**** BEGIN LOGGING AT Wed Feb 14 03:00:02 2018 Feb 14 03:00:12 I got to see my BBB Debian Distro on the server. Cool! Feb 14 03:00:20 I knew it could be done again and again. Feb 14 03:08:31 Otay...it works w/ the older version and the newer version. Wish I could have told more. O' well. Feb 14 03:11:39 Nothing like some servers and GUIs from the BBB! Feb 14 09:37:20 zmatt: does the 4.14 bone kernel have uio support Feb 14 09:42:58 zmatt: if so, how can i enable it? Feb 14 10:06:35 (yes, and simply by enabling it same as on 4.9-bone) Feb 14 10:07:11 I really don't understand why the beaglebone blue dts on -bone kernels doesn't simply enable pruss Feb 14 14:23:03 Hello, I have been trying to install cross compiler on my Ubuntu as mentioned in the this guide - https://sites.google.com/site/beagleboneblac/home/cross-compiling-linux-kernel-module . After having downloaded the u-boot files, the 'make tools' command is failing with an error of this sort : *** Configuration file ".config" not found! *** Please run some configurator (e.g. "make oldconfig" or *** "make menuconfig" or "make xconfi Feb 14 14:25:31 [contd] it gives me a bunch of options to set the specifications of the ARM device which I believe requires knowledge of how ARM works but I am a beginner so I can't follow. Kindly, help me with the setup of the cross compiler tools on my machine running Ubuntu 17.10 Feb 14 14:58:42 Hello. This is my first time in GSOC. I am doing a major in Electronics and specializing on VLSI Design and Embedded Systems. I feel BBB is where I can really contribute and the ideas list is also very interesting. I haven't written much codes for ARM apart from tinkering with a RPi3 for sometime. Can I get a pointer to where should I start ARM development in a Ubuntu Machine? Thanks in advance.:) Feb 14 15:04:41 zmatt: hey using your uio library, how can I load programs onto the pru Feb 14 15:11:23 zmatt: sorry, i meant, data not program,s Feb 14 15:11:53 like if I wanted to write the pwm duty cycle into my pru code which command should I use? Feb 14 15:12:10 parichay: actually, the first thing would be to fix your irc conenction. you're always leaving way to early :) Feb 14 15:14:04 parichay: and actually compiling for arm is not really different from compiling for x86, given the assumption that you are staying in userland Feb 14 15:16:26 LetoThe2nd: Hey. I am sorry, I didn't know how irc works but now I learnt some of the basics. I hope there won't be issues like that again. Feb 14 15:16:54 parichay: its just that if you leave after 5 minutes you probably are never around to get an answer :) Feb 14 15:18:31 LetoThe2nd: okay. Actually I am unable to install the cross compiler tools on my system, that is my problem. Feb 14 15:20:03 I started with this https://github.com/jadonk/gsoc-application/tree/master/ExampleEntryJasonKridner Feb 14 15:20:23 https://sites.google.com/site/beagleboneblac/home/cross-compiling-linux-kernel-module - here I am unable to the step 'make tools' Feb 14 15:21:41 parichay: yeah i saw your question. actually the step that you are referring to in the latter is specifically for kernel-crosscompilation. Feb 14 15:22:18 parichay: plus, that link refers to some rather unfortunate choice of arago toolchain Feb 14 15:23:16 parichay: i mean, ubuntu has a neat arm cross toolchain in the repos. Feb 14 15:24:13 https://launchpad.net/gcc-arm-embedded/+download I found this one, should i go for this one? Feb 14 15:24:36 parichay: no, that is for bare metal stuff Feb 14 15:24:55 just try an apt-cache search gnueabi :-) Feb 14 15:26:14 There are bunch of compilers for C, i don't know which one to choose Feb 14 15:26:15 that should happily offer you gcc-arm-linux-gnueabi Feb 14 15:26:23 which is kinda the default Feb 14 15:27:05 okay thanks Feb 14 16:10:33 Hello all, I have two question that have been nagging me quite a bit: Feb 14 16:11:05 1) Can I use the uio_pruss interface on my BeagleBone Blue (running 4.9-ti-rt) instead of remoteproc? Feb 14 16:11:36 2) If I have to use remoteproc, can someone please give a brief idea of how to use remoteproc to load PRU firmware? Feb 14 16:19:46 I included #include "am33xx-pruss-uio.dtsi" in the boneblue.dts Feb 14 16:19:53 however it doesn't seem to work Feb 14 16:20:44 (because the file isn't there) Feb 14 16:35:14 LetoThe2nd: gcc-arm-linux-gnueabihf Feb 14 16:40:46 Hello all! I am running kernel 4.9-ti-rt on my beaglebone blue, and I wanted to know how i can enable uio_pruss??? Feb 14 16:44:28 There is no tutorial on how to enable uio_pruss Feb 14 16:44:35 and the ones that are there don't work for the blue Feb 14 16:44:49 is there anyone who knows how to sort out this mes Feb 14 16:44:51 mess?* Feb 14 16:55:10 remoteproc is just too complicated to use :p Feb 14 17:02:00 i screwed up my boot/uEnv.tct and it isn't booting up anymore, so I have to refresh the image Feb 14 17:02:03 reflash* Feb 14 17:05:06 LetoThe2nd: Voila! I was able to compile it with arm-linux-gnueabi-gcc-7 compiler and also run it on qemu-arm-static emulator. Thanks :) Feb 14 17:09:10 dlaad: yeah there are a lot of people asking for this Feb 14 17:09:37 it shouldn't actually be hard Feb 14 17:09:44 maybe I'll see if I can fix up a dtb for that Feb 14 17:15:34 zmatt: Hey thanks a lot man! Also, I noticed that the pru works on 4.9-bone but the pinmux is really screwed on that Feb 14 17:15:42 yeah that's stupid too Feb 14 17:16:29 there's also no good reason why the blue doesn't allow you to select between remoteproc-pru and uio-pruss via /boot/uEnv.txt like you can do on other beaglebones Feb 14 17:16:31 zmatt: so while testing on 4.9-bone, I tried using pypruss, but it gave a lot of segmentation faults. I guess its just really old (cuz i saw jkridner's issue) Feb 14 17:16:52 zmatt: yeah i tried chaining my uEnv and it got locked on the next boot Feb 14 17:16:58 hence i had to reflash Feb 14 17:17:24 yeah I know. you don't really need it though, equivalent functionality can be had in pure python Feb 14 17:17:49 zmatt: but i don't know nay library that implements uio_pruss in pure python?? Feb 14 17:17:53 I added a partial example of that to the "py-uio" project I once started, even though I don't really work on that anymore Feb 14 17:18:34 https://github.com/mvduin/py-uio Feb 14 17:18:45 zmatt: okay ill check it out Feb 14 17:19:18 you need to install etc/udev/rules.d/09-uio-pruss.rules to make an easy to find alias for uio Feb 14 17:19:33 zmatt: btw, if you do get around to modding the dtb (i really hope you do, cuz even rcn is not responding to my request), please let me know Feb 14 17:19:43 and then after reboot (or remove/reinsert the kernel module) pruss-test.py should work Feb 14 17:19:51 yeah I'm looking at it right now Feb 14 17:19:56 should be really easy actually Feb 14 17:22:17 sorry i got disconnected Feb 14 17:22:40 https://pastebin.com/raw/RRd4n98X Feb 14 17:23:19 zmatt: right so you're looking at modding the dtb for..? perhaps something that would allow to switch between rproc (when it gets less intimidating :p) and uio_pruss on all the latest kernels? Feb 14 17:26:27 oh i didn't miss anything when i phased out for those few :p Feb 14 17:27:47 zmatt: check out RCN's comment at the end: https://github.com/adafruit/adafruit-beaglebone-io-python/issues/232 Feb 14 17:31:59 https://github.com/mvduin/dtb-rebuilder/tree/4.9-ti Feb 14 17:32:11 I added a am335x-boneblue-uio.dts file Feb 14 17:33:09 to update this onto my blue, I have to git pull in dtb-ti-4.9 right? Feb 14 17:33:25 okay so it sounds like rcn is fixing the problem? Feb 14 17:33:57 I guess so, hopefully it doesn't go over the top off his head Feb 14 17:35:18 anyway, id like to try your file out, does it work on all 4.9 and 4.14 kernels? Feb 14 17:38:01 I have done absolutely zero testing, I don't even have a beaglebone blue Feb 14 17:38:31 but if it works on 4.9-ti, then at the very least the same trick (#including the am33xx-pruss-uio.dtsi I added) should also work on other kernels Feb 14 17:38:46 I'd expect the 4.9-ti dtb to work on 4.9-bone too Feb 14 17:39:06 well let me be the first :p, I can just create a new file with that name and add your contents in it .. ? Feb 14 17:39:23 I actually did try including that when i was modding it myself, but it wasn't able to find the file Feb 14 17:39:31 uhh, you could just pull from my git repo Feb 14 17:40:06 or just copy what I did in that commit, whatever you prefer Feb 14 17:51:36 zmatt: idk why my board is not sshing into via the usb, so i will have to disconnect fo r abit Feb 14 17:54:56 zmatt: so it threw the same error as it did when I tried it previously: fatal error: am33xx-pruss-uio.dtsi: No such file or directory Feb 14 17:58:38 uhm, I *added* that file Feb 14 17:59:45 wtf shit Feb 14 17:59:47 sorry Feb 14 18:07:19 zmatt: ill set the dtb= in uEnv.txt to this right> Feb 14 18:07:28 yes Feb 14 18:07:33 after installing the dtb Feb 14 18:12:37 it locked my board Feb 14 18:12:42 its not booting up Feb 14 18:12:55 hm Feb 14 18:13:19 I don't suppose you can check for error messages on the serial console? Feb 14 18:13:38 otherwise I guess just boot from an sd card and change your uEnv.txt back Feb 14 18:13:46 yep thats what im doing Feb 14 18:13:55 so we should wait for rcn? Feb 14 18:19:54 I just pushed a different idea Feb 14 18:20:05 maybe that variant of am33xx-pruss-uio.dtsi works Feb 14 18:24:08 well its pretty late here, ill try it ou ttomorrow Feb 14 19:30:12 question to the ti people, jkridner[m], is there an AM335x with temperature spec -40 to +85 dec C? Feb 14 19:35:50 thinkfat: not entirely sure, but https://www.ti.com/product/am3358 gives two ranges, -40 to 105 and 0 to 90 Feb 14 19:36:01 any ideas page for GSoC 18? Feb 14 19:36:46 thinkfat: though i only see an operating temperature range, so it's unclear what the two values are Feb 14 19:37:32 vagrantc: thanks a lot, I found this page here http://www.ti.com/product/AM3359/datasheet/device_comparison#SPRS7179076 Feb 14 19:38:23 vagrantc: it looks like all but the 3352 would support at least -40 to 90 by default, and I guess for the 3352 you would be able to order one that fits Feb 14 19:39:36 thinkfat: same with the 3357 Feb 14 19:40:19 vagrantc: yes indeed. so I guess it's safe to say all of them could handle the range if you order the right part Feb 14 19:41:40 this has slightly wider tolerances too http://www.ti.com/product/am3358-ep/description Feb 14 19:42:28 all that said, i have no real-world experience testing these tolerances :) Feb 14 19:43:27 iirc high temperature + high frequency does come with power-on-hours limitations, but I don't remember if that's already an issue for 85 ͏°C Feb 14 22:09:48 Im having trouble getting the uart to work using the gpio pins on a beagle bone black. I followed the instructions here: https://archlinuxarm.org/platforms/armv7/ti/beaglebone and then added the enable_uboot_overlays=1 and uboot_overlay_addr=path to uart dtbo file and for some reason only the ttyS0 is working for the console but none of the actual uart ports. Any idea what I'm missing? Feb 14 22:10:24 uhh Feb 14 22:10:44 Uboot version is 2017.09 Feb 14 22:12:15 an u-boot build from rcn or a mainline one? Feb 14 22:12:59 It's the one included with the arch Linux image Feb 14 22:13:02 and to have any chance of u-boot overlays working you'll need to be sure to also use a beaglebone.org kernel and have the overlays installed Feb 14 22:13:11 yeah so u-boot overlays will definitely not work then Feb 14 22:13:53 Is there something specific with the beagle bone.org kernel? Feb 14 22:14:11 you can probably get it working with arch's userspace I guess, but when using an unusual distro you should expect to have to put in extra effort Feb 14 22:15:54 yes, there is Feb 14 22:16:03 I have the overlay files in lib/firmware like the beaglebone image. Feb 14 22:16:06 at the very least it will have the appropriate device tree files Feb 14 22:16:53 Device tree files as in the ones in lib/firmware? Feb 14 22:17:00 no, /boot/dtbs Feb 14 22:18:11 I suppose you could try just using the dtbs from https://github.com/RobertCNelson/dtb-rebuilder Feb 14 22:18:22 pick whatever branch seems closest to the kernel you're using Feb 14 22:19:02 get an MLO + u-boot.img from https://rcn-ee.com/repos/bootloader/am335x_boneblack/ Feb 14 22:19:36 make sure all ingredients are in the same place as on a standard beaglebone image Feb 14 22:20:01 and have a serial console ready for debugging any boot issues :) Feb 14 22:23:30 I'll try that and see what happens. Thanks for the suggestions. Feb 15 00:01:33 Is there a tutorial on using the USB client port on the Beagleboard x15 through Windows to share internet connections? I'm trying to get my USB WiFi adapter working. Feb 15 00:21:56 I'm not sure I follow, turtguy. You want to share over USB in the meantime because you don't have the adapter plugged into the X15 working yet? Feb 15 00:28:29 Yes. ha Feb 15 00:28:50 I have to have internet to install the drivers Feb 15 00:29:22 can you use wired ethernet for that? Feb 15 00:29:55 Unfortunately, that isn't possible where I am located in the house. Feb 15 00:39:58 I'll be back, I'm going to restart and try some things Feb 15 01:07:39 I guess he doesn't have cat 6 100' cables like other people. Feb 15 01:29:02 this is why you install structured cabling when you move into a place Feb 15 01:29:36 point you do have there. Feb 15 01:57:23 if i send a 4 byte command via spi how can i listen for 8 bytes? Feb 15 02:00:31 spi always reads and writes simultaneously Feb 15 02:04:59 there's no way to just read data without also writing something. what you're supposed to write depends on the peripheral Feb 15 02:05:10 usually it's "don't care" Feb 15 02:07:39 but when i send 4 bytes i should get an 8byte response Feb 15 02:07:47 but it's only sending 4 Feb 15 02:08:10 spi doesn't work like that, every clock cycle a bit is transferred out on mosi and a bit is transferred in on miso Feb 15 02:08:44 you, the master, decide how much data is being transferred Feb 15 02:08:53 the spi slave device has no control over this Feb 15 02:10:32 (the kernel does allow you to perform transmit-only and receive-only transfers, but those are just optimized versions of "discard received data" and "transmit zeros") Feb 15 02:13:12 i found a commandline tool to send SPI and i am able to see it on the spi monitor on the sensor Feb 15 02:13:40 but the sensor is only reporting miso of like 4 bytes when response should be 8 Feb 15 02:25:09 if need to read 8 bytes of data, then transfer 8 bytes of data Feb 15 02:25:12 it's that simple Feb 15 02:25:33 if you need to send 4 bytes and *then* read 8 bytes of data, transfer 12 bytes of data Feb 15 02:25:54 what's the sensor? have a link to the datasheet? Feb 15 02:36:53 Bongobong: as zmatt said SPI is not an asynchronous port you control ALL data transfer so you send a 4 byte command you then have to send 8 dummy bytes to read the response Feb 15 02:42:55 yeah sorry, it is a vectornav vn-100 Feb 15 02:42:58 https://www.vectornav.com/docs/default-source/documentation/vn-100-documentation/vn-100-user-manual-(um001).pdf?sfvrsn=b49fe6b9_18 Feb 15 02:43:19 when i sent 12 bytes first 4 as the command trailing 8 as zeros i was able to see the 8 byte response Feb 15 02:48:12 wow that datasheet is really vague about the protocol Feb 15 02:48:51 unless I just overlooked something Feb 15 02:50:15 "See the Basic Communication chapter" Feb 15 02:50:27 that chapter doesn't exist. well done. Feb 15 02:56:10 lol yeah Feb 15 02:56:16 how would i take data from the spidev1.0 and echo it into hex? Feb 15 02:57:59 you said you had already found a tool to do spi transfers? Feb 15 02:59:07 it's also not hard to write one if you hadn't... there are probably libraries for it in most languages, and the low-level kernel interface isn't particularly hard to use either **** ENDING LOGGING AT Thu Feb 15 03:00:01 2018