[PowerPC] Handle SUBFIC in reg+reg -> reg+imm transformation

We initially missed the subtract-immediate in this transformation.
This patch just adds that.

Differential revision: https://reviews.llvm.org/D84659
This commit is contained in:
Nemanja Ivanovic 2020-08-24 16:22:44 -05:00
parent 9e32d7b6e7
commit c485343c83
2 changed files with 23 additions and 13 deletions

View File

@ -3760,6 +3760,20 @@ bool PPCInstrInfo::simplifyToLI(MachineInstr &MI, MachineInstr &DefMI,
} }
return false; return false;
} }
case PPC::SUBFIC:
case PPC::SUBFIC8: {
// Only transform this if the CARRY implicit operand is dead.
if (MI.getNumOperands() > 3 && !MI.getOperand(3).isDead())
return false;
int64_t Minuend = MI.getOperand(2).getImm();
if (isInt<16>(Minuend - SExtImm)) {
ReplaceWithLI = true;
Is64BitLI = Opc == PPC::SUBFIC8;
NewImm = Minuend - SExtImm;
break;
}
return false;
}
case PPC::RLDICL: case PPC::RLDICL:
case PPC::RLDICL_rec: case PPC::RLDICL_rec:
case PPC::RLDICL_32: case PPC::RLDICL_32:

View File

@ -8,37 +8,33 @@ define void @_ZN1m1nEv(%struct.m.2.5.8.11* %this) local_unnamed_addr nounwind al
; CHECK-LABEL: _ZN1m1nEv: ; CHECK-LABEL: _ZN1m1nEv:
; CHECK: # %bb.0: # %entry ; CHECK: # %bb.0: # %entry
; CHECK-NEXT: mflr r0 ; CHECK-NEXT: mflr r0
; CHECK-NEXT: std r29, -24(r1) # 8-byte Folded Spill
; CHECK-NEXT: std r30, -16(r1) # 8-byte Folded Spill ; CHECK-NEXT: std r30, -16(r1) # 8-byte Folded Spill
; CHECK-NEXT: std r0, 16(r1) ; CHECK-NEXT: std r0, 16(r1)
; CHECK-NEXT: stdu r1, -64(r1) ; CHECK-NEXT: stdu r1, -48(r1)
; CHECK-NEXT: mr r30, r3 ; CHECK-NEXT: mr r30, r3
; CHECK-NEXT: li r3, 4
; CHECK-NEXT: ld r4, 16(r30) ; CHECK-NEXT: ld r4, 16(r30)
; CHECK-NEXT: ld r5, 8(r30) ; CHECK-NEXT: ld r5, 8(r30)
; CHECK-NEXT: subfic r29, r3, 64 ; CHECK-NEXT: lwz r6, 36(r30)
; CHECK-NEXT: rldicl r3, r5, 60, 4 ; CHECK-NEXT: rldicl r5, r5, 60, 4
; CHECK-NEXT: sld r4, r4, r29 ; CHECK-NEXT: sldi r4, r4, 60
; CHECK-NEXT: lwz r5, 36(r30) ; CHECK-NEXT: or r4, r4, r5
; CHECK-NEXT: or r3, r4, r3 ; CHECK-NEXT: rlwinm r3, r4, 31, 0, 0
; CHECK-NEXT: rlwinm r3, r3, 31, 0, 0 ; CHECK-NEXT: clrlwi r4, r6, 31
; CHECK-NEXT: clrlwi r4, r5, 31
; CHECK-NEXT: or r4, r4, r3 ; CHECK-NEXT: or r4, r4, r3
; CHECK-NEXT: bl _ZN1llsE1d ; CHECK-NEXT: bl _ZN1llsE1d
; CHECK-NEXT: nop ; CHECK-NEXT: nop
; CHECK-NEXT: ld r3, 16(r30) ; CHECK-NEXT: ld r3, 16(r30)
; CHECK-NEXT: ld r4, 8(r30) ; CHECK-NEXT: ld r4, 8(r30)
; CHECK-NEXT: rldicl r4, r4, 60, 4 ; CHECK-NEXT: rldicl r4, r4, 60, 4
; CHECK-NEXT: sld r3, r3, r29 ; CHECK-NEXT: sldi r3, r3, 60
; CHECK-NEXT: or r3, r3, r4 ; CHECK-NEXT: or r3, r3, r4
; CHECK-NEXT: sldi r3, r3, 31 ; CHECK-NEXT: sldi r3, r3, 31
; CHECK-NEXT: clrldi r4, r3, 32 ; CHECK-NEXT: clrldi r4, r3, 32
; CHECK-NEXT: bl _ZN1llsE1d ; CHECK-NEXT: bl _ZN1llsE1d
; CHECK-NEXT: nop ; CHECK-NEXT: nop
; CHECK-NEXT: addi r1, r1, 64 ; CHECK-NEXT: addi r1, r1, 48
; CHECK-NEXT: ld r0, 16(r1) ; CHECK-NEXT: ld r0, 16(r1)
; CHECK-NEXT: ld r30, -16(r1) # 8-byte Folded Reload ; CHECK-NEXT: ld r30, -16(r1) # 8-byte Folded Reload
; CHECK-NEXT: ld r29, -24(r1) # 8-byte Folded Reload
; CHECK-NEXT: mtlr r0 ; CHECK-NEXT: mtlr r0
; CHECK-NEXT: blr ; CHECK-NEXT: blr
entry: entry: