[PATCH] D79754: [OpenMP][AMDGCN] Support OpenMP offloading for AMDGCN architecture - Part 1

[Saiyedul Islam via Phabricator via llvm-commits]

saiislam created this revision.
saiislam added reviewers: ronlieb, yaxunl, b-sumner, scchan, JonChesterfield.
Herald added subscribers: llvm-commits, cfe-commits, dexonsmith, guansong.
Herald added a reviewer: jdoerfert.
Herald added projects: clang, LLVM.
saiislam updated this revision to Diff 263314.
saiislam added a comment.

Fixed lint errors.


Allow AMDGCN as a GPU offloading target for OpenMP during compiler
invocation and allow setting CUDAMode for it.

Originally authored by Greg Rodgers (@gregrodgers).


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D79754

Files:
  clang/lib/AST/Decl.cpp
  clang/lib/Frontend/CompilerInvocation.cpp
  llvm/include/llvm/ADT/Triple.h


Index: llvm/include/llvm/ADT/Triple.h
===================================================================
--- llvm/include/llvm/ADT/Triple.h
+++ llvm/include/llvm/ADT/Triple.h
@@ -692,6 +692,9 @@
     return getArch() == Triple::nvptx || getArch() == Triple::nvptx64;
   }
 
+  /// Tests whether the target is AMDGCN
+  bool isAMDGCN() const { return getArch() == Triple::amdgcn; }
+
   bool isAMDGPU() const {
     return getArch() == Triple::r600 || getArch() == Triple::amdgcn;
   }
Index: clang/lib/Frontend/CompilerInvocation.cpp
===================================================================
--- clang/lib/Frontend/CompilerInvocation.cpp
+++ clang/lib/Frontend/CompilerInvocation.cpp
@@ -3099,7 +3099,8 @@
 
   // Set the flag to prevent the implementation from emitting device exception
   // handling code for those requiring so.
-  if ((Opts.OpenMPIsDevice && T.isNVPTX()) || Opts.OpenCLCPlusPlus) {
+  if ((Opts.OpenMPIsDevice && (T.isNVPTX() || T.isAMDGCN())) ||
+      Opts.OpenCLCPlusPlus) {
     Opts.Exceptions = 0;
     Opts.CXXExceptions = 0;
   }
@@ -3133,6 +3134,7 @@
             TT.getArch() == llvm::Triple::ppc64le ||
             TT.getArch() == llvm::Triple::nvptx ||
             TT.getArch() == llvm::Triple::nvptx64 ||
+            TT.getArch() == llvm::Triple::amdgcn ||
             TT.getArch() == llvm::Triple::x86 ||
             TT.getArch() == llvm::Triple::x86_64))
         Diags.Report(diag::err_drv_invalid_omp_target) << A->getValue(i);
@@ -3150,13 +3152,13 @@
           << Opts.OMPHostIRFile;
   }
 
-  // Set CUDA mode for OpenMP target NVPTX if specified in options
-  Opts.OpenMPCUDAMode = Opts.OpenMPIsDevice && T.isNVPTX() &&
+  // Set CUDA mode for OpenMP target NVPTX/AMDGCN if specified in options
+  Opts.OpenMPCUDAMode = Opts.OpenMPIsDevice && (T.isNVPTX() || T.isAMDGCN()) &&
                         Args.hasArg(options::OPT_fopenmp_cuda_mode);
 
-  // Set CUDA mode for OpenMP target NVPTX if specified in options
+  // Set CUDA mode for OpenMP target NVPTX/AMDGCN if specified in options
   Opts.OpenMPCUDAForceFullRuntime =
-      Opts.OpenMPIsDevice && T.isNVPTX() &&
+      Opts.OpenMPIsDevice && (T.isNVPTX() || T.isAMDGCN()) &&
       Args.hasArg(options::OPT_fopenmp_cuda_force_full_runtime);
 
   // Record whether the __DEPRECATED define was requested.
Index: clang/lib/AST/Decl.cpp
===================================================================
--- clang/lib/AST/Decl.cpp
+++ clang/lib/AST/Decl.cpp
@@ -3221,6 +3221,11 @@
       !(BuiltinID == Builtin::BIprintf || BuiltinID == Builtin::BImalloc))
     return 0;
 
+  if (Context.getTargetInfo().getTriple().isAMDGCN() &&
+      Context.BuiltinInfo.isPredefinedLibFunction(BuiltinID) &&
+      !(BuiltinID == Builtin::BIprintf || BuiltinID == Builtin::BImalloc))
+    return 0;
+
   return BuiltinID;
 }
 


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D79754.263314.patch
Type: text/x-patch
Size: 2831 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20200512/12ccbbf6/attachment.bin>