- Arch-projects - lists.archlinux.org

[arch-projects] [initscripts] new release
by Tom Gundersen 25 Sep '11

25 Sep '11

Hi guys, I intend to make a new initscripts release in the next week. Please let me know if I have missed some patches, or if you notice any problems with the current git. Changes: * We'll start complaining about an empty /usr * We pivot into a shutdown ramfs if it exists in /run/initramfs * /etc/profile.d/locale.sh is now a static file, rather than being generated on every boot (this means we need to use -f when updating) * rc.d is can now filter the output in a nice way * We load fragments from sysctl.d in addition to the sysctl.conf file * /etc/localtime is now a symlink Cheers, Tom

3 8

[arch-projects] [initscripts] [GIT] Arch Linux initscripts repository annotated tag 2011.09.1 created. 2011.09.1
by tomegun＠archlinux.org 24 Sep '11

24 Sep '11

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 "Arch Linux initscripts repository". The annotated tag, 2011.09.1 has been created at d41f011eb7aae9b8216df262dd5d391b8f1e10a1 (tag) tagging 71fe1590598381588502e58b60622b0b72d79454 (commit) replaces 2011.07.3 tagged by Tom Gundersen on Sun Sep 25 01:09:51 2011 +0200 - Log ----------------------------------------------------------------- * We'll start complaining about an empty /usr * We pivot into a shutdown ramfs if it exists in /run/initramfs * /etc/profile.d/locale.sh is now a static file, rather than being generated on every boot (this means we need to use -f when updating) * rc.d is can now filter the output in a nice way * We load fragments from sysctl.d in addition to the sysctl.conf file * /etc/localtime is now a symlink Dave Reisner (5): arch-tmpfiles: avoid subshell in creating empty files network: error out on missing or unknown interface rc.d: clarify error when [[ -x /etc/rc.d/$i ]] fails functions: clarify source of module loading timezone: create /etc/localtime as a symlink Gerardo Exequiel Pozzi (2): [initscript] Make locale.sh consistent with default rc.conf [initscripts] Remove too early /dev/null redirection Sebastien Luttringer (5): Let background daemons show errors on stderr Remove double case in hwclock daemon New smarter display of bad daemon names in rc.d rc.d: handle a set of options Load sysctl config files from sysctl.d Tom Gundersen (16): shutdown: pivot into a shutdown ramfs /usr: warn on separate usr Merge branch 'pivot-shutdown' /usr: move all executables from /sbin to /usr/sbin /usr: clarify warning a bit locale: make /etc/profile.d/locale.sh dynamic functions: fix typo in warning message locale.sh: added missing file to git boot.log: clean ANSI codes Clean up more ANSI codes from /var/log/boot (close FS#25682) sutdown: support both poweroff and reboot shutdown: rely on init being installed by the initrd timezone: use function in shutdown like in sysinit shutdown: move disabling of monitoring/swap kill: give proper error messages and increase timeout bootlog: clean some more ANSI codes ----------------------------------------------------------------------- hooks/post-receive -- Arch Linux initscripts repository

1 0

[arch-projects] [initscripts] [GIT] Arch Linux initscripts repository branch master updated. 2011.09.1
by tomegun＠archlinux.org 24 Sep '11

24 Sep '11

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 "Arch Linux initscripts repository". The branch, master has been updated via 71fe1590598381588502e58b60622b0b72d79454 (commit) via f90526071644423b67a9ba4b0421f94b890dee89 (commit) via 0093b7447019a9ca73c619c007fdd287329f9f52 (commit) via 5f7f87fa3259310454793521ac181028c31c1014 (commit) via 34c8a9f600d17010ed6b77ad685ed83a76322d3d (commit) via e0cd30462e62bad1420bd83d4408cc83ae202fda (commit) via ae92346c03bc60578bfb5bddef9dbbf1b4d71460 (commit) via 872444291d3ab4cbeecd4a24216531564986a01c (commit) via 5948612fe7191734ccc89570e808659ee2e6b3e3 (commit) via 227b71e67d19f73128862d20aac714537c8517df (commit) via 4e573b6b35d4468b4b3856a7784728f75ab2e73a (commit) via 1ce1675dd7c2edf89c58be70908df289c2b2665b (commit) via a74f312263b4154414dea1bc494818b1bcbb6b9b (commit) from 7cccf54fcb9fab459590b8b6fd5c51f2088e3b2e (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 71fe1590598381588502e58b60622b0b72d79454 Author: Tom Gundersen <teg(a)jklm.no> Date: Sat Sep 24 22:50:46 2011 +0200 bootlog: clean some more ANSI codes I was getting "^[71G", this fixed it. Signed-off-by: Tom Gundersen <teg(a)jklm.no> commit f90526071644423b67a9ba4b0421f94b890dee89 Author: Tom Gundersen <teg(a)jklm.no> Date: Sat Sep 24 22:23:35 2011 +0200 kill: give proper error messages and increase timeout With the new logic we will only wait for the timeout in case of problems, we therefore increase the timeout as this will make problems more obvious. There has been reports of problems with processes not being terminated, but hopefully this should help us reproduce them. Signed-off-by: Tom Gundersen <teg(a)jklm.no> commit 0093b7447019a9ca73c619c007fdd287329f9f52 Author: Tom Gundersen <teg(a)jklm.no> Date: Sat Sep 24 22:21:32 2011 +0200 shutdown: move disabling of monitoring/swap We do this before killing all processes. There is no reason to do it later, and this might give some processes a bit more time to terminate properly. Signed-off-by: Tom Gundersen <teg(a)jklm.no> commit 5f7f87fa3259310454793521ac181028c31c1014 Author: Tom Gundersen <teg(a)jklm.no> Date: Sat Sep 24 22:19:29 2011 +0200 timezone: use function in shutdown like in sysinit This uses the function introduced in 9b77b7d7b535c3c52c99a95374e1f800466c6414. Signed-off-by: Tom Gundersen <teg(a)jklm.no> commit 34c8a9f600d17010ed6b77ad685ed83a76322d3d Author: Gerardo Exequiel Pozzi <vmlinuz386(a)yahoo.com.ar> Date: Mon Sep 12 12:32:23 2011 -0300 [initscripts] Remove too early /dev/null redirection Since recent change in udev package [#1] (install: remove post-install) In case of booting without initramfs and there is no /dev/null: * if / is ro, devtmpfs mount will fail, in consecuence /dev will be mounted as tmpfs. * if / is rw, devtmpfs mount will success, but a regular file /dev/null in / will be created. [#1] http://projects.archlinux.org/svntogit/packages.git/commit/trunk?h=packages… Signed-off-by: Gerardo Exequiel Pozzi <vmlinuz386(a)yahoo.com.ar> commit e0cd30462e62bad1420bd83d4408cc83ae202fda Author: Gerardo Exequiel Pozzi <vmlinuz386(a)yahoo.com.ar> Date: Fri Aug 19 11:29:30 2011 -0300 [initscript] Make locale.sh consistent with default rc.conf rc.conf defaults LOCALE to en_US.UTF-8 Signed-off-by: Gerardo Exequiel Pozzi <vmlinuz386(a)yahoo.com.ar> commit ae92346c03bc60578bfb5bddef9dbbf1b4d71460 Author: Dave Reisner <dreisner(a)archlinux.org> Date: Sun Sep 4 15:41:53 2011 -0400 timezone: create /etc/localtime as a symlink Abstract out timezone setting into its own function, passing Country/Area as a parameter. Do some menial checking to make sure: - the zonefile actually exists, erroring when it doesn't - /etc/localtime is a symlink to the correct zoneinfo file, relinking when it isn't. Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> commit 872444291d3ab4cbeecd4a24216531564986a01c Author: Dave Reisner <dreisner(a)archlinux.org> Date: Sun Sep 4 10:12:18 2011 -0400 functions: clarify source of module loading This isn't _all_ your modules. Just the extras specified from rc.conf. Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> commit 5948612fe7191734ccc89570e808659ee2e6b3e3 Author: Sebastien Luttringer <seblu(a)seblu.net> Date: Wed Aug 31 11:32:51 2011 +0200 Load sysctl config files from sysctl.d This patch implement loading of sysctl config files as described in http://0pointer.de/public/systemd-man/sysctl.d.html This is a very conveniant way of configuring sysctl option for sysadmin which can drop sysctl config files inside this directory to enable some feature. Dropping a file like disableipv6.conf inside this directory will disable ipv6 $ cat disableipv6.conf net.ipv6.conf.all.disable_ipv6 = 1 There is atm no package which use this functionnality Signed-off-by: Sebastien Luttringer <seblu(a)seblu.net> commit 227b71e67d19f73128862d20aac714537c8517df Author: Sebastien Luttringer <seblu(a)seblu.net> Date: Mon Aug 8 21:38:56 2011 +0200 rc.d: handle a set of options rc.d can now take --started, --stopped, --auto, --noauto as option which help user to filter list of daemon for all actions As a corollary list command can now take a list of dameon to display All kind of arguments can be mixed to obtain the proper output. zsh and bash completion are updated Note: Output of help command exit 0 and is no more printed on stderr Signed-off-by: Sebastien Luttringer <seblu(a)seblu.net> commit 4e573b6b35d4468b4b3856a7784728f75ab2e73a Author: Sebastien Luttringer <seblu(a)seblu.net> Date: Mon Aug 8 20:50:51 2011 +0200 New smarter display of bad daemon names in rc.d Signed-off-by: Sebastien Luttringer <seblu(a)seblu.net> commit 1ce1675dd7c2edf89c58be70908df289c2b2665b Author: Sebastien Luttringer <seblu(a)seblu.net> Date: Mon Aug 8 20:13:20 2011 +0200 Remove double case in hwclock daemon Signed-off-by: Sebastien Luttringer <seblu(a)seblu.net> commit a74f312263b4154414dea1bc494818b1bcbb6b9b Author: Sebastien Luttringer <seblu(a)seblu.net> Date: Wed Jul 27 23:07:15 2011 +0200 Let background daemons show errors on stderr Daemon running in background should let error output to be printed. Standart output is still hided to have a correct printing in default cases. This will help to detect error in daemon runned in background. This will also remove have_daemon call which is already called in start_daemon. Signed-off-by: Sebastien Luttringer <seblu(a)seblu.net> ----------------------------------------------------------------------- Summary of changes: Makefile | 4 +- arch-sysctl | 29 +++++++++++++ bash-completion | 20 ++++---- functions | 75 +++++++++++++++++++++++---------- hwclock | 6 +-- locale.sh | 2 +- rc.d | 124 +++++++++++++++++++++++++++++++++++++++--------------- rc.multi | 4 +- rc.shutdown | 17 +++---- rc.sysinit | 6 +-- zsh-completion | 8 +--- 11 files changed, 199 insertions(+), 96 deletions(-) create mode 100755 arch-sysctl hooks/post-receive -- Arch Linux initscripts repository

1 0

[arch-projects] [initscripts][PATCH 1/2] rc.sysinit: support /etc/localtime as a symlink
by Dave Reisner 24 Sep '11

24 Sep '11

Signed-off-by: Dave Reisner <dreisner(a)archlinux.org> --- rc.sysinit | 10 ++++++++-- 1 files changed, 8 insertions(+), 2 deletions(-) diff --git a/rc.sysinit b/rc.sysinit index 9f215c4..17d3cc6 100755 --- a/rc.sysinit +++ b/rc.sysinit @@ -221,9 +221,15 @@ run_hook sysinit_postmount status "Activating Swap" swapon -a -[[ $TIMEZONE ]] && - status "Configuring Time Zone" \ +if [[ $TIMEZONE && -f /usr/share/zoneinfo/$TIMEZONE ]]; then + stat_busy "Configuring Time Zone" + if [[ ! -L /etc/localtime ]]; then cp --remove-destination "/usr/share/zoneinfo/$TIMEZONE" /etc/localtime + elif [[ ! /etc/localtime -ef /usr/share/zoneinfo/$TIMEZONE ]]; then + ln -sf "/usr/share/zoneinfo/$TIMEZONE" /etc/localtime + fi + (( $? == 0 )) && stat_done || stat_fail +fi RANDOM_SEED=/var/lib/misc/random-seed [[ -f $RANDOM_SEED ]] && -- 1.7.6.3

1 1

[arch-projects] [PATCH] [initscripts] Remove too early /dev/null redirection
by Gerardo Exequiel Pozzi 22 Sep '11

22 Sep '11

Since recent change in udev package [#1] (install: remove post-install) In case of booting without initramfs and there is no /dev/null: * if / is ro, devtmpfs mount will fail, in consecuence /dev will be mounted as tmpfs. * if / is rw, devtmpfs mount will success, but a regular file /dev/null in / will be created. [#1] http://projects.archlinux.org/svntogit/packages.git/commit/trunk?h=packages… Signed-off-by: Gerardo Exequiel Pozzi <vmlinuz386(a)yahoo.com.ar> --- rc.sysinit | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/rc.sysinit b/rc.sysinit index 9f215c4..01fe49f 100755 --- a/rc.sysinit +++ b/rc.sysinit @@ -15,7 +15,7 @@ printsep mountpoint -q /proc || mount -n -t proc proc /proc -o nosuid,noexec,nodev mountpoint -q /sys || mount -n -t sysfs sys /sys -o nosuid,noexec,nodev mountpoint -q /run || mount -n -t tmpfs run /run -o mode=0755,size=10M,nosuid,nodev -mountpoint -q /dev || mount -n -t devtmpfs udev /dev -o mode=0755,size=10M,nosuid &>/dev/null \ +mountpoint -q /dev || mount -n -t devtmpfs udev /dev -o mode=0755,size=10M,nosuid \ || mount -n -t tmpfs udev /dev -o mode=0755,size=10M,nosuid mkdir -p -m 1777 /run/lock mkdir -p /dev/{pts,shm} -- 1.7.6.1

2 2

[arch-projects] [devtools] RFC, Pull request
by Jan Steffens 20 Sep '11

20 Sep '11

On Sun, Sep 18, 2011 at 12:48 PM, Lukas Fleischer <archlinux(a)cryptocrack.de> wrote: > On Sat, Sep 17, 2011 at 04:43:33PM +0200, Jan Steffens wrote: >> >> Started rewriting: https://github.com/heftig/devtools/commits/master >> >> Comments welcome. > > There's a couple of more things that could be bashified (particularly > basename(1) and all the grep(1) and sed(1) invocations used in command > substitutions, since most of them can be replaced with bash parameter > expansion magic [1], [2], [3]). > > Apart from that, looks okay at a glance. > > [1] http://projects.archlinux.org/devtools.git/commit/?id=7952d6fb > [2] http://projects.archlinux.org/devtools.git/commit/?id=3c175e98 > [3] http://projects.archlinux.org/devtools.git/commit/?id=5688152f > Updated the patchset. All uses of basename and some of the others have been removed.

1 0

[arch-projects] [PATCH] Add an option to explicitly run namcap
by Jan Steffens 18 Sep '11

18 Sep '11

Installs namcap if needed, *after* building the package, contrary to the former way of having to have namcap installed, e.g. via makedepends. Signed-off-by: Jan Steffens <jan.steffens(a)gmail.com> --- makechrootpkg | 14 ++++++++++++-- 1 files changed, 12 insertions(+), 2 deletions(-) diff --git a/makechrootpkg b/makechrootpkg index a60c9fe..974628c 100755 --- a/makechrootpkg +++ b/makechrootpkg @@ -18,6 +18,7 @@ update_first='0' clean_first='0' install_pkg='' add_to_db=0 +run_namcap=0 chrootdir='' @@ -54,11 +55,12 @@ usage() { echo '-I <pkg> Install a package into the working copy of the chroot' echo '-l <copy> The directory to use as the working copy of the chroot' echo ' Useful for maintaining multiple copies.' + echo '-n Run namcap on the package' echo " Default: $default_copy" exit 1 } -while getopts 'hcudr:I:l:' arg; do +while getopts 'hcudr:I:l:n' arg; do case "${arg}" in h) usage ;; c) clean_first=1 ;; @@ -67,6 +69,7 @@ while getopts 'hcudr:I:l:' arg; do r) chrootdir="$OPTARG" ;; I) install_pkg="$OPTARG" ;; l) copy="$OPTARG" ;; + n) run_namcap=1 ;; *) MAKEPKG_ARGS="$MAKEPKG_ARGS -$arg $OPTARG" ;; esac done @@ -175,6 +178,8 @@ if [ "$REPACK" != "1" ]; then rm -rf "$copydir/build/"* fi +[ "$run_namcap" -eq "1" ] && touch "$copydir/build/RUN_NAMCAP" + # Read .makepkg.conf even if called via sudo if [ -n "${SUDO_USER}" ]; then makepkg_conf="/$(eval echo ~${SUDO_USER})/.makepkg.conf" @@ -257,7 +262,10 @@ cd /build export HOME=/build sudo -u nobody makepkg $MAKEPKG_ARGS || touch BUILD_FAILED [ -f BUILD_FAILED ] && exit 1 -which namcap &>/dev/null && namcap /build/PKGBUILD /pkgdest/*.pkg.tar.* > /build/namcap.log +if [ -f RUN_NAMCAP ]; then + pacman -S --needed --noconfirm namcap + namcap /build/PKGBUILD /pkgdest/*.pkg.tar.* > /build/namcap.log +fi exit 0 EOF ) > "$copydir/chrootbuild" @@ -303,3 +311,5 @@ if [ -e "${copydir}/build/BUILD_FAILED" ]; then rm "${copydir}/build/BUILD_FAILED" exit 1 fi + +# vim:set noet ts=8 sw=8: -- 1.7.6.1

2 4

[arch-projects] [MKINITCPIO][PATCH 1/2] Fix save build directory with presets
by Sebastien Luttringer 16 Sep '11

16 Sep '11

Current behaviour of -s option of mkinitcpio is bad when presets are used. mkinitcpio saves main process build directory which is always empty when preset option is used. mkinitcpio doesn't save preset sub processes build directories which is the expected behaviour. This patch fix the two previous points by: - inherit -s parameter inside preset sub processes. - force main process (in case of preset) to clean build directory Signed-off-by: Sebastien Luttringer <seblu(a)seblu.net> --- mkinitcpio | 6 ++++-- 1 files changed, 4 insertions(+), 2 deletions(-) diff --git a/mkinitcpio b/mkinitcpio index 49dabbd..53254d7 100755 --- a/mkinitcpio +++ b/mkinitcpio @@ -64,7 +64,7 @@ cleanup () { if [[ $TMPDIR ]]; then if (( $SAVELIST )); then - msg "build directory saved in %s" "$TMPDIR" + msg "Build directory saved in %s" "$TMPDIR" else rm -rf "$TMPDIR" fi @@ -198,6 +198,7 @@ if [[ $PRESET ]]; then # Use -b, -m and -v options specified earlier declare -a preset_mkopts preset_cmd [[ $BASEDIR ]] && preset_mkopts+=(-b "$BASEDIR") + (( SAVELIST )) && preset_mkopts+=(-s) (( QUIET )) || preset_mkopts+=(-v) # Build all images . "$PRESET" @@ -237,7 +238,8 @@ if [[ $PRESET ]]; then msg2 "${preset_cmd[*]}" "$0" "${preset_cmd[@]}" done - cleanup 0 + # Main process never need to keep builded directory + SAVELIST=0 cleanup 0 else die "Preset not found: \`%s'" "$PRESET" fi -- Sebastien "Seblu" Luttringer

5 9

[arch-projects] [MKINITCPIO][PATCH 1/2] Fix save build directory with presets
by Sebastien Luttringer 16 Sep '11

16 Sep '11

Current behaviour of -s option of mkinitcpio is bad when presets are used. mkinitcpio saves main process build directory which is always empty when preset option is used. mkinitcpio doesn't save preset sub processes build directories which is the expected behaviour. This patch fix the two previous points by: - inherit -s parameter inside preset sub processes. - force main process (in case of preset) to clean build directory Signed-off-by: Sebastien Luttringer <seblu(a)seblu.net> --- mkinitcpio | 6 ++++-- 1 files changed, 4 insertions(+), 2 deletions(-) diff --git a/mkinitcpio b/mkinitcpio index 49dabbd..f48a26d 100755 --- a/mkinitcpio +++ b/mkinitcpio @@ -64,7 +64,7 @@ cleanup () { if [[ $TMPDIR ]]; then if (( $SAVELIST )); then - msg "build directory saved in %s" "$TMPDIR" + msg "Build directory saved in %s" "$TMPDIR" else rm -rf "$TMPDIR" fi @@ -198,6 +198,7 @@ if [[ $PRESET ]]; then # Use -b, -m and -v options specified earlier declare -a preset_mkopts preset_cmd [[ $BASEDIR ]] && preset_mkopts+=(-b "$BASEDIR") + (( SAVELIST )) && preset_mkopts+=(-s) (( QUIET )) || preset_mkopts+=(-v) # Build all images . "$PRESET" @@ -237,7 +238,8 @@ if [[ $PRESET ]]; then msg2 "${preset_cmd[*]}" "$0" "${preset_cmd[@]}" done - cleanup 0 + # Main process never needs to keep build directory + SAVELIST=0 cleanup 0 else die "Preset not found: \`%s'" "$PRESET" fi -- Sebastien "Seblu" Luttringer

1 1

[arch-projects] [PATCH namcap 1/2] Remove PkgdirRule
by Dan McGee 16 Sep '11

16 Sep '11

It is now unused, and caused confusion regarding the sandbox directory stuff. Ideally it should never need to be used again, as all rules should operate directly on the tarball. Signed-off-by: Dan McGee <dan(a)archlinux.org> --- Namcap/ruleclass.py | 5 ----- README | 1 - namcap.py | 14 -------------- 3 files changed, 0 insertions(+), 20 deletions(-) diff --git a/Namcap/ruleclass.py b/Namcap/ruleclass.py index f34c657..15250ad 100644 --- a/Namcap/ruleclass.py +++ b/Namcap/ruleclass.py @@ -44,9 +44,4 @@ class TarballRule(AbstractRule): "The parent class of rules that process tarballs" pass -class PkgdirRule(AbstractRule): - """The parent class of rules that process pkgdirs - (extracted tarballs)""" - pass - # vim: set ts=4 sw=4 noet: diff --git a/README b/README index b52ea42..89f6fb5 100644 --- a/README +++ b/README @@ -84,7 +84,6 @@ defined in Namcap.rules.all_rules : * PkgInfoRule classes process any pkginfo object * PkgbuildRule classes process only PKGBUILDs * TarballRule classes process binary packages - * PkgdirRule classes (unused) process uncompressed binary packages Put the new rule in a module and make sure it is imported in Namcap/rules/__init__.py diff --git a/namcap.py b/namcap.py index dcebc55..cca5dd6 100755 --- a/namcap.py +++ b/namcap.py @@ -32,8 +32,6 @@ import types import Namcap.depends import Namcap.tags -sandbox_directory = '/tmp/namcap.' + str(os.getpid()) - # Functions def get_modules(): """Return all possible modules (rules)""" @@ -97,14 +95,6 @@ def process_realpackage(package, modules): if isinstance(rule, Namcap.ruleclass.PkgInfoRule): rule.analyze(pkginfo, None) - elif isinstance(rule, Namcap.ruleclass.PkgdirRule): - # If it's not extracted, extract it and then analyze the package - if not extracted: - os.mkdir(sandbox_directory) - for j in pkgtar.getmembers(): - pkgtar.extract(j, sandbox_directory) - extracted = 1 - rule = pkg.analyze(pkginfo, sandbox_directory) elif isinstance(rule, Namcap.ruleclass.PkgbuildRule): pass elif isinstance(rule, Namcap.ruleclass.TarballRule): @@ -126,10 +116,6 @@ def process_realpackage(package, modules): if info_reporting: show_messages(pkginfo["name"], 'I', infos) - # Clean up if we extracted anything - if extracted: - shutil.rmtree(sandbox_directory) - def process_pkginfo(pkginfo, modules): """Runs namcap checks of a single, non-split PacmanPackage object""" for i in modules: -- 1.7.6.1

2 2