Since 2006-10-28, someone has broken `Face-frob-property'

Nix nix at esperi.org.uk
Tue Dec 19 18:50:28 EST 2006


On 19 Dec 2006, Aidan Kehoe spake thusly:
>  Ar an naoú lá déag de mí na Nollaig, scríobh Nix: 
>  > (if so, damn, I'll have to actually do some *work* to diagnose this
>  > rather than leaping directly to the solution by magic. :) )
>
> The comp.unix.programmer FAQ approach to generating a stack trace and
> wrapping malloc() with something that calls that would probably give an
-------------- next part --------------
> exact idea of what?s calling what, irrespective of the malloc
> implementation. Dog-slow, though. 

Yep. And that's what dmalloc does. :)

> Tangentally, has anyone got experience running XEmacs under valgrind?

I've tried and failed a couple of times. Mind you I didn't try terribly
hard because the only time I needed it, inspection by eye found the bug
five minutes later.

> One would need to do the temacs -batch -l ../lisp/loadup.el run-temacs
> thing, I imagine, but it might help.

valgrind leak detection might help as well, but I'm not sure that XEmacs
tries too terribly hard to free() all its memory if it's going to die
anyway (even an Fgarbage_collect() or whatever at the very end might
legitimately not free everything; a good few roots are static variables
which won't get freed.)

The one advantage of valgrind leak detection is that it's not going
to get broken by people doing strange things with malloc() :)

-- 
`He accused the FSF of being "something of a hypocrit", which
 shows that he neither understands hypocrisy nor can spell.'
   --- jimmybgood


More information about the XEmacs-Beta mailing list