[SystemZ] Improvement of emitMemMemWrapper()

[jonpa]

Summary

It was discovered that an extra register COPY remained when expanding a (var-len) memory operation with a loop and there were another use of the involved address register(s) afterwards.

A simple fix for this is to COPY the address registers before the loop and use that new vreg instead. This handles the test cases and also seems clearly beneficial on SPEC:

Spill|Reload   :    613173     613123    -50
Copies         :   1018500    1016022  -2478

It doesn't seem to make a difference to do this COPY in all cases, even though it is only useful in the case of a register loop with other uses of the register.

[uweigand]

This looks like a nice improvement. LGTM.

llvm/lib/Target/SystemZ/SystemZISelLowering.cpp

 
 //===----------------------------------------------------------------------===//
 // Custom insertion
 //===----------------------------------------------------------------------===//
 
 // Force base value Base into a register before MI.  Return the register.
 static Register forceReg(MachineInstr &MI, MachineOperand &Base,
                          const SystemZInstrInfo *TII) {
-  if (Base.isReg())
-    return Base.getReg();
-
   MachineBasicBlock *MBB = MI.getParent();
   MachineFunction &MF = *MBB->getParent();
   MachineRegisterInfo &MRI = MF.getRegInfo();
 
+  if (Base.isReg()) {
+    // Copy Base into a new virtual register to help register coalescing in
+    // cases with multiple uses.
+    Register Reg = MRI.createVirtualRegister(&SystemZ::ADDR64BitRegClass);
+    BuildMI(*MBB, MI, MI.getDebugLoc(), TII->get(SystemZ::COPY), Reg)
+      .add(Base);
+    return Reg;
+  }
+
   Register Reg = MRI.createVirtualRegister(&SystemZ::ADDR64BitRegClass);
   BuildMI(*MBB, MI, MI.getDebugLoc(), TII->get(SystemZ::LA), Reg)
       .add(Base)
       .addImm(0)
       .addReg(0);
   return Reg;
 }
 

llvm/test/CodeGen/SystemZ/memset-05.ll

 }
 
 ; Test that identical target instructions get reused.
 define void @fun2(i8* %Addr, i32 %Len) {
 ; CHECK-LABEL: fun2:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    llgfr %r1, %r3
 ; CHECK-NEXT:    aghi %r1, -1
-; CHECK-NEXT:    cgije %r1, -1, .LBB2_5
+; CHECK-NEXT:    cgije %r1, -1, .LBB2_4
 ; CHECK-NEXT:  # %bb.1:
 ; CHECK-NEXT:    srlg %r0, %r1, 8
 ; CHECK-NEXT:    lgr %r3, %r2
-; CHECK-NEXT:    cgije %r0, 0, .LBB2_4
-; CHECK-NEXT:  # %bb.2:
-; CHECK-NEXT:    lgr %r3, %r2
-; CHECK-NEXT:  .LBB2_3: # =>This Inner Loop Header: Depth=1
+; CHECK-NEXT:    cgije %r0, 0, .LBB2_3
+; CHECK-NEXT:  .LBB2_2: # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    xc 0(256,%r3), 0(%r3)
 ; CHECK-NEXT:    la %r3, 256(%r3)
-; CHECK-NEXT:    brctg %r0, .LBB2_3
-; CHECK-NEXT:  .LBB2_4:
+; CHECK-NEXT:    brctg %r0, .LBB2_2
+; CHECK-NEXT:  .LBB2_3:
 ; CHECK-NEXT:    exrl %r1, .Ltmp1
-; CHECK-NEXT:  .LBB2_5:
-; CHECK-NEXT:    cgije %r1, -1, .LBB2_10
-; CHECK-NEXT:  # %bb.6:
+; CHECK-NEXT:  .LBB2_4:
+; CHECK-NEXT:    cgije %r1, -1, .LBB2_8
+; CHECK-NEXT:  # %bb.5:
 ; CHECK-NEXT:    srlg %r0, %r1, 8
 ; CHECK-NEXT:    lgr %r3, %r2
-; CHECK-NEXT:    cgije %r0, 0, .LBB2_9
-; CHECK-NEXT:  # %bb.7:
-; CHECK-NEXT:    lgr %r3, %r2
-; CHECK-NEXT:  .LBB2_8: # =>This Inner Loop Header: Depth=1
+; CHECK-NEXT:    cgije %r0, 0, .LBB2_7
+; CHECK-NEXT:  .LBB2_6: # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    xc 0(256,%r3), 0(%r3)
 ; CHECK-NEXT:    la %r3, 256(%r3)
-; CHECK-NEXT:    brctg %r0, .LBB2_8
-; CHECK-NEXT:  .LBB2_9:
+; CHECK-NEXT:    brctg %r0, .LBB2_6
+; CHECK-NEXT:  .LBB2_7:
 ; CHECK-NEXT:    exrl %r1, .Ltmp1
-; CHECK-NEXT:  .LBB2_10:
+; CHECK-NEXT:  .LBB2_8:
 ; CHECK-NEXT:    cgibe %r1, -1, 0(%r14)
-; CHECK-NEXT:  .LBB2_11:
+; CHECK-NEXT:  .LBB2_9:
 ; CHECK-NEXT:    srlg %r0, %r1, 8
-; CHECK-NEXT:    cgije %r0, 0, .LBB2_13
-; CHECK-NEXT:  .LBB2_12: # =>This Inner Loop Header: Depth=1
+; CHECK-NEXT:    cgije %r0, 0, .LBB2_11
+; CHECK-NEXT:  .LBB2_10: # =>This Inner Loop Header: Depth=1
 ; CHECK-NEXT:    xc 0(256,%r2), 0(%r2)
 ; CHECK-NEXT:    la %r2, 256(%r2)
-; CHECK-NEXT:    brctg %r0, .LBB2_12
-; CHECK-NEXT:  .LBB2_13:
+; CHECK-NEXT:    brctg %r0, .LBB2_10
+; CHECK-NEXT:  .LBB2_11:
 ; CHECK-NEXT:    exrl %r1, .Ltmp0
 ; CHECK-NEXT:    br %r14
   tail call void @llvm.memset.p0i8.i32(i8* %Addr, i8 0, i32 %Len, i1 false)
   tail call void @llvm.memset.p0i8.i32(i8* %Addr, i8 0, i32 %Len, i1 false)
   tail call void @llvm.memset.p0i8.i32(i8* %Addr, i8 0, i32 %Len, i1 false)
   ret void
 }