RE: [Mingw-users] Really slow executables generated by Msys/Mingw gcc

[Danny S. <dan...@cl...>]

> -----Original Message-----
> From: Danny Smith 
> Sent: Tuesday, March 21, 2006 3:40 PM
> 
> Care to provide a testcase so we all know what your talking about?

<blush> I apologize.  You did send a testcase in original posting.  

OK.   Looking at assembler code, the difference is that 

(1)current version of mingw doesn't use -ffast-math (replacing library
calls by inline asm) by default.  This is a safe default since library
calls have greater range/accuracy near extremes than naked asm. Also
libray calls may have side effect of setting fp errs.
(2) Not inlining sin -> fsin, cos -> fcos prevents other optimizations/
(3) GCC-3.4.5 is not smart enough even with --ffast-math to replace
fsin/fcos calls with fsincos calls when advantageous (it is in your
testcase). GCC-4.1 is.

With your testcase and and gcc-3.4.5  and switches -O2 -ffast-math
time is reported as about 20-22 on my Athlon 2000,


With your testcase and and gcc-4.2 and switches -O2 -ffast-math
time is reported as about 7-8  on my Athlon 2000,

With your testcase and and msvc6 with option /O2 
time is reported as 10-11.

HTH

Danny



 

> > 
>