to your support provider?
Thanks lots guys. (1 Reply)
Hello Guys,
I need a little help here. I have been studying crash dump and per what I am reading, you can dedicate a slice to use as a dump device. Now when you dedicate this slice, do you have to :
1) create a mount point?
2) add entry in /etc/vfstab?
3) is this slice wu or wm?
4) should... (3 Replies)
Hi Experts,
I have configured HP-UX Service Guard cluster and it dumps crash every time i reboot a cluster node. Can anyone please help me to prevent these unnecessary crash dumps at the time of rebooting SG cluster node?
Thanks in advance.
Vaishey (2 Replies)
Discussion started by: Vaishey
2 Replies
LEARN ABOUT BSD
crash
CRASH(L) CRASH(L)
NAME
crash - analyze kernel dump or active system image
SYNOPSIS
crash [ aps ] [ -bdvtiz ] [ -s sfile ] [ -c cfile ] [ -u addr ]
DESCRIPTION
Synopsis of options -
-b Brief mode; skip general display of processes
-d Crash dump contains swap image. (default?)
-v Verbose; dump much information about each proc [future]
-t TTY structs to be dumpped
-i Incore inode table to be printed
-c FILE Provide non-standard file name for system image input
-s FILE Provide non-standard symboltable input
-u ADDR Trace a process other than currently selected one
-z Interrupt Trace displayed
aps Print PS & PC at time of interupt (doesn't do anything)
crash examines a dump of unix which it looks for in the file sysdump. It prints out the contents of the general registers, the kernel
stack and a traceback through the kernel stack. If an aps is specified, the ps and pc at time of interrupt are also printed out. The dump
of the stack commences from a "reasonable" address and all addresses are relocated to virtual addresses by using the value of kdsa6 found
in the dump.
The following options may be specified.
-b Brief mode; skip general display of processes. Only the currently selected process will be traced.
-c cfile
If the -c argument is found, the following argument is taken to be the name of a file containing the system image. The default is
"/usr/sys/core".
-s sfile
If the -s argument is found the following argument is taken to be the name of a file, containing a symbol table which should be used
in interpreting text addresses. The default is "/unix".
-u addr
Force a process to be displayed as if it were active when the crash occurred. addr is the octal address of the proc table entry.
FILES
/unix system namelist
/dev/swap swap device
/usr/sys/core core file
SEE ALSO crash(8), adb(1), ps(1), rstat(1)DIAGNOSTICS
Missing arg Unable to Open file Format Error in symbolfile
BUGS
Care should be used in running crash on "/dev/mem". Things can change while crash is running; the picture it gives is only a close approx-
imation to reality.
AUTHOR
Mike Muuss, JHU EE
John Stewart, Teledyne Geotech
March 1983 CRASH(L)