From: "Justin Mason" <[EMAIL PROTECTED]>
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
>
> Loren Wilton writes:
> > > from maillog
> > > Deep recursion on subroutine
"Mail::SpamAssassin::Message::Node::finish"
> > > at
/usr/local/lib/perl5/site_perl/5.8.5/Mail/SpamAssassin/Message/Node.pm
> > > line 659
> >
> > As far as I know (which may be wrong) the deep recursion thing isn't
related
> > to either bayes or awl expiry.  I seem to recall that the devs had a bug
on
> > this, but nobody could ever reproduce the problem to fix it.  Don't
recall
> > if the bug is still open or got dumped for being unreproducable.
> >
> > IF you can track this down to a particular message or set of
circumstances,
> > and IF it isn't either a bayes or awl expiry run, then it should be
reported
> > in BZ along with some appropriate documentation or test cases.
>
> this one's different from the usual cases.  The usual case is DB_File
> "deep recursion" errors, which typically indicates a broken perl
> distribution for that platform; not much we can do about that.
>
> This is deep recursion in *our* code, so we should be able to fix it,
> hence it's worth opening a bz about.

Um, another good bz might be to note that the standard debug message
should include the Message-ID field for the message that causes it.
That would give folks submitting bugs a chance, however small, of
feeding the actual bad message to you. (With procmail users if SA
were to add to the "deep recursion" debug message a header "SA-Blew"
it would be easy to clone the message is SA concludes its barfing
session. But really, we don't want to make it TOO easy to get good
debug cases, do we?)

(But then, seriously, I do have messages that go "twang", SA ends
with no error, no markup, and no anything else yet they test properly
via "spamc <bad" as well as "spamassassin -t" and bare spamassassin.
So the above idea is not a cure all. It would merely help a lot.)

{o_o}   <- knows full well how hard it is to get really good debug
        messages built. There never seems to be enough data. And when
        there might have been the users never report it.


Reply via email to