https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80239
Thomas Preud'homme <thopre01 at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|UNCONFIRMED |RESOLVED
Resolution|--- |INVALID
--- Comment #6 from Thomas Preud'homme <thopre01 at gcc dot gnu.org> ---
(In reply to Thomas Preud'homme from comment #5)
> (In reply to Thomas Preud'homme from comment #1)
> >
> > I believe the problem is that the equivalence stays with NO_REGS as register
> > class instead of being given that of a register (since it's going to be
> > reloaded into a register). My patch follows that path and as said above
> > works just after r235184 but the performance remains worse on recent trunk.
> >
> > When looking at code generated by recent trunk, the issue mentionned in the
> > original post is not there which suggest that something else is at play.
>
> I bisect this issue down to r239561 and reverting that commit on top of
> current trunk is enough to bring the performance back up. I'll still push
> the patch I mentioned above forward but given the circumstances I will wait
> for GCC 8 stage 1 for doing so.
>
> I'll post a testcase of the issue as soon as I have understood what is the
> source of the slowdown with r239561 in current trunk.
Alright, that was due to a change in alignment. Adding a nop before anything in
main's prologue brings back performance. Thus closing the bug.
