llvm/test/Transforms/LoopVectorize/float-minmax-instruction-flag.ll

; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt < %s -passes=loop-vectorize -force-vector-width=4 -S | FileCheck %s

; The function finds the smallest value from a float vector.
; Check if vectorization is enabled by instruction flag `fcmp nnan`.

define float @minloop(ptr nocapture readonly %arg) {
; CHECK-LABEL: @minloop(
; CHECK-NEXT:  top:
; CHECK-NEXT:    [[T:%.*]] = load float, ptr [[ARG:%.*]], align 4
; CHECK-NEXT:    br label [[LOOP:%.*]]
; CHECK:       loop:
; CHECK-NEXT:    [[T1:%.*]] = phi i64 [ [[T7:%.*]], [[LOOP]] ], [ 1, [[TOP:%.*]] ]
; CHECK-NEXT:    [[T2:%.*]] = phi float [ [[T6:%.*]], [[LOOP]] ], [ [[T]], [[TOP]] ]
; CHECK-NEXT:    [[T3:%.*]] = getelementptr float, ptr [[ARG]], i64 [[T1]]
; CHECK-NEXT:    [[T4:%.*]] = load float, ptr [[T3]], align 4
; CHECK-NEXT:    [[T5:%.*]] = fcmp nnan olt float [[T2]], [[T4]]
; CHECK-NEXT:    [[T6]] = select i1 [[T5]], float [[T2]], float [[T4]]
; CHECK-NEXT:    [[T7]] = add i64 [[T1]], 1
; CHECK-NEXT:    [[T8:%.*]] = icmp eq i64 [[T7]], 65537
; CHECK-NEXT:    br i1 [[T8]], label [[OUT:%.*]], label [[LOOP]]
; CHECK:       out:
; CHECK-NEXT:    [[T6_LCSSA:%.*]] = phi float [ [[T6]], [[LOOP]] ]
; CHECK-NEXT:    ret float [[T6_LCSSA]]
;
top:
  %t = load float, ptr %arg
  br label %loop

loop:
  %t1 = phi i64 [ %t7, %loop ], [ 1, %top ]
  %t2 = phi float [ %t6, %loop ], [ %t, %top ]
  %t3 = getelementptr float, ptr %arg, i64 %t1
  %t4 = load float, ptr %t3, align 4
  %t5 = fcmp nnan olt float %t2, %t4
  %t6 = select i1 %t5, float %t2, float %t4
  %t7 = add i64 %t1, 1
  %t8 = icmp eq i64 %t7, 65537
  br i1 %t8, label %out, label %loop

out:
  ret float %t6
}

; Function attributes are not used for FP min/max vectorization.

define float @minloopattr(ptr nocapture readonly %arg) #0 {
; CHECK-LABEL: @minloopattr(
; CHECK-NEXT:  top:
; CHECK-NEXT:    [[T:%.*]] = load float, ptr [[ARG:%.*]], align 4
; CHECK-NEXT:    br label [[LOOP:%.*]]
; CHECK:       loop:
; CHECK-NEXT:    [[T1:%.*]] = phi i64 [ [[T7:%.*]], [[LOOP]] ], [ 1, [[TOP:%.*]] ]
; CHECK-NEXT:    [[T2:%.*]] = phi float [ [[T6:%.*]], [[LOOP]] ], [ [[T]], [[TOP]] ]
; CHECK-NEXT:    [[T3:%.*]] = getelementptr float, ptr [[ARG]], i64 [[T1]]
; CHECK-NEXT:    [[T4:%.*]] = load float, ptr [[T3]], align 4
; CHECK-NEXT:    [[T5:%.*]] = fcmp olt float [[T2]], [[T4]]
; CHECK-NEXT:    [[T6]] = select i1 [[T5]], float [[T2]], float [[T4]]
; CHECK-NEXT:    [[T7]] = add i64 [[T1]], 1
; CHECK-NEXT:    [[T8:%.*]] = icmp eq i64 [[T7]], 65537
; CHECK-NEXT:    br i1 [[T8]], label [[OUT:%.*]], label [[LOOP]]
; CHECK:       out:
; CHECK-NEXT:    [[T6_LCSSA:%.*]] = phi float [ [[T6]], [[LOOP]] ]
; CHECK-NEXT:    ret float [[T6_LCSSA]]
;
top:
  %t = load float, ptr %arg
  br label %loop

loop:
  %t1 = phi i64 [ %t7, %loop ], [ 1, %top ]
  %t2 = phi float [ %t6, %loop ], [ %t, %top ]
  %t3 = getelementptr float, ptr %arg, i64 %t1
  %t4 = load float, ptr %t3, align 4
  %t5 = fcmp olt float %t2, %t4
  %t6 = select i1 %t5, float %t2, float %t4
  %t7 = add i64 %t1, 1
  %t8 = icmp eq i64 %t7, 65537
  br i1 %t8, label %out, label %loop

out:
  ret float %t6
}

; Check if vectorization is prevented without the flag or attribute.

define float @minloopnovec(ptr nocapture readonly %arg) {
; CHECK-LABEL: @minloopnovec(
; CHECK-NEXT:  top:
; CHECK-NEXT:    [[T:%.*]] = load float, ptr [[ARG:%.*]], align 4
; CHECK-NEXT:    br label [[LOOP:%.*]]
; CHECK:       loop:
; CHECK-NEXT:    [[T1:%.*]] = phi i64 [ [[T7:%.*]], [[LOOP]] ], [ 1, [[TOP:%.*]] ]
; CHECK-NEXT:    [[T2:%.*]] = phi float [ [[T6:%.*]], [[LOOP]] ], [ [[T]], [[TOP]] ]
; CHECK-NEXT:    [[T3:%.*]] = getelementptr float, ptr [[ARG]], i64 [[T1]]
; CHECK-NEXT:    [[T4:%.*]] = load float, ptr [[T3]], align 4
; CHECK-NEXT:    [[T5:%.*]] = fcmp olt float [[T2]], [[T4]]
; CHECK-NEXT:    [[T6]] = select i1 [[T5]], float [[T2]], float [[T4]]
; CHECK-NEXT:    [[T7]] = add i64 [[T1]], 1
; CHECK-NEXT:    [[T8:%.*]] = icmp eq i64 [[T7]], 65537
; CHECK-NEXT:    br i1 [[T8]], label [[OUT:%.*]], label [[LOOP]]
; CHECK:       out:
; CHECK-NEXT:    [[T6_LCSSA:%.*]] = phi float [ [[T6]], [[LOOP]] ]
; CHECK-NEXT:    ret float [[T6_LCSSA]]
;
top:
  %t = load float, ptr %arg
  br label %loop

loop:
  %t1 = phi i64 [ %t7, %loop ], [ 1, %top ]
  %t2 = phi float [ %t6, %loop ], [ %t, %top ]
  %t3 = getelementptr float, ptr %arg, i64 %t1
  %t4 = load float, ptr %t3, align 4
  %t5 = fcmp olt float %t2, %t4
  %t6 = select i1 %t5, float %t2, float %t4
  %t7 = add i64 %t1, 1
  %t8 = icmp eq i64 %t7, 65537
  br i1 %t8, label %out, label %loop

out:
  ret float %t6
}

; This test is checking that we don't vectorize when only one of the required flags is set.

define float @minloopmissingnsz(ptr nocapture readonly %arg) #1 {
; CHECK-LABEL: @minloopmissingnsz(
; CHECK-NEXT:  top:
; CHECK-NEXT:    [[T:%.*]] = load float, ptr [[ARG:%.*]], align 4
; CHECK-NEXT:    br label [[LOOP:%.*]]
; CHECK:       loop:
; CHECK-NEXT:    [[T1:%.*]] = phi i64 [ [[T7:%.*]], [[LOOP]] ], [ 1, [[TOP:%.*]] ]
; CHECK-NEXT:    [[T2:%.*]] = phi float [ [[T6:%.*]], [[LOOP]] ], [ [[T]], [[TOP]] ]
; CHECK-NEXT:    [[T3:%.*]] = getelementptr float, ptr [[ARG]], i64 [[T1]]
; CHECK-NEXT:    [[T4:%.*]] = load float, ptr [[T3]], align 4
; CHECK-NEXT:    [[T5:%.*]] = fcmp olt float [[T2]], [[T4]]
; CHECK-NEXT:    [[T6]] = select i1 [[T5]], float [[T2]], float [[T4]]
; CHECK-NEXT:    [[T7]] = add i64 [[T1]], 1
; CHECK-NEXT:    [[T8:%.*]] = icmp eq i64 [[T7]], 65537
; CHECK-NEXT:    br i1 [[T8]], label [[OUT:%.*]], label [[LOOP]]
; CHECK:       out:
; CHECK-NEXT:    [[T6_LCSSA:%.*]] = phi float [ [[T6]], [[LOOP]] ]
; CHECK-NEXT:    ret float [[T6_LCSSA]]
;
top:
  %t = load float, ptr %arg
  br label %loop

loop:
  %t1 = phi i64 [ %t7, %loop ], [ 1, %top ]
  %t2 = phi float [ %t6, %loop ], [ %t, %top ]
  %t3 = getelementptr float, ptr %arg, i64 %t1
  %t4 = load float, ptr %t3, align 4
  %t5 = fcmp olt float %t2, %t4
  %t6 = select i1 %t5, float %t2, float %t4
  %t7 = add i64 %t1, 1
  %t8 = icmp eq i64 %t7, 65537
  br i1 %t8, label %out, label %loop

out:
  ret float %t6
}

; This would assert on FMF propagation.

define void @not_a_min_max() {
; CHECK-LABEL: @not_a_min_max(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    br label [[LOOP:%.*]]
; CHECK:       loop:
; CHECK-NEXT:    [[F9_S0_V0:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[ADD:%.*]], [[LOOP]] ]
; CHECK-NEXT:    [[T14:%.*]] = icmp eq i32 [[F9_S0_V0]], 5
; CHECK-NEXT:    [[T15:%.*]] = select reassoc nnan ninf nsz contract afn i1 [[T14]], float 0x36A0000000000000, float 0.000000e+00
; CHECK-NEXT:    [[ADD]] = add nuw nsw i32 [[F9_S0_V0]], 1
; CHECK-NEXT:    br i1 false, label [[END:%.*]], label [[LOOP]]
; CHECK:       end:
; CHECK-NEXT:    ret void
;
entry:
  br label %loop

loop:
  %f9.s0.v0 = phi i32 [ 0, %entry ], [ %add, %loop ]
  %t14 = icmp eq i32 %f9.s0.v0, 5
  %t15 = select reassoc nnan ninf nsz contract afn i1 %t14, float 0x36A0000000000000, float 0.0
  %add = add nuw nsw i32 %f9.s0.v0, 1
  br i1 false, label %end, label %loop

end:
  ret void
}

attributes #0 = { "no-nans-fp-math"="true" "no-signed-zeros-fp-math"="true" }
attributes #1 = { "no-nans-fp-math"="true" }