kadmind - SIGSEGV

Ken Raeburn raeburn at MIT.EDU
Wed Dec 1 17:54:28 EST 2004


On Dec 1, 2004, at 17:43, Michael Tautschnig wrote:
> Another segfault:
> Any ideas?

Problem is, a crash in mallopt and such usually means the free pool is 
corrupted, which could've happened any time, and usually isn't caused 
directly by the call into the malloc code when the crash actually 
happens.

Could you try running kadmind using Electric Fence or valgrind?  It has 
a "-nofork" option that should make that easier.

Ken



More information about the Kerberos mailing list