<div dir="ltr"> <div class="gmail_extra"> <div class="gmail_quote">On Thu, Sep 11, 2014 at 3:53 PM, Dieter Nützel <<a href="mailto:Dieter@nuetzel-hh.de" target="_blank">Dieter@nuetzel-hh.de</a>> wrote: <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Am 12.09.2014 00:31, schrieb Jason Ekstrand:<div><div class="h5"> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> On Thu, Sep 11, 2014 at 2:55 PM, Dieter Nützel <<a href="mailto:Dieter@nuetzel-hh.de" target="_blank">Dieter@nuetzel-hh.de</a>> wrote: <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> Am 15.08.2014 04:50, schrieb Jason Ekstrand: <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> On Aug 14, 2014 7:13 PM, "Dieter Nützel" <<a href="mailto:Dieter@nuetzel-hh.de" target="_blank">Dieter@nuetzel-hh.de</a>> wrote: <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> Am 15.08.2014 02:36, schrieb Dave Airlie: <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> On 08/02/2014 02:11 PM, Jason Ekstrand wrote: <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> Most format conversion operations required by GL can be </blockquote></blockquote></blockquote></blockquote></blockquote> performed by <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> converting one channel at a time, shuffling the channels </blockquote></blockquote></blockquote></blockquote></blockquote> around, and <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> optionally filling missing channels with zeros and ones. </blockquote></blockquote></blockquote></blockquote></blockquote> This adds a <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> function to do just that in a general, yet efficient, way. v2: * Add better comments including full docs for functions * Don't use __typeof__ * Use inline helpers instead of writing out conversions </blockquote></blockquote></blockquote></blockquote></blockquote> by hand, <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> * Force full loop unrolling for better performance </blockquote></blockquote></blockquote> This file seems to anger gcc a lot. It seems to take upwards of a minute or two to compile here. gcc 4.8.3 on 32-bit x86. Dave. </blockquote> For me (on our poor little Duron 1800/2 GB) it ran ~5 </blockquote> minutes... <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> gcc 4.8.1 on 32-bit x86. </blockquote> If we'd like, the way the macros are set up, it would be easy to change it so that we do less unrolling in the cases where we are actually doing substantial format conversion and wouldn't notice the extra logic quite as much. I'll play with it a bit tomorrow or next week and see how how much of a hit we would actually take if we unrolled a little less in places. --Jason Ekstrand </blockquote> Ping. In a second it took 11+ minutes , here... </blockquote> 11 minutes! What system are you running? and are you using -03 or something? Yes, we can do something to cut it down, but it will probably require a configure flag; the question is what flag. --Jason </blockquote> </div></div> See above, the old children's system... ;-) -O2 -m32 -march=athlon-mp -mtune=athlon-mp -m3dnow -msse -mmmx -mfpmath=sse,387 -pipe Bad? - Worked for ages on AthlonMP....8-) Maybe it is bad on Duron (the MP thing, much smaller cache and better GCC), now. Dieter </blockquote></div> </div><div class="gmail_extra">Yeah, my recommendation would be hacking the macros to not unroll and keep the patch locally. If you've got a better idea as to how to organize the code so the compiler likes it, I'm open as long as we don't loose performance. </div><div class="gmail_extra">--Jason </div></div>