Re: [PATCH] gdb: make "start" breakpoint inferior-specific

[Simon Marchi via Gdb-patches <gdb-patches@sourceware.org>]

On 11/9/22 08:11, Andrew Burgess wrote:
> Simon Marchi <simark@simark.ca> writes:
> 
>>>
>>> In general I'm happy enough with the actual GDB change you propose here,
>>> but like Bruno, I also wondered if it would be better to add inferior
>>> specific breakpoints as an actual thing, rather than relying on the
>>> condition logic like you do here?
>>>
>>> I wasn't sure how you'd feel about this idea, so I didn't spend too
>>> long, but below is a VERY rough proof of concept patch, that appies on
>>> top of yours, that adds inferior specific breakpoints.  You temporary
>>> breakpoint now becomes:
>>>
>>>   tbreak -qualified main inferior 1
>>>
>>> But this functionality would also be available to a user if they wanted
>>> it, which might be nice.
>>
>> I think it's a good idea, but I would prefer if that wasn't a
>> prerequesite for this patch, since adding this feature can become
>> somewhat of a rabbit hole.  Functionally, a condition will essentially
>> behave the same, I think it's robust, and I would like to get rid of the
>> flakiness in the testsuite sooner than later :).
> 
> That's fine - I assume you would be happy if you code was updated to
> make use of inferior specific breakpoints if such a feature appeared in
> the future?

(re-adding gdb-patches, not sure where it got removed, maybe I messed up
my initial reply)

Absolutely, I would be fine.  What matters is that the user-visible
behavior works.


Simon