[llvm-bugs] [Bug 43463] New: HexagonOptAddrMode::changeStore - null dereference warning

[via llvm-bugs]

https://bugs.llvm.org/show_bug.cgi?id=43463

            Bug ID: 43463
           Summary: HexagonOptAddrMode::changeStore - null dereference
                    warning
           Product: libraries
           Version: trunk
          Hardware: PC
                OS: Windows NT
            Status: NEW
          Severity: enhancement
          Priority: P
         Component: Backend: Hexagon
          Assignee: unassignedbugs at nondot.org
          Reporter: llvm-dev at redking.me.uk
                CC: eugene.zelenko at gmail.com, kparzysz at quicinc.com,
                    llvm-bugs at lists.llvm.org

https://llvm.org/reports/scan-build/report-HexagonOptAddrMode.cpp-add-1-1.html#EndPath

MachineInstrBuilder MIB;
if (ImmOpNum == 0) {
  if (HII->getAddrMode(*OldMI) == HexagonII::BaseRegOffset) {
    short NewOpCode = HII->changeAddrMode_rr_ur(*OldMI);
    assert(NewOpCode >= 0 && "Invalid New opcode\n");
    MIB = BuildMI(*BB, InsertPt, OldMI->getDebugLoc(), HII->get(NewOpCode));
    MIB.add(OldMI->getOperand(1));
    MIB.add(OldMI->getOperand(2));
    MIB.add(ImmOp);
    MIB.add(OldMI->getOperand(3));
    OpStart = 4;
  } else if (HII->getAddrMode(*OldMI) == HexagonII::BaseImmOffset) {
    short NewOpCode = HII->changeAddrMode_io_abs(*OldMI);
    assert(NewOpCode >= 0 && "Invalid New opcode\n");
    MIB = BuildMI(*BB, InsertPt, OldMI->getDebugLoc(), HII->get(NewOpCode));
    const GlobalValue *GV = ImmOp.getGlobal();
    int64_t Offset = ImmOp.getOffset() + OldMI->getOperand(1).getImm();
    MIB.addGlobalAddress(GV, Offset, ImmOp.getTargetFlags());
    MIB.add(OldMI->getOperand(2));
    OpStart = 3;
  }
  Changed = true;
  LLVM_DEBUG(dbgs() << "[Changing]: " << *OldMI << "\n");
  LLVM_DEBUG(dbgs() << "[TO]: " << *MIB << "\n");
}

clang static analyzer is warning that Changed could be set to true even if the
MI builder is still empty (null). Should the Changed = true be put inside both
the if/else cases?

-- 
You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20190926/b9811f7d/attachment.html>