[RFA 0/9] Radically simplify the complaint system

Tom Tromey tom@tromey.com
Mon May 28 22:19:00 GMT 2018


>>>>> "Pedro" == Pedro Alves <palves@redhat.com> writes:

[ QUIT ]
Pedro> And then when that happens, 'complaints.c'::series is not
Pedro> restored back [because the clear_complaints call is skipped,
Pedro> because it is not exception safe], so all the following
Pedro> complaints are issued with SHORT_FIRST_MESSAGE:

Hah, oops.  I have a series now to remove SHORT_FIRST_MESSAGE.  It still
won't be perfect here -- there's no cleanup object still and so you
could in theory see too many complaints if you C-c at the wrong time.
I don't think that is super important though.

Pedro> I guess I just see complaint output similar to "set debug foo" output,
Pedro> where I also wish we were more consistent in printing some leading
Pedro> prefix for each debug knob, like, "set debug infrun 1" producing
Pedro> output messages prefixed with "infrun:" or something like that.

This makes sense to me too.  I was tempted to just remove it entirely,
but it is nice to be able to throttle the messages, since ordinarily
there are many of the same kind in a given objfile.

Pedro> Not sure what you mean here.  I was just thinking of something like
Pedro> simply making sure to include "DWARF reader:" in all complaint calls in
Pedro> dwarf2read.c.

Ok, I see.  I thought you meant something where symfile would tell the
complaints system about the module in question.

Tom



More information about the Gdb-patches mailing list