[llvm] [BOLT][AArch64] Fixed indirect call instrumentation snippet (PR #141918)
Alexey Moksyakov via llvm-commits
llvm-commits at lists.llvm.org
Fri Nov 7 00:09:06 PST 2025
https://github.com/yavtuk updated https://github.com/llvm/llvm-project/pull/141918
>From 0a63443fd53bfd0fe7d54ca4b59ecb1f71aebddf Mon Sep 17 00:00:00 2001
From: Moksyakov Alexey <moksyakov.alexey at huawei.com>
Date: Thu, 29 May 2025 07:28:13 +0000
Subject: [PATCH 1/8] [bolt][aarch64] Fixed indirect call instrumentation
snippet
Indirect call instrumentation snippet uses x16 register in exit
handler to go to destination target
__bolt_instr_ind_call_handler_func:
msr nzcv, x1
ldp x0, x1, [sp], #16
ldr x16, [sp], #16
ldp x0, x1, [sp], #16
br x16 <-----
Depend on compiler x16 register used to store smtng in cross function calling.
This patch adds the instrumentation snippet by calling instrumentation
runtime library through indirect call instruction and adding the wrapper
to store/load target value and the register for original indirect instruction.
Example:
mov x16, foo
infirectCall:
adrp x8, Label
add x8, x8, #:lo12:Lable
blr x8
Before:
Instrumented indirect call:
stp x0, x1, [sp, #-16]!
mov x0, x8
movk x1, #0x0, lsl #48
movk x1, #0x0, lsl #32
movk x1, #0x0, lsl #16
movk x1, #0x0
stp x0, x1, [sp, #-16]!
adrp x0, __bolt_instr_ind_call_handler_func
add x0, x0, #:lo12:__bolt_instr_ind_call_handler_func
blr x0
__bolt_instr_ind_call_handler: (exit snippet)
msr nzcv, x1
ldp x0, x1, [sp], #16
ldr x16, [sp], #16
ldp x0, x1, [sp], #16
br x16 <- overwrites the original value in X16
__bolt_instr_ind_call_handler_func: (entry snippet)
stp x0, x1, [sp, #-16]!
mrs x1, nzcv
adrp x0, __bolt_instr_ind_call_handler
add x0, x0, x0, #:lo12:__bolt_instr_ind_call_handler
ldr x0, [x0]
cmp x0, #0x0
b.eq __bolt_instr_ind_call_handler
str x30, [sp, #-16]!
blr x0 <--- runtime lib store/load all regs
ldr x30, [sp], #16
b __bolt_instr_ind_call_handler
_________________________________________________________________________
After:
mov x16, foo
infirectCall:
adrp x8, Label
add x8, x8, #:lo12:Lable
blr x8
Instrumented indirect call:
stp x0, x1, [sp, #-16]!
mov x0, x8
movk x1, #0x0, lsl #48
movk x1, #0x0, lsl #32
movk x1, #0x0, lsl #16
movk x1, #0x0
stp x0, x0, [sp, #-16]!
adrp x8, __bolt_instr_ind_call_handler_func
add x8, x8, #:lo12:__bolt_instr_ind_call_handler_func
str x30, [sp, #-16]!
blr x8 <--- call trampoline instr lib
ldr x30, [sp], #16
ldp x0, x1, [sp], #16
mov x8, x0 <---- restore original target
ldp x0, x1, [sp], #16
blr x8 <--- original indirect call instruction
// don't touch regs besides x0, x1
__bolt_instr_ind_call_handler: (exit snippet)
ldr x1, sp, 16
msr nzcv, x1
ldp x0, x1, [sp], #16
ret <---- return to original function with indirect call
__bolt_instr_ind_call_handler_func: (entry snippet)
stp x0, x1, [sp, #-16]!
mrs x1, nzcv
str x1, [sp, #-16]!
adrp x0, __bolt_instr_ind_call_handler
add x0, x0, #:lo12:__bolt_instr_ind_call_handler
ldr x0, [x0]
cmp x0, #0x0
b.eq __bolt_instr_ind_call_handler
str x30, [sp, #-16]!
blr x0 <--- runtime lib store/load all regs
ldr x30, [sp], #16
b __bolt_instr_ind_call_handler
---
bolt/include/bolt/Core/MCPlusBuilder.h | 5 +
bolt/lib/Passes/Instrumentation.cpp | 18 ++-
.../Target/AArch64/AArch64MCPlusBuilder.cpp | 126 +++++++++++++-----
bolt/runtime/instr.cpp | 4 +-
4 files changed, 111 insertions(+), 42 deletions(-)
diff --git a/bolt/include/bolt/Core/MCPlusBuilder.h b/bolt/include/bolt/Core/MCPlusBuilder.h
index b233452985502..a8a3a58dba836 100644
--- a/bolt/include/bolt/Core/MCPlusBuilder.h
+++ b/bolt/include/bolt/Core/MCPlusBuilder.h
@@ -511,6 +511,11 @@ class MCPlusBuilder {
llvm_unreachable("not implemented");
}
+ virtual void createDirectBranch(MCInst &Inst, const MCSymbol *Target,
+ MCContext *Ctx) {
+ llvm_unreachable("not implemented");
+ }
+
virtual MCPhysReg getX86R11() const { llvm_unreachable("not implemented"); }
virtual unsigned getShortBranchOpcode(unsigned Opcode) const {
diff --git a/bolt/lib/Passes/Instrumentation.cpp b/bolt/lib/Passes/Instrumentation.cpp
index fbf889279f1c0..e84acd00da369 100644
--- a/bolt/lib/Passes/Instrumentation.cpp
+++ b/bolt/lib/Passes/Instrumentation.cpp
@@ -293,9 +293,12 @@ void Instrumentation::instrumentIndirectTarget(BinaryBasicBlock &BB,
BinaryBasicBlock::iterator &Iter,
BinaryFunction &FromFunction,
uint32_t From) {
- auto L = FromFunction.getBinaryContext().scopeLock();
- const size_t IndCallSiteID = Summary->IndCallDescriptions.size();
- createIndCallDescription(FromFunction, From);
+ size_t IndCallSiteID;
+ {
+ auto L = FromFunction.getBinaryContext().scopeLock();
+ IndCallSiteID = Summary->IndCallDescriptions.size();
+ createIndCallDescription(FromFunction, From);
+ }
BinaryContext &BC = FromFunction.getBinaryContext();
bool IsTailCall = BC.MIB->isTailCall(*Iter);
@@ -305,9 +308,12 @@ void Instrumentation::instrumentIndirectTarget(BinaryBasicBlock &BB,
: IndCallHandlerExitBBFunction->getSymbol(),
IndCallSiteID, &*BC.Ctx);
- Iter = BB.eraseInstruction(Iter);
- Iter = insertInstructions(CounterInstrs, BB, Iter);
- --Iter;
+ if (!BC.isAArch64()) {
+ Iter = BB.eraseInstruction(Iter);
+ Iter = insertInstructions(CounterInstrs, BB, Iter);
+ --Iter;
+ } else
+ Iter = insertInstructions(CounterInstrs, BB, Iter);
}
bool Instrumentation::instrumentOneTarget(
diff --git a/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp b/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
index 9d5a578cfbdff..4895df33bec81 100644
--- a/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
+++ b/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
@@ -1966,6 +1966,15 @@ class AArch64MCPlusBuilder : public MCPlusBuilder {
convertJmpToTailCall(Inst);
}
+ void createDirectBranch(MCInst &Inst, const MCSymbol *Target,
+ MCContext *Ctx) override {
+ Inst.setOpcode(AArch64::B);
+ Inst.clear();
+ Inst.addOperand(MCOperand::createExpr(getTargetExprFor(
+ Inst, MCSymbolRefExpr::create(Target, MCSymbolRefExpr::VK_None, *Ctx),
+ *Ctx, 0)));
+ }
+
bool analyzeBranch(InstructionIterator Begin, InstructionIterator End,
const MCSymbol *&TBB, const MCSymbol *&FBB,
MCInst *&CondBranch,
@@ -2328,21 +2337,26 @@ class AArch64MCPlusBuilder : public MCPlusBuilder {
}
InstructionListType createInstrumentedIndCallHandlerExitBB() const override {
- InstructionListType Insts(5);
// Code sequence for instrumented indirect call handler:
+ // ldr x1, [sp, #16]
// msr nzcv, x1
// ldp x0, x1, [sp], #16
- // ldr x16, [sp], #16
- // ldp x0, x1, [sp], #16
- // br x16
- setSystemFlag(Insts[0], AArch64::X1);
- createPopRegisters(Insts[1], AArch64::X0, AArch64::X1);
- // Here we load address of the next function which should be called in the
- // original binary to X16 register. Writing to X16 is permitted without
- // needing to restore.
- loadReg(Insts[2], AArch64::X16, AArch64::SP);
- createPopRegisters(Insts[3], AArch64::X0, AArch64::X1);
- createIndirectBranch(Insts[4], AArch64::X16, 0);
+ // ret
+
+ InstructionListType Insts;
+
+ Insts.emplace_back();
+ loadReg(Insts.back(), AArch64::X1, AArch64::SP);
+
+ Insts.emplace_back();
+ setSystemFlag(Insts.back(), AArch64::X1);
+
+ Insts.emplace_back();
+ createPopRegisters(Insts.back(), AArch64::X0, AArch64::X1);
+
+ Insts.emplace_back();
+ createReturn(Insts.back());
+
return Insts;
}
@@ -2418,39 +2432,69 @@ class AArch64MCPlusBuilder : public MCPlusBuilder {
MCSymbol *HandlerFuncAddr,
int CallSiteID,
MCContext *Ctx) override {
- InstructionListType Insts;
// Code sequence used to enter indirect call instrumentation helper:
- // stp x0, x1, [sp, #-16]! createPushRegisters
+ // stp x0, x1, [sp, #-16]! createPushRegisters (1)
// mov target x0 convertIndirectCallToLoad -> orr x0 target xzr
// mov x1 CallSiteID createLoadImmediate ->
// movk x1, #0x0, lsl #48
// movk x1, #0x0, lsl #32
// movk x1, #0x0, lsl #16
// movk x1, #0x0
- // stp x0, x1, [sp, #-16]!
- // bl *HandlerFuncAddr createIndirectCall ->
+ // stp x0, x1, [sp, #-16]! (2)
// adr x0 *HandlerFuncAddr -> adrp + add
- // blr x0
+ // str x30, [sp, #-16]! (3)
+ // blr x0 (__bolt_instr_ind_call_handler_func)
+ // ldr x30, sp, #16 (3)
+ // ldp x0, x1, [sp], #16 (2)
+ // mov x0, x0 ; move target address to used register
+ // ldp x0, x1, [sp], #16 (1)
+
+ InstructionListType Insts;
Insts.emplace_back();
- createPushRegisters(Insts.back(), AArch64::X0, AArch64::X1);
+ createPushRegisters(Insts.back(), getIntArgRegister(0),
+ getIntArgRegister(1));
Insts.emplace_back(CallInst);
- convertIndirectCallToLoad(Insts.back(), AArch64::X0);
+ convertIndirectCallToLoad(Insts.back(), getIntArgRegister(0));
InstructionListType LoadImm =
createLoadImmediate(getIntArgRegister(1), CallSiteID);
Insts.insert(Insts.end(), LoadImm.begin(), LoadImm.end());
Insts.emplace_back();
- createPushRegisters(Insts.back(), AArch64::X0, AArch64::X1);
+ createPushRegisters(Insts.back(), getIntArgRegister(0),
+ getIntArgRegister(1));
Insts.resize(Insts.size() + 2);
- InstructionListType Addr =
- materializeAddress(HandlerFuncAddr, Ctx, AArch64::X0);
+ InstructionListType Addr = materializeAddress(
+ HandlerFuncAddr, Ctx, CallInst.getOperand(0).getReg());
assert(Addr.size() == 2 && "Invalid Addr size");
std::copy(Addr.begin(), Addr.end(), Insts.end() - Addr.size());
+
+ Insts.emplace_back();
+ storeReg(Insts.back(), AArch64::LR, getSpRegister(/*Size*/ 8));
+
+ Insts.emplace_back();
+ createIndirectCallInst(Insts.back(), false,
+ CallInst.getOperand(0).getReg());
+
Insts.emplace_back();
- createIndirectCallInst(Insts.back(), isTailCall(CallInst), AArch64::X0);
+ loadReg(Insts.back(), AArch64::LR, getSpRegister(/*Size*/ 8));
- // Carry over metadata including tail call marker if present.
- stripAnnotations(Insts.back());
- moveAnnotations(std::move(CallInst), Insts.back());
+ Insts.emplace_back();
+ createPopRegisters(Insts.back(), getIntArgRegister(0),
+ getIntArgRegister(1));
+
+ // move x0 to indirect call register
+ Insts.emplace_back();
+ Insts.back().setOpcode(AArch64::ORRXrs);
+ Insts.back().insert(Insts.back().begin(),
+ MCOperand::createReg(CallInst.getOperand(0).getReg()));
+ Insts.back().insert(Insts.back().begin() + 1,
+ MCOperand::createReg(AArch64::XZR));
+ Insts.back().insert(Insts.back().begin() + 2,
+ MCOperand::createReg(getIntArgRegister(0)));
+ Insts.back().insert(Insts.back().begin() + 3, MCOperand::createImm(0));
+
+ Insts.emplace_back();
+ createPopRegisters(Insts.back(), getIntArgRegister(0),
+ getIntArgRegister(1));
return Insts;
}
@@ -2472,30 +2516,44 @@ class AArch64MCPlusBuilder : public MCPlusBuilder {
// ldr x30, [sp], #16
// b IndCallHandler
InstructionListType Insts;
+
Insts.emplace_back();
- createPushRegisters(Insts.back(), AArch64::X0, AArch64::X1);
+ createPushRegisters(Insts.back(), getIntArgRegister(0),
+ getIntArgRegister(1));
+
Insts.emplace_back();
getSystemFlag(Insts.back(), getIntArgRegister(1));
+
+ Insts.emplace_back();
+ storeReg(Insts.back(), getIntArgRegister(1), getSpRegister(/*Size*/ 8));
+
Insts.emplace_back();
Insts.emplace_back();
InstructionListType Addr =
- materializeAddress(InstrTrampoline, Ctx, AArch64::X0);
+ materializeAddress(InstrTrampoline, Ctx, getIntArgRegister(0));
std::copy(Addr.begin(), Addr.end(), Insts.end() - Addr.size());
assert(Addr.size() == 2 && "Invalid Addr size");
+
Insts.emplace_back();
- loadReg(Insts.back(), AArch64::X0, AArch64::X0);
+ loadReg(Insts.back(), getIntArgRegister(0), getIntArgRegister(0));
+
InstructionListType cmpJmp =
- createCmpJE(AArch64::X0, 0, IndCallHandler, Ctx);
+ createCmpJE(getIntArgRegister(0), 0, IndCallHandler, Ctx);
Insts.insert(Insts.end(), cmpJmp.begin(), cmpJmp.end());
+
Insts.emplace_back();
- storeReg(Insts.back(), AArch64::LR, AArch64::SP);
+ storeReg(Insts.back(), AArch64::LR, getSpRegister(/*Size*/ 8));
+
Insts.emplace_back();
Insts.back().setOpcode(AArch64::BLR);
- Insts.back().addOperand(MCOperand::createReg(AArch64::X0));
+ Insts.back().addOperand(MCOperand::createReg(getIntArgRegister(0)));
+
Insts.emplace_back();
- loadReg(Insts.back(), AArch64::LR, AArch64::SP);
+ loadReg(Insts.back(), AArch64::LR, getSpRegister(/*Size*/ 8));
+
Insts.emplace_back();
- createDirectCall(Insts.back(), IndCallHandler, Ctx, /*IsTailCall*/ true);
+ createDirectBranch(Insts.back(), IndCallHandler, Ctx);
+
return Insts;
}
diff --git a/bolt/runtime/instr.cpp b/bolt/runtime/instr.cpp
index ae356e71cbe41..a174b982cbb84 100644
--- a/bolt/runtime/instr.cpp
+++ b/bolt/runtime/instr.cpp
@@ -1668,7 +1668,7 @@ extern "C" __attribute((naked)) void __bolt_instr_indirect_call()
#if defined(__aarch64__)
// clang-format off
__asm__ __volatile__(SAVE_ALL
- "ldp x0, x1, [sp, #288]\n"
+ "ldp x0, x1, [sp, #320]\n"
"bl instrumentIndirectCall\n"
RESTORE_ALL
"ret\n"
@@ -1705,7 +1705,7 @@ extern "C" __attribute((naked)) void __bolt_instr_indirect_tailcall()
#if defined(__aarch64__)
// clang-format off
__asm__ __volatile__(SAVE_ALL
- "ldp x0, x1, [sp, #288]\n"
+ "ldp x0, x1, [sp, #320]\n"
"bl instrumentIndirectCall\n"
RESTORE_ALL
"ret\n"
>From 4b0560fb10485e561563325d491c964c6a1d3768 Mon Sep 17 00:00:00 2001
From: Alexey Moksyakov <yavtuk at yandex.ru>
Date: Wed, 5 Nov 2025 14:22:29 +0300
Subject: [PATCH 2/8] [bolt][aarch64] Refactor insturmentation code snippet.
NFC
Signed-off-by: Moksyakov Alexey <moksyakov.alexey at huawei.com>
---
.../Target/AArch64/AArch64MCPlusBuilder.cpp | 19 ++++++++++---------
1 file changed, 10 insertions(+), 9 deletions(-)
diff --git a/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp b/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
index 4895df33bec81..38ed4a390f747 100644
--- a/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
+++ b/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
@@ -1971,8 +1971,7 @@ class AArch64MCPlusBuilder : public MCPlusBuilder {
Inst.setOpcode(AArch64::B);
Inst.clear();
Inst.addOperand(MCOperand::createExpr(getTargetExprFor(
- Inst, MCSymbolRefExpr::create(Target, MCSymbolRefExpr::VK_None, *Ctx),
- *Ctx, 0)));
+ Inst, MCSymbolRefExpr::create(Target, *Ctx), *Ctx, 0)));
}
bool analyzeBranch(InstructionIterator Begin, InstructionIterator End,
@@ -2346,13 +2345,14 @@ class AArch64MCPlusBuilder : public MCPlusBuilder {
InstructionListType Insts;
Insts.emplace_back();
- loadReg(Insts.back(), AArch64::X1, AArch64::SP);
+ loadReg(Insts.back(), getIntArgRegister(1), AArch64::SP);
Insts.emplace_back();
- setSystemFlag(Insts.back(), AArch64::X1);
+ setSystemFlag(Insts.back(), getIntArgRegister(1));
Insts.emplace_back();
- createPopRegisters(Insts.back(), AArch64::X0, AArch64::X1);
+ createPopRegisters(Insts.back(), getIntArgRegister(0),
+ getIntArgRegister(1));
Insts.emplace_back();
createReturn(Insts.back());
@@ -2503,11 +2503,12 @@ class AArch64MCPlusBuilder : public MCPlusBuilder {
createInstrumentedIndCallHandlerEntryBB(const MCSymbol *InstrTrampoline,
const MCSymbol *IndCallHandler,
MCContext *Ctx) override {
- // Code sequence used to check whether InstrTampoline was initialized
+ // Code sequence used to check whether InstrTrampoline was initialized
// and call it if so, returns via IndCallHandler
// stp x0, x1, [sp, #-16]!
// mrs x1, nzcv
- // adr x0, InstrTrampoline -> adrp + add
+ // adrp x0, InstrTrampoline
+ // add x0, x0, #lo12:InstrTrampoline
// ldr x0, [x0]
// subs x0, x0, #0x0
// b.eq IndCallHandler
@@ -2537,9 +2538,9 @@ class AArch64MCPlusBuilder : public MCPlusBuilder {
Insts.emplace_back();
loadReg(Insts.back(), getIntArgRegister(0), getIntArgRegister(0));
- InstructionListType cmpJmp =
+ InstructionListType CmpJmp =
createCmpJE(getIntArgRegister(0), 0, IndCallHandler, Ctx);
- Insts.insert(Insts.end(), cmpJmp.begin(), cmpJmp.end());
+ Insts.insert(Insts.end(), CmpJmp.begin(), CmpJmp.end());
Insts.emplace_back();
storeReg(Insts.back(), AArch64::LR, getSpRegister(/*Size*/ 8));
>From f5d4f48978635a4b766622ce04cb2f0871508157 Mon Sep 17 00:00:00 2001
From: Alexey Moksyakov <yavtuk at yandex.ru>
Date: Wed, 5 Nov 2025 14:25:43 +0300
Subject: [PATCH 3/8] [bolt][aarch64] Extend checking for indirect call
instrumentation test
---
.../AArch64/instrumentation-ind-call.c | 40 ++++++++++++++++++-
1 file changed, 39 insertions(+), 1 deletion(-)
diff --git a/bolt/test/runtime/AArch64/instrumentation-ind-call.c b/bolt/test/runtime/AArch64/instrumentation-ind-call.c
index f9056da333b4e..f66ac5a351cea 100644
--- a/bolt/test/runtime/AArch64/instrumentation-ind-call.c
+++ b/bolt/test/runtime/AArch64/instrumentation-ind-call.c
@@ -15,9 +15,47 @@ int main() {
REQUIRES: system-linux,bolt-runtime
RUN: %clang %cflags %s -o %t.exe -Wl,-q -no-pie -fpie
+RUN: llvm-objdump --disassemble-symbols=main %t.exe \
+RUN: | FileCheck %s --check-prefix=CHECKINDIRECTREG
+
+CHECKINDIRECTREG: mov w0, #0xa
+CHECKINDIRECTREG-NEXT: mov w1, #0x14
+CHECKINDIRECTREG-NEXT: blr x8
RUN: llvm-bolt %t.exe --instrument --instrumentation-file=%t.fdata \
-RUN: -o %t.instrumented
+RUN: -o %t.instrumented \
+RUN: | FileCheck %s --check-prefix=CHECK-INSTR-LOG
+
+CHECK-INSTR-LOG: BOLT-INSTRUMENTER: Number of indirect call site descriptors: 1
+
+RUN: llvm-objdump --disassemble-symbols=main %t.instrumented \
+RUN: | FileCheck %s --check-prefix=CHECK-INSTR-INDIRECTREG
+
+CHECK-INSTR-INDIRECTREG: mov w0, #0xa
+CHECK-INSTR-INDIRECTREG-NEXT: mov w1, #0x14
+// store current values
+CHECK-INSTR-INDIRECTREG-NEXT: stp x0, x1, {{.*}}
+// store the indirect target address in x0
+CHECK-INSTR-INDIRECTREG-NEXT: mov x0, x8
+// load callsite id into x1
+CHECK-INSTR-INDIRECTREG-NEXT: movk x1, {{.*}}
+CHECK-INSTR-INDIRECTREG-NEXT: movk x1, {{.*}}
+CHECK-INSTR-INDIRECTREG-NEXT: movk x1, {{.*}}
+CHECK-INSTR-INDIRECTREG-NEXT: movk x1, {{.*}}
+CHECK-INSTR-INDIRECTREG-NEXT: stp x0, x1, {{.*}}
+CHECK-INSTR-INDIRECTREG-NEXT: adrp x8, {{.*}}
+CHECK-INSTR-INDIRECTREG-NEXT: add x8, {{.*}}
+// store return address, used by library handler
+CHECK-INSTR-INDIRECTREG-NEXT: str x30, {{.*}}
+// call instrumentation library handler function
+CHECK-INSTR-INDIRECTREG-NEXT: blr x8
+// restore registers saved before
+CHECK-INSTR-INDIRECTREG-NEXT: ldr x30, {{.*}}
+CHECK-INSTR-INDIRECTREG-NEXT: ldp x0, x1, {{.*}}
+CHECK-INSTR-INDIRECTREG-NEXT: mov x8, x0
+CHECK-INSTR-INDIRECTREG-NEXT: ldp x0, x1, {{.*}}
+// original indirect call instruction
+CHECK-INSTR-INDIRECTREG-NEXT: blr x8
# Instrumented program needs to finish returning zero
RUN: %t.instrumented | FileCheck %s -check-prefix=CHECK-OUTPUT
>From a7e1e2c801bbf9d506fe56babafda9bf1971ee63 Mon Sep 17 00:00:00 2001
From: Alexey Moksyakov <yavtuk at yandex.ru>
Date: Wed, 5 Nov 2025 16:21:35 +0300
Subject: [PATCH 4/8] [bolt][aarch64] Sync comment with code snippet for
indirect call. NFC
Signed-off-by: Moksyakov Alexey <moksyakov.alexey at huawei.com>
---
bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp | 1 +
1 file changed, 1 insertion(+)
diff --git a/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp b/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
index 38ed4a390f747..69df630ff17f1 100644
--- a/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
+++ b/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
@@ -2507,6 +2507,7 @@ class AArch64MCPlusBuilder : public MCPlusBuilder {
// and call it if so, returns via IndCallHandler
// stp x0, x1, [sp, #-16]!
// mrs x1, nzcv
+ // str x1, [sp, #-16]!
// adrp x0, InstrTrampoline
// add x0, x0, #lo12:InstrTrampoline
// ldr x0, [x0]
>From 9eede6946bb6e40027c3a804cfc5ad77f91ba018 Mon Sep 17 00:00:00 2001
From: Moksyakov Alexey <moksyakov.alexey at huawei.com>
Date: Thu, 6 Nov 2025 13:55:30 +0000
Subject: [PATCH 5/8] [bolt][aarch64] Change indirect call handler snippet
Signed-off-by: Moksyakov Alexey <moksyakov.alexey at huawei.com>
---
.../Target/AArch64/AArch64MCPlusBuilder.cpp | 27 ++++++++++++-------
1 file changed, 17 insertions(+), 10 deletions(-)
diff --git a/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp b/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
index 69df630ff17f1..7526796d16b5c 100644
--- a/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
+++ b/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
@@ -2509,8 +2509,7 @@ class AArch64MCPlusBuilder : public MCPlusBuilder {
// mrs x1, nzcv
// str x1, [sp, #-16]!
// adrp x0, InstrTrampoline
- // add x0, x0, #lo12:InstrTrampoline
- // ldr x0, [x0]
+ // ldr x0, [x0, #lo12:InstrTrampoline]
// subs x0, x0, #0x0
// b.eq IndCallHandler
// str x30, [sp, #-16]!
@@ -2529,15 +2528,23 @@ class AArch64MCPlusBuilder : public MCPlusBuilder {
Insts.emplace_back();
storeReg(Insts.back(), getIntArgRegister(1), getSpRegister(/*Size*/ 8));
- Insts.emplace_back();
- Insts.emplace_back();
- InstructionListType Addr =
- materializeAddress(InstrTrampoline, Ctx, getIntArgRegister(0));
- std::copy(Addr.begin(), Addr.end(), Insts.end() - Addr.size());
- assert(Addr.size() == 2 && "Invalid Addr size");
+ // load handler address
+ MCInst InstAdrp;
+ InstAdrp.setOpcode(AArch64::ADRP);
+ InstAdrp.addOperand(MCOperand::createReg(getIntArgRegister(0)));
+ InstAdrp.addOperand(MCOperand::createImm(0));
+ setOperandToSymbolRef(InstAdrp, /* OpNum */ 1, InstrTrampoline,
+ /* Addend */ 0, Ctx, ELF::R_AARCH64_ADR_GOT_PAGE);
+ Insts.emplace_back(InstAdrp);
- Insts.emplace_back();
- loadReg(Insts.back(), getIntArgRegister(0), getIntArgRegister(0));
+ MCInst InstLoad;
+ InstLoad.setOpcode(AArch64::LDRXui);
+ InstLoad.addOperand(MCOperand::createReg(getIntArgRegister(0)));
+ InstLoad.addOperand(MCOperand::createReg(getIntArgRegister(0)));
+ InstLoad.addOperand(MCOperand::createImm(0));
+ setOperandToSymbolRef(InstLoad, /* OpNum */ 2, InstrTrampoline,
+ /* Addend */ 0, Ctx, ELF::R_AARCH64_LD64_GOT_LO12_NC);
+ Insts.emplace_back(InstLoad);
InstructionListType CmpJmp =
createCmpJE(getIntArgRegister(0), 0, IndCallHandler, Ctx);
>From cc807f967e42357f77f9138ab849345da389c5c1 Mon Sep 17 00:00:00 2001
From: Moksyakov Alexey <moksyakov.alexey at huawei.com>
Date: Thu, 6 Nov 2025 14:46:29 +0000
Subject: [PATCH 6/8] [bolt][aarch64] Change indirect call instrumentation
snippet
Remove redundant load/store instructions in entryBB/exitBB
indirect call instrumentation snippet, move msr/mrs to
SAVE_ALL/RESTORE_ALL
Signed-off-by: Moksyakov Alexey <moksyakov.alexey at huawei.com>
---
bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp | 16 ----------------
bolt/runtime/instr.cpp | 4 ++--
bolt/runtime/sys_aarch64.h | 6 ++++--
3 files changed, 6 insertions(+), 20 deletions(-)
diff --git a/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp b/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
index 7526796d16b5c..d83bc401f9929 100644
--- a/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
+++ b/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
@@ -2337,19 +2337,11 @@ class AArch64MCPlusBuilder : public MCPlusBuilder {
InstructionListType createInstrumentedIndCallHandlerExitBB() const override {
// Code sequence for instrumented indirect call handler:
- // ldr x1, [sp, #16]
- // msr nzcv, x1
// ldp x0, x1, [sp], #16
// ret
InstructionListType Insts;
- Insts.emplace_back();
- loadReg(Insts.back(), getIntArgRegister(1), AArch64::SP);
-
- Insts.emplace_back();
- setSystemFlag(Insts.back(), getIntArgRegister(1));
-
Insts.emplace_back();
createPopRegisters(Insts.back(), getIntArgRegister(0),
getIntArgRegister(1));
@@ -2506,8 +2498,6 @@ class AArch64MCPlusBuilder : public MCPlusBuilder {
// Code sequence used to check whether InstrTrampoline was initialized
// and call it if so, returns via IndCallHandler
// stp x0, x1, [sp, #-16]!
- // mrs x1, nzcv
- // str x1, [sp, #-16]!
// adrp x0, InstrTrampoline
// ldr x0, [x0, #lo12:InstrTrampoline]
// subs x0, x0, #0x0
@@ -2522,12 +2512,6 @@ class AArch64MCPlusBuilder : public MCPlusBuilder {
createPushRegisters(Insts.back(), getIntArgRegister(0),
getIntArgRegister(1));
- Insts.emplace_back();
- getSystemFlag(Insts.back(), getIntArgRegister(1));
-
- Insts.emplace_back();
- storeReg(Insts.back(), getIntArgRegister(1), getSpRegister(/*Size*/ 8));
-
// load handler address
MCInst InstAdrp;
InstAdrp.setOpcode(AArch64::ADRP);
diff --git a/bolt/runtime/instr.cpp b/bolt/runtime/instr.cpp
index a174b982cbb84..90ace77718603 100644
--- a/bolt/runtime/instr.cpp
+++ b/bolt/runtime/instr.cpp
@@ -1668,7 +1668,7 @@ extern "C" __attribute((naked)) void __bolt_instr_indirect_call()
#if defined(__aarch64__)
// clang-format off
__asm__ __volatile__(SAVE_ALL
- "ldp x0, x1, [sp, #320]\n"
+ "ldp x0, x1, [sp, #304]\n"
"bl instrumentIndirectCall\n"
RESTORE_ALL
"ret\n"
@@ -1705,7 +1705,7 @@ extern "C" __attribute((naked)) void __bolt_instr_indirect_tailcall()
#if defined(__aarch64__)
// clang-format off
__asm__ __volatile__(SAVE_ALL
- "ldp x0, x1, [sp, #320]\n"
+ "ldp x0, x1, [sp, #304]\n"
"bl instrumentIndirectCall\n"
RESTORE_ALL
"ret\n"
diff --git a/bolt/runtime/sys_aarch64.h b/bolt/runtime/sys_aarch64.h
index 77c9cfcc99f98..1875e0620678e 100644
--- a/bolt/runtime/sys_aarch64.h
+++ b/bolt/runtime/sys_aarch64.h
@@ -18,10 +18,12 @@
"stp x24, x25, [sp, #-16]!\n" \
"stp x26, x27, [sp, #-16]!\n" \
"stp x28, x29, [sp, #-16]!\n" \
- "str x30, [sp,#-16]!\n"
+ "mrs x29, nzcv\n" \
+ "stp x29, x30, [sp, #-16]!\n"
// Mirrors SAVE_ALL
#define RESTORE_ALL \
- "ldr x30, [sp], #16\n" \
+ "ldp x29, x30, [sp], #16\n" \
+ "msr nzcv, x29\n" \
"ldp x28, x29, [sp], #16\n" \
"ldp x26, x27, [sp], #16\n" \
"ldp x24, x25, [sp], #16\n" \
>From 1d9511cb4dfd998880d02de48b2f5899517b0566 Mon Sep 17 00:00:00 2001
From: Alexey Moksyakov <yavtuk at yandex.ru>
Date: Thu, 6 Nov 2025 19:03:10 +0300
Subject: [PATCH 7/8] [bolt][aarch64] set unused for set/get system flag
---
bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp b/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
index d83bc401f9929..3d962244a0da6 100644
--- a/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
+++ b/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
@@ -38,14 +38,14 @@ using namespace bolt;
namespace {
-static void getSystemFlag(MCInst &Inst, MCPhysReg RegName) {
+[[maybe_unused]] static void getSystemFlag(MCInst &Inst, MCPhysReg RegName) {
Inst.setOpcode(AArch64::MRS);
Inst.clear();
Inst.addOperand(MCOperand::createReg(RegName));
Inst.addOperand(MCOperand::createImm(AArch64SysReg::NZCV));
}
-static void setSystemFlag(MCInst &Inst, MCPhysReg RegName) {
+[[maybe_unused]] static void setSystemFlag(MCInst &Inst, MCPhysReg RegName) {
Inst.setOpcode(AArch64::MSR);
Inst.clear();
Inst.addOperand(MCOperand::createImm(AArch64SysReg::NZCV));
>From 411c273a016e60c5142e4a185017bce4fd7a7444 Mon Sep 17 00:00:00 2001
From: Alexey Moksyakov <yavtuk at yandex.ru>
Date: Fri, 7 Nov 2025 11:08:53 +0300
Subject: [PATCH 8/8] [bolt][aarch64] Remove redundant stp/ldp instr for
indirect call instrumentation
---
bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp | 10 ----------
1 file changed, 10 deletions(-)
diff --git a/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp b/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
index 3d962244a0da6..11ab54661a948 100644
--- a/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
+++ b/bolt/lib/Target/AArch64/AArch64MCPlusBuilder.cpp
@@ -2337,15 +2337,10 @@ class AArch64MCPlusBuilder : public MCPlusBuilder {
InstructionListType createInstrumentedIndCallHandlerExitBB() const override {
// Code sequence for instrumented indirect call handler:
- // ldp x0, x1, [sp], #16
// ret
InstructionListType Insts;
- Insts.emplace_back();
- createPopRegisters(Insts.back(), getIntArgRegister(0),
- getIntArgRegister(1));
-
Insts.emplace_back();
createReturn(Insts.back());
@@ -2497,7 +2492,6 @@ class AArch64MCPlusBuilder : public MCPlusBuilder {
MCContext *Ctx) override {
// Code sequence used to check whether InstrTrampoline was initialized
// and call it if so, returns via IndCallHandler
- // stp x0, x1, [sp, #-16]!
// adrp x0, InstrTrampoline
// ldr x0, [x0, #lo12:InstrTrampoline]
// subs x0, x0, #0x0
@@ -2508,10 +2502,6 @@ class AArch64MCPlusBuilder : public MCPlusBuilder {
// b IndCallHandler
InstructionListType Insts;
- Insts.emplace_back();
- createPushRegisters(Insts.back(), getIntArgRegister(0),
- getIntArgRegister(1));
-
// load handler address
MCInst InstAdrp;
InstAdrp.setOpcode(AArch64::ADRP);
More information about the llvm-commits
mailing list