**** BEGIN LOGGING AT Wed Oct 14 02:59:59 2015 Oct 14 03:06:28 dhirc: I'd ask on E2E http://e2e.ti.com/support/arm/sitara_arm/ Oct 14 03:17:37 do the rcn console images have the usb network stuff? Oct 14 04:21:15 ? Oct 14 04:37:26 I think my bbb may be haunted... Oct 14 05:34:26 ugh no wicd in the rcn debian image lol Oct 14 05:59:53 hi, I am facing issues in booting beaglebone with custom linux kernel image. Oct 14 06:00:41 I am able to bootup with prebuilt debian img(SD card) Oct 14 06:01:08 After that, i deleted prebuilt kernel in rootfs and replaced with my custom zImage. Oct 14 06:01:36 But i see that board is still booting with prebuilt img only. any pointers will be of great help Oct 14 06:02:19 santosh_ari: do you have access to the debug UART? Oct 14 06:02:38 yes, i am able to see the boot logs Oct 14 06:03:16 i have verified linux version with uname -a Oct 14 06:07:40 as I'm not sure how your image looks, you should verify if it's also on the FAT partition Oct 14 06:15:36 it is ext4 (similar to prebuilt debian img) Oct 14 06:16:25 boot log says, current device is mm0, error ready uenvcmd Oct 14 06:17:05 sorry.. ... boot log says, current device is mm1, error reading uenvcmd Oct 14 06:17:22 so switching partition. Current device is mmc0 Oct 14 06:18:02 apart from replacing my zimage, should i make any changes in uEnv file? Oct 14 06:22:10 ugh wtf is loading cape-universal Oct 14 06:27:40 I don't know. I build my kernels and have them packaged as deb packages. That takes care of updating relevant files Oct 14 06:31:03 https://github.com/RobertCNelson/boot-scripts/blob/master/boot/am335x_evm.sh#L213 Oct 14 06:31:06 no idea what calls it though Oct 14 06:31:13 but loading capes by default is annoying Oct 14 06:33:01 i dont get the point :/ Oct 14 06:33:55 tbr: i see, can you pls share the link about how to create a deb package. Oct 14 06:35:28 should be this one https://eewiki.net/display/linuxonarm/BeagleBone+Black#BeagleBoneBlack-LinuxKernel Oct 14 06:36:08 in the checkout there are probably explanations how to do it Oct 14 06:47:46 tbr: ok .. thankyou! Oct 14 08:24:45 nyt: where is the video?? Oct 14 12:20:33 Not that there's really any new information, but http://beagleboard.org/blog/2015-10-14-beagleboard-x15/ Oct 14 12:20:35 should be publishing the schematics and layout momentarily. Oct 14 12:21:48 nice Oct 14 12:22:13 +1 Oct 14 12:22:17 so far not many people who ask questions about it, but I'd expect that to go up Oct 14 13:04:38 BeagleBoard-X15 BOM and schematics link Oct 14 13:05:16 can anybody give the link of BeagleBoard-X15 BOM and schematics Oct 14 13:05:37 I'm waiting on a slightly cleaned-up version still :'( Oct 14 13:06:17 can you give the link of BeagleBoard-X15 BOM and schematics Oct 14 13:06:38 which part of not available yet, but soon, did you not understand? Oct 14 13:08:08 I just had the breaks put on the version I was about to release this morning, but the PR still went out. :( Oct 14 13:15:15 Hi! I am having trouble accessing my board through the USB. Oct 14 13:16:21 I am running Linux and my BBB is running the Debian that came with it. Oct 14 13:20:11 xikuuky: do you see the device appear over USB? Oct 14 13:21:19 Yes, I can access the START.htm page. Oct 14 13:22:19 ok, so what is the exact nature of your problem? Oct 14 13:22:59 Well, my browser won't load anything on the 192.168.7.2 page. No cloud9IDE, nothing. Oct 14 13:24:59 Extra info, I have installed the extra dev rules. Oct 14 13:25:23 are you sure it's not your browser? adblocker? noscript? etc Oct 14 13:26:55 Yeah, I've done this before. This problem is mind boggling to me. Oct 14 13:27:31 btw: if you're still on the original image, then updating that might be a good idea Oct 14 13:27:37 latest stable image is from march Oct 14 13:31:42 Alas I have no µSD. Oct 14 13:32:44 buy one :) Oct 14 13:32:51 4GB is cheap Oct 14 13:34:07 "the BeagleBoard-X15 provides nearly four three times the performance per core" Oct 14 13:34:23 They're not available in my country. Oct 14 13:35:14 nyt: anything about performance per watt? Oct 14 13:35:48 i was just copy/pasting from jkridner's link, grammar issue there Oct 14 13:35:54 Defiant: Each core is like a board increasing performance. Oct 14 13:44:50 jkridner, doesn't cape-universaln loading mess with your bonescript stuf? Oct 14 14:08:36 hi guys Oct 14 14:09:20 does any one using Debian with kernel 4.2.3 on his BBB? Oct 14 14:10:06 because i'm having some problem with the wifi usb Oct 14 14:10:20 what kind of trouble? Oct 14 14:10:29 hrm not with WiFi.. but shoot with your errors or issues and we might be able to help Oct 14 14:10:50 1 ) if i use an usb hub i don't reconize the usb wifi Oct 14 14:11:03 you giving it enough voltage/current? Oct 14 14:11:06 2) if i connect direclty to the usb port Oct 14 14:11:09 ie powered Oct 14 14:11:15 yep Oct 14 14:11:29 does lsusb even see a device on the port? Oct 14 14:11:30 i have power to bbb and tu usb port Oct 14 14:11:33 5v Oct 14 14:11:35 recognized or not Oct 14 14:12:07 yes with lsusb is can see the hub and othe usb connected to him Oct 14 14:12:18 but not the wifi Oct 14 14:12:40 then it sounds like you need a module compiled or added to the kernel Oct 14 14:12:56 just at first thought Oct 14 14:13:05 http://gowthamgowtham.blogspot.co.at/2013/11/mediatekralink-wifi-adapter-in.html Oct 14 14:13:11 i had follow this tutorial Oct 14 14:13:30 is it text only (i'm non-graphical atm) Oct 14 14:13:32 ? Oct 14 14:13:33 and also i had see that with this new kernel Oct 14 14:14:22 I had trouble with my sdio based r8723bs device on 4.2.3 at first because of some firmware blobs Oct 14 14:14:36 does that dongle use external firmware? Oct 14 14:14:54 yes Oct 14 14:15:05 but i see on debian reference Oct 14 14:15:27 that now they have had this ralink driver Oct 14 14:15:55 https://packages.debian.org/wheezy/firmware-ralink Oct 14 14:15:57 this one Oct 14 14:16:19 have you tried adding the blob into the kernel staticly? Oct 14 14:16:35 sry i don't know what is it blob Oct 14 14:16:47 the firmware binary Oct 14 14:16:57 ie ra5766.bin Oct 14 14:17:00 etc etc Oct 14 14:17:03 ok Oct 14 14:17:18 no i had use the apitude install Oct 14 14:17:21 ahh Oct 14 14:17:46 and also Oct 14 14:17:49 I usually build my own kernels so i won't be as helpful Oct 14 14:17:57 ahh Oct 14 14:18:13 i like keeping them trimmed down Oct 14 14:18:36 i'm keep learning stuff :D Oct 14 14:18:49 good Oct 14 14:18:54 one day may be i will be able to follow your way Oct 14 14:19:16 I've been building kernels since 96 so I've been at it a while Oct 14 14:19:18 er 94 Oct 14 14:19:24 wow i cant type today Oct 14 14:19:30 eheh Oct 14 14:19:35 but for example Oct 14 14:19:43 if i have a driver installed Oct 14 14:19:48 and i plug the usb Oct 14 14:20:02 and is not reconized by the sistem Oct 14 14:20:13 it's means that the driver are not good right? Oct 14 14:20:16 some work without external firmware.. some need it Oct 14 14:20:28 it all depends on hardware, driver, and udev Oct 14 14:20:36 and kernel infrastructure Oct 14 14:20:46 because if i run lsmod Oct 14 14:21:02 ,if i understand if for check the installed driver, Oct 14 14:21:13 if you run ifconfig -a do you see an interface for it Oct 14 14:21:21 no Oct 14 14:21:26 i don't see it Oct 14 14:21:26 and does dmesg | grep -i ralink show anything Oct 14 14:21:42 no Oct 14 14:22:07 it mean that are not installed right? Oct 14 14:22:18 loaded Oct 14 14:22:36 the drivers could be there and just not loading Oct 14 14:22:52 wyhats that url again? let me read it for a min Oct 14 14:22:59 Elray, with ralink, most often the kernel just missing your device's new "pid/vid"... Oct 14 14:23:16 http://gowthamgowtham.blogspot.co.at/2013/11/mediatekralink-wifi-adapter-in.html Oct 14 14:23:22 rcn-ee: yeah those are annoying drivers Oct 14 14:23:31 this are the step that i had follow Oct 14 14:23:55 Elray, just use v4.1.x-ti, we've backported that driver from mainline.. Oct 14 14:24:14 https://wiki.debian.org/rt2860sta Oct 14 14:24:28 oh Oct 14 14:24:34 for 3.8/3.14 you need to install the module via: "sudo apt-get install mt7601u-modules-`uname -r`" Oct 14 14:24:35 you are using the pi instructs Oct 14 14:24:38 ok so should work on 4.1x Oct 14 14:24:58 mt7601... ouch Oct 14 14:25:13 bad driver? Oct 14 14:25:24 bad wiki instructions.. Oct 14 14:25:25 no.. just tricky Oct 14 14:25:31 that too Oct 14 14:25:37 they are for a pi Oct 14 14:25:45 i have the 4.2.3 version of the kernel Oct 14 14:25:57 should i make a downgrade? Oct 14 14:26:11 Elray, nope.. Oct 14 14:26:20 what's the lsusb of the device? Oct 14 14:26:34 hey rcn, is there a purpose to loading cape-universal? It seems like it's just going to get in the way of bonescript, pybbio, beaglebone gem, and anything else that loads dtb as needed Oct 14 14:26:58 nyt, it allows userspace control of gpio/devices/etc.. Oct 14 14:27:02 Bus 001 Device 007: ID 148f:7601 Ralink Technology, Corp. MT7601U Wireless Adapter Oct 14 14:27:02 Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub Oct 14 14:27:28 doesnt config-pin do this without having to load the overlay? Oct 14 14:28:04 rcn-ee have you some suggest for my problem? Oct 14 14:28:07 Elray, double check you ahve "CONFIG_MT7601U" enabled.. (for v4.1.x i backported v4.2.x's version, so a downgrade wont chagne anything) Oct 14 14:28:27 nyt, config-pin needs the 'universal" overlay loaded... Oct 14 14:28:33 ah Oct 14 14:28:46 and it takes "all" the available pins.. ;) Oct 14 14:28:49 why not have it load it then if it isnt? Oct 14 14:28:59 this is my lsmod retourn : Module Size Used by Oct 14 14:28:59 snd_usb_audio 126380 0 Oct 14 14:28:59 snd_hwdep 5214 1 snd_usb_audio Oct 14 14:28:59 snd_usbmidi_lib 18550 1 snd_usb_audio Oct 14 14:28:59 snd_seq_midi 5191 0 Oct 14 14:28:59 snd_seq_midi_event 5886 1 snd_seq_midi Oct 14 14:28:59 snd_rawmidi 18360 2 snd_usbmidi_lib,snd_seq_midi Oct 14 14:29:06 oops Oct 14 14:30:26 this is my lsmod output Oct 14 14:30:28 omap_rng 4264 0 Oct 14 14:30:28 rng_core 7151 1 omap_rng Oct 14 14:30:28 drm_kms_helper 103740 4 tda998x,tilcdc Oct 14 14:30:28 uio_pdrv_genirq 3321 0 Oct 14 14:30:28 uio 8146 1 uio_pdrv_genirq Oct 14 14:30:28 leds_gpio 3186 0 Oct 14 14:30:28 mt7601Usta 479753 0 Oct 14 14:30:52 how can i check if the driver is enable? Oct 14 14:31:19 Elray, uname -r ? mt7601Usta should not be loaded.. Oct 14 14:31:51 uname -r = 4.2.3-bone2 Oct 14 14:31:56 ah ok Oct 14 14:32:26 Elray, did you actually plug in the usb wifi device "before" you went off to install the 3rd party driver? Oct 14 14:32:49 yep Oct 14 14:32:55 is it wrong? Oct 14 14:33:13 it should have worked out of the box.. that's all.. Oct 14 14:33:39 i know :D Oct 14 14:33:39 nyt, cape-universal is a stepping stone to extcon: https://github.com/pantoniou/linux-beagle-track-mainline/commit/4dd38fe9c0ef1a0db76cacb8f09bef9b36fcd9ff Oct 14 14:33:53 but it's not :P Oct 14 14:34:21 nyt, the end go, tell usespace what you want the pins to do.. Oct 14 14:34:52 so eventually you want everything using config-pin in userspace land? Oct 14 14:36:21 because at the moment it seems to just add an extra layer of cruft :( at least it's easier to toggle it now Oct 14 14:36:46 i think an improvement might be to not load the universal cape until something that requires it is called Oct 14 14:36:49 nyt, yeap.. yeah right now it's an extra layer.. with extcon the big overlay isn't needed.. Oct 14 14:36:55 but i may not have the full picture Oct 14 14:37:26 the machinekit guys use cape-universal when configuring their printing devices thru machinekit/linuxcnc Oct 14 14:39:22 why not add it to uEnv.txt as commented but default disabled? ;) then it wont get in the way of other scripting stuff Oct 14 14:39:36 Elray, did you have the firmware under: /lib/firmware/mt7601u.bin Oct 14 14:40:36 umm no Oct 14 14:40:51 _av500_, i got a couple of pics, but ill sneak some video on friday when I'm back. the work was for an 'escape the room' type game so they dont want much info on it published Oct 14 14:41:00 Elray, well that's kind of needed.. Oct 14 14:41:57 Elray, https://github.com/rcn-ee/mt7601u/blob/master/src/mcu/bin/MT7601.bin Oct 14 14:43:55 ok i'm woking on it :) Oct 14 14:45:26 rcn, anyway, just my $0.02, default disabled might get in the way less, but if that's going away eventually in favor of extcon, then it shouldn't matter? I think if the universal cape was default enabled like it is now in one of the official images it might be ugly Oct 14 14:48:05 shuld i need to reboot? Oct 14 14:50:29 Elray, yeap.. Oct 14 14:50:35 no change Oct 14 14:50:41 same situation Oct 14 14:50:53 nyt, it was very helpful on bringup of v4.1.x ;) Oct 14 14:52:34 ? Oct 14 14:52:50 Elray: any luck? Oct 14 14:52:59 I had to go run and fix something Oct 14 14:53:00 no Oct 14 14:53:14 was using 4.1.x on the 'official' debian images without any issues and no universal cape Oct 14 14:53:22 and everything was awesome ;) Oct 14 14:53:23 what was it helpful on? Oct 14 14:53:43 i just tried the jessie images and saw i couldn't load anything and was like wtf Oct 14 14:54:15 Elray: dmesg | grep -i firm Oct 14 14:54:18 anything? Oct 14 14:54:31 <_av500_> nyt: thx Oct 14 14:55:03 [ 19.961474] mt7601u 1-1:1.0: Direct firmware load for mt7601u.bin failed with error -2 Oct 14 14:55:14 ah hhaaa Oct 14 14:55:33 just a sec Oct 14 14:55:35 i suppose it's not a good news :D Oct 14 14:55:43 not necessarily Oct 14 14:57:18 which v* firmware did you download Oct 14 14:57:27 you need a specific rev for that card i think Oct 14 14:57:50 Elray, make sure it's the correct name... Oct 14 14:58:00 the repo has an upper case M.. Oct 14 14:58:20 ok Oct 14 14:59:26 hopefully you used the raw link: https://github.com/rcn-ee/mt7601u/raw/master/src/mcu/bin/MT7601.bin Oct 14 15:00:04 yep Oct 14 15:00:10 sry Oct 14 15:00:14 i change the name Oct 14 15:00:22 4.1.6-810 is the version i keep seeing referenced Oct 14 15:00:57 and if i do again dmseg | grep -i firm Oct 14 15:01:02 this is the resoult Oct 14 15:01:08 [ 19.961474] mt7601u 1-1:1.0: Direct firmware load for mt7601u.bin failed with error -2 Oct 14 15:01:08 [ 615.895401] mt7601u 1-1:1.0: Direct firmware load for mt7601u.bin failed with error -2 Oct 14 15:01:57 okay odd.. my mt7601u devices are at home, so i'll look at it tonight.. Oct 14 15:02:08 Elray: look up a google search for mt7601u.bin 4.1.6-810 Oct 14 15:02:16 see if that one comes up with a binary Oct 14 15:02:18 if so try it Oct 14 15:02:25 https://www.raspberrypi.org/forums/viewtopic.php?f=28&t=119095 Oct 14 15:02:30 i'm trying to do this Oct 14 15:02:37 let's see what happen Oct 14 15:02:42 ok Oct 14 15:03:12 you can also look at viewtopic.php?p=808093# Oct 14 15:03:18 which might be same thread Oct 14 15:03:24 hard to tell from my console Oct 14 15:04:32 Elray, porjo's firmware is the one on my github... Oct 14 15:05:22 [ 19.571043] mt7601u 1-1:1.0: Firmware Version: 0.1.00 Build: 7640 Build time: 201302052146____ Oct 14 15:05:27 rcn-ee: so how will things eventually work with extcon? Oct 14 15:05:35 ooh Oct 14 15:05:41 Elray: working? Oct 14 15:05:43 no Oct 14 15:06:06 if i run ifconfig i don't see ra0 Oct 14 15:06:15 make sure you use -a Oct 14 15:06:25 just in case it is not "up" Oct 14 15:06:44 Elray, "ra0" is mt7601's staged driver.. the mainline version in v4.1.x will use "wlanX".. Oct 14 15:08:04 nyt, i still need to test extcon more, but it's all in mainline solution.... Oct 14 15:08:20 it work! Oct 14 15:08:28 cool Oct 14 15:08:36 good going Oct 14 15:08:39 :) Oct 14 15:08:43 thanks a lot guy! Oct 14 15:09:18 no problem here.. we've all had to mess with firmware i imagine Oct 14 15:09:24 but Oct 14 15:09:31 last question Oct 14 15:10:01 now i have try direclty with the usb of bbb Oct 14 15:10:23 and work but why if i plug it in the hub Oct 14 15:10:29 it's no more visible? Oct 14 15:10:39 so what will be the proper way to configure pins then? Currently I have the gem building the necessary dts file on the fly then loading the dtb Oct 14 15:10:58 bonescript seems to do similar Oct 14 15:11:05 and py bbio has prebuilt dtbs i think Oct 14 15:11:16 Elray: that I do not know Oct 14 15:11:21 i dont use hubs Oct 14 15:11:25 these methods obv dont work with the universal cape loaded, so want to add code to do it the 'right' way if its changing Oct 14 15:11:29 it's a pity :D Oct 14 15:14:33 nyt, for v4.1.x either do the overlay like you are doing or cape-universal... extcon will be an issue down the road. ;) Oct 14 15:15:35 so the official images arent getting updated with this anytime soon i take it? Oct 14 15:16:36 nyt, we still have pru issues uio_pruss -> remotepro_pruss... Oct 14 15:17:40 nyt, and the multiple dtc verion's make things fun... wheezy = 3.8 and jessie=4.1. ;) Oct 14 15:17:57 im testing with your jessie images Oct 14 15:18:00 * Tenkawa grumbles about devicetree Oct 14 15:18:04 i was running the official image earlier with 4.1 Oct 14 15:18:08 and it was fine Oct 14 15:18:32 so i should just check to see if cape-universal is loaded then unload it? ;) Oct 14 15:18:37 (in code) Oct 14 15:19:17 nyt, you can trying unloading it... but unloading is still not 100%.. ;) Oct 14 15:19:23 heh Oct 14 15:19:59 i still think it should be a default disabled option in uenv ;) Oct 14 15:20:20 _av500_ here? Oct 14 15:20:55 Elray: so it works without the hub however not with right? Oct 14 15:21:05 yea Oct 14 15:21:20 isn't it wild? Oct 14 15:21:34 Elray: tried that from a cold shutdown and powerup? Oct 14 15:21:42 does the hub showup? Oct 14 15:21:51 it could be current related Oct 14 15:22:10 the hub is visible and also the usb sound card that i have connected is visible Oct 14 15:22:18 interesting Oct 14 15:22:27 does dmesg show anything Oct 14 15:22:34 nyt, i need to 'tweak' a few things to disable it... by default config-pin will load an overlay when one isn't loaded.. for 4.1.x it'll be the wrong one as it doesn't know to check a few things.. Oct 14 15:24:14 Elray, is it a powered hub? Oct 14 15:24:18 yes Oct 14 15:24:29 Elray, for a quick fix... stick in a spare usb-flash drive... Oct 14 15:24:40 then you can remove/insert the wifi usb at will.. Oct 14 15:24:41 in the hub? Oct 14 15:24:47 ahh Oct 14 15:24:49 "something" needs to be plugged into the hub at bootup.. Oct 14 15:25:34 ook i'm try Oct 14 15:30:34 umm Oct 14 15:30:34 no Oct 14 15:30:48 i'm try with a usb memory Oct 14 15:31:07 and also that one doesn't not appear Oct 14 15:31:56 also after the shotdown Oct 14 15:32:15 rcn-ee: so with cape-universal, how would I go about changing a gpio pin to output mode? just call config-pin or is there another way? Oct 14 15:32:36 config-pin Px.yz gout Oct 14 15:33:21 config-pin Px.yz hi/low/in Oct 14 15:33:26 can you control slew rates on gpio input? What about controlling analog pins? Oct 14 15:33:34 you can check modes via: config-pin -l Px.yz Oct 14 15:33:36 and pwm Oct 14 15:33:38 ah Oct 14 15:33:46 we leave analog alone. ;) Oct 14 15:34:54 nyt, you can see custom pinmuxnig here: https://github.com/beagleboard/bb.org-overlays/blob/master/src/arm/univ-all-00A0.dts#L1904 Oct 14 15:35:28 gpio,gpio_pu, gpio_pd everything is init as an input... Oct 14 15:36:59 so eventually config-pin will be the main interface to configure stuff for 4.1, and I should ditch loading overlays and use this? Oct 14 15:38:24 nyt, if you see unversal loaded in the slots, you should try to use config-pin.. As it makes things really easy to change pinmuxing.. Oct 14 15:39:01 k Oct 14 15:39:33 doesnt look like it has the ability to set slew rates for gpio stuff atm Oct 14 15:39:42 btw, config-pin is just a wrapper around /sys/class/gpio... Oct 14 15:40:24 nyt, btw, ping charles on that email thread.. we should have slew rate controls ... Oct 14 15:41:54 Is there a systemd service I can make mine depend on to ensure that the beaglebone's eeprom is enumerated before it executes? Oct 14 15:42:46 rcne-ee i solved Oct 14 15:43:11 Elray: what was the fix? Oct 14 15:43:14 i discover that some ports doesn't work Oct 14 15:43:21 ahh Oct 14 15:43:24 bad hub? Oct 14 15:43:34 cina forever :D Oct 14 15:44:34 could be that there are a limeted number of port supported by bbb? Oct 14 15:46:00 Elray, unless you have external power to the hub.. you have very limited power from the bbb's usb port.. Oct 14 15:46:16 every "port" requires 100mA... Oct 14 15:46:38 that one that i'm using has indipende 5v alimentation Oct 14 15:46:43 then when you insert a device (usb 2.0) you can get 500mA... Oct 14 15:50:10 yay... now I got my boxes in the order I want and now I need to decide how to "organize" them Oct 14 15:50:50 _av500_, got some pics for you when youre around Oct 14 15:52:06 Tenkawa wich box? Oct 14 15:52:33 Elray: bbb's and pi2's Oct 14 15:56:50 a lot of toy :D Oct 14 15:56:57 yep Oct 14 16:00:48 how are you planning to use it? Oct 14 16:01:15 Elray: no idea yet... mostly as generic servers Oct 14 16:02:02 I have many experimental boxes Oct 14 16:02:19 look cool :D Oct 14 16:05:31 Tenkawa have you some experience with Jack? Oct 14 16:06:43 the audio daemon? Oct 14 16:07:08 if so no. Oct 14 16:08:16 yea that one Oct 14 16:09:37 that ugly guy will be my next enemy Oct 14 16:09:43 :) Oct 14 16:09:56 ugh at the beagleboard list being moderated lol Oct 14 16:10:00 takes forever for my messages to show up Oct 14 16:11:17 ahah Oct 14 16:28:02 nyt, "register"... ;) Oct 14 16:28:29 it's one person manually approving till you register.. Oct 14 16:28:47 register where? Oct 14 16:29:23 join group? Oct 14 16:29:24 for the beagleboard list... Oct 14 16:29:25 yeap.. Oct 14 16:29:45 joined Oct 14 17:11:27 hi! I have installed linux and I want it on my bone black. I have followed all instruction given on the site. the problem I am getting is that my 4 user leds are blinking simultaneously please help me Oct 14 17:14:36 adil, which revision of the bbb b or c? Oct 14 17:15:18 bone black Oct 14 17:17:32 adil, rev b or c? Oct 14 17:18:04 b Oct 14 17:18:24 that's 2gb emmc.. did you try to load the 4gb image? Oct 14 17:18:55 yes I tried it but I was unable to upload the 2 gb then I tried the 4 gb Oct 14 17:19:31 adil, flash this 2gb image: http://elinux.org/Beagleboard:BeagleBoneBlack_Debian#BBB_.28All_Revs.29_eMMC_Flashers_2 Oct 14 17:19:51 ok let me try Oct 14 17:23:22 i just snagged some maker kits for rat shack for like $15 last night Oct 14 17:23:26 got an a6 and an a5s Oct 14 17:23:34 and 2 rev c's for $35 Oct 14 17:23:40 awesome clearance lol Oct 14 17:35:32 rcn-ee, I hope its done now cause all 4 user leds are steady on. thanks alot Oct 14 17:35:59 adil, yeap it's done. ;) Oct 14 17:36:40 now can I run linux by just ejecting the card the plug the usb again? Oct 14 17:37:09 adil, remove power, remove card and then insert your power source... it'll take upto 30 seconds on the first bootup as it's generating ssh keys.. Oct 14 17:37:24 ok thanks Oct 14 17:37:42 that was a great help thanks once again Oct 14 17:43:17 rcn-ee, can you please tell me how can I open linux from beagle bone, bbb has installed the driver but I cant see anything in mycomputer Oct 14 17:43:58 what do you mean open linux? Oct 14 17:44:06 adil, are you running windows/mac/linux on your "desktop pc"? Oct 14 17:44:12 you should be able to ssh to it at 192.168.7.2 Oct 14 17:44:15 windows Oct 14 17:44:41 did a usb flash drive popup? Oct 14 17:44:50 look under 'my computer'.. Oct 14 17:46:47 nothing appears under my computer Oct 14 17:47:47 did you plug it into the computer via the small usb-otg connector next to the ethernet jack? Oct 14 17:49:02 no, I have not plugged in the usb plug on that connector but the next to it Oct 14 17:49:55 adil, so how... where you planning on opening linux on the device? Oct 14 17:50:56 I have plugged in the usb plug and I hope that bbb can show me something under my computer but I could not see anything Oct 14 17:54:39 bummer, where did dhirc go, as TI fixed the link to the 2.00.00 version of the Linux SDK for AM335x Oct 14 17:55:07 hello Oct 14 17:55:15 anyone here play with node? Oct 14 18:18:18 rcn-ee: lol, nice buffer management -> https://e2e.ti.com/support/arm/sitara_arm/f/791/p/446385/1655708#1655708 Oct 14 18:19:39 laughs. ;) Oct 14 18:25:46 1st time BBB user Oct 14 18:26:05 Having some difficulty getting started. I followed the main page but still not sure if it's working... Oct 14 18:26:36 Anyone available to give me a hand? Oct 14 18:27:18 sorry, I'ld to keep my two hands Oct 14 18:31:50 lol penny for your thoughts then? Oct 14 18:32:09 *gives defiant a penny* Oct 14 18:33:17 So I can view the devices folder tree in windows explorer but I cannot connect via 192.168.7.2 Oct 14 18:35:03 connect as in browser to http://192....? Oct 14 18:42:28 correct Oct 14 18:46:04 does ssh work? eg. with Putty ? Oct 14 18:46:14 or ping Oct 14 18:47:16 rcn-ee: TI apparently "fixed" the problem, with the minor side effect of corrupting the DMA pointers if it hits a race condition ( pics -> https://goo.gl/jjPmUL ) Oct 14 18:47:46 (he had set up a camera overlooking a farm of devices since he couldn't find a reliable way to trigger it) Oct 14 18:49:29 zmatt: lol Oct 14 18:50:14 I'm not sure I'd personally see that as a good trade-off to reduce tearing Oct 14 18:50:25 fixing the buffer mismanagement sounds like a better idea :P Oct 14 18:51:59 qdh required I feel ... Oct 14 18:52:28 product shipping is an incredible motivator Oct 14 18:53:39 so I can SSH into it but I don't know what I'm doing in the shell... I was hoping to use jscript and bonescript to control it Oct 14 18:59:10 SpaceCowboy: ok well that narrows down whats wrong... Oct 14 19:21:33 I'm trying to run a PRU example program. I've followed the example here: http://boxysean.com/blog/2012/08/12/first-steps-with-the-beaglebone-pru/ Oct 14 19:21:57 currently, I get the error "prussdrv_open open failed" when executing the first program Oct 14 19:22:16 lsmod shows the uio_pruss is enabled too. Oct 14 19:22:43 anyone have any suggestions as to what's wrong? Oct 14 19:27:39 permissions? correct io#s provisioned in DT? Oct 14 19:28:53 so what kind of io is the x15 going to have? Oct 14 19:29:48 nyt, too many... Oct 14 19:30:05 i dont see any headers on the pic? Oct 14 19:30:41 am i missing something/ Oct 14 19:30:47 nyt, 2nd picture: http://www.elinux.org/Beagleboard:BeagleBoard-X15 Oct 14 19:30:57 high speed connectors.. (pci-e/sata/etc...) Oct 14 19:31:03 ah underneath Oct 14 19:31:23 estimated msrp? Oct 14 19:31:41 keep scrolling ;) Oct 14 19:32:32 nyt, the core is pretty spendy: https://store.ti.com/XAM5728AABCXE.aspx Oct 14 19:33:11 interesting Oct 14 19:37:13 veremit, I was running as root. Running this command from someone else's guide fixed it. "echo BB-BONE-PRU-01 >/sys/devices/bone_capemgr.9/slots" Oct 14 19:37:55 ah Oct 14 19:39:43 I had assumed that one of the scripts the first guide had me run would have added the PRU to the device tree, but I guess not? Oct 14 19:40:37 froggyman.. the methods change quite frequently with each release/version of the board & software Oct 14 19:41:20 veremit, I suppose that makes sense then. Oct 14 19:41:58 hopefully you can move forward now :D Oct 14 19:43:59 veremit, I think so. I'm sure i'll have plenty of future problems Oct 14 19:44:01 :D Oct 14 19:44:12 hehe well thats what makes it fun(tm) :D Oct 14 19:54:10 veremit, have you done much PRU dev? Oct 14 19:55:19 not I yet, froggy Oct 14 19:56:39 ahh Oct 14 19:57:19 my beagle is currently doing something quite shameful lol Oct 14 19:57:30 oh? Oct 14 19:57:34 collecting dust? Oct 14 19:57:45 no.. not quite -that- bad .. lol Oct 14 19:58:21 dhcp, named, nfs, cifs .. shortly vpn too Oct 14 19:59:28 but I will get to pru some day Oct 14 19:59:36 just need a new beagle .. :D Oct 14 20:01:23 get an x15. has 4 pru cores Oct 14 20:01:41 tbr... they're on sale now? finally? Oct 14 20:01:58 veremit: just got announced publicly Oct 14 20:02:14 how astronomical is theprice then?! Lol Oct 14 20:02:22 250 or such Oct 14 20:02:35 also it will be the usual unobtainium for a while Oct 14 20:03:19 just finished a crazy bbb project Oct 14 20:03:23 running an escape the room game w/ it Oct 14 20:03:32 looks like it's doing the rounds on the news pages Oct 14 20:03:56 tbr .. if cco have anything to do with it . .. will be unobtainium for the foreseeable .. Oct 14 20:04:42 reading hall effect sensors, magnetic door sensors, triggering leds, playing sounds and video, controlling hue lights, triggering h-bridge to open/close doors, button input, web server for controlling a few tablets, etc Oct 14 20:05:05 tis pretty insane board though Oct 14 20:12:19 nyt: and all without blowing up the board, again Oct 14 20:12:35 yeah nfc why it ate shit the first time Oct 14 20:13:10 all i did was add some resistors to the door sensors after Oct 14 20:13:24 which was a simple circuit, gpio input with pullup to ground Oct 14 20:15:02 i was testing kernel stuff when it crashed, wonder if it didnt like one of the pins not having the overlay up for it or something Oct 14 20:15:04 i got nothing otherwise Oct 14 20:15:34 its been up for a few days now Oct 14 20:16:07 just be sure *nothing* is connected to beagle pins if its unpowered Oct 14 20:16:35 I blew one up that way... Oct 14 20:17:33 I suspect even a few series resistors would prevent damage occurring Oct 14 20:17:47 but these things seem sensitive Oct 14 20:18:15 not like the trusty PIC or arduino/atmel/etc Oct 14 20:19:22 i power cycled it with stuff plugged in before it died Oct 14 20:19:30 if the am335x weren't so capable .. I'd still be doin things with an auxilliary controller and doing interface over i2c or another bus Oct 14 20:19:37 but nothing that should have had voltage on it Oct 14 20:19:52 ih well .. Oct 14 20:20:07 i was compiling dtc stuff when it died was up for a good 20 minutes since last power cycle Oct 14 20:20:09 "should" .. ;) Oct 14 20:20:20 well i went back with a meter and it was all clean Oct 14 20:20:24 so i have no idea Oct 14 20:20:49 anyway, everything has a few thousand ohm resistors in front of it now at least Oct 14 20:21:18 damn .. 3x usb3 host ports.... Oct 14 20:21:26 right Oct 14 20:21:32 impressive Oct 14 20:21:39 i want one but have no use for it at that price point Oct 14 20:21:59 what can those dsp cores achieve I wonder... Oct 14 20:22:17 the ti sharc stuff is cool Oct 14 20:22:27 ugh so Oct 14 20:22:34 the solar panels for my oil level sensor all got fogged :( Oct 14 20:22:39 gonna go try to polish em up Oct 14 20:22:42 no .. at ~250 thats 10 quad-core Rpi's ... Oct 14 20:23:00 I hate to say it .. but .. wow, thats much more computing power :D Oct 14 20:23:06 lol 10? Oct 14 20:23:14 thought they were still like $35 Oct 14 20:23:19 still Oct 14 20:23:21 ah depends on £/$ lol Oct 14 20:23:22 quad core rpi is nice Oct 14 20:23:23 and cheap Oct 14 20:23:32 going to port my code to it Oct 14 20:23:58 or .. hrm .. 3x or 4x quad-core Wandboards :D that's a nice cluster Oct 14 20:24:08 nuc ;) Oct 14 20:25:03 but yeah . I got 3 pi's not in use currently .. pi1's Oct 14 20:25:18 I should put an lcd on one .. oh wait .. no wifi Oct 14 20:26:10 lcd+wifi .. but then it costs more than the cpu board lol Oct 14 20:26:21 im using an rpi2 for a bt audio sink for a little stereo system in my kitchen Oct 14 20:26:23 thats about it so far Oct 14 20:26:25 and I can get a cheapo pad for that Oct 14 20:26:43 what you using for bt? Oct 14 20:26:49 $5 bt dongle Oct 14 20:26:55 usb Oct 14 20:26:56 work ok? Oct 14 20:27:20 yea Oct 14 20:27:22 no issues at all Oct 14 20:27:25 can share the codes if you want Oct 14 20:28:01 another time perhaps :) Oct 14 20:28:52 just a suggestion, doesn't seem like there is a way to update this but the quickstart guide for the bb view should be this http://www.element14.com/community/message/111707/l/problems-with-43-lcd-and-bbb#111707 Oct 14 20:29:08 and not that pdf which just has a link to the site Oct 14 20:29:22 BB-VIEW is rubbish Oct 14 20:29:42 far better solutions for 4.3" lcd Oct 14 20:29:48 and angstrom is obsolete for bbb now Oct 14 20:31:38 work bought the 4d cape .. not sure whether it was 4" or 7" Oct 14 20:31:56 yeah i think thats the good one Oct 14 20:32:08 wife used it in a pip boy she made Oct 14 20:32:24 just works Oct 14 20:34:11 ahh Oct 14 20:34:13 well shit Oct 14 20:34:32 ok forget the view, that's just a toy I was going to play with Oct 14 20:35:25 if I wanted to write node apps what would you suggest? what flavor of linux? Oct 14 20:36:24 hmm the 'view looks ok .. just the ele14 'version' Oct 14 20:36:59 doesn't make a difference .. but the current 'flavour' the beagle family are using is debian Oct 14 20:37:42 dtzitz, the 4.3 bbb-view lcd has the red/blue reversed.. Oct 14 20:38:08 http://www.element14.com/community/community/designcenter/single-board-computers/next-gen_beaglebone/blog/2015/05/20/beaglebone-black-with-the-bbview-and-the-new-beaglebone-debian-image Oct 14 20:38:10 you can use a later kernel image, BUT you need to rebuild the 3.8 branch with this line un-commented... Oct 14 20:38:28 https://github.com/RobertCNelson/bb-kernel/blob/am33x-v3.8/patch.sh#L1134 Oct 14 20:39:33 looks a bit like a mess as usual though .. hardware with no software support lol Oct 14 20:40:45 veremit, "broken" hardware with "some" softare support.. idiots reversed 8 bit red with 8bit blue.. Oct 14 20:40:59 wow Oct 14 20:41:00 lol Oct 14 20:41:35 rcn-ee .. "software issue" :P Oct 14 20:42:08 veremit, memcpy red to blue in software... yeah.. it's hardware... Oct 14 20:42:11 "nobody said it was red bits .. so we just wired them through .. " Oct 14 20:42:33 thats pretty grim lol Oct 14 20:42:43 but, we have a 3.8 (patched) kernel that works with it.. Oct 14 20:42:51 haven't bothered to patch 4.1 with it thou.. Oct 14 20:43:03 can't shoe-horn it in? Oct 14 20:43:39 I guess I am coming back to this tomorrow Oct 14 20:43:43 out of time today Oct 14 20:43:46 :) Oct 14 20:43:48 thanks for the pointers Oct 14 20:44:12 dtzitz, if you still can return it... and pick up 4dsystems' 4.3 lcd.. its' fully supported.. Oct 14 20:44:22 I am going to look at ubuntu and... maybe theres something for the click shield that will give me hdmi out? Oct 14 20:44:40 rcn-ee, not worried about it, work bought a bunch of stuff Oct 14 20:44:41 dtzitz, ubuntu/debian share the same kernel for beaglboard.org.. Oct 14 20:45:08 they bought the element14 starter kit looks like Oct 14 20:45:16 ubuntu trusty's userspace is a little painful as it's older.. but wily will be out next week and the new lts's will be opened.. Oct 14 20:46:26 rcn-ee, you have a personal preference for OS? Oct 14 20:47:16 dtzitz, i'm a little bias as i'm building/supporting debian for beagleboard.org... Oct 14 20:47:38 ubuntu was "fun" to use in the past, but for developers it's gotten too "corporate"... Oct 14 20:48:36 word, well then that's what I will go with Oct 14 20:57:45 rcn-ee.. you couldn't be pursuaded to more of the recent flavours? Oct 14 20:58:53 veremit, https://mjg59.dreamwidth.org/37113.html Oct 14 21:01:18 rcn-ee, so why should we care about canonical??? Oct 14 21:01:40 veremit, ubuntu=canocial.. Oct 14 21:02:06 For debian, we have a 3.8 based wheezy image, and a 4.1 based jessie image.. Oct 14 21:02:36 you can run 4.1 in wheezy and 3.8 in jessie. ;) Oct 14 21:03:08 then we have images for stretch, but that's not going to release for another 2 years.. Oct 14 21:03:35 but jessie is old .. Oct 14 21:03:40 its just not obsolete yet ;d Oct 14 21:03:49 veremit, then run stretch. ;) Oct 14 21:04:32 veremit, http://elinux.org/BeagleBoardDebian#Debian_Testing_.28stretch.29 Oct 14 21:06:34 althuogh I have to admit .. only reason I use debian is for pre-built packages .. don't wanna be compiling all the time a-la gentoo .. unless I get a really slick cross-compiler going the Proper gentoo way .. 0sigh- Oct 14 21:08:17 oops coffee .. brb Oct 14 21:30:09 mmmm coffeeeee... Oct 14 21:51:06 veremit: i just got distcc working on gentoo a little while back.. it's awesome :) Oct 14 21:51:47 cc0_been using it for clusters for years :) Oct 14 21:52:02 next step is getting cross-compiling to work with it :) Oct 14 21:54:23 cc0_ .. you can use it for kernel compiling too with a bit of Makefile bodgery .. but there are a few edge cases where it doesn't work .. eg building Thunderbird I've found. Oct 14 21:55:30 makes a difference when you got a couple of i7's on the network :D Oct 15 01:26:10 argh, jumping back into public ROM instead of rebooting is much much faster when using TFTP boot (avoids eth phy reset), but I can Oct 15 01:26:21 can't get it to work if my app ever enables the MMU Oct 15 01:26:30 even though I take care to disable it again, flush all caches, etc Oct 15 01:30:30 maybe the ROM uses the MMU to map itself in Oct 15 01:30:42 pubrom never enables the MMU Oct 15 01:30:57 in fact, on reboot it actually loads the program from TFTP again Oct 15 01:31:01 but then my program crashes Oct 15 01:31:24 and I've narrowed it down: enabling the MMU per se isn Oct 15 01:31:29 isn't the problem Oct 15 01:31:50 clocks in a different state then it needs? Oct 15 01:32:15 dcache is Oct 15 01:32:30 I doubt it, I don't touch them Oct 15 01:32:41 also, why would that be solved by not enabling data cache? :P Oct 15 01:33:35 secret secure mode side effects? Oct 15 01:33:55 I already put my meddling with secrom in an #if 0 Oct 15 01:37:20 meddling too much with secrom usually gets you punished with an MPU Security Violation reset anyway Oct 15 01:38:06 there's also no way to disable the resetout pulse is there? since that's the root issue for me Oct 15 01:39:59 oh well, reboot isn't that slow either, don't really want to dive into the arcane bowels of the caches again Oct 15 01:43:21 not so much you messing with it Oct 15 01:43:34 well I am Oct 15 01:43:40 was thinking may be the secure rom did something and the regular rom assumes that has happened Oct 15 01:44:10 yeah but I'm not changing all that much, and pubrom is pretty self-contained Oct 15 01:44:45 as evidenced by the fact it doesn't crash in pubrom, pubrom still successfully downloads my app via TFTP Oct 15 01:45:07 but once I try to enable mmu and caches again, something from the past seems to return to haunt me Oct 15 01:46:01 ah Oct 15 01:46:02 even though I've flushed everything both right before jumping back to pubrom and again before enabling the mmu/caches again Oct 15 01:46:32 totally unrelated - know of any docs on the PRU mailboxes off hand? the standard docs list them but doesn't describe them Oct 15 01:47:03 PRU mailboxes? I didn't think PRU had any, they're just using the system mailbox Oct 15 01:47:08 do the flushing ops block? Oct 15 01:47:26 Oh is that it... 'k Oct 15 01:47:27 they do if you follow them by asm( "dsb" ); Oct 15 01:47:51 zmatt: are you using DRAM? Oct 15 01:47:54 nope Oct 15 01:48:35 checked for any uninitialized accesses? Oct 15 01:48:44 by which you mean.. ? Oct 15 01:48:47 like stack variables sitting in SRAM Oct 15 01:48:57 prehaps the SRAM doesn't get cleared somehow Oct 15 01:49:09 the SRAM doesn't get cleared Oct 15 01:49:22 at all, under any circumstance, unless you do it yourself Oct 15 01:49:31 it's filled with random garbage on cold boot Oct 15 01:50:06 my stack is in an uninitialized section, I clear bss myself Oct 15 01:51:31 I've had trouble of this sort before... I eventually gave up and learned to just always reset the SoC Oct 15 01:51:48 I don't mind resetting the SoC, I just don't want to reset that ethernet phy :/ Oct 15 01:51:52 depends on what you mean by 'clear'. SRAM gets scribbled on by ROM Oct 15 01:52:07 cut the trace Oct 15 01:52:19 it *requires* reset on power up Oct 15 01:52:38 it should have been connected to a gpio Oct 15 01:53:16 especially since the reset pulse is currently very marginal, which may be the cause of the random phy problems some people have Oct 15 01:53:21 sure? some of them will let you use MDIO to initiate a reset Oct 15 01:53:31 the datasheet is very clear about it Oct 15 01:53:37 with very specific timing requirements Oct 15 01:53:44 hmmm that part sucks Oct 15 01:54:22 that phy is an insult to the ethernet subsystem it's connected to :/ Oct 15 01:55:15 what exactly are you doing that requires ethernet? Oct 15 01:55:24 TFTP booting Oct 15 01:55:30 recompile, reset Oct 15 01:55:40 edit, recompile, reset Oct 15 01:55:40 edit, recompile, reset Oct 15 01:55:41 etc Oct 15 01:55:57 having pubrom grab the executable from my HD is very convenient Oct 15 01:56:11 I mean in the final product Oct 15 01:56:22 you could always have U-boot TFTP things ;) Oct 15 01:56:39 that would magically make reboot faster? :P Oct 15 01:57:04 this isn't for a final product, I'm doing tests that are easier to do on baremetal than under an OS Oct 15 01:57:04 for development... Oct 15 01:57:16 oh Oct 15 01:58:04 note that the reboot time I'm complaining about is the occasional outlier of 2 seconds Oct 15 01:58:07 usually it's instant Oct 15 01:58:27 i wonder if you can move the reset to a GPIO Oct 15 01:58:32 xacto knife and.... Oct 15 01:58:50 in theory anything is possible, but it doesn't look like my idea of fun :P Oct 15 01:59:14 there are also some control module registers that affect the external reset pulse timing I think Oct 15 02:01:40 as for final products... basically all of them use ethernet **** ENDING LOGGING AT Thu Oct 15 02:59:59 2015