[ARM] Add macro fusion for AES instructions.

Summary:
This patch adds a macro fusion using CodeGen/MacroFusion.cpp to pair AES
instructions back to back and adds FeatureFuseAES to enable the feature.

Reviewers: evandro, javed.absar, rengolin, t.p.northover

Reviewed By: javed.absar

Subscribers: aemerson, mgorny, kristof.beyls, llvm-commits

Differential Revision: https://reviews.llvm.org/D34142

llvm-svn: 305988

llvm/lib/Target/ARM/ARM.td

@@ -100,7 +100,8 @@ def FeatureRAS : SubtargetFeature<"ras", "HasRAS", "true",
                 "Enable Reliability, Availability and Serviceability extensions">;
 def FeatureFPAO : SubtargetFeature<"fpao", "HasFPAO", "true",
                 "Enable fast computation of positive address offsets">;
-
+def FeatureFuseAES : SubtargetFeature<"fuse-aes", "HasFuseAES", "true",
+                                      "CPU fuses AES crypto operations">;
 
 // Cyclone has preferred instructions for zeroing VFP registers, which can
 // execute in 0 cycles.

llvm/lib/Target/ARM/ARMMacroFusion.cpp

@@ -0,0 +1,57 @@
+//===- ARMMacroFusion.cpp - ARM Macro Fusion ----------------------===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+//
+/// \file This file contains the ARM implementation of the DAG scheduling
+///  mutation to pair instructions back to back.
+//
+//===----------------------------------------------------------------------===//
+
+#include "ARMMacroFusion.h"
+#include "ARMSubtarget.h"
+#include "llvm/CodeGen/MacroFusion.h"
+#include "llvm/Target/TargetInstrInfo.h"
+
+namespace llvm {
+
+/// \brief Check if the instr pair, FirstMI and SecondMI, should be fused
+/// together. Given SecondMI, when FirstMI is unspecified, then check if
+/// SecondMI may be part of a fused pair at all.
+static bool shouldScheduleAdjacent(const TargetInstrInfo &TII,
+                                   const TargetSubtargetInfo &TSI,
+                                   const MachineInstr *FirstMI,
+                                   const MachineInstr &SecondMI) {
+  const ARMSubtarget &ST = static_cast<const ARMSubtarget&>(TSI);
+
+  // Assume wildcards for unspecified instrs.
+  unsigned FirstOpcode =
+    FirstMI ? FirstMI->getOpcode()
+	    : static_cast<unsigned>(ARM::INSTRUCTION_LIST_END);
+  unsigned SecondOpcode = SecondMI.getOpcode();
+
+  if (ST.hasFuseAES())
+    // Fuse AES crypto operations.
+    switch(SecondOpcode) {
+    // AES encode.
+    case ARM::AESMC :
+      return FirstOpcode == ARM::AESE ||
+             FirstOpcode == ARM::INSTRUCTION_LIST_END;
+    // AES decode.
+    case ARM::AESIMC:
+      return FirstOpcode == ARM::AESD ||
+             FirstOpcode == ARM::INSTRUCTION_LIST_END;
+    }
+
+  return false;
+}
+
+std::unique_ptr<ScheduleDAGMutation> createARMMacroFusionDAGMutation () {
+  return createMacroFusionDAGMutation(shouldScheduleAdjacent);
+}
+
+} // end namespace llvm

llvm/lib/Target/ARM/ARMMacroFusion.h

@@ -0,0 +1,24 @@
+//===- ARMMacroFusion.h - ARM Macro Fusion ------------------------===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+//
+/// \file This file contains the ARM definition of the DAG scheduling mutation
+/// to pair instructions back to back.
+//
+//===----------------------------------------------------------------------===//
+
+#include "llvm/CodeGen/MachineScheduler.h"
+
+namespace llvm {
+
+/// Note that you have to add:
+///   DAG.addMutation(createARMMacroFusionDAGMutation());
+/// to ARMPassConfig::createMachineScheduler() to have an effect.
+std::unique_ptr<ScheduleDAGMutation> createARMMacroFusionDAGMutation();
+
+} // llvm

llvm/lib/Target/ARM/ARMSubtarget.h

@@ -285,6 +285,10 @@ class ARMSubtarget : public ARMGenSubtargetInfo {
   /// HasFPAO - if true, processor  does positive address offset computation faster
   bool HasFPAO = false;
 
+  /// HasFuseAES - if true, processor executes back to back AES instruction
+  /// pairs faster.
+  bool HasFuseAES = false;
+
   /// If true, if conversion may decide to leave some instructions unpredicated.
   bool IsProfitableToUnpredicate = false;
 
@@ -561,6 +565,10 @@ class ARMSubtarget : public ARMGenSubtargetInfo {
   bool hasD16() const { return HasD16; }
   bool hasFullFP16() const { return HasFullFP16; }
 
+  bool hasFuseAES() const { return HasFuseAES; }
+  /// \brief Return true if the CPU supports any kind of instruction fusion.
+  bool hasFusion() const { return hasFuseAES(); }
+
   const Triple &getTargetTriple() const { return TargetTriple; }
 
   bool isTargetDarwin() const { return TargetTriple.isOSDarwin(); }

llvm/lib/Target/ARM/ARMTargetMachine.cpp

@@ -17,6 +17,7 @@
 #include "ARMRegisterBankInfo.h"
 #endif
 #include "ARMSubtarget.h"
+#include "ARMMacroFusion.h"
 #include "ARMTargetMachine.h"
 #include "ARMTargetObjectFile.h"
 #include "ARMTargetTransformInfo.h"
@@ -394,13 +395,19 @@ class ARMPassConfig : public TargetPassConfig {
   createMachineScheduler(MachineSchedContext *C) const override {
     ScheduleDAGMILive *DAG = createGenericSchedLive(C);
     // add DAG Mutations here.
+    const ARMSubtarget &ST = C->MF->getSubtarget<ARMSubtarget>();
+    if (ST.hasFusion())
+      DAG->addMutation(createARMMacroFusionDAGMutation());
     return DAG;
   }
 
   ScheduleDAGInstrs *
   createPostMachineScheduler(MachineSchedContext *C) const override {
     ScheduleDAGMI *DAG = createGenericSchedPostRA(C);
     // add DAG Mutations here.
+    const ARMSubtarget &ST = C->MF->getSubtarget<ARMSubtarget>();
+    if (ST.hasFusion())
+      DAG->addMutation(createARMMacroFusionDAGMutation());
     return DAG;
   }
 

llvm/lib/Target/ARM/CMakeLists.txt

@@ -49,6 +49,7 @@ add_llvm_target(ARMCodeGen
   ARMLoadStoreOptimizer.cpp
   ARMMCInstLower.cpp
   ARMMachineFunctionInfo.cpp
+  ARMMacroFusion.cpp
   ARMRegisterInfo.cpp
   ARMOptimizeBarriersPass.cpp
   ARMSelectionDAGInfo.cpp

llvm/test/CodeGen/ARM/misched-fusion-aes.ll

@@ -0,0 +1,203 @@
+; RUN: llc %s -o - -mtriple=armv8 -mattr=+crypto,+fuse-aes -enable-misched -disable-post-ra | FileCheck %s
+
+declare <16 x i8> @llvm.arm.neon.aese(<16 x i8> %d, <16 x i8> %k)
+declare <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %d)
+declare <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %d, <16 x i8> %k)
+declare <16 x i8> @llvm.arm.neon.aesimc(<16 x i8> %d)
+
+define void @aesea(<16 x i8>* %a0, <16 x i8>* %b0, <16 x i8>* %c0, <16 x i8> %d, <16 x i8> %e) {
+  %d0 = load <16 x i8>, <16 x i8>* %a0
+  %a1 = getelementptr inbounds <16 x i8>, <16 x i8>* %a0, i64 1
+  %d1 = load <16 x i8>, <16 x i8>* %a1
+  %a2 = getelementptr inbounds <16 x i8>, <16 x i8>* %a0, i64 2
+  %d2 = load <16 x i8>, <16 x i8>* %a2
+  %a3 = getelementptr inbounds <16 x i8>, <16 x i8>* %a0, i64 3
+  %d3 = load <16 x i8>, <16 x i8>* %a3
+  %k0 = load <16 x i8>, <16 x i8>* %b0
+  %e00 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %d0, <16 x i8> %k0)
+  %f00 = call <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %e00)
+  %e01 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %d1, <16 x i8> %k0)
+  %f01 = call <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %e01)
+  %e02 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %d2, <16 x i8> %k0)
+  %f02 = call <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %e02)
+  %e03 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %d3, <16 x i8> %k0)
+  %f03 = call <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %e03)
+  %b1 = getelementptr inbounds <16 x i8>, <16 x i8>* %b0, i64 1
+  %k1 = load <16 x i8>, <16 x i8>* %b1
+  %e10 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %f00, <16 x i8> %k1)
+  %f10 = call <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %e00)
+  %e11 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %f01, <16 x i8> %k1)
+  %f11 = call <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %e01)
+  %e12 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %f02, <16 x i8> %k1)
+  %f12 = call <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %e02)
+  %e13 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %f03, <16 x i8> %k1)
+  %f13 = call <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %e03)
+  %b2 = getelementptr inbounds <16 x i8>, <16 x i8>* %b0, i64 2
+  %k2 = load <16 x i8>, <16 x i8>* %b2
+  %e20 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %f10, <16 x i8> %k2)
+  %f20 = call <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %e10)
+  %e21 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %f11, <16 x i8> %k2)
+  %f21 = call <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %e11)
+  %e22 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %f12, <16 x i8> %k2)
+  %f22 = call <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %e12)
+  %e23 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %f13, <16 x i8> %k2)
+  %f23 = call <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %e13)
+  %b3 = getelementptr inbounds <16 x i8>, <16 x i8>* %b0, i64 3
+  %k3 = load <16 x i8>, <16 x i8>* %b3
+  %e30 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %f20, <16 x i8> %k3)
+  %f30 = call <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %e20)
+  %e31 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %f21, <16 x i8> %k3)
+  %f31 = call <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %e21)
+  %e32 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %f22, <16 x i8> %k3)
+  %f32 = call <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %e22)
+  %e33 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %f23, <16 x i8> %k3)
+  %f33 = call <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %e23)
+  %g0 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %f30, <16 x i8> %d)
+  %h0 = xor <16 x i8> %g0, %e
+  %g1 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %f31, <16 x i8> %d)
+  %h1 = xor <16 x i8> %g1, %e
+  %g2 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %f32, <16 x i8> %d)
+  %h2 = xor <16 x i8> %g2, %e
+  %g3 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %f33, <16 x i8> %d)
+  %h3 = xor <16 x i8> %g3, %e
+  store <16 x i8> %h0, <16 x i8>* %c0
+  %c1 = getelementptr inbounds <16 x i8>, <16 x i8>* %c0, i64 1
+  store <16 x i8> %h1, <16 x i8>* %c1
+  %c2 = getelementptr inbounds <16 x i8>, <16 x i8>* %c0, i64 2
+  store <16 x i8> %h2, <16 x i8>* %c2
+  %c3 = getelementptr inbounds <16 x i8>, <16 x i8>* %c0, i64 3
+  store <16 x i8> %h3, <16 x i8>* %c3
+  ret void
+
+; CHECK-LABEL: aesea:
+; CHECK: aese.8 [[QA:q[0-9][0-9]?]], {{q[0-9][0-9]?}}
+; CHECK-NEXT: aesmc.8 {{q[0-9][0-9]?}}, [[QA]]
+; CHECK: aese.8 [[QB:q[0-9][0-9]?]], {{q[0-9][0-9]?}}
+; CHECK-NEXT: aesmc.8 {{q[0-9][0-9]?}}, [[QB]]
+; CHECK: aese.8 [[QC:q[0-9][0-9]?]], {{q[0-9][0-9]?}}
+; CHECK-NEXT: aesmc.8 {{q[0-9][0-9]?}}, [[QC]]
+; CHECK: aese.8 [[QD:q[0-9][0-9]?]], {{q[0-9][0-9]?}}
+; CHECK-NEXT: aesmc.8 {{q[0-9][0-9]?}}, [[QD]]
+; CHECK: aese.8 [[QE:q[0-9][0-9]?]], {{q[0-9][0-9]?}}
+; CHECK-NEXT: aesmc.8 {{q[0-9][0-9]?}}, [[QE]]
+; CHECK: aese.8 [[QF:q[0-9][0-9]?]], {{q[0-9][0-9]?}}
+; CHECK-NEXT: aesmc.8 {{q[0-9][0-9]?}}, [[QF]]
+; CHECK: aese.8 {{q[0-9][0-9]?}}, {{q[0-9][0-9]?}}
+; CHECK: aese.8 [[QG:q[0-9][0-9]?]], {{q[0-9][0-9]?}}
+; CHECK-NEXT: aesmc.8 {{q[0-9][0-9]?}}, [[QG]]
+; CHECK: aese.8 {{q[0-9][0-9]?}}, {{q[0-9][0-9]?}}
+; CHECK: aese.8 [[QH:q[0-9][0-9]?]], {{q[0-9][0-9]?}}
+; CHECK-NEXT: aesmc.8 {{q[0-9][0-9]?}}, [[QH]]
+}
+
+define void @aesda(<16 x i8>* %a0, <16 x i8>* %b0, <16 x i8>* %c0, <16 x i8> %d, <16 x i8> %e) {
+  %d0 = load <16 x i8>, <16 x i8>* %a0
+  %a1 = getelementptr inbounds <16 x i8>, <16 x i8>* %a0, i64 1
+  %d1 = load <16 x i8>, <16 x i8>* %a1
+  %a2 = getelementptr inbounds <16 x i8>, <16 x i8>* %a0, i64 2
+  %d2 = load <16 x i8>, <16 x i8>* %a2
+  %a3 = getelementptr inbounds <16 x i8>, <16 x i8>* %a0, i64 3
+  %d3 = load <16 x i8>, <16 x i8>* %a3
+  %k0 = load <16 x i8>, <16 x i8>* %b0
+  %e00 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %d0, <16 x i8> %k0)
+  %f00 = call <16 x i8> @llvm.arm.neon.aesimc(<16 x i8> %e00)
+  %e01 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %d1, <16 x i8> %k0)
+  %f01 = call <16 x i8> @llvm.arm.neon.aesimc(<16 x i8> %e01)
+  %e02 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %d2, <16 x i8> %k0)
+  %f02 = call <16 x i8> @llvm.arm.neon.aesimc(<16 x i8> %e02)
+  %e03 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %d3, <16 x i8> %k0)
+  %f03 = call <16 x i8> @llvm.arm.neon.aesimc(<16 x i8> %e03)
+  %b1 = getelementptr inbounds <16 x i8>, <16 x i8>* %b0, i64 1
+  %k1 = load <16 x i8>, <16 x i8>* %b1
+  %e10 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %f00, <16 x i8> %k1)
+  %f10 = call <16 x i8> @llvm.arm.neon.aesimc(<16 x i8> %e00)
+  %e11 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %f01, <16 x i8> %k1)
+  %f11 = call <16 x i8> @llvm.arm.neon.aesimc(<16 x i8> %e01)
+  %e12 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %f02, <16 x i8> %k1)
+  %f12 = call <16 x i8> @llvm.arm.neon.aesimc(<16 x i8> %e02)
+  %e13 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %f03, <16 x i8> %k1)
+  %f13 = call <16 x i8> @llvm.arm.neon.aesimc(<16 x i8> %e03)
+  %b2 = getelementptr inbounds <16 x i8>, <16 x i8>* %b0, i64 2
+  %k2 = load <16 x i8>, <16 x i8>* %b2
+  %e20 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %f10, <16 x i8> %k2)
+  %f20 = call <16 x i8> @llvm.arm.neon.aesimc(<16 x i8> %e10)
+  %e21 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %f11, <16 x i8> %k2)
+  %f21 = call <16 x i8> @llvm.arm.neon.aesimc(<16 x i8> %e11)
+  %e22 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %f12, <16 x i8> %k2)
+  %f22 = call <16 x i8> @llvm.arm.neon.aesimc(<16 x i8> %e12)
+  %e23 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %f13, <16 x i8> %k2)
+  %f23 = call <16 x i8> @llvm.arm.neon.aesimc(<16 x i8> %e13)
+  %b3 = getelementptr inbounds <16 x i8>, <16 x i8>* %b0, i64 3
+  %k3 = load <16 x i8>, <16 x i8>* %b3
+  %e30 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %f20, <16 x i8> %k3)
+  %f30 = call <16 x i8> @llvm.arm.neon.aesimc(<16 x i8> %e20)
+  %e31 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %f21, <16 x i8> %k3)
+  %f31 = call <16 x i8> @llvm.arm.neon.aesimc(<16 x i8> %e21)
+  %e32 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %f22, <16 x i8> %k3)
+  %f32 = call <16 x i8> @llvm.arm.neon.aesimc(<16 x i8> %e22)
+  %e33 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %f23, <16 x i8> %k3)
+  %f33 = call <16 x i8> @llvm.arm.neon.aesimc(<16 x i8> %e23)
+  %g0 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %f30, <16 x i8> %d)
+  %h0 = xor <16 x i8> %g0, %e
+  %g1 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %f31, <16 x i8> %d)
+  %h1 = xor <16 x i8> %g1, %e
+  %g2 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %f32, <16 x i8> %d)
+  %h2 = xor <16 x i8> %g2, %e
+  %g3 = call <16 x i8> @llvm.arm.neon.aesd(<16 x i8> %f33, <16 x i8> %d)
+  %h3 = xor <16 x i8> %g3, %e
+  store <16 x i8> %h0, <16 x i8>* %c0
+  %c1 = getelementptr inbounds <16 x i8>, <16 x i8>* %c0, i64 1
+  store <16 x i8> %h1, <16 x i8>* %c1
+  %c2 = getelementptr inbounds <16 x i8>, <16 x i8>* %c0, i64 2
+  store <16 x i8> %h2, <16 x i8>* %c2
+  %c3 = getelementptr inbounds <16 x i8>, <16 x i8>* %c0, i64 3
+  store <16 x i8> %h3, <16 x i8>* %c3
+  ret void
+
+; CHECK-LABEL: aesda:
+; CHECK: aesd.8 [[QA:q[0-9][0-9]?]], {{q[0-9][0-9]?}}
+; CHECK-NEXT: aesimc.8 {{q[0-9][0-9]?}}, [[QA]]
+; CHECK: aesd.8 [[QB:q[0-9][0-9]?]], {{q[0-9][0-9]?}}
+; CHECK-NEXT: aesimc.8 {{q[0-9][0-9]?}}, [[QB]]
+; CHECK: aesd.8 [[QC:q[0-9][0-9]?]], {{q[0-9][0-9]?}}
+; CHECK-NEXT: aesimc.8 {{q[0-9][0-9]?}}, [[QC]]
+; CHECK: aesd.8 [[QD:q[0-9][0-9]?]], {{q[0-9][0-9]?}}
+; CHECK-NEXT: aesimc.8 {{q[0-9][0-9]?}}, [[QD]]
+; CHECK: aesd.8 [[QE:q[0-9][0-9]?]], {{q[0-9][0-9]?}}
+; CHECK-NEXT: aesimc.8 {{q[0-9][0-9]?}}, [[QE]]
+; CHECK: aesd.8 [[QF:q[0-9][0-9]?]], {{q[0-9][0-9]?}}
+; CHECK-NEXT: aesimc.8 {{q[0-9][0-9]?}}, [[QF]]
+; CHECK: aesd.8 {{q[0-9][0-9]?}}, {{q[0-9][0-9]?}}
+; CHECK: aesd.8 [[QG:q[0-9][0-9]?]], {{q[0-9][0-9]?}}
+; CHECK-NEXT: aesimc.8 {{q[0-9][0-9]?}}, [[QG]]
+; CHECK: aesd.8 {{q[0-9][0-9]?}}, {{q[0-9][0-9]?}}
+; CHECK: aesd.8 [[QH:q[0-9][0-9]?]], {{q[0-9][0-9]?}}
+; CHECK-NEXT: aesimc.8 {{q[0-9][0-9]?}}, [[QH]]
+}
+
+define void @aes_load_store(<16 x i8> *%p1, <16 x i8> *%p2 , <16 x i8> *%p3) {
+entry:
+  %x1 = alloca <16 x i8>, align 16
+  %x2 = alloca <16 x i8>, align 16
+  %x3 = alloca <16 x i8>, align 16
+  %x4 = alloca <16 x i8>, align 16
+  %x5 = alloca <16 x i8>, align 16
+  %in1 = load <16 x i8>, <16 x i8>* %p1, align 16
+  store <16 x i8> %in1, <16 x i8>* %x1, align 16
+  %aese1 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %in1, <16 x i8> %in1) #2
+  store <16 x i8> %aese1, <16 x i8>* %x2, align 16
+  %in2 = load <16 x i8>, <16 x i8>* %p2, align 16
+  %aesmc1= call <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %aese1) #2
+  store <16 x i8> %aesmc1, <16 x i8>* %x3, align 16
+  %aese2 = call <16 x i8> @llvm.arm.neon.aese(<16 x i8> %in1, <16 x i8> %in2) #2
+  store <16 x i8> %aese2, <16 x i8>* %x4, align 16
+  %aesmc2= call <16 x i8> @llvm.arm.neon.aesmc(<16 x i8> %aese2) #2
+  store <16 x i8> %aesmc2, <16 x i8>* %x5, align 16
+  ret void
+
+; CHECK-LABEL: aes_load_store:
+; CHECK: aese.8 [[QA:q[0-9][0-9]?]], {{q[0-9][0-9]?}}
+; CHECK-NEXT: aesmc.8 {{q[0-9][0-9]?}}, [[QA]]
+; CHECK: aese.8 [[QB:q[0-9][0-9]?]], {{q[0-9][0-9]?}}
+; CHECK-NEXT: aesmc.8 {{q[0-9][0-9]?}}, [[QB]]
+}