vendor/llvm/llvm-trunk-r301441

3 files changed, 161 insertions, 12 deletions

diff --git a/test/Transforms/LoopVectorize/X86/float-induction-x86.ll b/test/Transforms/LoopVectorize/X86/float-induction-x86.ll
index 08ac64e52d4fe..b5e914500fb4a 100644
--- a/test/Transforms/LoopVectorize/X86/float-induction-x86.ll
+++ b/test/Transforms/LoopVectorize/X86/float-induction-x86.ll
@@ -82,5 +82,68 @@ for.end:                                          ; preds = %for.end.loopexit, %
   ret void
 }
 
+; AUTO_VEC-LABEL: @external_use_with_fast_math(
+; AUTO_VEC-NEXT:  entry:
+; AUTO_VEC-NEXT:    [[TMP0:%.*]] = icmp sgt i64 %n, 1
+; AUTO_VEC-NEXT:    [[SMAX:%.*]] = select i1 [[TMP0]], i64 %n, i64 1
+; AUTO_VEC:         br i1 {{.*}}, label %for.body, label %min.iters.checked
+; AUTO_VEC:       min.iters.checked:
+; AUTO_VEC-NEXT:    [[N_VEC:%.*]] = and i64 [[SMAX]], 9223372036854775792
+; AUTO_VEC:         br i1 {{.*}}, label %for.body, label %vector.body
+; AUTO_VEC:       middle.block:
+; AUTO_VEC:         [[TMP11:%.*]] = add nsw i64 [[N_VEC]], -1
+; AUTO_VEC-NEXT:    [[CAST_CMO:%.*]] = sitofp i64 [[TMP11]] to double
+; AUTO_VEC-NEXT:    [[TMP12:%.*]] = fmul fast double [[CAST_CMO]], 3.000000e+00
+; AUTO_VEC-NEXT:    br i1 {{.*}}, label %for.end, label %for.body
+; AUTO_VEC:       for.end:
+; AUTO_VEC-NEXT:    [[J_LCSSA:%.*]] = phi double [ [[TMP12]], %middle.block ], [ %j, %for.body ]
+; AUTO_VEC-NEXT:    ret double [[J_LCSSA]]
+;
+define double @external_use_with_fast_math(double* %a, i64 %n) {
+entry:
+  br label %for.body
+
+for.body:
+  %i = phi i64 [ 0, %entry ], [%i.next, %for.body]
+  %j = phi double [ 0.0, %entry ], [ %j.next, %for.body ]
+  %tmp0 = getelementptr double, double* %a, i64 %i
+  store double %j, double* %tmp0
+  %i.next = add i64 %i, 1
+  %j.next = fadd fast double %j, 3.0
+  %cond = icmp slt i64 %i.next, %n
+  br i1 %cond, label %for.body, label %for.end
+
+for.end:
+  %tmp1 = phi double [ %j, %for.body ]
+  ret double %tmp1
+}
+
+; AUTO_VEC-LABEL: @external_use_without_fast_math(
+; AUTO_VEC:       for.body:
+; AUTO_VEC:         [[J:%.*]] = phi double [ 0.000000e+00, %entry ], [ [[J_NEXT:%.*]], %for.body ]
+; AUTO_VEC:         [[J_NEXT]] = fadd double [[J]], 3.000000e+00
+; AUTO_VEC:         br i1 {{.*}}, label %for.body, label %for.end
+; AUTO_VEC:       for.end:
+; AUTO_VEC-NEXT:    ret double [[J]]
+;
+define double @external_use_without_fast_math(double* %a, i64 %n) {
+entry:
+  br label %for.body
+
+for.body:
+  %i = phi i64 [ 0, %entry ], [%i.next, %for.body]
+  %j = phi double [ 0.0, %entry ], [ %j.next, %for.body ]
+  %tmp0 = getelementptr double, double* %a, i64 %i
+  store double %j, double* %tmp0
+  %i.next = add i64 %i, 1
+  %j.next = fadd double %j, 3.0
+  %cond = icmp slt i64 %i.next, %n
+  br i1 %cond, label %for.body, label %for.end
+
+for.end:
+  %tmp1 = phi double [ %j, %for.body ]
+  ret double %tmp1
+}
+
 attributes #0 = { "no-nans-fp-math"="true" }
 attributes #1 = { "no-nans-fp-math"="false" }
diff --git a/test/Transforms/LoopVectorize/induction.ll b/test/Transforms/LoopVectorize/induction.ll
index 0d7d9fe0c1b8a..6507166dd1f25 100644
--- a/test/Transforms/LoopVectorize/induction.ll
+++ b/test/Transforms/LoopVectorize/induction.ll
@@ -309,59 +309,59 @@ for.end:
 ;
 ; CHECK-LABEL: @scalarize_induction_variable_05(
 ; CHECK: vector.body:
-; CHECK:   %index = phi i32 [ 0, %vector.ph ], [ %index.next, %pred.udiv.continue2 ]
+; CHECK:   %index = phi i32 [ 0, %vector.ph ], [ %index.next, %pred.udiv.continue{{[0-9]+}} ]
 ; CHECK:   %[[I0:.+]] = add i32 %index, 0
 ; CHECK:   getelementptr inbounds i32, i32* %a, i32 %[[I0]]
 ; CHECK: pred.udiv.if:
 ; CHECK:   udiv i32 {{.*}}, %[[I0]]
-; CHECK: pred.udiv.if1:
+; CHECK: pred.udiv.if{{[0-9]+}}:
 ; CHECK:   %[[I1:.+]] = add i32 %index, 1
 ; CHECK:   udiv i32 {{.*}}, %[[I1]]
 ;
 ; UNROLL-NO_IC-LABEL: @scalarize_induction_variable_05(
 ; UNROLL-NO-IC: vector.body:
-; UNROLL-NO-IC:   %index = phi i32 [ 0, %vector.ph ], [ %index.next, %pred.udiv.continue11 ]
+; UNROLL-NO-IC:   %index = phi i32 [ 0, %vector.ph ], [ %index.next, %pred.udiv.continue{{[0-9]+}} ]
 ; UNROLL-NO-IC:   %[[I0:.+]] = add i32 %index, 0
 ; UNROLL-NO-IC:   %[[I2:.+]] = add i32 %index, 2
 ; UNROLL-NO-IC:   getelementptr inbounds i32, i32* %a, i32 %[[I0]]
 ; UNROLL-NO-IC:   getelementptr inbounds i32, i32* %a, i32 %[[I2]]
 ; UNROLL-NO-IC: pred.udiv.if:
 ; UNROLL-NO-IC:   udiv i32 {{.*}}, %[[I0]]
-; UNROLL-NO-IC: pred.udiv.if6:
+; UNROLL-NO-IC: pred.udiv.if{{[0-9]+}}:
 ; UNROLL-NO-IC:   %[[I1:.+]] = add i32 %index, 1
 ; UNROLL-NO-IC:   udiv i32 {{.*}}, %[[I1]]
-; UNROLL-NO-IC: pred.udiv.if8:
+; UNROLL-NO-IC: pred.udiv.if{{[0-9]+}}:
 ; UNROLL-NO-IC:   udiv i32 {{.*}}, %[[I2]]
-; UNROLL-NO-IC: pred.udiv.if10:
+; UNROLL-NO-IC: pred.udiv.if{{[0-9]+}}:
 ; UNROLL-NO-IC:   %[[I3:.+]] = add i32 %index, 3
 ; UNROLL-NO-IC:   udiv i32 {{.*}}, %[[I3]]
 ;
 ; IND-LABEL: @scalarize_induction_variable_05(
 ; IND: vector.body:
-; IND:   %index = phi i32 [ 0, %vector.ph ], [ %index.next, %pred.udiv.continue2 ]
+; IND:   %index = phi i32 [ 0, %vector.ph ], [ %index.next, %pred.udiv.continue{{[0-9]+}} ]
 ; IND:   %[[E0:.+]] = sext i32 %index to i64
 ; IND:   getelementptr inbounds i32, i32* %a, i64 %[[E0]]
 ; IND: pred.udiv.if:
 ; IND:   udiv i32 {{.*}}, %index
-; IND: pred.udiv.if1:
+; IND: pred.udiv.if{{[0-9]+}}:
 ; IND:   %[[I1:.+]] = or i32 %index, 1
 ; IND:   udiv i32 {{.*}}, %[[I1]]
 ;
 ; UNROLL-LABEL: @scalarize_induction_variable_05(
 ; UNROLL: vector.body:
-; UNROLL:   %index = phi i32 [ 0, %vector.ph ], [ %index.next, %pred.udiv.continue11 ]
+; UNROLL:   %index = phi i32 [ 0, %vector.ph ], [ %index.next, %pred.udiv.continue{{[0-9]+}} ]
 ; UNROLL:   %[[I2:.+]] = or i32 %index, 2
 ; UNROLL:   %[[E0:.+]] = sext i32 %index to i64
 ; UNROLL:   %[[G0:.+]] = getelementptr inbounds i32, i32* %a, i64 %[[E0]]
 ; UNROLL:   getelementptr i32, i32* %[[G0]], i64 2
 ; UNROLL: pred.udiv.if:
 ; UNROLL:   udiv i32 {{.*}}, %index
-; UNROLL: pred.udiv.if6:
+; UNROLL: pred.udiv.if{{[0-9]+}}:
 ; UNROLL:   %[[I1:.+]] = or i32 %index, 1
 ; UNROLL:   udiv i32 {{.*}}, %[[I1]]
-; UNROLL: pred.udiv.if8:
+; UNROLL: pred.udiv.if{{[0-9]+}}:
 ; UNROLL:   udiv i32 {{.*}}, %[[I2]]
-; UNROLL: pred.udiv.if10:
+; UNROLL: pred.udiv.if{{[0-9]+}}:
 ; UNROLL:   %[[I3:.+]] = or i32 %index, 3
 ; UNROLL:   udiv i32 {{.*}}, %[[I3]]
 
diff --git a/test/Transforms/LoopVectorize/phi-cost.ll b/test/Transforms/LoopVectorize/phi-cost.ll
new file mode 100644
index 0000000000000..5ccea66c76af4
--- /dev/null
+++ b/test/Transforms/LoopVectorize/phi-cost.ll
@@ -0,0 +1,86 @@
+; REQUIRES: asserts
+; RUN: opt < %s -loop-vectorize -force-vector-width=2 -force-vector-interleave=1 -instcombine -debug-only=loop-vectorize -disable-output -print-after=instcombine 2>&1 | FileCheck %s
+
+target datalayout = "e-m:e-i64:64-i128:128-n32:64-S128"
+
+; CHECK-LABEL: phi_two_incoming_values
+; CHECK:       LV: Found an estimated cost of 1 for VF 2 For instruction: %i = phi i64 [ %i.next, %if.end ], [ 0, %entry ]
+; CHECK:       LV: Found an estimated cost of 1 for VF 2 For instruction: %tmp5 = phi i32 [ %tmp1, %for.body ], [ %tmp4, %if.then ]
+; CHECK:       vector.body:
+; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, %vector.ph ], [ [[INDEX_NEXT:%.*]], %vector.body ]
+; CHECK:         [[WIDE_LOAD:%.*]] = load <2 x i32>, <2 x i32>* {{.*}}
+; CHECK:         [[TMP5:%.*]] = icmp sgt <2 x i32> [[WIDE_LOAD]], zeroinitializer
+; CHECK-NEXT:    [[TMP6:%.*]] = add <2 x i32> [[WIDE_LOAD]], <i32 1, i32 1>
+; CHECK-NEXT:    [[PREDPHI:%.*]] = select <2 x i1> [[TMP5]], <2 x i32> [[TMP6]], <2 x i32> [[WIDE_LOAD]]
+; CHECK:         store <2 x i32> [[PREDPHI]], <2 x i32>* {{.*}}
+; CHECK-NEXT:    [[INDEX_NEXT]] = add i64 [[INDEX]], 2
+;
+define void @phi_two_incoming_values(i32* %a, i32* %b, i64 %n) {
+entry:
+  br label %for.body
+
+for.body:
+  %i = phi i64 [ %i.next, %if.end ], [ 0, %entry ]
+  %tmp0 = getelementptr inbounds i32, i32* %a, i64 %i
+  %tmp1 = load i32, i32* %tmp0, align 4
+  %tmp2 = getelementptr inbounds i32, i32* %b, i64 %i
+  %tmp3 = icmp sgt i32 %tmp1, 0
+  br i1 %tmp3, label %if.then, label %if.end
+
+if.then:
+  %tmp4 = add i32 %tmp1, 1
+  br label %if.end
+
+if.end:
+  %tmp5 = phi i32 [ %tmp1, %for.body ], [ %tmp4, %if.then ]
+  store i32 %tmp5, i32* %tmp2, align 4
+  %i.next = add i64 %i, 1
+  %cond = icmp eq i64 %i, %n
+  br i1 %cond, label %for.end, label %for.body
+
+for.end:
+  ret void
+}
+
+; CHECK-LABEL: phi_three_incoming_values
+; CHECK:       LV: Found an estimated cost of 1 for VF 2 For instruction: %i = phi i64 [ %i.next, %if.end ], [ 0, %entry ]
+; CHECK:       LV: Found an estimated cost of 2 for VF 2 For instruction: %tmp8 = phi i32 [ 9, %for.body ], [ 3, %if.then ], [ %tmp7, %if.else ]
+; CHECK:       vector.body:
+; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, %vector.ph ], [ [[INDEX_NEXT:%.*]], %vector.body ]
+; CHECK:         [[PREDPHI:%.*]] = select <2 x i1> {{.*}}, <2 x i32> <i32 3, i32 3>, <2 x i32> <i32 9, i32 9>
+; CHECK:         [[PREDPHI7:%.*]] = select <2 x i1> {{.*}}, <2 x i32> {{.*}}, <2 x i32> [[PREDPHI]]
+; CHECK:         store <2 x i32> [[PREDPHI7]], <2 x i32>* {{.*}}
+; CHECK-NEXT:    [[INDEX_NEXT]] = add i64 [[INDEX]], 2
+;
+define void @phi_three_incoming_values(i32* %a, i32* %b, i64 %n) {
+entry:
+  br label %for.body
+
+for.body:
+  %i = phi i64 [ %i.next, %if.end ], [ 0, %entry ]
+  %tmp0 = getelementptr inbounds i32, i32* %a, i64 %i
+  %tmp1 = load i32, i32* %tmp0, align 4
+  %tmp2 = getelementptr inbounds i32, i32* %b, i64 %i
+  %tmp3 = load i32, i32* %tmp2, align 4
+  %tmp4 = icmp sgt i32 %tmp1, %tmp3
+  br i1 %tmp4, label %if.then, label %if.end
+
+if.then:
+  %tmp5 = icmp sgt i32 %tmp1, 19
+  br i1 %tmp5, label %if.end, label %if.else
+
+if.else:
+  %tmp6 = icmp slt i32 %tmp3, 4
+  %tmp7 = select i1 %tmp6, i32 4, i32 5
+  br label %if.end
+
+if.end:
+  %tmp8 = phi i32 [ 9, %for.body ], [ 3, %if.then ], [ %tmp7, %if.else ]
+  store i32 %tmp8, i32* %tmp0, align 4
+  %i.next = add i64 %i, 1
+  %cond = icmp eq i64 %i, %n
+  br i1 %cond, label %for.end, label %for.body
+
+for.end:
+  ret void
+}