**** BEGIN LOGGING AT Mon Jul 10 03:00:03 2017 Jul 10 03:00:39 Is there a book you can recommend that goes over some of the basic unix stuff like you were telling me about. I have tried to be a "unix" guy in the past, but just get bogged down in it all. I am a slow reader... :-( that is impressive. I would be into it more, but my wife keeps me from totally nerding out. Jul 10 03:01:12 hm for me it is learning by doing, aka failing upwards Jul 10 03:01:34 try some podcasts on your commute Jul 10 03:01:46 OK, you tube too? Jul 10 03:01:49 sure Jul 10 03:02:19 huh... so the LEDs are off, that means it is done? Jul 10 03:02:35 so much for 45 mins... Jul 10 03:02:41 should be, take out the sd card and start Jul 10 03:02:46 ok Jul 10 03:03:41 there are bounding happy lights... let me putty in and see the version Jul 10 03:03:52 yea Jul 10 03:05:53 i puttied in, what is a command for the version? Jul 10 03:06:31 uname -r Jul 10 03:06:59 4.4.54-ti-r93 Jul 10 03:07:46 and /cat/etc/issue Jul 10 03:07:49 sorry Jul 10 03:07:54 cat /etc/issue Jul 10 03:08:39 no worries, its all unix to me Jul 10 03:08:46 March 19th Jul 10 03:08:50 it's a winner! Jul 10 03:08:56 you might want to get comfortable with the Arch Linux wiki Jul 10 03:09:02 and it did not take 45 minutes to flash Jul 10 03:09:10 search there for all the basic stuff, it's really great Jul 10 03:10:48 wiki.archlinux.org Jul 10 03:12:14 ? Jul 10 03:13:21 yeah and with google you do e.g. "find linux version site:archlinux.org" Jul 10 03:14:12 often you get good up-to-date info Jul 10 03:15:12 joe___: lights went off? that doesn't sound good Jul 10 03:15:31 normally all leds stay solid on when the flasher is done Jul 10 03:15:32 all good it had finished flashing Jul 10 03:15:35 oh Jul 10 03:15:37 when the flash has completed, that's supposed to happen Jul 10 03:15:45 my lights also went out Jul 10 03:15:50 i think... Jul 10 03:16:05 they always went off the last 3 times I did it Jul 10 03:16:05 ... i read that somewhere Jul 10 03:16:10 hum, maybe it's been changed, odd Jul 10 03:16:31 what was the version you flashed z? Jul 10 03:17:21 I haven't used one of rcn's flashers in ages myself Jul 10 03:18:06 this was a flash from an SD card Jul 10 03:18:38 that is what a flasher is yes Jul 10 03:20:14 LoL, yeah I suck at this stuff... Jul 10 03:20:27 the debian systems we use once started as based on one of rcn's images, but that was a long time ago and not much of rcn's customizations remain :) we generally flashed via usb, although recently I also made a custom flasher card Jul 10 03:20:39 This is the uEnv from the new version --- should I connect the PRUDAQ now? https://www.irccloud.com/pastebin/lA8bEBmk/ Jul 10 03:20:41 You are right though, the guide I read says the LEDs should all light up Jul 10 03:21:01 or should I also disable HDMI first? Jul 10 03:21:27 I am now running 4.4.76-bone-rt-r18 Jul 10 03:22:02 ah yes, disable hdmi by uncommenting lines 31-32 Jul 10 03:23:55 (otherwise you can't configure the pins for use by pru) Jul 10 03:32:10 Hey guys? Jul 10 03:32:13 Anyone on here? Jul 10 03:32:30 so I downloaded the LV development env to my BBB, and looks good to go so far! Jul 10 03:32:55 hello Jul 10 03:33:02 LabVIEW Jul 10 03:33:06 ==LV Jul 10 03:34:12 Hi there joe___ Jul 10 03:34:20 Could you help me out? Jul 10 03:35:12 just state your problem so that people can read it Jul 10 03:36:14 yup, false is REALLY good at this stuff Jul 10 03:36:21 I am a EE Jul 10 03:36:33 lol no I am not I am here to get help haha Jul 10 03:39:18 it's all relative... Jul 10 03:39:21 ;-) Jul 10 03:39:54 I wonder why prudaq needs an -rt kernel though, considering the part that's actually real-time is handled by PRUSS Jul 10 03:40:42 well ok I had thought that this means PRUSS Jul 10 03:40:53 completely unrelated Jul 10 03:41:14 well ok ... will this hurt though? Jul 10 03:42:43 i'll put the prudaq on the bbb now Jul 10 03:43:14 the (much) better preemptability of -rt kernel doesn't come free, it makes the kernel as a whole less efficient Jul 10 03:43:44 got it. Once I got the board running I can change kernels Jul 10 03:43:50 so if the main job of linux is processing the large volume of data from the ADC, then yes an -rt kernel might actually hurt Jul 10 03:44:18 understood. We actually will not have that much data, but the timing is really important Jul 10 03:44:20 (rt kernels sometimes also cause kernel instability) Jul 10 03:46:07 and even with -rt and correctly configured process priorities (-rt is essentially useless if you forget about the latter), latencies on linux are still much much worse than in pruss Jul 10 03:47:12 thanks for the info! pruss all the way then. We want to record ion mobility spectrometry data ... I want to see if prudaq can do the job Jul 10 03:47:22 at least for protoytyping Jul 10 03:47:23 a while ago I did a test to measure the latency of irqs delivered to userspace (via uio) and found a 44-88 μs latency with some outliers Jul 10 03:47:44 that's not too bad Jul 10 03:47:46 switching to an -rt kernel barely improved the typical latencies but seemed to get rid of the outliers Jul 10 03:47:53 thanks! Jul 10 03:48:19 not too bad? phrased differently, that's 44000-88000 cpu cycles :P Jul 10 03:48:37 Yeah, but I had worse numbers in mind Jul 10 03:49:58 pru on the other hand, if it's waiting for an event then it can react in about 30 ns (time from input pin to output pin, mostly dominated by the signal propagation delays and i/o logic inside the SoC) Jul 10 03:50:46 thats what I want Jul 10 03:51:05 did you compare the prudaq to e.g. the PI-Plates ADC? Jul 10 03:51:30 I didn't compare prudaq to anything, I don't have a prudaq Jul 10 03:51:47 (nor an rpi) Jul 10 03:51:54 I see. Jul 10 03:55:16 pi-plates is really slow it seems Jul 10 03:55:52 and then there is http://www.waveshare.com/wiki/High-Precision_AD/DA_Board Jul 10 03:55:55 50 kHz bandwidth limit Jul 10 03:56:06 yeah Jul 10 03:57:37 everything goes via a microcontroller connected via spi to the rpi.... yeah okay Jul 10 03:58:01 yeah it's weird cannot imagine using this beyond debugging your raspi stuff I guess Jul 10 03:58:11 the nice thing about pru is that its integration into the SoC means it has awesome bandwidth to offload data Jul 10 03:58:47 oh that problem comes next yes Jul 10 03:59:07 what problem? Jul 10 03:59:29 prudaq is based on top of beaglelogic right, which already solved the data offloading Jul 10 03:59:32 ? Jul 10 04:00:44 yeah true, that's what I mean ... getting started with beaglelogic Jul 10 04:12:40 if you're using beaglelogic then I definitely don't see a reason to use an -rt kernel, all data goes from pruss to linux via a big ring buffer in memory anyway Jul 10 04:18:26 (if you need really low latency instead of high bandwidth then it would make more sense to run custom code on the PRU instead of using beaglelogic, I saw that prudaq has example code for that too) Jul 10 04:21:32 afk Jul 10 04:23:27 that makes a lot of sense Jul 10 05:24:46 Heyo. So here's my problem now. I've successfully downloaded the .img for my BeagleBoard Xm and now I've run putty to access it and an error popped up like this: Jul 10 05:24:48 https://usercontent.irccloud-cdn.com/file/9W8uUNZC/image.png Jul 10 05:34:31 how did you write the image to your SD card? Jul 10 05:34:58 and which image did you download exactly? Jul 10 05:39:59 Heyo. I downloaded this one: Debian 8.6 2016-11-06 4GB SD LXQT image for BeagleBoard-xM Jul 10 05:40:15 I used Win32 DiskImager. Jul 10 05:41:05 The BeagleBoard I'm using is an xm, it's the red square one ... right? Jul 10 05:41:57 ok so with the bone kernel, I still will have to install the stuff in https://github.com/beagleboard/am335x_pru_package, right? Jul 10 05:56:00 Nitro525_: xm is red and square, yes Jul 10 05:58:49 fasle: that repo contains the pru assembler (pasm), a library for using uio_pruss (libprussdrv), examples and docs. none of which really have to be *installed* as such :) Jul 10 05:59:32 ok just stuck a little with https://www.irccloud.com/pastebin/tR1MupII/ Jul 10 05:59:33 unless you need to compile code which expects libprussdrv to be installed globally Jul 10 06:00:10 the .9/slots is the old systematic right? Jul 10 06:00:15 that just means the script is obsolete garbage Jul 10 06:00:57 yeah thought so. So I just have to start the overlay myself, I guess in uEnv.txt as well Jul 10 06:01:02 no attempt should even be made to load an overlay, since you've already enabled uio_pruss via /boot/uEnv.txt and will need to do pinmux using config-pin Jul 10 06:01:33 thanks that was it Jul 10 06:01:54 so you can just setup.sh entirely Jul 10 06:02:43 *just skip Jul 10 06:07:03 I ran "sudo config-pin overlay prudac-00A0.dts" Jul 10 06:09:04 Does anyone know how I'm supposed to proceed with my problem? :/ Jul 10 06:10:19 @Nitro try to format the card again, check with a partitioning tool if you really have no other partitions on the drive? Jul 10 06:12:25 fasle: unless you first disabled cape-universal I would expect that failed? (which is why I said "no attempt should even be made to load an overlay") Jul 10 06:12:33 It's disk1 https://usercontent.irccloud-cdn.com/file/fYp97CIk/image.png Jul 10 06:13:18 Well there were no errors or anything. Sorry I thought the command would mean that it uses this overlay to automatically to the pin config Jul 10 06:13:25 let me read the docs... Jul 10 06:13:31 fasle: check kernel log? Jul 10 06:14:27 fasle: loading the overlay probably should work if cape-universal is disabled, but someone here recently had trouble with that (also with prudaq) while going the cape-universal route worked, hence my suggestion to try that Jul 10 06:17:32 Nitro525_: just to be sure you could try flashing with https://etcher.io/ ... clearly something went wrong since u-boot can't find the kernel, but I'm not sure what (and I'm not familiar at all with the xM) Jul 10 06:18:40 I wonder if it gets u-boot from sd card or from internal flash... do you get the same messages if you try to boot without sd card? Jul 10 06:19:29 Let me try booting without the sd card. Jul 10 06:20:16 Here's what popped up zmatt, https://usercontent.irccloud-cdn.com/file/qH5PJSv0/image.png Jul 10 06:21:51 ok, so u-boot definitely comes from internal storage Jul 10 06:22:27 that might be an issue, perhaps it's too old to be able to boot current images Jul 10 06:22:53 presumably there's a way to override that (similar to powering on a beaglebone with S2 button held down) Jul 10 06:23:48 all this is just a guess though Jul 10 06:28:24 So this is the desired pinmux right? https://www.irccloud.com/pastebin/hZSJ6Ufc/ Jul 10 06:33:13 are you asking if the overlay is right? presumably if it weren't then prudaq never worked and someone would have noticed Jul 10 06:35:10 (I never use raw hex values for pinmux myself since that is just plain unreadable) Jul 10 06:35:53 yeah, my problem is much simpler - I do not know how to proceed. Reading about all this. Is there are schema I can load into the config-pin tool or do I have to write that based on the pin functions in the prudaq docs? Jul 10 06:37:02 check the available pin modes for, e.g. P8.27 Jul 10 06:37:24 I think config-pin -l P8.27 or maybe -s, not sure Jul 10 06:38:24 here https://www.irccloud.com/pastebin/LfTPgZSB/ Jul 10 06:40:18 https://hastebin.com/pilucuyidi.bash Jul 10 06:40:24 would be my guess Jul 10 06:46:54 The bash script gave me errors, so I tried one of the operations manually. Same error https://www.irccloud.com/pastebin/NuNtCF1n/ Jul 10 06:47:21 but when I cd to that directory, I can cat the state file Jul 10 06:50:35 Well, this works: sudo config-pin P8.10 in- Jul 10 06:54:39 odd Jul 10 06:54:48 you did disable hdmi ? Jul 10 06:54:54 yes Jul 10 06:55:09 maybe a more recent universal overlay version? Jul 10 06:55:13 maybe check the current pinmux with my show-pins utility -> https://github.com/mvduin/bbb-pin-utils Jul 10 06:56:45 nice tool https://www.irccloud.com/pastebin/xchMMQWw/ Jul 10 06:59:13 and my current uEnv https://www.irccloud.com/pastebin/tPb1mwvL/ Jul 10 06:59:19 I'll restart the unit Jul 10 07:02:40 hdmi clearly didn't get disabled Jul 10 07:05:02 oh the disable_uboot_overlay_video=1 is not enough? Jul 10 07:07:37 I have no idea, I'm totally unfamiliar with u-boot overlays, I just followed the lead of the comments :) Jul 10 07:07:49 after a reboot, your script works without errors. The new show pins here https://www.irccloud.com/pastebin/l5OuhQMa/ Jul 10 07:09:16 yuck, cape-universal results in really messy output :) Jul 10 07:09:44 (tip: pipe output through sort to get it sorted by expander header pin) Jul 10 07:09:48 *expansion Jul 10 07:09:55 zmatt, what's an S2 button? Jul 10 07:10:07 Also I've found a post with someone having a similar problem Jul 10 07:10:16 and one user replied that the solution is with board validation? Jul 10 07:11:03 yeah just saw on your website https://www.irccloud.com/pastebin/PIQan6hn/ Jul 10 07:11:57 Nitro525_: S2 is a button on the beaglebone family of boards that makes sure ROM loads u-boot from sd card rather than internal storage... I have no idea if something similar exists on the xM or whether it's applicable here Jul 10 07:13:47 fasle: to see what I mean by "messy", this is the output I get on the beaglebone where I'm currently experimenting with pruss -> https://pastebin.com/raw/Xj4yhwtK Jul 10 07:14:23 yes. pretty :) Jul 10 07:15:05 (the difference is because cape-universal claims *all* pins, and uses pointlessly verbose DT node names as usual) Jul 10 07:16:14 (and requests all gpios) Jul 10 07:18:31 cat /sys/devices/platform/bone_capemgr/slots has all 4 slots empty so far, correct? Jul 10 07:18:58 (the HDMI ------L is absent on each of them) Jul 10 07:20:34 I really don't get why they give pinmux nodes names like "pinmux_P8_39_pruin_pin". the "pinmux_" prefix and the "_pin" suffix both redundant since you already know they're pinmux nodes since they're child nodes of the pinmux controller node... Jul 10 07:20:50 Hey zmatt, the beagleboard isn't part of the beaglebone family I think. There are only 2 buttons here, user and reset. Jul 10 07:21:58 Nitro525_: I know it isn't. I just referred to that particular *functionality*. note again that I'm *not* familiar with the xM Jul 10 07:22:12 Oh oh. Alright alright, thanks. Jul 10 07:23:10 fasle: hdmi was there before? odd Jul 10 07:23:38 no sorry just wanted to clarify that it is off in the current config Jul 10 07:24:00 fasle: I would expect that u-boot overlays don't show up in slots Jul 10 07:24:08 yeah ok Jul 10 08:57:12 fasle: please don't send private messages Jul 10 10:44:03 do people put heatsinks on their BBBs? Jul 10 10:57:14 I haven't heard of it -yet- .. Jul 10 11:28:53 Hi All, I'm using bb-kernel build script , how do I apply my changes to the code regardless of the current code at remote linux repository? Jul 10 14:51:15 Friends, I just loaded the latest Debian (2017-07-01) on my BBB C. When I run sudo node-red, node-red locks-up after loading palette node message. Any ideas? Jul 10 14:53:19 hi here Jul 10 14:58:23 o/ Jul 10 18:00:11 hi, having some trouble connecting to Wifi with BBBW. can anyone help me please? Jul 10 18:43:29 Has anyone used yocto/poky to build pasm so they could then use it to build an PRUSS app? Jul 10 19:16:42 jstashluk: why use yocto to build pasm? Jul 10 19:24:21 s/to build pasm?//g Jul 10 19:24:37 oops, ? still needed... Jul 10 20:37:06 zmatt: Is there a better way to install and let Yocto know where the pasm executable is located? Jul 11 01:54:12 Good morning BeaglePeeps. Jul 11 01:55:53 I've figured out that the model of my BeagleBoard It's RevC2. Jul 11 01:56:07 But the materials are outdated, links to websites are gone and I don't know where to get its software? Jul 11 02:01:19 beagleboard.org Jul 11 02:03:58 https://beagleboard.org/latest-images **** ENDING LOGGING AT Tue Jul 11 03:00:03 2017