Re: [PATCH 0/5] Fixes for per-inferior settings and $_gdb_setting_str()

[Andrew Burgess <aburgess@redhat.com>]

Simon Marchi <simark@simark.ca> writes:

> On 4/17/23 14:09, Simon Marchi via Gdb-patches wrote:
>> On 4/4/23 08:45, Andrew Burgess via Gdb-patches wrote:
>>> While working on another patch I spotted that
>>> $_gdb_setting_str("args") wasn't working correctly.  Turns out args,
>>> cwd, and inferior-tty all have the same problems.  This series
>>> resolves these issues.
>> 
>> That rang a bell, it's because I have fixed some of these downstream in
>> ROCgdb, and haven't sent it upstream yet.  The downstream commit is
>> here:
>> 
>> https://github.com/ROCm-Developer-Tools/ROCgdb/commit/54d0a79614071fd62df5e2dfe793ff26cc7e31e6
>> 
>> I'll take a look at your patches, I presume that they will do mostly the
>> same thing.  And if my patch does more (according to the commit message,
>> it fixes "remote exec-file" and "tdesc filename"), I can send what
>> remains after.
>
> So, my patch adds some tests for gdb.parameter (Python), MI and the with
> command.  However, they are sprinkled in different files.  I like that
> you have a test file specifically to exercise the settings that are
> per-inferior.  Your test could probably be augmented to exercise all
> these ways to get a setting.
>
> You series looks good to me, up to you if you want to merge it as-is or
> improve it based on what I had.

I reworked the tests taking onboard some of the ideas from your tests,
and pushed this series.

Do you plan to upstream your patch anytime soon?  I had been planning to
look at other per-inferior settings once this series was merged, but I
don't want to tread on your toes any more than I already have...

Thanks,
Andrew