On Wed, May 10, 2006 at 05:39:36PM +0530, Abhijit Pawar wrote: > That was a very good piece of information. I really got the OOM now. One > small doubt though. > > It describes that OOM selects the process to kill after some initial > checking about the failures in last 5 sec , 10 sec and etc... > Then it pick up the fresh process with large VM. > Now in this time between selecting the process and killing the process , is > every other thing locked? > > What if the Process that requested the mem got some mem in the mean time.. > > I guess there will be one more round of checking before actually killing the > process. > > I may be wrong .. but this is what I deduce from the article... > > Also this will be a very rare phenomenon as the HIGH MEM zone will always > give us the enough mem with KSWAPDEMON. > > Hope I am not wrong here. :) > > Thanks. > -- > Abhijit. > Killing a process is a very extreme solution. The OOM process selects a process (depending on its age and its memory size) : in fact, it tries to select the process the user wants to kill. Once the process is selected, he _will_ be killed : there won't be no more checks. You should look at the source code http://sosdg.org/~coywolf/lxr/source/mm/oom_kill.c It's not too big, not too complex, it's well documented : easy to understand. -- tyler tyler@xxxxxxxx ___________________________________________________________________________ Faites de Yahoo! votre page d'accueil sur le web pour retrouver directement vos services préférés : vérifiez vos nouveaux mails, lancez vos recherches et suivez l'actualité en temps réel. Rendez-vous sur http://fr.yahoo.com/set -- Kernelnewbies: Help each other learn about the Linux kernel. Archive: http://mail.nl.linux.org/kernelnewbies/ FAQ: http://kernelnewbies.org/faq/
