Skip to content

Commit 0b3bc39

Browse files
fhahnfhahn
committed
[AArch64] Change cost of (s|z)ext <4 x i8> -> <4 x i32> to 2.
Increase the cost for zext <4 x i8> -> <4 x i32> to 2 to match codegen, which currently uses 2 instructions (commonly ushll.8h, ushll.4s). Example lowering: https://llvm.godbolt.org/z/58TEoxh4v
1 parent 4a158f6 commit 0b3bc39

File tree

6 files changed

+38
-36
lines changed

6 files changed

+38
-36
lines changed

‎llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp‎

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -3161,6 +3161,8 @@ InstructionCost AArch64TTIImpl::getCastInstrCost(unsigned Opcode, Type *Dst,
31613161
{ISD::ZERO_EXTEND, MVT::v4i64, MVT::v4i16, 3},
31623162
{ISD::SIGN_EXTEND, MVT::v4i64, MVT::v4i32, 2},
31633163
{ISD::ZERO_EXTEND, MVT::v4i64, MVT::v4i32, 2},
3164+
{ISD::SIGN_EXTEND, MVT::v4i32, MVT::v4i8, 2},
3165+
{ISD::ZERO_EXTEND, MVT::v4i32, MVT::v4i8, 2},
31643166
{ISD::SIGN_EXTEND, MVT::v8i32, MVT::v8i8, 3},
31653167
{ISD::ZERO_EXTEND, MVT::v8i32, MVT::v8i8, 3},
31663168
{ISD::SIGN_EXTEND, MVT::v8i32, MVT::v8i16, 2},

‎llvm/test/Analysis/CostModel/AArch64/arith-widening.ll‎

Lines changed: 18 additions & 18 deletions
Original file line numberDiff line numberDiff line change
@@ -293,15 +293,15 @@ define void @extaddv4(<4 x i8> %i8, <4 x i16> %i16, <4 x i32> %i32, <4 x i64> %i
293293
; CHECK-NEXT: Cost Model: Found costs of 1 for: %zl1_8_16 = zext <4 x i8> %i8 to <4 x i16>
294294
; CHECK-NEXT: Cost Model: Found costs of 1 for: %zl2_8_16 = zext <4 x i8> %i8 to <4 x i16>
295295
; CHECK-NEXT: Cost Model: Found costs of 1 for: %azl_8_16 = add <4 x i16> %zl1_8_16, %zl2_8_16
296-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %sw_8_32 = sext <4 x i8> %i8 to <4 x i32>
296+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sw_8_32 = sext <4 x i8> %i8 to <4 x i32>
297297
; CHECK-NEXT: Cost Model: Found costs of 1 for: %asw_8_32 = add <4 x i32> %i32, %sw_8_32
298-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %sl1_8_32 = sext <4 x i8> %i8 to <4 x i32>
299-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %sl2_8_32 = sext <4 x i8> %i8 to <4 x i32>
298+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sl1_8_32 = sext <4 x i8> %i8 to <4 x i32>
299+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sl2_8_32 = sext <4 x i8> %i8 to <4 x i32>
300300
; CHECK-NEXT: Cost Model: Found costs of 1 for: %asl_8_32 = add <4 x i32> %sl1_8_32, %sl2_8_32
301-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %zw_8_32 = zext <4 x i8> %i8 to <4 x i32>
301+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zw_8_32 = zext <4 x i8> %i8 to <4 x i32>
302302
; CHECK-NEXT: Cost Model: Found costs of 1 for: %azw_8_32 = add <4 x i32> %i32, %zw_8_32
303-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %zl1_8_32 = zext <4 x i8> %i8 to <4 x i32>
304-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %zl2_8_32 = zext <4 x i8> %i8 to <4 x i32>
303+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zl1_8_32 = zext <4 x i8> %i8 to <4 x i32>
304+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zl2_8_32 = zext <4 x i8> %i8 to <4 x i32>
305305
; CHECK-NEXT: Cost Model: Found costs of 1 for: %azl_8_32 = add <4 x i32> %zl1_8_32, %zl2_8_32
306306
; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %sw_8_64 = sext <4 x i8> %i8 to <4 x i64>
307307
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %asw_8_64 = add <4 x i64> %i64, %sw_8_64
@@ -988,15 +988,15 @@ define void @extsubv4(<4 x i8> %i8, <4 x i16> %i16, <4 x i32> %i32, <4 x i64> %i
988988
; CHECK-NEXT: Cost Model: Found costs of 1 for: %zl1_8_16 = zext <4 x i8> %i8 to <4 x i16>
989989
; CHECK-NEXT: Cost Model: Found costs of 1 for: %zl2_8_16 = zext <4 x i8> %i8 to <4 x i16>
990990
; CHECK-NEXT: Cost Model: Found costs of 1 for: %azl_8_16 = sub <4 x i16> %zl1_8_16, %zl2_8_16
991-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %sw_8_32 = sext <4 x i8> %i8 to <4 x i32>
991+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sw_8_32 = sext <4 x i8> %i8 to <4 x i32>
992992
; CHECK-NEXT: Cost Model: Found costs of 1 for: %asw_8_32 = sub <4 x i32> %i32, %sw_8_32
993-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %sl1_8_32 = sext <4 x i8> %i8 to <4 x i32>
994-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %sl2_8_32 = sext <4 x i8> %i8 to <4 x i32>
993+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sl1_8_32 = sext <4 x i8> %i8 to <4 x i32>
994+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sl2_8_32 = sext <4 x i8> %i8 to <4 x i32>
995995
; CHECK-NEXT: Cost Model: Found costs of 1 for: %asl_8_32 = sub <4 x i32> %sl1_8_32, %sl2_8_32
996-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %zw_8_32 = zext <4 x i8> %i8 to <4 x i32>
996+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zw_8_32 = zext <4 x i8> %i8 to <4 x i32>
997997
; CHECK-NEXT: Cost Model: Found costs of 1 for: %azw_8_32 = sub <4 x i32> %i32, %zw_8_32
998-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %zl1_8_32 = zext <4 x i8> %i8 to <4 x i32>
999-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %zl2_8_32 = zext <4 x i8> %i8 to <4 x i32>
998+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zl1_8_32 = zext <4 x i8> %i8 to <4 x i32>
999+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zl2_8_32 = zext <4 x i8> %i8 to <4 x i32>
10001000
; CHECK-NEXT: Cost Model: Found costs of 1 for: %azl_8_32 = sub <4 x i32> %zl1_8_32, %zl2_8_32
10011001
; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %sw_8_64 = sext <4 x i8> %i8 to <4 x i64>
10021002
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %asw_8_64 = sub <4 x i64> %i64, %sw_8_64
@@ -1683,15 +1683,15 @@ define void @extmulv4(<4 x i8> %i8, <4 x i16> %i16, <4 x i32> %i32, <4 x i64> %i
16831683
; CHECK-NEXT: Cost Model: Found costs of 1 for: %zl1_8_16 = zext <4 x i8> %i8 to <4 x i16>
16841684
; CHECK-NEXT: Cost Model: Found costs of 1 for: %zl2_8_16 = zext <4 x i8> %i8 to <4 x i16>
16851685
; CHECK-NEXT: Cost Model: Found costs of 1 for: %azl_8_16 = mul <4 x i16> %zl1_8_16, %zl2_8_16
1686-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %sw_8_32 = sext <4 x i8> %i8 to <4 x i32>
1686+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sw_8_32 = sext <4 x i8> %i8 to <4 x i32>
16871687
; CHECK-NEXT: Cost Model: Found costs of 1 for: %asw_8_32 = mul <4 x i32> %i32, %sw_8_32
1688-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %sl1_8_32 = sext <4 x i8> %i8 to <4 x i32>
1689-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %sl2_8_32 = sext <4 x i8> %i8 to <4 x i32>
1688+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sl1_8_32 = sext <4 x i8> %i8 to <4 x i32>
1689+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sl2_8_32 = sext <4 x i8> %i8 to <4 x i32>
16901690
; CHECK-NEXT: Cost Model: Found costs of 1 for: %asl_8_32 = mul <4 x i32> %sl1_8_32, %sl2_8_32
1691-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %zw_8_32 = zext <4 x i8> %i8 to <4 x i32>
1691+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zw_8_32 = zext <4 x i8> %i8 to <4 x i32>
16921692
; CHECK-NEXT: Cost Model: Found costs of 1 for: %azw_8_32 = mul <4 x i32> %i32, %zw_8_32
1693-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %zl1_8_32 = zext <4 x i8> %i8 to <4 x i32>
1694-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %zl2_8_32 = zext <4 x i8> %i8 to <4 x i32>
1693+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zl1_8_32 = zext <4 x i8> %i8 to <4 x i32>
1694+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zl2_8_32 = zext <4 x i8> %i8 to <4 x i32>
16951695
; CHECK-NEXT: Cost Model: Found costs of 1 for: %azl_8_32 = mul <4 x i32> %zl1_8_32, %zl2_8_32
16961696
; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %sw_8_64 = sext <4 x i8> %i8 to <4 x i64>
16971697
; CHECK-NEXT: Cost Model: Found costs of RThru:28 CodeSize:1 Lat:1 SizeLat:1 for: %asw_8_64 = mul <4 x i64> %i64, %sw_8_64

‎llvm/test/Analysis/CostModel/AArch64/cast.ll‎

Lines changed: 4 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -41,8 +41,8 @@ define void @ext() {
4141
; CHECK-NEXT: Cost Model: Found costs of 1 for: %z2i32i64 = zext <2 x i32> undef to <2 x i64>
4242
; CHECK-NEXT: Cost Model: Found costs of 1 for: %s4i8i16 = sext <4 x i8> undef to <4 x i16>
4343
; CHECK-NEXT: Cost Model: Found costs of 1 for: %z4i8i16 = zext <4 x i8> undef to <4 x i16>
44-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = sext <4 x i8> undef to <4 x i32>
45-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %z4i8i32 = zext <4 x i8> undef to <4 x i32>
44+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s4i8i32 = sext <4 x i8> undef to <4 x i32>
45+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z4i8i32 = zext <4 x i8> undef to <4 x i32>
4646
; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s4i8i64 = sext <4 x i8> undef to <4 x i64>
4747
; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %z4i8i64 = zext <4 x i8> undef to <4 x i64>
4848
; CHECK-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = sext <4 x i16> undef to <4 x i32>
@@ -883,8 +883,8 @@ define i32 @load_extends() {
883883
; CHECK-NEXT: Cost Model: Found costs of 0 for: %r11 = zext i32 %loadi32 to i64
884884
; CHECK-NEXT: Cost Model: Found costs of 1 for: %v0 = sext <8 x i8> %loadv8i8 to <8 x i16>
885885
; CHECK-NEXT: Cost Model: Found costs of 1 for: %v1 = zext <8 x i8> %loadv8i8 to <8 x i16>
886-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %v2 = sext <4 x i8> %loadv4i8 to <4 x i32>
887-
; CHECK-NEXT: Cost Model: Found costs of 1 for: %v3 = zext <4 x i8> %loadv4i8 to <4 x i32>
886+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %v2 = sext <4 x i8> %loadv4i8 to <4 x i32>
887+
; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %v3 = zext <4 x i8> %loadv4i8 to <4 x i32>
888888
; CHECK-NEXT: Cost Model: Found costs of 1 for: %v4 = sext <2 x i8> %loadv2i8 to <2 x i64>
889889
; CHECK-NEXT: Cost Model: Found costs of 1 for: %v5 = zext <2 x i8> %loadv2i8 to <2 x i64>
890890
; CHECK-NEXT: Cost Model: Found costs of 1 for: %v6 = sext <4 x i16> %loadv4i16 to <4 x i32>

‎llvm/test/Analysis/CostModel/AArch64/free-widening-casts.ll‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -582,7 +582,7 @@ define <8 x i16> @neg_dissimilar_operand_kind_0(<8 x i8> %a, <8 x i8> %b) {
582582
}
583583

584584
; COST-LABEL: neg_dissimilar_operand_kind_1
585-
; COST-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %tmp0 = zext <4 x i8> %a to <4 x i32>
585+
; COST-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %tmp0 = zext <4 x i8> %a to <4 x i32>
586586
; COST-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %tmp1 = zext <4 x i16> %b to <4 x i32>
587587
define <4 x i32> @neg_dissimilar_operand_kind_1(<4 x i8> %a, <4 x i16> %b) {
588588
%tmp0 = zext <4 x i8> %a to <4 x i32>

‎llvm/test/Analysis/CostModel/AArch64/sve-cast.ll‎

Lines changed: 12 additions & 12 deletions
Original file line numberDiff line numberDiff line change
@@ -42,8 +42,8 @@ define void @ext() {
4242
; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z2i32i64 = zext <2 x i32> undef to <2 x i64>
4343
; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s4i8i16 = sext <4 x i8> undef to <4 x i16>
4444
; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z4i8i16 = zext <4 x i8> undef to <4 x i16>
45-
; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = sext <4 x i8> undef to <4 x i32>
46-
; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z4i8i32 = zext <4 x i8> undef to <4 x i32>
45+
; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s4i8i32 = sext <4 x i8> undef to <4 x i32>
46+
; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z4i8i32 = zext <4 x i8> undef to <4 x i32>
4747
; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s4i8i64 = sext <4 x i8> undef to <4 x i64>
4848
; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %z4i8i64 = zext <4 x i8> undef to <4 x i64>
4949
; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = sext <4 x i16> undef to <4 x i32>
@@ -184,8 +184,8 @@ define void @ext() {
184184
; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z2i32i64 = zext <2 x i32> undef to <2 x i64>
185185
; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i8i16 = sext <4 x i8> undef to <4 x i16>
186186
; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z4i8i16 = zext <4 x i8> undef to <4 x i16>
187-
; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = sext <4 x i8> undef to <4 x i32>
188-
; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z4i8i32 = zext <4 x i8> undef to <4 x i32>
187+
; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s4i8i32 = sext <4 x i8> undef to <4 x i32>
188+
; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z4i8i32 = zext <4 x i8> undef to <4 x i32>
189189
; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i8i64 = sext <4 x i8> undef to <4 x i64>
190190
; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z4i8i64 = zext <4 x i8> undef to <4 x i64>
191191
; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = sext <4 x i16> undef to <4 x i32>
@@ -255,8 +255,8 @@ define void @ext() {
255255
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z2i32i64 = zext <2 x i32> undef to <2 x i64>
256256
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i8i16 = sext <4 x i8> undef to <4 x i16>
257257
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z4i8i16 = zext <4 x i8> undef to <4 x i16>
258-
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = sext <4 x i8> undef to <4 x i32>
259-
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z4i8i32 = zext <4 x i8> undef to <4 x i32>
258+
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s4i8i32 = sext <4 x i8> undef to <4 x i32>
259+
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z4i8i32 = zext <4 x i8> undef to <4 x i32>
260260
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i8i64 = sext <4 x i8> undef to <4 x i64>
261261
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z4i8i64 = zext <4 x i8> undef to <4 x i64>
262262
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = sext <4 x i16> undef to <4 x i32>
@@ -1809,8 +1809,8 @@ define i32 @load_extends() #0 {
18091809
; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r11 = zext i32 %loadi32 to i64
18101810
; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %v0 = sext <8 x i8> %loadv8i8 to <8 x i16>
18111811
; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %v1 = zext <8 x i8> %loadv8i8 to <8 x i16>
1812-
; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %v2 = sext <4 x i8> %loadv4i8 to <4 x i32>
1813-
; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %v3 = zext <4 x i8> %loadv4i8 to <4 x i32>
1812+
; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %v2 = sext <4 x i8> %loadv4i8 to <4 x i32>
1813+
; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %v3 = zext <4 x i8> %loadv4i8 to <4 x i32>
18141814
; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %v4 = sext <2 x i8> %loadv2i8 to <2 x i64>
18151815
; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %v5 = zext <2 x i8> %loadv2i8 to <2 x i64>
18161816
; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %v6 = sext <4 x i16> %loadv4i16 to <4 x i32>
@@ -1899,8 +1899,8 @@ define i32 @load_extends() #0 {
18991899
; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r11 = zext i32 %loadi32 to i64
19001900
; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %v0 = sext <8 x i8> %loadv8i8 to <8 x i16>
19011901
; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %v1 = zext <8 x i8> %loadv8i8 to <8 x i16>
1902-
; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %v2 = sext <4 x i8> %loadv4i8 to <4 x i32>
1903-
; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %v3 = zext <4 x i8> %loadv4i8 to <4 x i32>
1902+
; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %v2 = sext <4 x i8> %loadv4i8 to <4 x i32>
1903+
; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %v3 = zext <4 x i8> %loadv4i8 to <4 x i32>
19041904
; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %v4 = sext <2 x i8> %loadv2i8 to <2 x i64>
19051905
; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %v5 = zext <2 x i8> %loadv2i8 to <2 x i64>
19061906
; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %v6 = sext <4 x i16> %loadv4i16 to <4 x i32>
@@ -1944,8 +1944,8 @@ define i32 @load_extends() #0 {
19441944
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r11 = zext i32 %loadi32 to i64
19451945
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %v0 = sext <8 x i8> %loadv8i8 to <8 x i16>
19461946
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %v1 = zext <8 x i8> %loadv8i8 to <8 x i16>
1947-
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %v2 = sext <4 x i8> %loadv4i8 to <4 x i32>
1948-
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %v3 = zext <4 x i8> %loadv4i8 to <4 x i32>
1947+
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %v2 = sext <4 x i8> %loadv4i8 to <4 x i32>
1948+
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %v3 = zext <4 x i8> %loadv4i8 to <4 x i32>
19491949
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %v4 = sext <2 x i8> %loadv2i8 to <2 x i64>
19501950
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %v5 = zext <2 x i8> %loadv2i8 to <2 x i64>
19511951
; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %v6 = sext <4 x i16> %loadv4i16 to <4 x i32>

‎llvm/test/Transforms/SLPVectorizer/AArch64/vecreduceadd.ll‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -883,7 +883,7 @@ entry:
883883

884884

885885
; COST-LABEL: Function: mla_v4i8_i32
886-
; COST: Cost: '-6'
886+
; COST: Cost: '-4'
887887
define i32 @mla_v4i8_i32(ptr %x, ptr %y) "target-features"="+dotprod" {
888888
; CHECK-LABEL: @mla_v4i8_i32(
889889
; CHECK-NEXT: entry:

0 commit comments

Comments
 (0)