**** BEGIN LOGGING AT Thu Apr 19 02:59:59 2012 Apr 19 03:01:32 it was booting ok, then I tried to use zcat on it, after than nothing will boot, even after I made a new sd Apr 19 03:21:13 I tried saveenv to erase the nand but nothing happened Apr 19 03:21:19 still the same Apr 19 03:21:52 this is with a new image too Apr 19 03:29:50 hello ? Apr 19 03:31:24 cbt? Apr 19 03:31:42 ? Apr 19 03:32:07 anyone here ? guess not ? Apr 19 05:14:26 <_av500_> no Apr 19 05:14:44 ?? Apr 19 05:15:03 <_av500_> ok, yes Apr 19 05:16:34 _av500_, can u look at this? https://groups.google.com/forum/?fromgroups#!topic/beagleboard/ArKFc3HS7_o Apr 19 05:16:56 I have a feeling something is hosed hardware wise :| Apr 19 05:21:24 <_av500_> sorry, no idea Apr 19 05:21:40 <_av500_> to verify the HW, boot an angstrom setup Apr 19 05:22:29 _av500_: why are you so helpful lately? Apr 19 05:22:37 :) Apr 19 05:23:47 _av500_ both angstrom and android hang about the same place with that usb init stuff Apr 19 05:24:10 _av500_ is cool :P Apr 19 05:28:08 i keep forgetting the xM Rev A had a NAND still Apr 19 05:41:10 emeb_mac: looks like you are working on a cool project Apr 19 05:41:43 mranostay: which one? :) Apr 19 05:42:07 <_av500_> the cool one Apr 19 05:42:19 <_av500_> you mean you have 2 cool projects? Apr 19 05:42:39 _av500_: all my projects are cool. Except the ones that are hot. Apr 19 05:42:55 <_av500_> mechanical stress must kill you Apr 19 05:43:05 which is most of them, now that summer is coming to AZ Apr 19 05:43:23 heh Apr 19 05:43:57 Seems like I always end up working on high-power RF amps just as the weather gets warm here... Apr 19 05:44:07 * koen yawns Apr 19 05:44:17 emeb_mac: space heaters in AZ? :) Apr 19 05:44:23 Office full of power supplies & analyzers. Apr 19 05:44:51 * emeb_mac nudges koen Apr 19 05:45:28 koen: liked the post in the google group about leading by example Apr 19 05:46:02 so TI has a warehouse in ND? Apr 19 05:46:44 Closer to Thief River Falls? Apr 19 05:47:02 <_av500_> emeb_mac: which post? Apr 19 05:47:31 _av500_: koen posted diffs for the dvi audio driver Apr 19 05:47:33 emeb_mac: I know some people want to plan digikey face to face meetings during ski season :) Apr 19 05:47:45 emeb_mac: probably. my tracker says Grand Forks ND Apr 19 05:47:51 middle of nowhere :) Apr 19 05:48:21 koen: Maybe Nordic. Not a lot of downhill in that area I'd say. Apr 19 05:48:58 I knew some folks from Egan who loved the winter snowmobile touring though. Apr 19 05:51:23 mranostay: yep - looks like it's just a hop-skip from Grand Forks to TRF. Apr 19 05:52:23 yay! my (beaglebone) OS can context-switch Apr 19 05:53:20 <_av500_> like a BOSs Apr 19 05:54:00 it can also crash Apr 19 05:54:11 and sometimes it crashes and context-switches at the SAME TIME Apr 19 05:54:56 como un jefe. Apr 19 06:03:13 anybody have an error when trying to install python-pyserial with opkg? Apr 19 06:04:21 yes, command not found Apr 19 06:05:07 hm, not here, it collected errors: Cannot install package python-pyserial Apr 19 06:06:01 preceded by unknown package Apr 19 06:08:31 any thoughts? Apr 19 06:11:11 it is an unknown package? Apr 19 06:11:24 that's what it says Apr 19 06:12:19 and you won't accept that? Apr 19 06:12:52 well I accept that it thinks it's an unknown package Apr 19 06:13:15 but others have seemed to not have troubles Apr 19 06:14:50 did you do opkg update? Apr 19 06:16:13 no Apr 19 06:16:16 I'll do that Apr 19 06:18:54 well that was embarrassing Apr 19 06:18:55 thanks koen Apr 19 06:19:09 <_av500_> koen: why does it still no do that by itselfß Apr 19 06:19:11 <_av500_> ? Apr 19 06:19:19 <_av500_> at least in the error msg Apr 19 06:25:05 _av500_: I asked the opkg developers and they said "we don't know, that's how it has always been" Apr 19 06:25:39 mdp: from a quick look the AVS patchset doesn't have ABB support, only SR class3, ABB is class 1.5 iirc Apr 19 06:36:26 _av500_: they care for our running gags Apr 19 07:28:10 * aholler moans about the weather Apr 19 07:28:32 weather shrugs Apr 19 07:29:10 * dm8tbr moos Apr 19 09:21:51 Hi Apr 19 09:22:32 can any body help me in how to cross compile opencv for beagleboard-xM Apr 19 09:22:37 please Apr 19 09:22:47 use angstrom, it has opencv already Apr 19 09:23:13 I am using angstrom Apr 19 09:24:01 so Apr 19 09:24:28 i know it has opencv but what is the procesure to compile any opencv sample code Apr 19 09:25:30 Ifrom net i came to know that u have to cross compile Apr 19 09:25:56 I am very new to this Apr 19 09:26:55 you dont have to cross compile Apr 19 09:27:20 you can install all the opencv stuff with angstrom and then compile your app that uses opencv on the board Apr 19 09:27:27 then how to compile and run on beagleboard it self Apr 19 09:27:39 well, I am not going to teach you how to compile code here Apr 19 09:27:40 can u please help me Apr 19 09:27:55 bascially, the same way you would compile on your PC Apr 19 09:27:58 like you would compile x86 code on an x86 machine Apr 19 09:28:14 it does not have gcc Apr 19 09:28:21 think of the beagle as a small "computer" :) Apr 19 09:28:25 install gcc then :) Apr 19 09:28:48 by opkg install gcc command Apr 19 09:28:52 also, these questions have been asked before Apr 19 09:29:14 yeah, google might have thousands of hits for that issue Apr 19 09:30:04 thanks a lot Apr 19 09:34:44 av500: what's this 'google' you keep talking about? Apr 19 09:37:17 koen: clean your glasses, I did not mention google Apr 19 09:37:30 there are many other fine search engines out there Apr 19 09:37:39 like quackquackfly Apr 19 09:40:18 * koen cleans glasses Apr 19 09:40:24 ogra_: what's this 'google' you keep talking about? Apr 19 09:40:26 Hi there, is anyone got a good how to that explains how to implement SPI on beagleboard? Apr 19 09:43:53 Hi friend Apr 19 09:45:15 I was trying to find procedure to compile a c program on beagleboard but unsuccessful Apr 19 09:45:49 can u direct me to some proper link or .. Apr 19 09:45:59 [11:27:54] bascially, the same way you would compile on your PC Apr 19 09:47:24 pradeep: are you trying cross compiling or you want to compile directly on the board? Apr 19 09:48:43 as u told I am thinking for compile directly on board Apr 19 09:49:15 pradeep_: are you being purposely obtuse, youve been given the answer 3 times now Apr 19 09:50:16 pradeep_: gcc hello.c -o hello_bin Apr 19 09:50:36 pradeep_: chmoad a+x hello_bin Apr 19 09:50:48 pradeep_: ./hello_gin Apr 19 09:51:09 pradeep_: ./hello_bin Apr 19 09:51:16 hello gin! Apr 19 09:51:18 once I have beagleboard running, go to gedit create hello world program. then ... Apr 19 09:51:51 pradeep_: like on any other linux computer Apr 19 09:52:15 hi all :) after some check with osilloscope i think that my lcd can work..so the problem i think is software...so i have some question. How can i get what /dev/fbN is used to lcd header? Apr 19 09:52:15 yeah ,i m trying again Apr 19 09:52:26 pradeep_: then what? Apr 19 09:52:44 koen: btw, what is that 2E: MAC on the beagle? Apr 19 09:52:55 my netgear router thinks it is bogus :) Apr 19 09:53:18 stuk_gen: /deb/fb0 most likely Apr 19 09:53:34 av500: isn't that the one based on the die ID? Apr 19 09:53:47 could be Apr 19 09:54:05 av500: i have /dev/fb0 fb1 fb2 and fb.... how can i get SURE that lcd header are on fb0? Apr 19 09:54:21 lcd headers are on the baord Apr 19 09:54:23 board Apr 19 09:54:48 av500: ok but where is the connection from /dev/fb0 to lcd header? is there some file configuration? Apr 19 09:55:14 reads DSS documentation Apr 19 09:55:16 read Apr 19 09:55:58 linux/Documentation/arm/OMAP/DSS Apr 19 09:58:12 koen: based on die id maybe, but it should still be valid Apr 19 10:02:54 koen: If the second bit transmitted on the wire is a 1, then the address is considered locally administered. If the second bit is a 0, then address is globally administered and the first 24-bits must be a globally unique OUI assigned by the IEEE. Apr 19 10:03:01 av500: is correct that i have in /sys/devices/omapdss a display0 dvi and display1 tv ? Apr 19 10:03:06 koen, yes, that's what I noticed…no ABB in there. I briefly reviewed the older ABB patches so I would have expected it to jump out at me if it were there Apr 19 10:03:18 av500: why isn't there a display lcd? Apr 19 10:03:18 koen: now, is it msb or lsb 1st :) Apr 19 10:03:35 stuk_gen: because "lcd" is just a name Apr 19 10:04:03 av500: ok...but beagleboard xM have a dvi port, s-video tv port, and lcd header Apr 19 10:04:22 av500: i'm especting display ... dvi,tv and lcd no? Apr 19 10:04:28 3 display* Apr 19 10:04:35 no Apr 19 10:04:49 you obviosly did not read the schematics Apr 19 10:04:51 or the SRM Apr 19 10:05:40 av500: no i'm not look the schematic...so i have to know that lcd header and dvi port have the same pins? Apr 19 10:07:18 stuk_gen: you are not curious? Apr 19 10:18:28 totally unrelated: http://www.lyricsvip.com/Violent-Femmes/Lack-of-Knowledge-Lyrics.html Apr 19 10:20:11 hi Apr 19 10:20:40 I tried to compile Apr 19 10:21:44 it is giving following errors -sh Apr 19 10:22:00 -sh: gcc:not found Apr 19 10:22:35 pradeep: you didn't install Toolchain dev Apr 19 10:23:03 Have you compiled your distrib using Narcissus?$= Apr 19 10:23:19 yes Apr 19 10:23:44 strictly speaking narcissus doesn't compile is the CS sense of the word, only in the publishing sense of the word Apr 19 10:24:43 koen: CS = common sense? Apr 19 10:25:01 ok you mean, it's just a website :) Apr 19 10:25:31 pradeep: is your board connected on internet throught Ethernet port? Apr 19 10:25:43 it created two images one distribution and other toolchain Apr 19 10:26:01 ok, but you don't need toolchain curently Apr 19 10:26:15 ok Apr 19 10:26:23 try "aptitude install gcc" on ur board Apr 19 10:26:27 yes it is connected Apr 19 10:27:48 aptitude? Apr 19 10:28:00 it's the same of apt-get Apr 19 10:28:06 * koen starts looking for some hard liquor to survive today on #beagle Apr 19 10:28:12 but safer in my mind Apr 19 10:29:30 hi Apr 19 10:30:11 i-sh: aptitude: not found Apr 19 10:30:25 and apt-get? Apr 19 10:30:57 * lpi is joining koen to his table Apr 19 10:31:26 pradeep_: opkg update; opkg list Apr 19 10:32:01 av500: I forgot that I was working on ubuntu ^^ Apr 19 10:32:15 yes, ubuntu does that to people... Apr 19 10:32:48 scary: http://www.heise.de/imgs/18/8/2/1/2/6/2/USB-Power-c433ad30a541bc3f.png Apr 19 10:32:53 it kills kitten too... Apr 19 10:33:46 mainline kernel has sound on bbxm - that's pleasant Apr 19 10:34:15 av500, wow, so i can sell my welder and just use my laptop and a USB cable in the future ? Apr 19 10:34:18 awesome ! Apr 19 10:35:52 i wonder if you can make it 20V 10A with a Y-cable :) Apr 19 10:36:14 You can use this stuff to start your car Apr 19 10:36:15 well, usual hub has 4 ports, so..... Apr 19 10:36:31 400W would be enough to power a PC Apr 19 10:37:37 hi apt-get not : found Apr 19 10:38:15 [12:31:25] pradeep_: opkg update; opkg list Apr 19 10:40:25 it gives so many thing starting with perl -module-unicore-lib.... Apr 19 10:40:28 av500: not if he's running fedora ;) Apr 19 10:42:17 https://groups.google.com/forum/?fromgroups#!topic/beagleboard/UcaeHFWUgzY Apr 19 10:43:18 sorry Apr 19 10:43:27 what for? Apr 19 10:44:14 just Apr 19 10:44:31 to bring attention Apr 19 10:44:57 * mdp reads "perl" and "unicorns" Apr 19 10:44:59 mmm Apr 19 10:45:00 forgiven Apr 19 10:45:22 "sorry" is the new "ping" ? Apr 19 10:45:26 panto, that's good news Apr 19 10:45:31 hi mdp Apr 19 10:45:46 pleasantly surprised Apr 19 10:45:48 hi panto Apr 19 10:45:58 now moving to the video part Apr 19 10:46:03 now what i should do? Apr 19 10:46:09 read Apr 19 10:46:12 * panto scrambles to find the hdmi to dvi-d connector Apr 19 10:46:36 here, once again: https://groups.google.com/forum/?fromgroups#!topic/beagleboard/UcaeHFWUgzY Apr 19 10:46:40 panto, I can fax you serveral Apr 19 10:46:44 or several Apr 19 10:47:57 severe faxing Apr 19 10:48:17 kindly fax me severely the codez Apr 19 10:49:01 panto, if I could fit the connector in 8KB I'd ship a 27c64 with it through .gr customs Apr 19 10:49:15 panto, you'll get it in ~3.2 months Apr 19 10:49:18 I was able to cross compile hello world and easily run on BB-xm Apr 19 10:50:14 can u suggest if i want to cross compile opencv and run on BB-xm Apr 19 10:51:59 lol Apr 19 10:53:16 pradeep: you dont need to compile opencv Apr 19 10:53:21 its already compiled Apr 19 10:54:08 if this was a ph0rum, opencv would have a sticky Apr 19 11:12:12 mdp: 'sticky' is dutch 70s slang for a joint Apr 19 11:12:37 hrm Apr 19 11:13:06 appropriate for opencv questions :) Apr 19 11:13:10 indeed Apr 19 11:14:02 koen, with your access to TI know-how, have you been given any updates on the USB DMA situation? Apr 19 11:14:47 no Apr 19 11:14:53 koen, I was reminded by one of the posts to the list Apr 19 11:15:13 just that OMAPN, where N>4 has silicon bugs Apr 19 11:15:34 nothing on software or AM parts Apr 19 11:15:47 I used to be cced on the internal debug progress..since I was the one to finally document the issue(s) in detail Apr 19 11:15:48 I did see that RMK sent a patchset to convert hsmmc to dmaengine Apr 19 11:15:54 but no longer Apr 19 11:16:15 yes, that was expected to happen Apr 19 11:16:51 AIUI that is a huge step in getting 'mainline' mmc working on bone Apr 19 11:17:02 the other TI asked him to handle the sdma->dmaengine conversion Apr 19 11:17:42 * koen looks for board-am335x.c in mainline Apr 19 11:17:50 yeah, we have to write an edma dmaengine driver now though Apr 19 11:17:55 koen, lol :) Apr 19 11:18:32 some people said I should push to mainline Apr 19 11:18:41 but I have trouble finding am335x support Apr 19 11:18:46 I must be doing something wrong Apr 19 11:19:01 koen, khilman and I were talking last night, and with VH's latest round of patches you can now actually boot the bone off of mainline Apr 19 11:19:27 first time this is possible with an under review patchset..pretty cool Apr 19 11:19:36 it is Apr 19 11:20:20 trouble is, you need to distract boris houndleroy with some scooby snacks cause he'll think this means "it's done" Apr 19 11:20:49 I have trouble communicating about mainline on bone Apr 19 11:21:21 me too Apr 19 11:21:31 it is hard to get the point across to non-software people about the progress being made while managing expectations Apr 19 11:21:47 it's difficult to soundbite the situation and the good reasons why the path in is slow Apr 19 11:22:49 but yeah, mr houndleroy will badger me about using 'mainline' this week Apr 19 11:24:13 on a good news front, we at least got the blessing to shift one person away from am37x upstream to help with am335x Apr 19 11:24:25 specifically because of this Apr 19 11:25:17 I'm afraid mr. houndleroy wil think this means the finish line is 2 weeks away :) Apr 19 11:30:06 koen, I found that the summary of the situation is in another mail. basically, the debug continues but is at the point where ajay will need to go to dallas Apr 19 11:30:22 koen, that should give you an idea of the status Apr 19 11:43:25 can any one help me Apr 19 11:44:14 i am unable to find any proper way to run opencv on beagleboard Apr 19 11:46:00 even i am not able to compile hello world on BB Apr 19 11:46:23 I can do this by cross compiling Apr 19 11:46:44 but not cross compile opencv Apr 19 11:46:55 please Apr 19 11:47:02 please Apr 19 11:47:16 http://narcissus.angstrom-distribution.org/ Apr 19 11:47:24 there's a opencv option, use this Apr 19 11:47:58 I have done that Apr 19 11:48:13 i have created distribution Apr 19 11:50:10 oO( sounds a bit like "i have made fire." ) Apr 19 11:52:04 please help Apr 19 11:52:55 any idea how to use I2C_READ_REPEATED_START ob beagleboard?? Apr 19 11:53:12 on Apr 19 11:54:22 hello Apr 19 11:54:44 is any one there to help me Apr 19 11:54:49 ? Apr 19 11:55:40 pradeep: obviously, at the moment nobody here is able to help you. please stop asking every few minutes now, come bak in a few hours. Apr 19 11:55:47 why do you want to compile opencv when it;s already available as binaries? Apr 19 11:56:38 bitbake koen; opkg install koen; ./koen Apr 19 11:57:21 XorA: MACHINETYPE=irc-supporter Apr 19 11:57:44 i find only c codes in that not binary codes Apr 19 11:58:34 * XorA is staying away from this as he doesnt even know what opencv is, I lost track at OpenGL :-D Apr 19 11:59:00 XorA: some vision thing. super fancy, many people love buzzwording it ;) Apr 19 11:59:07 beaglebone system question: /var/lock & run are mounted on tmpfs, but I do not see the entry in /etc/fstab. Which script does that? Is it somehow configured in sytemd? Apr 19 11:59:27 LetoThe2nd: so the computer watching me watching my pr0n, erm no thanks :-D Apr 19 11:59:56 XorA: yeah, nobody wants to see YOUR pr0n, i guess. Apr 19 12:00:38 any idea how to use I2C_READ_REPEATED_START ob beagleboard?? Apr 19 12:07:43 Hi, I'm new to the BeagleBone, so bear with me. (and still pretty new to Linux) :) I have a expansion board with a touch screen, but don't know how to change the P8 and P9 Mux modes. I'm running Fedora 15 on the BeagleBone. Could someone point me in the right direction? Should I switch to a different Linux distribution? Apr 19 12:10:34 yes, switch to angstrom, that has the best support for expansionboards Apr 19 12:10:38 thomas_fogh: somewhere in /sys/debug/xyz should be the possibility to change muxing. Apr 19 12:11:05 LetoThe2nd: that's answering the question, not solving the problem :) Apr 19 12:11:31 * koen suspects the actual question is "how do I make this expansionboard work" Apr 19 12:11:32 koen: hehe Apr 19 12:11:47 LetoThe2nd: Thanks! but I don't have a /sys/debug directory... Apr 19 12:11:49 http://mywiki.wooledge.org/XyProblem Apr 19 12:12:25 thomas_fogh: then either read up on debugs, and/or do what koen said ;) Apr 19 12:19:44 LetoThe2nd: I don't want to you to help me make the expansion board work, but help me understand the basics... Apr 19 12:21:02 thomas_fogh: well than, reading up on what debugfs is and how to use it certainly will help you understanding basics. Apr 19 12:21:33 thomas_fogh: http://en.wikipedia.org/wiki/Debugfs Apr 19 12:22:07 mdp: I'm trying to find a booting config, here's another failed boot: http://fpaste.org/6ZVw/ Apr 19 12:23:26 LetoThe2nd: debugfs is useless to get lcd+ts working Apr 19 12:23:43 since linux cannot to runtime init of pdata Apr 19 12:23:51 koen: i have never stated that it is helpful for lcd/ts. Apr 19 12:24:18 thomas_fogh: you need to install angstrom and look at how the angstrom kernel handles expansionboards Apr 19 12:26:12 koen: Thanks, I'll look into the angstrom distribution. Don't know why my boss installed Fedora 15, when there's a BeagleBone distribution available... Apr 19 12:26:34 koen: see, now angstrom is "The BeagleBon Distribution" Apr 19 12:29:00 LetoThe2nd: Well, closer to a BeagleBone distribution than Fedora 15. Apr 19 12:30:41 the kernel is the secret sauce Apr 19 12:31:04 * mdp considers that opencv may well be the beowulf cluster of a new linux generation Apr 19 12:33:07 mhhh.... sauce Apr 19 12:35:13 LetoThe2nd: angstrom is what comes with the beaglebone box, so yes, it is the beaglebone distro Apr 19 12:36:19 koen: linux can happilly do runtime pdata initalisation Apr 19 12:36:54 XorA: but not easily, you need to compile stuff :) Apr 19 12:37:17 koen: nope, create a mfd device that doesnt probe your device until its handed a binary blob Apr 19 12:37:27 agreed, there's no blessed userspace api :) Apr 19 12:37:57 XorA: echo new_mfd > /sys/class/something? Apr 19 12:38:06 or does it involve a .c file that needs to get compiled? Apr 19 12:38:27 mdp: when you say booting bone off of mainline, are you talking linux mainline? Apr 19 12:38:30 koen: no-one does it not as its not really sane, but in theory it would be quite simple Apr 19 12:38:51 koen: we have sysfs, we have mfd class add_children call Apr 19 12:39:52 travalas, sorry, linux kernel mainline, correct Apr 19 12:40:01 travalas, it's not useful for users, of course Apr 19 12:42:33 XorA: there's a few different way to get at that... Apr 19 12:43:13 mdp: yeah, I was listing one possible Apr 19 12:43:19 XorA, for i2c devices, there's already a logical place to extend this with the new_device interface that is quite dumb atm Apr 19 12:43:58 I looked a bit at doing the same for spi devices, but quite frankly, it all hinges on having a standard way to instantiate free form platform data for any of these devices Apr 19 12:44:13 there's very few i2c/spi devices that have any value without platform data :) Apr 19 12:44:44 but I agree, in concept the implementation is simple. Apr 19 12:45:34 mdp: yeah I was just trying to get across there is no actual requirement for pdata to be const at boot Apr 19 12:46:16 right, I think you guys are just talking past each other..koen's just lamenting that there's no user framework for this today Apr 19 12:46:36 to be really useful to the wider community, it needs to be somethign driven upstream Apr 19 12:47:27 hi, can dsp write directly to framebuffer ? Apr 19 12:47:32 I'm already frightened of somebody producing a "cape mfd" or similar one-off crap :) Apr 19 12:48:10 oooh Apr 19 12:48:14 excellent suggestion Apr 19 12:48:17 cape mfd device Apr 19 12:48:25 * mdp deletes the logs Apr 19 12:48:56 * mdp flings uv eraseable eproms at koen Apr 19 12:51:57 beaglebone system question: /var/lock & run are mounted on tmpfs, but I do not see the entry in /etc/fstab. Which script does that? Is it somehow configured in sytemd? Apr 19 12:53:09 XorA, outside of the interface to shove a platform data blob descriptor into the kernel and attach it to a platform/i2c/spi device is the interesting question of what tools do you use to generate that blob? Apr 19 12:54:11 XorA, I was thinking that for runtime instantiation like this, there's no sense reinventing the wheel..you simple use a DT fragment, dtc, .dtb output and post process if necessary..shove that in Apr 19 12:54:48 even on a DT enabled kernel, there's a use case to not have to reboot to add a device Apr 19 12:55:34 hotpluggin spi and i2c? Apr 19 12:56:22 I'm using avr-devices with bitbanged usb for that ;) Apr 19 12:56:23 not necessarily Apr 19 12:56:46 but we've been hotplugging platform devices by instantiating code in an fpga for a long time on other platforms Apr 19 12:56:48 and so forth Apr 19 12:57:26 there's tons of reasons it shouldn't be forced to reboot to add something Apr 19 13:00:31 aholler, I've been hacking on using an msp430 as a tool to emulate some popular i2c/spi ICs…so that is a hotplug case Apr 19 13:05:46 aholler, that boot log looks crazy bad Apr 19 13:13:05 ok,i'm reading more document for know how can i connect and run my lcd on beagleboard xM. I read a module omap-dss but in sys/devices/platform i have a omapdss and in dmesg i not found anything that cointains omapdss...so can do this the problem? Apr 19 13:35:06 stuk_gen: its the same DSS that outputs to DVI or LCD Apr 19 13:35:13 its even using the same pin Apr 19 13:35:29 all you need to do is to tell it to use the correct resolution and timing for your LCD Apr 19 13:35:38 *same pins Apr 19 13:35:56 av500: i try fbset and have correct resolution for all fb0 fb1 fb2 framebuffer Apr 19 13:37:58 jkridner_: ping Apr 19 13:38:05 pong Apr 19 13:38:49 mdp: I'd be happy to have the boot-only case handled initially. Apr 19 13:39:29 the ability to add a device at run-time is a care-about, but I think that improving drivers with solutions like 'new_device' in sysfs could be a workable approach for that. Apr 19 13:40:22 using device tree fragments would be ideal, but I wouldn't want to hold up until that was solved architecturally. Apr 19 13:40:41 \o/ DT Apr 19 13:41:05 koen: pong Apr 19 13:43:53 so Apr 19 13:44:05 if I change the code to match the schematics the DVI board stops working Apr 19 13:44:35 and I', 99% sure the bug is in my code Apr 19 13:49:42 jkridner_, well you already have the boot-only case in process with the massive DT effort in ARM Apr 19 13:50:02 so you particular use case is met Apr 19 13:51:01 it just isn't in a usable state. Apr 19 13:51:06 jkridner_: I don't meaning using fragments in the way you were thinking Apr 19 13:51:20 jkridner_: you can boot boards with DT Apr 19 13:51:38 but I can't get to the same level of functionality as with koen's kernel. Apr 19 13:51:56 yeah, just need to throw more people at it Apr 19 13:53:26 some problems want to be solved in the short-term too..since we are talking a 1 year horizon to have full functionality in mainline Apr 19 13:53:47 1+ years :) Apr 19 13:53:53 texas years or metric years? Apr 19 13:54:15 Is it possible to have a member of a global struct be a module parameter? Apr 19 13:54:17 :) Apr 19 13:55:06 koen, may need to translate to stardates Apr 19 14:01:37 jkridner_, there are patches posted constantly to expand DT support…they just need people to review and test on platforms that matter to them Apr 19 14:02:25 hi everyone! I'm currently trying to build an angstrom-based beaglebone image with a customized kernel. after building the original systemd-image, I ran bitbake virtual/kernel -c clean, -c menuconfig, -c compile, -c deploy. the resulting uImage in the deploy folder is still the one from the systemd build. what am I missing here? Apr 19 14:05:17 Why when you configure a pin (let's say UART4_RXD) to be pinmuxed to its uart function, why does the 0x20 bit (which normallly says its an input or an output) not seem to be reflected when you look at the pinmux debug fs entry? Apr 19 14:05:33 right now it says PIN_OUTPUT ... but it's the uart RX and it's working so obviously it's not rdiving it as an output Apr 19 14:05:44 does the PIN_OUTPUT vs. PIN_INPUT only matter if it's configured as a GPIO? Apr 19 14:07:48 Somebody know if it's obligatory to copy the MLO file first on the SDCard? I've read a post that said that the OMAP is looking for a file called "MLO" on the partition, and it doesn't matter the order of copying Apr 19 14:07:52 +s Apr 19 14:08:39 indeed Apr 19 14:08:43 its not obligatory Apr 19 14:09:23 ok, is it since the beginning of times, or it was fixed by revisions of the board? Apr 19 14:09:46 I follow the better safe than sorry credo and always put it on first Apr 19 14:10:06 lpi: it was the 11th commandment Apr 19 14:10:20 just that Moses had no idea what "MLO" meant Apr 19 14:10:32 of these 15.. ooops10 commandments? Apr 19 14:11:14 av500: thanks! Apr 19 14:11:32 * lpi is getting his holy Bible Apr 19 14:11:46 get a wholy beagle instead Apr 19 14:12:10 can someone tell me how to make beagleboard i2c in REPEATED_START mode?? please Apr 19 14:12:32 which file to modify? Apr 19 14:13:50 what does i2c repeated start do? Apr 19 14:14:07 wooly beagles Apr 19 14:14:32 why would i want a dog with holes in it? Apr 19 14:14:48 to read continuously ( for example 2 bytes of data) without stop bit for the first byte Apr 19 14:16:19 ah, dunno then sorry Apr 19 14:37:35 repeated start is a way to send a request and get the response in the same sort of i2c "frame" Apr 19 14:38:02 the purpose of it is to not allow any other master to squeeze in between and screw you up. Apr 19 14:38:09 is it from a different address? Apr 19 14:38:15 no Apr 19 14:38:22 ah Apr 19 14:38:32 yeah Apr 19 14:38:48 what hiwaeng said Apr 19 14:39:14 I'd guess that the bone driver supports this and the place I'd start is the i2c-dev docs, assuming you're playing in userspace right now Apr 19 14:39:55 ugh, what would cause msleep() to never return? Apr 19 14:39:56 can you prevent anything else from squeezing in without realtime support? Apr 19 14:40:07 mru: a time machine Apr 19 14:41:48 a sleeping pill Apr 19 14:42:05 or a poisoned apple Apr 19 14:42:17 try to kiss it Apr 19 14:42:23 an omap-kernel Apr 19 14:43:30 thurbad: I'm not sure the answer to that, but, keep one thing in mind here: somebody else squeezing in only matters if you're in a multi-master i2c bus Apr 19 14:44:04 if there' sonly one master (for instance the bone itself) then it's not really an issue; you just have to make sure YOUR software (which you control) does the right thing and doesn't stomp on itself or other pieces of sofwtare running on the system Apr 19 14:44:50 but there could be several devices on the bus Apr 19 14:44:57 and different drivers reading from them Apr 19 14:45:01 no? Apr 19 14:45:06 yeah Apr 19 14:45:13 true, and I suppose it would have to be playing with the same address to cause a problem? Apr 19 14:45:21 no Apr 19 14:45:26 hmm Apr 19 14:45:31 haha Apr 19 14:45:40 i meant yeah there coudl be multiple drivers all accessing the same bus Apr 19 14:45:54 but there probably wouldn't be multiple driver accessing the same devices on said bus Apr 19 14:46:26 There would be a low level 'bus driver' to keep them apart Apr 19 14:46:31