[PATCH]: Further updates to ada-* files
Mon Jun 28 14:47:00 GMT 2004
>>> Lets start with the obvious, can you fix the ARI problems (Run
>>> 'gdb_ari.sh -Wari' from http://sources.redhat.com/gdb/ari/)?
>>> Looking over the code, ada-tasks.c is going to be a problem. Can we
>>> ignore that for now (removing it?) and just focus on getting basic
>>> language support enabled? Once that's done, we can start working
>>> through the underlying issues that prompted ACT to add ada-tasks.c.
> Right. I disabled it (i.e., it's not compiled) in the latest version
> of the patch I sent, and do not expect it to be reviewed. We've
> decided that now is as good a time as any to do a much-needed major
> revision of ada-tasks.c; I'm not sure right now if I want to remove
> the source entirely from the repository. Depends on how major a
> change I need.
Lets remove it. The question is how big a thread model change GDB is
going to get (and regardless of Ada's task model) and how much having
that dead code would confuse things.
> The "underlying issues" are that Ada has tasks and they are (in general)
> built on top of, but in principle distinct from threads. As for why to
> put it in a separate file---well it's messy, due to the fact that
> tasks map onto threads differently on every single confounded platform,
> it seems.
It's messy because the underlying thread model is bug---d.
>>> Did that previous ChangeLog problem get fixed?
> Not yet, but I haven't forgotten it. It must go in, but I fear it will be
> of no help for the current task.
More information about the Gdb-patches