#!/usr/bin/env python # Copyright 2019 The Chromium Authors. All rights reserved. # Use of this source code is governed by a BSD-style license that can be # found in the LICENSE file. """Extracts an LLD partition from an ELF file."""
import argparse import hashlib import math import os import struct import subprocess import sys import tempfile
def _ComputeNewBuildId(old_build_id, file_path): """
Computes the new build-id from old build-id and file_path.
Args:
old_build_id: Original build-id in bytearray.
file_path: Path to output ELF file.
Returns:
New build id with the same length as |old_build_id|. """
m = hashlib.sha256()
m.update(old_build_id)
m.update(os.path.basename(file_path).encode('utf-8'))
hash_bytes = m.digest() # In case build_id is longer than hash computed, repeat the hash # to the desired length first.
id_size = len(old_build_id)
hash_size = len(hash_bytes) return (hash_bytes * (id_size // hash_size + 1))[:id_size]
def _ExtractPartition(objcopy, input_elf, output_elf, partition): """
Extracts a partition from an ELF file.
For partitions other than main partition, we need to rewrite
the .note.gnu.build-id section so that the build-id remains
unique.
Note:
- `objcopy` does not modify build-id when partitioning the
combined ELF file by default.
- The new build-id is calculated as hash of original build-id and partitioned ELF file name.
Args:
objcopy: Path to objcopy binary.
input_elf: Path to input ELF file.
output_elf: Path to output ELF file.
partition: Partition to extract from combined ELF file. None when
extracting main partition. """ ifnot partition: # main partition # We do not overwrite build-id on main partition to allow the expected # partition build ids to be synthesized given a libchrome.so binary, # if necessary.
subprocess.check_call(
[objcopy, '--extract-main-partition', input_elf, output_elf]) return
with tempfile.TemporaryDirectory() as tempdir:
temp_elf = os.path.join(tempdir, 'obj_without_id.so')
old_build_id_file = os.path.join(tempdir, 'old_build_id')
new_build_id_file = os.path.join(tempdir, 'new_build_id')
# Dump out build-id section and remove original build-id section from # ELF file.
subprocess.check_call([
objcopy, '--extract-partition',
partition, # Note: Not using '--update-section' here as it is not supported # by llvm-objcopy. '--remove-section',
build_id_section, '--dump-section', '{}={}'.format(build_id_section, old_build_id_file),
input_elf,
temp_elf,
])
with open(old_build_id_file, 'rb') as f:
note_content = f.read()
# .note section has following format according to <elf/external.h> # typedef struct { # unsigned char namesz[4]; /* Size of entry's owner string */ # unsigned char descsz[4]; /* Size of the note descriptor */ # unsigned char type[4]; /* Interpretation of the descriptor */ # char name[1]; /* Start of the name+desc data */ # } Elf_External_Note; # `build-id` rewrite is only required on Android platform, # where we have partitioned lib. # Android platform uses little-endian. # <: little-endian # 4x: Skip 4 bytes # L: unsigned long, 4 bytes
descsz, = struct.Struct('<4xL').unpack_from(note_content)
prefix = note_content[:-descsz]
build_id = note_content[-descsz:]
with open(new_build_id_file, 'wb') as f:
f.write(prefix + _ComputeNewBuildId(build_id, output_elf))
# Write back the new build-id section.
subprocess.check_call([
objcopy, '--add-section', '{}={}'.format(build_id_section, new_build_id_file), # Add alloc section flag, or else the section will be removed by # objcopy --strip-all when generating unstripped lib file. '--set-section-flags', '{}={}'.format(build_id_section, 'alloc'),
temp_elf,
output_elf,
])
def main():
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument( '--partition',
help='Name of partition if not the main partition',
metavar='PART')
parser.add_argument( '--objcopy',
required=True,
help='Path to llvm-objcopy binary',
metavar='FILE')
parser.add_argument( '--unstripped-output',
required=True,
help='Unstripped output file',
metavar='FILE')
parser.add_argument( '--stripped-output',
required=True,
help='Stripped output file',
metavar='FILE')
parser.add_argument('--dwp', help='Path to dwp binary', metavar='FILE')
parser.add_argument('input', help='Input file')
args = parser.parse_args()
if args.dwp:
dwp_args = [
args.dwp, '-e', args.unstripped_output, '-o',
args.unstripped_output + '.dwp'
] # Suppress output here because it doesn't seem to be useful. The most # common error is a segfault, which will happen if files are missing.
subprocess.check_output(dwp_args, stderr=subprocess.STDOUT)
Die Informationen auf dieser Webseite wurden
nach bestem Wissen sorgfältig zusammengestellt. Es wird jedoch weder Vollständigkeit, noch Richtigkeit,
noch Qualität der bereit gestellten Informationen zugesichert.
Bemerkung:
Die farbliche Syntaxdarstellung und die Messung sind noch experimentell.