v8/test/cctest/parsing/test-scanner-streams.cc

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

910 lines
33 KiB
C++
Raw Normal View History

Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
// Copyright 2016 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/heap/factory-inl.h"
#include "src/objects/objects-inl.h"
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
#include "src/parsing/scanner-character-streams.h"
#include "src/parsing/scanner.h"
#include "test/cctest/cctest.h"
namespace {
// Implement ExternalSourceStream based on const char**.
// This will take each string as one chunk. The last chunk must be empty.
class ChunkSource : public v8::ScriptCompiler::ExternalSourceStream {
public:
explicit ChunkSource(const char** chunks) : current_(0) {
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
do {
chunks_.push_back(
{reinterpret_cast<const uint8_t*>(*chunks), strlen(*chunks)});
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
chunks++;
} while (chunks_.back().len > 0);
}
explicit ChunkSource(const char* chunks) : current_(0) {
do {
chunks_.push_back(
{reinterpret_cast<const uint8_t*>(chunks), strlen(chunks)});
chunks += strlen(chunks) + 1;
} while (chunks_.back().len > 0);
}
ChunkSource(const uint8_t* data, size_t char_size, size_t len,
bool extra_chunky)
: current_(0) {
// If extra_chunky, we'll use increasingly large chunk sizes. If not, we'll
// have a single chunk of full length. Make sure that chunks are always
// aligned to char-size though.
size_t chunk_size = extra_chunky ? char_size : len;
for (size_t i = 0; i < len; i += chunk_size, chunk_size += char_size) {
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
chunks_.push_back({data + i, i::Min(chunk_size, len - i)});
}
chunks_.push_back({nullptr, 0});
}
~ChunkSource() override = default;
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
bool SetBookmark() override { return false; }
void ResetToBookmark() override {}
size_t GetMoreData(const uint8_t** src) override {
DCHECK_LT(current_, chunks_.size());
Chunk& next = chunks_[current_++];
uint8_t* chunk = new uint8_t[next.len];
if (next.len > 0) {
i::MemMove(chunk, next.ptr, next.len);
}
*src = chunk;
return next.len;
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
}
private:
struct Chunk {
const uint8_t* ptr;
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
size_t len;
};
std::vector<Chunk> chunks_;
size_t current_;
};
// Checks that Lock() / Unlock() pairs are balanced. Not thread-safe.
class LockChecker {
public:
LockChecker() : lock_depth_(0) {}
~LockChecker() { CHECK_EQ(0, lock_depth_); }
void Lock() const { lock_depth_++; }
void Unlock() const {
CHECK_GT(lock_depth_, 0);
lock_depth_--;
}
bool IsLocked() const { return lock_depth_ > 0; }
int LockDepth() const { return lock_depth_; }
protected:
mutable int lock_depth_;
};
class TestExternalResource : public v8::String::ExternalStringResource,
public LockChecker {
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
public:
explicit TestExternalResource(uint16_t* data, int length)
: LockChecker(), data_(data), length_(static_cast<size_t>(length)) {}
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
const uint16_t* data() const override {
CHECK(IsLocked());
return data_;
}
size_t length() const override { return length_; }
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
bool IsCacheable() const override { return false; }
void Lock() const override { LockChecker::Lock(); }
void Unlock() const override { LockChecker::Unlock(); }
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
private:
uint16_t* data_;
size_t length_;
};
class TestExternalOneByteResource
: public v8::String::ExternalOneByteStringResource,
public LockChecker {
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
public:
TestExternalOneByteResource(const char* data, size_t length)
: data_(data), length_(length) {}
const char* data() const override {
CHECK(IsLocked());
return data_;
}
size_t length() const override { return length_; }
bool IsCacheable() const override { return false; }
void Lock() const override { LockChecker::Lock(); }
void Unlock() const override { LockChecker::Unlock(); }
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
private:
const char* data_;
size_t length_;
};
// A test string with all lengths of utf-8 encodings.
const char unicode_utf8[] =
"abc" // 3x ascii
"\xc3\xa4" // a Umlaut, code point 228
"\xe2\xa8\xa0" // >> (math symbol), code point 10784
"\xf0\x9f\x92\xa9" // best character, code point 128169,
// as utf-16 surrogates: 55357 56489
"def"; // 3x ascii again.
const uint16_t unicode_ucs2[] = {97, 98, 99, 228, 10784, 55357,
56489, 100, 101, 102, 0};
i::Handle<i::String> NewExternalTwoByteStringFromResource(
i::Isolate* isolate, TestExternalResource* resource) {
i::Factory* factory = isolate->factory();
// String creation accesses the resource.
resource->Lock();
i::Handle<i::String> uc16_string(
factory->NewExternalStringFromTwoByte(resource).ToHandleChecked());
resource->Unlock();
return uc16_string;
}
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
} // anonymous namespace
TEST(Utf8StreamAsciiOnly) {
const char* chunks[] = {"abc", "def", "ghi", ""};
ChunkSource chunk_source(chunks);
std::unique_ptr<v8::internal::Utf16CharacterStream> stream(
v8::internal::ScannerStream::For(
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
&chunk_source, v8::ScriptCompiler::StreamedSource::UTF8));
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
// Read the data without dying.
v8::internal::uc32 c;
do {
c = stream->Advance();
} while (c != v8::internal::Utf16CharacterStream::kEndOfInput);
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
}
TEST(Utf8StreamMaxNonSurrogateCharCode) {
const char* chunks[] = {"\uffff\uffff", ""};
ChunkSource chunk_source(chunks);
std::unique_ptr<v8::internal::Utf16CharacterStream> stream(
v8::internal::ScannerStream::For(
&chunk_source, v8::ScriptCompiler::StreamedSource::UTF8));
// Read the correct character.
uint16_t max = unibrow::Utf16::kMaxNonSurrogateCharCode;
CHECK_EQ(max, static_cast<uint32_t>(stream->Advance()));
CHECK_EQ(max, static_cast<uint32_t>(stream->Advance()));
CHECK_EQ(i::Utf16CharacterStream::kEndOfInput, stream->Advance());
}
TEST(Utf8StreamBOM) {
// Construct test string w/ UTF-8 BOM (byte order mark)
char data[3 + arraysize(unicode_utf8)] = {"\xef\xbb\xbf"};
strncpy(data + 3, unicode_utf8, arraysize(unicode_utf8));
const char* chunks[] = {data, "\0"};
ChunkSource chunk_source(chunks);
std::unique_ptr<v8::internal::Utf16CharacterStream> stream(
v8::internal::ScannerStream::For(
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
&chunk_source, v8::ScriptCompiler::StreamedSource::UTF8));
// Read the data without tripping over the BOM.
for (size_t i = 0; unicode_ucs2[i]; i++) {
CHECK_EQ(unicode_ucs2[i], stream->Advance());
}
CHECK_EQ(v8::internal::Utf16CharacterStream::kEndOfInput, stream->Advance());
// Make sure seek works.
stream->Seek(0);
CHECK_EQ(unicode_ucs2[0], stream->Advance());
stream->Seek(5);
CHECK_EQ(unicode_ucs2[5], stream->Advance());
// Try again, but make sure we have to seek 'backwards'.
while (v8::internal::Utf16CharacterStream::kEndOfInput != stream->Advance()) {
// Do nothing. We merely advance the stream to the end of its input.
}
stream->Seek(5);
CHECK_EQ(unicode_ucs2[5], stream->Advance());
}
TEST(Utf8SplitBOM) {
// Construct chunks with a BOM split into two chunks.
char partial_bom[] = "\xef\xbb";
char data[1 + arraysize(unicode_utf8)] = {"\xbf"};
strncpy(data + 1, unicode_utf8, arraysize(unicode_utf8));
{
const char* chunks[] = {partial_bom, data, "\0"};
ChunkSource chunk_source(chunks);
std::unique_ptr<v8::internal::Utf16CharacterStream> stream(
v8::internal::ScannerStream::For(
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
&chunk_source, v8::ScriptCompiler::StreamedSource::UTF8));
// Read the data without tripping over the BOM.
for (size_t i = 0; unicode_ucs2[i]; i++) {
CHECK_EQ(unicode_ucs2[i], stream->Advance());
}
}
// And now with single-byte BOM chunks.
char bom_byte_1[] = "\xef";
char bom_byte_2[] = "\xbb";
{
const char* chunks[] = {bom_byte_1, bom_byte_2, data, "\0"};
ChunkSource chunk_source(chunks);
std::unique_ptr<v8::internal::Utf16CharacterStream> stream(
v8::internal::ScannerStream::For(
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
&chunk_source, v8::ScriptCompiler::StreamedSource::UTF8));
// Read the data without tripping over the BOM.
for (size_t i = 0; unicode_ucs2[i]; i++) {
CHECK_EQ(unicode_ucs2[i], stream->Advance());
}
}
}
TEST(Utf8SplitMultiBOM) {
// Construct chunks with a split BOM followed by another split BOM.
const char* chunks = "\xef\xbb\0\xbf\xef\xbb\0\xbf\0\0";
ChunkSource chunk_source(chunks);
std::unique_ptr<i::Utf16CharacterStream> stream(
v8::internal::ScannerStream::For(
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
&chunk_source, v8::ScriptCompiler::StreamedSource::UTF8));
// Read the data, ensuring we get exactly one of the two BOMs back.
CHECK_EQ(0xFEFF, stream->Advance());
CHECK_EQ(i::Utf16CharacterStream::kEndOfInput, stream->Advance());
}
TEST(Utf8AdvanceUntil) {
// Test utf-8 advancing until a certain char.
const char line_term = '\n';
const size_t kLen = arraysize(unicode_utf8);
char data[kLen + 1];
strncpy(data, unicode_utf8, kLen);
data[kLen - 1] = line_term;
data[kLen] = '\0';
{
const char* chunks[] = {data, "\0"};
ChunkSource chunk_source(chunks);
std::unique_ptr<v8::internal::Utf16CharacterStream> stream(
v8::internal::ScannerStream::For(
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
&chunk_source, v8::ScriptCompiler::StreamedSource::UTF8));
int32_t res = stream->AdvanceUntil(
[](int32_t c0_) { return unibrow::IsLineTerminator(c0_); });
CHECK_EQ(line_term, res);
}
}
TEST(AdvanceMatchAdvanceUntil) {
// Test if single advance and advanceUntil behave the same
char data[] = {'a', 'b', '\n', 'c', '\0'};
{
const char* chunks[] = {data, "\0"};
ChunkSource chunk_source_a(chunks);
std::unique_ptr<v8::internal::Utf16CharacterStream> stream_advance(
v8::internal::ScannerStream::For(
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
&chunk_source_a, v8::ScriptCompiler::StreamedSource::UTF8));
ChunkSource chunk_source_au(chunks);
std::unique_ptr<v8::internal::Utf16CharacterStream> stream_advance_until(
v8::internal::ScannerStream::For(
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
&chunk_source_au, v8::ScriptCompiler::StreamedSource::UTF8));
int32_t au_c0_ = stream_advance_until->AdvanceUntil(
[](int32_t c0_) { return unibrow::IsLineTerminator(c0_); });
int32_t a_c0_ = '0';
while (!unibrow::IsLineTerminator(a_c0_)) {
a_c0_ = stream_advance->Advance();
}
// Check both advances methods have the same output
CHECK_EQ(a_c0_, au_c0_);
// Check if both set the cursor to the correct position by advancing both
// streams by one character.
a_c0_ = stream_advance->Advance();
au_c0_ = stream_advance_until->Advance();
CHECK_EQ(a_c0_, au_c0_);
}
}
TEST(Utf8AdvanceUntilOverChunkBoundaries) {
// Test utf-8 advancing until a certain char, crossing chunk boundaries.
// Split the test string at each byte and pass it to the stream. This way,
// we'll have a split at each possible boundary.
size_t len = strlen(unicode_utf8);
char buffer[arraysize(unicode_utf8) + 4];
for (size_t i = 1; i < len; i++) {
// Copy source string into buffer, splitting it at i.
// Then add three chunks, 0..i-1, i..strlen-1, empty.
strncpy(buffer, unicode_utf8, i);
strncpy(buffer + i + 1, unicode_utf8 + i, len - i);
buffer[i] = '\0';
buffer[len + 1] = '\n';
buffer[len + 2] = '\0';
buffer[len + 3] = '\0';
const char* chunks[] = {buffer, buffer + i + 1, buffer + len + 2};
ChunkSource chunk_source(chunks);
std::unique_ptr<v8::internal::Utf16CharacterStream> stream(
v8::internal::ScannerStream::For(
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
&chunk_source, v8::ScriptCompiler::StreamedSource::UTF8));
int32_t res = stream->AdvanceUntil(
[](int32_t c0_) { return unibrow::IsLineTerminator(c0_); });
CHECK_EQ(buffer[len + 1], res);
}
}
TEST(Utf8ChunkBoundaries) {
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
// Test utf-8 parsing at chunk boundaries.
// Split the test string at each byte and pass it to the stream. This way,
// we'll have a split at each possible boundary.
size_t len = strlen(unicode_utf8);
char buffer[arraysize(unicode_utf8) + 3];
for (size_t i = 1; i < len; i++) {
// Copy source string into buffer, splitting it at i.
// Then add three chunks, 0..i-1, i..strlen-1, empty.
strncpy(buffer, unicode_utf8, i);
strncpy(buffer + i + 1, unicode_utf8 + i, len - i);
buffer[i] = '\0';
buffer[len + 1] = '\0';
buffer[len + 2] = '\0';
const char* chunks[] = {buffer, buffer + i + 1, buffer + len + 2};
ChunkSource chunk_source(chunks);
std::unique_ptr<v8::internal::Utf16CharacterStream> stream(
v8::internal::ScannerStream::For(
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
&chunk_source, v8::ScriptCompiler::StreamedSource::UTF8));
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
for (size_t i = 0; unicode_ucs2[i]; i++) {
CHECK_EQ(unicode_ucs2[i], stream->Advance());
}
CHECK_EQ(v8::internal::Utf16CharacterStream::kEndOfInput,
stream->Advance());
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
}
}
TEST(Utf8SingleByteChunks) {
// Have each byte as a single-byte chunk.
size_t len = strlen(unicode_utf8);
char buffer[arraysize(unicode_utf8) + 4];
for (size_t i = 1; i < len - 1; i++) {
// Copy source string into buffer, make a single-byte chunk at i.
strncpy(buffer, unicode_utf8, i);
strncpy(buffer + i + 3, unicode_utf8 + i + 1, len - i - 1);
buffer[i] = '\0';
buffer[i + 1] = unicode_utf8[i];
buffer[i + 2] = '\0';
buffer[len + 2] = '\0';
buffer[len + 3] = '\0';
const char* chunks[] = {buffer, buffer + i + 1, buffer + i + 3,
buffer + len + 3};
ChunkSource chunk_source(chunks);
std::unique_ptr<v8::internal::Utf16CharacterStream> stream(
v8::internal::ScannerStream::For(
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
&chunk_source, v8::ScriptCompiler::StreamedSource::UTF8));
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
for (size_t j = 0; unicode_ucs2[j]; j++) {
CHECK_EQ(unicode_ucs2[j], stream->Advance());
}
CHECK_EQ(v8::internal::Utf16CharacterStream::kEndOfInput,
stream->Advance());
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
}
}
#define CHECK_EQU(v1, v2) CHECK_EQ(static_cast<int>(v1), static_cast<int>(v2))
void TestCharacterStream(const char* reference, i::Utf16CharacterStream* stream,
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
unsigned length, unsigned start, unsigned end) {
// Read streams one char at a time
unsigned i;
for (i = start; i < end; i++) {
CHECK_EQU(i, stream->pos());
CHECK_EQU(reference[i], stream->Advance());
}
CHECK_EQU(end, stream->pos());
CHECK_EQU(i::Utf16CharacterStream::kEndOfInput, stream->Advance());
CHECK_EQU(end + 1, stream->pos());
stream->Back();
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
// Pushback, re-read, pushback again.
while (i > end / 4) {
int32_t c0 = reference[i - 1];
CHECK_EQU(i, stream->pos());
stream->Back();
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
i--;
CHECK_EQU(i, stream->pos());
int32_t c1 = stream->Advance();
i++;
CHECK_EQU(i, stream->pos());
CHECK_EQ(c0, c1);
stream->Back();
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
i--;
CHECK_EQU(i, stream->pos());
}
// Seek + read streams one char at a time.
unsigned halfway = end / 2;
stream->Seek(stream->pos() + halfway - i);
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
for (i = halfway; i < end; i++) {
CHECK_EQU(i, stream->pos());
CHECK_EQU(reference[i], stream->Advance());
}
CHECK_EQU(i, stream->pos());
CHECK_LT(stream->Advance(), 0);
// Seek back, then seek beyond end of stream.
stream->Seek(start);
if (start < length) {
CHECK_EQU(stream->Advance(), reference[start]);
} else {
CHECK_LT(stream->Advance(), 0);
}
stream->Seek(length + 5);
CHECK_LT(stream->Advance(), 0);
}
void TestCloneCharacterStream(const char* reference,
i::Utf16CharacterStream* stream,
unsigned length) {
std::unique_ptr<i::Utf16CharacterStream> clone = stream->Clone();
unsigned i;
unsigned halfway = length / 2;
// Advance original half way.
for (i = 0; i < halfway; i++) {
CHECK_EQU(i, stream->pos());
CHECK_EQU(reference[i], stream->Advance());
}
// Test advancing original stream didn't affect the clone.
TestCharacterStream(reference, clone.get(), length, 0, length);
// Test advancing clone didn't affect original stream.
TestCharacterStream(reference, stream, length, i, length);
}
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
#undef CHECK_EQU
void TestCharacterStreams(const char* one_byte_source, unsigned length,
unsigned start = 0, unsigned end = 0) {
if (end == 0) end = length;
i::Isolate* isolate = CcTest::i_isolate();
i::Factory* factory = isolate->factory();
// 2-byte external string
std::unique_ptr<i::uc16[]> uc16_buffer(new i::uc16[length]);
i::Vector<const i::uc16> two_byte_vector(uc16_buffer.get(),
static_cast<int>(length));
{
for (unsigned i = 0; i < length; i++) {
uc16_buffer[i] = static_cast<i::uc16>(one_byte_source[i]);
}
TestExternalResource resource(uc16_buffer.get(), length);
i::Handle<i::String> uc16_string(
NewExternalTwoByteStringFromResource(isolate, &resource));
std::unique_ptr<i::Utf16CharacterStream> uc16_stream(
i::ScannerStream::For(isolate, uc16_string, start, end));
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
TestCharacterStream(one_byte_source, uc16_stream.get(), length, start, end);
// This avoids the GC from trying to free a stack allocated resource.
if (uc16_string->IsExternalString())
i::Handle<i::ExternalTwoByteString>::cast(uc16_string)
->SetResource(isolate, nullptr);
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
}
// 1-byte external string
i::Vector<const uint8_t> one_byte_vector =
i::OneByteVector(one_byte_source, static_cast<int>(length));
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
i::Handle<i::String> one_byte_string =
factory->NewStringFromOneByte(one_byte_vector).ToHandleChecked();
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
{
TestExternalOneByteResource one_byte_resource(one_byte_source, length);
i::Handle<i::String> ext_one_byte_string(
factory->NewExternalStringFromOneByte(&one_byte_resource)
.ToHandleChecked());
std::unique_ptr<i::Utf16CharacterStream> one_byte_stream(
i::ScannerStream::For(isolate, ext_one_byte_string, start, end));
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
TestCharacterStream(one_byte_source, one_byte_stream.get(), length, start,
end);
// This avoids the GC from trying to free a stack allocated resource.
if (ext_one_byte_string->IsExternalString())
i::Handle<i::ExternalOneByteString>::cast(ext_one_byte_string)
->SetResource(isolate, nullptr);
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
}
// 1-byte generic i::String
{
std::unique_ptr<i::Utf16CharacterStream> string_stream(
i::ScannerStream::For(isolate, one_byte_string, start, end));
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
TestCharacterStream(one_byte_source, string_stream.get(), length, start,
end);
}
// 2-byte generic i::String
{
i::Handle<i::String> two_byte_string =
factory->NewStringFromTwoByte(two_byte_vector).ToHandleChecked();
std::unique_ptr<i::Utf16CharacterStream> two_byte_string_stream(
i::ScannerStream::For(isolate, two_byte_string, start, end));
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
TestCharacterStream(one_byte_source, two_byte_string_stream.get(), length,
start, end);
}
// Streaming has no notion of start/end, so let's skip streaming tests for
// these cases.
if (start != 0 || end != length) return;
// 1-byte streaming stream, single + many chunks.
{
const uint8_t* data = one_byte_vector.begin();
const uint8_t* data_end = one_byte_vector.end();
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
ChunkSource single_chunk(data, 1, data_end - data, false);
std::unique_ptr<i::Utf16CharacterStream> one_byte_streaming_stream(
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
i::ScannerStream::For(&single_chunk,
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
v8::ScriptCompiler::StreamedSource::ONE_BYTE));
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
TestCharacterStream(one_byte_source, one_byte_streaming_stream.get(),
length, start, end);
ChunkSource many_chunks(data, 1, data_end - data, true);
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
one_byte_streaming_stream.reset(i::ScannerStream::For(
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
&many_chunks, v8::ScriptCompiler::StreamedSource::ONE_BYTE));
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
TestCharacterStream(one_byte_source, one_byte_streaming_stream.get(),
length, start, end);
}
// UTF-8 streaming stream, single + many chunks.
{
const uint8_t* data = one_byte_vector.begin();
const uint8_t* data_end = one_byte_vector.end();
ChunkSource chunks(data, 1, data_end - data, false);
std::unique_ptr<i::Utf16CharacterStream> utf8_streaming_stream(
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
i::ScannerStream::For(&chunks,
v8::ScriptCompiler::StreamedSource::UTF8));
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
TestCharacterStream(one_byte_source, utf8_streaming_stream.get(), length,
start, end);
ChunkSource many_chunks(data, 1, data_end - data, true);
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
utf8_streaming_stream.reset(i::ScannerStream::For(
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
&many_chunks, v8::ScriptCompiler::StreamedSource::UTF8));
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
TestCharacterStream(one_byte_source, utf8_streaming_stream.get(), length,
start, end);
}
// 2-byte streaming stream, single + many chunks.
{
const uint8_t* data =
reinterpret_cast<const uint8_t*>(two_byte_vector.begin());
const uint8_t* data_end =
reinterpret_cast<const uint8_t*>(two_byte_vector.end());
ChunkSource chunks(data, 2, data_end - data, false);
std::unique_ptr<i::Utf16CharacterStream> two_byte_streaming_stream(
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
i::ScannerStream::For(&chunks,
v8::ScriptCompiler::StreamedSource::TWO_BYTE));
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
TestCharacterStream(one_byte_source, two_byte_streaming_stream.get(),
length, start, end);
ChunkSource many_chunks(data, 2, data_end - data, true);
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
two_byte_streaming_stream.reset(i::ScannerStream::For(
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
&many_chunks, v8::ScriptCompiler::StreamedSource::TWO_BYTE));
Rework scanner-character-streams. - Smaller, more consistent streams API (Advance, Back, pos, Seek) - Remove implementations from the header, in favor of creation functions. Observe: - Performance: - All Utf16CharacterStream methods have an inlinable V8_LIKELY w/ a body of only a few instructions. I expect most calls to end up there. - There used to be performance problems w/ bookmarking, particularly with copying too much data on SetBookmark w/ UTF-8 streaming streams. All those copies are gone. - The old streaming streams implementation used to copy data even for 2-byte input. It no longer does. - The only remaining 'slow' method is the Seek(.) slow case for utf-8 streaming streams. I don't expect this to be called a lot; and even if, I expect it to be offset by the gains in the (vastly more frequent) calls to the other methods or the 'fast path'. - If it still bothers us, there are several ways to speed it up. - API & code cleanliness: - I want to remove the 'old' API in a follow-up CL, which should mostly delete code, or replace it 1:1. - In a 2nd follow-up I want to delete much of the UTF-8 handling in Blink for streaming streams. - The "bookmark" is now always implemented (and mostly very fast), so we should be able to use it for more things. - Testing & correctness: - The unit tests now cover all stream implementations, and are pretty good and triggering all the edge cases. - Vastly more DCHECKs of the invariants. BUG=v8:4947 Review-Url: https://codereview.chromium.org/2314663002 Cr-Commit-Position: refs/heads/master@{#39464}
2016-09-16 08:29:41 +00:00
TestCharacterStream(one_byte_source, two_byte_streaming_stream.get(),
length, start, end);
}
}
TEST(CharacterStreams) {
v8::Isolate* isolate = CcTest::isolate();
v8::HandleScope handles(isolate);
v8::Local<v8::Context> context = v8::Context::New(isolate);
v8::Context::Scope context_scope(context);
TestCharacterStreams("abcdefghi", 9);
TestCharacterStreams("abc\0\n\r\x7f", 7);
TestCharacterStreams("\0", 1);
TestCharacterStreams("", 0);
// 4k large buffer.
char buffer[4096 + 1];
for (unsigned i = 0; i < arraysize(buffer); i++) {
buffer[i] = static_cast<char>(i & 0x7F);
}
buffer[arraysize(buffer) - 1] = '\0';
TestCharacterStreams(buffer, arraysize(buffer) - 1);
TestCharacterStreams(buffer, arraysize(buffer) - 1, 576, 3298);
}
// Regression test for crbug.com/651333. Read invalid utf-8.
TEST(Regress651333) {
const uint8_t bytes[] =
"A\xf1"
"ad"; // Anad, with n == n-with-tilde.
const uint16_t unicode[] = {65, 65533, 97, 100};
// Run the test for all sub-strings 0..N of bytes, to make sure we hit the
// error condition in and at chunk boundaries.
for (size_t len = 0; len < arraysize(bytes); len++) {
// Read len bytes from bytes, and compare against the expected unicode
// characters. Expect kBadChar ( == Unicode replacement char == code point
// 65533) instead of the incorrectly coded Latin1 char.
ChunkSource chunks(bytes, 1, len, false);
std::unique_ptr<i::Utf16CharacterStream> stream(i::ScannerStream::For(
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
&chunks, v8::ScriptCompiler::StreamedSource::UTF8));
for (size_t i = 0; i < len; i++) {
CHECK_EQ(unicode[i], stream->Advance());
}
CHECK_EQ(i::Utf16CharacterStream::kEndOfInput, stream->Advance());
}
}
void TestChunkStreamAgainstReference(
const char* cases[],
const std::vector<std::vector<uint16_t>>& unicode_expected) {
for (size_t c = 0; c < unicode_expected.size(); ++c) {
ChunkSource chunk_source(cases[c]);
std::unique_ptr<i::Utf16CharacterStream> stream(i::ScannerStream::For(
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
&chunk_source, v8::ScriptCompiler::StreamedSource::UTF8));
for (size_t i = 0; i < unicode_expected[c].size(); i++) {
CHECK_EQ(unicode_expected[c][i], stream->Advance());
}
CHECK_EQ(i::Utf16CharacterStream::kEndOfInput, stream->Advance());
stream->Seek(0);
for (size_t i = 0; i < unicode_expected[c].size(); i++) {
CHECK_EQ(unicode_expected[c][i], stream->Advance());
}
CHECK_EQ(i::Utf16CharacterStream::kEndOfInput, stream->Advance());
}
}
TEST(Regress6377) {
const char* cases[] = {
"\xf0\x90\0" // first chunk - start of 4-byte seq
"\x80\x80" // second chunk - end of 4-byte seq
"a\0", // and an 'a'
"\xe0\xbf\0" // first chunk - start of 3-byte seq
"\xbf" // second chunk - one-byte end of 3-byte seq
"a\0", // and an 'a'
"\xc3\0" // first chunk - start of 2-byte seq
"\xbf" // second chunk - end of 2-byte seq
"a\0", // and an 'a'
"\xf0\x90\x80\0" // first chunk - start of 4-byte seq
"\x80" // second chunk - one-byte end of 4-byte seq
"a\xc3\0" // and an 'a' + start of 2-byte seq
"\xbf\0", // third chunk - end of 2-byte seq
};
const std::vector<std::vector<uint16_t>> unicode_expected = {
{0xD800, 0xDC00, 97}, {0xFFF, 97}, {0xFF, 97}, {0xD800, 0xDC00, 97, 0xFF},
};
CHECK_EQ(unicode_expected.size(), arraysize(cases));
TestChunkStreamAgainstReference(cases, unicode_expected);
}
TEST(Regress6836) {
const char* cases[] = {
// 0xC2 is a lead byte, but there's no continuation. The bug occurs when
// this happens near the chunk end.
"X\xc2Y\0",
// Last chunk ends with a 2-byte char lead.
"X\xc2\0",
// Last chunk ends with a 3-byte char lead and only one continuation
// character.
"X\xe0\xbf\0",
};
const std::vector<std::vector<uint16_t>> unicode_expected = {
{0x58, 0xFFFD, 0x59}, {0x58, 0xFFFD}, {0x58, 0xFFFD},
};
CHECK_EQ(unicode_expected.size(), arraysize(cases));
TestChunkStreamAgainstReference(cases, unicode_expected);
}
[unicode] Return (the correct) errors for overlong / surrogate sequences. This fix is two-fold: 1) Incremental UTF-8 decoding: Unify incorrect UTF-8 handling between V8 and Blink. Incremental UTF-8 decoding used to allow some overlong sequences / invalid code points which Blink treated as errors. This caused the decoder and the Blink UTF-8 decoder to produce a different number of bytes, resulting in random failures when scripts were streamed (especially, this was detected by the skipping inner functions feature which adds CHECKs against expected function positions). 2) Non-incremental UTF-8 decoding: return the correct amount of invalid characters. According to the encoding spec ( https://encoding.spec.whatwg.org/#utf-8-decoder ), the first byte of an overlong sequence / invalid code point generates an invalid character, and the rest of the bytes are not processed (i.e., pushed back to the byte stream). When they're handled, they will look like lonely continuation bytes, and will generate an invalid character each. As a result, an overlong 4-byte sequence should generate 4 invalid characters (not 1). This is a potentially breaking change, since the (non-incremental) UTF-8 decoding is exposed via the API (String::NewFromUtf8). The behavioral difference happens when the client is passing in invalid UTF-8 (containing overlong / surrogate sequences). However, afaict, this doesn't change the semantics of any JavaScript program: according to the ECMAScript spec, the program is a sequence of Unicode code points, and there's no way to invoke the UTF-8 decoding functionalities from inside JavaScript. Though, this changes the behavior of d8 when decoding source files which are invalid UTF-8. This doesn't change anything related to URI decoding (it already throws exceptions for overlong sequences / invalid code points). BUG: chromium:765608, chromium:758236, v8:5516 Bug: Change-Id: Ib029f6a8e87186794b092e4e8af32d01cee3ada0 Reviewed-on: https://chromium-review.googlesource.com/671020 Commit-Queue: Marja Hölttä <marja@chromium.org> Reviewed-by: Franziska Hinkelmann <franzih@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#48105}
2017-09-20 07:32:33 +00:00
TEST(TestOverlongAndInvalidSequences) {
const char* cases[] = {
// Overlong 2-byte sequence.
"X\xc0\xbfY\0",
// Another overlong 2-byte sequence.
"X\xc1\xbfY\0",
// Overlong 3-byte sequence.
"X\xe0\x9f\xbfY\0",
// Overlong 4-byte sequence.
"X\xf0\x89\xbf\xbfY\0",
// Invalid 3-byte sequence (reserved for surrogates).
"X\xed\xa0\x80Y\0",
// Invalid 4-bytes sequence (value out of range).
"X\xf4\x90\x80\x80Y\0",
};
const std::vector<std::vector<uint16_t>> unicode_expected = {
{0x58, 0xFFFD, 0xFFFD, 0x59},
{0x58, 0xFFFD, 0xFFFD, 0x59},
{0x58, 0xFFFD, 0xFFFD, 0xFFFD, 0x59},
{0x58, 0xFFFD, 0xFFFD, 0xFFFD, 0xFFFD, 0x59},
{0x58, 0xFFFD, 0xFFFD, 0xFFFD, 0x59},
{0x58, 0xFFFD, 0xFFFD, 0xFFFD, 0xFFFD, 0x59},
[unicode] Return (the correct) errors for overlong / surrogate sequences. This fix is two-fold: 1) Incremental UTF-8 decoding: Unify incorrect UTF-8 handling between V8 and Blink. Incremental UTF-8 decoding used to allow some overlong sequences / invalid code points which Blink treated as errors. This caused the decoder and the Blink UTF-8 decoder to produce a different number of bytes, resulting in random failures when scripts were streamed (especially, this was detected by the skipping inner functions feature which adds CHECKs against expected function positions). 2) Non-incremental UTF-8 decoding: return the correct amount of invalid characters. According to the encoding spec ( https://encoding.spec.whatwg.org/#utf-8-decoder ), the first byte of an overlong sequence / invalid code point generates an invalid character, and the rest of the bytes are not processed (i.e., pushed back to the byte stream). When they're handled, they will look like lonely continuation bytes, and will generate an invalid character each. As a result, an overlong 4-byte sequence should generate 4 invalid characters (not 1). This is a potentially breaking change, since the (non-incremental) UTF-8 decoding is exposed via the API (String::NewFromUtf8). The behavioral difference happens when the client is passing in invalid UTF-8 (containing overlong / surrogate sequences). However, afaict, this doesn't change the semantics of any JavaScript program: according to the ECMAScript spec, the program is a sequence of Unicode code points, and there's no way to invoke the UTF-8 decoding functionalities from inside JavaScript. Though, this changes the behavior of d8 when decoding source files which are invalid UTF-8. This doesn't change anything related to URI decoding (it already throws exceptions for overlong sequences / invalid code points). BUG: chromium:765608, chromium:758236, v8:5516 Bug: Change-Id: Ib029f6a8e87186794b092e4e8af32d01cee3ada0 Reviewed-on: https://chromium-review.googlesource.com/671020 Commit-Queue: Marja Hölttä <marja@chromium.org> Reviewed-by: Franziska Hinkelmann <franzih@chromium.org> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#48105}
2017-09-20 07:32:33 +00:00
};
CHECK_EQ(unicode_expected.size(), arraysize(cases));
TestChunkStreamAgainstReference(cases, unicode_expected);
}
TEST(RelocatingCharacterStream) {
ManualGCScope manual_gc_scope;
CcTest::InitializeVM();
i::Isolate* i_isolate = CcTest::i_isolate();
v8::HandleScope scope(CcTest::isolate());
const char* string = "abcd";
int length = static_cast<int>(strlen(string));
std::unique_ptr<i::uc16[]> uc16_buffer(new i::uc16[length]);
for (int i = 0; i < length; i++) {
uc16_buffer[i] = string[i];
}
i::Vector<const i::uc16> two_byte_vector(uc16_buffer.get(), length);
i::Handle<i::String> two_byte_string =
i_isolate->factory()
->NewStringFromTwoByte(two_byte_vector, i::AllocationType::kYoung)
.ToHandleChecked();
std::unique_ptr<i::Utf16CharacterStream> two_byte_string_stream(
i::ScannerStream::For(i_isolate, two_byte_string, 0, length));
CHECK_EQ('a', two_byte_string_stream->Advance());
CHECK_EQ('b', two_byte_string_stream->Advance());
CHECK_EQ(size_t{2}, two_byte_string_stream->pos());
i::String raw = *two_byte_string;
i_isolate->heap()->CollectGarbage(i::NEW_SPACE,
i::GarbageCollectionReason::kUnknown);
// GC moved the string.
CHECK_NE(raw, *two_byte_string);
CHECK_EQ('c', two_byte_string_stream->Advance());
CHECK_EQ('d', two_byte_string_stream->Advance());
}
TEST(RelocatingUnbufferedCharacterStream) {
ManualGCScope manual_gc_scope;
CcTest::InitializeVM();
i::Isolate* i_isolate = CcTest::i_isolate();
v8::HandleScope scope(CcTest::isolate());
const char16_t* string = u"abc\u2603";
int length = static_cast<int>(std::char_traits<char16_t>::length(string));
std::unique_ptr<i::uc16[]> uc16_buffer(new i::uc16[length]);
for (int i = 0; i < length; i++) {
uc16_buffer[i] = string[i];
}
i::Vector<const i::uc16> two_byte_vector(uc16_buffer.get(), length);
i::Handle<i::String> two_byte_string =
i_isolate->factory()
->NewStringFromTwoByte(two_byte_vector, i::AllocationType::kYoung)
.ToHandleChecked();
std::unique_ptr<i::Utf16CharacterStream> two_byte_string_stream(
i::ScannerStream::For(i_isolate, two_byte_string, 0, length));
// Seek to offset 2 so that the buffer_pos_ is not zero initially.
two_byte_string_stream->Seek(2);
CHECK_EQ('c', two_byte_string_stream->Advance());
CHECK_EQ(size_t{3}, two_byte_string_stream->pos());
i::String raw = *two_byte_string;
i_isolate->heap()->CollectGarbage(i::NEW_SPACE,
i::GarbageCollectionReason::kUnknown);
// GC moved the string and buffer was updated to the correct location.
CHECK_NE(raw, *two_byte_string);
// Check that we correctly moved based on buffer_pos_, not based on a position
// of zero.
CHECK_EQ(u'\u2603', two_byte_string_stream->Advance());
CHECK_EQ(size_t{4}, two_byte_string_stream->pos());
}
TEST(CloneCharacterStreams) {
v8::HandleScope handles(CcTest::isolate());
v8::Local<v8::Context> context = v8::Context::New(CcTest::isolate());
v8::Context::Scope context_scope(context);
i::Isolate* isolate = CcTest::i_isolate();
i::Factory* factory = isolate->factory();
const char* one_byte_source = "abcdefghi";
unsigned length = static_cast<unsigned>(strlen(one_byte_source));
// Check that cloning a character stream does not update
// 2-byte external string
std::unique_ptr<i::uc16[]> uc16_buffer(new i::uc16[length]);
i::Vector<const i::uc16> two_byte_vector(uc16_buffer.get(),
static_cast<int>(length));
{
for (unsigned i = 0; i < length; i++) {
uc16_buffer[i] = static_cast<i::uc16>(one_byte_source[i]);
}
TestExternalResource resource(uc16_buffer.get(), length);
i::Handle<i::String> uc16_string(
NewExternalTwoByteStringFromResource(isolate, &resource));
std::unique_ptr<i::Utf16CharacterStream> uc16_stream(
i::ScannerStream::For(isolate, uc16_string, 0, length));
CHECK(resource.IsLocked());
CHECK_EQ(1, resource.LockDepth());
std::unique_ptr<i::Utf16CharacterStream> cloned = uc16_stream->Clone();
CHECK_EQ(2, resource.LockDepth());
uc16_stream = std::move(cloned);
CHECK_EQ(1, resource.LockDepth());
TestCloneCharacterStream(one_byte_source, uc16_stream.get(), length);
// This avoids the GC from trying to free a stack allocated resource.
if (uc16_string->IsExternalString())
i::Handle<i::ExternalTwoByteString>::cast(uc16_string)
->SetResource(isolate, nullptr);
}
// 1-byte external string
i::Vector<const uint8_t> one_byte_vector =
i::OneByteVector(one_byte_source, static_cast<int>(length));
i::Handle<i::String> one_byte_string =
factory->NewStringFromOneByte(one_byte_vector).ToHandleChecked();
{
TestExternalOneByteResource one_byte_resource(one_byte_source, length);
i::Handle<i::String> ext_one_byte_string(
factory->NewExternalStringFromOneByte(&one_byte_resource)
.ToHandleChecked());
std::unique_ptr<i::Utf16CharacterStream> one_byte_stream(
i::ScannerStream::For(isolate, ext_one_byte_string, 0, length));
TestCloneCharacterStream(one_byte_source, one_byte_stream.get(), length);
// This avoids the GC from trying to free a stack allocated resource.
if (ext_one_byte_string->IsExternalString())
i::Handle<i::ExternalOneByteString>::cast(ext_one_byte_string)
->SetResource(isolate, nullptr);
}
// Relocatinable streams aren't clonable.
{
std::unique_ptr<i::Utf16CharacterStream> string_stream(
i::ScannerStream::For(isolate, one_byte_string, 0, length));
CHECK(!string_stream->can_be_cloned());
i::Handle<i::String> two_byte_string =
factory->NewStringFromTwoByte(two_byte_vector).ToHandleChecked();
std::unique_ptr<i::Utf16CharacterStream> two_byte_string_stream(
i::ScannerStream::For(isolate, two_byte_string, 0, length));
CHECK(!two_byte_string_stream->can_be_cloned());
}
// Chunk sources currently not cloneable.
{
const char* chunks[] = {"1234", "\0"};
ChunkSource chunk_source(chunks);
std::unique_ptr<i::Utf16CharacterStream> one_byte_streaming_stream(
i::ScannerStream::For(&chunk_source,
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
v8::ScriptCompiler::StreamedSource::ONE_BYTE));
CHECK(!one_byte_streaming_stream->can_be_cloned());
std::unique_ptr<i::Utf16CharacterStream> utf8_streaming_stream(
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
i::ScannerStream::For(&chunk_source,
v8::ScriptCompiler::StreamedSource::UTF8));
CHECK(!utf8_streaming_stream->can_be_cloned());
std::unique_ptr<i::Utf16CharacterStream> two_byte_streaming_stream(
i::ScannerStream::For(&chunk_source,
Reland "[Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher" This is a reland of 31228a69c36c65f1f2c9e7f2f5b433ff40bcd5a6 TBR=yangguo@chromium.org,verwaest@chromium.org Original change's description: > [Compile] Refactor BackgroundCompileTask to enable its use by CompilerDispatcher > > Splits background compilation data out of ScriptStreamingData and into > BackgroundCompileTask. Also makes BackgroundCompileTask no longer a sub-class > of ScriptStreamingTask, and instead have ScriptStreamingTask delegate to a > BackgroundCompileTask. > > As part of this change, we now create the CharacterStream on the main thread, > and therefore have to set the (thread-local) runtime_call_stats on the already > created CharacterStream when the BackgroundCompileTask is run on the background > thread. As such, changes to CharacterStream were needed to feed the > runtime_call_stats through appropriately. > > Deprecates Source::GetCachedData and StreamedSource::GetCachedData since they are > no longer used, and the streamed source never has cached data (streaming is > suppressed if cached data is available). Also removes Utf8ChunkedStream which > is dead code. > > BUG=v8:8041, v8:8015 > TBR=yangguo@chromium.org > > Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng > Change-Id: Ifcc723ebf930a1dc01135fcb70929d6168471cb3 > Reviewed-on: https://chromium-review.googlesource.com/1236353 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Toon Verwaest <verwaest@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56182} Bug: v8:8041, v8:8015 Change-Id: Ied5132c537d4c25c6e355f289c2a9cc1f8ff98e9 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Reviewed-on: https://chromium-review.googlesource.com/1242097 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Maya Lekova <mslekova@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#56201}
2018-09-21 21:40:51 +00:00
v8::ScriptCompiler::StreamedSource::TWO_BYTE));
CHECK(!two_byte_streaming_stream->can_be_cloned());
}
}