I was just talking to myself, because i'm tending to overcomplicate things ;)
Cheers, Ralf. Scott Hyndman wrote: > Who you calling stupid? ;) > > Yeah, I'll let you know. > > Scott > > -----Original Message----- > From: [EMAIL PROTECTED] on behalf of Ralf Bokelberg > Sent: Thu 1/19/2006 9:38 AM > To: Open Source Flash Mailing List > Cc: > Subject: Re: [osflash] hamtasc: stacktrace available > > I'd say kiss, since normaly i'm are not interested in the absolute > timing, but only need to know, where the hotspots are. Anyway, i'm eager > to see, what you come up with. > > Cheers, > Ralf. > > > Scott Hyndman wrote: > > >>Good point Nicolas, >> >>But to amend what you've said, in many cases it would require even more than >>the subtraction of only two getTimer() calls. Any function that calls any >>other function internally would have to have its time changes based on the >>number of internal function calls it performs. Since we have the stack data, >>we can have these times kind of bubble up, or just an internal function count >>or something...but this will add even more overhead. >> >>I'm going to give it a shot. I'll let everyone know, >> >>Scott >> >>-----Original Message----- >>From: [EMAIL PROTECTED] on behalf of Nicolas Cannasse >>Sent: Thu 1/19/2006 2:21 AM >>To: Open Source Flash Mailing List >>Cc: >>Subject: Re: [osflash] hamtasc: stacktrace available >> >> >> >>>very nice...now we just need someone to write a tool that will record >>>the entry & exit every time a function is called, as well as a timestamp >>>(from getTimer()), then process all this data to profile your app, >>>giving output similar to the gprof tool. >>> >>> % cumulative self self total >>>time seconds seconds calls ms/call ms/call name >>>17.7 3.72 3.72 13786208 0.00 0.00 Ns_DStringNAppend [8] >>> 6.1 >>> 5.00 1.28 107276 0.01 0.03 MakePath [10] >>> 2.9 5.60 0.60 1555972 0.00 0.00 Ns_DStringFree [35] >>> 2.7 6.18 0.58 1555965 0.00 0.00 Ns_DStringInit [36] >>> >>>2.3 6.67 0.49 1507858 0.00 0.00 ns_realloc [40] >>>[example pulled from random webpage] >>> >>>That would be helpful for finding bottlenecks in your code. >> >> >>The problem is that for very small function getting called often, the >>overhead of timestamp operations will be too much big and will greatly >>increase the weight of theses functions in the profile report. Unless >>you run some benchmark first to deduce the cost of the 2 getTimers and >>subtract it from the time spent in function. >> >>Nicolas >> >>_______________________________________________ >>osflash mailing list >>[email protected] >>http://osflash.org/mailman/listinfo/osflash_osflash.org >> >> >> >> >> >>------------------------------------------------------------------------ >> >>_______________________________________________ >>osflash mailing list >>[email protected] >>http://osflash.org/mailman/listinfo/osflash_osflash.org > > > _______________________________________________ > osflash mailing list > [email protected] > http://osflash.org/mailman/listinfo/osflash_osflash.org > > > > > > ------------------------------------------------------------------------ > > _______________________________________________ > osflash mailing list > [email protected] > http://osflash.org/mailman/listinfo/osflash_osflash.org _______________________________________________ osflash mailing list [email protected] http://osflash.org/mailman/listinfo/osflash_osflash.org
