[clang] [llvm] [AArch64] Add intrinsics for 9.6 crypto instructions (PR #165545)

via cfe-commits cfe-commits at lists.llvm.org
Wed Oct 29 04:33:11 PDT 2025


https://github.com/Lukacma created https://github.com/llvm/llvm-project/pull/165545

This patch add intrinsics for crpyto instructions defined in https://github.com/ARM-software/acle/pull/411 ACLE proposal

>From 44338e5c9cb5edf233cc6bd5f80e76fbd27098a0 Mon Sep 17 00:00:00 2001
From: Marian Lukac <Marian.Lukac at arm.com>
Date: Mon, 27 Oct 2025 12:04:35 +0000
Subject: [PATCH 1/2] [AArch64] Add intrinsics support for 9.6 crypto

---
 clang/include/clang/Basic/arm_sve.td          |  17 ++
 .../sve2p1-intrinsics/acle_sve2p1_crypto.c    | 215 ++++++++++++++++++
 clang/utils/TableGen/SveEmitter.cpp           |   9 +-
 llvm/include/llvm/IR/IntrinsicsAArch64.td     |  27 +++
 .../Target/AArch64/AArch64ISelDAGToDAG.cpp    |  64 +++++-
 .../lib/Target/AArch64/AArch64SVEInstrInfo.td |   4 +-
 .../AArch64/sve2p1-intrinsics-crypto.ll       | 155 +++++++++++++
 7 files changed, 477 insertions(+), 14 deletions(-)
 create mode 100644 clang/test/CodeGen/AArch64/sve2p1-intrinsics/acle_sve2p1_crypto.c
 create mode 100644 llvm/test/CodeGen/AArch64/sve2p1-intrinsics-crypto.ll

diff --git a/clang/include/clang/Basic/arm_sve.td b/clang/include/clang/Basic/arm_sve.td
index d2b7b78b9970f..605cf93d196ff 100644
--- a/clang/include/clang/Basic/arm_sve.td
+++ b/clang/include/clang/Basic/arm_sve.td
@@ -1841,6 +1841,23 @@ def SVPMULLT_PAIR_U64   : SInst<"svpmullt_pair[_{d}]",   "ddd", "Ul", MergeNone,
 def SVPMULLT_PAIR_N_U64 : SInst<"svpmullt_pair[_n_{d}]", "dda", "Ul", MergeNone, "aarch64_sve_pmullt_pair", [VerifyRuntimeMode]>;
 }
 
+let SVETargetGuard = "sve-aes2", SMETargetGuard = "sve-aes2,ssve-aes" in {
+def SVAESD_X2   : SInst<"svaesd_laneq[_{d}_x2]",   "22di", "Uc", MergeNone, "aarch64_sve_aesd_laneq_x2", [IsOverloadNone, VerifyRuntimeMode]>;
+def SVAESIMC_X2 : SInst<"svaesdimc_laneq[_{d}_x2]", "22di",  "Uc", MergeNone, "aarch64_sve_aesdimc_laneq_x2", [IsOverloadNone, VerifyRuntimeMode]>;
+def SVAESE_X2   : SInst<"svaese_laneq[_{d}_x2]",   "22di", "Uc", MergeNone, "aarch64_sve_aese_laneq_x2", [IsOverloadNone, VerifyRuntimeMode]>;
+def SVAESeMC_X2  : SInst<"svaesemc_laneq[_{d}_x2]",  "22di",  "Uc", MergeNone, "aarch64_sve_aesemc_laneq_x2", [IsOverloadNone, VerifyRuntimeMode]>;
+
+def SVAESD_X4   : SInst<"svaesd_laneq[_{d}_x4]",   "44di", "Uc", MergeNone, "aarch64_sve_aesd_laneq_x4", [IsOverloadNone, VerifyRuntimeMode]>;
+def SVAESIMC_X4 : SInst<"svaesdimc_laneq[_{d}_x4]", "44di",  "Uc", MergeNone, "aarch64_sve_aesdimc_laneq_x4", [IsOverloadNone, VerifyRuntimeMode]>;
+def SVAESE_X4   : SInst<"svaese_laneq[_{d}_x4]",   "44di", "Uc", MergeNone, "aarch64_sve_aese_laneq_x4", [IsOverloadNone, VerifyRuntimeMode]>;
+def SVAESeMC_X4  : SInst<"svaesemc_laneq[_{d}_x4]",  "44di",  "Uc", MergeNone, "aarch64_sve_aesemc_laneq_x4", [IsOverloadNone, VerifyRuntimeMode]>;
+
+def SVPMULL_U64   : SInst<"svpmull[_u64]",   "2dd", "Ul", MergeNone, "aarch64_sve_pmull", [IsOverloadNone, VerifyRuntimeMode]>;
+def SVPMULL_N_U64 : SInst<"svpmull[_n_u64]", "2da", "Ul", MergeNone, "aarch64_sve_pmull", [IsOverloadNone, VerifyRuntimeMode]>;
+def SVPMLAL_U64   : SInst<"svpmlal[_u64]",   "22dd", "Ul", MergeNone, "aarch64_sve_pmlal", [IsOverloadNone, VerifyRuntimeMode]>;
+def SVPMLAL_N_U64 : SInst<"svpmlal[_n_u64]", "22da", "Ul", MergeNone, "aarch64_sve_pmlal", [IsOverloadNone, VerifyRuntimeMode]>;
+}
+
 let SVETargetGuard = "sve-sha3", SMETargetGuard = "sme2p1,sve-sha3" in {
 def SVRAX1 : SInst<"svrax1[_{d}]",   "ddd", "lUl", MergeNone, "aarch64_sve_rax1", [IsOverloadNone, VerifyRuntimeMode]>;
 }
diff --git a/clang/test/CodeGen/AArch64/sve2p1-intrinsics/acle_sve2p1_crypto.c b/clang/test/CodeGen/AArch64/sve2p1-intrinsics/acle_sve2p1_crypto.c
new file mode 100644
index 0000000000000..bbe71d9db92cf
--- /dev/null
+++ b/clang/test/CodeGen/AArch64/sve2p1-intrinsics/acle_sve2p1_crypto.c
@@ -0,0 +1,215 @@
+// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64 -target-feature +sve -target-feature +sve-aes2 -O1 -Werror -Wall -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64 -target-feature +sve -target-feature +sve-aes2 -O1 -Werror -Wall -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64 -target-feature +sve -target-feature +sve-aes2 -O1 -Werror -Wall -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64 -target-feature +sve -target-feature +sve-aes2 -O1 -Werror -Wall -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
+
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64 -target-feature +sme -target-feature +sve-aes2 -target-feature +ssve-aes -O1 -Werror -Wall -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64 -target-feature +sme -target-feature +sve-aes2 -target-feature +ssve-aes -O1 -Werror -Wall -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64 -target-feature +sme -target-feature +sve-aes2 -target-feature +ssve-aes -O1 -Werror -Wall -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -fclang-abi-compat=latest -DSVE_OVERLOADED_FORMS -triple aarch64 -target-feature +sme -target-feature +sve-aes2 -target-feature +ssve-aes -O1 -Werror -Wall -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -triple aarch64 -target-feature +sve -target-feature +sve-aes2 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
+// REQUIRES: aarch64-registered-target
+
+#include <arm_sve.h>
+
+#ifdef SVE_OVERLOADED_FORMS
+// A simple used,unused... macro, long enough to represent any SVE builtin.
+#define SVE_ACLE_FUNC(A1,A2_UNUSED,A3,A4_UNUSED) A1##A3
+#else
+#define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
+#endif
+
+#ifdef __ARM_FEATURE_SME
+#define STREAMING __arm_streaming
+#else
+#define STREAMING
+#endif
+
+// CHECK-LABEL: @test_svaesd_u8_x2(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aesd.laneq.x2(<vscale x 16 x i8> [[OP1_COERCE0:%.*]], <vscale x 16 x i8> [[OP1_COERCE1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], i32 0)
+// CHECK-NEXT:    ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z17test_svaesd_u8_x211svuint8x2_tu11__SVUint8_t(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aesd.laneq.x2(<vscale x 16 x i8> [[OP1_COERCE0:%.*]], <vscale x 16 x i8> [[OP1_COERCE1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], i32 0)
+// CPP-CHECK-NEXT:    ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svuint8x2_t test_svaesd_u8_x2(svuint8x2_t op1, svuint8_t op2) STREAMING
+{
+  return SVE_ACLE_FUNC(svaesd_laneq,_u8_x2,,)(op1, op2, 0);
+}
+
+// CHECK-LABEL: @test_svaesdimc_u8_x2(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aesdimc.laneq.x2(<vscale x 16 x i8> [[OP1_COERCE0:%.*]], <vscale x 16 x i8> [[OP1_COERCE1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], i32 0)
+// CHECK-NEXT:    ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z20test_svaesdimc_u8_x211svuint8x2_tu11__SVUint8_t(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aesdimc.laneq.x2(<vscale x 16 x i8> [[OP1_COERCE0:%.*]], <vscale x 16 x i8> [[OP1_COERCE1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], i32 0)
+// CPP-CHECK-NEXT:    ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svuint8x2_t test_svaesdimc_u8_x2(svuint8x2_t op1, svuint8_t op2) STREAMING
+{
+  return SVE_ACLE_FUNC(svaesdimc_laneq,_u8_x2,,)(op1, op2, 0);
+}
+
+// CHECK-LABEL: @test_svaese_u8_x2(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aese.laneq.x2(<vscale x 16 x i8> [[OP1_COERCE0:%.*]], <vscale x 16 x i8> [[OP1_COERCE1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], i32 0)
+// CHECK-NEXT:    ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z17test_svaese_u8_x211svuint8x2_tu11__SVUint8_t(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aese.laneq.x2(<vscale x 16 x i8> [[OP1_COERCE0:%.*]], <vscale x 16 x i8> [[OP1_COERCE1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], i32 0)
+// CPP-CHECK-NEXT:    ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svuint8x2_t test_svaese_u8_x2(svuint8x2_t op1, svuint8_t op2) STREAMING
+{
+  return SVE_ACLE_FUNC(svaese_laneq,_u8_x2,,)(op1, op2, 0);
+}
+
+// CHECK-LABEL: @test_svaesemc_u8_x2(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aesemc.laneq.x2(<vscale x 16 x i8> [[OP1_COERCE0:%.*]], <vscale x 16 x i8> [[OP1_COERCE1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], i32 0)
+// CHECK-NEXT:    ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z19test_svaesemc_u8_x211svuint8x2_tu11__SVUint8_t(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aesemc.laneq.x2(<vscale x 16 x i8> [[OP1_COERCE0:%.*]], <vscale x 16 x i8> [[OP1_COERCE1:%.*]], <vscale x 16 x i8> [[OP2:%.*]], i32 0)
+// CPP-CHECK-NEXT:    ret { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svuint8x2_t test_svaesemc_u8_x2(svuint8x2_t op1, svuint8_t op2) STREAMING
+{
+  return SVE_ACLE_FUNC(svaesemc_laneq,_u8_x2,,)(op1, op2, 0);
+}
+
+// CHECK-LABEL: @test_svaesd_u8_x4(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aesd.laneq.x4(<vscale x 16 x i8> [[OP1_COERCE0:%.*]], <vscale x 16 x i8> [[OP1_COERCE1:%.*]], <vscale x 16 x i8> [[OP1_COERCE2:%.*]], <vscale x 16 x i8> [[OP1_COERCE3:%.*]], <vscale x 16 x i8> [[OP2:%.*]], i32 0)
+// CHECK-NEXT:    ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z17test_svaesd_u8_x411svuint8x4_tu11__SVUint8_t(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aesd.laneq.x4(<vscale x 16 x i8> [[OP1_COERCE0:%.*]], <vscale x 16 x i8> [[OP1_COERCE1:%.*]], <vscale x 16 x i8> [[OP1_COERCE2:%.*]], <vscale x 16 x i8> [[OP1_COERCE3:%.*]], <vscale x 16 x i8> [[OP2:%.*]], i32 0)
+// CPP-CHECK-NEXT:    ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svuint8x4_t test_svaesd_u8_x4(svuint8x4_t op1, svuint8_t op2) STREAMING
+{
+  return SVE_ACLE_FUNC(svaesd_laneq,_u8_x4,,)(op1, op2, 0);
+}
+
+// CHECK-LABEL: @test_svaesdimc_u8_x4(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aesdimc.laneq.x4(<vscale x 16 x i8> [[OP1_COERCE0:%.*]], <vscale x 16 x i8> [[OP1_COERCE1:%.*]], <vscale x 16 x i8> [[OP1_COERCE2:%.*]], <vscale x 16 x i8> [[OP1_COERCE3:%.*]], <vscale x 16 x i8> [[OP2:%.*]], i32 0)
+// CHECK-NEXT:    ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z20test_svaesdimc_u8_x411svuint8x4_tu11__SVUint8_t(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aesdimc.laneq.x4(<vscale x 16 x i8> [[OP1_COERCE0:%.*]], <vscale x 16 x i8> [[OP1_COERCE1:%.*]], <vscale x 16 x i8> [[OP1_COERCE2:%.*]], <vscale x 16 x i8> [[OP1_COERCE3:%.*]], <vscale x 16 x i8> [[OP2:%.*]], i32 0)
+// CPP-CHECK-NEXT:    ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svuint8x4_t test_svaesdimc_u8_x4(svuint8x4_t op1, svuint8_t op2) STREAMING
+{
+  return SVE_ACLE_FUNC(svaesdimc_laneq,_u8_x4,,)(op1, op2, 0);
+}
+
+// CHECK-LABEL: @test_svaese_u8_x4(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aese.laneq.x4(<vscale x 16 x i8> [[OP1_COERCE0:%.*]], <vscale x 16 x i8> [[OP1_COERCE1:%.*]], <vscale x 16 x i8> [[OP1_COERCE2:%.*]], <vscale x 16 x i8> [[OP1_COERCE3:%.*]], <vscale x 16 x i8> [[OP2:%.*]], i32 0)
+// CHECK-NEXT:    ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z17test_svaese_u8_x411svuint8x4_tu11__SVUint8_t(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aese.laneq.x4(<vscale x 16 x i8> [[OP1_COERCE0:%.*]], <vscale x 16 x i8> [[OP1_COERCE1:%.*]], <vscale x 16 x i8> [[OP1_COERCE2:%.*]], <vscale x 16 x i8> [[OP1_COERCE3:%.*]], <vscale x 16 x i8> [[OP2:%.*]], i32 0)
+// CPP-CHECK-NEXT:    ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svuint8x4_t test_svaese_u8_x4(svuint8x4_t op1, svuint8_t op2) STREAMING
+{
+  return SVE_ACLE_FUNC(svaese_laneq,_u8_x4,,)(op1, op2, 0);
+}
+
+// CHECK-LABEL: @test_svaesemc_u8_x4(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aesemc.laneq.x4(<vscale x 16 x i8> [[OP1_COERCE0:%.*]], <vscale x 16 x i8> [[OP1_COERCE1:%.*]], <vscale x 16 x i8> [[OP1_COERCE2:%.*]], <vscale x 16 x i8> [[OP1_COERCE3:%.*]], <vscale x 16 x i8> [[OP2:%.*]], i32 0)
+// CHECK-NEXT:    ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z19test_svaesemc_u8_x411svuint8x4_tu11__SVUint8_t(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aesemc.laneq.x4(<vscale x 16 x i8> [[OP1_COERCE0:%.*]], <vscale x 16 x i8> [[OP1_COERCE1:%.*]], <vscale x 16 x i8> [[OP1_COERCE2:%.*]], <vscale x 16 x i8> [[OP1_COERCE3:%.*]], <vscale x 16 x i8> [[OP2:%.*]], i32 0)
+// CPP-CHECK-NEXT:    ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP0]]
+//
+svuint8x4_t test_svaesemc_u8_x4(svuint8x4_t op1, svuint8_t op2) STREAMING
+{
+  return SVE_ACLE_FUNC(svaesemc_laneq,_u8_x4,,)(op1, op2, 0);
+}
+
+// CHECK-LABEL: @test_svpmull_u64(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.pmull(<vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
+// CHECK-NEXT:    ret { <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z16test_svpmull_u64u12__SVUint64_tS_(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.pmull(<vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    ret { <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP0]]
+//
+svuint64x2_t test_svpmull_u64(svuint64_t op1, svuint64_t op2) STREAMING
+{
+  return SVE_ACLE_FUNC(svpmull,_u64,,)(op1, op2);
+}
+
+// CHECK-LABEL: @test_svpmull_n_u64(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x i64> poison, i64 [[OP2:%.*]], i64 0
+// CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x i64> [[DOTSPLATINSERT]], <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.pmull(<vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[DOTSPLAT]])
+// CHECK-NEXT:    ret { <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z18test_svpmull_n_u64u12__SVUint64_tm(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x i64> poison, i64 [[OP2:%.*]], i64 0
+// CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x i64> [[DOTSPLATINSERT]], <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.pmull(<vscale x 2 x i64> [[OP1:%.*]], <vscale x 2 x i64> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    ret { <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP0]]
+//
+svuint64x2_t test_svpmull_n_u64(svuint64_t op1, uint64_t op2) STREAMING
+{
+  return SVE_ACLE_FUNC(svpmull,_n_u64,,)(op1, op2);
+}
+
+// CHECK-LABEL: @test_svpmlal_u64(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.pmlal(<vscale x 2 x i64> [[OP1_COERCE0:%.*]], <vscale x 2 x i64> [[OP1_COERCE1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i64> [[OP3:%.*]])
+// CHECK-NEXT:    ret { <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z16test_svpmlal_u6412svuint64x2_tu12__SVUint64_tS0_(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.pmlal(<vscale x 2 x i64> [[OP1_COERCE0:%.*]], <vscale x 2 x i64> [[OP1_COERCE1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i64> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    ret { <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP0]]
+//
+svuint64x2_t test_svpmlal_u64(svuint64x2_t op1, svuint64_t op2, svuint64_t op3) STREAMING
+{
+  return SVE_ACLE_FUNC(svpmlal,_u64,,)(op1, op2, op3);
+}
+
+// CHECK-LABEL: @test_svpmlal_n_u64(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x i64> poison, i64 [[OP3:%.*]], i64 0
+// CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x i64> [[DOTSPLATINSERT]], <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.pmlal(<vscale x 2 x i64> [[OP1_COERCE0:%.*]], <vscale x 2 x i64> [[OP1_COERCE1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i64> [[DOTSPLAT]])
+// CHECK-NEXT:    ret { <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z18test_svpmlal_n_u6412svuint64x2_tu12__SVUint64_tm(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x i64> poison, i64 [[OP3:%.*]], i64 0
+// CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x i64> [[DOTSPLATINSERT]], <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.pmlal(<vscale x 2 x i64> [[OP1_COERCE0:%.*]], <vscale x 2 x i64> [[OP1_COERCE1:%.*]], <vscale x 2 x i64> [[OP2:%.*]], <vscale x 2 x i64> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    ret { <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP0]]
+//
+svuint64x2_t test_svpmlal_n_u64(svuint64x2_t op1, svuint64_t op2, uint64_t op3) STREAMING
+{
+  return SVE_ACLE_FUNC(svpmlal,_n_u64,,)(op1, op2, op3);
+}
diff --git a/clang/utils/TableGen/SveEmitter.cpp b/clang/utils/TableGen/SveEmitter.cpp
index b1e94e0ce0975..bbeed4402da1b 100644
--- a/clang/utils/TableGen/SveEmitter.cpp
+++ b/clang/utils/TableGen/SveEmitter.cpp
@@ -272,9 +272,16 @@ class Intrinsic {
           Proto[I] == 'R' || Proto[I] == '@' || Proto[I] == '!')
         break;
 
+      if (Proto[I] == '2')
+        Param += 1;
+      if (Proto[I] == '4')
+        Param += 3;
+
       // Multivector modifier can be skipped
-      if (Proto[I] == '.')
+      if (Proto[I] == '.') {
+        Param -= 1; // Adjust for the increment at the top of the loop
         I += 2;
+      }
     }
     assert(I != Proto.size() && "Prototype has no splat operand");
     return Param;
diff --git a/llvm/include/llvm/IR/IntrinsicsAArch64.td b/llvm/include/llvm/IR/IntrinsicsAArch64.td
index b81edc385cd43..71949c93f83b0 100644
--- a/llvm/include/llvm/IR/IntrinsicsAArch64.td
+++ b/llvm/include/llvm/IR/IntrinsicsAArch64.td
@@ -4194,4 +4194,31 @@ let TargetPrefix = "aarch64" in {
   def int_aarch64_sme_fp8_fvdot_lane_za16_vg1x2  : SME_FP8_ZA_LANE_VGx2_Intrinsic;
   def int_aarch64_sme_fp8_fvdotb_lane_za32_vg1x4 : SME_FP8_ZA_LANE_VGx2_Intrinsic;
   def int_aarch64_sme_fp8_fvdott_lane_za32_vg1x4 : SME_FP8_ZA_LANE_VGx2_Intrinsic;
+  
+  // AES2
+  class SVE2_Crypto_LANE_X2_Intrinsic
+  : DefaultAttrsIntrinsic<[llvm_nxv16i8_ty, llvm_nxv16i8_ty],
+      [llvm_nxv16i8_ty, llvm_nxv16i8_ty, llvm_nxv16i8_ty, llvm_i32_ty],
+      [ImmArg<ArgIndex<3>>, IntrNoMem]>;
+  class SVE2_Crypto_LANE_X4_Intrinsic
+  : DefaultAttrsIntrinsic<[llvm_nxv16i8_ty, llvm_nxv16i8_ty, llvm_nxv16i8_ty, llvm_nxv16i8_ty],
+      [llvm_nxv16i8_ty, llvm_nxv16i8_ty, llvm_nxv16i8_ty, 
+       llvm_nxv16i8_ty, llvm_nxv16i8_ty, llvm_i32_ty],
+      [ImmArg<ArgIndex<5>>, IntrNoMem]>;
+
+  def int_aarch64_sve_aesd_laneq_x2 : SVE2_Crypto_LANE_X2_Intrinsic;
+  def int_aarch64_sve_aesdimc_laneq_x2 : SVE2_Crypto_LANE_X2_Intrinsic;
+  def int_aarch64_sve_aese_laneq_x2 : SVE2_Crypto_LANE_X2_Intrinsic;
+  def int_aarch64_sve_aesemc_laneq_x2 : SVE2_Crypto_LANE_X2_Intrinsic;
+
+  def int_aarch64_sve_aesd_laneq_x4 : SVE2_Crypto_LANE_X4_Intrinsic;
+  def int_aarch64_sve_aesdimc_laneq_x4 : SVE2_Crypto_LANE_X4_Intrinsic;
+  def int_aarch64_sve_aese_laneq_x4 : SVE2_Crypto_LANE_X4_Intrinsic;
+  def int_aarch64_sve_aesemc_laneq_x4 : SVE2_Crypto_LANE_X4_Intrinsic;
+
+  def int_aarch64_sve_pmull : DefaultAttrsIntrinsic<[llvm_nxv2i64_ty, llvm_nxv2i64_ty],
+      [llvm_nxv2i64_ty, llvm_nxv2i64_ty], [IntrNoMem]>;
+  def int_aarch64_sve_pmlal : DefaultAttrsIntrinsic<[llvm_nxv2i64_ty, llvm_nxv2i64_ty],
+      [llvm_nxv2i64_ty, llvm_nxv2i64_ty, llvm_nxv2i64_ty, llvm_nxv2i64_ty], [IntrNoMem]>;
 }
+
diff --git a/llvm/lib/Target/AArch64/AArch64ISelDAGToDAG.cpp b/llvm/lib/Target/AArch64/AArch64ISelDAGToDAG.cpp
index e7b2d20e2a6cb..2655aadb26c19 100644
--- a/llvm/lib/Target/AArch64/AArch64ISelDAGToDAG.cpp
+++ b/llvm/lib/Target/AArch64/AArch64ISelDAGToDAG.cpp
@@ -1974,26 +1974,28 @@ void AArch64DAGToDAGISel::SelectDestructiveMultiIntrinsic(SDNode *N,
   SDLoc DL(N);
   EVT VT = N->getValueType(0);
   unsigned FirstVecIdx = HasPred ? 2 : 1;
+  SDUse *OpsIter = N->op_begin() + 1; // Skip intrinsic ID
+  SmallVector<SDValue, 4> Ops;
 
-  auto GetMultiVecOperand = [=](unsigned StartIdx) {
-    SmallVector<SDValue, 4> Regs(N->ops().slice(StartIdx, NumVecs));
+  auto GetMultiVecOperand = [&]() {
+    SmallVector<SDValue, 4> Regs(OpsIter, OpsIter + NumVecs);
+    OpsIter += NumVecs;
     return createZMulTuple(Regs);
   };
 
-  SDValue Zdn = GetMultiVecOperand(FirstVecIdx);
+  if (HasPred)
+    Ops.push_back(*OpsIter++);
 
-  SDValue Zm;
+  Ops.push_back(GetMultiVecOperand());
   if (IsZmMulti)
-    Zm = GetMultiVecOperand(NumVecs + FirstVecIdx);
+    Ops.push_back(GetMultiVecOperand());
   else
-    Zm = N->getOperand(NumVecs + FirstVecIdx);
+    Ops.push_back(*OpsIter++);
 
+  // Append any remaining operands.
+  Ops.append(OpsIter, N->op_end());
   SDNode *Intrinsic;
-  if (HasPred)
-    Intrinsic = CurDAG->getMachineNode(Opcode, DL, MVT::Untyped,
-                                       N->getOperand(1), Zdn, Zm);
-  else
-    Intrinsic = CurDAG->getMachineNode(Opcode, DL, MVT::Untyped, Zdn, Zm);
+  Intrinsic = CurDAG->getMachineNode(Opcode, DL, MVT::Untyped, Ops);
   SDValue SuperReg = SDValue(Intrinsic, 0);
   for (unsigned i = 0; i < NumVecs; ++i)
     ReplaceUses(SDValue(N, i), CurDAG->getTargetExtractSubreg(
@@ -6248,6 +6250,46 @@ void AArch64DAGToDAGISel::Select(SDNode *Node) {
                AArch64::FMINNM_VG4_4Z4Z_S, AArch64::FMINNM_VG4_4Z4Z_D}))
         SelectDestructiveMultiIntrinsic(Node, 4, true, Op);
       return;
+    case Intrinsic::aarch64_sve_aese_laneq_x2:
+      SelectDestructiveMultiIntrinsic(Node, 2, false, AArch64::AESE_2ZZI_B);
+      return;
+    case Intrinsic::aarch64_sve_aesd_laneq_x2:
+      SelectDestructiveMultiIntrinsic(Node, 2, false, AArch64::AESD_2ZZI_B);
+      return;
+    case Intrinsic::aarch64_sve_aesemc_laneq_x2:
+      SelectDestructiveMultiIntrinsic(Node, 2, false, AArch64::AESEMC_2ZZI_B);
+      return;
+    case Intrinsic::aarch64_sve_aesdimc_laneq_x2:
+      SelectDestructiveMultiIntrinsic(Node, 2, false, AArch64::AESDIMC_2ZZI_B);
+      return;
+    case Intrinsic::aarch64_sve_aese_laneq_x4:
+      SelectDestructiveMultiIntrinsic(Node, 4, false, AArch64::AESE_4ZZI_B);
+      return;
+    case Intrinsic::aarch64_sve_aesd_laneq_x4:
+      SelectDestructiveMultiIntrinsic(Node, 4, false, AArch64::AESD_4ZZI_B);
+      return;
+    case Intrinsic::aarch64_sve_aesemc_laneq_x4:
+      SelectDestructiveMultiIntrinsic(Node, 4, false, AArch64::AESEMC_4ZZI_B);
+      return;
+    case Intrinsic::aarch64_sve_aesdimc_laneq_x4:
+      SelectDestructiveMultiIntrinsic(Node, 4, false, AArch64::AESDIMC_4ZZI_B);
+      return;
+    case Intrinsic::aarch64_sve_pmlal:
+      SelectDestructiveMultiIntrinsic(Node, 2, false, AArch64::PMLAL_2ZZZ_Q);
+      return;
+    case Intrinsic::aarch64_sve_pmull: {
+      SDLoc DL(Node);
+      SmallVector<SDValue, 4> Regs(Node->ops().slice(1, 2));
+      SDNode *Res =
+          CurDAG->getMachineNode(AArch64::PMULL_2ZZZ_Q, DL, MVT::Untyped, Regs);
+      SDValue SuperReg = SDValue(Res, 0);
+      for (unsigned I = 0; I < 2; I++)
+        ReplaceUses(SDValue(Node, I),
+                    CurDAG->getTargetExtractSubreg(AArch64::zsub0 + I, DL, VT,
+                                                   SuperReg));
+      CurDAG->RemoveDeadNode(Node);
+      return;
+    }
     case Intrinsic::aarch64_sve_fcvtzs_x2:
       SelectCVTIntrinsic(Node, 2, AArch64::FCVTZS_2Z2Z_StoS);
       return;
diff --git a/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td b/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td
index 3b268dcbca600..7098ac57073cd 100644
--- a/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td
+++ b/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td
@@ -4219,12 +4219,12 @@ let Predicates = [HasSVEAES2, HasNonStreamingSVE_or_SSVE_AES] in {
   def AESE_2ZZI_B    : sve_crypto_binary_multi2<0b000, "aese">;
   def AESD_2ZZI_B    : sve_crypto_binary_multi2<0b010, "aesd">;
   def AESEMC_2ZZI_B  : sve_crypto_binary_multi2<0b100, "aesemc">;
-  def AESDMIC_2ZZI_B : sve_crypto_binary_multi2<0b110, "aesdimc">;
+  def AESDIMC_2ZZI_B : sve_crypto_binary_multi2<0b110, "aesdimc">;
   // SVE_AES2 multi-vector instructions (x4)
   def AESE_4ZZI_B    : sve_crypto_binary_multi4<0b0000, "aese">;
   def AESD_4ZZI_B    : sve_crypto_binary_multi4<0b0100, "aesd">;
   def AESEMC_4ZZI_B  : sve_crypto_binary_multi4<0b1000, "aesemc">;
-  def AESDMIC_4ZZI_B : sve_crypto_binary_multi4<0b1100, "aesdimc">;
+  def AESDIMC_4ZZI_B : sve_crypto_binary_multi4<0b1100, "aesdimc">;
 
   // SVE_AES2 multi-vector polynomial multiply
   def PMLAL_2ZZZ_Q : sve_crypto_pmlal_multi<"pmlal">;
diff --git a/llvm/test/CodeGen/AArch64/sve2p1-intrinsics-crypto.ll b/llvm/test/CodeGen/AArch64/sve2p1-intrinsics-crypto.ll
new file mode 100644
index 0000000000000..5cbe27628e25d
--- /dev/null
+++ b/llvm/test/CodeGen/AArch64/sve2p1-intrinsics-crypto.ll
@@ -0,0 +1,155 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc -mtriple=aarch64-linux-gnu -mattr=+sve,+sve-aes2 < %s | FileCheck %s
+; RUN: llc -mtriple=aarch64-linux-gnu -mattr=+sve,+sve-aes2,+ssve-aes -force-streaming < %s | FileCheck %s
+
+;
+; AESE
+;
+
+define { <vscale x 16 x i8>, <vscale x 16 x i8> } @aese_x2(<vscale x 16 x i8> %a, <vscale x 16 x i8> %b, <vscale x 16 x i8> %c) {
+; CHECK-LABEL: aese_x2:
+; CHECK:       // %bb.0:
+; CHECK-NEXT:    // kill: def $z1 killed $z1 killed $z0_z1 def $z0_z1
+; CHECK-NEXT:    // kill: def $z0 killed $z0 killed $z0_z1 def $z0_z1
+; CHECK-NEXT:    aese { z0.b, z1.b }, { z0.b, z1.b }, z2.q[0]
+; CHECK-NEXT:    ret
+  %out = call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aese.laneq.x2(<vscale x 16 x i8> %a, <vscale x 16 x i8> %b, <vscale x 16 x i8> %c, i32 0)
+  ret { <vscale x 16 x i8>, <vscale x 16 x i8> } %out
+}
+
+define { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @aese_x4(<vscale x 16 x i8> %a, <vscale x 16 x i8> %b, <vscale x 16 x i8> %c, <vscale x 16 x i8> %d) {
+; CHECK-LABEL: aese_x4:
+; CHECK:       // %bb.0:
+; CHECK-NEXT:    // kill: def $z3 killed $z3 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
+; CHECK-NEXT:    // kill: def $z2 killed $z2 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
+; CHECK-NEXT:    // kill: def $z1 killed $z1 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
+; CHECK-NEXT:    // kill: def $z0 killed $z0 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
+; CHECK-NEXT:    aese { z0.b - z3.b }, { z0.b - z3.b }, z2.q[0]
+; CHECK-NEXT:    ret
+  %out= call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> }  @llvm.aarch64.sve.aese.laneq.x4(<vscale x 16 x i8> %a,
+                                                                                                    <vscale x 16 x i8> %b, <vscale x 16 x i8> %c,
+                                                                                                    <vscale x 16 x i8> %d, <vscale x 16 x i8> %c , i32 0)
+  ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } %out
+}
+
+;
+; AESD
+;
+
+define { <vscale x 16 x i8>, <vscale x 16 x i8> } @aesd_x2(<vscale x 16 x i8> %a, <vscale x 16 x i8> %b, <vscale x 16 x i8> %c) {
+; CHECK-LABEL: aesd_x2:
+; CHECK:       // %bb.0:
+; CHECK-NEXT:    // kill: def $z1 killed $z1 killed $z0_z1 def $z0_z1
+; CHECK-NEXT:    // kill: def $z0 killed $z0 killed $z0_z1 def $z0_z1
+; CHECK-NEXT:    aesd { z0.b, z1.b }, { z0.b, z1.b }, z2.q[0]
+; CHECK-NEXT:    ret
+  %out= call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aesd.laneq.x2(<vscale x 16 x i8> %a, <vscale x 16 x i8> %b, <vscale x 16 x i8> %c, i32 0)
+  ret { <vscale x 16 x i8>, <vscale x 16 x i8> } %out
+}
+
+define { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @aesd_x4(<vscale x 16 x i8> %a, <vscale x 16 x i8> %b, <vscale x 16 x i8> %c, <vscale x 16 x i8> %d) {
+; CHECK-LABEL: aesd_x4:
+; CHECK:       // %bb.0:
+; CHECK-NEXT:    // kill: def $z3 killed $z3 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
+; CHECK-NEXT:    // kill: def $z2 killed $z2 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
+; CHECK-NEXT:    // kill: def $z1 killed $z1 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
+; CHECK-NEXT:    // kill: def $z0 killed $z0 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
+; CHECK-NEXT:    aesd { z0.b - z3.b }, { z0.b - z3.b }, z2.q[0]
+; CHECK-NEXT:    ret
+  %out= call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> }  @llvm.aarch64.sve.aesd.laneq.x4(<vscale x 16 x i8> %a,
+                                                                                                    <vscale x 16 x i8> %b, <vscale x 16 x i8> %c,
+                                                                                                    <vscale x 16 x i8> %d, <vscale x 16 x i8> %c , i32 0)
+  ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } %out
+}
+
+
+;
+; AESEMC
+;
+
+define { <vscale x 16 x i8>, <vscale x 16 x i8> } @aesemc_x2(<vscale x 16 x i8> %a, <vscale x 16 x i8> %b, <vscale x 16 x i8> %c) {
+; CHECK-LABEL: aesemc_x2:
+; CHECK:       // %bb.0:
+; CHECK-NEXT:    // kill: def $z1 killed $z1 killed $z0_z1 def $z0_z1
+; CHECK-NEXT:    // kill: def $z0 killed $z0 killed $z0_z1 def $z0_z1
+; CHECK-NEXT:    aesemc { z0.b, z1.b }, { z0.b, z1.b }, z2.q[0]
+; CHECK-NEXT:    ret
+  %out= call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aesemc.laneq.x2(<vscale x 16 x i8> %a, <vscale x 16 x i8> %b, <vscale x 16 x i8> %c, i32 0)
+  ret { <vscale x 16 x i8>, <vscale x 16 x i8> } %out
+}
+
+define { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @aesemc_x4(<vscale x 16 x i8> %a, <vscale x 16 x i8> %b, <vscale x 16 x i8> %c, <vscale x 16 x i8> %d) {
+; CHECK-LABEL: aesemc_x4:
+; CHECK:       // %bb.0:
+; CHECK-NEXT:    // kill: def $z3 killed $z3 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
+; CHECK-NEXT:    // kill: def $z2 killed $z2 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
+; CHECK-NEXT:    // kill: def $z1 killed $z1 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
+; CHECK-NEXT:    // kill: def $z0 killed $z0 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
+; CHECK-NEXT:    aesemc { z0.b - z3.b }, { z0.b - z3.b }, z2.q[0]
+; CHECK-NEXT:    ret
+  %out= call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> }  @llvm.aarch64.sve.aesemc.laneq.x4(<vscale x 16 x i8> %a,
+                                                                                                    <vscale x 16 x i8> %b, <vscale x 16 x i8> %c,
+                                                                                                    <vscale x 16 x i8> %d, <vscale x 16 x i8> %c , i32 0)
+  ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } %out
+}
+
+;
+; AESDIMC
+;
+
+define { <vscale x 16 x i8>, <vscale x 16 x i8> } @aesdimc_x2(<vscale x 16 x i8> %a, <vscale x 16 x i8> %b, <vscale x 16 x i8> %c) {
+; CHECK-LABEL: aesdimc_x2:
+; CHECK:       // %bb.0:
+; CHECK-NEXT:    // kill: def $z1 killed $z1 killed $z0_z1 def $z0_z1
+; CHECK-NEXT:    // kill: def $z0 killed $z0 killed $z0_z1 def $z0_z1
+; CHECK-NEXT:    aesdimc { z0.b, z1.b }, { z0.b, z1.b }, z2.q[0]
+; CHECK-NEXT:    ret
+  %out= call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.aesdimc.laneq.x2(<vscale x 16 x i8> %a, <vscale x 16 x i8> %b, <vscale x 16 x i8> %c, i32 0)
+  ret { <vscale x 16 x i8>, <vscale x 16 x i8> } %out
+}
+
+define { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @aesdimc_x4(<vscale x 16 x i8> %a, <vscale x 16 x i8> %b, <vscale x 16 x i8> %c, <vscale x 16 x i8> %d) {
+; CHECK-LABEL: aesdimc_x4:
+; CHECK:       // %bb.0:
+; CHECK-NEXT:    // kill: def $z3 killed $z3 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
+; CHECK-NEXT:    // kill: def $z2 killed $z2 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
+; CHECK-NEXT:    // kill: def $z1 killed $z1 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
+; CHECK-NEXT:    // kill: def $z0 killed $z0 killed $z0_z1_z2_z3 def $z0_z1_z2_z3
+; CHECK-NEXT:    aesdimc { z0.b - z3.b }, { z0.b - z3.b }, z2.q[0]
+; CHECK-NEXT:    ret
+  %out= call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> }  @llvm.aarch64.sve.aesdimc.laneq.x4(<vscale x 16 x i8> %a,
+                                                                                                    <vscale x 16 x i8> %b, <vscale x 16 x i8> %c,
+                                                                                                    <vscale x 16 x i8> %d, <vscale x 16 x i8> %c , i32 0)
+  ret { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } %out
+}
+
+;
+; PMULL
+;
+
+define { <vscale x 2 x i64>, <vscale x 2 x i64> } @pmull_i64(<vscale x 2 x i64> %a, <vscale x 2 x i64> %b) {
+; CHECK-LABEL: pmull_i64:
+; CHECK:       // %bb.0:
+; CHECK-NEXT:    pmull { z0.q, z1.q }, z0.d, z1.d
+; CHECK-NEXT:    ret
+  %out = call { <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.pmull(<vscale x 2 x i64> %a,
+                                                                       <vscale x 2 x i64> %b)
+  ret { <vscale x 2 x i64>, <vscale x 2 x i64> } %out
+}
+
+
+;
+; PMLAL
+;
+
+define { <vscale x 2 x i64>, <vscale x 2 x i64> } @pmlal_i64(<vscale x 2 x i64> %a, <vscale x 2 x i64> %b, <vscale x 2 x i64> %c, <vscale x 2 x i64> %d) {
+; CHECK-LABEL: pmlal_i64:
+; CHECK:       // %bb.0:
+; CHECK-NEXT:    // kill: def $z1 killed $z1 killed $z0_z1 def $z0_z1
+; CHECK-NEXT:    // kill: def $z0 killed $z0 killed $z0_z1 def $z0_z1
+; CHECK-NEXT:    pmlal { z0.q, z1.q }, z2.d, z3.d
+; CHECK-NEXT:    ret
+  %out = call { <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.pmlal(<vscale x 2 x i64> %a, <vscale x 2 x i64> %b, <vscale x 2 x i64> %c,
+                                                                       <vscale x 2 x i64> %d)
+  ret { <vscale x 2 x i64>, <vscale x 2 x i64> } %out
+}

>From d86b9ebd30bfcb1a6f1dd8d644ea7db46eedcf7d Mon Sep 17 00:00:00 2001
From: Marian Lukac <Marian.Lukac at arm.com>
Date: Tue, 28 Oct 2025 14:50:50 +0000
Subject: [PATCH 2/2] Remove unnecessary code

---
 llvm/lib/Target/AArch64/AArch64ISelDAGToDAG.cpp | 1 -
 1 file changed, 1 deletion(-)

diff --git a/llvm/lib/Target/AArch64/AArch64ISelDAGToDAG.cpp b/llvm/lib/Target/AArch64/AArch64ISelDAGToDAG.cpp
index 2655aadb26c19..217fb2adffaac 100644
--- a/llvm/lib/Target/AArch64/AArch64ISelDAGToDAG.cpp
+++ b/llvm/lib/Target/AArch64/AArch64ISelDAGToDAG.cpp
@@ -1973,7 +1973,6 @@ void AArch64DAGToDAGISel::SelectDestructiveMultiIntrinsic(SDNode *N,
 
   SDLoc DL(N);
   EVT VT = N->getValueType(0);
-  unsigned FirstVecIdx = HasPred ? 2 : 1;
   SDUse *OpsIter = N->op_begin() + 1; // Skip intrinsic ID
   SmallVector<SDValue, 4> Ops;
 



More information about the cfe-commits mailing list