polyhedron 2005 results for llvm svn

Project Infrastructure LLVM Dev List Archives
1

Anton,
   I didn't have to time to do fresh set of benchmarks for
llvm-gfortran from 2.4 and current gcc trunk. I'll do that
this weekend. However I suspect I'll see similar behavior
as before. Compared to FSF gcc 4.2, llvm-gfortran will show
improvements but it won't be as impressive compared to
gcc trunk. Is there a projected date for supporting the
vectorization in llvm for the gcc 4.2 front-end?
            Jack
ps I may go ahead an release a fink llvm-gcc42 package
for 2.4 with the patch from llvm svn and disabling the
warning flags in CC1_SPEC of darwin.h so that llvm-gfortran
gets some usage.

2

Anton,
  I didn't have to time to do fresh set of benchmarks for
llvm-gfortran from 2.4 and current gcc trunk. I'll do that
this weekend. However I suspect I'll see similar behavior
as before. Compared to FSF gcc 4.2, llvm-gfortran will show
improvements but it won't be as impressive compared to
gcc trunk. Is there a projected date for supporting the
vectorization in llvm for the gcc 4.2 front-end?

Whenever LLVM vectorization supports vectorization it will be part of llvm optimizer, which means it will be independent of llvm-gcc. Meanwhile, it is a good idea to open bugzilla reports, @ llvm.org, from polyhedron test cases where gcc's vectorization helps significantly.

           Jack
ps I may go ahead an release a fink llvm-gcc42 package
for 2.4 with the patch from llvm svn and disabling the
warning flags in CC1_SPEC of darwin.h so that llvm-gfortran
gets some usage.

I sent you a patch in separate email that may help gfortran. Try it first.