NVPTX -> OpenCL

hpvm/docs/compilation.md

@@ -5,7 +5,7 @@ Compilation of an HPVM program involves the following steps:
 2. `opt` takes (`main.ll`) and invoke the GenHPVM pass on it, which converts the HPVM-C function calls to HPVM intrinsics. This generates the HPVM textual representation (`main.hpvm.ll`).
 3. `opt` takes the HPVM textual representation (`main.hpvm.ll`) and invokes the following passes in sequence: 
     * BuildDFG: Converts the textual representation to the internal HPVM representation.
-    * LocalMem and DFG2LLVM_NVPTX: Invoked only when GPU target is selected. Generates the kernel module (`main.kernels.ll`) and the portion of the host code that invokes the kernel into the host module (`main.host.ll`).
+    * LocalMem and DFG2LLVM_OpenCL: Invoked only when GPU target is selected. Generates the kernel module (`main.kernels.ll`) and the portion of the host code that invokes the kernel into the host module (`main.host.ll`).
     * DFG2LLVM_CPU: Generates either all, or the remainder of the host module (`main.host.ll`) depending on the chosen target.
     * ClearDFG: Deletes the internal HPVM representation from memory.
 4. `clang` is used to to compile any remaining project files that would be later linked with the host module.

hpvm/lib/Transforms/CMakeLists.txt

 add_subdirectory(BuildDFG)
 add_subdirectory(ClearDFG)
-add_subdirectory(DFG2LLVM_NVPTX)
+add_subdirectory(DFG2LLVM_OpenCL)
 add_subdirectory(DFG2LLVM_CPU)
 add_subdirectory(GenHPVM)
 add_subdirectory(LocalMem)

hpvm/lib/Transforms/DFG2LLVM_CPU/DFG2LLVM_CPU.cpp

@@ -1426,7 +1426,7 @@ void CGT_CPU::codeGen(DFLeafNode *N) {
     switch (N->getTag()) {
     case hpvm::GPU_TARGET:
       // A leaf node should not have an cpu function for GPU
-      // by design of DFG2LLVM_NVPTX backend
+      // by design of DFG2LLVM_OpenCL backend
       assert(!(N->hasCPUGenFuncForTarget(hpvm::GPU_TARGET)) &&
              "Leaf node not expected to have GPU GenFunc");
       break;

hpvm/lib/Transforms/DFG2LLVM_NVPTX/CMakeLists.txt → hpvm/lib/Transforms/DFG2LLVM_OpenCL/CMakeLists.txt

@@ -4,9 +4,9 @@ endif()
 
 set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -DLLVM_BUILD_DIR=${PROJECT_BINARY_DIR}")
 
-add_llvm_library( LLVMDFG2LLVM_NVPTX
+add_llvm_library( LLVMDFG2LLVM_OpenCL
   MODULE
-  DFG2LLVM_NVPTX.cpp
+  DFG2LLVM_OpenCL.cpp
 
   DEPENDS
   intrinsics_gen

hpvm/lib/Transforms/DFG2LLVM_NVPTX/LLVMBuild.txt → hpvm/lib/Transforms/DFG2LLVM_OpenCL/LLVMBuild.txt

-;===- ./lib/Transforms/DFG2LLVM_NVPTX/LLVMBuild.txt ------------*- Conf -*--===;
+;===- ./lib/Transforms/DFG2LLVM_OpenCL/LLVMBuild.txt ------------*- Conf -*--===;
 ;
 ;                     The LLVM Compiler Infrastructure
 ;
@@ -17,5 +17,5 @@
 
 [component_0]
 type = Library
-name = DFG2LLVM_NVPTX
+name = DFG2LLVM_OpenCL
 parent = Transforms

hpvm/test/benchmarks/hpvm-cava/Makefile

@@ -65,7 +65,7 @@ ifeq ($(TARGET),seq)
   HPVM_OPTFLAGS += -hpvm-timers-cpu
 else
   DEVICE = GPU_TARGET
-  HPVM_OPTFLAGS = -load LLVMBuildDFG.so -load LLVMLocalMem.so -load LLVMDFG2LLVM_NVPTX.so -load LLVMDFG2LLVM_CPU.so -load LLVMClearDFG.so -localmem -dfg2llvm-nvptx -dfg2llvm-cpu -clearDFG
+  HPVM_OPTFLAGS = -load LLVMBuildDFG.so -load LLVMLocalMem.so -load LLVMDFG2LLVM_OpenCL.so -load LLVMDFG2LLVM_CPU.so -load LLVMClearDFG.so -localmem -dfg2llvm-nvptx -dfg2llvm-cpu -clearDFG
   HPVM_OPTFLAGS += -hpvm-timers-cpu -hpvm-timers-ptx
 endif
   TESTGEN_OPTFLAGS += -hpvm-timers-gen

hpvm/test/benchmarks/parboil/common/mk/hpvm.mk

@@ -22,7 +22,7 @@ ifeq ($(TARGET),seq)
   HPVM_OPTFLAGS = -load LLVMBuildDFG.so -load LLVMDFG2LLVM_CPU.so -load LLVMClearDFG.so -dfg2llvm-cpu -clearDFG
 else
   DEVICE = GPU_TARGET
-  HPVM_OPTFLAGS = -load LLVMBuildDFG.so -load LLVMLocalMem.so -load LLVMDFG2LLVM_NVPTX.so -load LLVMDFG2LLVM_CPU.so -load LLVMClearDFG.so -localmem -dfg2llvm-nvptx -dfg2llvm-cpu -clearDFG
+  HPVM_OPTFLAGS = -load LLVMBuildDFG.so -load LLVMLocalMem.so -load LLVMDFG2LLVM_OpenCL.so -load LLVMDFG2LLVM_CPU.so -load LLVMClearDFG.so -localmem -dfg2llvm-nvptx -dfg2llvm-cpu -clearDFG
 endif
 
 CFLAGS += -DDEVICE=$(DEVICE)

hpvm/test/benchmarks/pipeline/Makefile

@@ -52,7 +52,7 @@ ifeq ($(TARGET),seq)
   HPVM_OPTFLAGS += -hpvm-timers-cpu
 else
   DEVICE = GPU_TARGET
-  HPVM_OPTFLAGS = -load LLVMBuildDFG.so -load LLVMLocalMem.so -load LLVMDFG2LLVM_NVPTX.so -load LLVMDFG2LLVM_CPU.so -load LLVMClearDFG.so -localmem -dfg2llvm-nvptx -dfg2llvm-cpu -clearDFG
+  HPVM_OPTFLAGS = -load LLVMBuildDFG.so -load LLVMLocalMem.so -load LLVMDFG2LLVM_OpenCL.so -load LLVMDFG2LLVM_CPU.so -load LLVMClearDFG.so -localmem -dfg2llvm-nvptx -dfg2llvm-cpu -clearDFG
   HPVM_OPTFLAGS += -hpvm-timers-cpu -hpvm-timers-ptx
 endif
   TESTGEN_OPTFLAGS += -hpvm-timers-gen

hpvm/test/benchmarks/template/Makefile

@@ -56,7 +56,7 @@ ifeq ($(TARGET),seq)
   HPVM_OPTFLAGS += -hpvm-timers-cpu
 else
   DEVICE = GPU_TARGET
-  HPVM_OPTFLAGS = -load LLVMBuildDFG.so -load LLVMLocalMem.so -load LLVMDFG2LLVM_NVPTX.so -load LLVMDFG2LLVM_CPU.so -load LLVMClearDFG.so -localmem -dfg2llvm-nvptx -dfg2llvm-cpu -clearDFG
+  HPVM_OPTFLAGS = -load LLVMBuildDFG.so -load LLVMLocalMem.so -load LLVMDFG2LLVM_OpenCL.so -load LLVMDFG2LLVM_CPU.so -load LLVMClearDFG.so -localmem -dfg2llvm-nvptx -dfg2llvm-cpu -clearDFG
   HPVM_OPTFLAGS += -hpvm-timers-cpu -hpvm-timers-ptx
 endif
   TESTGEN_OPTFLAGS += -hpvm-timers-gen

hpvm/test/regressionTests/DFG2LLVM_NVPTX/ThreeLevel.atomic.genvisc.ll

-; RUN: opt -load LLVMBuildDFG.so -load LLVMLocalMem.so -load LLVMDFG2LLVM_NVPTX.so -S -localmem -dfg2llvm-nvptx <  %s | FileCheck %s
+; RUN: opt -load LLVMBuildDFG.so -load LLVMLocalMem.so -load LLVMDFG2LLVM_OpenCL.so -S -localmem -dfg2llvm-nvptx <  %s | FileCheck %s
 ; ModuleID = 'ThreeLevel.atomic.ll'
 source_filename = "ThreeLevel.constmem.c"
 target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"

hpvm/test/regressionTests/DFG2LLVM_NVPTX/ThreeLevel.genvisc.ll

-; RUN: opt -load LLVMBuildDFG.so -load LLVMLocalMem.so -load LLVMDFG2LLVM_NVPTX.so -S -localmem -dfg2llvm-nvptx <  %s | FileCheck %s
+; RUN: opt -load LLVMBuildDFG.so -load LLVMLocalMem.so -load LLVMDFG2LLVM_OpenCL.so -S -localmem -dfg2llvm-nvptx <  %s | FileCheck %s
 ; ModuleID = 'ThreeLevel.ll'
 source_filename = "ThreeLevel.c"
 target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"

hpvm/test/regressionTests/DFG2LLVM_NVPTX/ThreeLevel.opt.genvisc.ll

-; RUN: opt -load LLVMBuildDFG.so -load LLVMLocalMem.so -load LLVMDFG2LLVM_NVPTX.so -S -localmem -dfg2llvm-nvptx <  %s | FileCheck %s
+; RUN: opt -load LLVMBuildDFG.so -load LLVMLocalMem.so -load LLVMDFG2LLVM_OpenCL.so -S -localmem -dfg2llvm-nvptx <  %s | FileCheck %s
 ; ModuleID = 'ThreeLevel.opt.ll'
 source_filename = "ThreeLevel.opt.c"
 target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"