Summary: | gnome-extra/gnome-power-manager lists two batteries but system has only one | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Priit Laes (IRC: plaes) <plaes> |
Component: | [OLD] GNOME | Assignee: | Freedesktop bugs <freedesktop-bugs> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | andrew, freedesktop-bugs, wingrunr21 |
Priority: | High | ||
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: | hal-0.5.10-sysfs-battery-fixes.patch |
Description
Priit Laes (IRC: plaes)
2007-12-25 14:42:54 UTC
Not an issue with gpm, its hal that lists it twice and thats caused by the duplication in /proc/acpi (disable your /proc/acpi, yes, Im aware, no acpid, but thats unneeded). *** Bug 212859 has been marked as a duplicate of this bug. *** (In reply to comment #1) > Not an issue with gpm, its hal that lists it twice and thats caused by the > duplication in /proc/acpi (disable your /proc/acpi, yes, Im aware, no acpid, > but thats unneeded). > That is all well and good but it does not fix the underlying issue. The patches referenced in the following blog post should be applied to the ebuild so that hal can correctly coexist with both a sysfs and procfs battery interface. Blog: http://dkukawka.blogspot.com/2008/01/hal-sysfs-acpi-batteries-fixed.html in fact acpid still works even when removing the deprecated stuff from /proc that would make it stop showing 2 batteries so I can't see a good reason to keep the deprecated interfaces. (In reply to comment #4) > in fact acpid still works even when removing the deprecated stuff from /proc > that would make it stop showing 2 batteries so I can't see a good reason to > keep the deprecated interfaces. > Bug 212859 has a patch that can be applied to Hal 0.5.10 to fix this problem without removing the /proc stuff. http://bugs.gentoo.org/show_bug.cgi?id=212859 reopening since using only sysfs interface had issues I didn't figured out before. Created attachment 150331 [details, diff]
hal-0.5.10-sysfs-battery-fixes.patch
attaching patch from the duped bug.
@gentopia, hey guys, could we get your opinion on this bug+patch ? it looks like it is quiet the same patch i proposed for BUG #213469. So yes for me it works perfectly, with and without /proc enabled additionaly to /sysfs gnome-power-manager realizes if the Batterystatus changes, and shuts down whenever the Battery goes close to empty... (In reply to comment #9) > it looks like it is quiet the same patch i proposed for BUG #213469. > > So yes for me it works perfectly, with and without /proc enabled additionaly to > /sysfs > > gnome-power-manager realizes if the Batterystatus changes, and shuts down > whenever the Battery goes close to empty... > It is the same patch. Gilles just now realized what I tried to tell him before. We needed a patch that fixes the issue. Simply removing the /proc interfaces was a workaround and a poor one at that. This patch works fine for me as well (as I posted earlier). (In reply to comment #10) > It is the same patch. Gilles just now realized what I tried to tell him > before. We needed a patch that fixes the issue. Simply removing the /proc > interfaces was a workaround and a poor one at that. This patch works fine for > me as well (as I posted earlier). > be more explicit next time ;) Disabling /proc interface wasn't enough because the sysfs interface isn't polled which results in g-p-m not updating battery status anymore. Now I'd really like to hear from gentopia people because I don't want to break hal :) > be more explicit next time ;)
> Disabling /proc interface wasn't enough because the sysfs interface isn't
> polled which results in g-p-m not updating battery status anymore. Now I'd
> really like to hear from gentopia people because I don't want to break hal :)
>
I'm not sure this is going to matter much longer since hal 0.5.11 is in RC2 status.
Well, until a new Hal _is_ released, this bug is a serious piss off for anyone who has upgraded to Linux 2.6.24 and has unwittingly agreed to setting CONFIG_ACPI_SYSFS_POWER. The upstream bug seems to be: http://bugs.freedesktop.org/show_bug.cgi?id=13669 which is reported fixed, but obviously is not in a release yet. If that is the origin of the patch suggested, then it really would be excellent if it could be included in Gentoo's hal-0.5.10 package as a revbump. Various other distro reports that may be relevant: https://bugs.launchpad.net/ubuntu/+source/gnome-power-manager/+bug/194052 https://bugs.launchpad.net/ubuntu/+source/gnome-power-manager/+bug/177570 https://bugzilla.novell.com/show_bug.cgi?id=342808 http://bugs.archlinux.org/task/9605 but whatever. The workaround I just employed was to revert CONFIG_ACPI_SYSFS_POWER to 'n' but that seems a bit backward looking. AfC You could just apply the linked patch. Create an overlay out of the hal 0.5.10 ebuild and add an epatch line src_unpack() I think this has been solved by now... So closing. |