**** BEGIN LOGGING AT Thu Dec 31 02:59:58 2015 Dec 31 05:18:14 yeah wow by bbb just seemed to die too for no reason. i booted it up, cloud9 didn't started, so i typed reboot in the console, and now it wont boot.... Dec 31 05:22:15 sometimes hitting the reset button *works* aka turns the user LEDs on.. but it ends up with them all turning except for the cpu light which flashes faintly Dec 31 05:26:57 that's so weird the only thing I changed from the last successful boot was adding a workspace directory for cloud9 which didn't want to boot since the folder wasn't there Dec 31 06:33:51 are you able to ssh in? Dec 31 08:53:04 7 Dec 31 08:57:59 8 Dec 31 09:06:10 9 Dec 31 13:42:10 tbr: can't even ping, i'm gonna flash it Dec 31 17:28:04 I've got some BBBs that are randomly hard crashing without any trace. No panics, in the logs. No panics on the serial console (logged on a separate computer). Dec 31 17:28:28 The kernel is vanilla 4.1.10 from the linux-stable git repo. Dec 31 17:29:01 What (if any) are my options for getting a stack trace on the crash? Dec 31 17:29:55 why aren't you using a beagle kernel? Dec 31 17:30:41 I'm on the cusp of just removing things from the linux config until it runs stable -- but that doesn't me much towards a "why" Dec 31 17:30:49 I don't think the patches are major Dec 31 17:31:18 I'd be up for doing a comparison, but I was also having this issue with the bbb kernel many months back. Dec 31 17:31:48 hmm Dec 31 17:32:06 I've had a mainline-derived kernel from the 3.14 days and my uptime is .. wait .. Dec 31 17:32:10 *Most* of my reasons for going vanilla were driven by the mess that was the BBB device tree directory (arch/arm/boot/dts/*) Dec 31 17:32:26 [ 7:17am] root@beagleblack ~ # uptime Dec 31 17:32:26 17:32:13 up 247 days, 19:20, 2 users, load average: 0.41, 0.24, 0.12 Dec 31 17:32:26 [ 5:32pm] root@beagleblack ~ # uname -a Dec 31 17:32:26 Linux beagleblack 3.13.6-bone7.2 #1 SMP Fri Mar 21 03:59:54 GMT 2014 armv7l GNU/Linux Dec 31 17:32:26 [ 5:32pm] root@beagleblack ~ # Dec 31 17:33:00 that's based on the RN kernels .. bit bloaty but mainline-derived Dec 31 17:33:21 nothing stopping you paring that down, mind Dec 31 17:33:22 Also, last I used that repo, it was being constantly rebased on the remote and made it a mess to deal with Dec 31 17:33:47 where are those patches pulled from? I'm assuming from TI somewhere/how Dec 31 17:33:48 well the linus tree is rebased periodically, isn't it? Dec 31 17:34:11 perhaps not given its situation Dec 31 17:34:16 never in my experience (both linux, and linux-stable) Dec 31 17:34:18 but its not alien to git Dec 31 17:34:35 * veremit nudges rcn-ee Dec 31 17:34:53 rcn-ee: Early happy new years :) Dec 31 17:35:02 happy new year veremit Dec 31 17:35:31 near 7hr to go this end lol. how goes? Dec 31 17:36:01 veremit, pretty good, i have that banana pc, but work keeps adding more projects. ;) Dec 31 17:36:16 damnit! you're not working today? Dec 31 17:36:37 the banana pi wasn't too bad to get going iirc Dec 31 17:36:54 now that the ddram clock is better. ;) Dec 31 17:36:57 but those sunxi chips do have a strange boot process Dec 31 17:37:11 ah I saw patches for the orange pi :p Dec 31 17:37:28 I wanna see a lemon pi next .. lol :) Dec 31 17:37:58 anyhow .. was draggin ya into the discussion about kernels .. since that's your thing .. :p Dec 31 17:38:22 one of D: Dec 31 17:38:46 yeah, ment orange ;) Dec 31 17:38:52 with the h3.. Dec 31 17:38:54 figured so Dec 31 17:38:56 yup Dec 31 17:39:00 not a bad chip Dec 31 17:39:21 runs desktop quite nice .. have you spun it up yet? Dec 31 17:39:47 I think I'm running the deb9 desktop from that guys page on the forums .. did I link ya before? Dec 31 17:40:05 done a bunch of them via buildroot Dec 31 17:41:26 kaak: no joy with https://github.com/beagleboard/linux ? Dec 31 17:41:41 I'mlookin up the TI one .. but I think beagle is derived from that anyway Dec 31 17:43:00 wow found an ancient openembedded version on elinux .. lol Dec 31 17:45:14 I've found it much easier to work in linux-mainline/stable all-around. From branching, to configuring, to modifying, etc. My kernel is part of a CI/CD process and maintainability (from all aspects) is a large part of that process Dec 31 17:45:32 ah Dec 31 17:45:49 you probably wanna grab/monitor patches yourself then Dec 31 17:46:51 you're not getting an overheating problem, perhaps? Dec 31 17:46:57 Yeah, I think that is slowly becoming the answer -- but I'm not sure I want to admit that to myself just yet :`( Dec 31 17:47:00 nothing should stall the kernel without a dump Dec 31 17:47:25 smells like hardware hang of some kind Dec 31 17:47:34 which could be related to clocks or all sorts Dec 31 17:47:52 It is inside an all metal enclosure. But I've also seen this during environmental tests that run between -40c and 80c Dec 31 17:48:18 kaak.. that's one instance where handing it over to a 3rd party reduces the work .. theoretically you should still ahve some traceability though Dec 31 17:48:50 hmm .. stock bbb .. not one of the 'industrial' variants/custom board? Dec 31 17:49:08 industrial from circuitco Dec 31 17:49:21 ok Dec 31 17:49:33 had to ask lol Dec 31 17:50:34 Where are the TI patches published? Dec 31 17:50:50 kaak, pure 4.1.10? Dec 31 17:51:29 rcn-ee: yeah. with some minor tweaks to the config Dec 31 17:51:39 thinks kaak, should retest with 4.1.11, which "includes" https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/commit/?id=29a6161885e2fe7638e271a406521ba9ddd9e89c Dec 31 17:51:54 which fixes the nasty... "4.1.x randomly reboots for no reason bug" ;) Dec 31 17:52:28 musb .. our friend! lol Dec 31 17:53:01 rcn-ee: :) -- I've been keeping tabs on the git log for the musb subdirectory -- hoping to catch a glimmer of hope about its stability issues Dec 31 17:53:42 are there any email lists, or forums where the musb devel collaborates at about issues and potential fixes? Dec 31 17:54:04 kaak, here or the beagleboard user group.. Dec 31 17:54:45 kaak, since we fixed that bug in mainline and backported it, i haven't seen any random reboots.. (well my testers only run for 7 days between reboots, (for the next rcX)) Dec 31 17:54:50 http://rcn-ee.homeip.net:81/farm/uptime/test-bbb-1.log Dec 31 17:55:47 While on the topic, I found this tool very nice for tracking downtime events: https://launchpad.net/downtimed Dec 31 17:56:09 i'm just calling uptime -p ;) Dec 31 17:56:16 looks interesting.. ;) Dec 31 17:57:16 kaak, how does downtimed fair with the broken rtc clock on bootup? do you still get usable results. ? Dec 31 17:57:35 I've been using it for tracking uptime and downtimes and using the data to get uptime percentages for tracking problems Dec 31 18:00:12 I've got an RTC onboard, so it may be less useful in general -- but It works by just touching files (atimes) in its lib directory Dec 31 18:01:23 rcn-ee .. if you find a "cheap" arm board with onboard rtc .. I'd like to know about it... Dec 31 18:01:51 although at the moment .. I'll probably just fling an old atom at the task I want .. Dec 31 18:02:04 i really just need to find a cheap i2c/rtc/battery module and order about 100 of them.. for all my arm's .;) Dec 31 18:02:16 it's built in tools generate logs like so: http://pastebin.com/8gxNXPLz -- and it's database is easily parsable for more custom uses and calculations Dec 31 18:02:30 rcn-ee .. there's loads from china .. I bought two different types Dec 31 18:02:40 one for Pi, one for others Dec 31 18:02:48 * veremit rummages the desk Dec 31 18:02:59 only a couple of $ each Dec 31 18:04:20 I've been pretty happy with the ds1307 from maxim. I just disable the built in one, and added it in the device tree for i2c-2 and my devices always come up with the right time. Dec 31 18:04:24 thanks veremit didn't think about hte pi ;) Dec 31 18:04:37 I grab time series data with them, so having the correct time is pretty critical for me Dec 31 18:06:24 rcn-ee.. here's one type .. http://www.miniinthebox.com/i2c-ds1307-real-time-clock-module-for-arduino-tiny-rtc-2560-uno-r3_p508780.html?litb_from=sysmail Dec 31 18:07:01 veremit, there's some really cheap ones ;) http://www.gearbest.com/boards-shields/pp_248286.html?currency=USD&gclid=CjwKEAiAwZO0BRDvxs_1w-qFnhkSJABo10ggCSoIS8-o9F2e_jEvnsPxW-ErOTv2YEbeSfv94RLrshoC_Lvw_wcB Dec 31 18:07:43 here's the other I bought .. works fine witht the Pi .. http://www.banggood.com/RTC-Real-Time-Clock-Module-For-Raspberry-Pi-p-965976.html Dec 31 18:08:32 exactly Dec 31 18:08:47 the pi one plugs directly onto the expansion header Dec 31 18:09:14 kaak .. the new upower ones last much longer Dec 31 18:09:32 depends whether you're on battery or power. Dec 31 18:09:55 coin cell modules, better to get the newer lower power chips Dec 31 18:10:21 they're down to some stupid uA->pA consumption Dec 31 18:10:28 nA** Dec 31 18:10:35 pA would be mental lol Dec 31 18:12:45 I've got a coin cell (~10mAh) and based off the data in the sheet, I calculated ~5 years run time off the battery. It's battery powered state was something like 10-50nA Dec 31 18:14:08 not bad :) Dec 31 18:14:41 I got a ds1307 in a breadboard here Dec 31 18:14:57 should tag it to the atmega thats next to it .. lol Dec 31 18:16:28 just looked it up again, its data-retention current draw is given as 10 - 100 nA Dec 31 18:17:38 yea its not far different to the ds3231 Dec 31 19:38:11 kaak is that the max or typical value? Use the max value if you want to know. You also need the draw over temperature. Dec 31 19:40:53 GenTooMan: min and max -- I just went with the max for all my estimates and applied a hot temperature de-rating. Dec 31 19:41:25 It came back with multiple years of off-run-time, and I was pretty happy with that for my application Dec 31 19:41:54 kaak indeed that's what you have to do. Dec 31 19:42:38 always use worst case or you will always be disappointed. Dec 31 20:34:53 hello Dec 31 20:35:24 any electrical engineers in the house? Dec 31 20:36:37 anyways I was wondering , I am an electrical eng. doing mostly hardware validation and looking to get into embedded programming Dec 31 20:37:11 do you guys think doing a project with a beagle board can give me leverage in the industry? Dec 31 20:38:54 to quote tbr: http://www.catb.org/esr/faqs/smart-questions.html Dec 31 20:39:19 \o/ Dec 31 20:39:31 :] Dec 31 20:40:46 * tbr totally thinks it can leverage the IoT synergy with embedded application development agility Dec 31 20:41:05 IoT == twitter.com/IoToilets Dec 31 20:41:29 uhoh /o\ Dec 31 20:42:28 it's "da shit" Dec 31 20:43:40 or not Dec 31 22:52:05 Hi all - are there any good pointes to overcoming issues of flashing my bb green? I'm trying to flash the emmc, only it won't take... Dec 31 22:52:37 what image are you using? Dec 31 22:52:54 snappy ubuntu... Dec 31 22:52:56 the lateeset Dec 31 22:53:44 ah cant vouch for that Dec 31 22:54:05 the debian ones are good I know that .. so you could do a test-write with one of those first .. Dec 31 22:54:08 I suspect the ubuntu image :/ Dec 31 22:54:16 ok Dec 31 22:54:33 have you done a md5sum/etc on the downloaded file to check its complete? Dec 31 22:55:00 then its just a matter of dd'ing/etc to a uSD successfully and , Boot. Dec 31 22:55:18 theoretically :) Dec 31 22:55:34 yep, I can boot from the ubuntu written sd... Dec 31 22:55:43 it just won't flash the emmc Dec 31 22:56:06 does it have a script or somesuch? Dec 31 22:56:29 the flasher images I've always used, just blow the image straight in Dec 31 22:56:50 although, granted, that's just a script too :) Dec 31 22:57:07 yep - beagleboard.org/latest-images talks about the /boot/uEnv.txt file Dec 31 22:57:15 I added one... with the uncommented code... Dec 31 22:57:19 still no luck Dec 31 22:57:58 hmm .. if you cat /proc/cmdline does it pick up the changes? or is it the wrong uEnv.txt .. Dec 31 22:58:13 (you'd be surprised lol) Dec 31 22:59:16 not sure what you mean by that Dec 31 23:05:07 in some cases, there's a dummy uEnv.txt .. and /boot/ isn't the actual boot device .. so just check what's mounted and where :) Dec 31 23:05:34 historically there have been some fudges to the boot sequences Jan 01 01:47:38 Hi. I've imaged an sd with the wheezy debian 7.9 and when I boot from it all I get is a mouse on a black screen. Anything I should look at or just go back a version? Jan 01 01:49:39 well you get a mouse .. so that's something .. **** ENDING LOGGING AT Fri Jan 01 02:59:58 2016