Files
PX4-Autopilot/Tools/msg/templates/ucdr/msg.h.em
T
Jacob DahlandBalduin 791ba46833 refactor(msg): address CDR fields by offset in the generated serializers (#29546)
The generated ucdr_serialize/ucdr_deserialize functions advanced
buf.iterator and buf.offset after every field. Because the memcpy
through buf.iterator may alias the ucdrBuffer itself, the compiler had
to reload the iterator, store it back and update the offset in memory
for each field: four extra instructions per field, in every topic
forwarded by uxrce_dds_client.

Compute the CDR offset of each field in the generator instead, copy
relative to the position at entry, and advance the buffer once by the
topic size. The wire format and the final buffer state are unchanged; a
host test serializing and deserializing random data for all 259 topics
with the old and new headers gives identical bytes, structs and buffer
positions.

px4_fmu-v6x_default: 1929304 -> 1919176 bytes of text, -10128 B.

Assisted-by: Claude:claude-fable-5-1

Co-authored-by: Balduin <balduin@auterion.com>
2026-10-31 12:12:55 -06:00

184 lines
6.5 KiB
Plaintext

@###############################################
@#
@# EmPy template
@#
@###############################################
@# generates CDR serialization & deserialization methods
@#
@# Context:
@# - file_name_in (String) Source file
@# - spec (msggen.MsgSpec) Parsed specification of the .msg file
@# - search_path (dict) search paths for genmsg
@# - topics (List of String) topic names
@###############################################
/****************************************************************************
*
* Copyright (C) 2013-2022 PX4 Development Team. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. 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.
* 3. Neither the name PX4 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.
*
****************************************************************************/
@{
import genmsg.msgs
import re
from px_generate_uorb_topic_helper import * # this is in Tools/
topic = name_snake_case
uorb_struct = '%s_s'%name_snake_case
# get fields, struct size and paddings
def add_fields(msg_fields, name_prefix='', offset=0):
fields = []
for field in msg_fields:
if not field.is_header:
field_size = sizeof_field_type(field)
type_name = field.type
# detect embedded types
sl_pos = type_name.find('/')
if (sl_pos >= 0):
package = type_name[:sl_pos]
type_name = type_name[sl_pos + 1:]
# detect arrays
a_pos = type_name.find('[')
array_size = 1
if (a_pos >= 0):
# field is array
array_size = int(type_name[a_pos+1:-1])
type_name = type_name[:a_pos]
if sl_pos >= 0: # nested type
children_fields = get_children_fields(field.base_type, search_path)
for i in range(array_size):
sub_name_prefix = name_prefix+field.name
if array_size > 1:
sub_name_prefix += '['+str(i)+']'
sub_fields, offset = add_fields(children_fields, sub_name_prefix+'.', offset)
fields.extend(sub_fields)
else:
assert field_size > 0
# note: the maximum alignment for XCDR is 8 and for XCDR2 it is 4
padding = (field_size - (offset % field_size)) & (field_size - 1)
fields.append((type_name, name_prefix+field.name, field_size * array_size, padding))
offset += array_size * field_size + padding
return fields, offset
fields, struct_size = add_fields(spec.parsed_fields())
}@
// auto-generated file
#pragma once
#include <ucdr/microcdr.h>
#include <string.h>
#include <uORB/topics/@(topic).h>
@##############################
@# Includes for dependencies
@##############################
@{
for field in spec.parsed_fields():
if (not field.is_builtin):
if not field.is_header:
(package, name) = genmsg.names.package_resource_name(field.base_type)
package = package or spec.package # convert '' to package
name_snake = re.sub(r'(?<!^)(?=[A-Z])', '_', name).lower()
print('#include <uORB/ucdr/%s.h>'%(name_snake))
}@
static inline constexpr int ucdr_topic_size_@(topic)()
{
return @(struct_size);
}
static inline bool ucdr_serialize_@(topic)(const void* data, ucdrBuffer& buf, int64_t time_offset = 0)
{
const @(uorb_struct)& topic = *static_cast<const @(uorb_struct)*>(data);
// Each field is written at a fixed offset from the current position and the
// buffer is advanced once at the end, so the compiler does not have to reload
// the iterator after every memcpy (which may alias it) and store it back.
uint8_t* const out = buf.iterator;
@{
offset = 0
for field_type, field_name, field_size, padding in fields:
offset += padding # padding bytes are skipped, not written
print('\tstatic_assert(sizeof(topic.{0}) == {1}, "size mismatch");'.format(field_name, field_size))
if field_type == 'uint64' and field_name == 'timestamp':
print('\tconst uint64_t timestamp_adjusted = topic.timestamp + time_offset;')
print('\tmemcpy(out + {1}, &timestamp_adjusted, sizeof(topic.{0}));'.format(field_name, offset))
elif field_type == 'uint64' and field_name == 'timestamp_sample':
print('\tconst uint64_t timestamp_sample_adjusted = topic.timestamp_sample + time_offset;')
print('\tmemcpy(out + {1}, &timestamp_sample_adjusted, sizeof(topic.{0}));'.format(field_name, offset))
else:
print('\tmemcpy(out + {1}, &topic.{0}, sizeof(topic.{0}));'.format(field_name, offset))
offset += field_size
assert offset == struct_size
}@
buf.iterator += @(struct_size);
buf.offset += @(struct_size);
return true;
}
static inline bool ucdr_deserialize_@(topic)(ucdrBuffer& buf, @(uorb_struct)& topic, int64_t time_offset = 0)
{
const uint8_t* const in = buf.iterator;
@{
offset = 0
for field_type, field_name, field_size, padding in fields:
offset += padding # padding bytes are skipped, not read
print('\tstatic_assert(sizeof(topic.{0}) == {1}, "size mismatch");'.format(field_name, field_size))
print('\tmemcpy(&topic.{0}, in + {1}, sizeof(topic.{0}));'.format(field_name, offset))
if field_type == 'uint64' and (field_name == 'timestamp' or field_name == 'timestamp_sample'):
print('\tif (topic.{0} == 0) topic.{0} = hrt_absolute_time();'.format(field_name, field_name))
print('\telse topic.{0} = math::min(topic.{0} - time_offset, hrt_absolute_time());'.format(field_name, field_name))
offset += field_size
assert offset == struct_size
}@
buf.iterator += @(struct_size);
buf.offset += @(struct_size);
return true;
}