Re: [Cegcc-devel] Performance considerations

2007-09-23 Thread Pedro Alves
Danny Backx wrote: > I don't have the DLL mentioned on that MSDN page on my PDA. (That makes > sense, the page says it only applies to Windows Embedded CE which I > don't think I have.) > > The functions are all in our coredll.def, but without the ordinal. Would > that be the problem ? No, as th

Re: [Cegcc-devel] Performance considerations

2007-09-23 Thread Danny Backx
On Thu, 2007-09-20 at 00:33 +0100, Pedro Alves wrote: > I doubt it. There shouldn't be any difference in calling system functions. > It's just a bl anyway. > > Your example is about floating point arithmetic. Is Xwords doing it a lot? > > Pgeorges, is you chess engine doing a lot of fp math? >

Re: [Cegcc-devel] Performance considerations

2007-09-20 Thread pgeorges
Pedro Alves a écrit : > Danny Backx escreveu: >> On Wed, 2007-09-19 at 17:51 +0200, pgeorges wrote: Crosswords, compiled with cegcc, runs noticably slower than compiled with the m$ compiler. Eric House also reported this. I've not investigated this yet. We may have inefficiencies oth

Re: [Cegcc-devel] Performance considerations

2007-09-19 Thread Pedro Alves
Danny Backx escreveu: > On Wed, 2007-09-19 at 17:51 +0200, pgeorges wrote: >>> Crosswords, compiled with cegcc, runs noticably slower than compiled >>> with the m$ compiler. Eric House also reported this. I've not >>> investigated this yet. We may have inefficiencies other than the code >>> generat

Re: [Cegcc-devel] Performance considerations

2007-09-19 Thread Danny Backx
On Wed, 2007-09-19 at 17:51 +0200, pgeorges wrote: > > Crosswords, compiled with cegcc, runs noticably slower than compiled > > with the m$ compiler. Eric House also reported this. I've not > > investigated this yet. We may have inefficiencies other than the code > > generated by the compiler. Some

Re: [Cegcc-devel] Performance considerations

2007-09-19 Thread pgeorges
Danny Backx a écrit : > On Wed, 2007-09-19 at 11:51 +0200, pgeorges wrote: > I ported some chess engines to Pocket PC with cegecc and noticed the number of calculated positions per second is about 30% lower with my ports than the equivalent made by other people who used Microsof

Re: [Cegcc-devel] Performance considerations

2007-09-19 Thread Danny Backx
On Wed, 2007-09-19 at 11:51 +0200, pgeorges wrote: > >> I ported some chess engines to Pocket PC with cegecc and noticed the > >> number of calculated positions per second is about 30% lower with my > >> ports than the equivalent made by other people who used Microsoft > >> compilers. > Playing

Re: [Cegcc-devel] Performance considerations

2007-09-19 Thread pgeorges
Pedro Alves a écrit : > pgeorges wrote: >> Hi, >> >> I ported some chess engines to Pocket PC with cegecc and noticed the >> number of calculated positions per second is about 30% lower with my >> ports than the equivalent made by other people who used Microsoft >> compilers. >> Years ago I made

Re: [Cegcc-devel] Performance considerations

2007-09-15 Thread Pedro Alves
pgeorges wrote: > Hi, > > I ported some chess engines to Pocket PC with cegecc and noticed the > number of calculated positions per second is about 30% lower with my > ports than the equivalent made by other people who used Microsoft compilers. > Years ago I made some comparisons between gcc and

[Cegcc-devel] Performance considerations

2007-09-15 Thread pgeorges
Hi, I ported some chess engines to Pocket PC with cegecc and noticed the number of calculated positions per second is about 30% lower with my ports than the equivalent made by other people who used Microsoft compilers. Years ago I made some comparisons between gcc and icc under Linux and notice