Mailing List Archive: exim: dev

testsuite

Index | Next | Previous | View Threaded

eximusers at downhill

Nov 14, 2009, 7:45 AM

Post #1 of 4 (714 views)
Permalink

testsuite

Hello,
I have given the test-suite a testrun.

Stuff it breaks on:

* umask. I am running with umask 0002 and therefore aux-fixed/* is g+w
in my CVS checkout. Also the files generated at runtime in aux-var/
are generated g+w.

Solution:
find aux-fixed -exec chmod g-w {} +
umask 0022

* locales. The very first thing runtest does is parsing ifconfig
output. Mine was localized.
Solution:
env LC_ALL=C ...

* charsets.
I usually build exim with HEADERS_CHARSET="UTF-8". Which seems to
break e.g. Basic/0178.
Solution:
Do not set HEADERS_CHARSET="UTF-8".

cu andreas
--
`What a good friend you are to him, Dr. Maturin. His other friends are
so grateful to you.'
`I sew his ears on from time to time, sure'

--
## List details at http://lists.exim.org/mailman/listinfo/exim-dev Exim details at http://www.exim.org/ ##

johannes at sipsolutions

Nov 14, 2009, 7:52 AM

Post #2 of 4 (680 views)
Permalink

Re: testsuite [In reply to]

On Sat, 2009-11-14 at 16:45 +0100, Andreas Metzler wrote:
> Hello,
> I have given the test-suite a testrun.
>
> Stuff it breaks on:

...

And, apart from the segfault that was fixed, that's it? That seems
good :)

johannes

Attachments:

signature.asc (0.78 KB)

eximusers at downhill

Nov 14, 2009, 8:01 AM

Post #3 of 4 (667 views)
Permalink

Re: testsuite [In reply to]

On 2009-11-14 Andreas Metzler <eximusers [at] downhill> wrote:

> Hello,
> I have given the test-suite a testrun.
[...]

Hello,

updating the testsuite for 4.70 is a PITA. ;-)

* Exim now inserts (re-sorts or has stopped re-sorting - I do not know
what) Date and Message-ID headers at different places. This breaks
*many* of the tests. I think this was done intentionally andis
related to DKIM, correct? I have pressed (u)pdate for these cases
and should be able to get a patch from there.

* The debug log has gotten a little bit more verbose, logging GnuTLS
compile-time and current versions. Patch attached.

* PDKIM also adds to the debug log, adding a newline followed by
PDKIM <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
The newline is cumbersome. Are there strong reasons against:

---------------------------------------------------
--- ./src/pdkim/pdkim.c.orig 2009-11-14 14:21:38.000000000 +0100
+++ ./src/pdkim/pdkim.c 2009-11-14 15:03:29.000000000 +0100
@@ -1283,7 +1283,7 @@ DLLEXPORT int pdkim_feed_finish(pdkim_ct
#ifdef PDKIM_DEBUG
if (ctx->debug_stream)
fprintf(ctx->debug_stream,
- "\nPDKIM <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n");
+ "PDKIM <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<\n");
#endif
}
---------------------------------------------------
The addition of the nonempty line is easily handled by attached patch.

cu and- more stuff to follow - reas
--
`What a good friend you are to him, Dr. Maturin. His other friends are
so grateful to you.'
`I sew his ears on from time to time, sure'

Attachments:

drop-gnutls-version-info.diff (0.47 KB)

drop-pdkim-debugging.diff (0.44 KB)

eximusers at downhill

Nov 15, 2009, 1:34 AM

Post #4 of 4 (669 views)
Permalink

Re: testsuite [In reply to]

On 2009-11-14 Johannes Berg <johannes [at] sipsolutions> wrote:
> On Sat, 2009-11-14 at 16:45 +0100, Andreas Metzler wrote:
> > Hello,
> > I have given the test-suite a testrun.
> >
> > Stuff it breaks on:

> ...

> And, apart from the segfault that was fixed, that's it? That seems
> good :)

Hello,

I have not run IPv6, *SQL, cyrus-sasl, dovecot-server and LDAP tests.
I have not got IPv6 connectoion here and I do not have the respective
servers fed with the required data running.

However afaict the segfault is really the only new issue in 4.70 I
have stumbled upon. There are some testcases that fail (0533) but
these are not new issues, they also fail for me with 4.69 or 4.68.

cu andreas

--
`What a good friend you are to him, Dr. Maturin. His other friends are
so grateful to you.'
`I sew his ears on from time to time, sure'

--
## List details at http://lists.exim.org/mailman/listinfo/exim-dev Exim details at http://www.exim.org/ ##

Index | Next | Previous | View Threaded

Interested in having your list archived? Contact Gossamer Threads