Primary Git Repository for the Zephyr Project. Zephyr is a new generation, scalable, optimized, secure RTOS for multiple hardware architectures.
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.
 
 
 
 
 
 

427 lines
15 KiB

# Copyright (c) 2024-2025 The Linux Foundation
# SPDX-License-Identifier: Apache-2.0
import logging
import os
import pickle
import re
import subprocess
import sys
from collections import namedtuple
from pathlib import Path
import list_boards
import list_hardware
import list_shields
import yaml
import zephyr_module
from gen_devicetree_rest import VndLookup
from runners.core import ZephyrBinaryRunner
ZEPHYR_BASE = Path(__file__).parents[2]
ZEPHYR_BINDINGS = ZEPHYR_BASE / "dts/bindings"
EDT_PICKLE_PATHS = [
"zephyr/edt.pickle",
"hello_world/zephyr/edt.pickle" # for board targets using sysbuild
]
RUNNERS_YAML_PATHS = [
"zephyr/runners.yaml",
"hello_world/zephyr/runners.yaml" # for board targets using sysbuild
]
logger = logging.getLogger(__name__)
class DeviceTreeUtils:
_compat_description_cache = {}
@classmethod
def get_first_sentence(cls, text):
"""Extract the first sentence from a text block (typically a node description).
Args:
text: The text to extract the first sentence from.
Returns:
The first sentence found in the text, or the entire text if no sentence
boundary is found.
"""
if not text:
return ""
text = text.replace('\n', ' ')
# Split by double spaces to get paragraphs
paragraphs = text.split(' ')
first_paragraph = paragraphs[0].strip()
# Look for a period followed by a space in the first paragraph
period_match = re.search(r'(.*?)\.(?:\s|$)', first_paragraph)
if period_match:
return period_match.group(1).strip()
# If no period in the first paragraph, return the entire first paragraph
return first_paragraph
@classmethod
def get_cached_description(cls, node):
"""Get the cached description for a devicetree node.
Args:
node: A devicetree node object with matching_compat and description attributes.
Returns:
The cached description for the node's compatible, creating it if needed.
"""
return cls._compat_description_cache.setdefault(
node.matching_compat,
cls.get_first_sentence(node.description)
)
def guess_file_from_patterns(directory, patterns, name, extensions):
for pattern in patterns:
for ext in extensions:
matching_file = next(directory.glob(pattern.format(name=name, ext=ext)), None)
if matching_file:
return matching_file
return None
def guess_image(board_or_shield):
img_exts = ["jpg", "jpeg", "webp", "png"]
patterns = [
"**/{name}.{ext}",
"**/*{name}*.{ext}",
"**/*.{ext}",
]
img_file = guess_file_from_patterns(
board_or_shield.dir, patterns, board_or_shield.name, img_exts
)
return (img_file.relative_to(ZEPHYR_BASE)).as_posix() if img_file else None
def guess_doc_page(board_or_shield):
patterns = [
"doc/index.{ext}",
"**/{name}.{ext}",
"**/*{name}*.{ext}",
"**/*.{ext}",
]
doc_file = guess_file_from_patterns(
board_or_shield.dir, patterns, board_or_shield.name, ["rst"]
)
return doc_file
def gather_board_build_info(twister_out_dir):
"""Gather EDT objects and runners info for each board from twister output directory.
Args:
twister_out_dir: Path object pointing to twister output directory
Returns:
A tuple of two dictionaries:
- A dictionary mapping board names to a dictionary of board targets and their EDT.
objects.
The structure is: {board_name: {board_target: edt_object}}
- A dictionary mapping board names to a dictionary of board targets and their runners
info.
The structure is: {board_name: {board_target: runners_info}}
"""
board_devicetrees = {}
board_runners = {}
if not twister_out_dir.exists():
return board_devicetrees, board_runners
# Find all build_info.yml files in twister-out
build_info_files = list(twister_out_dir.glob("*/**/build_info.yml"))
for build_info_file in build_info_files:
edt_pickle_file = None
for pickle_path in EDT_PICKLE_PATHS:
maybe_file = build_info_file.parent / pickle_path
if maybe_file.exists():
edt_pickle_file = maybe_file
break
if not edt_pickle_file:
continue
runners_yaml_file = None
for runners_yaml_path in RUNNERS_YAML_PATHS:
maybe_file = build_info_file.parent / runners_yaml_path
if maybe_file.exists():
runners_yaml_file = maybe_file
break
try:
with open(build_info_file) as f:
build_info = yaml.safe_load(f)
board_info = build_info.get('cmake', {}).get('board', {})
board_name = board_info.get('name')
qualifier = board_info.get('qualifiers', '')
revision = board_info.get('revision', '')
board_target = board_name
if revision != '':
board_target = f"{board_target}@{revision}"
if qualifier:
board_target = f"{board_target}/{qualifier}"
with open(edt_pickle_file, 'rb') as f:
edt = pickle.load(f)
board_devicetrees.setdefault(board_name, {})[board_target] = edt
if runners_yaml_file:
with open(runners_yaml_file) as f:
runners_yaml = yaml.safe_load(f)
board_runners.setdefault(board_name, {})[board_target] = (
runners_yaml
)
except Exception as e:
logger.error(f"Error processing build info file {build_info_file}: {e}")
return board_devicetrees, board_runners
def run_twister_cmake_only(outdir, vendor_filter):
"""Run twister in cmake-only mode to generate build info files.
Args:
outdir: Directory where twister should output its files
vendor_filter: Limit build info to boards from listed vendors
"""
twister_cmd = [
sys.executable,
f"{ZEPHYR_BASE}/scripts/twister",
"-T", "samples/hello_world/",
"-M",
*[arg for path in EDT_PICKLE_PATHS for arg in ('--keep-artifacts', path)],
*[arg for path in RUNNERS_YAML_PATHS for arg in ('--keep-artifacts', path)],
"--cmake-only",
"--outdir", str(outdir),
]
if vendor_filter:
for vendor in vendor_filter:
twister_cmd += ["--vendor", vendor]
else:
twister_cmd += ["--all"]
minimal_env = {
'PATH': os.environ.get('PATH', ''),
'ZEPHYR_BASE': str(ZEPHYR_BASE),
'HOME': os.environ.get('HOME', ''),
'PYTHONPATH': os.environ.get('PYTHONPATH', '')
}
try:
subprocess.run(twister_cmd, check=True, cwd=ZEPHYR_BASE, env=minimal_env)
except subprocess.CalledProcessError as e:
logger.warning(f"Failed to run Twister, list of hw features might be incomplete.\n{e}")
def get_catalog(generate_hw_features=False, hw_features_vendor_filter=None):
"""Get the board catalog.
Args:
generate_hw_features: If True, run twister to generate hardware features information.
hw_features_vendor_filter: If generate_hw_features is True, limit hardware feature
information generation to boards from this list of vendors.
"""
import tempfile
vnd_lookup = VndLookup(ZEPHYR_BASE / "dts/bindings/vendor-prefixes.txt", [])
module_settings = {
"arch_root": [ZEPHYR_BASE],
"board_root": [ZEPHYR_BASE],
"soc_root": [ZEPHYR_BASE],
}
for module in zephyr_module.parse_modules(ZEPHYR_BASE):
for key in module_settings:
root = module.meta.get("build", {}).get("settings", {}).get(key)
if root is not None:
module_settings[key].append(Path(module.project) / root)
Args = namedtuple("args", ["arch_roots", "board_roots", "soc_roots", "board_dir", "board"])
args_find_boards = Args(
arch_roots=module_settings["arch_root"],
board_roots=module_settings["board_root"],
soc_roots=module_settings["soc_root"],
board_dir=[],
board=None,
)
boards = list_boards.find_v2_boards(args_find_boards)
shields = list_shields.find_shields(args_find_boards)
systems = list_hardware.find_v2_systems(args_find_boards)
board_catalog = {}
shield_catalog = {}
board_devicetrees = {}
board_runners = {}
if generate_hw_features:
logger.info("Running twister in cmake-only mode to get Devicetree files for all boards")
with tempfile.TemporaryDirectory() as tmp_dir:
run_twister_cmake_only(tmp_dir, hw_features_vendor_filter)
board_devicetrees, board_runners = gather_board_build_info(Path(tmp_dir))
else:
logger.info("Skipping generation of supported hardware features.")
for board in boards.values():
vendor = board.vendor or "others"
socs = {soc.name for soc in board.socs}
full_name = board.full_name or board.name
doc_page = guess_doc_page(board)
supported_features = {}
# Use pre-gathered build info and DTS files
if board.name in board_devicetrees:
for board_target, edt in board_devicetrees[board.name].items():
features = {}
for node in edt.nodes:
if node.binding_path is None:
continue
binding_path = Path(node.binding_path)
is_custom_binding = False
if binding_path.is_relative_to(ZEPHYR_BINDINGS):
binding_type = binding_path.relative_to(ZEPHYR_BINDINGS).parts[0]
else:
binding_type = "misc"
is_custom_binding = True
if node.matching_compat is None:
continue
# skip "zephyr,xxx" compatibles
if node.matching_compat.startswith("zephyr,"):
continue
description = DeviceTreeUtils.get_cached_description(node)
title = node.title
filename = node.filename
lineno = node.lineno
locations = set()
if Path(filename).is_relative_to(ZEPHYR_BASE):
filename = Path(filename).relative_to(ZEPHYR_BASE)
if filename.parts[0] == "boards":
locations.add("board")
else:
locations.add("soc")
existing_feature = features.get(binding_type, {}).get(
node.matching_compat
)
node_info = {
"filename": str(filename),
"lineno": lineno,
"dts_path": Path(node.filename),
"binding_path": Path(node.binding_path),
}
node_list_key = "okay_nodes" if node.status == "okay" else "disabled_nodes"
if existing_feature:
locations.update(existing_feature["locations"])
existing_feature.setdefault(node_list_key, []).append(node_info)
continue
feature_data = {
"description": description,
"title": title,
"custom_binding": is_custom_binding,
"locations": locations,
"okay_nodes": [],
"disabled_nodes": [],
}
feature_data[node_list_key].append(node_info)
features.setdefault(binding_type, {})[node.matching_compat] = feature_data
# Store features for this specific target
supported_features[board_target] = features
board_runner_info = {}
if board.name in board_runners:
# Assume all board targets have the same runners so only consider the runners
# for the first board target.
r = list(board_runners[board.name].values())[0]
board_runner_info["runners"] = r.get("runners")
board_runner_info["flash-runner"] = r.get("flash-runner")
board_runner_info["debug-runner"] = r.get("debug-runner")
# Grab all the twister files for this board and use them to figure out all the archs it
# supports.
archs = set()
pattern = f"{board.name}*.yaml"
for twister_file in board.dir.glob(pattern):
try:
with open(twister_file) as f:
board_data = yaml.safe_load(f)
archs.add(board_data.get("arch"))
except Exception as e:
logger.error(f"Error parsing twister file {twister_file}: {e}")
if doc_page and doc_page.is_relative_to(ZEPHYR_BASE):
doc_page_path = doc_page.relative_to(ZEPHYR_BASE).as_posix()
else:
doc_page_path = None
board_catalog[board.name] = {
"name": board.name,
"full_name": full_name,
"doc_page": doc_page_path,
"vendor": vendor,
"archs": list(archs),
"socs": list(socs),
"revision_default": board.revision_default,
"supported_features": supported_features,
"image": guess_image(board),
# runners
"supported_runners": board_runner_info.get("runners", []),
"flash_runner": board_runner_info.get("flash-runner", ""),
"debug_runner": board_runner_info.get("debug-runner", ""),
}
socs_hierarchy = {}
for soc in systems.get_socs():
family = soc.family or "<no family>"
series = soc.series or "<no series>"
socs_hierarchy.setdefault(family, {}).setdefault(series, []).append(soc.name)
available_runners = {}
for runner in ZephyrBinaryRunner.get_runners():
available_runners[runner.name()] = {
"name": runner.name(),
"commands": runner.capabilities().commands,
}
for shield in shields:
doc_page = guess_doc_page(shield)
if doc_page and doc_page.is_relative_to(ZEPHYR_BASE):
doc_page_path = doc_page.relative_to(ZEPHYR_BASE).as_posix()
else:
doc_page_path = None
shield_catalog[shield.name] = {
"name": shield.name,
"full_name": shield.full_name or shield.name,
"vendor": shield.vendor or "others",
"doc_page": doc_page_path,
"image": guess_image(shield),
}
return {
"boards": board_catalog,
"shields": shield_catalog,
"vendors": {**vnd_lookup.vnd2vendor, "others": "Other/Unknown"},
"socs": socs_hierarchy,
"runners": available_runners,
}