[GISel] Add combiners for G_INTTOPTR and G_PTRTOINT

[mkitzan]

Summary

Patch adds two new GICombinerRules, one for G_INTTOPTR and one for G_PTRTOINT. The G_INTTOPTR elides ptr2int(int2ptr(x)) to a copy of x, if the cast is within the same address space. The G_PTRTOINT elides int2ptr(ptr2int(x)) to a copy of x. Patch additionally adds new combiner tests for the AArch64 target to test these new combiner rules.

[paquette]

llvm/include/llvm/Target/GlobalISel/Combine.td

293

Maybe put these under identity_combines?

[mkitzan]

Moved the two new combiner rules to identity_combines.

[paquette]

LGTM

[mkitzan]

Great, thanks for the review! I don't have commit access. Would be great if someone could commit on my behalf.

[aditya_nandakumar]

I can commit it for you.

[mkitzan]

Thanks, Aditya!

[aditya_nandakumar]

Committed in 2144a3bdbba4

llvm/include/llvm/CodeGen/GlobalISel/CombinerHelper.h

public:
 
   /// Reduce a shift by a constant to an unmerge and a shift on a half sized
   /// type. This will not produce a shift smaller than \p TargetShiftSize.
   bool matchCombineShiftToUnmerge(MachineInstr &MI, unsigned TargetShiftSize,
                                  unsigned &ShiftVal);
   bool applyCombineShiftToUnmerge(MachineInstr &MI, const unsigned &ShiftVal);
   bool tryCombineShiftToUnmerge(MachineInstr &MI, unsigned TargetShiftAmount);
 
+  /// Transform IntToPtr(PtrToInt(x)) to x if cast is in the same address space.
+  bool matchCombineI2PToP2I(MachineInstr &MI, Register &Reg);
+  bool applyCombineI2PToP2I(MachineInstr &MI, Register &Reg);
+
+  /// Transform PtrToInt(IntToPtr(x)) to x.
+  bool matchCombineP2IToI2P(MachineInstr &MI, Register &Reg);
+  bool applyCombineP2IToI2P(MachineInstr &MI, Register &Reg);
+
   /// Return true if any explicit use operand on \p MI is defined by a
   /// G_IMPLICIT_DEF.
   bool matchAnyExplicitUseIsUndef(MachineInstr &MI);
 
   /// Return true if all register explicit use operands on \p MI are defined by
   /// a G_IMPLICIT_DEF.
   bool matchAllExplicitUsesAreUndef(MachineInstr &MI);
 

llvm/include/llvm/Target/GlobalISel/Combine.td

 def simplify_add_to_sub_matchinfo: GIDefMatchData<"std::tuple<Register, Register>">;
 def simplify_add_to_sub: GICombineRule <
   (defs root:$root, simplify_add_to_sub_matchinfo:$info),
   (match (wip_match_opcode G_ADD):$root,
     [{ return Helper.matchSimplifyAddToSub(*${root}, ${info}); }]),
   (apply [{ return Helper.applySimplifyAddToSub(*${root}, ${info});}])
 >;
 
+// Fold int2ptr(ptr2int(x)) -> x
+def p2i_to_i2p_matchinfo: GIDefMatchData<"Register">;
+def p2i_to_i2p: GICombineRule<
+  (defs root:$root, p2i_to_i2p_matchinfo:$info),
+  (match (wip_match_opcode G_INTTOPTR):$root,
+    [{ return Helper.matchCombineI2PToP2I(*${root}, ${info}); }]),
+  (apply [{ return Helper.applyCombineI2PToP2I(*${root}, ${info}); }])
+>;
+
+// Fold ptr2int(int2ptr(x)) -> x
+def i2p_to_p2i_matchinfo: GIDefMatchData<"Register">;
+def i2p_to_p2i: GICombineRule<
+  (defs root:$root, i2p_to_p2i_matchinfo:$info),
+  (match (wip_match_opcode G_PTRTOINT):$root,
+    [{ return Helper.matchCombineP2IToI2P(*${root}, ${info}); }]),
+  (apply [{ return Helper.applyCombineP2IToI2P(*${root}, ${info}); }])
+>;
+
 // FIXME: These should use the custom predicate feature once it lands.
 def undef_combines : GICombineGroup<[undef_to_fp_zero, undef_to_int_zero,
                                      undef_to_negative_one,
                                      propagate_undef_any_op,
                                      propagate_undef_all_ops,
                                      propagate_undef_shuffle_mask,
                                      erase_undef_store]>;
 
 def identity_combines : GICombineGroup<[select_same_val, right_identity_zero,
                                         binop_same_val, binop_left_to_zero,
                                         binop_right_to_zero]>;
 
 def trivial_combines : GICombineGroup<[copy_prop, mul_to_shl]>;
 def all_combines : GICombineGroup<[trivial_combines, ptr_add_immed_chain,
     combines_for_extload, combine_indexed_load_store, undef_combines,
-    identity_combines, simplify_add_to_sub]>;
+    identity_combines, simplify_add_to_sub, p2i_to_i2p, i2p_to_p2i]>;

llvm/lib/CodeGen/GlobalISel/CombinerHelper.cpp

bool CombinerHelper::tryCombineShiftToUnmerge(MachineInstr &MI,
   if (matchCombineShiftToUnmerge(MI, TargetShiftAmount, ShiftAmt)) {
     applyCombineShiftToUnmerge(MI, ShiftAmt);
     return true;
   }
 
   return false;
 }
 
+bool CombinerHelper::matchCombineI2PToP2I(MachineInstr &MI, Register &Reg) {
+  assert(MI.getOpcode() == TargetOpcode::G_INTTOPTR && "Expected a G_INTTOPTR");
+  Register DstReg = MI.getOperand(0).getReg();
+  LLT DstTy = MRI.getType(DstReg);
+  Register SrcReg = MI.getOperand(1).getReg();
+  return mi_match(SrcReg, MRI,
+                  m_GPtrToInt(m_all_of(m_SpecificType(DstTy), m_Reg(Reg))));
+}
+
+bool CombinerHelper::applyCombineI2PToP2I(MachineInstr &MI, Register &Reg) {
+  assert(MI.getOpcode() == TargetOpcode::G_INTTOPTR && "Expected a G_INTTOPTR");
+  Register DstReg = MI.getOperand(0).getReg();
+  Builder.setInstr(MI);
+  Builder.buildCopy(DstReg, Reg);
+  MI.eraseFromParent();
+  return true;
+}
+
+bool CombinerHelper::matchCombineP2IToI2P(MachineInstr &MI, Register &Reg) {
+  assert(MI.getOpcode() == TargetOpcode::G_PTRTOINT && "Expected a G_PTRTOINT");
+  Register SrcReg = MI.getOperand(1).getReg();
+  return mi_match(SrcReg, MRI, m_GIntToPtr(m_Reg(Reg)));
+}
+
+bool CombinerHelper::applyCombineP2IToI2P(MachineInstr &MI, Register &Reg) {
+  assert(MI.getOpcode() == TargetOpcode::G_PTRTOINT && "Expected a G_PTRTOINT");
+  Register DstReg = MI.getOperand(0).getReg();
+  Builder.setInstr(MI);
+  Builder.buildZExtOrTrunc(DstReg, Reg);
+  MI.eraseFromParent();
+  return true;
+}
+
 bool CombinerHelper::matchAnyExplicitUseIsUndef(MachineInstr &MI) {
   return any_of(MI.explicit_uses(), [this](const MachineOperand &MO) {
     return MO.isReg() &&
            getOpcodeDef(TargetOpcode::G_IMPLICIT_DEF, MO.getReg(), MRI);
   });
 }
 
 bool CombinerHelper::matchAllExplicitUsesAreUndef(MachineInstr &MI) {

llvm/test/CodeGen/AArch64/GlobalISel/combine-inttoptr-ptrtoint.mir

+# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
+# RUN: llc -run-pass=aarch64-prelegalizer-combiner -verify-machineinstrs -mtriple aarch64-unknown-unknown %s -o - | FileCheck %s
+# RUN: llc -debugify-and-strip-all-safe -run-pass=aarch64-prelegalizer-combiner -verify-machineinstrs -mtriple aarch64-unknown-unknown %s -o - | FileCheck %s
+---
+name:            test_combine_inttoptr_same_addressspace
+body:             |
+  bb.1:
+  liveins: $x0, $x1
+
+    ; CHECK-LABEL: name: test_combine_inttoptr_same_addressspace
+    ; CHECK: [[COPY:%[0-9]+]]:_(p64) = COPY $x0
+    ; CHECK: $x1 = COPY [[COPY]](p64)
+    %0:_(p64) = COPY $x0
+    %1:_(s64) = G_PTRTOINT %0
+    %2:_(p64) = G_INTTOPTR %1
+    $x1 = COPY %2
+...
+---
+name:            test_combine_inttoptr_diff_addressspace
+body:             |
+  bb.1:
+  liveins: $x0, $x1
+
+    ; CHECK-LABEL: name: test_combine_inttoptr_diff_addressspace
+    ; CHECK: [[COPY:%[0-9]+]]:_(p0) = COPY $x0
+    ; CHECK: [[PTRTOINT:%[0-9]+]]:_(s64) = G_PTRTOINT [[COPY]](p0)
+    ; CHECK: [[INTTOPTR:%[0-9]+]]:_(p64) = G_INTTOPTR [[PTRTOINT]](s64)
+    ; CHECK: $x1 = COPY [[INTTOPTR]](p64)
+    %0:_(p0) = COPY $x0
+    %1:_(s64) = G_PTRTOINT %0
+    %2:_(p64) = G_INTTOPTR %1
+    $x1 = COPY %2
+...

llvm/test/CodeGen/AArch64/GlobalISel/combine-ptrtoint.mir

+# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
+# RUN: llc -run-pass=aarch64-prelegalizer-combiner -verify-machineinstrs -mtriple aarch64-unknown-unknown %s -o - | FileCheck %s
+# RUN: llc -debugify-and-strip-all-safe -run-pass=aarch64-prelegalizer-combiner -verify-machineinstrs -mtriple aarch64-unknown-unknown %s -o - | FileCheck %s
+---
+name:            test_combine_ptrtoint
+body:             |
+  bb.1:
+  liveins: $x0, $x1
+
+    ; CHECK-LABEL: name: test_combine_ptrtoint
+    ; CHECK: [[COPY:%[0-9]+]]:_(s64) = COPY $x0
+    ; CHECK: $x1 = COPY [[COPY]](s64)
+    %0:_(s64) = COPY $x0
+    %1:_(p64) = G_INTTOPTR %0
+    %2:_(s64) = G_PTRTOINT %1
+    $x1 = COPY %2
+...