On Wednesday 20 of May 2009, Anshul Jain wrote:
On Wed, May 20, 2009 at 2:20 PM, Lubos Lunak
wrote: One possible and very likely answer is that you are just not measuring properly, since it's very non-trivial to measure memory usage on Linux and 99.9% people get it wrong. See http://ktown.kde.org/~seli/memory/ .
If you are sure your numbers are right, then the easiest option would be finding out what uses that memory.
PS: It is possible to get Exmap from home:/llunak:/mix repo, but it crashes on 64bit for an unknown reason.
Yes, you're right...measuring memory under Linux is not an easy task since it involves a ton of parameters. Now what did surprise me was the difference in memory usage between 11.0 and 11.1 using the *same* utility, i.e. htop. I can understand htop being a ways off in its measurement...but a delta measurement between 11.0 and 11.1 should be atleast some indicator...right?
No. Doing something wrong the same way twice doesn't necessarily mean the results will be wrong the same way.
I tried to narrow down the specific app that consumed memory...and didn't see anything going out of control. X, kded, plasma, krunner etc used pretty much the same amount of memory as they did in 11.0...except that the overall usage was considerably higher. Maybe its the kernel? I wouldn't know...just guessing thats all.
I can't tell you much more than what I've already said, maybe except that in benchmarking there is not much room for guessing. -- Lubos Lunak KDE developer -------------------------------------------------------------- SUSE LINUX, s.r.o. e-mail: l.lunak@suse.cz , l.lunak@kde.org Lihovarska 1060/12 tel: +420 284 084 672 190 00 Prague 9 fax: +420 284 028 951 Czech Republic http://www.suse.cz -- To unsubscribe, e-mail: opensuse-kde+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-kde+help@opensuse.org