**** BEGIN LOGGING AT Wed Jun 27 02:59:58 2012 Jun 27 07:31:05 lyakh: \o/ Jun 27 07:34:45 av500: hi, what caused such your great joy? :-) Jun 27 07:35:20 [ANNOUNCEMENT] (tentative) Android generic V4L2 camera HAL Jun 27 07:35:25 gm Jun 27 07:35:29 av500: hehe :) Jun 27 07:35:35 av500 lol Jun 27 07:36:05 libcamera.so liboecamera.so liboevendorcamera.so Jun 27 07:38:23 hi ant Jun 27 07:48:45 morgen woglinde Jun 27 07:49:00 Hello everyone, i'm relatively new to linux and i have some problems with the BeagleBone and USB camera (uEye). The camera appear correctly when i do lsusb and dmesg, but it seems I can't access to it. I use the command ueyesetid (to set the id of the uEye cameras connected) but it says no camera is connected. I heard of some configurations to make when rebuilding a kernel : CONFIG_MUSB_PIO_ONLY=y to disable the DMA (what is it ? Jun 27 07:49:52 av500: let me know when you manage to get that to run on your systems ;-) Jun 27 07:58:28 seb-workfly CONFIG_MUSB_PIO_ONLY is linux-kernel configuration option Jun 27 08:09:36 hmhm why does _omap3 defintion on codec-engine dont works for sakoman multi-omap3 Jun 27 08:09:58 koen: reply to Tim, strike until the iron is hot ;) Jun 27 08:12:03 hm ah found it Jun 27 08:15:22 gm djwillis Jun 27 08:24:01 lol ka6sox on heise.de Jun 27 08:27:53 hi Jun 27 08:27:58 quellarchiv, I have to use that word more often Jun 27 08:28:30 koen is it dutch? Jun 27 08:28:32 hi mahi Jun 27 08:28:50 k we have the project the beagleboard is finalised with medical demo Jun 27 08:29:03 woglinde: no, german Jun 27 08:29:04 tell me the features Jun 27 08:29:16 woglinde: dutch would be 'bronarchief' Jun 27 08:29:23 mahi we should tell you the features of your app? Jun 27 08:30:02 koen so you use it comperable to "zeitgeist" Jun 27 08:30:08 gi ericb Jun 27 08:30:38 woglinde: funny you should mention zeitgeist Jun 27 08:31:18 woglinde: I used it to describe the updated yocto site at collab summit and davest stopped with his talk to comment on me using 'zeitgeist' Jun 27 08:31:28 koen I only heard the us guys like it Jun 27 08:31:37 * koen is well versed in the art of derailing a presentation Jun 27 08:32:18 an ubuntu made a stupid app/lib with the name Jun 27 08:33:32 not ubuntu Jun 27 08:33:41 gnome developers too in love with launchpad Jun 27 08:36:02 that was one of the big hurdles into getting it accepted into gnome proper Jun 27 08:36:29 people who haven't drunk the shuttleworth koolaid hate bzr, launchpad and all that nonsense Jun 27 08:53:05 angstrom don't use /etc/network/interface ? Jun 27 08:53:22 depends what you have install Jun 27 08:53:38 if you have nm or conman installed its maybee not used Jun 27 08:54:03 lyakh: once I figure out what it does... Jun 27 08:54:24 av500: sure, feel free to ask :) Jun 27 08:54:40 its a hal for v4l2 based cameras Jun 27 08:54:57 av500: exactly Jun 27 08:54:57 woglinde: mmm and how can i know what i've installed? XD Jun 27 08:54:57 so if I have a cam that v4l2 supports it can be used in android Jun 27 08:55:03 yep Jun 27 08:55:19 hmm, but that was in TI android before, no? Jun 27 08:55:24 for omap3 at least Jun 27 08:55:30 well, at some point... Jun 27 08:55:42 but if was a ti extension Jun 27 08:55:45 but that was also hacked in in a ti-specific way Jun 27 08:55:46 not generic android Jun 27 08:55:48 right Jun 27 08:55:50 yes Jun 27 08:55:53 for the omap3 isp etc.. Jun 27 08:56:07 and with some hardware-specific stuff Jun 27 08:56:22 well, no, don't think it was for isp, it was for a usb cam, I think Jun 27 08:56:33 samsung also had something similar Jun 27 08:56:45 they had it for their SoC camera interface Jun 27 08:56:53 for ti is was not for usb Jun 27 08:56:59 but also with lots of hardware-specific hacks Jun 27 08:57:00 we used it with a cam on CCDC Jun 27 08:57:11 in versions that I saw Jun 27 08:57:15 they had two HALs Jun 27 08:57:20 one for ISP Jun 27 08:57:26 that one didn't use V4L Jun 27 08:57:29 and one for USB Jun 27 08:57:34 that one was for V4L Jun 27 08:57:51 and it was broken in all recent enough android versions Jun 27 08:59:23 hey Jun 27 08:59:58 where can I make a ticket about a bitbake recipe ? Jun 27 09:00:11 ntp-ssl's URI is now 404 Jun 27 09:00:22 tell koen Jun 27 09:00:49 stuk_gen opkg help Jun 27 09:00:49 mmm last image that i've build have a problem with dhcp Jun 27 09:01:03 woglinde: opkg why? Jun 27 09:02:05 -> 10:54 < stuk_gen> woglinde: mmm and how can i know what i've installed? XD Jun 27 09:02:09 a ok Jun 27 09:02:14 there's a connman Jun 27 09:02:15 :) Jun 27 09:02:58 so connman don't use /etc/network/interface Jun 27 09:05:50 oa wtf /usr/lib/connman/ O_O a lot of ethernet_!"£!%%!" cool Jun 27 09:06:11 stuk_gen - welcome to the dbus era !@#$!@#% :) Jun 27 09:07:11 guyzmo: dbus? why? i use dbus but not see that file Jun 27 09:08:13 a simple way to understand what file is using at now? Jun 27 09:12:58 btw, it's in /var/lib/connman Jun 27 09:14:02 guyzmo: i have 20 files name ethernet_nnnnnnnnnn_cable Jun 27 09:15:13 stuk_gen - nnnnnn should be your mac address in lower case Jun 27 09:15:28 so I assume you have 20 interfaces :D Jun 27 09:15:46 great! thanks Jun 27 09:15:55 guyzmo: no i have only eth0 and usb0 Jun 27 09:16:05 i don't understand why there's a lot O_O Jun 27 09:16:29 maybe you changed your mac address ? Jun 27 09:16:35 me..not Jun 27 09:19:38 or connman has a bug Jun 27 09:19:39 if you dont used a fix one, the driver chooses a random one Jun 27 09:19:40 :-p Jun 27 09:20:07 woglinde - well, the driver shall use the mac address hardcoded in the NIC firmware Jun 27 09:20:08 something have a script to put static ip in the setting? Jun 27 09:20:19 nobody on internet using static... Jun 27 09:20:49 guyzmo if I remember correctly the smc usb ones dont have a hardcodecd mac Jun 27 09:21:39 woglinde - they found that buying an EEPROM is too expensive ? :P Jun 27 09:22:01 woglinde: mmm Jun 27 09:22:13 woglinde: so every time the mac change? O_O Jun 27 09:22:25 fuckkkk Jun 27 09:22:27 ardcoded in the NIC firmware - lol Jun 27 09:22:28 its change XD Jun 27 09:22:29 +h Jun 27 09:22:41 every time that i reboot the mac change XD Jun 27 09:23:11 then it's running in demo mode Jun 27 09:23:22 ynezz: what's you mean? Jun 27 09:23:49 you've to pay for fixed MAC :) Jun 27 09:24:04 me? Jun 27 09:24:20 when i buy the board....thay can put a mac on it?!? Jun 27 09:25:00 so the cip ethernet doesn't have a mac Jun 27 09:25:04 incredible Jun 27 09:25:28 stuk_gen google for setting mac via driver options Jun 27 09:27:31 now i understand the dhcp... Jun 27 09:27:41 but its bugged and get a random ip in the last build Jun 27 09:40:21 Hello everyone, i'm relatively new to linux and i have some problems with the BeagleBone and USB camera (uEye). The camera appear correctly when i do lsusb and dmesg, but it seems I can't access to it. I use the command ueyesetid (to set the id of the uEye cameras connected) but it says no camera is connected. I heard of some configurations to make when rebuilding a kernel : CONFIG_MUSB_PIO_ONLY=y to disable the DMA (what is it ? Jun 27 09:40:55 woglinde: and is there a way to use connman to use a default config without follow the mac? Jun 27 09:45:29 stuk_gen I have no clue about conman Jun 27 09:45:36 I mostly use if/up down Jun 27 09:46:15 seb-workfly CONFIG_MUSB_PIO_ONLY=y is a linux kernel configuration-option Jun 27 09:46:33 which you have to set before building the kernel Jun 27 09:47:30 woglinde: what's is if/up down? Jun 27 09:48:27 stuk_gen the commands which are using /etc/network/interfaces Jun 27 09:49:46 with that setting after i don't use connman but /etc/init.d/network start stop restart and its base on /etc/network/interfaces? Jun 27 09:50:03 yes Jun 27 09:51:05 interesting...but i don't understand tha name USB_PIO... Jun 27 09:51:17 what's is relate to networking? Jun 27 09:52:16 seb-workfly: which kernel? Jun 27 09:52:38 ahahah Jun 27 09:52:48 woglinde: what are you understand XD Jun 27 09:53:06 woglinde: i think you tell to me Jun 27 09:54:48 MUSB is the chipset/platform used for usb on omap3/beagle Jun 27 09:55:06 one of them Jun 27 09:56:10 install a gui and you might be able to use connman Jun 27 09:56:51 don't forget, connman is for embedded systems, therefor it requires a gui ;) Jun 27 09:59:01 and python Jun 27 09:59:38 aholler I thought you can configure it via dbus-send Jun 27 10:01:25 that isn't usable by humans. It's the same state as bluez has had for some years until bluez-tools was developed Jun 27 10:02:37 woglinde: dbus send does not counnt Jun 27 10:02:41 so just wait until a future gsoc is over ;) Jun 27 10:02:47 thats like cat foo > /dev/something Jun 27 10:03:29 also, how does dbus send allow me to preconfigure a setup? Jun 27 10:03:42 where does that config data end up in? Jun 27 10:10:45 av500 /var/lib/conman Jun 27 10:10:46 haha Jun 27 10:10:49 or so Jun 27 10:11:06 as I said I dont have a clue about conman besides its using dbus Jun 27 10:14:55 woglinde: np, I dont blame you :) Jun 27 10:20:03 dbus-send blabla Jun 27 10:20:58 its incredible...when i'm see the finish...another problem give me kiss in the ass ahah Jun 27 10:21:03 kick* Jun 27 10:21:32 your problems are small Jun 27 10:22:26 Hello there Jun 27 10:22:36 hello again Jun 27 10:23:01 i am trying to Sync two BeagleXm Video over omapfbply with netsync Jun 27 10:23:10 basically a wall Jun 27 10:23:19 can any one guide me Jun 27 10:23:59 Hi all, How to install VNC in beablebone Jun 27 10:24:34 i need the source code of beagleboard xm Jun 27 10:24:59 ee bingo Jun 27 10:25:23 means what i cant understand Jun 27 10:25:47 any one of omapfbplay Knowledge ? Jun 27 10:25:48 ash_: your statement doesn't make sense. the BBXM isn't a software Jun 27 10:26:13 common dude Leave him he must be novice Jun 27 10:26:45 VNC(virtual network computer) Jun 27 10:26:48 any one omapfbplay ???? Jun 27 10:27:34 how to download the code from beagle board xm Jun 27 10:27:42 kp av500 might know Jun 27 10:27:54 but I'd recommend to ask him smart questions Jun 27 10:27:56 thanks for pointer Jun 27 10:28:06 solution pls.. Jun 27 10:28:12 ash_ ftp? ssh? mouting the sdcard in your host? Jun 27 10:28:22 KP_: yes, omapfbplay can do that? Jun 27 10:28:38 Ya i know omapfbplay can do that Jun 27 10:28:42 you need to compile it with netsync support on both sides Jun 27 10:28:53 even i have downloaded the Soruce of same Jun 27 10:28:55 then you start the master with the number of slaves Jun 27 10:28:59 av500 what did you use for the videowand at fosdem? Jun 27 10:29:04 and then you start the slaves Jun 27 10:29:10 its done on Anstrom by default Jun 27 10:29:14 once all slaves are started, it should start to play Jun 27 10:29:27 woglinde: omapfbplay Jun 27 10:29:32 hehe okay Jun 27 10:29:41 can u guide me the commands line Arguments Jun 27 10:30:38 i know that Files are pick locally then netsync is use to sync the time and based on time all Files are Sync Jun 27 10:30:53 yes Jun 27 10:31:06 so basically two problem Jun 27 10:31:07 I dont know the exact arguments, but there are only a few Jun 27 10:31:25 ok Jun 27 10:31:39 any idea about Scaling and croping Jun 27 10:32:07 it scales to full screen Jun 27 10:32:19 if you need more fancy stuff, you need to add that Jun 27 10:32:26 in the v4l2 options Jun 27 10:32:40 ya but to create a video wall we must crop the required Video from File Jun 27 10:32:50 ah Jun 27 10:32:55 you need to preprocess that Jun 27 10:33:05 using e.g. libav/ffmpeg Jun 27 10:33:13 preprocess what ? Jun 27 10:33:16 the video Jun 27 10:33:24 no no its there in the Source by Default Jun 27 10:33:32 no Jun 27 10:33:36 i have check the code its there to Scale Jun 27 10:33:43 omapfbplay does not handle cropping or scaling Jun 27 10:33:50 for a 2x2 wall, you need to create 4 videos Jun 27 10:33:54 its Does Jun 27 10:33:57 each being 1/4 of the original Jun 27 10:34:04 wait let me show you the code Jun 27 10:34:26 no no omapfbply can Scale and Crop Jun 27 10:34:45 struct frame_format { unsigned width, height; unsigned disp_x, disp_y; unsigned disp_w, disp_h; unsigned y_stride, uv_stride; enum PixelFormat pixfmt; }; Jun 27 10:34:52 look at this Struct Jun 27 10:35:07 unsigned disp_x, disp_y; are Start Disp Area Jun 27 10:35:22 yes Jun 27 10:35:25 nsigned disp_w, disp_h; are Disp Width Jun 27 10:35:36 and Jun 27 10:35:49 set_scale(&dp, &frame_fmt, flags); Jun 27 10:35:53 in Source Jun 27 10:35:57 so its there Jun 27 10:36:12 well fine then Jun 27 10:36:17 lunch now... Jun 27 10:36:25 humm Jun 27 10:36:27 ok Jun 27 10:36:45 where are u from ? Jun 27 10:37:41 he is in the same country than me Jun 27 10:38:42 hehe Jun 27 10:38:50 and u r from which country :) Jun 27 10:40:02 djeez Jun 27 10:40:12 'you' has an 'y' 'o' and 'u' Jun 27 10:40:24 'are' has an 'a' 'r' and 'e' Jun 27 10:40:52 KP_: either fix your keyboard or stop using 12 year old speak Jun 27 10:40:58 (or both) Jun 27 10:41:39 12a ago people knew how to write you and are Jun 27 10:42:18 maybee KP_ is ten Jun 27 10:42:22 did't get you ? Jun 27 10:42:22 who knows Jun 27 10:42:27 whts the matter ? Jun 27 10:42:44 spelling Jun 27 10:42:53 KP_ we are not in the bronx here bro Jun 27 10:43:29 maybee KP_ is ten Jun 27 10:43:37 then please check your Spell as well Jun 27 10:44:10 i am more into R&D Stuff less into Chat Culture Jun 27 10:44:43 i am here to Fix Stuff and to learn chating culture .... please get my point Jun 27 10:45:03 *not to learn chat Jun 27 10:45:13 500 pts Jun 27 10:45:15 KP_ - writing good english helps for good communication between people so other people can understand what you write, that's not a matter of chat culture Jun 27 10:45:34 l33t sp34k is part of chat culture :p Jun 27 10:45:42 c00l Jun 27 10:45:50 ok dude understood and taken your point Jun 27 10:46:00 use some ph0rums Jun 27 10:46:01 now please come to the point Jun 27 10:46:36 do you have any idea about omapfbplay ? Jun 27 10:50:12 personnally I don't, never used that Jun 27 11:03:17 KP_: splitting the video before passing it to omapfbplay is the best option Jun 27 11:03:27 KP_: that's how all the beagleboard videowalls are doing it Jun 27 11:04:00 * koen stabs touchscreen controllers Jun 27 11:16:51 Does someone use an USB camera uEye ? I detect it on the beaglebone (kernel 3.2.14 and Angstrom) but i can't use it, it's not detected as a connected camera by the related daemon Jun 27 11:18:51 oh my god Jun 27 11:19:02 he is asking it the third time Jun 27 11:19:13 and didnt react on the other answers Jun 27 11:20:38 Sorry, i saw your answers just now, i forgot the chat. Thanks for your answers about MUSB. Jun 27 11:23:59 I'll try to stay in front of it now ^^. Can someone tell me what is the DMA and what is the point of disabling it with CONFIG_MUSB_PIO_ONLY=y in the case of an USB camera please ? Jun 27 11:24:32 direct Jun 27 11:24:35 memory Jun 27 11:24:39 address(ing)? Jun 27 11:25:33 hm latest driver version is 4.0 Jun 27 11:25:39 on the ids site Jun 27 11:25:47 i think it's direct memory access, but i don't understand what it doest exactly Jun 27 11:25:57 where did you find to disable dma for musb? Jun 27 11:26:31 in a google group someone tells : I have been having problems with USB cameras on the Beaglebone. One workaround I found that partially works is to disable DMA by setting: CONFIG_MUSB_PIO_ONLY=y Jun 27 11:26:33 how to install VNC(virtual network computer) in beaglebone Jun 27 11:26:42 solution pls..... Jun 27 11:27:10 beats me Jun 27 11:27:18 i need some specific drivers for ARM (i got it and the daemon runs correctly) Jun 27 11:27:22 Tamilarasi opkg help Jun 27 11:27:48 the problem the camera is not recognized by the daemon Jun 27 11:28:10 seb-workfly ask the manufactor I guess Jun 27 11:29:30 hm readme says Jun 27 11:29:33 check /usr/local/share/ueye/ueyeethd/ueyeethd.conf Jun 27 11:29:33 IDS dont support ARM, and the stemmer-imaging support don't know too, i asked them already and i think they are looking for a solution now, but i don't even know the cause of my problem Jun 27 11:29:54 get another camera Jun 27 11:29:56 did it already, it's empty Jun 27 11:30:11 that's a solution yes ^^ Jun 27 11:30:17 thats maybee the problem Jun 27 11:30:20 that it is empty Jun 27 11:30:35 try to install the stuff on an i386 machine Jun 27 11:30:48 and look than into that file, when the camera is runnning Jun 27 11:33:11 I never would buy a camera which do not support v4l interface Jun 27 11:37:43 ;Ni1 ; Generated by Nickel v1.1.0 (see chaoslizard.org/devel/bohr). Jun 27 11:37:55 always the same thing Jun 27 11:38:07 and the amera is detected on my PC Jun 27 11:38:54 what does CONFIG_MUSB_PIO_ONLY=y actually do ? Jun 27 11:56:29 seb-workfly disable DMA for usb Jun 27 11:56:46 which normaly means higher cpu load and maybee speed decreases Jun 27 12:02:09 hello! How to compile kernel for bb-xM with ethernet support Jun 27 12:02:36 bitbake virtual/kernel Jun 27 12:02:47 mine kernel (both from beagleboard git repo and vanilla) doesn't detect smsc chip, so i suppose the kernel is modified Jun 27 12:04:12 stuk_gen: so I need to build openembedded, right? Jun 27 12:05:51 i'm using it Jun 27 12:06:07 this kernel is patched, or is it vanilla? Jun 27 12:06:19 or you dont know Jun 27 12:09:41 i don't know...i only know that works Jun 27 12:10:03 cool, Ill try Jun 27 12:13:34 aleek o.O the official kernel support the smsc chipset, dont know where you got a non working kernel Jun 27 12:14:26 wonit works, but It doesnt detect the chip Jun 27 12:14:37 it works* Jun 27 12:14:42 woglinde: ^ Jun 27 12:14:45 what are you running on your xm? Jun 27 12:14:57 but I dont have usb0 device Jun 27 12:15:52 woglinde: I diffed my kernel's booting log, and original and I have no "smsc95xx" device detected which is plugged to ehci Jun 27 12:15:52 did you try ifconfig -a? Jun 27 12:15:58 woglinde: yes I did Jun 27 12:16:18 again what distribution you are running on your xm? Jun 27 12:16:24 none Jun 27 12:16:33 vanilla kernel + buildroot generated rootfs Jun 27 12:16:45 o.O Jun 27 12:16:49 ok thanks but what is the advantages ? ^^ Jun 27 12:17:05 I tried also kernel from https://gitorious.org/beagleboard-validation Jun 27 12:17:07 seb-workfly that your usb cam might work Jun 27 12:17:20 and I downloaded config from /proc/config.gz Jun 27 12:17:25 also doesnt work ;) Jun 27 12:17:30 ok thank you i'll try ^ Jun 27 12:18:26 some guys in beagle mailing list tell to include usb eth gadget but this is very different driver Jun 27 12:18:34 and it doesnt work either Jun 27 12:18:53 aleek pastebin your kernel-config Jun 27 12:19:44 aleek otherwise try the stuff from here Jun 27 12:19:45 http://downloads.angstrom-distribution.org/demo/beagleboard/ Jun 27 12:19:48 that should work Jun 27 12:20:05 woglinde: yes that should work, but it doesnt have IPv6 which I need ;) Jun 27 12:20:11 woglinde: http://wklej.org/id/780671/ Jun 27 12:20:55 original kernel provided with the board works very well actually Jun 27 12:21:17 that is the angstrom kernel Jun 27 12:21:20 sheesh Jun 27 12:21:22 some people Jun 27 12:21:38 hm strange you have the drivers compiled in Jun 27 12:22:00 aleek which vanilla kernel version? Jun 27 12:22:05 you might need patches Jun 27 12:22:15 * koen looks at his beagleboards Jun 27 12:22:25 network working with ipv6 out of the box Jun 27 12:22:30 this kernel is from https://gitorious.org/beagleboard-validation and the config is taken from *working* kernel from /proc/config.gz :) Jun 27 12:22:37 koen haha I thought that Jun 27 12:22:56 I tried also vanilla 3.1 with omap2plus config with some modifications Jun 27 12:23:08 aleek: follow the instructions at http://www.angstrom-distribution.org/building-angstrom Jun 27 12:23:13 that will build a known good kernel Jun 27 12:23:22 cool! :) Jun 27 12:23:43 if you have a working cross-compiler and know what a 'patch' is: https://github.com/beagleboard/kernel/ Jun 27 12:24:14 * koen is amazed at the number of people claiming they *need* to build their own kernel but don't know what a patch is Jun 27 12:24:39 yes, I know :D Jun 27 12:24:55 koen is gstreamer-ti working with kernel 3.2? Jun 27 12:25:24 koen: I have a systemd-service the starts something using the sgx. On 'reboot', /etc/init.d/pvr-init runs before my systemd service is stopped, and It Crashes. got an idea how I can make my services shut down first? Jun 27 12:25:26 koen: did you see the guy on the ml who said he'd written a compiler? Jun 27 12:26:10 woglinde: as well as it worked with other kernels Jun 27 12:26:11 koen: both my services are WantedBy=multi-user.target Jun 27 12:26:12 mru: I did Jun 27 12:26:36 tasslehoff: Before=/After= Jun 27 12:26:44 tasslehoff: but sgx will crash anyway :) Jun 27 12:27:16 * mru wonders if 'develop' is the new word for 'port' Jun 27 12:27:47 koen hm and whats with ti-codec-engine release 3.x? Jun 27 12:28:12 woglinde: for me run only with mplayer Jun 27 12:28:22 with qt phonon i have a little problem Jun 27 12:32:53 koen: nah.it doesn't if I add a "sleep 2" before the modprobes in pvr-init (not as hard, at least) Jun 27 12:34:36 * guyzmo happy Jun 27 12:34:44 I got a working kernel \\o o// \o/ Jun 27 12:34:53 guyzmo lol Jun 27 12:35:50 I have beaten the mystics of bitbake Jun 27 12:36:24 o.O Jun 27 12:43:12 btw, I don't understand, in a previous build (on my old build machine) I had the crontab file fixed, and now with a fresh setup-tools setup on the new build machine, the crontab file is erroneous again Jun 27 12:50:29 I'm having trouble using the HP Disk Formatter on W7. Running as Administrator, I get no choices in the drive select dropdown Jun 27 12:51:27 I have the SD card that came with my A5 Bone in an MMC slot, with Windows showing an unreadable E:\ drive (it's the FAT partition) Jun 27 12:51:57 who says it's FAT? Jun 27 12:52:19 mikeg dont use windows it always fools you Jun 27 12:52:27 I ripped the E:\ drive into an img file (~70MB) and read it with a Linux VM Jun 27 12:52:43 mikeg: little we can do to fix windows Jun 27 12:52:51 use the linuxvm to access your mmc reader Jun 27 12:53:44 Yeah that'd be good, but even when VMWare Player shows the SmartCard reader as shared with the guest, nothing appears in the VM Jun 27 12:54:00 try virtualbox Jun 27 12:54:09 VirtualBox can't even find the MMC Jun 27 12:54:29 you might need to install usb2.0 guest package Jun 27 12:54:50 or windows host usb2.0 package Jun 27 12:54:52 hm Jun 27 12:54:57 let me see if this exist Jun 27 12:55:08 Yeah I have the "Extension Pack" installed Jun 27 12:55:12 okay Jun 27 12:55:14 sorry Jun 27 12:55:33 No worries, thanks for trying :) Jun 27 12:55:34 than I am out of clues Jun 27 12:55:39 ah I have one Jun 27 12:55:46 use a usb/cdrom distro Jun 27 12:56:03 and than the mkcard.sh script Jun 27 12:56:34 Yeah I think I might be able to get it to work by running linux or by trying a USB SD Card adapter Jun 27 13:52:30 why http://beagleboard.org/project/ubuntu/ links to http://elinux.org/BeagleBoardUbuntu as the homepage? why not https://wiki.ubuntu.com/ARM ? Jun 27 13:56:20 ubuntu page is more generic Jun 27 13:56:34 elinux page is ubuntu on bb specific Jun 27 13:57:08 do you know if ubuntu supports the image on elinux? Jun 27 13:57:43 I mean if there is any wrong with the image or bug Jun 27 13:57:52 use the ubuntu one Jun 27 13:58:17 if there is bug ubuntu will not do anything about it Jun 27 13:58:27 it took them only 3 months to fix mysql bug Jun 27 13:58:40 they release every 0.6 years :) Jun 27 13:58:55 as in from point where there was CVE number assigned and they were notified Jun 27 13:58:56 jacekowski, there was a mysql bug on beagle ? Jun 27 13:59:04 ogra_: no Jun 27 13:59:20 but i'm talking about ubuntu approach to getting things done right Jun 27 13:59:25 angs: http://beagleboard.org/project/ubuntu/ is a PROJECT that somebody entered Jun 27 13:59:32 its not official info Jun 27 13:59:37 or week from point where everyone on internet knew about the bug Jun 27 14:00:07 av500: it is more clear for me now, thank you Jun 27 14:13:00 hi rcn-ee Jun 27 14:13:11 hey woglinde how's it going? Jun 27 14:13:21 okay Jun 27 14:13:32 still trying gstreamer-ti Jun 27 14:15:36 woglinde, ??? Jun 27 14:18:34 ka6sox as head of webos opensource devs Jun 27 14:18:46 * ka6sox falls over. Jun 27 14:19:16 here in the article in english Jun 27 14:19:16 woglinde, got link? Jun 27 14:19:18 http://www.h-online.com/open/news/item/HP-releases-Community-Edition-of-webOS-as-open-source-1626782.html Jun 27 14:19:52 in accurate.... Jun 27 14:19:57 oh well.... Jun 27 14:20:45 somebody is wrong on the internet...... Jun 27 14:21:11 I *must* stay up till its fixed? Jun 27 14:21:59 a few more translations and I could teach a pig to fly and sing... Jun 27 14:22:26 hm the german article has 52 comments Jun 27 14:24:41 Quesiton Jun 27 14:24:50 Answer Jun 27 14:24:53 is there a particular kernel module I need for gst-dsp to work? Jun 27 14:25:00 yes Jun 27 14:25:17 dsp-bridge driver Jun 27 14:25:26 instead of syslink Jun 27 14:25:26 Is that the Ti one? Jun 27 14:25:36 kernel patch or something else? Jun 27 14:25:37 they are both from ti Jun 27 14:25:51 omap3 has no syslink Jun 27 14:25:52 Sh0rtWave you need kernel .32 Jun 27 14:25:58 ir has dspbridge or dsplink Jun 27 14:25:58 args Jun 27 14:26:03 sorry Jun 27 14:26:03 it* Jun 27 14:26:05 I'm using a pre-configured kernel at the moment. Jun 27 14:26:07 to many names Jun 27 14:26:18 which seems to have led me wrong as much as buildroot did. Jun 27 14:27:24 got source for 3.2.8, that should have everything in it Jun 27 14:27:41 aight, let me configure this then Jun 27 14:27:46 where should I look for that driver? Jun 27 14:28:43 in the kernel's menuconfig Jun 27 14:30:58 hrm Jun 27 14:31:00 no dice? Jun 27 14:31:26 yes Jun 27 14:31:31 in the kernel congig Jun 27 14:31:33 config Jun 27 14:31:49 av500 the driver is still in 3.2.8? Jun 27 14:32:25 I have no idea Jun 27 14:32:29 Yes, but the question is where to look for it? Jun 27 14:32:36 its used in meego and meego was mainlin, no? :) Jun 27 14:33:01 ah right flipec Jun 27 14:33:08 hm dont knoww Jun 27 14:33:17 meego had patches too Jun 27 14:33:33 unpossible!!!!!1111! Jun 27 14:33:42 av500: funny, I have plasma-active onmy beaglebone right now :) Jun 27 14:33:49 hah Jun 27 14:33:51 meego is finish for "the patches are upstream" Jun 27 14:34:00 av500: heh Jun 27 14:37:43 koen, is the active plasma contained within the warp field? Jun 27 14:41:53 found it Jun 27 14:43:10 hm looks like its in staging Jun 27 14:43:20 http://lkml.org/lkml/2012/4/20/554 Jun 27 14:44:08 yeah I found it already Jun 27 14:44:11 kernel building now Jun 27 14:44:15 correction: built Jun 27 14:44:18 you might need patches Jun 27 14:46:38 we'll see Jun 27 14:48:43 I would like to use a usb wifi adapter that has ralink 2880. on beaglebone A5. I installed kernel-module-rt2800usb package. I do not see any difference in the lsmod > http://pastebin.com/tjhCAgAg do I need to install any other package? Jun 27 14:50:05 yay, DSP initialized Jun 27 14:50:36 tho I do see some noise about a deprecated ioctl Jun 27 14:51:46 Sh0rtWave, procwrap_detach? Jun 27 14:51:55 hi Jun 27 14:52:26 please any one help to my problem Jun 27 14:52:57 please kindly response to my problem Jun 27 14:52:59 which problem Jun 27 14:53:25 we can't, we don't know your problem Jun 27 14:53:46 Sh0rtWave did you installed tidsp-binaries and dsp-tools yet? Jun 27 14:53:56 dm8tbr the problem is the problem Jun 27 14:54:09 now be nice and read this: http://www.catb.org/~esr/faqs/smart-questions.html Jun 27 14:54:39 dm8tbr that might be the solution for his problem Jun 27 14:54:41 the link might solve /that/ problem ;) Jun 27 14:54:48 *g* Jun 27 14:54:54 :D Jun 27 14:55:58 I interfaced GPS receiver to beagleboard c4 as iam getting NMEA data sucessfully when iam opening the maps on beagle board iam getting the exact location on the map Jun 27 14:56:41 so you want the wrong location? ;) Jun 27 14:56:42 so no problem all works Jun 27 14:56:57 sorry iam not the location Jun 27 14:57:08 but I wonder how google maps reads the gps device Jun 27 14:57:14 is your gps in a building? Jun 27 14:57:20 not getting the exact location on the map Jun 27 14:57:34 i kept out side Jun 27 14:57:46 near a building? atleast 20-30ft away? Jun 27 14:57:52 yes Jun 27 14:57:52 santosh567 what program do you use for localization? Jun 27 14:58:30 I hate webdevs Jun 27 14:58:34 I have ported Angstrom OS on beagleboard c4 Jun 27 14:58:41 o.O Jun 27 14:58:49 really? that's big news. ;) Jun 27 14:59:00 (sorry, I had to say it out loud, there's one getting me crazy right now) Jun 27 14:59:01 angstorm was there all the time for bb c4 Jun 27 14:59:16 guyzmo shoot him? Jun 27 14:59:26 woglinde - if only I could.. Jun 27 14:59:30 I interfaced GPS receiver to beagleboard c4 as iam getting NMEA data sucessfully when iam opening the maps on beagle board iam not the getting the exact location on the map Jun 27 14:59:35 santosh567 again what program do you use for localization? Jun 27 14:59:43 i have installed gpsd also Jun 27 14:59:47 ah Jun 27 14:59:53 yeah gpsd Jun 27 15:00:08 how long did you stay outside? Jun 27 15:00:32 locking from coldstate may need half an hour Jun 27 15:00:40 depending on the gps device Jun 27 15:00:49 use cgps on console to see if you have a fix Jun 27 15:00:54 when i opening the maps at back in terminal it is getting gpsd port not set,gpsd address port not set Jun 27 15:01:04 lol Jun 27 15:01:11 now you see your problem Jun 27 15:01:21 you configured either gpsd not right Jun 27 15:01:31 or the programm which uses gpsd Jun 27 15:01:40 the valid is getting after 10 mins Jun 27 15:01:50 yes Jun 27 15:01:53 than fix your map programm Jun 27 15:01:57 what ever this is Jun 27 15:02:01 and as I said try cgps Jun 27 15:02:06 how should i must configue ? Jun 27 15:02:23 i will try cgps 2maro Jun 27 15:02:25 if cgps dont works other programs which using gpsd dont work also Jun 27 15:02:38 do we have patch for this beagle kernel bug: http://pastebin.com/hAgsRSep ? Jun 27 15:02:45 can u give ur email id please Jun 27 15:02:46 try telnet localhost gpsd Jun 27 15:02:49 no Jun 27 15:02:54 is the kernel 3.2.21 available for beaglebone? Jun 27 15:03:05 you can google yourself how to connect to gpsd Jun 27 15:03:06 angstrom dist? Jun 27 15:03:37 no Jun 27 15:03:45 santosh567 what no? Jun 27 15:03:45 The gamin build has version 3.2.8 I believe Jun 27 15:03:49 http://blog.galemin.com/2011/06/buildroot-2011-05-all-in-one-for-beagleboard-xm/ Jun 27 15:03:54 santosh567 you cannt use google Jun 27 15:04:25 sanstosh567 or you cannt read the docs at http://catb.org/gpsd/ ? Jun 27 15:04:35 i think we can use openstreet maps Jun 27 15:04:54 gpsd has nothing to do with openstreetmap Jun 27 15:05:18 ok Jun 27 15:05:49 try telnet localhost gpsd means Jun 27 15:06:14 okay have to go till later Jun 27 15:06:50 woglinde, 3.4.4, oneiric, gst-dsp is playing back bigbuckbunny... Jun 27 15:07:07 on my xM.. (took awhile to downlolad the file) Jun 27 15:07:34 hello sir woglinde can u give ur email id Jun 27 15:09:15 how to acess the bluetooth gps receiver to BeagleBoard c4 Jun 27 15:09:44 can any have idea Jun 27 15:11:08 Anybody hooked up GPS GS407 with BeagleBone Jun 27 15:11:18 Cause I'm about to do this right now Jun 27 15:11:26 following a Netduino example Jun 27 15:11:27 and? Jun 27 15:11:49 I will appreciate assistance if someone has personal experience Jun 27 15:14:10 how to acess the bluetooth gps receiver to BeagleBoard c4 Jun 27 15:15:43 Use usb/bluetooth? Jun 27 15:16:58 usb bluetooth Jun 27 15:17:43 i have interfaced usb bluetooth dongle to BB c4 Jun 27 15:18:27 now i want to access my bluetooth gps receiver from BB c4 how any steps or procedure Jun 27 15:19:04 joelagnel can u tell me Jun 27 15:19:28 Ok so Jun 27 15:20:07 can any one tell what is cgps? Jun 27 15:21:00 can any one tell what is cgps? Jun 27 15:21:08 can any one tell what is cgps? Jun 27 15:21:27 can you repeat that one more time? Jun 27 15:21:53 joelagnel: is your github tree up to date? Jun 27 15:22:07 (and hasn't anyone told you github is verboten at TI? ;) ) Jun 27 15:22:19 what? Jun 27 15:22:36 I guess robclark didn't get that memo either... Jun 27 15:22:51 different part of TI Jun 27 15:23:04 I guess things went way too smoothly Jun 27 15:23:45 koen, "the other TI" Jun 27 15:24:39 Can anyone help me i have interfaced USB gps receiver to beagleboard c4 Jun 27 15:25:20 ehh Jun 27 15:25:26 what is the problem Jun 27 15:25:31 do you have google ban? Jun 27 15:25:39 i am getting the data NMEA sucessfully after 10 mins but it is not showing the exact location the maps Jun 27 15:26:03 As iam using angstrom Os Jun 27 15:26:23 so mgmt "kindly" suggested not to use github, but keep everything internal Jun 27 15:26:42 you know, kill off all motivation Jun 27 15:26:52 can't have things working, you know Jun 27 15:27:13 i don't care what management says Jun 27 15:27:22 so far it has served me well Jun 27 15:27:39 +1 Jun 27 15:28:01 As iam using angstrom OS and iam using the tangogps for maps Jun 27 15:28:01 if they want to fire me over some silly administrative detail, so be it Jun 27 15:28:39 hah Jun 27 15:28:44 So Jun 27 15:29:02 i am getting the data NMEA sucessfully after 10 mins but it is not showing the exact location the maps Jun 27 15:29:03 does Ti hate github because it makes them more open and spills their corporate beans or do just hate the "here's mine, gimme yours" culture? Jun 27 15:29:23 Can anyone help me i have interfaced USB gps receiver to beagleboard c4 Jun 27 15:29:39 i am getting the data NMEA sucessfully after 10 mins but it is not showing the exact location the maps Jun 27 15:29:58 its getting errors like gpsd port not set Jun 27 15:30:00 there is no reason to repeat yourself Jun 27 15:30:01 joelagnel: btw, if 720p works we should make that the default for the dvi cape as well Jun 27 15:30:11 santosh567: have you tried googling for that error? Jun 27 15:31:00 i cant understang how to set the gpsd address port Jun 27 15:37:16 koen, 720p works but the pattern of white lines still appears Jun 27 15:37:19 do you have a board? Jun 27 15:37:33 even an older one would show you the pattern. Jun 27 15:37:37 man I tell you Jun 27 15:37:56 I fail to see how people can say shit like "my boss approved your invoice" to me when I have a clear email chain of me doing *exactly* what they wanted. Jun 27 15:38:05 WTF does that even mean? Jun 27 15:38:12 Like I have to have his approval for money he fucking owes me? Jun 27 15:38:23 Anyway Jun 27 15:46:28 I have interfaced USB gps receiver to theBB c4 and iam using Angstrom OS ,i am getting the data NMEA sucessfully after 10 mins but it is not showing the exact location the maps Jun 27 15:47:19 san please go to your boss and get you fired Jun 27 15:47:32 when i opened the maps its showing like in terminal Jun 27 15:48:05 santosh please read about gpsd and what is does and come back Jun 27 15:48:17 gconf repo_name not set REPOLIST == NULL *** on_drawingarea1_configure_event(): pixmap created ###### ERROR - Could not open OUTFILE: No such file or directory aieee: outfile=NULL connection to gpsd SUCCEEDED *** on_drawingarea1_configure_event(): pixmap created * mouse drag +8events * mouse drag +8events * mouse drag +8events *** on_drawingarea1_configure_event(): pixmap created * mouse drag +8events *** reset_gpsd_io(): connectio Jun 27 15:48:27 joelagnel: I have an hdmi cape, yes Jun 27 15:48:29 oh man Jun 27 15:48:34 oh cool, ok Jun 27 15:49:01 koen, I have seen that the pattern of lines changes with the timings (they get closer and merge), but nothing as of why this happens Jun 27 15:49:13 interference? Jun 27 15:49:27 even with standard timings, the picture has those lines Jun 27 15:49:30 the color is ok. Jun 27 15:49:48 koen, hmm, well its a fixed pattern so don't think so Jun 27 15:50:02 hw team is schematic/checking routing now i understand Jun 27 15:50:23 *re-checking schematic Jun 27 15:51:30 the input to the hdmi chip looks ok. even with all zeros in the fb, you'd see white lines on a black background Jun 27 15:51:32 funny how people do want to remove resistors of the beaglebone, but not rebuilt a kernel Jun 27 15:52:00 it could be a clock line or something leaking into the signals Jun 27 15:52:11 or you put the chip in nxp-partymode Jun 27 15:52:21 ;) party never gets over Jun 27 15:53:38 * koen orders a hit on the TI person behind the 4.3" capes Jun 27 15:53:43 WTF Jun 27 15:53:50 "everything is working, we have software ready" Jun 27 15:53:59 haven't seen sw yet Jun 27 15:54:13 and the schematic swaps gnd with SCL Jun 27 15:54:31 good fucking job on that Jun 27 15:55:24 but that project was from before I joined Jun 27 15:55:37 quick question for PRPLague regarding trainer xM direction jumper #2 - does it control the direction of all pins on U4 chip or just the one pair of pins? I wanted to add a reset button for trainer xM, but could only get it to reset when grounding pin 26 on the expansion header and not TP7 Jun 27 15:55:42 nowadays if someone says "I have it working" I say "send me the codez and exact steps" Jun 27 15:57:00 I could always just wire a little push button to pin 26 in the proto area :) but I was just wondering abou the direction jumper 2 :) Jun 27 16:08:47 ls Jun 27 16:08:59 is the beagle bone board designed such that a gigE PHY could be installed in place of the current PHY so gigabit would work? Jun 27 16:09:03 no such file or directory Jun 27 16:48:10 linuxjacques, no Jun 27 16:49:01 bradfa, are the board traces not able to handle the bandwidth? Jun 27 16:49:43 or is it missing lines (IIRC gigabit has more) Jun 27 16:50:17 linuxjacques, I believe the pin mux doesn't put both mii and gmii on the same balls Jun 27 16:50:22 but check that in the trm Jun 27 16:50:55 bradfa, my real goal is to find the max network performance of the AM3359, so a "real" dev board may be in order :-\ Jun 27 16:51:02 bradfa, thanks for the pointers Jun 27 16:51:43 linuxjacques, I think you'll have to spin a board then, I don't think any existing am335x boards I've seen do gigE Jun 27 16:52:02 * bradfa checks datasheet (not trm, pin mux is in datasheet, sorry for confusion) Jun 27 16:52:44 oh, in that case, I wonder how well tested the gigE functionality is ... Jun 27 16:53:21 well, if it's any indication, don't run rgmii, there's errata on the clock line Jun 27 16:53:45 needs external clock for rmii (and rgmii? I have to look again) Jun 27 16:53:47 heh, good point. I dl'd the errata but haven't looked at it yet Jun 27 16:53:53 definitely read it Jun 27 16:53:58 will do Jun 27 16:54:10 points about usb and ethernet that are needed to understand if you spin a board Jun 27 16:55:15 yeah, my requirement is high performance USB and ethernet - basically DMA data in over USB and send it out the ethernet port as fast as possible Jun 27 16:55:34 if you run Linux, and use TI's kernel, enjoy USB :) Jun 27 16:55:40 with possibly some processing of the data in between Jun 27 16:55:55 yep, this would be linux Jun 27 16:57:48 linuxjacques, actually... I might be wrong about the pin mux Jun 27 16:59:10 linuxjacques, in fact, I think I am wrong Jun 27 16:59:35 hmmm Jun 27 16:59:49 you may be able to do rgmii over the same pins Jun 27 17:00:12 but, whether the lengths are matched correctly and pinned out on the phy side, I dunno Jun 27 17:00:22 but an external clock would be required? Jun 27 17:00:24 if you mail the list and ask, I'm sure Gerald will answer quick Jun 27 17:00:49 yeah, I'll compose an email later today Jun 27 17:01:41 errata is only for rmii, not rgmii Jun 27 17:01:56 ah, ok Jun 27 17:02:05 but there's another errata for rgmii where if your phy doesn't support a delay mode, then you have to deal with things Jun 27 17:02:22 I'm not an Ethernet expert, but it looks like a pain in the butt Jun 27 17:03:03 I'm going to venture finding a PHY that drops in will be more difficult than anything else though Jun 27 17:03:13 google IO live Jun 27 17:04:35 I haven't even looked into that aspect yet Jun 27 17:05:13 and high frequencies and blue wires don't mix well Jun 27 17:06:31 green wires work better than blue ones at high frequencies Jun 27 17:06:36 ;) Jun 27 17:06:38 :-D Jun 27 17:07:30 linuxjacques, if you spin a board with gigE, let me know, would consider buying one to test at work with Jun 27 17:07:44 even better if it has both gigE setup Jun 27 17:07:51 or two gigE setup Jun 27 17:07:54 to be more clear Jun 27 17:08:19 bradfa, I will keep you posted. It's a bit of a catch-22 though, if we knew the gigE worked well, we would be more likely to spin a board :-0 Jun 27 17:08:21 :-D Jun 27 17:09:03 (it's a bit of a time crunch issue) Jun 27 17:09:17 there's no time crunches in engineering :P Jun 27 17:09:40 need to check data sheet but maybe one of these might help? http://www.phytec.com/products/som/Cortex-A8/phyCORE-AM335x.html Jun 27 17:10:25 yep! Jun 27 17:10:27 bradfa, wow, I gotta check that out Jun 27 17:10:57 carrier board says 2 gigE sets over rgmii Jun 27 17:11:01 BTW, have you heard anything about the AM33x parts other than AM3359 and AM3358 being available? we don't need GPU Jun 27 17:11:07 they're coming Jun 27 17:11:45 I think late Q3 maybe, and for some reason I thought am3352 was next Jun 27 17:11:52 but I'm not a ti guy Jun 27 17:12:15 linuxjacques: http://goo.gl/GLa6n Jun 27 17:12:25 yeah we should probably just ask our rep Jun 27 17:12:55 mdp, nice! Jun 27 17:13:05 ask him when you can get delivery of those RTP parts Jun 27 17:13:47 linuxjacques, on that phytec, only 1 gigE interface, but it's called "Ethernet2" but that's probably all you want anyways Jun 27 17:14:22 bradfa, yep, we only hard require one Jun 27 17:14:29 mdp, so does that mean future beagles won't have the XAM unqualified parts on them? Jun 27 17:14:43 mdp, not that I expect you to know... Jun 27 17:14:55 I know nothing Jun 27 17:14:58 :) Jun 27 17:15:02 I exist only to paste public urls Jun 27 17:15:20 fair enough, you do a great job at it! :) Jun 27 17:15:25 heh Jun 27 17:15:41 bradfa, I've honed my skills for years...you now reap the benefits Jun 27 17:17:52 mdp, marketing must have made the list of partnumbers on that page, 3352 for zcz is both at the beginning of the list and at the end (granted different speed grades, but everything else is in order) Jun 27 17:18:01 mdp: heheh Jun 27 17:19:30 linuxjacques, also worth looking at: http://www.armkits.com/product/devkit8600.asp Jun 27 17:19:42 min order quantity of 5 Jun 27 17:20:30 * bradfa gets started with lithium battery, hopes not to create a fire on his desk! Jun 27 17:22:16 bradfa, russell crane *is* both the poster and in our marketing if that helps to close the loop ;) Jun 27 17:22:55 bradfa: cut one open to verify :) Jun 27 17:28:53 bradfa, thanks. sorry, someone was in my cube Jun 27 17:29:19 mranostay, does this remind you of when we were testing the battery charging stuff on that one bringup? worrying about fires in the lab while we were gone. ;) Jun 27 17:30:04 we're going to get the battery supplier to put an NTC inside, but our samples don't have one, so I get to cheat the sense with a resistor. Doesn't make me feel comfy... Jun 27 17:30:35 although mailing list threads on both battery power for shutting down and getting the PMIC to turn off power couldn't have come at a better time for me :) Jun 27 17:30:50 mdp: didn't i have to babysit that board? :) Jun 27 17:35:38 is GPIO interrupt detection still done with poll() or is there a proper interrupt mechanism now? Jun 27 17:39:30 mranostay, yeah Jun 27 17:40:24 did our fire insurance cover insane battery charger circuits? :) Jun 27 17:40:59 bah, small companies can't worry about those details ;) Jun 27 18:15:10 nemik: Now there's an interesting, and relevant question to my interests. Jun 27 18:16:13 define proper Jun 27 18:17:18 the barstool what holds up the drunk patron. Jun 27 18:18:13 djlewis: I'm still on the first pint... Jun 27 18:25:07 at least your beer is served as pints, ours comes in skinny contoured glasses, maybe 8oz Jun 27 18:25:16 I opt for the bottle Jun 27 18:28:08 even better, it's imperial pints Jun 27 18:29:02 bottle? Jun 27 18:29:09 how about the case? :) Jun 27 18:29:34 now we're talking Jun 27 18:39:27 mru: you sure #beagle isn't a AA meeting wating to happen? Jun 27 18:39:29 *waiting Jun 27 18:42:42 I'm back home now Jun 27 18:44:05 mranostay: naw. Jun 27 18:44:25 mranostay: #beagle is a source of alcoholism, not a support group for it. Jun 27 18:44:37 mranostay: heh Jun 27 18:45:01 Any AA group worth its salt would be like "#beagle? yeah, stay away from there. Get your friends to make sure you do." Jun 27 18:45:41 mru: done being the weird guy in the corner? Jun 27 18:45:52 mranostay: I was at the bar actually Jun 27 18:45:58 eating a burrito Jun 27 18:47:01 pub burritos? Jun 27 18:47:19 they do "mexican" food on wednesdays Jun 27 18:47:33 no actual mexicans involved Jun 27 18:47:36 but still pretty good Jun 27 18:47:40 unlike here :) Jun 27 18:47:54 so Taco Hell style basically? Jun 27 18:48:04 way better Jun 27 18:48:30 it's all cooked from scratch in their kitchen Jun 27 18:49:37 and then there's the barmaid factor... Jun 27 18:49:55 *g* Jun 27 18:52:10 pics or it didn't happen Jun 27 18:52:34 it's a pub, of course they have barmaids Jun 27 18:54:28 anyone knows if mark porter is available here in irc? Jun 27 18:54:50 matt porter is here... Jun 27 18:54:56 dunno about mark Jun 27 18:55:19 matt will do just fine... :D (I must be tired...) Jun 27 18:55:27 mdp: ^^ Jun 27 18:55:57 hi dwery Jun 27 18:56:25 it's you! :D Jun 27 18:56:30 sorta here Jun 27 18:56:45 I am me Jun 27 18:57:04 that's good, 'cause I'd hate for you to be me Jun 27 18:57:14 mdp: I have just found your framebuffer driver Jun 27 18:57:19 because if you were, then who would I be? Jun 27 18:57:25 mru, I'm 33% of the way there Jun 27 18:57:31 and I was wondering if there's a kernel image ready to be downloaded somewhere Jun 27 18:57:36 mru, could you be me then? Jun 27 18:58:35 dwery, koen's latest builds might have it...but I'm not sure he has the platform glue in place due to conflicts with capes... Jun 27 18:58:37 koen? Jun 27 18:58:49 I know the driver itself is there Jun 27 18:58:51 correct Jun 27 18:59:31 you need something like this: http://dominion.thruhere.net/koen/angstrom/beaglebone/0004-HACK-beaglebone-enable-st7735fb-when-using-BeBoPr-ca.patch Jun 27 19:00:03 ok, thanks. I also saw an email regarding DMA problems and mentioning a patch for that Jun 27 19:00:18 it's all integrated now Jun 27 19:00:51 the tree in the blog post is known working integrated as is. Jun 27 19:01:31 dwery, are you comfortable building a kernel? Jun 27 19:02:08 only from time to time... Jun 27 19:02:11 * dwery hides... Jun 27 19:02:16 https://github.com/ohporter/linux-koen/tree/linux-ti33x-psp-3.2-r6c+gitr09e9651bcf2ee8d86685f2a8075bc6557b1d3b91-st7735fb this one, right? Jun 27 19:02:48 koen, are you looking at hdmi? Jun 27 19:02:58 I hope you have everything you need Jun 27 19:03:19 dwery, yes Jun 27 19:03:38 ok, I'll compile and see what happens Jun 27 19:03:42 ty! Jun 27 19:04:23 dwery, the problem is that depending on which version you have...it may not work ;) adafruit has another incompatible version of the display out...I'm working on merging some changes to handle that version as well. Jun 27 19:04:46 however, it has to be manually selected Jun 27 19:04:52 I gave a quick look at adafruit's code.. it's mostly in the reset routing as far as I can see Jun 27 19:05:17 I'll eventually adapt it. I just need to play a bit with the display before moving into is target circuit Jun 27 19:05:53 the front and back porch settting are different is all Jun 27 19:06:25 I think their bargain basement vendor in .cn changes it without notice Jun 27 19:06:35 quite probable. have you found a way to identify the displays? Jun 27 19:07:03 nope, and I asked..and there's no way...only way to differentiate is the color of the tab on the protective overlay that you remove when it arrives ;) Jun 27 19:07:11 you can see my query on their ph0rums :( Jun 27 19:07:31 saw that.. :( I tried to find the cover in the wastebin... Jun 27 19:07:33 ask mranostay how much he appreciates that Jun 27 19:07:42 he was the first victim Jun 27 19:07:59 I think it was green with a red tab Jun 27 19:08:38 it had been so long since I got mine until I heard about a new one..that the notion of "tabs" in their comments was completely meaningless 4-5 months later Jun 27 19:09:27 I'm working on a couple things with it in my copious spare time this week so I should have some patch updates RSN Jun 27 19:10:13 spare what? what time? Jun 27 19:10:15 also moving to the xM so I can rebase to a mainline kernel Jun 27 19:10:26 prpplague: I need my wife to leave too ;) Jun 27 19:10:33 hehe Jun 27 19:10:47 ok, I'll tell how's going. the board itself is marked 0612 Jun 27 19:11:15 mdp: only thing i have done with the spare time since my wife has been gone is that i have successfully made the kitchen look like a tornado came through it Jun 27 19:12:04 prpplague: you should line up all your shoes with beer bottles in them for when she returns ;) Jun 27 19:12:24 dwery, next to CARD_CS? Jun 27 19:12:53 mdp: no, between IC1 and the miniSD socket Jun 27 19:13:08 mdp: i don't have that many shoes Jun 27 19:13:20 mdp: i was thinking of a large bottle pyramid Jun 27 19:13:30 prpplague: ahh, yes, you would need her shoe collection maybe ;) Jun 27 19:14:13 my other option is to quickly code something for spidev Jun 27 19:14:15 dwery, ok, maybe that's another way to identify the old and new, or whatever...I have 2912 between the CARD_CS and IC1 (the evil level shifter) Jun 27 19:14:22 mdp: hehe Jun 27 19:14:40 dwery, I will publicly humiliate you if you drive it from userspace...that's a promise Jun 27 19:14:57 * dwery hides under the display... Jun 27 19:15:03 there's another one: http://www.adafruit.com/images/medium/18tftbobback_MED.jpg Jun 27 19:15:29 it's marked with silk screen, mine is marked as reversed over copper Jun 27 19:16:04 cool, I have a feeling that might be a post tab removal ID method Jun 27 19:16:21 heh Jun 27 19:16:35 in any case, try running with that branch and defconfig as I documented in the blog post..there's a trivial patch to apply to support the new version in the meantime Jun 27 19:17:32 k Jun 27 19:17:51 I'll add a module parameter so it's easy to select from the cmdline Jun 27 19:18:26 nice Jun 27 19:18:30 *cough* DT will solve all this Jun 27 19:18:51 wonder if there's a way to autodetect Jun 27 19:18:53 is there yet another version? Jun 27 19:18:59 dwery: how? :) Jun 27 19:19:03 mranostay, this is just about the two versions Jun 27 19:19:05 there is no EEPROM Jun 27 19:19:08 no, no way to detect..period Jun 27 19:19:43 it's just a ST7735, nothing to read to see the display type Jun 27 19:20:50 you can pass in a different display type to the st7735fb driver via 1) platform data 2) module parameter/cmdline 3) ..down the road...DT Jun 27 19:21:26 k Jun 27 19:21:35 so I'm adding a _RED/_GREEN tab display type or whatever you can select from the list of displays Jun 27 19:22:30 trying to make time to finish this so it can go upstream and we get rid of some of these integration issues. Jun 27 19:24:37 ouch.. I wiped my toolchain the other day.. have to rebuild that too Jun 27 19:26:41 mranostay, does your display have that 4 digit number where the above image shows? Jun 27 19:27:38 -EATWORK Jun 27 19:27:51 * mranostay makes note for later Jun 27 19:29:43 mranostay, oh..work..hrm Jun 27 19:30:05 mranostay, thanks, that would help for documentation Jun 27 19:30:38 mdp: so do you and koen have the same number on yours? Jun 27 19:30:57 good question Jun 27 19:30:59 koen? Jun 27 19:34:09 mranostay, did you see koen's video where he did the IC1 bypass to run the st7735 at 16MHz? looks very nice Jun 27 19:35:30 http://www.youtube.com/watch?v=0C4lXej2H_I Jun 27 19:36:34 lol, he's got my fritzing diagram on dead trees there Jun 27 19:47:04 https://github.com/mranostay -- are you on a swing set? Jun 27 19:48:54 HeadAche: boat Jun 27 19:57:36 mdp: ready to build. you suggested some particular options in the .config? Jun 27 20:00:59 dwery, there's a defconfig in the top level directory of that kernel tree...that's koen's defconfig...in that branch, I've modified it to have st7735fb support enabled. so use it. Jun 27 20:01:08 ok, I used that one Jun 27 20:01:09 mranostay, ppopov's sailboat? Jun 27 20:03:13 uhuhh.. there's pm firmware now :D Jun 27 20:08:50 mdp: sven's Jun 27 20:09:14 ;) Jun 27 20:10:18 dwery, I have some doc series about 50% done to explain how to do git-base kernel devel/deployment, but I guess that won't help you right now Jun 27 20:10:30 another thing to finish Jun 27 20:10:32 don't worry, I know it Jun 27 20:10:47 wget the .bin and copy in place Jun 27 20:11:00 done Jun 27 20:11:00 stupid PM architecture :( Jun 27 20:12:22 Is there a tutorial to get the power consumption on beagleboard ( I mean, exactly what I have to do)? Can it be done via SW? Jun 27 20:25:01 mdp: booting... Jun 27 20:27:58 there's some garbage on the display, nice! Jun 27 20:28:51 dwery: normal Jun 27 20:30:18 it seems gnome is loading :) Jun 27 20:32:27 manOnSilverMount: in bbsrm there is mentioned a resistor to measure current across Jun 27 20:45:38 anyone know if implementation of the "hardware reset function" , a must for eMMC drivers? Jun 27 20:52:52 joelagnel: i've found many sd controllers and devices can get in an "unknown state" fairly easy. if you don't have a good reset function it can be very problematic for long term stability and to recover from minor errors that then become major ones Jun 27 20:54:20 prpplague, thanks, actually omap_hsmmc host driver doesn't implement the ops->hw_reset function Jun 27 20:54:58 so I'm wondering if its required, but as you said it probably is Jun 27 20:55:19 joelagnel: yea, i normally add some custom code for my projects for stability Jun 27 20:56:04 prpplague, ahh Jun 27 20:57:22 mranostay: something loads, but the graphics looks distorted. is that the display model problem or something else? Jun 27 21:02:10 mdp: I love them dead trees Jun 27 21:02:34 mdp: there should also be a yoctopus in the picture Jun 27 21:02:40 dwery, that means you have the "other display" Jun 27 21:02:43 koen, hehe Jun 27 21:02:56 koen, you're a killah! Jun 27 21:03:22 * dwery feels lucky Jun 27 21:04:15 it's roulette that doesn't involve a firearm..mostly harmless Jun 27 21:05:43 I'll try to fix it Jun 27 21:11:59 dwery, apply http://pastebin.com/WuWVHaP0 Jun 27 21:14:56 mdp: doesn't apply "patch unexpectedly ends in middle of line". partial cut and paste? Jun 27 21:20:55 mdp: applied manually Jun 27 21:20:58 much better now Jun 27 21:21:44 yeah, sorry, grabbed the hack from g+ so malformed Jun 27 21:22:02 so my display works with the same init sequence as yours Jun 27 21:22:18 as you can see, it's just the offsets Jun 27 21:22:25 dwery? Jun 27 21:22:31 adafruit's code has different sequences Jun 27 21:22:52 dwery, that's for another display type...but from what I hear...nobody has that yet Jun 27 21:23:24 I think they just got some dump of code from their vendor and dropped it in their github library Jun 27 21:23:50 so, according to their naming scheme, mine is ST7735R "red" Jun 27 21:24:20 in any case, I'm handling both of these in the updated driver using platform/DT/module options Jun 27 21:25:02 dwery, that's the same as mranostay's...and sounds like we have another way to identify them due to the sequence number placement on the PCB Jun 27 21:25:03 nice to know Jun 27 21:25:09 I'll may remove IC1 Jun 27 21:25:13 maybe* Jun 27 21:25:55 you can then run at 16MHz without a problem since that's just below the max freq for the st7735 Jun 27 21:26:41 just hack the platform data accordingly Jun 27 21:27:22 yep Jun 27 21:27:26 thanks for your help Jun 27 21:30:02 mru: proper interrupt meaning you get alerted somehow rather than having to rely on kernel polling the pin Jun 27 21:30:26 it seems like you don't understand what poll() does Jun 27 21:32:46 DESCRIPTION poll() performs a similar task to select(2): it waits for one of a set of file descriptors to become ready to perform I/O. Jun 27 21:32:52 dwery, np, glad to hear it works for you. Jun 27 21:33:20 how that wait is accomplished is up to the kernel Jun 27 21:33:34 typically it would wait for an interrupt from whatever device is behind the file descriptor Jun 27 21:34:14 mru, the name of poll() inevitably seem to make it a FAQ Jun 27 21:34:54 mdp: heh was that my malformed patch? :) Jun 27 21:35:05 mdp: only for people unable to read a man page Jun 27 21:35:11 mranostay, just fscking g+ failings Jun 27 21:35:18 mru, hehe Jun 27 21:35:18 * dwery pokes mranostay Jun 27 21:35:21 which admittedly is most of them Jun 27 21:35:26 dwery: yes? Jun 27 21:35:35 well, poll requires app to poll it every so often Jun 27 21:35:42 mranostay: just poking around Jun 27 21:36:30 poll may and will timeout Jun 27 21:37:29 oh c'mon... Jun 27 21:37:30 poll() means check for ready fds and optionally wait if there are none Jun 27 21:37:36 peek Jun 27 21:37:51 "Specifying a negative value in timeout means an infinite timeout." Jun 27 21:38:06 yes, that's what optionally means Jun 27 21:38:36 timeout is upper limit Jun 27 21:38:49 so if you set it to 10, that means it will timeout after maximum 10s Jun 27 21:38:52 it may timeout faster Jun 27 21:39:07 no, it can't _time out_ any sooner Jun 27 21:39:14 if it returns sooner, it wasn't a timeout Jun 27 21:39:18 The timeout argument specifies an upper limit on the time for which poll() will block, in milliseconds. Specifying a negative value in timeout means an infinite timeout. Jun 27 21:39:19 something became ready Jun 27 21:39:26 *upper limit* Jun 27 21:39:41 "upper limit on the time" aka "timeout" Jun 27 21:39:42 hope the kernel developer read that too Jun 27 21:40:07 what does an infinite upper limit mean to you? Jun 27 21:40:18 again shit documentation that doesn't clarify it 100% Jun 27 21:40:27 I think of an "8" lying on its side Jun 27 21:40:30 just like memcmp and mysql auth bug Jun 27 21:40:30 the documentation is perfectly clear Jun 27 21:40:31 looking at me Jun 27 21:41:08 mru: not to me, i understand that poll may return sooner and nothing says it's because of event on one of fd's Jun 27 21:41:41 you didn't read the "RETURN VALUE" section then Jun 27 21:42:16 there are 3 options: 1) positive => something is ready, 2) zero => timeout, 3) negative => error Jun 27 21:42:18 still not clear Jun 27 21:42:33 then you're just daft Jun 27 21:42:35 zero => timeout Jun 27 21:42:45 go back to what i pasted before Jun 27 21:42:50 *g* Jun 27 21:42:53 "A value of 0 indicates that the call timed out" Jun 27 21:42:53 *upper limit* Jun 27 21:43:10 time out, as in the upper limit was reached Jun 27 21:43:12 it doesn't mean it timed out because of maximum time limit Jun 27 21:43:25 reaching the limit is what time out means dammit Jun 27 21:43:31 it's the DEFINITION of timeout Jun 27 21:43:41 so there is 100% guarantee that it will never return 0 if i set timeout to negative value? Jun 27 21:43:43 there's a clarifying statement right after that ... the one that I pasted Jun 27 21:44:41 let's say that it timeouts earlier. this doesn't change a bit of the code that you have to write to properly handle all the cases. Jun 27 21:44:51 nope Jun 27 21:44:58 * mranostay hands mru a beer Jun 27 21:45:13 well, if it doesn't timeout earlier you may ignore timeout case Jun 27 21:45:15 let's use words with their normal meanings, shall we? Jun 27 21:45:37 you never ignore any return case. never. Jun 27 21:46:17 * mranostay trys to parse jacekowski's last statement Jun 27 21:46:37 hm new android Jun 27 21:47:13 i've just had a look at kernel code Jun 27 21:47:24 jacekowski 2.4? Jun 27 21:47:39 3.4.3 Jun 27 21:48:16 and it looks like it may return in case it gets an event Jun 27 21:48:32 you can get EINTR, yes Jun 27 21:48:34 but then it may decide that that event is not usefull for userspace Jun 27 21:49:29 so you may get 0 events but that's different 0 to timeout 0 but you can't tell a difference in userspace Jun 27 21:50:19 i'm not sure if glibc will not mask that somehow Jun 27 21:55:21 jacekowski: the man page does tell you about the signal exception path, btw Jun 27 22:01:05 always read the BUGS section Jun 27 22:01:31 hmm, that gives me an idea Jun 27 22:01:51 we should extend the ELF spec to require all buggy code to go in the .bugs section Jun 27 22:02:07 then secure systems can simply map that section non-executable Jun 27 22:04:52 * mranostay cuts mru from drinking tonight Jun 27 22:14:21 would it then require the entire executable be mapped to .bugs if the .bugs section is found to be incomplete? Jun 27 22:17:08 mranostay, that's just cruel. mranostay->meanie++ ... you're about to overflow the counter Jun 27 22:17:26 mdp: signed or unsigned? Jun 27 22:17:46 unsigned! Jun 27 22:17:51 the guy is ruthless Jun 27 22:17:55 is it a long long? Jun 27 22:19:56 Hi, I am researching the safe temp range for a bone, I saw a few notes on the google group stating 0-90C, does anyone know if that is while in operation or the max it can be exposed to before damaging components even if the board is powered off? Jun 27 22:20:13 check the data sheet Jun 27 22:21:30 at least 260C Jun 27 22:21:48 they all been exposed to that temp range Jun 27 22:23:19 the ds says Tjmax 105C for 80k POH Jun 27 22:23:27 90C for 100K POH Jun 27 22:23:59 isn't that operational? Jun 27 22:24:08 yes Jun 27 22:24:23 that's what people usually want to know Jun 27 22:24:50 storage temp is listed as -55C to 155C Jun 27 22:26:16 Yea the operational range should be no issue, I was more concerned with short term exposure, in this case to about 110C, while it was off Jun 27 22:26:31 if that is the case I should be OK Jun 27 22:26:36 yeah, should be fine Jun 27 22:26:57 awesome thanks for the help :) Jun 27 22:28:12 ds2, I'm upgrading him to the longest of longy long long integers Jun 27 22:37:49 see all this timeout talk is why i don't want to use poll and want a real interrupt Jun 27 22:41:30 the pulses i have to count are 2ms apart, i'm not sure fd/poll is fast enough for that Jun 27 22:47:48 uint1024_t? Jun 27 22:48:28 what are we counting the atoms in the universe? Jun 27 22:48:46 no, particles Jun 27 22:50:25 well i hope you are happy you just change something doing that :P Jun 27 22:50:57 got to keep the entropy pool filled Jun 27 22:54:50 nemik: you're still not understanding how poll works Jun 27 22:55:25 fine, use an interrupt Jun 27 22:56:27 fctrl() the fd Jun 27 22:56:27 doing anything in userspace with 2ms period ain't still gonna work Jun 27 22:56:52 then signal(SIGIO, yourinterrupthandler); Jun 27 22:56:54 there. Jun 27 22:57:15 * mru thinks poll() will be much faster than delivering a signal Jun 27 22:58:03 the request was for interrupts, not for something that works in a superior manner Jun 27 22:58:07 ;) Jun 27 23:36:03 handling a signal is much more complicated than delivering it ;) Jun 27 23:49:07 alright thanks for the suggestion ds2 Jun 27 23:50:23 mru: yea i might have to write a kernel driver for this, could that work then for 2ms? Jun 28 00:18:36 is a touchscreen supposed to have the x across the top any y down the left ? Jun 28 00:22:34 nemik: what kind of pulse and amplitude/duration? Jun 28 00:23:33 X and Y are purely how you want to define it Jun 28 00:24:50 ds2 I thought there was a standard for laying out touch screens, I could be mistaken Jun 28 00:42:33 yes, they are all rectangular Jun 28 00:44:11 heathkid: amplitude is good, can make it what i want. duration is short though only about 50ns Jun 28 00:44:23 it's Wiegand data Jun 28 00:45:05 I wouldn't count on a gpio triggering directly on that Jun 28 00:45:21 you probably need a high-speed detector that puts out a longer pulse Jun 28 00:45:54 actually sorry, it is 40uS Jun 28 00:46:00 way off Jun 28 00:46:57 I still suggest you check the datasheet and make sure the inputs can handle that Jun 28 00:47:33 really? an arduino can handle it, i think the am33xx should too Jun 28 00:47:55 fine, take your chances Jun 28 00:48:08 it's short enough that I'd check the specs just to be sure Jun 28 00:48:18 alright, i will Jun 28 00:48:42 gpios are mainly intended for low-speed operation Jun 28 00:49:16 yea i know, if it takes me writing a kernel module for it, that's OK. i can do request_irq there right? Jun 28 00:54:24 consider a phone with a tilt sensor Jun 28 00:54:35 what is X and what is Y? Jun 28 01:19:22 ds2 X is the x-axis of a touchscreen and y is the y-axis, guessing by the fact that there are x and y minimum values defined, I would figure that they would have standard Jun 28 01:23:37 Borillion: there is no X or Y axis on a touch screen. Jun 28 01:23:55 for a 4 wire setup, there are just 2 pairs of signals Jun 28 01:23:58 call them as you wish Jun 28 01:26:31 ds2, not a problem, issue is that the touch location doesn't line up with the shown pointer location Jun 28 01:26:47 use the tslib calibrator Jun 28 01:27:22 I should point out Im trying to use android Jun 28 01:27:33 android is broq. Jun 28 01:27:42 broq? Jun 28 01:27:49 broken? **** ENDING LOGGING AT Thu Jun 28 02:59:58 2012