This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: -fprofile-generate and -fprofile-use
- From: Zdenek Dvorak <rakdver at atrey dot karlin dot mff dot cuni dot cz>
- To: Peter Steinmetz <steinmtz at us dot ibm dot com>
- Cc: Steven Bosscher <stevenb at suse dot de>, gcc at gcc dot gnu dot org,girish_vaithees at yahoo dot com, hubicka at ucw dot cz, janis187 at us dot ibm dot com
- Date: Wed, 31 Aug 2005 11:03:40 +0200
- Subject: Re: -fprofile-generate and -fprofile-use
- References: <OFF66140F3.529AEE7E-ON8625706D.005867D0-8625706D.0058F1E9@us.ibm.com>
Hello,
> >A more likely source of performance degradation is that loop unrolling
> >is enabled when profiling, and loop unrolling is almost always a bad
> >pessimization on 32 bits x86 targets.
>
> To clarify, I was compiling with -funroll-loops and -fpeel-loops
> enabled in both cases.
>
> The FDO slowdown in my case was caused by the presence of some loop
> invariant code that was getting removed from the loop by the loop
> optimizer pass in the non-FDO case.
you may try adding -fmove-loop-invariants flag, which enables new
invariant motion pass.
Zdenek