Re: [PATCH 1/4] Change evpy_emit_event to accept a gdbpy_opt_borrowed_ref

[Simon Marchi <simark@simark.ca>]

On 9/14/26 12:59 PM, Tom Tromey wrote:
>>>>>> "Simon" == Simon Marchi <simark@simark.ca> writes:
> 
> Simon> Unrelated, but those two if could be collapsed into one with &&.
> 
> Yeah, I normally don't care about this kind of thing.
> 
> Simon> Also unrelated: I would change evregpy_no_listeners_p to
> Simon> evregpy_has_listeners_p, because double-negative is confusing.
> 
> Simon> I can make following patches to change those.
> 
> Alright.
> 
> Simon> Copying a review comment from Claude that seems important:
> [...]
> 
> Yeah.  I think there's a longstanding GCC bug about not warning about
> this.
> 
> Anyway I fixed it.
> 
> Simon> Otherwise, it also noted:
> 
> Simon>   - gdb/python/py-breakpoint.c:1309 and :1340: both calls now fit on one line.
> 
> I don't know which lines these are but the obvious-to-me choices end up
> at 81/82 characters and so can't be changed.

Sorry, I shouldn't have copy pasted this comment without checking it
first, my bad.

I did trust it blindly because I did see some spots in the series where
some lines could be unwrapped, so I assumed it was talking about this
spots.  For instance, the return statement in this function, although
it's in patch 3, not patch 1:

static gdbpy_ref<>
create_continue_event_object (ptid_t ptid)
{
  gdbpy_ref<> py_thr = py_get_event_thread (ptid);

  if (py_thr == nullptr)
    return nullptr;

  return create_thread_event_object (&continue_event_object_type,
				     py_thr);
}

Anyway, I'll be more careful in the future.

Simon