From: Subrata M. <su...@li...> - 2009-04-07 07:21:26
|
Hi Li, On Tue, 2009-04-07 at 11:15 +0800, Li Zefan wrote: > Subrata Modak wrote: > > Hi Li, > > > > On Fri, 2009-04-03 at 17:31 +0800, Li Zefan wrote: > >>>>> Could you help us: > >>>>> 1) Let us know the additional testing in Memory Controller that is > >>>>> needed in the backdrop of the new enhancement(s) made in 2.6.29. Would > >>>>> you like to provide some tests for them as well to LTP ? > >>>>> 2) Review the existing Memory Controller test cases in LTP and tell us > >>>>> how do you feel about them ? > >>>>> > >>>> I'm now co-operating with China members(cn.fujitsu.com) and maybe you know > >>>> memcg's test case was enhanced by them(Li Zefan at el.). > >>> Thanks for your reply. Li Zefan also works closely with LTP. And he is > >>> also involved in adding tests to LTP. I am happy to see that he is also > >>> working on the memcg´s test. > >>> > >>> Li, > >>> > >>> Please find your best time to share with LTP. > >>> > >> I have not test programs that test those new features in memory controller, > >> but I have a few test cases that reproduce (almost) all memcg bugs and > >> cgroup bugs that I found. > > > > I think they should be a good candidate for LTP. If you have hit bugs > > with them, then they will be good to track any future regression(s). > > That's why I wrote down the test programs, and they might trigger some > unknown races. > > > Feel free to send a patch adding them to appropriate place in LTP. > > > > I'll find time to do this. Thanks. One more thing. I do not want to annoy you often. But, this is regarding the CPUSETS tests that you said will be sending to LTP. Is there something ready yet ? I would prefer you to start a thread even with a small/single test case, and, then keep adding when you find time. Regards-- Subrata |