On Friday 09 May 2014 09:00 PM, Valdis.Kletnieks@xxxxxx wrote: > On Fri, 09 May 2014 17:39:21 +0530, Jay Aurabind said: > >> So shouldnt we assume that initial value (somewhere around 6K) should be >> enough since the maximum it went down is only till its 50% mark ? > > Depends. Are you of the "we haven't seen it before, so it can't happen" > school of programming, or the "it could plausibly happen, so we should > guard against it" school? > > Consider you get down to that 6K mark - and now you hit that 1K allocation > that you didn't bother cleaning up because "we've never seen it before". And > then you hit a hardIRQ that *also* didn't bother cleaning up their 1K allocation > that *that* kernel hacker "had never seen it before". What happens to your > system? And how long is it going to take for you to figure out why every > several weeks, your system dies with a totally different random memory > overlay? > Point taken. Thank you Valdis, Martin, Paul and Frank for sharing your thoughts. Cheers, Jay
Attachment:
signature.asc
Description: OpenPGP digital signature
_______________________________________________ Kernelnewbies mailing list Kernelnewbies@xxxxxxxxxxxxxxxxx http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
