[LoongArch] Eliminate the redundant sign extension of division (#107971)

If all incoming values of `div.d` are sign-extended and all users only
use the lower 32 bits, then convert them to W versions.

Fixes: #107946
(cherry picked from commit 0f47e3aebdd2a4a938468a272ea4224552dbf176)

diff --git a/llvm/lib/Target/LoongArch/LoongArchOptWInstrs.cpp b/llvm/lib/Target/LoongArch/LoongArchOptWInstrs.cpp
index abac69054f3b9178dfc70fc59a31835ecaf7de3b..ab90409fdf47d0bff37e1dde60448f61befcf71d 100644 (file)
--- a/llvm/lib/Target/LoongArch/LoongArchOptWInstrs.cpp
+++ b/llvm/lib/Target/LoongArch/LoongArchOptWInstrs.cpp
@@ -637,6 +637,19 @@ static bool isSignExtendedW(Register SrcReg, const LoongArchSubtarget &ST,
         break;
       }
       return false;
+    // If all incoming values are sign-extended and all users only use
+    // the lower 32 bits, then convert them to W versions.
+    case LoongArch::DIV_D: {
+      if (!AddRegToWorkList(MI->getOperand(1).getReg()))
+        return false;
+      if (!AddRegToWorkList(MI->getOperand(2).getReg()))
+        return false;
+      if (hasAllWUsers(*MI, ST, MRI)) {
+        FixableDef.insert(MI);
+        break;
+      }
+      return false;
+    }
     }
   }
 
@@ -651,6 +664,8 @@ static unsigned getWOp(unsigned Opcode) {
     return LoongArch::ADDI_W;
   case LoongArch::ADD_D:
     return LoongArch::ADD_W;
+  case LoongArch::DIV_D:
+    return LoongArch::DIV_W;
   case LoongArch::LD_D:
   case LoongArch::LD_WU:
     return LoongArch::LD_W;

diff --git a/llvm/test/CodeGen/LoongArch/ir-instruction/sdiv-udiv-srem-urem.ll b/llvm/test/CodeGen/LoongArch/ir-instruction/sdiv-udiv-srem-urem.ll
index c22acdb49690715f9e29e489c5e18e13b3653e93..c5af79157eaadc1b51f53244bb046ecf7c2b8bbd 100644 (file)
--- a/llvm/test/CodeGen/LoongArch/ir-instruction/sdiv-udiv-srem-urem.ll
+++ b/llvm/test/CodeGen/LoongArch/ir-instruction/sdiv-udiv-srem-urem.ll
@@ -191,8 +191,7 @@ define signext i32 @sdiv_si32_ui32_ui32(i32 %a, i32 %b) {
 ; LA64:       # %bb.0: # %entry
 ; LA64-NEXT:    addi.w $a1, $a1, 0
 ; LA64-NEXT:    addi.w $a0, $a0, 0
-; LA64-NEXT:    div.d $a0, $a0, $a1
-; LA64-NEXT:    addi.w $a0, $a0, 0
+; LA64-NEXT:    div.w $a0, $a0, $a1
 ; LA64-NEXT:    ret
 ;
 ; LA32-TRAP-LABEL: sdiv_si32_ui32_ui32:
@@ -208,12 +207,11 @@ define signext i32 @sdiv_si32_ui32_ui32(i32 %a, i32 %b) {
 ; LA64-TRAP:       # %bb.0: # %entry
 ; LA64-TRAP-NEXT:    addi.w $a1, $a1, 0
 ; LA64-TRAP-NEXT:    addi.w $a0, $a0, 0
-; LA64-TRAP-NEXT:    div.d $a0, $a0, $a1
+; LA64-TRAP-NEXT:    div.w $a0, $a0, $a1
 ; LA64-TRAP-NEXT:    bnez $a1, .LBB5_2
 ; LA64-TRAP-NEXT:  # %bb.1: # %entry
 ; LA64-TRAP-NEXT:    break 7
 ; LA64-TRAP-NEXT:  .LBB5_2: # %entry
-; LA64-TRAP-NEXT:    addi.w $a0, $a0, 0
 ; LA64-TRAP-NEXT:    ret
 entry:
   %r = sdiv i32 %a, %b