u-boot/tools/buildman/builder.py

1797 lines
73 KiB
Python
Raw Normal View History

# SPDX-License-Identifier: GPL-2.0+
# Copyright (c) 2013 The Chromium OS Authors.
#
# Bloat-o-meter code used here Copyright 2004 Matt Mackall <mpm@selenic.com>
#
import collections
from datetime import datetime, timedelta
import glob
import os
import re
import queue
import shutil
import signal
import string
import sys
import threading
import time
from buildman import builderthread
from buildman import toolchain
from patman import command
from patman import gitutil
from patman import terminal
from patman.terminal import tprint
buildman: Detect Kconfig loops Hex and int Kconfig options are supposed to have defaults. This is so we can configure U-Boot without having to enter particular values for the items that don't have specific values in the board's defconfig file. If this rule is not followed, then introducing a new Kconfig can produce a loop like this: Break things (BREAK_ME) [] (NEW) Error in reading or end of file. Break things (BREAK_ME) [] (NEW) Error in reading or end of file. The continues forever since buildman passes /dev/null to 'conf', and the build system just tries again. Eventually there is so much output that buildman runs out of memory. We can detect this situation by looking for a symbol (like 'BREAK_ME') which has no default (the '[]' above) and is marked as new. If this appears multiple times in the output, we know something is wrong. Add a filter function for the output which detects this situation. Allow it to return True to terminate the process. Implement this termination in cros_subprocess. With this we get a nice message: buildman --board sandbox -T0 Building current source for 1 boards (0 threads, 32 jobs per thread) sandbox: w+ sandbox +.config:66:warning: symbol value '' invalid for BREAK_ME + +Error in reading or end of file. +make[3]: *** [scripts/kconfig/Makefile:75: syncconfig] Terminated +make[2]: *** [Makefile:569: syncconfig] Terminated +make: *** [Makefile:177: sub-make] Terminated +(** did you define an int/hex Kconfig with no default? **) Signed-off-by: Simon Glass <sjg@chromium.org>
2021-10-20 03:43:24 +00:00
# This indicates an new int or hex Kconfig property with no default
# It hangs the build since the 'conf' tool cannot proceed without valid input.
#
# We get a repeat sequence of something like this:
# >>
# Break things (BREAK_ME) [] (NEW)
# Error in reading or end of file.
# <<
# which indicates that BREAK_ME has an empty default
RE_NO_DEFAULT = re.compile(b'\((\w+)\) \[] \(NEW\)')
"""
Theory of Operation
Please see README for user documentation, and you should be familiar with
that before trying to make sense of this.
Buildman works by keeping the machine as busy as possible, building different
commits for different boards on multiple CPUs at once.
The source repo (self.git_dir) contains all the commits to be built. Each
thread works on a single board at a time. It checks out the first commit,
configures it for that board, then builds it. Then it checks out the next
commit and builds it (typically without re-configuring). When it runs out
of commits, it gets another job from the builder and starts again with that
board.
Clearly the builder threads could work either way - they could check out a
commit and then built it for all boards. Using separate directories for each
commit/board pair they could leave their build product around afterwards
also.
The intent behind building a single board for multiple commits, is to make
use of incremental builds. Since each commit is built incrementally from
the previous one, builds are faster. Reconfiguring for a different board
removes all intermediate object files.
Many threads can be working at once, but each has its own working directory.
When a thread finishes a build, it puts the output files into a result
directory.
The base directory used by buildman is normally '../<branch>', i.e.
a directory higher than the source repository and named after the branch
being built.
Within the base directory, we have one subdirectory for each commit. Within
that is one subdirectory for each board. Within that is the build output for
that commit/board combination.
Buildman also create working directories for each thread, in a .bm-work/
subdirectory in the base dir.
As an example, say we are building branch 'us-net' for boards 'sandbox' and
'seaboard', and say that us-net has two commits. We will have directories
like this:
us-net/ base directory
01_g4ed4ebc_net--Add-tftp-speed-/
sandbox/
u-boot.bin
seaboard/
u-boot.bin
02_g4ed4ebc_net--Check-tftp-comp/
sandbox/
u-boot.bin
seaboard/
u-boot.bin
.bm-work/
00/ working directory for thread 0 (contains source checkout)
build/ build output
01/ working directory for thread 1
build/ build output
...
u-boot/ source directory
.git/ repository
"""
"""Holds information about a particular error line we are outputing
char: Character representation: '+': error, '-': fixed error, 'w+': warning,
'w-' = fixed warning
boards: List of Board objects which have line in the error/warning output
errline: The text of the error line
"""
ErrLine = collections.namedtuple('ErrLine', 'char,boards,errline')
# Possible build outcomes
OUTCOME_OK, OUTCOME_WARNING, OUTCOME_ERROR, OUTCOME_UNKNOWN = list(range(4))
# Translate a commit subject into a valid filename (and handle unicode)
trans_valid_chars = str.maketrans('/: ', '---')
BASE_CONFIG_FILENAMES = [
'u-boot.cfg', 'u-boot-spl.cfg', 'u-boot-tpl.cfg'
]
EXTRA_CONFIG_FILENAMES = [
'.config', '.config-spl', '.config-tpl',
'autoconf.mk', 'autoconf-spl.mk', 'autoconf-tpl.mk',
'autoconf.h', 'autoconf-spl.h','autoconf-tpl.h',
]
class Config:
"""Holds information about configuration settings for a board."""
def __init__(self, config_filename, target):
self.target = target
self.config = {}
for fname in config_filename:
self.config[fname] = {}
def Add(self, fname, key, value):
self.config[fname][key] = value
def __hash__(self):
val = 0
for fname in self.config:
for key, value in self.config[fname].items():
print(key, value)
val = val ^ hash(key) & hash(value)
return val
class Environment:
"""Holds information about environment variables for a board."""
def __init__(self, target):
self.target = target
self.environment = {}
def Add(self, key, value):
self.environment[key] = value
class Builder:
"""Class for building U-Boot for a particular commit.
Public members: (many should ->private)
already_done: Number of builds already completed
base_dir: Base directory to use for builder
checkout: True to check out source, False to skip that step.
This is used for testing.
col: terminal.Color() object
count: Number of commits to build
do_make: Method to call to invoke Make
fail: Number of builds that failed due to error
force_build: Force building even if a build already exists
force_config_on_failure: If a commit fails for a board, disable
incremental building for the next commit we build for that
board, so that we will see all warnings/errors again.
force_build_failures: If a previously-built build (i.e. built on
a previous run of buildman) is marked as failed, rebuild it.
git_dir: Git directory containing source repository
num_jobs: Number of jobs to run at once (passed to make as -j)
num_threads: Number of builder threads to run
out_queue: Queue of results to process
re_make_err: Compiled regular expression for ignore_lines
queue: Queue of jobs to run
threads: List of active threads
toolchains: Toolchains object to use for building
upto: Current commit number we are building (0.count-1)
warned: Number of builds that produced at least one warning
force_reconfig: Reconfigure U-Boot on each comiit. This disables
incremental building, where buildman reconfigures on the first
commit for a baord, and then just does an incremental build for
the following commits. In fact buildman will reconfigure and
retry for any failing commits, so generally the only effect of
this option is to slow things down.
in_tree: Build U-Boot in-tree instead of specifying an output
directory separate from the source code. This option is really
only useful for testing in-tree builds.
work_in_output: Use the output directory as the work directory and
don't write to a separate output directory.
thread_exceptions: List of exceptions raised by thread jobs
Private members:
_base_board_dict: Last-summarised Dict of boards
_base_err_lines: Last-summarised list of errors
_base_warn_lines: Last-summarised list of warnings
_build_period_us: Time taken for a single build (float object).
_complete_delay: Expected delay until completion (timedelta)
_next_delay_update: Next time we plan to display a progress update
(datatime)
_show_unknown: Show unknown boards (those not built) in summary
_start_time: Start time for the build
_timestamps: List of timestamps for the completion of the last
last _timestamp_count builds. Each is a datetime object.
_timestamp_count: Number of timestamps to keep in our list.
_working_dir: Base working directory containing all threads
_single_builder: BuilderThread object for the singer builder, if
threading is not being used
buildman: Detect Kconfig loops Hex and int Kconfig options are supposed to have defaults. This is so we can configure U-Boot without having to enter particular values for the items that don't have specific values in the board's defconfig file. If this rule is not followed, then introducing a new Kconfig can produce a loop like this: Break things (BREAK_ME) [] (NEW) Error in reading or end of file. Break things (BREAK_ME) [] (NEW) Error in reading or end of file. The continues forever since buildman passes /dev/null to 'conf', and the build system just tries again. Eventually there is so much output that buildman runs out of memory. We can detect this situation by looking for a symbol (like 'BREAK_ME') which has no default (the '[]' above) and is marked as new. If this appears multiple times in the output, we know something is wrong. Add a filter function for the output which detects this situation. Allow it to return True to terminate the process. Implement this termination in cros_subprocess. With this we get a nice message: buildman --board sandbox -T0 Building current source for 1 boards (0 threads, 32 jobs per thread) sandbox: w+ sandbox +.config:66:warning: symbol value '' invalid for BREAK_ME + +Error in reading or end of file. +make[3]: *** [scripts/kconfig/Makefile:75: syncconfig] Terminated +make[2]: *** [Makefile:569: syncconfig] Terminated +make: *** [Makefile:177: sub-make] Terminated +(** did you define an int/hex Kconfig with no default? **) Signed-off-by: Simon Glass <sjg@chromium.org>
2021-10-20 03:43:24 +00:00
_terminated: Thread was terminated due to an error
_restarting_config: True if 'Restart config' is detected in output
"""
class Outcome:
"""Records a build outcome for a single make invocation
Public Members:
rc: Outcome value (OUTCOME_...)
err_lines: List of error lines or [] if none
sizes: Dictionary of image size information, keyed by filename
- Each value is itself a dictionary containing
values for 'text', 'data' and 'bss', being the integer
size in bytes of each section.
func_sizes: Dictionary keyed by filename - e.g. 'u-boot'. Each
value is itself a dictionary:
key: function name
value: Size of function in bytes
config: Dictionary keyed by filename - e.g. '.config'. Each
value is itself a dictionary:
key: config name
value: config value
environment: Dictionary keyed by environment variable, Each
value is the value of environment variable.
"""
def __init__(self, rc, err_lines, sizes, func_sizes, config,
environment):
self.rc = rc
self.err_lines = err_lines
self.sizes = sizes
self.func_sizes = func_sizes
self.config = config
self.environment = environment
def __init__(self, toolchains, base_dir, git_dir, num_threads, num_jobs,
gnu_make='make', checkout=True, show_unknown=True, step=1,
buildman: allow more incremental building One use-case for buildman is to continually run it interactively after each small step in a large refactoring operation. This gives more immediate feedback than making a number of commits and then going back and testing them. For this to work well, buildman needs to be extremely fast. At present, a couple issues prevent it being as fast as it could be: 1) Each time buildman runs "make %_defconfig", it runs "make mrproper" first. This throws away all previous build results, requiring a from-scratch build. Optionally avoiding this would speed up the build, at the cost of potentially causing or missing some build issues. 2) A build tree is created per thread rather than per board. When a thread switches between building different boards, this often causes many files to be rebuilt due to changing config options. Using a separate build tree for each board would avoid this. This does put more strain on the system's disk cache, but it is worth it on my system at least. This commit adds two command-line options to implement the changes described above; -I ("--incremental") turns of "make mrproper" and -P ("--per-board-out-dir") creats a build directory per board rather than per thread. Tested: ./tools/buildman/buildman.py tegra ./tools/buildman/buildman.py -I -P tegra ./tools/buildman/buildman.py -b tegra_dev tegra ./tools/buildman/buildman.py -b tegra_dev -I -P tegra ... each once after deleting the buildman result/work directory, and once "incrementally" after a previous identical invocation. Signed-off-by: Stephen Warren <swarren@nvidia.com> Reviewed-by: Tom Rini <trini@konsulko.com> Acked-by: Simon Glass <sjg@chromium.org> # v1 Tested-by: Simon Glass <sjg@chromium.org> # v1 Acked-by: Simon Glass <sjg@chromium.org>
2016-04-11 16:48:44 +00:00
no_subdirs=False, full_path=False, verbose_build=False,
mrproper=False, per_board_out_dir=False,
config_only=False, squash_config_y=False,
warnings_as_errors=False, work_in_output=False,
test_thread_exceptions=False, adjust_cfg=None):
"""Create a new Builder object
Args:
toolchains: Toolchains object to use for building
base_dir: Base directory to use for builder
git_dir: Git directory containing source repository
num_threads: Number of builder threads to run
num_jobs: Number of jobs to run at once (passed to make as -j)
gnu_make: the command name of GNU Make.
checkout: True to check out source, False to skip that step.
This is used for testing.
show_unknown: Show unknown boards (those not built) in summary
step: 1 to process every commit, n to process every nth commit
no_subdirs: Don't create subdirectories when building current
source for a single board
full_path: Return the full path in CROSS_COMPILE and don't set
PATH
verbose_build: Run build with V=1 and don't use 'make -s'
mrproper: Always run 'make mrproper' when configuring
buildman: allow more incremental building One use-case for buildman is to continually run it interactively after each small step in a large refactoring operation. This gives more immediate feedback than making a number of commits and then going back and testing them. For this to work well, buildman needs to be extremely fast. At present, a couple issues prevent it being as fast as it could be: 1) Each time buildman runs "make %_defconfig", it runs "make mrproper" first. This throws away all previous build results, requiring a from-scratch build. Optionally avoiding this would speed up the build, at the cost of potentially causing or missing some build issues. 2) A build tree is created per thread rather than per board. When a thread switches between building different boards, this often causes many files to be rebuilt due to changing config options. Using a separate build tree for each board would avoid this. This does put more strain on the system's disk cache, but it is worth it on my system at least. This commit adds two command-line options to implement the changes described above; -I ("--incremental") turns of "make mrproper" and -P ("--per-board-out-dir") creats a build directory per board rather than per thread. Tested: ./tools/buildman/buildman.py tegra ./tools/buildman/buildman.py -I -P tegra ./tools/buildman/buildman.py -b tegra_dev tegra ./tools/buildman/buildman.py -b tegra_dev -I -P tegra ... each once after deleting the buildman result/work directory, and once "incrementally" after a previous identical invocation. Signed-off-by: Stephen Warren <swarren@nvidia.com> Reviewed-by: Tom Rini <trini@konsulko.com> Acked-by: Simon Glass <sjg@chromium.org> # v1 Tested-by: Simon Glass <sjg@chromium.org> # v1 Acked-by: Simon Glass <sjg@chromium.org>
2016-04-11 16:48:44 +00:00
per_board_out_dir: Build in a separate persistent directory per
board rather than a thread-specific directory
config_only: Only configure each build, don't build it
squash_config_y: Convert CONFIG options with the value 'y' to '1'
warnings_as_errors: Treat all compiler warnings as errors
work_in_output: Use the output directory as the work directory and
don't write to a separate output directory.
test_thread_exceptions: Uses for tests only, True to make the
threads raise an exception instead of reporting their result.
This simulates a failure in the code somewhere
adjust_cfg_list (list of str): List of changes to make to .config
file before building. Each is one of (where C is the config
option with or without the CONFIG_ prefix)
C to enable C
~C to disable C
C=val to set the value of C (val must have quotes if C is
a string Kconfig
"""
self.toolchains = toolchains
self.base_dir = base_dir
if work_in_output:
self._working_dir = base_dir
else:
self._working_dir = os.path.join(base_dir, '.bm-work')
self.threads = []
self.do_make = self.Make
self.gnu_make = gnu_make
self.checkout = checkout
self.num_threads = num_threads
self.num_jobs = num_jobs
self.already_done = 0
self.force_build = False
self.git_dir = git_dir
self._show_unknown = show_unknown
self._timestamp_count = 10
self._build_period_us = None
self._complete_delay = None
self._next_delay_update = datetime.now()
self._start_time = datetime.now()
self.force_config_on_failure = True
self.force_build_failures = False
self.force_reconfig = False
self._step = step
self.in_tree = False
self._error_lines = 0
self.no_subdirs = no_subdirs
self.full_path = full_path
self.verbose_build = verbose_build
self.config_only = config_only
self.squash_config_y = squash_config_y
self.config_filenames = BASE_CONFIG_FILENAMES
self.work_in_output = work_in_output
self.adjust_cfg = adjust_cfg
if not self.squash_config_y:
self.config_filenames += EXTRA_CONFIG_FILENAMES
buildman: Detect Kconfig loops Hex and int Kconfig options are supposed to have defaults. This is so we can configure U-Boot without having to enter particular values for the items that don't have specific values in the board's defconfig file. If this rule is not followed, then introducing a new Kconfig can produce a loop like this: Break things (BREAK_ME) [] (NEW) Error in reading or end of file. Break things (BREAK_ME) [] (NEW) Error in reading or end of file. The continues forever since buildman passes /dev/null to 'conf', and the build system just tries again. Eventually there is so much output that buildman runs out of memory. We can detect this situation by looking for a symbol (like 'BREAK_ME') which has no default (the '[]' above) and is marked as new. If this appears multiple times in the output, we know something is wrong. Add a filter function for the output which detects this situation. Allow it to return True to terminate the process. Implement this termination in cros_subprocess. With this we get a nice message: buildman --board sandbox -T0 Building current source for 1 boards (0 threads, 32 jobs per thread) sandbox: w+ sandbox +.config:66:warning: symbol value '' invalid for BREAK_ME + +Error in reading or end of file. +make[3]: *** [scripts/kconfig/Makefile:75: syncconfig] Terminated +make[2]: *** [Makefile:569: syncconfig] Terminated +make: *** [Makefile:177: sub-make] Terminated +(** did you define an int/hex Kconfig with no default? **) Signed-off-by: Simon Glass <sjg@chromium.org>
2021-10-20 03:43:24 +00:00
self._terminated = False
self._restarting_config = False
self.warnings_as_errors = warnings_as_errors
self.col = terminal.Color()
self._re_function = re.compile('(.*): In function.*')
self._re_files = re.compile('In file included from.*')
self._re_warning = re.compile('(.*):(\d*):(\d*): warning: .*')
self._re_dtb_warning = re.compile('(.*): Warning .*')
self._re_note = re.compile('(.*):(\d*):(\d*): note: this is the location of the previous.*')
self._re_migration_warning = re.compile(r'^={21} WARNING ={22}\n.*\n=+\n',
re.MULTILINE | re.DOTALL)
self.thread_exceptions = []
self.test_thread_exceptions = test_thread_exceptions
if self.num_threads:
self._single_builder = None
self.queue = queue.Queue()
self.out_queue = queue.Queue()
for i in range(self.num_threads):
t = builderthread.BuilderThread(
self, i, mrproper, per_board_out_dir,
test_exception=test_thread_exceptions)
t.setDaemon(True)
t.start()
self.threads.append(t)
t = builderthread.ResultThread(self)
t.setDaemon(True)
t.start()
self.threads.append(t)
else:
self._single_builder = builderthread.BuilderThread(
self, -1, mrproper, per_board_out_dir)
ignore_lines = ['(make.*Waiting for unfinished)', '(Segmentation fault)']
self.re_make_err = re.compile('|'.join(ignore_lines))
# Handle existing graceful with SIGINT / Ctrl-C
signal.signal(signal.SIGINT, self.signal_handler)
def __del__(self):
"""Get rid of all threads created by the builder"""
for t in self.threads:
del t
def signal_handler(self, signal, frame):
sys.exit(1)
def SetDisplayOptions(self, show_errors=False, show_sizes=False,
show_detail=False, show_bloat=False,
list_error_boards=False, show_config=False,
show_environment=False, filter_dtb_warnings=False,
filter_migration_warnings=False):
"""Setup display options for the builder.
Args:
show_errors: True to show summarised error/warning info
show_sizes: Show size deltas
show_detail: Show size delta detail for each board if show_sizes
show_bloat: Show detail for each function
list_error_boards: Show the boards which caused each error/warning
show_config: Show config deltas
show_environment: Show environment deltas
filter_dtb_warnings: Filter out any warnings from the device-tree
compiler
filter_migration_warnings: Filter out any warnings about migrating
a board to driver model
"""
self._show_errors = show_errors
self._show_sizes = show_sizes
self._show_detail = show_detail
self._show_bloat = show_bloat
self._list_error_boards = list_error_boards
self._show_config = show_config
self._show_environment = show_environment
self._filter_dtb_warnings = filter_dtb_warnings
self._filter_migration_warnings = filter_migration_warnings
def _AddTimestamp(self):
"""Add a new timestamp to the list and record the build period.
The build period is the length of time taken to perform a single
build (one board, one commit).
"""
now = datetime.now()
self._timestamps.append(now)
count = len(self._timestamps)
delta = self._timestamps[-1] - self._timestamps[0]
seconds = delta.total_seconds()
# If we have enough data, estimate build period (time taken for a
# single build) and therefore completion time.
if count > 1 and self._next_delay_update < now:
self._next_delay_update = now + timedelta(seconds=2)
if seconds > 0:
self._build_period = float(seconds) / count
todo = self.count - self.upto
self._complete_delay = timedelta(microseconds=
self._build_period * todo * 1000000)
# Round it
self._complete_delay -= timedelta(
microseconds=self._complete_delay.microseconds)
if seconds > 60:
self._timestamps.popleft()
count -= 1
def SelectCommit(self, commit, checkout=True):
"""Checkout the selected commit for this build
"""
self.commit = commit
if checkout and self.checkout:
gitutil.checkout(commit.hash)
def Make(self, commit, brd, stage, cwd, *args, **kwargs):
"""Run make
Args:
commit: Commit object that is being built
brd: Board object that is being built
stage: Stage that we are at (mrproper, config, build)
cwd: Directory where make should be run
args: Arguments to pass to make
kwargs: Arguments to pass to command.run_pipe()
"""
buildman: Detect Kconfig loops Hex and int Kconfig options are supposed to have defaults. This is so we can configure U-Boot without having to enter particular values for the items that don't have specific values in the board's defconfig file. If this rule is not followed, then introducing a new Kconfig can produce a loop like this: Break things (BREAK_ME) [] (NEW) Error in reading or end of file. Break things (BREAK_ME) [] (NEW) Error in reading or end of file. The continues forever since buildman passes /dev/null to 'conf', and the build system just tries again. Eventually there is so much output that buildman runs out of memory. We can detect this situation by looking for a symbol (like 'BREAK_ME') which has no default (the '[]' above) and is marked as new. If this appears multiple times in the output, we know something is wrong. Add a filter function for the output which detects this situation. Allow it to return True to terminate the process. Implement this termination in cros_subprocess. With this we get a nice message: buildman --board sandbox -T0 Building current source for 1 boards (0 threads, 32 jobs per thread) sandbox: w+ sandbox +.config:66:warning: symbol value '' invalid for BREAK_ME + +Error in reading or end of file. +make[3]: *** [scripts/kconfig/Makefile:75: syncconfig] Terminated +make[2]: *** [Makefile:569: syncconfig] Terminated +make: *** [Makefile:177: sub-make] Terminated +(** did you define an int/hex Kconfig with no default? **) Signed-off-by: Simon Glass <sjg@chromium.org>
2021-10-20 03:43:24 +00:00
def check_output(stream, data):
if b'Restart config' in data:
self._restarting_config = True
# If we see 'Restart config' following by multiple errors
if self._restarting_config:
m = RE_NO_DEFAULT.findall(data)
# Number of occurences of each Kconfig item
multiple = [m.count(val) for val in set(m)]
# If any of them occur more than once, we have a loop
if [val for val in multiple if val > 1]:
self._terminated = True
return True
return False
self._restarting_config = False
self._terminated = False
cmd = [self.gnu_make] + list(args)
result = command.run_pipe([cmd], capture=True, capture_stderr=True,
buildman: Detect Kconfig loops Hex and int Kconfig options are supposed to have defaults. This is so we can configure U-Boot without having to enter particular values for the items that don't have specific values in the board's defconfig file. If this rule is not followed, then introducing a new Kconfig can produce a loop like this: Break things (BREAK_ME) [] (NEW) Error in reading or end of file. Break things (BREAK_ME) [] (NEW) Error in reading or end of file. The continues forever since buildman passes /dev/null to 'conf', and the build system just tries again. Eventually there is so much output that buildman runs out of memory. We can detect this situation by looking for a symbol (like 'BREAK_ME') which has no default (the '[]' above) and is marked as new. If this appears multiple times in the output, we know something is wrong. Add a filter function for the output which detects this situation. Allow it to return True to terminate the process. Implement this termination in cros_subprocess. With this we get a nice message: buildman --board sandbox -T0 Building current source for 1 boards (0 threads, 32 jobs per thread) sandbox: w+ sandbox +.config:66:warning: symbol value '' invalid for BREAK_ME + +Error in reading or end of file. +make[3]: *** [scripts/kconfig/Makefile:75: syncconfig] Terminated +make[2]: *** [Makefile:569: syncconfig] Terminated +make: *** [Makefile:177: sub-make] Terminated +(** did you define an int/hex Kconfig with no default? **) Signed-off-by: Simon Glass <sjg@chromium.org>
2021-10-20 03:43:24 +00:00
cwd=cwd, raise_on_error=False, infile='/dev/null',
output_func=check_output, **kwargs)
if self._terminated:
# Try to be helpful
result.stderr += '(** did you define an int/hex Kconfig with no default? **)'
if self.verbose_build:
result.stdout = '%s\n' % (' '.join(cmd)) + result.stdout
result.combined = '%s\n' % (' '.join(cmd)) + result.combined
return result
def ProcessResult(self, result):
"""Process the result of a build, showing progress information
Args:
result: A CommandResult object, which indicates the result for
a single build
"""
col = terminal.Color()
if result:
target = result.brd.target
self.upto += 1
if result.return_code != 0:
self.fail += 1
elif result.stderr:
self.warned += 1
if result.already_done:
self.already_done += 1
if self._verbose:
terminal.print_clear()
boards_selected = {target : result.brd}
self.ResetResultSummary(boards_selected)
self.ProduceResultSummary(result.commit_upto, self.commits,
boards_selected)
else:
target = '(starting)'
# Display separate counts for ok, warned and fail
ok = self.upto - self.warned - self.fail
line = '\r' + self.col.build(self.col.GREEN, '%5d' % ok)
line += self.col.build(self.col.YELLOW, '%5d' % self.warned)
line += self.col.build(self.col.RED, '%5d' % self.fail)
line += ' /%-5d ' % self.count
remaining = self.count - self.upto
if remaining:
line += self.col.build(self.col.MAGENTA, ' -%-5d ' % remaining)
else:
line += ' ' * 8
# Add our current completion time estimate
self._AddTimestamp()
if self._complete_delay:
line += '%s : ' % self._complete_delay
line += target
terminal.print_clear()
tprint(line, newline=False, limit_to_line=True)
def _GetOutputDir(self, commit_upto):
"""Get the name of the output directory for a commit number
The output directory is typically .../<branch>/<commit>.
Args:
commit_upto: Commit number to use (0..self.count-1)
"""
if self.work_in_output:
return self._working_dir
commit_dir = None
if self.commits:
commit = self.commits[commit_upto]
subject = commit.subject.translate(trans_valid_chars)
# See _GetOutputSpaceRemovals() which parses this name
commit_dir = ('%02d_g%s_%s' % (commit_upto + 1,
commit.hash, subject[:20]))
elif not self.no_subdirs:
commit_dir = 'current'
if not commit_dir:
return self.base_dir
return os.path.join(self.base_dir, commit_dir)
def GetBuildDir(self, commit_upto, target):
"""Get the name of the build directory for a commit number
The build directory is typically .../<branch>/<commit>/<target>.
Args:
commit_upto: Commit number to use (0..self.count-1)
target: Target name
"""
output_dir = self._GetOutputDir(commit_upto)
if self.work_in_output:
return output_dir
return os.path.join(output_dir, target)
def GetDoneFile(self, commit_upto, target):
"""Get the name of the done file for a commit number
Args:
commit_upto: Commit number to use (0..self.count-1)
target: Target name
"""
return os.path.join(self.GetBuildDir(commit_upto, target), 'done')
def GetSizesFile(self, commit_upto, target):
"""Get the name of the sizes file for a commit number
Args:
commit_upto: Commit number to use (0..self.count-1)
target: Target name
"""
return os.path.join(self.GetBuildDir(commit_upto, target), 'sizes')
def GetFuncSizesFile(self, commit_upto, target, elf_fname):
"""Get the name of the funcsizes file for a commit number and ELF file
Args:
commit_upto: Commit number to use (0..self.count-1)
target: Target name
elf_fname: Filename of elf image
"""
return os.path.join(self.GetBuildDir(commit_upto, target),
'%s.sizes' % elf_fname.replace('/', '-'))
def GetObjdumpFile(self, commit_upto, target, elf_fname):
"""Get the name of the objdump file for a commit number and ELF file
Args:
commit_upto: Commit number to use (0..self.count-1)
target: Target name
elf_fname: Filename of elf image
"""
return os.path.join(self.GetBuildDir(commit_upto, target),
'%s.objdump' % elf_fname.replace('/', '-'))
def GetErrFile(self, commit_upto, target):
"""Get the name of the err file for a commit number
Args:
commit_upto: Commit number to use (0..self.count-1)
target: Target name
"""
output_dir = self.GetBuildDir(commit_upto, target)
return os.path.join(output_dir, 'err')
def FilterErrors(self, lines):
"""Filter out errors in which we have no interest
We should probably use map().
Args:
lines: List of error lines, each a string
Returns:
New list with only interesting lines included
"""
out_lines = []
if self._filter_migration_warnings:
text = '\n'.join(lines)
text = self._re_migration_warning.sub('', text)
lines = text.splitlines()
for line in lines:
if self.re_make_err.search(line):
continue
if self._filter_dtb_warnings and self._re_dtb_warning.search(line):
continue
out_lines.append(line)
return out_lines
def ReadFuncSizes(self, fname, fd):
"""Read function sizes from the output of 'nm'
Args:
fd: File containing data to read
fname: Filename we are reading from (just for errors)
Returns:
Dictionary containing size of each function in bytes, indexed by
function name.
"""
sym = {}
for line in fd.readlines():
try:
if line.strip():
size, type, name = line[:-1].split()
except:
tprint("Invalid line in file '%s': '%s'" % (fname, line[:-1]))
continue
if type in 'tTdDbB':
# function names begin with '.' on 64-bit powerpc
if '.' in name[1:]:
name = 'static.' + name.split('.')[0]
sym[name] = sym.get(name, 0) + int(size, 16)
return sym
def _ProcessConfig(self, fname):
"""Read in a .config, autoconf.mk or autoconf.h file
This function handles all config file types. It ignores comments and
any #defines which don't start with CONFIG_.
Args:
fname: Filename to read
Returns:
Dictionary:
key: Config name (e.g. CONFIG_DM)
value: Config value (e.g. 1)
"""
config = {}
if os.path.exists(fname):
with open(fname) as fd:
for line in fd:
line = line.strip()
if line.startswith('#define'):
values = line[8:].split(' ', 1)
if len(values) > 1:
key, value = values
else:
key = values[0]
value = '1' if self.squash_config_y else ''
if not key.startswith('CONFIG_'):
continue
elif not line or line[0] in ['#', '*', '/']:
continue
else:
key, value = line.split('=', 1)
if self.squash_config_y and value == 'y':
value = '1'
config[key] = value
return config
def _ProcessEnvironment(self, fname):
"""Read in a uboot.env file
This function reads in environment variables from a file.
Args:
fname: Filename to read
Returns:
Dictionary:
key: environment variable (e.g. bootlimit)
value: value of environment variable (e.g. 1)
"""
environment = {}
if os.path.exists(fname):
with open(fname) as fd:
for line in fd.read().split('\0'):
try:
key, value = line.split('=', 1)
environment[key] = value
except ValueError:
# ignore lines we can't parse
pass
return environment
def GetBuildOutcome(self, commit_upto, target, read_func_sizes,
read_config, read_environment):
"""Work out the outcome of a build.
Args:
commit_upto: Commit number to check (0..n-1)
target: Target board to check
read_func_sizes: True to read function size information
read_config: True to read .config and autoconf.h files
read_environment: True to read uboot.env files
Returns:
Outcome object
"""
done_file = self.GetDoneFile(commit_upto, target)
sizes_file = self.GetSizesFile(commit_upto, target)
sizes = {}
func_sizes = {}
config = {}
environment = {}
if os.path.exists(done_file):
with open(done_file, 'r') as fd:
try:
return_code = int(fd.readline())
except ValueError:
# The file may be empty due to running out of disk space.
# Try a rebuild
return_code = 1
err_lines = []
err_file = self.GetErrFile(commit_upto, target)
if os.path.exists(err_file):
with open(err_file, 'r') as fd:
err_lines = self.FilterErrors(fd.readlines())
# Decide whether the build was ok, failed or created warnings
if return_code:
rc = OUTCOME_ERROR
elif len(err_lines):
rc = OUTCOME_WARNING
else:
rc = OUTCOME_OK
# Convert size information to our simple format
if os.path.exists(sizes_file):
with open(sizes_file, 'r') as fd:
for line in fd.readlines():
values = line.split()
rodata = 0
if len(values) > 6:
rodata = int(values[6], 16)
size_dict = {
'all' : int(values[0]) + int(values[1]) +
int(values[2]),
'text' : int(values[0]) - rodata,
'data' : int(values[1]),
'bss' : int(values[2]),
'rodata' : rodata,
}
sizes[values[5]] = size_dict
if read_func_sizes:
pattern = self.GetFuncSizesFile(commit_upto, target, '*')
for fname in glob.glob(pattern):
with open(fname, 'r') as fd:
dict_name = os.path.basename(fname).replace('.sizes',
'')
func_sizes[dict_name] = self.ReadFuncSizes(fname, fd)
if read_config:
output_dir = self.GetBuildDir(commit_upto, target)
for name in self.config_filenames:
fname = os.path.join(output_dir, name)
config[name] = self._ProcessConfig(fname)
if read_environment:
output_dir = self.GetBuildDir(commit_upto, target)
fname = os.path.join(output_dir, 'uboot.env')
environment = self._ProcessEnvironment(fname)
return Builder.Outcome(rc, err_lines, sizes, func_sizes, config,
environment)
return Builder.Outcome(OUTCOME_UNKNOWN, [], {}, {}, {}, {})
def GetResultSummary(self, boards_selected, commit_upto, read_func_sizes,
read_config, read_environment):
"""Calculate a summary of the results of building a commit.
Args:
board_selected: Dict containing boards to summarise
commit_upto: Commit number to summarize (0..self.count-1)
read_func_sizes: True to read function size information
read_config: True to read .config and autoconf.h files
read_environment: True to read uboot.env files
Returns:
Tuple:
Dict containing boards which passed building this commit.
keyed by board.target
List containing a summary of error lines
Dict keyed by error line, containing a list of the Board
objects with that error
List containing a summary of warning lines
Dict keyed by error line, containing a list of the Board
objects with that warning
Dictionary keyed by board.target. Each value is a dictionary:
key: filename - e.g. '.config'
value is itself a dictionary:
key: config name
value: config value
Dictionary keyed by board.target. Each value is a dictionary:
key: environment variable
value: value of environment variable
"""
def AddLine(lines_summary, lines_boards, line, board):
line = line.rstrip()
if line in lines_boards:
lines_boards[line].append(board)
else:
lines_boards[line] = [board]
lines_summary.append(line)
board_dict = {}
err_lines_summary = []
err_lines_boards = {}
warn_lines_summary = []
warn_lines_boards = {}
config = {}
environment = {}
for board in boards_selected.values():
outcome = self.GetBuildOutcome(commit_upto, board.target,
read_func_sizes, read_config,
read_environment)
board_dict[board.target] = outcome
last_func = None
last_was_warning = False
for line in outcome.err_lines:
if line:
if (self._re_function.match(line) or
self._re_files.match(line)):
last_func = line
else:
is_warning = (self._re_warning.match(line) or
self._re_dtb_warning.match(line))
is_note = self._re_note.match(line)
if is_warning or (last_was_warning and is_note):
if last_func:
AddLine(warn_lines_summary, warn_lines_boards,
last_func, board)
AddLine(warn_lines_summary, warn_lines_boards,
line, board)
else:
if last_func:
AddLine(err_lines_summary, err_lines_boards,
last_func, board)
AddLine(err_lines_summary, err_lines_boards,
line, board)
last_was_warning = is_warning
last_func = None
tconfig = Config(self.config_filenames, board.target)
for fname in self.config_filenames:
if outcome.config:
for key, value in outcome.config[fname].items():
tconfig.Add(fname, key, value)
config[board.target] = tconfig
tenvironment = Environment(board.target)
if outcome.environment:
for key, value in outcome.environment.items():
tenvironment.Add(key, value)
environment[board.target] = tenvironment
return (board_dict, err_lines_summary, err_lines_boards,
warn_lines_summary, warn_lines_boards, config, environment)
def AddOutcome(self, board_dict, arch_list, changes, char, color):
"""Add an output to our list of outcomes for each architecture
This simple function adds failing boards (changes) to the
relevant architecture string, so we can print the results out
sorted by architecture.
Args:
board_dict: Dict containing all boards
arch_list: Dict keyed by arch name. Value is a string containing
a list of board names which failed for that arch.
changes: List of boards to add to arch_list
color: terminal.Colour object
"""
done_arch = {}
for target in changes:
if target in board_dict:
arch = board_dict[target].arch
else:
arch = 'unknown'
str = self.col.build(color, ' ' + target)
if not arch in done_arch:
str = ' %s %s' % (self.col.build(color, char), str)
done_arch[arch] = True
if not arch in arch_list:
arch_list[arch] = str
else:
arch_list[arch] += str
def ColourNum(self, num):
color = self.col.RED if num > 0 else self.col.GREEN
if num == 0:
return '0'
return self.col.build(color, str(num))
def ResetResultSummary(self, board_selected):
"""Reset the results summary ready for use.
Set up the base board list to be all those selected, and set the
error lines to empty.
Following this, calls to PrintResultSummary() will use this
information to work out what has changed.
Args:
board_selected: Dict containing boards to summarise, keyed by
board.target
"""
self._base_board_dict = {}
for board in board_selected:
self._base_board_dict[board] = Builder.Outcome(0, [], [], {}, {},
{})
self._base_err_lines = []
self._base_warn_lines = []
self._base_err_line_boards = {}
self._base_warn_line_boards = {}
self._base_config = None
self._base_environment = None
def PrintFuncSizeDetail(self, fname, old, new):
grow, shrink, add, remove, up, down = 0, 0, 0, 0, 0, 0
delta, common = [], {}
for a in old:
if a in new:
common[a] = 1
for name in old:
if name not in common:
remove += 1
down += old[name]
delta.append([-old[name], name])
for name in new:
if name not in common:
add += 1
up += new[name]
delta.append([new[name], name])
for name in common:
diff = new.get(name, 0) - old.get(name, 0)
if diff > 0:
grow, up = grow + 1, up + diff
elif diff < 0:
shrink, down = shrink + 1, down - diff
delta.append([diff, name])
delta.sort()
delta.reverse()
args = [add, -remove, grow, -shrink, up, -down, up - down]
if max(args) == 0 and min(args) == 0:
return
args = [self.ColourNum(x) for x in args]
indent = ' ' * 15
tprint('%s%s: add: %s/%s, grow: %s/%s bytes: %s/%s (%s)' %
tuple([indent, self.col.build(self.col.YELLOW, fname)] + args))
tprint('%s %-38s %7s %7s %+7s' % (indent, 'function', 'old', 'new',
'delta'))
for diff, name in delta:
if diff:
color = self.col.RED if diff > 0 else self.col.GREEN
msg = '%s %-38s %7s %7s %+7d' % (indent, name,
old.get(name, '-'), new.get(name,'-'), diff)
tprint(msg, colour=color)
def PrintSizeDetail(self, target_list, show_bloat):
"""Show details size information for each board
Args:
target_list: List of targets, each a dict containing:
'target': Target name
'total_diff': Total difference in bytes across all areas
<part_name>: Difference for that part
show_bloat: Show detail for each function
"""
targets_by_diff = sorted(target_list, reverse=True,
key=lambda x: x['_total_diff'])
for result in targets_by_diff:
printed_target = False
for name in sorted(result):
diff = result[name]
if name.startswith('_'):
continue
if diff != 0:
color = self.col.RED if diff > 0 else self.col.GREEN
msg = ' %s %+d' % (name, diff)
if not printed_target:
tprint('%10s %-15s:' % ('', result['_target']),
newline=False)
printed_target = True
tprint(msg, colour=color, newline=False)
if printed_target:
tprint()
if show_bloat:
target = result['_target']
outcome = result['_outcome']
base_outcome = self._base_board_dict[target]
for fname in outcome.func_sizes:
self.PrintFuncSizeDetail(fname,
base_outcome.func_sizes[fname],
outcome.func_sizes[fname])
def PrintSizeSummary(self, board_selected, board_dict, show_detail,
show_bloat):
"""Print a summary of image sizes broken down by section.
The summary takes the form of one line per architecture. The
line contains deltas for each of the sections (+ means the section
got bigger, - means smaller). The numbers are the average number
of bytes that a board in this section increased by.
For example:
powerpc: (622 boards) text -0.0
arm: (285 boards) text -0.0
Args:
board_selected: Dict containing boards to summarise, keyed by
board.target
board_dict: Dict containing boards for which we built this
commit, keyed by board.target. The value is an Outcome object.
show_detail: Show size delta detail for each board
show_bloat: Show detail for each function
"""
arch_list = {}
arch_count = {}
# Calculate changes in size for different image parts
# The previous sizes are in Board.sizes, for each board
for target in board_dict:
if target not in board_selected:
continue
base_sizes = self._base_board_dict[target].sizes
outcome = board_dict[target]
sizes = outcome.sizes
# Loop through the list of images, creating a dict of size
# changes for each image/part. We end up with something like
# {'target' : 'snapper9g45, 'data' : 5, 'u-boot-spl:text' : -4}
# which means that U-Boot data increased by 5 bytes and SPL
# text decreased by 4.
err = {'_target' : target}
for image in sizes:
if image in base_sizes:
base_image = base_sizes[image]
# Loop through the text, data, bss parts
for part in sorted(sizes[image]):
diff = sizes[image][part] - base_image[part]
col = None
if diff:
if image == 'u-boot':
name = part
else:
name = image + ':' + part
err[name] = diff
arch = board_selected[target].arch
if not arch in arch_count:
arch_count[arch] = 1
else:
arch_count[arch] += 1
if not sizes:
pass # Only add to our list when we have some stats
elif not arch in arch_list:
arch_list[arch] = [err]
else:
arch_list[arch].append(err)
# We now have a list of image size changes sorted by arch
# Print out a summary of these
for arch, target_list in arch_list.items():
# Get total difference for each type
totals = {}
for result in target_list:
total = 0
for name, diff in result.items():
if name.startswith('_'):
continue
total += diff
if name in totals:
totals[name] += diff
else:
totals[name] = diff
result['_total_diff'] = total
result['_outcome'] = board_dict[result['_target']]
count = len(target_list)
printed_arch = False
for name in sorted(totals):
diff = totals[name]
if diff:
# Display the average difference in this name for this
# architecture
avg_diff = float(diff) / count
color = self.col.RED if avg_diff > 0 else self.col.GREEN
msg = ' %s %+1.1f' % (name, avg_diff)
if not printed_arch:
tprint('%10s: (for %d/%d boards)' % (arch, count,
arch_count[arch]), newline=False)
printed_arch = True
tprint(msg, colour=color, newline=False)
if printed_arch:
tprint()
if show_detail:
self.PrintSizeDetail(target_list, show_bloat)
def PrintResultSummary(self, board_selected, board_dict, err_lines,
err_line_boards, warn_lines, warn_line_boards,
config, environment, show_sizes, show_detail,
show_bloat, show_config, show_environment):
"""Compare results with the base results and display delta.
Only boards mentioned in board_selected will be considered. This
function is intended to be called repeatedly with the results of
each commit. It therefore shows a 'diff' between what it saw in
the last call and what it sees now.
Args:
board_selected: Dict containing boards to summarise, keyed by
board.target
board_dict: Dict containing boards for which we built this
commit, keyed by board.target. The value is an Outcome object.
err_lines: A list of errors for this commit, or [] if there is
none, or we don't want to print errors
err_line_boards: Dict keyed by error line, containing a list of
the Board objects with that error
warn_lines: A list of warnings for this commit, or [] if there is
none, or we don't want to print errors
warn_line_boards: Dict keyed by warning line, containing a list of
the Board objects with that warning
config: Dictionary keyed by filename - e.g. '.config'. Each
value is itself a dictionary:
key: config name
value: config value
environment: Dictionary keyed by environment variable, Each
value is the value of environment variable.
show_sizes: Show image size deltas
show_detail: Show size delta detail for each board if show_sizes
show_bloat: Show detail for each function
show_config: Show config changes
show_environment: Show environment changes
"""
def _BoardList(line, line_boards):
"""Helper function to get a line of boards containing a line
Args:
line: Error line to search for
line_boards: boards to search, each a Board
Return:
List of boards with that error line, or [] if the user has not
requested such a list
"""
boards = []
board_set = set()
if self._list_error_boards:
for board in line_boards[line]:
if not board in board_set:
boards.append(board)
board_set.add(board)
return boards
def _CalcErrorDelta(base_lines, base_line_boards, lines, line_boards,
char):
"""Calculate the required output based on changes in errors
Args:
base_lines: List of errors/warnings for previous commit
base_line_boards: Dict keyed by error line, containing a list
of the Board objects with that error in the previous commit
lines: List of errors/warning for this commit, each a str
line_boards: Dict keyed by error line, containing a list
of the Board objects with that error in this commit
char: Character representing error ('') or warning ('w'). The
broken ('+') or fixed ('-') characters are added in this
function
Returns:
Tuple
List of ErrLine objects for 'better' lines
List of ErrLine objects for 'worse' lines
"""
better_lines = []
worse_lines = []
for line in lines:
if line not in base_lines:
errline = ErrLine(char + '+', _BoardList(line, line_boards),
line)
worse_lines.append(errline)
for line in base_lines:
if line not in lines:
errline = ErrLine(char + '-',
_BoardList(line, base_line_boards), line)
better_lines.append(errline)
return better_lines, worse_lines
def _CalcConfig(delta, name, config):
"""Calculate configuration changes
Args:
delta: Type of the delta, e.g. '+'
name: name of the file which changed (e.g. .config)
config: configuration change dictionary
key: config name
value: config value
Returns:
String containing the configuration changes which can be
printed
"""
out = ''
for key in sorted(config.keys()):
out += '%s=%s ' % (key, config[key])
return '%s %s: %s' % (delta, name, out)
def _AddConfig(lines, name, config_plus, config_minus, config_change):
"""Add changes in configuration to a list
Args:
lines: list to add to
name: config file name
config_plus: configurations added, dictionary
key: config name
value: config value
config_minus: configurations removed, dictionary
key: config name
value: config value
config_change: configurations changed, dictionary
key: config name
value: config value
"""
if config_plus:
lines.append(_CalcConfig('+', name, config_plus))
if config_minus:
lines.append(_CalcConfig('-', name, config_minus))
if config_change:
lines.append(_CalcConfig('c', name, config_change))
def _OutputConfigInfo(lines):
for line in lines:
if not line:
continue
if line[0] == '+':
col = self.col.GREEN
elif line[0] == '-':
col = self.col.RED
elif line[0] == 'c':
col = self.col.YELLOW
tprint(' ' + line, newline=True, colour=col)
def _OutputErrLines(err_lines, colour):
"""Output the line of error/warning lines, if not empty
Also increments self._error_lines if err_lines not empty
Args:
err_lines: List of ErrLine objects, each an error or warning
line, possibly including a list of boards with that
error/warning
colour: Colour to use for output
"""
if err_lines:
out_list = []
for line in err_lines:
boards = ''
names = [board.target for board in line.boards]
board_str = ' '.join(names) if names else ''
if board_str:
out = self.col.build(colour, line.char + '(')
out += self.col.build(self.col.MAGENTA, board_str,
bright=False)
out += self.col.build(colour, ') %s' % line.errline)
else:
out = self.col.build(colour, line.char + line.errline)
out_list.append(out)
tprint('\n'.join(out_list))
self._error_lines += 1
ok_boards = [] # List of boards fixed since last commit
warn_boards = [] # List of boards with warnings since last commit
err_boards = [] # List of new broken boards since last commit
new_boards = [] # List of boards that didn't exist last time
unknown_boards = [] # List of boards that were not built
for target in board_dict:
if target not in board_selected:
continue
# If the board was built last time, add its outcome to a list
if target in self._base_board_dict:
base_outcome = self._base_board_dict[target].rc
outcome = board_dict[target]
if outcome.rc == OUTCOME_UNKNOWN:
unknown_boards.append(target)
elif outcome.rc < base_outcome:
if outcome.rc == OUTCOME_WARNING:
warn_boards.append(target)
else:
ok_boards.append(target)
elif outcome.rc > base_outcome:
if outcome.rc == OUTCOME_WARNING:
warn_boards.append(target)
else:
err_boards.append(target)
else:
new_boards.append(target)
# Get a list of errors and warnings that have appeared, and disappeared
better_err, worse_err = _CalcErrorDelta(self._base_err_lines,
self._base_err_line_boards, err_lines, err_line_boards, '')
better_warn, worse_warn = _CalcErrorDelta(self._base_warn_lines,
self._base_warn_line_boards, warn_lines, warn_line_boards, 'w')
# Display results by arch
if any((ok_boards, warn_boards, err_boards, unknown_boards, new_boards,
worse_err, better_err, worse_warn, better_warn)):
arch_list = {}
self.AddOutcome(board_selected, arch_list, ok_boards, '',
self.col.GREEN)
self.AddOutcome(board_selected, arch_list, warn_boards, 'w+',
self.col.YELLOW)
self.AddOutcome(board_selected, arch_list, err_boards, '+',
self.col.RED)
self.AddOutcome(board_selected, arch_list, new_boards, '*', self.col.BLUE)
if self._show_unknown:
self.AddOutcome(board_selected, arch_list, unknown_boards, '?',
self.col.MAGENTA)
for arch, target_list in arch_list.items():
tprint('%10s: %s' % (arch, target_list))
self._error_lines += 1
_OutputErrLines(better_err, colour=self.col.GREEN)
_OutputErrLines(worse_err, colour=self.col.RED)
_OutputErrLines(better_warn, colour=self.col.CYAN)
_OutputErrLines(worse_warn, colour=self.col.YELLOW)
if show_sizes:
self.PrintSizeSummary(board_selected, board_dict, show_detail,
show_bloat)
if show_environment and self._base_environment:
lines = []
for target in board_dict:
if target not in board_selected:
continue
tbase = self._base_environment[target]
tenvironment = environment[target]
environment_plus = {}
environment_minus = {}
environment_change = {}
base = tbase.environment
for key, value in tenvironment.environment.items():
if key not in base:
environment_plus[key] = value
for key, value in base.items():
if key not in tenvironment.environment:
environment_minus[key] = value
for key, value in base.items():
new_value = tenvironment.environment.get(key)
if new_value and value != new_value:
desc = '%s -> %s' % (value, new_value)
environment_change[key] = desc
_AddConfig(lines, target, environment_plus, environment_minus,
environment_change)
_OutputConfigInfo(lines)
if show_config and self._base_config:
summary = {}
arch_config_plus = {}
arch_config_minus = {}
arch_config_change = {}
arch_list = []
for target in board_dict:
if target not in board_selected:
continue
arch = board_selected[target].arch
if arch not in arch_list:
arch_list.append(arch)
for arch in arch_list:
arch_config_plus[arch] = {}
arch_config_minus[arch] = {}
arch_config_change[arch] = {}
for name in self.config_filenames:
arch_config_plus[arch][name] = {}
arch_config_minus[arch][name] = {}
arch_config_change[arch][name] = {}
for target in board_dict:
if target not in board_selected:
continue
arch = board_selected[target].arch
all_config_plus = {}
all_config_minus = {}
all_config_change = {}
tbase = self._base_config[target]
tconfig = config[target]
lines = []
for name in self.config_filenames:
if not tconfig.config[name]:
continue
config_plus = {}
config_minus = {}
config_change = {}
base = tbase.config[name]
for key, value in tconfig.config[name].items():
if key not in base:
config_plus[key] = value
all_config_plus[key] = value
for key, value in base.items():
if key not in tconfig.config[name]:
config_minus[key] = value
all_config_minus[key] = value
for key, value in base.items():
new_value = tconfig.config.get(key)
if new_value and value != new_value:
desc = '%s -> %s' % (value, new_value)
config_change[key] = desc
all_config_change[key] = desc
arch_config_plus[arch][name].update(config_plus)
arch_config_minus[arch][name].update(config_minus)
arch_config_change[arch][name].update(config_change)
_AddConfig(lines, name, config_plus, config_minus,
config_change)
_AddConfig(lines, 'all', all_config_plus, all_config_minus,
all_config_change)
summary[target] = '\n'.join(lines)
lines_by_target = {}
for target, lines in summary.items():
if lines in lines_by_target:
lines_by_target[lines].append(target)
else:
lines_by_target[lines] = [target]
for arch in arch_list:
lines = []
all_plus = {}
all_minus = {}
all_change = {}
for name in self.config_filenames:
all_plus.update(arch_config_plus[arch][name])
all_minus.update(arch_config_minus[arch][name])
all_change.update(arch_config_change[arch][name])
_AddConfig(lines, name, arch_config_plus[arch][name],
arch_config_minus[arch][name],
arch_config_change[arch][name])
_AddConfig(lines, 'all', all_plus, all_minus, all_change)
#arch_summary[target] = '\n'.join(lines)
if lines:
tprint('%s:' % arch)
_OutputConfigInfo(lines)
for lines, targets in lines_by_target.items():
if not lines:
continue
tprint('%s :' % ' '.join(sorted(targets)))
_OutputConfigInfo(lines.split('\n'))
# Save our updated information for the next call to this function
self._base_board_dict = board_dict
self._base_err_lines = err_lines
self._base_warn_lines = warn_lines
self._base_err_line_boards = err_line_boards
self._base_warn_line_boards = warn_line_boards
self._base_config = config
self._base_environment = environment
# Get a list of boards that did not get built, if needed
not_built = []
for board in board_selected:
if not board in board_dict:
not_built.append(board)
if not_built:
tprint("Boards not built (%d): %s" % (len(not_built),
', '.join(not_built)))
def ProduceResultSummary(self, commit_upto, commits, board_selected):
(board_dict, err_lines, err_line_boards, warn_lines,
warn_line_boards, config, environment) = self.GetResultSummary(
board_selected, commit_upto,
read_func_sizes=self._show_bloat,
read_config=self._show_config,
read_environment=self._show_environment)
if commits:
msg = '%02d: %s' % (commit_upto + 1,
commits[commit_upto].subject)
tprint(msg, colour=self.col.BLUE)
self.PrintResultSummary(board_selected, board_dict,
err_lines if self._show_errors else [], err_line_boards,
warn_lines if self._show_errors else [], warn_line_boards,
config, environment, self._show_sizes, self._show_detail,
self._show_bloat, self._show_config, self._show_environment)
def ShowSummary(self, commits, board_selected):
"""Show a build summary for U-Boot for a given board list.
Reset the result summary, then repeatedly call GetResultSummary on
each commit's results, then display the differences we see.
Args:
commit: Commit objects to summarise
board_selected: Dict containing boards to summarise
"""
self.commit_count = len(commits) if commits else 1
self.commits = commits
self.ResetResultSummary(board_selected)
self._error_lines = 0
for commit_upto in range(0, self.commit_count, self._step):
self.ProduceResultSummary(commit_upto, commits, board_selected)
if not self._error_lines:
tprint('(no errors to report)', colour=self.col.GREEN)
def SetupBuild(self, board_selected, commits):
"""Set up ready to start a build.
Args:
board_selected: Selected boards to build
commits: Selected commits to build
"""
# First work out how many commits we will build
count = (self.commit_count + self._step - 1) // self._step
self.count = len(board_selected) * count
self.upto = self.warned = self.fail = 0
self._timestamps = collections.deque()
def GetThreadDir(self, thread_num):
"""Get the directory path to the working dir for a thread.
Args:
thread_num: Number of thread to check (-1 for main process, which
is treated as 0)
"""
if self.work_in_output:
return self._working_dir
return os.path.join(self._working_dir, '%02d' % max(thread_num, 0))
def _PrepareThread(self, thread_num, setup_git):
"""Prepare the working directory for a thread.
This clones or fetches the repo into the thread's work directory.
Optionally, it can create a linked working tree of the repo in the
thread's work directory instead.
Args:
thread_num: Thread number (0, 1, ...)
setup_git:
'clone' to set up a git clone
'worktree' to set up a git worktree
"""
thread_dir = self.GetThreadDir(thread_num)
builderthread.Mkdir(thread_dir)
git_dir = os.path.join(thread_dir, '.git')
# Create a worktree or a git repo clone for this thread if it
# doesn't already exist
if setup_git and self.git_dir:
src_dir = os.path.abspath(self.git_dir)
if os.path.isdir(git_dir):
# This is a clone of the src_dir repo, we can keep using
# it but need to fetch from src_dir.
tprint('\rFetching repo for thread %d' % thread_num,
newline=False)
gitutil.fetch(git_dir, thread_dir)
terminal.print_clear()
elif os.path.isfile(git_dir):
# This is a worktree of the src_dir repo, we don't need to
# create it again or update it in any way.
pass
elif os.path.exists(git_dir):
# Don't know what could trigger this, but we probably
# can't create a git worktree/clone here.
raise ValueError('Git dir %s exists, but is not a file '
'or a directory.' % git_dir)
elif setup_git == 'worktree':
tprint('\rChecking out worktree for thread %d' % thread_num,
newline=False)
gitutil.add_worktree(src_dir, thread_dir)
terminal.print_clear()
elif setup_git == 'clone' or setup_git == True:
tprint('\rCloning repo for thread %d' % thread_num,
newline=False)
gitutil.clone(src_dir, thread_dir)
terminal.print_clear()
else:
raise ValueError("Can't setup git repo with %s." % setup_git)
def _PrepareWorkingSpace(self, max_threads, setup_git):
"""Prepare the working directory for use.
Set up the git repo for each thread. Creates a linked working tree
if git-worktree is available, or clones the repo if it isn't.
Args:
max_threads: Maximum number of threads we expect to need. If 0 then
1 is set up, since the main process still needs somewhere to
work
setup_git: True to set up a git worktree or a git clone
"""
builderthread.Mkdir(self._working_dir)
if setup_git and self.git_dir:
src_dir = os.path.abspath(self.git_dir)
if gitutil.check_worktree_is_available(src_dir):
setup_git = 'worktree'
# If we previously added a worktree but the directory for it
# got deleted, we need to prune its files from the repo so
# that we can check out another in its place.
gitutil.prune_worktrees(src_dir)
else:
setup_git = 'clone'
# Always do at least one thread
for thread in range(max(max_threads, 1)):
self._PrepareThread(thread, setup_git)
def _GetOutputSpaceRemovals(self):
"""Get the output directories ready to receive files.
Figure out what needs to be deleted in the output directory before it
can be used. We only delete old buildman directories which have the
expected name pattern. See _GetOutputDir().
Returns:
List of full paths of directories to remove
"""
if not self.commits:
return
dir_list = []
for commit_upto in range(self.commit_count):
dir_list.append(self._GetOutputDir(commit_upto))
to_remove = []
for dirname in glob.glob(os.path.join(self.base_dir, '*')):
if dirname not in dir_list:
leaf = dirname[len(self.base_dir) + 1:]
m = re.match('[0-9]+_g[0-9a-f]+_.*', leaf)
if m:
to_remove.append(dirname)
return to_remove
def _PrepareOutputSpace(self):
"""Get the output directories ready to receive files.
We delete any output directories which look like ones we need to
create. Having left over directories is confusing when the user wants
to check the output manually.
"""
to_remove = self._GetOutputSpaceRemovals()
if to_remove:
tprint('Removing %d old build directories...' % len(to_remove),
newline=False)
for dirname in to_remove:
shutil.rmtree(dirname)
terminal.print_clear()
def BuildBoards(self, commits, board_selected, keep_outputs, verbose):
"""Build all commits for a list of boards
Args:
commits: List of commits to be build, each a Commit object
boards_selected: Dict of selected boards, key is target name,
value is Board object
keep_outputs: True to save build output files
verbose: Display build results as they are completed
Returns:
Tuple containing:
- number of boards that failed to build
- number of boards that issued warnings
- list of thread exceptions raised
"""
self.commit_count = len(commits) if commits else 1
self.commits = commits
self._verbose = verbose
self.ResetResultSummary(board_selected)
builderthread.Mkdir(self.base_dir, parents = True)
self._PrepareWorkingSpace(min(self.num_threads, len(board_selected)),
commits is not None)
self._PrepareOutputSpace()
tprint('\rStarting build...', newline=False)
self.SetupBuild(board_selected, commits)
self.ProcessResult(None)
self.thread_exceptions = []
# Create jobs to build all commits for each board
for brd in board_selected.values():
job = builderthread.BuilderJob()
job.board = brd
job.commits = commits
job.keep_outputs = keep_outputs
job.work_in_output = self.work_in_output
job.adjust_cfg = self.adjust_cfg
job.step = self._step
if self.num_threads:
self.queue.put(job)
else:
self._single_builder.RunJob(job)
if self.num_threads:
term = threading.Thread(target=self.queue.join)
term.setDaemon(True)
term.start()
while term.is_alive():
term.join(100)
# Wait until we have processed all output
self.out_queue.join()
tprint()
msg = 'Completed: %d total built' % self.count
if self.already_done:
msg += ' (%d previously' % self.already_done
if self.already_done != self.count:
msg += ', %d newly' % (self.count - self.already_done)
msg += ')'
duration = datetime.now() - self._start_time
if duration > timedelta(microseconds=1000000):
if duration.microseconds >= 500000:
duration = duration + timedelta(seconds=1)
duration = duration - timedelta(microseconds=duration.microseconds)
rate = float(self.count) / duration.total_seconds()
msg += ', duration %s, rate %1.2f' % (duration, rate)
tprint(msg)
if self.thread_exceptions:
tprint('Failed: %d thread exceptions' % len(self.thread_exceptions),
colour=self.col.RED)
return (self.fail, self.warned, self.thread_exceptions)