gcc-4.5.0 Success Story with Freecell Solver

[Shlomi Fish]

Hi Diego,

On Tuesday 15 Jun 2010 23:06:04 Diego Iastrubni wrote:
> Shlomi:
> 
> <code>
> svn co http://llvm.org/svn/llvm-project/llvm/trunk llvm
> mkdir -p llvm/tools/clang
> svn co http://llvm.org/svn/llvm-project/cfe/trunk llvm/tools/clang
> mkdir llvm/cmake-build
> cd llvm/cmake-build
> ccmake ../
> make
> export CC=$PWD/bin/clang
> export CXX=$PWD/bin/clang++
> </code>
> 

thanks for the recipe to build clang.

> Now re-compile your library + test tools.
> 
> (PS, If you want to fully bootstrap clang instead of just using stage1
> here, use this "ugly" shell script to get a "clang built with clang built
> clang", aka stage 3).
> 
> Please report how good/bad clang is for your case.
> 

Well, after I built clang (which took a while) I was able to compile Freecell 
Solver using it. The first thing I should note is that the compilation 
appeared to have been much slower than with gcc. 

Then I benchmarked the results. The benchmark-using-clang ran at 
105.567487955093 seconds instead of 90.6 seconds for gcc-4.5.0 - much worse! I 
don't rule out that my code has some gccisms which make the clang performance 
worse, but it still seems that gcc-4.5.0 is the superior compiler.

Regards,

	Shlomi Fish

-- 
-----------------------------------------------------------------
Shlomi Fish       http://www.shlomifish.org/
Stop Using MSIE - http://www.shlomifish.org/no-ie/

God considered inflicting XSLT as the tenth plague of Egypt, but then
decided against it because he thought it would be too evil.

Please reply to list if it's a mailing list post - http://shlom.in/reply .