[DebugInfo][AArch64] Recognise target specific instruction as mov instr

[avl]

Summary

This fix is for the problem from https://bugs.llvm.org/show_bug.cgi?id=38714.
Specifically, Simple Register Coalescing creates following conversion :

    
undef %0.sub_32:gpr64 = ORRWrs $wzr, %3:gpr32common, 0, debug-location !24;

It copies 32-bit value from gpr32 into gpr64. But Live DEBUG_VALUE analysis
is not able to create debug location record for that instruction. So the problem
is in that debug location for argc variable is incorrect. The fix is
to write custom isCopyInstrImpl() which would recognize the ORRWrs instr.

[efriedma]

llvm/lib/Target/AArch64/AArch64InstrInfo.cpp

5519

I don't think we'll produce an ORRXrs until after it's relevant here, but I guess it's safe enough.

You need to check that the instruction is actually a "mov"; it looks like you're correctly checking that the shift amount is zero, but you aren't checking that the other register is wzr/xzr.

llvm/test/CodeGen/AArch64/aarch64-mov-debug-locs.ll

24 ↗

(On Diff #199887)

Can you make an MIR testcase using -run-pass instead of running the complete llc pipeline?

[avl]

addressed comments: added check for WZR/XZR registers, changed test to run only LiveDebugValues pass.

[efriedma]

llvm/lib/Target/AArch64/AArch64InstrInfo.cpp

5521

The isReg/isImm checks aren't necessary; the legal operands are determined by the opcode.

[avl]

deleted redundant check.

[efriedma]

LGTM with one minor comment

llvm/lib/Target/AArch64/AArch64InstrInfo.cpp

5523

Please clean up unnecessary parentheses (you don't need parentheses around each comparison using "&&").

llvm/lib/Target/AArch64/AArch64InstrInfo.h

public:
   static bool isFalkorShiftExtFast(const MachineInstr &MI);
   /// Return true if the instructions is a SEH instruciton used for unwinding
   /// on Windows.
   static bool isSEHInstruction(const MachineInstr &MI);
 
 #define GET_INSTRINFO_HELPER_DECLS
 #include "AArch64GenInstrInfo.inc"
 
+protected:
+  /// If the specific machine instruction is a instruction that moves/copies
+  /// value from one register to another register return true along with
+  /// @Source machine operand and @Destination machine operand.
+  bool isCopyInstrImpl(const MachineInstr &MI, const MachineOperand *&Source,
+                       const MachineOperand *&Destination) const override;
+
 private:
   /// Sets the offsets on outlined instructions in \p MBB which use SP
   /// so that they will be valid post-outlining.
   ///
   /// \param MBB A \p MachineBasicBlock in an outlined function.
   void fixupPostOutline(MachineBasicBlock &MBB) const;
 
   void instantiateCondBranch(MachineBasicBlock &MBB, const DebugLoc &DL,

llvm/lib/Target/AArch64/AArch64InstrInfo.cpp

MachineBasicBlock::iterator AArch64InstrInfo::insertOutlinedCall(
   return CallPt;
 }
 
 bool AArch64InstrInfo::shouldOutlineFromFunctionByDefault(
   MachineFunction &MF) const {
   return MF.getFunction().hasMinSize();
 }
 
+bool AArch64InstrInfo::isCopyInstrImpl(
+    const MachineInstr &MI, const MachineOperand *&Source,
+    const MachineOperand *&Destination) const {
+
+  // AArch64::ORRWrs and AArch64::ORRXrs with WZR/XZR reg
+  // and zero immediate operands used as an alias for mov instruction.
+  if ((MI.getOpcode() == AArch64::ORRWrs) && MI.getOperand(1).isReg() &&
+      (MI.getOperand(1).getReg() == AArch64::WZR) && MI.getOperand(3).isImm() &&
+      (MI.getOperand(3).getImm() == 0x0)) {
+    Destination = &MI.getOperand(0);
+    Source = &MI.getOperand(2);
+    return true;
+  }
+
+  if ((MI.getOpcode() == AArch64::ORRXrs) && MI.getOperand(1).isReg() &&
+      (MI.getOperand(1).getReg() == AArch64::XZR) && MI.getOperand(3).isImm() &&
+      (MI.getOperand(3).getImm() == 0x0)) {
+    Destination = &MI.getOperand(0);
+    Source = &MI.getOperand(2);
+    return true;
+  }
+
+  return false;
+}
+
 #define GET_INSTRINFO_HELPERS
 #include "AArch64GenInstrInfo.inc"

llvm/test/CodeGen/AArch64/aarch64-mov-debug-locs.mir

+
+# RUN: llc -mtriple=aarch64-none-linux-gnu -verify-machineinstrs -run-pass=livedebugvalues %s -o - | FileCheck %s
+
+#
+# This test is for the problem from https://bugs.llvm.org/show_bug.cgi?id=38714.
+# 
+# Specifically, Simple Register Coalescing creates following conversion :
+#
+# undef %0.sub_32:gpr64 = ORRWrs $wzr, %3:gpr32common, 0, debug-location !24;
+#
+# It copies 32-bit value from gpr32 into gpr64. But later Live DEBUG_VALUE analysis 
+# is not able to create debug location record for that instruction. The fix is
+# to write custom isCopyInstrImpl() which would recognize the ORRWrs instr as 
+# move instruction.  
+#
+# The LLVM-IR below was produced by clang on the following C++ code:
+# 
+# int printf(const char* restrict format, ...);
+# 
+# 
+# int main( int argc, char **argv )
+# {
+#   for(int i = 1; i < argc; ++i) {
+#     printf("Argument %d: %s\n", i, argv[i]);
+#   }
+#   return 0;
+# }
+# 
+
+# CHECK-LABEL: main 
+# CHECK: DBG_VALUE $w0, $noreg 
+# CHECK: renamable $[[REG1:w[0-9]+]] = ORRWrs $wzr, killed renamable $w0, 0
+# CHECK: DBG_VALUE $[[REG1]], $noreg 
+
+
+--- |
+  ; ModuleID = '<stdin>'
+  source_filename = "test.c"
+  target datalayout = "e-m:e-i8:8:32-i16:16:32-i64:64-i128:128-n32:64-S128"
+  target triple = "aarch64-unknown-linux"
+  
+  @.str = private unnamed_addr constant [17 x i8] c"Argument %d: %s\0A\00", align 1
+  
+  ; Function Attrs: nounwind
+  define dso_local i32 @main(i32 %argc, i8** nocapture readonly %argv) local_unnamed_addr #0 !dbg !7 {
+  entry:
+    call void @llvm.dbg.value(metadata i32 %argc, metadata !15, metadata !DIExpression()), !dbg !19
+    call void @llvm.dbg.value(metadata i8** %argv, metadata !16, metadata !DIExpression()), !dbg !19
+    call void @llvm.dbg.value(metadata i32 1, metadata !17, metadata !DIExpression()), !dbg !20
+    %cmp5 = icmp sgt i32 %argc, 1, !dbg !21
+    br i1 %cmp5, label %for.body.preheader, label %for.cond.cleanup, !dbg !23
+  
+  for.body.preheader:                               ; preds = %entry
+    %wide.trip.count = zext i32 %argc to i64, !dbg !24
+    br label %for.body, !dbg !25
+  
+  for.cond.cleanup:                                 ; preds = %for.body, %entry
+    ret i32 0, !dbg !27
+  
+  for.body:                                         ; preds = %for.body, %for.body.preheader
+    %indvars.iv = phi i64 [ 1, %for.body.preheader ], [ %indvars.iv.next, %for.body ]
+    call void @llvm.dbg.value(metadata i64 %indvars.iv, metadata !17, metadata !DIExpression()), !dbg !20
+    %scevgep = getelementptr i8*, i8** %argv, i64 %indvars.iv, !dbg !25
+    %0 = load i8*, i8** %scevgep, align 8, !dbg !25, !tbaa !28
+    %tmp = trunc i64 %indvars.iv to i32
+    %call = tail call i32 (i8*, ...) @printf(i8* getelementptr inbounds ([17 x i8], [17 x i8]* @.str, i64 0, i64 0), i32 %tmp, i8* %0), !dbg !32
+    %indvars.iv.next = add nuw nsw i64 %indvars.iv, 1, !dbg !33
+    call void @llvm.dbg.value(metadata i32 undef, metadata !17, metadata !DIExpression(DW_OP_plus_uconst, 1, DW_OP_stack_value)), !dbg !20
+    %exitcond = icmp eq i64 %wide.trip.count, %indvars.iv.next, !dbg !21
+    br i1 %exitcond, label %for.cond.cleanup, label %for.body, !dbg !23, !llvm.loop !34
+  }
+  
+  ; Function Attrs: nounwind
+  declare dso_local i32 @printf(i8* nocapture readonly, ...) local_unnamed_addr #1
+  
+  ; Function Attrs: nounwind readnone speculatable
+  declare void @llvm.dbg.value(metadata, metadata, metadata) #2
+  
+  ; Function Attrs: nounwind
+  declare void @llvm.stackprotector(i8*, i8**) #3
+  
+  attributes #0 = { nounwind "correctly-rounded-divide-sqrt-fp-math"="false" "disable-tail-calls"="false" "less-precise-fpmad"="false" "min-legal-vector-width"="0" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-jump-tables"="false" "no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false" "no-trapping-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="generic" "target-features"="+neon" "unsafe-fp-math"="false" "use-soft-float"="false" }
+  attributes #1 = { nounwind "correctly-rounded-divide-sqrt-fp-math"="false" "disable-tail-calls"="false" "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false" "no-trapping-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="generic" "target-features"="+neon" "unsafe-fp-math"="false" "use-soft-float"="false" }
+  attributes #2 = { nounwind readnone speculatable }
+  attributes #3 = { nounwind }
+  
+  !llvm.dbg.cu = !{!0}
+  !llvm.module.flags = !{!3, !4, !5}
+  !llvm.ident = !{!6}
+  
+  !0 = distinct !DICompileUnit(language: DW_LANG_C99, file: !1, producer: "clang version 9.0.0 ", isOptimized: true, runtimeVersion: 0, emissionKind: FullDebug, enums: !2, nameTableKind: None)
+  !1 = !DIFile(filename: "test.c", directory: "test")
+  !2 = !{}
+  !3 = !{i32 2, !"Dwarf Version", i32 4}
+  !4 = !{i32 2, !"Debug Info Version", i32 3}
+  !5 = !{i32 1, !"wchar_size", i32 4}
+  !6 = !{!"clang version 9.0.0 "}
+  !7 = distinct !DISubprogram(name: "main", scope: !1, file: !1, line: 6, type: !8, scopeLine: 7, flags: DIFlagPrototyped, spFlags: DISPFlagDefinition | DISPFlagOptimized, unit: !0, retainedNodes: !14)
+  !8 = !DISubroutineType(types: !9)
+  !9 = !{!10, !10, !11}
+  !10 = !DIBasicType(name: "int", size: 32, encoding: DW_ATE_signed)
+  !11 = !DIDerivedType(tag: DW_TAG_pointer_type, baseType: !12, size: 64)
+  !12 = !DIDerivedType(tag: DW_TAG_pointer_type, baseType: !13, size: 64)
+  !13 = !DIBasicType(name: "char", size: 8, encoding: DW_ATE_unsigned_char)
+  !14 = !{!15, !16, !17}
+  !15 = !DILocalVariable(name: "argc", arg: 1, scope: !7, file: !1, line: 6, type: !10)
+  !16 = !DILocalVariable(name: "argv", arg: 2, scope: !7, file: !1, line: 6, type: !11)
+  !17 = !DILocalVariable(name: "i", scope: !18, file: !1, line: 8, type: !10)
+  !18 = distinct !DILexicalBlock(scope: !7, file: !1, line: 8, column: 3)
+  !19 = !DILocation(line: 0, scope: !7)
+  !20 = !DILocation(line: 0, scope: !18)
+  !21 = !DILocation(line: 8, column: 20, scope: !22)
+  !22 = distinct !DILexicalBlock(scope: !18, file: !1, line: 8, column: 3)
+  !23 = !DILocation(line: 8, column: 3, scope: !18)
+  !24 = !DILocation(line: 0, scope: !22)
+  !25 = !DILocation(line: 9, column: 36, scope: !26)
+  !26 = distinct !DILexicalBlock(scope: !22, file: !1, line: 8, column: 33)
+  !27 = !DILocation(line: 11, column: 3, scope: !7)
+  !28 = !{!29, !29, i64 0}
+  !29 = !{!"any pointer", !30, i64 0}
+  !30 = !{!"omnipotent char", !31, i64 0}
+  !31 = !{!"Simple C/C++ TBAA"}
+  !32 = !DILocation(line: 9, column: 5, scope: !26)
+  !33 = !DILocation(line: 8, column: 28, scope: !22)
+  !34 = distinct !{!34, !23, !35}
+  !35 = !DILocation(line: 10, column: 3, scope: !18)
+
+...
+---
+name:            main
+alignment:       2
+exposesReturnsTwice: false
+legalized:       false
+regBankSelected: false
+selected:        false
+failedISel:      false
+tracksRegLiveness: true
+hasWinCFI:       false
+registers:       []
+liveins:         
+  - { reg: '$w0', virtual-reg: '' }
+  - { reg: '$x1', virtual-reg: '' }
+frameInfo:       
+  isFrameAddressTaken: false
+  isReturnAddressTaken: false
+  hasStackMap:     false
+  hasPatchPoint:   false
+  stackSize:       48
+  offsetAdjustment: 0
+  maxAlignment:    8
+  adjustsStack:    true
+  hasCalls:        true
+  stackProtector:  ''
+  maxCallFrameSize: 0
+  cvBytesOfCalleeSavedRegisters: 0
+  hasOpaqueSPAdjustment: false
+  hasVAStart:      false
+  hasMustTailInVarArgFunc: false
+  localFrameSize:  0
+  savePoint:       ''
+  restorePoint:    ''
+fixedStack:      []
+stack:           
+  - { id: 0, name: '', type: spill-slot, offset: -8, size: 8, alignment: 8, 
+      stack-id: 0, callee-saved-register: '$lr', callee-saved-restored: true, 
+      debug-info-variable: '', debug-info-expression: '', debug-info-location: '' }
+  - { id: 1, name: '', type: spill-slot, offset: -16, size: 8, alignment: 8, 
+      stack-id: 0, callee-saved-register: '$fp', callee-saved-restored: true, 
+      debug-info-variable: '', debug-info-expression: '', debug-info-location: '' }
+  - { id: 2, name: '', type: spill-slot, offset: -24, size: 8, alignment: 8, 
+      stack-id: 0, callee-saved-register: '$x19', callee-saved-restored: true, 
+      debug-info-variable: '', debug-info-expression: '', debug-info-location: '' }
+  - { id: 3, name: '', type: spill-slot, offset: -32, size: 8, alignment: 8, 
+      stack-id: 0, callee-saved-register: '$x20', callee-saved-restored: true, 
+      debug-info-variable: '', debug-info-expression: '', debug-info-location: '' }
+  - { id: 4, name: '', type: spill-slot, offset: -40, size: 8, alignment: 8, 
+      stack-id: 0, callee-saved-register: '$x21', callee-saved-restored: true, 
+      debug-info-variable: '', debug-info-expression: '', debug-info-location: '' }
+  - { id: 5, name: '', type: spill-slot, offset: -48, size: 8, alignment: 8, 
+      stack-id: 0, callee-saved-register: '$x22', callee-saved-restored: true, 
+      debug-info-variable: '', debug-info-expression: '', debug-info-location: '' }
+constants:       []
+machineFunctionInfo: {}
+body:             |
+  bb.0.entry:
+    successors: %bb.1(0x40000000), %bb.3(0x40000000)
+    liveins: $w0, $x1, $x21, $x22, $x19, $x20, $lr
+  
+    DBG_VALUE $w0, $noreg, !15, !DIExpression(), debug-location !19
+    DBG_VALUE $w0, $noreg, !15, !DIExpression(), debug-location !19
+    DBG_VALUE $x1, $noreg, !16, !DIExpression(), debug-location !19
+    early-clobber $sp = frame-setup STPXpre killed $x22, killed $x21, $sp, -6 :: (store 8 into %stack.5), (store 8 into %stack.4)
+    frame-setup STPXi killed $x20, killed $x19, $sp, 2 :: (store 8 into %stack.3), (store 8 into %stack.2)
+    frame-setup STPXi $fp, killed $lr, $sp, 4 :: (store 8 into %stack.1), (store 8 into %stack.0)
+    $fp = frame-setup ADDXri $sp, 32, 0
+    frame-setup CFI_INSTRUCTION def_cfa $w29, 16
+    frame-setup CFI_INSTRUCTION offset $w30, -8, debug-location !23
+    frame-setup CFI_INSTRUCTION offset $w29, -16, debug-location !23
+    frame-setup CFI_INSTRUCTION offset $w19, -24, debug-location !23
+    frame-setup CFI_INSTRUCTION offset $w20, -32, debug-location !23
+    frame-setup CFI_INSTRUCTION offset $w21, -40, debug-location !23
+    frame-setup CFI_INSTRUCTION offset $w22, -48, debug-location !23
+    DBG_VALUE 1, $noreg, !17, !DIExpression(), debug-location !20
+    dead $wzr = SUBSWri renamable $w0, 2, 0, implicit-def $nzcv, debug-location !23
+    Bcc 11, %bb.3, implicit $nzcv, debug-location !23
+  
+  bb.1.for.body.preheader:
+    successors: %bb.2(0x80000000)
+    liveins: $w0, $x1
+  
+    $x21 = ADRP target-flags(aarch64-page) @.str
+    $x19 = ORRXrs $xzr, killed $x1, 0
+    DBG_VALUE $x19, $noreg, !16, !DIExpression(), debug-location !19
+    renamable $w22 = ORRWrs $wzr, killed renamable $w0, 0, implicit-def $x22, debug-location !24
+    $w20 = MOVZWi 1, 0, implicit-def $x20
+    renamable $x21 = ADDXri killed $x21, target-flags(aarch64-pageoff, aarch64-nc) @.str, 0
+  
+  bb.2.for.body:
+    successors: %bb.3(0x04000000), %bb.2(0x7c000000)
+    liveins: $x19, $x20, $x21, $x22
+  
+    DBG_VALUE $x20, $noreg, !17, !DIExpression(), debug-location !20
+    renamable $x2 = LDRXroX renamable $x19, renamable $x20, 0, 1, debug-location !25 :: (load 8 from %ir.scevgep, !tbaa !28)
+    $x0 = ORRXrs $xzr, $x21, 0, debug-location !32
+    $w1 = ORRWrs $wzr, $w20, 0, debug-location !32
+    BL @printf, csr_aarch64_aapcs, implicit-def dead $lr, implicit $sp, implicit killed $x0, implicit killed $w1, implicit killed $x2, implicit-def $sp, implicit-def dead $w0, debug-location !32
+    renamable $x20 = nuw nsw ADDXri killed renamable $x20, 1, 0, debug-location !33
+    DBG_VALUE $noreg, $noreg, !17, !DIExpression(DW_OP_plus_uconst, 1, DW_OP_stack_value), debug-location !20
+    $xzr = SUBSXrs renamable $x22, renamable $x20, 0, implicit-def $nzcv, implicit-def $nzcv, debug-location !23
+    Bcc 1, %bb.2, implicit killed $nzcv, debug-location !23
+  
+  bb.3.for.cond.cleanup:
+    $fp, $lr = frame-destroy LDPXi $sp, 4, debug-location !27 :: (load 8 from %stack.1), (load 8 from %stack.0)
+    $x20, $x19 = frame-destroy LDPXi $sp, 2, debug-location !27 :: (load 8 from %stack.3), (load 8 from %stack.2)
+    $w0 = ORRWrs $wzr, $wzr, 0, debug-location !27
+    early-clobber $sp, $x22, $x21 = frame-destroy LDPXpost $sp, 6, debug-location !27 :: (load 8 from %stack.5), (load 8 from %stack.4)
+    RET undef $lr, implicit killed $w0, debug-location !27
+
+...