**** BEGIN LOGGING AT Fri Apr 10 02:59:59 2015 Apr 10 11:47:05 alexanderhiam: ping Apr 10 13:25:10 Abhishek_: ping. Apr 10 13:25:46 Abhishek_: sorry. final year project was keeping me extremely busy :( Apr 10 13:48:31 Hi shubhangi Apr 10 13:52:17 for the machinekit usecase you wrote "For creating a sysfs interface, you would need a kernel driver. So do you mean to say that you wish to move the Machinekit implementation as a kernel-mode thing?" . the driver proposed for gsoc would Apr 10 13:52:29 suffice for it Apr 10 13:53:10 and i know a lot more need to be done on the usecase. a codesketch ? Apr 10 13:54:36 what i thought of was just libprussdrv in machinekit being replaced by the (new kernel driver+userspace lib) Apr 10 13:56:00 and since the userspace lib (a wrapper around sysfs provided by the driver) would have similar functions of libprussdrv, porting would be a little easier. Apr 10 13:56:31 Abhishek_: Apr 10 14:02:52 shubhangi: Are you able to understand how presently the comm between Machinekit and the PRUs take place? Apr 10 14:06:29 not fully. machinekit lacks documentation. have to understand a lot of code. was not able to through all of it as of now. Apr 10 14:12:55 I mean, the role of libprussdrv there Apr 10 15:04:35 geekswine: pong Apr 10 15:38:47 alexanderhiam: Can you please tell me what I should do during the current period. i.e. until 27 April Apr 10 15:44:54 kiran4399: mainly just hang around here and be ready to answer any questions that may come up Apr 10 15:48:51 alexanderhiam: OK... Apr 10 15:56:15 vvu I will check your comments. regarding your question, it is msc. Apr 10 16:00:52 alexanderhiam: i have wrote library for initializing ADC..can u have a look at it.. Apr 10 16:00:57 https://github.com/GeeksWine/BeagleArduino/blob/master/adc%20libraries/adc.c Apr 10 16:01:23 i have not enabled all step configuration registers... Apr 10 16:01:30 like delay and all !!! Apr 10 16:01:34 Abhishek_ : as far as i can understand, its role is to load the firmware, initial pru configs and then provides dram address to hal components ( https://github.com/machinekit/machinekit/blob/cb40ab0da23c59f580d0412534f54d2c0f372bc9/src/hal/components/hal_pru.c#L322-L330 ) . task are being written to (and current status read from) by directly accessing the Apr 10 16:01:34 memory addresses ( for ex https://github.com/machinekit/machinekit/blob/d90135a6f4117796de0c7b6ca3c02fa08322f4a7/src/hal/drivers/hal_pru_generic/stepgen.c#L639-L702 ) Apr 10 16:02:13 got busy with the project and all :( ending of the sem :( Apr 10 16:02:42 geekswine: that's a good start, but it's the step config stuff that I was really looking for. As it is nothing's going to be in the fifo because the sequencer isn't setup Apr 10 16:04:45 alexanderhiam: i was also pretty confuse that what do we need for what :/ Apr 10 16:04:56 like i understood for avg sampling and all.. Apr 10 16:06:08 shubhangi: so how will the same applications be done once the new system in place? Apr 10 16:16:58 alexanderhiam: did u see that ?? like step conf has around 16 reg..so confused which to use for this... Apr 10 16:19:56 any advice and comments pls ?? Apr 10 16:24:24 Abhishek_: hmm ..went through more code ... currently all communication is lots of memory handling operations .. read/write/updating pointers to new locations etc... independent of libprussdrv . The new framework can streamline this ... have to figureout exaclty how things will fall into place Apr 10 16:25:07 geekswine: that's why I asked for the ADC code, it's a lot more complicated than an ATmega ;) Apr 10 16:25:53 the sequencer is to allow you to do a whole bunch of different sampling techniques, it's up to you to decide what makes the most sense Apr 10 16:26:56 e.g. you could set up each step to sample a different channel, or you could set up just one step to sample the channel requested with analogRead() Apr 10 16:30:48 geekswine: PyBBIO used to confugre the ADC directly through the registers long ago... Apr 10 16:31:12 https://github.com/graycatlabs/PyBBIO/blob/8ad4b89c1665421ed9fd790d3c1d335d13a11139/bbio.py#L91-L111 and https://github.com/graycatlabs/PyBBIO/blob/8ad4b89c1665421ed9fd790d3c1d335d13a11139/bbio.py#L222-L236 Apr 10 16:31:31 I'm not saying that's the way to do it, just an example of one way Apr 10 16:32:10 * alexanderhiam is slightly embarrassed digging up that old messy code Apr 10 16:33:30 vvu I have replied to your comment and updated the proposal. Thanks! Apr 10 16:34:43 alexanderhiam: thanx for that..needed some kind of reference :D it will be very helpful !!! Apr 10 16:35:23 geekswine: you can also take a look at the ADC kernel driver Apr 10 16:36:38 https://github.com/beagleboard/linux/blob/3.14/drivers/iio/adc/ti_am335x_adc.c Apr 10 16:37:03 alexanderhiam: okk..thanks :) and hope i m not doing something wrong which may disappoint u :P Apr 10 16:38:53 geekswine: no way, as long as you're digging in and trying to figure it out I'm happy Apr 10 16:39:18 :D Apr 10 16:39:32 i wonder that i could explore this a month back :P lots of interesting stuffs :D Apr 10 17:01:25 alexanderhiam: should i post that github link in the comments section where u asked ?? Apr 10 17:53:59 arianepaola: thanx, when i reach home i will give you some links for thr codeless kext Apr 10 20:30:51 hmmmmm Apr 10 21:18:10 "Things that make you go hmmmm..." Apr 11 00:35:31 ds2: there ? Apr 11 00:37:48 * nerdboy quoting an AF Capt **** ENDING LOGGING AT Sat Apr 11 02:59:58 2015