**** BEGIN LOGGING AT Mon Apr 01 02:59:57 2019 Apr 01 13:46:33 Students: How are the proposals coming along? Apr 01 14:20:41 working on the kernel device tree as it require time for beggineer part and adding images and use cases in the proposal Apr 01 15:39:16 http://www.ti.com/lit/ds/symlink/cd74hc299.pdf i had found the another shift register with same specification will it be right for the project Apr 01 15:40:46 as the ic that was i was not able to find it listed was not availble Apr 01 15:41:08 It looks fine. All of the universal shift registers are going to have similar characteristics. Apr 01 15:43:55 can you guide me through the kernel driver development as i followed the derek molloy guides build some them myself studied about device tree from https://bootlin.com/doc/training/linux-kernel/linux-kernel-slides.pdf and their lecture on youtube now i got little understatning of it but the whole process is still not so clear to me. Apr 01 15:45:24 For the kernel module, we'll probably have you work with an existing project (like BeagleLogic) and use its driver and PRU interfacing as a template for your own work. Apr 01 15:45:42 The mentors are there to assist in getting you set up to do that sort of development. Apr 01 15:46:17 The device tree is only a small part of the overall puzzle. Apr 01 15:47:00 should i have to to develop pcb circuit for the project also as i know circuit desiging as well Apr 01 15:47:19 sry pcb board design Apr 01 15:53:25 You are welcome to, but the software is the more important aspect of the system. A breadboard design with jumper wires is adequate for GSoC. The time required for a PCB layout to be manufactured makes it difficult to fit it into the GSoC time period without the PCB fabrication becoming expensive. Apr 01 15:54:00 If you do a PCB layout, it should be with a tool that people can download freely and use for small designs, like Eagle PCB. Apr 01 15:54:25 A tool that can generate Gerber files for fabrication. Apr 01 15:55:09 That way, people can download your PCB design and enhance it for whatever purpose they need the bus for. Most people will just be using a breadboard to prototype interfacing projects, though. Apr 01 15:55:56 The PCB layout will also have to be available as open source hardware. Apr 01 15:56:16 thanks for the suggestion .so ,then will it be fine to setup a game controller like project for testing the things out at the end of the weeks in the gsoc project Apr 01 15:57:36 Sure, that should be fine. Apr 01 15:57:55 https://www.oshwa.org/sharing-best-practices/ Apr 01 15:58:07 Follow those guidelines for your design. Apr 01 16:00:56 I would suggest a project that requires a 16-bit or 32-bit bus of some sort. Maybe an even wider bus, if you can find a good project. Apr 01 16:01:17 Like interfacing with a memory chip of some sort. Apr 01 16:01:46 cool i will look into it Apr 01 16:04:47 That will also show off the ability to chain shift registers together. Apr 01 16:05:28 i was thinking to develop a library that can able to do communication at different baud rate with other microcontroller or modules is this idea seems fesible to you as the PRU have quite good execution time Apr 01 16:07:51 Sure, that would be very useful. You'll have to think about how you'd do that in the PRU firmware. Probably a nop loop of some sort to insert wait states to slow things down to different baud rates. Apr 01 16:09:53 as the shift register ic is and 4 bit bi directional logic level conveter or 8 bit is not available in fritzing so i am not able to generate circuit diagram for now i think i should explore other options what do you suggest for now. Apr 01 16:13:18 AdaFruit's Fritzing part libraries should have some bidirectional line-level converters you can use. Apr 01 16:13:48 They're pretty good about adding Fritzing parts for the items they have in stock. Apr 01 16:14:08 https://blog.adafruit.com/2012/08/27/how-to-install-and-use-the-adafruit-fritzing-library/ Apr 01 16:15:26 i will look into it ☺ Apr 01 16:15:49 with communication enable with the shift register can serve a good purpose for this project Apr 01 16:19:14 https://elinux.org/File:Pocket_gpio.jpg i have shown exposed gpio that can be used easily i/o is this sheems fine to you Apr 01 16:22:13 Yes, I looked at that. I made comments on your proposal, too. Apr 01 16:22:39 ok thanks for it Apr 01 16:23:04 You should estimate how many GPIOs that you'll need and pick a few for your proposal. You can limit yourself to GPIOs with a PRU mode, or you can try to go with regular GPIOs. Apr 01 16:23:30 The PRU firmware will be much easier if you stick with GPIOs that have PRU modes. Apr 01 16:23:41 But, you trade off flexibility if you do that. Apr 01 16:24:48 Most of the available GPIOs on the PocketBeagle have a PRU muxing mode on them, so it should be too difficult. Apr 01 16:25:30 If you were targeting the BeagleBone Black, for example, you might have to mux out some other feature to turn on the PRU pins. Apr 01 16:25:31 i was thinking of using only the 5 gpio that are left unmuxed and can be used as i/o easily that is pin 2,4,6,8,10 on pr2 Apr 01 16:26:00 * hendersa has to get to a meeting at work. Apr 01 16:26:17 Look over your pins, come up with a muxing plan, and add it into your proposal. Apr 01 16:26:24 We can look it over more later. Apr 01 16:26:47 ok thanks for the suggestion that you have givien it help me a lot **** ENDING LOGGING AT Tue Apr 02 02:59:57 2019