v8/test/unittests/assembler/turbo-assembler-arm64-unittest.cc
Jakob Gruber c7cb9beca1 Reland "Reland "[deoptimizer] Change deopt entries into builtins""
This is a reland of fbfa9bf4ec

The arm64 was missing proper codegen for CFI, thus sizes were off.

Original change's description:
> Reland "[deoptimizer] Change deopt entries into builtins"
>
> This is a reland of 7f58ced72e
>
> It fixes the different exit size emitted on x64/Atom CPUs due to
> performance tuning in TurboAssembler::Call. Additionally, add
> cctests to verify the fixed size exits.
>
> Original change's description:
> > [deoptimizer] Change deopt entries into builtins
> >
> > While the overall goal of this commit is to change deoptimization
> > entries into builtins, there are multiple related things happening:
> >
> > - Deoptimization entries, formerly stubs (i.e. Code objects generated
> >   at runtime, guaranteed to be immovable), have been converted into
> >   builtins. The major restriction is that we now need to preserve the
> >   kRootRegister, which was formerly used on most architectures to pass
> >   the deoptimization id. The solution differs based on platform.
> > - Renamed DEOPT_ENTRIES_OR_FOR_TESTING code kind to FOR_TESTING.
> > - Removed heap/ support for immovable Code generation.
> > - Removed the DeserializerData class (no longer needed).
> > - arm64: to preserve 4-byte deopt exits, introduced a new optimization
> >   in which the final jump to the deoptimization entry is generated
> >   once per Code object, and deopt exits can continue to emit a
> >   near-call.
> > - arm,ia32,x64: change to fixed-size deopt exits. This reduces exit
> >   sizes by 4/8, 5, and 5 bytes, respectively.
> >
> > On arm the deopt exit size is reduced from 12 (or 16) bytes to 8 bytes
> > by using the same strategy as on arm64 (recalc deopt id from return
> > address). Before:
> >
> >  e300a002       movw r10, <id>
> >  e59fc024       ldr ip, [pc, <entry offset>]
> >  e12fff3c       blx ip
> >
> > After:
> >
> >  e59acb35       ldr ip, [r10, <entry offset>]
> >  e12fff3c       blx ip
> >
> > On arm64 the deopt exit size remains 4 bytes (or 8 bytes in same cases
> > with CFI). Additionally, up to 4 builtin jumps are emitted per Code
> > object (max 32 bytes added overhead per Code object). Before:
> >
> >  9401cdae       bl <entry offset>
> >
> > After:
> >
> >  # eager deoptimization entry jump.
> >  f95b1f50       ldr x16, [x26, <eager entry offset>]
> >  d61f0200       br x16
> >  # lazy deoptimization entry jump.
> >  f95b2b50       ldr x16, [x26, <lazy entry offset>]
> >  d61f0200       br x16
> >  # the deopt exit.
> >  97fffffc       bl <eager deoptimization entry jump offset>
> >
> > On ia32 the deopt exit size is reduced from 10 to 5 bytes. Before:
> >
> >  bb00000000     mov ebx,<id>
> >  e825f5372b     call <entry>
> >
> > After:
> >
> >  e8ea2256ba     call <entry>
> >
> > On x64 the deopt exit size is reduced from 12 to 7 bytes. Before:
> >
> >  49c7c511000000 REX.W movq r13,<id>
> >  e8ea2f0700     call <entry>
> >
> > After:
> >
> >  41ff9560360000 call [r13+<entry offset>]
> >
> > Bug: v8:8661,v8:8768
> > Change-Id: I13e30aedc360474dc818fecc528ce87c3bfeed42
> > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2465834
> > Commit-Queue: Jakob Gruber <jgruber@chromium.org>
> > Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
> > Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
> > Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
> > Cr-Commit-Position: refs/heads/master@{#70597}
>
> Tbr: ulan@chromium.org, tebbi@chromium.org, rmcilroy@chromium.org
> Bug: v8:8661,v8:8768,chromium:1140165
> Change-Id: Ibcd5c39c58a70bf2b2ac221aa375fc68d495e144
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2485506
> Reviewed-by: Jakob Gruber <jgruber@chromium.org>
> Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
> Commit-Queue: Jakob Gruber <jgruber@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#70655}

Tbr: ulan@chromium.org, tebbi@chromium.org, rmcilroy@chromium.org
Bug: v8:8661
Bug: v8:8768
Bug: chromium:1140165
Change-Id: I471cc94fc085e527dc9bfb5a84b96bd907c2333f
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2488682
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#70672}
2020-10-21 06:01:38 +00:00

197 lines
6.2 KiB
C++

// Copyright 2018 the V8 project authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#include "src/codegen/arm64/macro-assembler-arm64-inl.h"
#include "src/codegen/macro-assembler.h"
#include "src/execution/simulator.h"
#include "src/utils/ostreams.h"
#include "test/common/assembler-tester.h"
#include "test/unittests/test-utils.h"
#include "testing/gtest-support.h"
namespace v8 {
namespace internal {
#define __ tasm.
// If we are running on android and the output is not redirected (i.e. ends up
// in the android log) then we cannot find the error message in the output. This
// macro just returns the empty string in that case.
#if defined(ANDROID) && !defined(V8_ANDROID_LOG_STDOUT)
#define ERROR_MESSAGE(msg) ""
#else
#define ERROR_MESSAGE(msg) msg
#endif
// Test the x64 assembler by compiling some simple functions into
// a buffer and executing them. These tests do not initialize the
// V8 library, create a context, or use any V8 objects.
class TurboAssemblerTest : public TestWithIsolate {};
TEST_F(TurboAssemblerTest, TestHardAbort) {
auto buffer = AllocateAssemblerBuffer();
TurboAssembler tasm(isolate(), AssemblerOptions{}, CodeObjectRequired::kNo,
buffer->CreateView());
__ set_root_array_available(false);
__ set_abort_hard(true);
__ CodeEntry();
__ Abort(AbortReason::kNoReason);
CodeDesc desc;
tasm.GetCode(isolate(), &desc);
buffer->MakeExecutable();
// We need an isolate here to execute in the simulator.
auto f = GeneratedCode<void>::FromBuffer(isolate(), buffer->start());
ASSERT_DEATH_IF_SUPPORTED({ f.Call(); }, ERROR_MESSAGE("abort: no reason"));
}
TEST_F(TurboAssemblerTest, TestCheck) {
auto buffer = AllocateAssemblerBuffer();
TurboAssembler tasm(isolate(), AssemblerOptions{}, CodeObjectRequired::kNo,
buffer->CreateView());
__ set_root_array_available(false);
__ set_abort_hard(true);
__ CodeEntry();
// Fail if the first parameter is 17.
__ Mov(w1, Immediate(17));
__ Cmp(w0, w1); // 1st parameter is in {w0}.
__ Check(Condition::ne, AbortReason::kNoReason);
__ Ret();
CodeDesc desc;
tasm.GetCode(isolate(), &desc);
buffer->MakeExecutable();
// We need an isolate here to execute in the simulator.
auto f = GeneratedCode<void, int>::FromBuffer(isolate(), buffer->start());
f.Call(0);
f.Call(18);
ASSERT_DEATH_IF_SUPPORTED({ f.Call(17); }, ERROR_MESSAGE("abort: no reason"));
}
struct MoveObjectAndSlotTestCase {
const char* comment;
Register dst_object;
Register dst_slot;
Register object;
Register offset_register = no_reg;
};
const MoveObjectAndSlotTestCase kMoveObjectAndSlotTestCases[] = {
{"no overlap", x0, x1, x2},
{"no overlap", x0, x1, x2, x3},
{"object == dst_object", x2, x1, x2},
{"object == dst_object", x2, x1, x2, x3},
{"object == dst_slot", x1, x2, x2},
{"object == dst_slot", x1, x2, x2, x3},
{"offset == dst_object", x0, x1, x2, x0},
{"offset == dst_object && object == dst_slot", x0, x1, x1, x0},
{"offset == dst_slot", x0, x1, x2, x1},
{"offset == dst_slot && object == dst_object", x0, x1, x0, x1}};
// Make sure we include offsets that cannot be encoded in an add instruction.
const int kOffsets[] = {0, 42, kMaxRegularHeapObjectSize, 0x101001};
template <typename T>
class TurboAssemblerTestWithParam : public TurboAssemblerTest,
public ::testing::WithParamInterface<T> {};
using TurboAssemblerTestMoveObjectAndSlot =
TurboAssemblerTestWithParam<MoveObjectAndSlotTestCase>;
TEST_P(TurboAssemblerTestMoveObjectAndSlot, MoveObjectAndSlot) {
const MoveObjectAndSlotTestCase test_case = GetParam();
TRACED_FOREACH(int32_t, offset, kOffsets) {
auto buffer = AllocateAssemblerBuffer();
TurboAssembler tasm(nullptr, AssemblerOptions{}, CodeObjectRequired::kNo,
buffer->CreateView());
__ CodeEntry();
__ Push(x0, padreg);
__ Mov(test_case.object, x1);
Register src_object = test_case.object;
Register dst_object = test_case.dst_object;
Register dst_slot = test_case.dst_slot;
Operand offset_operand(0);
if (test_case.offset_register == no_reg) {
offset_operand = Operand(offset);
} else {
__ Mov(test_case.offset_register, Operand(offset));
offset_operand = Operand(test_case.offset_register);
}
std::stringstream comment;
comment << "-- " << test_case.comment << ": MoveObjectAndSlot("
<< dst_object << ", " << dst_slot << ", " << src_object << ", ";
if (test_case.offset_register == no_reg) {
comment << "#" << offset;
} else {
comment << test_case.offset_register;
}
comment << ") --";
__ RecordComment(comment.str().c_str());
__ MoveObjectAndSlot(dst_object, dst_slot, src_object, offset_operand);
__ RecordComment("--");
// The `result` pointer was saved on the stack.
UseScratchRegisterScope temps(&tasm);
Register scratch = temps.AcquireX();
__ Pop(padreg, scratch);
__ Str(dst_object, MemOperand(scratch));
__ Str(dst_slot, MemOperand(scratch, kSystemPointerSize));
__ Ret();
CodeDesc desc;
tasm.GetCode(nullptr, &desc);
if (FLAG_print_code) {
Handle<Code> code =
Factory::CodeBuilder(isolate(), desc, CodeKind::FOR_TESTING).Build();
StdoutStream os;
code->Print(os);
}
buffer->MakeExecutable();
// We need an isolate here to execute in the simulator.
auto f = GeneratedCode<void, byte**, byte*>::FromBuffer(isolate(),
buffer->start());
byte* object = new byte[offset];
byte* result[] = {nullptr, nullptr};
f.Call(result, object);
// The first element must be the address of the object, and the second the
// slot addressed by `offset`.
EXPECT_EQ(result[0], &object[0]);
EXPECT_EQ(result[1], &object[offset]);
delete[] object;
}
}
INSTANTIATE_TEST_SUITE_P(TurboAssemblerTest,
TurboAssemblerTestMoveObjectAndSlot,
::testing::ValuesIn(kMoveObjectAndSlotTestCases));
#undef __
#undef ERROR_MESSAGE
} // namespace internal
} // namespace v8