[LLVM][AArch64] Add "u" variants of SVE saturating/rounding shift left intrinsics. #170503
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…t intrinsics. This follows the same path used for regular shift intrinsics whereby the "don't care about the result of inactive lanes" property of the associated _x builtins is preserved. In doing this we gain the ability to switch between reversed and movprfx variants depending on what best fits register allocation.
llvmbot
added
backend:AArch64
clang:frontend
Member
|
@llvm/pr-subscribers-llvm-transforms @llvm/pr-subscribers-backend-aarch64 Author: Paul Walker (paulwalker-arm) ChangesThis follows the same path used for regular shift intrinsics whereby the "don't care about the result of inactive lanes" property of the associated _x builtins is preserved. In doing this we gain the ability to switch between reversed and movprfx variants depending on what best fits register allocation. Patch is 182.16 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/170503.diff 11 Files Affected:
diff --git a/clang/include/clang/Basic/arm_sve.td b/clang/include/clang/Basic/arm_sve.td
index 8dc40a665bd9a..780891152845b 100644
--- a/clang/include/clang/Basic/arm_sve.td
+++ b/clang/include/clang/Basic/arm_sve.td
@@ -1293,25 +1293,25 @@ defm SVRSQRTE : SInstZPZ<"svrsqrte", "Ui", "aarch64_sve_ursqrte">;
//------------------------------------------------------------------------------
-multiclass SInstZPZxZ<string name, string types, string pat_v, string pat_n, string intrinsic, list<FlagType> flags=[]> {
- def _M : SInst<name # "[_{d}]", pat_v, types, MergeOp1, intrinsic, flags>;
- def _X : SInst<name # "[_{d}]", pat_v, types, MergeAny, intrinsic, flags>;
- def _Z : SInst<name # "[_{d}]", pat_v, types, MergeZero, intrinsic, flags>;
+multiclass SInstZPZxZ<string name, string types, string pat_v, string pat_n, string m_intrinsic, string x_intrinsic, list<FlagType> flags=[]> {
+ def _M : SInst<name # "[_{d}]", pat_v, types, MergeOp1, m_intrinsic, flags>;
+ def _X : SInst<name # "[_{d}]", pat_v, types, MergeAny, x_intrinsic, flags>;
+ def _Z : SInst<name # "[_{d}]", pat_v, types, MergeZero, m_intrinsic, flags>;
- def _N_M : SInst<name # "[_n_{d}]", pat_n, types, MergeOp1, intrinsic, flags>;
- def _N_X : SInst<name # "[_n_{d}]", pat_n, types, MergeAny, intrinsic, flags>;
- def _N_Z : SInst<name # "[_n_{d}]", pat_n, types, MergeZero, intrinsic, flags>;
+ def _N_M : SInst<name # "[_n_{d}]", pat_n, types, MergeOp1, m_intrinsic, flags>;
+ def _N_X : SInst<name # "[_n_{d}]", pat_n, types, MergeAny, x_intrinsic, flags>;
+ def _N_Z : SInst<name # "[_n_{d}]", pat_n, types, MergeZero, m_intrinsic, flags>;
}
let SVETargetGuard = "sve2|sme" in {
-defm SVQRSHL_S : SInstZPZxZ<"svqrshl", "csil", "dPdx", "dPdK", "aarch64_sve_sqrshl", [VerifyRuntimeMode]>;
-defm SVQRSHL_U : SInstZPZxZ<"svqrshl", "UcUsUiUl", "dPdx", "dPdK", "aarch64_sve_uqrshl", [VerifyRuntimeMode]>;
-defm SVQSHL_S : SInstZPZxZ<"svqshl", "csil", "dPdx", "dPdK", "aarch64_sve_sqshl", [VerifyRuntimeMode]>;
-defm SVQSHL_U : SInstZPZxZ<"svqshl", "UcUsUiUl", "dPdx", "dPdK", "aarch64_sve_uqshl", [VerifyRuntimeMode]>;
-defm SVRSHL_S : SInstZPZxZ<"svrshl", "csil", "dPdx", "dPdK", "aarch64_sve_srshl", [VerifyRuntimeMode]>;
-defm SVRSHL_U : SInstZPZxZ<"svrshl", "UcUsUiUl", "dPdx", "dPdK", "aarch64_sve_urshl", [VerifyRuntimeMode]>;
-defm SVSQADD : SInstZPZxZ<"svsqadd", "UcUsUiUl", "dPdx", "dPdK", "aarch64_sve_usqadd", [VerifyRuntimeMode]>;
-defm SVUQADD : SInstZPZxZ<"svuqadd", "csil", "dPdu", "dPdL", "aarch64_sve_suqadd", [VerifyRuntimeMode]>;
+defm SVQRSHL_S : SInstZPZxZ<"svqrshl", "csil", "dPdx", "dPdK", "aarch64_sve_sqrshl", "aarch64_sve_sqrshl_u", [VerifyRuntimeMode]>;
+defm SVQRSHL_U : SInstZPZxZ<"svqrshl", "UcUsUiUl", "dPdx", "dPdK", "aarch64_sve_uqrshl", "aarch64_sve_uqrshl_u", [VerifyRuntimeMode]>;
+defm SVQSHL_S : SInstZPZxZ<"svqshl", "csil", "dPdx", "dPdK", "aarch64_sve_sqshl", "aarch64_sve_sqshl_u", [VerifyRuntimeMode]>;
+defm SVQSHL_U : SInstZPZxZ<"svqshl", "UcUsUiUl", "dPdx", "dPdK", "aarch64_sve_uqshl", "aarch64_sve_uqshl_u", [VerifyRuntimeMode]>;
+defm SVRSHL_S : SInstZPZxZ<"svrshl", "csil", "dPdx", "dPdK", "aarch64_sve_srshl", "aarch64_sve_srshl_u", [VerifyRuntimeMode]>;
+defm SVRSHL_U : SInstZPZxZ<"svrshl", "UcUsUiUl", "dPdx", "dPdK", "aarch64_sve_urshl", "aarch64_sve_urshl_u", [VerifyRuntimeMode]>;
+defm SVSQADD : SInstZPZxZ<"svsqadd", "UcUsUiUl", "dPdx", "dPdK", "aarch64_sve_usqadd", "aarch64_sve_usqadd", [VerifyRuntimeMode]>;
+defm SVUQADD : SInstZPZxZ<"svuqadd", "csil", "dPdu", "dPdL", "aarch64_sve_suqadd", "aarch64_sve_suqadd", [VerifyRuntimeMode]>;
def SVABA_S : SInst<"svaba[_{d}]", "dddd", "csil" , MergeNone, "aarch64_sve_saba", [VerifyRuntimeMode]>;
def SVABA_U : SInst<"svaba[_{d}]", "dddd", "UcUsUiUl", MergeNone, "aarch64_sve_uaba", [VerifyRuntimeMode]>;
diff --git a/clang/test/CodeGen/AArch64/sve2-intrinsics/acle_sve2_qrshl.c b/clang/test/CodeGen/AArch64/sve2-intrinsics/acle_sve2_qrshl.c
index cb1793d98418a..5b47497286847 100644
--- a/clang/test/CodeGen/AArch64/sve2-intrinsics/acle_sve2_qrshl.c
+++ b/clang/test/CodeGen/AArch64/sve2-intrinsics/acle_sve2_qrshl.c
@@ -297,12 +297,12 @@ svuint64_t test_svqrshl_u64_m(svbool_t pg, svuint64_t op1, svint64_t op2)
// CHECK-LABEL: @test_svqrshl_s8_x(
// CHECK-NEXT: entry:
-// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.sqrshl.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]])
+// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.sqrshl.u.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]])
// CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
//
// CPP-CHECK-LABEL: @_Z17test_svqrshl_s8_xu10__SVBool_tu10__SVInt8_tS0_(
// CPP-CHECK-NEXT: entry:
-// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.sqrshl.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]])
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.sqrshl.u.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
//
svint8_t test_svqrshl_s8_x(svbool_t pg, svint8_t op1, svint8_t op2)
@@ -313,13 +313,13 @@ svint8_t test_svqrshl_s8_x(svbool_t pg, svint8_t op1, svint8_t op2)
// CHECK-LABEL: @test_svqrshl_s16_x(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.sqrshl.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]])
+// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.sqrshl.u.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]])
// CHECK-NEXT: ret <vscale x 8 x i16> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z18test_svqrshl_s16_xu10__SVBool_tu11__SVInt16_tS0_(
// CPP-CHECK-NEXT: entry:
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.sqrshl.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]])
+// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.sqrshl.u.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP1]]
//
svint16_t test_svqrshl_s16_x(svbool_t pg, svint16_t op1, svint16_t op2)
@@ -330,13 +330,13 @@ svint16_t test_svqrshl_s16_x(svbool_t pg, svint16_t op1, svint16_t op2)
// CHECK-LABEL: @test_svqrshl_s32_x(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.sqrshl.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]])
+// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.sqrshl.u.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]])
// CHECK-NEXT: ret <vscale x 4 x i32> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z18test_svqrshl_s32_xu10__SVBool_tu11__SVInt32_tS0_(
// CPP-CHECK-NEXT: entry:
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.sqrshl.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]])
+// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.sqrshl.u.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP1]]
//
svint32_t test_svqrshl_s32_x(svbool_t pg, svint32_t op1, svint32_t op2)
@@ -347,13 +347,13 @@ svint32_t test_svqrshl_s32_x(svbool_t pg, svint32_t op1, svint32_t op2)
// CHECK-LABEL: @test_svqrshl_s64_x(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.sqrshl.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
+// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.sqrshl.u.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
// CHECK-NEXT: ret <vscale x 2 x i64> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z18test_svqrshl_s64_xu10__SVBool_tu11__SVInt64_tS0_(
// CPP-CHECK-NEXT: entry:
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.sqrshl.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
+// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.sqrshl.u.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP1]]
//
svint64_t test_svqrshl_s64_x(svbool_t pg, svint64_t op1, svint64_t op2)
@@ -363,12 +363,12 @@ svint64_t test_svqrshl_s64_x(svbool_t pg, svint64_t op1, svint64_t op2)
// CHECK-LABEL: @test_svqrshl_u8_x(
// CHECK-NEXT: entry:
-// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.uqrshl.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]])
+// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.uqrshl.u.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]])
// CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
//
// CPP-CHECK-LABEL: @_Z17test_svqrshl_u8_xu10__SVBool_tu11__SVUint8_tu10__SVInt8_t(
// CPP-CHECK-NEXT: entry:
-// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.uqrshl.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]])
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.uqrshl.u.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
//
svuint8_t test_svqrshl_u8_x(svbool_t pg, svuint8_t op1, svint8_t op2)
@@ -379,13 +379,13 @@ svuint8_t test_svqrshl_u8_x(svbool_t pg, svuint8_t op1, svint8_t op2)
// CHECK-LABEL: @test_svqrshl_u16_x(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.uqrshl.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]])
+// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.uqrshl.u.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]])
// CHECK-NEXT: ret <vscale x 8 x i16> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z18test_svqrshl_u16_xu10__SVBool_tu12__SVUint16_tu11__SVInt16_t(
// CPP-CHECK-NEXT: entry:
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.uqrshl.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]])
+// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.uqrshl.u.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP1]]
//
svuint16_t test_svqrshl_u16_x(svbool_t pg, svuint16_t op1, svint16_t op2)
@@ -396,13 +396,13 @@ svuint16_t test_svqrshl_u16_x(svbool_t pg, svuint16_t op1, svint16_t op2)
// CHECK-LABEL: @test_svqrshl_u32_x(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.uqrshl.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]])
+// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.uqrshl.u.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]])
// CHECK-NEXT: ret <vscale x 4 x i32> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z18test_svqrshl_u32_xu10__SVBool_tu12__SVUint32_tu11__SVInt32_t(
// CPP-CHECK-NEXT: entry:
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.uqrshl.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]])
+// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.uqrshl.u.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP1]]
//
svuint32_t test_svqrshl_u32_x(svbool_t pg, svuint32_t op1, svint32_t op2)
@@ -413,13 +413,13 @@ svuint32_t test_svqrshl_u32_x(svbool_t pg, svuint32_t op1, svint32_t op2)
// CHECK-LABEL: @test_svqrshl_u64_x(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.uqrshl.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
+// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.uqrshl.u.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
// CHECK-NEXT: ret <vscale x 2 x i64> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z18test_svqrshl_u64_xu10__SVBool_tu12__SVUint64_tu11__SVInt64_t(
// CPP-CHECK-NEXT: entry:
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.uqrshl.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
+// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.uqrshl.u.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP1]]
//
svuint64_t test_svqrshl_u64_x(svbool_t pg, svuint64_t op1, svint64_t op2)
@@ -775,14 +775,14 @@ svuint64_t test_svqrshl_n_u64_m(svbool_t pg, svuint64_t op1, int64_t op2)
// CHECK-NEXT: entry:
// CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 16 x i8> poison, i8 [[OP2:%.*]], i64 0
// CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 16 x i8> [[DOTSPLATINSERT]], <vscale x 16 x i8> poison, <vscale x 16 x i32> zeroinitializer
-// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.sqrshl.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[DOTSPLAT]])
+// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.sqrshl.u.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[DOTSPLAT]])
// CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
//
// CPP-CHECK-LABEL: @_Z19test_svqrshl_n_s8_xu10__SVBool_tu10__SVInt8_ta(
// CPP-CHECK-NEXT: entry:
// CPP-CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 16 x i8> poison, i8 [[OP2:%.*]], i64 0
// CPP-CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 16 x i8> [[DOTSPLATINSERT]], <vscale x 16 x i8> poison, <vscale x 16 x i32> zeroinitializer
-// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.sqrshl.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[DOTSPLAT]])
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.sqrshl.u.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[DOTSPLAT]])
// CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
//
svint8_t test_svqrshl_n_s8_x(svbool_t pg, svint8_t op1, int8_t op2)
@@ -795,7 +795,7 @@ svint8_t test_svqrshl_n_s8_x(svbool_t pg, svint8_t op1, int8_t op2)
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x i16> poison, i16 [[OP2:%.*]], i64 0
// CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x i16> [[DOTSPLATINSERT]], <vscale x 8 x i16> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.sqrshl.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[DOTSPLAT]])
+// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.sqrshl.u.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[DOTSPLAT]])
// CHECK-NEXT: ret <vscale x 8 x i16> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z20test_svqrshl_n_s16_xu10__SVBool_tu11__SVInt16_ts(
@@ -803,7 +803,7 @@ svint8_t test_svqrshl_n_s8_x(svbool_t pg, svint8_t op1, int8_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
// CPP-CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x i16> poison, i16 [[OP2:%.*]], i64 0
// CPP-CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x i16> [[DOTSPLATINSERT]], <vscale x 8 x i16> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.sqrshl.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[DOTSPLAT]])
+// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.sqrshl.u.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[DOTSPLAT]])
// CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP1]]
//
svint16_t test_svqrshl_n_s16_x(svbool_t pg, svint16_t op1, int16_t op2)
@@ -816,7 +816,7 @@ svint16_t test_svqrshl_n_s16_x(svbool_t pg, svint16_t op1, int16_t op2)
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x i32> poison, i32 [[OP2:%.*]], i64 0
// CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x i32> [[DOTSPLATINSERT]], <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.sqrshl.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[DOTSPLAT]])
+// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.sqrshl.u.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[DOTSPLAT]])
// CHECK-NEXT: ret <vscale x 4 x i32> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z20test_svqrshl_n_s32_xu10__SVBool_tu11__SVInt32_ti(
@@ -824,7 +824,7 @@ svint16_t test_svqrshl_n_s16_x(svbool_t pg, svint16_t op1, int16_t ...
[truncated]
|
Member
|
@llvm/pr-subscribers-llvm-ir Author: Paul Walker (paulwalker-arm) ChangesThis follows the same path used for regular shift intrinsics whereby the "don't care about the result of inactive lanes" property of the associated _x builtins is preserved. In doing this we gain the ability to switch between reversed and movprfx variants depending on what best fits register allocation. Patch is 182.16 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/170503.diff 11 Files Affected:
diff --git a/clang/include/clang/Basic/arm_sve.td b/clang/include/clang/Basic/arm_sve.td
index 8dc40a665bd9a..780891152845b 100644
--- a/clang/include/clang/Basic/arm_sve.td
+++ b/clang/include/clang/Basic/arm_sve.td
@@ -1293,25 +1293,25 @@ defm SVRSQRTE : SInstZPZ<"svrsqrte", "Ui", "aarch64_sve_ursqrte">;
//------------------------------------------------------------------------------
-multiclass SInstZPZxZ<string name, string types, string pat_v, string pat_n, string intrinsic, list<FlagType> flags=[]> {
- def _M : SInst<name # "[_{d}]", pat_v, types, MergeOp1, intrinsic, flags>;
- def _X : SInst<name # "[_{d}]", pat_v, types, MergeAny, intrinsic, flags>;
- def _Z : SInst<name # "[_{d}]", pat_v, types, MergeZero, intrinsic, flags>;
+multiclass SInstZPZxZ<string name, string types, string pat_v, string pat_n, string m_intrinsic, string x_intrinsic, list<FlagType> flags=[]> {
+ def _M : SInst<name # "[_{d}]", pat_v, types, MergeOp1, m_intrinsic, flags>;
+ def _X : SInst<name # "[_{d}]", pat_v, types, MergeAny, x_intrinsic, flags>;
+ def _Z : SInst<name # "[_{d}]", pat_v, types, MergeZero, m_intrinsic, flags>;
- def _N_M : SInst<name # "[_n_{d}]", pat_n, types, MergeOp1, intrinsic, flags>;
- def _N_X : SInst<name # "[_n_{d}]", pat_n, types, MergeAny, intrinsic, flags>;
- def _N_Z : SInst<name # "[_n_{d}]", pat_n, types, MergeZero, intrinsic, flags>;
+ def _N_M : SInst<name # "[_n_{d}]", pat_n, types, MergeOp1, m_intrinsic, flags>;
+ def _N_X : SInst<name # "[_n_{d}]", pat_n, types, MergeAny, x_intrinsic, flags>;
+ def _N_Z : SInst<name # "[_n_{d}]", pat_n, types, MergeZero, m_intrinsic, flags>;
}
let SVETargetGuard = "sve2|sme" in {
-defm SVQRSHL_S : SInstZPZxZ<"svqrshl", "csil", "dPdx", "dPdK", "aarch64_sve_sqrshl", [VerifyRuntimeMode]>;
-defm SVQRSHL_U : SInstZPZxZ<"svqrshl", "UcUsUiUl", "dPdx", "dPdK", "aarch64_sve_uqrshl", [VerifyRuntimeMode]>;
-defm SVQSHL_S : SInstZPZxZ<"svqshl", "csil", "dPdx", "dPdK", "aarch64_sve_sqshl", [VerifyRuntimeMode]>;
-defm SVQSHL_U : SInstZPZxZ<"svqshl", "UcUsUiUl", "dPdx", "dPdK", "aarch64_sve_uqshl", [VerifyRuntimeMode]>;
-defm SVRSHL_S : SInstZPZxZ<"svrshl", "csil", "dPdx", "dPdK", "aarch64_sve_srshl", [VerifyRuntimeMode]>;
-defm SVRSHL_U : SInstZPZxZ<"svrshl", "UcUsUiUl", "dPdx", "dPdK", "aarch64_sve_urshl", [VerifyRuntimeMode]>;
-defm SVSQADD : SInstZPZxZ<"svsqadd", "UcUsUiUl", "dPdx", "dPdK", "aarch64_sve_usqadd", [VerifyRuntimeMode]>;
-defm SVUQADD : SInstZPZxZ<"svuqadd", "csil", "dPdu", "dPdL", "aarch64_sve_suqadd", [VerifyRuntimeMode]>;
+defm SVQRSHL_S : SInstZPZxZ<"svqrshl", "csil", "dPdx", "dPdK", "aarch64_sve_sqrshl", "aarch64_sve_sqrshl_u", [VerifyRuntimeMode]>;
+defm SVQRSHL_U : SInstZPZxZ<"svqrshl", "UcUsUiUl", "dPdx", "dPdK", "aarch64_sve_uqrshl", "aarch64_sve_uqrshl_u", [VerifyRuntimeMode]>;
+defm SVQSHL_S : SInstZPZxZ<"svqshl", "csil", "dPdx", "dPdK", "aarch64_sve_sqshl", "aarch64_sve_sqshl_u", [VerifyRuntimeMode]>;
+defm SVQSHL_U : SInstZPZxZ<"svqshl", "UcUsUiUl", "dPdx", "dPdK", "aarch64_sve_uqshl", "aarch64_sve_uqshl_u", [VerifyRuntimeMode]>;
+defm SVRSHL_S : SInstZPZxZ<"svrshl", "csil", "dPdx", "dPdK", "aarch64_sve_srshl", "aarch64_sve_srshl_u", [VerifyRuntimeMode]>;
+defm SVRSHL_U : SInstZPZxZ<"svrshl", "UcUsUiUl", "dPdx", "dPdK", "aarch64_sve_urshl", "aarch64_sve_urshl_u", [VerifyRuntimeMode]>;
+defm SVSQADD : SInstZPZxZ<"svsqadd", "UcUsUiUl", "dPdx", "dPdK", "aarch64_sve_usqadd", "aarch64_sve_usqadd", [VerifyRuntimeMode]>;
+defm SVUQADD : SInstZPZxZ<"svuqadd", "csil", "dPdu", "dPdL", "aarch64_sve_suqadd", "aarch64_sve_suqadd", [VerifyRuntimeMode]>;
def SVABA_S : SInst<"svaba[_{d}]", "dddd", "csil" , MergeNone, "aarch64_sve_saba", [VerifyRuntimeMode]>;
def SVABA_U : SInst<"svaba[_{d}]", "dddd", "UcUsUiUl", MergeNone, "aarch64_sve_uaba", [VerifyRuntimeMode]>;
diff --git a/clang/test/CodeGen/AArch64/sve2-intrinsics/acle_sve2_qrshl.c b/clang/test/CodeGen/AArch64/sve2-intrinsics/acle_sve2_qrshl.c
index cb1793d98418a..5b47497286847 100644
--- a/clang/test/CodeGen/AArch64/sve2-intrinsics/acle_sve2_qrshl.c
+++ b/clang/test/CodeGen/AArch64/sve2-intrinsics/acle_sve2_qrshl.c
@@ -297,12 +297,12 @@ svuint64_t test_svqrshl_u64_m(svbool_t pg, svuint64_t op1, svint64_t op2)
// CHECK-LABEL: @test_svqrshl_s8_x(
// CHECK-NEXT: entry:
-// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.sqrshl.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]])
+// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.sqrshl.u.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]])
// CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
//
// CPP-CHECK-LABEL: @_Z17test_svqrshl_s8_xu10__SVBool_tu10__SVInt8_tS0_(
// CPP-CHECK-NEXT: entry:
-// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.sqrshl.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]])
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.sqrshl.u.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
//
svint8_t test_svqrshl_s8_x(svbool_t pg, svint8_t op1, svint8_t op2)
@@ -313,13 +313,13 @@ svint8_t test_svqrshl_s8_x(svbool_t pg, svint8_t op1, svint8_t op2)
// CHECK-LABEL: @test_svqrshl_s16_x(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.sqrshl.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]])
+// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.sqrshl.u.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]])
// CHECK-NEXT: ret <vscale x 8 x i16> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z18test_svqrshl_s16_xu10__SVBool_tu11__SVInt16_tS0_(
// CPP-CHECK-NEXT: entry:
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.sqrshl.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]])
+// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.sqrshl.u.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP1]]
//
svint16_t test_svqrshl_s16_x(svbool_t pg, svint16_t op1, svint16_t op2)
@@ -330,13 +330,13 @@ svint16_t test_svqrshl_s16_x(svbool_t pg, svint16_t op1, svint16_t op2)
// CHECK-LABEL: @test_svqrshl_s32_x(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.sqrshl.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]])
+// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.sqrshl.u.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]])
// CHECK-NEXT: ret <vscale x 4 x i32> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z18test_svqrshl_s32_xu10__SVBool_tu11__SVInt32_tS0_(
// CPP-CHECK-NEXT: entry:
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.sqrshl.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]])
+// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.sqrshl.u.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP1]]
//
svint32_t test_svqrshl_s32_x(svbool_t pg, svint32_t op1, svint32_t op2)
@@ -347,13 +347,13 @@ svint32_t test_svqrshl_s32_x(svbool_t pg, svint32_t op1, svint32_t op2)
// CHECK-LABEL: @test_svqrshl_s64_x(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.sqrshl.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
+// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.sqrshl.u.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
// CHECK-NEXT: ret <vscale x 2 x i64> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z18test_svqrshl_s64_xu10__SVBool_tu11__SVInt64_tS0_(
// CPP-CHECK-NEXT: entry:
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.sqrshl.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
+// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.sqrshl.u.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP1]]
//
svint64_t test_svqrshl_s64_x(svbool_t pg, svint64_t op1, svint64_t op2)
@@ -363,12 +363,12 @@ svint64_t test_svqrshl_s64_x(svbool_t pg, svint64_t op1, svint64_t op2)
// CHECK-LABEL: @test_svqrshl_u8_x(
// CHECK-NEXT: entry:
-// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.uqrshl.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]])
+// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.uqrshl.u.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]])
// CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
//
// CPP-CHECK-LABEL: @_Z17test_svqrshl_u8_xu10__SVBool_tu11__SVUint8_tu10__SVInt8_t(
// CPP-CHECK-NEXT: entry:
-// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.uqrshl.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]])
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.uqrshl.u.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[OP2:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
//
svuint8_t test_svqrshl_u8_x(svbool_t pg, svuint8_t op1, svint8_t op2)
@@ -379,13 +379,13 @@ svuint8_t test_svqrshl_u8_x(svbool_t pg, svuint8_t op1, svint8_t op2)
// CHECK-LABEL: @test_svqrshl_u16_x(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.uqrshl.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]])
+// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.uqrshl.u.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]])
// CHECK-NEXT: ret <vscale x 8 x i16> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z18test_svqrshl_u16_xu10__SVBool_tu12__SVUint16_tu11__SVInt16_t(
// CPP-CHECK-NEXT: entry:
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.uqrshl.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]])
+// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.uqrshl.u.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[OP2:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP1]]
//
svuint16_t test_svqrshl_u16_x(svbool_t pg, svuint16_t op1, svint16_t op2)
@@ -396,13 +396,13 @@ svuint16_t test_svqrshl_u16_x(svbool_t pg, svuint16_t op1, svint16_t op2)
// CHECK-LABEL: @test_svqrshl_u32_x(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.uqrshl.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]])
+// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.uqrshl.u.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]])
// CHECK-NEXT: ret <vscale x 4 x i32> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z18test_svqrshl_u32_xu10__SVBool_tu12__SVUint32_tu11__SVInt32_t(
// CPP-CHECK-NEXT: entry:
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.uqrshl.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]])
+// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.uqrshl.u.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[OP2:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP1]]
//
svuint32_t test_svqrshl_u32_x(svbool_t pg, svuint32_t op1, svint32_t op2)
@@ -413,13 +413,13 @@ svuint32_t test_svqrshl_u32_x(svbool_t pg, svuint32_t op1, svint32_t op2)
// CHECK-LABEL: @test_svqrshl_u64_x(
// CHECK-NEXT: entry:
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.uqrshl.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
+// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.uqrshl.u.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
// CHECK-NEXT: ret <vscale x 2 x i64> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z18test_svqrshl_u64_xu10__SVBool_tu12__SVUint64_tu11__SVInt64_t(
// CPP-CHECK-NEXT: entry:
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.uqrshl.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
+// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.uqrshl.u.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
// CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP1]]
//
svuint64_t test_svqrshl_u64_x(svbool_t pg, svuint64_t op1, svint64_t op2)
@@ -775,14 +775,14 @@ svuint64_t test_svqrshl_n_u64_m(svbool_t pg, svuint64_t op1, int64_t op2)
// CHECK-NEXT: entry:
// CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 16 x i8> poison, i8 [[OP2:%.*]], i64 0
// CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 16 x i8> [[DOTSPLATINSERT]], <vscale x 16 x i8> poison, <vscale x 16 x i32> zeroinitializer
-// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.sqrshl.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[DOTSPLAT]])
+// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.sqrshl.u.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[DOTSPLAT]])
// CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
//
// CPP-CHECK-LABEL: @_Z19test_svqrshl_n_s8_xu10__SVBool_tu10__SVInt8_ta(
// CPP-CHECK-NEXT: entry:
// CPP-CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 16 x i8> poison, i8 [[OP2:%.*]], i64 0
// CPP-CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 16 x i8> [[DOTSPLATINSERT]], <vscale x 16 x i8> poison, <vscale x 16 x i32> zeroinitializer
-// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.sqrshl.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[DOTSPLAT]])
+// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.aarch64.sve.sqrshl.u.nxv16i8(<vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> [[OP1:%.*]], <vscale x 16 x i8> [[DOTSPLAT]])
// CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]]
//
svint8_t test_svqrshl_n_s8_x(svbool_t pg, svint8_t op1, int8_t op2)
@@ -795,7 +795,7 @@ svint8_t test_svqrshl_n_s8_x(svbool_t pg, svint8_t op1, int8_t op2)
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x i16> poison, i16 [[OP2:%.*]], i64 0
// CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x i16> [[DOTSPLATINSERT]], <vscale x 8 x i16> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.sqrshl.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[DOTSPLAT]])
+// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.sqrshl.u.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[DOTSPLAT]])
// CHECK-NEXT: ret <vscale x 8 x i16> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z20test_svqrshl_n_s16_xu10__SVBool_tu11__SVInt16_ts(
@@ -803,7 +803,7 @@ svint8_t test_svqrshl_n_s8_x(svbool_t pg, svint8_t op1, int8_t op2)
// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
// CPP-CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x i16> poison, i16 [[OP2:%.*]], i64 0
// CPP-CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x i16> [[DOTSPLATINSERT]], <vscale x 8 x i16> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.sqrshl.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[DOTSPLAT]])
+// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.aarch64.sve.sqrshl.u.nxv8i16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> [[OP1:%.*]], <vscale x 8 x i16> [[DOTSPLAT]])
// CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP1]]
//
svint16_t test_svqrshl_n_s16_x(svbool_t pg, svint16_t op1, int16_t op2)
@@ -816,7 +816,7 @@ svint16_t test_svqrshl_n_s16_x(svbool_t pg, svint16_t op1, int16_t op2)
// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x i32> poison, i32 [[OP2:%.*]], i64 0
// CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x i32> [[DOTSPLATINSERT]], <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.sqrshl.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[DOTSPLAT]])
+// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.sqrshl.u.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[OP1:%.*]], <vscale x 4 x i32> [[DOTSPLAT]])
// CHECK-NEXT: ret <vscale x 4 x i32> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z20test_svqrshl_n_s32_xu10__SVBool_tu11__SVInt32_ti(
@@ -824,7 +824,7 @@ svint16_t test_svqrshl_n_s16_x(svbool_t pg, svint16_t op1, int16_t ...
[truncated]
|
c-rhodes
approved these changes
Dec 4, 2025
Collaborator
c-rhodes
left a comment
c-rhodes
left a comment
There was a problem hiding this comment.
couple of minor comments, otherwise LGTM cheers
Collaborator
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/116/builds/21846 Here is the relevant piece of the build log for the reference |
Collaborator
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/3/builds/25852 Here is the relevant piece of the build log for the reference |
Collaborator
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/129/builds/34436 Here is the relevant piece of the build log for the reference |
honeygoyal
pushed a commit
to honeygoyal/llvm-project
that referenced
this pull request
Dec 9, 2025
…t intrinsics. (llvm#170503) This follows the same path used for regular shift intrinsics whereby the "don't care about the result of inactive lanes" property of the associated _x builtins is preserved. In doing this we gain the ability to switch between reversed and movprfx variants depending on what best fits register allocation.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This follows the same path used for regular shift intrinsics whereby the "don't care about the result of inactive lanes" property of the associated _x builtins is preserved. In doing this we gain the ability to switch between reversed and movprfx variants depending on what best fits register allocation.