Bug#366008: hal: hibernate functionality will break power/ibooks
On Tue, May 09, 2006 at 10:07:47AM +0200, Johannes Berg wrote: > On Tue, 2006-05-09 at 10:03 +0200, Sjoerd Simons wrote: > > > Actually i already patched hal to have > > power_management.can_suspend_to_disk = false on pmu systems. But your right, > > programs can still call that hal interface (although i hoped no actually > > do). > No, actually, I could suspend to disk with hibernate, just not to RAM. Oh right, i was somewhat confused. Yeah this can be fixed easily by placing the hibernate command after the check for pmu systems, will fix in the next package :) Thanks! Sjoerd -- I put up my thumb... and it blotted out the planet Earth. -- Neil Armstrong -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#366008: hal: hibernate functionality will break power/ibooks
On Wed, May 03, 2006 at 11:30:41AM +0200, Johannes Berg wrote: > Package: hal > Version: 0.5.8-1 > Severity: normal > > Hi, > > On power/ibook systems that have hibernate installed, hibernate to ram > will not work properly because it tries to use something like > > echo -n mem > /sys/power/state > > which is broken. I've submitted a kernel patch to stop mem and standby > from showing up in /sys/power/state and am running that patch, which > means that I was just lucky that it didn't kill my system a few minutes > ago! > > Now, the scripts in other places work around this by > disallowing /sys/power/state for pmu based systems because those allow > crashing the system through that file. > > In any case, the point here is that this check should also be applied to > the new patch in order to not break pmu based systems that happen to > have hibernate installed. Actually i already patched hal to have power_management.can_suspend_to_disk = false on pmu systems. But your right, programs can still call that hal interface (although i hoped no actually do). I'll try to make the patch more watertight for the next package :) Sjoerd -- If it wasn't for Newton, we wouldn't have to eat bruised apples. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#366008: hal: hibernate functionality will break power/ibooks
On Tue, 2006-05-09 at 10:03 +0200, Sjoerd Simons wrote: > Actually i already patched hal to have > power_management.can_suspend_to_disk = false on pmu systems. But your right, > programs can still call that hal interface (although i hoped no actually do). No, actually, I could suspend to disk with hibernate, just not to RAM. johannes signature.asc Description: This is a digitally signed message part
Bug#366008: hal: hibernate functionality will break power/ibooks
Package: hal Version: 0.5.8-1 Severity: normal Hi, On power/ibook systems that have hibernate installed, hibernate to ram will not work properly because it tries to use something like echo -n mem > /sys/power/state which is broken. I've submitted a kernel patch to stop mem and standby from showing up in /sys/power/state and am running that patch, which means that I was just lucky that it didn't kill my system a few minutes ago! Now, the scripts in other places work around this by disallowing /sys/power/state for pmu based systems because those allow crashing the system through that file. In any case, the point here is that this check should also be applied to the new patch in order to not break pmu based systems that happen to have hibernate installed. Yes, I know that the whole situation with PMU based systems sucks majorly. I'm working on fixing it, but it isn't trivial. A first step was to actually not make echo -n mem > /sys/power/state crash on PMU based systems by disallowing it completely. johannes -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
