August 2014 - Arch-projects - lists.archlinux.org

[arch-projects] [mkinitcpio] [PATCH] Add sd-resume hook.
by Ivan Shapovalov 31 Oct '14

31 Oct '14

This depends on the systemd hook. It adds systemd-hibernate-resume@.service and systemd-hibernate-resume-generator which perform userspace parsing of resume= kernel parameter, allowing to specify the resume device by its persistent path ("resume=/dev/disk/by-foo/bar") or fstab-like specifier ("resume=FOO=bar"), just like the root= parameter. --- Given that relevant functionality has just landed in systemd git, it makes sense to pipeline things a bit (even if actual applying will be delayed until systemd-217). Also, does this need to be a separate hook or can be merged into 'systemd' hook itself? install/sd-resume | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) create mode 100644 install/sd-resume diff --git a/install/sd-resume b/install/sd-resume new file mode 100644 index 0000000..2995a8e --- /dev/null +++ b/install/sd-resume @@ -0,0 +1,22 @@ +#!/bin/bash + +build() { + map add_binary \ + /usr/lib/systemd/systemd-hibernate-resume \ + /usr/lib/systemd/system-generators/systemd-hibernate-resume-generator + + map add_systemd_unit \ + systemd-hibernate-resume@.service \ + local-fs-pre.target +} + +help() { + cat <<HELPEOF +This hook adds builtin systemd facility for resuming from disk. It allows to +specify the resume device by its persistent path (resume=/dev/disk/by-foo/bar) +or in fstab-like way (resume=FOO=bar), just like it's possible with the root= +parameter. +HELPEOF +} + +# vim: set ft=sh ts=4 sw=4 et: -- 2.1.0

3 7

[arch-projects] [mkinitcpio] [GIT] The official mkinitcpio repository branch master updated. v18-2-gb209e8c
by thomas＠archlinux.org 31 Aug '14

31 Aug '14

This is an automated email from the git hooks/post-receive script. It was generated because a ref change was pushed to the repository containing the project "The official mkinitcpio repository". The branch, master has been updated via b209e8c5e965965c50878fb57a628bca83a06f8d (commit) from ea4c4154205372154457c794513ae46b61ea4e4c (commit) Those revisions listed above that are new to this repository have not appeared on any other notification email; so we list those revisions in full, below. - Log ----------------------------------------------------------------- commit b209e8c5e965965c50878fb57a628bca83a06f8d Author: Thomas Bächler <thomas(a)archlinux.org> Date: Sun Aug 31 09:13:29 2014 +0200 Add 'hostdata' hook This hook adds host-specific configuration to the image. Currently, it adds the hostname and machine-id files for usage with journald. ----------------------------------------------------------------------- Summary of changes: install/hostdata | 14 ++++++++++++++ 1 file changed, 14 insertions(+) create mode 100644 install/hostdata hooks/post-receive -- The official mkinitcpio repository

1 0

[arch-projects] [namcap] [PATCH] Add py_mtime rule
by Kyle Keen 28 Aug '14

28 Aug '14

Signed-off-by: Kyle Keen <keenerd(a)gmail.com> --- Namcap/rules/__init__.py | 1 + Namcap/rules/py_mtime.py | 129 +++++++++++++++++++++++++++++++++++++++++++++++ Namcap/util.py | 28 ++++++++++ namcap-tags | 3 ++ 4 files changed, 161 insertions(+) create mode 100644 Namcap/rules/py_mtime.py diff --git a/Namcap/rules/__init__.py b/Namcap/rules/__init__.py index f7780d2..8dc4e68 100644 --- a/Namcap/rules/__init__.py +++ b/Namcap/rules/__init__.py @@ -42,6 +42,7 @@ from . import ( missingbackups, perllocal, permissions, + py_mtime, rpath, scrollkeeper, shebangdepends, diff --git a/Namcap/rules/py_mtime.py b/Namcap/rules/py_mtime.py new file mode 100644 index 0000000..aaff238 --- /dev/null +++ b/Namcap/rules/py_mtime.py @@ -0,0 +1,129 @@ +# +# namcap rules - py_mtime +# Copyright (C) 2013 Kyle Keen <keenerd(a)gmail.com> +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 2 of the License, or +# (at your option) any later version. +# +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with this program; if not, write to the Free Software +# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +# + +""" +Check for py timestamps that are ahead of pyc/pyo timestamps +""" + +import os +from Namcap.util import load_mtree +from Namcap.ruleclass import * + +def _quick_filter(names): + "can this package be skipped outright" + if not names: + return True + found_py = any(n.endswith('.py') for n in names) + found_pyc = any(n.endswith('.pyc') for n in names) + found_pyo = any(n.endswith('.pyo') for n in names) + if found_py and found_pyc: + return False + if found_py and found_pyo: + return False + return True + +def _tar_timestamps(tar): + "takes a tar object" + return dict((m.name, m.mtime) for m in tar.getmembers()) + +def _mtree_timestamps(tar): + "takes a tar object" + return dict((h, a['time']) for h,a in load_mtree(tar) if 'time' in a) + +def _generic_timestamps(tar): + "works for mtree and tar" + if '.MTREE' in tar.getnames(): + return _mtree_timestamps(tar) + return _tar_timestamps(tar) + +def _try_mtree(tar): + "returns True if good, False if bad, None if N/A" + if '.MTREE' not in tar.getnames(): + return None + stamps = _mtree_timestamps(tar) + if _quick_filter(stamps.keys()): + return True + return not _mtime_filter(stamps) + +def _try_tar(tar): + "returns True if good, False if bad" + names = tar.getnames() + if _quick_filter(names): + return True + mtimes = _tar_timestamps(tar) + return not _mtime_filter(mtimes) + +def _split_all(path): + "like os.path.split but splits every directory" + p2 = path + dirs = [] + while p2 and p2 != '/': + p2,p3 = os.path.split(p2) + dirs.insert(0, p3) + #dirs.insert(0, '/') + return dirs + +def _source_py(path): + "given a pyc/pyo, return the source path" + if not path.endswith('.pyc') and not path.endswith('.pyo'): + return None + path = path[:-1] + # handle py2 + if '__pycache__' not in path: + return path + # handle py3 + splitup = _split_all(path) + if splitup[-2] != '__pycache__': + return None + splitup.pop(-2) + f = splitup[-1] + f = f.split('.') + f.pop(-2) + splitup[-1] = '.'.join(f) + return os.path.join(*splitup) + +def _mtime_filter(mtimes): + "return list of bad py file names" + bad = [] + for name, mt2 in mtimes.items(): + if not name.endswith('.pyc') and not name.endswith('.pyo'): + continue + source_name = _source_py(name) + if source_name not in mtimes: + continue + mt1 = mtimes[source_name] + if mt1 > mt2: + bad.append(source_name) + return bad + +class package(TarballRule): + name = "py_mtime" + description = "Check for py timestamps that are ahead of pyc/pyo timestamps" + def analyze(self, pkginfo, tar): + mtree_status = _try_mtree(tar) + tar_status = _try_tar(tar) + if mtree_status == False and tar_status: + # mtree only + self.warning = [('py-mtime-mtree-warning', ())] + elif not tar_status: + # tar or both + self.errors = [('py-mtime-tar-error', ())] + self.infos = [('py-mtime-file-name %s', f[1:]) for f in _mtime_filter(_generic_timestamps(tar))] + +# vim: set ts=4 sw=4 noet: diff --git a/Namcap/util.py b/Namcap/util.py index 21d7163..0613202 100644 --- a/Namcap/util.py +++ b/Namcap/util.py @@ -20,6 +20,7 @@ import os import re import stat +import gzip def _read_carefully(path, readcall): if not os.path.isfile(path): @@ -77,4 +78,31 @@ def script_type(path): clean_filename = lambda s: re.search(r"/tmp/namcap\.[0-9]*/(.*)", s).group(1) +def _mtree_line(line): + "returns head, {key:value}" + # todo, un-hex the escaped chars + head,_,kvs = line.partition(' ') + kvs = dict(kv.split('=') for kv in kvs.split(' ')) + return head, kvs + +def load_mtree(tar): + "takes a tar object, returns (path, {attributes})" + if '.MTREE' not in tar.getnames(): + raise StopIteration + zfile = tar.extractfile('.MTREE') + text = gzip.open(zfile).read().decode("utf-8") + defaults = {} + for line in text.split('\n'): + if not line: + continue + if line.startswith('#'): + continue + head, kvs = _mtree_line(line) + if head == '/set': + defaults = kvs + attr = {} + attr.update(defaults) + attr.update(kvs) + yield head, attr + # vim: set ts=4 sw=4 noet: diff --git a/namcap-tags b/namcap-tags index d638478..8b67330 100644 --- a/namcap-tags +++ b/namcap-tags @@ -67,6 +67,9 @@ perllocal-pod-present %s :: perllocal.pod found in %s. pkgname-in-description :: Description should not contain the package name. potential-non-fhs-info-page %s :: Potential non-FHS info page (%s) found. potential-non-fhs-man-page %s :: Potential non-FHS man page (%s) found. +py-mtime-mtree-warning :: Found .py file newer (sub-second) than associated .pyc/pyo. +py-mtime-tar-error :: Found .py file newer than associated .pyc/pyo. +py-mtime-file-name %s :: Python script (%s) is newer than associated .pyc/pyo. script-link-detected %s in %s :: Script link detected (%s) in file %s scrollkeeper-dir-exists %s :: Scrollkeeper directory exists (%s). Remember to not run scrollkeeper till post_{install,upgrade,remove}. site-ruby :: Found usr/lib/ruby/site_ruby in package, usr/lib/ruby/vendor_ruby should be used instead. -- 2.0.1

2 1

[arch-projects] [namcap] [PATCH] Add py_mtime rule
by keenerd 27 Aug '14

27 Aug '14

This is a patch inspired by https://www.archlinux.org/todo/fix-bad-mtime-pycpyo-files-for-python-packag… It checks packages with py/pyc/pyo files for time conflicts. When a py file is newer than a pyc/pyo python locks up a little while figuring out what to do. Sometime it is our fault for copying files in the wrong order or doing a sed after the pyc was generated. Sometimes it is upstream's fault; many setup.py scripts copy them in the wrong order. The test checks both the tar file and the mtree logs. If there is a time error in the tar file, it will affect every installation. If the error only appears in the mtree file, then the package will be fine but there is a chance that you won't get as lucky the next time the package is built. The tar case generates an error and the mtree case generates a warning. My apologies for including the patch as an attachment. I tried to send this with 'git send-email' but every time Mailmain said "Message rejected by filter rule match" -Kyle http://kmkeen.com

1 0

[arch-projects] [mkinitcpio] [RFC] initrd: systemd-{networkd, resolved}
by Timofey Titovets 27 Aug '14

27 Aug '14

I resend mail from systemd-devel, because i think it make sense in mkinitcpio, i think it can be added by systemd hook. I can try write a patch, but before, i want to know opinions of arch developers. ---------- Forwarded message ---------- From: Lennart Poettering <lennart(a)poettering.net> Date: 2014-08-26 3:57 GMT+03:00 Subject: Re: [systemd-devel] [RFC] initrd: systemd-{networkd,resolved} To: Timofey Titovets <nefelim4ag(a)gmail.com> Копия: systemd-devel(a)lists.freedesktop.org On Sat, 23.08.14 23:26, Timofey Titovets (nefelim4ag(a)gmail.com) wrote: > Good time of day list, i just want to hear your opinions - pro and con > > Pre story: > Before i install bunch of systemd-{networkd,resolved} and > wpa_supplicant (for connecting to HotSpots) > > After i spend some time to think about network handling by networkd > and resolved, about some cases and i have idea, what if try to > implement early network connection in initrd? > Pro: > spherical cow - If mount rootfs fail to boot, or mount, we have > network access to restore system. > If system headless we can get remote access as case above and i think it cool. > It useful for nfs mount. > > Yes, this implies what we use systemd, as init in boot process. > What i miss? This is good idea or i must burn him in hell. One of the reasons systemd-networkd exists is to make it useful for running in the initrd, so that we can boot from networked volumes. It should actually pretty much work already, even though it cannot deserialize its state after the initrd transition. Lennart -- Lennart Poettering, Red Hat -- Have a nice day, Timofey.

1 0

[arch-projects] [netctl] KillMode in netctl{,@}.service
by Leonid Isaev 15 Aug '14

15 Aug '14

Hi, I am wondering what's the purpose of KillMode=none in netctl.service and netctl@.service. Is it an artifact of some sort? I think it's better to not include a KillMode line (so the default KillMode=control-group is used) because then systemd would clean processes left by "network stop" (should there be any)... Thanks, -- Leonid Isaev GPG fingerprints: DA92 034D B4A8 EC51 7EA6 20DF 9291 EE8A 043C B8C4 C0DF 20D0 C075 C3F1 E1BE 775A A7AE F6CB 164B 5A6D

2 1

[arch-projects] [netctl] [GIT] Network control utility based on systemd annotated tag 1.9 created. 1.9
by jouke＠archlinux.org 05 Aug '14

05 Aug '14

This is an automated email from the git hooks/post-receive script. It was generated because a ref change was pushed to the repository containing the project "Network control utility based on systemd". The annotated tag, 1.9 has been created at b74615dbd0180fd9fd020cdab3a2b92bcd64c6fa (tag) tagging 8b3178a79ffbf82313a316ba90270192e374f281 (commit) replaces 1.8 tagged by Jouke Witteveen on Tue Aug 5 14:27:39 2014 +0200 - Log ----------------------------------------------------------------- Release 1.9 -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQEcBAABAgAGBQJT4M3EAAoJEFa2H7+CSSwVYcsIAODOkloKQ6wBIU2YTcpr3D9v 3vHalBgDads5lUHaPTEhsm5DH0/9OxdTN5Z8+tu1ADnlbD/yT/DaYSfSh7zZgX6s Wv+v0ebuEUiRVMMyqNXf5xa5+PQvebc6zhGzQS9Btz+rUCld/uLpRSEkKfGgv3jp hPyjtCZwlZS2G+8f0pZY1f5ojkQz8OcR6loCEtpevylTEdWjVt06ArreL4CrZCMS kl8CDzBSfToDrkpg+Z45zV7y47kennK5gvzWUuirUJryu8Q8iRe5u41mG8hysSOP UvEmnhshcF6+4ZsuNNHllvPqFweZFk2heIVfLLq8chnKog/4Nfpt/LznGA50B+Q= =cuTa -----END PGP SIGNATURE----- Jouke Witteveen (6): Document support for IPv6 with dhcpcd Update tunnel connections Disable IPv6 on IP6=no Add a connection type for openvswitch interfaces Code quality 1.9 release updates ----------------------------------------------------------------------- hooks/post-receive -- Network control utility based on systemd

1 0

[arch-projects] [netctl] [GIT] Network control utility based on systemd branch master updated. 1.8-6-g8b3178a
by jouke＠archlinux.org 05 Aug '14

05 Aug '14

This is an automated email from the git hooks/post-receive script. It was generated because a ref change was pushed to the repository containing the project "Network control utility based on systemd". The branch, master has been updated via 8b3178a79ffbf82313a316ba90270192e374f281 (commit) via 810cdb650116037caf2eaa45b7f9d553b897b270 (commit) via 1d490769d87875bbc03de1c1c7b198e4f9f42b91 (commit) via 3fce1e6f67cd52afa4470c1ee7d42112d9d9405b (commit) from 0f51aa966b1afb835425f8a4c91ec38f4f36babc (commit) Those revisions listed above that are new to this repository have not appeared on any other notification email; so we list those revisions in full, below. - Log ----------------------------------------------------------------- commit 8b3178a79ffbf82313a316ba90270192e374f281 Author: Jouke Witteveen <j.witteveen(a)gmail.com> Date: Tue Aug 5 14:27:18 2014 +0200 1.9 release updates commit 810cdb650116037caf2eaa45b7f9d553b897b270 Author: Jouke Witteveen <j.witteveen(a)gmail.com> Date: Tue Aug 5 13:57:14 2014 +0200 Code quality commit 1d490769d87875bbc03de1c1c7b198e4f9f42b91 Author: Jouke Witteveen <j.witteveen(a)gmail.com> Date: Sun Jul 27 22:11:42 2014 +0200 Add a connection type for openvswitch interfaces Open vSwitch support code contributed by: Jonathan Hudson <jh+arch(a)daria.co.uk> commit 3fce1e6f67cd52afa4470c1ee7d42112d9d9405b Author: Jouke Witteveen <j.witteveen(a)gmail.com> Date: Sat Jul 5 15:56:12 2014 +0200 Disable IPv6 on IP6=no ----------------------------------------------------------------------- Summary of changes: Makefile | 2 +- NEWS | 4 ++++ docs/examples/bonding | 4 ++-- docs/examples/openvswitch | 5 +++++ docs/netctl.profile.5.txt | 9 +++++++++ src/lib/connections/openvswitch | 39 +++++++++++++++++++++++++++++++++++++++ src/lib/dhcp/dhcpcd | 4 +--- src/lib/ip | 3 ++- src/netctl-auto | 2 +- 9 files changed, 64 insertions(+), 8 deletions(-) create mode 100644 docs/examples/openvswitch create mode 100644 src/lib/connections/openvswitch hooks/post-receive -- Network control utility based on systemd

1 0

[arch-projects] [mkinitcpio] [GIT] The official mkinitcpio repository branch master updated. v18-1-gea4c415
by dreisner＠archlinux.org 05 Aug '14

05 Aug '14

This is an automated email from the git hooks/post-receive script. It was generated because a ref change was pushed to the repository containing the project "The official mkinitcpio repository". The branch, master has been updated via ea4c4154205372154457c794513ae46b61ea4e4c (commit) from a35c698c16759c3c05c692fd10f140d186053177 (commit) Those revisions listed above that are new to this repository have not appeared on any other notification email; so we list those revisions in full, below. - Log ----------------------------------------------------------------- commit ea4c4154205372154457c794513ae46b61ea4e4c Author: Dave Reisner <dreisner(a)archlinux.org> Date: Mon Aug 4 08:31:37 2014 -0400 avoid compound conditional leading to spurious "errors" As seen: https://bbs.archlinux.org/viewtopic.php?id=185204 https://bbs.archlinux.org/viewtopic.php?id=185265 ----------------------------------------------------------------------- Summary of changes: functions | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) hooks/post-receive -- The official mkinitcpio repository

1 0