Bugzilla Bug 99446

The acpid ebuild does not populate /etc/acpi or /etc/acpi/events with any
action
triggers or scripts. On x86/amd64, I believe it should include at least actions
for the power and sleep buttons, as well as variables in /etc/conf.d to
determine what to do on these keypresses (e.g. shutdown or hibernate on power
button).

There should also be separate packages with acpid scripts for various
non-standard acpi events like those sent by laptop keyboard Fn keypresses.

Forcing the user to look for and manually install the necessary acpid scripts
just to get acpi button functionality is unacceptable in my opinion.

sys-power/acpid _does_ install a default handler for the power button. The
other
ACPI events are specific to the machine in question, which is also why acpid
itself doesn't include such scripts.

The Gentoo ebuild includes a basic skeleton which users can populate with the
behaviour of their choice.

The acpi sleep button and the lid button are standard, as far as I know:

blackbird events # cat /etc/acpi/events/power
event=button[ /]power
action=/etc/acpi/powerbtn.sh
blackbird events # cat /etc/acpi/events/sleep
event=button[ /]sleep
action=/etc/acpi/suspend.sh 3
blackbird acpi # cat events/lid
event=button[ /]lid
action=/etc/acpi/lid.sh

Even if some systems fire different ACPI events on these buttons, it doesn't
hurt to include handlers for these. The handlers need not be active by default,
but they should be there: a novice user has no idea that they need to populate
/etc/acpi with scripts of their own for these events to work. The scripts could
be shipped as a separate package, but they need to be available.

acpid does include the power button action; sorry, I was wrong about that.

why not grab the scripts here:
http://packages.ubuntu.com/hoary/admin/acpi-support

and install them into the docdir ?

Good idea. I'll take a closer look at those soon.

*** Bug 105389 has been marked as a duplicate of this bug. ***

Can we please see some progress on this?

(In reply to comment #6)
> Can we please see some progress on this?

Sure. Feel free to attach a working ebuild for the acpi-support package which
integrates it into the existing gentoo power management framework.

Created an attachment (id=89769) [details]
acpi-support-0.85.ebuild

This still needs a lot of work, but I'm interested in comments.

Sleep and hibernate functions are implemented through hibernate-script.
There are vendor-specific dependencies, I haven't gone through them all yet.

I'd like to implement/finish up a few features to make this more comprehensive:
-a service script to set power policy on components when starting
-cpufreq management
-NIC, wireless, display, hdd power management policy
-better integration with hibernate-script

(In reply to comment #8)
> This still needs a lot of work, but I'm interested in comments.

Cool, thank you for working on this. I'll try to find time to review it and
comment on it after my exams last week. It would be nice if we could finally
get this into shape and added to portage. :)

Currently there's a bug in the package that prevents acpid from receiving an
event when acpi_fakekey presses a button (e.g. `acpi_fakekey 142` should press
the sleep button). The button keypress is sent to /dev/input/eventX, but acpid
doesn't receive it (it should receive an event through /proc/acpi/event).

... never mind, this seems to be a problem with my machine.

On further examination this is not a problem with my machine but a known
regression in acpi_fakekey. I'll have to find a way to fix it before we can go
on with this package.

Created an attachment (id=90474) [details]
acpi-support-0.85.ebuild

OK, I have disabled acpi_fakekey for the ACPI buttons and am calling the
scripts directly instead for those. I've worked a bit on the package, please
test it and let me know if you see anyting wrong with the scripts or the
organization.

Also a tiny update to the ebuild (added a doc install)

The latest ebuild version is at

http://horizon.ath.cx/packages/acpi-support-0.85.ebuild

Also I'm not sure how to proceed with the version number. This package was spun
off ubuntu's acpi-support, but I've heavily modified it, so I'm not sure what
numbering scheme to apply or whether to rename the package to avoid confusion.

(In reply to comment #14)
> The latest ebuild version is at
> 
> http://horizon.ath.cx/packages/acpi-support-0.85.ebuild
> 
> Also I'm not sure how to proceed with the version number. This package was spun
> off ubuntu's acpi-support, but I've heavily modified it, so I'm not sure what
> numbering scheme to apply or whether to rename the package to avoid confusion.
> 

Why not create a brand new package like, say, acpi-scripts-0.0.1 or
acpi-scripts-gentoo-0.0.1 or something?

I get the following error when starting the init script provided:

cat: /var/lib/acpi-support/system-manufacturer: No such file or directory
cat: /var/lib/acpi-support/system-product-name: No such file or directory
cat: /var/lib/acpi-support/system-version: No such file or directory
cat: /var/lib/acpi-support/bios-version: No such file or directory

compiled and works fine on amd64.

emerge --info
Portage 2.1.2_rc4-r9 (default-linux/amd64/2006.1, gcc-4.1.1, glibc-2.5-r0,
2.6.19-suspend2-r1 x86_64)
=================================================================
System uname: 2.6.19-suspend2-r1 x86_64 AMD Turion(tm) 64 Mobile Technology
MT-37
Gentoo Base System version 1.12.8
Timestamp of tree: Sat, 13 Jan 2007 13:50:01 +0000
distcc 2.18.3 x86_64-pc-linux-gnu (protocols 1 and 2) (default port 3632)
[enabled]
ccache version 2.4 [disabled]
dev-java/java-config: 1.3.7, 2.0.31-r1
dev-lang/python:     2.4.4
dev-python/pycrypto: 2.0.1-r5
dev-util/ccache:     2.4-r6
sys-apps/sandbox:    1.2.18.1
sys-devel/autoconf:  2.13, 2.61
sys-devel/automake:  1.4_p6, 1.5, 1.6.3, 1.7.9-r1, 1.8.5-r3, 1.9.6-r2, 1.10
sys-devel/binutils:  2.17
sys-devel/gcc-config: 1.3.14
sys-devel/libtool:   1.5.22
virtual/os-headers:  2.6.19
ACCEPT_KEYWORDS="amd64 ~amd64"
AUTOCLEAN="yes"
CBUILD="x86_64-pc-linux-gnu"
CFLAGS="-march=athlon64 -msse3 -O2 -pipe"
CHOST="x86_64-pc-linux-gnu"
CONFIG_PROTECT="/etc /usr/kde/3.5/env /usr/kde/3.5/share/config
/usr/kde/3.5/shutdown /usr/share/X11/xkb /usr/share/config"
CONFIG_PROTECT_MASK="/etc/env.d /etc/env.d/java/ /etc/gconf
/etc/java-config/vms/ /etc/revdep-rebuild /etc/splash /etc/terminfo
/etc/texmf/web2c"
CXXFLAGS="-march=athlon64 -msse3 -O2 -pipe"
DISTDIR="/usr/portage/distfiles"
FEATURES="autoconfig distcc distlocks metadata-transfer parallel-fetch sandbox
sfperms strict"
GENTOO_MIRRORS="http://gentoo.osuosl.org/
ftp://distro.ibiblio.org/pub/linux/distributions/gentoo
http://distro.ibiblio.org/pub/linux/distributions/gentoo
ftp://ftp.gtlib.gatech.edu/pub/gentoo http://www.gtlib.gatech.edu/pub/gentoo
http://gentoo.cites.uiuc.edu/pub/gentoo/
ftp://gentoo.cites.uiuc.edu/pub/gentoo/"
LANG="en_US.UTF-8"
LINGUAS="en"
MAKEOPTS="-j5"
PKGDIR="/usr/portage/packages"
PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress
--force --whole-file --delete --delete-after --stats --timeout=180
--exclude=/distfiles --exclude=/local --exclude=/packages"
PORTAGE_TMPDIR="/var/tmp"
PORTDIR="/usr/portage"
PORTDIR_OVERLAY="/usr/local/portage"
SYNC="rsync://rsync.namerica.gentoo.org/gentoo-portage"
USE="X a52 aac accessibility acl acpi ada addbookmarks aiglx alias alsa amd64
amr arts audiofile autoreplace avahi bash-completion berkdb bitmap-fonts
bittorrent bluetooth bzip2 cairo calendar cjk clearcase cli connectionstatus
contactnotes cpudetection cracklib crypt cscope css cups curl cvs dbus dhcp
dlloader dmi doc dri dv dvd dvdr dvdread eap-tls emacs encode examples exif fam
fame fbsplash ffmpeg finger flac fortran ftp gadu gcc64 gcj gd gdbm gif glep
gnomecanvas gnutls gpm graphviz groupwise gs gstreamer gtk hal haskell
highlight history iconv idn ieee1394 imagemagick imap imlib intl ipv6 irc irda
irmc isdnlog jabber jack java javamail javascript jingle jpeg jpeg2k kde kipi
kqemu lame latex libcaca libg++ lirc lm_sensors log4j lua lucene lzo mad mbrola
md5sum mikmod mjpeg mng modplug mono mp3 mp4 mpeg musepack musicbrainz ncurses
netjack network nfs nls nowlistening nptl nptlonly nvidia objc objc++ objc-gc
offensive ogg openal openexr opengl oss pager pam pascal patch pcmcia pcre pdf
perforce perl php pmu png posix ppds pppd python qt3 qt3support qt4 quicktime
radius readline reflection rhino rle ruby samba sametime sdl seamonkey session
slp sndfile sockets spell spl sql sqlite ssl statistics subversion svg tcl tcpd
texteffect theora threads tiff timidity tk translator truetype truetype-fonts
type1-fonts udev unicode usb v4l v4l2 vcd vim vim-pager vim-with-x
visualization vorbis vorbis-psy webpresence wifi winpopup wmf xcb xinerama xml
xorg xpm xv xvid yahoo yv12 zeroconf zip zlib" ALSA_CARDS="intel8x0"
ALSA_PCM_PLUGINS="adpcm alaw asym copy dmix dshare dsnoop empty extplug file
hooks iec958 ioplug ladspa lfloat linear meter mulaw multi null plug rate route
share shm softvol" ELIBC="glibc" INPUT_DEVICES="keyboard synaptics mouse evdev"
KERNEL="linux" LINGUAS="en" LIRC_DEVICES="sir" USERLAND="GNU" VIDEO_CARDS="vga
nv nvidia"
Unset:  CTARGET, EMERGE_DEFAULT_OPTS, INSTALL_MASK, LC_ALL, LDFLAGS,
PORTAGE_RSYNC_EXTRA_OPTS

oops, other then the init script problem mentioned in comment #16.

To fix the init script : i just did

mkdir /var/log/acpi-support/
touch /var/log/acpi-support/bios-version
touch /var/log/acpi-support/system-manufacturer
touch /var/log/acpi-support/system-product-name
touch /var/log/acpi-support/system-version

Probably the ebuild should be updated to create these.

Created an attachment (id=111555) [details]
acpi-support-0.91.ebuild

New ebuild: now uses ubuntu's original tarball and patches it. Please test.

Created an attachment (id=111557) [details]
acpi-support-0.91-gentoo.patch

Patch for 0.91 (goes into $overlay/sys-power/acpi-support/files)

The latest ebuilds will now be found at:

http://www.gentoo-sunrise.org/sunrise/browser/sunrise/sys-power/acpi-support

acpi-support 0.91 w/patch worked fine here on amd64.

Revision bump to 0.95 (I'm guessing this will ship with feisty), some more
integration, new link (sunrise has moved):

http://overlays.gentoo.org/proj/sunrise/browser/sunrise/sys-power/acpi-support

The current acpi-support package has a minor bug where acpi-fakekey sends its
events to the wrong device (in my case, it sent it to the power button) : a
patch can be found at http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=433771 ,
which fixed the problem for me.

Andrey, are you trying to push your patches to upstream?

(In reply to comment #26)
> Andrey, are you trying to push your patches to upstream?
> 

I'm not so far, because my changes are very minor and Gentoo-centric. I'll
generalize them and break out a patch to send upstream in a few days, along
with making an ebuild for 0.109 which is the latest version that will probably
ship with Hardy.

I'm taking this bug since I'm now (almost) a dev and have been de facto
maintaining this in Sunrise for two years and no one else has expressed
interest.

By the way, I've done a bit of research about this and it seems that Fedora and
others uses a hal-based solution which Ubuntu and Debian are adopting and
trying to deprecate acpi-support.

(In reply to comment #28)
> By the way, I've done a bit of research about this and it seems that Fedora and
> others uses a hal-based solution which Ubuntu and Debian are adopting and
> trying to deprecate acpi-support.
> 

Ah, thanks. I'll have to track down that hal-based package and turn my
attention to porting that to Gentoo.

For further reference:

https://wiki.ubuntu.com/PMUtilsSpec

The function of acpi-support is nebulous in ubuntu. The suspend/hibernate
functionality is now being handled by the cross-platform pm-utils package which
acpi-support needs to depend on. The handlers for vendor-specific acpi events
are still provided by acpi-support but it seems that the plan is to provide
hal/udev rules for all of them.

(In reply to comment #29)
> (In reply to comment #28)
> > By the way, I've done a bit of research about this and it seems that Fedora and
> > others uses a hal-based solution which Ubuntu and Debian are adopting and
> > trying to deprecate acpi-support.
> > 
> 
> Ah, thanks. I'll have to track down that hal-based package and turn my
> attention to porting that to Gentoo.
> 

There is nothing to track down really. They are porting the drivers to use the
generic input devices  from kernel to userspace. A great example of such is for
asus-laptop module which is gonna be done up in 2.6.29 hopefully, patches are
already avaliable and can be applied against 2.6.28 for testing.

http://patchwork.kernel.org/patch/3305/
http://patchwork.kernel.org/patch/3043/

(In reply to comment #30)
> For further reference:
> 
> https://wiki.ubuntu.com/PMUtilsSpec
> 
> The function of acpi-support is nebulous in ubuntu. The suspend/hibernate
> functionality is now being handled by the cross-platform pm-utils package which
> acpi-support needs to depend on. The handlers for vendor-specific acpi events
> are still provided by acpi-support but it seems that the plan is to provide
> hal/udev rules for all of them.
> 

With the new generic input layer being used hal will get all its info from the
kernel on how to configure the keyboard for X use. No need to create any custom
rules.