**** BEGIN LOGGING AT Mon Sep 17 02:59:59 2018 Sep 17 03:35:03 I think I finally have a working server. Yea! Sep 17 03:35:18 Issue number one = I did not type up the software Sep 17 03:35:26 Issue number two = I only corrected it Sep 17 03:36:01 Now... Sep 17 03:36:43 Is there a working method for the MotorCape outside of the older models of using the chips onboard on this Cape? Sep 17 03:37:52 I know from what ________ told me that getting config-pin is a way to use it. Sep 17 03:37:54 But... Sep 17 03:37:57 I would like a new method. Sep 17 03:38:08 Can you point me in the correct direction? Sep 17 03:40:53 gpio.h is a file on the main Linux page on GitHub.com. But...there are supposedly new way/methods to "skinning a cat." Sep 17 03:40:58 So, help me help you. Sep 17 03:40:59 ... Sep 17 03:41:23 <<< crying to help! Sep 17 03:42:05 I guess I am looking at this incorrectly. I should just figure it out. I understand now. Sep 17 03:42:07 yea! Sep 17 03:44:08 https://training.ti.com/sitara-arm-processors-boot-camp-introduction-sitara-linux-sdk?cu=14788 is a start I guess. Sep 17 03:44:18 Off to it! Sep 17 03:54:55 "Instead of using a method that works, I want to do it some other way that I think exists, even though I don't know what" Sep 17 03:57:53 using config-pin is an alternative to using an overlay, which I suggested since the overlay inexplicably configures the motor direction controls to be leds rather than gpios as you'd expect, hence is incompatible with most libraries to control gpios Sep 17 03:58:39 (and using config-pin is easier than fixing the overlay) Sep 17 04:04:20 I know you are right but it should be a learning experience for me. Sep 17 04:04:30 Let me learn how-to. Sep 17 04:04:44 how to what? I have no idea what goal you are trying to achieve Sep 17 04:05:00 So, change the overlay and then update u-boot, right? Sep 17 04:05:16 there is absolutely no reason to mess with u-boot Sep 17 04:05:20 Oh. Sep 17 04:05:44 So, what is the deal w/ this overlay? Can I change it and not update the EEPROM? Sep 17 04:06:02 I have literally *just explained* what the deal is with this overlay Sep 17 04:06:18 Right. It is more difficult to change it. Sep 17 04:06:22 I understand this fact. Sep 17 04:06:49 But...why not use difficulties for fun? I would like to learn about how to change this overlay. Sep 17 04:07:26 and the cape eeprom merely contains information used by u-boot to know the name of the overlay file to load Sep 17 04:07:38 the overlay itself is simply loaded from /lib/firmware Sep 17 04:07:46 Oh. Sep 17 04:08:01 I figured we had to update everything since u-boot overlays is now the way. Sep 17 04:08:30 So, I just change the overlay to suit my needs in /lib/firmware and that is it? Sep 17 04:08:32 and why not? because you typically already have trouble enough simply getting things to work at all. if you're going to try making a working overlay and inevitably find it doesn't work, you'll have no idea whether the problem is with your overlay or with your test software Sep 17 04:09:05 Hey. Sep 17 04:09:09 I know. Sep 17 04:09:35 so my suggestion would be to focus on first getting the cape to work, and if you really want to you could try meddling with overlays after that Sep 17 04:09:53 Oh. I have no doubt that this Cape works. Sep 17 04:10:14 I am not just trying to get the Cape to manage my motors, i.e. even though ultimately it will. Sep 17 04:10:16 I've never expressed any doubt about that Sep 17 04:10:52 with "getting the cape to work" I meant "make software that successfully uses the cape" Sep 17 04:10:58 Oh. Sep 17 04:11:01 I understand. Sep 17 04:11:04 Now. Sep 17 04:11:10 Sheesh. Sep 17 04:11:21 You think I could not use a library to make things go. Sep 17 04:11:29 Python! Sep 17 04:12:08 Hey zmatt: Did you ever see the Chip maker on this Cape? Sep 17 04:12:24 I forgot to take a photo. Sep 17 04:12:39 no. why does it matter? Sep 17 04:12:56 It matters b/c each chip interacts differently w/ the end use. Sep 17 04:13:20 For example: some chips have to have specific software to make them work. Sep 17 04:13:25 Like this, see. Sep 17 04:14:46 I found out that the L298 needs to have a pin driven low before it can be driven high. And when the driven high pin is working, the motor can then be told which way to turn. Sep 17 04:15:27 So... Sep 17 04:15:45 the motor cape was designed to have, per motor output, one direction-pin and one enable-pin. the enable-pin is typically pwm-controlled to be able to vary the motor speed Sep 17 04:16:09 I usually work w/ geared motors. Sep 17 04:16:40 Okay. Sep 17 04:17:06 if you don't care about controlling motor speed then you can just control the enable-pin as gpio instead. in that case you don't need to use config-pin at all, just removing the overlay suffices Sep 17 04:17:52 So, one motor can go forward at a speed of 50% (PWM) and I can change that forward motion by making my pin the opposite of what it was when it was moving forward. Sep 17 04:17:53 ... Sep 17 04:17:55 That works. Sep 17 04:18:32 I looked in my uEnv.txt file and all I had was u-boot overlays enabled w/ capeuniversal. Sep 17 04:19:14 ... okay, and? Sep 17 04:19:28 I thought I would have had that specific .dts file in u-boot overlays. Sep 17 04:19:42 The one you showed me from whenever. Sep 17 04:20:05 https://github.com/RobertCNelson/bb.org-overlays/blob/master/src/arm/BBORG_MOTOR-00A2.dts is the file. Sep 17 04:20:14 like I've already said _multiple times_, the overlay for a cape is loaded automatically by u-boot Sep 17 04:20:21 I know. Sep 17 04:20:36 then why do you still ask this question? Sep 17 04:20:39 B/c of what you are saying, the overlay is in /lib/firmware. Sep 17 04:21:06 yes, /lib/firmware/BBORG_MOTOR-00A2.dtbo Sep 17 04:21:34 But...I just cannot understand why making an attempt to adjust the overlay to fit my needs on this Cape is not more exciting. Sep 17 04:21:35 ... Sep 17 04:21:48 B/c it is. Making overlays is more exciting no matter what you think. Sep 17 04:22:35 zmatt: how did you find out about overlays in the first place? School or occupational learning? Sep 17 04:22:50 no Sep 17 04:22:59 Fine. Sep 17 04:23:03 Not my business. Sep 17 04:23:04 I know. Sep 17 04:23:16 I just look at stuff and figure it out Sep 17 04:23:20 Oh. Sep 17 04:24:14 zmatt: Can you please direct me to writing up overlays? Sep 17 04:24:31 and I already explained why I think it's a mistake for you to try to make an overlay at this point instead of first getting the cape to work Sep 17 04:24:47 and I'm done repeating myself to today. afk. Sep 17 04:24:51 *for today Sep 17 04:24:51 Fine. I will get the Cape to work first. How is that? Sep 17 04:25:13 to today works too. Sep 17 04:25:46 Okay, okay. I will erase the lib/firmware/ad;fasdkasd;fsd.dts and set off on my journey. Sep 17 05:55:42 hello, can some one help me getting internet acess on the pocket beagle. I read the pages online, however none have worked so far. I am using ubuntu 18.04 Sep 17 05:56:48 grepart: so you want to share internet access from your ubuntu system to the usb-connected pocketbeagle? Sep 17 06:01:44 grepart: hello? Sep 17 06:22:19 Hi Sep 17 06:22:37 @zmatt Yes Sep 17 06:23:16 hello, can some one help me getting internet acess on the pocket beagle. I read the pages online, however none have worked so far. I am using ubuntu 18.04 Sep 17 06:23:42 I want to share internet access from your ubuntu system to the usb-connected pocketbeagle? Sep 17 06:24:51 I don't use gnome networkmanager myself, but from what I understand: make sure the dnsmasq-base package is installed, use nm-connection-editor to configure the usb interface to share the internet connection, make sure it is configured to use the ip 192.168.7.1 instead of 10.42.0.1 Sep 17 06:25:07 I did that Sep 17 06:25:37 then log into the pocketbeagle and do: sudo ip route add default via 192.168.7.1 Sep 17 06:25:49 already done that too Sep 17 06:26:00 and it doesn't suffice? that is odd Sep 17 06:26:13 it send the ping but is not receiving any packets back Sep 17 06:26:53 Do I need to change anything from the host system? Sep 17 06:28:03 can you use pastebin.com (or similar paste site) to share the output of the following commands on both host and pocketbeagle: Sep 17 06:28:11 ip route Sep 17 06:28:23 Also from the host system when I do ifconfig, it shows two connections. one at inet 192.168.7.1 and the other at inet 192.168.6.1 Sep 17 06:28:36 iptables -t nat -L Sep 17 06:28:45 actually the latter command is only relevant on the host Sep 17 06:29:35 yes the .6.1 one is used for Mac and Linux hosts Sep 17 06:30:20 it uses the usb-standard ethernet device class, while .7.1 uses microsoft's proprietary rndis protocol because microsoft didn't feel like implementing the standard (as usual) Sep 17 06:30:40 oh sorry Sep 17 06:30:42 windows host Sep 17 06:30:44 linux host Sep 17 06:30:54 so eh, .6.1 should be configured, not .7.1 Sep 17 06:31:08 but presumably you have that, since otherwise you wouldn't be able to communicate with the pocketbeagle Sep 17 06:31:51 either that or it actually uses .7.x on linux also, but I think it should be .6.x (linux supports both protocols so it could use either in principle) Sep 17 06:32:01 https://pastebin.com/P10YPdVX This is the output from ifconfig Sep 17 06:32:06 from ubuntu Sep 17 06:32:12 that's not what I asked for Sep 17 06:32:32 I asked for "ip route" on both ubuntu and pocketbeagle, and "iptables -t nat -L" on ubuntu Sep 17 06:32:53 the latter needs sudo actually Sep 17 06:33:50 oh it actually uses both connections on linux, that makes sense actually... not sure why I thought it would be either-or Sep 17 06:34:30 Hi Sep 17 06:34:42 I am sorry I got disonnected again Sep 17 06:34:46 https://pastebin.com/a6Ue0Xdq Sep 17 06:34:55 This is the nestat -anr from the BBB Sep 17 06:35:01 sorry pocketbeagle* Sep 17 06:35:09 is it really too much to just give the output of the commands I asked for? Sep 17 06:35:28 I got disconnected Sep 17 06:35:33 "ip route" on both ubuntu and pocketbeagle and "sudo iptables -t nat -L" on ubuntu Sep 17 06:36:11 although right away it looks like there's no default route on the pocketbeagle Sep 17 06:39:31 https://pastebin.com/0hg2egGT Sep 17 06:39:58 The outputs of ip route and iptables Sep 17 06:40:08 ok so it does have a default route, apparently I'm just not familiar with the output of netstat -r (which is why I asked for ip route) Sep 17 06:40:26 no internet sharing has been configured on ubuntu though Sep 17 06:40:39 are you sure you configured it to share your connection? Sep 17 06:40:59 what am I supposed to do to configure it on the ubuntu Sep 17 06:41:00 ? Sep 17 06:41:56 I I used the following to set up on ubuntu Sep 17 06:41:58 https://pastebin.com/u5AhEP1y Sep 17 06:42:42 okay so earlier when I gave a brief outline of steps and you said "I did that" you were lying Sep 17 06:42:51 since you did something totally different Sep 17 06:44:00 that nat table entry you added is clearly no longer present, presumably since the firewall/nat is managed by networkmanager Sep 17 06:44:03 this is ubuntu, use the gui Sep 17 06:44:29 okay can you please repeat? Sep 17 06:44:34 I am sorry about that. Sep 17 06:44:48 what do I do on ubuntu? Sep 17 06:47:23 help please, I keep losing connection and so am unable to read messages. Please tell me what do on the ubuntu machine Sep 17 06:47:54 open nm-connection-editor, find the interface that uses .7.1 as address, configure it to share internet connection but make sure it's still configured to use 192.168.7.1 as IP address instead of the default of 10.42.0.1 (I'm not sure what this looks like in the gui, I haven't been able to find screenshots) Sep 17 06:48:28 and maybe download some irc client to use instead of the web-chat, that should help give you a more stable connection Sep 17 06:48:49 getting frequently disconnected is frustrating for both you and anyone trying to help you Sep 17 06:49:01 I am not using GUI Sep 17 06:49:17 on ubuntu you probably should be Sep 17 06:49:18 I am using the terminal to set this up. Sep 17 06:49:29 in that case use nmcli Sep 17 06:49:36 but I don't know the commands for that Sep 17 06:49:56 okay I will try that Sep 17 06:50:16 how should I configure the interface to share the connection though? Sep 17 06:50:35 it shouldn't be hard to find the setting in the gui Sep 17 06:50:48 that's usually the benefit of GUIs :P Sep 17 06:51:04 I am using Mate Sep 17 06:51:39 as long as it uses gnome-networkmanager it doesn't matter Sep 17 06:52:13 I am using mate not gnome Sep 17 06:52:13 also make sure the dnsmasq-base package is installed, it's required for enabling internet sharing in networkmanager Sep 17 06:52:27 it almost certainly still uses gnome-networkmanager Sep 17 06:56:26 https://pastebin.com/UgVr3vWe Sep 17 06:56:41 now it should be connected right? Sep 17 07:00:07 I don't know what you think I can tell from just this output (which is the same as your ip route was previously) Sep 17 07:00:18 I cant find it Sep 17 07:00:44 there's something weird going on with that enx606405fd9f22 interface btw, since it also has a link-local address configured Sep 17 07:00:47 when I open the network manager there is no interface that has the ip .7.1 Sep 17 07:01:33 and what is weird about that? Sep 17 07:02:09 because most network managers only configure a link-local ipv4 address if they're configured to use dhcp but fail to acquire an address Sep 17 07:02:45 at this point I suggest disconnecting the pocketbeagle and reconnecting it to get rid of the stuff you've been doing from the commandline Sep 17 07:02:52 start fresh Sep 17 07:04:02 I have already done that like a milion times. But okay I will do it one more time Sep 17 07:05:05 maybe reboot ubuntu as well Sep 17 07:07:06 that seems excessive Sep 17 07:08:40 grepart_: that you've done it before is irrelevant since both your words and the output of ip route seem to indicate you've been manually fiddling with the configuration after connecting the pocketbeagle and the goal of disconnecting and reconnecting it is to undo that Sep 17 07:10:36 hi sir i have phasing one problem can you please solve that issue . problem is i connected mongo OH board is connected to beagleboneblack through usb at that time mongo OH is restarting every time Sep 17 07:11:05 what is the problem of restarting Sep 17 07:11:08 what on earth is a "mongo OH board" Sep 17 07:12:50 if it requires more than 500 mA of current then you won't be able to connect it directly to the beaglebone, you'll need to insert a powered usb hub in between Sep 17 07:13:28 oho kkk Sep 17 07:14:02 thank you for valuable information Sep 17 07:14:14 (and even then it's not guaranteed since in general a usb port isn't required to be able to supply more than 500 mA) Sep 17 07:18:25 and one more sir i made ppp connection fron mango OH to Beagleboneblack than that time internet communication is very slow ,when i was used usb communication than internet is very good Sep 17 07:19:08 okay Sep 17 07:21:20 is it possible to connect ehternet communication from mango OH to Beagleboneblack Sep 17 07:21:47 sure? assuming whatever the "mango OH" is has ethernet Sep 17 07:22:01 not uart communication like similar as dongle Sep 17 07:22:09 (I'm finding zero relevant google hits so I have no idea what it is) Sep 17 07:22:28 I have no idea what you're trying to say Sep 17 19:31:25 Hey guys, I'm running kernel 4.9.78-bone9 and am trying to figure out where to get source to this version (or how to patch the regular kernel source to get this version) Sep 17 19:32:49 branch am33x-v4.9 of https://github.com/RobertCNelson/bb-kernel Sep 17 19:33:19 the readme however claims it shouldn't be used for building older git tags, and refers to some other repo for that, although I have no idea why Sep 17 19:34:01 okay, I'll start there, thanks Sep 17 19:34:11 you can also find the patched kernel tree for that exact version at https://github.com/RobertCNelson/linux-stable-rcn-ee/tree/4.9.78-bone9 Sep 17 19:34:26 but if you want to reproduce the kernel, using the build scripts is probably better Sep 17 19:34:30 much preferred, I'l do that. Thanks a bunch Sep 17 19:34:54 well, I've got some stuff to try Sep 17 19:35:03 the build scripts will directly produce debs for you Sep 17 19:35:39 build scripts in https://github.com/RobertCNelson/bb-kernel? Sep 17 19:35:44 yeah Sep 17 19:35:55 I'll try both and see what works. Thanks. Sep 17 19:36:08 https://pastebin.com/eLhrp1Hg Sep 17 19:36:13 some remarks I made about the process Sep 17 19:37:12 typically you'll want to use the latest kernel in a series (e.g. 4.9-bone in this case) anyway, to ensure you have the latest bugfixes Sep 17 19:46:38 Okay, building. We'll see how that goes. Thanks again. Sep 17 20:03:56 how can I implement an ocp barrier in C ? Sep 17 21:16:47 CoffeeBreakfast: read from any register of the same peripheral Sep 17 21:17:44 (making sure the read is not optimized away by the compiler, typically accomplished by performing the access through a volatile-qualified pointer) Sep 17 21:20:13 CoffeeBreakfast: note that if you're accessing a peripheral from linux userspace (via /dev/mem or uio) then the memory will be mapped as strongly-ordered and ocp barriers will be unnecessary. it may however be wise to avoid relying on that. (I personally actually patch the kernel to use device-type mappings instead of strongly-ordered ones, same as what's being used in the kernel) Sep 17 21:29:11 thanks, zmatt. Another q: https://github.com/mvduin/py-uio/blob/master/src/uio.py#L192 -> the equivalent in C ? Sep 17 21:31:39 (almost finishing the C version of that) Sep 17 21:49:59 uint x = 1; write( fd, &x, sizeof x ) Sep 17 21:50:59 or uint32_t if you want to be precise (although uint is 4 bytes on any linux system so *shrug*) Sep 17 21:53:41 thanks! Sep 17 21:57:49 It's alive! Sep 17 21:58:54 obviously It's a very rude version of your code, in C Sep 17 22:01:01 I generally use C++ myself Sep 17 22:04:19 I prefer C++ over Python, when OOP Sep 17 22:05:03 I don't like python *at all*. py-uio is my first non-trivial python project and it may very well be my last Sep 17 22:05:39 awesome job in such case Sep 18 01:40:23 So... Sep 18 01:40:54 For the .dtbo file in /lib/firmware for the MotorCape, do I just erase it and then add it later if needed? Sep 18 01:41:02 Or... Sep 18 01:41:10 Do I add it to uEnv.txt? Sep 18 01:41:27 i.e. sudo rm xxxxxx.xxxx.dtbo. Sep 18 01:41:52 Off to test it! Sep 18 01:43:15 I do not have any text in that file. Sep 18 01:43:24 Maybe this is the issue. Sep 18 01:46:23 I just used the cmd: sudo rm xxxxx.xxxxx.dtbo and removed it. Sep 18 01:46:28 I am rebooting. Sep 18 01:49:36 Hello set_ Sep 18 01:49:58 Coffee! Sep 18 01:50:03 Guess what? Sep 18 01:51:57 zmatt: echo "1000" > /sys/bus/iio/devices/iio\:device1/out_voltage0_raw works great... but printf("32000\n"); and pipe from that c program to out_voltage0_raw doesn't work Sep 18 01:52:02 set_ what? Sep 18 01:52:12 I got this darn MotorCape! Sep 18 01:52:42 I am typing up some software now but the GPIO is not registering as output right now. Sep 18 01:54:49 Well... Sep 18 01:55:15 I got the software to print, "I like donkey face!" but the motor no go. Sep 18 01:55:18 Sheesh. Sep 18 01:57:46 I'm having troubles too Sep 18 01:57:51 Oh! Sep 18 01:57:57 What are you doing? Sep 18 01:58:42 an arbitrary waveform generator Sep 18 01:58:48 Damn! Sep 18 01:59:02 I got my LED to light up but the motor is not moving just yet. Sep 18 01:59:23 I would have to look up the awg. Sep 18 01:59:30 I know nothing about it. Sep 18 02:02:16 I receive an error for a GPIO channel and output. Sep 18 02:03:30 So, either the PWM channel works and runs the LED or the GPIO channel states that it has not been setup(). Sep 18 02:03:32 Oh well. Sep 18 02:14:10 https://pastebin.com/gWcP0DGZ is what I tried but I have been unsuccessful. I am going to try another library real quickly. Sep 18 02:29:24 real quickly = forever and a day! **** ENDING LOGGING AT Tue Sep 18 03:00:00 2018