Re: Fwd: Fails to bring up a back trace

[Basile Starynkevitch]

On 11/30/23 05:58, tomas@tuxteam.de wrote:
> On Thu, Nov 30, 2023 at 10:43:45AM +1300, Davin Pearson wrote:
>
> [...]
>
> > Yet it fails to bring up an error backtrace.
> I fear this is too scant a context to feed educated guesses. But
> this, at least, looks wrong to me:
>
> > I also tried the following command:
> >
> > emacs --geometry 2048x1024 --eval "(setq debug-on-error)" --debug-init &
> ... it should be "(setq debug-on-error t)".


A possible approach to backtracking on Linux might be to improve or 
enhance GNU emacs to use Ian Taylor's libbacktrace open source library. 
The libbacktrace source code is on 
https://github.com/ianlancetaylor/libbacktrace

On Linux, mixing that library with the dladdr function (see 
https://man7.org/linux/man-pages/man3/dladdr.3.html) gives human 
readable backtraces.

And this is used both in recent GCC compilers (see https://gcc.gnu.org/ 
- the source code of libbacktrace is incorporated in GCC 13...) and in 
the RefPerSys open source inference engine project on 
https://github.com/RefPerSys/RefPerSys/

It seems that the latest GNU emacs snapshot commit e87644baa3239ce5 is 
using GNU libc backtracing functions, but not (like GCC does) the 
libbacktrace from Ian Taylor.

For details, contact me (Basile Starynkevitch, in France near Paris) by 
email to basile@starynkevitch.net

Regards

--
Basile Starynkevitch
 <basile@starynkevitch.net>
(only mine opinions / les opinions sont miennes uniquement)
92340 Bourg-la-Reine, France
web page: starynkevitch.net/Basile/