This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [Patch] Splitting memory references during unrolling (resubmission)


> 
> could you please measure the effect of this patch on some benchmark
> on i686 and post results to the mailing list?
> 
> I suspect this optimization might spoil the code there; if this
> turned out to be the case, you would have to restrict it to work
> only on architectures where it is profitable.
> 
> Zdenek

I could not find any significant performance differences while running 
the SPEC benchmarks on the patch and mainline with -O3 and funroll-loops
flags (on i686).

I can add a new flag to enable this optimization if you still have 
concerns 
about it's effect.

Revital


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]