Re: [PATCH][gdb/symtab] Redo "Fix assert in set_length"

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

On 12/5/23 21:59, Tom Tromey wrote:
>>>>>> "Tom" == Tom de Vries via Gdb-patches <gdb-patches@sourceware.org> writes:
> 
> Tom> Hi,
> Tom> This reverts commit 1c04f72368c ("[gdb/symtab] Fix assert in set_length"), due
> Tom> to a regression reported in PR29572, and implements a different fix for PR29453.
> 
> Tom> The fix is to not use the CU table in a .debug_names section to construct
> Tom> all_comp_units, but instead use create_all_comp_units, and then verify the CU
> Tom> table from .debug_names.  This also fixes PR25969, so remove the KFAIL.
> 
> Tom> Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=29572
> Tom> Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=25969
> 
> Tom> Any comments?
> 
> I've come around on this patch and I think you should check it in.
> 
> My main concern was the time cost, but you showed that this isn't
> important; and anyway gdb is probably going to map these sections and do
> more stuff with them later anyhow.
> 
> So, I think you should check this in.
> 

The patch no longer applied cleanly, so I've updated it.

Also I came across a finalize_all_units call in dwarf2_read_debug_names 
that was no longer necessary, so I've removed that one.

Reposted here ( 
https://sourceware.org/pipermail/gdb-patches/2023-December/204835.html ) 
and committed.

Thanks,
- Tom

> Approved-By: Tom Tromey <tom@tromey.com>
> 
> Tom