Re: GDB 13 Release 2022-09-11 update

[Tom de Vries <tdevries@suse.de>]

On 9/20/22 01:26, Tom Tromey wrote:
>>> My impression is that .gdb_index never really worked correctly for
>>> Ada.
> 
> Tom> Hmm, I committed 7ab96794115 ("[gdb/symtab] Enable ada .gdb_index").
> 
> Wow, I had completely forgotten about this.
> I see I even wrote a corresponding documentation patch.
> 
>>> So, while that is a regression of sorts, since the feature isn't really
>>> usable for Ada in general,
> 
> Tom> Um, can you elaborate a bit on why you think that?
> 
> Loss of memory.
> 
> Is there a bug open for this?

Yes, https://sourceware.org/bugzilla/show_bug.cgi?id=29179 .

> Anyway, I will look at it when I look
> into the other issues with the new DWARF reader.
> 

OK.  FWIW, I don't mind if we drop this feature if it's difficult to 
repair or maintain going forward, or too expensive in terms of index 
size, I'm just trying to make sure that we don't silently regress.

> I looked and I can't find my patch to change how .debug_names is
> constructed.  Hopefully that won't be too hard to rewrite.  However, I
> did find a patch that I had forgotten about that changes .debug_names so
> that it can write out a BFD, solving the string table problem (though I
> still didn't write the reader side of this).

Hm, interesting.  Maybe you could post that patch to some PR?

Thanks,
- Tom