Bernhard Reiter wrote: > Am Montag, 28. Februar 2011 08:07:49 schrieb Rolf Eike Beer:
>> [41887.926269] Out of memory: kill process 3543 (browserd) score >> 157581312 >> or a child >> [41887.926483] Killed process 3543 (browserd) > >> [42006.473602] Out of memory: kill process 1508 (browserd) score >> 143491072 >> or a child >> [42006.473815] Killed process 1508 (browserd) > >> [42036.066650] Killed process 1397 (hildon-thumbnai) > >> [42041.402557] Out of memory: kill process 4286 (akonadi_control) score >> 1501828 or a child >> [42041.402770] Killed process 4288 (akonadiserver) > > The last process it killed was akonadiserver. No idea why it killed the > others > first. The systems certainly will work fine until it goes into the > swapping > condition. So memory consumption of the processes is something we need to > look at. Well, that's the general policy decision of the kernel. Since the kernel is a bit older (2.6.28) that may be even a rather stupid one, i.e. kill the process that allocates the memory. Or even more likely the one that references it's previously allocated memory for the first time. You could read about "overcommitment" if you are interested in the dirty details. Eike _______________________________________________ Kde-mobile-users mailing list [email protected] https://mail.kde.org/mailman/listinfo/kde-mobile-users
