**** BEGIN LOGGING AT Wed Jul 05 03:00:06 2017 Jul 05 09:11:03 hi, i have a problem with initialising wifi on my BBB. The usb wifi adapter works and is correctly configured but there appears to be a race condition when the device starts up which means that sometimes the wlan0 interface fails to be initialised. Jul 05 09:12:51 i set up the wifi using connmanctl and i think it's conflicting with another system maybe systemd or udev Jul 05 09:13:15 has anyone else experienced this or anyone who knows how to find out what's going on? Jul 05 09:21:46 Hi, I have an original BeagleBoard B6 which is not powering on correctly as far as i can tell Jul 05 09:22:03 The power light comes on but non of the other LEDs come on when plugged in Jul 05 09:22:17 Does anyone have any ideas as to what could be wrong? Jul 05 09:37:24 Hi, I have a BBB wireless. After some problems I have changed a line in the eMMC, the problem now is that the BBB seems to flash in loop, so I can not communicate with it. How to do a factory reset ? Jul 05 09:42:57 reflash using a flasher sd card Jul 05 09:44:09 spuz: I'm really not familiar with connman, but presumably you can configure it to automatically bring up the wlan interface? that's how it would be done with any other network manager at least Jul 05 09:45:13 (it should not matter whether connman starts before or after the wlan interface becomes available, any decent network manager can deal with hotplugged network interfaces) Jul 05 09:54:02 I reflash using a flasher sd cart and now it is ok, thanks ! Jul 05 11:05:35 hello Jul 05 11:09:55 Hi All, what is the alternative to the eMMC it's obsolete ,can i use MTFC4GACAJCN-1M? Jul 05 11:10:01 in BBB Jul 05 11:11:50 you already asked this a month ago, and got a detailed answer Jul 05 11:12:41 yes, but i see there is extra pin called SD, something regarding new fast mode HS400 (but it also compatible to HS200) Jul 05 11:12:49 so does it make any diffrence? Jul 05 11:13:06 neither HS200 nor HS400 is supported, but eMMC will still support the older modes as well Jul 05 11:13:06 sorry pin name DS Jul 05 11:13:33 so I can be positive that this eMMC part is compatible with old mode? Jul 05 11:13:35 so DS is not used, you can leave it not-connected Jul 05 11:13:57 zmatt: connman is already attemping to bring wlan0 up on startup, it simply fails for some reason some of the time. Other times it does come up successfully. if I try to unplug and replug the usb wifi adapter after boot, it will not recover if it failed to start initially Jul 05 11:15:30 zmatt : how can i verify this part supports the "old mode", is there any name for the old mode? Jul 05 11:15:41 HS Jul 05 11:16:08 I think... dunno actually if that is the name or even if it has a name as such Jul 05 11:16:31 spuz: sounds like connman simply sucks Jul 05 11:17:52 cyberguy_: yeah, "high speed" or "high speed SDR" to distinguish it from high speed DDR Jul 05 11:18:59 more important to check is that it supports 3.3V I/O, unless you change the schematic to use 1.8V I/O supply for the pins used for eMMC Jul 05 11:19:16 zmatt: is there any other option for setting up wifi in debian? Jul 05 11:19:27 spuz: I use systemd-networkd Jul 05 11:20:11 there are various other options Jul 05 11:20:42 (systemd-networkd requires debian stretch, the version of systemd used in jessie is too ancient) Jul 05 11:22:15 zmatt : another question , can i compile linux-net for BBB ? is there anything similar to this? Jul 05 11:23:37 sure you can try to merge that branch or cherry-pick the specific commits you need. and no I'm not in the mood to give a tutorial on how to use git Jul 05 11:30:28 so i need to merge linux-next with linux mainline? and then try to apply patches Jul 05 11:30:33 ? Jul 05 11:39:57 Has anyone experienced flashing with Uniflash ? or node-beagle-boot ? Jul 05 11:57:28 Hi, I have 2 BBB wireless, and I have flashed both with the same image. The problem is that with one BBB I don't have wlan0 when ifconfig. Do you think it is a hardware problem ? Jul 05 11:58:46 zmatt: well, I tried to set up systemd and networkd and now the device fails to even boot... Jul 05 11:58:56 zmatt : are you fimiliar with a patch to remove the dependency with the BBB EEPROM ? Jul 05 11:58:59 so reflash needed I think Jul 05 12:00:06 spuz: what do you mean by "set up systemd" ? Jul 05 12:01:36 gege: some BBB wireless have misprogrammed EEPROM (production issue) making it think it's a normal BBB. do you see an eth0 interface? Jul 05 12:02:52 spuz: switching network manager is definitely something that requires a bit of care to avoid losing network connectivity altogether, though I have no idea how you managed to get it to fail to boot entirely Jul 05 12:03:01 cyberguy_: no Jul 05 12:03:39 cyberguy_: if we were to do a custom design based on the BBB I would make sure that the EEPROM is retained :P Jul 05 12:04:03 it shouldn't be hard to remove the dependency though Jul 05 12:05:31 spuz: having a serial debug cable is definitely recommended to ensure you can log in even if it has no network connectivity, although I have done the switch successfully without using the debug console Jul 05 12:09:01 spuz: when flash with the SD card the EEPROM should be the same as the EEPROM of the BBB which work isn't it ? When ifconfig I have just info about lo, usb0 and usb1. Jul 05 12:09:50 the EEPROM just stores board type and serial number and is write-protected Jul 05 12:10:05 you're confusing the EEPROM with the eMMC Jul 05 12:10:26 the EEPROM is programmed in factory Jul 05 12:11:02 Ah ok, so what should i do if the EEPROM is misprogrammed ? Jul 05 12:11:16 if you're not seeing eth0 then it's not this issue Jul 05 12:11:55 (in which case I have no idea what could be the issue) Jul 05 12:13:21 zmatt: i just followed these instructions. I have no idea why it's not booting either: https://forum.manjaro.org/t/how-to-use-systemd-networkd-to-manage-your-wifi/1557/13 Jul 05 12:13:47 when i say not booting, i meant not setting up the usb0 or usb1 networks which is how i am trying to connect to the device Jul 05 12:13:52 i don't have a serial cable Jul 05 12:14:24 gege, i'm not sure what you're asking. could you say that again? Jul 05 12:14:36 gege meant to ask me probably Jul 05 12:14:52 ah, ok Jul 05 12:15:08 oups, yes sorry Jul 05 12:15:25 usb0/usb1 might not get automagically setup anymore if you switch network manager, it depends a bit on how they are being set up (I have no idea since I don't use usb networking) Jul 05 12:17:10 they can of course be set up with appropriate config Jul 05 12:17:24 So I should have wlan0 (like on the BBB which works), or at least eth0 if the EEPROM is programmed as if it was not wireless, but in this case I don't have any of these two. Jul 05 12:17:49 gege: can you pastebin the kernel log? Jul 05 12:17:53 (output of dmesg) Jul 05 12:20:31 all the output ? Jul 05 12:21:18 well I'm guessing that if you can figure out which part is important then you already would have located the issue :P Jul 05 12:21:28 [ 0.000000] Booting Linux on physical CPU 0x0 [ 0.000000] Initializing cgroup subsys cpuset [ 0.000000] Initializing cgroup subsys cpu [ 0.000000] Initializing cgroup subsys cpuacct [ 0.000000] Linux version 4.4.54-ti-r93 (root@a4-imx6q-wandboard-2gb) (gccrsion 4.9.2 (Debian 4.9.2-10) ) #1 SMP Fri Mar 17 13:08:22 UTC 2017 [ 0.000000] CPU: ARMv7 Processor [413fc082] revision 2 (ARMv7), cr=10c5387 [ 0.000000] CPU: PI Jul 05 12:21:31 pastebin Jul 05 12:21:35 pastebin.com Jul 05 12:21:52 don't paste anything multiline in the chat, it is unreadable and just gets truncated anyway Jul 05 12:22:48 yeah ok, i didn't know pastebin, seams lot more adapted, sorry Jul 05 12:22:51 https://pastebin.com/YDw0VyjE Jul 05 12:24:03 there are tons of similar sites of course, e.g. hastebin.com and paste.debian.net ) Jul 05 12:24:50 so it knows it's a BBB wireless (line 9) Jul 05 12:26:18 btw there seems to be something wonky with your copy-paste... each line that's sufficiently long has a few chars missing Jul 05 12:26:53 columns 78-80 Jul 05 12:27:51 can you pastebin the log of one that is working okay? Jul 05 12:28:34 line 253 seems suspicious, although I recall having seen it in other people's pastes Jul 05 12:28:42 on my way Jul 05 12:28:43 (I don't have a BBB wireless so I can't check myself) Jul 05 12:33:52 sorry needed to start it again : https://pastebin.com/nUhHZ305 Jul 05 12:36:56 ok so that line is indeed present in both cases Jul 05 12:37:07 then I have no idea Jul 05 12:37:40 hum Jul 05 12:39:10 zmatt: i'm now working on the u-boot, to remove all the i2c reading from the EEPROM, can i printf there? or the console is not yet initialized? Jul 05 12:41:16 (thanks anyway for the time you've spent on it) What is the kernel log exactly ? Jul 05 12:41:42 cyberguy_: if you simply tried it then 1. you'd learn the answer 2. wouldn't waste other people's time Jul 05 12:42:17 gege: a log of messages from the kernel :) Jul 05 12:44:07 Ok ok, in fact i just didn't know the translation in english, thanks then, I continue prospecting Jul 05 12:48:24 zmatt : i'm not sure.. i thought maybe i'm doing something wrong here... because i didn't see anyting.... Jul 05 12:48:32 zmatt : how do you debug this if you can't print anything.... Jul 05 12:49:01 jtag Jul 05 12:56:18 zmatt: i've reflashed the device. can you point me to a guide for setting up usb wifi on the BBB that doesn't use connman? Jul 05 12:56:58 no idea... I've never really needed a guide so I've also never looked for one :) Jul 05 12:58:02 zmatt : jtag? what programming tool ? code composer? Jul 05 13:01:51 for example, or openocd + gdb Jul 05 13:02:20 (I'm assuming that with "programming tool" you meant debugger) Jul 05 13:03:23 so i need to assemble the jtag connector? Jul 05 13:03:51 if you want to use jtag, soldering on the jtag header would certainly be recommended yes :P Jul 05 13:04:30 what about the jtag emulator ? is it only loaded after the kernel is up? Jul 05 13:04:38 the usb jtag emulator i mean Jul 05 13:05:34 ? Jul 05 13:05:53 I have no idea what you're referring to Jul 05 13:15:25 Does someone know if there is a cooling system for beagleboard black as accessory? Jul 05 13:16:23 o.O Jul 05 13:16:38 what are you doing to your poor beaglebone to necessitate cooling? Jul 05 13:18:25 well, it is just to be in the safe side as we are going to use it in quit hot industrial environment Jul 05 13:18:35 ah ok Jul 05 13:19:25 I've never heard of a cooling system specifically for the beaglebone, but presumably there are generic solutions Jul 05 13:19:54 I've once seen tiny adhesive cpu heatsinks for the rpi... no idea if they actually do something useful though :P Jul 05 13:21:53 http://www.onemansanthology.com/blog/dfrobot-self-adhesive-pure-copper-heatsink-for-raspberry-pi/ Jul 05 13:23:13 it it compatible to bbb too? Jul 05 13:23:18 is it Jul 05 13:24:29 I mean in term of size Jul 05 13:24:43 I was about to suggest installing the appropriate drivers :P Jul 05 13:25:02 is it not plug and play ?!? Jul 05 13:25:06 :) Jul 05 13:25:45 What do you mean with appropriate drivers? Jul 05 13:26:07 I was being silly :P Jul 05 13:26:41 ah ok :) Jul 05 13:26:58 the size of the heatsink is listed, I'm sure you can determine the size of the AM3358 yourself (by using a measuring tape or checking the datasheet) Jul 05 13:27:28 note that I have no experience with this heatsink, it's just a random one I happen to know existed... I'm sure there are tons more Jul 05 13:30:03 ok, actually I thought about this solution at the beginning, I was thinking of using a fan cooled heatsink too Jul 05 13:30:28 if you really want to be sure you might need to do a thermal analysis of some sort Jul 05 13:31:14 TI also has some resources about this topic, e.g. http://processors.wiki.ti.com/index.php/AM335x_Thermal_Considerations Jul 05 13:32:57 thank you, I saw that, I was wondering if there is something already made for bbb Jul 05 14:40:59 hi Jul 05 16:13:53 hi there ! Jul 05 16:14:29 fast question, do you know of a cape providing one or more ethernet ports ? Jul 05 16:15:28 <_av500_> nope Jul 05 16:15:33 I would like to implement a VPN server with BBB, I need at least 2 ethernet ports Jul 05 16:16:18 USB Ethernet Jul 05 16:16:27 dot1q Jul 05 16:16:27 ... use-to-Ethernet is an option i considered Jul 05 16:16:32 etc etc... Jul 05 16:16:45 usb-to-Ethernet, yes Jul 05 16:17:07 if i don't find a cape i will do that Jul 05 16:22:59 dot1q ... umm, not an expert on that, if you mean to give two IP to the same Ethernet port, no, i prefer not to do that. Jul 05 16:24:08 because VPN is for a company i don't want to mix LAN and extraLAN stuff as much as I can Jul 05 16:26:52 thank you for your help, bye ! Jul 05 16:44:59 (different board...) Jul 05 16:55:31 hi zmatt Jul 05 17:28:45 * MathOnNapkins wonders if anyone is going to have to scrub the logs from yesterday Jul 05 17:49:38 * bradfa scrubs some logs using napkins, gets splinter, goes to find coffee Jul 05 18:05:57 How can I unload a device tree overlay? Jul 05 18:06:31 I can load it using "echo myname > /sys/devices/platform/bone_capemgr/slots" Jul 05 18:08:48 reboot? Jul 05 18:09:20 ds2: of course. But I thought there would be a trick to unload it without rebooting. Jul 05 18:17:26 echo -SLOT >/sys/devices/platform/bone_capemgr/slots Jul 05 18:17:29 where SLOT is the slot number Jul 05 18:18:04 zmatt: nice Jul 05 18:18:18 I recommend saving any open documents first, unloading overlays can sometimes be a bit prone to causing kernel panics Jul 05 18:19:02 unloading overlays also doesn't undo pinmux Jul 05 18:19:33 (although you can then load another overlay to change the pinmux again) Jul 05 18:24:27 hello Jul 05 18:25:17 has anyone know if it is possible to display kernel boot logs to the HDMI output with a beaglebone black? Jul 05 18:30:52 jla: well, sorta... obviously you won't have hdmi output until the appropriate driver loads Jul 05 18:32:13 and by using the console framebuffer? Jul 05 18:32:48 and compiling kernel with framebuffer lcd? Jul 05 18:32:51 yes I'm talking about the framebuffer console (which will be active by default) Jul 05 18:32:53 At the moment prussdrv_open() does not work and I don't know why. zmatt: You told me it is also possible to use PRUSS without any device overlays. Is there a how to available? Jul 05 18:33:16 compiling in the appropriate drivers may result in output sooner yes Jul 05 18:33:49 no one has tried? Jul 05 18:33:51 (tilcdc) Jul 05 18:33:55 no idea Jul 05 18:34:22 I mean, tried what? Jul 05 18:34:52 NTQ: have you gotten rid of the big honkin' errors in your kernel log? Jul 05 18:34:55 to output kernel log to the HDMI ouput Jul 05 18:35:16 zmatt: looks clean now Jul 05 18:35:52 jla: kernel log will automatically go there once the drivers for tilcdc and hdmi load. the default log level may suppress most log messages though, you can tweak that using kernel parameters Jul 05 18:36:04 omitting the "quiet" parameter will already help a lot probably Jul 05 18:36:18 (it may however also slow down boot considerably) Jul 05 18:36:43 NTQ: can you pastebin your current /boot/uEnv.txt ? Jul 05 18:37:31 ok thank you Jul 05 18:37:53 zmatt: of course: https://hastebin.com/utonodomam.ini Jul 05 18:39:02 zmatt: And dmesg: https://hastebin.com/sisuzexaxe.vbs Jul 05 18:39:12 I was about to ask :D Jul 05 18:39:30 I know what you want :-D Jul 05 18:39:33 lol, hastebin is failing nicely at syntax guessing Jul 05 18:39:51 yeah Jul 05 18:39:54 tip about hastebin: you can get it to use different syntax highlighting by changing the suffix Jul 05 18:40:02 e.g. https://hastebin.com/sisuzexaxe.txt has no highlighting Jul 05 18:40:08 nice Jul 05 18:42:14 odd, it kinda looks like u-boot overlays aren't working at all Jul 05 18:42:50 wait, you're running from sd card? Jul 05 18:44:54 can you try powering on while holding the S2 button down? Jul 05 18:46:13 yes Jul 05 18:46:53 (and then check dmesg again, or lsmod) Jul 05 18:47:19 I thought pressing that button changes between onboard flash and sd card Jul 05 18:47:39 it removes eMMC from the boot order Jul 05 18:48:10 by "boot order" I mean where bootrom will try to locate u-boot Jul 05 18:48:19 hasta la vista: https://hastebin.com/rafinalici.txt Jul 05 18:49:43 haha, now I get a bus error. I was not aware of that behaviour with eMMC and device overlays Jul 05 18:49:47 uhh Jul 05 18:49:51 do you have a BBG or BBB ? Jul 05 18:49:55 BBG Jul 05 18:50:15 compare line 8 of your two pastes Jul 05 18:50:36 any chance the u-boot on your sd card is ancient? Jul 05 18:50:46 wait, this is something we can easily check Jul 05 18:50:50 strange Jul 05 18:52:05 dd if=/dev/mmcblk0 skip=768 count=1 | strings Jul 05 18:52:19 should probably dredge up the u-boot version on your sd card Jul 05 18:52:49 zmatt: U-Boot 2017.07-rc1-00002-g4d87f7 Jul 05 18:52:57 same with mmcblk1 for the u-boot version on eMMC (which is what is actually used if you power on without holding down S2) Jul 05 18:53:38 mmcblk1: U-Boot 2016.03-00001-gd12d09f fo Jul 05 18:53:54 ok so that one is definitely too old to support overlays Jul 05 18:54:26 I'm not sure about the 2017.07-rc1 .. my gut feeling is that it ought to be recent enough Jul 05 18:54:35 what I don't understand is why it seems to be misidentifying your hardware Jul 05 18:55:49 oh wait 2017.07 is right now, lol, yes that will be recent eonugh Jul 05 18:55:51 *enough Jul 05 18:56:10 ;-) Jul 05 18:57:15 I think I need a beginner course to understand all these things that happen on my bbg. It's very interesting but I want to understand what is going on. Most websites only come only with simple examples. Jul 05 18:57:44 ^^ Jul 05 18:58:33 to be fair, odd things seem to be happening on your system Jul 05 18:59:18 I don't suppose you have a serial debug cable to see what u-boot is printing? Jul 05 18:59:56 unfortunality not. Jul 05 19:01:15 can you pastebin the output of lsmod ? Jul 05 19:01:21 I have a serial adapter to usb but I that's not a good one. Jul 05 19:01:38 http://paste.debian.net/974996/ Jul 05 19:02:29 uio_pruss did load.... Jul 05 19:02:37 do you have any /dev/uio* devices? Jul 05 19:03:11 yes, uio0 to uio7 Jul 05 19:03:38 sounds like it should actually be working now Jul 05 19:04:01 (i.e. the bus error might be a software issue) Jul 05 19:04:14 Maybe from the dts file? Jul 05 19:04:28 or simply from the pruss oder host code? Jul 05 19:05:14 I suspect line 8 might just be misleading due to u-boot overlays, I don't see any other evidence that it's mistaken your green for a black Jul 05 19:05:46 i.e. if it did then there would be log messages about lcdc Jul 05 19:06:05 so I think the u-boot on your SD card is working fine Jul 05 19:06:42 ok Jul 05 19:07:00 to avoid having to keep S2 pressed on every power-up you can either upgrade or wipe the u-boot on eMMC Jul 05 19:07:46 (if wiped then bootrom will check sd card as alternative) Jul 05 19:07:57 Oh, I get a conflict on P8.30 in dmesg Jul 05 19:08:00 maybe I wipe it. Jul 05 19:09:14 this suffices to "wipe" u-boot from eMMC (or rather, render it unrecognizable for bootrom): dd if=/dev/zero of=/dev/mmcblk1 seek=256 count=1 Jul 05 19:09:47 if you don't care about any of the content on eMMC anymore you can also simply wipe it entirely with: blkdiscard /dev/mmcblk1 Jul 05 19:10:56 can you pastebin the dmesg with pin conflict message? Jul 05 19:11:09 I saw one in the first log you gave but that was because you tried to load the overlay twice Jul 05 19:11:41 this is the whole dmesg: http://paste.debian.net/974997/ Jul 05 19:12:15 yep, you tried to load it twice again Jul 05 19:12:17 Yes, sometimes I load it twice because I do not know if I loaded it already. Jul 05 19:12:34 well that's what the pin conflict is complaining about Jul 05 19:12:39 ok Jul 05 19:12:50 I thought it was something about this bus error Jul 05 19:13:00 nope Jul 05 19:13:01 After some reboots there is no bus error. Jul 05 19:13:06 that is weird Jul 05 19:13:08 That's weird Jul 05 19:13:17 ^^ Jul 05 19:13:26 might also be related with the unbalanced pm_runtime calls, although I'm not sure why Jul 05 19:13:35 I actually sent rcn patches for that today Jul 05 19:14:23 "platform 4a300000.pruss: enabled after unload, idling" ... that's also sounding really weird Jul 05 19:14:53 enabled after unload? why is it unloading? Jul 05 19:15:07 Why are there no other people complaining about this? Jul 05 19:15:09 based on timestamps that's happening when you're loading that overlay Jul 05 19:15:49 I hate to suggest this, but maybe consider enabling cape-universal and using config-pin to setup your pins instead of using an overlay? Jul 05 19:16:17 ah, yes. config-pin is the magic word. I will google examples for this Jul 05 19:17:27 I think it has --help Jul 05 19:18:25 Do I have to change anythin for this in uEnv.txt? Jul 05 19:18:45 uncomment line 42 (enable_uboot_cape_universal=1) Jul 05 19:20:16 do I have to disable other lines? Jul 05 19:20:29 nope Jul 05 19:21:51 So I can set the pin modes of the pins I want to use in my PRU source like in the dts file and then it's working? Jul 05 19:22:06 you set the pin modes using config-pin Jul 05 19:22:31 of course Jul 05 19:22:40 (instead of using an overlay) Jul 05 19:22:46 apart from that there's no difference Jul 05 19:22:51 cool Jul 05 19:23:50 Who are pins named? like P8_30? Jul 05 19:23:56 *how Jul 05 19:25:44 For example when I want to use pr1_pru1_pru_r31_6: Is that correct? "config-pin P8_39 pruin" Jul 05 19:26:04 I've never used config-pin Jul 05 19:26:15 oh ok Jul 05 19:26:32 based on a quick glance at its source it seems to be flexible Jul 05 19:26:47 P8_39 or P8.39 should work Jul 05 19:27:36 30 I mean Jul 05 19:27:43 or whatever :P Jul 05 19:29:42 nice Jul 05 19:34:22 Can the pru also change pin modes at runtime? Jul 05 19:35:29 nope Jul 05 19:35:47 although the edio peripheral in pruss can afaik do bidirectional gpio Jul 05 19:37:04 wow. it's working Jul 05 19:37:08 that won't be quite as efficient as the direct pru i/o though (but probably still faster than using normal gpio) Jul 05 19:37:10 no bus error Jul 05 19:37:12 \o/ Jul 05 19:37:16 woo Jul 05 19:39:26 \o/ Jul 05 19:49:50 Now I finally can modify my fpga project to work with the beaglebone Jul 05 21:19:43 In the debian image: why does /etc/apt-apt.conf.d/03-proxy-oracle exist? This seems very wrong. Jul 05 21:20:36 why does /etc/apt-apt.conf.d/03-proxy-https exist? This also seems very wrong Jul 06 02:11:00 * plaw high 5's pdp7 **** ENDING LOGGING AT Thu Jul 06 03:00:02 2017