[PATCH] D28853: [SelectionDAG] Improve knownbits handling of UMIN/UMAX (PR31293)

[Simon Pilgrim via Phabricator via llvm-commits]

This revision was automatically updated to reflect the committed changes.
Closed by commit rL292528: [SelectionDAG] Improve knownbits handling of UMIN/UMAX (PR31293) (authored by RKSimon).

Changed prior to commit:
  https://reviews.llvm.org/D28853?vs=84818&id=85043#toc

Repository:
  rL LLVM

https://reviews.llvm.org/D28853

Files:
  llvm/trunk/lib/CodeGen/SelectionDAG/SelectionDAG.cpp
  llvm/trunk/test/CodeGen/X86/known-bits-vector.ll


Index: llvm/trunk/test/CodeGen/X86/known-bits-vector.ll
===================================================================
--- llvm/trunk/test/CodeGen/X86/known-bits-vector.ll
+++ llvm/trunk/test/CodeGen/X86/known-bits-vector.ll
@@ -481,22 +481,14 @@
 ; X32:       # BB#0:
 ; X32-NEXT:    vpminud {{\.LCPI.*}}, %xmm0, %xmm0
 ; X32-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,0,3,3]
-; X32-NEXT:    vpblendw {{.*#+}} xmm1 = xmm0[0],mem[1],xmm0[2],mem[3],xmm0[4],mem[5],xmm0[6],mem[7]
-; X32-NEXT:    vpsrld $16, %xmm0, %xmm0
-; X32-NEXT:    vpblendw {{.*#+}} xmm0 = xmm0[0],mem[1],xmm0[2],mem[3],xmm0[4],mem[5],xmm0[6],mem[7]
-; X32-NEXT:    vaddps {{\.LCPI.*}}, %xmm0, %xmm0
-; X32-NEXT:    vaddps %xmm0, %xmm1, %xmm0
+; X32-NEXT:    vcvtdq2ps %xmm0, %xmm0
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: knownbits_umin_shuffle_uitofp:
 ; X64:       # BB#0:
 ; X64-NEXT:    vpminud {{.*}}(%rip), %xmm0, %xmm0
 ; X64-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,0,3,3]
-; X64-NEXT:    vpblendw {{.*#+}} xmm1 = xmm0[0],mem[1],xmm0[2],mem[3],xmm0[4],mem[5],xmm0[6],mem[7]
-; X64-NEXT:    vpsrld $16, %xmm0, %xmm0
-; X64-NEXT:    vpblendw {{.*#+}} xmm0 = xmm0[0],mem[1],xmm0[2],mem[3],xmm0[4],mem[5],xmm0[6],mem[7]
-; X64-NEXT:    vaddps {{.*}}(%rip), %xmm0, %xmm0
-; X64-NEXT:    vaddps %xmm0, %xmm1, %xmm0
+; X64-NEXT:    vcvtdq2ps %xmm0, %xmm0
 ; X64-NEXT:    retq
   %1 = call <4 x i32> @llvm.x86.sse41.pminud(<4 x i32> %a0, <4 x i32> <i32 65535, i32 -1, i32 -1, i32 262143>)
   %2 = shufflevector <4 x i32> %1, <4 x i32> undef, <4 x i32> <i32 0, i32 0, i32 3, i32 3>
@@ -511,14 +503,12 @@
 ; X32:       # BB#0:
 ; X32-NEXT:    vpmaxud {{\.LCPI.*}}, %xmm0, %xmm0
 ; X32-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[1,1,2,2]
-; X32-NEXT:    vpsrad $31, %xmm0, %xmm0
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: knownbits_umax_shuffle_ashr:
 ; X64:       # BB#0:
 ; X64-NEXT:    vpmaxud {{.*}}(%rip), %xmm0, %xmm0
 ; X64-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[1,1,2,2]
-; X64-NEXT:    vpsrad $31, %xmm0, %xmm0
 ; X64-NEXT:    retq
   %1 = call <4 x i32> @llvm.x86.sse41.pmaxud(<4 x i32> %a0, <4 x i32> <i32 65535, i32 -1, i32 -1, i32 262143>)
   %2 = shufflevector <4 x i32> %1, <4 x i32> undef, <4 x i32> <i32 1, i32 1, i32 2, i32 2>
Index: llvm/trunk/lib/CodeGen/SelectionDAG/SelectionDAG.cpp
===================================================================
--- llvm/trunk/lib/CodeGen/SelectionDAG/SelectionDAG.cpp
+++ llvm/trunk/lib/CodeGen/SelectionDAG/SelectionDAG.cpp
@@ -2687,12 +2687,42 @@
     KnownOne = KnownOne2.byteSwap();
     break;
   }
-  case ISD::SMIN:
-  case ISD::SMAX:
-  case ISD::UMIN:
+  case ISD::UMIN: {
+    computeKnownBits(Op.getOperand(0), KnownZero, KnownOne, DemandedElts,
+                     Depth + 1);
+    computeKnownBits(Op.getOperand(1), KnownZero2, KnownOne2, DemandedElts,
+                     Depth + 1);
+
+    // UMIN - we know that the result will have the maximum of the
+    // known zero leading bits of the inputs.
+    unsigned LeadZero = KnownZero.countLeadingOnes();
+    LeadZero = std::max(LeadZero, KnownZero2.countLeadingOnes());
+
+    KnownZero &= KnownZero2;
+    KnownOne &= KnownOne2;
+    KnownZero |= APInt::getHighBitsSet(BitWidth, LeadZero);
+    break;
+  }
   case ISD::UMAX: {
     computeKnownBits(Op.getOperand(0), KnownZero, KnownOne, DemandedElts,
                      Depth + 1);
+    computeKnownBits(Op.getOperand(1), KnownZero2, KnownOne2, DemandedElts,
+                     Depth + 1);
+
+    // UMAX - we know that the result will have the maximum of the
+    // known one leading bits of the inputs.
+    unsigned LeadOne = KnownOne.countLeadingOnes();
+    LeadOne = std::max(LeadOne, KnownOne2.countLeadingOnes());
+
+    KnownZero &= KnownZero2;
+    KnownOne &= KnownOne2;
+    KnownOne |= APInt::getHighBitsSet(BitWidth, LeadOne);
+    break;
+  }
+  case ISD::SMIN:
+  case ISD::SMAX: {
+    computeKnownBits(Op.getOperand(0), KnownZero, KnownOne, DemandedElts,
+                     Depth + 1);
     // If we don't know any bits, early out.
     if (!KnownOne && !KnownZero)
       break;


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D28853.85043.patch
Type: text/x-patch
Size: 4049 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20170119/f9f8b5c8/attachment.bin>