>>>>> "SC" == Stephen Campbell <steve@avalon.dartmouth.edu> writes:
SC> You are correct that it means EX_USAGE, but there is nothing in the
SC> debug log about this incident. Why did you suspect EX_USAGE? I can
SC> find nothing invalid in that command.
I suspect EX_USAGE because I've seen it before, and I have some of
sysexits.h memorized. I have never been able to determine why sendmail
likes to exit with this status, although combing the sources might help.
Combing Sendmail 8.7.3, which I happen to have the source out for, I see
the following three places where it would exit with EX_USAGE without
printing anything to stderr:
When it's processing arguments and falls through the case statement.
After it does delivery, when errors were encountered during delivery:
/*
** Scan argv and deliver the message to everyone.
*/
sendtoargv(av, CurEnv);
/* if we have had errors sofar, arrange a meaningful exit stat */
if (Errors > 0 && ExitStat == EX_OK)
ExitStat = EX_USAGE;
If something blows in a rewrite rule.
Do any of these look possible? If your problem is repeatable, you might
consider turning up the sendmail debuging level and see if anything comes
out of it.
- J<
References:
|
|
