Red Hat Bugzilla – Bug 154972
unexplained SIGSEGV death in SIGSEGV signal handler
Last modified: 2007-11-30 17:07:17 EST
Split off from Bug #154221
With bug #154221 fixed, we can get a core dump for this case.
Why it happens remains unclear.
Created attachment 113213 [details]
reproducer, compile on 2.1AS and run on RHEL4
The problem here turns out to be an interaction between bug #154221 and the
exec-shield implementation on x86. Fixing that bug in a way that interacts well
with the exec-shield code alleviates the spurious faults.
See bug #154221 comment #23 for a patch that addresses this problem.
This patch is still under review by the RHEL4 kernel team and not verified to be
without other consequences.
The chances of hitting this problem should be drastically reduced by booting
with the exec-shield=0 parameter. Using the test case here, I have not been
able to reproduce the problem when exec-shield=0 is set at boot time. However,
I cannot be sure that the problem could never arise in that configuration.
An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on the solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.