On Feb 13, 2008, at 20:57, Steven Miller wrote: > Very weird, when running kadmin.local under valgrind, > it does NOT segfault. I am including the valgrind > output.
Here, valgrind is complaining about some uninitialized fields, and in the gdb stack trace, the program was copying data based on one of those same fields having a nonzero value. If the uninitialized length field was nonzero, and the data pointer was null or garbage, the copy would fail. So, it looks like valgrind did dig up a little more information on this bug. Thanks, I'll see if I can track down what's going on. Ken ________________________________________________ Kerberos mailing list Kerberos@mit.edu https://mailman.mit.edu/mailman/listinfo/kerberos
