[minicoredumper] Invalid stack when all memory regions are dumped

[Abhi Arjun]

I bumped up  stack size of minicoredumper by modifying main() of corestripper.c to 128MB. Still I see the issue. 

Pls note that, I have removed compression section from recept file completely, as I was seeing  below errors while trying to decompress.

debug.txt file is sane. I don't see any debugs recursively printed there.
Regards,
Abhi

root# gzip -d core.tar.gz 
root# ls
core  core.tar  debug.txt
root# tar -xvf core.tar
core
tar: core: invalid sparse archive member
tar: Exiting with failure status due to previous errors


Jul 19, 2021, 22:00 by abhiarjun at tutanota.com:

>
> Thanks for your reply. 
> I just use a simple test program.  Stack size is 8KB
> Regards,
> Abhi
>
> ~/mini$ cat cd.c
> #include <stdio.h>
> #include <stdlib.h>
> #include <assert.h>
>
> int main()
> {
>     int a=10;
>     printf("Test Program\n");
>     assert(0);
> }
> ~/mini$ ulimit -a
> stack size              (kbytes, -s) 8192
>
>
>
>  
>
> -- 
> Sent with Tutanota, the secure & ad-free mailbox: 
> https://tutanota.com
>
>
> Jul 19, 2021, 21:09 by john.ogness at linutronix.de:
>
>> Hello Abhi Arjun,
>>
>> On 2021-07-08, Abhi Arjun <abhiarjun at tutanota.com> wrote:
>>
>>> Im the recept file, under maps section if only "[vdso]" is present
>>> then I get proper stack in gdb. However if I add "*" to include all
>>> memory regions (that's my requirement), then stack becomes
>>> invalid/corrupted as shown below. I just have main() function in the
>>> program.
>>>
>>
>> I could not reproduce the problem. When using a wildcard, the matching
>> function uses recursion. Maybe you are having stack space issues?
>>
>> Can you provide an example program that causes this?
>>
>> John Ogness
>>
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linutronix.de/pipermail/minicoredumper/attachments/20210719/e2c70443/attachment.htm>