Geoff Hutchison writes:
> Well, I'm still not to the bottom of things. It seems like there may
> be several bugs all intertwined... I attach the VERBOSE=-vvv output
> from make check. It may prove a little more informative (or not). The
> failed wordlist test makes me wonder if records are either going in
> incorrectly or if some are just not going in...
>
> Here's the coredump for wordskip. It looks like it's working on an
> null or otherwise battered list. If I try to cast *this in gdb or
> access any of the methods, it won't work.
This is driving me nuts :-} My best guess is that it's a Berkeley DB
configuration problem. I've checked the berkeley db code and the fact that
mutexes are not found on your architecture should not be a problem (used
fcntl instead).
From all the messages of the tests it boils down to a single problem :
it cannot write to a Berkeley DB file. I'm ready to spend an hour on your
machine to try to figure this out, any time of your convinience (I'm currently
in Boston).
Alternatively you could turn on all the debug and tracing capabilities of
Berkeley DB (--enable-debug + --enable-diagnostic).
Cheers,
--
Loic Dachary http://www.dachary.org/ [EMAIL PROTECTED]
24 av Secretan http://www.senga.org/ [EMAIL PROTECTED]
75019 Paris Tel: 33 1 42 45 09 16 [EMAIL PROTECTED]
------------------------------------
To unsubscribe from the htdig3-dev mailing list, send a message to
[EMAIL PROTECTED]
You will receive a message to confirm this.
