On Sun, 28 Mar 2010 21:41:51 +0300 Tom Hacohen <[email protected]> said:
> On Sat, Mar 27, 2010 at 7:23 AM, Carsten Haitzler <[email protected]>wrote: > > > valgrind is happy with elm's own genlist tests.... what are you doing so > > differently? > > > I just ran: > valgrind --leak-check=full elementary_test > Did nothing but looking at the main page (and then closing it with the 'X'). > And I got: > > ==16449== LEAK SUMMARY: > ==16449== definitely lost: 768 bytes in 3 blocks > ==16449== indirectly lost: 6,340 bytes in 317 blocks > ==16449== possibly lost: 186,566 bytes in 985 blocks > ==16449== still reachable: 526,725 bytes in 5,238 blocks > > So I'm not really sure valgrind is happy at all :P (Maybe my e version isn't > new enough, I'm using 46848, but even that, would that leak so much?) totally diifferent issue to brian's and as such the definitely lost line and indirectly lost are all that really count and they will be pretty much constant as they are 1-off allocs for things used during all of runtime that on shutdown have pointers lost. you need to actually understand memory and allocation before quoting valgrind numbers. -- ------------- Codito, ergo sum - "I code, therefore I am" -------------- The Rasterman (Carsten Haitzler) [email protected] ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ enlightenment-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
