Markus Armbruster writes: > Lluís Vilanova <vilan...@ac.upc.edu> writes: >> Markus Armbruster writes: >> >>> Lluís Vilanova <vilan...@ac.upc.edu> writes: >>>> Also removes old "trace-event", "trace-file" and "info trace-events" HMP >>>> commands. >> >>> We gain the ability to control trace via QMP, but lose the ability to >>> control it via HMP, correct? >> >> Right. I can keep the HMP commands, but doing so requires exposing the >> internal >> trace event identifier number in the QMP interface. Also, "trace-file" >> cannot be >> implemented on top of the current QMP commands, and is specific to the >> "simple" >> tracing backend.
> Would rough feature-parity with QMP be feasible with new HMP commands on > top of the QMP interfaces, completely ignoring HMP backward > compatibility? Except for "trace-file", it would be quite easy, yes. For "trace-file", we can either keep the current HMP implementation, or add new QMP commands to support it (I'd rather devote my time to something else, though). Thanks, Lluis -- "And it's much the same thing with knowledge, for whenever you learn something new, the whole world becomes that much richer." -- The Princess of Pure Reason, as told by Norton Juster in The Phantom Tollbooth
