Bug 1262797

Summary: Unable to generate complete backtrace using coredump
Product: Red Hat Enterprise Linux 7 Reporter: jigar <jraising>
Component: abrtAssignee: abrt <abrt-devel-list>
Status: CLOSED INSUFFICIENT_DATA QA Contact: BaseOS QE - Apps <qe-baseos-apps>
Severity: medium Docs Contact:
Priority: unspecified    
Version: 7.1CC: jberan, jraising, mhabrnal, mkyral
Target Milestone: rc   
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2017-11-13 15:05:11 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On:    
Bug Blocks: 1298243, 1420851    

Description jigar 2015-09-14 11:47:11 UTC 
Description of problem: When trying to load the crash dump in gdb captured by abrt, following error messages are reported:

warning: Corrupted shared library list: 0x7f1a549fa730 != 0x15062e0

warning: Corrupted shared library list: 0x7f1a549fa730 != 0x15062e0

Following the error, the complete backtrace is not getting generated.

How reproducible: Sometimes

Steps to Reproduce:
Load the attached coredump in gdb for /usr/bin/kdeinit4 and try to generate a backtrace.

Actual results: Corrupted shared Library does not allow generation of complete backtrace.

Expected results: It should be able to generate complete backtrace. 

Additional info: Customer is using NVIDIA drivers, can it lead to package mismatches or corrupting shared library list?
Comment 2 Jakub Filak 2015-09-14 12:07:42 UTC 
Thank you for the report! I am afraid that ABRT cannot do much in this matter. It might happen that the kdeinit4 process messes up its own memory in the way that the core file cannot be successfully loaded by gdb. Can you please turn of abrt, generate several core files using pure kernel core file handler and try to generate backtrace from each of the core files?

# systemctl stop abrt-ccpp
# mkdir /var/tmp/cores/
# echo "/var/tmp/cores/core.%P" > /proc/sys/kernel/core_pattern
... reproduce the crash
# cd /var/tmp/cores/
# for c in $(ls core*); do gdb -c $c; done

I do not think NVIDIA kernel drivers can break user space process memory. You can try to run kdeinit4 in valgrind to detect memory errors, but that might be extremely time consuming operation.
Comment 3 Matej Habrnal 2017-11-13 15:05:11 UTC 
I am not able to reproduce this problem and it looks like it is third party software issue hence I am closing it.

Please feel free to reopen this bugzilla if you are able to reproduce it.