[CostModel][X86] Adjust truncate SSE/AVX legalized costs based on llvm-mca reports.

Update truncation costs based on the worst case costs from the script in D103695.

Move to using legalized types wherever possible, which allows us to prune the cost tables.

diff --git a/llvm/lib/Target/X86/X86TargetTransformInfo.cpp b/llvm/lib/Target/X86/X86TargetTransformInfo.cpp
index 1d802e0..30156f4 100644 (file)
--- a/llvm/lib/Target/X86/X86TargetTransformInfo.cpp
+++ b/llvm/lib/Target/X86/X86TargetTransformInfo.cpp
@@ -1887,10 +1887,17 @@ InstructionCost X86TTIImpl::getCastInstrCost(unsigned Opcode, Type *Dst,
     { ISD::SIGN_EXTEND, MVT::v4i64,  MVT::v4i32,  2 },
     { ISD::ZERO_EXTEND, MVT::v4i64,  MVT::v4i32,  2 },
 
-    { ISD::TRUNCATE,    MVT::v2i8,   MVT::v2i32,  1 },
-    { ISD::TRUNCATE,    MVT::v2i16,  MVT::v2i64,  1 },
-    { ISD::TRUNCATE,    MVT::v4i32,  MVT::v4i64,  1 },
     { ISD::TRUNCATE,    MVT::v8i1,   MVT::v8i32,  2 },
+
+    { ISD::TRUNCATE,    MVT::v16i8,  MVT::v8i16,  1 },
+    { ISD::TRUNCATE,    MVT::v16i8,  MVT::v4i32,  1 },
+    { ISD::TRUNCATE,    MVT::v16i8,  MVT::v2i64,  1 },
+    { ISD::TRUNCATE,    MVT::v16i8,  MVT::v8i32,  4 },
+    { ISD::TRUNCATE,    MVT::v16i8,  MVT::v4i64,  4 },
+    { ISD::TRUNCATE,    MVT::v8i16,  MVT::v4i32,  1 },
+    { ISD::TRUNCATE,    MVT::v8i16,  MVT::v2i64,  1 },
+    { ISD::TRUNCATE,    MVT::v8i16,  MVT::v4i64,  5 },
+    { ISD::TRUNCATE,    MVT::v4i32,  MVT::v4i64,  1 },
     { ISD::TRUNCATE,    MVT::v8i16,  MVT::v8i32,  2 },
 
     { ISD::FP_EXTEND,   MVT::v8f64,  MVT::v8f32,  3 },
@@ -1955,16 +1962,11 @@ InstructionCost X86TTIImpl::getCastInstrCost(unsigned Opcode, Type *Dst,
     { ISD::TRUNCATE,    MVT::v16i1,  MVT::v16i64, 11 },
 
     { ISD::TRUNCATE,    MVT::v16i8,  MVT::v16i16, 2 }, // and+extract+packuswb
-    { ISD::TRUNCATE,    MVT::v2i8,   MVT::v2i32,  2 }, // and+packusdw+packuswb
-    { ISD::TRUNCATE,    MVT::v8i8,   MVT::v8i32,  4 },
+    { ISD::TRUNCATE,    MVT::v16i8,  MVT::v8i32,  5 },
     { ISD::TRUNCATE,    MVT::v8i16,  MVT::v8i32,  5 },
-    { ISD::TRUNCATE,    MVT::v4i8,   MVT::v4i64,  4 },
-    { ISD::TRUNCATE,    MVT::v4i16,  MVT::v4i64,  3 }, // and+extract+2*packusdw
+    { ISD::TRUNCATE,    MVT::v16i8,  MVT::v4i64,  5 },
+    { ISD::TRUNCATE,    MVT::v8i16,  MVT::v4i64,  3 }, // and+extract+2*packusdw
     { ISD::TRUNCATE,    MVT::v4i32,  MVT::v4i64,  2 },
-    { ISD::TRUNCATE,    MVT::v8i8,   MVT::v8i64, 11 },
-    { ISD::TRUNCATE,    MVT::v8i16,  MVT::v8i64,  9 },
-    { ISD::TRUNCATE,    MVT::v8i32,  MVT::v8i64,  3 },
-    { ISD::TRUNCATE,    MVT::v16i8,  MVT::v16i64, 11 },
 
     { ISD::SINT_TO_FP,  MVT::v4f32,  MVT::v4i1,   3 },
     { ISD::SINT_TO_FP,  MVT::v4f64,  MVT::v4i1,   3 },
@@ -2039,15 +2041,9 @@ InstructionCost X86TTIImpl::getCastInstrCost(unsigned Opcode, Type *Dst,
     { ISD::TRUNCATE,    MVT::v2i1,   MVT::v2i16,  1 }, // PMOVXZWQ
     { ISD::TRUNCATE,    MVT::v4i1,   MVT::v4i8,   1 }, // PMOVXZBD
 
-    { ISD::TRUNCATE,    MVT::v2i8,   MVT::v2i16,  1 },
-    { ISD::TRUNCATE,    MVT::v4i8,   MVT::v4i16,  1 },
-    { ISD::TRUNCATE,    MVT::v8i8,   MVT::v8i16,  1 },
-    { ISD::TRUNCATE,    MVT::v4i8,   MVT::v4i32,  1 },
-    { ISD::TRUNCATE,    MVT::v4i16,  MVT::v4i32,  1 },
-    { ISD::TRUNCATE,    MVT::v8i8,   MVT::v8i32,  3 },
-    { ISD::TRUNCATE,    MVT::v8i16,  MVT::v8i32,  3 },
-    { ISD::TRUNCATE,    MVT::v16i16, MVT::v16i32, 6 },
-    { ISD::TRUNCATE,    MVT::v2i8,   MVT::v2i64,  1 }, // PSHUFB
+    { ISD::TRUNCATE,    MVT::v16i8,  MVT::v4i32,  2 },
+    { ISD::TRUNCATE,    MVT::v8i16,  MVT::v4i32,  2 },
+    { ISD::TRUNCATE,    MVT::v16i8,  MVT::v2i64,  2 },
 
     { ISD::SINT_TO_FP,  MVT::f32,    MVT::i32,    1 },
     { ISD::SINT_TO_FP,  MVT::f64,    MVT::i32,    1 },
@@ -2166,21 +2162,17 @@ InstructionCost X86TTIImpl::getCastInstrCost(unsigned Opcode, Type *Dst,
     { ISD::TRUNCATE,    MVT::v4i1,   MVT::v4i8,   2 }, // PUNPCKLBW+WD
     { ISD::TRUNCATE,    MVT::v8i1,   MVT::v8i8,   1 }, // PUNPCKLBW
 
-    { ISD::TRUNCATE,    MVT::v2i8,   MVT::v2i16,  2 }, // PAND+PACKUSWB
-    { ISD::TRUNCATE,    MVT::v4i8,   MVT::v4i16,  2 }, // PAND+PACKUSWB
-    { ISD::TRUNCATE,    MVT::v8i8,   MVT::v8i16,  2 }, // PAND+PACKUSWB
+    { ISD::TRUNCATE,    MVT::v16i8,  MVT::v8i16,  2 }, // PAND+PACKUSWB
     { ISD::TRUNCATE,    MVT::v16i8,  MVT::v16i16, 3 },
-    { ISD::TRUNCATE,    MVT::v2i8,   MVT::v2i32,  3 }, // PAND+2*PACKUSWB
-    { ISD::TRUNCATE,    MVT::v2i16,  MVT::v2i32,  1 },
-    { ISD::TRUNCATE,    MVT::v4i8,   MVT::v4i32,  3 },
-    { ISD::TRUNCATE,    MVT::v4i16,  MVT::v4i32,  3 },
-    { ISD::TRUNCATE,    MVT::v8i8,   MVT::v8i32,  4 },
+    { ISD::TRUNCATE,    MVT::v16i8,  MVT::v4i32,  3 }, // PAND+2*PACKUSWB
     { ISD::TRUNCATE,    MVT::v16i8,  MVT::v16i32, 7 },
+    { ISD::TRUNCATE,    MVT::v2i16,  MVT::v2i32,  1 },
+    { ISD::TRUNCATE,    MVT::v8i16,  MVT::v4i32,  3 },
     { ISD::TRUNCATE,    MVT::v8i16,  MVT::v8i32,  5 },
-    { ISD::TRUNCATE,    MVT::v16i16, MVT::v16i32, 10 },
-    { ISD::TRUNCATE,    MVT::v2i8,   MVT::v2i64,  4 }, // PAND+3*PACKUSWB
-    { ISD::TRUNCATE,    MVT::v2i16,  MVT::v2i64,  2 }, // PSHUFD+PSHUFLW
-    { ISD::TRUNCATE,    MVT::v2i32,  MVT::v2i64,  1 }, // PSHUFD
+    { ISD::TRUNCATE,    MVT::v16i16, MVT::v16i32,10 },
+    { ISD::TRUNCATE,    MVT::v16i8,  MVT::v2i64,  4 }, // PAND+3*PACKUSWB
+    { ISD::TRUNCATE,    MVT::v8i16,  MVT::v2i64,  2 }, // PSHUFD+PSHUFLW
+    { ISD::TRUNCATE,    MVT::v4i32,  MVT::v2i64,  1 }, // PSHUFD
   };
 
   // Attempt to map directly to (simple) MVT types to let us match custom entries.

diff --git a/llvm/test/Analysis/CostModel/X86/arith-fix.ll b/llvm/test/Analysis/CostModel/X86/arith-fix.ll
index abab6f4..3b50f6b 100644 (file)
--- a/llvm/test/Analysis/CostModel/X86/arith-fix.ll
+++ b/llvm/test/Analysis/CostModel/X86/arith-fix.ll
@@ -38,17 +38,17 @@ define i32 @smul(i32 %arg) {
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V4I64 = call <4 x i64> @llvm.smul.fix.v4i64(<4 x i64> undef, <4 x i64> undef, i32 3)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V8I64 = call <8 x i64> @llvm.smul.fix.v8i64(<8 x i64> undef, <8 x i64> undef, i32 3)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I32 = call i32 @llvm.smul.fix.i32(i32 undef, i32 undef, i32 3)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V4I32 = call <4 x i32> @llvm.smul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 66 for instruction: %V8I32 = call <8 x i32> @llvm.smul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 132 for instruction: %V16I32 = call <16 x i32> @llvm.smul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 31 for instruction: %V4I32 = call <4 x i32> @llvm.smul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 62 for instruction: %V8I32 = call <8 x i32> @llvm.smul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 124 for instruction: %V16I32 = call <16 x i32> @llvm.smul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call i16 @llvm.smul.fix.i16(i16 undef, i16 undef, i32 3)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V8I16 = call <8 x i16> @llvm.smul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 66 for instruction: %V16I16 = call <16 x i16> @llvm.smul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 132 for instruction: %V32I16 = call <32 x i16> @llvm.smul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 140 for instruction: %V32I16 = call <32 x i16> @llvm.smul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call i8 @llvm.smul.fix.i8(i8 undef, i8 undef, i32 3)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I8 = call <16 x i8> @llvm.smul.fix.v16i8(<16 x i8> undef, <16 x i8> undef, i32 3)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V32I8 = call <32 x i8> @llvm.smul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 84 for instruction: %V64I8 = call <64 x i8> @llvm.smul.fix.v64i8(<64 x i8> undef, <64 x i8> undef, i32 3)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V32I8 = call <32 x i8> @llvm.smul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V64I8 = call <64 x i8> @llvm.smul.fix.v64i8(<64 x i8> undef, <64 x i8> undef, i32 3)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'smul'
@@ -57,17 +57,17 @@ define i32 @smul(i32 %arg) {
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V4I64 = call <4 x i64> @llvm.smul.fix.v4i64(<4 x i64> undef, <4 x i64> undef, i32 3)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V8I64 = call <8 x i64> @llvm.smul.fix.v8i64(<8 x i64> undef, <8 x i64> undef, i32 3)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I32 = call i32 @llvm.smul.fix.i32(i32 undef, i32 undef, i32 3)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I32 = call <4 x i32> @llvm.smul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V8I32 = call <8 x i32> @llvm.smul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 100 for instruction: %V16I32 = call <16 x i32> @llvm.smul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I32 = call <4 x i32> @llvm.smul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V8I32 = call <8 x i32> @llvm.smul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V16I32 = call <16 x i32> @llvm.smul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call i16 @llvm.smul.fix.i16(i16 undef, i16 undef, i32 3)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I16 = call <8 x i16> @llvm.smul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V16I16 = call <16 x i16> @llvm.smul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V32I16 = call <32 x i16> @llvm.smul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V8I16 = call <8 x i16> @llvm.smul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V16I16 = call <16 x i16> @llvm.smul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V32I16 = call <32 x i16> @llvm.smul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call i8 @llvm.smul.fix.i8(i8 undef, i8 undef, i32 3)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I8 = call <16 x i8> @llvm.smul.fix.v16i8(<16 x i8> undef, <16 x i8> undef, i32 3)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I8 = call <32 x i8> @llvm.smul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V64I8 = call <64 x i8> @llvm.smul.fix.v64i8(<64 x i8> undef, <64 x i8> undef, i32 3)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32I8 = call <32 x i8> @llvm.smul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V64I8 = call <64 x i8> @llvm.smul.fix.v64i8(<64 x i8> undef, <64 x i8> undef, i32 3)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'smul'
@@ -77,12 +77,12 @@ define i32 @smul(i32 %arg) {
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V8I64 = call <8 x i64> @llvm.smul.fix.v8i64(<8 x i64> undef, <8 x i64> undef, i32 3)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I32 = call i32 @llvm.smul.fix.i32(i32 undef, i32 undef, i32 3)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I32 = call <4 x i32> @llvm.smul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 53 for instruction: %V8I32 = call <8 x i32> @llvm.smul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 106 for instruction: %V16I32 = call <16 x i32> @llvm.smul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 57 for instruction: %V8I32 = call <8 x i32> @llvm.smul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 114 for instruction: %V16I32 = call <16 x i32> @llvm.smul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call i16 @llvm.smul.fix.i16(i16 undef, i16 undef, i32 3)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I16 = call <8 x i16> @llvm.smul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V16I16 = call <16 x i16> @llvm.smul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V32I16 = call <32 x i16> @llvm.smul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 53 for instruction: %V16I16 = call <16 x i16> @llvm.smul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 106 for instruction: %V32I16 = call <32 x i16> @llvm.smul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call i8 @llvm.smul.fix.i8(i8 undef, i8 undef, i32 3)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16I8 = call <16 x i8> @llvm.smul.fix.v16i8(<16 x i8> undef, <16 x i8> undef, i32 3)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V32I8 = call <32 x i8> @llvm.smul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
@@ -100,8 +100,8 @@ define i32 @smul(i32 %arg) {
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 62 for instruction: %V16I32 = call <16 x i32> @llvm.smul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call i16 @llvm.smul.fix.i16(i16 undef, i16 undef, i32 3)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I16 = call <8 x i16> @llvm.smul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V16I16 = call <16 x i16> @llvm.smul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V32I16 = call <32 x i16> @llvm.smul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V16I16 = call <16 x i16> @llvm.smul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 66 for instruction: %V32I16 = call <32 x i16> @llvm.smul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call i8 @llvm.smul.fix.i8(i8 undef, i8 undef, i32 3)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V16I8 = call <16 x i8> @llvm.smul.fix.v16i8(<16 x i8> undef, <16 x i8> undef, i32 3)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V32I8 = call <32 x i8> @llvm.smul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
@@ -171,17 +171,17 @@ define i32 @smul(i32 %arg) {
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V4I64 = call <4 x i64> @llvm.smul.fix.v4i64(<4 x i64> undef, <4 x i64> undef, i32 3)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V8I64 = call <8 x i64> @llvm.smul.fix.v8i64(<8 x i64> undef, <8 x i64> undef, i32 3)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I32 = call i32 @llvm.smul.fix.i32(i32 undef, i32 undef, i32 3)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I32 = call <4 x i32> @llvm.smul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 94 for instruction: %V8I32 = call <8 x i32> @llvm.smul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 188 for instruction: %V16I32 = call <16 x i32> @llvm.smul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I32 = call <4 x i32> @llvm.smul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V8I32 = call <8 x i32> @llvm.smul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 180 for instruction: %V16I32 = call <16 x i32> @llvm.smul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call i16 @llvm.smul.fix.i16(i16 undef, i16 undef, i32 3)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 35 for instruction: %V8I16 = call <8 x i16> @llvm.smul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 70 for instruction: %V16I16 = call <16 x i16> @llvm.smul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 140 for instruction: %V32I16 = call <32 x i16> @llvm.smul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 39 for instruction: %V8I16 = call <8 x i16> @llvm.smul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 78 for instruction: %V16I16 = call <16 x i16> @llvm.smul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 148 for instruction: %V32I16 = call <32 x i16> @llvm.smul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call i8 @llvm.smul.fix.i8(i8 undef, i8 undef, i32 3)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16I8 = call <16 x i8> @llvm.smul.fix.v16i8(<16 x i8> undef, <16 x i8> undef, i32 3)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32I8 = call <32 x i8> @llvm.smul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V64I8 = call <64 x i8> @llvm.smul.fix.v64i8(<64 x i8> undef, <64 x i8> undef, i32 3)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V32I8 = call <32 x i8> @llvm.smul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 84 for instruction: %V64I8 = call <64 x i8> @llvm.smul.fix.v64i8(<64 x i8> undef, <64 x i8> undef, i32 3)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'smul'
@@ -190,17 +190,17 @@ define i32 @smul(i32 %arg) {
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V4I64 = call <4 x i64> @llvm.smul.fix.v4i64(<4 x i64> undef, <4 x i64> undef, i32 3)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V8I64 = call <8 x i64> @llvm.smul.fix.v8i64(<8 x i64> undef, <8 x i64> undef, i32 3)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I32 = call i32 @llvm.smul.fix.i32(i32 undef, i32 undef, i32 3)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I32 = call <4 x i32> @llvm.smul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V8I32 = call <8 x i32> @llvm.smul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 100 for instruction: %V16I32 = call <16 x i32> @llvm.smul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I32 = call <4 x i32> @llvm.smul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V8I32 = call <8 x i32> @llvm.smul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V16I32 = call <16 x i32> @llvm.smul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call i16 @llvm.smul.fix.i16(i16 undef, i16 undef, i32 3)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I16 = call <8 x i16> @llvm.smul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V16I16 = call <16 x i16> @llvm.smul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V32I16 = call <32 x i16> @llvm.smul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V8I16 = call <8 x i16> @llvm.smul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V16I16 = call <16 x i16> @llvm.smul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V32I16 = call <32 x i16> @llvm.smul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call i8 @llvm.smul.fix.i8(i8 undef, i8 undef, i32 3)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I8 = call <16 x i8> @llvm.smul.fix.v16i8(<16 x i8> undef, <16 x i8> undef, i32 3)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I8 = call <32 x i8> @llvm.smul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V64I8 = call <64 x i8> @llvm.smul.fix.v64i8(<64 x i8> undef, <64 x i8> undef, i32 3)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32I8 = call <32 x i8> @llvm.smul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V64I8 = call <64 x i8> @llvm.smul.fix.v64i8(<64 x i8> undef, <64 x i8> undef, i32 3)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'smul'
@@ -210,12 +210,12 @@ define i32 @smul(i32 %arg) {
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V8I64 = call <8 x i64> @llvm.smul.fix.v8i64(<8 x i64> undef, <8 x i64> undef, i32 3)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I32 = call i32 @llvm.smul.fix.i32(i32 undef, i32 undef, i32 3)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I32 = call <4 x i32> @llvm.smul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 53 for instruction: %V8I32 = call <8 x i32> @llvm.smul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 106 for instruction: %V16I32 = call <16 x i32> @llvm.smul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 57 for instruction: %V8I32 = call <8 x i32> @llvm.smul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 114 for instruction: %V16I32 = call <16 x i32> @llvm.smul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call i16 @llvm.smul.fix.i16(i16 undef, i16 undef, i32 3)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I16 = call <8 x i16> @llvm.smul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V16I16 = call <16 x i16> @llvm.smul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V32I16 = call <32 x i16> @llvm.smul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 53 for instruction: %V16I16 = call <16 x i16> @llvm.smul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 106 for instruction: %V32I16 = call <32 x i16> @llvm.smul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call i8 @llvm.smul.fix.i8(i8 undef, i8 undef, i32 3)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16I8 = call <16 x i8> @llvm.smul.fix.v16i8(<16 x i8> undef, <16 x i8> undef, i32 3)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V32I8 = call <32 x i8> @llvm.smul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
@@ -272,17 +272,17 @@ define i32 @umul(i32 %arg) {
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V4I64 = call <4 x i64> @llvm.umul.fix.v4i64(<4 x i64> undef, <4 x i64> undef, i32 3)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V8I64 = call <8 x i64> @llvm.umul.fix.v8i64(<8 x i64> undef, <8 x i64> undef, i32 3)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.umul.fix.i32(i32 undef, i32 undef, i32 3)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V4I32 = call <4 x i32> @llvm.umul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 58 for instruction: %V8I32 = call <8 x i32> @llvm.umul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 116 for instruction: %V16I32 = call <16 x i32> @llvm.umul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V4I32 = call <4 x i32> @llvm.umul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 54 for instruction: %V8I32 = call <8 x i32> @llvm.umul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 108 for instruction: %V16I32 = call <16 x i32> @llvm.umul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call i16 @llvm.umul.fix.i16(i16 undef, i16 undef, i32 3)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V8I16 = call <8 x i16> @llvm.umul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 58 for instruction: %V16I16 = call <16 x i16> @llvm.umul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 116 for instruction: %V32I16 = call <32 x i16> @llvm.umul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 124 for instruction: %V32I16 = call <32 x i16> @llvm.umul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call i8 @llvm.umul.fix.i8(i8 undef, i8 undef, i32 3)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I8 = call <16 x i8> @llvm.umul.fix.v16i8(<16 x i8> undef, <16 x i8> undef, i32 3)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I8 = call <32 x i8> @llvm.umul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V64I8 = call <64 x i8> @llvm.umul.fix.v64i8(<64 x i8> undef, <64 x i8> undef, i32 3)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32I8 = call <32 x i8> @llvm.umul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V64I8 = call <64 x i8> @llvm.umul.fix.v64i8(<64 x i8> undef, <64 x i8> undef, i32 3)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'umul'
@@ -291,17 +291,17 @@ define i32 @umul(i32 %arg) {
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V4I64 = call <4 x i64> @llvm.umul.fix.v4i64(<4 x i64> undef, <4 x i64> undef, i32 3)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V8I64 = call <8 x i64> @llvm.umul.fix.v8i64(<8 x i64> undef, <8 x i64> undef, i32 3)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.umul.fix.i32(i32 undef, i32 undef, i32 3)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I32 = call <4 x i32> @llvm.umul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V8I32 = call <8 x i32> @llvm.umul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 100 for instruction: %V16I32 = call <16 x i32> @llvm.umul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I32 = call <4 x i32> @llvm.umul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V8I32 = call <8 x i32> @llvm.umul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V16I32 = call <16 x i32> @llvm.umul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call i16 @llvm.umul.fix.i16(i16 undef, i16 undef, i32 3)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I16 = call <8 x i16> @llvm.umul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V16I16 = call <16 x i16> @llvm.umul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V32I16 = call <32 x i16> @llvm.umul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V8I16 = call <8 x i16> @llvm.umul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V16I16 = call <16 x i16> @llvm.umul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V32I16 = call <32 x i16> @llvm.umul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call i8 @llvm.umul.fix.i8(i8 undef, i8 undef, i32 3)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I8 = call <16 x i8> @llvm.umul.fix.v16i8(<16 x i8> undef, <16 x i8> undef, i32 3)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I8 = call <32 x i8> @llvm.umul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V64I8 = call <64 x i8> @llvm.umul.fix.v64i8(<64 x i8> undef, <64 x i8> undef, i32 3)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32I8 = call <32 x i8> @llvm.umul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V64I8 = call <64 x i8> @llvm.umul.fix.v64i8(<64 x i8> undef, <64 x i8> undef, i32 3)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'umul'
@@ -311,12 +311,12 @@ define i32 @umul(i32 %arg) {
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V8I64 = call <8 x i64> @llvm.umul.fix.v8i64(<8 x i64> undef, <8 x i64> undef, i32 3)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.umul.fix.i32(i32 undef, i32 undef, i32 3)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I32 = call <4 x i32> @llvm.umul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 53 for instruction: %V8I32 = call <8 x i32> @llvm.umul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 106 for instruction: %V16I32 = call <16 x i32> @llvm.umul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 57 for instruction: %V8I32 = call <8 x i32> @llvm.umul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 114 for instruction: %V16I32 = call <16 x i32> @llvm.umul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call i16 @llvm.umul.fix.i16(i16 undef, i16 undef, i32 3)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I16 = call <8 x i16> @llvm.umul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V16I16 = call <16 x i16> @llvm.umul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V32I16 = call <32 x i16> @llvm.umul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 53 for instruction: %V16I16 = call <16 x i16> @llvm.umul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 106 for instruction: %V32I16 = call <32 x i16> @llvm.umul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call i8 @llvm.umul.fix.i8(i8 undef, i8 undef, i32 3)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16I8 = call <16 x i8> @llvm.umul.fix.v16i8(<16 x i8> undef, <16 x i8> undef, i32 3)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V32I8 = call <32 x i8> @llvm.umul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
@@ -334,8 +334,8 @@ define i32 @umul(i32 %arg) {
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 62 for instruction: %V16I32 = call <16 x i32> @llvm.umul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call i16 @llvm.umul.fix.i16(i16 undef, i16 undef, i32 3)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I16 = call <8 x i16> @llvm.umul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V16I16 = call <16 x i16> @llvm.umul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V32I16 = call <32 x i16> @llvm.umul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V16I16 = call <16 x i16> @llvm.umul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 66 for instruction: %V32I16 = call <32 x i16> @llvm.umul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call i8 @llvm.umul.fix.i8(i8 undef, i8 undef, i32 3)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V16I8 = call <16 x i8> @llvm.umul.fix.v16i8(<16 x i8> undef, <16 x i8> undef, i32 3)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V32I8 = call <32 x i8> @llvm.umul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
@@ -405,17 +405,17 @@ define i32 @umul(i32 %arg) {
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V4I64 = call <4 x i64> @llvm.umul.fix.v4i64(<4 x i64> undef, <4 x i64> undef, i32 3)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V8I64 = call <8 x i64> @llvm.umul.fix.v8i64(<8 x i64> undef, <8 x i64> undef, i32 3)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.umul.fix.i32(i32 undef, i32 undef, i32 3)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I32 = call <4 x i32> @llvm.umul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 94 for instruction: %V8I32 = call <8 x i32> @llvm.umul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 188 for instruction: %V16I32 = call <16 x i32> @llvm.umul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I32 = call <4 x i32> @llvm.umul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V8I32 = call <8 x i32> @llvm.umul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 180 for instruction: %V16I32 = call <16 x i32> @llvm.umul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call i16 @llvm.umul.fix.i16(i16 undef, i16 undef, i32 3)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 35 for instruction: %V8I16 = call <8 x i16> @llvm.umul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 70 for instruction: %V16I16 = call <16 x i16> @llvm.umul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 140 for instruction: %V32I16 = call <32 x i16> @llvm.umul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 39 for instruction: %V8I16 = call <8 x i16> @llvm.umul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 78 for instruction: %V16I16 = call <16 x i16> @llvm.umul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 148 for instruction: %V32I16 = call <32 x i16> @llvm.umul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call i8 @llvm.umul.fix.i8(i8 undef, i8 undef, i32 3)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16I8 = call <16 x i8> @llvm.umul.fix.v16i8(<16 x i8> undef, <16 x i8> undef, i32 3)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32I8 = call <32 x i8> @llvm.umul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V64I8 = call <64 x i8> @llvm.umul.fix.v64i8(<64 x i8> undef, <64 x i8> undef, i32 3)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V32I8 = call <32 x i8> @llvm.umul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 84 for instruction: %V64I8 = call <64 x i8> @llvm.umul.fix.v64i8(<64 x i8> undef, <64 x i8> undef, i32 3)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'umul'
@@ -424,17 +424,17 @@ define i32 @umul(i32 %arg) {
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V4I64 = call <4 x i64> @llvm.umul.fix.v4i64(<4 x i64> undef, <4 x i64> undef, i32 3)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V8I64 = call <8 x i64> @llvm.umul.fix.v8i64(<8 x i64> undef, <8 x i64> undef, i32 3)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.umul.fix.i32(i32 undef, i32 undef, i32 3)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I32 = call <4 x i32> @llvm.umul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V8I32 = call <8 x i32> @llvm.umul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 100 for instruction: %V16I32 = call <16 x i32> @llvm.umul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I32 = call <4 x i32> @llvm.umul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V8I32 = call <8 x i32> @llvm.umul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V16I32 = call <16 x i32> @llvm.umul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call i16 @llvm.umul.fix.i16(i16 undef, i16 undef, i32 3)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I16 = call <8 x i16> @llvm.umul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V16I16 = call <16 x i16> @llvm.umul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V32I16 = call <32 x i16> @llvm.umul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V8I16 = call <8 x i16> @llvm.umul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V16I16 = call <16 x i16> @llvm.umul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V32I16 = call <32 x i16> @llvm.umul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call i8 @llvm.umul.fix.i8(i8 undef, i8 undef, i32 3)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I8 = call <16 x i8> @llvm.umul.fix.v16i8(<16 x i8> undef, <16 x i8> undef, i32 3)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I8 = call <32 x i8> @llvm.umul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V64I8 = call <64 x i8> @llvm.umul.fix.v64i8(<64 x i8> undef, <64 x i8> undef, i32 3)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32I8 = call <32 x i8> @llvm.umul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V64I8 = call <64 x i8> @llvm.umul.fix.v64i8(<64 x i8> undef, <64 x i8> undef, i32 3)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'umul'
@@ -444,12 +444,12 @@ define i32 @umul(i32 %arg) {
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V8I64 = call <8 x i64> @llvm.umul.fix.v8i64(<8 x i64> undef, <8 x i64> undef, i32 3)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I32 = call i32 @llvm.umul.fix.i32(i32 undef, i32 undef, i32 3)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I32 = call <4 x i32> @llvm.umul.fix.v4i32(<4 x i32> undef, <4 x i32> undef, i32 3)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 53 for instruction: %V8I32 = call <8 x i32> @llvm.umul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 106 for instruction: %V16I32 = call <16 x i32> @llvm.umul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 57 for instruction: %V8I32 = call <8 x i32> @llvm.umul.fix.v8i32(<8 x i32> undef, <8 x i32> undef, i32 3)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 114 for instruction: %V16I32 = call <16 x i32> @llvm.umul.fix.v16i32(<16 x i32> undef, <16 x i32> undef, i32 3)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I16 = call i16 @llvm.umul.fix.i16(i16 undef, i16 undef, i32 3)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I16 = call <8 x i16> @llvm.umul.fix.v8i16(<8 x i16> undef, <8 x i16> undef, i32 3)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V16I16 = call <16 x i16> @llvm.umul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V32I16 = call <32 x i16> @llvm.umul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 53 for instruction: %V16I16 = call <16 x i16> @llvm.umul.fix.v16i16(<16 x i16> undef, <16 x i16> undef, i32 3)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 106 for instruction: %V32I16 = call <32 x i16> @llvm.umul.fix.v32i16(<32 x i16> undef, <32 x i16> undef, i32 3)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = call i8 @llvm.umul.fix.i8(i8 undef, i8 undef, i32 3)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16I8 = call <16 x i8> @llvm.umul.fix.v16i8(<16 x i8> undef, <16 x i8> undef, i32 3)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V32I8 = call <32 x i8> @llvm.umul.fix.v32i8(<32 x i8> undef, <32 x i8> undef, i32 3)

diff --git a/llvm/test/Analysis/CostModel/X86/arith-overflow.ll b/llvm/test/Analysis/CostModel/X86/arith-overflow.ll
index 1f5d773..81bf82e 100644 (file)
--- a/llvm/test/Analysis/CostModel/X86/arith-overflow.ll
+++ b/llvm/test/Analysis/CostModel/X86/arith-overflow.ll
@@ -994,17 +994,17 @@ define i32 @smul(i32 %arg) {
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 58 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.smul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 116 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.smul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.smul.with.overflow.i32(i32 undef, i32 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.smul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 58 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 116 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 27 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.smul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 54 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 108 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.smul.with.overflow.i16(i16 undef, i16 undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.smul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 58 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 116 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 124 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.smul.with.overflow.i8(i8 undef, i8 undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.smul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.smul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 84 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.smul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'smul'
@@ -1013,17 +1013,17 @@ define i32 @smul(i32 %arg) {
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.smul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.smul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.smul.with.overflow.i32(i32 undef, i32 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.smul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 100 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.smul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.smul.with.overflow.i16(i16 undef, i16 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.smul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.smul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.smul.with.overflow.i8(i8 undef, i8 undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.smul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.smul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 84 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.smul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'smul'
@@ -1033,12 +1033,12 @@ define i32 @smul(i32 %arg) {
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 108 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.smul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.smul.with.overflow.i32(i32 undef, i32 undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.smul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 56 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 112 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 60 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 120 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.smul.with.overflow.i16(i16 undef, i16 undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.smul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 56 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 112 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.smul.with.overflow.i8(i8 undef, i8 undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.smul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 52 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
@@ -1056,8 +1056,8 @@ define i32 @smul(i32 %arg) {
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 62 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.smul.with.overflow.i16(i16 undef, i16 undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.smul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 66 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.smul.with.overflow.i8(i8 undef, i8 undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.smul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 29 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
@@ -1127,17 +1127,17 @@ define i32 @smul(i32 %arg) {
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.smul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.smul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.smul.with.overflow.i32(i32 undef, i32 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 47 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.smul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 94 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 188 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 45 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.smul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 90 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 180 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.smul.with.overflow.i16(i16 undef, i16 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 35 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.smul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 70 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 140 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 39 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.smul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 78 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 148 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.smul.with.overflow.i8(i8 undef, i8 undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.smul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 84 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.smul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.smul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'smul'
@@ -1146,17 +1146,17 @@ define i32 @smul(i32 %arg) {
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.smul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.smul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.smul.with.overflow.i32(i32 undef, i32 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.smul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 100 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.smul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.smul.with.overflow.i16(i16 undef, i16 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.smul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.smul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.smul.with.overflow.i8(i8 undef, i8 undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.smul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.smul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 84 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.smul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'smul'
@@ -1166,12 +1166,12 @@ define i32 @smul(i32 %arg) {
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 108 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.smul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.smul.with.overflow.i32(i32 undef, i32 undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.smul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 56 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 112 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 60 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.smul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 120 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.smul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.smul.with.overflow.i16(i16 undef, i16 undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.smul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 56 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.smul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 112 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.smul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.smul.with.overflow.i8(i8 undef, i8 undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.smul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 52 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.smul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
@@ -1232,17 +1232,17 @@ define i32 @umul(i32 %arg) {
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 50 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.umul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 100 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.umul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.umul.with.overflow.i32(i32 undef, i32 undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.umul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 56 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 112 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.umul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 52 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 104 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.umul.with.overflow.i16(i16 undef, i16 undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.umul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 56 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 112 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 120 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.umul.with.overflow.i8(i8 undef, i8 undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.umul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 60 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.umul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.umul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'umul'
@@ -1251,17 +1251,17 @@ define i32 @umul(i32 %arg) {
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 36 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.umul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 72 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.umul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.umul.with.overflow.i32(i32 undef, i32 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.umul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.umul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.umul.with.overflow.i16(i16 undef, i16 undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.umul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.umul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 72 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.umul.with.overflow.i8(i8 undef, i8 undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.umul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 60 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.umul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.umul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'umul'
@@ -1271,12 +1271,12 @@ define i32 @umul(i32 %arg) {
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.umul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.umul.with.overflow.i32(i32 undef, i32 undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.umul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 52 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 104 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 56 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 112 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.umul.with.overflow.i16(i16 undef, i16 undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.umul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 52 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 104 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.umul.with.overflow.i8(i8 undef, i8 undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.umul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
@@ -1294,8 +1294,8 @@ define i32 @umul(i32 %arg) {
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 60 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.umul.with.overflow.i16(i16 undef, i16 undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.umul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.umul.with.overflow.i8(i8 undef, i8 undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.umul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 25 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
@@ -1365,17 +1365,17 @@ define i32 @umul(i32 %arg) {
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.umul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.umul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.umul.with.overflow.i32(i32 undef, i32 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.umul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 92 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 184 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.umul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 176 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.umul.with.overflow.i16(i16 undef, i16 undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.umul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 136 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.umul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 144 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.umul.with.overflow.i8(i8 undef, i8 undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.umul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; SLM-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.umul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; SLM-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.umul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'umul'
@@ -1384,17 +1384,17 @@ define i32 @umul(i32 %arg) {
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 36 for instruction: %V4I64 = call { <4 x i64>, <4 x i1> } @llvm.umul.with.overflow.v4i64(<4 x i64> undef, <4 x i64> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 72 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.umul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.umul.with.overflow.i32(i32 undef, i32 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.umul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 96 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.umul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.umul.with.overflow.i16(i16 undef, i16 undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.umul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.umul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 72 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.umul.with.overflow.i8(i8 undef, i8 undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.umul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
-; GLM-NEXT:  Cost Model: Found an estimated cost of 60 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.umul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)
+; GLM-NEXT:  Cost Model: Found an estimated cost of 68 for instruction: %V64I8 = call { <64 x i8>, <64 x i1> } @llvm.umul.with.overflow.v64i8(<64 x i8> undef, <64 x i8> undef)
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'umul'
@@ -1404,12 +1404,12 @@ define i32 @umul(i32 %arg) {
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V8I64 = call { <8 x i64>, <8 x i1> } @llvm.umul.with.overflow.v8i64(<8 x i64> undef, <8 x i64> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I32 = call { i32, i1 } @llvm.umul.with.overflow.i32(i32 undef, i32 undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V4I32 = call { <4 x i32>, <4 x i1> } @llvm.umul.with.overflow.v4i32(<4 x i32> undef, <4 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 52 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 104 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 56 for instruction: %V8I32 = call { <8 x i32>, <8 x i1> } @llvm.umul.with.overflow.v8i32(<8 x i32> undef, <8 x i32> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 112 for instruction: %V16I32 = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> undef, <16 x i32> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I16 = call { i16, i1 } @llvm.umul.with.overflow.i16(i16 undef, i16 undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V8I16 = call { <8 x i16>, <8 x i1> } @llvm.umul.with.overflow.v8i16(<8 x i16> undef, <8 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 52 for instruction: %V16I16 = call { <16 x i16>, <16 x i1> } @llvm.umul.with.overflow.v16i16(<16 x i16> undef, <16 x i16> undef)
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 104 for instruction: %V32I16 = call { <32 x i16>, <32 x i1> } @llvm.umul.with.overflow.v32i16(<32 x i16> undef, <32 x i16> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %I8 = call { i8, i1 } @llvm.umul.with.overflow.i8(i8 undef, i8 undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V16I8 = call { <16 x i8>, <16 x i1> } @llvm.umul.with.overflow.v16i8(<16 x i8> undef, <16 x i8> undef)
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V32I8 = call { <32 x i8>, <32 x i1> } @llvm.umul.with.overflow.v32i8(<32 x i8> undef, <32 x i8> undef)

diff --git a/llvm/test/Analysis/CostModel/X86/arith.ll b/llvm/test/Analysis/CostModel/X86/arith.ll
index ed679a3..40f7815 100644 (file)
--- a/llvm/test/Analysis/CostModel/X86/arith.ll
+++ b/llvm/test/Analysis/CostModel/X86/arith.ll
@@ -947,8 +947,8 @@ define i32 @mul(i32 %arg) {
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = mul <4 x i8> undef, undef
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I8 = mul <8 x i8> undef, undef
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = mul <16 x i8> undef, undef
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V32I8 = mul <32 x i8> undef, undef
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V64I8 = mul <64 x i8> undef, undef
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32I8 = mul <32 x i8> undef, undef
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V64I8 = mul <64 x i8> undef, undef
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'mul'
@@ -965,12 +965,12 @@ define i32 @mul(i32 %arg) {
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = mul <16 x i16> undef, undef
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = mul <32 x i16> undef, undef
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = mul i8 undef, undef
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I8 = mul <2 x i8> undef, undef
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I8 = mul <4 x i8> undef, undef
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I8 = mul <8 x i8> undef, undef
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = mul <2 x i8> undef, undef
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = mul <4 x i8> undef, undef
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I8 = mul <8 x i8> undef, undef
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = mul <16 x i8> undef, undef
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V32I8 = mul <32 x i8> undef, undef
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V64I8 = mul <64 x i8> undef, undef
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32I8 = mul <32 x i8> undef, undef
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V64I8 = mul <64 x i8> undef, undef
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'mul'
@@ -987,9 +987,9 @@ define i32 @mul(i32 %arg) {
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = mul <16 x i16> undef, undef
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = mul <32 x i16> undef, undef
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = mul i8 undef, undef
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I8 = mul <2 x i8> undef, undef
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I8 = mul <4 x i8> undef, undef
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I8 = mul <8 x i8> undef, undef
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = mul <2 x i8> undef, undef
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = mul <4 x i8> undef, undef
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I8 = mul <8 x i8> undef, undef
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = mul <16 x i8> undef, undef
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V32I8 = mul <32 x i8> undef, undef
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V64I8 = mul <64 x i8> undef, undef
@@ -1097,12 +1097,12 @@ define i32 @mul(i32 %arg) {
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = mul <16 x i16> undef, undef
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = mul <32 x i16> undef, undef
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = mul i8 undef, undef
-; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = mul <2 x i8> undef, undef
-; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = mul <4 x i8> undef, undef
-; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I8 = mul <8 x i8> undef, undef
+; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V2I8 = mul <2 x i8> undef, undef
+; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I8 = mul <4 x i8> undef, undef
+; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I8 = mul <8 x i8> undef, undef
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = mul <16 x i8> undef, undef
-; SLM-NEXT:  Cost Model: Found an estimated cost of 18 for instruction: %V32I8 = mul <32 x i8> undef, undef
-; SLM-NEXT:  Cost Model: Found an estimated cost of 36 for instruction: %V64I8 = mul <64 x i8> undef, undef
+; SLM-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V32I8 = mul <32 x i8> undef, undef
+; SLM-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V64I8 = mul <64 x i8> undef, undef
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'mul'
@@ -1119,12 +1119,12 @@ define i32 @mul(i32 %arg) {
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16I16 = mul <16 x i16> undef, undef
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V32I16 = mul <32 x i16> undef, undef
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = mul i8 undef, undef
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I8 = mul <2 x i8> undef, undef
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I8 = mul <4 x i8> undef, undef
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I8 = mul <8 x i8> undef, undef
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = mul <2 x i8> undef, undef
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = mul <4 x i8> undef, undef
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I8 = mul <8 x i8> undef, undef
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16I8 = mul <16 x i8> undef, undef
-; GLM-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V32I8 = mul <32 x i8> undef, undef
-; GLM-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V64I8 = mul <64 x i8> undef, undef
+; GLM-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32I8 = mul <32 x i8> undef, undef
+; GLM-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V64I8 = mul <64 x i8> undef, undef
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'mul'
@@ -1141,9 +1141,9 @@ define i32 @mul(i32 %arg) {
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16I16 = mul <16 x i16> undef, undef
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32I16 = mul <32 x i16> undef, undef
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = mul i8 undef, undef
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I8 = mul <2 x i8> undef, undef
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I8 = mul <4 x i8> undef, undef
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I8 = mul <8 x i8> undef, undef
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = mul <2 x i8> undef, undef
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = mul <4 x i8> undef, undef
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8I8 = mul <8 x i8> undef, undef
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V16I8 = mul <16 x i8> undef, undef
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V32I8 = mul <32 x i8> undef, undef
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V64I8 = mul <64 x i8> undef, undef

diff --git a/llvm/test/Analysis/CostModel/X86/cast.ll b/llvm/test/Analysis/CostModel/X86/cast.ll
index 7343921..087fef4 100644 (file)
--- a/llvm/test/Analysis/CostModel/X86/cast.ll
+++ b/llvm/test/Analysis/CostModel/X86/cast.ll
@@ -14,7 +14,7 @@ define i32 @add(i32 %arg) {
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %C = trunc <4 x i32> undef to <4 x i1>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %D = zext <8 x i1> undef to <8 x i32>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %E = sext <8 x i1> undef to <8 x i32>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %F = trunc <8 x i32> undef to <8 x i1>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %F = trunc <8 x i32> undef to <8 x i1>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %G = zext i1 undef to i32
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %H = trunc i32 undef to i1
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
@@ -25,7 +25,7 @@ define i32 @add(i32 %arg) {
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %C = trunc <4 x i32> undef to <4 x i1>
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %D = zext <8 x i1> undef to <8 x i32>
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %E = sext <8 x i1> undef to <8 x i32>
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %F = trunc <8 x i32> undef to <8 x i1>
+; SSE41-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %F = trunc <8 x i32> undef to <8 x i1>
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %G = zext i1 undef to i32
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %H = trunc i32 undef to i1
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
@@ -101,12 +101,12 @@ define i32 @zext_sext(<8 x i1> %in) {
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %D3 = zext <16 x i16> undef to <16 x i32>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %D4 = zext <16 x i8> undef to <16 x i32>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %D5 = zext <16 x i1> undef to <16 x i32>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %E = trunc <4 x i64> undef to <4 x i32>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %E = trunc <4 x i64> undef to <4 x i32>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %F = trunc <8 x i32> undef to <8 x i16>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %F1 = trunc <16 x i16> undef to <16 x i8>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %F2 = trunc <8 x i32> undef to <8 x i8>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %F3 = trunc <4 x i64> undef to <4 x i8>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %G = trunc <8 x i64> undef to <8 x i32>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %F2 = trunc <8 x i32> undef to <8 x i8>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %F3 = trunc <4 x i64> undef to <4 x i8>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %G = trunc <8 x i64> undef to <8 x i32>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %G1 = trunc <16 x i32> undef to <16 x i16>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %G2 = trunc <16 x i32> undef to <16 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
@@ -131,13 +131,13 @@ define i32 @zext_sext(<8 x i1> %in) {
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %D3 = zext <16 x i16> undef to <16 x i32>
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %D4 = zext <16 x i8> undef to <16 x i32>
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %D5 = zext <16 x i1> undef to <16 x i32>
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %E = trunc <4 x i64> undef to <4 x i32>
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %F = trunc <8 x i32> undef to <8 x i16>
+; SSE41-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %E = trunc <4 x i64> undef to <4 x i32>
+; SSE41-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %F = trunc <8 x i32> undef to <8 x i16>
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %F1 = trunc <16 x i16> undef to <16 x i8>
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %F2 = trunc <8 x i32> undef to <8 x i8>
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %F3 = trunc <4 x i64> undef to <4 x i8>
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %G = trunc <8 x i64> undef to <8 x i32>
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %G1 = trunc <16 x i32> undef to <16 x i16>
+; SSE41-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %F2 = trunc <8 x i32> undef to <8 x i8>
+; SSE41-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %F3 = trunc <4 x i64> undef to <4 x i8>
+; SSE41-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %G = trunc <8 x i64> undef to <8 x i32>
+; SSE41-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %G1 = trunc <16 x i32> undef to <16 x i16>
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %G2 = trunc <16 x i32> undef to <16 x i8>
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
@@ -164,10 +164,10 @@ define i32 @zext_sext(<8 x i1> %in) {
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %E = trunc <4 x i64> undef to <4 x i32>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %F = trunc <8 x i32> undef to <8 x i16>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %F1 = trunc <16 x i16> undef to <16 x i8>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %F2 = trunc <8 x i32> undef to <8 x i8>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %F3 = trunc <4 x i64> undef to <4 x i8>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %G = trunc <8 x i64> undef to <8 x i32>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %G1 = trunc <16 x i32> undef to <16 x i16>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %F2 = trunc <8 x i32> undef to <8 x i8>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %F3 = trunc <4 x i64> undef to <4 x i8>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %G = trunc <8 x i64> undef to <8 x i32>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %G1 = trunc <16 x i32> undef to <16 x i16>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %G2 = trunc <16 x i32> undef to <16 x i8>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
@@ -197,7 +197,7 @@ define i32 @zext_sext(<8 x i1> %in) {
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %F2 = trunc <8 x i32> undef to <8 x i8>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %F3 = trunc <4 x i64> undef to <4 x i8>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %G = trunc <8 x i64> undef to <8 x i32>
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %G1 = trunc <16 x i32> undef to <16 x i16>
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %G1 = trunc <16 x i32> undef to <16 x i16>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %G2 = trunc <16 x i32> undef to <16 x i8>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;

diff --git a/llvm/test/Analysis/CostModel/X86/fptosi.ll b/llvm/test/Analysis/CostModel/X86/fptosi.ll
index 904320b..a5aa5bd 100644 (file)
--- a/llvm/test/Analysis/CostModel/X86/fptosi.ll
+++ b/llvm/test/Analysis/CostModel/X86/fptosi.ll
@@ -121,8 +121,8 @@ define i32 @fptosi_double_i16(i32 %arg) {
 ; SSE42-LABEL: 'fptosi_double_i16'
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = fptosi double undef to i16
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I16 = fptosi <2 x double> undef to <2 x i16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I16 = fptosi <4 x double> undef to <4 x i16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = fptosi <8 x double> undef to <8 x i16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I16 = fptosi <4 x double> undef to <4 x i16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I16 = fptosi <8 x double> undef to <8 x i16>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'fptosi_double_i16'
@@ -149,8 +149,8 @@ define i32 @fptosi_double_i16(i32 %arg) {
 ; SLM-LABEL: 'fptosi_double_i16'
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = fptosi double undef to i16
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I16 = fptosi <2 x double> undef to <2 x i16>
-; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I16 = fptosi <4 x double> undef to <4 x i16>
-; SLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = fptosi <8 x double> undef to <8 x i16>
+; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I16 = fptosi <4 x double> undef to <4 x i16>
+; SLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I16 = fptosi <8 x double> undef to <8 x i16>
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
   %I16 = fptosi double undef to i16
@@ -165,21 +165,21 @@ define i32 @fptosi_double_i8(i32 %arg) {
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = fptosi double undef to i8
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = fptosi <2 x double> undef to <2 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I8 = fptosi <4 x double> undef to <4 x i8>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I8 = fptosi <8 x double> undef to <8 x i8>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I8 = fptosi <8 x double> undef to <8 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'fptosi_double_i8'
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = fptosi double undef to i8
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I8 = fptosi <2 x double> undef to <2 x i8>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = fptosi <4 x double> undef to <4 x i8>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I8 = fptosi <8 x double> undef to <8 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I8 = fptosi <4 x double> undef to <4 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I8 = fptosi <8 x double> undef to <8 x i8>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'fptosi_double_i8'
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = fptosi double undef to i8
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I8 = fptosi <2 x double> undef to <2 x i8>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I8 = fptosi <4 x double> undef to <4 x i8>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I8 = fptosi <8 x double> undef to <8 x i8>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I8 = fptosi <8 x double> undef to <8 x i8>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'fptosi_double_i8'
@@ -199,8 +199,8 @@ define i32 @fptosi_double_i8(i32 %arg) {
 ; SLM-LABEL: 'fptosi_double_i8'
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = fptosi double undef to i8
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I8 = fptosi <2 x double> undef to <2 x i8>
-; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = fptosi <4 x double> undef to <4 x i8>
-; SLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I8 = fptosi <8 x double> undef to <8 x i8>
+; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I8 = fptosi <4 x double> undef to <4 x i8>
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I8 = fptosi <8 x double> undef to <8 x i8>
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
   %I8 = fptosi double undef to i8
@@ -337,8 +337,8 @@ define i32 @fptosi_float_i16(i32 %arg) {
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = fptosi float undef to i16
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I16 = fptosi <2 x float> undef to <2 x i16>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I16 = fptosi <4 x float> undef to <4 x i16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = fptosi <8 x float> undef to <8 x i16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16I16 = fptosi <16 x float> undef to <16 x i16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = fptosi <8 x float> undef to <8 x i16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V16I16 = fptosi <16 x float> undef to <16 x i16>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'fptosi_float_i16'
@@ -346,7 +346,7 @@ define i32 @fptosi_float_i16(i32 %arg) {
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I16 = fptosi <2 x float> undef to <2 x i16>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I16 = fptosi <4 x float> undef to <4 x i16>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I16 = fptosi <8 x float> undef to <8 x i16>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16I16 = fptosi <16 x float> undef to <16 x i16>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V16I16 = fptosi <16 x float> undef to <16 x i16>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'fptosi_float_i16'
@@ -354,7 +354,7 @@ define i32 @fptosi_float_i16(i32 %arg) {
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I16 = fptosi <2 x float> undef to <2 x i16>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I16 = fptosi <4 x float> undef to <4 x i16>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I16 = fptosi <8 x float> undef to <8 x i16>
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I16 = fptosi <16 x float> undef to <16 x i16>
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I16 = fptosi <16 x float> undef to <16 x i16>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX512-LABEL: 'fptosi_float_i16'
@@ -369,8 +369,8 @@ define i32 @fptosi_float_i16(i32 %arg) {
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = fptosi float undef to i16
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I16 = fptosi <2 x float> undef to <2 x i16>
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I16 = fptosi <4 x float> undef to <4 x i16>
-; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = fptosi <8 x float> undef to <8 x i16>
-; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16I16 = fptosi <16 x float> undef to <16 x i16>
+; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = fptosi <8 x float> undef to <8 x i16>
+; SLM-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V16I16 = fptosi <16 x float> undef to <16 x i16>
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
   %I16 = fptosi float undef to i16
@@ -386,7 +386,7 @@ define i32 @fptosi_float_i8(i32 %arg) {
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = fptosi float undef to i8
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = fptosi <2 x float> undef to <2 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I8 = fptosi <4 x float> undef to <4 x i8>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I8 = fptosi <8 x float> undef to <8 x i8>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I8 = fptosi <8 x float> undef to <8 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I8 = fptosi <16 x float> undef to <16 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
@@ -394,7 +394,7 @@ define i32 @fptosi_float_i8(i32 %arg) {
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = fptosi float undef to i8
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I8 = fptosi <2 x float> undef to <2 x i8>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I8 = fptosi <4 x float> undef to <4 x i8>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I8 = fptosi <8 x float> undef to <8 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I8 = fptosi <8 x float> undef to <8 x i8>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I8 = fptosi <16 x float> undef to <16 x i8>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
@@ -426,7 +426,7 @@ define i32 @fptosi_float_i8(i32 %arg) {
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = fptosi float undef to i8
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I8 = fptosi <2 x float> undef to <2 x i8>
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I8 = fptosi <4 x float> undef to <4 x i8>
-; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I8 = fptosi <8 x float> undef to <8 x i8>
+; SLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I8 = fptosi <8 x float> undef to <8 x i8>
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I8 = fptosi <16 x float> undef to <16 x i8>
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;

diff --git a/llvm/test/Analysis/CostModel/X86/fptoui.ll b/llvm/test/Analysis/CostModel/X86/fptoui.ll
index d09a74b..ed1c420 100644 (file)
--- a/llvm/test/Analysis/CostModel/X86/fptoui.ll
+++ b/llvm/test/Analysis/CostModel/X86/fptoui.ll
@@ -121,8 +121,8 @@ define i32 @fptoui_double_i16(i32 %arg) {
 ; SSE42-LABEL: 'fptoui_double_i16'
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = fptoui double undef to i16
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I16 = fptoui <2 x double> undef to <2 x i16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I16 = fptoui <4 x double> undef to <4 x i16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = fptoui <8 x double> undef to <8 x i16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I16 = fptoui <4 x double> undef to <4 x i16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I16 = fptoui <8 x double> undef to <8 x i16>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'fptoui_double_i16'
@@ -149,8 +149,8 @@ define i32 @fptoui_double_i16(i32 %arg) {
 ; SLM-LABEL: 'fptoui_double_i16'
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = fptoui double undef to i16
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I16 = fptoui <2 x double> undef to <2 x i16>
-; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I16 = fptoui <4 x double> undef to <4 x i16>
-; SLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I16 = fptoui <8 x double> undef to <8 x i16>
+; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I16 = fptoui <4 x double> undef to <4 x i16>
+; SLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8I16 = fptoui <8 x double> undef to <8 x i16>
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
   %I16 = fptoui double undef to i16
@@ -165,21 +165,21 @@ define i32 @fptoui_double_i8(i32 %arg) {
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = fptoui double undef to i8
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = fptoui <2 x double> undef to <2 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V4I8 = fptoui <4 x double> undef to <4 x i8>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I8 = fptoui <8 x double> undef to <8 x i8>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V8I8 = fptoui <8 x double> undef to <8 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'fptoui_double_i8'
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = fptoui double undef to i8
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I8 = fptoui <2 x double> undef to <2 x i8>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = fptoui <4 x double> undef to <4 x i8>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I8 = fptoui <8 x double> undef to <8 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I8 = fptoui <4 x double> undef to <4 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I8 = fptoui <8 x double> undef to <8 x i8>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'fptoui_double_i8'
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = fptoui double undef to i8
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I8 = fptoui <2 x double> undef to <2 x i8>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I8 = fptoui <4 x double> undef to <4 x i8>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I8 = fptoui <8 x double> undef to <8 x i8>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I8 = fptoui <8 x double> undef to <8 x i8>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'fptoui_double_i8'
@@ -199,8 +199,8 @@ define i32 @fptoui_double_i8(i32 %arg) {
 ; SLM-LABEL: 'fptoui_double_i8'
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = fptoui double undef to i8
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I8 = fptoui <2 x double> undef to <2 x i8>
-; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4I8 = fptoui <4 x double> undef to <4 x i8>
-; SLM-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8I8 = fptoui <8 x double> undef to <8 x i8>
+; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4I8 = fptoui <4 x double> undef to <4 x i8>
+; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8I8 = fptoui <8 x double> undef to <8 x i8>
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
   %I8 = fptoui double undef to i8
@@ -337,8 +337,8 @@ define i32 @fptoui_float_i16(i32 %arg) {
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = fptoui float undef to i16
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I16 = fptoui <2 x float> undef to <2 x i16>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I16 = fptoui <4 x float> undef to <4 x i16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = fptoui <8 x float> undef to <8 x i16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16I16 = fptoui <16 x float> undef to <16 x i16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = fptoui <8 x float> undef to <8 x i16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V16I16 = fptoui <16 x float> undef to <16 x i16>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'fptoui_float_i16'
@@ -346,7 +346,7 @@ define i32 @fptoui_float_i16(i32 %arg) {
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I16 = fptoui <2 x float> undef to <2 x i16>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I16 = fptoui <4 x float> undef to <4 x i16>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I16 = fptoui <8 x float> undef to <8 x i16>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16I16 = fptoui <16 x float> undef to <16 x i16>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V16I16 = fptoui <16 x float> undef to <16 x i16>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'fptoui_float_i16'
@@ -354,7 +354,7 @@ define i32 @fptoui_float_i16(i32 %arg) {
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I16 = fptoui <2 x float> undef to <2 x i16>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I16 = fptoui <4 x float> undef to <4 x i16>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8I16 = fptoui <8 x float> undef to <8 x i16>
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16I16 = fptoui <16 x float> undef to <16 x i16>
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16I16 = fptoui <16 x float> undef to <16 x i16>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX512-LABEL: 'fptoui_float_i16'
@@ -369,8 +369,8 @@ define i32 @fptoui_float_i16(i32 %arg) {
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I16 = fptoui float undef to i16
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2I16 = fptoui <2 x float> undef to <2 x i16>
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4I16 = fptoui <4 x float> undef to <4 x i16>
-; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I16 = fptoui <8 x float> undef to <8 x i16>
-; SLM-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16I16 = fptoui <16 x float> undef to <16 x i16>
+; SLM-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8I16 = fptoui <8 x float> undef to <8 x i16>
+; SLM-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V16I16 = fptoui <16 x float> undef to <16 x i16>
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
   %I16 = fptoui float undef to i16
@@ -386,7 +386,7 @@ define i32 @fptoui_float_i8(i32 %arg) {
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %I8 = fptoui float undef to i8
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2I8 = fptoui <2 x float> undef to <2 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I8 = fptoui <4 x float> undef to <4 x i8>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I8 = fptoui <8 x float> undef to <8 x i8>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8I8 = fptoui <8 x float> undef to <8 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I8 = fptoui <16 x float> undef to <16 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
@@ -394,7 +394,7 @@ define i32 @fptoui_float_i8(i32 %arg) {
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = fptoui float undef to i8
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I8 = fptoui <2 x float> undef to <2 x i8>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I8 = fptoui <4 x float> undef to <4 x i8>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I8 = fptoui <8 x float> undef to <8 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I8 = fptoui <8 x float> undef to <8 x i8>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I8 = fptoui <16 x float> undef to <16 x i8>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
@@ -426,7 +426,7 @@ define i32 @fptoui_float_i8(i32 %arg) {
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %I8 = fptoui float undef to i8
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2I8 = fptoui <2 x float> undef to <2 x i8>
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4I8 = fptoui <4 x float> undef to <4 x i8>
-; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8I8 = fptoui <8 x float> undef to <8 x i8>
+; SLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8I8 = fptoui <8 x float> undef to <8 x i8>
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16I8 = fptoui <16 x float> undef to <16 x i8>
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;

diff --git a/llvm/test/Analysis/CostModel/X86/intrinsic-cost-kinds.ll b/llvm/test/Analysis/CostModel/X86/intrinsic-cost-kinds.ll
index e4da1f8..dddd3b5 100644 (file)
--- a/llvm/test/Analysis/CostModel/X86/intrinsic-cost-kinds.ll
+++ b/llvm/test/Analysis/CostModel/X86/intrinsic-cost-kinds.ll
@@ -48,7 +48,7 @@ declare void @llvm.memcpy.p0i8.p0i8.i32(i8*, i8*, i32, i1)
 define void @umul(i32 %a, i32 %b, <16 x i32> %va, <16 x i32> %vb) {
 ; THRU-LABEL: 'umul'
 ; THRU-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %s = call { i32, i1 } @llvm.umul.with.overflow.i32(i32 %a, i32 %b)
-; THRU-NEXT:  Cost Model: Found an estimated cost of 112 for instruction: %v = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> %va, <16 x i32> %vb)
+; THRU-NEXT:  Cost Model: Found an estimated cost of 104 for instruction: %v = call { <16 x i32>, <16 x i1> } @llvm.umul.with.overflow.v16i32(<16 x i32> %va, <16 x i32> %vb)
 ; THRU-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
 ;
 ; LATE-LABEL: 'umul'

diff --git a/llvm/test/Analysis/CostModel/X86/min-legal-vector-width.ll b/llvm/test/Analysis/CostModel/X86/min-legal-vector-width.ll
index 3c5e9d5..46293ab 100644 (file)
--- a/llvm/test/Analysis/CostModel/X86/min-legal-vector-width.ll
+++ b/llvm/test/Analysis/CostModel/X86/min-legal-vector-width.ll
@@ -216,9 +216,9 @@ define void @sext512() "min-legal-vector-width"="512" {
 define void @trunc256() "min-legal-vector-width"="256" {
 ; VEC256-LABEL: 'trunc256'
 ; VEC256-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %A = trunc <8 x i64> undef to <8 x i32>
-; VEC256-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %B = trunc <8 x i64> undef to <8 x i16>
-; VEC256-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %C = trunc <8 x i64> undef to <8 x i8>
-; VEC256-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %D = trunc <16 x i32> undef to <16 x i16>
+; VEC256-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %B = trunc <8 x i64> undef to <8 x i16>
+; VEC256-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %C = trunc <8 x i64> undef to <8 x i8>
+; VEC256-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %D = trunc <16 x i32> undef to <16 x i16>
 ; VEC256-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %E = trunc <16 x i32> undef to <16 x i8>
 ; VEC256-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %F = trunc <32 x i16> undef to <32 x i8>
 ; VEC256-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
@@ -547,9 +547,9 @@ define i32 @trunc_vXi1() "min-legal-vector-width"="256" {
 ; AVX-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i1>
 ; AVX-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i1>
 ; AVX-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i1>
-; AVX-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i1>
-; AVX-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i1>
-; AVX-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i1>
+; AVX-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i1>
+; AVX-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i1>
+; AVX-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i1>
 ; AVX-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i16 = trunc <2 x i16> undef to <2 x i1>
 ; AVX-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i16 = trunc <4 x i16> undef to <4 x i1>
 ; AVX-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %V8i16 = trunc <8 x i16> undef to <8 x i1>

diff --git a/llvm/test/Analysis/CostModel/X86/rem.ll b/llvm/test/Analysis/CostModel/X86/rem.ll
index 2a12701..314ca6f 100644 (file)
--- a/llvm/test/Analysis/CostModel/X86/rem.ll
+++ b/llvm/test/Analysis/CostModel/X86/rem.ll
@@ -741,8 +741,8 @@ define i32 @srem_constpow2() {
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 396 for instruction: %V32i16 = srem <32 x i16> undef, <i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = srem i8 undef, 16
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 143 for instruction: %V16i8 = srem <16 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 286 for instruction: %V32i8 = srem <32 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 572 for instruction: %V64i8 = srem <64 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V32i8 = srem <32 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 576 for instruction: %V64i8 = srem <64 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'srem_constpow2'
@@ -760,8 +760,8 @@ define i32 @srem_constpow2() {
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 396 for instruction: %V32i16 = srem <32 x i16> undef, <i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = srem i8 undef, 16
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 143 for instruction: %V16i8 = srem <16 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 286 for instruction: %V32i8 = srem <32 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 572 for instruction: %V64i8 = srem <64 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V32i8 = srem <32 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 576 for instruction: %V64i8 = srem <64 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'srem_constpow2'
@@ -779,8 +779,8 @@ define i32 @srem_constpow2() {
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 180 for instruction: %V32i16 = srem <32 x i16> undef, <i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = srem i8 undef, 16
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 69 for instruction: %V16i8 = srem <16 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 138 for instruction: %V32i8 = srem <32 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 276 for instruction: %V64i8 = srem <64 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 140 for instruction: %V32i8 = srem <32 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 280 for instruction: %V64i8 = srem <64 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'srem_constpow2'
@@ -874,8 +874,8 @@ define i32 @srem_constpow2() {
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 184 for instruction: %V32i16 = srem <32 x i16> undef, <i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256>
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = srem i8 undef, 16
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 71 for instruction: %V16i8 = srem <16 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
-; SLM-NEXT:  Cost Model: Found an estimated cost of 142 for instruction: %V32i8 = srem <32 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
-; SLM-NEXT:  Cost Model: Found an estimated cost of 284 for instruction: %V64i8 = srem <64 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
+; SLM-NEXT:  Cost Model: Found an estimated cost of 144 for instruction: %V32i8 = srem <32 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
+; SLM-NEXT:  Cost Model: Found an estimated cost of 288 for instruction: %V64i8 = srem <64 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'srem_constpow2'
@@ -893,8 +893,8 @@ define i32 @srem_constpow2() {
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 180 for instruction: %V32i16 = srem <32 x i16> undef, <i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256, i16 2, i16 4, i16 8, i16 16, i16 32, i16 64, i16 128, i16 256>
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = srem i8 undef, 16
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 69 for instruction: %V16i8 = srem <16 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
-; GLM-NEXT:  Cost Model: Found an estimated cost of 138 for instruction: %V32i8 = srem <32 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
-; GLM-NEXT:  Cost Model: Found an estimated cost of 276 for instruction: %V64i8 = srem <64 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
+; GLM-NEXT:  Cost Model: Found an estimated cost of 140 for instruction: %V32i8 = srem <32 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
+; GLM-NEXT:  Cost Model: Found an estimated cost of 280 for instruction: %V64i8 = srem <64 x i8> undef, <i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16, i8 2, i8 4, i8 8, i8 16>
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'srem_constpow2'
@@ -1036,8 +1036,8 @@ define i32 @srem_uniformconstpow2() {
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V32i16 = srem <32 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = srem i8 undef, 16
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16i8 = srem <16 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32i8 = srem <32 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V64i8 = srem <64 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V32i8 = srem <32 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V64i8 = srem <64 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'srem_uniformconstpow2'
@@ -1055,8 +1055,8 @@ define i32 @srem_uniformconstpow2() {
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V32i16 = srem <32 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = srem i8 undef, 16
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16i8 = srem <16 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32i8 = srem <32 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V64i8 = srem <64 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V32i8 = srem <32 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V64i8 = srem <64 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'srem_uniformconstpow2'
@@ -1074,8 +1074,8 @@ define i32 @srem_uniformconstpow2() {
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V32i16 = srem <32 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = srem i8 undef, 16
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16i8 = srem <16 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32i8 = srem <32 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V64i8 = srem <64 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V32i8 = srem <32 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V64i8 = srem <64 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'srem_uniformconstpow2'
@@ -1169,8 +1169,8 @@ define i32 @srem_uniformconstpow2() {
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V32i16 = srem <32 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = srem i8 undef, 16
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16i8 = srem <16 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
-; SLM-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V32i8 = srem <32 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
-; SLM-NEXT:  Cost Model: Found an estimated cost of 84 for instruction: %V64i8 = srem <64 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
+; SLM-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V32i8 = srem <32 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
+; SLM-NEXT:  Cost Model: Found an estimated cost of 88 for instruction: %V64i8 = srem <64 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; GLM-LABEL: 'srem_uniformconstpow2'
@@ -1188,8 +1188,8 @@ define i32 @srem_uniformconstpow2() {
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V32i16 = srem <32 x i16> undef, <i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16, i16 16>
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %I8 = srem i8 undef, 16
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16i8 = srem <16 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
-; GLM-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32i8 = srem <32 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
-; GLM-NEXT:  Cost Model: Found an estimated cost of 76 for instruction: %V64i8 = srem <64 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
+; GLM-NEXT:  Cost Model: Found an estimated cost of 40 for instruction: %V32i8 = srem <32 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
+; GLM-NEXT:  Cost Model: Found an estimated cost of 80 for instruction: %V64i8 = srem <64 x i8> undef, <i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16, i8 16>
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; BTVER2-LABEL: 'srem_uniformconstpow2'

diff --git a/llvm/test/Analysis/CostModel/X86/slm-arith-costs.ll b/llvm/test/Analysis/CostModel/X86/slm-arith-costs.ll
index d57da8c..6d169bd 100644 (file)
--- a/llvm/test/Analysis/CostModel/X86/slm-arith-costs.ll
+++ b/llvm/test/Analysis/CostModel/X86/slm-arith-costs.ll
@@ -47,11 +47,11 @@ entry:
 
 define <2 x i8> @slm-costs_8_v2_mul(<2 x i8> %a, <2 x i8> %b)  {
 ; SLM-LABEL: 'slm-costs_8_v2_mul'
-; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %res = mul nsw <2 x i8> %a, %b
+; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %res = mul nsw <2 x i8> %a, %b
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret <2 x i8> %res
 ;
 ; GLM-LABEL: 'slm-costs_8_v2_mul'
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %res = mul nsw <2 x i8> %a, %b
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %res = mul nsw <2 x i8> %a, %b
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret <2 x i8> %res
 ;
 entry:
@@ -61,11 +61,11 @@ entry:
 
 define <4 x i8> @slm-costs_8_v4_mul(<4 x i8> %a, <4 x i8> %b)  {
 ; SLM-LABEL: 'slm-costs_8_v4_mul'
-; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %res = mul nsw <4 x i8> %a, %b
+; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %res = mul nsw <4 x i8> %a, %b
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret <4 x i8> %res
 ;
 ; GLM-LABEL: 'slm-costs_8_v4_mul'
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %res = mul nsw <4 x i8> %a, %b
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %res = mul nsw <4 x i8> %a, %b
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret <4 x i8> %res
 ;
 entry:
@@ -177,11 +177,11 @@ entry:
 
 define <8 x i8> @slm-costs_8_v8_mul(<8 x i8> %a, <8 x i8> %b)  {
 ; SLM-LABEL: 'slm-costs_8_v8_mul'
-; SLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %res = mul nsw <8 x i8> %a, %b
+; SLM-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %res = mul nsw <8 x i8> %a, %b
 ; SLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret <8 x i8> %res
 ;
 ; GLM-LABEL: 'slm-costs_8_v8_mul'
-; GLM-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %res = mul nsw <8 x i8> %a, %b
+; GLM-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %res = mul nsw <8 x i8> %a, %b
 ; GLM-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret <8 x i8> %res
 ;
 entry:

diff --git a/llvm/test/Analysis/CostModel/X86/sse-itoi.ll b/llvm/test/Analysis/CostModel/X86/sse-itoi.ll
index bf0262e..9d9e044 100644 (file)
--- a/llvm/test/Analysis/CostModel/X86/sse-itoi.ll
+++ b/llvm/test/Analysis/CostModel/X86/sse-itoi.ll
@@ -418,17 +418,11 @@ define void @sext_v4i8_to_v4i16(<4 x i8>* %a) {
 }
 
 define void @truncate_v16i32_to_v16i16(<16 x i32>* %a) {
-; SSE2-LABEL: 'truncate_v16i32_to_v16i16'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %1 = load <16 x i32>, <16 x i32>* %a, align 64
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %2 = trunc <16 x i32> %1 to <16 x i16>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: store <16 x i16> %2, <16 x i16>* undef, align 4
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
-;
-; SSE41-LABEL: 'truncate_v16i32_to_v16i16'
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %1 = load <16 x i32>, <16 x i32>* %a, align 64
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %2 = trunc <16 x i32> %1 to <16 x i16>
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: store <16 x i16> %2, <16 x i16>* undef, align 4
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
+; CHECK-LABEL: 'truncate_v16i32_to_v16i16'
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %1 = load <16 x i32>, <16 x i32>* %a, align 64
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %2 = trunc <16 x i32> %1 to <16 x i16>
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: store <16 x i16> %2, <16 x i16>* undef, align 4
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
 ;
   %1 = load <16 x i32>, <16 x i32>* %a
   %2 = trunc <16 x i32> %1 to <16 x i16>
@@ -437,17 +431,11 @@ define void @truncate_v16i32_to_v16i16(<16 x i32>* %a) {
 }
 
 define void @truncate_v8i32_to_v8i16(<8 x i32>* %a) {
-; SSE2-LABEL: 'truncate_v8i32_to_v8i16'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %1 = load <8 x i32>, <8 x i32>* %a, align 32
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %2 = trunc <8 x i32> %1 to <8 x i16>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: store <8 x i16> %2, <8 x i16>* undef, align 4
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
-;
-; SSE41-LABEL: 'truncate_v8i32_to_v8i16'
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %1 = load <8 x i32>, <8 x i32>* %a, align 32
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %2 = trunc <8 x i32> %1 to <8 x i16>
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: store <8 x i16> %2, <8 x i16>* undef, align 4
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
+; CHECK-LABEL: 'truncate_v8i32_to_v8i16'
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %1 = load <8 x i32>, <8 x i32>* %a, align 32
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %2 = trunc <8 x i32> %1 to <8 x i16>
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: store <8 x i16> %2, <8 x i16>* undef, align 4
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
 ;
   %1 = load <8 x i32>, <8 x i32>* %a
   %2 = trunc <8 x i32> %1 to <8 x i16>
@@ -464,7 +452,7 @@ define void @truncate_v4i32_to_v4i16(<4 x i32>* %a) {
 ;
 ; SSE41-LABEL: 'truncate_v4i32_to_v4i16'
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %1 = load <4 x i32>, <4 x i32>* %a, align 16
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %2 = trunc <4 x i32> %1 to <4 x i16>
+; SSE41-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %2 = trunc <4 x i32> %1 to <4 x i16>
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: store <4 x i16> %2, <4 x i16>* undef, align 4
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
 ;
@@ -490,13 +478,13 @@ define void @truncate_v16i32_to_v16i8(<16 x i32>* %a) {
 define void @truncate_v8i32_to_v8i8(<8 x i32>* %a) {
 ; SSE2-LABEL: 'truncate_v8i32_to_v8i8'
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %1 = load <8 x i32>, <8 x i32>* %a, align 32
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %2 = trunc <8 x i32> %1 to <8 x i8>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %2 = trunc <8 x i32> %1 to <8 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: store <8 x i8> %2, <8 x i8>* undef, align 4
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
 ;
 ; SSE41-LABEL: 'truncate_v8i32_to_v8i8'
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %1 = load <8 x i32>, <8 x i32>* %a, align 32
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %2 = trunc <8 x i32> %1 to <8 x i8>
+; SSE41-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %2 = trunc <8 x i32> %1 to <8 x i8>
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: store <8 x i8> %2, <8 x i8>* undef, align 4
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
 ;
@@ -515,7 +503,7 @@ define void @truncate_v4i32_to_v4i8(<4 x i32>* %a) {
 ;
 ; SSE41-LABEL: 'truncate_v4i32_to_v4i8'
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %1 = load <4 x i32>, <4 x i32>* %a, align 16
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %2 = trunc <4 x i32> %1 to <4 x i8>
+; SSE41-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %2 = trunc <4 x i32> %1 to <4 x i8>
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: store <4 x i8> %2, <4 x i8>* undef, align 4
 ; SSE41-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
 ;
@@ -539,17 +527,11 @@ define void @truncate_v16i16_to_v16i8(<16 x i16>* %a) {
 }
 
 define void @truncate_v8i16_to_v8i8(<8 x i16>* %a) {
-; SSE2-LABEL: 'truncate_v8i16_to_v8i8'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %1 = load <8 x i16>, <8 x i16>* %a, align 16
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %2 = trunc <8 x i16> %1 to <8 x i8>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: store <8 x i8> %2, <8 x i8>* undef, align 4
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
-;
-; SSE41-LABEL: 'truncate_v8i16_to_v8i8'
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %1 = load <8 x i16>, <8 x i16>* %a, align 16
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %2 = trunc <8 x i16> %1 to <8 x i8>
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: store <8 x i8> %2, <8 x i8>* undef, align 4
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
+; CHECK-LABEL: 'truncate_v8i16_to_v8i8'
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %1 = load <8 x i16>, <8 x i16>* %a, align 16
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %2 = trunc <8 x i16> %1 to <8 x i8>
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: store <8 x i8> %2, <8 x i8>* undef, align 4
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
 ;
   %1 = load <8 x i16>, <8 x i16>* %a
   %2 = trunc <8 x i16> %1 to <8 x i8>
@@ -558,17 +540,11 @@ define void @truncate_v8i16_to_v8i8(<8 x i16>* %a) {
 }
 
 define void @truncate_v4i16_to_v4i8(<4 x i16>* %a) {
-; SSE2-LABEL: 'truncate_v4i16_to_v4i8'
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %1 = load <4 x i16>, <4 x i16>* %a, align 8
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %2 = trunc <4 x i16> %1 to <4 x i8>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: store <4 x i8> %2, <4 x i8>* undef, align 4
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
-;
-; SSE41-LABEL: 'truncate_v4i16_to_v4i8'
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %1 = load <4 x i16>, <4 x i16>* %a, align 8
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %2 = trunc <4 x i16> %1 to <4 x i8>
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: store <4 x i8> %2, <4 x i8>* undef, align 4
-; SSE41-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
+; CHECK-LABEL: 'truncate_v4i16_to_v4i8'
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %1 = load <4 x i16>, <4 x i16>* %a, align 8
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %2 = trunc <4 x i16> %1 to <4 x i8>
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: store <4 x i8> %2, <4 x i8>* undef, align 4
+; CHECK-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
 ;
   %1 = load <4 x i16>, <4 x i16>* %a
   %2 = trunc <4 x i16> %1 to <4 x i8>

diff --git a/llvm/test/Analysis/CostModel/X86/trunc.ll b/llvm/test/Analysis/CostModel/X86/trunc.ll
index a74ab5c..b169e9a 100644 (file)
--- a/llvm/test/Analysis/CostModel/X86/trunc.ll
+++ b/llvm/test/Analysis/CostModel/X86/trunc.ll
@@ -15,17 +15,17 @@ define i32 @trunc_vXi32() {
 ; SSE-LABEL: 'trunc_vXi32'
 ; SSE-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i64 = trunc i64 undef to i32
 ; SSE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i32>
-; SSE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i32>
-; SSE-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i32>
-; SSE-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i32>
+; SSE-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i32>
+; SSE-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i32>
+; SSE-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i32>
 ; SSE-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'trunc_vXi32'
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i64 = trunc i64 undef to i32
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i32>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i32>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i32>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i32>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i32>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i32>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'trunc_vXi32'
@@ -48,8 +48,8 @@ define i32 @trunc_vXi32() {
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i64 = trunc i64 undef to i32
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i32>
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i32>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i32>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i32>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i32>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i32>
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
   %i64 = trunc i64 undef to i32
@@ -64,76 +64,76 @@ define i32 @trunc_vXi16() {
 ; SSE2-LABEL: 'trunc_vXi16'
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i64 = trunc i64 undef to i16
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i16>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i16>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i16>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i16>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i16>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i16>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i16>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i16>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i16>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i32 = trunc i32 undef to i16
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i16>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i16>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i16>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i16>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i16>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i16>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'trunc_vXi16'
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i64 = trunc i64 undef to i16
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i16>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i16>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i16>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i16>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i16>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i16>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i16>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i16>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i16>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i32 = trunc i32 undef to i16
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i16>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i16>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i16>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i16>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i16>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i16>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'trunc_vXi16'
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i64 = trunc i64 undef to i16
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 44 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i16>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i32 = trunc i32 undef to i16
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i16>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i16>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i16>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'trunc_vXi16'
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i64 = trunc i64 undef to i16
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i16>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i16>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i16>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i16>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i16>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i16>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i16>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i16>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i32 = trunc i32 undef to i16
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i16>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i16>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i16>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i16>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i16>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i16>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i16>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i16>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX2-LABEL: 'trunc_vXi16'
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i64 = trunc i64 undef to i16
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i16>
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i16>
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i16>
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i16>
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i16>
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i16>
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i16>
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i16>
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i16>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i32 = trunc i32 undef to i16
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i16>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i16>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i16>
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i16>
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i16>
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i16>
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i16>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX512-LABEL: 'trunc_vXi16'
@@ -155,15 +155,15 @@ define i32 @trunc_vXi16() {
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i64 = trunc i64 undef to i16
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i16>
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i16>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i16>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i16>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 38 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i16>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i16>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 13 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i16>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 26 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i16>
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i32 = trunc i32 undef to i16
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i16>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i16>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i16>
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i16>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i16>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i16>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i16>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i16>
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
   %i64 = trunc i64 undef to i16
@@ -186,94 +186,94 @@ define i32 @trunc_vXi8() {
 ; SSE2-LABEL: 'trunc_vXi8'
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i64 = trunc i64 undef to i8
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i8>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i8>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i8>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 39 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i8>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 78 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i8>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 156 for instruction: %V64i64 = trunc <64 x i64> undef to <64 x i8>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i8>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i8>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i8>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i8>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V64i64 = trunc <64 x i64> undef to <64 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i32 = trunc i32 undef to i8
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i8>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i8>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i8>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i8>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i8>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i8>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i16 = trunc i16 undef to i8
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2i16 = trunc <2 x i16> undef to <2 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4i16 = trunc <4 x i16> undef to <4 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8i16 = trunc <8 x i16> undef to <8 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16i16 = trunc <16 x i16> undef to <16 x i8>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V32i16 = trunc <32 x i16> undef to <32 x i8>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V64i16 = trunc <64 x i16> undef to <64 x i8>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32i16 = trunc <32 x i16> undef to <32 x i8>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64i16 = trunc <64 x i16> undef to <64 x i8>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSSE3-LABEL: 'trunc_vXi8'
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i64 = trunc i64 undef to i8
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i8>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i8>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 19 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i8>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 39 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i8>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 78 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i8>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 156 for instruction: %V64i64 = trunc <64 x i64> undef to <64 x i8>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i8>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i8>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i8>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i8>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V64i64 = trunc <64 x i64> undef to <64 x i8>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i32 = trunc i32 undef to i8
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i8>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i8>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i8>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i8>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i8>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i8>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i8>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i8>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i8>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i16 = trunc i16 undef to i8
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2i16 = trunc <2 x i16> undef to <2 x i8>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4i16 = trunc <4 x i16> undef to <4 x i8>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8i16 = trunc <8 x i16> undef to <8 x i8>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16i16 = trunc <16 x i16> undef to <16 x i8>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V32i16 = trunc <32 x i16> undef to <32 x i8>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V64i16 = trunc <64 x i16> undef to <64 x i8>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32i16 = trunc <32 x i16> undef to <32 x i8>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64i16 = trunc <64 x i16> undef to <64 x i8>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; SSE42-LABEL: 'trunc_vXi8'
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i64 = trunc i64 undef to i8
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i8>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i8>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i8>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i8>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i8>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 60 for instruction: %V64i64 = trunc <64 x i64> undef to <64 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V64i64 = trunc <64 x i64> undef to <64 x i8>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i32 = trunc i32 undef to i8
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i8>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i8>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i8>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i8>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i8>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i8>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i16 = trunc i16 undef to i8
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i16 = trunc <2 x i16> undef to <2 x i8>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i16 = trunc <4 x i16> undef to <4 x i8>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8i16 = trunc <8 x i16> undef to <8 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2i16 = trunc <2 x i16> undef to <2 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4i16 = trunc <4 x i16> undef to <4 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8i16 = trunc <8 x i16> undef to <8 x i8>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16i16 = trunc <16 x i16> undef to <16 x i8>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V32i16 = trunc <32 x i16> undef to <32 x i8>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V64i16 = trunc <64 x i16> undef to <64 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32i16 = trunc <32 x i16> undef to <32 x i8>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64i16 = trunc <64 x i16> undef to <64 x i8>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
 ;
 ; AVX1-LABEL: 'trunc_vXi8'
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i64 = trunc i64 undef to i8
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i8>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i8>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i8>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i8>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i8>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V64i64 = trunc <64 x i64> undef to <64 x i8>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i8>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i8>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i8>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i8>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 41 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i8>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 82 for instruction: %V64i64 = trunc <64 x i64> undef to <64 x i8>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i32 = trunc i32 undef to i8
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i8>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i8>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i8>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i8>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i8>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i8>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i8>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i8>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i16 = trunc i16 undef to i8
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i16 = trunc <2 x i16> undef to <2 x i8>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i16 = trunc <4 x i16> undef to <4 x i8>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8i16 = trunc <8 x i16> undef to <8 x i8>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2i16 = trunc <2 x i16> undef to <2 x i8>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4i16 = trunc <4 x i16> undef to <4 x i8>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8i16 = trunc <8 x i16> undef to <8 x i8>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16i16 = trunc <16 x i16> undef to <16 x i8>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V32i16 = trunc <32 x i16> undef to <32 x i8>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V64i16 = trunc <64 x i16> undef to <64 x i8>
@@ -283,10 +283,10 @@ define i32 @trunc_vXi8() {
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i64 = trunc i64 undef to i8
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i8>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i8>
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i8>
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i8>
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i8>
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V64i64 = trunc <64 x i64> undef to <64 x i8>
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i8>
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i8>
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 33 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i8>
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 66 for instruction: %V64i64 = trunc <64 x i64> undef to <64 x i8>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i32 = trunc i32 undef to i8
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i8>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i8>
@@ -353,23 +353,23 @@ define i32 @trunc_vXi8() {
 ;
 ; BTVER2-LABEL: 'trunc_vXi8'
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i64 = trunc i64 undef to i8
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i8>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i8>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i8>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i8>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i8>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V64i64 = trunc <64 x i64> undef to <64 x i8>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i8>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i8>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i8>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 20 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i8>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 41 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i8>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 82 for instruction: %V64i64 = trunc <64 x i64> undef to <64 x i8>
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i32 = trunc i32 undef to i8
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i8>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i8>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i8>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i8>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i8>
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i8>
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 15 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i8>
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 30 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i8>
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i16 = trunc i16 undef to i8
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i16 = trunc <2 x i16> undef to <2 x i8>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i16 = trunc <4 x i16> undef to <4 x i8>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8i16 = trunc <8 x i16> undef to <8 x i8>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2i16 = trunc <2 x i16> undef to <2 x i8>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4i16 = trunc <4 x i16> undef to <4 x i8>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8i16 = trunc <8 x i16> undef to <8 x i8>
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V16i16 = trunc <16 x i16> undef to <16 x i8>
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V32i16 = trunc <32 x i16> undef to <32 x i8>
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V64i16 = trunc <64 x i16> undef to <64 x i8>
@@ -406,25 +406,25 @@ define i32 @trunc_vXi1() {
 ; SSE2-LABEL: 'trunc_vXi1'
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i64 = trunc i64 undef to i1
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i1>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i1>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i1>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i1>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i1>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V64i64 = trunc <64 x i64> undef to <64 x i1>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i1>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i1>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i1>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i1>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V64i64 = trunc <64 x i64> undef to <64 x i1>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i32 = trunc i32 undef to i1
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i1>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i1>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i1>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i1>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i1>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i1>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i1>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i1>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i1>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i1>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i16 = trunc i16 undef to i1
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2i16 = trunc <2 x i16> undef to <2 x i1>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i16 = trunc <4 x i16> undef to <4 x i1>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %V8i16 = trunc <8 x i16> undef to <8 x i1>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16i16 = trunc <16 x i16> undef to <16 x i1>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32i16 = trunc <32 x i16> undef to <32 x i1>
-; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64i16 = trunc <64 x i16> undef to <64 x i1>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16i16 = trunc <16 x i16> undef to <16 x i1>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32i16 = trunc <32 x i16> undef to <32 x i1>
+; SSE2-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64i16 = trunc <64 x i16> undef to <64 x i1>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i8 = trunc i8 undef to i1
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2i8 = trunc <2 x i8> undef to <2 x i1>
 ; SSE2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4i8 = trunc <4 x i8> undef to <4 x i1>
@@ -437,25 +437,25 @@ define i32 @trunc_vXi1() {
 ; SSSE3-LABEL: 'trunc_vXi1'
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i64 = trunc i64 undef to i1
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i1>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i1>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i1>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i1>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i1>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V64i64 = trunc <64 x i64> undef to <64 x i1>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i1>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i1>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i1>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i1>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 128 for instruction: %V64i64 = trunc <64 x i64> undef to <64 x i1>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i32 = trunc i32 undef to i1
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i1>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i1>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i1>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i1>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i1>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i1>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i1>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i1>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 24 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i1>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 48 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i1>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i16 = trunc i16 undef to i1
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V2i16 = trunc <2 x i16> undef to <2 x i1>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i16 = trunc <4 x i16> undef to <4 x i1>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %V8i16 = trunc <8 x i16> undef to <8 x i1>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16i16 = trunc <16 x i16> undef to <16 x i1>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32i16 = trunc <32 x i16> undef to <32 x i1>
-; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64i16 = trunc <64 x i16> undef to <64 x i1>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16i16 = trunc <16 x i16> undef to <16 x i1>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32i16 = trunc <32 x i16> undef to <32 x i1>
+; SSSE3-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64i16 = trunc <64 x i16> undef to <64 x i1>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i8 = trunc i8 undef to i1
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V2i8 = trunc <2 x i8> undef to <2 x i1>
 ; SSSE3-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4i8 = trunc <4 x i8> undef to <4 x i1>
@@ -468,25 +468,25 @@ define i32 @trunc_vXi1() {
 ; SSE42-LABEL: 'trunc_vXi1'
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i64 = trunc i64 undef to i1
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %V2i64 = trunc <2 x i64> undef to <2 x i1>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i1>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i1>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 7 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i1>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 14 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i1>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 28 for instruction: %V64i64 = trunc <64 x i64> undef to <64 x i1>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V4i64 = trunc <4 x i64> undef to <4 x i1>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V8i64 = trunc <8 x i64> undef to <8 x i1>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V16i64 = trunc <16 x i64> undef to <16 x i1>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V32i64 = trunc <32 x i64> undef to <32 x i1>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 64 for instruction: %V64i64 = trunc <64 x i64> undef to <64 x i1>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i32 = trunc i32 undef to i1
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i1>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i1>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i1>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i1>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 6 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i1>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 12 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i1>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i1>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i1>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i1>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 32 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i1>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i16 = trunc i16 undef to i1
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i16 = trunc <2 x i16> undef to <2 x i1>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i16 = trunc <4 x i16> undef to <4 x i1>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %V8i16 = trunc <8 x i16> undef to <8 x i1>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V16i16 = trunc <16 x i16> undef to <16 x i1>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V32i16 = trunc <32 x i16> undef to <32 x i1>
-; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V64i16 = trunc <64 x i16> undef to <64 x i1>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 4 for instruction: %V16i16 = trunc <16 x i16> undef to <16 x i1>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V32i16 = trunc <32 x i16> undef to <32 x i1>
+; SSE42-NEXT:  Cost Model: Found an estimated cost of 16 for instruction: %V64i16 = trunc <64 x i16> undef to <64 x i1>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i8 = trunc i8 undef to i1
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i8 = trunc <2 x i8> undef to <2 x i1>
 ; SSE42-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i8 = trunc <4 x i8> undef to <4 x i1>
@@ -508,9 +508,9 @@ define i32 @trunc_vXi1() {
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i1>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i1>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i1>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i1>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i1>
-; AVX1-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i1>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i1>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i1>
+; AVX1-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i1>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i16 = trunc i16 undef to i1
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i16 = trunc <2 x i16> undef to <2 x i1>
 ; AVX1-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i16 = trunc <4 x i16> undef to <4 x i1>
@@ -539,9 +539,9 @@ define i32 @trunc_vXi1() {
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i1>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i1>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 2 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i1>
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i1>
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i1>
-; AVX2-NEXT:  Cost Model: Found an estimated cost of 22 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i1>
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 8 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i1>
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i1>
+; AVX2-NEXT:  Cost Model: Found an estimated cost of 34 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i1>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i16 = trunc i16 undef to i1
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i16 = trunc <2 x i16> undef to <2 x i1>
 ; AVX2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i16 = trunc <4 x i16> undef to <4 x i1>
@@ -632,9 +632,9 @@ define i32 @trunc_vXi1() {
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i32 = trunc <2 x i32> undef to <2 x i1>
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %V4i32 = trunc <4 x i32> undef to <4 x i1>
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %V8i32 = trunc <8 x i32> undef to <8 x i1>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 11 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i1>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 23 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i1>
-; BTVER2-NEXT:  Cost Model: Found an estimated cost of 46 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i1>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 10 for instruction: %V16i32 = trunc <16 x i32> undef to <16 x i1>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 21 for instruction: %V32i32 = trunc <32 x i32> undef to <32 x i1>
+; BTVER2-NEXT:  Cost Model: Found an estimated cost of 42 for instruction: %V64i32 = trunc <64 x i32> undef to <64 x i1>
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: %i16 = trunc i16 undef to i1
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V2i16 = trunc <2 x i16> undef to <2 x i1>
 ; BTVER2-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %V4i16 = trunc <4 x i16> undef to <4 x i1>

diff --git a/llvm/test/Transforms/SLPVectorizer/X86/arith-fix.ll b/llvm/test/Transforms/SLPVectorizer/X86/arith-fix.ll
index 375fae8..87ee673 100644 (file)
--- a/llvm/test/Transforms/SLPVectorizer/X86/arith-fix.ll
+++ b/llvm/test/Transforms/SLPVectorizer/X86/arith-fix.ll
@@ -149,70 +149,22 @@ define void @smul_v8i64() {
 
 define void @smul_v16i32() {
 ; SSE-LABEL: @smul_v16i32(
-; SSE-NEXT:    [[A0:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @a32, i32 0, i64 0), align 4
-; SSE-NEXT:    [[A1:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @a32, i32 0, i64 1), align 4
-; SSE-NEXT:    [[A2:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @a32, i32 0, i64 2), align 4
-; SSE-NEXT:    [[A3:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @a32, i32 0, i64 3), align 4
-; SSE-NEXT:    [[A4:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @a32, i32 0, i64 4), align 4
-; SSE-NEXT:    [[A5:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @a32, i32 0, i64 5), align 4
-; SSE-NEXT:    [[A6:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @a32, i32 0, i64 6), align 4
-; SSE-NEXT:    [[A7:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @a32, i32 0, i64 7), align 4
-; SSE-NEXT:    [[A8:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @a32, i32 0, i64 8), align 4
-; SSE-NEXT:    [[A9:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @a32, i32 0, i64 9), align 4
-; SSE-NEXT:    [[A10:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @a32, i32 0, i64 10), align 4
-; SSE-NEXT:    [[A11:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @a32, i32 0, i64 11), align 4
-; SSE-NEXT:    [[A12:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @a32, i32 0, i64 12), align 4
-; SSE-NEXT:    [[A13:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @a32, i32 0, i64 13), align 4
-; SSE-NEXT:    [[A14:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @a32, i32 0, i64 14), align 4
-; SSE-NEXT:    [[A15:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @a32, i32 0, i64 15), align 4
-; SSE-NEXT:    [[B0:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @b32, i32 0, i64 0), align 4
-; SSE-NEXT:    [[B1:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @b32, i32 0, i64 1), align 4
-; SSE-NEXT:    [[B2:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @b32, i32 0, i64 2), align 4
-; SSE-NEXT:    [[B3:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @b32, i32 0, i64 3), align 4
-; SSE-NEXT:    [[B4:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @b32, i32 0, i64 4), align 4
-; SSE-NEXT:    [[B5:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @b32, i32 0, i64 5), align 4
-; SSE-NEXT:    [[B6:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @b32, i32 0, i64 6), align 4
-; SSE-NEXT:    [[B7:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @b32, i32 0, i64 7), align 4
-; SSE-NEXT:    [[B8:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @b32, i32 0, i64 8), align 4
-; SSE-NEXT:    [[B9:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @b32, i32 0, i64 9), align 4
-; SSE-NEXT:    [[B10:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @b32, i32 0, i64 10), align 4
-; SSE-NEXT:    [[B11:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @b32, i32 0, i64 11), align 4
-; SSE-NEXT:    [[B12:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @b32, i32 0, i64 12), align 4
-; SSE-NEXT:    [[B13:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @b32, i32 0, i64 13), align 4
-; SSE-NEXT:    [[B14:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @b32, i32 0, i64 14), align 4
-; SSE-NEXT:    [[B15:%.*]] = load i32, i32* getelementptr inbounds ([16 x i32], [16 x i32]* @b32, i32 0, i64 15), align 4
-; SSE-NEXT:    [[R0:%.*]] = call i32 @llvm.smul.fix.i32(i32 [[A0]], i32 [[B0]], i32 3)
-; SSE-NEXT:    [[R1:%.*]] = call i32 @llvm.smul.fix.i32(i32 [[A1]], i32 [[B1]], i32 3)
-; SSE-NEXT:    [[R2:%.*]] = call i32 @llvm.smul.fix.i32(i32 [[A2]], i32 [[B2]], i32 3)
-; SSE-NEXT:    [[R3:%.*]] = call i32 @llvm.smul.fix.i32(i32 [[A3]], i32 [[B3]], i32 3)
-; SSE-NEXT:    [[R4:%.*]] = call i32 @llvm.smul.fix.i32(i32 [[A4]], i32 [[B4]], i32 3)
-; SSE-NEXT:    [[R5:%.*]] = call i32 @llvm.smul.fix.i32(i32 [[A5]], i32 [[B5]], i32 3)
-; SSE-NEXT:    [[R6:%.*]] = call i32 @llvm.smul.fix.i32(i32 [[A6]], i32 [[B6]], i32 3)
-; SSE-NEXT:    [[R7:%.*]] = call i32 @llvm.smul.fix.i32(i32 [[A7]], i32 [[B7]], i32 3)
-; SSE-NEXT:    [[R8:%.*]] = call i32 @llvm.smul.fix.i32(i32 [[A8]], i32 [[B8]], i32 3)
-; SSE-NEXT:    [[R9:%.*]] = call i32 @llvm.smul.fix.i32(i32 [[A9]], i32 [[B9]], i32 3)
-; SSE-NEXT:    [[R10:%.*]] = call i32 @llvm.smul.fix.i32(i32 [[A10]], i32 [[B10]], i32 3)
-; SSE-NEXT:    [[R11:%.*]] = call i32 @llvm.smul.fix.i32(i32 [[A11]], i32 [[B11]], i32 3)
-; SSE-NEXT:    [[R12:%.*]] = call i32 @llvm.smul.fix.i32(i32 [[A12]], i32 [[B12]], i32 3)
-; SSE-NEXT:    [[R13:%.*]] = call i32 @llvm.smul.fix.i32(i32 [[A13]], i32 [[B13]], i32 3)
-; SSE-NEXT:    [[R14:%.*]] = call i32 @llvm.smul.fix.i32(i32 [[A14]], i32 [[B14]], i32 3)
-; SSE-NEXT:    [[R15:%.*]] = call i32 @llvm.smul.fix.i32(i32 [[A15]], i32 [[B15]], i32 3)
-; SSE-NEXT:    store i32 [[R0]], i32* getelementptr inbounds ([16 x i32], [16 x i32]* @c32, i32 0, i64 0), align 4
-; SSE-NEXT:    store i32 [[R1]], i32* getelementptr inbounds ([16 x i32], [16 x i32]* @c32, i32 0, i64 1), align 4
-; SSE-NEXT:    store i32 [[R2]], i32* getelementptr inbounds ([16 x i32], [16 x i32]* @c32, i32 0, i64 2), align 4
-; SSE-NEXT:    store i32 [[R3]], i32* getelementptr inbounds ([16 x i32], [16 x i32]* @c32, i32 0, i64 3), align 4
-; SSE-NEXT:    store i32 [[R4]], i32* getelementptr inbounds ([16 x i32], [16 x i32]* @c32, i32 0, i64 4), align 4
-; SSE-NEXT:    store i32 [[R5]], i32* getelementptr inbounds ([16 x i32], [16 x i32]* @c32, i32 0, i64 5), align 4
-; SSE-NEXT:    store i32 [[R6]], i32* getelementptr inbounds ([16 x i32], [16 x i32]* @c32, i32 0, i64 6), align 4
-; SSE-NEXT:    store i32 [[R7]], i32* getelementptr inbounds ([16 x i32], [16 x i32]* @c32, i32 0, i64 7), align 4
-; SSE-NEXT:    store i32 [[R8]], i32* getelementptr inbounds ([16 x i32], [16 x i32]* @c32, i32 0, i64 8), align 4
-; SSE-NEXT:    store i32 [[R9]], i32* getelementptr inbounds ([16 x i32], [16 x i32]* @c32, i32 0, i64 9), align 4
-; SSE-NEXT:    store i32 [[R10]], i32* getelementptr inbounds ([16 x i32], [16 x i32]* @c32, i32 0, i64 10), align 4
-; SSE-NEXT:    store i32 [[R11]], i32* getelementptr inbounds ([16 x i32], [16 x i32]* @c32, i32 0, i64 11), align 4
-; SSE-NEXT:    store i32 [[R12]], i32* getelementptr inbounds ([16 x i32], [16 x i32]* @c32, i32 0, i64 12), align 4
-; SSE-NEXT:    store i32 [[R13]], i32* getelementptr inbounds ([16 x i32], [16 x i32]* @c32, i32 0, i64 13), align 4
-; SSE-NEXT:    store i32 [[R14]], i32* getelementptr inbounds ([16 x i32], [16 x i32]* @c32, i32 0, i64 14), align 4
-; SSE-NEXT:    store i32 [[R15]], i32* getelementptr inbounds ([16 x i32], [16 x i32]* @c32, i32 0, i64 15), align 4
+; SSE-NEXT:    [[TMP1:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([16 x i32]* @a32 to <4 x i32>*), align 4
+; SSE-NEXT:    [[TMP2:%.*]] = load <4 x i32>, <4 x i32>* bitcast (i32* getelementptr inbounds ([16 x i32], [16 x i32]* @a32, i32 0, i64 4) to <4 x i32>*), align 4
+; SSE-NEXT:    [[TMP3:%.*]] = load <4 x i32>, <4 x i32>* bitcast (i32* getelementptr inbounds ([16 x i32], [16 x i32]* @a32, i32 0, i64 8) to <4 x i32>*), align 4
+; SSE-NEXT:    [[TMP4:%.*]] = load <4 x i32>, <4 x i32>* bitcast (i32* getelementptr inbounds ([16 x i32], [16 x i32]* @a32, i32 0, i64 12) to <4 x i32>*), align 4
+; SSE-NEXT:    [[TMP5:%.*]] = load <4 x i32>, <4 x i32>* bitcast ([16 x i32]* @b32 to <4 x i32>*), align 4
+; SSE-NEXT:    [[TMP6:%.*]] = load <4 x i32>, <4 x i32>* bitcast (i32* getelementptr inbounds ([16 x i32], [16 x i32]* @b32, i32 0, i64 4) to <4 x i32>*), align 4
+; SSE-NEXT:    [[TMP7:%.*]] = load <4 x i32>, <4 x i32>* bitcast (i32* getelementptr inbounds ([16 x i32], [16 x i32]* @b32, i32 0, i64 8) to <4 x i32>*), align 4
+; SSE-NEXT:    [[TMP8:%.*]] = load <4 x i32>, <4 x i32>* bitcast (i32* getelementptr inbounds ([16 x i32], [16 x i32]* @b32, i32 0, i64 12) to <4 x i32>*), align 4
+; SSE-NEXT:    [[TMP9:%.*]] = call <4 x i32> @llvm.smul.fix.v4i32(<4 x i32> [[TMP1]], <4 x i32> [[TMP5]], i32 3)
+; SSE-NEXT:    [[TMP10:%.*]] = call <4 x i32> @llvm.smul.fix.v4i32(<4 x i32> [[TMP2]], <4 x i32> [[TMP6]], i32 3)
+; SSE-NEXT:    [[TMP11:%.*]] = call <4 x i32> @llvm.smul.fix.v4i32(<4 x i32> [[TMP3]], <4 x i32> [[TMP7]], i32 3)
+; SSE-NEXT:    [[TMP12:%.*]] = call <4 x i32> @llvm.smul.fix.v4i32(<4 x i32> [[TMP4]], <4 x i32> [[TMP8]], i32 3)
+; SSE-NEXT:    store <4 x i32> [[TMP9]], <4 x i32>* bitcast ([16 x i32]* @c32 to <4 x i32>*), align 4
+; SSE-NEXT:    store <4 x i32> [[TMP10]], <4 x i32>* bitcast (i32* getelementptr inbounds ([16 x i32], [16 x i32]* @c32, i32 0, i64 4) to <4 x i32>*), align 4
+; SSE-NEXT:    store <4 x i32> [[TMP11]], <4 x i32>* bitcast (i32* getelementptr inbounds ([16 x i32], [16 x i32]* @c32, i32 0, i64 8) to <4 x i32>*), align 4
+; SSE-NEXT:    store <4 x i32> [[TMP12]], <4 x i32>* bitcast (i32* getelementptr inbounds ([16 x i32], [16 x i32]* @c32, i32 0, i64 12) to <4 x i32>*), align 4
 ; SSE-NEXT:    ret void
 ;
 ; SLM-LABEL: @smul_v16i32(