Python Coredump on FreeBSD

[Harald Schneider]

Pattern and strings can grow very large. That means pattern something like

"Unknown local
part|(user|mailbox|account|recipient|Addressee).*?(unknown|invalid|not
found|not here)|(no such|doesn't
have|invalid|unknown).*?(account|user|mailbox|recipient|addressee)"

for the pattern and 5 - 500 KB for the strings being searched.

Indeed there is an issue with the stacksize on FreeBSD wich also forces Zope
to crash. I found a patch fpr Python 2.1.3, which
should fix this issue:
http://zwiki.org/PatchPython213

At the moment, I rebuilt Python from source with that patch applied. If this
doesn't solve my problem, I'll try to mimmize stack usage by unsing smaller
regex patterns and search strings ... oh boy ... :-)

I'll keep you posted - thanks!

--Harald

"Master of the Unixverse" <ah at abraxas.devsoft.com> schrieb im Newsbeitrag
news:slrnbc009f.aa.ah at abraxas.devsoft.com...
> In article <ox2cnT5extyQQCKjXTWcpg at speakeasy.net>, A.M. Kuchling wrote:
> > On Mon, 12 May 2003 16:20:08 +0200,
> > Harald Schneider <h_schneider at marketmix.com> wrote:
> >> When I call
> >> re.compile(.......).search(.....)
> >> in sequence a few times, then the regular expression module re causes
this
> >> coredump. It's implementation must differ from other platforms. It
looks
> >
> > What is the pattern being compiled, and how large is the string being
> > searched?  There are patterns that consume a lot of C stack space while
> > searching.  There's a fixed recursion limit that's intended to prevent
core
> > dumps, and it's possible the recursion limit is set incorrectly for
FreeBSD.
>
> The application is multithreaded and there is indeed a "problem" with
> threads on FreeBSD vs. Linux here:  the default stacksize that a thread
> gets is different bewteen Linux and FreeBSD.  Linux threads have more
> default stacksize.  No idea about Windows.  So this may very well be
> the reason.
>
> Anton Hartl, ah AT muc DOT de