| """ |
| BitBake Utility Functions |
| """ |
| |
| # Copyright (C) 2004 Michael Lauer |
| # |
| # SPDX-License-Identifier: GPL-2.0-only |
| # |
| |
| import re, fcntl, os, string, stat, shutil, time |
| import sys |
| import errno |
| import logging |
| import bb |
| import bb.msg |
| import multiprocessing |
| import fcntl |
| import importlib |
| import importlib.machinery |
| import importlib.util |
| import itertools |
| import subprocess |
| import glob |
| import fnmatch |
| import traceback |
| import errno |
| import signal |
| import collections |
| import copy |
| import ctypes |
| from subprocess import getstatusoutput |
| from contextlib import contextmanager |
| from ctypes import cdll |
| |
| logger = logging.getLogger("BitBake.Util") |
| python_extensions = importlib.machinery.all_suffixes() |
| |
| |
| def clean_context(): |
| return { |
| "os": os, |
| "bb": bb, |
| "time": time, |
| } |
| |
| def get_context(): |
| return _context |
| |
| |
| def set_context(ctx): |
| _context = ctx |
| |
| # Context used in better_exec, eval |
| _context = clean_context() |
| |
| class VersionStringException(Exception): |
| """Exception raised when an invalid version specification is found""" |
| |
| def explode_version(s): |
| r = [] |
| alpha_regexp = re.compile(r'^([a-zA-Z]+)(.*)$') |
| numeric_regexp = re.compile(r'^(\d+)(.*)$') |
| while (s != ''): |
| if s[0] in string.digits: |
| m = numeric_regexp.match(s) |
| r.append((0, int(m.group(1)))) |
| s = m.group(2) |
| continue |
| if s[0] in string.ascii_letters: |
| m = alpha_regexp.match(s) |
| r.append((1, m.group(1))) |
| s = m.group(2) |
| continue |
| if s[0] == '~': |
| r.append((-1, s[0])) |
| else: |
| r.append((2, s[0])) |
| s = s[1:] |
| return r |
| |
| def split_version(s): |
| """Split a version string into its constituent parts (PE, PV, PR)""" |
| s = s.strip(" <>=") |
| e = 0 |
| if s.count(':'): |
| e = int(s.split(":")[0]) |
| s = s.split(":")[1] |
| r = "" |
| if s.count('-'): |
| r = s.rsplit("-", 1)[1] |
| s = s.rsplit("-", 1)[0] |
| v = s |
| return (e, v, r) |
| |
| def vercmp_part(a, b): |
| va = explode_version(a) |
| vb = explode_version(b) |
| while True: |
| if va == []: |
| (oa, ca) = (0, None) |
| else: |
| (oa, ca) = va.pop(0) |
| if vb == []: |
| (ob, cb) = (0, None) |
| else: |
| (ob, cb) = vb.pop(0) |
| if (oa, ca) == (0, None) and (ob, cb) == (0, None): |
| return 0 |
| if oa < ob: |
| return -1 |
| elif oa > ob: |
| return 1 |
| elif ca is None: |
| return -1 |
| elif cb is None: |
| return 1 |
| elif ca < cb: |
| return -1 |
| elif ca > cb: |
| return 1 |
| |
| def vercmp(ta, tb): |
| (ea, va, ra) = ta |
| (eb, vb, rb) = tb |
| |
| r = int(ea or 0) - int(eb or 0) |
| if (r == 0): |
| r = vercmp_part(va, vb) |
| if (r == 0): |
| r = vercmp_part(ra, rb) |
| return r |
| |
| def vercmp_string(a, b): |
| """ Split version strings and compare them """ |
| ta = split_version(a) |
| tb = split_version(b) |
| return vercmp(ta, tb) |
| |
| def vercmp_string_op(a, b, op): |
| """ |
| Compare two versions and check if the specified comparison operator matches the result of the comparison. |
| This function is fairly liberal about what operators it will accept since there are a variety of styles |
| depending on the context. |
| """ |
| res = vercmp_string(a, b) |
| if op in ('=', '=='): |
| return res == 0 |
| elif op == '<=': |
| return res <= 0 |
| elif op == '>=': |
| return res >= 0 |
| elif op in ('>', '>>'): |
| return res > 0 |
| elif op in ('<', '<<'): |
| return res < 0 |
| elif op == '!=': |
| return res != 0 |
| else: |
| raise VersionStringException('Unsupported comparison operator "%s"' % op) |
| |
| def explode_deps(s): |
| """ |
| Take an RDEPENDS style string of format: |
| "DEPEND1 (optional version) DEPEND2 (optional version) ..." |
| and return a list of dependencies. |
| Version information is ignored. |
| """ |
| r = [] |
| l = s.split() |
| flag = False |
| for i in l: |
| if i[0] == '(': |
| flag = True |
| #j = [] |
| if not flag: |
| r.append(i) |
| #else: |
| # j.append(i) |
| if flag and i.endswith(')'): |
| flag = False |
| # Ignore version |
| #r[-1] += ' ' + ' '.join(j) |
| return r |
| |
| def explode_dep_versions2(s, *, sort=True): |
| """ |
| Take an RDEPENDS style string of format: |
| "DEPEND1 (optional version) DEPEND2 (optional version) ..." |
| and return a dictionary of dependencies and versions. |
| """ |
| r = collections.OrderedDict() |
| l = s.replace(",", "").split() |
| lastdep = None |
| lastcmp = "" |
| lastver = "" |
| incmp = False |
| inversion = False |
| for i in l: |
| if i[0] == '(': |
| incmp = True |
| i = i[1:].strip() |
| if not i: |
| continue |
| |
| if incmp: |
| incmp = False |
| inversion = True |
| # This list is based on behavior and supported comparisons from deb, opkg and rpm. |
| # |
| # Even though =<, <<, ==, !=, =>, and >> may not be supported, |
| # we list each possibly valid item. |
| # The build system is responsible for validation of what it supports. |
| if i.startswith(('<=', '=<', '<<', '==', '!=', '>=', '=>', '>>')): |
| lastcmp = i[0:2] |
| i = i[2:] |
| elif i.startswith(('<', '>', '=')): |
| lastcmp = i[0:1] |
| i = i[1:] |
| else: |
| # This is an unsupported case! |
| raise VersionStringException('Invalid version specification in "(%s" - invalid or missing operator' % i) |
| lastcmp = (i or "") |
| i = "" |
| i.strip() |
| if not i: |
| continue |
| |
| if inversion: |
| if i.endswith(')'): |
| i = i[:-1] or "" |
| inversion = False |
| if lastver and i: |
| lastver += " " |
| if i: |
| lastver += i |
| if lastdep not in r: |
| r[lastdep] = [] |
| r[lastdep].append(lastcmp + " " + lastver) |
| continue |
| |
| #if not inversion: |
| lastdep = i |
| lastver = "" |
| lastcmp = "" |
| if not (i in r and r[i]): |
| r[lastdep] = [] |
| |
| if sort: |
| r = collections.OrderedDict(sorted(r.items(), key=lambda x: x[0])) |
| return r |
| |
| def explode_dep_versions(s): |
| """ |
| Take an RDEPENDS style string of format: |
| "DEPEND1 (optional version) DEPEND2 (optional version) ..." |
| skip null value and items appeared in dependency string multiple times |
| and return a dictionary of dependencies and versions. |
| """ |
| r = explode_dep_versions2(s) |
| for d in r: |
| if not r[d]: |
| r[d] = None |
| continue |
| if len(r[d]) > 1: |
| bb.warn("explode_dep_versions(): Item %s appeared in dependency string '%s' multiple times with different values. explode_dep_versions cannot cope with this." % (d, s)) |
| r[d] = r[d][0] |
| return r |
| |
| def join_deps(deps, commasep=True): |
| """ |
| Take the result from explode_dep_versions and generate a dependency string |
| """ |
| result = [] |
| for dep in deps: |
| if deps[dep]: |
| if isinstance(deps[dep], list): |
| for v in deps[dep]: |
| result.append(dep + " (" + v + ")") |
| else: |
| result.append(dep + " (" + deps[dep] + ")") |
| else: |
| result.append(dep) |
| if commasep: |
| return ", ".join(result) |
| else: |
| return " ".join(result) |
| |
| def _print_trace(body, line): |
| """ |
| Print the Environment of a Text Body |
| """ |
| error = [] |
| # print the environment of the method |
| min_line = max(1, line-4) |
| max_line = min(line + 4, len(body)) |
| for i in range(min_line, max_line + 1): |
| if line == i: |
| error.append(' *** %.4d:%s' % (i, body[i-1].rstrip())) |
| else: |
| error.append(' %.4d:%s' % (i, body[i-1].rstrip())) |
| return error |
| |
| def better_compile(text, file, realfile, mode = "exec", lineno = 0): |
| """ |
| A better compile method. This method |
| will print the offending lines. |
| """ |
| try: |
| cache = bb.methodpool.compile_cache(text) |
| if cache: |
| return cache |
| # We can't add to the linenumbers for compile, we can pad to the correct number of blank lines though |
| text2 = "\n" * int(lineno) + text |
| code = compile(text2, realfile, mode) |
| bb.methodpool.compile_cache_add(text, code) |
| return code |
| except Exception as e: |
| error = [] |
| # split the text into lines again |
| body = text.split('\n') |
| error.append("Error in compiling python function in %s, line %s:\n" % (realfile, e.lineno)) |
| if hasattr(e, "lineno"): |
| error.append("The code lines resulting in this error were:") |
| # e.lineno: line's position in reaflile |
| # lineno: function name's "position -1" in realfile |
| # e.lineno - lineno: line's relative position in function |
| error.extend(_print_trace(body, e.lineno - lineno)) |
| else: |
| error.append("The function causing this error was:") |
| for line in body: |
| error.append(line) |
| error.append("%s: %s" % (e.__class__.__name__, str(e))) |
| |
| logger.error("\n".join(error)) |
| |
| e = bb.BBHandledException(e) |
| raise e |
| |
| def _print_exception(t, value, tb, realfile, text, context): |
| error = [] |
| try: |
| exception = traceback.format_exception_only(t, value) |
| error.append('Error executing a python function in %s:\n' % realfile) |
| |
| # Strip 'us' from the stack (better_exec call) unless that was where the |
| # error came from |
| if tb.tb_next is not None: |
| tb = tb.tb_next |
| |
| textarray = text.split('\n') |
| |
| linefailed = tb.tb_lineno |
| |
| tbextract = traceback.extract_tb(tb) |
| tbformat = traceback.format_list(tbextract) |
| error.append("The stack trace of python calls that resulted in this exception/failure was:") |
| error.append("File: '%s', lineno: %s, function: %s" % (tbextract[0][0], tbextract[0][1], tbextract[0][2])) |
| error.extend(_print_trace(textarray, linefailed)) |
| |
| # See if this is a function we constructed and has calls back into other functions in |
| # "text". If so, try and improve the context of the error by diving down the trace |
| level = 0 |
| nexttb = tb.tb_next |
| while nexttb is not None and (level+1) < len(tbextract): |
| error.append("File: '%s', lineno: %s, function: %s" % (tbextract[level+1][0], tbextract[level+1][1], tbextract[level+1][2])) |
| if tbextract[level][0] == tbextract[level+1][0] and tbextract[level+1][2] == tbextract[level][0]: |
| # The code was possibly in the string we compiled ourselves |
| error.extend(_print_trace(textarray, tbextract[level+1][1])) |
| elif tbextract[level+1][0].startswith("/"): |
| # The code looks like it might be in a file, try and load it |
| try: |
| with open(tbextract[level+1][0], "r") as f: |
| text = f.readlines() |
| error.extend(_print_trace(text, tbextract[level+1][1])) |
| except: |
| error.append(tbformat[level+1]) |
| else: |
| error.append(tbformat[level+1]) |
| nexttb = tb.tb_next |
| level = level + 1 |
| |
| error.append("Exception: %s" % ''.join(exception)) |
| |
| # If the exception is from spawning a task, let's be helpful and display |
| # the output (which hopefully includes stderr). |
| if isinstance(value, subprocess.CalledProcessError) and value.output: |
| error.append("Subprocess output:") |
| error.append(value.output.decode("utf-8", errors="ignore")) |
| finally: |
| logger.error("\n".join(error)) |
| |
| def better_exec(code, context, text = None, realfile = "<code>", pythonexception=False): |
| """ |
| Similiar to better_compile, better_exec will |
| print the lines that are responsible for the |
| error. |
| """ |
| import bb.parse |
| if not text: |
| text = code |
| if not hasattr(code, "co_filename"): |
| code = better_compile(code, realfile, realfile) |
| try: |
| exec(code, get_context(), context) |
| except (bb.BBHandledException, bb.parse.SkipRecipe, bb.data_smart.ExpansionError, bb.process.ExecutionError): |
| # Error already shown so passthrough, no need for traceback |
| raise |
| except Exception as e: |
| if pythonexception: |
| raise |
| (t, value, tb) = sys.exc_info() |
| try: |
| _print_exception(t, value, tb, realfile, text, context) |
| except Exception as e2: |
| logger.error("Exception handler error: %s" % str(e2)) |
| |
| e = bb.BBHandledException(e) |
| raise e |
| |
| def simple_exec(code, context): |
| exec(code, get_context(), context) |
| |
| def better_eval(source, locals, extraglobals = None): |
| ctx = get_context() |
| if extraglobals: |
| ctx = copy.copy(ctx) |
| for g in extraglobals: |
| ctx[g] = extraglobals[g] |
| return eval(source, ctx, locals) |
| |
| @contextmanager |
| def fileslocked(files): |
| """Context manager for locking and unlocking file locks.""" |
| locks = [] |
| if files: |
| for lockfile in files: |
| locks.append(bb.utils.lockfile(lockfile)) |
| |
| try: |
| yield |
| finally: |
| for lock in locks: |
| bb.utils.unlockfile(lock) |
| |
| def lockfile(name, shared=False, retry=True, block=False): |
| """ |
| Use the specified file as a lock file, return when the lock has |
| been acquired. Returns a variable to pass to unlockfile(). |
| Parameters: |
| retry: True to re-try locking if it fails, False otherwise |
| block: True to block until the lock succeeds, False otherwise |
| The retry and block parameters are kind of equivalent unless you |
| consider the possibility of sending a signal to the process to break |
| out - at which point you want block=True rather than retry=True. |
| """ |
| basename = os.path.basename(name) |
| if len(basename) > 255: |
| root, ext = os.path.splitext(basename) |
| basename = root[:255 - len(ext)] + ext |
| |
| dirname = os.path.dirname(name) |
| mkdirhier(dirname) |
| |
| name = os.path.join(dirname, basename) |
| |
| if not os.access(dirname, os.W_OK): |
| logger.error("Unable to acquire lock '%s', directory is not writable", |
| name) |
| sys.exit(1) |
| |
| op = fcntl.LOCK_EX |
| if shared: |
| op = fcntl.LOCK_SH |
| if not retry and not block: |
| op = op | fcntl.LOCK_NB |
| |
| while True: |
| # If we leave the lockfiles lying around there is no problem |
| # but we should clean up after ourselves. This gives potential |
| # for races though. To work around this, when we acquire the lock |
| # we check the file we locked was still the lock file on disk. |
| # by comparing inode numbers. If they don't match or the lockfile |
| # no longer exists, we start again. |
| |
| # This implementation is unfair since the last person to request the |
| # lock is the most likely to win it. |
| |
| try: |
| lf = open(name, 'a+') |
| fileno = lf.fileno() |
| fcntl.flock(fileno, op) |
| statinfo = os.fstat(fileno) |
| if os.path.exists(lf.name): |
| statinfo2 = os.stat(lf.name) |
| if statinfo.st_ino == statinfo2.st_ino: |
| return lf |
| lf.close() |
| except OSError as e: |
| if e.errno == errno.EACCES or e.errno == errno.ENAMETOOLONG: |
| logger.error("Unable to acquire lock '%s', %s", |
| e.strerror, name) |
| sys.exit(1) |
| try: |
| lf.close() |
| except Exception: |
| pass |
| pass |
| if not retry: |
| return None |
| |
| def unlockfile(lf): |
| """ |
| Unlock a file locked using lockfile() |
| """ |
| try: |
| # If we had a shared lock, we need to promote to exclusive before |
| # removing the lockfile. Attempt this, ignore failures. |
| fcntl.flock(lf.fileno(), fcntl.LOCK_EX|fcntl.LOCK_NB) |
| os.unlink(lf.name) |
| except (IOError, OSError): |
| pass |
| fcntl.flock(lf.fileno(), fcntl.LOCK_UN) |
| lf.close() |
| |
| def _hasher(method, filename): |
| import mmap |
| |
| with open(filename, "rb") as f: |
| try: |
| with mmap.mmap(f.fileno(), 0, access=mmap.ACCESS_READ) as mm: |
| for chunk in iter(lambda: mm.read(8192), b''): |
| method.update(chunk) |
| except ValueError: |
| # You can't mmap() an empty file so silence this exception |
| pass |
| return method.hexdigest() |
| |
| |
| def md5_file(filename): |
| """ |
| Return the hex string representation of the MD5 checksum of filename. |
| """ |
| import hashlib |
| return _hasher(hashlib.new('MD5', usedforsecurity=False), filename) |
| |
| def sha256_file(filename): |
| """ |
| Return the hex string representation of the 256-bit SHA checksum of |
| filename. |
| """ |
| import hashlib |
| return _hasher(hashlib.sha256(), filename) |
| |
| def sha1_file(filename): |
| """ |
| Return the hex string representation of the SHA1 checksum of the filename |
| """ |
| import hashlib |
| return _hasher(hashlib.sha1(), filename) |
| |
| def sha384_file(filename): |
| """ |
| Return the hex string representation of the SHA384 checksum of the filename |
| """ |
| import hashlib |
| return _hasher(hashlib.sha384(), filename) |
| |
| def sha512_file(filename): |
| """ |
| Return the hex string representation of the SHA512 checksum of the filename |
| """ |
| import hashlib |
| return _hasher(hashlib.sha512(), filename) |
| |
| def preserved_envvars_exported(): |
| """Variables which are taken from the environment and placed in and exported |
| from the metadata""" |
| return [ |
| 'BB_TASKHASH', |
| 'HOME', |
| 'LOGNAME', |
| 'PATH', |
| 'PWD', |
| 'SHELL', |
| 'USER', |
| 'LC_ALL', |
| 'BBSERVER', |
| ] |
| |
| def preserved_envvars(): |
| """Variables which are taken from the environment and placed in the metadata""" |
| v = [ |
| 'BBPATH', |
| 'BB_PRESERVE_ENV', |
| 'BB_ENV_PASSTHROUGH', |
| 'BB_ENV_PASSTHROUGH_ADDITIONS', |
| ] |
| return v + preserved_envvars_exported() |
| |
| def filter_environment(good_vars): |
| """ |
| Create a pristine environment for bitbake. This will remove variables that |
| are not known and may influence the build in a negative way. |
| """ |
| |
| removed_vars = {} |
| for key in list(os.environ): |
| if key in good_vars: |
| continue |
| |
| removed_vars[key] = os.environ[key] |
| del os.environ[key] |
| |
| # If we spawn a python process, we need to have a UTF-8 locale, else python's file |
| # access methods will use ascii. You can't change that mode once the interpreter is |
| # started so we have to ensure a locale is set. Ideally we'd use C.UTF-8 but not all |
| # distros support that and we need to set something. |
| os.environ["LC_ALL"] = "en_US.UTF-8" |
| |
| if removed_vars: |
| logger.debug("Removed the following variables from the environment: %s", ", ".join(removed_vars.keys())) |
| |
| return removed_vars |
| |
| def approved_variables(): |
| """ |
| Determine and return the list of variables which are approved |
| to remain in the environment. |
| """ |
| if 'BB_PRESERVE_ENV' in os.environ: |
| return os.environ.keys() |
| approved = [] |
| if 'BB_ENV_PASSTHROUGH' in os.environ: |
| approved = os.environ['BB_ENV_PASSTHROUGH'].split() |
| approved.extend(['BB_ENV_PASSTHROUGH']) |
| else: |
| approved = preserved_envvars() |
| if 'BB_ENV_PASSTHROUGH_ADDITIONS' in os.environ: |
| approved.extend(os.environ['BB_ENV_PASSTHROUGH_ADDITIONS'].split()) |
| if 'BB_ENV_PASSTHROUGH_ADDITIONS' not in approved: |
| approved.extend(['BB_ENV_PASSTHROUGH_ADDITIONS']) |
| return approved |
| |
| def clean_environment(): |
| """ |
| Clean up any spurious environment variables. This will remove any |
| variables the user hasn't chosen to preserve. |
| """ |
| if 'BB_PRESERVE_ENV' not in os.environ: |
| good_vars = approved_variables() |
| return filter_environment(good_vars) |
| |
| return {} |
| |
| def empty_environment(): |
| """ |
| Remove all variables from the environment. |
| """ |
| for s in list(os.environ.keys()): |
| os.unsetenv(s) |
| del os.environ[s] |
| |
| def build_environment(d): |
| """ |
| Build an environment from all exported variables. |
| """ |
| import bb.data |
| for var in bb.data.keys(d): |
| export = d.getVarFlag(var, "export", False) |
| if export: |
| os.environ[var] = d.getVar(var) or "" |
| |
| def _check_unsafe_delete_path(path): |
| """ |
| Basic safeguard against recursively deleting something we shouldn't. If it returns True, |
| the caller should raise an exception with an appropriate message. |
| NOTE: This is NOT meant to be a security mechanism - just a guard against silly mistakes |
| with potentially disastrous results. |
| """ |
| extra = '' |
| # HOME might not be /home/something, so in case we can get it, check against it |
| homedir = os.environ.get('HOME', '') |
| if homedir: |
| extra = '|%s' % homedir |
| if re.match('(/|//|/home|/home/[^/]*%s)$' % extra, os.path.abspath(path)): |
| return True |
| return False |
| |
| def remove(path, recurse=False, ionice=False): |
| """Equivalent to rm -f or rm -rf""" |
| if not path: |
| return |
| if recurse: |
| for name in glob.glob(path): |
| if _check_unsafe_delete_path(name): |
| raise Exception('bb.utils.remove: called with dangerous path "%s" and recurse=True, refusing to delete!' % name) |
| # shutil.rmtree(name) would be ideal but its too slow |
| cmd = [] |
| if ionice: |
| cmd = ['ionice', '-c', '3'] |
| subprocess.check_call(cmd + ['rm', '-rf'] + glob.glob(path)) |
| return |
| for name in glob.glob(path): |
| try: |
| os.unlink(name) |
| except OSError as exc: |
| if exc.errno != errno.ENOENT: |
| raise |
| |
| def prunedir(topdir, ionice=False): |
| """ Delete everything reachable from the directory named in 'topdir'. """ |
| # CAUTION: This is dangerous! |
| if _check_unsafe_delete_path(topdir): |
| raise Exception('bb.utils.prunedir: called with dangerous path "%s", refusing to delete!' % topdir) |
| remove(topdir, recurse=True, ionice=ionice) |
| |
| # |
| # Could also use return re.compile("(%s)" % "|".join(map(re.escape, suffixes))).sub(lambda mo: "", var) |
| # but thats possibly insane and suffixes is probably going to be small |
| # |
| def prune_suffix(var, suffixes, d): |
| """ |
| See if var ends with any of the suffixes listed and |
| remove it if found |
| """ |
| for suffix in suffixes: |
| if suffix and var.endswith(suffix): |
| return var[:-len(suffix)] |
| return var |
| |
| def mkdirhier(directory): |
| """Create a directory like 'mkdir -p', but does not complain if |
| directory already exists like os.makedirs |
| """ |
| |
| try: |
| os.makedirs(directory) |
| except OSError as e: |
| if e.errno != errno.EEXIST or not os.path.isdir(directory): |
| raise e |
| |
| def movefile(src, dest, newmtime = None, sstat = None): |
| """Moves a file from src to dest, preserving all permissions and |
| attributes; mtime will be preserved even when moving across |
| filesystems. Returns true on success and false on failure. Move is |
| atomic. |
| """ |
| |
| #print "movefile(" + src + "," + dest + "," + str(newmtime) + "," + str(sstat) + ")" |
| try: |
| if not sstat: |
| sstat = os.lstat(src) |
| except Exception as e: |
| logger.warning("movefile: Stating source file failed...", e) |
| return None |
| |
| destexists = 1 |
| try: |
| dstat = os.lstat(dest) |
| except: |
| dstat = os.lstat(os.path.dirname(dest)) |
| destexists = 0 |
| |
| if destexists: |
| if stat.S_ISLNK(dstat[stat.ST_MODE]): |
| try: |
| os.unlink(dest) |
| destexists = 0 |
| except Exception as e: |
| pass |
| |
| if stat.S_ISLNK(sstat[stat.ST_MODE]): |
| try: |
| target = os.readlink(src) |
| if destexists and not stat.S_ISDIR(dstat[stat.ST_MODE]): |
| os.unlink(dest) |
| os.symlink(target, dest) |
| #os.lchown(dest,sstat[stat.ST_UID],sstat[stat.ST_GID]) |
| os.unlink(src) |
| return os.lstat(dest) |
| except Exception as e: |
| logger.warning("movefile: failed to properly create symlink:", dest, "->", target, e) |
| return None |
| |
| renamefailed = 1 |
| # os.rename needs to know the dest path ending with file name |
| # so append the file name to a path only if it's a dir specified |
| srcfname = os.path.basename(src) |
| destpath = os.path.join(dest, srcfname) if os.path.isdir(dest) \ |
| else dest |
| |
| if sstat[stat.ST_DEV] == dstat[stat.ST_DEV]: |
| try: |
| bb.utils.rename(src, destpath) |
| renamefailed = 0 |
| except Exception as e: |
| if e.errno != errno.EXDEV: |
| # Some random error. |
| logger.warning("movefile: Failed to move", src, "to", dest, e) |
| return None |
| # Invalid cross-device-link 'bind' mounted or actually Cross-Device |
| |
| if renamefailed: |
| didcopy = 0 |
| if stat.S_ISREG(sstat[stat.ST_MODE]): |
| try: # For safety copy then move it over. |
| shutil.copyfile(src, destpath + "#new") |
| bb.utils.rename(destpath + "#new", destpath) |
| didcopy = 1 |
| except Exception as e: |
| logger.warning('movefile: copy', src, '->', dest, 'failed.', e) |
| return None |
| else: |
| #we don't yet handle special, so we need to fall back to /bin/mv |
| a = getstatusoutput("/bin/mv -f " + "'" + src + "' '" + dest + "'") |
| if a[0] != 0: |
| logger.warning("movefile: Failed to move special file:" + src + "' to '" + dest + "'", a) |
| return None # failure |
| try: |
| if didcopy: |
| os.lchown(destpath, sstat[stat.ST_UID], sstat[stat.ST_GID]) |
| os.chmod(destpath, stat.S_IMODE(sstat[stat.ST_MODE])) # Sticky is reset on chown |
| os.unlink(src) |
| except Exception as e: |
| logger.warning("movefile: Failed to chown/chmod/unlink", dest, e) |
| return None |
| |
| if newmtime: |
| os.utime(destpath, (newmtime, newmtime)) |
| else: |
| os.utime(destpath, (sstat[stat.ST_ATIME], sstat[stat.ST_MTIME])) |
| newmtime = sstat[stat.ST_MTIME] |
| return newmtime |
| |
| def copyfile(src, dest, newmtime = None, sstat = None): |
| """ |
| Copies a file from src to dest, preserving all permissions and |
| attributes; mtime will be preserved even when moving across |
| filesystems. Returns true on success and false on failure. |
| """ |
| #print "copyfile(" + src + "," + dest + "," + str(newmtime) + "," + str(sstat) + ")" |
| try: |
| if not sstat: |
| sstat = os.lstat(src) |
| except Exception as e: |
| logger.warning("copyfile: stat of %s failed (%s)" % (src, e)) |
| return False |
| |
| destexists = 1 |
| try: |
| dstat = os.lstat(dest) |
| except: |
| dstat = os.lstat(os.path.dirname(dest)) |
| destexists = 0 |
| |
| if destexists: |
| if stat.S_ISLNK(dstat[stat.ST_MODE]): |
| try: |
| os.unlink(dest) |
| destexists = 0 |
| except Exception as e: |
| pass |
| |
| if stat.S_ISLNK(sstat[stat.ST_MODE]): |
| try: |
| target = os.readlink(src) |
| if destexists and not stat.S_ISDIR(dstat[stat.ST_MODE]): |
| os.unlink(dest) |
| os.symlink(target, dest) |
| os.lchown(dest,sstat[stat.ST_UID],sstat[stat.ST_GID]) |
| return os.lstat(dest) |
| except Exception as e: |
| logger.warning("copyfile: failed to create symlink %s to %s (%s)" % (dest, target, e)) |
| return False |
| |
| if stat.S_ISREG(sstat[stat.ST_MODE]): |
| try: |
| srcchown = False |
| if not os.access(src, os.R_OK): |
| # Make sure we can read it |
| srcchown = True |
| os.chmod(src, sstat[stat.ST_MODE] | stat.S_IRUSR) |
| |
| # For safety copy then move it over. |
| shutil.copyfile(src, dest + "#new") |
| bb.utils.rename(dest + "#new", dest) |
| except Exception as e: |
| logger.warning("copyfile: copy %s to %s failed (%s)" % (src, dest, e)) |
| return False |
| finally: |
| if srcchown: |
| os.chmod(src, sstat[stat.ST_MODE]) |
| os.utime(src, (sstat[stat.ST_ATIME], sstat[stat.ST_MTIME])) |
| |
| else: |
| #we don't yet handle special, so we need to fall back to /bin/mv |
| a = getstatusoutput("/bin/cp -f " + "'" + src + "' '" + dest + "'") |
| if a[0] != 0: |
| logger.warning("copyfile: failed to copy special file %s to %s (%s)" % (src, dest, a)) |
| return False # failure |
| try: |
| os.lchown(dest, sstat[stat.ST_UID], sstat[stat.ST_GID]) |
| os.chmod(dest, stat.S_IMODE(sstat[stat.ST_MODE])) # Sticky is reset on chown |
| except Exception as e: |
| logger.warning("copyfile: failed to chown/chmod %s (%s)" % (dest, e)) |
| return False |
| |
| if newmtime: |
| os.utime(dest, (newmtime, newmtime)) |
| else: |
| os.utime(dest, (sstat[stat.ST_ATIME], sstat[stat.ST_MTIME])) |
| newmtime = sstat[stat.ST_MTIME] |
| return newmtime |
| |
| def break_hardlinks(src, sstat = None): |
| """ |
| Ensures src is the only hardlink to this file. Other hardlinks, |
| if any, are not affected (other than in their st_nlink value, of |
| course). Returns true on success and false on failure. |
| |
| """ |
| try: |
| if not sstat: |
| sstat = os.lstat(src) |
| except Exception as e: |
| logger.warning("break_hardlinks: stat of %s failed (%s)" % (src, e)) |
| return False |
| if sstat[stat.ST_NLINK] == 1: |
| return True |
| return copyfile(src, src, sstat=sstat) |
| |
| def which(path, item, direction = 0, history = False, executable=False): |
| """ |
| Locate `item` in the list of paths `path` (colon separated string like $PATH). |
| If `direction` is non-zero then the list is reversed. |
| If `history` is True then the list of candidates also returned as result,history. |
| If `executable` is True then the candidate has to be an executable file, |
| otherwise the candidate simply has to exist. |
| """ |
| |
| if executable: |
| is_candidate = lambda p: os.path.isfile(p) and os.access(p, os.X_OK) |
| else: |
| is_candidate = lambda p: os.path.exists(p) |
| |
| hist = [] |
| paths = (path or "").split(':') |
| if direction != 0: |
| paths.reverse() |
| |
| for p in paths: |
| next = os.path.join(p, item) |
| hist.append(next) |
| if is_candidate(next): |
| if not os.path.isabs(next): |
| next = os.path.abspath(next) |
| if history: |
| return next, hist |
| return next |
| |
| if history: |
| return "", hist |
| return "" |
| |
| @contextmanager |
| def umask(new_mask): |
| """ |
| Context manager to set the umask to a specific mask, and restore it afterwards. |
| """ |
| current_mask = os.umask(new_mask) |
| try: |
| yield |
| finally: |
| os.umask(current_mask) |
| |
| def to_boolean(string, default=None): |
| """ |
| Check input string and return boolean value True/False/None |
| depending upon the checks |
| """ |
| if not string: |
| return default |
| |
| normalized = string.lower() |
| if normalized in ("y", "yes", "1", "true"): |
| return True |
| elif normalized in ("n", "no", "0", "false"): |
| return False |
| else: |
| raise ValueError("Invalid value for to_boolean: %s" % string) |
| |
| def contains(variable, checkvalues, truevalue, falsevalue, d): |
| """Check if a variable contains all the values specified. |
| |
| Arguments: |
| |
| variable -- the variable name. This will be fetched and expanded (using |
| d.getVar(variable)) and then split into a set(). |
| |
| checkvalues -- if this is a string it is split on whitespace into a set(), |
| otherwise coerced directly into a set(). |
| |
| truevalue -- the value to return if checkvalues is a subset of variable. |
| |
| falsevalue -- the value to return if variable is empty or if checkvalues is |
| not a subset of variable. |
| |
| d -- the data store. |
| """ |
| |
| val = d.getVar(variable) |
| if not val: |
| return falsevalue |
| val = set(val.split()) |
| if isinstance(checkvalues, str): |
| checkvalues = set(checkvalues.split()) |
| else: |
| checkvalues = set(checkvalues) |
| if checkvalues.issubset(val): |
| return truevalue |
| return falsevalue |
| |
| def contains_any(variable, checkvalues, truevalue, falsevalue, d): |
| """Check if a variable contains any values specified. |
| |
| Arguments: |
| |
| variable -- the variable name. This will be fetched and expanded (using |
| d.getVar(variable)) and then split into a set(). |
| |
| checkvalues -- if this is a string it is split on whitespace into a set(), |
| otherwise coerced directly into a set(). |
| |
| truevalue -- the value to return if checkvalues is a subset of variable. |
| |
| falsevalue -- the value to return if variable is empty or if checkvalues is |
| not a subset of variable. |
| |
| d -- the data store. |
| """ |
| val = d.getVar(variable) |
| if not val: |
| return falsevalue |
| val = set(val.split()) |
| if isinstance(checkvalues, str): |
| checkvalues = set(checkvalues.split()) |
| else: |
| checkvalues = set(checkvalues) |
| if checkvalues & val: |
| return truevalue |
| return falsevalue |
| |
| def filter(variable, checkvalues, d): |
| """Return all words in the variable that are present in the checkvalues. |
| |
| Arguments: |
| |
| variable -- the variable name. This will be fetched and expanded (using |
| d.getVar(variable)) and then split into a set(). |
| |
| checkvalues -- if this is a string it is split on whitespace into a set(), |
| otherwise coerced directly into a set(). |
| |
| d -- the data store. |
| """ |
| |
| val = d.getVar(variable) |
| if not val: |
| return '' |
| val = set(val.split()) |
| if isinstance(checkvalues, str): |
| checkvalues = set(checkvalues.split()) |
| else: |
| checkvalues = set(checkvalues) |
| return ' '.join(sorted(checkvalues & val)) |
| |
| |
| def get_referenced_vars(start_expr, d): |
| """ |
| :return: names of vars referenced in start_expr (recursively), in quasi-BFS order (variables within the same level |
| are ordered arbitrarily) |
| """ |
| |
| seen = set() |
| ret = [] |
| |
| # The first entry in the queue is the unexpanded start expression |
| queue = collections.deque([start_expr]) |
| # Subsequent entries will be variable names, so we need to track whether or not entry requires getVar |
| is_first = True |
| |
| empty_data = bb.data.init() |
| while queue: |
| entry = queue.popleft() |
| if is_first: |
| # Entry is the start expression - no expansion needed |
| is_first = False |
| expression = entry |
| else: |
| # This is a variable name - need to get the value |
| expression = d.getVar(entry, False) |
| ret.append(entry) |
| |
| # expandWithRefs is how we actually get the referenced variables in the expression. We call it using an empty |
| # data store because we only want the variables directly used in the expression. It returns a set, which is what |
| # dooms us to only ever be "quasi-BFS" rather than full BFS. |
| new_vars = empty_data.expandWithRefs(expression, None).references - set(seen) |
| |
| queue.extend(new_vars) |
| seen.update(new_vars) |
| return ret |
| |
| |
| def cpu_count(): |
| return multiprocessing.cpu_count() |
| |
| def nonblockingfd(fd): |
| fcntl.fcntl(fd, fcntl.F_SETFL, fcntl.fcntl(fd, fcntl.F_GETFL) | os.O_NONBLOCK) |
| |
| def process_profilelog(fn, pout = None): |
| # Either call with a list of filenames and set pout or a filename and optionally pout. |
| if not pout: |
| pout = fn + '.processed' |
| |
| with open(pout, 'w') as pout: |
| import pstats |
| if isinstance(fn, list): |
| p = pstats.Stats(*fn, stream=pout) |
| else: |
| p = pstats.Stats(fn, stream=pout) |
| p.sort_stats('time') |
| p.print_stats() |
| p.print_callers() |
| p.sort_stats('cumulative') |
| p.print_stats() |
| |
| pout.flush() |
| |
| # |
| # Was present to work around multiprocessing pool bugs in python < 2.7.3 |
| # |
| def multiprocessingpool(*args, **kwargs): |
| |
| import multiprocessing.pool |
| #import multiprocessing.util |
| #multiprocessing.util.log_to_stderr(10) |
| # Deal with a multiprocessing bug where signals to the processes would be delayed until the work |
| # completes. Putting in a timeout means the signals (like SIGINT/SIGTERM) get processed. |
| def wrapper(func): |
| def wrap(self, timeout=None): |
| return func(self, timeout=timeout if timeout is not None else 1e100) |
| return wrap |
| multiprocessing.pool.IMapIterator.next = wrapper(multiprocessing.pool.IMapIterator.next) |
| |
| return multiprocessing.Pool(*args, **kwargs) |
| |
| def exec_flat_python_func(func, *args, **kwargs): |
| """Execute a flat python function (defined with def funcname(args):...)""" |
| # Prepare a small piece of python code which calls the requested function |
| # To do this we need to prepare two things - a set of variables we can use to pass |
| # the values of arguments into the calling function, and the list of arguments for |
| # the function being called |
| context = {} |
| funcargs = [] |
| # Handle unnamed arguments |
| aidx = 1 |
| for arg in args: |
| argname = 'arg_%s' % aidx |
| context[argname] = arg |
| funcargs.append(argname) |
| aidx += 1 |
| # Handle keyword arguments |
| context.update(kwargs) |
| funcargs.extend(['%s=%s' % (arg, arg) for arg in kwargs.keys()]) |
| code = 'retval = %s(%s)' % (func, ', '.join(funcargs)) |
| comp = bb.utils.better_compile(code, '<string>', '<string>') |
| bb.utils.better_exec(comp, context, code, '<string>') |
| return context['retval'] |
| |
| def edit_metadata(meta_lines, variables, varfunc, match_overrides=False): |
| """Edit lines from a recipe or config file and modify one or more |
| specified variable values set in the file using a specified callback |
| function. Lines are expected to have trailing newlines. |
| Parameters: |
| meta_lines: lines from the file; can be a list or an iterable |
| (e.g. file pointer) |
| variables: a list of variable names to look for. Functions |
| may also be specified, but must be specified with '()' at |
| the end of the name. Note that the function doesn't have |
| any intrinsic understanding of :append, :prepend, :remove, |
| or overrides, so these are considered as part of the name. |
| These values go into a regular expression, so regular |
| expression syntax is allowed. |
| varfunc: callback function called for every variable matching |
| one of the entries in the variables parameter. The function |
| should take four arguments: |
| varname: name of variable matched |
| origvalue: current value in file |
| op: the operator (e.g. '+=') |
| newlines: list of lines up to this point. You can use |
| this to prepend lines before this variable setting |
| if you wish. |
| and should return a four-element tuple: |
| newvalue: new value to substitute in, or None to drop |
| the variable setting entirely. (If the removal |
| results in two consecutive blank lines, one of the |
| blank lines will also be dropped). |
| newop: the operator to use - if you specify None here, |
| the original operation will be used. |
| indent: number of spaces to indent multi-line entries, |
| or -1 to indent up to the level of the assignment |
| and opening quote, or a string to use as the indent. |
| minbreak: True to allow the first element of a |
| multi-line value to continue on the same line as |
| the assignment, False to indent before the first |
| element. |
| To clarify, if you wish not to change the value, then you |
| would return like this: return origvalue, None, 0, True |
| match_overrides: True to match items with _overrides on the end, |
| False otherwise |
| Returns a tuple: |
| updated: |
| True if changes were made, False otherwise. |
| newlines: |
| Lines after processing |
| """ |
| |
| var_res = {} |
| if match_overrides: |
| override_re = r'(_[a-zA-Z0-9-_$(){}]+)?' |
| else: |
| override_re = '' |
| for var in variables: |
| if var.endswith('()'): |
| var_res[var] = re.compile(r'^(%s%s)[ \\t]*\([ \\t]*\)[ \\t]*{' % (var[:-2].rstrip(), override_re)) |
| else: |
| var_res[var] = re.compile(r'^(%s%s)[ \\t]*[?+:.]*=[+.]*[ \\t]*(["\'])' % (var, override_re)) |
| |
| updated = False |
| varset_start = '' |
| varlines = [] |
| newlines = [] |
| in_var = None |
| full_value = '' |
| var_end = '' |
| |
| def handle_var_end(): |
| prerun_newlines = newlines[:] |
| op = varset_start[len(in_var):].strip() |
| (newvalue, newop, indent, minbreak) = varfunc(in_var, full_value, op, newlines) |
| changed = (prerun_newlines != newlines) |
| |
| if newvalue is None: |
| # Drop the value |
| return True |
| elif newvalue != full_value or (newop not in [None, op]): |
| if newop not in [None, op]: |
| # Callback changed the operator |
| varset_new = "%s %s" % (in_var, newop) |
| else: |
| varset_new = varset_start |
| |
| if isinstance(indent, int): |
| if indent == -1: |
| indentspc = ' ' * (len(varset_new) + 2) |
| else: |
| indentspc = ' ' * indent |
| else: |
| indentspc = indent |
| if in_var.endswith('()'): |
| # A function definition |
| if isinstance(newvalue, list): |
| newlines.append('%s {\n%s%s\n}\n' % (varset_new, indentspc, ('\n%s' % indentspc).join(newvalue))) |
| else: |
| if not newvalue.startswith('\n'): |
| newvalue = '\n' + newvalue |
| if not newvalue.endswith('\n'): |
| newvalue = newvalue + '\n' |
| newlines.append('%s {%s}\n' % (varset_new, newvalue)) |
| else: |
| # Normal variable |
| if isinstance(newvalue, list): |
| if not newvalue: |
| # Empty list -> empty string |
| newlines.append('%s ""\n' % varset_new) |
| elif minbreak: |
| # First item on first line |
| if len(newvalue) == 1: |
| newlines.append('%s "%s"\n' % (varset_new, newvalue[0])) |
| else: |
| newlines.append('%s "%s \\\n' % (varset_new, newvalue[0])) |
| for item in newvalue[1:]: |
| newlines.append('%s%s \\\n' % (indentspc, item)) |
| newlines.append('%s"\n' % indentspc) |
| else: |
| # No item on first line |
| newlines.append('%s " \\\n' % varset_new) |
| for item in newvalue: |
| newlines.append('%s%s \\\n' % (indentspc, item)) |
| newlines.append('%s"\n' % indentspc) |
| else: |
| newlines.append('%s "%s"\n' % (varset_new, newvalue)) |
| return True |
| else: |
| # Put the old lines back where they were |
| newlines.extend(varlines) |
| # If newlines was touched by the function, we'll need to return True |
| return changed |
| |
| checkspc = False |
| |
| for line in meta_lines: |
| if in_var: |
| value = line.rstrip() |
| varlines.append(line) |
| if in_var.endswith('()'): |
| full_value += '\n' + value |
| else: |
| full_value += value[:-1] |
| if value.endswith(var_end): |
| if in_var.endswith('()'): |
| if full_value.count('{') - full_value.count('}') >= 0: |
| continue |
| full_value = full_value[:-1] |
| if handle_var_end(): |
| updated = True |
| checkspc = True |
| in_var = None |
| else: |
| skip = False |
| for (varname, var_re) in var_res.items(): |
| res = var_re.match(line) |
| if res: |
| isfunc = varname.endswith('()') |
| if isfunc: |
| splitvalue = line.split('{', 1) |
| var_end = '}' |
| else: |
| var_end = res.groups()[-1] |
| splitvalue = line.split(var_end, 1) |
| varset_start = splitvalue[0].rstrip() |
| value = splitvalue[1].rstrip() |
| if not isfunc and value.endswith('\\'): |
| value = value[:-1] |
| full_value = value |
| varlines = [line] |
| in_var = res.group(1) |
| if isfunc: |
| in_var += '()' |
| if value.endswith(var_end): |
| full_value = full_value[:-1] |
| if handle_var_end(): |
| updated = True |
| checkspc = True |
| in_var = None |
| skip = True |
| break |
| if not skip: |
| if checkspc: |
| checkspc = False |
| if newlines and newlines[-1] == '\n' and line == '\n': |
| # Squash blank line if there are two consecutive blanks after a removal |
| continue |
| newlines.append(line) |
| return (updated, newlines) |
| |
| |
| def edit_metadata_file(meta_file, variables, varfunc): |
| """Edit a recipe or config file and modify one or more specified |
| variable values set in the file using a specified callback function. |
| The file is only written to if the value(s) actually change. |
| This is basically the file version of edit_metadata(), see that |
| function's description for parameter/usage information. |
| Returns True if the file was written to, False otherwise. |
| """ |
| with open(meta_file, 'r') as f: |
| (updated, newlines) = edit_metadata(f, variables, varfunc) |
| if updated: |
| with open(meta_file, 'w') as f: |
| f.writelines(newlines) |
| return updated |
| |
| |
| def edit_bblayers_conf(bblayers_conf, add, remove, edit_cb=None): |
| """Edit bblayers.conf, adding and/or removing layers |
| Parameters: |
| bblayers_conf: path to bblayers.conf file to edit |
| add: layer path (or list of layer paths) to add; None or empty |
| list to add nothing |
| remove: layer path (or list of layer paths) to remove; None or |
| empty list to remove nothing |
| edit_cb: optional callback function that will be called after |
| processing adds/removes once per existing entry. |
| Returns a tuple: |
| notadded: list of layers specified to be added but weren't |
| (because they were already in the list) |
| notremoved: list of layers that were specified to be removed |
| but weren't (because they weren't in the list) |
| """ |
| |
| import fnmatch |
| |
| def remove_trailing_sep(pth): |
| if pth and pth[-1] == os.sep: |
| pth = pth[:-1] |
| return pth |
| |
| approved = bb.utils.approved_variables() |
| def canonicalise_path(pth): |
| pth = remove_trailing_sep(pth) |
| if 'HOME' in approved and '~' in pth: |
| pth = os.path.expanduser(pth) |
| return pth |
| |
| def layerlist_param(value): |
| if not value: |
| return [] |
| elif isinstance(value, list): |
| return [remove_trailing_sep(x) for x in value] |
| else: |
| return [remove_trailing_sep(value)] |
| |
| addlayers = layerlist_param(add) |
| removelayers = layerlist_param(remove) |
| |
| # Need to use a list here because we can't set non-local variables from a callback in python 2.x |
| bblayercalls = [] |
| removed = [] |
| plusequals = False |
| orig_bblayers = [] |
| |
| def handle_bblayers_firstpass(varname, origvalue, op, newlines): |
| bblayercalls.append(op) |
| if op == '=': |
| del orig_bblayers[:] |
| orig_bblayers.extend([canonicalise_path(x) for x in origvalue.split()]) |
| return (origvalue, None, 2, False) |
| |
| def handle_bblayers(varname, origvalue, op, newlines): |
| updated = False |
| bblayers = [remove_trailing_sep(x) for x in origvalue.split()] |
| if removelayers: |
| for removelayer in removelayers: |
| for layer in bblayers: |
| if fnmatch.fnmatch(canonicalise_path(layer), canonicalise_path(removelayer)): |
| updated = True |
| bblayers.remove(layer) |
| removed.append(removelayer) |
| break |
| if addlayers and not plusequals: |
| for addlayer in addlayers: |
| if addlayer not in bblayers: |
| updated = True |
| bblayers.append(addlayer) |
| del addlayers[:] |
| |
| if edit_cb: |
| newlist = [] |
| for layer in bblayers: |
| res = edit_cb(layer, canonicalise_path(layer)) |
| if res != layer: |
| newlist.append(res) |
| updated = True |
| else: |
| newlist.append(layer) |
| bblayers = newlist |
| |
| if updated: |
| if op == '+=' and not bblayers: |
| bblayers = None |
| return (bblayers, None, 2, False) |
| else: |
| return (origvalue, None, 2, False) |
| |
| with open(bblayers_conf, 'r') as f: |
| (_, newlines) = edit_metadata(f, ['BBLAYERS'], handle_bblayers_firstpass) |
| |
| if not bblayercalls: |
| raise Exception('Unable to find BBLAYERS in %s' % bblayers_conf) |
| |
| # Try to do the "smart" thing depending on how the user has laid out |
| # their bblayers.conf file |
| if bblayercalls.count('+=') > 1: |
| plusequals = True |
| |
| removelayers_canon = [canonicalise_path(layer) for layer in removelayers] |
| notadded = [] |
| for layer in addlayers: |
| layer_canon = canonicalise_path(layer) |
| if layer_canon in orig_bblayers and not layer_canon in removelayers_canon: |
| notadded.append(layer) |
| notadded_canon = [canonicalise_path(layer) for layer in notadded] |
| addlayers[:] = [layer for layer in addlayers if canonicalise_path(layer) not in notadded_canon] |
| |
| (updated, newlines) = edit_metadata(newlines, ['BBLAYERS'], handle_bblayers) |
| if addlayers: |
| # Still need to add these |
| for addlayer in addlayers: |
| newlines.append('BBLAYERS += "%s"\n' % addlayer) |
| updated = True |
| |
| if updated: |
| with open(bblayers_conf, 'w') as f: |
| f.writelines(newlines) |
| |
| notremoved = list(set(removelayers) - set(removed)) |
| |
| return (notadded, notremoved) |
| |
| def get_collection_res(d): |
| collections = (d.getVar('BBFILE_COLLECTIONS') or '').split() |
| collection_res = {} |
| for collection in collections: |
| collection_res[collection] = d.getVar('BBFILE_PATTERN_%s' % collection) or '' |
| |
| return collection_res |
| |
| |
| def get_file_layer(filename, d, collection_res={}): |
| """Determine the collection (as defined by a layer's layer.conf file) containing the specified file""" |
| if not collection_res: |
| collection_res = get_collection_res(d) |
| |
| def path_to_layer(path): |
| # Use longest path so we handle nested layers |
| matchlen = 0 |
| match = None |
| for collection, regex in collection_res.items(): |
| if len(regex) > matchlen and re.match(regex, path): |
| matchlen = len(regex) |
| match = collection |
| return match |
| |
| result = None |
| bbfiles = (d.getVar('BBFILES_PRIORITIZED') or '').split() |
| bbfilesmatch = False |
| for bbfilesentry in bbfiles: |
| if fnmatch.fnmatchcase(filename, bbfilesentry): |
| bbfilesmatch = True |
| result = path_to_layer(bbfilesentry) |
| break |
| |
| if not bbfilesmatch: |
| # Probably a bbclass |
| result = path_to_layer(filename) |
| |
| return result |
| |
| |
| # Constant taken from http://linux.die.net/include/linux/prctl.h |
| PR_SET_PDEATHSIG = 1 |
| |
| class PrCtlError(Exception): |
| pass |
| |
| def signal_on_parent_exit(signame): |
| """ |
| Trigger signame to be sent when the parent process dies |
| """ |
| signum = getattr(signal, signame) |
| # http://linux.die.net/man/2/prctl |
| result = cdll['libc.so.6'].prctl(PR_SET_PDEATHSIG, signum) |
| if result != 0: |
| raise PrCtlError('prctl failed with error code %s' % result) |
| |
| # |
| # Manually call the ioprio syscall. We could depend on other libs like psutil |
| # however this gets us enough of what we need to bitbake for now without the |
| # dependency |
| # |
| _unamearch = os.uname()[4] |
| IOPRIO_WHO_PROCESS = 1 |
| IOPRIO_CLASS_SHIFT = 13 |
| |
| def ioprio_set(who, cls, value): |
| NR_ioprio_set = None |
| if _unamearch == "x86_64": |
| NR_ioprio_set = 251 |
| elif _unamearch[0] == "i" and _unamearch[2:3] == "86": |
| NR_ioprio_set = 289 |
| elif _unamearch == "aarch64": |
| NR_ioprio_set = 30 |
| |
| if NR_ioprio_set: |
| ioprio = value | (cls << IOPRIO_CLASS_SHIFT) |
| rc = cdll['libc.so.6'].syscall(NR_ioprio_set, IOPRIO_WHO_PROCESS, who, ioprio) |
| if rc != 0: |
| raise ValueError("Unable to set ioprio, syscall returned %s" % rc) |
| else: |
| bb.warn("Unable to set IO Prio for arch %s" % _unamearch) |
| |
| def set_process_name(name): |
| from ctypes import cdll, byref, create_string_buffer |
| # This is nice to have for debugging, not essential |
| try: |
| libc = cdll.LoadLibrary('libc.so.6') |
| buf = create_string_buffer(bytes(name, 'utf-8')) |
| libc.prctl(15, byref(buf), 0, 0, 0) |
| except: |
| pass |
| |
| def disable_network(uid=None, gid=None): |
| """ |
| Disable networking in the current process if the kernel supports it, else |
| just return after logging to debug. To do this we need to create a new user |
| namespace, then map back to the original uid/gid. |
| """ |
| libc = ctypes.CDLL('libc.so.6') |
| |
| # From sched.h |
| # New user namespace |
| CLONE_NEWUSER = 0x10000000 |
| # New network namespace |
| CLONE_NEWNET = 0x40000000 |
| |
| if uid is None: |
| uid = os.getuid() |
| if gid is None: |
| gid = os.getgid() |
| |
| ret = libc.unshare(CLONE_NEWNET | CLONE_NEWUSER) |
| if ret != 0: |
| logger.debug("System doesn't suport disabling network without admin privs") |
| return |
| with open("/proc/self/uid_map", "w") as f: |
| f.write("%s %s 1" % (uid, uid)) |
| with open("/proc/self/setgroups", "w") as f: |
| f.write("deny") |
| with open("/proc/self/gid_map", "w") as f: |
| f.write("%s %s 1" % (gid, gid)) |
| |
| def export_proxies(d): |
| """ export common proxies variables from datastore to environment """ |
| import os |
| |
| variables = ['http_proxy', 'HTTP_PROXY', 'https_proxy', 'HTTPS_PROXY', |
| 'ftp_proxy', 'FTP_PROXY', 'no_proxy', 'NO_PROXY', |
| 'GIT_PROXY_COMMAND'] |
| exported = False |
| |
| for v in variables: |
| if v in os.environ.keys(): |
| exported = True |
| else: |
| v_proxy = d.getVar(v) |
| if v_proxy is not None: |
| os.environ[v] = v_proxy |
| exported = True |
| |
| return exported |
| |
| |
| def load_plugins(logger, plugins, pluginpath): |
| def load_plugin(name): |
| logger.debug('Loading plugin %s' % name) |
| spec = importlib.machinery.PathFinder.find_spec(name, path=[pluginpath] ) |
| if spec: |
| mod = importlib.util.module_from_spec(spec) |
| spec.loader.exec_module(mod) |
| return mod |
| |
| logger.debug('Loading plugins from %s...' % pluginpath) |
| |
| expanded = (glob.glob(os.path.join(pluginpath, '*' + ext)) |
| for ext in python_extensions) |
| files = itertools.chain.from_iterable(expanded) |
| names = set(os.path.splitext(os.path.basename(fn))[0] for fn in files) |
| for name in names: |
| if name != '__init__': |
| plugin = load_plugin(name) |
| if hasattr(plugin, 'plugin_init'): |
| obj = plugin.plugin_init(plugins) |
| plugins.append(obj or plugin) |
| else: |
| plugins.append(plugin) |
| |
| |
| class LogCatcher(logging.Handler): |
| """Logging handler for collecting logged messages so you can check them later""" |
| def __init__(self): |
| self.messages = [] |
| logging.Handler.__init__(self, logging.WARNING) |
| def emit(self, record): |
| self.messages.append(bb.build.logformatter.format(record)) |
| def contains(self, message): |
| return (message in self.messages) |
| |
| def is_semver(version): |
| """ |
| Is the version string following the semver semantic? |
| |
| https://semver.org/spec/v2.0.0.html |
| """ |
| regex = re.compile( |
| r""" |
| ^ |
| (0|[1-9]\d*)\.(0|[1-9]\d*)\.(0|[1-9]\d*) |
| (?:-( |
| (?:0|[1-9]\d*|\d*[a-zA-Z-][0-9a-zA-Z-]*) |
| (?:\.(?:0|[1-9]\d*|\d*[a-zA-Z-][0-9a-zA-Z-]*))* |
| ))? |
| (?:\+( |
| [0-9a-zA-Z-]+ |
| (?:\.[0-9a-zA-Z-]+)* |
| ))? |
| $ |
| """, re.VERBOSE) |
| |
| if regex.match(version) is None: |
| return False |
| |
| return True |
| |
| # Wrapper around os.rename which can handle cross device problems |
| # e.g. from container filesystems |
| def rename(src, dst): |
| try: |
| os.rename(src, dst) |
| except OSError as err: |
| if err.errno == 18: |
| # Invalid cross-device link error |
| shutil.move(src, dst) |
| else: |
| raise err |
| |
| @contextmanager |
| def environment(**envvars): |
| """ |
| Context manager to selectively update the environment with the specified mapping. |
| """ |
| backup = dict(os.environ) |
| try: |
| os.environ.update(envvars) |
| yield |
| finally: |
| for var in envvars: |
| if var in backup: |
| os.environ[var] = backup[var] |
| elif var in os.environ: |
| del os.environ[var] |
| |
| def is_local_uid(uid=''): |
| """ |
| Check whether uid is a local one or not. |
| Can't use pwd module since it gets all UIDs, not local ones only. |
| """ |
| if not uid: |
| uid = os.getuid() |
| with open('/etc/passwd', 'r') as f: |
| for line in f: |
| line_split = line.split(':') |
| if len(line_split) < 3: |
| continue |
| if str(uid) == line_split[2]: |
| return True |
| return False |