**** BEGIN LOGGING AT Mon Nov 24 03:00:33 2014 Nov 24 03:29:56 hi Nov 24 03:30:30 i'm following the instructions all the way down in this webpage: http://processors.wiki.ti.com/index.php/Using_the_WL18xx_Cape_with_BeagleBone_Black Nov 24 03:31:18 but when it comes to the testing phase on a real Beaglebone-black + TI WL18xxMOD cape boards, I met this error: Nov 24 03:31:36 # ifconfig wlan0 up Nov 24 03:31:36 ifconfig: SIOCSIFFLAGS: Invalid argument Nov 24 03:32:12 does anybody happen to meet the same error? any suggestion? Nov 24 03:33:54 run # lsmod shows this: Nov 24 03:33:56 # lsmod Nov 24 03:33:56 Module Size Used by Nov 24 03:33:56 wl18xx 65985 0 Nov 24 03:33:57 wlcore 141586 1 wl18xx Nov 24 03:33:57 mac80211 294664 2 wl18xx,wlcore Nov 24 03:33:58 cfg80211 172361 3 mac80211,wl18xx,wlcore Nov 24 03:33:59 wlcore_sdio 4139 0 Nov 24 03:34:03 compat 4192 4 cfg80211,mac80211,wl18xx,wlcore_sdio Nov 24 04:11:45 ifconfig -a ' shows what? Nov 24 05:47:45 hi guys does anyone know why does 'rfkill' do a 'hard block' on wifi dongles? I have Realtek Nov 24 08:35:47 Hi, I read the SRM and see that it says the GPIO pins are 3.3V (as expected) and plenty of warnings not to apply 5V or apply power to pins while the board is off, but I'm curious if there's any guarantee that they might be 3.43V tolerant Nov 24 08:36:10 read the TRM Nov 24 08:36:36 okay so I searched for the TRM and couldn't find anything other than the SRM and some schematics Nov 24 08:36:47 I'm not sure what document you're talking about exactly Nov 24 08:36:58 rule of thumb: there is a (latch-up) protection diode on the io pins to VDD. if that gets cunducting, that's your limit Nov 24 08:37:06 TRM = technical reference manual Nov 24 08:37:11 the one of the processor Nov 24 08:37:19 yes I searched for technical--oh, the processor Nov 24 08:37:52 I'm not really an EE guy; I need to read up a lot so I should probably get to doing that Nov 24 08:37:59 btw: why will you have 3.43V on the input? Nov 24 08:38:20 I'm interfacing with some hardware that does logic at 3.43V or some reason Nov 24 08:38:44 that sounds illogical Nov 24 08:39:03 it's probably 3.3V with a power supply that is slightly off Nov 24 08:39:17 I've found multiple documents that say it's 3.43V Nov 24 08:39:19 weird I know Nov 24 08:39:25 o_0 Nov 24 08:39:26 I haven't tried measuring it myself yet though Nov 24 08:39:28 what hw is that? Nov 24 08:39:39 The GameCube's controller ports Nov 24 08:39:58 eh.. Nov 24 08:40:06 then i'd say the documents are wrong Nov 24 08:40:13 that's mass produced electronics Nov 24 08:40:13 could be Nov 24 08:40:16 it's not official Nov 24 08:40:19 it will adhere to standard Nov 24 08:40:26 none of the stuff I've found is official Nov 24 08:40:34 so either 1.2V, 1.8V, 3.3V or 5V logic Nov 24 08:40:44 anything inbetween is non-sensical in this application Nov 24 08:41:03 (non-sensical = it makes production more expensive) Nov 24 08:41:11 sure Nov 24 08:41:25 so, i wouldnt worry about the 3.43V thing Nov 24 08:41:28 ok I found the CPU manual and it seems to say that maximum tolerated voltage is 3.465V Nov 24 08:41:37 so it should be ok one way or the other Nov 24 08:41:44 I should still measure it first though Nov 24 08:42:19 3.43V is 3.3V+4%, that's about what you get if you have a cheap power supply and measure it with an even cheaper multimeter Nov 24 08:42:51 jpfau: is the 3.465V the absolute maximum rating? Nov 24 08:43:33 I think that's what this is saying Nov 24 08:43:56 so yeah...with fluctuations that might not be good? I should put some time into learning EE stuff Nov 24 08:45:42 i wouldnt worry about it Nov 24 08:47:03 the voltage specs are usually such that you can interface systems with different powersupplies that have the same nominal voltage without additional circtuitry Nov 24 08:48:01 ok, thanks Nov 24 08:48:10 I'm gonna get some help from a friend who knows this stuff way better than I do Nov 24 14:38:40 Hi, I'm running fine a 5,7" LCD touchscreen @ 640x480 with ubuntu distro by adding 10-monitor.conf file into X11 settings (otherwise larger portion of screen is truncated due to native 800x480 res.), with an Angstrom distro I'm not able to set the 640x480 res. with the same method: Is Angstrom using different video mode/device ? Any suggestion please ? Nov 24 14:40:22 (LCD is a cape one) Nov 24 14:49:02 Does anyone know where the repository for the code is that is running on the beaglebone black rev C? It looks like the github repo is out of date (at least for uboot) Nov 24 14:50:12 there is the uboot git Nov 24 15:21:51 Can somebody guide me on how to connect my beaglebone black to my Ubuntu Laptop using LAN cable? Nov 24 15:22:06 USB connect is not an option because I broke one and I use the other one for my internet modem Nov 24 15:31:48 I am running ubuntu 14.04 on a bbb with a touchscreen but can't get the touchscreen to work, I installed tslib and when I cat /dev/input/event0 and touch the screen it spills data. Sometime I am messing with TSLIB variables and the mouse appears on lxde and moves if I touch the screen but uncalibrated Nov 24 15:32:44 when i run ts_calibrate I get "selected device is not a touchscreen I understand" Nov 24 15:32:50 zenx: google for tslib Nov 24 15:33:00 zenx: there are steps to be taken to calibrate it first Nov 24 15:33:15 also, tslib does not provide a normal mouse, the underlying system has to support it Nov 24 15:33:21 KotH: I am setting the TSLIB variables accordingly Nov 24 15:33:26 dunno how X is configured though Nov 24 15:33:36 KotH: yes the qt i am using should be compiled with the tslib plugin Nov 24 15:34:06 how X is configured for the touchscreen is a mistery for me Nov 24 15:34:08 er... there is a difference between runing qt with qws (what you probably are trying to do) and running qt with x11 Nov 24 15:34:18 as i said, google Nov 24 15:34:24 I am running qt5, no qws i think Nov 24 15:34:27 there are lots of howtos and #exactsteps Nov 24 15:34:30 KotH: ok thanks Nov 24 15:34:43 and I am running qt5 with framebuffer Nov 24 15:34:46 not X Nov 24 15:35:19 ah.. so that would be qws :) Nov 24 15:35:53 then google for "tslib calibrate" and follow those steps Nov 24 15:36:00 thank you Nov 24 15:36:12 after that, it should work Nov 24 17:02:57 hi all Nov 24 17:04:29 What is the best way to play with 96 gpio on the beagle board black ? I have not found cape to extend gpio ... Nov 24 17:11:14 What is the best way to play with 96 gpio on the beagle board black ? I have not found cape to extend gpio ... Nov 24 18:34:38 Is it possible to run X-applications without having the binary blob for the gpu? Nov 24 20:51:40 <_av500_> sw-gl Nov 24 20:51:46 <_av500_> should be doable Nov 24 21:29:14 Hello Nov 24 21:29:35 I'm trying to power a BeagleBone Black from an external 5V PSU... connected to pins Nov 24 21:30:22 Documentation says I should be able to connect on P9, PIN 1 for GND & PIN 5 for VDD_5V Nov 24 21:30:47 But it doesn't work. Do I need to do something else? Nov 24 21:31:04 I'd prefer not to use the barrel DC connector, nor the USB connector. Nov 24 23:14:59 I made a kernel module for the timer capture hardware on the BBB: https://github.com/ddrown/pps-gmtimer Nov 24 23:15:24 #* PPS 0 4 273 13 -157ns[ -291ns] +/- 32ns Nov 24 23:15:59 ur readme looks nice :) Nov 24 23:16:09 heh Nov 24 23:16:16 can't tell if serious... Nov 24 23:16:55 looks good for someone who wants to jump with the stuff. Nov 24 23:17:14 not me, haven't work with timer on bbb Nov 24 23:17:52 yeah, probably a pretty short list of people interested in it Nov 24 23:18:31 but it does have some interesting info out of it: interrupt latency is around 6us Nov 24 23:19:48 hmm Nov 24 23:23:30 Hello, Is blacklib still available ? can it be used with beaglebone ? Nov 24 23:24:05 thing: what's blacklib? Nov 24 23:24:52 Sorry, I read on the net its beagleboard library to create PWM, on GPIOs Nov 24 23:28:50 hmm Nov 24 23:30:19 can we drive little dc motor directly from GPIO ? Nov 24 23:31:53 most of the GPIO can drive 6mA, is that enough? Nov 24 23:32:07 I think so. Nov 24 23:32:15 thanks. Nov 25 00:34:15 Hello everyone Nov 25 00:36:47 What do I need to do to see boot messages... currently when I boot I can only connect when I see the login prompt and so its letting me only connect after the boot process is complete I think. Im connecting via usb serial connection Nov 25 00:42:17 get a proper serial debug cable .. and connect to debug header :) Nov 25 00:43:57 veremit: thank you. Is is possible over usb cable? Nov 25 00:46:06 http://elinux.org/Beagleboard:BeagleBone_Black_Accessories#Standard_FTDI_Cable Nov 25 00:47:22 veremit: thanks - taking a look Nov 25 00:48:28 the initial boot messages are sent to a simple terminal device before any device drivers are loaded. Nov 25 00:49:18 veremit: that makes total sense... so by the time my usb connection happens its post boot Nov 25 00:49:28 yuup Nov 25 00:50:13 veremit: I am trying to debug a custom kernel not booting... do you think my failed boot would have messages in /var/log... so that even if I can boot I could mount the drive in another machine and inspect the logs? Nov 25 00:51:26 if it gets far enough to write a log .. maybe .. :) Nov 25 00:51:58 if the kernel borks.. you're not gonna be opening any filesystems ;) Nov 25 00:52:14 veremit: hrmm... that sounds about right Nov 25 00:53:08 veremit: Maybe you'll know this.. is there a difference between uImage and zImage files? Nov 25 00:53:09 kernel-level .. you're gonna want that serial console, trust me :) Nov 25 00:53:26 veremit: ordering now :-) Nov 25 00:53:34 paulfm: it would for certain kinds of boot problems Nov 25 00:53:34 But not all boot problems. Serial console would be the best way to debug Nov 25 00:54:15 I think uImage has a bit of uboot attached to it .. where zImage is just plain compressed kernel Nov 25 00:55:49 ddrown: thank you - I am definitely getting one... I need boot messages bigtime Nov 25 00:56:35 http://chaitutechbits.blogspot.co.uk/2010/12/what-is-difference-between-zimage-bz.html Nov 25 00:56:55 veremit: aha... hrmm.. I did a custom kernel and ended up with uImage and zImage files... the web how to sid to use the zImage but its not booting...Ill try the uImage Nov 25 00:57:15 hah.. handy. Sorry - I should have googled that Nov 25 00:57:24 you'll need a modified uboot for a zimage .. its doable Nov 25 00:59:41 veremit: What I did was put the zImage in the boot partition.. and edited uEnv to use the new Image... and also copied modules to the modules dir... but I see my modules dir is like myimage-3.18.3-dts-dirty... something like that... should my kernel know how to ding that specific module dir? Nov 25 01:00:08 ding = find :-) Nov 25 01:00:10 kernel will find its own modules, if it follows the right naming convention Nov 25 01:00:18 lol... typo city - sorry Nov 25 01:00:38 if it can't find .. it just won't load any :) Nov 25 01:00:53 I see when I do mkimage -l ...it has a similar string... but its truncated... maybe just for display Nov 25 01:01:12 veremit: ahh... good point.. so with no module it should still boot? Nov 25 01:01:49 usually .. unless something fundamental is missing Nov 25 01:02:05 veremit: so the bare minimum is to put the new kernel in the boot partition and update uEnv.txt. Does anything else need to happen? Nov 25 01:02:06 eg. filesystem support built as a module :P Nov 25 01:02:43 veremit: haha that would be a good one. I did leave all the defaults (only added aufs3 support) Nov 25 01:03:23 hrmm... I think I need to try the uImage.. maybe Ill get luckt Nov 25 01:03:26 lucky Nov 25 01:04:24 paulfm: did you rebuild your initrd? Nov 25 01:04:37 ddrown: uhm.. no I didn't Nov 25 01:04:56 ddrown: .. I'm feeling a little stupid right now Nov 25 01:05:12 hmm what's using an initrd? Nov 25 01:05:14 paulfm: I think mkinitramfs is the command Nov 25 01:05:32 don't remembr doing that .. except maybe for debian.. Nov 25 01:05:47 I am sunning debian Nov 25 01:05:49 running Nov 25 01:06:17 initrd_file=initrd.img Nov 25 01:06:17 loadinitrd=load mmc ${mmcdev}:${mmcpart} ${initrd_addr} ${initrd_file}; setenv initrd_size ${filesize} Nov 25 01:06:46 ^ defaults from uEnv.txt Nov 25 01:06:48 ddrown: yow.. are those args for mkinitramfs ? Nov 25 01:06:57 ddrown: ahh... Nov 25 01:07:00 no, that's just to show it's loading initrd.img from /boot/uboot/ Nov 25 01:07:42 hmmm looking into how I regenerate a new initrd Nov 25 01:08:47 mkinitramfs -o initrd-new.img [optional kernel version] Nov 25 01:09:00 lsinitramfs initrd-new.img Nov 25 01:09:03 to show the contents Nov 25 01:09:32 ddrown: awesome! trying it now Nov 25 01:09:56 do I need to be in root kernel source dir?... Nov 25 01:10:02 nope Nov 25 01:10:11 ok giving it a go! Nov 25 01:10:12 it should pick everything up from /lib/modules/[kernel version] Nov 25 01:10:32 aahh... so I need to run this on bbb Nov 25 01:10:40 right Nov 25 01:10:47 trying.. Nov 25 01:12:19 is there a way in debian to configure low-power modes of the am3358 on the beaglebone black? Nov 25 01:12:56 what is the power consumption of the prus? i've looked in the datasheet and reference manual for the am3358 and can't find any information on pru power consumption. Nov 25 01:14:01 yates: You keep asking those questions here. Nov 25 01:15:32 ddrown: so It generated a new inited image... but when I list it... it only shows old /lib/modules/3.8.13-bone modules dir Nov 25 01:15:59 you imply that the PRUs have a separate power consumption figure .. I don't think it can be distinguished from the main chip power requirements Nov 25 01:16:07 ddrown: my new 3.18.3-dts.... modules dir is not listed in the file listing Nov 25 01:16:35 paulfm: by default, mkinitramfs uses the currently running kernel Nov 25 01:16:42 ddrown: do I need to specify it as an arg maybe... the usage says the kernel version is optional Nov 25 01:16:51 you'd have to give it the other kernel version on the arg Nov 25 01:17:08 ddrown: aha! Nov 25 01:17:34 ddrown: I thought it may pick up all in the modules dir... wishful thinking Nov 25 01:24:01 ddrown: well that worked!.... going to try the new kernel again with the new initrd (might be my lucky day) Nov 25 02:31:26 hrmm... no luck booting.. serial cable here I come Nov 25 02:53:27 hello **** ENDING LOGGING AT Tue Nov 25 02:59:57 2014