This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug gcov-profile/45890] Coverage may be weakened by r164986
- From: "rguenther at suse dot de" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: Tue, 04 Oct 2011 10:18:18 +0000
- Subject: [Bug gcov-profile/45890] Coverage may be weakened by r164986
- Auto-submitted: auto-generated
- References: <bug-45890-4@http.gcc.gnu.org/bugzilla/>
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45890
--- Comment #7 from rguenther at suse dot de <rguenther at suse dot de> 2011-10-04 10:18:18 UTC ---
On Tue, 4 Oct 2011, hubicka at ucw dot cz wrote:
> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45890
>
> --- Comment #6 from Jan Hubicka <hubicka at ucw dot cz> 2011-10-04 10:04:37 UTC ---
> > Where do we do it right now? At profile instrumentation stage, right?
> > I'd say if we don't need SSA form we should do coverage instrumentation
> > right after CFG construction, no? Before into-SSA we already have
> > stuff like OMP lowering. But yes, just after into-SSA would be
> > an improvement.
>
> We could also re-instantiate the former profiling pass before SSA for coverage,
> but it seems to me that there is no real reason to destroy line info before
> we get into-SSA and things will be a bit smoother if we do instrumentation
> at SSA form only. But I don't really have any strong opinions here.
Yeah, I suppose you are right.