This is the mail archive of the gcc@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: GCC trunk SPEC2000 performance


> On Wed, Jun 19, 2002 at 10:43:40PM -0400, David Edelsohn wrote:
> > 2) "peak" (-O3 -funroll-all-loops -fprefetch=loop-arrays) produces slower
> > code than base (-O2) for many tests.
> 
> Peak cannot be considered "peak" with -funroll-all-loops
> as opposed to -funroll-loops.  That forces the compiler 
> to unroll a loop against its better judgement.

Yes, it is not meant to give best perofmrance, as the code is bloated
seriously.  I like it to point out the code issue problems anyway.
It is nice to see that GCC generated code is getting smaller gradulary,
that was not the case with older releases.

Honza
> 
> It's good for correctness testing, however.
> 
> 
> r~


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