[Polly] Added OpenCL Runtime to GPURuntime Library for GPGPU CodeGen
authorSiddharth Bhat <siddu.druid@gmail.com>
Fri, 5 May 2017 07:54:49 +0000 (07:54 +0000)
committerSiddharth Bhat <siddu.druid@gmail.com>
Fri, 5 May 2017 07:54:49 +0000 (07:54 +0000)
commit51904ae35aada8243afe53ac58e14fc8788661ee
treef61ac3f00759f0dd7f3748a5388d5706a0b67b31
parent07a17719592767c959c7f524c94871972bdea7c6
[Polly] Added OpenCL Runtime to GPURuntime Library for GPGPU CodeGen

Summary:
When compiling for GPU, one can now choose to compile for OpenCL or CUDA,
with the corresponding polly-gpu-runtime flag (libopencl / libcudart). The
GPURuntime library (GPUJIT) has been extended with the OpenCL Runtime library
for that purpose, correctly choosing the corresponding library calls to the
option chosen when compiling (via different initialization calls).

Additionally, a specific GPU Target architecture can now be chosen with -polly-gpu-arch (only nvptx64 implemented thus far).

Reviewers: grosser, bollu, Meinersbur, etherzhhb, singam-sanjay

Reviewed By: grosser, Meinersbur

Subscribers: singam-sanjay, llvm-commits, pollydev, nemanjai, mgorny, yaxunl, Anastasia

Tags: #polly

Differential Revision: https://reviews.llvm.org/D32431

llvm-svn: 302215
polly/CMakeLists.txt
polly/include/polly/CodeGen/PPCGCodeGeneration.h [new file with mode: 0644]
polly/include/polly/LinkAllPasses.h
polly/lib/CodeGen/PPCGCodeGeneration.cpp
polly/lib/Support/RegisterPasses.cpp
polly/test/GPGPU/cuda-managed-memory-simple.ll
polly/test/GPGPU/size-cast.ll
polly/tools/CMakeLists.txt
polly/tools/GPURuntime/GPUJIT.c
polly/tools/GPURuntime/GPUJIT.h