- Arch-projects - lists.archlinux.org

[arch-projects] [devtools] [GIT] The official devtools repository branch master updated. 20140510-8-g12a1300
by dreisner＠archlinux.org 22 Sep '14

22 Sep '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 devtools repository". The branch, master has been updated via 12a1300694272928e70408f4ce2fe85bacd19eb4 (commit) via 625e6bd41206c5018b1a9e11c5cc2b3aef55a66e (commit) via 6dce935b99389713cf5b8a5b698b1abd6670c916 (commit) via 98841eb694fa3cd575fde60e562586cf33415c5b (commit) via be00fcd47b9fa6ebdab7173372ab219514e0b88d (commit) via 651f8f834ff4076ece781d276cdc60137c961081 (commit) via d83805bc54b04d8c63982784e1a220a2bd87e3b0 (commit) via af6c0a0f6a7644f583f2d83ba7a61e82b77bdc37 (commit) from 905198295dcb4fffcd16edff3fd7052a4cfc5788 (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 12a1300694272928e70408f4ce2fe85bacd19eb4 Author: Luke Shumaker <LukeShu(a)sbcglobal.net> Date: Sat May 10 22:22:06 2014 -0400 makechrootpkg: Use the btrfs mountpoint/subvolume check consistently. Commit 59e348fc3c5dd086331d884a6dd76fb43a92b7eb added a btrfs subvolume check, but only used it in create_chroot(); it missed clean_temporary(). commit 625e6bd41206c5018b1a9e11c5cc2b3aef55a66e Author: Luke Shumaker <LukeShu(a)sbcglobal.net> Date: Sat May 10 22:22:05 2014 -0400 mkarchroot: Correct "Usage:" text. The "app" hasn't been an option since arch-nspawn was created. commit 6dce935b99389713cf5b8a5b698b1abd6670c916 Author: Luke Shumaker <LukeShu(a)sbcglobal.net> Date: Sat May 10 22:22:04 2014 -0400 archbuild: Fix a variable name mistake. ($copydir -> $copy) It tried to lock `$copydir.lock`, which was the ONLY mention of $copydir in the entire file. Surely it meant `$copy.lock`; the line was probably originally copy/pasted from makechrootpkg or similar, where $copydir is used. commit 98841eb694fa3cd575fde60e562586cf33415c5b Author: Dave Reisner <dreisner(a)archlinux.org> Date: Sat May 10 09:31:26 2014 -0400 common: remove unreachable statement We run from a non-interactive shell, so the exec which is inevitably called will replace the current process and 'die' will never run under any circumstances. This also fixes a bug with the su fallback which would cause multiple arguments to be concatenated without any whitespace between them. commit be00fcd47b9fa6ebdab7173372ab219514e0b88d Author: Dave Reisner <dreisner(a)archlinux.org> Date: Sun May 4 15:43:50 2014 -0400 makechrootpkg: use a simpler/safer expression with eval commit 651f8f834ff4076ece781d276cdc60137c961081 Author: Luke Shumaker <LukeShu(a)sbcglobal.net> Date: Sat May 10 22:22:03 2014 -0400 arch-nspawn: Fix a grammar mistake (a/an) in an error message. commit d83805bc54b04d8c63982784e1a220a2bd87e3b0 Author: Dave Reisner <dreisner(a)archlinux.org> Date: Sun Aug 31 09:43:00 2014 -0400 nspawn: skip machine registration No need to do this since we don't run an init system in the build root. commit af6c0a0f6a7644f583f2d83ba7a61e82b77bdc37 Author: Dave Reisner <dreisner(a)archlinux.org> Date: Sun Jun 1 23:54:20 2014 -0400 common.sh: propagate error through trap_exit Fixes a breakage introduced in 6db31cc16a80442 which leads to errors being masked from makechrootpkg. ----------------------------------------------------------------------- Summary of changes: arch-nspawn.in | 7 ++----- archbuild.in | 2 +- lib/common.sh | 6 +++--- makechrootpkg.in | 4 ++-- mkarchroot.in | 2 +- 5 files changed, 9 insertions(+), 12 deletions(-) hooks/post-receive -- The official devtools repository

1 0

[arch-projects] [PATCH V2] [devtools] Fix soname checking
by Doug Newgard 21 Sep '14

21 Sep '14

This makes soname checking much more robust and is explicit when a soname change is found. Thanks to Bluewind for pointing me to find-libprovides. --- checkpkg.in | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/checkpkg.in b/checkpkg.in index ccbbecd..3630567 100644 --- a/checkpkg.in +++ b/checkpkg.in @@ -63,12 +63,11 @@ for _pkgname in "${pkgname[@]}"; do sdiff -s "$TEMPDIR/filelist-$_pkgname-old" "$TEMPDIR/filelist-$_pkgname" - if diff "$TEMPDIR/filelist-$_pkgname"{-old,} | grep '\.so' &>/dev/null; then - mkdir -p "$TEMPDIR/pkg" - bsdtar -x -C "$TEMPDIR" -f "$pkgfile" #> /dev/null - comm -13 <(sort "$TEMPDIR/filelist-$_pkgname-old") <(sort "$TEMPDIR/filelist-$_pkgname") | grep .so$ | while read i; do - echo "${i}: " "$(objdump -p "$TEMPDIR/$i" | grep SONAME)" - done + find-libprovides "$TEMPDIR/$oldpkg" | sort > "$TEMPDIR/libraries-$_pkgname-old" + find-libprovides "$pkgfile" | sort > "$TEMPDIR/libraries-$_pkgname" + if ! cmp "$TEMPDIR/libraries-$_pkgname-old" "$TEMPDIR/libraries-$_pkgname" &> /dev/null; then + msg "Sonames differ in $_pkgname!" + sdiff -s "$TEMPDIR/libraries-$_pkgname-old" "$TEMPDIR/libraries-$_pkgname" else msg "No soname differences for $_pkgname." fi -- 2.1.0

1 2

[arch-projects] [PATCH] [devtools] Fix soname checking
by Doug Newgard 21 Sep '14

21 Sep '14

This makes soname checking much more robust and is explicit when a soname change is found. --- checkpkg.in | 18 ++++++++++++++---- 1 file changed, 14 insertions(+), 4 deletions(-) diff --git a/checkpkg.in b/checkpkg.in index ccbbecd..6078e59 100644 --- a/checkpkg.in +++ b/checkpkg.in @@ -64,10 +64,20 @@ for _pkgname in "${pkgname[@]}"; do sdiff -s "$TEMPDIR/filelist-$_pkgname-old" "$TEMPDIR/filelist-$_pkgname" if diff "$TEMPDIR/filelist-$_pkgname"{-old,} | grep '\.so' &>/dev/null; then - mkdir -p "$TEMPDIR/pkg" - bsdtar -x -C "$TEMPDIR" -f "$pkgfile" #> /dev/null - comm -13 <(sort "$TEMPDIR/filelist-$_pkgname-old") <(sort "$TEMPDIR/filelist-$_pkgname") | grep .so$ | while read i; do - echo "${i}: " "$(objdump -p "$TEMPDIR/$i" | grep SONAME)" + mkdir -p "$TEMPDIR/pkg" "$TEMPDIR/pkg-old" + bsdtar -x -C "$TEMPDIR/pkg" -f "$pkgfile" #> /dev/null + bsdtar -x -C "$TEMPDIR/pkg-old" -f "$oldpkg" #> /dev/null + comm -2 "$TEMPDIR/filelist-$_pkgname-old" "$TEMPDIR/filelist-$_pkgname" | grep '\.so$' | while read i; do + oldsoname="$(objdump -p "$TEMPDIR/pkg-old/$i" | awk '/SONAME/ {print $2}')" + if [[ -e "$TEMPDIR/pkg/$i" ]]; then + newsoname="$(objdump -p "$TEMPDIR/pkg/$i" | awk '/SONAME/ {print $2}')" + else + newsoname="MISSING" + fi + if [[ $oldsoname != $newsoname ]]; then + msg "SONAMES differ!" + printf "$i\n\tOld: %s\n\tNew: %s\n" "$oldsoname" "$newsoname" + fi done else msg "No soname differences for $_pkgname." -- 2.1.0

1 0

[arch-projects] [dbscripts][PATCH] disallow packages without a valid PACKAGER
by Dave Reisner 10 Sep '14

10 Sep '14

Shame. --- current list of packages without a PACKAGER: https://paste.xinu.at/eL2/ db-functions | 7 +++++++ db-update | 3 +++ 2 files changed, 10 insertions(+) diff --git a/db-functions b/db-functions index bb49894..7e1b3aa 100644 --- a/db-functions +++ b/db-functions @@ -297,6 +297,13 @@ getpkgarch() { echo "$_ver" } +check_packager() { + local _packager + + _packager=$(_grep_pkginfo "$1" "packager") + [[ $_packager && $_packager != 'Unknown Packager' ]] +} + getpkgfile() { if [[ ${#} -ne 1 ]]; then error 'No canonical package found!' diff --git a/db-update b/db-update index 576fe2b..cdc35a5 100755 --- a/db-update +++ b/db-update @@ -51,6 +51,9 @@ for repo in ${repos[@]}; do if ! check_pkgrepos "${pkg}"; then die "Package ${repo}/${pkg##*/} already exists in another repository" fi + if ! check_packager "${pkg}"; then + die "Package ${repo}/${pkg##*/} does not have a valid packager" + fi done if ! check_splitpkgs ${repo} ${pkgs[@]}; then die "Missing split packages for ${repo}" -- 2.1.0

1 0

[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