**** BEGIN LOGGING AT Wed Apr 11 02:59:58 2012 Apr 11 03:37:26 in case you guys have never seen this (I hadn't) http://www.face-rec.org/databases/ Apr 11 04:01:25 so, i'm having the same issue building the cloud9-image this guy was having 4 months ago: http://pastebin.com/fRJgx85C Apr 11 04:01:28 what gives? Apr 11 04:01:48 found this chatlog which suggested a bbappend file with some sed-fu but didnt seem to help http://logs.nslu2-linux.org/livelogs/oe/oe.20111217.txt Apr 11 04:02:35 when i try to connect my beaglebone to my computer (running ubuntu 10.04), nothing is showing up in the /dev/ folder. does anyone know how to fix this? Apr 11 04:15:30 does anyone know where to get the kernel headers for compiling modules on the beaglebone? Apr 11 05:13:33 mycle_vername: have you tried building against the angstrom kernel built with bitbake/oe Apr 11 07:01:23 does anyone know of anywhere that has stock of the xm? Apr 11 07:08:14 <_av500_> check the retailers listed on the website Apr 11 07:13:36 special comp usually has them when no one else does... but they're a little higher price and their shipping has been flaky in my experience Apr 11 07:14:29 nevermind.. they're out of stock too Apr 11 07:26:41 pretty much everyone except farnell uk is out of stock Apr 11 07:27:58 at the lovely price of usd$190 Apr 11 07:28:10 plus whatever shipping is Apr 11 07:37:57 morning beagles Apr 11 07:59:00 and bagles? Apr 11 07:59:21 nah, I hate bagles :-D Apr 11 07:59:41 hrhr Apr 11 08:51:46 any1 know to to setup serial ports on the bone? I need tty5 to be open and getting data from a ardrino... The bash "dmesg | grep tty*" give me "console [ttyO0] enabled" Apr 11 08:52:40 its running : Linux omap 3.2.0-psp1 Apr 11 08:52:40 harleyman: google "serial port beaglebone" -> third or fourth hit: http://www.gigamegablog.com/2012/01/22/beaglebone-coding-101-using-the-serial-and-analog-pins/ Apr 11 08:53:05 I will have a look Apr 11 08:53:57 tty5 would be not a serial port Apr 11 08:54:35 why ? Apr 11 08:54:51 So I have to use like tty1 ? Apr 11 08:55:00 ttyO5 would be an omap serial, ttyS5 would be a 8250 serial port etc etc Apr 11 08:55:40 tty1 or tty5 would just be virtual terminals. Apr 11 08:55:52 oki.. Apr 11 08:56:03 does that cpu even have 5 serial ports? Apr 11 08:56:26 if not, i only need one. i'm willing to sell those i don't need.! Apr 11 08:56:41 <_av500_> XorA: I think 2,3 and 4 are fake Apr 11 08:57:00 <_av500_> somebody mixed up 1,5 with 1-5 Apr 11 08:57:17 DISPLAYPORT = '/dev/ttyO2' from http://www.gigamegablog.com/2012/01/22/beaglebone-coding-101-using-the-serial-and-analog-pins/ Apr 11 08:57:26 also keep in mind reference voltages... Apr 11 08:57:57 do I have to set the serial port (TTYO2) up when booting Apr 11 08:58:05 harleyman: do not only skim the article for cool things to copy/paste. really read it. Apr 11 08:58:28 oki.. Apr 11 08:58:32 dm8tbr: you mean like 230V/400V AC50Hz? Apr 11 08:58:33 I see UART1 and UART2 on the connectors :-D Apr 11 08:59:19 LetoThe2nd: well I prefer to measure stuff in V/m and A/m, but that's just me and my wireless faible. ;) Apr 11 08:59:47 dm8tbr: wireless is just smoke and mirrors. Apr 11 09:00:05 * LetoThe2nd always prefers a good old plain honest wire, with two ends. Apr 11 09:00:45 * XorA introduces the mobius network patch :-D Apr 11 09:01:11 XorA: including a quatum tunnel for ipv6 packets? Apr 11 09:01:29 if you like Apr 11 09:04:54 LetoThe2nd: yeah, smoke will start coming out of your pockets (more exactly the gadgets in those) at 2kV/m and 2A/m - don't underestimate a pissed of wireless guy with Nicola as his role model. :-þ Apr 11 09:05:21 so let me understand it rigtig.. cat /sys/kernel/debug/omap_mux/uart1_rxd is uart1 and is = ttyO1 Apr 11 09:07:17 the /sys/kernel/debug/omap_mux only got uart0 and uart1 Apr 11 09:07:38 but there is 6 uart on the bone Apr 11 09:07:43 that's just muxing... Apr 11 09:11:57 so I have to define uart2-3-4-5 ? Apr 11 09:15:11 something.. http://elinux.org/BeagleBoardPinMux Apr 11 09:15:16 dm8tbr: yeah, i know the powers of wireless :) Apr 11 09:16:34 goddam it does have 5 UARTs Apr 11 09:16:49 <_av500_> XorA: thinking of running a BBS :) Apr 11 09:17:06 _av500_: I could write it as a systemd plugin :_D Apr 11 09:19:01 XorA : yeah.. one is USB Apr 11 09:32:21 XorA : "The Beaglebone has 6 serial UARTs. One of those, UART0, is connected to the USB port, but that leaves us with 5 to play with" Apr 11 09:33:05 10 minute job then :-D Apr 11 09:52:52 XorA: and you can use the PRUs to add even more serial ports (4-8) Apr 11 09:55:15 <_av500_> and bitbang a few more with GPIO Apr 11 09:56:43 * koen has seen enough git mergetool -t meld for today Apr 11 09:57:56 solder on a few more FTDI chips as well :-D Apr 11 09:58:05 koen: take it to the next level then. git mergetool -t mold. Apr 11 09:58:16 <_av500_> mould Apr 11 09:58:35 darth? Apr 11 09:59:53 I think I have all beaglebone patches rebased ontop of the new vendor kernel Apr 11 10:03:21 Have any of you experienced constantly red D13 both with OTG and 5V external power? My board was working but after connecting and disconnecting a few times this problem occured Apr 11 10:04:27 my board is a beagleboard xM Apr 11 10:05:04 rev C Apr 11 10:06:19 The voltage from my supply is 5,13V Apr 11 10:11:48 try with exactly 5V Apr 11 10:12:24 <_av500_> 5.0000000000000000V Apr 11 10:12:51 so board is destroyed? Apr 11 10:13:00 not nice to the board Apr 11 10:13:25 red led means the overvoltage protection has kicked in Apr 11 10:13:59 How do I use serial? Do I connect my adapter then power on the board and connect? I keep getting a message from my software saying device or resource busy Apr 11 10:14:31 <_av500_> what adapter? Apr 11 10:14:35 <_av500_> what os? Apr 11 10:14:45 kvarley what exactly do you want to do over serial Apr 11 10:14:47 ahhhh Apr 11 10:14:47 koen, thanks for maintaining a useful kernel Apr 11 10:14:48 sorry Apr 11 10:15:02 I have got it now guys _av500_ and woglinde I had a zombie session Apr 11 10:15:46 kvarley what? Apr 11 10:15:55 you disabled the getty? Apr 11 10:16:26 I tried to connect Apr 11 10:16:29 Then closed the window Apr 11 10:16:36 But the sesion remained active because I used screen Apr 11 10:16:45 So I just killed that session and started again and am now in Apr 11 10:17:03 so you wanted a serial console Apr 11 10:17:04 okay Apr 11 10:17:14 _av500_ Found my resolution error in seconds xD Apr 11 10:17:45 <_av500_> so what was it? Apr 11 10:18:11 "Unable to read uEnv.txt from mmc 0:1" Apr 11 10:18:28 I'll try with uEnv now Apr 11 10:21:03 kvarley is a nice customer fixes all his problems himself Apr 11 10:21:43 I was moaning for the last few days tho xD Apr 11 10:26:34 I've tried my board with voltages from 4,5 to 5,2 volts the red LED is always on. Apr 11 10:27:20 Atle: Check your PSU outputs a DC which is regulated Apr 11 10:27:31 I guess the over protection circuit is damaged Apr 11 10:28:08 I used it first with the suggested 5V 2A PS Apr 11 10:28:54 this worked for some time, but after connecting / disconnecting the red LED started glowing Apr 11 10:29:15 hhmm Apr 11 10:29:25 The red light is also with when I use the OTG Apr 11 10:30:31 arrg, correction the red light is also on when I use the OTG to power the board Apr 11 10:42:57 My uEnv.txt is loaded but my resolution does not change. Can anybody help? The serial output and uEnv.txt is pasted here: http://paste.ubuntu.com/924673/ Apr 11 10:49:31 cat /proc/cmdline Apr 11 10:49:35 and look sat dmesg Apr 11 10:50:51 <_av500_> mute kernel is mute Apr 11 10:51:45 <_av500_> maybe drop the "quiet" Apr 11 10:56:05 koen: http://paste.ubuntu.com/924687/ Apr 11 10:57:42 I connected my 5v power to the OTG port and got my board working, but the red light is still there. Anyway the board is booting, so I'm happy. Apr 11 10:58:00 kvarley: the touchbook has a builtin lcd, so you can't change the resolution Apr 11 10:58:19 koen: I'm using the superjumbo image provided by AlwaysInnovating Apr 11 10:58:33 koen: Running on the beagleboard-xm so it can be changed Apr 11 10:59:38 Is there anyway I can pass the boot parameters from the serial port? Apr 11 11:02:00 o.O Apr 11 11:02:43 Well my script clearly isn't working Apr 11 11:03:20 <_av500_> [ 0.000000] Kernel command line: console=tty1 mpurate=auto buddy=none camera=lbcm3m1 vram=12M omapfb.vram=0:8M,1:4M omapfb.mode=dvi:1024x768MR-24@60 omapdss.def_disp=dvi root=/dev/mmcblk0p2 rw rootfstype=ext3 rootwai Apr 11 11:03:34 <_av500_> that is the command line Apr 11 11:03:43 hm no serial Apr 11 11:03:46 console Apr 11 11:05:02 _av500_: That proves that my uEnv isn't being applied then? Apr 11 11:06:48 <_av500_> the content of your uEnv is wrong Apr 11 11:06:54 Ah ok Apr 11 11:06:58 <_av500_> uEnv and boot.scr do not have the same format Apr 11 11:07:04 <_av500_> boot.scr is a script Apr 11 11:07:09 <_av500_> so it has commands like setenv Apr 11 11:07:16 <_av500_> uEvnt holds key value pairs Apr 11 11:07:28 <_av500_> and aholler messed that up Apr 11 11:07:32 So if I copy the output cat /proc/cmdline and just change the resolution that should work? Apr 11 11:09:36 <_av500_> copy to where? Apr 11 11:09:37 <_av500_> no Apr 11 11:09:48 <_av500_> google for uenv.txt and see the default contents Apr 11 11:10:08 my version did understood boot.scr too, but the init looked too complicated for jkridner_ ;) Apr 11 11:10:47 <_av500_> aholler: still, tell kvarley what to put in uenv.txt Apr 11 11:11:03 Found something here http://elinux.org/BeagleBoardUbuntu#Beagle_Bx.2FCx Apr 11 11:11:24 don't know, depends on u-boot. printenv should help Apr 11 11:12:33 _av500_: fwiw you can do uenvcmd= to emulate boot.scr Apr 11 11:13:01 Ok, so aholler if I copy the output of printenv into uEnv.txt it should in theory work if I amend the resolution there? Apr 11 11:13:04 kvarley: google for uEnv.txt Apr 11 11:13:10 no Apr 11 11:16:17 0.o Apr 11 11:16:25 It...Works Apr 11 11:16:49 Thanks _av500_ , aholler, koen, and all Apr 11 11:17:03 gg knows it all ;) Apr 11 11:17:16 <_av500_> kvarley: so what did you do in the end? Apr 11 11:18:08 Used my serial to find out my uBoot version to figure out whether it was boot.scr or uEnv.txt that I needed Apr 11 11:18:47 Then make uEnv.txt in the first partition of my card, with the following in it: "dvimode=1280x720-16@60 uenvcmd=run loaduimage; run mmcboot" Apr 11 11:31:16 mdp: do you know which IP block this refers to: [ 5.717010] Unhandled fault: external abort on non-linefetch (0x1018) at 0x40305014 Apr 11 11:31:19 ? Apr 11 11:34:02 <_av500_> GPTIMER12 Apr 11 11:37:03 heh Apr 11 11:37:20 "I have no recollection of that event" Apr 11 11:37:39 <_av500_> alien abduction? Apr 11 11:39:17 koen, that's a fault accessing the sram Apr 11 11:39:53 koen, given that PM has to fiddle with and use sram, I'd wager it's due to a clock not being reenabled or similar Apr 11 11:40:23 I will wager 100% of Tartarus' next paycheck Apr 11 11:40:36 I guessed it was a clock problem, I was curious which ipblock it was :) Apr 11 11:41:30 if I want to see whats on com5 on my bone I tryed a "ttylog -b 19200 /dev/ttyO5" with no lock :-( Apr 11 11:42:11 maybe I'm missing a setting somewhere... :-/ Apr 11 11:42:19 <_av500_> pinmux? Apr 11 11:42:39 can be.. Apr 11 11:42:51 its a default ubuntu omap install Apr 11 11:43:48 _av500_ : if you want to setup and read the com5.. how will you do it ? Apr 11 11:45:37 I would use stty Apr 11 11:45:47 if no kernel magic is needed Apr 11 11:45:54 to configure it right for the board Apr 11 11:46:22 stty ? thats a set tty ... Apr 11 11:47:10 man is your friend Apr 11 11:47:22 woglinde : that what I asking about... do I need some sort of kernel magic to use com5 Apr 11 11:47:35 or can I use the stty Apr 11 11:47:52 root@omap:~# man stty Apr 11 11:47:53 bash: man: command not found Apr 11 11:47:53 root@omap:~# Apr 11 11:48:33 stty --help Apr 11 11:48:38 better! Apr 11 11:51:13 stty exist on your desktop too Apr 11 11:51:23 if you dont use the os with the cryptic help search Apr 11 11:51:34 tty on a win7 ? Apr 11 11:52:38 as I said -> if you dont use the os with the cryptic help search Apr 11 11:52:57 but when you install cygwin with man you have man too Apr 11 11:57:10 thats why.. i did't Apr 11 12:09:24 so in the "/sys/kernel/debug/omap_mux/" -thats what can be used? I see uart0 and uart1 only Apr 11 12:10:01 use the source luke Apr 11 12:11:12 the kernel source...I guess Apr 11 12:13:35 right Apr 11 12:18:59 I will use the kernel-source then... Apr 11 12:37:30 <_av500_> koen: optware? Apr 11 12:43:13 _av500_: I'm not touching that with a 10 foot pole Apr 11 12:52:22 yessss!! its working ;-) Apr 11 12:53:02 got data in minicom from my modul Apr 11 12:53:15 so what did you do? Apr 11 12:53:53 kill a chicken and dance the kasatczok Apr 11 12:54:03 moved the modul to uart1 and waiting to compile kernel Apr 11 12:55:35 lol Apr 11 12:55:55 the quick fix.. Apr 11 12:56:54 mdp: I suspect it's the sgx block that is causing the linefetch aborts Apr 11 12:57:02 mdp: or rather the devmem2 magic :) Apr 11 12:58:21 "abort on non-linefetch" means something did an uncached access of an address without anything responding Apr 11 12:58:53 either the address is unmapped or some clock is disabled Apr 11 13:06:05 no-one answer koen and fill his log with errors :-D Apr 11 13:18:08 you guys, still getting this error: http://pastebin.com/fRJgx85C (not my pastebin, someone elses from months ago, I get the same thing) Apr 11 13:19:04 <_av500_> recycling pastebins is ecological! Apr 11 13:19:12 =D Apr 11 13:20:03 recycling it because it was in a chatlog from #oe (i think) where there was some discussion of it Apr 11 13:20:09 here: http://logs.nslu2-linux.org/livelogs/oe/oe.20111217.txt Apr 11 13:20:14 nodejs-native 0.4 doesn't exist anymore Apr 11 13:20:23 so if you're getting that error, your checkout is old Apr 11 13:20:41 i pulled it from git just last night, hmm let me double check Apr 11 13:20:53 the error was the same, it failed on tar because the directory was screwed up Apr 11 13:29:21 confirmed, it's the sgx block Apr 11 13:29:36 <_av500_> unblokc it Apr 11 13:29:41 <_av500_> unblock even Apr 11 13:30:00 ES_REVISION="$(devmem2 0x56000014 | sed -e s:0x00010205:8: | tail -n1 | awk -F': ' '{print $2}')" Apr 11 13:38:37 wtf Apr 11 13:42:09 what is the clean way to read back the image from mtd partition with "ubifs" image? Apr 11 13:44:37 mru: the sgx people said "we don't want to expose the version into sysfs because we might read a wrong memory area" Apr 11 13:44:59 mru: I still don't get how using devmem2 in userspace can be a preferred solution Apr 11 13:45:39 because clearly that could never read the wrong address Apr 11 13:46:09 exactly Apr 11 13:46:28 I bet there's even a #define for the addresses Apr 11 13:46:40 since the driver needs to know about them anyway Apr 11 13:50:05 any1 try doing a full-image backup of the uSD card ? Apr 11 13:50:41 the boss asked me.. "what if the card is broken".. Apr 11 13:50:53 koen: everything is hardcoded in that driver, I don't think it actually bothers to check its running on the right hardware Apr 11 13:52:12 harleyman: use dd Apr 11 13:52:22 good idea ;) Apr 11 14:09:31 hi! has anyone confirmed that the dsp bridge driver actually works on a 3.0 kernel? The driver seems to load accoridng to the startup messages, but I can't seem to interact with it. Just checking if even trying is futile ;) Apr 11 14:10:21 <_av500_> define "can't" Apr 11 14:11:51 for example running a dsp "ping" program gives "DSP device detected!! \nDSPNode_Allocate failed 0xffffffff" Apr 11 14:19:40 koen, seeing that devmem hackery in the sdk scripts the first time led me to rant for hours Apr 11 14:19:52 I've now accepted that this is the status quo Apr 11 14:20:30 koen: why not write your own driver just to return version in sysfs? Apr 11 14:30:38 _av500_: however, strangely, currently the dspbridge module doesn't seem to be loaded. And I can't find a .ko for it, so I'll recompile the kernel again to see if it makes a difference before attempting anything else Apr 11 14:35:26 <_av500_> stippeng: module could be builtin Apr 11 14:35:31 <_av500_> check /proc/devices Apr 11 14:37:05 XorA: I don't work for TI anymore, they should fix that *$(@*($# themselves Apr 11 14:37:30 koen: :-D Apr 11 14:43:03 koen, nothing to be fixed…"it works" Apr 11 14:43:21 unpossible to argue with that Apr 11 14:45:22 _av500_: indeed it is! "248 DspBridge". Can I pass arguments to built in modules in the same way I can pass arguments to modprobe? Apparently the DSP howto says that I should pass the path to the dsp base image to the dspbridge module. Apr 11 14:47:46 <_av500_> yes Apr 11 14:48:35 _av500_: ok, thanks, I'll try that Apr 11 14:59:46 koen: here's the output from my build: http://pastebin.com/Qu0V7NAT Apr 11 15:01:03 i'd pulled the tree from github Apr 11 15:04:03 f00bard: ./oebb.sh update Apr 11 15:23:34 ssvb: thanks for the updated cpuburns Apr 11 15:27:01 koen: ok, ran update, now it's doing something different than before, good sign! do i need to do a clean build though? Apr 11 15:34:18 no Apr 11 15:37:02 heh Apr 11 15:37:02 root@beaglebone:~# systemd-analyze blame 3930ms pvr-init.service Apr 11 15:37:15 4 seconds to do some modprobes and devmem2s Apr 11 15:37:17 awesome Apr 11 15:40:41 koen, you have to have sleeps in there Apr 11 15:40:50 you wouldn't want an actual syncronization mechanism Apr 11 15:41:10 heh thats 26s faster than when I first ran that Apr 11 15:41:20 Russ: I ripped out the sleeps since modprobe will now wait for the devicenode to appear before exiting Apr 11 15:41:56 Russ: I still fail to explain that to the TI engineers writing those scripts somehow :) Apr 11 15:42:02 is it devmem2 doing mmap over and over that is taking a long time? Apr 11 15:42:16 koen: I submitted a patch like that 3 years ago Apr 11 15:42:34 in which case it might be nice if devmem2 could stdin as input and process each line Apr 11 15:42:49 koen: admittedly the omap4 team are probably different bunch of guys working on identicle code :-D Apr 11 15:43:37 XorA: a year ago I did the hdvpss loading with 3 systemd units and 2 udev rules so everything was async with callbacks Apr 11 15:43:58 iirc it went from ~60s to ~12s Apr 11 15:54:04 ugh, now it fails fetching kbd-1.15.2.tar.bz2 from kernel.org. the site doesnt even have any bz2s Apr 11 16:18:13 kernel.org (got) changed a bit last year Apr 11 16:52:05 rcn-ee: you around? Apr 11 16:52:39 I've been tryin Apr 11 16:53:09 the latest precise images on my XM and it seems USB and ethernet are no longer working Apr 11 16:53:36 this was after updating the kernel with the march 28th updates Apr 11 16:54:06 i will try a fresh precise image when i get home tonight Apr 11 16:55:12 also, have you ever tried to use pivot_root to move a RFS? Apr 11 16:59:54 <_av500_> that image does not seem to be too precise Apr 11 17:00:22 smplman, strange, it works here.. is the light on the ethernet adapter? Apr 11 17:00:48 rcn-ee: no the light is not on Apr 11 17:01:11 it could be something i did in the kernel upgrade Apr 11 17:01:20 do you have serial access? if you do, what does "sudo rmmod smsc95xx" "sudo modprobe smsc95xx".. Apr 11 17:01:50 i do have serial access, but I'm not with my hardware atm Apr 11 17:01:56 then 'sudo /etc/init.d/networking restart"... Apr 11 17:02:12 ah.. do you know what kernel you upgrade too? Apr 11 17:02:26 3.2-x7 possibly Apr 11 17:03:06 what would be the best way to setup an external usb drive for the rfs? Apr 11 17:03:24 i have an sad in an adapter that i have been wanting to use Apr 11 17:03:29 ssd* Apr 11 17:04:11 would it be easiest to leave the boot partition on the SD and then copy the rfs to the usb drive Apr 11 17:04:24 the easiest would be to use the "netinstall" from the start, as you can pick which drive to install too, and the script at the end of install detects that and sets uEnv.txt correctly.. Apr 11 17:05:03 but, a straight copy would be fine.. just edit /etc/fstab and uEnv.txt to use "/dev/sda1" instead of "/dev/mmcblk0p2" Apr 11 17:05:04 sweet, i saw something where u-boot would boot an entire usb drive with the kernel Apr 11 17:05:26 ahh, see i was thinking that the rfs location was compiled into the kernel Apr 11 17:05:36 they might be getting closer now, i still use an sd card as a jump point between the two.. Apr 11 17:06:08 <_av500_> rcn-ee: gpio polarity? Apr 11 17:06:18 <_av500_> for the eth+hub powrron? Apr 11 17:06:27 cool, i know i asked without my hardware but you are usually not online at night when i do my tinkering Apr 11 17:06:48 anything i find tonight i will post to the google group Apr 11 17:06:56 nah, that should be long taken care of.. ( i have an xm C and xm A in the lab here working with that image..) i'm thinking it's that random smsc95xx didn't want to start failure that only gruemaster and i ever see.. Apr 11 17:06:58 thanks again Apr 11 17:07:24 <_av500_> rcn-ee: must be a .deb based infection :) Apr 11 17:07:31 * GrueMaster hears his name, lumbers oout of his cave. Apr 11 17:08:06 GrueMaster, did you give up on that xM who's smsc95xx that randomly would not start? Apr 11 17:08:45 I had to. No one on my side would support me, and no one here confirmed the bug. Apr 11 17:09:15 ah, i think smplman might have hit it to.. so that makes 3 of us... Apr 11 17:09:45 i will verify tonight Apr 11 17:09:56 <_av500_> GrueMaster: I never saw that Apr 11 17:09:56 i need to start bringing my hardware to work with me Apr 11 17:10:29 cool thanks smplman, for my one board, by rmmod, then modprobeing, then restarting networking it seemed to come up (light turned back on) etc.. but it was random.. Apr 11 17:12:18 The bug is https://bugs.launchpad.net/ubuntu/+source/u-boot-linaro/+bug/838200 Apr 11 17:12:50 any improvements on sgx in precise? or are we still waiting on TI? Apr 11 17:13:52 checks.. http://software-dl.ti.com/dsps/dsps_public_sw/sdo_sb/targetcontent/gfxsdk/ humm, newer release... Apr 11 17:14:02 btw, I have never had the loss of USB. Just networking. And the rmmod/insmod trick does indeed help (although not always and not a solution when doing a netboot). Apr 11 17:14:24 http://processors.wiki.ti.com/index.php/RN_4_06_00_01 yay ddx 1.7 Apr 11 17:14:27 rcn-ee: that breaks X, we need to poke prabu about that Apr 11 17:14:46 koen, i wouldn't expect less ;) Apr 11 17:14:51 rcn-ee: it's partially drm now Apr 11 17:15:00 so close to the omap4 stuff, yet so far apart Apr 11 17:15:25 i wonder if bolts up to v3.3/3.4 drm.. ;) Apr 11 17:16:43 nah, i better go back to the bone kernel, for a bit.. koen did you notice: http://arago-project.org/git/projects/?p=linux-am33x.git;a=commit;h=e8004dad869568692ca2a45c04e8464ce48cc4d7 Apr 11 17:17:10 looks like 3 new musb patches overall, haven't given them a test yet.. Apr 11 17:17:13 <_av500_> omap5 wont have musb any more Apr 11 17:17:22 <_av500_> what will we rant about all day Apr 11 17:17:37 av500, does it have sata? ;) Apr 11 17:17:44 <_av500_> yes Apr 11 17:17:49 <_av500_> and usb3 og Apr 11 17:17:52 <_av500_> otg Apr 11 17:17:52 _av500_: we can rant about whatever usb it does have Apr 11 17:17:53 rcn-ee: haven't noticed that yet Apr 11 17:17:59 crap.. nothing to rant. what will we do.. Apr 11 17:18:05 <_av500_> mru: of course Apr 11 17:18:31 rcn-ee: I rebased all patches today Apr 11 17:18:41 rcn-ee: cpufreq and cpuidle work now Apr 11 17:19:29 i saw that too, so that firmware now actually works? ;) Apr 11 17:19:42 well Apr 11 17:19:45 resume doesn't work Apr 11 17:19:54 <_av500_> and suspend? Apr 11 17:20:11 it seems to suspend Apr 11 17:20:16 if resuming is not required, all it takes to suspend is pulling the power plug Apr 11 17:22:04 koen: fwiw, PM firmware testing was part of the latest am-sdk Apr 11 17:34:29 koen, any progress on the USB DMA problem on am335x? Running everything in PIO mode is a drag Apr 11 17:35:01 mdp said "RSN" Apr 11 17:35:08 :) Apr 11 17:41:25 usb3? without m? Apr 11 17:47:34 mdp, mranostay: https://github.com/koenkooi/linux/tree/linux-ti33x-psp-3.2-r10a+gitr88dcc46017b970660ff113c1bb4e7d832c5a3693 Apr 11 17:47:50 mmm...thanks Apr 11 17:48:25 mdp: that has the old version of your spifb Apr 11 17:50:23 ok, I don't want to issue a pull request until I incorporate and refactor for the 3 lcd versions properly Apr 11 18:50:18 koen: feeling any better about the UI changes? Apr 11 18:50:38 slightly Apr 11 18:50:53 the trick is to just stop caring Apr 11 18:51:33 koen: not a matter of caring, it is a matter of using it Apr 11 18:52:09 koen: Apr 11 18:52:13 http://pastebin.com/TX0sYsEP Apr 11 18:57:42 koen, release your anger Apr 11 19:04:05 mdp: as a tarball? Apr 11 19:04:25 sounds nasty Apr 11 19:04:47 * mru gives koen some feathers, just in case Apr 11 19:06:17 koen: ^^^ regarding the resume failure, that's a known issue. Apr 11 19:06:24 check http://processors.wiki.ti.com/index.php/AM335x-PSP_04.06.00.07_Release_Notes#Known_Issues Apr 11 19:07:21 Search the page for 90962 Apr 11 19:08:11 fwiw, we observed the problem on 1/3 of the boards we tried Apr 11 19:12:28 cehh: good to see that the issue is public now Apr 11 19:12:55 * koen needs to do a calculation when TI will run out of AM335x ES1.0 silicon Apr 11 19:14:24 koen: and is that sooner or later then the targeted life time build of Bone's ;) Apr 11 19:15:40 hmm, will there be ever bones with a usb-powered ftdi and a normal vid? Apr 11 19:15:52 vid is programmable Apr 11 19:17:55 what do most people using the beagleboard as a springboard for design do about graphics and emissions? Apr 11 19:18:21 cut down on the beans Apr 11 19:18:38 ah... ~.~ Apr 11 19:19:31 and yes that could address both points Apr 11 19:19:36 ;) Apr 11 19:19:58 if you need displays, it is next to trivial to add one on there Apr 11 19:20:03 aholler: the current ones are ftdi powered and have a VID recognized by linux Apr 11 19:23:33 ftdi-powered? Apr 11 19:26:27 I thought if the bone isn't powered, the ftdi isn't powered too. Apr 11 19:27:23 not sure why someone wants this. Apr 11 19:30:38 the ftdi is powered over usb Apr 11 19:31:09 ah, ok. Apr 11 19:35:27 is there a definitive page on how to build beaglebone firmware? going to start from scratch again, i guess Apr 11 19:35:44 like, which git repo i should be pulling from, etc. Apr 11 19:35:58 it's not a firmware Apr 11 19:36:01 anyway: http://www.angstrom-distribution.org/building-angstrom Apr 11 19:40:39 I installed lxde on my beagleboard in angstrom and now videos have stopped scaling in gnome-mplayer Apr 11 19:40:46 Any ideas as to why? Apr 11 19:49:37 hi koen, if you have tried the initial version of my cortex-a8 cpuburn earlier, then the results are already a bit old :) Apr 11 19:49:59 I have uploaded a slightly better version Apr 11 19:52:40 also looks like it attracted some unwanted attention from the Intel camp, so I added some clarifications at the end of the blog post Apr 11 19:54:34 ssvb: I'm using the one I downloaded ~2 hours ago Apr 11 19:54:48 on a usb powered beaglebone: Apr 11 19:54:49 idle: 35 Apr 11 19:54:49 cpuburn-neon: 46 Apr 11 19:54:50 cpuburn-neona8: 51 Apr 11 19:55:09 celsius Apr 11 19:57:01 koen: I have updated it roughly half an hour ago, so that was probably an old version Apr 11 19:57:14 :) Apr 11 19:58:39 the new version works in thumb mode and has some code cleanups, these are the changes - https://github.com/ssvb/ssvb.github.com/commit/c890e37ede76bfcfa2df807b27518f6762ef7c58#files/2012-04-10/ssvb-cpuburn-a8.S Apr 11 20:00:42 seems to add ~100mA in my tests Apr 11 20:06:44 koen: does the temperature sensor in beaglebone provide trustworthy results? Apr 11 20:27:52 anyone ever try to load openpandora on an XM? Apr 11 20:28:14 ssvb: external IR thermometer Apr 11 21:03:52 koen uhd updates done Apr 11 21:03:56 need to do gnuradio Apr 11 21:04:20 and figure out how to submit them without making you burst a blood vessl Apr 11 21:24:20 Hi. I accidentally put 5V into a GPIO pin and my BB won't start at all. Is it totally toast, or is there any hope that I can resurrect it? Apr 11 21:42:31 rcn-ee: is there a precise version of the net install? Apr 11 21:46:27 MattRichardson: that may be a question of RMA for the beagleboard list Apr 11 21:47:00 rcn-ee: never mind, looking at the script I answered my own question Apr 11 21:47:53 djlewis: OK thanks. Judging by the warnings in the SRM, I think I'm SOL. Apr 11 21:48:33 MattRichardson: I suppose you did unhook the board from everything and let it sit for a while? Apr 11 21:48:45 <_av500_> gm Apr 11 21:48:54 _av500_: howdy Apr 11 21:50:05 djlewis: yeah, overnight. Still when I plug it in, I get a super-quick flash of the PWR led, then nothing. Apr 11 21:50:59 * djlewis tries to hold back the tears :( Apr 11 21:51:12 sweet little beagle Apr 11 21:51:54 djlewis: Yes, it served me well. I ordered a couple new ones. I'll miss the little guy. Apr 11 21:52:24 <_av500_> you can at least reuse the power led Apr 11 21:52:30 MattRichardson: you do the board voltage tests in the SRM? Apr 11 21:53:00 rcn-ee: BTW, the net install boots insanely fast Apr 11 21:53:22 djlewis: No, looking for that now, but don't see anything Apr 11 21:55:41 smp4488, yeah isn't it nice? ;) ah, looks like cannoical updated the image again.. Apr 11 22:09:38 rcn-ee: is the net-install a minimal or full? Apr 11 22:30:40 smp4488, it's completely up to you... you can make it minmal, full, or even crazy full... Apr 11 23:09:27 Hm Apr 11 23:09:48 no way to discover what DDRs are connected, so it has to be set at compiletime, right? Apr 11 23:14:31 XMPPwocky: yep, there's no SPD on these boards Apr 11 23:14:42 ;( Apr 11 23:23:51 has anyone gotten "connman" wifi tethering to work using the standard beaglebone build? I can get wifi up and running, but not tethering. Apr 11 23:26:57 am I crazy or is ks0108 support part of the linux kernel? Apr 11 23:27:01 out of the box? Apr 11 23:39:46 well, I have a little ks0108 126x64 monochrome lcd and I want to get it hooked up to show the linux console, etc. I think that I have to get a framebuffer driver running for it? Could someone maybe point me in the general direction I need to look to do more research? Apr 11 23:46:18 git grep ks0108 Apr 11 23:47:02 ? Apr 11 23:48:15 http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git;a=blob;f=Documentation/auxdisplay/ks0108;hb=HEAD Apr 11 23:49:52 that's what I was looking at and that's why I came here to ask more about it Apr 11 23:50:24 it's one of those things that once you get it you get it, but until then it makes 0 sense Apr 11 23:51:47 the kernel driver for it assumes it's hooked to a pc parallel port Apr 11 23:52:48 he just searched a general direction. replacing parport-pins with gpios shouldn't be a problem. Apr 11 23:53:24 okay. So it would need to be modified? Apr 11 23:54:04 yep, you could either create a gpio-parport driver..or gut the auxdisplay driver itself Apr 11 23:54:42 or you can get even crazier by dumping the whole auxdisplay model the author went after…move it to a deferred i/o style FB driver and be able to use standard fbdev apps Apr 11 23:54:58 looks like a lot of fun to me Apr 11 23:55:45 aholler, no argument there…neat project Apr 11 23:56:30 hmm, not sure if it's usable. I've used one of those ks* attached to an avr some years ago ;) Apr 11 23:57:31 it might not be that fast for the console Apr 11 23:57:43 well, you don't have to use console Apr 11 23:58:04 I generally don't run fbcon on my 1.8" spi lcd either :) Apr 11 23:58:17 hmm, okay. Is using the framebuffer driver a matter of kernel parameters or something u-boot would take care of? Apr 11 23:58:41 isaacbw, FB is a kernel framework Apr 11 23:58:52 you would have to rewrite that driver to use FB Apr 11 23:59:40 you should read up on that auxdisplay framework, what it was intended for (simple aux display to desktop systems, high latency stuff) and if that's fine, just stick with that Apr 11 23:59:55 hack it to work with gpio instead of parport write/read calls Apr 12 00:02:22 awesome, thanks. I think I'll read through this: http://linuxconsole.sourceforge.net/fbdev/HOWTO/ Apr 12 00:03:04 take a look at http://lxr.linux.no/linux+v3.3/Documentation/auxdisplay/* too Apr 12 00:04:07 there's also a userspace driver library for low-end lcds too that's popular Apr 12 00:05:17 any idea what it's called? Apr 12 00:05:22 screw it, I'll use the starterware bootloader :? Apr 12 00:05:45 lcdproc is one but I thought there was something more generic as a library Apr 12 00:05:54 they may only support character lcds though Apr 12 00:06:22 mm, okay Apr 12 00:07:39 thanks for the help Apr 12 00:09:07 maybe lcd4linux too Apr 12 00:09:12 worth at least seeing what they do Apr 12 00:09:22 you've got plenty of options Apr 12 00:10:00 kkeller posted on the list an example of bitbanging a nokia display from javascript even..yet another option… Apr 12 00:10:48 Arduino's libraries have a bunch of LCD stuff, might be handy for reference Apr 12 00:11:52 yeah, that's what kkeller simply ported with simple wrappers to "bonescript" fwiw Apr 12 00:12:18 I did the same for the adafruit 1.8" lcd..used the arduino lib for the painful magic init reference Apr 12 00:49:37 is the microsd card that ships with the beaglebone partitioned? I'm assuming I can only see the boot partition when I mount the beaglebone via usb Apr 12 01:29:15 isaacbw: yes, you'll see them all on a linux box Apr 12 01:29:25 boot partition is fat Apr 12 01:31:57 anyone else feel like James Bond when downloading stuff from TI? Apr 12 01:32:25 "(e) I will NOT USE or TRANSFER this Software/Tool for use in any sensitive NUCLEAR, CHEMICAL or BIOLOGICAL WEAPONS, or MISSILE TECHNOLOGY end-uses unless authorized by the U.S. Government by regulation or specific license." Apr 12 01:34:12 yeah i have crap like that when i order a beagle, but for crytographic munitions, probably because openssl is on the sd card >_> Apr 12 01:35:10 because of COURSE terrorists would not break such a promise after they get the hardware... Apr 12 01:35:24 clearly a most effective measure! :) Apr 12 01:37:20 not to mention any terrorist with the resources to build a nuke clearly needs a SoC to make it work Apr 12 02:21:58 moving my finger down on my touchscreen with an app running that shows the touch position, move to the left. Does this mean the touchscreen input is rotated counter clockwise? Apr 12 02:38:12 sounds like it Apr 12 02:42:26 things I have been finding are ugly fixes http://code.google.com/p/android-x86/issues/attachmentText?id=474&aid=4740002000&name=new+file&token=a--x3N0_g2xgMMec7ijJnRlyHik%3A1334195779372 Apr 12 02:43:58 eh, what on earth is the point of android-x86 Apr 12 02:44:01 I've seen it before Apr 12 02:44:09 it seems pretty silly Apr 12 02:44:28 no idea, so you can run it on laptops and pc's Apr 12 02:44:42 again, WHHHHHYYYYYYY Apr 12 02:44:45 heh Apr 12 02:44:54 but its the patches im interested in they used to fix this touch screen Apr 12 02:45:25 you have android running on the beagle? Apr 12 02:45:37 except there are things like apple-hid.c that I see, and dont understand why they are being edited or touched/needed Apr 12 02:45:40 yes I do Apr 12 02:45:48 ICS with hardware acceleration Apr 12 02:46:17 Beagle xm A2 Apr 12 02:50:25 isaacbw, I agree about the why of x86 I just use debian **** ENDING LOGGING AT Thu Apr 12 02:59:58 2012