Re: CVS commit: src

[Robert Elz <kre%munnari.OZ.AU@localhost>]

    Date:        Tue, 26 Oct 2021 15:37:56 +0200
    From:        Joerg Sonnenberger <joerg%bec.de@localhost>
    Message-ID:  <YXgEtDQsRGtjxJ61%bec.de@localhost>

  | Personally, I would prefer to just kill -pg support completely, but
  | that's a separate discussion.

Yes, it is.

  | I don't think per-file profiling is that useful

Nor do I in general.   However, doing routine call counting on mcount
(or anything it calls) would simply be absurd.   If you want to
know how mant times mcount has been called, you simply sum all the other
counters.   (pc sampling, however it is done, detecting mcount is fine).

I agree that forcing inlining is a good idea, and should be done, but
I also feel that disabling profiling of mcount is also the right thing
to do.   I don't think we need a general ability to disable profiling on
any random file/function but if that is the easiest and cleanest way to
be able to disable profiling of mcount then I can accept it - provided
that's all it is actually used for,

kre