Hi, today did some quick experiments in order to see if preforking gives us any improvements regarding memory usage. I launched 5 instances of a very simple pygtk activity (http://dev.laptop.org/git?p=users/marco/edit-activity;a=summary) and executed 'free' before and after, along with ps_mem.py.
I would like to do more interesting experiments in the near future (different activities, etc), but any comments about this data will be appreciated. From my unexperienced look, we are saving around 3.8MB per instance. http://www.pixelbeat.org/scripts/ps_mem.py Normal: total used free shared buffers cached Mem: 237836 183240 54596 0 0 79420 -/+ buffers/cache: 103820 134016 8.6 MiB + 10.3 MiB = 18.9 MiB Edit <db74d9451 9.0 MiB + 10.4 MiB = 19.4 MiB Edit <f28c569d2 9.0 MiB + 10.4 MiB = 19.4 MiB Edit <f6f10dcc6 9.0 MiB + 10.4 MiB = 19.4 MiB Edit <f0b1f10b7 9.0 MiB + 10.4 MiB = 19.4 MiB Edit <672d9cb6f total used free shared buffers cached Mem: 237836 228948 8888 0 0 81940 -/+ buffers/cache: 147008 90828 Preforking: total used free shared buffers cached Mem: 237836 183396 54440 0 0 80324 -/+ buffers/cache: 103072 134764 4.4 MiB + 13.7 MiB = 18.0 MiB Edit <040689fe6 4.8 MiB + 13.8 MiB = 18.6 MiB Edit <e97d30e3a 4.8 MiB + 13.8 MiB = 18.6 MiB Edit <0d80eebf8 4.9 MiB + 13.8 MiB = 18.6 MiB Edit <ceff1fd08 4.9 MiB + 13.8 MiB = 18.6 MiB Edit <dd5d72fb3 total used free shared buffers cached Mem: 237836 208924 28912 0 0 80780 -/+ buffers/cache: 128144 109692 Thanks, Tomeu _______________________________________________ Sugar mailing list [email protected] http://lists.laptop.org/listinfo/sugar
