blob: 50f6b4158d5e64fec7f8dd2e9f906990471c85f8 [file] [log] [blame]
/*
* Copyright (C) 2016 The Android Open Source Project
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "jni_macro_assembler_arm64.h"
#include "entrypoints/quick/quick_entrypoints.h"
#include "indirect_reference_table.h"
#include "jni/jni_env_ext.h"
#include "jni/local_reference_table.h"
#include "lock_word.h"
#include "managed_register_arm64.h"
#include "offsets.h"
#include "thread.h"
using namespace vixl::aarch64; // NOLINT(build/namespaces)
namespace art HIDDEN {
namespace arm64 {
#ifdef ___
#error "ARM64 Assembler macro already defined."
#else
#define ___ asm_.GetVIXLAssembler()->
#endif
#define reg_x(X) Arm64Assembler::reg_x(X)
#define reg_w(W) Arm64Assembler::reg_w(W)
#define reg_d(D) Arm64Assembler::reg_d(D)
#define reg_s(S) Arm64Assembler::reg_s(S)
// The AAPCS64 requires 16-byte alignment. This is the same as the Managed ABI stack alignment.
static constexpr size_t kAapcs64StackAlignment = 16u;
static_assert(kAapcs64StackAlignment == kStackAlignment);
// STP signed offset for W-register can encode any 4-byte aligned offset smaller than this cutoff.
static constexpr size_t kStpWOffsetCutoff = 256u;
// STP signed offset for X-register can encode any 8-byte aligned offset smaller than this cutoff.
static constexpr size_t kStpXOffsetCutoff = 512u;
// STP signed offset for S-register can encode any 4-byte aligned offset smaller than this cutoff.
static constexpr size_t kStpSOffsetCutoff = 256u;
// STP signed offset for D-register can encode any 8-byte aligned offset smaller than this cutoff.
static constexpr size_t kStpDOffsetCutoff = 512u;
Arm64JNIMacroAssembler::~Arm64JNIMacroAssembler() {
}
void Arm64JNIMacroAssembler::FinalizeCode() {
___ FinalizeCode();
}
void Arm64JNIMacroAssembler::GetCurrentThread(ManagedRegister dest) {
___ Mov(reg_x(dest.AsArm64().AsXRegister()), reg_x(TR));
}
void Arm64JNIMacroAssembler::GetCurrentThread(FrameOffset offset) {
StoreToOffset(TR, SP, offset.Int32Value());
}
// See Arm64 PCS Section 5.2.2.1.
void Arm64JNIMacroAssembler::IncreaseFrameSize(size_t adjust) {
if (adjust != 0u) {
CHECK_ALIGNED(adjust, kStackAlignment);
AddConstant(SP, -adjust);
cfi().AdjustCFAOffset(adjust);
}
}
// See Arm64 PCS Section 5.2.2.1.
void Arm64JNIMacroAssembler::DecreaseFrameSize(size_t adjust) {
if (adjust != 0u) {
CHECK_ALIGNED(adjust, kStackAlignment);
AddConstant(SP, adjust);
cfi().AdjustCFAOffset(-adjust);
}
}
ManagedRegister Arm64JNIMacroAssembler::CoreRegisterWithSize(ManagedRegister m_src, size_t size) {
DCHECK(size == 4u || size == 8u) << size;
Arm64ManagedRegister src = m_src.AsArm64();
// Switch between X and W registers using the `XRegister` and `WRegister` enumerations.
static_assert(W0 == static_cast<std::underlying_type_t<XRegister>>(X0));
static_assert(W30 == static_cast<std::underlying_type_t<XRegister>>(X30));
static_assert(WSP == static_cast<std::underlying_type_t<XRegister>>(SP));
static_assert(WZR == static_cast<std::underlying_type_t<XRegister>>(XZR));
if (src.IsXRegister()) {
if (size == 8u) {
return m_src;
}
auto id = static_cast<std::underlying_type_t<XRegister>>(src.AsXRegister());
return Arm64ManagedRegister::FromWRegister(enum_cast<WRegister>(id));
} else {
CHECK(src.IsWRegister());
if (size == 4u) {
return m_src;
}
auto id = static_cast<std::underlying_type_t<WRegister>>(src.AsWRegister());
return Arm64ManagedRegister::FromXRegister(enum_cast<XRegister>(id));
}
}
void Arm64JNIMacroAssembler::AddConstant(XRegister rd, int32_t value, Condition cond) {
AddConstant(rd, rd, value, cond);
}
void Arm64JNIMacroAssembler::AddConstant(XRegister rd,
XRegister rn,
int32_t value,
Condition cond) {
if ((cond == al) || (cond == nv)) {
// VIXL macro-assembler handles all variants.
___ Add(reg_x(rd), reg_x(rn), value);
} else {
// temp = rd + value
// rd = cond ? temp : rn
UseScratchRegisterScope temps(asm_.GetVIXLAssembler());
temps.Exclude(reg_x(rd), reg_x(rn));
Register temp = temps.AcquireX();
___ Add(temp, reg_x(rn), value);
___ Csel(reg_x(rd), temp, reg_x(rd), cond);
}
}
void Arm64JNIMacroAssembler::StoreWToOffset(StoreOperandType type,
WRegister source,
XRegister base,
int32_t offset) {
switch (type) {
case kStoreByte:
___ Strb(reg_w(source), MEM_OP(reg_x(base), offset));
break;
case kStoreHalfword:
___ Strh(reg_w(source), MEM_OP(reg_x(base), offset));
break;
case kStoreWord:
___ Str(reg_w(source), MEM_OP(reg_x(base), offset));
break;
default:
LOG(FATAL) << "UNREACHABLE";
}
}
void Arm64JNIMacroAssembler::StoreToOffset(XRegister source, XRegister base, int32_t offset) {
CHECK_NE(source, SP);
___ Str(reg_x(source), MEM_OP(reg_x(base), offset));
}
void Arm64JNIMacroAssembler::StoreSToOffset(SRegister source, XRegister base, int32_t offset) {
___ Str(reg_s(source), MEM_OP(reg_x(base), offset));
}
void Arm64JNIMacroAssembler::StoreDToOffset(DRegister source, XRegister base, int32_t offset) {
___ Str(reg_d(source), MEM_OP(reg_x(base), offset));
}
void Arm64JNIMacroAssembler::Store(FrameOffset offs, ManagedRegister m_src, size_t size) {
Store(Arm64ManagedRegister::FromXRegister(SP), MemberOffset(offs.Int32Value()), m_src, size);
}
void Arm64JNIMacroAssembler::Store(ManagedRegister m_base,
MemberOffset offs,
ManagedRegister m_src,
size_t size) {
Arm64ManagedRegister base = m_base.AsArm64();
Arm64ManagedRegister src = m_src.AsArm64();
if (src.IsNoRegister()) {
CHECK_EQ(0u, size);
} else if (src.IsWRegister()) {
CHECK_EQ(4u, size);
StoreWToOffset(kStoreWord, src.AsWRegister(), base.AsXRegister(), offs.Int32Value());
} else if (src.IsXRegister()) {
CHECK_EQ(8u, size);
StoreToOffset(src.AsXRegister(), base.AsXRegister(), offs.Int32Value());
} else if (src.IsSRegister()) {
StoreSToOffset(src.AsSRegister(), base.AsXRegister(), offs.Int32Value());
} else {
CHECK(src.IsDRegister()) << src;
StoreDToOffset(src.AsDRegister(), base.AsXRegister(), offs.Int32Value());
}
}
void Arm64JNIMacroAssembler::StoreRawPtr(FrameOffset offs, ManagedRegister m_src) {
Arm64ManagedRegister src = m_src.AsArm64();
CHECK(src.IsXRegister()) << src;
StoreToOffset(src.AsXRegister(), SP, offs.Int32Value());
}
void Arm64JNIMacroAssembler::StoreStackPointerToThread(ThreadOffset64 tr_offs, bool tag_sp) {
UseScratchRegisterScope temps(asm_.GetVIXLAssembler());
Register scratch = temps.AcquireX();
___ Mov(scratch, reg_x(SP));
if (tag_sp) {
___ Orr(scratch, scratch, 0x2);
}
___ Str(scratch, MEM_OP(reg_x(TR), tr_offs.Int32Value()));
}
// Load routines.
void Arm64JNIMacroAssembler::LoadImmediate(XRegister dest, int32_t value, Condition cond) {
if ((cond == al) || (cond == nv)) {
___ Mov(reg_x(dest), value);
} else {
// temp = value
// rd = cond ? temp : rd
if (value != 0) {
UseScratchRegisterScope temps(asm_.GetVIXLAssembler());
temps.Exclude(reg_x(dest));
Register temp = temps.AcquireX();
___ Mov(temp, value);
___ Csel(reg_x(dest), temp, reg_x(dest), cond);
} else {
___ Csel(reg_x(dest), reg_x(XZR), reg_x(dest), cond);
}
}
}
void Arm64JNIMacroAssembler::LoadWFromOffset(LoadOperandType type,
WRegister dest,
XRegister base,
int32_t offset) {
switch (type) {
case kLoadSignedByte:
___ Ldrsb(reg_w(dest), MEM_OP(reg_x(base), offset));
break;
case kLoadSignedHalfword:
___ Ldrsh(reg_w(dest), MEM_OP(reg_x(base), offset));
break;
case kLoadUnsignedByte:
___ Ldrb(reg_w(dest), MEM_OP(reg_x(base), offset));
break;
case kLoadUnsignedHalfword:
___ Ldrh(reg_w(dest), MEM_OP(reg_x(base), offset));
break;
case kLoadWord:
___ Ldr(reg_w(dest), MEM_OP(reg_x(base), offset));
break;
default:
LOG(FATAL) << "UNREACHABLE";
}
}
// Note: We can extend this member by adding load type info - see
// sign extended A64 load variants.
void Arm64JNIMacroAssembler::LoadFromOffset(XRegister dest, XRegister base, int32_t offset) {
CHECK_NE(dest, SP);
___ Ldr(reg_x(dest), MEM_OP(reg_x(base), offset));
}
void Arm64JNIMacroAssembler::LoadSFromOffset(SRegister dest, XRegister base, int32_t offset) {
___ Ldr(reg_s(dest), MEM_OP(reg_x(base), offset));
}
void Arm64JNIMacroAssembler::LoadDFromOffset(DRegister dest, XRegister base, int32_t offset) {
___ Ldr(reg_d(dest), MEM_OP(reg_x(base), offset));
}
void Arm64JNIMacroAssembler::Load(Arm64ManagedRegister dest,
XRegister base,
int32_t offset,
size_t size) {
if (dest.IsNoRegister()) {
CHECK_EQ(0u, size) << dest;
} else if (dest.IsWRegister()) {
CHECK_EQ(4u, size) << dest;
___ Ldr(reg_w(dest.AsWRegister()), MEM_OP(reg_x(base), offset));
} else if (dest.IsXRegister()) {
CHECK_NE(dest.AsXRegister(), SP) << dest;
if (size == 1u) {
___ Ldrb(reg_w(dest.AsOverlappingWRegister()), MEM_OP(reg_x(base), offset));
} else if (size == 4u) {
___ Ldr(reg_w(dest.AsOverlappingWRegister()), MEM_OP(reg_x(base), offset));
} else {
CHECK_EQ(8u, size) << dest;
___ Ldr(reg_x(dest.AsXRegister()), MEM_OP(reg_x(base), offset));
}
} else if (dest.IsSRegister()) {
___ Ldr(reg_s(dest.AsSRegister()), MEM_OP(reg_x(base), offset));
} else {
CHECK(dest.IsDRegister()) << dest;
___ Ldr(reg_d(dest.AsDRegister()), MEM_OP(reg_x(base), offset));
}
}
void Arm64JNIMacroAssembler::Load(ManagedRegister m_dst, FrameOffset src, size_t size) {
return Load(m_dst.AsArm64(), SP, src.Int32Value(), size);
}
void Arm64JNIMacroAssembler::Load(ManagedRegister m_dst,
ManagedRegister m_base,
MemberOffset offs,
size_t size) {
return Load(m_dst.AsArm64(), m_base.AsArm64().AsXRegister(), offs.Int32Value(), size);
}
void Arm64JNIMacroAssembler::LoadRawPtrFromThread(ManagedRegister m_dst, ThreadOffset64 offs) {
Arm64ManagedRegister dst = m_dst.AsArm64();
CHECK(dst.IsXRegister()) << dst;
LoadFromOffset(dst.AsXRegister(), TR, offs.Int32Value());
}
// Copying routines.
void Arm64JNIMacroAssembler::MoveArguments(ArrayRef<ArgumentLocation> dests,
ArrayRef<ArgumentLocation> srcs,
ArrayRef<FrameOffset> refs) {
size_t arg_count = dests.size();
DCHECK_EQ(arg_count, srcs.size());
DCHECK_EQ(arg_count, refs.size());
auto get_mask = [](ManagedRegister reg) -> uint64_t {
Arm64ManagedRegister arm64_reg = reg.AsArm64();
if (arm64_reg.IsXRegister()) {
size_t core_reg_number = static_cast<size_t>(arm64_reg.AsXRegister());
DCHECK_LT(core_reg_number, 31u); // xSP, xZR not allowed.
return UINT64_C(1) << core_reg_number;
} else if (arm64_reg.IsWRegister()) {
size_t core_reg_number = static_cast<size_t>(arm64_reg.AsWRegister());
DCHECK_LT(core_reg_number, 31u); // wSP, wZR not allowed.
return UINT64_C(1) << core_reg_number;
} else if (arm64_reg.IsDRegister()) {
size_t fp_reg_number = static_cast<size_t>(arm64_reg.AsDRegister());
DCHECK_LT(fp_reg_number, 32u);
return (UINT64_C(1) << 32u) << fp_reg_number;
} else {
DCHECK(arm64_reg.IsSRegister());
size_t fp_reg_number = static_cast<size_t>(arm64_reg.AsSRegister());
DCHECK_LT(fp_reg_number, 32u);
return (UINT64_C(1) << 32u) << fp_reg_number;
}
};
// More than 8 core or FP reg args are very rare, so we do not optimize for
// that case by using LDP/STP, except for situations that arise even with low
// number of arguments. We use STP for the non-reference spilling which also
// covers the initial spill for native reference register args as they are
// spilled as raw 32-bit values. We also optimize loading args to registers
// with LDP, whether references or not, except for the initial non-null
// reference which we do not need to load at all.
// Collect registers to move while storing/copying args to stack slots.
// Convert processed references to `jobject`.
uint64_t src_regs = 0u;
uint64_t dest_regs = 0u;
for (size_t i = 0; i != arg_count; ++i) {
const ArgumentLocation& src = srcs[i];
const ArgumentLocation& dest = dests[i];
const FrameOffset ref = refs[i];
if (ref != kInvalidReferenceOffset) {
DCHECK_EQ(src.GetSize(), kObjectReferenceSize);
DCHECK_EQ(dest.GetSize(), static_cast<size_t>(kArm64PointerSize));
} else {
DCHECK_EQ(src.GetSize(), dest.GetSize());
}
if (dest.IsRegister()) {
// Note: For references, `Equals()` returns `false` for overlapping W and X registers.
if (ref != kInvalidReferenceOffset &&
src.IsRegister() &&
src.GetRegister().AsArm64().AsOverlappingXRegister() ==
dest.GetRegister().AsArm64().AsXRegister()) {
// Just convert to `jobject`. No further processing is needed.
CreateJObject(dest.GetRegister(), ref, src.GetRegister(), /*null_allowed=*/ i != 0u);
} else if (src.IsRegister() && src.GetRegister().Equals(dest.GetRegister())) {
// Nothing to do.
} else {
if (src.IsRegister()) {
src_regs |= get_mask(src.GetRegister());
}
dest_regs |= get_mask(dest.GetRegister());
}
} else if (ref != kInvalidReferenceOffset) {
if (src.IsRegister()) {
// Note: We can clobber `src` here as the register cannot hold more than one argument.
ManagedRegister src_x =
CoreRegisterWithSize(src.GetRegister(), static_cast<size_t>(kArm64PointerSize));
CreateJObject(src_x, ref, src.GetRegister(), /*null_allowed=*/ i != 0u);
Store(dest.GetFrameOffset(), src_x, dest.GetSize());
} else {
CreateJObject(dest.GetFrameOffset(), ref, /*null_allowed=*/ i != 0u);
}
} else {
if (src.IsRegister()) {
static_assert(kStpWOffsetCutoff == kStpSOffsetCutoff);
static_assert(kStpXOffsetCutoff == kStpDOffsetCutoff);
if (i + 1u != arg_count &&
srcs[i + 1u].IsRegister() &&
srcs[i + 1u].GetSize() == dest.GetSize() &&
src.GetRegister().AsArm64().IsGPRegister() ==
srcs[i + 1u].GetRegister().AsArm64().IsGPRegister() &&
refs[i + 1u] == kInvalidReferenceOffset &&
!dests[i + 1u].IsRegister() &&
dests[i + 1u].GetFrameOffset().SizeValue() ==
dest.GetFrameOffset().SizeValue() + dest.GetSize() &&
IsAlignedParam(dest.GetFrameOffset().SizeValue(), dest.GetSize()) &&
dest.GetFrameOffset().SizeValue() <
(dest.GetSize() == 8u ? kStpXOffsetCutoff : kStpWOffsetCutoff)) {
DCHECK_EQ(dests[i + 1u].GetSize(), dest.GetSize());
Arm64ManagedRegister src_reg = src.GetRegister().AsArm64();
Arm64ManagedRegister src2_reg = srcs[i + 1u].GetRegister().AsArm64();
DCHECK_EQ(dest.GetSize() == 8u, src_reg.IsXRegister() || src_reg.IsDRegister());
DCHECK_EQ(dest.GetSize() == 8u, src2_reg.IsXRegister() || src2_reg.IsDRegister());
if (src_reg.IsWRegister()) {
___ Stp(reg_w(src_reg.AsWRegister()),
reg_w(src2_reg.AsWRegister()),
MEM_OP(sp, dest.GetFrameOffset().SizeValue()));
} else if (src_reg.IsXRegister()) {
___ Stp(reg_x(src_reg.AsXRegister()),
reg_x(src2_reg.AsXRegister()),
MEM_OP(sp, dest.GetFrameOffset().SizeValue()));
} else if (src_reg.IsSRegister()) {
___ Stp(reg_s(src_reg.AsSRegister()),
reg_s(src2_reg.AsSRegister()),
MEM_OP(sp, dest.GetFrameOffset().SizeValue()));
} else {
DCHECK(src_reg.IsDRegister());
___ Stp(reg_d(src_reg.AsDRegister()),
reg_d(src2_reg.AsDRegister()),
MEM_OP(sp, dest.GetFrameOffset().SizeValue()));
}
++i;
} else {
Store(dest.GetFrameOffset(), src.GetRegister(), dest.GetSize());
}
} else {
Copy(dest.GetFrameOffset(), src.GetFrameOffset(), dest.GetSize());
}
}
}
// Fill destination registers.
// There should be no cycles, so this simple algorithm should make progress.
while (dest_regs != 0u) {
uint64_t old_dest_regs = dest_regs;
for (size_t i = 0; i != arg_count; ++i) {
const ArgumentLocation& src = srcs[i];
const ArgumentLocation& dest = dests[i];
const FrameOffset ref = refs[i];
if (!dest.IsRegister()) {
continue; // Stored in first loop above.
}
auto can_process = [&](ManagedRegister dest_reg) {
uint64_t dest_reg_mask = get_mask(dest_reg);
if ((dest_reg_mask & dest_regs) == 0u) {
return false; // Equals source, or already filled in one of previous iterations.
}
if ((dest_reg_mask & src_regs) != 0u) {
return false; // Cannot clobber this register yet.
}
return true;
};
if (!can_process(dest.GetRegister())) {
continue;
}
if (src.IsRegister()) {
if (ref != kInvalidReferenceOffset) {
CreateJObject(dest.GetRegister(), ref, src.GetRegister(), /*null_allowed=*/ i != 0u);
} else {
Move(dest.GetRegister(), src.GetRegister(), dest.GetSize());
}
src_regs &= ~get_mask(src.GetRegister()); // Allow clobbering source register.
} else if (i + 1u != arg_count &&
(i != 0u || ref == kInvalidReferenceOffset) && // Not for non-null reference.
dests[i + 1u].IsRegister() &&
dest.GetRegister().AsArm64().IsGPRegister() ==
dests[i + 1u].GetRegister().AsArm64().IsGPRegister() &&
!srcs[i + 1u].IsRegister() &&
srcs[i + 1u].GetSize() == src.GetSize() &&
srcs[i + 1u].GetFrameOffset().SizeValue() ==
src.GetFrameOffset().SizeValue() + src.GetSize() &&
IsAlignedParam(src.GetFrameOffset().SizeValue(), src.GetSize()) &&
can_process(dests[i + 1u].GetRegister())) {
Arm64ManagedRegister dest_reg = dest.GetRegister().AsArm64();
Arm64ManagedRegister dest2_reg = dests[i + 1u].GetRegister().AsArm64();
DCHECK(ref == kInvalidReferenceOffset || dest_reg.IsXRegister());
DCHECK(refs[i + 1u] == kInvalidReferenceOffset || dest2_reg.IsXRegister());
if (dest_reg.IsDRegister()) {
DCHECK_EQ(dest.GetSize(), 8u);
DCHECK_EQ(dests[i + 1u].GetSize(), 8u);
___ Ldp(reg_d(dest_reg.AsDRegister()),
reg_d(dest2_reg.AsDRegister()),
MEM_OP(sp, src.GetFrameOffset().SizeValue()));
} else if (dest_reg.IsSRegister()) {
DCHECK_EQ(dest.GetSize(), 4u);
DCHECK_EQ(dests[i + 1u].GetSize(), 4u);
___ Ldp(reg_s(dest_reg.AsSRegister()),
reg_s(dest2_reg.AsSRegister()),
MEM_OP(sp, src.GetFrameOffset().SizeValue()));
} else if (src.GetSize() == 8u) {
DCHECK_EQ(dest.GetSize(), 8u);
DCHECK_EQ(dests[i + 1u].GetSize(), 8u);
___ Ldp(reg_x(dest_reg.AsXRegister()),
reg_x(dest2_reg.AsXRegister()),
MEM_OP(sp, src.GetFrameOffset().SizeValue()));
} else {
DCHECK_EQ(dest.GetSize(), ref != kInvalidReferenceOffset ? 8u : 4u);
DCHECK_EQ(dests[i + 1u].GetSize(), refs[i + 1u] != kInvalidReferenceOffset ? 8u : 4u);
auto to_w = [](Arm64ManagedRegister reg) {
return reg_w(reg.IsXRegister() ? reg.AsOverlappingWRegister() : reg.AsWRegister());
};
___ Ldp(to_w(dest_reg), to_w(dest2_reg), MEM_OP(sp, src.GetFrameOffset().SizeValue()));
auto to_mr_w = [](Arm64ManagedRegister reg) {
return Arm64ManagedRegister::FromWRegister(reg.AsOverlappingWRegister());
};
if (ref != kInvalidReferenceOffset) {
CreateJObject(dest_reg, ref, to_mr_w(dest_reg), /*null_allowed=*/ true);
}
if (refs[i + 1u] != kInvalidReferenceOffset) {
CreateJObject(dest2_reg, refs[i + 1u], to_mr_w(dest2_reg), /*null_allowed=*/ true);
}
}
dest_regs &= ~get_mask(dest2_reg); // Destination register was filled.
++i; // Proceed to mark the other destination register as filled.
} else {
if (ref != kInvalidReferenceOffset) {
CreateJObject(
dest.GetRegister(), ref, ManagedRegister::NoRegister(), /*null_allowed=*/ i != 0u);
} else {
Load(dest.GetRegister(), src.GetFrameOffset(), dest.GetSize());
}
}
dest_regs &= ~get_mask(dest.GetRegister()); // Destination register was filled.
}
CHECK_NE(old_dest_regs, dest_regs);
DCHECK_EQ(0u, dest_regs & ~old_dest_regs);
}
}
void Arm64JNIMacroAssembler::Move(ManagedRegister m_dst, ManagedRegister m_src, size_t size) {
Arm64ManagedRegister dst = m_dst.AsArm64();
if (kIsDebugBuild) {
// Check that the destination is not a scratch register.
UseScratchRegisterScope temps(asm_.GetVIXLAssembler());
if (dst.IsXRegister()) {
CHECK(!temps.IsAvailable(reg_x(dst.AsXRegister())));
} else if (dst.IsWRegister()) {
CHECK(!temps.IsAvailable(reg_w(dst.AsWRegister())));
} else if (dst.IsSRegister()) {
CHECK(!temps.IsAvailable(reg_s(dst.AsSRegister())));
} else {
CHECK(!temps.IsAvailable(reg_d(dst.AsDRegister())));
}
}
Arm64ManagedRegister src = m_src.AsArm64();
if (!dst.Equals(src)) {
if (dst.IsXRegister()) {
if (size == 4) {
CHECK(src.IsWRegister());
___ Mov(reg_w(dst.AsOverlappingWRegister()), reg_w(src.AsWRegister()));
} else {
if (src.IsXRegister()) {
___ Mov(reg_x(dst.AsXRegister()), reg_x(src.AsXRegister()));
} else {
___ Mov(reg_x(dst.AsXRegister()), reg_x(src.AsOverlappingXRegister()));
}
}
} else if (dst.IsWRegister()) {
CHECK(src.IsWRegister()) << src;
___ Mov(reg_w(dst.AsWRegister()), reg_w(src.AsWRegister()));
} else if (dst.IsSRegister()) {
CHECK(src.IsSRegister()) << src;
___ Fmov(reg_s(dst.AsSRegister()), reg_s(src.AsSRegister()));
} else {
CHECK(dst.IsDRegister()) << dst;
CHECK(src.IsDRegister()) << src;
___ Fmov(reg_d(dst.AsDRegister()), reg_d(src.AsDRegister()));
}
}
}
void Arm64JNIMacroAssembler::Move(ManagedRegister m_dst, size_t value) {
Arm64ManagedRegister dst = m_dst.AsArm64();
DCHECK(dst.IsXRegister());
___ Mov(reg_x(dst.AsXRegister()), value);
}
void Arm64JNIMacroAssembler::Copy(FrameOffset dest, FrameOffset src, size_t size) {
DCHECK(size == 4 || size == 8) << size;
UseScratchRegisterScope temps(asm_.GetVIXLAssembler());
Register scratch = (size == 8) ? temps.AcquireX() : temps.AcquireW();
___ Ldr(scratch, MEM_OP(reg_x(SP), src.Int32Value()));
___ Str(scratch, MEM_OP(reg_x(SP), dest.Int32Value()));
}
void Arm64JNIMacroAssembler::SignExtend(ManagedRegister mreg, size_t size) {
Arm64ManagedRegister reg = mreg.AsArm64();
CHECK(size == 1 || size == 2) << size;
CHECK(reg.IsWRegister()) << reg;
if (size == 1) {
___ Sxtb(reg_w(reg.AsWRegister()), reg_w(reg.AsWRegister()));
} else {
___ Sxth(reg_w(reg.AsWRegister()), reg_w(reg.AsWRegister()));
}
}
void Arm64JNIMacroAssembler::ZeroExtend(ManagedRegister mreg, size_t size) {
Arm64ManagedRegister reg = mreg.AsArm64();
CHECK(size == 1 || size == 2) << size;
CHECK(reg.IsWRegister()) << reg;
if (size == 1) {
___ Uxtb(reg_w(reg.AsWRegister()), reg_w(reg.AsWRegister()));
} else {
___ Uxth(reg_w(reg.AsWRegister()), reg_w(reg.AsWRegister()));
}
}
void Arm64JNIMacroAssembler::VerifyObject(ManagedRegister /*src*/, bool /*could_be_null*/) {
// TODO: not validating references.
}
void Arm64JNIMacroAssembler::VerifyObject(FrameOffset /*src*/, bool /*could_be_null*/) {
// TODO: not validating references.
}
void Arm64JNIMacroAssembler::Jump(ManagedRegister m_base, Offset offs) {
Arm64ManagedRegister base = m_base.AsArm64();
CHECK(base.IsXRegister()) << base;
UseScratchRegisterScope temps(asm_.GetVIXLAssembler());
Register scratch = temps.AcquireX();
___ Ldr(scratch, MEM_OP(reg_x(base.AsXRegister()), offs.Int32Value()));
___ Br(scratch);
}
void Arm64JNIMacroAssembler::Call(ManagedRegister m_base, Offset offs) {
Arm64ManagedRegister base = m_base.AsArm64();
CHECK(base.IsXRegister()) << base;
___ Ldr(lr, MEM_OP(reg_x(base.AsXRegister()), offs.Int32Value()));
___ Blr(lr);
}
void Arm64JNIMacroAssembler::CallFromThread(ThreadOffset64 offset) {
// Call *(TR + offset)
___ Ldr(lr, MEM_OP(reg_x(TR), offset.Int32Value()));
___ Blr(lr);
}
void Arm64JNIMacroAssembler::CreateJObject(ManagedRegister m_out_reg,
FrameOffset spilled_reference_offset,
ManagedRegister m_in_reg,
bool null_allowed) {
Arm64ManagedRegister out_reg = m_out_reg.AsArm64();
Arm64ManagedRegister in_reg = m_in_reg.AsArm64();
CHECK(in_reg.IsNoRegister() || in_reg.IsWRegister()) << in_reg;
CHECK(out_reg.IsXRegister()) << out_reg;
if (null_allowed) {
UseScratchRegisterScope temps(asm_.GetVIXLAssembler());
Register scratch = temps.AcquireX();
// Null values get a jobject value null. Otherwise, the jobject is
// the address of the spilled reference.
// e.g. out_reg = (in == 0) ? 0 : (SP+spilled_reference_offset)
if (in_reg.IsNoRegister()) {
in_reg = Arm64ManagedRegister::FromWRegister(out_reg.AsOverlappingWRegister());
LoadWFromOffset(kLoadWord, in_reg.AsWRegister(), SP, spilled_reference_offset.Int32Value());
}
___ Add(scratch, reg_x(SP), spilled_reference_offset.Int32Value());
___ Cmp(reg_w(in_reg.AsWRegister()), 0);
___ Csel(reg_x(out_reg.AsXRegister()), scratch, xzr, ne);
} else {
AddConstant(out_reg.AsXRegister(), SP, spilled_reference_offset.Int32Value(), al);
}
}
void Arm64JNIMacroAssembler::CreateJObject(FrameOffset out_off,
FrameOffset spilled_reference_offset,
bool null_allowed) {
UseScratchRegisterScope temps(asm_.GetVIXLAssembler());
Register scratch = temps.AcquireX();
if (null_allowed) {
Register scratch2 = temps.AcquireW();
___ Ldr(scratch2, MEM_OP(reg_x(SP), spilled_reference_offset.Int32Value()));
___ Add(scratch, reg_x(SP), spilled_reference_offset.Int32Value());
// Null values get a jobject value null. Otherwise, the jobject is
// the address of the spilled reference.
// e.g. scratch = (scratch == 0) ? 0 : (SP+spilled_reference_offset)
___ Cmp(scratch2, 0);
___ Csel(scratch, scratch, xzr, ne);
} else {
___ Add(scratch, reg_x(SP), spilled_reference_offset.Int32Value());
}
___ Str(scratch, MEM_OP(reg_x(SP), out_off.Int32Value()));
}
void Arm64JNIMacroAssembler::DecodeJNITransitionOrLocalJObject(ManagedRegister m_reg,
JNIMacroLabel* slow_path,
JNIMacroLabel* resume) {
constexpr uint64_t kGlobalOrWeakGlobalMask = IndirectReferenceTable::GetGlobalOrWeakGlobalMask();
constexpr uint64_t kIndirectRefKindMask = IndirectReferenceTable::GetIndirectRefKindMask();
constexpr size_t kGlobalOrWeakGlobalBit = WhichPowerOf2(kGlobalOrWeakGlobalMask);
Register reg = reg_w(m_reg.AsArm64().AsWRegister());
___ Tbnz(reg.X(), kGlobalOrWeakGlobalBit, Arm64JNIMacroLabel::Cast(slow_path)->AsArm64());
___ And(reg.X(), reg.X(), ~kIndirectRefKindMask);
___ Cbz(reg.X(), Arm64JNIMacroLabel::Cast(resume)->AsArm64()); // Skip load for null.
___ Ldr(reg, MEM_OP(reg.X()));
}
void Arm64JNIMacroAssembler::TryToTransitionFromRunnableToNative(
JNIMacroLabel* label, [[maybe_unused]] ArrayRef<const ManagedRegister> scratch_regs) {
constexpr uint32_t kNativeStateValue = Thread::StoredThreadStateValue(ThreadState::kNative);
constexpr uint32_t kRunnableStateValue = Thread::StoredThreadStateValue(ThreadState::kRunnable);
constexpr ThreadOffset64 thread_flags_offset = Thread::ThreadFlagsOffset<kArm64PointerSize>();
constexpr ThreadOffset64 thread_held_mutex_mutator_lock_offset =
Thread::HeldMutexOffset<kArm64PointerSize>(kMutatorLock);
UseScratchRegisterScope temps(asm_.GetVIXLAssembler());
Register scratch = temps.AcquireW();
Register scratch2 = temps.AcquireW();
// CAS release, old_value = kRunnableStateValue, new_value = kNativeStateValue, no flags.
vixl::aarch64::Label retry;
___ Bind(&retry);
static_assert(thread_flags_offset.Int32Value() == 0); // LDXR/STLXR require exact address.
___ Ldxr(scratch, MEM_OP(reg_x(TR)));
___ Mov(scratch2, kNativeStateValue);
// If any flags are set, go to the slow path.
static_assert(kRunnableStateValue == 0u);
___ Cbnz(scratch, Arm64JNIMacroLabel::Cast(label)->AsArm64());
___ Stlxr(scratch, scratch2, MEM_OP(reg_x(TR)));
___ Cbnz(scratch, &retry);
// Clear `self->tlsPtr_.held_mutexes[kMutatorLock]`.
___ Str(xzr, MEM_OP(reg_x(TR), thread_held_mutex_mutator_lock_offset.Int32Value()));
}
void Arm64JNIMacroAssembler::TryToTransitionFromNativeToRunnable(
JNIMacroLabel* label,
[[maybe_unused]] ArrayRef<const ManagedRegister> scratch_regs,
[[maybe_unused]] ManagedRegister return_reg) {
constexpr uint32_t kNativeStateValue = Thread::StoredThreadStateValue(ThreadState::kNative);
constexpr uint32_t kRunnableStateValue = Thread::StoredThreadStateValue(ThreadState::kRunnable);
constexpr ThreadOffset64 thread_flags_offset = Thread::ThreadFlagsOffset<kArm64PointerSize>();
constexpr ThreadOffset64 thread_held_mutex_mutator_lock_offset =
Thread::HeldMutexOffset<kArm64PointerSize>(kMutatorLock);
constexpr ThreadOffset64 thread_mutator_lock_offset =
Thread::MutatorLockOffset<kArm64PointerSize>();
UseScratchRegisterScope temps(asm_.GetVIXLAssembler());
Register scratch = temps.AcquireW();
Register scratch2 = temps.AcquireW();
// CAS acquire, old_value = kNativeStateValue, new_value = kRunnableStateValue, no flags.
vixl::aarch64::Label retry;
___ Bind(&retry);
static_assert(thread_flags_offset.Int32Value() == 0); // LDAXR/STXR require exact address.
___ Ldaxr(scratch, MEM_OP(reg_x(TR)));
___ Mov(scratch2, kNativeStateValue);
// If any flags are set, or the state is not Native, go to the slow path.
// (While the thread can theoretically transition between different Suspended states,
// it would be very unexpected to see a state other than Native at this point.)
___ Cmp(scratch, scratch2);
___ B(ne, Arm64JNIMacroLabel::Cast(label)->AsArm64());
static_assert(kRunnableStateValue == 0u);
___ Stxr(scratch, wzr, MEM_OP(reg_x(TR)));
___ Cbnz(scratch, &retry);
// Set `self->tlsPtr_.held_mutexes[kMutatorLock]` to the mutator lock.
___ Ldr(scratch.X(), MEM_OP(reg_x(TR), thread_mutator_lock_offset.Int32Value()));
___ Str(scratch.X(), MEM_OP(reg_x(TR), thread_held_mutex_mutator_lock_offset.Int32Value()));
}
void Arm64JNIMacroAssembler::SuspendCheck(JNIMacroLabel* label) {
UseScratchRegisterScope temps(asm_.GetVIXLAssembler());
Register scratch = temps.AcquireW();
___ Ldr(scratch, MEM_OP(reg_x(TR), Thread::ThreadFlagsOffset<kArm64PointerSize>().Int32Value()));
___ Tst(scratch, Thread::SuspendOrCheckpointRequestFlags());
___ B(ne, Arm64JNIMacroLabel::Cast(label)->AsArm64());
}
void Arm64JNIMacroAssembler::ExceptionPoll(JNIMacroLabel* label) {
UseScratchRegisterScope temps(asm_.GetVIXLAssembler());
Register scratch = temps.AcquireX();
___ Ldr(scratch, MEM_OP(reg_x(TR), Thread::ExceptionOffset<kArm64PointerSize>().Int32Value()));
___ Cbnz(scratch, Arm64JNIMacroLabel::Cast(label)->AsArm64());
}
void Arm64JNIMacroAssembler::DeliverPendingException() {
// Pass exception object as argument.
// Don't care about preserving X0 as this won't return.
// Note: The scratch register from `ExceptionPoll()` may have been clobbered.
___ Ldr(reg_x(X0), MEM_OP(reg_x(TR), Thread::ExceptionOffset<kArm64PointerSize>().Int32Value()));
___ Ldr(lr,
MEM_OP(reg_x(TR),
QUICK_ENTRYPOINT_OFFSET(kArm64PointerSize, pDeliverException).Int32Value()));
___ Blr(lr);
// Call should never return.
___ Brk();
}
std::unique_ptr<JNIMacroLabel> Arm64JNIMacroAssembler::CreateLabel() {
return std::unique_ptr<JNIMacroLabel>(new (asm_.GetAllocator()) Arm64JNIMacroLabel());
}
void Arm64JNIMacroAssembler::Jump(JNIMacroLabel* label) {
CHECK(label != nullptr);
___ B(Arm64JNIMacroLabel::Cast(label)->AsArm64());
}
void Arm64JNIMacroAssembler::TestGcMarking(JNIMacroLabel* label, JNIMacroUnaryCondition cond) {
CHECK(label != nullptr);
UseScratchRegisterScope temps(asm_.GetVIXLAssembler());
Register test_reg;
DCHECK_EQ(Thread::IsGcMarkingSize(), 4u);
if (kUseBakerReadBarrier) {
// TestGcMarking() is used in the JNI stub entry when the marking register is up to date.
if (kIsDebugBuild && emit_run_time_checks_in_debug_mode_) {
Register temp = temps.AcquireW();
asm_.GenerateMarkingRegisterCheck(temp);
}
test_reg = reg_w(MR);
} else {
test_reg = temps.AcquireW();
int32_t is_gc_marking_offset = Thread::IsGcMarkingOffset<kArm64PointerSize>().Int32Value();
___ Ldr(test_reg, MEM_OP(reg_x(TR), is_gc_marking_offset));
}
switch (cond) {
case JNIMacroUnaryCondition::kZero:
___ Cbz(test_reg, Arm64JNIMacroLabel::Cast(label)->AsArm64());
break;
case JNIMacroUnaryCondition::kNotZero:
___ Cbnz(test_reg, Arm64JNIMacroLabel::Cast(label)->AsArm64());
break;
default:
LOG(FATAL) << "Not implemented unary condition: " << static_cast<int>(cond);
UNREACHABLE();
}
}
void Arm64JNIMacroAssembler::TestMarkBit(ManagedRegister m_ref,
JNIMacroLabel* label,
JNIMacroUnaryCondition cond) {
DCHECK(kUseBakerReadBarrier);
Register ref = reg_x(m_ref.AsArm64().AsOverlappingXRegister());
UseScratchRegisterScope temps(asm_.GetVIXLAssembler());
Register scratch = temps.AcquireW();
___ Ldr(scratch, MEM_OP(ref, mirror::Object::MonitorOffset().SizeValue()));
static_assert(LockWord::kMarkBitStateSize == 1u);
switch (cond) {
case JNIMacroUnaryCondition::kZero:
___ Tbz(scratch, LockWord::kMarkBitStateShift, Arm64JNIMacroLabel::Cast(label)->AsArm64());
break;
case JNIMacroUnaryCondition::kNotZero:
___ Tbnz(scratch, LockWord::kMarkBitStateShift, Arm64JNIMacroLabel::Cast(label)->AsArm64());
break;
default:
LOG(FATAL) << "Not implemented unary condition: " << static_cast<int>(cond);
UNREACHABLE();
}
}
void Arm64JNIMacroAssembler::TestByteAndJumpIfNotZero(uintptr_t address, JNIMacroLabel* label) {
UseScratchRegisterScope temps(asm_.GetVIXLAssembler());
Register scratch = temps.AcquireX();
___ Mov(scratch, address);
___ Ldrb(scratch.W(), MEM_OP(scratch, 0));
___ Cbnz(scratch.W(), Arm64JNIMacroLabel::Cast(label)->AsArm64());
}
void Arm64JNIMacroAssembler::Bind(JNIMacroLabel* label) {
CHECK(label != nullptr);
___ Bind(Arm64JNIMacroLabel::Cast(label)->AsArm64());
}
void Arm64JNIMacroAssembler::BuildFrame(size_t frame_size,
ManagedRegister method_reg,
ArrayRef<const ManagedRegister> callee_save_regs) {
// Setup VIXL CPURegList for callee-saves.
CPURegList core_reg_list(CPURegister::kRegister, kXRegSize, 0);
CPURegList fp_reg_list(CPURegister::kVRegister, kDRegSize, 0);
for (auto r : callee_save_regs) {
Arm64ManagedRegister reg = r.AsArm64();
if (reg.IsXRegister()) {
core_reg_list.Combine(reg_x(reg.AsXRegister()).GetCode());
} else {
DCHECK(reg.IsDRegister());
fp_reg_list.Combine(reg_d(reg.AsDRegister()).GetCode());
}
}
size_t core_reg_size = core_reg_list.GetTotalSizeInBytes();
size_t fp_reg_size = fp_reg_list.GetTotalSizeInBytes();
// Increase frame to required size.
DCHECK_ALIGNED(frame_size, kStackAlignment);
// Must at least have space for Method* if we're going to spill it.
DCHECK_GE(frame_size,
core_reg_size + fp_reg_size + (method_reg.IsRegister() ? kXRegSizeInBytes : 0u));
IncreaseFrameSize(frame_size);
// Save callee-saves.
asm_.SpillRegisters(core_reg_list, frame_size - core_reg_size);
asm_.SpillRegisters(fp_reg_list, frame_size - core_reg_size - fp_reg_size);
if (method_reg.IsRegister()) {
// Write ArtMethod*
DCHECK(X0 == method_reg.AsArm64().AsXRegister());
StoreToOffset(X0, SP, 0);
}
}
void Arm64JNIMacroAssembler::RemoveFrame(size_t frame_size,
ArrayRef<const ManagedRegister> callee_save_regs,
bool may_suspend) {
// Setup VIXL CPURegList for callee-saves.
CPURegList core_reg_list(CPURegister::kRegister, kXRegSize, 0);
CPURegList fp_reg_list(CPURegister::kVRegister, kDRegSize, 0);
for (auto r : callee_save_regs) {
Arm64ManagedRegister reg = r.AsArm64();
if (reg.IsXRegister()) {
core_reg_list.Combine(reg_x(reg.AsXRegister()).GetCode());
} else {
DCHECK(reg.IsDRegister());
fp_reg_list.Combine(reg_d(reg.AsDRegister()).GetCode());
}
}
size_t core_reg_size = core_reg_list.GetTotalSizeInBytes();
size_t fp_reg_size = fp_reg_list.GetTotalSizeInBytes();
// For now we only check that the size of the frame is large enough to hold spills and method
// reference.
DCHECK_GE(frame_size, core_reg_size + fp_reg_size);
DCHECK_ALIGNED(frame_size, kAapcs64StackAlignment);
cfi().RememberState();
// Restore callee-saves.
asm_.UnspillRegisters(core_reg_list, frame_size - core_reg_size);
asm_.UnspillRegisters(fp_reg_list, frame_size - core_reg_size - fp_reg_size);
// Emit marking register refresh even with all GCs as we are still using the
// register due to nterp's dependency.
if (kReserveMarkingRegister) {
vixl::aarch64::Register mr = reg_x(MR); // Marking Register.
vixl::aarch64::Register tr = reg_x(TR); // Thread Register.
if (may_suspend) {
// The method may be suspended; refresh the Marking Register.
___ Ldr(mr.W(), MemOperand(tr, Thread::IsGcMarkingOffset<kArm64PointerSize>().Int32Value()));
} else {
// The method shall not be suspended; no need to refresh the Marking Register.
// The Marking Register is a callee-save register and thus has been
// preserved by native code following the AAPCS64 calling convention.
// The following condition is a compile-time one, so it does not have a run-time cost.
if (kIsDebugBuild) {
// The following condition is a run-time one; it is executed after the
// previous compile-time test, to avoid penalizing non-debug builds.
if (emit_run_time_checks_in_debug_mode_) {
// Emit a run-time check verifying that the Marking Register is up-to-date.
UseScratchRegisterScope temps(asm_.GetVIXLAssembler());
Register temp = temps.AcquireW();
// Ensure we are not clobbering a callee-save register that was restored before.
DCHECK(!core_reg_list.IncludesAliasOf(temp.X()))
<< "core_reg_list should not contain scratch register X" << temp.GetCode();
asm_.GenerateMarkingRegisterCheck(temp);
}
}
}
}
// Decrease frame size to start of callee saved regs.
DecreaseFrameSize(frame_size);
// Return to LR.
___ Ret();
// The CFI should be restored for any code that follows the exit block.
cfi().RestoreState();
cfi().DefCFAOffset(frame_size);
}
void Arm64JNIMacroAssembler::LoadLocalReferenceTableStates(ManagedRegister jni_env_reg,
ManagedRegister previous_state_reg,
ManagedRegister current_state_reg) {
constexpr size_t kLRTSegmentStateSize = sizeof(jni::LRTSegmentState);
DCHECK_EQ(kLRTSegmentStateSize, kWRegSizeInBytes);
const MemberOffset previous_state_offset = JNIEnvExt::LrtPreviousStateOffset(kArm64PointerSize);
const MemberOffset current_state_offset = JNIEnvExt::LrtSegmentStateOffset(kArm64PointerSize);
DCHECK_EQ(previous_state_offset.SizeValue() + kLRTSegmentStateSize,
current_state_offset.SizeValue());
___ Ldp(
reg_w(previous_state_reg.AsArm64().AsWRegister()),
reg_w(current_state_reg.AsArm64().AsWRegister()),
MemOperand(reg_x(jni_env_reg.AsArm64().AsXRegister()), previous_state_offset.Int32Value()));
}
void Arm64JNIMacroAssembler::StoreLocalReferenceTableStates(ManagedRegister jni_env_reg,
ManagedRegister previous_state_reg,
ManagedRegister current_state_reg) {
constexpr size_t kLRTSegmentStateSize = sizeof(jni::LRTSegmentState);
DCHECK_EQ(kLRTSegmentStateSize, kWRegSizeInBytes);
const MemberOffset previous_state_offset = JNIEnvExt::LrtPreviousStateOffset(kArm64PointerSize);
const MemberOffset current_state_offset = JNIEnvExt::LrtSegmentStateOffset(kArm64PointerSize);
DCHECK_EQ(previous_state_offset.SizeValue() + kLRTSegmentStateSize,
current_state_offset.SizeValue());
// Set the current segment state together with restoring the cookie.
___ Stp(
reg_w(previous_state_reg.AsArm64().AsWRegister()),
reg_w(current_state_reg.AsArm64().AsWRegister()),
MemOperand(reg_x(jni_env_reg.AsArm64().AsXRegister()), previous_state_offset.Int32Value()));
}
#undef ___
} // namespace arm64
} // namespace art