Follow-up of D78082 (x86-64).
This change avoids dynamic relocations in `xray_instr_map` for ARM/AArch64/powerpc64le.
MIPS64 cannot use 64-bit PC-relative addresses because R_MIPS_PC64 is not defined.
Because MIPS32 shares the same code, for simplicity, we don't use PC-relative addresses for MIPS32 as well.
Tested on AArch64 Linux and ppc64le Linux.
Reviewed By: ianlevesque
Differential Revision: https://reviews.llvm.org/D78590
// When |Enable|==false, we set back the first instruction in the sled to be
// B #32
- uint32_t *FirstAddress = reinterpret_cast<uint32_t *>(Sled.Address);
+ uint32_t *FirstAddress = reinterpret_cast<uint32_t *>(Sled.address());
uint32_t *CurAddress = FirstAddress + 1;
if (Enable) {
*CurAddress = uint32_t(PatchOpcodes::PO_LdrW0_12);
// When |Enable|==false, we set back the first instruction in the sled to be
// B #20
- uint32_t *FirstAddress = reinterpret_cast<uint32_t *>(Sled.Address);
+ uint32_t *FirstAddress = reinterpret_cast<uint32_t *>(Sled.address());
uint32_t *CurAddress = FirstAddress + 1;
if (Enable) {
CurAddress =
unsigned char Version;
unsigned char Padding[13]; // Need 32 bytes
uint64_t address() const {
-#ifndef __x86_64__
- // R_MIPS_PC64 does not exist. Use absolute address even for version 2.
- return Address;
-#else
- // TODO Eventually all targets but MIPS64 should take this branch.
if (Version < 2)
return Address;
// The target address is relative to the location of the Address variable.
return reinterpret_cast<uint64_t>(&Address) + Address;
-#endif
}
#elif SANITIZER_WORDSIZE == 32
uint32_t Address;
unsigned char AlwaysInstrument;
unsigned char Version;
unsigned char Padding[5]; // Need 16 bytes
- uint32_t address() const { return Address; }
+ uint32_t address() const {
+ if (Version < 2)
+ return Address;
+ // The target address is relative to the location of the Address variable.
+ return reinterpret_cast<uint32_t>(&Address) + Address;
+ }
#else
#error "Unsupported word size."
#endif
bool patchFunctionEntry(const bool Enable, uint32_t FuncId,
const XRaySledEntry &Sled,
void (*Trampoline)()) XRAY_NEVER_INSTRUMENT {
- const uint64_t Address = Sled.Address;
+ const uint64_t Address = Sled.address();
if (Enable) {
// lis 0, FuncId[16..32]
// li 0, FuncId[0..15]
bool patchFunctionExit(const bool Enable, uint32_t FuncId,
const XRaySledEntry &Sled) XRAY_NEVER_INSTRUMENT {
- const uint64_t Address = Sled.Address;
+ const uint64_t Address = Sled.address();
if (Enable) {
// lis 0, FuncId[16..32]
// li 0, FuncId[0..15]
MCSection *InstMap = nullptr;
MCSection *FnSledIndex = nullptr;
const Triple &TT = TM.getTargetTriple();
- // Version 2 uses a PC-relative address on all supported targets.
- bool PCRel = TT.isX86();
+ // Use PC-relative addresses on all targets except MIPS (MIPS64 cannot use
+ // PC-relative addresses because R_MIPS_PC64 does not exist).
+ bool PCRel = !TT.isMIPS();
if (TT.isOSBinFormatELF()) {
auto LinkedToSym = cast<MCSymbolELF>(CurrentFnSym);
auto Flags = ELF::SHF_ALLOC | ELF::SHF_LINK_ORDER;
EmitToStreamer(*OutStreamer, MCInstBuilder(AArch64::HINT).addImm(0));
OutStreamer->emitLabel(Target);
- recordSled(CurSled, MI, Kind);
+ recordSled(CurSled, MI, Kind, 2);
}
void AArch64AsmPrinter::LowerHWASAN_CHECK_MEMACCESS(const MachineInstr &MI) {
emitNops(NoopsInSledCount);
OutStreamer->emitLabel(Target);
- recordSled(CurSled, MI, Kind);
+ recordSled(CurSled, MI, Kind, 2);
}
void ARMAsmPrinter::LowerPATCHABLE_FUNCTION_ENTER(const MachineInstr &MI)
OutContext)));
EmitToStreamer(*OutStreamer, MCInstBuilder(PPC::MTLR8).addReg(PPC::X0));
OutStreamer->emitLabel(EndOfSled);
- recordSled(BeginOfSled, *MI, SledKind::FUNCTION_ENTER);
+ recordSled(BeginOfSled, *MI, SledKind::FUNCTION_ENTER, 2);
break;
}
case TargetOpcode::PATCHABLE_RET: {
EmitToStreamer(*OutStreamer, RetInst);
if (IsConditional)
OutStreamer->emitLabel(FallthroughLabel);
- recordSled(BeginOfSled, *MI, SledKind::FUNCTION_EXIT);
+ recordSled(BeginOfSled, *MI, SledKind::FUNCTION_EXIT, 2);
break;
}
case TargetOpcode::PATCHABLE_FUNCTION_EXIT:
; CHECK-NEXT: ret\r
}\r
; CHECK-LABEL: xray_instr_map\r
-; CHECK-LABEL: Lxray_sleds_start0:\r
-; CHECK: .xword .Lxray_sled_0\r
-; CHECK: .xword .Lxray_sled_1\r
+; CHECK-LABEL: .Lxray_sleds_start0:\r
+; CHECK-NEXT: .Ltmp2:\r
+; CHECK: .xword .Lxray_sled_0-.Ltmp2\r
+; CHECK: .Ltmp3:\r
+; CHECK-NEXT: .xword .Lxray_sled_1-.Ltmp3\r
; CHECK-LABEL: Lxray_sleds_end0:\r
; CHECK-LABEL: xray_fn_idx\r
; CHECK: .xword .Lxray_sleds_start0\r
; CHECK-NEXT: nop\r
; CHECK-NEXT: nop\r
; CHECK-NEXT: nop\r
-; CHECK-LABEL: .Ltmp2:\r
+; CHECK-LABEL: .Ltmp4:\r
; CHECK: .p2align 2\r
; CHECK-LABEL: Lxray_sled_3:\r
; CHECK-NEXT: b #32\r
; CHECK-NEXT: nop\r
; CHECK-NEXT: nop\r
; CHECK-NEXT: nop\r
-; CHECK-LABEL: .Ltmp3:\r
+; CHECK-LABEL: .Ltmp5:\r
%retval = tail call i32 @callee()\r
; CHECK: b callee\r
ret i32 %retval\r
; CHECK-NEXT: nop
; CHECK-NEXT: nop
; CHECK-NEXT: nop
-; CHECK-LABEL: Ltmp2:
+; CHECK-LABEL: Ltmp4:
; CHECK: .p2align 2
; CHECK-LABEL: Lxray_sled_3:
; CHECK-NEXT: b #20
; CHECK-NEXT: nop
; CHECK-NEXT: nop
; CHECK-NEXT: nop
-; CHECK-LABEL: Ltmp3:
+; CHECK-LABEL: Ltmp5:
%retval = tail call i32 @callee()
; CHECK: b {{.*}}callee
ret i32 %retval
; CHECK-NEXT: nop
; CHECK-NEXT: mtlr 0
}
-; CHECK-LABEL: xray_instr_map,"awo",@progbits,foo{{$}}
+; CHECK-LABEL: xray_instr_map,"ao",@progbits,foo{{$}}
; CHECK: .Lxray_sleds_start0:
-; CHECK-NEXT: .quad .Ltmp0
+; CHECK-NEXT: .Ltmp3:
+; CHECK-NEXT: .quad .Ltmp0-.Ltmp3
; CHECK-NEXT: .quad foo
; CHECK-NEXT: .byte 0x00
; CHECK-NEXT: .byte 0x01
-; CHECK-NEXT: .byte 0x00
+; CHECK-NEXT: .byte 0x02
; CHECK-NEXT: .space 13
-; CHECK-NEXT: .quad .Ltmp2
+; CHECK-NEXT: .Ltmp4:
+; CHECK-NEXT: .quad .Ltmp2-.Ltmp4
; CHECK-NEXT: .quad foo
; CHECK-NEXT: .byte 0x01
; CHECK-NEXT: .byte 0x01
-; CHECK-NEXT: .byte 0x00
+; CHECK-NEXT: .byte 0x02
; CHECK-NEXT: .space 13
; CHECK-NEXT: .Lxray_sleds_end0:
; CHECK-LABEL: xray_fn_idx,"awo",@progbits,foo{{$}}
}
define i32 @caller() nounwind noinline uwtable "function-instrument"="xray-always" {
-; CHECK-LABEL: .Ltmp3:
-; CHECK: b .Ltmp4
+; CHECK-LABEL: .Ltmp5:
+; CHECK-NEXT: b .Ltmp6
; CHECK-NEXT: nop
; CHECK-NEXT: std 0, -8(1)
; CHECK-NEXT: mflr 0
; CHECK-NEXT: bl __xray_FunctionEntry
; CHECK-NEXT: nop
; CHECK-NEXT: mtlr 0
-; CHECK-LABEL: .Ltmp4:
+; CHECK-LABEL: .Ltmp6:
%retval = tail call i32 @callee()
ret i32 %retval
-; CHECK-LABEL: .Ltmp5:
+; CHECK-LABEL: .Ltmp7:
; CHECK: blr
; CHECK-NEXT: nop
; CHECK-NEXT: std 0, -8(1)