You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

160 lines
4.5 KiB

#!/usr/bin/env python
# Copyright 2017, The Android Open Source Project
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from __future__ import print_function
try:
from os import fstat, stat, remove
from sys import exit
from argparse import ArgumentParser, FileType
from ctypes import sizeof, Structure, c_char, c_int
from struct import pack, calcsize
import zlib
except Exception as e:
print("some module is needed:" + str(e))
exit(-1)
dt_head_info_fmt = '4sII'
dt_entry_fmt = 'Q4I2Q'
dtimg_version = 1
dtb_count = 1
def write32(output, value):
output.write(chr(value & 255)) ; value=value // 256
output.write(chr(value & 255)) ; value=value // 256
output.write(chr(value & 255)) ; value=value // 256
output.write(chr(value & 255))
def compress(filename, input, output):
output.write('\037\213\010')
output.write(chr(0))
statval = stat(filename)
write32(output, 0)
output.write('\002')
output.write('\003')
crcval = zlib.crc32("")
compobj = zlib.compressobj(9, zlib.DEFLATED, -zlib.MAX_WBITS,
zlib.DEF_MEM_LEVEL, 0)
while True:
data = input.read(1024)
if data == "":
break
crcval = zlib.crc32(data, crcval)
output.write(compobj.compress(data))
output.write(compobj.flush())
write32(output, crcval)
write32(output, statval.st_size)
def dtb_compress(dtb_file):
try:
outputname = dtb_file + '.gz'
input = open(dtb_file, 'rb')
output = open(outputname, 'wb')
compress(dtb_file, input, output)
input.close()
output.close()
except Exception as e:
print('dtb_compress error:' + str(e))
exit(-1)
return outputname
class dt_head_info(Structure):
_fields_ = [('magic', c_char * 4),
('version', c_int),
('dt_count', c_int)]
class dt_entry_t(Structure):
_fields_ = [('dtb_size', c_int),
('dtb_offset', c_int)]
def align_page_size(offset, pagesize):
return (pagesize - (offset % pagesize))
def write_head_info(head_info, args):
args.output.write(pack(dt_head_info_fmt,
head_info.magic,
head_info.version,
head_info.dt_count))
def write_dtb_entry_t(dt_entry, args):
args.output.write(pack(dt_entry_fmt,
0, # reserved
dt_entry.dtb_size,
0, # reserved
dt_entry.dtb_offset,
0, # reserved
0, # reserved
0)) # reserved
def write_padding(args, padding):
for i in range(0, padding):
args.output.write('\x00')
def write_dtb(args):
dtb_file = args.dtb
out_dtb = dtb_file
if args.compress == True:
out_dtb = dtb_compress(dtb_file)
try:
dtb_offset = calcsize(dt_head_info_fmt) + \
calcsize(dt_entry_fmt) + \
4
padding = align_page_size(dtb_offset, args.pagesize)
dtb_size = stat(out_dtb).st_size
dtb_size_padding = align_page_size(dtb_size, args.pagesize)
dt_entry = dt_entry_t(dtb_size + dtb_size_padding,
dtb_offset + padding)
write_dtb_entry_t(dt_entry, args)
args.output.write(pack('I', 0)) # SUCCESS code number
write_padding(args, padding)
with open(out_dtb, 'rb') as dtb_fd:
args.output.write(dtb_fd.read(dtb_size))
write_padding(args, dtb_size_padding)
except Exception as e:
print('write dtb error:' + str(e))
exit(-1)
def clean_gz_file(args):
try:
if args.compress != True:
return
remove(args.dtb + '.gz')
except Exception as e:
print('clean gz file error:' + str(e))
exit(-1)
def parse_cmdline():
parser = ArgumentParser()
parser.add_argument('-c', '--compress', help='compress dtb or not',
action='store_true')
parser.add_argument('-d', '--dtb', help='path to the dtb', type=str,
required=True)
parser.add_argument('-s', '--pagesize', help='align page size',
type=int, choices=[2**i for i in range(11,15)],
default=2048)
parser.add_argument('-o', '--output', help='output file name',
type=FileType('wb'), required=True)
return parser.parse_args()
def main():
args = parse_cmdline()
dtimg_head_info = dt_head_info('HSDT', dtimg_version, dtb_count)
write_head_info(dtimg_head_info, args)
write_dtb(args)
clean_gz_file(args)
if __name__ == '__main__':
main()