node/tools/js2c.py
Keyhan Vakil 190596c189
src: register external references for source code
Currently we use external strings for internalized builtin source code.
However when a snapshot is taken, any external string whose resource
is not registered is flattened into a SeqString (see ref). The
result is that module source code stored in the snapshot does not
use external strings after deserialization. This patch registers an
external string resource for each internalized builtin's source. The
savings are substantial: ~1.9 MB of heap memory per isolate, or ~44%
of an otherwise empty isolate's heap usage:

```console
$ node --expose-gc -p 'gc(),process.memoryUsage().heapUsed'
4190968
$ ./node --expose-gc -p 'gc(),process.memoryUsage().heapUsed'
2327536
```

The savings can be even higher for user snapshots which may include
more internal modules.

The existing UnionBytes implementation was ill-suited, because it only
created an external string resource when ToStringChecked was called,
but we need to register the external string resources before the
isolate even exists. We change UnionBytes to no longer own the data,
and shift ownership of the data to a new external resource class
called StaticExternalByteResource.  StaticExternalByteResource are
either statically allocated (for internalized builtin code) or owned
by the static `externalized_builtin_sources` map, so they will only be
destructed when static resources are destructed. We change JS2C to emit
statements to register a string resource for each internalized builtin.

Refs: d2c8fbe9cc/src/snapshot/serializer.cc (L633)
PR-URL: https://github.com/nodejs/node/pull/47055
Reviewed-By: Joyee Cheung <joyeec9h3@gmail.com>
Reviewed-By: Chengzhong Wu <legendecas@gmail.com>
2023-05-12 11:57:20 +02:00

270 lines
7.9 KiB
Python
Executable file

#!/usr/bin/env python
#
# Copyright 2006-2008 the V8 project authors. All rights reserved.
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are
# met:
#
# * Redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer.
# * Redistributions in binary form must reproduce the above
# copyright notice, this list of conditions and the following
# disclaimer in the documentation and/or other materials provided
# with the distribution.
# * Neither the name of Google Inc. nor the names of its
# contributors may be used to endorse or promote products derived
# from this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
"""
This is a utility for converting JavaScript source code into uint16_t[],
that are used for embedding JavaScript code into the Node.js binary.
"""
import argparse
import os
import re
import functools
import codecs
import utils
def ReadFile(filename):
if is_verbose:
print(filename)
with codecs.open(filename, "r", "utf-8") as f:
lines = f.read()
return lines
TEMPLATE = """
#include "env-inl.h"
#include "node_builtins.h"
#include "node_external_reference.h"
#include "node_internals.h"
namespace node {{
namespace builtins {{
{0}
namespace {{
const ThreadsafeCopyOnWrite<BuiltinSourceMap> global_source_map {{
BuiltinSourceMap{{ {1} }}
}};
}}
void BuiltinLoader::LoadJavaScriptSource() {{
source_ = global_source_map;
}}
void RegisterExternalReferencesForInternalizedBuiltinCode(
ExternalReferenceRegistry* registry) {{
{2}
}}
UnionBytes BuiltinLoader::GetConfig() {{
return UnionBytes(&{3});
}}
}} // namespace builtins
}} // namespace node
"""
ONE_BYTE_STRING = """
static const uint8_t {0}[] = {{
{1}
}};
static StaticExternalOneByteResource {2}({0}, {3}, nullptr);
"""
TWO_BYTE_STRING = """
static const uint16_t {0}[] = {{
{1}
}};
static StaticExternalTwoByteResource {2}({0}, {3}, nullptr);
"""
INITIALIZER = '{{"{0}", UnionBytes(&{1}) }},'
REGISTRATION = 'registry->Register(&{0});'
CONFIG_GYPI_ID = 'config_raw'
CONFIG_GYPI_RESOURCE_ID = 'config_resource'
SLUGGER_RE = re.compile(r'[.\-/]')
is_verbose = False
def GetDefinition(var, source, resource_var, step=30):
template = ONE_BYTE_STRING
code_points = [ord(c) for c in source]
if any(c > 127 for c in code_points):
template = TWO_BYTE_STRING
# Treat non-ASCII as UTF-8 and encode as UTF-16 Little Endian.
encoded_source = bytearray(source, 'utf-16le')
code_points = [
encoded_source[i] + (encoded_source[i + 1] * 256)
for i in range(0, len(encoded_source), 2)
]
# For easier debugging, align to the common 3 char for code-points.
elements_s = ['%3s' % x for x in code_points]
# Put no more then `step` code-points in a line.
slices = [elements_s[i:i + step] for i in range(0, len(elements_s), step)]
lines = [','.join(s) for s in slices]
array_content = ',\n'.join(lines)
length = len(code_points)
definition = template.format(var, array_content, resource_var, length)
return definition
def AddModule(filename, definitions, initializers, registrations):
code = ReadFile(filename)
name = NormalizeFileName(filename)
slug = SLUGGER_RE.sub('_', name)
var = slug + '_raw'
resource_var = slug + '_resource'
definition = GetDefinition(var, code, resource_var)
initializer = INITIALIZER.format(name, resource_var)
registration = REGISTRATION.format(resource_var)
definitions.append(definition)
initializers.append(initializer)
registrations.append(registration)
def NormalizeFileName(filename):
split = filename.split('/')
if split[0] == 'deps':
split = ['internal'] + split
else: # `lib/**/*.js` so drop the 'lib' part
split = split[1:]
if len(split):
filename = '/'.join(split)
return os.path.splitext(filename)[0]
def JS2C(source_files, target):
# Build source code lines
definitions = []
initializers = []
registrations = []
for filename in source_files['.js']:
AddModule(filename, definitions, initializers, registrations)
for filename in source_files['.mjs']:
AddModule(filename, definitions, initializers, registrations)
config_def = handle_config_gypi(source_files['config.gypi'])
definitions.append(config_def)
# Emit result
definitions = ''.join(definitions)
initializers = '\n '.join(initializers)
registrations = '\n '.join(registrations)
out = TEMPLATE.format(definitions, initializers,
registrations, CONFIG_GYPI_RESOURCE_ID)
write_if_chaged(out, target)
def handle_config_gypi(config_filename):
# if its a gypi file we're going to want it as json
# later on anyway, so get it out of the way now
config = ReadFile(config_filename)
config = jsonify(config)
config_def = GetDefinition(CONFIG_GYPI_ID, config, CONFIG_GYPI_RESOURCE_ID)
return config_def
def jsonify(config):
# 1. string comments
config = re.sub(r'#.*?\n', '', config)
# 2. join multiline strings
config = re.sub(r"'$\s+'", '', config, flags=re.M)
# 3. normalize string literals from ' into "
config = re.sub('\'', '"', config)
# 4. turn pseudo-booleans strings into Booleans
config = re.sub('"true"', 'true', config)
config = re.sub('"false"', 'false', config)
return config
def write_if_chaged(content, target):
if os.path.exists(target):
with open(target, 'rt') as existing:
old_content = existing.read()
else:
old_content = ''
if old_content == content:
os.utime(target, None)
return
with open(target, "wt") as output:
output.write(content)
def SourceFileByExt(files_by_ext, filename):
"""
:type files_by_ext: dict
:type filename: str
:rtype: dict
"""
ext = os.path.splitext(filename)[-1]
files_by_ext.setdefault(ext, []).append(filename)
return files_by_ext
def main():
parser = argparse.ArgumentParser(
description='Convert code files into `uint16_t[]`s',
fromfile_prefix_chars='@'
)
parser.add_argument('--target', help='output file')
parser.add_argument(
'--directory',
default=None,
help='input file directory')
parser.add_argument(
'--root',
default=None,
help='root directory containing the sources')
parser.add_argument('--verbose', action='store_true', help='output file')
parser.add_argument('sources', nargs='*', help='input files')
options = parser.parse_args()
global is_verbose
is_verbose = options.verbose
sources = options.sources
if options.root is not None:
os.chdir(options.root)
if options.directory is not None:
js_files = utils.SearchFiles(options.directory, 'js')
mjs_files = utils.SearchFiles(options.directory, 'mjs')
sources = js_files + mjs_files + options.sources
source_files = functools.reduce(SourceFileByExt, sources, {})
# Should have exactly 3 types: `.js`, `.mjs` and `.gypi`
assert len(source_files) == 3
# Currently config.gypi is the only `.gypi` file allowed
assert len(source_files['.gypi']) == 1
assert os.path.basename(source_files['.gypi'][0]) == 'config.gypi'
source_files['config.gypi'] = source_files.pop('.gypi')[0]
JS2C(source_files, options.target)
if __name__ == "__main__":
main()