On Fri, 2016-07-08 at 14:38 +0300, Stepan Zakharov wrote: > Thanks. I will look into that option. Yes, callgrind and kcachegrind are very easy to use/very precise/..., but the price to pay is the slowness. Other profilers having much less overhead might be good enough.
> But I can also split my task. > And right now I want to understand the boundaries of a tool. At least > approximate. > To my knowledge, the size of the in-memory data of callgrind does not increase a lot if you run a long time/run many times the same instructions. The callgrind memory use will rather increase with the amount of different pairs of 'caller->callee' that you have, which probably depends more on the algorithms that you run rather than e.g. the amount of data being processed by the algorithm. That being said, you can also e.g. take a dump of the callgrind recording let's say every 24hours, so that in case of crash, you still have a significant part of the data. Philippe > ------------------------------------------------------------------------------ Attend Shape: An AT&T Tech Expo July 15-16. Meet us at AT&T Park in San Francisco, CA to explore cutting-edge tech and listen to tech luminaries present their vision of the future. This family event has something for everyone, including kids. Get more information and register today. http://sdm.link/attshape _______________________________________________ Valgrind-users mailing list Valgrind-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/valgrind-users
