[NVPTX][NewPM] Re-enable NVVMReflectPass

Disabled alongside NVVMIntrRangePass in https://reviews.llvm.org/D96166,
but turns out NVVMIntrRangePass was the issue.

Reviewed By: tra

Differential Revision: https://reviews.llvm.org/D96291

diff --git a/llvm/lib/Target/NVPTX/NVPTXTargetMachine.cpp b/llvm/lib/Target/NVPTX/NVPTXTargetMachine.cpp
index 2158838..3a31b3b 100644 (file)
--- a/llvm/lib/Target/NVPTX/NVPTXTargetMachine.cpp
+++ b/llvm/lib/Target/NVPTX/NVPTXTargetMachine.cpp
@@ -222,17 +222,16 @@ void NVPTXTargetMachine::registerPassBuilderCallbacks(PassBuilder &PB,
         return false;
       });
 
-  // FIXME: these passes are causing numerical discrepancies, investigate and
-  // re-enable.
-
-  // PB.registerPipelineStartEPCallback(
-  //     [this, DebugPassManager](ModulePassManager &PM,
-  //                              PassBuilder::OptimizationLevel Level) {
-  //       FunctionPassManager FPM(DebugPassManager);
-  //       FPM.addPass(NVVMReflectPass(Subtarget.getSmVersion()));
-  //       FPM.addPass(NVVMIntrRangePass(Subtarget.getSmVersion()));
-  //       PM.addPass(createModuleToFunctionPassAdaptor(std::move(FPM)));
-  //     });
+  PB.registerPipelineStartEPCallback(
+      [this, DebugPassManager](ModulePassManager &PM,
+                               PassBuilder::OptimizationLevel Level) {
+        FunctionPassManager FPM(DebugPassManager);
+        FPM.addPass(NVVMReflectPass(Subtarget.getSmVersion()));
+        // FIXME: NVVMIntrRangePass is causing numerical discrepancies,
+        // investigate and re-enable.
+        // FPM.addPass(NVVMIntrRangePass(Subtarget.getSmVersion()));
+        PM.addPass(createModuleToFunctionPassAdaptor(std::move(FPM)));
+      });
 }
 
 TargetTransformInfo

diff --git a/llvm/test/CodeGen/NVPTX/nvvm-reflect-arch.ll b/llvm/test/CodeGen/NVPTX/nvvm-reflect-arch.ll
index faab355..8e8d866 100644 (file)
--- a/llvm/test/CodeGen/NVPTX/nvvm-reflect-arch.ll
+++ b/llvm/test/CodeGen/NVPTX/nvvm-reflect-arch.ll
@@ -1,10 +1,9 @@
 ; Libdevice in recent CUDA versions relies on __CUDA_ARCH reflecting GPU type.
 ; Verify that __nvvm_reflect() is replaced with an appropriate value.
 ;
-; FIXME: fix pass and re-enable under new PM
-; RUN: opt %s -S -nvvm-reflect -O2 -enable-new-pm=0 -mtriple=nvptx64 \
+; RUN: opt %s -S -nvvm-reflect -O2 -mtriple=nvptx64 \
 ; RUN:   | FileCheck %s --check-prefixes=COMMON,SM20
-; RUN: opt %s -S -nvvm-reflect -O2 -enable-new-pm=0 -mtriple=nvptx64 -mcpu=sm_35 \
+; RUN: opt %s -S -nvvm-reflect -O2 -mtriple=nvptx64 -mcpu=sm_35 \
 ; RUN:   | FileCheck %s --check-prefixes=COMMON,SM35
 
 @"$str" = private addrspace(1) constant [12 x i8] c"__CUDA_ARCH\00"