Re: [Texmacs-dev] RE: Texmacs segfault

[David Allouche]

On Mon, Oct 14, 2002 at 01:20:09AM +0400, Igor V. Kovalenko wrote:
> 
> I do think there is some race in the code, may be more than one.
> I still have a crash that disappears when I run "texmacs -d myfile.tm"
> instead of plain "texmacs myfile.tm". May be a trace of some asynchronous
> initialization dependence bug. Traces show that some "eb->" dereference
> a NULL pointer, might be that some edit box not created yet. Sorry, it
> is hard to reproduce, but I can try on request...

That is weird... to the best of my knowledge, TeXmacs is entirely
devoid of any kind of asynchronous code. Unless I am mistaken, you can
rule out race conditions.

So, the other good candidate (except compiler bugs) is memory
corruption. There are good tools to track memory corruptions bugs...
Maybe you could try to use valgrind, for example.

I am going to run a texmacs compiled with g++-3.2 with symbol enabled
so I can do a bit of post mortem too (assuming I have the same problem
as you).

-- 
David Allouche         | GNU TeXmacs -- Writing is a pleasure
Free software engineer |    http://www.texmacs.org
   http://ddaa.net     |    http://alqua.com/tmresources
   address@hidden  |    address@hidden
TeXmacs is NOT a LaTeX front-end and is unrelated to emacs.