[X86] Promote 16-bit CTTZ_ZERO_UNDEF to 32-bit variant

Related to PR50172.

Protects us against regressions after we will start doing cttz(zext(x)) -> zext(cttz(x)) transformation in the middle-end.

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D101662

diff --git a/llvm/lib/Target/X86/X86ISelLowering.cpp b/llvm/lib/Target/X86/X86ISelLowering.cpp
index b3d0900..691398b 100644 (file)
--- a/llvm/lib/Target/X86/X86ISelLowering.cpp
+++ b/llvm/lib/Target/X86/X86ISelLowering.cpp
@@ -364,8 +364,13 @@ X86TargetLowering::X86TargetLowering(const X86TargetMachine &TM,
   // encoding.
   setOperationPromotedToType(ISD::CTTZ           , MVT::i8   , MVT::i32);
   setOperationPromotedToType(ISD::CTTZ_ZERO_UNDEF, MVT::i8   , MVT::i32);
-  if (!Subtarget.hasBMI()) {
-    setOperationAction(ISD::CTTZ           , MVT::i16  , Custom);
+
+  if (Subtarget.hasBMI()) {
+    // Promote the i16 zero undef variant and force it on up to i32 when tzcnt
+    // is enabled.
+    setOperationPromotedToType(ISD::CTTZ_ZERO_UNDEF, MVT::i16, MVT::i32);
+  } else {
+    setOperationAction(ISD::CTTZ, MVT::i16, Custom);
     setOperationAction(ISD::CTTZ           , MVT::i32  , Custom);
     setOperationAction(ISD::CTTZ_ZERO_UNDEF, MVT::i16  , Legal);
     setOperationAction(ISD::CTTZ_ZERO_UNDEF, MVT::i32  , Legal);

diff --git a/llvm/test/CodeGen/X86/clz.ll b/llvm/test/CodeGen/X86/clz.ll
index f8d0336..0c12135 100644 (file)
--- a/llvm/test/CodeGen/X86/clz.ll
+++ b/llvm/test/CodeGen/X86/clz.ll
@@ -56,12 +56,14 @@ define i16 @cttz_i16(i16 %x)  {
 ;
 ; X86-CLZ-LABEL: cttz_i16:
 ; X86-CLZ:       # %bb.0:
-; X86-CLZ-NEXT:    tzcntw {{[0-9]+}}(%esp), %ax
+; X86-CLZ-NEXT:    tzcntl {{[0-9]+}}(%esp), %eax
+; X86-CLZ-NEXT:    # kill: def $ax killed $ax killed $eax
 ; X86-CLZ-NEXT:    retl
 ;
 ; X64-CLZ-LABEL: cttz_i16:
 ; X64-CLZ:       # %bb.0:
-; X64-CLZ-NEXT:    tzcntw %di, %ax
+; X64-CLZ-NEXT:    tzcntl %edi, %eax
+; X64-CLZ-NEXT:    # kill: def $ax killed $ax killed $eax
 ; X64-CLZ-NEXT:    retq
   %tmp = call i16 @llvm.cttz.i16( i16 %x, i1 true )
   ret i16 %tmp