Bug 383312

Summary: k45.suse.de: Machine with Beta6 (#2) kernel hangs after gtod_infinite testrun was killed
Product: [SUSE Linux Enterprise Real Time Extension] SUSE Linux Enterprise Real Time 10 SP2 (SLERT 10 SP2) Reporter: Felix Foerster <ffoerster>
Component: kernelAssignee: Ihno Krumreich <ihno>
Status: RESOLVED DUPLICATE QA Contact: Erik Hamera <erik.hamera>
Severity: Major    
Priority: P5 - None CC: ffoerster
Version: BETA6   
Target Milestone: ---   
Hardware: x86-64   
OS: Other   
Whiteboard:
Found By: --- Services Priority:
Business Priority: Blocker: ---
Marketing QA Status: --- IT Deployment: ---

Description Felix Foerster 2008-04-24 12:06:09 UTC 
The machine was running gtod_infinite as the last test before it hangs. 
The test was called by the ctcs2 framework with the current slert-ltp-realtime.tcf from Daniel. 
The test took longer than 180 sec so it was terminated by ctcs2. 
gtod_infinite and gtod_infinite.cyclictest both exited cleanly, according to the log files. 
But shortly after that, the machine hang.
Ctcs2 couldn't remove it's interrupt-flag in the log directory and the logfiles for the next testcase where not created. That indicates that the machine hangs before executing the next testcase (gtod_latency).
The sysrqs didn't show up in the log, so i don't know whether they where working then or not ( no serial console ).

model name : Intel(R) Pentium(R) D CPU 2.80GHz
CPUS:  2
Kernel:  2.6.22.19-0.7-rt #1 SMP PREEMPT RT 2008-04-19 06:21:37 +0200 x86_64
Comment 1 Felix Foerster 2008-04-24 14:08:44 UTC 
On the second run of the slert-ltp-realtime.tcf the gtod_infinite testcase keeps running and ctcs2 doesn't kill it. But it is possible to kill it by hand.
Comment 2 Felix Foerster 2008-04-24 14:31:58 UTC 
Update for the second run. Now the machine hangs again ! But this time it took about 20 minutes before that happened.
Comment 3 Felix Foerster 2008-04-25 10:56:53 UTC 
On the third run ctcs2 again wasen't able to terminate gtod_infinite on the testmachine. This problem maybe related to Bug 380751 - "pthread_cond_latency hangs while several runs", but thats about pthread_cond_latency and not gtod_infinite. The fact that the test can be killed by hand but not by ctcs2 seems  to be related to Bug 376781.
But the problem with the hang is not related to any mentioned bug.
I will update to beta7 and try to reproduce the bug.
Comment 4 Felix Foerster 2008-04-28 15:34:46 UTC 
Same problem on KOTD 2.6.22.19-SLERT10_BRANCH_20080426032956-rt, machine still hangs after running gtod_infinite.
Comment 5 Ihno Krumreich 2008-04-29 00:01:41 UTC 
Does sysrequest work?
Comment 6 Felix Foerster 2008-04-29 11:17:46 UTC 
No sysrq's work. But i have no serial console on that machine.
Comment 7 Daniel Gollub 2008-05-08 18:03:36 UTC 
Felix, this is with k45.suse.de - right?
Comment 8 Felix Foerster 2008-05-09 10:05:50 UTC 
Yes it is K45.
Comment 9 Ihno Krumreich 2008-08-18 07:29:22 UTC 
Is this bug still valid?
Comment 10 Felix Foerster 2008-08-18 09:47:29 UTC 
The problem with gtod is solved, also the problem with the longing of ctcs2. The bug can be closed !
Comment 11 Daniel Gollub 2008-08-18 09:52:02 UTC 
So this is a dup. of bug 389143  - right?

*** This bug has been marked as a duplicate of bug 389143 ***