diff options
| author | Hans Wennborg <hans@hanshq.net> | 2018-09-04 09:29:18 +0000 |
|---|---|---|
| committer | Hans Wennborg <hans@hanshq.net> | 2018-09-04 09:29:18 +0000 |
| commit | 8111e8d5127bbd07dee4e43914bff34ab894b269 (patch) | |
| tree | d3e4d5a2f5bd6baf10d62f414ab8d3123e7de170 /lib | |
| parent | 1a558539efc93d166ead9439a13c5226359c394e (diff) | |
Merging r340959:
------------------------------------------------------------------------
r340959 | mareko | 2018-08-29 22:03:00 +0200 (Wed, 29 Aug 2018) | 9 lines
AMDGPU: Handle 32-bit address wraparounds for SMRD opcodes
Summary: This fixes GPU hangs with OpenGL bindless handle arithmetic.
Reviewers: arsenm, nhaehnle
Subscribers: kzhuravl, jvesely, wdng, yaxunl, dstuttard, tpr, t-tye, llvm-commits
Differential Revision: https://reviews.llvm.org/D51203
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_70@341351 91177308-0d34-0410-b5e6-96231b3b80d8
Diffstat (limited to 'lib')
| -rw-r--r-- | lib/Target/AMDGPU/AMDGPUISelDAGToDAG.cpp | 6 |
1 files changed, 5 insertions, 1 deletions
diff --git a/lib/Target/AMDGPU/AMDGPUISelDAGToDAG.cpp b/lib/Target/AMDGPU/AMDGPUISelDAGToDAG.cpp index f25f4d4693e..7cb0e12a680 100644 --- a/lib/Target/AMDGPU/AMDGPUISelDAGToDAG.cpp +++ b/lib/Target/AMDGPU/AMDGPUISelDAGToDAG.cpp @@ -1451,7 +1451,11 @@ bool AMDGPUDAGToDAGISel::SelectSMRD(SDValue Addr, SDValue &SBase, SDValue &Offset, bool &Imm) const { SDLoc SL(Addr); - if (CurDAG->isBaseWithConstantOffset(Addr)) { + // A 32-bit (address + offset) should not cause unsigned 32-bit integer + // wraparound, because s_load instructions perform the addition in 64 bits. + if ((Addr.getValueType() != MVT::i32 || + Addr->getFlags().hasNoUnsignedWrap()) && + CurDAG->isBaseWithConstantOffset(Addr)) { SDValue N0 = Addr.getOperand(0); SDValue N1 = Addr.getOperand(1); |