Use Julia's library binaries

IntelLabs / ParallelAccelerator.jl

The ParallelAccelerator package, part of the High Performance Scripting project at Intel Labs

BSD 2-Clause "Simplified" License

294 stars 32 forks source link

Use Julia's library binaries #15

Closed ehsantn closed 8 years ago

ehsantn commented 8 years ago

Julia uses many libraries like OpenBLAS for fast implementation of math routines. ParallelAccelerator needs to find out which ones are installed and where, in order to use them.

Example: for GEMM, ParallelAccelerator looks for MKL or OpenBLAS in system's PATH, but Julia carries it in some binary (libopenblas64_.so in binary installation of v0.4 Julia).

ehsantn commented 8 years ago

This doesn't seem to be possible in a portable way since we generate a shared library and we can't access other loaded shared libraries.