On 11/10/2014 12:32 AM, David Gerard wrote: > Here are five plots of master (as of last night), and five of > diizy/memmgr (as of last night). Only thing running was lmms in a > terminal. > > It's a lot more consistent. Bottom cluster is master, top cluster is > diizy/memmgr. I also set the scale to 750 instead of 2000 to make it a > bit clearer. >
Ok then, we're back to square one then. Note to all: benchmarks need to be done with no other software running, otherwise the noise from other software will muddle the benchmarking. Now the thing to do would be to go through the commits in the memmgr branch one by one, and find out if there's a clear spot where the CPU usage increases... ------------------------------------------------------------------------------ _______________________________________________ LMMS-devel mailing list LMMS-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/lmms-devel
