AMDGPU: Prevent uniform loops from becoming infinite

Summary:
Uniform loops where the branch leaving the loop is predicated on VCCNZ
must be skipped if EXEC = 0, otherwise they will be infinite.

Reviewers: tstellarAMD, arsenm

Subscribers: arsenm, llvm-commits

Differential Revision: http://reviews.llvm.org/D18137

llvm-svn: 263658

1 files changed, 6 insertions, 0 deletions

diff --git a/llvm/lib/Target/AMDGPU/SILowerControlFlow.cpp b/llvm/lib/Target/AMDGPU/SILowerControlFlow.cpp
index 7dd0d7bc8f7..2a645d15ca2 100644
--- a/llvm/lib/Target/AMDGPU/SILowerControlFlow.cpp
+++ b/llvm/lib/Target/AMDGPU/SILowerControlFlow.cpp
@@ -137,6 +137,12 @@ bool SILowerControlFlow::shouldSkip(MachineBasicBlock *From,
          NumInstr < SkipThreshold && I != E; ++I) {
 
       if (I->isBundle() || !I->isBundled())
+        // When a uniform loop is inside non-uniform control flow, the branch
+        // leaving the loop might be an S_CBRANCH_VCCNZ, which is never taken
+        // when EXEC = 0. We should skip the loop lest it becomes infinite.
+        if (I->getOpcode() == AMDGPU::S_CBRANCH_VCCNZ)
+          return true;
+
         if (++NumInstr >= SkipThreshold)
           return true;
     }