[llvm] e8413ac - [AArch64] Expand some vector of i64 reductions on NEON
Cameron McInally via llvm-commits
llvm-commits at lists.llvm.org
Wed Sep 23 14:01:48 PDT 2020
Author: Cameron McInally
Date: 2020-09-23T16:01:24-05:00
New Revision: e8413ac97f6ca2b1897cc9555ad9b0194456629f
URL: https://github.com/llvm/llvm-project/commit/e8413ac97f6ca2b1897cc9555ad9b0194456629f
DIFF: https://github.com/llvm/llvm-project/commit/e8413ac97f6ca2b1897cc9555ad9b0194456629f.diff
LOG: [AArch64] Expand some vector of i64 reductions on NEON
With the exception of VECREDUCE_ADD, there are no NEON instructions to support vector of i64 reductions. This patch removes the Custom lowerings for those and adds some test coverage to confirm.
Differential Revision: https://reviews.llvm.org/D88161
Added:
Modified:
llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
llvm/test/CodeGen/AArch64/aarch64-addv.ll
llvm/test/CodeGen/AArch64/vecreduce-umax-legalization.ll
Removed:
################################################################################
diff --git a/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp b/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
index 86354f2c06bc..772f88d4f9d1 100644
--- a/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
+++ b/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
@@ -889,26 +889,30 @@ AArch64TargetLowering::AArch64TargetLowering(const TargetMachine &TM,
setOperationAction(ISD::MUL, MVT::v4i32, Custom);
setOperationAction(ISD::MUL, MVT::v2i64, Custom);
+ // Saturates
for (MVT VT : { MVT::v8i8, MVT::v4i16, MVT::v2i32,
MVT::v16i8, MVT::v8i16, MVT::v4i32, MVT::v2i64 }) {
- // Vector reductions
- setOperationAction(ISD::VECREDUCE_ADD, VT, Custom);
- setOperationAction(ISD::VECREDUCE_SMAX, VT, Custom);
- setOperationAction(ISD::VECREDUCE_SMIN, VT, Custom);
- setOperationAction(ISD::VECREDUCE_UMAX, VT, Custom);
- setOperationAction(ISD::VECREDUCE_UMIN, VT, Custom);
-
- // Saturates
setOperationAction(ISD::SADDSAT, VT, Legal);
setOperationAction(ISD::UADDSAT, VT, Legal);
setOperationAction(ISD::SSUBSAT, VT, Legal);
setOperationAction(ISD::USUBSAT, VT, Legal);
}
+
+ // Vector reductions
for (MVT VT : { MVT::v4f16, MVT::v2f32,
MVT::v8f16, MVT::v4f32, MVT::v2f64 }) {
setOperationAction(ISD::VECREDUCE_FMAX, VT, Custom);
setOperationAction(ISD::VECREDUCE_FMIN, VT, Custom);
}
+ for (MVT VT : { MVT::v8i8, MVT::v4i16, MVT::v2i32,
+ MVT::v16i8, MVT::v8i16, MVT::v4i32 }) {
+ setOperationAction(ISD::VECREDUCE_ADD, VT, Custom);
+ setOperationAction(ISD::VECREDUCE_SMAX, VT, Custom);
+ setOperationAction(ISD::VECREDUCE_SMIN, VT, Custom);
+ setOperationAction(ISD::VECREDUCE_UMAX, VT, Custom);
+ setOperationAction(ISD::VECREDUCE_UMIN, VT, Custom);
+ }
+ setOperationAction(ISD::VECREDUCE_ADD, MVT::v2i64, Custom);
setOperationAction(ISD::ANY_EXTEND, MVT::v4i32, Legal);
setTruncStoreAction(MVT::v2i32, MVT::v2i16, Expand);
diff --git a/llvm/test/CodeGen/AArch64/aarch64-addv.ll b/llvm/test/CodeGen/AArch64/aarch64-addv.ll
index 2fd57f70a65d..f1e0f299bd49 100644
--- a/llvm/test/CodeGen/AArch64/aarch64-addv.ll
+++ b/llvm/test/CodeGen/AArch64/aarch64-addv.ll
@@ -33,6 +33,7 @@ define i32 @add_S( <4 x i32>* %arr) {
define i64 @add_D(<2 x i64>* %arr) {
; CHECK-LABEL: add_D
; CHECK-NOT: addv
+; CHECK: addp {{d[0-9]+}}, {{v[0-9]+}}.2d
%bin.rdx = load <2 x i64>, <2 x i64>* %arr
%r = call i64 @llvm.experimental.vector.reduce.add.v2i64(<2 x i64> %bin.rdx)
ret i64 %r
diff --git a/llvm/test/CodeGen/AArch64/vecreduce-umax-legalization.ll b/llvm/test/CodeGen/AArch64/vecreduce-umax-legalization.ll
index 99b6ebe2f1da..7c9415211681 100644
--- a/llvm/test/CodeGen/AArch64/vecreduce-umax-legalization.ll
+++ b/llvm/test/CodeGen/AArch64/vecreduce-umax-legalization.ll
@@ -9,6 +9,7 @@ declare i32 @llvm.experimental.vector.reduce.umax.v1i32(<1 x i32> %a)
declare i64 @llvm.experimental.vector.reduce.umax.v1i64(<1 x i64> %a)
declare i128 @llvm.experimental.vector.reduce.umax.v1i128(<1 x i128> %a)
+declare i64 @llvm.experimental.vector.reduce.umax.v2i64(<2 x i64> %a)
declare i8 @llvm.experimental.vector.reduce.umax.v3i8(<3 x i8> %a)
declare i8 @llvm.experimental.vector.reduce.umax.v9i8(<9 x i8> %a)
declare i32 @llvm.experimental.vector.reduce.umax.v3i32(<3 x i32> %a)
@@ -82,6 +83,19 @@ define i128 @test_v1i128(<1 x i128> %a) nounwind {
ret i128 %b
}
+; No i64 vector support for UMAX.
+define i64 @test_v2i64(<2 x i64> %a) nounwind {
+; CHECK-LABEL: test_v2i64:
+; CHECK: // %bb.0:
+; CHECK-NEXT: mov x8, v0.d[1]
+; CHECK-NEXT: fmov x9, d0
+; CHECK-NEXT: cmp x9, x8
+; CHECK-NEXT: csel x0, x9, x8, hi
+; CHECK-NEXT: ret
+ %b = call i64 @llvm.experimental.vector.reduce.umax.v2i64(<2 x i64> %a)
+ ret i64 %b
+}
+
define i8 @test_v3i8(<3 x i8> %a) nounwind {
; CHECK-LABEL: test_v3i8:
; CHECK: // %bb.0:
More information about the llvm-commits
mailing list