[PATCH 1/2] new memory-changed MI notification.

Tom Tromey tromey@redhat.com
Mon Oct 15 19:07:00 GMT 2012


>>>>> "Doug" == Douglas Evans <dje@google.com> writes:

Doug> For my own education, is this suppression just an optimization, or is
Doug> there a correctness issue here?
Doug> I can imagine that it's an optimization, why notify the frontend
Doug> something changed when it's the frontend that requested the change.
Doug> But there is *zero* documentation in mi-main.h on *why* struct
Doug> mi_suppress_notification exists, so it's hard to tell. :-(
Doug> [I realize your patch is just adding an entry, but I'd like to learn
Doug> what the reason for it is.]

I don't know about this case in particular, but in many cases in MI, a
command will have a result record that encodes the exact same data that
an async response would give.  On the other hand, if a CLI command is
invoked, then there is no relevant response record, and so the async
response must be generated.

I think it is an optimization.  Offhand I can't think of a correctness
issue here, but I guess I wouldn't be surprised if there is one lurking
for some command or another.

Tom



More information about the Gdb-patches mailing list