[PowerPC] Only legalize FNEARBYINT with unsafe fp math

Commit 0f0330a78709 legalized these nodes on PPC without consideration of
unsafe math which means that we get inexact exceptions raised for nearbyint.
Since this doesn't conform to the standard, switch this legalization to depend
on unsafe fp math.

3 files changed, 260 insertions, 81 deletions

diff --git a/llvm/lib/Target/PowerPC/PPCISelLowering.cpp b/llvm/lib/Target/PowerPC/PPCISelLowering.cpp
index 9fab29dded7..df2234eea28 100644
--- a/llvm/lib/Target/PowerPC/PPCISelLowering.cpp
+++ b/llvm/lib/Target/PowerPC/PPCISelLowering.cpp
@@ -782,17 +782,22 @@ PPCTargetLowering::PPCTargetLowering(const PPCTargetMachine &TM,
       }
       setOperationAction(ISD::EXTRACT_VECTOR_ELT, MVT::v2f64, Legal);
 
+      // The nearbyint variants are not allowed to raise the inexact exception
+      // so we can only code-gen them with unsafe math.
+      if (TM.Options.UnsafeFPMath) {
+        setOperationAction(ISD::FNEARBYINT, MVT::f64, Legal);
+        setOperationAction(ISD::FNEARBYINT, MVT::f32, Legal);
+      }
+
       setOperationAction(ISD::FFLOOR, MVT::v2f64, Legal);
       setOperationAction(ISD::FCEIL, MVT::v2f64, Legal);
       setOperationAction(ISD::FTRUNC, MVT::v2f64, Legal);
       setOperationAction(ISD::FNEARBYINT, MVT::v2f64, Legal);
       setOperationAction(ISD::FROUND, MVT::v2f64, Legal);
-      setOperationAction(ISD::FNEARBYINT, MVT::f64, Legal);
       setOperationAction(ISD::FROUND, MVT::f64, Legal);
 
       setOperationAction(ISD::FNEARBYINT, MVT::v4f32, Legal);
       setOperationAction(ISD::FROUND, MVT::v4f32, Legal);
-      setOperationAction(ISD::FNEARBYINT, MVT::f32, Legal);
       setOperationAction(ISD::FROUND, MVT::f32, Legal);
 
       setOperationAction(ISD::MUL, MVT::v2f64, Legal);
diff --git a/llvm/test/CodeGen/PowerPC/scalar-rounding-ops.ll b/llvm/test/CodeGen/PowerPC/scalar-rounding-ops.ll
index 94625fcf551..7f18853b8b9 100644
--- a/llvm/test/CodeGen/PowerPC/scalar-rounding-ops.ll
+++ b/llvm/test/CodeGen/PowerPC/scalar-rounding-ops.ll
@@ -342,14 +342,14 @@ declare i64 @llvm.llround.i64.f32(float)
 
 define dso_local double @test_nearbyint(double %d) local_unnamed_addr {
 ; BE-LABEL: test_nearbyint:
-; BE:       # %bb.0: # %entry
-; BE-NEXT:    xsrdpic f1, f1
-; BE-NEXT:    blr
+; BE:    # %bb.0: # %entry
+; BE:    bl nearbyint
+; BE:    blr
 ;
 ; CHECK-LABEL: test_nearbyint:
-; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    xsrdpic f1, f1
-; CHECK-NEXT:    blr
+; CHECK:    # %bb.0: # %entry
+; CHECK:    bl nearbyint
+; CHECK:    blr
 ;
 ; FAST-LABEL: test_nearbyint:
 ; FAST:       # %bb.0: # %entry
@@ -364,14 +364,14 @@ declare double @llvm.nearbyint.f64(double)
 
 define dso_local float @test_nearbyintf(float %f) local_unnamed_addr {
 ; BE-LABEL: test_nearbyintf:
-; BE:       # %bb.0: # %entry
-; BE-NEXT:    xsrdpic f1, f1
-; BE-NEXT:    blr
+; BE:    # %bb.0: # %entry
+; BE:    bl nearbyint
+; BE:    blr
 ;
 ; CHECK-LABEL: test_nearbyintf:
-; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    xsrdpic f1, f1
-; CHECK-NEXT:    blr
+; CHECK:    # %bb.0: # %entry
+; CHECK:    bl nearbyintf
+; CHECK:    blr
 ;
 ; FAST-LABEL: test_nearbyintf:
 ; FAST:       # %bb.0: # %entry
diff --git a/llvm/test/CodeGen/PowerPC/vector-constrained-fp-intrinsics.ll b/llvm/test/CodeGen/PowerPC/vector-constrained-fp-intrinsics.ll
index ecd19fd677c..253e74cf0bf 100644
--- a/llvm/test/CodeGen/PowerPC/vector-constrained-fp-intrinsics.ll
+++ b/llvm/test/CodeGen/PowerPC/vector-constrained-fp-intrinsics.ll
@@ -5101,20 +5101,34 @@ entry:
 define <1 x float> @constrained_vector_nearbyint_v1f32() #0 {
 ; PC64LE-LABEL: constrained_vector_nearbyint_v1f32:
 ; PC64LE:       # %bb.0: # %entry
+; PC64LE-NEXT:    mflr 0
+; PC64LE-NEXT:    std 0, 16(1)
+; PC64LE-NEXT:    stdu 1, -32(1)
 ; PC64LE-NEXT:    addis 3, 2, .LCPI80_0@toc@ha
-; PC64LE-NEXT:    lfs 0, .LCPI80_0@toc@l(3)
-; PC64LE-NEXT:    xsrdpic 0, 0
-; PC64LE-NEXT:    xscvdpspn 0, 0
+; PC64LE-NEXT:    lfs 1, .LCPI80_0@toc@l(3)
+; PC64LE-NEXT:    bl nearbyintf
+; PC64LE-NEXT:    nop
+; PC64LE-NEXT:    xscvdpspn 0, 1
 ; PC64LE-NEXT:    xxsldwi 34, 0, 0, 1
+; PC64LE-NEXT:    addi 1, 1, 32
+; PC64LE-NEXT:    ld 0, 16(1)
+; PC64LE-NEXT:    mtlr 0
 ; PC64LE-NEXT:    blr
 ;
 ; PC64LE9-LABEL: constrained_vector_nearbyint_v1f32:
 ; PC64LE9:       # %bb.0: # %entry
+; PC64LE9-NEXT:    mflr 0
+; PC64LE9-NEXT:    std 0, 16(1)
+; PC64LE9-NEXT:    stdu 1, -32(1)
 ; PC64LE9-NEXT:    addis 3, 2, .LCPI80_0@toc@ha
-; PC64LE9-NEXT:    lfs 0, .LCPI80_0@toc@l(3)
-; PC64LE9-NEXT:    xsrdpic 0, 0
-; PC64LE9-NEXT:    xscvdpspn 0, 0
+; PC64LE9-NEXT:    lfs 1, .LCPI80_0@toc@l(3)
+; PC64LE9-NEXT:    bl nearbyintf
+; PC64LE9-NEXT:    nop
+; PC64LE9-NEXT:    xscvdpspn 0, 1
 ; PC64LE9-NEXT:    xxsldwi 34, 0, 0, 1
+; PC64LE9-NEXT:    addi 1, 1, 32
+; PC64LE9-NEXT:    ld 0, 16(1)
+; PC64LE9-NEXT:    mtlr 0
 ; PC64LE9-NEXT:    blr
 entry:
   %nearby = call <1 x float> @llvm.experimental.constrained.nearbyint.v1f32(
@@ -5127,19 +5141,50 @@ entry:
 define <2 x double> @constrained_vector_nearbyint_v2f64() #0 {
 ; PC64LE-LABEL: constrained_vector_nearbyint_v2f64:
 ; PC64LE:       # %bb.0: # %entry
+; PC64LE-NEXT:    mflr 0
+; PC64LE-NEXT:    std 0, 16(1)
+; PC64LE-NEXT:    stdu 1, -64(1)
 ; PC64LE-NEXT:    addis 3, 2, .LCPI81_0@toc@ha
-; PC64LE-NEXT:    addi 3, 3, .LCPI81_0@toc@l
-; PC64LE-NEXT:    lxvd2x 0, 0, 3
-; PC64LE-NEXT:    xxswapd 0, 0
-; PC64LE-NEXT:    xvrdpic 34, 0
+; PC64LE-NEXT:    lfd 1, .LCPI81_0@toc@l(3)
+; PC64LE-NEXT:    bl nearbyint
+; PC64LE-NEXT:    nop
+; PC64LE-NEXT:    li 3, 48
+; PC64LE-NEXT:    # kill: def $f1 killed $f1 def $vsl1
+; PC64LE-NEXT:    stxvd2x 1, 1, 3 # 16-byte Folded Spill
+; PC64LE-NEXT:    addis 3, 2, .LCPI81_1@toc@ha
+; PC64LE-NEXT:    lfs 1, .LCPI81_1@toc@l(3)
+; PC64LE-NEXT:    bl nearbyint
+; PC64LE-NEXT:    nop
+; PC64LE-NEXT:    li 3, 48
+; PC64LE-NEXT:    # kill: def $f1 killed $f1 def $vsl1
+; PC64LE-NEXT:    lxvd2x 0, 1, 3 # 16-byte Folded Reload
+; PC64LE-NEXT:    xxmrghd 34, 1, 0
+; PC64LE-NEXT:    addi 1, 1, 64
+; PC64LE-NEXT:    ld 0, 16(1)
+; PC64LE-NEXT:    mtlr 0
 ; PC64LE-NEXT:    blr
 ;
 ; PC64LE9-LABEL: constrained_vector_nearbyint_v2f64:
 ; PC64LE9:       # %bb.0: # %entry
+; PC64LE9-NEXT:    mflr 0
+; PC64LE9-NEXT:    std 0, 16(1)
+; PC64LE9-NEXT:    stdu 1, -48(1)
 ; PC64LE9-NEXT:    addis 3, 2, .LCPI81_0@toc@ha
-; PC64LE9-NEXT:    addi 3, 3, .LCPI81_0@toc@l
-; PC64LE9-NEXT:    lxvx 0, 0, 3
-; PC64LE9-NEXT:    xvrdpic 34, 0
+; PC64LE9-NEXT:    lfd 1, .LCPI81_0@toc@l(3)
+; PC64LE9-NEXT:    bl nearbyint
+; PC64LE9-NEXT:    nop
+; PC64LE9-NEXT:    addis 3, 2, .LCPI81_1@toc@ha
+; PC64LE9-NEXT:    # kill: def $f1 killed $f1 def $vsl1
+; PC64LE9-NEXT:    stxv 1, 32(1) # 16-byte Folded Spill
+; PC64LE9-NEXT:    lfs 1, .LCPI81_1@toc@l(3)
+; PC64LE9-NEXT:    bl nearbyint
+; PC64LE9-NEXT:    nop
+; PC64LE9-NEXT:    lxv 0, 32(1) # 16-byte Folded Reload
+; PC64LE9-NEXT:    # kill: def $f1 killed $f1 def $vsl1
+; PC64LE9-NEXT:    xxmrghd 34, 1, 0
+; PC64LE9-NEXT:    addi 1, 1, 48
+; PC64LE9-NEXT:    ld 0, 16(1)
+; PC64LE9-NEXT:    mtlr 0
 ; PC64LE9-NEXT:    blr
 entry:
   %nearby = call <2 x double> @llvm.experimental.constrained.nearbyint.v2f64(
@@ -5152,50 +5197,80 @@ entry:
 define <3 x float> @constrained_vector_nearbyint_v3f32() #0 {
 ; PC64LE-LABEL: constrained_vector_nearbyint_v3f32:
 ; PC64LE:       # %bb.0: # %entry
-; PC64LE-NEXT:    addis 3, 2, .LCPI82_2@toc@ha
-; PC64LE-NEXT:    addis 4, 2, .LCPI82_1@toc@ha
-; PC64LE-NEXT:    lfs 0, .LCPI82_2@toc@l(3)
-; PC64LE-NEXT:    lfs 1, .LCPI82_1@toc@l(4)
+; PC64LE-NEXT:    mflr 0
+; PC64LE-NEXT:    stfd 30, -16(1) # 8-byte Folded Spill
+; PC64LE-NEXT:    stfd 31, -8(1) # 8-byte Folded Spill
+; PC64LE-NEXT:    std 0, 16(1)
+; PC64LE-NEXT:    stdu 1, -48(1)
 ; PC64LE-NEXT:    addis 3, 2, .LCPI82_0@toc@ha
-; PC64LE-NEXT:    xsrdpic 0, 0
-; PC64LE-NEXT:    lfs 2, .LCPI82_0@toc@l(3)
+; PC64LE-NEXT:    lfs 1, .LCPI82_0@toc@l(3)
+; PC64LE-NEXT:    bl nearbyintf
+; PC64LE-NEXT:    nop
+; PC64LE-NEXT:    addis 3, 2, .LCPI82_1@toc@ha
+; PC64LE-NEXT:    fmr 31, 1
+; PC64LE-NEXT:    lfs 1, .LCPI82_1@toc@l(3)
+; PC64LE-NEXT:    bl nearbyintf
+; PC64LE-NEXT:    nop
+; PC64LE-NEXT:    addis 3, 2, .LCPI82_2@toc@ha
+; PC64LE-NEXT:    fmr 30, 1
+; PC64LE-NEXT:    lfs 1, .LCPI82_2@toc@l(3)
+; PC64LE-NEXT:    bl nearbyintf
+; PC64LE-NEXT:    nop
+; PC64LE-NEXT:    xscvdpspn 0, 30
 ; PC64LE-NEXT:    addis 3, 2, .LCPI82_3@toc@ha
-; PC64LE-NEXT:    xsrdpic 1, 1
-; PC64LE-NEXT:    addi 3, 3, .LCPI82_3@toc@l
-; PC64LE-NEXT:    xsrdpic 2, 2
-; PC64LE-NEXT:    xscvdpspn 0, 0
 ; PC64LE-NEXT:    xscvdpspn 1, 1
+; PC64LE-NEXT:    addi 3, 3, .LCPI82_3@toc@l
 ; PC64LE-NEXT:    xxsldwi 34, 0, 0, 1
-; PC64LE-NEXT:    xscvdpspn 0, 2
+; PC64LE-NEXT:    xscvdpspn 0, 31
 ; PC64LE-NEXT:    xxsldwi 35, 1, 1, 1
-; PC64LE-NEXT:    vmrglw 2, 3, 2
+; PC64LE-NEXT:    vmrglw 2, 2, 3
 ; PC64LE-NEXT:    lvx 3, 0, 3
 ; PC64LE-NEXT:    xxsldwi 36, 0, 0, 1
 ; PC64LE-NEXT:    vperm 2, 4, 2, 3
+; PC64LE-NEXT:    addi 1, 1, 48
+; PC64LE-NEXT:    ld 0, 16(1)
+; PC64LE-NEXT:    lfd 31, -8(1) # 8-byte Folded Reload
+; PC64LE-NEXT:    lfd 30, -16(1) # 8-byte Folded Reload
+; PC64LE-NEXT:    mtlr 0
 ; PC64LE-NEXT:    blr
 ;
 ; PC64LE9-LABEL: constrained_vector_nearbyint_v3f32:
 ; PC64LE9:       # %bb.0: # %entry
+; PC64LE9-NEXT:    mflr 0
+; PC64LE9-NEXT:    stfd 30, -16(1) # 8-byte Folded Spill
+; PC64LE9-NEXT:    stfd 31, -8(1) # 8-byte Folded Spill
+; PC64LE9-NEXT:    std 0, 16(1)
+; PC64LE9-NEXT:    stdu 1, -48(1)
 ; PC64LE9-NEXT:    addis 3, 2, .LCPI82_0@toc@ha
-; PC64LE9-NEXT:    lfs 0, .LCPI82_0@toc@l(3)
+; PC64LE9-NEXT:    lfs 1, .LCPI82_0@toc@l(3)
+; PC64LE9-NEXT:    bl nearbyintf
+; PC64LE9-NEXT:    nop
 ; PC64LE9-NEXT:    addis 3, 2, .LCPI82_1@toc@ha
+; PC64LE9-NEXT:    fmr 31, 1
 ; PC64LE9-NEXT:    lfs 1, .LCPI82_1@toc@l(3)
+; PC64LE9-NEXT:    bl nearbyintf
+; PC64LE9-NEXT:    nop
 ; PC64LE9-NEXT:    addis 3, 2, .LCPI82_2@toc@ha
-; PC64LE9-NEXT:    xsrdpic 0, 0
-; PC64LE9-NEXT:    lfs 2, .LCPI82_2@toc@l(3)
+; PC64LE9-NEXT:    fmr 30, 1
+; PC64LE9-NEXT:    lfs 1, .LCPI82_2@toc@l(3)
+; PC64LE9-NEXT:    bl nearbyintf
+; PC64LE9-NEXT:    nop
+; PC64LE9-NEXT:    xscvdpspn 0, 1
+; PC64LE9-NEXT:    xxsldwi 34, 0, 0, 1
+; PC64LE9-NEXT:    xscvdpspn 0, 30
+; PC64LE9-NEXT:    xxsldwi 35, 0, 0, 1
+; PC64LE9-NEXT:    xscvdpspn 0, 31
 ; PC64LE9-NEXT:    addis 3, 2, .LCPI82_3@toc@ha
 ; PC64LE9-NEXT:    addi 3, 3, .LCPI82_3@toc@l
-; PC64LE9-NEXT:    xsrdpic 1, 1
-; PC64LE9-NEXT:    xsrdpic 2, 2
-; PC64LE9-NEXT:    xscvdpspn 0, 0
-; PC64LE9-NEXT:    xscvdpspn 1, 1
-; PC64LE9-NEXT:    xscvdpspn 2, 2
-; PC64LE9-NEXT:    xxsldwi 36, 0, 0, 1
-; PC64LE9-NEXT:    xxsldwi 35, 1, 1, 1
-; PC64LE9-NEXT:    xxsldwi 34, 2, 2, 1
 ; PC64LE9-NEXT:    vmrglw 2, 3, 2
 ; PC64LE9-NEXT:    lxvx 35, 0, 3
+; PC64LE9-NEXT:    xxsldwi 36, 0, 0, 1
 ; PC64LE9-NEXT:    vperm 2, 4, 2, 3
+; PC64LE9-NEXT:    addi 1, 1, 48
+; PC64LE9-NEXT:    ld 0, 16(1)
+; PC64LE9-NEXT:    lfd 31, -8(1) # 8-byte Folded Reload
+; PC64LE9-NEXT:    lfd 30, -16(1) # 8-byte Folded Reload
+; PC64LE9-NEXT:    mtlr 0
 ; PC64LE9-NEXT:    blr
 entry:
   %nearby = call <3 x float> @llvm.experimental.constrained.nearbyint.v3f32(
@@ -5208,31 +5283,72 @@ entry:
 define <3 x double> @constrained_vector_nearby_v3f64() #0 {
 ; PC64LE-LABEL: constrained_vector_nearby_v3f64:
 ; PC64LE:       # %bb.0: # %entry
-; PC64LE-NEXT:    addis 3, 2, .LCPI83_1@toc@ha
-; PC64LE-NEXT:    addi 3, 3, .LCPI83_1@toc@l
-; PC64LE-NEXT:    lxvd2x 0, 0, 3
+; PC64LE-NEXT:    mflr 0
+; PC64LE-NEXT:    std 0, 16(1)
+; PC64LE-NEXT:    stdu 1, -80(1)
+; PC64LE-NEXT:    li 3, 64
+; PC64LE-NEXT:    stxvd2x 63, 1, 3 # 16-byte Folded Spill
 ; PC64LE-NEXT:    addis 3, 2, .LCPI83_0@toc@ha
 ; PC64LE-NEXT:    lfd 1, .LCPI83_0@toc@l(3)
-; PC64LE-NEXT:    xxswapd 0, 0
-; PC64LE-NEXT:    xsrdpic 3, 1
-; PC64LE-NEXT:    xvrdpic 2, 0
-; PC64LE-NEXT:    xxswapd 1, 2
-; PC64LE-NEXT:    # kill: def $f2 killed $f2 killed $vsl2
-; PC64LE-NEXT:    # kill: def $f1 killed $f1 killed $vsl1
+; PC64LE-NEXT:    bl nearbyint
+; PC64LE-NEXT:    nop
+; PC64LE-NEXT:    li 3, 48
+; PC64LE-NEXT:    # kill: def $f1 killed $f1 def $vsl1
+; PC64LE-NEXT:    stxvd2x 1, 1, 3 # 16-byte Folded Spill
+; PC64LE-NEXT:    addis 3, 2, .LCPI83_1@toc@ha
+; PC64LE-NEXT:    lfs 1, .LCPI83_1@toc@l(3)
+; PC64LE-NEXT:    bl nearbyint
+; PC64LE-NEXT:    nop
+; PC64LE-NEXT:    li 3, 48
+; PC64LE-NEXT:    # kill: def $f1 killed $f1 def $vsl1
+; PC64LE-NEXT:    lxvd2x 0, 1, 3 # 16-byte Folded Reload
+; PC64LE-NEXT:    addis 3, 2, .LCPI83_2@toc@ha
+; PC64LE-NEXT:    xxmrghd 63, 0, 1
+; PC64LE-NEXT:    lfd 1, .LCPI83_2@toc@l(3)
+; PC64LE-NEXT:    bl nearbyint
+; PC64LE-NEXT:    nop
+; PC64LE-NEXT:    xxswapd 0, 63
+; PC64LE-NEXT:    li 3, 64
+; PC64LE-NEXT:    xxlor 2, 63, 63
+; PC64LE-NEXT:    lxvd2x 63, 1, 3 # 16-byte Folded Reload
+; PC64LE-NEXT:    fmr 3, 1
+; PC64LE-NEXT:    fmr 1, 0
+; PC64LE-NEXT:    addi 1, 1, 80
+; PC64LE-NEXT:    ld 0, 16(1)
+; PC64LE-NEXT:    mtlr 0
 ; PC64LE-NEXT:    blr
 ;
 ; PC64LE9-LABEL: constrained_vector_nearby_v3f64:
 ; PC64LE9:       # %bb.0: # %entry
+; PC64LE9-NEXT:    mflr 0
+; PC64LE9-NEXT:    std 0, 16(1)
+; PC64LE9-NEXT:    stdu 1, -64(1)
 ; PC64LE9-NEXT:    addis 3, 2, .LCPI83_0@toc@ha
-; PC64LE9-NEXT:    lfd 0, .LCPI83_0@toc@l(3)
+; PC64LE9-NEXT:    lfd 1, .LCPI83_0@toc@l(3)
+; PC64LE9-NEXT:    stxv 63, 48(1) # 16-byte Folded Spill
+; PC64LE9-NEXT:    bl nearbyint
+; PC64LE9-NEXT:    nop
 ; PC64LE9-NEXT:    addis 3, 2, .LCPI83_1@toc@ha
-; PC64LE9-NEXT:    addi 3, 3, .LCPI83_1@toc@l
-; PC64LE9-NEXT:    xsrdpic 3, 0
-; PC64LE9-NEXT:    lxvx 0, 0, 3
-; PC64LE9-NEXT:    xvrdpic 2, 0
-; PC64LE9-NEXT:    xxswapd 1, 2
+; PC64LE9-NEXT:    # kill: def $f1 killed $f1 def $vsl1
+; PC64LE9-NEXT:    stxv 1, 32(1) # 16-byte Folded Spill
+; PC64LE9-NEXT:    lfs 1, .LCPI83_1@toc@l(3)
+; PC64LE9-NEXT:    bl nearbyint
+; PC64LE9-NEXT:    nop
+; PC64LE9-NEXT:    lxv 0, 32(1) # 16-byte Folded Reload
+; PC64LE9-NEXT:    addis 3, 2, .LCPI83_2@toc@ha
+; PC64LE9-NEXT:    # kill: def $f1 killed $f1 def $vsl1
+; PC64LE9-NEXT:    xxmrghd 63, 0, 1
+; PC64LE9-NEXT:    lfd 1, .LCPI83_2@toc@l(3)
+; PC64LE9-NEXT:    bl nearbyint
+; PC64LE9-NEXT:    nop
+; PC64LE9-NEXT:    fmr 3, 1
+; PC64LE9-NEXT:    xxswapd 1, 63
+; PC64LE9-NEXT:    xscpsgndp 2, 63, 63
+; PC64LE9-NEXT:    lxv 63, 48(1) # 16-byte Folded Reload
 ; PC64LE9-NEXT:    # kill: def $f1 killed $f1 killed $vsl1
-; PC64LE9-NEXT:    # kill: def $f2 killed $f2 killed $vsl2
+; PC64LE9-NEXT:    addi 1, 1, 64
+; PC64LE9-NEXT:    ld 0, 16(1)
+; PC64LE9-NEXT:    mtlr 0
 ; PC64LE9-NEXT:    blr
 entry:
   %nearby = call <3 x double> @llvm.experimental.constrained.nearbyint.v3f64(
@@ -5245,28 +5361,86 @@ entry:
 define <4 x double> @constrained_vector_nearbyint_v4f64() #0 {
 ; PC64LE-LABEL: constrained_vector_nearbyint_v4f64:
 ; PC64LE:       # %bb.0: # %entry
+; PC64LE-NEXT:    mflr 0
+; PC64LE-NEXT:    std 0, 16(1)
+; PC64LE-NEXT:    stdu 1, -80(1)
+; PC64LE-NEXT:    li 3, 64
+; PC64LE-NEXT:    stxvd2x 63, 1, 3 # 16-byte Folded Spill
 ; PC64LE-NEXT:    addis 3, 2, .LCPI84_0@toc@ha
-; PC64LE-NEXT:    addis 4, 2, .LCPI84_1@toc@ha
-; PC64LE-NEXT:    addi 3, 3, .LCPI84_0@toc@l
-; PC64LE-NEXT:    lxvd2x 0, 0, 3
-; PC64LE-NEXT:    addi 3, 4, .LCPI84_1@toc@l
-; PC64LE-NEXT:    lxvd2x 1, 0, 3
-; PC64LE-NEXT:    xxswapd 0, 0
-; PC64LE-NEXT:    xxswapd 1, 1
-; PC64LE-NEXT:    xvrdpic 34, 0
-; PC64LE-NEXT:    xvrdpic 35, 1
+; PC64LE-NEXT:    lfd 1, .LCPI84_0@toc@l(3)
+; PC64LE-NEXT:    bl nearbyint
+; PC64LE-NEXT:    nop
+; PC64LE-NEXT:    li 3, 48
+; PC64LE-NEXT:    # kill: def $f1 killed $f1 def $vsl1
+; PC64LE-NEXT:    stxvd2x 1, 1, 3 # 16-byte Folded Spill
+; PC64LE-NEXT:    addis 3, 2, .LCPI84_1@toc@ha
+; PC64LE-NEXT:    lfd 1, .LCPI84_1@toc@l(3)
+; PC64LE-NEXT:    bl nearbyint
+; PC64LE-NEXT:    nop
+; PC64LE-NEXT:    li 3, 48
+; PC64LE-NEXT:    # kill: def $f1 killed $f1 def $vsl1
+; PC64LE-NEXT:    lxvd2x 0, 1, 3 # 16-byte Folded Reload
+; PC64LE-NEXT:    addis 3, 2, .LCPI84_2@toc@ha
+; PC64LE-NEXT:    xxmrghd 63, 1, 0
+; PC64LE-NEXT:    lfd 1, .LCPI84_2@toc@l(3)
+; PC64LE-NEXT:    bl nearbyint
+; PC64LE-NEXT:    nop
+; PC64LE-NEXT:    li 3, 48
+; PC64LE-NEXT:    # kill: def $f1 killed $f1 def $vsl1
+; PC64LE-NEXT:    stxvd2x 1, 1, 3 # 16-byte Folded Spill
+; PC64LE-NEXT:    addis 3, 2, .LCPI84_3@toc@ha
+; PC64LE-NEXT:    lfd 1, .LCPI84_3@toc@l(3)
+; PC64LE-NEXT:    bl nearbyint
+; PC64LE-NEXT:    nop
+; PC64LE-NEXT:    li 3, 48
+; PC64LE-NEXT:    vmr 2, 31
+; PC64LE-NEXT:    # kill: def $f1 killed $f1 def $vsl1
+; PC64LE-NEXT:    lxvd2x 0, 1, 3 # 16-byte Folded Reload
+; PC64LE-NEXT:    li 3, 64
+; PC64LE-NEXT:    lxvd2x 63, 1, 3 # 16-byte Folded Reload
+; PC64LE-NEXT:    xxmrghd 35, 1, 0
+; PC64LE-NEXT:    addi 1, 1, 80
+; PC64LE-NEXT:    ld 0, 16(1)
+; PC64LE-NEXT:    mtlr 0
 ; PC64LE-NEXT:    blr
 ;
 ; PC64LE9-LABEL: constrained_vector_nearbyint_v4f64:
 ; PC64LE9:       # %bb.0: # %entry
+; PC64LE9-NEXT:    mflr 0
+; PC64LE9-NEXT:    std 0, 16(1)
+; PC64LE9-NEXT:    stdu 1, -64(1)
 ; PC64LE9-NEXT:    addis 3, 2, .LCPI84_0@toc@ha
-; PC64LE9-NEXT:    addi 3, 3, .LCPI84_0@toc@l
-; PC64LE9-NEXT:    lxvx 0, 0, 3
+; PC64LE9-NEXT:    lfd 1, .LCPI84_0@toc@l(3)
+; PC64LE9-NEXT:    stxv 63, 48(1) # 16-byte Folded Spill
+; PC64LE9-NEXT:    bl nearbyint
+; PC64LE9-NEXT:    nop
 ; PC64LE9-NEXT:    addis 3, 2, .LCPI84_1@toc@ha
-; PC64LE9-NEXT:    addi 3, 3, .LCPI84_1@toc@l
-; PC64LE9-NEXT:    xvrdpic 34, 0
-; PC64LE9-NEXT:    lxvx 0, 0, 3
-; PC64LE9-NEXT:    xvrdpic 35, 0
+; PC64LE9-NEXT:    # kill: def $f1 killed $f1 def $vsl1
+; PC64LE9-NEXT:    stxv 1, 32(1) # 16-byte Folded Spill
+; PC64LE9-NEXT:    lfd 1, .LCPI84_1@toc@l(3)
+; PC64LE9-NEXT:    bl nearbyint
+; PC64LE9-NEXT:    nop
+; PC64LE9-NEXT:    lxv 0, 32(1) # 16-byte Folded Reload
+; PC64LE9-NEXT:    addis 3, 2, .LCPI84_2@toc@ha
+; PC64LE9-NEXT:    # kill: def $f1 killed $f1 def $vsl1
+; PC64LE9-NEXT:    xxmrghd 63, 1, 0
+; PC64LE9-NEXT:    lfd 1, .LCPI84_2@toc@l(3)
+; PC64LE9-NEXT:    bl nearbyint
+; PC64LE9-NEXT:    nop
+; PC64LE9-NEXT:    addis 3, 2, .LCPI84_3@toc@ha
+; PC64LE9-NEXT:    # kill: def $f1 killed $f1 def $vsl1
+; PC64LE9-NEXT:    stxv 1, 32(1) # 16-byte Folded Spill
+; PC64LE9-NEXT:    lfd 1, .LCPI84_3@toc@l(3)
+; PC64LE9-NEXT:    bl nearbyint
+; PC64LE9-NEXT:    nop
+; PC64LE9-NEXT:    lxv 0, 32(1) # 16-byte Folded Reload
+; PC64LE9-NEXT:    vmr 2, 31
+; PC64LE9-NEXT:    lxv 63, 48(1) # 16-byte Folded Reload
+; PC64LE9-NEXT:    # kill: def $f1 killed $f1 def $vsl1
+; PC64LE9-NEXT:    xxmrghd 35, 1, 0
+; PC64LE9-NEXT:    addi 1, 1, 64
+; PC64LE9-NEXT:    ld 0, 16(1)
+; PC64LE9-NEXT:    mtlr 0
 ; PC64LE9-NEXT:    blr
 entry:
   %nearby = call <4 x double> @llvm.experimental.constrained.nearbyint.v4f64(