- Arch-projects - lists.archlinux.org

[arch-projects] [PATCH] rc.d: declare ENV as an array
by Dave Reisner 27 May '11

27 May '11

Signed-off-by: Dave Reisner <d(a)falconindy.com> --- Sorry to sneak this in -- I should have noticed it earlier. This is simply a continuation of my crusade against eval, which would happily go to town with word splitting. As this conflicts with Eric's patch, one of us will need to refactor slightly. dave rc.d | 12 ++++++------ 1 files changed, 6 insertions(+), 6 deletions(-) diff --git a/rc.d b/rc.d index 55871b1..af29574 100755 --- a/rc.d +++ b/rc.d @@ -46,13 +46,13 @@ case $1 in shift # set same environment variables as init runlevel=$(/sbin/runlevel) - ENV="PATH='/bin:/usr/bin:/sbin:/usr/sbin'" - ENV+=" PREVLEVEL='${runlevel:0:1}'" - ENV+=" RUNLEVEL='${runlevel:2:1}'" - ENV+=" CONSOLE='${CONSOLE:-/dev/console}'" - ENV+=" TERM='${TERM}'" + ENV=("PATH=/bin:/usr/bin:/sbin:/usr/sbin" + "PREVLEVEL=${runlevel%% *}" + "RUNLEVEL=${runlevel##* }" + "CONSOLE=${CONSOLE:-/dev/console}" + "TERM=$TERM") for i; do - [[ -x "/etc/rc.d/$i" ]] && cd / && eval /usr/bin/env -i $ENV "/etc/rc.d/$i" "$action" + [[ -x "/etc/rc.d/$i" ]] && cd / && /usr/bin/env -i "${ENV[@]}" "/etc/rc.d/$i" "$action" (( ret += !! $? )) # clamp exit value to 0/1 done esac -- 1.7.5.2

2 1

[arch-projects] [initscripts][PATCH] rc.d: Add error message when daemon script is not found
by Eric Bélanger 27 May '11

27 May '11

Signed-off-by: Eric Bélanger <snowmaniscool(a)gmail.com> --- rc.d | 7 ++++++- 1 files changed, 6 insertions(+), 1 deletions(-) diff --git a/rc.d b/rc.d index 55871b1..97f266a 100755 --- a/rc.d +++ b/rc.d @@ -52,7 +52,12 @@ case $1 in ENV+=" CONSOLE='${CONSOLE:-/dev/console}'" ENV+=" TERM='${TERM}'" for i; do - [[ -x "/etc/rc.d/$i" ]] && cd / && eval /usr/bin/env -i $ENV "/etc/rc.d/$i" "$action" + if [[ -x "/etc/rc.d/$i" ]]; then + cd / + eval /usr/bin/env -i $ENV "/etc/rc.d/$i" "$action" + else + printf "${C_OTHER}:: ${C_FAIL}Error: ${C_DONE}Daemon script $i does not exist.\n" + fi (( ret += !! $? )) # clamp exit value to 0/1 done esac -- 1.7.5.2

2 1

[arch-projects] [initscripts][mkinitcpio][module-init-tools][udev] dropping load-modules.sh
by Tom Gundersen 24 May '11

24 May '11

Hi guys, I have been working on the remaining bits before we can drop our load-modules.sh from udev. I asked around on #udev to see what they thought of moving some of our functionality upstream. Kay Sievers rightly pointed out that the functionality belongs in module-init-tools, not in udev. I made a few patches against module-init-tools which simplifies what we have to do in initscripts/mkinitcpio, and I'd like to ask your opinion about this before submitting upstream. Kernel commandline blacklisting: modprobe will parse /proc/kcmdline in addition to the configuration files to blacklist modules. The syntax is "modprobe.blacklist=mod1,mod2,mod3". rc.conf blacklist compatibility: This is only relevant for initscripts. modprobe will read /run/modprobe.d in addition to /etc/modprobe.d for configuration files, so at boot we can generate a blacklist file directly in /run based on rc.conf. I'm still working on some improvements to this patch that are needed before it can be upstreamed (if two files with the same name exist in different directories one needs to be ignored), but the functionality we care about is already there. Any comments? Does this sound like enough to remove the blacklisting logic provided by modules-load.sh? For an idea of what initscripts will do at boot once this is in modules-init-tools, have a look here: <https://github.com/teg/initscripts-arch/blob/ccac6d93278fe4ce2e8cbf9923c089…>. Cheers, Tom

1 0

[arch-projects] [PATCH 0/2] iproute2 migration
by Dave Reisner 24 May '11

24 May '11

I'm resubmitting my work of migrating to iproute2 from net-tools at Tom's request. It's pretty much the same, but with the feedback from Dan, I've added some extremely simple deprecation logic to yell at the user when he or she is still using the net-tools variables. The new logic uses some fairly simple (and generic) declarations: interface, address, netmask, and gateway. If interface is undefined, we fire off the deprecation warning, as this is the _only_ variable required to establish a connection via iproute2. If address is undefined, we skip reading netmask and gateway, and assume usage of dhcp. If address _is_ defined, we assert the need for gateway and netmask. I've opted to make this warning fairly common, firing it both on bringing up, and bringing down interfaces, as I think this needs to be made very clear, given the severity of the change. Along with this change, we would need to do a little bit of juggling with coreutils and yp-tools in order to make net-tools truly optional: - coreutils is currently built without hostname. enable this, removing it from net-tools. - yp-tools is currently in extra. we would need to bring this into core in order to support setting a domainname, and remove the symlinks to hostname from the net-tools package. As always, comments, criticisms, and tomatoes are welcome. dave

5 14

[arch-projects] [initscripts] [GIT] Arch Linux initscripts repository branch master updated. 2011.05.2-17-gdfdf8b6
by tomegun＠archlinux.org 24 May '11

24 May '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 dfdf8b683633d89a9c7c7560397c57a8a1594147 (commit) via 18adeacb36967463330060237725ffbdfd47e2a7 (commit) via a3314c769768d8c40b46767d26f9783d4f359358 (commit) via 395803ff628e41940bbaaf5f1ffbcf591b007102 (commit) via d543c7aa098c214fa5e55a497353dda47b95d3ad (commit) via d4d9f1f97f0fbf798eafbb5c1c1976c342605eb5 (commit) via 1ba11e7f78af5742dbc211d9202db4549558ca44 (commit) via d1be877eb0ad3be634cee08aada8f8d1265023a4 (commit) via 9888b262c2deb32064b30eb637ece9e424356777 (commit) from 68c61ed4ce2eddf29c4f047133406d72162fd396 (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 dfdf8b683633d89a9c7c7560397c57a8a1594147 Merge: a3314c7 18adeac Author: Tom Gundersen <teg(a)jklm.no> Date: Mon May 23 01:31:48 2011 +0200 Merge remote-tracking branch 'seblu/master' commit 18adeacb36967463330060237725ffbdfd47e2a7 Author: Sebastien Luttringer <seblu(a)seblu.net> Date: Mon May 23 01:13:57 2011 +0200 rc.d zsh completion detect running daemons This patch adds functionalities of bash completions about running demons Signed-off-by: Sebastien Luttringer <seblu(a)seblu.net> commit a3314c769768d8c40b46767d26f9783d4f359358 Author: Dave Reisner <d(a)falconindy.com> Date: Thu Apr 28 22:42:11 2011 -0400 network: deprecate net-tools, add iproute2 support Provide large warnings when net-tools functionality is used. Add documentation in rc.conf for the new iproute2 based config. Signed-off-by: Tom Gundersen <teg(a)jklm.no> commit 395803ff628e41940bbaaf5f1ffbcf591b007102 Author: Clément Démoulins <clement(a)archivel.fr> Date: Wed May 18 22:16:02 2011 +0200 Add a zsh completion script for the command rc.d. seblu: Fix typo, add vim modeline Signed-off-by: Sebastien Luttringer <seblu(a)seblu.net> commit d543c7aa098c214fa5e55a497353dda47b95d3ad Author: Sebastien Luttringer <seblu(a)seblu.net> Date: Sat May 21 21:12:28 2011 +0200 rc.d bash completion detect running daemon This patch adds sexy features from Auguste Pop bash completion patch. Mainly, detect if a daemon is running and complete accordingly Signed-off-by: Sebastien Luttringer <seblu(a)seblu.net> commit d4d9f1f97f0fbf798eafbb5c1c1976c342605eb5 Author: Sebastien Luttringer <seblu(a)seblu.net> Date: Thu May 5 02:50:11 2011 +0200 Add rc.d bash completion Signed-off-by: Sebastien Luttringer <seblu(a)seblu.net> commit 1ba11e7f78af5742dbc211d9202db4549558ca44 Author: Sebastien Luttringer <seblu(a)seblu.net> Date: Wed May 4 23:35:25 2011 +0200 Improve coloring on light and transparent term This patch improve coloring in light and transparent term. It uses tput if available or default color code. It remplace \033 by \e. It's a kind of rollback from commit 65f410, because initscripts are now in bash, and argument "\e isn't recongized in all shells" is not an issue anymore. This is not done in locale.sh script which can be sourced by others shell, so we try to be the more compliant as possible Signed-off-by: Sebastien Luttringer <seblu(a)seblu.net> commit d1be877eb0ad3be634cee08aada8f8d1265023a4 Author: Sebastien Luttringer <seblu(a)seblu.net> Date: Thu May 5 01:42:07 2011 +0200 Add TERM var to rc.d cleaned env We need to do this to allow a correct detection of terminal colors in functions sourced by rc.d scripts Signed-off-by: Sebastien Luttringer <seblu(a)seblu.net> commit 9888b262c2deb32064b30eb637ece9e424356777 Author: Sebastien Luttringer <seblu(a)seblu.net> Date: Thu May 5 01:26:44 2011 +0200 Add syntax to vim modeline in rc.d script With new name of script, vim doesn't reconize correctly file format. We need to set it explicitly. Signed-off-by: Sebastien Luttringer <seblu(a)seblu.net> ----------------------------------------------------------------------- Summary of changes: Makefile | 4 +- PKGBUILD | 7 ++- bash-completion | 22 ++++++++++ functions | 47 +++++++++++++--------- network | 118 ++++++++++++++++++++++++++++++++++++++++-------------- rc.conf | 40 +++++-------------- rc.d | 3 +- rc.sysinit | 6 +- zsh-completion | 37 +++++++++++++++++ 9 files changed, 196 insertions(+), 88 deletions(-) create mode 100644 bash-completion create mode 100644 zsh-completion hooks/post-receive -- Arch Linux initscripts repository

1 0

[arch-projects] [netcfg] doing a netcfg release ?
by Rémy Oudompheng 22 May '11

22 May '11

Hello, I see the current netcfg release is more than a year old. It seems wise to do a new release with the current state of the git repository, which has accumulated several patches over time. If nobody objects, I can write some release notes and do the packaging. Regards, Rémy.

4 4

[arch-projects] [initscripts] [GIT] Arch Linux initscripts repository branch master updated. 2011.05.2-8-g68c61ed
by tomegun＠archlinux.org 22 May '11

22 May '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 68c61ed4ce2eddf29c4f047133406d72162fd396 (commit) from e0e1420a3506a5988b8f76139788f0ffea298e0c (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 68c61ed4ce2eddf29c4f047133406d72162fd396 Author: Dave Reisner <d(a)falconindy.com> Date: Thu Apr 28 22:42:03 2011 -0400 network/rc.sysinit: remove dead code Signed-off-by: Tom Gundersen <teg(a)jklm.no> ----------------------------------------------------------------------- Summary of changes: network | 5 ----- rc.sysinit | 2 -- 2 files changed, 0 insertions(+), 7 deletions(-) hooks/post-receive -- Arch Linux initscripts repository

1 0

[arch-projects] [initscripts] simple rc.d bash completion script v2
by Auguste Pop 20 May '11

20 May '11

this is another version of the bash completion script. removed an unnecessary echo and allows auto completion for not well-adopted usage like rc.d status foo. here's the patch: commit 47c48b15047ae46dd2e3d56559dad17d57f78c91 Author: Auguste Pop <auguste(a)gmail.com> Date: Fri May 20 22:13:31 2011 +0800 basic rc.d bash completion diff --git a/Makefile b/Makefile index ad8a058..a7f3ad2 100644 --- a/Makefile +++ b/Makefile @@ -1,5 +1,5 @@ VER := $(shell git describe) -DIRS := /etc/rc.d /etc/conf.d /etc/rc.d/functions.d /etc/cron.hourly /sbin /etc/modprobe.d +DIRS := /etc/rc.d /etc/conf.d /etc/rc.d/functions.d /etc/cron.hourly /sbin /etc/modprobe.d /etc/bash_completion.d minilogd: minilogd.o @@ -13,6 +13,7 @@ install: minilogd installdirs install -m644 -t $(DESTDIR)/etc/rc.d functions install -m755 -t $(DESTDIR)/etc/rc.d hwclock network netfs install -m755 -t $(DESTDIR)/sbin minilogd rc.d modprobe-blacklist + install -m644 bash_completion $(DESTDIR)/etc/bash_completion.d/rc.d ln -s /run/initscripts/modprobe-blacklist.conf $(DESTDIR)/etc/modprobe.d/arch-blacklist.conf clean: diff --git a/bash_completion b/bash_completion new file mode 100644 index 0000000..9d766fe --- /dev/null +++ b/bash_completion @@ -0,0 +1,56 @@ +#!/bin/bash + +_rc_d_getdaemons () +{ + local -a files=("$1"/*) + echo "${files[@]##*/}" +} + +_rc_d_arraydiff () +{ + local -A arr + + for elem in "${!1}"; do + arr[$elem]=1 + done + for elem in "${!2}"; do + unset arr[$elem] + done + echo "${!arr[@]}" +} + +_rc_d () +{ + local command commands available started + local cur words cword + + commands="list help start stop restart" + available=($(_rc_d_getdaemons /etc/rc.d)) + started=($(_rc_d_getdaemons /run/daemons)) + + _get_comp_words_by_ref -n =: cur words cword + [[ $cword -gt 1 ]] && command=${words[1]} + + if [[ -z "$command" ]]; then + COMPREPLY=($(compgen -W "$commands" -- $cur)) + else + case "$command" in + start) + COMPREPLY=($(compgen -W "$(_rc_d_arraydiff \ + available[@] started[@])" -X "functions*" -- $cur)) + ;; + stop|restart) + COMPREPLY=($(compgen -W "${started[*]}" -- $cur)) + ;; + list|help) + COMPREPLY=() + ;; + *) + COMPREPLY=($(compgen -W "${available[*]}" \ + -X "functions*" --$cur)) + ;; + esac + fi +} + +complete -F _rc_d rc.d

1 0

[arch-projects] [initscripts][RFC] managing /tmp
by Tom Gundersen 20 May '11

20 May '11

Hi guys, Thanks to all of you who answered my many previous RFC's, I will follow up on them with code soon :-) In the last initscripts release I inedvertently introduced a bug where /tmp is not cleared on boot if it is mounted on a non-tmpfs drive (it was only cleared if it lived on the root partition). This lead to a long discussion in the bug report about possible use-cases for /tmp. Among other things it was requested that we should stop clearing /tmp or make it configurable. As always, I'm opposed to adding configuration options just for the sake of it, so I propose something different. First of all, this is what I took away from the discussion: 1) /tmp should always be mounted as tmpfs (there are some really obscure scenarios when it should not be (if you have individual files larger than 256GB on it)) 2) if /tmp is not mounted as tmpfs it should always be cleared at boot That said, some users disagree, and want to /tmp to be preserved on reboot. Usually I would deny such a request, but I figured it would actually simplify things, so why not. This is my proposal: 1) We advice users to mount /tmp as tmpfs (possibly with the proviso that the partition they would normally have used for /tmp is turned into swap space, and the size option of /tmp is set to the size of the old partition), this could be added to the standard fstab if we wish. 2) We never touch /tmp at boot. This makes no difference if the users follow our advice. 3) If the user wants to mount /tmp as a different fs, we advice them to put "rm -rf /tmp/* /tmp/.*" in rc.local, or manage /tmp in any other way they want (e.g. by using the tmpwatch package). What do you think? Cheers, Tom

4 11

[arch-projects] [initscripts] [GIT] Arch Linux initscripts repository branch master updated. 2011.05.2-7-ge0e1420
by tomegun＠archlinux.org 19 May '11

19 May '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 e0e1420a3506a5988b8f76139788f0ffea298e0c (commit) via b6e17433c8838acdc48ccfe726325118792a171e (commit) via aacbe897cc822d1daaa44cb754d56666e4f07e84 (commit) via cc1b8a39309fef45d092140fa130d0e82c696a44 (commit) via 99d64f5789f01243903be4b32078ec38bdce75fa (commit) via 417ab8ce6c830042375da4ded4374a9f30a5387e (commit) from c6b3853de29e5bb42adda4a2f62a08d87ea15740 (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 e0e1420a3506a5988b8f76139788f0ffea298e0c Author: Gerardo Exequiel Pozzi <vmlinuz386(a)yahoo.com.ar> Date: Wed May 18 23:32:36 2011 +0200 add killall5 omit pid functionality (me) Update the offsets so can be merged now without any issues and respect the indentation. Implements FS#10536 - [initscripts] internal: save a pid from killall5 All credits to: Kurt J. Bosch Signed-off-by: Gerardo Exequiel Pozzi <vmlinuz386(a)yahoo.com.ar> Signed-off-by: Tom Gundersen <teg(a)jklm.no> commit b6e17433c8838acdc48ccfe726325118792a171e Author: Gerardo Exequiel Pozzi <vmlinuz386(a)yahoo.com.ar> Date: Wed May 18 23:24:38 2011 +0200 use mount -n for devpts and shm Signed-off-by: Gerardo Exequiel Pozzi <vmlinuz386(a)yahoo.com.ar> Signed-off-by: Tom Gundersen <teg(a)jklm.no> commit aacbe897cc822d1daaa44cb754d56666e4f07e84 Author: Thomas Bächler <thomas(a)archlinux.org> Date: Tue May 17 20:13:39 2011 +0200 Fix typo Signed-off-by: Tom Gundersen <teg(a)jklm.no> commit cc1b8a39309fef45d092140fa130d0e82c696a44 Author: Tom Gundersen <teg(a)jklm.no> Date: Wed May 18 23:08:13 2011 +0200 fixup: also clear /tmp if it is a mountpoint eda191cf78365cc3d1 inadvertently introduced a change so that /tmp was not cleared if it was a mountpoint. This restores the old behaviour. It is strongly adviced to not monut anything but a tmpfs on /tmp. This would avoid the issue. Signed-off-by: Tom Gundersen <teg(a)jklm.no> commit 99d64f5789f01243903be4b32078ec38bdce75fa Author: Tom Gundersen <teg(a)jklm.no> Date: Mon May 9 22:15:43 2011 +0200 udev: generate blacklist on boot udev will read /etc/modprobe.d/*.conf and blacklist all modules that are listed as blacklist <module> We parse rc.conf at boot and generate such a .conf file. It cannot be written to /etc this early, so we save it to /run. A symlink exists in /etc to get the desired functionality. With this patch (and an analogous one in mkinitcpio) load-modules.sh can be removed from the udev package without loss of significant functionality. Setting MOD_AUTOLOAD will then no longer take effect, nor will kernel parameters. Thouhgh, kernel parameters might still affect the initramfs as the implementation is independent. Original-idea-by: Benjamen Richer <br(a)waldteufel-online.net> Based-on-patch-by: David Reisner <d(a)falconindy.com> Signed-off-by: Tom Gundersen <teg(a)jklm.no> commit 417ab8ce6c830042375da4ded4374a9f30a5387e Author: Tom Gundersen <teg(a)jklm.no> Date: Mon May 16 13:36:18 2011 +0200 makefile: do not make functions executable This was a bug in the original conversion to makefile (installing functions twice, the last time with wrong permissions). Signed-off-by: Tom Gundersen <teg(a)jklm.no> ----------------------------------------------------------------------- Summary of changes: Makefile | 9 +++++---- functions | 16 ++++++++++++++-- modprobe-blacklist | 25 +++++++++++++++++++++++++ rc.conf | 2 +- rc.sysinit | 11 +++++++---- 5 files changed, 52 insertions(+), 11 deletions(-) create mode 100755 modprobe-blacklist hooks/post-receive -- Arch Linux initscripts repository

1 0