This is the mail archive of the
mailing list for the GCC project.
Re: PATCH: Darwin / Mac OS X native support, revised
- To: Richard Henderson <rth at redhat dot com>
- Subject: Re: PATCH: Darwin / Mac OS X native support, revised
- From: Stan Shebs <shebs at apple dot com>
- Date: Fri, 06 Apr 2001 17:44:58 -0700
- CC: gcc-patches at gcc dot gnu dot org
- References: <3ACE46FB.FEE99A33@apple.com> <20010406165045.A5026@redhat.com>
Richard Henderson wrote:
> On Fri, Apr 06, 2001 at 03:45:14PM -0700, Stan Shebs wrote:
> > I believe that I've now accounted for all the feedback...
> > + #undef FUNCTION_PROFILER
> > + #define FUNCTION_PROFILER(FILE, LABELNO) \
> > + darwin_output_function_profiler (FILE, LABELNO);
> I did mention how you should look at PROFILE_HOOK, didn't I?
> You probably want something practically identical to the
> aix definition. You can probably even remove the ABI check
> in output_profile_hook and just call the same function.
I did look at it, but for reasons unknown to me, the calling
conventions for Darwin mcount are rather different from the
system's normal calling conventions, and PROFILE_HOOK isn't
general enough to handle this; it assumes that mcount will be
called using standard library conventions for the target system.
It may be possible to change the implementation of mcount in a
future rev of Darwin, but not today.