**** BEGIN LOGGING AT Tue Dec 22 03:00:27 2020
Dec 22 03:00:27 <boinkers> mattb0000ne: Are you still wanting to learn more? I suggest we compare articles, no?
Dec 22 03:01:40 <boinkers> https://ww1.microchip.com/downloads/en/DeviceDoc/AR1020-AR1021-LINUX-SPI-I2C-V102.pdf is what you need to know if you want your Cape to work, i.e. outside of the BBAI caveats.
Dec 22 03:04:11 <boinkers> Okay, from what I am reading, this is where I can come in handy (I guess), we need to cross-compile this source "arch/arm/mach-s3c6410/mach-smdk6410.c" but we will most likely need to change it some.
Dec 22 03:04:22 <boinkers> There is no way it is 100% acurate.
Dec 22 03:05:00 <boinkers> Hello?
Dec 22 03:05:27 <boinkers> I am staying patient now to wait for more communication.
Dec 22 03:14:24 <GenTooMan> mattb0000ne I believe the driver is for multiple screens, does it use the I2C bus for identification and IO allocation?
Dec 22 03:19:26 <boinkers> SPI or I2C and I think it can use UART too.
Dec 22 03:19:46 <boinkers> Blah!
Dec 22 03:19:55 <boinkers> I think he thinks we are mad at him.
Dec 22 03:20:56 <boinkers> Hopefully, he looks at the https://ww1.microchip.com/downloads  PDF for sake's sake.
Dec 22 03:21:20 <boinkers> That was not supposed to be a hyperlink.
Dec 22 03:22:36 <boinkers> Anyway, it would be a neat project to make work. I got the cape. I do not have the darn display though.
Dec 22 03:23:15 <boinkers> Those tiny wires in ribbon format are beyond me. Neat package, well kept, and tidy!
Dec 22 03:25:44 <boinkers> five wires? No way.
Dec 22 03:26:11 <boinkers> For i2c, five wires is all it is saying so far for communication.
Dec 22 03:28:27 <GenTooMan> no he probably is AFK Away From Keyboard
Dec 22 03:28:35 <boinkers> What section is he on?
Dec 22 03:28:43 <boinkers> Same page may work better.
Dec 22 03:29:19 <boinkers> I see a bunch of ideas for SPI too.
Dec 22 03:29:41 <boinkers> but...there was something about kconfig. Is that a thing w/ the BBB?
Dec 22 03:31:04 <boinkers> The MicroChip link I sent in for the PDF is only  pages. I think he has it covered.
Dec 22 03:31:24 <boinkers> 26 pages.
Dec 22 03:31:32 <boinkers> Ubuntu hates my expanded keyboard.
Dec 22 03:31:42 <GenTooMan> I have no idea but he's not here at this time
Dec 22 03:32:07 <boinkers> GenTooMan: What the hell, man?
Dec 22 03:32:27 <boinkers> Sir, I thought we were rallying for the Cape ideas?
Dec 22 03:33:39 <boinkers> i2c is not visited on the AI like it is on the BBB. I do not think that the i2c peripherals are even located near one another.
Dec 22 03:36:09 <boinkers> "HOLDING"
Dec 22 03:36:11 <boinkers> Ha.
Dec 22 03:37:00 <boinkers> GenTooMan: Do you really hate giving out flashcards as much as I think you do?
Dec 22 03:37:10 <boinkers> for me, for me?
Dec 22 03:41:41 <GenTooMan> the I2c bus is used to configure the BBB io to support the LCD likely shutting off the HDMI driver and setting up the other IO to operate.
Dec 22 03:44:23 <boinkers> Oh.
Dec 22 03:44:34 <boinkers> See. I was unaware of that factoid.
Dec 22 03:45:05 <boinkers> I have no idea how to do that stuff.
Dec 22 03:45:33 <boinkers> It is like a switch or source defined?
Dec 22 03:45:57 <GenTooMan> normally you don't
Dec 22 03:46:15 <GenTooMan> the BBB sets things up for you that's part of the use of the i2c devices on capes
Dec 22 03:46:27 <boinkers> Right. Right.
Dec 22 03:46:39 <boinkers> But, there is another i2c device.
Dec 22 03:46:52 <boinkers> I think three actually.
Dec 22 03:46:56 <GenTooMan> however I believe mattb0000ne removed the driver for all LCD connected to the bus.
Dec 22 03:47:06 <boinkers> Hmm?
Dec 22 03:47:13 <boinkers> What do you mean?
Dec 22 03:47:20 <GenTooMan> when likely he needed to look for the hook to set the display output for the device connected.
Dec 22 03:47:33 <boinkers> Barrg.
Dec 22 03:47:41 <GenTooMan> he removed a driver, which is likely for a bus connected LCD (like he was using).
Dec 22 03:47:51 <boinkers> Oh.
Dec 22 03:47:57 <boinkers> That is stressful.
Dec 22 03:48:10 <GenTooMan> no it's learning! you can't learn by making mistakes.
Dec 22 03:48:17 <boinkers> So, one of the drivers onboard the Ha.
Dec 22 03:48:24 <GenTooMan> err you can't learn without making mistakes
Dec 22 03:48:28 <GenTooMan> :D
Dec 22 03:48:28 <boinkers> Ha.
Dec 22 03:48:43 <boinkers> Hey. So, one of the drivers onboard the actual Cape was removed?
Dec 22 03:49:19 <boinkers> Smiles, smiles.
Dec 22 03:49:22 <GenTooMan> I believe that's what he did, when what he likely needed was a way to make sure that the LCD information was being relayed to the driver for the LCD
Dec 22 03:49:44 <boinkers> Oh. b/c it was looped to other drivers?
Dec 22 03:50:07 <GenTooMan> that's when having console access is helpful dmsg is handy to see what is happening on startup.
Dec 22 03:50:29 <boinkers> I just do not get it. What about that link I showed? Did the kconfig not work?
Dec 22 03:50:51 <boinkers> and the cross-compile w/ changes for the board in question, AI?
Dec 22 03:51:14 <GenTooMan> honestly we have no idea what it did because we don't have the dmsg information. that tells you what the kernel thinks it has and what it intends to do with it.
Dec 22 03:51:27 <boinkers> I mean, there are enough peripherals onboard the AI to...oh.
Dec 22 03:51:33 <boinkers> Okay.
Dec 22 03:52:17 <boinkers> So, even w/ a TTL to USB cable, no go?
Dec 22 03:52:28 <boinkers> Oh. AI. Cape. Got it.
Dec 22 03:53:00 <boinkers> What about that TTL to USB cable for the three pin adapter?
Dec 22 03:53:41 <boinkers> Would there still be no way to get info. from the Cape from the AI?
Dec 22 03:54:50 <boinkers> Blindly in the dark or purchase a new one?
Dec 22 03:55:01 <boinkers> Or add the driver back on?
Dec 22 03:56:20 <boinkers> So, one driver becomes missing, the kernel now does not have a way to compute what is happening on the Cape, and the AI is held useless?
Dec 22 03:56:50 <GenTooMan> I suspect it just can't configure the LCD.
Dec 22 03:56:56 <boinkers> I guess.
Dec 22 03:57:06 <GenTooMan> IE the driver starts by configuring the registers for driving the LCD
Dec 22 03:57:29 <boinkers> He could always get a fanatic to attach the driver back on!
Dec 22 03:57:55 <GenTooMan> it gets the information how to set up the data bus and signals from an I2C device, however if it didn't "just work" to start with ...
Dec 22 03:58:05 <GenTooMan> Is the AI IO still a mess?
Dec 22 03:58:16 <boinkers> I do not know that answer.
Dec 22 03:58:23 <boinkers> Probably.
Dec 22 03:58:57 <GenTooMan> I remember zmatt saying that things weren't great in that respect, so the issue may have nothing to do with the LCD but the basic stuff written for the AI is not complete.
Dec 22 03:59:31 <boinkers> Why in the Sam Keniston did no one say, "No, do not do that right now. Hold out."
Dec 22 04:00:28 <boinkers> "No, do not remove the driver from the Cape!"
Dec 22 04:01:41 <boinkers> I mean, I understand a rig but how is one driver being removed supposed to clear up a device issue?
Dec 22 04:02:00 <boinkers> It was a short in the dark?
Dec 22 04:02:03 <boinkers> Ha!
Dec 22 04:02:50 <boinkers> I really wanted to compile and do some diggin'.
Dec 22 04:02:54 <boinkers> Oh well.
Dec 22 04:03:55 <boinkers> "Probing will commence in five seconds."
Dec 22 04:03:56 <boinkers> Ha.
Dec 22 04:04:33 <boinkers> Oh well. 10:00 again, GenTooMan...
Dec 22 04:04:41 <boinkers> Sheesh.
Dec 22 04:05:02 <boinkers> That was stressful for me. I feel bad for mattb0000ne.
Dec 22 04:05:43 <boinkers> Do it! Do not do it! Why did you do it?
Dec 22 04:05:45 <boinkers> Argh!
Dec 22 04:06:11 <boinkers> When someone told me to ruin my Cape, I said, "Nope!"
Dec 22 04:06:31 <boinkers> After throwing one thing out, I realized the parts are all interconnected!
Dec 22 04:06:49 <boinkers> Who knows? That part may come back to haunt me in my PJs.
Dec 22 04:06:52 <boinkers> Over and out!
Dec 22 04:15:30 * GenTooMan decides getting to bed on time is good. So LO that means you are 1 hour behind me.
Dec 22 06:38:16 <sakethr98[m]> Hello Folks! Saketh here.(after a very long time)
Dec 22 06:38:49 <sakethr98[m]> I would like to confirm if it's okay to travel with beagle boards in  cabin baggage while on flights
Dec 22 06:39:06 <jkridner[m]> I do so all the time.
Dec 22 06:39:33 <jkridner[m]> no different than traveling with a powered down phone.
Dec 22 06:40:26 <sakethr98[m]> Thank you.
Dec 22 07:34:23 <zmatt> jkridner[m]: no, it's far less problematic than traveling with a phone, since a phone contains a li-ion battery :P
Dec 22 08:15:27 <sakethr98[m]> Haha
Dec 22 09:56:04 <TonyTheLion> @zmatt: I compiled and installed kernel 4.19.94-4.14-rt+, but I am still getting the same error. I enabled the driver, and fixed the dtsi interrupts. Is this a supported kernel, as you were saying last week?
Dec 22 09:57:17 <zmatt> uhh that's not a validly formatted version number, you probably copy-pasted that wrong
Dec 22 09:57:50 <TonyTheLion> that's what 'uname -r' gives me
Dec 22 09:57:50 <zmatt> and I don't actually remember what your issue was :P
Dec 22 09:58:08 <zmatt> that makes no sense, what on earth did you install?
Dec 22 09:58:55 <TonyTheLion> oh, it gives me "tps65217 0-0024: Failed to read IRQ status: -13" when I try to use the charger driver to tell me the USB was unplugged/plugged in when its asleep "echo mem > /sys/power/state".
Dec 22 09:59:09 <zmatt> oh right, that weirdness
Dec 22 09:59:35 <zmatt> what do you mean by "enabled the driver, and fixed the dtsi interrupts" ?
Dec 22 09:59:50 <zmatt> and what kernel did you install?
Dec 22 10:00:00 <TonyTheLion> enabled the driver means, setting to "=m" in the kernel config
Dec 22 10:00:05 <zmatt> which driver?
Dec 22 10:00:23 <TonyTheLion> tps65217-charger
Dec 22 10:01:14 <zmatt> oh it's not compiled-in by default... weird, ok
Dec 22 10:01:48 <TonyTheLion> kernel is branch 4.19-rt from https://github.com/beagleboard/linux.git
Dec 22 10:02:34 <zmatt> I have no idea what that will get you.. and why -rt ?  and if you want to customize the kernel, use the kernel build repo which automates the whole process of building the .deb for you
Dec 22 10:02:53 <zmatt> https://pastebin.com/eLhrp1Hg  my notes on customizing a beaglebone kernel
Dec 22 10:03:16 <zmatt> you'll want to use the repo for -ti kernel series and select branch ti-linux-4.19.y
Dec 22 10:03:40 <TonyTheLion> ah that is helpful
Dec 22 10:03:46 <TonyTheLion> what is -rt for?
Dec 22 10:04:55 <zmatt> it's built with preempt_rt, see rt.wiki.kernel.org
Dec 22 10:05:30 <zmatt> unless you explicitly need it I'd recommend against it
Dec 22 10:06:20 <zmatt> since unless you're purposefully making use of it, it'll just be a less efficient and probably buggier kernel
Dec 22 10:06:28 <TonyTheLion> I'll have to check if we explicitly need it, not entirely sure
Dec 22 10:08:03 <zmatt> are you already using realtime scheduling and performance analysis showed that peak latency is too high?
Dec 22 10:11:59 <TonyTheLion> no
Dec 22 10:12:38 <zmatt> well then that answers your question ;P
Dec 22 10:12:45 <TonyTheLion> :)
Dec 22 12:08:33 <garry> can anyone help me getting familiar to the platform
Dec 22 15:20:12 <kremlin> m
Dec 22 16:56:04 <ayjay_t> is the kernel using gpio driver for any essential function?
Dec 22 16:56:07 <ayjay_t> in the base bbb image?
Dec 22 16:56:24 <ayjay_t> or systemd
Dec 22 21:57:35 <boinkers> GenTooMan: I think I got the source to account for stoppage times but it is backwards for now. So, it runs until the stop halts execution but then it starts the source from the beginning. Boo!
Dec 22 21:58:37 <boinkers> So, if I am three inches to the left on the slide when the interrupt happens, it starts the source from that point forward in the while loop.
Dec 22 21:59:30 <boinkers> Source runs >>> the interrupt happens b/c of the extent of the slide >>> the program starts from scratch
Dec 22 21:59:51 <boinkers> So, I need to account for the transition from the interrupt.
Dec 22 22:01:45 <boinkers> So, would it be a nested loop that accounts for this transition?
Dec 22 22:03:28 <boinkers> So, through the script >>> interrupt >>> starts where it left off? That is not easy w/out some idea that I do not currently have now.
Dec 22 22:37:44 <boinkers> So, I make a .py file w/ a function and call that function when I need to start from scratch after the interrupt. Okay. I think I am moving toward something.
Dec 22 22:40:21 <ayjay_t> any docs on the current state of BBB GPU?
Dec 22 22:52:13 <boinkers> ayjay_t: I did not find that section on the TRM. Is that an added component?
Dec 22 22:54:56 <boinkers> I am looking through the SRM, too.
Dec 22 22:56:36 <boinkers> Anyway, I will keep looking. For me and my knowledge as of now, I do not know anything about the GPU.
Dec 22 22:56:39 <boinkers> Sorry.
Dec 22 22:59:24 <boinkers> Is that the SGX?
Dec 22 23:02:11 <boinkers> Man...the new SRM on github is nice.
Dec 22 23:02:21 <boinkers> Way to go, fellows/ladies.
Dec 23 00:05:39 <ayjay_t> alright so, GPU docs, can I safely turn off GPIO driver
Dec 23 00:05:46 <ayjay_t> Qs on my stack
Dec 23 00:06:15 <ayjay_t> also good luck to boinkers
Dec 23 00:55:49 <mattb0000ne> how do you get a script executed at start up
Dec 23 01:07:37 <boinkers> mattB000ne: Would you like me to tell you or would you want to learn about .service files somewhere else?
Dec 23 01:13:53 <mattb0000ne> if you have a link to a good resource i will take it
Dec 23 01:14:05 <boinkers> mattB0000ne: Okay. Please hold. I will get one.
Dec 23 01:15:31 <boinkers> Hello...I usually use this one but it is not pleasing to the eyes. It is very useful. https://www.freedesktop.org/software/systemd/man/systemd.service.html#Examples
Dec 23 01:16:09 <boinkers> Also, .service files can be placed in /etc/systemd/system/ for instance. You can look in that dir. to see what else is located in that specific dir.
Dec 23 01:16:23 <boinkers> Some people put them in /lib/systemd/system/ though.
Dec 23 01:16:27 <boinkers> It really depends.
Dec 23 01:17:40 <boinkers> mattb0000ne: If you are looking for a short script to run a simple .py or .cpp file, I can show you one on pastebin.com.
Dec 23 01:18:56 <boinkers> Six or seven lines and booting on boot! Also, you will need to use systemctl to start, stop, enable, disable your .service files.
Dec 23 01:21:07 <boinkers> Please just let me know if you need an example w/ required commands for starting on boot.
Dec 23 01:47:14 <mattb0000ne> I will manage set_
Dec 23 01:47:45 <boinkers> Okay. Did you get it to boot w/ the file running at boot?
Dec 23 01:48:16 <boinkers> Or...whatever you are making run on boot?
Dec 23 01:48:57 <mattb0000ne> https://www.tecmint.com/auto-execute-linux-scripts-during-reboot-or-startup/#:~:text=Executing%20Linux%20Scripts%20During%20Reboot%20or%20Startup%20There,cron%20scheduler%20also%20allows%20the%20use%20of%20%40reboot.
Dec 23 01:52:11 <boinkers> Oh. Cron.
Dec 23 01:52:58 <boinkers> I thought you wanted it to work on boot?
Dec 23 01:53:23 <boinkers> I thought w/ Cron one would have to sign in.
Dec 23 01:54:19 <boinkers> in the beginning, I also wondered why my scripts were not in working order during boot w/out sign in. I was using cron. This maybe why? I guess. Who knows?
Dec 23 01:54:58 <mattb0000ne> I do not think services would do what I needed
Dec 23 01:55:33 <boinkers> Okay.
Dec 23 02:00:45 <mattb0000ne> thank you for the effort
Dec 23 02:02:12 <boinkers> No isse.
Dec 23 02:09:25 <boinkers> Sorry. no issue.
Dec 23 02:20:10 <zmatt> mattb0000ne: https://pastebin.com/KXVdTNrL   example of a startup service
Dec 23 02:21:04 <mattb0000ne> zmatt: thanks
Dec 23 02:21:38 <zmatt> mattb0000ne: cron is for things that need to be executed periodically.. and it's also a kinda obsolete since systemd also supports running services periodically (via timer units)
Dec 23 02:22:04 <mattb0000ne> zmatt: what about the use of rc.local that no good?
Dec 23 02:22:57 <zmatt> mattb0000ne: rc.local is a legacy thing, it's executed by rc-local.service (conditional on /etc/rc.local existing)
Dec 23 02:23:29 <zmatt> so it's like a custom systemd service, except without the flexibility of being able to configure it to your needs :P
Dec 23 02:23:36 <mattb0000ne> ok
Dec 23 02:24:28 <zmatt> but my service example should be straightforward enough I hope
Dec 23 02:24:49 <mattb0000ne> zmatt: I am trying to spare set_ feelings but yes
Dec 23 02:27:14 <zmatt> an alternative to making a system service is making a user service: https://pastebin.com/uqY2DcGs
Dec 23 02:28:32 <zmatt> mostly similar, except it has more restrictions (e.g. it will always run under the debian account), but it avoids having to put stuff in /etc/ and you can control the service without sudo
Dec 23 02:30:56 <mattb0000ne> zmatt do you ever fiddle with the usb3
Dec 23 02:30:58 <mattb0000ne> port
Dec 23 02:31:21 <zmatt> I haven't fiddled much with the bbai in general
Dec 23 02:31:57 <mattb0000ne> I have it plugged into a usb hub
Dec 23 02:32:06 <mattb0000ne> and it works fine since it is powered
Dec 23 02:32:28 <mattb0000ne> I was following this thread https://www.element14.com/community/thread/75059/l/usb-30-support-for-beagle-bone-ai
Dec 23 02:32:49 <mattb0000ne> when I run that script it disconnects rather than allow for the detection of more devices
Dec 23 02:34:20 <mattb0000ne> still have activity
Dec 23 02:34:24 <mattb0000ne> but I lose ssh
Dec 23 02:36:18 <zmatt> what do you mean "more devices", you said it already worked?
Dec 23 02:37:34 <zmatt> and having to do this manually is indicative of some kernel logic being missing/broken for the bbai's hardware setup, normally it should switch automatically
Dec 23 02:38:22 <mattb0000ne> I want it to see the camera
Dec 23 02:38:28 <mattb0000ne> which is also connected
Dec 23 02:38:31 <mattb0000ne> maybe it is
Dec 23 02:38:44 <zmatt> you mean it sees the hub but not any downstream devices?
Dec 23 02:38:44 <mattb0000ne> let me check
Dec 23 02:38:48 <mattb0000ne> running the software now
Dec 23 02:39:18 <mattb0000ne> the different rXX of the kernels are those just bug fixes ?
Dec 23 02:39:31 <zmatt> patches or config changes
Dec 23 02:43:22 <boinkers> My feelings?
Dec 23 02:43:27 <boinkers> Do not worry.
Dec 23 02:44:47 <boinkers> I am watching the .org's way of doing EXPO!
Dec 23 02:45:01 <boinkers> I think it is an older one since COVID-19 kicked in.
Dec 23 02:47:16 <boinkers> Embedded World 2020!
Dec 23 02:47:48 <mattb0000ne> set_ you are the Beagle ambassodor but you confuse the hell out of me
Dec 23 02:51:40 <boinkers> What do you mean?
Dec 23 02:51:48 <boinkers> Oh. .service files?
Dec 23 02:52:09 <boinkers> Do not worry. It is a southern charm I have.
Dec 23 02:52:09 <boinkers> Ha.
Dec 23 02:54:11 <boinkers> I have been looking over pwm instances and source from libs. I am not close to solving the project w/out help. I think I may bow out of helping. I am not gaining any ground right now on the pwm lib for BBB.
Dec 23 02:55:06 <boinkers> I have found many but they are all older and require about five or six files outside of the PWM lib itself that needs to be in python.
Dec 23 02:55:47 <boinkers> I keep finding file structure type libs. that do not make me understand what exactly is going on in their libs.
Dec 23 02:57:09 <mattb0000ne> what are you wokring on
Dec 23 02:57:24 <mattb0000ne> I can never understand what the hell you are doing becuase I have no context
Dec 23 02:57:29 <mattb0000ne> i take it this is not the drone
Dec 23 02:59:38 <mattb0000ne> ok so that is a big no on detecting the camera through the port
Dec 23 02:59:41 <mattb0000ne> boooooooooooooooooooooooo
**** ENDING LOGGING AT Wed Dec 23 02:59:56 2020