**** BEGIN LOGGING AT Tue Jul 28 02:59:58 2015 Jul 28 12:39:22 gm everyone! Jul 28 12:46:36 vmayoral imuguruza: Hello! Jul 28 14:04:22 neemo: ... ping? Jul 28 15:08:07 neemo: FYI it's pretty essential that you push everything you have before the meeting tomorrow, and the sooner the better so nerdboy and I can look it over beforehand Jul 28 15:45:40 alexanderhiam :ping Jul 28 16:38:05 apaar__: pong Jul 28 16:42:09 alexanderhiam :finally done benchmarking but was looking at wether pru_bridge is realtime or not.To do that i took time diff in pru and passed it using the bridge itself .Is this a ok way of checking? Jul 28 16:43:10 you mean you used the PRU timer? What did you measure with it? Jul 28 16:43:24 is that code on gh? Jul 28 16:43:42 yes i used the pru timer https://github.com/Apaar/PRU-Bridge/blob/master/src/firmware/test.c#L17 Jul 28 16:45:24 you got some indentation weirdness Jul 28 16:46:25 yep i code on codeblocks which takes tab as 3 spaces and then changes on nano take tab as 6 spaces :( it is like that all over the place Jul 28 16:47:04 you're also never resetting the timer, so it'll be wrong if current is before it rolls over and last is after Jul 28 16:47:14 tab as 3 spaces?! Jul 28 16:47:29 all editors should have settings for that Jul 28 16:47:54 what are you trying to time? Jul 28 16:48:26 oh, you're just measuring jitter? Jul 28 16:48:33 there were some doubts on mem access by pru and driver at the same time introducing nan realtime nature.. Jul 28 16:48:35 yep :) Jul 28 16:49:07 it looks like there is no jitter but i have tried only 11000 bytes at a time Jul 28 16:49:28 that'll give you a measure of the jitter in the PRU, as long as you're accounting for the timer overflow Jul 28 16:50:10 also the last thing should only mess up the first value(which i ignore) right? Jul 28 16:50:14 timer overflow?/ Jul 28 16:52:27 the timer is a fixed number of bits (32 I assume?), so the maximum value it can hold is 2^32-1. By default it probably rolls back to 0 after that, i.e. it's counting up modulo 2**32 Jul 28 16:53:01 some timers can also be configured to start counting down once they reach their maximum value, which would really screw up your timing Jul 28 16:53:23 oh ok i will look into that.not encountered it till now :P Jul 28 16:55:04 one method would be to reset the accumulator register to 0 at the start of the each loop instead of saving the current value, then you just need the value at the end of the loop Jul 28 16:56:15 cool Jul 28 16:57:18 one this is done most of the basic framework/api's will be operational :) Jul 28 17:03:29 awesome Jul 28 18:51:06 DiegoTc: do you konw if jkridner going to show up these days or no, would be much better if I use these days in the his modifications. Jul 28 19:01:39 ebadawy: good question, trying to catch him also Jul 28 19:01:55 _av500_: do you know something about Jason? Jul 28 19:04:18 he's about 6 feet tall... Jul 28 19:07:46 you remind me of a very close friend to me nerdboy Jul 28 19:07:54 he is not now! Jul 28 20:12:00 why, got something against comedy? Jul 28 20:22:21 nerdboy: no, just kidding :) Jul 28 20:24:19 tried fish-slapping your friend? Jul 28 20:25:23 lol, always wanted to. but unfortunately no :( Jul 28 21:07:26 * nerdboy goes for trimslice kernel and taco tuesday **** ENDING LOGGING AT Wed Jul 29 02:59:58 2015