[Llvm-bgq-discuss] bgclang r220548-20141024 on vesta/mira

[Hal Finkel]

Hi everyone,

I've updated our bgclang install on vesta and mira to r220548-20141024.

This new build picks up some nice upstream improvements and bug fixes and some local improvements and fixes.

 - support for __attribute__((align_value(n))) and #pragma omp simd aligned(n) for specifying pointer alignments

 - better support for QPX vectorization of math functions with -ffast-math is enabled (also, this build uses the scalar SLEEF functions for exp/exp2/exp10 by default because they're much faster than the libm versions).

 - bgclang++11 includes a work-around for std::locale("") being broken (because libc's newlocale is broken when provided with "") -- hopefully boost.filesystem will work better now.

 - A work-around (via a wrapper header) is provided for spi/include/kernel/process.h so you shouldn't need to worry about the missing __INLINE__ on Kernel_GetJobID in the system header any more.

For those maintaining their own installs, RPMs (and SRPMs) are available on the trac page: https://trac.alcf.anl.gov/projects/llvm-bgq

Thanks again,
Hal

-- 
Hal Finkel
Assistant Computational Scientist
Leadership Computing Facility
Argonne National Laboratory
_______________________________________________
llvm-bgq-discuss mailing list
llvm-bgq-discuss at lists.alcf.anl.gov
https://lists.alcf.anl.gov/mailman/listinfo/llvm-bgq-discuss