-f option with kinit

Xu, Qiang (FXSGSC) Qiang.Xu at fujixerox.com
Thu Oct 2 23:17:17 EDT 2008


> -----Original Message-----
> From: Ken Raeburn [mailto:raeburn at MIT.EDU]
> Sent: Friday, October 03, 2008 10:40 AM
> To: Xu, Qiang (FXSGSC)
> Cc: krbdev at mit.edu
> Subject: Re: -f option with kinit
>
> There shouldn't be; it just sets a flag in the request.

Agree, I am also confused by this behavior.

> What's the -M option?

This -M option is a Xerox modification for the password flag, indicating the username needs to be converted to UTF-8 encoding (but no conversion of password, as per what you have instructed me last year). By default, the option should be -X.

> A stack trace from the core dump may help indicate what's wrong.

Just as you said.

Previously I didn't find where the core file lies. Now I have managed to locate it. And GDB shows the stack trace is:
=========================================================================
Core was generated by `./kinit -f 120117097110 at SESSWIN2003.COM -M 070097105114049050051'.
Program terminated with signal 11, Segmentation fault.
#0  0x42d04829 in strcat () from /lib/libc.so.6
(gdb) bt
#0  0x42d04829 in strcat () from /lib/libc.so.6
#1  0x0804ad47 in parse_options (argc=5, argv=0xbf8301a0, opts=0xbf8301a0,
    progname=0xbf831457 "kinit") at kinit.c:659
#2  0x0804b5f6 in main (argc=5, argv=0xbf830284) at kinit.c:1458
(gdb)
=========================================================================
Fortunately, this falls in the part of Xerox modification (#ifdef XEROX_MODIFICATIONS_KERBEROS). So I think I can fix the defect of the modifcation later.

Thank you,
Xu Qiang




More information about the krbdev mailing list