**** BEGIN LOGGING AT Tue Feb 14 02:59:56 2012 Feb 14 03:03:06 you sure you picked the right machine type? Feb 14 03:10:08 jay6981: yep I picked beagleboard Feb 14 03:10:58 if you have a known good one I'd be interested in trying it Feb 14 05:04:01 hello, getting ssh_exchange_identification: Connection closed by remote host Feb 14 05:04:10 when trying to ssh into my new beaglebone Feb 14 05:04:23 cloud9 interface works fine. Feb 14 05:04:41 using usb with mac lion Feb 14 05:04:49 anyone have this issue and get past it? Feb 14 05:08:12 yes Feb 14 05:08:28 lion vs ftdi driver Feb 14 05:08:44 ftdi driver worked? Feb 14 05:08:47 lion not? Feb 14 05:08:49 i "got past it" by powering it with a 5v dc jack and plugging in the ethernet Feb 14 05:09:02 no, the ftdi driver doesn't quite work well with lion & ip-over-usb Feb 14 05:09:17 weird "firewall" issues which are not actually firewall issues Feb 14 05:09:26 =( serial doesn't work either Feb 14 05:09:33 there is a hack to make serial work Feb 14 05:09:54 i haven't tried it but i found it a couple of days ago just randomly googling Feb 14 05:10:05 switching to ethernet was simpler for me :) Feb 14 05:10:32 cool. thanks. Feb 14 05:11:08 it grabs beaglebone.local and beaglebone. if possible, which makes it kind of awesome Feb 14 05:14:00 in fact there is right now a beaglebone.hq.netapp.com sitting there waiting for someone to walk by it Feb 14 07:58:22 netapp goes beaglebone? Feb 14 08:15:14 Sup. Feb 14 08:16:02 Any idea of how likely TI are to supply you with a 'evaluation' BGA chip, i.e. 1 piece at $<100? Feb 14 08:16:20 We're talking theoretically of course. Feb 14 08:50:11 <_av500_> easy Feb 14 08:50:15 <_av500_> just ask Feb 14 09:11:28 Yeah. Okay. I doubt I have any hope for what I'm after currently since there exists no public domumentation for it Feb 14 09:12:14 But thanks av500; that actually says more than it does on the surface Feb 14 09:12:27 <_av500_> lunra: for what? Feb 14 09:12:33 .... *cough* OMAP5 Feb 14 09:12:34 <_av500_> what chip do you want? Feb 14 09:12:38 <_av500_> ah Feb 14 09:12:46 <_av500_> as I said, talk to TI Feb 14 09:13:00 <_av500_> if what you want to do is interesting to them, anything can happen Feb 14 09:13:32 Well, as a hobbyist, no. My plan is to lie; hey, other hobbyists have done such, but not generally with such a er, prestigious company. Feb 14 09:13:58 <_av500_> and you will solder that one single OMAP5 to where? Feb 14 09:14:02 I forgot I generally use this nick here Feb 14 09:14:11 Custom board, manufactured somewhere. Feb 14 09:14:24 <_av500_> ah, its you Feb 14 09:14:35 Yeah. Was I better off hiding? :3 Feb 14 09:15:30 <_av500_> no Feb 14 09:15:47 <_av500_> why not wait for the omap5 panda and just order one? Feb 14 09:15:55 <_av500_> do you think you can make yours faster than TI? Feb 14 09:17:08 No, I think I can make mine -better- than TI ;) Feb 14 09:18:53 Not to sound arrogant, but there are some things that bother me about the Panda/Beagle. Any idea/rumour about the potential existence of designs for one anyway? Feb 14 09:20:03 I've got some contacts who work in the military that are gonna help me with board design, if you're wondering. They do this kinda stuff for a living so they can teach me how to make traces not ring and all Feb 14 09:21:43 <_av500_> since you have no idea what your are up to, I won't try to stop you :) Feb 14 09:23:21 You mean "Since you have no idea what you're doing" or "Since I have no idea what you're doing"? Feb 14 09:23:42 <_av500_> make your own mixture Feb 14 09:24:08 malglico. I have no idea whether you are insulting me or not. Feb 14 09:24:27 <_av500_> lunra: :) Feb 14 09:24:40 does anyone have a known good SD image (for a BB xM rev C)? I'm having trouble getting mine to boot Feb 14 09:24:57 <_av500_> the angstrom demo images.... Feb 14 09:25:10 yes, not having much luck with those :( Feb 14 09:25:36 well, actually I used the narcissus minimal image since I don't have a 4GB SD card handy ATM Feb 14 09:26:46 hmm actually it's only 137MB 7zipd Feb 14 09:26:53 unlikely it would be larger than 2GB Feb 14 09:26:55 Mainly I plan to make the board larger and such to make hacking around easier. And to have all connectors be DIP. Talked with le design guy, he says it's possible to have long traces as long as you're careful with them. It's only conversational at this point. But so is 'Horseboard' or whatever the hell they decide to call it, as far as I am aware. Feb 14 09:27:49 the limitation on trace length is almost entireley dependent on how fast a signal you want to put down it Feb 14 09:28:01 and what sort of termination you put at the other end Feb 14 09:28:47 Yeah, that was mentioned, things like RAM clock would have to be paid attention to Feb 14 09:29:35 <_av500_> RAM is POP on top Feb 14 09:29:41 <_av500_> the trace length is 0 Feb 14 09:29:45 I thought this stuff stacked the RAM ontop of teh CPU anyway Feb 14 09:29:52 <_av500_> unless you go for the DDR3 omap5 Feb 14 09:29:53 Not for the 5432? Feb 14 09:29:53 right Feb 14 09:30:03 <_av500_> ompa5 will have pop or DDR3 Feb 14 09:30:20 the 5432 (the DDR3) one is not PoP Feb 14 09:31:05 I dunno, you're right, I'll wait for TI's board. Feb 14 09:31:56 <_av500_> lunra: it all depends on your skills and knowledge in that area Feb 14 09:32:07 <_av500_> if you did such boarss in the past, go ahead Feb 14 09:32:16 <_av500_> if it's your first, I would reconsider Feb 14 09:32:29 especially a brand new chip Feb 14 09:32:35 chock full of sillicon bugs to screw you over Feb 14 09:32:36 Aware. Feb 14 09:32:56 <_av500_> also i wont be cheap Feb 14 09:32:58 <_av500_> it Feb 14 09:33:11 <_av500_> making a few prototypre PCBs will set you back quite some money Feb 14 09:33:13 Limited; but where would one start making 'such boards'? Hey can you 'recommend a one day DIY soldering project involving 1000 pin BGAs? I can get those at Radio Shack/etc right?' Feb 14 09:33:27 (Joking in a lighthearted manner there, no nastiness implied!) Feb 14 09:37:24 possibly some good news from greece @ http://pesn.com/2012/02/13/9602039_Hope_from_Athens_found_in_Cold_Fusion/ Feb 14 09:38:23 can enable OpenFabs ... Feb 14 09:38:48 needed by OpenSoftware and OpenHardware. Feb 14 09:42:04 which will again make OpenDistribution necessary. :-) Feb 14 09:45:57 etheretic1: with all those "free energy" projects, we just need one to succeed to Feb 14 09:46:15 dwery: yes! Feb 14 09:46:35 to start hacking... Feb 14 09:46:51 I just wonder which one.. Feb 14 09:46:52 it's key to world peace & abundance. Feb 14 09:47:06 geez, those scams are still getting around Feb 14 09:47:08 and huge wars in the middle east :D Feb 14 09:48:18 It amazes me that in the day and age of the Internet, scammers like that *still* get investors Feb 14 09:48:53 Russ: it's no scam. read eugene mallowes description about what happened at MIT in 1987 (keywords: cold fusion experiment Feb 14 09:49:01 Russ: it amazes me that I still receive emails where peaople wants to give me billions of $$$ Feb 14 09:49:32 <_av500_> dwery: maybe that is the same people Feb 14 09:49:47 I doubt, I checked their IPs :D Feb 14 09:49:47 <_av500_> they use the free energy to earn millions and then give to you Feb 14 09:50:02 <_av500_> all part of world peace & abundance Feb 14 09:50:05 performed by hot fusion physcisists who receives billions of $ for hf "research". Feb 14 09:50:09 that's sounds like a nice plane. they're very kind to me Feb 14 09:50:28 <_av500_> I just wonder what the price of bread will be once everybody is a billionaire Feb 14 09:50:43 <_av500_> might ask the people in the Valley though.... Feb 14 09:50:47 _av500_: 0. Feb 14 09:51:15 etheretic1, yes, and the oil companies bought out that carburetor that gets 200mpg to keep it off the markt Feb 14 09:51:24 money doesn't really exist anymore. Feb 14 09:51:34 <_av500_> for my wallet, true statement Feb 14 09:51:42 Russ: amongst other things, yes. Feb 14 09:52:54 http://ecatnews.com/?p=1987 Is that some kind of sick coincidence? Feb 14 09:53:09 imagine being able to make anything you want, as long as it's not harmful to the biosphere... Feb 14 09:53:15 page = 1987 and an event in 1987 was described, however this is not it Feb 14 09:53:43 lunra: was it 88? Feb 14 09:53:46 wait, not harmful to the biosphere, are we talking about the same andrea rossi? Feb 14 09:54:06 Russ: maybe. Feb 14 09:54:16 and also the guy who did the electricty from waste heat scam? Feb 14 09:54:22 * _av500_ prefers Signor Rossi Feb 14 09:55:00 It's tinfoil hat time, google is playing vague on my search queries Feb 14 09:55:26 cf will be an intermediary step anyway - important, because it will prove (nearly) free energy is possible in principle. Feb 14 09:56:00 etheretic1, He was a scammer when he did petroldragon, the waste heat thing, and now the "energy catalyzer" Feb 14 09:56:17 100% inefficiency is considered impossible, but what of 99.99999...%/// Feb 14 09:56:45 lunra: if that's the case, conspiracy theorism has become an epidemic. 8-) Feb 14 09:57:14 Which comment are you referring to, the one about google, or efficiency? Feb 14 09:58:06 Russ: the disinfo machine against rossi and other f.e. proponents - who are running it, do you think? Feb 14 09:58:21 clearly I'm a paid shill Feb 14 09:58:26 * _av500_ too Feb 14 09:58:39 Fucking magnets, how do they work? ;) Feb 14 09:58:40 <_av500_> and my disinfo machine is broken Feb 14 09:58:44 entrained, more likely. Feb 14 09:58:46 <_av500_> still I doubt it Feb 14 09:59:15 etheretic1, read, enjoy, http://forums.randi.org/showthread.php?t=198040 Feb 14 09:59:34 it's not as if "The Laws of Physics" are hewn in stone. Feb 14 09:59:34 btw I heard of an USA company that does energy from waste a ciuple of weeks ago Feb 14 09:59:41 couple* Feb 14 09:59:51 randi is/was an idiot. Feb 14 10:00:05 etheretic1, yes, and paid by the oil companies Feb 14 10:00:12 I heard about something that breaks down carbon from trash in a simplified cracking process, a few years back, I guess the oil companies ate them. Feb 14 10:00:13 etheretic1, too bad he doesn't post a single message in that thread Feb 14 10:00:59 http://pesn.com/2011/07/23/9501875_Number-1_Breakthrough_Solid_State_Generator/ Sounds like magic. Get out the wands~ Feb 14 10:02:03 http://amasci.com/pathsk2.txt - using this as a diagnostic manual, quite a few here are suffering from P.S. Feb 14 10:02:29 not premenstrual syndrome... :D Feb 14 10:03:10 sigh, so now we've moved into UFO's and psi too, I think we can end the off topic discussion on that note Feb 14 10:03:45 Eeyup. Feb 14 10:03:46 it may be wise. Feb 14 10:03:59 'till the next news come in Feb 14 10:04:13 31 hr till elc! Feb 14 10:10:17 <_av500_> Russ: we have been at that point in the discussion here in the past Feb 14 10:10:59 was lunra talking about making a DIP adapter for a BGA? Feb 14 10:11:06 11:11 Feb 14 10:14:19 Russ: I think just a board with 0.1" pitch expansion headers Feb 14 10:14:23 Russ: No. Feb 14 10:14:30 Lol, what Dar1us said Feb 14 10:14:39 I say "just" but.. Feb 14 10:14:49 DIP adapter for BGA... >_< Feb 14 10:14:52 ok, I just wanted to add that dip pins are traces too Feb 14 10:15:00 Yep. Feb 14 10:18:54 all wires are including the ones in the chip, just depends on the frequency as to how much of a problem it will be Feb 14 10:22:46 forget I brought it up... Feb 14 10:31:17 <_av500_> 0.1" pitch isfor wimps, I demand screw terminals Feb 14 10:56:49 _av500_: why settle for anything less than what Dr. Frankenstein used to jazz some life into his monster? bit.ly/yeNcEE is surely what you want! :-) Feb 14 11:02:56 www.tuflet.com is bloody close to what i want re. linux-based navscomp, btw. Feb 14 11:06:21 add some IP67-ness, sprinkle with GPS/AIS/NMEA, and voila. Feb 14 11:39:52 do I need uimage on the boot partition? looking at the setup it appears that this is instead loaded off the ext2 partition Feb 14 11:51:21 <_av500_> depends on your setup Feb 14 11:55:47 * tasslehoffwrk put it on ext2 and never looked back Feb 14 13:30:26 anyone has an idea why my Beaglebone hangs on boot: http://nopaste.info/045b76b2c1.html (latest angstrom cloud9 image) ? An older version seems to work. Feb 14 15:13:49 Hello all. Does anyone know if the BeagleBoard xM comes with a battery so the date and time settings can persist over multiple boots? Feb 14 15:20:05 <_av500_> no Feb 14 15:20:08 <_av500_> no battery Feb 14 15:22:52 Geminizer: It does not come with a battery, but there is a place to attach one to keep date and time values persistent. Feb 14 15:41:21 jkridner... thanks for the tip Feb 14 15:47:19 * mranostay yawns Feb 14 16:45:55 If I rebuild u-boot for BeagleBone, which files besides the MLO and u-boot.img do I have to replace on the SD card? Feb 14 17:03:51 Mark__: that should be it Feb 14 17:10:10 is there any beagle fpga solution? ... i looked at beagle tracker... doesnt seem complete or/ready to buy k Feb 14 17:23:54 arcanescu: I've been trying to get the Beagle Tracker into production for more than a year now. Feb 14 17:24:09 Trouble is that all the folks I've talked to have higher priorities. Feb 14 18:33:59 Is there a way to not use get the bootloader to directly execute your program and not boot linux at all? .... bypassing the OS overhead on beagle? has anyone tried something like that>? Feb 14 18:34:51 im actually working on that right now Feb 14 18:35:20 I got a C program to run within u-boot Feb 14 18:36:21 jeeeesus Feb 14 18:36:23 holy holy how how ? Feb 14 18:36:41 look up u-boot standalone programs Feb 14 18:37:12 I guess you can run simple programs within u-boot, I'm just running a simple memory test Feb 14 18:37:26 okay... any intresting examples links you have to share would be intresting... i might not be looking to where you might be looking Feb 14 18:37:35 (more advanced than the built-in mtest function) Feb 14 18:37:50 the thing is i made a DVB-S plugin.... and apparently its slow... so im wondering why a 1ghz processor cant do that Feb 14 18:38:16 this plugin is in gstreamer btw i could easily make it standalone... but would it make a difference if it ran it through u-boot Feb 14 18:38:53 go to your u-boot source folder/examples/standalone Feb 14 18:38:59 there is a lot of information there Feb 14 18:39:20 thanks mark__ will do Feb 14 18:39:23 arcanescu: what do you think does the OS, if you call it overhead? Feb 14 18:39:58 well yea aholler: the memorymanagment process handling bladi bla di bla Feb 14 18:41:04 aholler: so thats why i call it overhead... i even tried raising priority but that didnt do much good Feb 14 18:41:28 anyone know why my angstrom kernel fails to boot after I rebuild my u-boot image? I assume I have to rebuild the kernel, I'm just a little lost to what needs to change. Feb 14 18:45:37 fyi, for anyone qualified and/or interested, i am looking for someone to do a write up on using openocd to debug linux kernel modules Feb 14 18:45:54 anyone interested, give me a shout at danders@tincantools.com Feb 14 18:49:29 prpplague: I'll will gladly read it :D Feb 14 18:50:10 dwery: hehe Feb 14 18:50:10 dwery: did I show you https://plus.google.com/100242854243155306943/posts/17K4JKvHUVE already? Feb 14 19:00:56 arcanescu: if you have tested changing the priority, what else runs? Feb 14 19:01:54 can't be much, if it could work without an os at all Feb 14 19:02:05 aholler: meaning? Feb 14 19:03:37 I'm just trying to imaging what you are doing ;) Feb 14 19:03:42 aholler: there are just 2 blocks Feb 14 19:03:59 one is called energy dispersal and the other is reed solomon enecoding rs(204,188) Feb 14 19:04:13 energy dispersal = xor the data with a shift register output Feb 14 19:04:39 in order to get around the processor issues with that i craeted a look up table... instead of shifting the register everytime Feb 14 19:04:46 and why should raising the priority might help, if nothing else runs? Feb 14 19:04:56 the problem comes when i add RS encoding Feb 14 19:05:07 it runs fine... but it is too slow for realtime Feb 14 19:05:25 like for example here are the times ive noted on a quad pc and beagle : Feb 14 19:05:45 5945798519 ns (0.099M) and 42855468750 ns (0.71 M) beagle Feb 14 19:06:01 prpplague: there is this zy1000 thing not sure if you are aware, based on openocd, the us distributor is I guess having a write up, check with them if you like Feb 14 19:06:56 aholler: are you saying i should try to run something else with it to confirm the priority is at max? Feb 14 19:07:06 no Feb 14 19:07:52 i mean there is another project i saw doing the same thing with an FPGA with a 50Mhz clock and 4Mhz internal for clocking the data out Feb 14 19:08:31 so im totally confused on what might be wrong other than the processor being slow Feb 14 19:08:40 you should try to find out what is slow Feb 14 19:08:52 it is the reedsolomon encoding block Feb 14 19:09:02 and ive replaced it now in such a way that it is using lookup tables Feb 14 19:09:13 mnt_real: yea i don't think they'd like me using their docs, hehe Feb 14 19:10:15 mnt_real: plus i don't see any detail docs on their site Feb 14 19:10:18 * prpplague looks again Feb 14 19:11:08 mnt_real: yea i don't see any detailed docs on using openocd for linux kernel debugging Feb 14 19:12:26 prpplague: I'd think they wouldn't mind especially since the doc will most likely mention the probe that they are selling, you want me ask them for you ? Feb 14 19:12:53 arcanescu: I just wanted to say that changing the priority doesn't make much sense if nothing else runs. Feb 14 19:12:55 *cough* again , i don;t think they want _me_ to use their docs as i am a competator Feb 14 19:13:12 prpplague: ok ;) Feb 14 19:13:16 Hhehehe Feb 14 19:13:18 aholler: oh! .. yeanothing else is running Feb 14 19:13:52 kernels are easy enough to debug without jtag :P Feb 14 19:14:11 ds2: indeed, but i've had a bunch of requests for docs on using gdb via jtag Feb 14 19:14:20 ds2: i just don't have the time to write it up Feb 14 19:15:07 prpplague: IIRC, openocd provides a GDB server, doesn't it? Feb 14 19:15:14 ds2: yea Feb 14 19:15:23 then what is there to write up? Feb 14 19:15:31 a gdb tutorial Feb 14 19:15:43 ;) Feb 14 19:15:45 what's wrong with the one gdb provides? :D Feb 14 19:16:02 ds2: a howto fo using openocd with gdb and kernel debugging Feb 14 19:16:20 ds2: apparrently it isn't enough for people to get going with openocd Feb 14 19:16:32 I think what prpplague needs someone explains how yo add the symbol info for module debugging etc when remote connect Feb 14 19:16:33 uh... okay Feb 14 19:16:41 there are some non-intuitive things you do when debugging the kernel Feb 14 19:16:44 why do you need symbols? Feb 14 19:17:19 if you are desperate enough to need a jtag, symbols are the last of your problems Feb 14 19:17:44 I think he wants to have doc that explains how to tell gdb client where to find the symbol info ... but yea I agree with what you just said ;) Feb 14 19:17:54 prefers just a memory dump via jtag to symbolic debugging Feb 14 19:18:28 prpplague, I'll have to talk to you about this at ELC, interesting topic :) Feb 14 19:18:32 * mdp goes to the airport Feb 14 19:18:38 heheh Feb 14 19:18:47 mdp: see ya there Feb 14 19:19:44 mdp: I thought you wrote that LinuxScope thing, quite a stretch from just dumping memory ;) Feb 14 19:19:50 sometimes I wish my pc would have an accessible jtag. Feb 14 19:20:03 for kernel debugging ;) Feb 14 19:20:05 smm Feb 14 19:20:23 it's easier to setup than anything else Feb 14 19:20:58 ds2: basically there are jtag docs, openocd docs, gdb docs, but there isn't a really good single doc that ties the core pieces together Feb 14 19:21:12 ds2: just looking for a decent write up of the overall process Feb 14 19:21:54 prpplague: I see. seems like spoon feeding Feb 14 19:22:16 ds2: yes, indeed from your perspective it most certainly is Feb 14 19:23:32 btw, anyone has aturioal about useing kvm for kernel-debugging? ;) Feb 14 19:24:44 \o/ working javascript ADC readout Feb 14 19:25:27 what's with that all .js hype? Feb 14 19:25:43 koen: wow, no! that's very nice! Feb 14 19:29:19 ynezz: my ELC presentation is in js Feb 14 19:29:46 would anyone happen to know where BeagleBoard xMs are in stock in the US ? Feb 14 19:29:47 koen: js or nodejs? Feb 14 19:30:25 ds2: nodejs as server, js in the browser Feb 14 19:31:17 I see Feb 14 19:31:57 koen: a variation of jkridner's scope demo from the other BB's? Feb 14 19:32:39 ds2: yes Feb 14 19:33:43 ds2: https://github.com/koenkooi/bonescript/commits/ELC2012-koen Feb 14 19:34:15 I'll look at it after I get mine done :-/ Feb 14 19:35:16 ds2: you may have a lot to do to catch up to that :) Feb 14 19:37:25 djlewis: let's just say, I'll be happy once I figure out my demo-less talk to go with the slides Feb 14 19:37:37 no time to even compete or get close! Feb 14 19:38:17 bbl Feb 14 19:41:50 koen: do you have a script that gets the pinout graphics out of the eeprom? Feb 14 19:44:06 dwery: yes, https://github.com/koenkooi/bonescript/blob/ELC2012-koen/ELC2012-koen/js/eeprom-web.js Feb 14 19:44:28 dwery: it generates an svg with jquery-svg Feb 14 19:44:53 dwery: I'll convert into a standalone app that also lists the muxmode when I get back Feb 14 19:44:56 pretty lovely. circuitco might benefits from this Feb 14 19:45:31 dwery: guess why Jason and I are doing this :) Feb 14 19:45:37 :D Feb 14 20:56:24 koen, looks pretty Feb 14 21:00:26 koen: but smells bad Feb 14 21:01:07 would smell better if written in ruby Feb 14 21:01:34 with sprinkles? Feb 14 21:01:47 mdp: hey i thought you left for the airport Feb 14 21:01:50 jkridner only likes json cause he thinks it's named after him Feb 14 21:02:03 I did, sitting on my plane Feb 14 21:02:10 fiddling with slideware Feb 14 21:02:48 bah, beaglebone ethernet dropping ~20% of incoming packets Feb 14 21:03:05 mdp: hehe Feb 14 21:04:04 * mdp is unable to use enet on his A1 bone Feb 14 21:04:09 * mranostay waves from Yocto Day Feb 14 21:04:38 mranostay, find a suggestion box, and suggest a new project name Feb 14 21:06:39 heh mdp always causing trouble Feb 14 21:23:24 Anyone know how to change the default standalone load address? Feb 14 21:28:35 mark__: less vague? Feb 14 21:28:41 load address of what Feb 14 21:29:42 in a self guided missle? Feb 14 21:30:02 djlewis: be nice :) Feb 14 21:30:38 sorry, i been watching the syrian crisis on msn.com Feb 14 21:30:54 did that ever end? Feb 14 21:36:11 Does anyboyd here have a good understanding of the data ram types in the PRU and how they interact with main system memory? I'm tryign to figure the best approach out to getting a PRU job to send data to main memory Feb 14 21:36:54 SilicaGel: via a DMA engine? Feb 14 21:37:04 that's what I'm wondering Feb 14 21:37:07 do you really need the DMA at all? Feb 14 21:37:24 Each PRU gets its own ram, normally ... and then there's also a SHARED pru static ram Feb 14 21:37:34 Can't the main CPU just access that through either l3 or l4 ? Feb 14 21:37:58 hmmm no clue that is a few levels above my knowledge Feb 14 21:38:23 a lot of the examples do in fact use DMA to do basically copies between PRU ram and main sdram Feb 14 21:38:31 and I'm sur ethe performanc eof that is much better Feb 14 21:38:39 but i'm not sure it is actually required Feb 14 21:52:11 reads like your are searching for peek/poke ;) Feb 14 21:52:51 the secret sausage of c64's ;) Feb 14 21:53:47 sorry, I don't have any dsp-knowledge at all ;) Feb 14 21:54:25 i'm actually looking at the bone's sitara processor which has two built-in PRUs Feb 14 21:54:34 but I asked here because it's really the same organization as the OMAP based beagles Feb 14 21:54:39 just bigger srams Feb 14 21:55:08 hmm, a PRU is what? Feb 14 21:55:19 It's sort of like a microcontroller built into it Feb 14 21:55:26 other vendors have out M0 into their cores Feb 14 21:55:47 T.I. uses this PRU which is their own invention, it's a fast risc processor for doing uber real time stuff Feb 14 21:56:02 it's also very useful for another reason Feb 14 21:56:15 and they have their own memory? Feb 14 21:56:35 people talk about using the PRU to do power management Feb 14 21:56:40 let it shut down the main CPU Feb 14 21:56:45 then watch for events to tell it when to wake up Feb 14 21:56:56 Yeah. The PRUSS is two PRUs (like two processors) Feb 14 21:57:01 each one has its own program memory Feb 14 21:57:04 and its own data memory Feb 14 21:57:14 and there's a "shared" memory as well Feb 14 21:57:29 thanks for that explanation Feb 14 21:57:31 we're talking small though. I think the pru data memory is like 8KBytes and the shared is 12K Feb 14 21:57:46 The PRUs on the AM3358/9 are 200 MHz ish but can be slowed down Feb 14 21:58:12 you can implement things like extra UARTs in the PRUs, or even a whoel CAN controller, or whatever Feb 14 21:58:38 people here didn't seem to like the PRU because it was TI proprietary Feb 14 21:58:45 Yeah. Feb 14 21:58:46 they'd rather have a M3 core Feb 14 21:58:50 Or an M0 Feb 14 21:59:00 it seemed like TI might not stick with the PRU for long Feb 14 21:59:01 I would rather have one of those as well, but the PRU is there Feb 14 21:59:11 People say that, but I'm not sure they know what they are talking about Feb 14 21:59:37 I guess I can see the PRU eventually going away but i don't know. Feb 14 22:00:02 people say a lot, indeed Feb 14 22:00:09 the thing is Feb 14 22:00:15 the PRU might be annoying for hobbyists Feb 14 22:00:22 but that's not who this is all targeted toward. Feb 14 22:00:59 SilicaGel: so the PRUSS can control the GPIOs? Feb 14 22:01:10 It can, yeah Feb 14 22:01:16 yeah, the am335x seems more geared towards industrial apps Feb 14 22:01:19 Generally speaking, It can see the global memory map. Feb 14 22:01:25 You know how pinmux works right? Feb 14 22:01:36 yep Feb 14 22:01:38 Well there's a secondary pinmux in front of the PRUs so what the PRU can see is more limited Feb 14 22:01:48 but it can see GPIO pins Feb 14 22:01:53 here's another crazy thing i bet you didn't know Feb 14 22:01:55 that's interesting Feb 14 22:01:59 the PRU susystem has another UART in it Feb 14 22:02:00 has anyone managed to get ubuntu working with audio input nad output? Feb 14 22:02:01 its own aurt Feb 14 22:02:04 dedicated to it Feb 14 22:02:13 sweet little PRU.... Feb 14 22:02:17 here kitty kitty :D Feb 14 22:02:20 I still can't seem to get audio input to work following the documentation on the wiki Feb 14 22:02:49 I've hear is somewhat poorly documented, right? Feb 14 22:02:54 heard* Feb 14 22:03:01 TheAlphaNerd: what are you using for audio input device? Feb 14 22:03:07 beagleboard xm Feb 14 22:03:09 hm well I don't know. There are a bunch of examples but they're complicated. Feb 14 22:03:11 audio input jack Feb 14 22:03:16 The assembly language is fairly well documented Feb 14 22:03:18 like a mic or what Feb 14 22:03:19 there's no C for it Feb 14 22:03:27 right now I'm just using an ipod Feb 14 22:03:32 so its a line level signal Feb 14 22:03:36 ok, that should be line level Feb 14 22:03:38 just trying to get it to work Feb 14 22:03:40 yep Feb 14 22:03:46 it works with angstrom Feb 14 22:03:49 run alsamixer Feb 14 22:03:53 I have Feb 14 22:03:55 unmuted everything Feb 14 22:03:59 max'd volumed Feb 14 22:04:04 no dice Feb 14 22:04:19 this is with the kernel + build form the wiki Feb 14 22:04:23 SilicaGel: interesting anyway Feb 14 22:04:28 angstrom build works out of the box Feb 14 22:04:32 I'll have to dig deeper someday Feb 14 22:04:38 does alsamixer show your audio devices? Feb 14 22:04:42 linaro can't even boot the mixer Feb 14 22:04:51 boot a mixer? Feb 14 22:04:58 start alsamixer Feb 14 22:05:05 buy a kitchenaid Feb 14 22:05:06 they boot instantly Feb 14 22:05:11 TheAlphaNerd: i could never get audio working on my xm in ubuntu Feb 14 22:05:12 :) Feb 14 22:05:26 smplman: there is a build of ubuntu from ccrma Feb 14 22:05:29 called satellite Feb 14 22:05:30 for the xm Feb 14 22:05:34 that has audio in / out working Feb 14 22:05:48 and a bunch of apps already compiled included pure data, c sound, and super collider Feb 14 22:05:58 but the kernel was compiled without a sync module for midi sync Feb 14 22:06:06 and its an old kernel so I can't exactly recompile Feb 14 22:06:07 :( Feb 14 22:16:48 * etheretic1 muses "this channel seems ripe for a no-holds-barred linux-audio-stinks-to-high-heaven- rant." Feb 14 22:17:59 maybe someone would use audio if it would go over hdmi Feb 14 22:18:03 Oh, I dunno, I was using my BB revC2 with audio I/O for ham radio apps. Feb 14 22:18:35 but attaching a hdmi and audi-cable is just to uncomfortable ;) Feb 14 22:20:00 I'm going to embed the beagle inside of an amplifier… for a headless electronic music speaker Feb 14 22:24:09 Hmmmmm Feb 14 22:24:24 there is no HDMI on the Beagle! Feb 14 22:24:54 there is on the xm Feb 14 22:24:58 not on the bone though I don't think Feb 14 22:25:06 its DVI with an HDMI plug methinks Feb 14 22:25:15 there is HDMI on the beagleboard Feb 14 22:25:51 for me the connector looks pretty like a standard hdmi-thingy Feb 14 22:25:54 only as a connector, not a standard Feb 14 22:25:59 ds2: when i read that for some reason i got the "there is no sanctuary" scene from logans run played in my head... Feb 14 22:26:19 well, the beaglenoard xM has a HDMI connector on it Feb 14 22:26:21 no cec, no audio, no hdcp Feb 14 22:26:42 ds2: http://youtu.be/q0Amt30_QVQ Feb 14 22:26:49 emeb: hey bud Feb 14 22:26:53 yes, the beagleboards after B models do have a dvi mode hdmi connector Feb 14 22:27:06 I never said that, I just said that, at least I, never used audio on the beagle because that would need a second cable Feb 14 22:27:07 prpplague: hi! What's new? Feb 14 22:27:13 prpplague: heh Feb 14 22:27:19 emeb: hi Feb 14 22:27:22 emeb: not a darn thing Feb 14 22:27:30 prpplague: when's your flight? Feb 14 22:27:32 is that hard disk seeking in the background audio? Feb 14 22:27:33 djlewis: duuuude. Feb 14 22:27:45 prpplague: Nothing new is a good thing some days. Feb 14 22:27:55 emeb: the sigrok guys were wanting to talk to you about maybe making a version of your fpga board designed as an logic analyzer Feb 14 22:28:08 prpplague: send 'em on over. Feb 14 22:28:36 emeb: hehe, join in over at #sigrok when you have time , biot and uwe_ are the ones to talk to Feb 14 22:28:56 * emeb is hoping to some day get an FPGA board into production. So far it's been kinda dry... Feb 14 22:28:59 Russ: for the logans run video? Feb 14 22:29:02 ya Feb 14 22:29:16 Russ: probably not, considering when logan's run was made.... Feb 14 22:29:48 hmm, something like a beagle-based open logic-sniffer would be nice ;) Feb 14 22:30:04 aholler: hence the sigrok stuff... Feb 14 22:30:23 emeb: thought you had one? Feb 14 22:30:28