**** BEGIN LOGGING AT Tue Mar 06 02:59:58 2012 Mar 06 03:00:16 mine take 8 hours on my vm Mar 06 03:34:33 the build just failed with Mar 06 03:34:35 | /usr/bin/ld: cannot find -lbz2 Mar 06 03:34:35 | collect2: ld returned 1 exit status Mar 06 03:34:57 what is bz2 and how can i get it? Mar 06 03:41:27 is it bzip2? cause I already have it installed Mar 06 03:41:29 bzip2 is already the newest version. Mar 06 03:43:38 what os are you building on? ubuntu? you can use this to get all the packages OE requires http://mc1322x.devl.org/ubuntu.html Mar 06 03:43:45 the "apt-get'able Openembedded" part Mar 06 03:47:15 yeah ubuntu Mar 06 03:48:27 i did sudo apt-get install openembedded-minimal Mar 06 03:48:34 but it could not find the package Mar 06 03:48:42 you need to add the repo Mar 06 03:49:46 I followed this instructions here to setup my host Mar 06 03:49:47 http://www.openembedded.org/wiki/OEandYourDistro Mar 06 03:49:54 did the debian stuff Mar 06 03:49:59 and the part for ubuntu Mar 06 03:51:42 add deb http://oss.leggewie.org/oe/deb ./ to your /etc/apt/sources.list Mar 06 03:51:55 then apt-get install openembedded-minimal Mar 06 03:52:22 and then apt-get install chrpath python-dev, and then it compiled for me Mar 06 03:54:22 already have chrpath and python-dev Mar 06 03:54:43 I will try your instructions tommorrow Mar 06 03:54:48 thank you Mar 06 03:54:59 I have to work tomorrow so it is time to sleep Mar 06 03:55:15 thank you W1N9Zr0 for your help Mar 06 04:57:29 hi ...can any one let me know how to allocate extra 128K mem at boot time Mar 06 05:26:10 another drive by IRC! Mar 06 05:26:26 whoa a whole another 128k? :) Mar 06 05:27:26 mranostay: and on 12 other channels Mar 06 05:29:17 nice Mar 06 05:42:09 <_av500_> for the record, that r-pi guy is an asshole Mar 06 05:43:20 _av500_: raspberry-pi i assume? Mar 06 05:43:37 <_av500_> yep Mar 06 05:43:59 <_av500_> "somebody is making a boatload of cash from the beagle and panda" Mar 06 05:44:03 heh a friend of mine said he was going to order one. i was trying to talk him into getting a real board :) Mar 06 05:44:20 _av500_: "give me money instead" :) Mar 06 05:44:32 <_av500_> "r-pi has the most advanced multimedia of all of them" Mar 06 05:44:43 <_av500_> it does not even play mpeg2.... Mar 06 05:45:19 <_av500_> koen: please explain that ferrari, you 1% money monger! Mar 06 05:46:19 ew Broadcom chip Mar 06 05:46:22 eff that Mar 06 05:48:18 "Yes. The GPU binary also contains the first stage bootloader." <-- not sure i follow that Mar 06 05:51:39 ah he works for Broadcom no wonder Mar 06 05:52:53 <_av500_> i dont hold that against him Mar 06 05:53:00 <_av500_> bb people work for ti Mar 06 05:53:15 <_av500_> but he should know prices Mar 06 05:53:34 <_av500_> and how expensive 8 layer with micro vias is Mar 06 05:53:48 <_av500_> which he does not need due to his ancient design Mar 06 05:54:58 ARM11 hehe Mar 06 05:57:33 _av500_: i'm betting you can't hand anything more than SD if you are decoding Mar 06 05:58:13 _av500_: hehe Mar 06 05:58:25 * mranostay runs maven and watchs it download the internet... Mar 06 06:00:30 <_av500_> it has a hw decoder for video Mar 06 06:00:37 <_av500_> like every chinese soc for years Mar 06 06:06:17 _av500_: i avoided commenting on alison's post Mar 06 06:17:57 well,i guess,arm11 guy can't understand cortex a8 and a-9 feelings. Mar 06 06:18:21 arm11 is soo old Mar 06 06:23:08 ARM11's w/o full power management is a total waste of silicon Mar 06 06:29:03 <_av500_> hitlin37: he knows, but i guess he has to use what broadcom offers... Mar 06 06:38:23 it has no compitition with panda and beagle,when i got panda during the middle of last year,there was hardly any device out in the market on omap4.panda was deleivred way ahead of time,same was with beagle. Mar 06 07:18:43 _av500_: the bom and pricing for beagle is public Mar 06 07:24:03 he is awaken! :P Mar 06 07:24:07 * mranostay waves to koen Mar 06 07:24:22 koen: how is the ferrari? Mar 06 07:24:34 do you have a trailer hookup on it? Mar 06 07:26:18 av500: er? :) Mar 06 07:29:16 "....I mean, $175 for a Pandaboard or $100 for a Beagleboard? Somebody’s got to be amassing a pile of cash there....." Mar 06 07:45:35 ~~ Mar 06 08:11:09 koen is piling money? Mar 06 08:11:49 yes Mar 06 08:11:56 cool. Mar 06 08:11:57 he is the 1% Mar 06 08:12:25 he had to leave TI because he was making sooo much money on the side Mar 06 08:12:25 koen: do you make funny buildings with the coins? like an eiffel tower of 1cents? Mar 06 08:13:58 whats the old rockstar saying: "i spent most of my money on drugs, alcohol, fast cars and beautiful women. i just wasted the rest." Mar 06 08:41:57 * mranostay notes he probably should be sleeping now Mar 06 08:42:12 * mranostay protests at koen's flat Mar 06 08:42:16 * av500 only sleeps at his desk Mar 06 08:46:25 office mikado. but we've had smoke party yesterday :) Mar 06 08:46:25 damn startup work schedule :) Mar 06 08:46:46 wtf is a smoke party? Mar 06 08:46:49 mranostay: no, I strictly sleep 9 to 5 only Mar 06 08:50:37 mranostay: you power something, and then it smokes. Mar 06 08:51:00 one of our engineers had fun with tantal capacitors yesterday ;) Mar 06 08:51:48 ok not what i think of smoke Mar 06 08:51:52 * mranostay lives in CA :) Mar 06 08:52:23 not that i do anything like that of course :) Mar 06 08:52:36 not that i would care. Mar 06 08:53:19 * LetoThe2nd once visited a rock concert where one of the guitar players had replaced the marshall logo on his amp with a huge "thank you for POT smoking" sticker :) Mar 06 08:53:46 * mranostay scans the channel for his boss Mar 06 08:53:53 LetoThe2nd: heh Mar 06 08:54:47 av500: zoning out? Mar 06 08:55:55 <_av500_> nah Mar 06 08:56:12 <_av500_> i can type irc in my sleep Mar 06 08:58:39 using the slirc client ;) Mar 06 09:28:18 koen: "...We"re shipping F14 because it"s fairly well tested (the OLPC ARM release for the XO 1.75 is based on it) and it uses upstart rather than systemd (boots with a smaller footprint)...." Mar 06 09:29:08 how to see results in beaglebone Mar 06 09:30:25 how to see sum of two numbers from beaglebone on windows pc.... Mar 06 09:31:55 av500 hav u got any idea? Mar 06 09:32:26 I have to admit I dont even understand what you are asking Mar 06 09:33:35 av...I am just asking how to read some register value and display it on PC from beaglebone Mar 06 09:34:56 running what, using what tools... be precise. Mar 06 09:35:08 I am using CCS 5, adding two numbers in .C file the project is built and running..........but m not able to see result Mar 06 09:39:35 can u tell me, in which register of AM3359 result will be stored and how to read and display it on PC Mar 06 09:40:44 check your assembler source, it will tell you where its stored (i assume that when you talk about registers, you mean the gp registers of the cpu) Mar 06 09:40:47 register? are you programming in C or in assembly? Mar 06 09:41:20 av500: probably clicking funny singlestep buttons without knowing what they do ;) Mar 06 09:41:39 or just use printf. Mar 06 09:41:51 in C :)...OK tell me how to read 'c' = a+b. Mar 06 09:41:51 av500: LOL@upstart comment Mar 06 09:41:59 I know printf Mar 06 09:42:50 but..........printf is used for standard output device...i.e. monitor of PC. how it will work on beaglebone Mar 06 09:43:03 gloria_: you are messing things up. either we talk about assembly, and registers - or about c. and if we are talking about c, stop moaning about registers. Mar 06 09:43:22 gloria_: as i said, use printf. you obvisouly don't know what it does ;) Mar 06 09:43:36 gloria_: have it print to the serial console. Mar 06 09:43:40 * koen is watching a presentation about syslog on youtube Mar 06 09:43:53 * koen wonders how we ended up in that situation Mar 06 09:45:58 gloria_: don't get me wrong - i'm neither CCS user nor do i know anything about it. but i know that you can output by printf. Mar 06 09:47:23 * LetoThe2nd tends to interpret that as "information overflow" Mar 06 09:49:25 . Mar 06 09:51:27 koen: how we ended up with having unix information in youtube? Mar 06 09:51:55 LetoThe2nd: when using CCS, there is no console Mar 06 09:52:10 you have to jump thru hoops to get a printf Mar 06 09:52:33 av500: then there's always sprintf into some buffer you can watch. Mar 06 09:52:36 but then, who needs CCS to try c=a+b; Mar 06 09:52:43 LetoThe2nd: indeed Mar 06 09:53:24 no idea anyways why one thinks that c = a + b has got something to do with registers. Mar 06 09:53:39 koen: I need a version of the internet that filter all r-pi content Mar 06 09:53:50 av500: have it replaced with nudes? Mar 06 09:53:52 I can stop lol'ing all day long Mar 06 09:54:04 LetoThe2nd: I'll send you the content Mar 06 09:54:09 kind of reverse parental filtering ;) Mar 06 09:55:33 "...omg there will be F14 only for the r-pi, what about *security*...." Mar 06 10:08:35 av500: the r-pi news is getting annoying TBH Mar 06 10:11:52 you don't say :) Mar 06 10:32:09 av500: I can bear hype, but all the "ARM11 is tech from the future!!!!one!!!" is just irritating Mar 06 10:33:14 koen: well the problem is that the numbers 8 or 9, as in cortex-aX are smaller than 11. Mar 06 10:57:15 Any one here know how to work with i2c? Mar 06 10:58:33 nope, but we've got massive experience with metaquestions ;) Mar 06 10:59:11 I'm sure that's the case, I've got massive experience with meta-info :-) Mar 06 10:59:35 Sometimes google is really useless :-) Mar 06 11:00:06 setup muxing values correctly, enable i2c device in board file and for the rest see kernel documentation Mar 06 11:00:30 something like that, nothing special Mar 06 11:01:50 If it were linux.... But I'm having to use Windows Embedded Compact 7 where all of this is already configured and it's more a less a wild goose chase hunting down the values need to call whatever function is needed at the time. Mar 06 11:02:39 good luck with that :) Mar 06 11:02:42 But that was exactly how I was able to access gpio and I had no big problems with spi Mar 06 11:02:53 but i2c is really kicking my a$$ Mar 06 11:03:40 I heard that luck isn't part of but rather an enormous amount of masochism and high threshold for pain but it pays the bills ;-) Mar 06 11:04:24 what kind of issues are you having with i2c? Mar 06 11:06:56 I searched for OMAP_DEVICE_I2CX (where x is 1,2 or 3) from %WINCEROOT% all the way down through every and in every file but was unable to find the value for it and it's also not an environment variable. I have no idea what it could be or what it's value is and it's needed for the function Void* I2COpen(UINT devId). I've already tried some values but all I get back is zero and not a handle like I need. Mar 06 11:07:41 * av500 winces Mar 06 11:07:55 AKF_WEC7_Devel: you are most likely the only one with WINCE here Mar 06 11:08:06 but if you send us the code, we can have a look Mar 06 11:09:42 Yeah I know... My question is really about the identifier omap_device_i2c1. If isn't defined/declared in code and isn't set as an environment variable, then how can it be used in the code? Mar 06 11:10:32 Because the examples I have in code of I2COpen being successfully called use this identifier. Mar 06 11:10:54 maybe its in a lib? Mar 06 11:11:13 That's a very good point Mar 06 11:12:24 I'll try searching through them as well. Thanks for the help. Mar 06 11:22:15 bring it on: http://petenelson.co.uk/2012/03/why-the-raspberry-pi-will-save-the-uk/ Mar 06 11:28:19 av500: so r-pi is like devicetree! Mar 06 11:28:37 DT will also save the UK? Mar 06 11:28:52 we have to be carefull it not becoming an empire again Mar 06 11:28:57 * dm8tbr wonders when ElReg will start referring to it as the JesusBoard Mar 06 11:29:00 dt will save the world Mar 06 11:29:13 dm8tbr: lets start now Mar 06 11:39:28 * _tasslehoff_ started the day by doing 'sudo rm -rf /etc/rc*.d/*' on his laptop, but is feeling better now. Mar 06 11:44:29 Hello. I'm using gdb on rev B4 beagleboard. As I'm trying to look up source code with list command, it only gives me path to a source file and line number i.e. "60 in ../src/filename.cpp". It's the same when I do the step command. It's extremely inconvenient. Is there any way to make it show sourcecode? Mar 06 11:46:08 It uses Angstrom by the way Mar 06 11:48:36 with recent angstrom releases the -dbg packages include source as well Mar 06 12:14:31 I have just brought DEVKIT8000 beagle board, I am trying to learn Embdedd linux on it. Mar 06 12:15:15 as i am new in this world. what basic programs should i try to get better hands on Mar 06 12:15:17 I'd start by learning to spell Mar 06 12:17:32 devkit8000 is not a beagleboard Mar 06 12:18:56 oh sorry that i didnt know. but as much i know it has linux 2.6 ported with angstorm linux Mar 06 12:19:50 heh Mar 06 12:20:07 the devkit people never bothered to contribute back devkit8000 support to angstrom Mar 06 12:20:16 Sachin_: depends on what you want to do with it Mar 06 12:21:04 oh is it. but on [http://elinux.org/BeagleBoard#DevKit8000] it is saying that it is clone of beagle Mar 06 12:22:38 quantumi : Purpose of buying devkit is leaning embedded linux Mar 06 12:23:00 according to that please suggest Mar 06 12:23:50 well. are you interested in robotics? the I'd start by trying to interface it with a motor controller Mar 06 12:24:27 if it's mobile development that you're after I'd start by getting Android up and running Mar 06 12:24:59 otherwise, I'm not sure what you mean by learning embedded linux Mar 06 12:25:03 Yes Really i was just browsing through ROS and http://www.hervanta.com/stuff/Beaglebot Mar 06 12:27:01 Sachin_: I only recently received my board where the plan was to create a guitar effect pedal, so my next steps is to get the kernel to echo audio at the lowest latency possible Mar 06 12:27:39 Robotics or your project , those are big goals Mar 06 12:28:36 but as i am new to this, i will start with first LED driver once i got folder file framework then will think of big project. Mar 06 12:29:21 ah yes blinking the leds could be a fun first task Mar 06 12:29:58 tried that as well but understand what should be used for delay, Mar 06 12:30:24 read HR TIMER and used it, once i start my module then board gets hang Mar 06 12:30:42 with HR TIMER delay. Mar 06 12:31:07 I created a bash script and controlledthe leds via /sys/class/leds Mar 06 12:31:45 not sure if they're exposed through /sys on DEVKIT though, only tried with beaglebone Mar 06 12:33:06 it is supported on devkit Mar 06 12:34:20 my question i walked through provided led driver, but didnt understand that flow , it has created lot of structured and they are just pointing from here to there Mar 06 13:00:00 so I did a fresh build of the beaglebone OS. But I don't seem to have a fully working GCC toolchain. Mar 06 13:00:25 how else did it build then? Mar 06 13:01:05 I mean on the board itself. Mar 06 13:01:34 I'm trying to compile some wifi drivers and I'm missing as and stdlib amoung other things Mar 06 13:02:13 you have to explicitly build the target toolchain Mar 06 13:02:26 or if you have net opkg install it Mar 06 13:02:45 yeah, I've got net and working opkg. Mar 06 13:02:52 I just can't seem to find the assembler Mar 06 13:03:38 opkg update Mar 06 13:03:54 opkg install task-native-sdk Mar 06 13:04:00 or so Mar 06 13:04:09 that's what I was looking for Mar 06 13:05:24 I'm guessing there is something I can do in bitbake to get that out of the box too? Mar 06 13:05:35 bitbake task-native-sdk Mar 06 13:05:37 ? Mar 06 13:10:11 so with bitbake to add stuff to the image, I just call bitbake multiple times and then systemd-image? Mar 06 13:11:55 no Mar 06 13:12:04 you need that list what you want in the image recipe Mar 06 13:20:13 still appear to be missing fixdep now. Mar 06 13:36:11 I guess kernel-dev would be useful. Mar 06 13:39:26 djeez, yocto is turning out the be the ubuntu of openembedded Mar 06 13:39:52 it attracts people too stupid to google or run find -name "foo*" Mar 06 13:40:21 oh what, Yocto? Ubuntu??? sounds great sign me up Mar 06 13:40:30 koen: do they also threaten people not to use "Yocto" unofficially? Mar 06 13:40:38 I also love useless UIs Mar 06 13:40:59 jwinnebeck: yocto is useless even without a ui Mar 06 13:40:59 av500: sort of Mar 06 13:41:16 ha Mar 06 13:41:28 av500: you can get a prompt answer to your questions if you include a "but is yocto" statement in the mail Mar 06 13:41:31 Why are you guys such Ubuntu haters though? I didn't think Ubuntu was bad now? Mar 06 13:41:43 jwinnebeck: it jumped the shark a while ago Mar 06 13:41:50 jwinnebeck: i dont hate ubuntu Mar 06 13:41:50 except for Unity... I can ignore that because I only use Ubuntu on server Mar 06 13:41:53 jwinnebeck: apart from being completely unsuited for embedded Mar 06 13:42:06 oh well yeah... unsuited for embedded Mar 06 13:42:07 jwinnebeck: in fact I dont care about non-linux OSd Mar 06 13:42:09 OSs Mar 06 13:42:20 but it hasn't become somehow bad for server now, has it? Mar 06 13:42:21 it is Lord Humongus, devourer of memory Mar 06 13:42:58 not to mention that it's impossible to build from scratch almost Mar 06 13:43:28 I installed Ubuntu 12.04 on a virtualbox VM the other day, and virtualbox did not end up well at all. I have 64 bit guest and host on 8GB RAM system, gave 4GB of RAM to VM and virtualbox crashes with an "out of memory" error the first time, the second time it took out the host as well Mar 06 13:44:08 Actually the think that both impresses me the most and that I hate the most about Angstrom is that it builds everything from the ground up Mar 06 13:44:52 it's awesome because it finally seems to fix a lot of our concerns about GPL software, it's not because sometimes it is 2 hours to compile hello world Mar 06 13:45:32 “Yeah. It’s not entirely clear to me why the Beagleboard is so expensive. Somebody in that Beagleboard value chain has got to be making a pile of money – I mean, $175 for a Pandaboard or $100 for a Beagleboard? Somebody’s got to be amassing a pile of cash there, because that’s a $10 chip in that device. I don’t know why they’re so expensive." Mar 06 13:45:56 lol Mar 06 13:46:03 Oh the beagleboard is just 1 chip now... Mar 06 13:46:07 i thought beagleboard was revolutionary considering other devkit was sold, say, $5000 a piece Mar 06 13:46:08 OK Mar 06 13:46:14 Yeah it is Mar 06 13:46:37 beagleboard, although I never bought one, was a major motivator for me to look into ARM Mar 06 13:46:46 xxiao: /join #rpi, there's something for 25 bucks Mar 06 13:47:04 ynezz: what's that? A10? Mar 06 13:47:14 haha, no ARM11 Mar 06 13:47:18 brm brm bcm! Mar 06 13:47:37 yeah a broadcom ARM11 Mar 06 13:47:43 arm11 on stereoids :p Mar 06 13:47:57 are asteroids, depends on the POV Mar 06 13:48:03 s/are/or/ Mar 06 13:48:12 so it's 70% the speed of the beaglebone if you follow the paper specs Mar 06 13:48:17 “In terms of general purpose computing, it’s got this 700MHz ARM11, and our benchmark shows it’s about 20 per cent slower than a Beagleboard for general purpose computing. But, you know, it’s a quarter of the price – somewhere between a sixth and a quarter of the price – so yeah, I expect that our first customers are going to be Beagleboard-type customers.” Mar 06 13:48:27 since it's 1ghz I think they said but it's not superscalar like A8 Mar 06 13:49:14 oh, n/m it's a 700mhz core, so it's half the speed Mar 06 13:49:18 what is a BB type of customer? Mar 06 13:49:21 "on paper" Mar 06 13:49:45 jwinnebeck: arm marketed the A8 as twice the performance per clock compared to arm11 Mar 06 13:49:52 by all means, RPI looks like a good thin-client device Mar 06 13:49:53 Yeah that's why I said "on paper" Mar 06 13:49:58 2 instructions per clock Mar 06 13:50:22 so the bone is 1440 MIPS, the rpi is 700 (I thought it was ghz until just a second ago...) Mar 06 13:51:41 what I don't know about the superscalar in A8 is how often it's able to do both instructions at once Mar 06 13:51:49 jwinnebeck: I'm going out on a limb, but the lack of NEON is why ARM11 sucks Mar 06 13:51:51 so wouldn't 2x speed over ARM11 be the best-case theoretical? Mar 06 13:52:06 no Mar 06 13:52:24 or yes Mar 06 13:52:26 or maybe Mar 06 13:52:32 Well, koen, supposedly I just heard that vfp on A8 sucks horribly, so if you do 32-bit float with NEON it's fast but in 64-bit vfp it's 10 cycles Mar 06 13:52:48 But then I also heard it's something like 20+ cycles to do integer division, so I don't get it Mar 06 13:53:03 it makes my head hurt so I just trust gcc and don't worry about premature optimization... Mar 06 13:53:21 jwinnebeck, heh trust in gcc is often fatal :) Mar 06 13:53:33 It's not when you're not writing "tight loops" Mar 06 13:53:46 jwinnebeck: things like pixman, libjpeg-turbo, bluez and libav have neon optimizations Mar 06 13:53:52 I have to forget about it to save my sanity Mar 06 13:54:13 I did a c++ flops test on A8 with vfp and got 43 MFLOPS don't know if that's good or not but that's what I got. Mar 06 13:54:22 I was asking myself the other day, should I use float or double in one case, and I started reading about vfp and neon and my head spun Mar 06 13:54:31 RPI Broadcom Dual-Core VideoCore IV Media Co-Processor Mar 06 13:54:38 jwinnebeck: so for a simple "load picture in browser, scroll" it will be more than twice as fast as an arm11 with the same MHz Mar 06 13:55:02 ya... Mar 06 13:55:32 I knew that in Intel, floats can be "slower" than doubles, but realistically they are the same speed... Mar 06 13:55:38 doesn't seem to be the case with cortex A8 Mar 06 13:55:50 although I haven't benchmarked it, just looking at paper Mar 06 13:56:25 jwinnebeck: if someone sits down and integrates the media accel into xbmc the r-pi would make a half-decent media box Mar 06 13:56:44 jwinnebeck: since xbmc rendering is completely in opengl Mar 06 13:57:07 any plan to run angstrom on rpi? ubuntu is not doing it, redhat is the only one doing some support on arm11 Mar 06 13:57:29 arm11 makes me want to barf Mar 06 13:57:41 xxiao: "run" as in take make? Mar 06 13:57:44 type Mar 06 13:57:59 beaglebone (really sitara) is "it"... finally crosses the threshold of super cheap and super fast Mar 06 13:58:14 cortex A8 system < $100 == win Mar 06 13:58:28 av500: i'd say support like BB, anyway i don't really care, i doon't need a rpi Mar 06 13:58:57 xxiao: is it much more work than to add their kernel and set armv6? Mar 06 13:59:14 maybe yocto can jumo at it Mar 06 13:59:16 jump Mar 06 13:59:18 gah Mar 06 14:00:32 I asked the yocto people to drop they god awful beagleboard support and move to the pi Mar 06 14:00:47 but that upset the TI people on the yocto board Mar 06 14:01:33 hmm the RPI doesn't sound very open Mar 06 14:01:40 "To get the full SoC documentation you would need to sign an NDA with Broadcom" Mar 06 14:02:02 jwinnebeck: "who cares, it's $25?!?!?!" Mar 06 14:02:08 "Will it fit in an Altoids tin?" -- " it’s a little too big to fit the tin" Mar 06 14:02:13 oh man I didn't realize that Mar 06 14:02:18 that's #1 beaglebone feature Mar 06 14:02:31 I thought the RPI was smaller than bone by a bit Mar 06 14:02:52 jwinnebeck: exactly, where there is broadcom, i run away Mar 06 14:03:04 One thing that's nice is that it has HDMI Mar 06 14:03:59 But for what I'm using bone for, I don't need a display (or want one) Mar 06 14:04:18 jwinnebeck: of course it has HDMI, it' a *computer* Mar 06 14:04:27 a *computer* Mar 06 14:04:31 although I still dream of having an HDMI system, maybe android, on my TV, maybe with panda or beagleboard Mar 06 14:04:31 drum roll Mar 06 14:04:33 there are quite some arm9/arm11 with good codec support Mar 06 14:04:49 jwinnebeck: we are stting up a panda here as a upnp renderer Mar 06 14:05:09 upnp "renderer" that is new to me what does that mean? Mar 06 14:05:22 upnp has server, controller and renderer Mar 06 14:05:41 server holds the data, controller browses data and controlls renderer to render the data Mar 06 14:05:57 Is this the same as DLNA? Mar 06 14:05:58 server = nas box, controller = android app, renderer = panda Mar 06 14:06:06 jwinnebeck: DLNA is a trademark Mar 06 14:06:17 where it sends data but the end device decodes it? Mar 06 14:06:19 so we already owe somebody money :) Mar 06 14:06:23 haha Mar 06 14:06:50 So UPnP AV is a "competitor" of DLNA, or are they the same thing with two different terms Mar 06 14:06:53 upnp has "universal" in its name, so its useless Mar 06 14:07:14 so dlna "standardises" upnp, by e.g. demanding mpeg2 in TS support and ignoring MKV Mar 06 14:07:35 OK... so they are related Mar 06 14:07:39 yes Mar 06 14:07:47 DLNA takes the U out of UPNP Mar 06 14:07:52 I tried DLNA with my TV/bluray player and was extremely dissapointed Mar 06 14:08:00 yes, thats the one Mar 06 14:08:25 I thought that the DLNA would be my nice "solution" to the problem of wanting to play music in the living room Mar 06 14:08:39 well music and show slideshows of vacation photos on TV for guests Mar 06 14:08:39 oh, it can be Mar 06 14:08:46 it was pretty much useless Mar 06 14:08:48 as long as all the 3 parts play together nicely Mar 06 14:09:07 well it "worked". It worked like how I could "browse the web" on my Windows CE device 6 years ago Mar 06 14:09:12 yes Mar 06 14:09:25 and then there is appl with airplay that does not care for a standard and just makes it work Mar 06 14:09:42 first it took like 5-10 seconds for it to render the JPEG image, and it couldn't start rendering the next image while displaying the current Mar 06 14:09:47 so it was horribly slow Mar 06 14:09:59 jwinnebeck: yes, "smart" TVs tend to be like that Mar 06 14:10:05 and getting to actually to do anything besides play ONE music file or ONE jpeg was a pain Mar 06 14:10:16 damn, TI confirms, no way to route a DMTIMER into the PRU. Rats. Mar 06 14:10:26 Well I didn't buy a smart TV because I don't believe in that, but my bluray player was "smart" Mar 06 14:10:35 mostly because they are all "smart" now Mar 06 14:10:56 The smart TV thing... I bought an HDTV last year, it was like $400 to "upgrade" to a smart TV Mar 06 14:11:06 I'm thinking a TV might last 10 years Mar 06 14:11:07 spare the money Mar 06 14:11:11 exactly Mar 06 14:11:16 tv is dumb display Mar 06 14:11:18 I can buy a COMPUTER for <$400 Mar 06 14:11:19 is a Mar 06 14:11:27 hook it up and upgrade it every year or whatever Mar 06 14:11:42 I don't want a 10 year old *shitty* computer built *into* the TV for $400.... Mar 06 14:12:21 So there is this thing like the android HDMI dongle thing, that makes a lot more sense because you can upgrade it and it's also way cheaper, and small enough to be hidden Mar 06 14:13:14 and when it's obsolete in 2 years when Cortex B5000 comes out with 3ghz oct-core for $2.50, because that's how TI rolls, you can replace it Mar 06 14:13:30 the power for cost in these ARM chips is staggering... and climbing very fast Mar 06 14:13:42 exactly Mar 06 14:13:50 so buy the dumbest tv you can find Mar 06 14:14:08 yeah that's what I did, I bought the top of the line dumb TV for I think $800 for 46 inch Mar 06 14:14:48 the same size with "smart" and "LED" was like $1800 because LED was somehow "so friggin cool" at the time... the picture is the same the only point to LED is to save a little power, and you don't save $500 in electricty! Mar 06 14:14:57 http://www.alwaysinnovating.com/home/index.htm Mar 06 14:15:14 yeah that's the HDMI dongle I was seeing Mar 06 14:15:58 wow it's $50-$99 expected MSRP with 1-1.5ghz A9 in it! Mar 06 14:16:12 and dual core Mar 06 14:16:23 and wifi Mar 06 14:16:25 that's insane Mar 06 14:16:32 tv is for twits anyway. Mar 06 14:16:45 well that is true, I could get rid of TV, but for my wife Mar 06 14:17:05 * etheretic3 tsks Mar 06 14:17:40 But for some reason she thinks that Kitchen Nightmares > staring at code in IDE Mar 06 14:17:42 I don't know why Mar 06 14:18:26 actually, I like that show anyway, so I can't say anything bad about it Mar 06 14:18:34 the lorax is a good movie for those with kids Mar 06 14:19:01 * etheretic hasn't had a boob tube for over 10 years. Mar 06 14:20:00 the idea to watch something an editor has chosen to send before i watch it is repulsive. Mar 06 14:21:00 no coincidence it's called "programming". Mar 06 14:21:11 * av500 finds it tedious to watch unedited stuff Mar 06 14:21:38 av500: there's the internet... :D Mar 06 14:21:44 ? Mar 06 14:22:20 Yeah I don't know if I agree with the editing part... Mar 06 14:22:43 obviously you want a cut and edited show, you don't want to watch 8 hours of grips having lunch and all that Mar 06 14:23:06 jwinnebeck: no, all these directors with their huge paychecks are overrated Mar 06 14:23:10 av500: it's like bill hicks/beaches - "i prefer to stay at home listening to music *i* like." Mar 06 14:23:37 still you have to find that music Mar 06 14:23:51 or you buy all newly released CDs Mar 06 14:23:54 progarchives.org Mar 06 14:24:51 404 Mar 06 14:25:58 er, .com Mar 06 14:26:37 looks "edited" Mar 06 14:26:47 a well-made film is a work of art Mar 06 14:26:54 emphasis on well Mar 06 14:27:04 made in a well? Mar 06 14:29:49 if that's what it takes Mar 06 14:30:04 http://www.imdb.com/title/tt0097303/ Mar 06 14:36:10 * mranostay awakens Mar 06 14:39:20 mranostay: gm Mar 06 14:39:31 well rested? Mar 06 14:40:20 so so Mar 06 14:59:53 mru: did you see my pm? Mar 06 15:14:03 Hello. As I connect Jlink to my BB, it sets reset pin to 1. Is there a way to fix it? Also, supply voltage is always to low. Mar 06 15:18:05 Gaxx: why not ask the jlink people? Mar 06 15:19:47 av500: too logic Mar 06 15:28:43 <_tasslehoff_> Setting the STOPON_PWRON bit in u-boot causes my board to behave weird after an 8-second reboot. The board resets (x-loader starts again) 4-5 times before it boots normally. On regular reboot/poweron all is normal. Mar 06 15:32:01 then dont set it? Mar 06 15:37:52 <_tasslehoff_> av500: I kinda want to, since we want to use that button for recovering from Mar 06 15:43:27 <_tasslehoff_> hmm. item 23 in the tps silicon errata seems to explain it. Mar 06 16:15:48 <_tasslehoff_> yep. for reference in case someone ends up in these irclogs when researching: after doing an 8-second PWRON restart, one must wait 8 additional seconds before setting the PWRON bit again, or the device will poweroff/restart. Workaround: None (except waiting 8 seconds). Mar 06 16:16:14 koen: you switched to reading mailing lists online, correct? I've been trying for a number of months now, but am about ready to resubscribe to some. I'm just too lazy to go to webpages, such as the Beagle's Google Group. Just lazy, I guess. Mar 06 16:18:26 wmat: gmane's nntp interface is a great way to read mailing lists Mar 06 16:19:01 mru: I should try that. It appeals more to my innate laziness. Mar 06 16:19:16 * av500 looks at the >1mio unread on his nntp lkml entry :) Mar 06 16:34:34 Did the USB issues on the beaglebone get fixed? I was surprised to see that USB appears to be fully working on my bone now, so I am happy. I haven't tested it in awhile Mar 06 16:40:45 wmat: I use nntp via gmane Mar 06 16:41:00 wmat: it works, as long as you don't switch computers a lot :) Mar 06 16:41:17 Holy Crap, do you think the geniuses in india fixed the USB re-enumeration problem?!!? Yay!! Mar 06 16:41:50 I guess so, I can't find anything online about the bug or the fix, except for one page about building gentoo that said to hack it with echo F > /proc/driver/musb_hdrc.1 Mar 06 16:42:00 I was about to try that hack but figured I should test my bone first, and it is working Mar 06 16:42:12 so this new kernel 3.2 must have fixed it Mar 06 16:42:37 Well Mar 06 16:42:42 The only change to our kernel is the PRU clock fix, that can't have anything to do with it SilicaGel, can it? Mar 06 16:42:50 since the clocks are in a tree Mar 06 16:42:51 Like Ferris Bueller said, Life moves pretty fast. If you don't stop and look around once in a while, you could miss it. Mar 06 16:43:04 nope that wouldn't Mar 06 16:43:20 SilicaGel: 3.2 fixed the usb replug issue, yes Mar 06 16:43:26 YAY! Mar 06 16:43:28 yeah I am so happy Mar 06 16:43:28 :D :D :D Mar 06 16:44:02 Now we can safely use the USB port on our project Mar 06 16:44:56 jwinnebeck: I wouldn't go that far Mar 06 16:45:03 jwinnebeck: the driver still sucks ass Mar 06 16:45:12 oh, in what way does it suck Mar 06 16:45:34 dma issues Mar 06 16:45:42 issues when dma is disabled Mar 06 16:45:45 ISO issues Mar 06 16:45:51 that's vague... what is the impact of it? it doesn't work or it's just slow? Mar 06 16:46:04 locking issues because there are 2 musb controllers, the driver was written for a single one Mar 06 16:46:22 i don't think even these DAQ devices use isochronous mode Mar 06 16:46:25 though they'd benefit from it. Mar 06 16:47:00 USB sucks, film at 11 Mar 06 16:47:22 [+1] Mar 06 16:47:29 you know, it used to be "film at 12" but TI made everybody sign a non disclosure Mar 06 16:47:32 if only we knew people who worked at mentor graphics, so we can ask them about that MUSB IP block.... Mar 06 16:47:34 Does anyone have experience with using musb as a device? Mar 06 16:47:47 SilicaGel: no it was 10 and then daylight saving came Mar 06 16:47:53 haha Mar 06 16:47:54 koen, I think they're all shit-canned by now Mar 06 16:48:06 shit-caned? Mar 06 16:48:17 what does mentor graphics have to do with this Mar 06 16:48:27 SilicaGel: they oversold TI Mar 06 16:48:39 promised TI a working usb Mar 06 16:48:49 hardware or software? Mar 06 16:48:52 av500: yes. They whack them with long bludgeoning devices made out of poop. Shit caning. Mar 06 16:48:59 SilicaGel: MUSB == mentor usb Mar 06 16:49:08 jwinnebeck: hw, the sw is simple, right? Mar 06 16:49:09 SilicaGel: it's not TI IP :) Mar 06 16:49:42 av500: well the SW is apparently not simple enough if there are plugging issues and DMA and ISO issues... unless the DMA/ISO is in reference to the hardware part Mar 06 16:49:59 jwinnebeck: sorry, forgot the ":)" Mar 06 16:50:03 oh Mar 06 16:50:14 why would TI buy a USB controller from Mentor? That sounds really weird to me Mar 06 16:50:21 also also Mar 06 16:50:36 SilicaGel: why would they make their own? Mar 06 16:50:41 its not a core competency Mar 06 16:50:45 they make DSPs Mar 06 16:50:49 thats what counts! Mar 06 16:50:52 I would think TI would have a USB chip by now... Mar 06 16:51:00 or pick one that doesn't suck Mar 06 16:51:00 SilicaGel, why would that kid stick his hand in the electrical outlet... Mar 06 16:51:04 koen: Have you tried a UBI/UBIFS image on beagle on a recent kernel? Mar 06 16:51:11 Tartarus: nope Mar 06 16:51:13 k Mar 06 16:51:19 Tartarus: 3.0.x is the most recent thing I tried Mar 06 16:52:10 I heard that the SD/MMC slot is still "sucking" as well on the bone, is that why I/O seems so slow, or is it because of the Kingston card that came with it Mar 06 16:52:13 OE and TI docs differ and my quick pass at using the OE docs by hand gave me problems Mar 06 16:52:51 Tartarus: I've been using the OE way for ages, but you know that Mar 06 16:53:00 Tartarus: since you fixed the bugs in my implementation :) Mar 06 16:53:03 Yes, and it even worked for me ages ago :) Mar 06 16:53:21 we seem to have this conversation every 3 months or so Mar 06 16:54:14 Yeah, and now I'm back at actually using the image this time around Mar 06 16:54:19 might even make further progress Mar 06 16:55:52 Of course, maybe i'm getting the layout wrong too, damned things... Mar 06 17:00:04 OK, yeah, user error :) OE instructions work as expected Mar 06 18:20:50 jkridner: ping - you there? Mar 06 18:24:17 Hi, i've been looking at the gpio.txt file to clarify GPIO interrupts, but I don't understand the end of this sentence : Mar 06 18:24:34 Write these strings to select the signal edge(s) that will make poll(2) on the "value" file return. Mar 06 18:25:02 they will make poll(2) do what exactly ? Mar 06 18:28:07 spow maybe this will help you: https://www.ridgerun.com/developer/wiki/index.php/How_to_use_GPIO_signals Mar 06 18:28:35 YAY, PRU internal eCAP works Mar 06 18:28:44 Not that there was really much doubt. The doubt was 'can i figure out how to make it work' Mar 06 18:28:48 * dwery pats SilicaGel Mar 06 18:28:55 twice. Mar 06 18:28:55 This means: I have a clock! Mar 06 18:29:30 Next question is: is it worth routing that eCAP interrupt to HOST1 as an event? Or just poll its status over the PRUSS internal mux32 Mar 06 18:30:41 HOST1 is.. ? Mar 06 18:30:44 kkeller: thanks, i'll ignore gpio.txt altogether and refer to the direct user-space solution :) Mar 06 18:31:06 dwery: there's a whole wacky interrupt to channel to HOST mapping (2 mappings) in the PRU Mar 06 18:31:13 ok Mar 06 18:31:37 dwery: the picture explains it better than words, http://processors.wiki.ti.com/index.php/PRU_Interrupt_Controller#Interrupt_Mapping Mar 06 18:32:09 If you map an interrupt to a channel that is mapped to a HOST that's either HOST0 or HOST1, then that "event" is visible right in R31 Mar 06 18:32:16 meaning the PRU can check the status of that event very efficiently Mar 06 18:32:25 ok, interesting Mar 06 18:32:28 yes, I'd map it Mar 06 18:32:53 the problem is the two host interrupts are a finite resource Mar 06 18:33:19 but HOST0 and HOST1 cannot interrupt the ARM core, right? Mar 06 18:33:24 correct Mar 06 18:33:32 uhm.. Mar 06 18:33:36 2 through 9 are for mapping to PRU interrupts visible to the MPU Mar 06 18:33:43 This is confusing as F**K I know haha Mar 06 18:33:47 but I think I understand it Mar 06 18:33:55 and HOST0 and HOST1 can take multiple sources Mar 06 18:34:04 RIGHT Mar 06 18:34:17 but if there are multipel sources, then taht means that after you get the event, you have to poll around to figure out which one caused it Mar 06 18:34:35 i think that both options are usefult and interesting Mar 06 18:34:48 sometimes I might want the CPU to react Mar 06 18:35:36 Yeah. So far all I have done in having the PRU interrupt the MPU is to send an event Mar 06 18:35:46 for some reason I don't understand, it's event 19+16 Mar 06 18:35:51 I'm tryin to figure out our mapping now Mar 06 18:36:12 ok I see why now. Mar 06 18:36:22 PRUSS1_EVTOUT0 is the CPU's IRQ20 Mar 06 18:40:32 an SPI master in the PRU shouldn't be hard to do... Mar 06 18:40:43 once you have a clock, I agree!!! Mar 06 18:41:16 yeah Mar 06 18:41:25 those bits will be so banged by the time I'm done with them they won't even look like bits any more Mar 06 18:41:32 I'll bang those bits, and they'll stay banged Mar 06 18:41:43 :D Mar 06 19:31:40 A Nebraska woman has sold a three-year-old McDonald’s Chicken McNugget that resembles President George Washington for $8,100 on eBay. <-- we're in the wrong business Mar 06 19:31:46 we should be making chicken mcnuggets shaped like famous things Mar 06 19:36:26 I want to know the one who bought it... Mar 06 19:36:45 * mru goes off to design a che guevara bone cape Mar 06 19:44:00 mru: is that a kind of chicken mcnugget? Mar 06 19:48:32 most cpus actually have chicken bits Mar 06 20:00:44 probably more chicken bits than actual nuggets have :P Mar 06 20:36:32 Does anyone know if it's possible to exit secure mode via JTAG on a cortex-a8? Mar 06 20:36:54 you'll need gptimer12 for that... Mar 06 20:37:04 * dwery could not resist. Mar 06 20:43:05 Hello Beagle Team Mar 06 20:44:20 we are a team? Mar 06 20:44:23 Can Anybody have any idea of Angstrom OE ? Mar 06 20:44:35 perhaps Mar 06 20:45:46 Can anybody tell me how much it will take to build Angstrom OE? Mar 06 20:47:38 read the manual? Mar 06 20:48:24 what is "Angstrom OE"? Mar 06 20:48:31 I only know "Angstrom" Mar 06 20:48:34 or "OE" Mar 06 20:48:49 * koen checks http://www.angstrom-distribution.org/ Mar 06 20:49:01 "Our name is Ångström, not OpenEmbedded Ångström" Mar 06 20:49:11 maybe that needs some more permutations for people to get the point Mar 06 20:49:35 * mranostay knows how to annoy koen at the next conference Mar 06 20:49:38 :) Mar 06 20:55:48 mranostay: if you really want to annoy him call it OE Armstrong Mar 06 20:56:20 hehe i think he would punch me Mar 06 20:56:43 armstrong is my favourite Mar 06 20:58:00 guys sorry for that Mar 06 20:58:27 AJ__: you don't need to apologize! Mar 06 20:58:28 no need to apologise for providing entertainment Mar 06 20:58:46 you had no idea you were pushing a hot button Mar 06 20:59:17 so I'm going to the arm office tomorrow, do I get a free beer if I wear an intel t-shirt? Mar 06 20:59:24 SilicaGel: that's like arguing that TI should make their own SoC interconnect instead of the ones they purchase and integrate Mar 06 20:59:43 Holy Crap! Mar 06 20:59:49 They didn't make that themselves, either?!?! Mar 06 21:00:02 nope Mar 06 21:00:14 What a bunch of slackers! Mar 06 21:00:16 sonicmx Mar 06 21:00:17 they even tell you who made it Mar 06 21:00:50 I want to know about Getting Started With C6Run On Beagleboard Mar 06 21:02:31 angst-rom :P Mar 06 21:02:37 indeed Mar 06 21:03:27 AJ__, have you checked out the processors.wiki.ti.com pages? Mar 06 21:03:30 like this one Mar 06 21:03:33 mythtv-theme-childish Mar 06 21:03:34 er Mar 06 21:03:52 processors.wiki.ti.com/index.php/C6Run_FAQ Mar 06 21:04:04 be careful with that paste button Mar 06 21:04:20 its the line ending within the paste that really gets you Mar 06 21:04:38 Russ , yes I did Mar 06 21:04:54 But I have few questions Mar 06 21:05:00 ok Mar 06 21:05:52 I have working Angstrom sd card, Mar 06 21:07:03 I want to use DSP for Image processing task..I am planning to use C6Run for that..How much time need to Build Base Image for Beagleboard? Mar 06 21:07:19 It says >24 hours in tutorial Mar 06 21:07:29 sounds right. Mar 06 21:08:18 also how about the space ? I read in one forum near about 20 GB space required for that..IS it rrue? Mar 06 21:08:21 true? Mar 06 21:08:31 yes Mar 06 21:08:42 okay.. Mar 06 21:10:06 24 hours is a bit high, depending on your build system Mar 06 21:10:43 okay.. sounds good Mar 06 21:11:35 so building dsplink and codec engine (CE) is next step after Setting up OpenEmbedded ? Mar 06 21:12:27 I think you have to get the CE binary from TI, don't you Mar 06 21:13:36 anyone used the rotary encoder drivers before and know a good resource for getting the driver setup in the board file? I have two quadrature encoders I'd like to setup, but I'm not sure how to fill in with more than 1. Mar 06 21:14:34 thurbad, I am not sure about that..But Angstrom has dsplink and CE already Mar 06 21:15:39 yes, butI don't think you can get source for CE Mar 06 21:17:09 you mean I need to download C6000 Code Generation Tool , right/ Mar 06 21:17:45 ? Mar 06 21:28:35 Thaks for help.. Mar 06 21:46:07 hm eCAP clock is going twice as fast as I thin it should be Mar 06 21:46:45 the eCAP internal to the PRUSS should be running off of CORE_CLKOUTM4 Mar 06 21:46:49 seems to not be the case Mar 06 21:48:07 I wonder if it's actually being clocked by L3LS_GCLK even though it's internal to the PRUSS Mar 06 21:49:26 got to go, more on this later Mar 06 21:58:21 hi Mar 06 21:58:52 i'm on my new internet service, but it won't let me connect directly to freenode, it says I need to use SASL authentication Mar 06 21:59:23 but the web gateway works fine so i'll use that Mar 06 22:01:38 offtopic i am engaged, planning wedding next year Mar 06 22:01:46 SilicaGel: twice as fat? We shall bang and shift those bits! Mar 06 22:01:57 fast*\ Mar 06 22:02:40 snowrichard: i get that too when I try to connect from my verizon hotspot Mar 06 22:02:54 its sprint network Mar 06 22:03:18 wireless? Mar 06 22:03:34 yeah 3g hotspot, novatel 2200 Mar 06 22:04:09 figures I think att blocks irc on my android Mar 06 22:04:11 its the size of a credit card, about 1/8 inch thick Mar 06 22:04:52 so i could almost put it in my wallet lol Mar 06 22:05:34 page load from google mail from one page to next is about a second Mar 06 22:05:46 it was taking 5 on the satellite Mar 06 22:06:51 i have the sprint hotspot and can also use tethering on the android Mar 06 22:07:16 so 2.5 + 4.0 GB / month at full speed Mar 06 22:08:29 slows to 256kbps if i go over the 2.5 on the sprint Mar 06 22:10:49 snowrichard: live in the boonies? Mar 06 22:11:27 just out side the range of cable (500 ft from road) out of range for dsl 17,200 cable feet or so l Mar 06 22:12:26 so i've terminated this satellite service with hughesnet Mar 06 22:13:00 it will be on till 22nd or i ship the modem and antenna back Mar 06 22:15:19 getting coffee brb Mar 06 22:24:51 hello, can some one help me with debain and beagle bone? Mar 06 22:26:22 why would you want to reverse a bain? Mar 06 22:28:08 hello, can some one help me with debain and beagle bone? Mar 06 22:28:37 yes repeat yourself why don't ya Mar 06 22:29:03 reverse a bain? Mar 06 22:29:22 I dont get it Mar 06 22:30:27 * mranostay turns off troll mode Mar 06 22:30:37 just ask your question why don't ya Mar 06 22:30:57 I see Mar 06 22:31:32 i instaled debian on beagle bone folowing thins instructions Mar 06 22:31:34 http://elinux.org/BeagleBoardDebian Mar 06 22:32:00 but im left with a no bootable system and no kernel Mar 06 22:32:00 OK Mar 06 22:32:01 and? Mar 06 22:32:34 Well, what does your failure look like? Use pastebin :) Mar 06 22:33:42 there is no falure, the istalation finish sucsesfuly but as said in the instructions im left with a no buteble system and no kernel Mar 06 22:34:02 how can I instal the kernel after the debian instlation Mar 06 22:37:41 sso? Mar 06 22:39:13 can some one help? Mar 06 22:40:55 Well, when you try and start the system Mar 06 22:40:57 What happens Mar 06 22:41:20 Or are you saying you didn't install u-boot and MLO and a kernel onto your SD card, from other instructions? Mar 06 22:43:34 yes I did Mar 06 22:43:36 but Mar 06 22:43:45 after Mar 06 22:43:57 the debian instalation that iis Mar 06 22:43:57 lost Mar 06 22:44:17 cpould it bee because I choose entire disk when partitioning? Mar 06 22:46:12 mario_: you did this on a SD card right :) Mar 06 22:48:43 G'day. Mar 06 22:49:11 I'm working on a project which needs a high-speed (>50 MS/sec) ADC. Mar 06 22:49:22 I'd also like to use a BeagleBoard. Mar 06 22:50:10 The best technique I've seen to attach an ADC to a BeagleBoard (or rather, a BeagleBoard-xM) is i2c, which is severely underpowered for what I want. Mar 06 22:50:25 Any suggestions for how I could achieve that? Mar 06 22:52:38 jashank: what are you interfacing to the ADC? Mar 06 22:54:52 mranostay: A 400 MHz receiver, then a 0.5dB NF LNA, then an antenna. Mar 06 22:55:33 I'd like to try building a software-defined receiver, but if I can't get the ADC sampling fast enough, I'll have to stick with hardware receivers. Mar 06 22:56:17 jashank: i2C is too slow, I'd try SPI Mar 06 22:56:40 but I'm not sure it's fast enough. Mar 06 22:57:33 the other option is to write code for the PRUSS Mar 06 22:58:14 SPI maxes out at 70MHz. Mar 06 22:58:37 dwery: What's a PRUSS? Mar 06 22:59:34 Ah, a Programmable Realtime Unit Subsystem. Mar 06 23:00:49 One comment I saw suggested using uPP, but I don't think the OMAP3530 supports uPP. Mar 06 23:04:30 jashank: I'd start by finding the ADC Mar 06 23:04:58 Okay. Mar 06 23:05:02 no uuPP afaik Mar 06 23:05:13 but there's a memory mapped bus interface Mar 06 23:06:07 (with addr and data muxed) Mar 06 23:23:12 hi what is the difference between beaglebone A5 and beaglebone A3? Mar 06 23:25:05 the software that ships on the sd card.. pretty much Mar 06 23:25:49 a4 had a problem and a3 was essentially reissued Mar 06 23:27:56 thank u for the explanation, so basically, they r the same right? Mar 06 23:28:22 beagleboneee: yes Mar 06 23:28:22 hardware wise, as far as I'm aware Mar 06 23:29:06 could u also tell me how many external devices can I connect to a beaglebone by using the communication interfaces? Mar 06 23:29:36 there are a lot of interfaces, it depends on the type of the devices Mar 06 23:31:44 I am not familiar with the interfaces, it is obvious that one devices to the usb port, but how about the JTAG port? there are two JTAG ports on the left and right side of the board, I wonder how many devices could be connected externally to the beaglebone Mar 06 23:32:00 those are not JTAG ports Mar 06 23:32:36 what port are they? Mar 06 23:32:45 they're called CAPE connectors, and you can connect different capes to them, http://circuitco.com/support/index.php?title=BeagleBone_Capes Mar 06 23:33:03 or you can connect your own device.. there are serial, i2c and spi ports on those connectors Mar 06 23:33:38 great Mar 06 23:33:49 check the referencr manual Mar 06 23:34:08 thank you very much for the answers Mar 06 23:34:17 I will read the manual Mar 06 23:34:27 have a nice night Mar 06 23:34:45 see you Mar 06 23:35:35 http://circuitco.com/support/index.php?title=BeagleBone_MSTP anyone has bought that cape? Mar 06 23:38:49 I am planning to buy that one on http://search.digikey.com/us/en/cat/programmers-development-systems/general-embedded-dev-boards-and-kits-mcu-dsp-fpga-cpld/2621773?k=beaglebone Mar 06 23:38:58 would u recommend it? Mar 06 23:39:21 or would u recommend any another capes? Mar 06 23:39:23 you'll be satisfied Mar 06 23:39:33 capes are add ons to the beaglebone Mar 06 23:39:47 some knowledge of linux is required too Mar 06 23:40:38 ok thanks Mar 06 23:52:09 hi Mar 06 23:52:16 had to run some errands Mar 07 01:41:46 How can I make the build system build the uImage again for me? I tried MACHINE=beaglebone ./oebb.sh bitbake virtual/kernel but is said everything is uptodate and there is nothing to do Mar 07 01:42:15 NOTE: Tasks Summary: Attempted 1161 tasks of which 1161 didn't need to be rerun and all succeeded Mar 07 01:42:53 I went to the deploy directory and deleted the uImage by hand then did the previous command I mentioned and it did not recreate the uImage Mar 07 01:43:20 bitbake -f -c virtual/kernel Mar 07 01:43:29 I do not want to recompile/relink the uImage I just want to re-generate it Mar 07 01:43:42 then bitbake -c deploy virtual/kernel Mar 07 01:44:15 the first one should have had -c compile Mar 07 01:44:40 what is -c for? Mar 07 01:44:45 command Mar 07 01:44:52 i imagine -f is force? Mar 07 01:45:01 something like that, yeah Mar 07 01:45:20 ok let me give bitbake -c deploy virtual/kernel a try Mar 07 01:46:36 the recompile step doesn't take that long if you don't change options/source... assuming you didn't remove the temporary 'work' files, which is done by the normal sequence Mar 07 01:46:44 it will take some time my vm is really slow but thank you thurbad Mar 07 02:32:44 SilicaGe Mar 07 02:33:16 hrm Mar 07 02:34:08 thurbad: bitbake -c deploy virtual/kernel did not do the trick :-( Mar 07 02:34:46 NOTE: Tasks Summary: Attempted 405 tasks of which 405 didn't need to be rerun and all succeeded. Mar 07 02:35:08 no umage got re-deployed Mar 07 02:35:22 try with an additional -f then Mar 07 02:40:49 thurbad: i assume -f ignores the stage stamp? Mar 07 02:42:12 not entirely sure what the -f is semantically but it forces it to rerun steps that it wouldn't otherwise run Mar 07 02:43:44 * SilicaGel12 Do Not Eat **** ENDING LOGGING AT Wed Mar 07 02:59:58 2012