**** BEGIN LOGGING AT Wed Aug 14 02:59:56 2019
Aug 14 04:45:13 <nexgen> Hello
Aug 14 04:45:59 <nexgen> Please let me know, how to flash uboot to BBB
Aug 14 05:15:24 <nexgen> thanks
Aug 14 16:06:46 <Abhishek_1> pratimugale : It looks like there are a bunch of options to configure the kernel build at compile time, and basically you have to use the serial console as well
Aug 14 16:06:58 <Abhishek_1> I.e. connect a USB-UART converter
Aug 14 16:07:19 <Abhishek_1> To the 6-pin connector of the BBB
Aug 14 16:13:43 <zeekhuge[m]> pratimugale: can yoy describe what exactly is the problem with packaging?
Aug 14 16:15:52 <pratimugale[m]> Abhishek_:  I don't have a USB-UART converter at the moment, I'll try to get one asap. Currently I am working on the documentation. Also I haven't understood how to go about the debugging part itself i.e. how the converter will be used.
Aug 14 16:16:56 <pratimugale[m]> <zeekhuge[m] "pratimugale: can yoy describe wh"> zeekhuge: Will it be better for developers to install the project from source itself?
Aug 14 16:24:58 <pratimugale[m]> because there are not many binaries that are to be installed.
Aug 14 16:26:44 <pratimugale[m]> Many guides on the net on packaging start with "get the upstream tarball", due to which I'm a little confused on how to edit the files in the debian/ directory
Aug 14 16:30:23 <vaishnav98_[m]> Hi all
Aug 14 16:31:49 <embden_new[m]> hi
Aug 14 16:32:51 <pratimugale[m]> hi all
Aug 14 16:34:24 <cwicks[m]> Hello students and mentors
Aug 14 16:35:44 <embden_new[m]> hello cwicks
Aug 14 16:36:16 <cwicks[m]> Thanks for the roll call students
Aug 14 16:36:53 <cwicks[m]> Mentors. Please chime in. Who’s here?
Aug 14 16:38:13 <cwicks[m]> @pratimugale  can you go first today please
Aug 14 16:38:27 <embden_new[m]> I can call mine: sstabellini  drhunter  julieng  - we are starting
Aug 14 16:38:55 <pratimugale[m]> cwicks: sure
Aug 14 16:39:13 <drhunter[m]> Here
Aug 14 16:39:18 <cwicks[m]> Thanks embden_new. For some reason I can’t get this riot app to tag people today.
Aug 14 16:40:57 <pratimugale[m]> This week I tested the API on other debian images (4.14.71 only) and solved the problem I was having that the pruss_api driver wasn't being probed on its own (after starting the rpmsg firmware). Currently I'm doing the documentation and developing the wiki page: https://github.com/pratimugale/PRUSS-Bindings/wiki. On the side I have learnt most of the debian packaging; I don't exactly understand how to edit the
Aug 14 16:40:58 <pratimugale[m]> files in the debian/ directory after using dh_make.
Aug 14 16:41:03 <vaishnav98_[m]> jkridner ravikp7 : Hi
Aug 14 16:41:58 <pratimugale[m]> cwicks zeekhuge Abhishek_ ^
Aug 14 16:43:19 <cwicks[m]> Welcome Hendersa
Aug 14 16:43:31 * hendersa waves.
Aug 14 16:45:09 <cwicks[m]> Pratim whose helping you with your blocker  in the Debian question
Aug 14 16:45:31 <cwicks[m]> Mentors any questions on Pratim’s update?
Aug 14 16:46:10 <pratimugale[m]> cwicks: I am discussing it with zeekhuge
Aug 14 16:46:26 <cwicks[m]> Embden you are next.
Aug 14 16:46:54 <embden_new[m]> This week I was trying to figure out what to do till/how to prepare for the final evaluation. So, basically, we want to tidy up the code and test the work of the crossbar.
Aug 14 16:46:55 <embden_new[m]> Right now, I am testing the ethernet controller - bb-x15 can't ping itself from dom0 for now. The blocker is mainly the fact that it is not clear how to test the work of the crossbar and to what extent to change the code, since, the proper implementation of the crossbar support requires a very good understanding of xen (because leads to architectural changes (adding secondary interrupt controllers support in
Aug 14 16:46:56 <embden_new[m]> general)). So, I, basically, during the next week want to make code look better and more reasonable.
Aug 14 16:48:30 <drhunter[m]> embden. It may be simpler to test something like a general purpose timer running in user space on dom0. This will need to be routed via crossbar and avoid further complexity in ethernet driver
Aug 14 16:49:39 <embden_new[m]> drhunter:  ok, but could you tell me how to test it in few words?
Aug 14 16:50:08 <embden_new[m]> I can see that date output changes for example
Aug 14 16:50:26 <embden_new[m]> I don't have c library compiled for the platform
Aug 14 16:50:57 <drhunter[m]> at the moment, no :)  It is an idea for something simpler that will take an interrupt from a peripheral which will be routed via crossbar to the xen distributor.
Aug 14 16:51:53 <embden_new[m]> Also, I can see Linux output about eth0 going up and down
Aug 14 16:52:21 <drhunter[m]> If you are getting something with eth0 then we are at least alive.
Aug 14 16:52:52 <drhunter[m]> So maybe that will be sufficient. I didn't want you to have to debug a complex driver.
Aug 14 16:53:30 <drhunter[m]> Have you checked if there is an interrupt involved in this process? Just in case it is just polling a register.
Aug 14 16:54:30 <embden_new[m]> drhunter:  not yet
Aug 14 16:55:40 <drhunter[m]> I don't think we have any other questions at the moment. Do you agree Denis?
Aug 14 16:56:10 <jkridner[m]> sorry i didn’t make the meeting time.
Aug 14 16:56:53 <embden_new[m]> drhunter:  it would be interesting to know more about the upcoming evaluation
Aug 14 16:57:34 <embden_new[m]> drhunter:  at the same time when I ping from one port to another I can see that led are synchoriously  blinking .
Aug 14 16:57:56 <embden_new[m]> so, probably, interrupt might be stuck somewhere
Aug 14 16:58:04 <vaishnav98_[m]> <jkridner[m] "sorry i didn’t make the meeting "> jkridner Hi
Aug 14 16:58:46 <julieng> embden_new[m]: A lot of network drivers are polling because it is more performant. I agree with drhunter[m] that the timer is probably a best solution.
Aug 14 16:59:12 <julieng> Alternatively, you can count the number of interrupts received in Xen and print them when pressing a key.
Aug 14 16:59:22 <julieng> Such as the one dump the domain state
Aug 14 16:59:35 <cwicks[m]> Embden. It may be helpful to take a look at your original scope and milestones and communicate with your mentors regarding the flyer status maybe with email to align and agree on current status and expected completion. So you have clear agree on what is “finished”
Aug 14 17:00:59 <cwicks[m]> Typo. Not flyer. Project.
Aug 14 17:02:14 <cwicks[m]> I don’t want to slow down the communication with embden. So y’all keep up your discussion.
Aug 14 17:02:47 <cwicks[m]> Vaishnav. Can you give your update.
Aug 14 17:03:08 <embden_new[m]> julieng: how to test the timer?
Aug 14 17:03:23 <vaishnav98_[m]> jkridner cwicks ravikp7: this week I was able to write some tests for verifying the greybus peripherals created when a manifest is loaded and provided documentation for running them, I have also tried to bring in support for waveform click by updating to the latest driver sources but still no new iio device is being created even though the driver probe happens without any error messages. Currently I am trying out
Aug 14 17:03:24 <vaishnav98_[m]> gb-netlink as suggested by cfriedt.
Aug 14 17:03:25 <vaishnav98_[m]> Also I would like to hear jkridner 's and ravikp7 's views on what all tasks to currently focus on and how to prepare the project for final evaluation.
Aug 14 17:03:52 <drhunter[m]> embden, lets discuss after the meeting.
Aug 14 17:04:38 <embden_new[m]> drhunter:  ok
Aug 14 17:07:26 <jkridner[m]> the insclick/rmclick experience for the various modules and a reasonable set of setup instructions are critical to me.
Aug 14 17:08:33 <jkridner[m]> as many of the clicks from the list as possible.
Aug 14 17:09:03 <zeekhuge[m]> pratimugale : what edits are you talking about? I dont see instructions to any edit in the readme
Aug 14 17:09:44 <vaishnav98_[m]> jkridner: for clicks with platform data how should insclick instantiate the device ? Currently I am making use of the Mikrobus driver for doing it , is it okay?
Aug 14 17:11:07 <vaishnav98_[m]> <jkridner[m] "as many of the clicks from the l"> jkridner: currently support for 5 out of the 8 click I have received is added, will try my level best to add support for the remaining clicks
Aug 14 17:11:32 <jkridner[m]> if it is needed, just think about the per-click work required. that’s the thing to document.
Aug 14 17:12:34 <jkridner[m]> maybe document the out-of-tree build process as well as integration, at least into the standard beagle images.
Aug 14 17:13:02 * jkridner[m] wonders how big the click support is to add to the images.
Aug 14 17:15:05 <vaishnav98_[m]> <jkridner[m] "maybe document the out-of-tree b"> jkridner: okay I'll do that , jkridner will greybus drivers be enabled in future builds? currently I'm testing 4.14.108 and later versions.
Aug 14 17:15:13 <pratimugale[m]> <jkridner[m] "and upstreaming into distros eve"> zeekhuge: I was talking about editing the file/templates like 'rules', 'preinst.ex', 'postinst.ex' made after running dh_make
Aug 14 17:16:00 <pratimugale[m]> I replied to a previous message by mistake
Aug 14 17:17:03 <zeekhuge[m]> pratimugale: Okay. So whats the problem with packaging?
Aug 14 17:20:52 <jkridner[m]> would be great, but secondary.
Aug 14 17:20:56 <pratimugale[m]> zeekhuge: For example, I don't understand how to write instructions to install prussd.py while doing the packaging
Aug 14 17:21:22 <pratimugale[m]> where to write*
Aug 14 17:22:00 <zeekhuge[m]> Just add a recipe "install" In the makefile which will do all the setup. And then use the checkinstall utility.
Aug 14 17:23:57 <vaishnav98_[m]> <jkridner[m] "would be great, but secondary. "> jkridner : sure :) , also currently gbsim supports only one click device per interface (SPI/I2C) , so two SPI based clicks won't work at the same time , how can I deal with this issue
Aug 14 17:25:02 <pratimugale[m]> zeekhuge: This Makefile would contain the same instructions as in install.sh right ->https://github.com/pratimugale/PRUSS-Bindings/blob/a745477870845164217842f1db682bf6df5d784c/install.sh#L30?
Aug 14 17:25:39 <zeekhuge[m]> If install.sh moves all the files at their appropriate positions, then yes.
Aug 14 17:26:17 <zeekhuge[m]> It would result in a deb file which then can be used with dpkg to install it.
Aug 14 17:28:45 <pratimugale[m]> zeekhuge: Okay, I'm doing it right now and will let you know the results
Aug 14 17:28:47 <jkridner[m]> any needs on default image integration for any students?
Aug 14 17:30:23 <cwicks[m]> I’m sorry I have to leave a few minutes early.
Aug 14 17:32:25 <jkridner[m]> cwicks: thanks for you continued support!!!
Aug 14 17:32:39 * jkridner[m] waves bye to cwicks 
Aug 14 17:33:39 <jkridner[m]> vaishnav98_: cfriedt wanted to know if you could add the property suggestions he had in mind to manifesto.
Aug 14 17:33:43 <vaishnav98_[m]> <vaishnav98_[m] "jkridner : sure :) , also curren"> jkridner: can you provide some feedback on how deal with this issue
Aug 14 17:34:33 <jkridner[m]> oh, yikes. that's a tough one. I think the real solution is a gbsim instance per mikrobus port. :-(
Aug 14 17:34:51 <jkridner[m]> each one start up managing just its own port.
Aug 14 17:37:04 <vaishnav98_[m]> <jkridner[m] "vaishnav98_: cfriedt wanted to k"> jkridner: I think I can add it to manifests and get it to the point where the data is received at the greybus drivers, but I'm confused at how this data will then be passed to the subsequent click device drivers
Aug 14 17:37:51 <vaishnav98_[m]> <jkridner[m] "each one start up managing just "> jkridner: thanks for the suggestion, I'll try that and update you
Aug 14 18:04:48 <embden_new[m]> drhunter:  so, what about timer testing?
Aug 14 18:07:18 <drhunter[m]> <embden_new[m] "drhunter:  so, what about timer "> Not had time to look at how Linux can access GP timers yet.
Aug 14 18:07:55 <embden_new[m]> ok, I thought you knew already
Aug 14 18:17:52 <embden_new[m]> https://pastebin.com/4dvUYuNJ
Aug 14 18:17:52 <embden_new[m]> https://pastebin.com/NC6DWeRr
Aug 14 18:17:53 <embden_new[m]> might be useful drhunter julieng
Aug 14 19:18:18 <embden_new[m]> drhunter:  I also thought about writing into gpio
Aug 14 19:21:33 <embden_new[m]> and thought I have no c library, I think I can write a kernel module then drhunter  julieng ?
Aug 14 19:46:21 <embden_new[m]> drhunter:  ds2  julieng  hm, an interesting error occurred on its own https://pastebin.com/01ijdmPh
Aug 14 19:47:20 <embden_new[m]> and the console is still alive and works
**** ENDING LOGGING AT Thu Aug 15 02:59:58 2019