**** BEGIN LOGGING AT Tue Aug 28 03:00:03 2018
Aug 28 03:37:20 <CoffeeBreakfast> Anyone knows an app for visualizing/exploring the kernel's data structures??
Aug 28 03:43:52 <CoffeeBreakfast> Like CodeViz, but for data structures
Aug 28 03:47:02 <set_> not me.
Aug 28 03:47:31 <CoffeeBreakfast> Hello set_ !
Aug 28 03:48:22 <set_> Coffee: Hello!
Aug 28 03:48:35 <set_> I can look if you are tired of searching.
Aug 28 03:50:49 <set_> https://github.com/jarun/dslib might be of interest.
Aug 28 03:51:31 <set_> https://0xax.gitbooks.io/linux-insides/DataStructures/ might be helpful too.
Aug 28 04:17:32 <CoffeeBreakfast> DDD maybe
Aug 28 05:28:19 <zmatt> just close your eyes, clear your mind, and then visualize a big plate of spaghetti
Aug 28 05:54:27 <MaheenK>  I have connected 2 webcams with my BBB. They are appearing in lsusb. I am using fswebcam to test the images. first camera(video0) is capuring image and i can view it. But second camera (video1) is not capturing images. I am getting following error while running fswebcam command 'Timed out waiting for frame!" than "No Frame captured"'. Please let me know about solving this issue
Aug 28 05:56:42 <zmatt> disabling usb autosuspend is generally worth trying, some devices don't like autosuspend and get confused
Aug 28 05:58:22 <MaheenK> let me try
Aug 28 05:59:30 <zmatt> you can disable autosuspend for a single device by writing "on" to /sys/bus/usb/devices/.../power/control  (where ... identifies the device)
Aug 28 06:01:16 <zmatt> you can disable autosuspend for all newly plugged-in devices by writing -1 to /sys/module/usbcore/parameters/autosuspend
Aug 28 06:02:03 <zmatt> (the latter can be made persistent by adding   usbcore.autosuspend=-1   to the "cmdline" variable in /boot/uEnv.txt)
Aug 28 06:02:28 <MaheenK> i think i did it once. in uEnv
Aug 28 06:03:42 <zmatt> if disabling autosuspend doesn't help, it may be worth to see if inserting a usb hub between the bbb and the device helps, especially a self-powered (rather than bus-powered) hub
Aug 28 06:05:27 <MaheenK> bus powered hub is connected between bbb and webcam
Aug 28 08:31:26 <thinkfat> gm
Aug 28 08:35:19 <thinkfat> ds2: that's actually a pretty good idea. There should be a gerber plot of the outline somewhere. usually they're separate
Aug 28 08:35:42 <thinkfat> ds2: any hint what program to use to convert it?
Aug 28 08:36:10 <thinkfat> jkridner: dimensioned drawing - not machine consumable.
Aug 28 08:45:06 <thinkfat> ds2: hm. EdgeCuts gives me the outline. but not he connectors...
**** BEGIN LOGGING AT Tue Aug 28 09:19:08 2018
Aug 28 13:55:21 <kcid42_> Hi, I have a question about dts files. I need an extra pin on my uart for half duplex RS485, and there is no pre-build file available. When I add the gpio pin manually, i get this errors in dmesg:
Aug 28 13:55:31 <kcid42_> [    1.394970] pinctrl-single 44e10800.pinmux: pin PIN95 already requested by ocp:helper; cannot claim for ocp:P9_19_pinmux
Aug 28 13:55:37 <kcid42_> [    1.406085] pinctrl-single 44e10800.pinmux: pin-95 (ocp:P9_19_pinmux) status -22
Aug 28 13:55:39 <kcid42_> [    1.406085] pinctrl-single 44e10800.pinmux: pin-95 (ocp:P9_19_pinmux) status -22
Aug 28 13:56:04 <kcid42_> i don't get it, as I am not loading any other dts files, I disabled all of them
Aug 28 13:56:12 <kcid42_> do you guys have any idea?
Aug 28 14:34:13 <zmatt> kcid42_: what does your dts file look like exactly? can you pastebin it?
Aug 28 14:34:43 <zmatt> it sounds like a conflict with cape-universal, but I'm wondering why you have a conflict only on that pin
Aug 28 14:35:30 <zmatt> why are you using an overlay in the first place, is the gpio needed by the kernel driver?
Aug 28 14:39:39 <kremlin> m
Aug 28 15:00:54 <ds2> gerbv
Aug 28 15:03:51 <kcid42_> zmatt: created a pastebin of the dts here: https://pastebin.com/uRyc4D2L
Aug 28 15:04:44 <kcid42_> The io pin has been multiplexed, and I want to use it as a GPIO. In my understanding, dts is the way to go. I am running a 4.9 kernel, and a lot of older options don't work anymore there
Aug 28 15:05:17 <zmatt> oh, P9.19 .. that pin is configured to i2c by default for cape autodetection
Aug 28 15:05:39 <zmatt> you will probably want to use a different pin unless you've completely run out of options
Aug 28 15:06:31 <zmatt> if you really want to use it as gpio, you can either use config-pin to reconfigure it to gpio at runtime, or have your dts explicitly disable its normal use in the main dt
Aug 28 15:07:22 <kcid42_> ok, that explains a lot. Other pins are working fine indeed
Aug 28 15:07:33 <zmatt> using an overlay (a dts) actually *is* the old way.  usually people use cape-universal (via its "config-pin" utility) on current images
Aug 28 15:07:34 <kcid42_> We are building a custom cape, and the first batches of PCB are in production now, so I rather not change the pin
Aug 28 15:07:39 <zmatt> overlays are still supported too though
Aug 28 15:08:15 <kcid42_> ah, it's hard to find with all the different guides from different dates. I will look into config pin then
Aug 28 15:09:08 <zmatt> basically the options are: 1. runtime configuration using cape-universal (default on current images)  2. overlay applied by u-boot (this will implicitly disable cape-universal)  3. overlay loaded at runtime by kernel (deprecated)  4. custom main dts entirely
Aug 28 15:09:19 <kcid42_> thanks for the help, we are building this thing, if you are interested: https://www.instagram.com/energyfloors/
Aug 28 15:09:33 <zmatt> older images only supported options 3 and 4
Aug 28 15:11:46 <zmatt> I used to work in the same office as the "sustainable dance club" which also had such people-powered floor tiles.  the whole idea is dumb as hell but hey, people like it ;)
Aug 28 15:13:16 <kcid42_> Haha, indeed. The new floors use solar power, which is more effective in powering stuff.
Aug 28 15:13:42 <kcid42_> I am now at the sustainable dance club / energy floors office, does that mean you are now at the RDM building?
Aug 28 15:13:52 <zmatt> putting solar panels on roofs makes a lot more sense than putting them on floors
Aug 28 15:14:26 <kcid42_> technically, yes. But if you want to engage people in thinking about power, letting them play games on them is actually a pretty good idea
Aug 28 15:14:29 <zmatt> ah you're the same people?  and no, never heard of "the RDM building"
Aug 28 15:15:18 <kcid42_> yeah, same people. This is a new concept we are working on
Aug 28 15:15:46 <zmatt> sustainable dance club used to be in (or at least have some sort of presence in) the office of Rinnic/Vaude in Maassluis
Aug 28 15:17:04 <kcid42_> ah, that is even longer ago. We are now in Heijplaat, Rotterdam, which is a old marine building site named RDM
Aug 28 15:17:22 <kcid42_> Nice coincidence that you are helping me now with these IO pins
Aug 28 15:17:36 <zmatt> it's a small world
Aug 28 15:17:47 <kcid42_> haha, it is
Aug 28 15:18:55 <zmatt> btw if you're making overlays, you may be interested in my https://github.com/mvduin/overlay-utils
Aug 28 15:19:09 <zmatt> it's designed to make writing overlays a bit less annoying
Aug 28 15:19:39 <kcid42_> ok, will try that, after I tried config-pin
Aug 28 15:19:45 <kcid42_> I just tried config-pin, but as before, I get an error|
Aug 28 15:19:56 <kcid42_> bash: /sys/devices/platform/ocp/ocp*P9_19_pinmux/state: No such file or directory Cannot write pinmux file: /sys/devices/platform/ocp/ocp*P9_19_pinmux/state
Aug 28 15:20:21 <zmatt> ohhh
Aug 28 15:20:40 <kcid42_> I was previously assuming that these mappings are not available as config-pin was outdated, but when it is not, something else is wrong
Aug 28 15:20:46 <zmatt> I just noticed your kernel error is actually about the P9_19_pinmux definition in the base dts failing rather than your overlay failing
Aug 28 15:20:51 <zmatt> i.e. in the conflict, your overlay won
Aug 28 15:20:57 <zmatt> that's why config-pin won't work
Aug 28 15:21:18 <kcid42_> let me disable my custom dts then
Aug 28 15:22:37 <zmatt> being able to use config-pin usually requires you don't have any custom overlays enabled at all.  P9.19 and P9.20 are somewhat special since their cape-universal-style pinmux helpers are in the base dts instead
Aug 28 15:23:02 <zmatt> (this itself is a fairly recent development, they used to be hard-configured to i2c rather than being available for reconfiguration at all)
Aug 28 15:23:26 <WilhelmVonWeiner> Is there a way to boot an image over tftp?
Aug 28 15:23:41 <zmatt> kcid42_: beware that regardless of what you do in dts, the bootloader will still configure those pins to i2c and attempt transfers
Aug 28 15:23:50 <zmatt> WilhelmVonWeiner: yes
Aug 28 15:24:10 <zmatt> WilhelmVonWeiner: both via ethernet and via rndis
Aug 28 15:25:18 <WilhelmVonWeiner> Is there a tutorial page somewhere
Aug 28 15:25:41 <zmatt> nope, it's a pretty advanced thing and requires building a custom u-boot
Aug 28 15:27:32 <WilhelmVonWeiner> Custom u-boot? I can already fetch files over tftp at least
Aug 28 15:29:21 <kcid42_> Great, config-pin is indeed working with the universal cape
Aug 28 15:29:35 <zmatt> WilhelmVonWeiner: but that's assuming u-boot is already on the beaglebone I presume?
Aug 28 15:29:55 <kcid42_> the pin is tied to a rs485 rts signal, so some jitter at boot is fine
Aug 28 15:30:01 <kcid42_> as long as it works after
Aug 28 15:31:19 <WilhelmVonWeiner> zmatt: yeah, but I've had the beaglebone for a long time and don't remember all the modifications I may have made to it
Aug 28 15:31:35 <zmatt> WilhelmVonWeiner: I mean, it depends on what you're trying to do exactly.  if you don't mind relying on the u-boot installed on eMMC and a suitable config file (also on eMMC) to make it perform the tftp boot then you don't need a custom-built u-boot
Aug 28 15:31:58 <zmatt> if you want to be able to boot a completely blank board via tftp, including serving u-boot itself via tftp, then you do need one
Aug 28 15:32:07 <WilhelmVonWeiner> Would you know what this config file would have to contain
Aug 28 15:32:32 <WilhelmVonWeiner> the board has u-boot installed, I had some """issues""" with `parted`
Aug 28 15:33:04 <zmatt> so, what sort of setup is your end-goal exactly?
Aug 28 15:33:18 <WilhelmVonWeiner> Some kind of image on the eMMC
Aug 28 15:33:24 <zmatt> so why not boot from sd card?
Aug 28 15:33:29 <zmatt> and reflash eMMC ?
Aug 28 15:34:41 <WilhelmVonWeiner> Can't boot from SD for some reason
Aug 28 15:34:47 <WilhelmVonWeiner> Think the physical port is buggered
Aug 28 15:35:25 <zmatt> are you sure about that?  try powering on while holding down the S2 button (the button closest to the card slot).  let go of the S2 button once the power led turns on
Aug 28 15:37:57 <WilhelmVonWeiner> The power LED doesn't turn on, because it doesn't like the card. "Card did not respond to voltage select!
Aug 28 15:38:48 <zmatt> what you're saying makes no sense.  the power led turning on isn't dependent on the card slot, and it can't print a message if the board isn't powered
Aug 28 15:39:49 <zmatt> also, if you powered it on in the way I described, it won't load u-boot from eMMC.  this means that if the card isn't readable, it will print nothing (or rather, it will periodically print "CCC" due to attempting xmodem boot)
Aug 28 15:40:01 <WilhelmVonWeiner> I'm not trying to trick you or something. If you hold down the s2, the serial output will give that error for the MicroSD card and the power LED won't turn on
Aug 28 15:41:04 <zmatt> you're describing two impossible things.  unless you mean it powers off after printing the message? (but the fact it prints anything still implies you didn't properly hold down the S2 button during power-up)
Aug 28 15:41:47 <WilhelmVonWeiner> there's only one way to hold down a button
Aug 28 15:42:20 <zmatt> well it's a tiny fiddly button, I've had cases where I thought I held it down but in fact it wasn't really pressed down (at least at the moment I plugged in power)
Aug 28 15:42:24 <WilhelmVonWeiner> As I said, I assume the physical sd card slot is buggered, because it "did not respond to voltage select"
Aug 28 15:43:16 <zmatt> that fact message is printed at all implies either you didn't power on with S2 held down, or it managed to load u-boot from sd card (in which case the card slot can't be buggered)
Aug 28 15:43:55 <zmatt> those cases can be distinguished easily by not having a card inserted
Aug 28 15:43:57 <WilhelmVonWeiner> means u-boot is on the eMMC
Aug 28 15:44:16 <zmatt> if you then power on with S2 held down, it will not print any u-boot message at all
Aug 28 15:45:05 <WilhelmVonWeiner> would the board being powered over USB matter?
Aug 28 15:45:37 <zmatt> no it doesn't matter how it's powered, as long as the S2 button is already held down at the moment power is applied to the board (i.e. the power led turn on)
Aug 28 15:49:29 <WilhelmVonWeiner> god I didn't realise a new board was 90 pounds
Aug 28 15:50:10 <zmatt> for a bbb? that sounds overpriced
Aug 28 15:50:10 <ds2> ?????? what kind of board is that heavy?
Aug 28 15:50:14 <zmatt> lol
Aug 28 15:50:30 <WilhelmVonWeiner> Definitely sounds overpriced
Aug 28 15:50:31 <ds2> lots of gold?
Aug 28 15:50:40 <WilhelmVonWeiner> 70 lowball according to amazon/ebay
Aug 28 15:51:21 <zmatt> the prices I'm seeing is 60-70 euro
Aug 28 15:51:44 <WilhelmVonWeiner> where?
Aug 28 15:53:05 <zmatt> various dutch online shops that appeared as sponsored links at the top of the google search results
Aug 28 15:54:30 <zmatt> element14 gives €46,59 ex VAT
Aug 28 15:55:10 <WilhelmVonWeiner> Yeah weird
Aug 28 15:55:23 <WilhelmVonWeiner> that same one popped up on Amazon but my initial search revealed higher prices
Aug 28 15:55:41 <zmatt> or £40.89 for UK
Aug 28 15:55:45 <WilhelmVonWeiner> maybe in this case I clicked a direct link.
Aug 28 15:55:57 <WilhelmVonWeiner> I bought the one I have now for 14 pounds lol
Aug 28 15:56:22 <zmatt> well I guess that explains the mystery of why stuff isn't working :P
Aug 28 15:56:58 <zmatt> I'm guessing the seller failed to include "broken card slot" in the description?
Aug 28 15:57:04 <set_> Hey!
Aug 28 15:57:05 <kcid42_> I soldered a sdcard holder on a raspberry pi once after breaking it, that's not too hard either, and only cost 4 euros
Aug 28 15:57:12 <set_> Get me a motor driver, too!
Aug 28 15:57:59 <WilhelmVonWeiner> no, the card slot didn't used to be broken. I've had it for n years, moving from box to box to shelf. The only problem here is my cavalier attitude to `parted`
Aug 28 15:58:11 <zmatt> parted doesn't break a card slot though
Aug 28 15:58:44 <zmatt> (although I can't say for sure whether it is in fact broken)
Aug 28 16:00:03 <kcid42_> thanks for the help guys, bye
Aug 28 16:02:26 <WilhelmVonWeiner> Well the slot was irrelevant at the time
Aug 28 16:02:40 <WilhelmVonWeiner> well only relevant in the sense I was going to partition the card in it
Aug 28 16:03:30 <set_> Hey! Does anyone know what is going on w/ the new Motor Cape? I see the seller pushed back the date and lead times are high.
Aug 28 16:04:18 <set_> I am trying to get my hands dirty w/ some more batteries and motors!
Aug 28 16:06:42 <set_> Also...while this is off subject, I have a capacitor that is mia online. I cannot find the datasheet!
Aug 28 16:06:56 <WilhelmVonWeiner> zmatt: Do you have any source I can start with for booting or flashing some kind of Linux image over ethernet?
Aug 28 16:07:24 <set_> 100 10v UT is what the cap reads on the top.
Aug 28 16:07:48 <zmatt> WilhelmVonWeiner: maybe try https://github.com/ravikp7/node-beagle-boot ?
Aug 28 16:08:01 <zmatt> just found it while googling
Aug 28 16:13:22 <set_> If anyone gets bored of doing their own thing, send rations and support the missing capacitor.
Aug 28 16:14:02 <set_> I have been using this motor driver w/ the BBB and the dang thing has this odd capacitor on it.
Aug 28 16:31:34 <set_> ...
Aug 28 16:31:46 <set_> anyway...I know that question or bunch of blabber is off subject. No issue.
Aug 28 17:03:07 <WilhelmVonWeiner> you know what
Aug 28 17:03:12 <WilhelmVonWeiner> nothing happens on holding s2
Aug 28 17:03:18 <WilhelmVonWeiner> board doesn't load nothing or do nothing
Aug 28 17:03:26 <zmatt> with sd card inserted?
Aug 28 17:03:38 <WilhelmVonWeiner> yeah
Aug 28 17:03:51 <WilhelmVonWeiner> and me holding the pins right up against the card just to be sure
Aug 28 17:04:11 <zmatt> what do you mean by that?
Aug 28 17:04:37 <zmatt> also, that means I was right earlier when I said you weren't actually powering up with s2 held down based on the results you reported ;)
Aug 28 17:05:23 <zmatt> but okay, that excludes the possibility that the u-boot on eMMC is the reason it doesn't boot from sd card
Aug 28 17:05:29 <WilhelmVonWeiner> I think I was but was reading the same stuff from the terminal thinking it'd scrolled
Aug 28 17:05:37 <zmatt> lol
Aug 28 17:06:04 <WilhelmVonWeiner> Nah I still get some sd card issue
Aug 28 17:06:05 <WilhelmVonWeiner> omap_hsmmc_send_cmd: timedout waiting on cmd inhibit to clear
Aug 28 17:06:15 <WilhelmVonWeiner> mmc_read_data: timedout waiting for status!
Aug 28 17:06:22 <WilhelmVonWeiner> in reverse order.
Aug 28 17:06:48 <zmatt> that's from booting from eMMC ?
Aug 28 17:07:12 <zmatt> can you copy-paste the actual output to pastebin.com ?
Aug 28 17:07:28 <WilhelmVonWeiner> whilst holding my sd in place, I get that, and sure one sec
Aug 28 17:07:52 <WilhelmVonWeiner> https://ghostbin.com/paste/mzgkb
Aug 28 17:08:45 <zmatt> so it actually manages to do some communication, but incredibly flaky
Aug 28 17:09:08 <zmatt> so I guess it might actually be some form of physical damage to the card slot
Aug 28 17:09:37 <zmatt> ahh, it automatically attempts tftp boot... that explains why you brought it up
Aug 28 17:09:56 <zmatt> I didn't know the default boot script does that nowadays (it didn't used to)
Aug 28 17:09:59 <WilhelmVonWeiner> Yeah, seeing as sd is dead ethernet comms was the first thing that came to mind
Aug 28 17:10:30 <zmatt> actually, you said you made a mistake with parted right?
Aug 28 17:10:40 <WilhelmVonWeiner> yep
Aug 28 17:11:12 <zmatt> I wonder if you actually wiped the eMMC, or merely deleted the partition
Aug 28 17:11:24 <zmatt> since in the latter case, creating the partition would restore the system to being bootable
Aug 28 17:11:57 <WilhelmVonWeiner> does u-boot provide utilities to help achieve that?
Aug 28 17:12:04 <zmatt> actually never mind, you didn't wipe eMMC since u-boot is still there
Aug 28 17:12:26 <zmatt> I'm inclined to assume u-boot will have a command for that
Aug 28 17:14:01 <WilhelmVonWeiner> you know what
Aug 28 17:14:10 <WilhelmVonWeiner> there's a usbboot command
Aug 28 17:14:29 <WilhelmVonWeiner> Going to explore that
Aug 28 17:17:14 <zmatt> oh wait, see if this works:  ums 0 mmc 1
Aug 28 17:17:28 <zmatt> hopefully the ums command is compiled into the u-boot you have
Aug 28 17:17:42 <zmatt> this makes the eMMC available as usb mass storage block device
Aug 28 17:17:55 <zmatt> which makes it really easy to then fix the partition table
Aug 28 17:18:26 <zmatt> (you could even reflash it entirely, but u-boot's ums is slooooooooow so it wouldn't be fun)
Aug 28 17:18:39 <zmatt> (unless it got much faster since last time I used it)
Aug 28 17:18:43 <WilhelmVonWeiner> unfortunately, `Unknown command 'ums'`
Aug 28 17:18:47 <zmatt> well that sucks
Aug 28 17:20:06 <zmatt> if you do find a command to create a partition, be sure to make it start at 4MB from the start of the disk (i.e. start sector 8192)
Aug 28 17:20:25 <zmatt> hopefully the ext4 filesystem is still there intact
Aug 28 17:46:24 <WilhelmVonWeiner> Thanks for the support attempts.
Aug 28 18:56:22 <WilhelmVonWeiner> This tool has got me sorted: https://github.com/ungureanuvladvictor/BBBlfs
Aug 28 20:33:54 <zmatt> WilhelmVonWeiner: beware that that project is unmaintained and the image included with that tool no longer works on recent beaglebones (unsupported eMMC version)
Aug 28 20:34:45 <zmatt> I actually came across that node-beagle-boot project while googling for bbblfs, hoped it would be a more modern replacement (but haven't really looked at it for more than a second)
Aug 28 20:58:13 <thinkfat_> bofff. nobody seems to stock capacitors these days
Aug 28 20:58:19 <thinkfat_> impossil
Aug 28 20:58:23 <thinkfat_> ble
Aug 28 20:58:41 <thinkfat_> looks like I get to scrape some of an older board.
Aug 28 20:58:58 <thinkfat_> off of...
Aug 28 21:00:10 <thinkfat_> 50V 4.7µF x7r in 1206 - nothing at mouser or digikey, lead time 42 weeks.
Aug 28 23:55:06 <puzzles> Capacitors are in a worldwide shortage.
Aug 28 23:59:25 <waveforms> puzzles: Why?
Aug 28 23:59:41 <puzzles> Materials shortage, I think.
Aug 28 23:59:54 <waveforms> Oh I was thinking that.
Aug 29 00:00:05 <waveforms> I hope they get that sorted soon.
Aug 29 00:00:05 <puzzles> Digi-Key put out a bulletin on it a while back.
Aug 29 00:00:12 <waveforms> Wow.
Aug 29 00:00:23 <puzzles> Me too. I make a lot of prototypes at work.
Aug 29 00:01:47 <waveforms> Yeah.
Aug 29 00:02:14 <waveforms> I'm finally working on a passion project that I feel has real potential.
Aug 29 00:03:08 <waveforms> I'm not using caps yet, but you never know when/if you'll need something when a new idea comes.
Aug 29 00:03:47 <zmatt> "many analysts do not believe that lead times will return to pre-shortage levels until 2020"
Aug 29 00:03:50 <zmatt> fun
Aug 29 00:22:50 <puzzles> Glad we have a decent stock at work!
Aug 29 00:33:32 <GenTooMan> well the ceramics require specialized materials used in them often. I'm sure the auto industry is not helping nor is the cell phone companies. They order such large quantities.
**** ENDING LOGGING AT Wed Aug 29 03:00:03 2018