Date: Tue, 19 Jun 2012 17:14:25 +0200 From: Fred Morcos <fred.morcos@gmail.com> To: Michel Talon <talon@lpthe.jussieu.fr> Cc: Wojciech Puchar <wojtek@wojtek.tensor.gdynia.pl>, David Brodbeck <gull@gull.us>, FreeBSD Questions <freebsd-questions@freebsd.org> Subject: Re: Why Clang Message-ID: <CAH3a3KWKNF5Bt-8=KgtbMh=rV6GfUO7OaeE6-SutxkcRe8cG3Q@mail.gmail.com> In-Reply-To: <402199FE-380B-41B6-866B-7D5D66C457D5@lpthe.jussieu.fr> References: <402199FE-380B-41B6-866B-7D5D66C457D5@lpthe.jussieu.fr>
next in thread | previous in thread | raw e-mail | index | archive | help
I would also guess that the base system is stuck with gcc ~4.1 due to the GPLv3-ization of later gcc version. Is that correct? On Tue, Jun 19, 2012 at 4:43 PM, Michel Talon <talon@lpthe.jussieu.fr> wrot= e: > David Brodbeck said: >> Another way of looking at it is after 25 years of optimization GCC is >> unable to beat a new compiler that's had almost none... > Unfortunately this affirmation is blatantly false, recent gcc produce cod= e > much faster than clang. I give here an example which i like, a monte carl= o computation for a spin lattice. > Everything runs on my macbook. > > lilas% clang -v > Apple clang version 2.1 (tags/Apple/clang-163.7.1) (based on LLVM 3.0svn) > Target: x86_64-apple-darwin11.4.0 > lilas% clang -O4 test.c -lf2c > lilas% time ./a.out > ... > > real =A0 =A00m2.359s > user =A0 =A00m2.341s > sys =A0 =A0 0m0.003s > > lilas% /usr/local/bin/gcc -v > =85 > gcc version 4.6.1 (GCC) > > lilas% /usr/local/bin/gcc -O3 test.c -lf2c > lilas% time ./a.out > =85 > > real =A0 =A00m1.241s > user =A0 =A00m1.234s > sys =A0 =A0 0m0.003s > > So gcc gives an executable running twice faster than clang, basically, wh= en both compilers > are run at maximal optimization. To show the effectiveness of the optimiz= er, here is the running > time without any optimization: > > lilas% /usr/local/bin/gcc =A0test.c -lf2c > lilas% time ./a.out > =85 > > real =A0 =A00m6.895s > user =A0 =A00m6.889s > sys =A0 =A0 0m0.005s > > What this demonstrates is that for programs which do real computations, o= ptimization is > *very* important, and gcc is now very good (i have not shown the numbers = but they match the Intel compiler) > while clang is at the level gcc was ten years ago. So i fully agree with = Wojciech Puchar, the move to clang > is only driven by anti GPL propaganda which is frankly completely stupid,= since in any events, gcc > does not contaminate the binaries it produces (except when using contamin= ated accompanying libraries > e.g. for C++). Of course, when compiling FreeBSD kernel or similar progra= ms which do little computation > there is no harm using clang. I suspect that the price is higher for prog= rams like mencoder which require > the highest efficiency. > > I will not comment on the better error messages coming from clang, this c= ould be a more serious argument. > > -- > > Michel Talon > talon@lpthe.jussieu.fr > > > > >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAH3a3KWKNF5Bt-8=KgtbMh=rV6GfUO7OaeE6-SutxkcRe8cG3Q>