[Webkit-unassigned] [Bug 48212] SVG FELighting performance issues

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=48212





--- Comment #9 from Zoltan Herczeg <zherczeg at webkit.org>  2010-10-27 09:47:13 PST ---
> You don't need to recompile for using callgrind, and it is very useful to see how much time is spend in which function. Valgrind is a tool chain, and consists of different tools, not only cachegrind (I guess you were just thinking about this). kcachegrind2 on the other hand is just a graphical tool and makes it quiet easy to visualize the output of callgrind.

I like valgrind, I even wrote valgrind tools myself. Valgrind using a recompiler called libVEX. It reads the machine instructions from the binary, convert them to a high level assembly language, where a tool can modify this stream (thus it is really easy to insert calls into the stream, that is what most tools do). Later, this high level assembly language is compiled back to binary form on-the-fly by libVEX again. If your tool does not touch the stream, you get a 3 times slower binary code. Hence it is useless for performance related things. But you can follow the control flow (inserting calls before function entries, and such), you can insert checks before each memory reads/writes (memcheck), and so on. If you need info, but you don't care about speed -> valgring (libVEX) is your tool. If speed matters, use cycle accurate simulators (we also have one for arm: http://www.inf.u-szeged.hu/xeemu/)

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.