Re: GCC 4.6 performance regressions

2011-02-11 Thread Quentin Neill
On Thu, Feb 10, 2011 at 3:13 AM, Jonathan Wakely wrote: > On 10 February 2011 05:18, Quentin Neill wrote: >> On Wed, Feb 9, 2011 at 2:42 AM, Jonathan Wakely >> wrote: >>> On 9 February 2011 08:34, Sebastian Pop wrote: For example x264 defines CFLAGS="-O4 -ffast-math $CFLAGS", and so >>

Re: GCC 4.6 performance regressions

2011-02-10 Thread Jonathan Wakely
On 10 February 2011 05:18, Quentin Neill wrote: > On Wed, Feb 9, 2011 at 2:42 AM, Jonathan Wakely wrote: >> On 9 February 2011 08:34, Sebastian Pop wrote: >>> >>> For example x264 defines CFLAGS="-O4 -ffast-math $CFLAGS", and so >>> building this benchmark with CFLAGS="-O2" would have no effect. >

Re: GCC 4.6 performance regressions

2011-02-09 Thread Quentin Neill
On Wed, Feb 9, 2011 at 2:42 AM, Jonathan Wakely wrote: > On 9 February 2011 08:34, Sebastian Pop wrote: >> >> For example x264 defines CFLAGS="-O4 -ffast-math $CFLAGS", and so >> building this benchmark with CFLAGS="-O2" would have no effect. > > Why not? > > Ignoring the fact -O3 is the highest l

Re: GCC 4.6 performance regressions

2011-02-09 Thread Michael Larabel
Hi Everyone, I've been following the thread already from the start (I'm on the list). Matthew Tippett (CC'ed now as well) and I have already been working on some ways to help further and one of us should have some more information to be presented shortly. If any of you have any other question

Re: GCC 4.6 performance regressions

2011-02-09 Thread Richard Guenther
On Wed, Feb 9, 2011 at 7:20 AM, Tony Poppleton wrote: >> While I appreciate Phoronix as a booster site, their benchmarking >> practice often seems very dodgy; I'd take the results with a large grain >> of salt > > The main reason I posted the link in the first place was because it > was reflec

Re: GCC 4.6 performance regressions

2011-02-09 Thread Jakub Jelinek
On Wed, Feb 09, 2011 at 08:42:05AM +, Jonathan Wakely wrote: > On 9 February 2011 08:34, Sebastian Pop wrote: > > > > For example x264 defines CFLAGS="-O4 -ffast-math $CFLAGS", and so > > building this benchmark with CFLAGS="-O2" would have no effect. > > Why not? > > Ignoring the fact -O3 is

Re: GCC 4.6 performance regressions

2011-02-09 Thread Jonathan Wakely
On 9 February 2011 08:34, Sebastian Pop wrote: > > For example x264 defines CFLAGS="-O4 -ffast-math $CFLAGS", and so > building this benchmark with CFLAGS="-O2" would have no effect. Why not? Ignoring the fact -O3 is the highest level for GCC, the manual says: "If you use multiple -O options, wit

Re: GCC 4.6 performance regressions

2011-02-09 Thread Sebastian Pop
On Wed, Feb 9, 2011 at 00:20, Tony Poppleton wrote: >> While I appreciate Phoronix as a booster site, their benchmarking >> practice often seems very dodgy; I'd take the results with a large grain >> of salt > > The main reason I posted the link in the first place was because it > was reflecti

Re: GCC 4.6 performance regressions

2011-02-09 Thread Jonathan Wakely
On 9 February 2011 06:20, Tony Poppleton wrote: > > Out of interest, has their been much communication in the past between > GCC and Phoronix to address any of these issues in their previous > benchmarks? I signed up to their forum to point out that snapshots have additional checking turned on by

Re: GCC 4.6 performance regressions

2011-02-08 Thread Tony Poppleton
> While I appreciate Phoronix as a booster site, their benchmarking > practice often seems very dodgy; I'd take the results with a large grain > of salt The main reason I posted the link in the first place was because it was reflecting my own emperical evidence for the application I am working

Re: GCC 4.6 performance regressions

2011-02-08 Thread Miles Bader
Jonathan Wakely writes: >> Because phoronix uses make -j the compile times are highly random. > > Don't they know how to use 'time' to measure something more useful? > I wouldn't be entirely surprised, last time I looked they didn't seem > to know to use --enable-checking=release when comparing co

Re: GCC 4.6 performance regressions

2011-02-08 Thread Jonathan Wakely
On 8 February 2011 22:49, Sebastian Pop wrote: > > Because phoronix uses make -j the compile times are highly random. Don't they know how to use 'time' to measure something more useful? I wouldn't be entirely surprised, last time I looked they didn't seem to know to use --enable-checking=release w

Re: GCC 4.6 performance regressions

2011-02-08 Thread Sebastian Pop
On Tue, Feb 8, 2011 at 16:14, Xinliang David Li wrote: > What are the base option set used in all the comparison? O2, O3?  Some The flags are those set by the Makefiles of the different benchmarks (as downloaded from the web). Setting different flags with CFLAGS=... is painful. > of the build t

Re: GCC 4.6 performance regressions

2011-02-08 Thread Xinliang David Li
What are the base option set used in all the comparison? O2, O3? Some of the build time results look weired -- e.g., adding -march speeds up *compile time* by 35%. David On Tue, Feb 8, 2011 at 8:08 AM, Tony Poppleton wrote: > Hi, > > The following article has a fairly comprehensive set of bench

Re: GCC 4.6 performance regressions

2011-02-08 Thread Jeff Law
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 02/08/11 09:08, Tony Poppleton wrote: > Hi, > > The following article has a fairly comprehensive set of benchmarks run > against all the current stable releases of GCC as well as 4.6.0. >http://www.phoronix.com/scan.php?page=article&item=intel_