Steve Langasek <[EMAIL PROTECTED]> writes: > Hey, so I guess forgetting that you've created a zero-length > /etc/krb5.keytab for testing on a system that you're later trying to > convert to use pam_krb5 is a good way to motivate oneself to find and > fix this bug...
> It turns out that the problem is buggy handling of fread() in > src/lib/krb5/keytab/kt_file.c; the code assumes that fread only returns > 0 upon error, when it can, of course, also return 0 if there are 0 bytes > to be read. :) So it sets kerror to 0 (errno), closes the file but > doesn't completely clean up, and the FILE* gets double-freed. > Possible patch attached, reassigned and downgraded. Yup, thanks, this looks right to me. I'll apply this to the next release, unless Sam has any objections. -- Russ Allbery ([EMAIL PROTECTED]) <http://www.eyrie.org/~eagle/> -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
