**** BEGIN LOGGING AT Thu Nov 01 02:59:58 2012 Nov 01 03:08:36 Swaziland? Nov 01 03:09:30 alan_o: how goes the slides? Nov 01 03:10:05 I sent them to Craig Nov 01 03:10:20 All done except some pictures of my demo, which is not done yet :) Nov 01 03:10:22 getting there Nov 01 03:10:26 got my cape mostly soldered up Nov 01 03:10:38 7805, and tested the regulator Nov 01 03:10:43 scared to plug it into the bone :) Nov 01 03:10:51 want to get one other thing working first Nov 01 03:11:00 before I unplug all these breadboard wires Nov 01 03:11:12 7805? heheh Nov 01 03:11:33 Well, I've got one of those 4-way AA battery holders from Radio Shack Nov 01 03:11:39 little too hot for the bone to handle Nov 01 03:11:58 I'd think that is way too cold for the bone w/the 7805 Nov 01 03:12:03 7805's are 2V overhead Nov 01 03:12:22 4 AA Alk. is 6.1 fresh out of the box; 4AA NiCd/NiMH is 4.8V Nov 01 03:12:39 hmmmm Nov 01 03:12:57 well crap..... Nov 01 03:13:07 * alan_o is not a hardware guy... clearly Nov 01 03:13:09 got LDO? Nov 01 03:13:20 ok, that's not a problem. I can run off a 9v Nov 01 03:13:33 9V cells aren't made to output a lot of current Nov 01 03:13:41 bone is >100mA if you use ethernet Nov 01 03:13:43 alan_o: i'm a sw guy but knows hw more than average java guy :P Nov 01 03:13:50 that gives you maybe 30mins of power if you are lucky Nov 01 03:14:06 shouldn't run ethernet Nov 01 03:14:11 use 8 cells Nov 01 03:14:14 no eth, no ftdi Nov 01 03:14:33 9v's are for like 40mA max drain if you want reasonable life Nov 01 03:14:39 mranostay: I'm in the same camp Nov 01 03:15:34 I don't have an LDO Nov 01 03:15:43 yeah, I did the math on the 9v and it wasn't favorable Nov 01 03:15:51 didn't take into account drain Nov 01 03:15:57 I just need it to run for an hour :) Nov 01 03:16:02 use 2 of the 4AA's in a series Nov 01 03:16:10 and just burn off extra power on the 7805 then Nov 01 03:16:25 or get a RC battery pack...those come in 7.2V varieties Nov 01 03:16:31 If I had 6 AA's Nov 01 03:16:33 wallyworld or ratshack should have them Nov 01 03:16:39 that gets me to 9 Nov 01 03:16:52 I think radshack has a 6-way AA Nov 01 03:16:52 that'd work Nov 01 03:17:15 or better yet Nov 01 03:17:15 alan_o, and a power cord is not acceptable? Nov 01 03:17:23 get one of those "USB batteries" Nov 01 03:17:23 Well, it's going to be remote Nov 01 03:17:30 that gives you 5V out of the box Nov 01 03:17:37 yeah, I thought about the USB battery option Nov 01 03:17:54 but that runs the FTDI if I do that Nov 01 03:18:10 no, Nov 01 03:18:13 cut up a USB cable Nov 01 03:18:15 alan_o, you just going to send craig and updated set of slides or is that it? Nov 01 03:18:17 oh, well yeah Nov 01 03:18:20 and feed 5V into it directly Nov 01 03:18:21 s/and/an Nov 01 03:19:01 mdp: Craig said it's fine and that he'd get them from me after the talk. Nov 01 03:19:31 Seems like the LF talks a tough game early on then walks it back. Nov 01 03:19:37 I think I sent mine on the way out the door last time Nov 01 03:19:39 yeah Nov 01 03:19:45 the first thing I saw said, "must submit or be cancelled" Nov 01 03:19:53 alan_o: jack of all trades master of none here :) Nov 01 03:19:53 now it's all, "you guys are what make this conference" Nov 01 03:20:09 which has both helped and hurt me in jobs :) Nov 01 03:20:37 alan_o, that was my experience with things as LF took over elc Nov 01 03:21:00 it was super relaxed when it was just celf/tbird doing the logistics Nov 01 03:21:20 just dump them on elinux whenever ;) Nov 01 03:21:46 things were better when tbird did things Nov 01 03:21:54 I like tbird a lot Nov 01 03:22:14 so ds2, are you coming? Nov 01 03:22:24 alan_o: no time or budget for travel Nov 01 03:22:31 aww man.... :( Nov 01 03:22:45 up to eyeballs with stuff Nov 01 03:22:48 Well, ELC-SF is close enough. Nov 01 03:23:01 yes Nov 01 03:23:03 alan_o, yeah, tbird used to mail a few people asking, "hey, it's past the deadline but are you interested in doing a session?!?" Nov 01 03:23:07 hehe Nov 01 03:23:13 too close (temporally) Nov 01 03:23:47 who ever selected the date needs to be given a talking to Nov 01 03:23:50 ds2: yeah, I'm in the same boat. I'm turning good stuff away at this point, then this ELC stuff has caused me to put what I have on hold. I think it's worth it though. Nov 01 03:23:56 ds2: vote by mail! Nov 01 03:24:16 mranostay: mail who? Nov 01 03:24:35 last year it was too early already Nov 01 03:24:36 of course i tossed by vote ( ;) @ mdp) Nov 01 03:24:40 *my Nov 01 03:25:10 ds2, that must be an LF schedule thing Nov 01 03:25:26 it was always two month later like clockwork Nov 01 03:25:39 yep... used to end of Apr after ESC Nov 01 03:25:43 right Nov 01 03:26:45 should be about 3 weeks for the CFP Nov 01 03:26:48 maybe it'll be a bust Nov 01 03:27:01 arm techcon was sad Nov 01 03:27:15 it'll at least have the locals Nov 01 03:27:20 and that counts for a lot Nov 01 03:27:23 mdp: for ELC SF you mean? Nov 01 03:27:29 mranostay, yes Nov 01 03:27:38 3 weeks before E-ELC would make my airfare ungodly :) Nov 01 03:27:56 i wouldn't be going if my talk wasn't selected :) Nov 01 03:28:37 I mean, it's kinda ridiculous that one will end and the SF one will have to open the CFP almost immediately-ish Nov 01 03:29:04 maybe CELF should seperate from LF Nov 01 03:29:15 hmm i should submit one for ELC as well Nov 01 03:29:29 mranostay, travel is cheaper for you Nov 01 03:30:04 mranostay, taser cape! Nov 01 03:30:38 mdp: yeah well not if have to take days off for it :) Nov 01 03:31:46 mranostay: company not sending you? Nov 01 03:33:17 You know what grinds my gears? Music out of a piezo speaker. Nov 01 03:33:40 people sit there and listen to music on their cell phones. Do they not hear that it sounds like crap? Nov 01 03:33:57 alan_o: i work in a completely different field atm Nov 01 03:34:06 "cloud networking" Nov 01 03:34:19 oh really? I thought that was joking around... Nov 01 03:34:43 I was next to some server IT guys at lunch. Listening to them talk about their work problems. Nov 01 03:34:46 cloudy networks? storm coming? Nov 01 03:36:02 I was thinking about how much farther beneath them I am in the layering of hw/sw, and then how many layers there are beneath me. Nov 01 03:38:23 alan_o: needed a linux bringup guy and i needed out of Alabama it all works out :) Nov 01 03:38:52 so you did the bringup, and now you work on Node? Nov 01 03:40:33 * alan_o can't even make the dang uart work.... again :( Nov 01 03:40:45 UARTs are overrated Nov 01 03:40:52 console over GPIO Nov 01 03:41:00 bitbang! Nov 01 03:41:21 just put my tongue on the gpio pin... should be all anyone needs. Nov 01 03:44:14 why do we off-by-one in the DT? Nov 01 03:44:21 is uart0 not a valid identifier? Nov 01 03:44:36 Anyone know if regmap-i2c in kernel can handle special kind of i2c read requirements? For example if I need to switch to a particular page before doing a read, can regmap handle this? Nov 01 03:44:37 datsheet says uart0, it shows up as ttyO0, why uart1 in the dts? Nov 01 03:44:47 alan_o: no i work on the build as well.. i would quit if i had to do Node :) Nov 01 03:44:53 <-- barely kidding :) Nov 01 03:44:55 And, also the special case of writing to password protected registers which need a write to a password register? Nov 01 03:45:05 hi joelagnel__ Nov 01 03:45:11 mranostay: hey there Nov 01 03:45:35 alan_o: pull up the SRM! Nov 01 03:45:52 *page switching requiring writing to a certain page register over i2c as well Nov 01 03:46:30 mranostay: srm matches the trm, which is off by one from the dts Nov 01 03:47:33 alan_o: http://knowyourmeme.com/photos/181367-close-enough--2 Nov 01 03:48:13 I guess there are limitations. It'd be nice to hook pre and post functions to regmap to handle these case, still keeping the read/write code to a minimum which is the point of regmap Nov 01 03:49:38 mranostay: hehe Nov 01 03:56:52 alan_o: my close enough stage depends on the time of night Nov 01 04:25:38 dammit even Grub is GPLv3 now? Nov 01 04:27:59 ugh.. grub Nov 01 04:28:40 "don't edit this file, even though you see exactly what you need to change.. edit some other file that we're not going to tell you about which is extra convoluted" Nov 01 04:28:52 "and varies between distros" Nov 01 04:29:03 "and has lots of conditional sections" Nov 01 04:29:07 loves to explain this 'limitation' to x86 tomorrow :) Nov 01 04:29:35 since ppc and mips are so 'limited' from running a JVM :) Nov 01 04:29:37 hehe... we need to start working 'limitation' in Nov 01 04:29:46 I thought that was a feature Nov 01 04:29:52 new meme starting? :) Nov 01 04:29:58 maybe... Nov 01 04:30:44 so no pc jvms from IBM? Nov 01 04:30:48 they used to do that Nov 01 04:30:51 s/pc/ppc/ Nov 01 04:31:35 since Oracle? Nov 01 04:32:03 I don't keep up Nov 01 04:32:10 I haven't looked since Oracle Nov 01 04:34:04 ugh.... dts didn't fail on pintctrl-names (extra t) Nov 01 04:34:14 didn't fail compile Nov 01 04:34:19 failed to work :) Nov 01 04:34:53 alan_o: the real question do you really need java? :) Nov 01 04:35:08 learn a real programming language :) Nov 01 04:35:10 Well how else can you make things cross platform? Nov 01 04:35:23 C is just sugar for asm Nov 01 04:35:26 it's not webscale Nov 01 04:35:43 eh maybe 5% of the time Nov 01 04:35:54 huh? Nov 01 04:36:07 most C code is platform independant Nov 01 04:36:18 I was being funny (apparently unsuccessfully) Nov 01 04:36:20 if you are talking arch at least Nov 01 04:36:33 I agree, make stuff in C and C++ Nov 01 04:36:33 <-- beer time here :) Nov 01 04:36:51 I know the C++ will get stones thrown at me in this channel, but it is the right tool for a lot of jobs Nov 01 04:37:40 uart!!!! Nov 01 04:38:05 first had to figure out that pinctrl was overwriting my devmem2 Nov 01 04:38:11 remind me to stab you in BCN :) Nov 01 04:38:19 wonder how often that gets polled... Nov 01 04:38:25 for C++? Nov 01 04:38:28 yes :) Nov 01 04:38:32 How else you gonna make a GUI? Nov 01 04:38:35 gtk? Nov 01 04:38:38 that's a nightmare Nov 01 04:38:42 really? Nov 01 04:38:54 Well, I think so, compared to doing it in C++ with Qt Nov 01 04:38:56 of course most of my GUI work has been in Python... Nov 01 04:38:57 or with Gtkmm Nov 01 04:39:02 yeah, OO Nov 01 04:39:15 Objects work good for that stuff Nov 01 04:40:14 If I'm making a userspace app of any real size, I use C++. Nov 01 04:40:17 except python doesn't ******* thread Nov 01 04:40:34 thread? Nov 01 04:40:39 can't make threads in py? Nov 01 04:40:43 that doesn't seem right Nov 01 04:40:46 not real ones Nov 01 04:40:49 but what do I know Nov 01 04:40:49 oh Nov 01 04:40:51 green threads Nov 01 04:41:19 * mranostay googles Nov 01 04:41:26 yes that would be the term Nov 01 04:41:30 I don't know enough to ague either way on that one. I know blackdown when they did the Linux Java port did green first then eventually moved to native Nov 01 04:41:38 seems like they thought native was better Nov 01 04:41:41 also them do opcode counts for switching.. Nov 01 04:41:45 but that was 13 years ago Nov 01 04:42:19 by first job was python coding so it has infected me since :) Nov 01 04:42:27 hehe Nov 01 04:42:51 my first post-college non-intern job was MFC on this nightmarish codebase. Nov 01 04:43:07 took the whole team a week (more?) to get it built on my machine Nov 01 04:43:22 some people had left, lots of knowledge lost Nov 01 04:43:42 "I think you need msxml, but what version....." Nov 01 04:43:45 unreal ma Nov 01 04:43:46 man Nov 01 04:44:33 There was one engineer there, whenever stuff would crash, would just check for the pointer that was null and skip the code that crashed. Nov 01 04:44:39 alan_o: my next job was EA as a lab bunny/intern.. that never graduated... :) Nov 01 04:44:53 never figured out _why_ NULLs were getting passed around. Nov 01 04:45:03 so as a result we just lost data left and right Nov 01 04:45:07 what a mess Nov 01 04:45:17 EA where? Nov 01 04:45:24 ewwwwwwwwwwww grub Nov 01 04:45:26 we have EA here (Madden and NCAA) Nov 01 04:45:30 alan_o: Embedded Alley Nov 01 04:45:38 oh Nov 01 04:45:41 not EA Games ;) Nov 01 04:45:54 alan_o: worked in the same office in Ohio as mdp Nov 01 04:45:59 so you're 25 on your 4th job? Nov 01 04:46:46 alan_o: counting acquisitions? Nov 01 04:46:56 I don't count those Nov 01 04:47:03 mranostay: you were only at EA? did you do any of the other emblinux companies? Nov 01 04:47:07 3rd and 26 Nov 01 04:47:59 you listed 3, and you worked 2 places after that Nov 01 04:48:04 no two Nov 01 04:48:19 The two on your linkedin? Nov 01 04:48:41 EA was bought by Mentor Nov 01 04:48:49 oh ok Nov 01 04:48:57 bought...killed... murdered Nov 01 04:49:15 no comment Nov 01 04:49:20 no comment Nov 01 04:49:27 ahahahahaahhaahahhhahah Nov 01 04:49:53 mranostay, I have no one to blame but myself, ultimately ;) Nov 01 04:50:15 i'm not totally complaining :) Nov 01 04:50:34 only the MGC part sucked Nov 01 04:50:41 MGC job part, that is Nov 01 04:53:22 i shouldn't even say this here. but i did absolutely nothing for the last year :D Nov 01 04:57:26 alan_o: so your talk has a demo? Nov 01 04:58:18 I hope :) Nov 01 04:58:43 I think I'll get there Nov 01 04:58:44 #wirenestbomb Nov 01 04:58:50 haha Nov 01 04:58:54 I'm soldering right now Nov 01 04:59:22 i totally expect worst case my first demo being to TSA Nov 01 04:59:41 does it look like a wirenestbomb? Nov 01 04:59:52 no minus the disconnected tube Nov 01 05:00:20 i fit the domestic terrorist profile :) Nov 01 05:07:46 alan_o: your long hair isn't helping :) Nov 01 05:07:57 at least mdp is a boy scout Nov 01 05:08:21 mranostay: bring a prayer mat with you Nov 01 05:09:06 ds2: can i yell "allahu akbar" too? Nov 01 05:09:16 sure, why not Nov 01 05:09:29 hello, for the beaglebone readme page:http://beagleboard.org/static/beaglebone/latest/README.htm Step 4, the code has an error, and didn't run success. Nov 01 05:09:32 and be sure to order a halah meal too Nov 01 05:10:55 mdp is a boy scout? or just looking? Nov 01 05:10:57 The revised version you find at jcqiu.wordpress.com Nov 01 05:12:06 Jie: best bet is to post that to the mailing list, and cc Jason Kridner Nov 01 05:12:48 alan_o: asst. scoutmaster Nov 01 05:13:26 alan_o: ok, thanks Nov 01 05:14:01 mranostay: hey, me too Nov 01 05:14:56 well emeb_mac welcome to the machine Nov 01 05:15:19 thx mranostay Nov 01 05:15:30 what's shakin' this lovely evening? Nov 01 05:15:47 go trickertreating? Nov 01 05:16:26 apartments with external gates :) Nov 01 05:16:36 keeps the punk kids away :) Nov 01 05:17:37 hardly anyone in our area does it anymore. Nov 01 05:17:54 Last time we gave out candy we had like 5 groups come to the door all night. Nov 01 05:18:36 scumbag parents... don't take candy from strangers... allows you to go out once a year to get candy from stangers :-_ Nov 01 05:18:42 so the last few years we just turned out the lights and played "not home" Nov 01 05:18:50 yeah. No trust any longer. Nov 01 05:19:34 i scare my neighbors i think Nov 01 05:19:42 Highest turnout ever for me this year Nov 01 05:20:20 in any apartment they have seen me carry guns to my car, loud music, excessive drinking with friends, and general bad stuff :) Nov 01 05:20:44 And then there's that cat. Nov 01 05:20:53 what? no special plant parts being incinerated, mranostay? Nov 01 05:21:19 no comment Nov 01 05:21:39 mranostay: you need carriers for your guns that don't look like guns Nov 01 05:21:49 not to keep from offending people Nov 01 05:22:05 i used to carry my ar-15 to my car in Ohio without a case :) Nov 01 05:22:06 but to keep people from knowing that you have guns in your house when you're not home. Nov 01 05:22:22 Yeah, I guess I say that and do understand that it cuts both ways Nov 01 05:22:24 alan_o: then people would think he's a bass guitarist or a violin player Nov 01 05:22:40 alan_o: i had a cop neighbor too :) Nov 01 05:22:51 you can have guns in CA? Nov 01 05:23:02 yes just not as open :) Nov 01 05:23:18 some are stored in Ohio :) Nov 01 05:23:34 btw, I re-read my comment and it reads like I'm your dad or something.... don't read it that way :) Nov 01 05:23:54 it is unlikey i think :) Nov 01 05:24:22 I have a friend who says, "I'm old enough to be your dad; in fact, what's your mom's name?" Nov 01 05:24:33 alan_o: I haven't use the google group before. should I join the group, and then able to post it to the mailing list? Nov 01 05:24:36 * alan_o is only 5 years older than mranostay Nov 01 05:24:46 Jie: something like that Nov 01 05:24:52 I don't remember the exact steps Nov 01 05:24:52 alan_o: yeah but look 10 years younger :P Nov 01 05:25:20 * emeb_mac is old enough he doesn't talk about his age any longer. Nov 01 05:26:15 i'm starting not to get carded with a day or two of not shaving Nov 01 05:26:36 it's all downhill from there. Nov 01 05:26:39 I'm the same way but s/shaving/bathing/ Nov 01 05:26:49 * emeb_mac holds nose Nov 01 05:27:18 my face for the record Nov 01 05:27:34 j/k, but no shave + no bath + long hair + the lumberjack clothes that I wear, is kind of scary Nov 01 05:27:41 I like to go into the bank like that Nov 01 05:27:42 hmmmm not bathing in FL Nov 01 05:27:58 it helps if you're carrying an ax. Nov 01 05:28:11 alan_o: do you live in a unabomber shack? Nov 01 05:28:14 alan_o: have a FL CC? :) Nov 01 05:28:25 why not add a hockey mask and chain saw? Nov 01 05:28:33 I do not have an FL CC, if it means what I think you mean Nov 01 05:28:38 I _should_ Nov 01 05:28:43 * mranostay notes he can carry in pretty much every state but the one he lives in :) Nov 01 05:28:43 and I'm a lazy idiot for not Nov 01 05:28:55 alan_o: i have one :) Nov 01 05:28:57 emeb_mac: no, I live in a house farm. Nov 01 05:29:08 koen: I think this patch is critical for mcasp, but hasn't been pushed yet though it was supposed to be a long time back. this could be why we have broken mcasp, both mdp and vaibhav submitted patches for this Nov 01 05:29:28 would be a nice thing to include in the beaglebone kernels: https://patchwork.kernel.org/patch/1499271/ Nov 01 05:29:42 Haven't tested yet if this fixes my audio issues Nov 01 05:29:59 anyone playing fedora on beaglebone? Nov 01 05:30:29 Jie: are you trying to make mranostay mad? Nov 01 05:30:35 s/pushed/included in mainline/ Nov 01 05:30:39 * joelagnel__ zzzz Nov 01 05:31:04 * alan_o is afraid to plug his cape in now that it's all soldered Nov 01 05:31:09 what's all this fedora on beagle stuff lately. Nov 01 05:31:20 alan_o: let out the magic smoke! Nov 01 05:31:33 alan_o: .... I'm not kidding Nov 01 05:31:40 (in the manner of Emily Littella) Nov 01 05:31:40 Jie: why fedora? Nov 01 05:31:53 mranostay: reminds me of the 'lost' series :) Nov 01 05:32:22 alan_o: i was going to do my demo with a battery cape ask me about that at E-ELC :) Nov 01 05:32:23 mranostay: I goona do a project base on fedroa Nov 01 05:33:22 why? Nov 01 05:33:27 so is that battery cape UL approved? Nov 01 05:33:36 mranostay: battery cape is just so big. Nov 01 05:34:02 mranostay: student project. programming GPIO on beaglebone of Fedora 17 Nov 01 05:34:18 Do it on Angstrom that comes with it Nov 01 05:34:20 it already works Nov 01 05:34:26 Exact steps are on the website Nov 01 05:34:30 why make it hard? Nov 01 05:34:33 for no gain? Nov 01 05:35:00 alan_o: teacher's order! other wise gonna fail.... Nov 01 05:35:04 oh Nov 01 05:35:15 so there's your answer emeb_mac Nov 01 05:35:27 :) Nov 01 05:35:43 Yeah, so there's some downloads on one of the web pages for it. Nov 01 05:35:47 then just fail Nov 01 05:35:50 go forth and burn sd cards Nov 01 05:36:03 alan_o: could u give me the link about u said worked on Angstrom Nov 01 05:36:23 google for gpio beaglebone Nov 01 05:36:28 there are tons of blog articles Nov 01 05:36:38 /sys/class/gpio Nov 01 05:36:46 export, read, write Nov 01 05:37:05 stupid teachers Nov 01 05:37:32 "I can't be bothered to fully understand something, so you have to follow my ignorance or fail" Nov 01 05:38:19 well, the other side of that is...... "shut up and do as your told and I don't want to hear any backtalk" Nov 01 05:38:45 another bad way to go. Nov 01 05:39:00 you talk as though you don't remember school :) Nov 01 05:39:06 does there have beaglebone fedora 17 image now? I can't find it Nov 01 05:39:21 alan_o: I've tried to block it. Nov 01 05:39:50 still have those nightmares about being unprepared for midterms though. Nov 01 05:39:54 Jie: http://elinux.org/BeagleBone#Fedora Nov 01 05:40:02 first result for "beaglebone fedora" Nov 01 05:40:10 yay - google is working again! Nov 01 05:40:30 alan_o: I went to there, only image for beagleboard xm Nov 01 05:40:49 Jie: http://beagleboard.org/static/beaglebone/latest/README.htm Nov 01 05:40:58 might not be a supported fedora configuration Nov 01 05:41:26 http://elinux.org/BeagleBoardFedora Nov 01 05:42:14 alan_o: the readme page I already went throughed Nov 01 05:42:19 * mranostay splashs holy water on Jie Nov 01 05:42:34 need garlic too? Nov 01 05:44:18 I just touch this stuff for a while, Nov 01 05:44:22 * alan_o disconnects his bone, hoping for the best.... Nov 01 05:44:42 Thanks Nov 01 05:46:24 alan_o: when is the flight? Nov 01 05:47:53 Sat 3pm Nov 01 05:48:11 booting up Nov 01 05:48:18 * alan_o wipes brow Nov 01 05:48:19 alan_o: what airline? Nov 01 05:48:24 American Nov 01 05:48:36 one stop in miami, 1.5h layover, then zzzzz all the way to BCN Nov 01 05:49:10 ouch Nov 01 05:49:30 ouch? I think it's a pretty nice flight Nov 01 05:49:33 no JFK nonsense Nov 01 05:49:42 what is your touchdown time? Nov 01 05:49:46 the layover is early, so the ZZZ can be uninterupted Nov 01 05:50:07 830AM Sunday Nov 01 05:50:56 * mdp is now adjust to left coast time Nov 01 05:52:06 oh you got a few days Nov 01 05:52:09 plenty of time Nov 01 05:52:26 i so going to be jetlagged for my talk :) Nov 01 05:52:30 oh yeah... except being in Miami on Friday. Nov 01 05:52:36 mranostay: been to Europe before? Nov 01 05:52:43 taking the redeye? Nov 01 05:52:53 The key is this... sleep on the plane Nov 01 05:52:55 hello. I am on my 3rd day of trying things on bb xm. Does anyone have any benchmark on performance? I am using the SD card configuration it was shipped with Nov 01 05:52:56 as much as you can Nov 01 05:53:04 and to me it seems to run very slow Nov 01 05:53:10 alan_o, yeah, right...so easy to sleep Nov 01 05:53:24 not just UIs. but even downloading files seems extremely slow Nov 01 05:53:25 bulbage: incinerate that card. buy a decent card and use that Nov 01 05:53:28 then... when you get there, stay AWAKE until 9 or 930. It's hard, but the next day you're mostly on schedule Nov 01 05:53:33 that card is slower then molasses in siberia Nov 01 05:53:46 mdp: it's hard but you have to try, that's my experience. Nov 01 05:53:53 lots of red-eyes to UK Nov 01 05:54:12 mdp: you're tall right? Nov 01 05:54:18 you look tall in your picture Nov 01 05:54:26 alan_ol, mranostay, good luck with your talks. looking forward to watching the live video stream :) Nov 01 05:54:31 ds2: what card do you suggest? Nov 01 05:54:40 alan_o: is he next to mranostay in that picture? Nov 01 05:54:50 alan_o, I once did the 8am arrival in paris thing on monday (couldn't leave on saturday due to commitment) ... I spent the entire flight telling myself I had to go to sleep or I would be miserable throughout the all-day meetings... Nov 01 05:55:09 alan_o, well, I managed to keep myself awake all the flight..sucked Nov 01 05:55:13 bulbage: I have had good luck with local store brand... so I don't have a reocmmendation... almost any retail card will be faster then that card Nov 01 05:55:27 mdp: oh yeah, you can't work the next day. Nov 01 05:55:29 then the usual miserable 1.5 hour cab ride across to the other side of paris Nov 01 05:55:30 blech Nov 01 05:55:33 ds2: what pic? Nov 01 05:55:44 the next day is blown, but the day after that.... is _almost_ ok. Nov 01 05:55:49 mranostay: donno - you look tall in your picture Nov 01 05:55:50 for me anyway Nov 01 05:55:57 alan_o, there have been times when my comfort was less important than what I had to be around for here Nov 01 05:56:12 mranostay, ds2 linkedin profile Nov 01 05:56:17 oh Nov 01 05:56:38 alan_o, 6'1" so tall enough to be uncomfortable on an intl flight Nov 01 05:56:54 mdp: yeah.no doubt. I have a set of those eye covers. They look stupid but they work. Nov 01 05:57:00 it's mostly seat width for my shoulders quite honestly Nov 01 05:57:02 Virgin Atlantic gave them to me Nov 01 05:57:08 yeah, me too Nov 01 05:57:13 I'm 6-1 with wide shoulders Nov 01 05:57:36 you look wider than me in the picture, and I have a lot of trouble with that Nov 01 05:57:43 usually my ass goes painful to numb and back so I can't fall asleep Nov 01 05:57:43 especially 3 guys in a row Nov 01 05:57:54 i'm short so unless i have a fat guy next to me i'm fine Nov 01 05:58:08 mranostay: yeah, planes were made for small guys Nov 01 05:58:34 alan_o, yeah, probably worst was 3 of us guys...none fat..sitting in a row..all with wide shoulders Nov 01 05:58:51 we all had to turn at an angle from CLE->SFO that night Nov 01 05:58:53 *sigh* Nov 01 05:59:06 ds2: you mean as a general purpose computer? I am a bit confused by your comments. Nov 01 05:59:29 hey good news...my new distractions are here! Nov 01 05:59:40 ?? Nov 01 06:00:17 launchpad lm4f120s have arrived! Nov 01 06:00:23 mdp: Last one I was on, I was in asile, and there was this big burly guy in the middle. I was touching his shoulder and getting hit with the drink cart. Nov 01 06:00:29 StellarisWare in ROM, baby! Nov 01 06:01:16 alan_o, I usually go aisle and sprawl my legs out there...I've gotten good at reading at subconciously drawing my legs in and out as the cart passes Nov 01 06:01:19 installed my wireless module... still works... Nov 01 06:02:01 this is an rpi killer Nov 01 06:02:23 I do asile unless it's a redeye, then do window to lean on it (doesn't always work). Nov 01 06:02:31 mdp: did you say you're in CA? Nov 01 06:02:54 no , not traveling..at home in OH Nov 01 06:03:16 you said something about Left Coast time Nov 01 06:03:19 I'm just keeping odd hours Nov 01 06:03:24 ohh right Nov 01 06:03:32 just because I'm hacking at 2am local time Nov 01 06:03:57 ds2: so is this http://www.youtube.com/watch?v=mcy929FV8Qg false advertising? Nov 01 06:04:02 oh, I keep forgetting you're not CDT in OH Nov 01 06:04:24 I guess not advertising since it is not by anyone from bb.org Nov 01 06:04:31 that should be fine Nov 01 06:04:41 the problem is they are using a slow card Nov 01 06:04:45 nothing wrong w/the sw Nov 01 06:05:08 alan_o, EDT goes pretty far west of here Nov 01 06:05:22 yeah, it's a wedge, iirc Nov 01 06:05:32 * alan_o googles the tz map Nov 01 06:05:45 indianapolis and louisville are still in EDT Nov 01 06:06:01 I guess I find it strange that on beagleboard chat room I am being told to not use the beagleboard. Nov 01 06:06:39 I thought I would be reading: yes use the board but this is the right way to speed it up... Nov 01 06:06:43 but most of TN is not. And I always think of TN being southeast and OH being midwest, but TN is actually west of OH Nov 01 06:06:53 right Nov 01 06:07:04 funny the way the map works in the mind Nov 01 06:07:05 alan_o: ohio is weird state :) Nov 01 06:07:11 High in the middle Nov 01 06:07:14 OH is a fine state Nov 01 06:07:14 kinda the midwest Nov 01 06:07:21 friendly folks there Nov 01 06:07:22 sure not the east coast Nov 01 06:07:29 ds2, where? Nov 01 06:07:31 ds2: hey i know i lived there for 24 years :) Nov 01 06:07:38 only complaint is the damn turnpike ;) Nov 01 06:07:51 mdp: friendlier folks then the Chicago ones Nov 01 06:07:53 just no damn jobs or good pay :) Nov 01 06:07:59 ds2, not saying much Nov 01 06:08:05 mranostay: where, east or OH/ Nov 01 06:08:06 ? Nov 01 06:08:07 mranostay, remote remote remote ;) Nov 01 06:08:25 of course, I only know of OH near the turnpike Nov 01 06:08:37 I know a lot of remote people, but nobody seems to want to _hire_ remote people Nov 01 06:08:38 ds2, the fastest road in OH Nov 01 06:08:41 <_av500_> yo Nov 01 06:08:43 maybe I'm not l33t enough yet. Nov 01 06:08:48 ho! Nov 01 06:08:55 mdp: my plan in 5-10 years :) Nov 01 06:09:55 alan_o, I just had a lucky set of circumstances Nov 01 06:10:23 mdp: wasn't EA a fully distributed company? Nov 01 06:10:30 well ,wait, not lucky Nov 01 06:10:40 ds2: CA and OH Nov 01 06:10:41 I controlled my ability to go live where I wanted to Nov 01 06:10:59 http://www.youtube.com/watch?v=kyGSaHr00xo another one claiming "laptop like" performance Nov 01 06:11:06 mranostay, it was CA ,AZ, .gr., .jp, .ru at first Nov 01 06:11:14 then I moved AZ->OH Nov 01 06:11:15 oh... thought you guys had resources in .ru and AZ too? Nov 01 06:11:29 yes, Tartarus and myself in AZ Nov 01 06:11:37 like the founder was from AZ Nov 01 06:11:38 mdp: don't forget NH Nov 01 06:11:41 ah okay Nov 01 06:11:51 oh, yes, and NH Nov 01 06:11:54 you guys got the guy out in NH? Nov 01 06:12:08 4 founders Nov 01 06:12:17 NH, AZ, and 2 in CA Nov 01 06:12:32 NH moved to CA and is there Nov 01 06:12:37 mranostay works with him now ;) Nov 01 06:12:44 mdp: where in AZ? Nov 01 06:12:53 emeb_mac: I was in phx Nov 01 06:13:05 * emeb_mac is in mesa Nov 01 06:13:12 Tartarus is/was in tuc but is moving now Nov 01 06:13:20 ohhh different guy from NH Nov 01 06:13:23 NH/MA Nov 01 06:13:30 emeb_mac: cool..home was in chandler Nov 01 06:13:32 lotta folks moving out of TUS these days it seems. Nov 01 06:14:10 mdp: and now OH? probably gained 5lbs in H20 Nov 01 06:14:35 originally from here Nov 01 06:14:41 it's worked out quite well Nov 01 06:14:50 mdp: me too - Cincinnati. Nov 01 06:14:59 everybody is from here! ;) Nov 01 06:15:23 one of those funny things in the various job attractor areas Nov 01 06:15:26 oh why oh why oh why oh... Nov 01 06:15:50 I am glad I bought the board and joined this chat room...I am finding out a lot about people's opinions and chatter Nov 01 06:16:00 it's social! Nov 01 06:16:09 ok, time for me to crash...bbt Nov 01 06:16:15 I guess whenever I feel lonely I will join this chat room Nov 01 06:16:22 they got social networks, we got social embedded computing. Nov 01 06:16:31 and start talking a storm about my travels and opinions Nov 01 06:16:43 it's only bandwidth. Nov 01 06:16:49 I am wondering where people go for technical support? Nov 01 06:16:52 hmmm... Nov 01 06:17:00 ask a technical question. Nov 01 06:17:10 maybe someone will answer. Nov 01 06:17:14 where is cincinnati? Nov 01 06:17:27 SW OH Nov 01 06:17:29 is that technical for this chat room? Nov 01 06:17:51 now I get the feeling your being sacastic. Nov 01 06:18:02 s/sacastic/sarcastic/ Nov 01 06:18:28 heh Nov 01 06:18:29 maybe it is getting late or maybe it is because I just was told in bb chat room to not use the bb xm Nov 01 06:18:41 no, use the xM Nov 01 06:18:42 avoid the bone Nov 01 06:18:51 i am using the xm Nov 01 06:18:55 * mranostay tilts head Nov 01 06:18:58 <_av500_> maybe rpi could help here? Nov 01 06:20:18 _av500_: remind me to stab you at E-ELC :-) Nov 01 06:20:20 but not loving the speed. mine seems very sluggish. I tried ubuntu 12.04 and it was worse. I just did opkg update/ugrade Nov 01 06:20:25 still pretty slow box Nov 01 06:20:52 is there any benchmark data? Nov 01 06:21:01 here's solid evidence of what platform to use: http://www.google.co.uk/trends/explore#q=raspberry%20pi%2C%20beagleboard%2C%20beaglebone%2C%20arduino&cmpt=q Nov 01 06:21:07 <_av500_> mranostay: damn, being stabbed in 2 channels at the same time is tough Nov 01 06:22:41 _av500_: what is your blood type? Nov 01 06:25:06 another disappointed customer. Nov 01 06:25:19 use a faster card Nov 01 06:25:32 a leaner OS? Nov 01 06:25:47 if that is too hard to understand, prehaps you should wait for the leapfrog board or the presk00l board Nov 01 06:26:12 I'm sure Disney will have one out soon. Nov 01 06:27:34 well panto-man Nov 01 06:27:43 hi mr Nov 01 06:27:45 i see the swing shift is here Nov 01 06:27:46 mranostay, Nov 01 06:27:53 damn autocomplete Nov 01 06:28:30 emeb_mac: This is more chatty than usual Nov 01 06:28:58 usually there's traffic but somewhat more technical Nov 01 06:29:05 It's too bad about bulbage Nov 01 06:29:56 yes Nov 01 06:30:20 still didn't quite get what his question was Nov 01 06:30:33 reading the log Nov 01 06:30:39 ds2 told him to get a new SD card Nov 01 06:30:53 bulbage thought "card" ment his xM Nov 01 06:31:23 ds2 said "you should incinerate that card" after bulbage said "booted from the SD card and ..." Nov 01 06:31:50 another Cool Hand Luke situation. Nov 01 06:31:50 then ds2 said get a store brand, and bulbage thought he meant "buy a PC" Nov 01 06:32:03 "What we have here is a failure to communicate" Nov 01 06:32:03 emeb_mac: which scene? Nov 01 06:32:05 oh Nov 01 06:32:07 that :0 Nov 01 06:32:14 I was gonna say, I know that movie well Nov 01 06:32:37 Only saw it once. Good one. The prison seemed so... innocent. Nov 01 06:32:48 yeah Nov 01 06:32:54 a lot like summer camp Nov 01 06:33:00 exactly. Nov 01 06:33:17 Filmed in FL by the way.... Nov 01 06:33:37 If prison were really like that, there'd be less people in prison Nov 01 06:33:49 ? Nov 01 06:34:03 maybe more - nice vacation. Nov 01 06:34:04 because while it seemed summer-campy, 8-10 hours a day of cutting brush by the road would be a pretty good deterent Nov 01 06:34:12 there's that. Nov 01 06:34:17 most of the movie isn't focused on the work Nov 01 06:34:36 but it's 6 days a week cutting brush, then on Sunday you can fight. Nov 01 06:34:46 or the chain gang at the beginning of O Brother Where Art Thou. Nov 01 06:35:17 only saw taht one once. Liked it. Nov 01 06:35:47 it's pretty funny. Nov 01 06:35:57 One of the Cohen Bros. better ones. Nov 01 06:36:12 right up there with Raising Arizona and Fargo. Nov 01 06:36:26 I've not seen either of those Nov 01 06:37:25 Nov 01 06:37:28 ok, I'm out of here. I have to get some sleep if I want to accomplish anything tomorrow Nov 01 06:37:45 emeb_mac: yeah, I don't see as many movies as I'd like Nov 01 06:38:14 but it was a good night. My cape works, no magic smoke... All the rest is just software (and mechanical) Nov 01 06:38:22 what cape? Nov 01 06:38:31 for my ELC presentation Nov 01 06:38:56 http://elceurope2012.sched.org/event/3ac1543b2161c0127d0a68d2491c576c#.UJIZASOsReg Nov 01 06:39:19 Nice! Nov 01 06:39:44 It's been a long week :) Nov 01 06:40:00 I bet. Getting ready for show & tell is always hectic. Nov 01 06:40:18 alan_o: we only have to be sober Monday remember :) Nov 01 06:40:43 Thats right man.... Monday at 12PM CET, vacation starts :) Nov 01 06:40:58 It'll be nice to be "off" for a few days Nov 01 06:41:04 alan_o: hotel bar? :) Nov 01 06:41:25 I'll be wherever the crowd is Nov 01 06:43:36 ok,laters all Nov 01 06:43:52 good idea Nov 01 06:43:53 laters Nov 01 08:50:55 koen: I'm trying out angstrom-v2012.12-yocto1.3 with the same mlo/u-boot/kernel I had before. After 'Starting kernel ...' all is silent on the console. Any idea why that could be? Has changed? Nov 01 08:52:37 tasslehoff try the old setup Nov 01 08:53:09 what does AM33XX_PULL_ENBL mean? Nov 01 08:55:48 woglinde: hm? the old setup? Nov 01 08:55:50 if i don't want to use pullups, do i need to change it to AM33XX_PIN_OUTPUT? Nov 01 08:59:44 tasslehoff the one which worked before Nov 01 08:59:55 dont tell me you did not save that Nov 01 09:01:18 woglinde: heh. we have a stable release, but I'm trying to move forward Nov 01 09:01:57 so I want to make the new setup work Nov 01 09:03:02 good moaning Nov 01 09:03:11 moan away brother Nov 01 09:03:39 * dm8tbr also decided to indulge in some mild trolling at UDS/LCE today Nov 01 09:04:03 tasslehoff good try this old setup Nov 01 09:04:13 maybee your hw got borked or so Nov 01 09:04:30 dm8tbr about what? Nov 01 09:05:02 * dm8tbr is wearing a MER-project t-shirt today Nov 01 09:05:15 haha better maemo Nov 01 09:05:42 woglinde: nope. I can switch to the old setup and it works fine, so there's something else at play here. will try a different kernel. Nov 01 09:05:57 tasslehoff good so this is clear Nov 01 09:06:04 better double checking first Nov 01 09:06:15 actually meego redone proper, nowadays Nov 01 09:41:16 tasslehoff: there is something about compiler versions Nov 01 09:41:26 and newer gcc failing to compile kernel properly Nov 01 09:41:53 dm8tbr: how many apps in the store? Nov 01 09:42:45 av500 ah right yes Nov 01 09:42:56 4.5 or 4.6 was last version Nov 01 09:43:04 av500: apps? store? Nov 01 10:05:30 av500: ah, thanks. Nov 01 12:32:32 xxiao: you was right: https://semiaccurate.com/2012/11/01/freescale-layoffs-target-one-particular-product/ Nov 01 12:42:41 av500/xxiao thats sad news Nov 01 12:46:42 hmm, how to cheer you up? Nov 01 12:46:53 cheap chinese chips are so yummy! Nov 01 12:47:02 hmm, I need food Nov 01 12:47:04 bbl Nov 01 12:47:57 http://mashable.com/2012/11/01/kim-dotcom-mega/ Nov 01 12:48:01 finally Nov 01 13:09:20 av500: so what exactly is Freescale's "areas of high-value and high-growth", I wonder? Nov 01 13:23:03 wmat good question Nov 01 13:27:39 * xxiao is checking various boards on the bench, more and more will be useless due to EOL from the chip makers Nov 01 13:29:42 av500: freescale acquired A15 license but gave up on implementation a while ago, which was a sign Nov 01 13:36:46 jkridner, ping? Nov 01 13:36:55 panto: pong Nov 01 13:37:02 hi there Nov 01 13:37:11 hi, nice to chat with you... Nov 01 13:37:16 thanks for your efforts on capebus!! Nov 01 13:37:17 I don't know if you're following the flame-fest in lkml Nov 01 13:37:24 I'm not. Nov 01 13:37:27 jkridner, hey, that's what you guys are paying me :) Nov 01 13:37:27 which one? Nov 01 13:37:42 [PATCH 0/3] capebus moving omap_devices to mach-omap2 Nov 01 13:37:49 wmat: no idea Nov 01 13:38:15 looks like we need an intervention Nov 01 13:38:20 k. Nov 01 13:38:26 either way works Nov 01 13:38:37 I'm not thin skinned Nov 01 13:38:56 k... I'm trying to get my browser to launch. Unity died on me. Nov 01 13:39:03 and I'm annoyed at the hypocrisy Nov 01 13:39:42 jkridner: it's 20+ emails by now Nov 01 13:39:56 av500: you'll get a good laugh at the suggestions Nov 01 13:40:03 k. Mac's Internet is stupid slow for some reason. Nov 01 13:40:18 this Mountain Lion upgrade has been nothing but trouble. Nov 01 13:40:24 koen: ? Nov 01 13:40:30 jkridner, agreed Nov 01 13:40:40 it's like mac-os went downhill after snow leopard Nov 01 13:40:47 jkridner: install Linux Nov 01 13:41:04 Apple ][ was good ..... Nov 01 13:41:16 av500: parse the cape eeproms in userspace, make it download "drivers" over usb, reboot Nov 01 13:41:28 av500: "drivers" being use in a general sense, not linux drivers Nov 01 13:41:56 the other suggestion was to patch all i2c drivers to enable autodetect Nov 01 13:41:58 wmat: I'm typing here on a Linux machine... Nov 01 13:42:10 but, Unity crashed, so I don't have a way to open a browser. Nov 01 13:42:13 well... Nov 01 13:42:13 jkridner: don't confuse ubuntu with linux Nov 01 13:42:21 I do, but I haven't found the link to Chrome yet. Nov 01 13:42:38 why not use a real distro, like gentoo? Nov 01 13:42:45 a source based one? Nov 01 13:42:56 I started the bootstrap process, but something failed in my kernel build... Nov 01 13:42:57 * koen sells av500 some more pixy dust Nov 01 13:43:02 yummy Nov 01 13:43:06 I'll have to get back to it. Nov 01 13:43:07 * wmat looks for the lid to the can of worms Nov 01 13:43:53 * jkridner found /opt/google/chrome Nov 01 13:44:13 its a trap! Nov 01 13:45:32 * panto dreads to contemplate the s**tstorm when he (eventually) posts his scheduler patches Nov 01 13:49:08 hmmm... I don't seem to have the e-mail chain. Last one I had was an RFC. Nov 01 13:49:47 panto: this is different than the RFC you sent about 20 hours ago, right? Nov 01 13:50:15 it's a very small patch series that just moves a few drivers from capebus to arch/arm/mach-omap2 Nov 01 13:50:33 search for "[PATCH 0/3] capebus moving omap_devices to mach-omap2" Nov 01 13:52:35 panto: was linux-omap in copy? Nov 01 13:52:40 yes Nov 01 13:54:03 weird. Nov 01 13:55:18 http://www.spinics.net/lists/linux-omap/msg80990.html isn't in my Inbox. Nov 01 13:55:35 * mdp wakes up Nov 01 13:55:46 wait, yes it is, but that one doesn't have a spam war. Nov 01 13:55:51 er, war. Nov 01 13:56:16 panto, omg, don't bring me into that by mentioning the 6502 cape in the thread ;) Nov 01 13:56:24 too late Nov 01 13:56:41 hmmm.... and that is still a different subject than you mentioned. Nov 01 13:57:07 everyone get's flamb?'d today Nov 01 13:57:33 jkridner, hang on Nov 01 13:57:51 http://www.spinics.net/lists/linux-omap/msg80979.html Nov 01 13:58:19 panto, yeah, well, I'm not clarifying what it is in the middle of that storm Nov 01 13:59:13 jkridner, that's it Nov 01 14:01:00 it only started because the __init in omap_build_device is bogus and leads to crashes Nov 01 14:01:03 imagine that Nov 01 14:01:35 koen: nice thread Nov 01 14:01:36 ah... seems that they are finally learning the issues. Nov 01 14:01:44 are they? Nov 01 14:01:50 I like the "beat up your coworkers" bit Nov 01 14:01:55 and the mention of musb Nov 01 14:01:56 well, it is coming to some sort of head. Nov 01 14:01:57 well, patches are welcome for fixing the DT core Nov 01 14:02:20 yeah, that thread really went off the deepend when bitching about totally unrealted driver support Nov 01 14:03:41 I'm sure customers will love to assemble dtsi fragments and change u-boot everytime they plug a new board in Nov 01 14:03:42 now I want a mention of 6502 and musb in the same email Nov 01 14:03:46 lol Nov 01 14:03:48 yes Nov 01 14:03:50 we love that Nov 01 14:03:57 good grief Nov 01 14:04:16 why cant somebody tell the kernel people how products are actually shipped? Nov 01 14:04:18 av500, and the question..."that's a remoteproc...so it uses omap mailbox?!?" Nov 01 14:04:24 * mdp faceplants Nov 01 14:04:33 mdp, imagine how we feel Nov 01 14:04:34 that you NEVER ship a SW update that relies on something already present on the device Nov 01 14:04:47 mdp: thats balbi for you Nov 01 14:05:06 we always ship ALL relevant SW together Nov 01 14:05:20 so we can flash on HW version and test it Nov 01 14:05:30 av500, perfectionists Nov 01 14:05:35 yeah Nov 01 14:05:47 av500, or pragmatists Nov 01 14:05:52 yep Nov 01 14:06:05 our current g9 ships in ~20 different HW configs Nov 01 14:06:15 * bradfa hasn't been paying attention to linux-omap ml, now has some reading to enjoy! :) Nov 01 14:06:25 bradfa: we aim to please! Nov 01 14:06:47 koen, it's good timing, I just finished a book and need something to read Nov 01 14:06:51 panto, in all of that thread, I did repeatedly see an opening to fix it in the kernel...just not with your exact implementation Nov 01 14:07:11 first rule of mainline club is to not get married to your draft implementation Nov 01 14:07:23 mdp: I don't think we're married to capebus, but we do want a solution in the kernel Nov 01 14:07:41 well, then go back and read what was suggested Nov 01 14:07:48 not the bootloader one Nov 01 14:07:55 koen: yes Nov 01 14:07:58 in the kernel Nov 01 14:08:10 ha wait Nov 01 14:08:14 propose a microkernel Nov 01 14:08:15 mdp, fixing DT is going to be a magnitude more difficult Nov 01 14:08:21 and a cape module :) Nov 01 14:09:05 and still you won't avoid the core issue of all the devices that you want to created be part of 'something' Nov 01 14:09:10 be it cape, or something else Nov 01 14:09:44 its a bit like compound usb devices Nov 01 14:09:50 just inside the SoC Nov 01 14:09:58 not exactly Nov 01 14:09:59 panto, then you need to explain why addressing it in the DT core support/design won't work Nov 01 14:10:31 I didn't say it won't work Nov 01 14:10:38 panto, for one, I think you should have never had support in there for !DT devices...it just clouds issue..so there you have people picking at something that's unimportant Nov 01 14:10:53 mdp, I understand that now Nov 01 14:10:58 panto, you just said fixing DT won't solve the compound device problem Nov 01 14:11:05 wait Nov 01 14:11:11 let's take from the top Nov 01 14:11:20 there are a multitude of issues Nov 01 14:11:40 yep, and you had 1-2 acks for some of them Nov 01 14:11:52 some of the supporting patches you sent Nov 01 14:12:08 yes, those went in just fine Nov 01 14:12:20 ok, what is the issue with the capes Nov 01 14:12:25 let's drop the bus monicker Nov 01 14:12:37 it's different things to different people Nov 01 14:12:48 for one...I think name choices have more importance that even the reviewers will admit Nov 01 14:13:14 for most, the cape is just a way to connect external peripherals to the SoC interfaces Nov 01 14:13:48 in that case, which is the generic cape case, there's no need for a driver, only a specially crafted cape node Nov 01 14:13:58 two...there's subtle ton coming through on several of the responses....I noticed "SDK" mentioned a lot...it shows a certain mindset behind it Nov 01 14:14:16 yeah, what is SDK? Nov 01 14:14:19 panto, so it's like asoc Nov 01 14:14:24 just "userland"? Nov 01 14:14:26 I'm a bit confused. Is the da8xx_lcdc the lcd controller on Bone's SoC? Nov 01 14:14:31 tomba, yes Nov 01 14:14:44 av500, I *think* he's referring to our TI blessed SDK...thinking that's what's driving this or something Nov 01 14:14:59 panto: then I don't quite understand why it needs to be added late. It's always present on a Bone, right? Nov 01 14:15:08 tomba: but not used Nov 01 14:15:19 the same pins could be PWMs Nov 01 14:15:20 tomba: pinmux resources Nov 01 14:15:22 or GPIOs Nov 01 14:15:28 or PRU pins Nov 01 14:15:30 tomba, different configuration for different panels Nov 01 14:15:34 panto: hrm, right Nov 01 14:15:39 tomba, the pinctrl framework allows them to be muxed at driver probe time...or later Nov 01 14:15:48 you can't know until you find out what kind of cape you're dealing with Nov 01 14:16:02 panto: well, different panels should be handled later. the LCDC is still there, and it's the same. Nov 01 14:16:16 the driver is not built that way Nov 01 14:16:25 you fill up, and tell it to go Nov 01 14:16:48 getting the configuration wrong, there's no way to change it later Nov 01 14:16:53 yes, I guessed that much =). the driver is not good for the job, then. Nov 01 14:17:03 panto, so I guess what I'm thinking is you might want to take the approach of providing an change to the device model and DT core that provides a framework where some of the stuff being done in asoc can be moved to a generic layer Nov 01 14:17:35 then instead of having the omap guys bitching I'll get every arch out there Nov 01 14:17:53 almost, but I get the ppc, arm people bitching at least Nov 01 14:18:12 OMAP DSS doesn't work quite right either, but I'm planning to fix it so that the DSS itself can always be loaded, and the panel devices can be added later, via hotplug if such is available. Nov 01 14:18:26 there's parallels with v4l changes that are in progress to better handle compound devices... Nov 01 14:18:30 bone needs a DSS driver :) Nov 01 14:19:12 av500: the common panel/display framework should help there. whenever that happens. Nov 01 14:19:16 mdp, I'm not out to duplicate things that are already out there Nov 01 14:19:25 I didn't say duplicate Nov 01 14:19:34 anyway, I could handle being offered anything like you say Nov 01 14:19:40 handle it in a generic way Nov 01 14:19:51 a VM inside the kernel? Nov 01 14:19:55 most generic :) Nov 01 14:19:55 but not understanding the real problems we're trying to solve it's another Nov 01 14:20:04 panto: yep Nov 01 14:20:49 av500, what if I told you you have to use one of the methods proposed to handle your add-on boards for your design? Nov 01 14:21:05 uhm? Nov 01 14:21:21 well, no Nov 01 14:21:37 oh come on Nov 01 14:21:46 as I said before, a boot loader that mangles a DT will be part of what we ship Nov 01 14:21:46 don't you have a USB port? Nov 01 14:22:11 av500, that's been standard practice in powerpc land from the beginning according to DT Nov 01 14:22:13 so we would have a boot1 in flash and a boot2 aka DT mangler aka prekernel as sw update together wit ha kernel Nov 01 14:22:29 so, semantically, nothing changed Nov 01 14:22:39 just that we split knowledge over 2 codebases Nov 01 14:23:26 and, if we made devices like we did in the past, with hotpluggable capes, DT would need to be mangled at runtime Nov 01 14:23:26 jkridner, still around? Nov 01 14:23:37 yeah... Nov 01 14:23:43 gen7 had like 7 different accessories to be plugged Nov 01 14:23:43 still trying to get a handle I can reply to. Nov 01 14:23:55 av500: and gen9 has 9, etc? Nov 01 14:23:56 using LCD, I2C, UART, PWM abnd USB Nov 01 14:23:56 It makes no sense to me that I'm missing only a select # of e-mails from linux-omap Nov 01 14:23:59 so like your cpaes Nov 01 14:24:02 just at runtime Nov 01 14:24:13 * mdp wants gen11 Nov 01 14:24:28 cause it goes to 11? Nov 01 14:25:02 bakers dozen Nov 01 14:25:20 11 is banker's dozen Nov 01 14:25:25 panto, can't be beat Nov 01 14:26:20 today was the first time I didn't see one of those monthly reminder e-mails in a while. Nov 01 14:26:35 jkridner: you are blocked Nov 01 14:26:38 * panto has a knack of kicking the hornets nest every time Nov 01 14:26:41 posted nudity maybe? Nov 01 14:27:04 big lol at the "autodetect" SPI btw Nov 01 14:27:23 for some its autodetect, for other "turn on the death ray" by accident Nov 01 14:27:42 "accident" Nov 01 14:27:59 I was just probing register, your honour Nov 01 14:28:01 av500, yeah SPI devices are not know for handling random accesses all that well Nov 01 14:28:02 +s Nov 01 14:28:38 in fact, the mistake is to offer all these IO pins, it should have only USB, like any good PC Nov 01 14:28:51 or pcie Nov 01 14:28:56 yep Nov 01 14:29:04 or lightning connector Nov 01 14:29:18 or ................. cloud Nov 01 14:29:18 av500, agreed, just regular USB, m'kay? no that enhanced musb, please Nov 01 14:29:26 oooh, cloud Nov 01 14:29:43 thurbad: no, oort cloud Nov 01 14:30:09 panto, perhaps you can fix the SPI "spec" ;) Nov 01 14:30:27 mdp: and probing is so cool for SPI write only devices Nov 01 14:30:41 maybe the onboard camera can detect LEDs going on Nov 01 14:30:51 funny Nov 01 14:31:00 yep, I'm starting a runtime probe implementation for the st7735 lcd controller now! Nov 01 14:31:02 or a fluctuation in the power consumption Nov 01 14:31:12 I had an old boss that asked me of a way to detect if a chip had possibly burned out on the board Nov 01 14:31:33 smell sensor? Nov 01 14:31:33 I told he should put an electronic nose on the board and I could program it to smell the smoke Nov 01 14:31:38 :) Nov 01 14:31:41 :) Nov 01 14:31:48 lol Nov 01 14:31:52 that sounds like a project from my kiddie electronics kit Nov 01 14:31:52 nose cape FTW! Nov 01 14:31:58 "The Electronic Nose" Nov 01 14:32:02 yep Nov 01 14:32:04 or an ozone detector Nov 01 14:32:15 panto: have you never heard of the "lp0 on fire" error message? Nov 01 14:32:19 yep Nov 01 14:32:25 HCF Nov 01 14:35:32 ugh... so aggravating not to be able to hit reply to http://www.spinics.net/lists/linux-omap/msg81034.html Nov 01 14:35:44 I just don't understand why I'm getting some messages and not others. Nov 01 14:47:23 jkridner: >60$? Nov 01 14:51:47 anyone know magic to reply to e-mail you didn't get? Nov 01 14:53:07 jkridner, lkml.org has a feature where you can forward (but it's really a bounce in my terminology) any lkml post to a mail. Nov 01 14:55:24 jkridner: use a proper mailer that lets you write arbitrary headers Nov 01 14:56:26 * mdp imagines mru wearing a mutt t-shirt at an Outlook conference Nov 01 14:57:14 mdp: they would not understand Nov 01 14:57:23 the opposite would work Nov 01 14:57:41 mru: in that case, have of em would wear one :) Nov 01 15:13:41 what does AM33XX_PULL_ENBL mean? if i don't want to use pullups, do i need to change it to AM33XX_PIN_OUTPUT? Nov 01 15:13:52 why cant ti make cool shit like this: https://lh4.googleusercontent.com/-K8J-Jp_aJCY/UJKMyAa-VWI/AAAAAAAAvIc/vHxqR8oNW3Q/s1178/12+-+1 Nov 01 15:14:02 upside down preferably Nov 01 15:14:25 av500: that looks like a toy Nov 01 15:14:58 rly? Nov 01 15:15:31 yeah, think of other things which round.... Frisbees and umm.... Nov 01 15:15:36 yeah, it looks like a simon with only one button Nov 01 15:15:45 and 2 wheels Nov 01 15:15:52 simon with only one button would be elame Nov 01 15:16:04 *lame Nov 01 15:16:17 JViz: do you have the TRM? Nov 01 15:16:27 alan_o: what's TRM? Nov 01 15:16:34 Technical Reference Manual Nov 01 15:16:47 bedtime reading Nov 01 15:16:49 alan_o: yeah Nov 01 15:16:53 JViz: you could use morse code Nov 01 15:17:01 AM335x ARM® CortexTM-A8 Microprocessors Nov 01 15:17:01 (MPUs) Nov 01 15:17:02 Technical Reference Manual Nov 01 15:17:07 Google for that Nov 01 15:17:09 av500: bitbang it! Nov 01 15:17:12 Then once you have it....... Nov 01 15:17:14 alan_o: no, not that manual Nov 01 15:17:25 mru: but I need one long and one short bit Nov 01 15:17:43 JViz: Then look at 9.3.1.50 Nov 01 15:17:44 bang it on the end, makes it shorter Nov 01 15:17:49 bang it on the side, makes it longer Nov 01 15:17:56 alan_o: thank you Nov 01 15:18:11 JViz: That table is how the mux is set up, with pull up/down enable/disable and rx enable Nov 01 15:18:15 and slew Nov 01 15:18:16 mru: bang galore! Nov 01 15:18:27 JViz: anyway, the thing you're looking at corresponds to that register description Nov 01 15:19:43 Went to pay my credit card bill... Due on the 4th (Sunday, whatever sense that makes), bank can't do a transfer that will get there before the 5th.. Bank and CC are both the same. Nov 01 15:19:56 Called on the phone and the automated thing says "you have no balance" Nov 01 15:20:01 what a joke Nov 01 15:20:09 now I get to sit on hold..... ugh. Nov 01 15:20:36 oh, ringing!!! Nov 01 15:22:41 * av500 does not get that US credit card thing Nov 01 15:22:51 its only good to make banks rich Nov 01 15:23:45 av500: no doubt, but try to buy a plane ticket Nov 01 15:23:56 no, I get that Nov 01 15:24:17 its just funny how you cannot make simple bank transfers Nov 01 15:24:57 yet put people on the moon - allegedly Nov 01 15:25:50 oh no.... you're not one of _those_ are you? Nov 01 15:26:33 the shadows looks fake Nov 01 15:33:06 oooh Nov 01 15:33:15 alan cox is weighing in Nov 01 15:35:04 koen: on what? Nov 01 15:35:08 koen: greetings btw Nov 01 15:35:19 koen: ready for some beer in barcelona? Nov 01 15:35:33 my name is not koen, but yes Nov 01 15:35:44 * dm8tbr still has beer at LCE in copenhagen Nov 01 15:36:01 av500: hehe Nov 01 15:36:16 dm8tbr: you bum Nov 01 15:36:34 * dm8tbr listens to a presentation about arm power probe by andy green Nov 01 15:36:47 turned out to be EE101 Nov 01 15:37:02 so, I'll need a beer Nov 01 15:38:15 panto: has, it's goop! Nov 01 15:38:33 mainline all the capes! Nov 01 15:38:45 meh Nov 01 15:39:01 as I said, I have a knack of causing s**tstorms Nov 01 15:39:08 np Nov 01 15:39:14 we all benefit from it Nov 01 15:39:25 if I tried to play shy.... Nov 01 15:39:27 nah... Nov 01 15:40:49 btw, this reminds me of my youth Nov 01 15:40:56 when I was even more obnoxious Nov 01 15:42:20 dm8tbr: hehe Nov 01 15:43:02 dm8tbr: hehe, just FYI, i did the first of power measurement presentations at LC, hehe Nov 01 15:43:16 :) Nov 01 15:43:40 * dm8tbr wonders how many EEs in the room and how many bored Nov 01 15:43:52 dm8tbr: lol Nov 01 15:44:37 dm8tbr: i'm picking up circuit design (admittedly cut/paste style) so am not bored at all. Nov 01 15:45:14 oh. ah. does anyone know where i can get a library part for orcad, for a 96FBGA DDR3 RAM IC? save me some time.... Nov 01 15:45:59 panto, av500 which mailing list are you referring to (I assume you're referring to a ML) Nov 01 15:46:08 lkml Nov 01 15:46:20 talking about capes on lkml? nice! Nov 01 15:47:02 more like flamed to a crisp Nov 01 15:47:12 lkcl: orcad is not open source, we dont support that here Nov 01 15:47:20 dm8tbr: https://plus.google.com/101339419642360856354/posts/NWyzygqCMp8 Nov 01 15:47:27 dm8tbr: well worth the money to purchase Nov 01 15:47:50 av500: *sigh* i know. i've done the kicad part already. but kicad simply isn't up to the job of dealing with the layout of 1600mhz DDR3 RAM. Nov 01 15:47:52 instruments are overrated, we can simulate everything in the cloud Nov 01 15:47:59 is that the measurement book? (I'm on a mobile phone) Nov 01 15:48:08 lkcl: that DRAM is not open source either! Nov 01 15:48:15 prpplague: almost ready for beer Nov 01 15:48:35 dm8tbr: yea Nov 01 15:48:40 koen: dandy Nov 01 15:48:48 koen: "we have much to discuss" Nov 01 15:48:50 prpplague: how much beaconboards do you have left? Nov 01 15:48:58 av500: i can always make lots of money from hardware, doing beagle-like boards and fund free software projects afterwards Nov 01 15:49:05 koen: maybe 10 or so Nov 01 15:49:19 koen: clint have any left? Nov 01 15:49:23 prpplague: can you bring 2 or so to ELC-E for panto? Nov 01 15:49:33 prpplague: I have no idea, I'd have to ask clint Nov 01 15:49:34 panto needs guiding light Nov 01 15:49:35 should be able to Nov 01 15:49:38 a beacon in the dark Nov 01 15:49:43 prpplague: awesome! Nov 01 15:49:49 panto: you at elce? Nov 01 15:49:54 yep Nov 01 15:49:54 koen: they might be rev1 which has a slightly different pin out Nov 01 15:49:59 cool Nov 01 15:49:59 koen: i'll have to check Nov 01 15:51:40 * koen isn't going to open his email client for the next few hours Nov 01 15:52:33 koen: looks like i have 13 left Nov 01 15:52:52 uuh shiny, html5 output from the energy probe. ;-p Nov 01 15:59:50 dm8tbr: BNC to html5? Nov 01 16:00:26 is there a good guide for setting up a sysfs for a device Nov 01 16:00:53 koen: i already promised a few of the beacons to peter huwe, but i should be able to spare a few for panto Nov 01 16:01:24 panto: what are you going to be doing with them? just tinkering? Nov 01 16:02:31 fester: see other device and copy :) Nov 01 16:02:42 prpplague, annoying the maintainers :) Nov 01 16:03:09 av500: html5 to gardena Nov 01 16:03:14 panto: hehe, dandy Nov 01 16:04:22 the LED on the beacon need an SELinux policy Nov 01 16:04:27 hehe Nov 01 16:04:27 it can really hurt the eye Nov 01 16:04:41 panto / koen beaglebone version of the beacon for next year? Nov 01 16:05:02 prpplague: there is a contest going on Nov 01 16:05:10 propose it Nov 01 16:05:13 av500: oh? Nov 01 16:05:14 * panto submits his nose cape design Nov 01 16:05:28 * av500 makes a taste cape Nov 01 16:05:55 av500: url? Nov 01 16:06:04 prpplague: all over teh interwebs Nov 01 16:06:07 read your g+ Nov 01 16:06:14 av500: hmm, didn't see it Nov 01 16:06:16 not just the #bacon Nov 01 16:06:17 * prpplague googles Nov 01 16:06:23 see boris' posts Nov 01 16:06:32 or that krinder dude Nov 01 16:06:37 hello all, i'm trying to use the spi frame buffer on a beaglebone to connect to an lcd. i've followed matt porter's work exactly but i don't see any action on spi1 and i don't have a /dev/fb0... does anyone have any experience with lcd over spi? Nov 01 16:06:47 sec Nov 01 16:06:58 http://26-26-54.hardwarebug.org/131 Nov 01 16:06:59 av500: hehe Nov 01 16:07:01 av500: found it Nov 01 16:07:03 mdp: ^^^^^^ Nov 01 16:07:18 hehe Nov 01 16:07:41 av500: i don't think jkridner was too impressed with the usefulness of the beacon Nov 01 16:07:53 av500: so i doubt it would be too interesting for their contest, hehe Nov 01 16:08:13 does anyone know how to get the kernel to create the /dev/fb0 framebuffer? Nov 01 16:10:10 prpplague: I found the LED very usefull Nov 01 16:10:38 * prpplague uses his beacon board for all kinds of debugging Nov 01 16:10:44 rabryn: you followed mdp's exact steps? Nov 01 16:10:49 * av500 uses his to show CPU load Nov 01 16:11:00 rabryn: which kernel? Nov 01 16:11:19 3.2.28 Nov 01 16:12:07 so the one that comes with it? Nov 01 16:12:21 is spi1 set up in your board file? Nov 01 16:12:27 send link to mdp's steps Nov 01 16:12:31 i believe i followed them exactly, initially i simply worked with the spi1 port and was able to send/receive data via loopback. but i'm not sure what's supposed to happend now... Nov 01 16:13:00 i verified the st7735fb support is enabled in the kernel Nov 01 16:13:09 (integrated, not module) Nov 01 16:13:13 ok Nov 01 16:13:24 is st77xxx turned on in the board file? Nov 01 16:13:48 arch/arm/mach-omap2/board-am33xx-evm.c Nov 01 16:13:49 rabryn, it's entirely possible that you have the "other display" which was not supported on the version of the driver documented on my blog Nov 01 16:13:59 I have not updated thing to refer to the latest yet...but... Nov 01 16:14:22 rabryn: if you go get the github.com/beagleboard/kernel Nov 01 16:14:59 beaglebone-3.6 or 3.7 branch has the latest st7735fb driver that supports both green and red tab versions of the display...plus a big pile of bug fixes you'll want Nov 01 16:15:13 in board-am335xevm.c i modified the am335x_spi1_slave_info to use adafruit_tft18... was there something else? Nov 01 16:15:49 rabryn, I don't recommend trying that tree. Nov 01 16:15:56 do you know if you have a green or red tab? Nov 01 16:16:17 it's the color of the protective film that comes on the st7735 breakout board Nov 01 16:16:20 that's the only way to tell Nov 01 16:16:38 understood, is the beaglebone-3.6 or 3.7 trees you suggest "super experimental" or can they be used for production? Nov 01 16:17:00 rabryn, I use beaglebone-3.6 for production, I cannot comment on 3.7 Nov 01 16:17:08 actually, i'm not using the st7735, i'm using a formike display that employs the same interface (3 wire spi). Nov 01 16:17:33 mdp, are you matt porter? Nov 01 16:17:34 av500: i'm having trouble finding the code for one, do you have a link? Nov 01 16:17:44 rabryn, I am today! Nov 01 16:17:50 LOL! NICE! Nov 01 16:18:05 rabryn: I'm sorry, did you just say you aren't using st7735 but are trying to use my driver? Nov 01 16:18:21 fester: linux kernel source code Nov 01 16:18:21 what luck... i tried watching your video at the embedded linux conference but it cuts out at 17 minutes. :-( Nov 01 16:18:31 mdp, prove it! say something that only matt porter would know... Nov 01 16:18:46 rabryn: 17minutes is all we can stand Nov 01 16:18:57 that is correct, i'm actually trying to use a FORMIKE KWH014TG01-F01. Nov 01 16:19:18 panto, I know the KISS song that played at your wedding Nov 01 16:19:44 geese Nov 01 16:19:46 av500, +1 Nov 01 16:19:50 :) Nov 01 16:19:51 he's the right matt porter people Nov 01 16:19:55 ((applause)) Nov 01 16:19:55 mdp: all the other guests would know that too Nov 01 16:19:56 ok Nov 01 16:20:02 calling off the helicopter Nov 01 16:20:06 av500: like for example the pwm sysfs interface Nov 01 16:20:14 mru, hrm Nov 01 16:20:20 I can't prove I'm me Nov 01 16:20:29 maybe there was only 1 guest Nov 01 16:20:35 menage a trois... Nov 01 16:20:44 rabryn: you have to write a driver for your part Nov 01 16:20:47 dont go into details Nov 01 16:21:04 my exact steps require using an adafruit lcd 1.8" tft with the st7735 controller Nov 01 16:21:12 but you could use the driver as a starting point Nov 01 16:21:17 mdp, since you're here... when should i expect to see the kernel try to speak with the lcd via spi? i've have a logic analyzer hooked up to spi1 but i don't see any movement on the pins... when would that normally happen (if i'm using the 3.6 kernel for instance) Nov 01 16:21:23 rabryn: lucky you, you get to make a driver! Nov 01 16:21:25 * alan_o is jealous Nov 01 16:21:49 rabryn, at driver probe time Nov 01 16:21:51 .... I learnt Bazaar (not hard once you know Subversion).... Nov 01 16:21:59 alan_o, i'm a little scared... Nov 01 16:22:30 panto, so we can at least isolate me to being one of the guests at your wedding Nov 01 16:22:30 rabryn, don't be... it was a bit hard for us all at first Nov 01 16:22:35 but then we started to like it Nov 01 16:22:42 rabryn: don't be scared. Start with the other one, like mdp said. Nov 01 16:22:46 ok, so that fact that i'm not seeing anything happening now is probably because of an issue with this particular kernel? Nov 01 16:23:27 mdp, yes, but now everyone knows that a KISS song played at my wedding Nov 01 16:23:36 rabryn, in general you don't need to be scared, you can't really cause any bodily injury or health problems if you make a mistake here. and it's even unlikely to cause h/w damage unless you over current the pins Nov 01 16:23:39 I was hoping this could be our little secret :( Nov 01 16:24:11 panto, it's still a hot topic here..."Greeks love KISS" ;) Nov 01 16:24:35 mdp, thanks, it's the banging of my head on the wall that might cause the injury... i suppose that's why i'm scared. :-) Nov 01 16:24:57 rabryn: I wear a helmet ;) Nov 01 16:25:28 nice Nov 01 16:25:36 mdp even knows that you can short output gpio pins directly to ground and be ok :) Nov 01 16:25:48 soft ground Nov 01 16:25:53 rabryn, when you're banging your head a bit, you're almost there... just swap the order and you're bitbanging Nov 01 16:25:56 alan_o, don't encourage people!!! Nov 01 16:26:02 av500: frozen tundra! Nov 01 16:26:20 alan_o, you'll laugh at this...I'm sitting here with a label maker, labeling my nest of wires to dismantle it Nov 01 16:26:21 rabryn: in any case, wrap yourself in antostatic bags Nov 01 16:26:25 leave eye slits Nov 01 16:26:29 and a breathing tube Nov 01 16:26:32 you guys are so much help! Nov 01 16:26:40 we are Nov 01 16:26:46 alan_o, in hopes that it can be reassembled and work in yurup Nov 01 16:27:10 mdp: oh man.. I go nuts doing that stuff, I have to meticulously check every wire a hundred times Nov 01 16:27:28 av500: i found this http://www.mjmwired.net/kernel/Documentation/hwmon/sysfs-interface but i'm not seeing where a device is initialized in sysfs or where the driver updates when a file is updated in the fs Nov 01 16:27:38 I was replacing a couple resistors....each time I moved something...regression test! Nov 01 16:27:52 mdp: yeah, that's where I'm at. Nov 01 16:28:24 alan_o, I have my address/data buses to the bone on some attached ribbon cable which makes it easy to reassemble Nov 01 16:28:34 oh, that's not bad Nov 01 16:28:34 rabryn: dont be scares, we are from the government and we are here to help Nov 01 16:28:47 I'm going to carry the breadboard separate from the bone in the protocape thing Nov 01 16:29:02 I like that protocape board now Nov 01 16:29:08 you should get one of those attached breadboards from Adafruit. I was all "I don't need extra stuff" but I can see how it would help for transport of proto sttuff Nov 01 16:29:32 mdp: the breadboard protcape? Nov 01 16:29:36 alan_o, except that one of those wouldn't begin to hold all these parts Nov 01 16:29:43 sorry, not protocape Nov 01 16:29:48 it's the cape breakout board from CCO Nov 01 16:30:05 yeah, breadboards are never big enough. Nov 01 16:30:05 hmm, when I plug stuff into the breadcape, does DT update itself? Nov 01 16:30:16 av500: hah Nov 01 16:30:19 does there happen to be a yocto recipe for the beaglebone kernel at github by chance? Nov 01 16:30:21 av500, it only starts your breadmaker Nov 01 16:30:44 koen might know, he's a yocto developer Nov 01 16:31:05 mdp: I got the one from Adafruit. I like the CCO ok but it was too much money. (Im' so cheap) Nov 01 16:31:19 I need some help rotating my screen on my phone(which is very similiar to beagleboard and use same cpu). I found that echo "1" > /sys/class/..../rotate should rotate my screen. But I get a compliang about "Invalid configuration". and something about overlay 0 horizontally not in display area. Nov 01 16:31:33 rabryn: that's right, koen maintains a yocto-approved project. Nov 01 16:31:34 so I need ot change teh reoslution I think, but how? Nov 01 16:31:39 * alan_o gets out his stamp of approval Nov 01 16:32:05 alan_o, this guy: http://beagleboardtoys.com/wiki/index.php?title=BeagleBone_Breakout Nov 01 16:32:32 alan_o, I actually needed it for day job work anyway Nov 01 16:32:44 oh, that one Nov 01 16:33:30 it made it easy to swap in a different bone to my interface too when I thought I blew one up Nov 01 16:33:48 yeah, that's a pretty good use case Nov 01 16:33:56 oh, crap...I forgot the sysboot pin stuff in my lessons learned..duh Nov 01 16:34:03 have to send craig an update Nov 01 16:34:08 lessons learned..... I should have one of those Nov 01 16:34:23 1. never fight a land war in Asia Nov 01 16:34:24 alan_o, I have 3 slides of that stuff already Nov 01 16:34:42 alan_o, um, hey, did you get a copy of my slides? that's the first one Nov 01 16:34:49 2. Never engage in a battle of wits with a Sicilian when death is on the line Nov 01 16:35:09 this is all inconceivable Nov 01 16:35:09 is it really? Nov 01 16:35:13 hahahahahah Nov 01 16:35:24 no, I don't have your slidew Nov 01 16:35:30 slides Nov 01 16:35:44 snooping my interwebs I'm sure Nov 01 16:35:58 yeah.. this wireless stuff isn't just for play Nov 01 16:36:15 Artox: what phone? Nov 01 16:36:23 and running what on it? Nov 01 16:36:29 gta04 Nov 01 16:36:33 running opensuse Nov 01 16:36:34 ah Nov 01 16:36:36 oh Nov 01 16:36:43 linux-3.4 Nov 01 16:36:46 uh Nov 01 16:36:55 Fireworks! Nov 01 16:36:58 eh Nov 01 16:38:52 Artox: GT04 is actually shipping? Nov 01 16:39:04 yes Nov 01 16:39:10 mine arrived 3 weeks ago or so Nov 01 16:39:37 and ofr osme testing I'Äd like a rotated screen and dont know how Nov 01 16:41:11 . Nov 01 16:49:36 Artox: screen roration on omap3 is not for the faint of heart Nov 01 16:49:43 there are some constraints in DSS Nov 01 16:49:59 as for pixel clocks etc Nov 01 17:26:23 hello! I have a doubt. Googling hasn't helped me understand the difference between the boot process with u-boot.bin and u-boot.img on a Beagleboard xM rev C for Angstrom dist. Can somebody help me understand the diffference? Nov 01 17:27:03 my board does not boot with u-boot.img but boots up with u-boot.bin Nov 01 17:27:17 and are they different? Nov 01 17:27:41 I don't know, but the device does not boot Nov 01 17:28:00 it stays at booting the kernel message Nov 01 17:28:02 .img is a u-boot image Nov 01 17:28:05 has a u-boot header on it Nov 01 17:28:11 .bin is raw Nov 01 17:28:20 so uboot can boot uboot Nov 01 17:28:21 newer x-loaders (MLO) want .img Nov 01 17:28:33 ah Nov 01 17:29:06 so what should I change to get it to boot? Any additional stuff requires to be done? Nov 01 17:29:24 but you have a booting one Nov 01 17:29:33 yah Nov 01 17:29:35 yeah Nov 01 17:29:38 use the booting one :) Nov 01 17:29:42 * alan_o wipes hands Nov 01 17:29:51 ok :) Nov 01 17:30:59 it is pretty old version though and the MAC address for ethernet over USB keeps on changing after every boot on the booting version Nov 01 17:31:22 so configuring internet everytime is annoying Nov 01 17:31:50 There's a kernel config command line option for setting the mac address Nov 01 17:32:03 than update to a version where you can set the mac adresse via cmdline or module option Nov 01 17:32:20 alan_o not with older smsc drivers Nov 01 17:33:07 oh Nov 01 17:33:44 is there a command for update (like ubuntu apt-get upgrade)? Nov 01 17:33:56 opkg update Nov 01 17:34:01 opkg upgrade Nov 01 17:34:02 no Nov 01 17:34:05 ok great! Nov 01 17:34:10 avoid opkg upgrade Nov 01 17:34:14 might not update your kernel Nov 01 17:34:21 ok sure! Nov 01 17:34:21 just use a newer image Nov 01 17:34:32 it's more likely to break your system than fix anything Nov 01 17:36:26 thurbad but koen fights strongly for clean updates Nov 01 17:37:18 I have another doubt....I am seeing the uEnv.txt file. It has some stuff regarding dvimode ...should I comment it since I do not have a monitor connected to my beagleboard? Nov 01 17:37:21 I'm not denying that Nov 01 17:37:38 the file is here: http://elinux.org/BeagleBoardUbuntu#RootStock:_Build_an_Ubuntu_root_file_system Nov 01 17:37:39 help-seeker dont matters Nov 01 17:37:50 ok Nov 01 17:57:43 does anyone have working a spi? Nov 01 17:57:54 old kernel or new? Nov 01 17:57:59 new Nov 01 17:58:01 yes Nov 01 17:58:06 next question Nov 01 17:58:10 :) Nov 01 17:58:12 did the wiring requirements change from the old to the new? Nov 01 17:58:16 no Nov 01 17:58:25 but the DT stuff has to be set up right in the new Nov 01 17:58:34 DT stuff? Nov 01 17:58:38 devicetree Nov 01 17:58:41 define new kernel? Nov 01 17:58:51 alan_o: ^^^^ Nov 01 17:58:51 the holy grale Nov 01 17:58:53 version, where you got it from, etc... Nov 01 17:59:30 Linux beaglebone 3.2.23 #1 Wed Oct 17 12:20:27 EDT 2012 armv7l GNU/Linux Nov 01 18:00:07 whoa..... Nov 01 18:00:12 :) Nov 01 18:00:16 you're living in the past! Nov 01 18:00:21 old kernel.. so forget my talk about DT Nov 01 18:00:23 I got an LCD7 and had to cut one of the traces Nov 01 18:00:34 so you cut your traces and SPI no worky anymore Nov 01 18:00:42 did you cut SPI1_sclk? :) Nov 01 18:00:43 no, it wasn't working to begin with Nov 01 18:00:54 i cut a trace to one of the buttons on the lcd Nov 01 18:00:54 koen / panto beacon boards are packed and ready Nov 01 18:01:01 * panto bows Nov 01 18:01:05 thanks Nov 01 18:01:27 panto: i've allocated two for you Nov 01 18:01:30 panto: that enough? Nov 01 18:01:32 the lcd7 apparently uses the pins for both spi0 and spi1 Nov 01 18:01:39 sure Nov 01 18:01:43 panto: dandy Nov 01 18:02:00 panto: feedback on a version for bone would be appreciated Nov 01 18:02:12 don't sweat it Nov 01 18:02:26 i only saw one pin being use for it on spi1, and it was a button, so... Nov 01 18:03:03 then commented out the line that disabled the spi1 pinmux in the kernel, and rebuilt it Nov 01 18:03:34 but i seem to only get garbage over spi1 Nov 01 18:03:45 it does communicate at the right time, but the data is all wrong Nov 01 18:04:50 Someone is spending money to advertise. I have a beaglebone capes add covering the bottom of an emailed add from Mouser. Nov 01 18:05:08 and a cape design contest Nov 01 18:05:13 djlewis: email ads are cheap Nov 01 18:05:26 if you see it during superbowl though..... Nov 01 18:05:41 av500 haha Nov 01 18:06:41 superbowl ads are ego strokers Nov 01 18:06:41 why do you say I have an old kernel? Nov 01 18:06:58 because the new kernel is 3.7 Nov 01 18:07:40 with DT Nov 01 18:07:58 and ponies Nov 01 18:08:02 janet jackson at superbowl was the best anyway Nov 01 18:08:48 I wonder if ponies and boobs have DT bindings Nov 01 18:09:18 that must be a very new kernel Nov 01 18:09:31 yes why not Nov 01 18:10:16 compatible = "hasbro,my-little-pony"; Nov 01 18:11:20 lol Nov 01 18:17:30 gm sakoman Nov 01 18:23:46 panto: you know there is an HR bot right? :) Nov 01 18:24:13 * mranostay waves to BeagleBot Nov 01 18:25:03 I'm sure I can argue that the only person I can sexually harass in the office is my wife... if she lets me :) Nov 01 18:25:14 lol Nov 01 18:27:27 I loved the discussion at the start of the contract with the HR person Nov 01 18:28:05 do tell Nov 01 18:28:10 you sexually harassed the HR Nov 01 18:28:19 naughty boy Nov 01 18:28:20 "no, I'm pretty sure there won't be a need for training about harassment in the workplace" Nov 01 18:28:39 till later Nov 01 18:28:45 "yes, it's just me... in an office... alone... forever alone" Nov 01 18:28:47 i have yet to take a sexual harrassment class Nov 01 18:28:57 cause you'll fail Nov 01 18:29:05 is that standard in th US now? Nov 01 18:29:12 panto: the name really is misleading Nov 01 18:30:46 come on Nov 01 18:30:48 admit it Nov 01 18:30:55 you'd love to live in the mad men era Nov 01 18:31:21 minus the suits Nov 01 18:31:36 no suit, no hot secretary Nov 01 18:33:04 i would like to be able to drink at work as well :) Nov 01 19:02:15 simple mmu question. Are page tables absolutely neccesary to utilize mmu and cache Nov 01 19:05:16 so a flat section translation(4096 entries of 1MB each) would also require additional 4096x256 pages Nov 01 19:05:34 if it was all ram and all 4k pages Nov 01 19:08:11 cortex Nov 01 19:08:24 that failed.. Nov 01 19:13:17 so i'm trying to build kernel 3.6 from the beaglebone-3.6 branch at github with a compiler i built with yocto (gcc is 4.6.4) and i get this error... "selected processor does not support requested special purpose register -- `mrs r2,cpsr'" anyone have a clue? Nov 01 19:14:04 i used the beaglebone config out of the "configs" directory... Nov 01 19:16:59 after some googling i found a patch that changed '-Wa,-march=all' to '-Wa,-march=armv7-a' in arch/arm/boot/compressed/Makefile, i made that change and the kernel built but it wouldn't boot... Nov 01 19:17:24 panto: can we really have everything enabled in a .dts? Nov 01 19:17:28 rabryn yes you proably need gcc-4.5 for compiling Nov 01 19:17:34 ...per Tony's assertion. Nov 01 19:18:04 tony said disabled Nov 01 19:18:04 jkrinder which toolchain now builds a bootable kernel? Nov 01 19:21:55 woglinde, where do you get the compiler you're using? do you build it by hand? Nov 01 19:24:04 woglinde: I'm told: http://www.beagleboard.org/irclogs/index.php?date=2012-10-30#T13:10:32 but I haven't confirmed yet. Nov 01 19:24:41 mdp: sorry, the word "enabled" is overloaded. Nov 01 19:24:52 ok Nov 01 19:24:55 mdp: seems like every word you try to use has multiple meanings. :( Nov 01 19:25:27 I mean, can it be done by listing everything in a .dts. It seems to me it cannot and Tony goes on about adding the ability to dynamically alter the device tree. Nov 01 19:25:30 we already have every cape in the dts regardless Nov 01 19:25:42 mdp: but, that doesn't scale... Nov 01 19:25:56 jkridner, that's not even part of the discussion right now Nov 01 19:25:58 we know that Nov 01 19:25:59 we want people to be able to do capes independent of the image releases. Nov 01 19:26:19 do you want tony to change dtc compile linker and define a new object format for it all? Nov 01 19:26:28 work on one battle Nov 01 19:26:45 I can't figure out what we are arguing about. Nov 01 19:27:20 tony suggested a way to handle this in the kernel DT core code *without* calling private omap hwmod calls directly from drivers/* like capebus does Nov 01 19:27:34 that's what everybody has suggested as one alternative from the beginning Nov 01 19:28:03 that may have been lost in the noise of tryin to educate people on the basic use case and why doing it in the bootloader is undesirable Nov 01 19:28:17 rabryn -> 20:24 < jkridner> woglinde: I'm told: http://www.beagleboard.org/irclogs/index.php?date=2012-10-30#T13:10:32 but I haven't confirmed yet. Nov 01 19:28:44 woglinde, got it, thanks. Nov 01 19:39:02 mdp: so, you think panto should take an approach that does not call the private omap hwmod calls from drivers/* ? Nov 01 19:39:28 of course, it's merely an implementation detail Nov 01 19:39:48 my preference is to take a path that leads to getting in the kernel Nov 01 19:40:53 this is why I talk a lot on our mainline calls about "semantics" and comments I need to incorporate that are just semantics rather than functional changes Nov 01 19:40:54 <_av500_> would a trained boot-loader-writing monkey help here? Nov 01 19:41:03 _av500_, damn skippy it would Nov 01 19:41:06 I'm confused still, but you are making it sound like there is a clear task that enables getting capebus implemented that is not excessive work for panto. is that true? Nov 01 19:41:50 jkrider, this is just one comment, but yes, he's been offered an acceptable path...at least to the omap maintainer Nov 01 19:42:38 jkridner, it's very predictable when something in drivers/* calls directly to an arch/mach-omap2/* call these days...that's simply not allowed Nov 01 19:42:48 nobody is going to tell you great job if you do that Nov 01 19:43:23 that's why numerous people said fix it in the DT core...only tony offered some specifics as to what he thinks would be good Nov 01 19:43:38 I'll leave it to you guys to figure out Nov 01 19:43:53 my project works great on beaglebone-3.6 ;) Nov 01 19:43:56 k, i'll wait for panto/koen to tell me about the challenges. Nov 01 19:49:06 jkridner: it's time. if you want to work in mainline it requires time to do rewrites and incorporate changes...consider other ideas Nov 01 19:49:31 working code is the first 1% of the task at least Nov 01 19:49:32 yes, but how much time. Nov 01 19:49:44 generally 10x what you expect Nov 01 19:50:02 I have 0 feel for how much effort is involved in the requested change. Nov 01 19:52:35 jkridner, well, considering that it involves convincing the DT core maintainers on dt-discuss, figure a significant amount of time to arrive at a consensus Nov 01 19:53:12 it's never as sexy as making stuff work Nov 01 19:53:56 yeah, once it works, I rarely care anymore :) Nov 01 19:54:06 heh Nov 01 19:54:18 mranostay: "close enough" Nov 01 19:55:23 jkridner, recommended reading...today's 30 kernel developers in 30 weeks with steven rostedt Nov 01 19:55:30 read his advice to developers Nov 01 19:56:01 I like him Nov 01 19:56:12 s/today's/this week's/ Nov 01 19:56:15 http://goo.gl/5tkWi Nov 01 19:56:53 alan_o, I also appreciate his comments on why he lives where he lives ;) Nov 01 19:57:20 I haven't read it. I'll check it out Nov 01 19:57:25 once I get back from the garage Nov 01 19:57:32 * alan_o goes to find a drill and an awl Nov 01 19:57:42 alan_o, he's a good guy Nov 01 19:58:25 I don't see anything relevant in https://www.linux.com/news/special-feature/linux-developers/658135-30-linux-kernel-developers-in-30-weeks-steven-rostedt Nov 01 19:58:30 I mean, he gave us ktest.pl, which is enough accomplishment for a lifetime..except for the perl-ness Nov 01 19:58:39 lol, are you serious Nov 01 19:59:24 "One thing I've learned while working on the -rt patch is that you can not push things into the kernel. They need to be pulled in. Don't come around saying that you have the best new feature and everyone should just throw flower petals at your heels. You need to convince the current maintainers that what you have will benefit them." Nov 01 20:00:10 sure... I think it has to be seen by the upper maintainers what our issue is and have them guide us down the path of producing a viable solution. Nov 01 20:01:43 that's why I introduce a lot of am335x improvements as da8xx cleanups first ;) Nov 01 20:02:15 everybody likes having existing drivers poor code cleaned/fixed Nov 01 20:02:47 your use case is sound, just need to work the implementation details Nov 01 20:03:18 jkridner, you've got to prepare yourself for the need to spend serious devel time upstream on the dtc tools themselves though Nov 01 20:11:48 hey, we spoke earlier about the spi framebuffer... you mentioned that i'd have to write my own driver so i started taking a look at yours as a template. i put a kprint in the st7735fb_probe function and it's not getting called when i modprobe it. Nov 01 20:12:27 i'm new device driver development, was wondering if you had any suggestions. i think once i can see the spi lines twiddle on the scope i'm home free... Nov 01 20:12:40 at least for a while. Nov 01 20:13:05 3.2 or 3.6? Nov 01 20:13:43 back Nov 01 20:14:05 jkridner, let me scroll back Nov 01 20:14:18 3.2, i tried the 3.6 but it wouldn't compile, at which point i realized my gcc is 4.6.4 and it was recommended i use 4.5... i really don't want to go down that rabbit hole right now. Nov 01 20:15:00 I use a cslite 2010.09 Nov 01 20:15:23 ubuntu-packaged linaro 4.6.x ish worked for me too Nov 01 20:15:25 jkridner, around? Nov 01 20:15:35 yup Nov 01 20:15:42 but I'm using cslite because it exposes some bugs in stuff I'm working on Nov 01 20:16:09 rabryn, on 3.2 you must have the driver compile in or as a module Nov 01 20:16:23 right... i'm using yocto that built my glibc, rfs, etc. not sure any of that matters though Nov 01 20:16:42 then you must have the spi device info with proper modalias configured in board-am335xevm.c Nov 01 20:16:57 ok, so what's the question Nov 01 20:17:11 initially i had it compiled in, then i made a module so i could load/unload it easier. Nov 01 20:17:38 I just build in for devel since it's fast to reboot Nov 01 20:18:02 mdp, i'm sure i do, let me double check. Nov 01 20:18:03 and my kind of bugs generally lock up the whole system Nov 01 20:18:35 rabryn, then you need to make sure the spi1 data is bring registered Nov 01 20:18:45 two things have to happen in the big picture Nov 01 20:19:05 1) driver inits and registers itself with the ids it will match against Nov 01 20:19:07 panto: from me? I'm trying to figure out if you find the current suggestions reasonable and if you'll be able to take any actions. Nov 01 20:19:19 i need to know if we are stuck. Nov 01 20:19:23 2) device gets registered in the board file and advertises the id is provides for matching Nov 01 20:19:35 ok, let me do a wrap-up Nov 01 20:19:56 rabryn: so then the two match in the driver core which calls the probe Nov 01 20:20:06 so one of the two is not registered Nov 01 20:20:07 with the current state of things (ie. without something like capebus) nothing works Nov 01 20:20:16 rabryn: or your id is not matching Nov 01 20:20:21 so check all that Nov 01 20:20:30 panto, "nothing works" ;) Nov 01 20:20:34 mdp, you're referring to the modalias right? i have am335x_spi1_slave_info's modalias set to "adafruit_tft18" Nov 01 20:20:46 mdp, well, from the cape side it doesn't Nov 01 20:20:46 panto, you mean it works statically ;) Nov 01 20:21:01 yeah, if you modify the dts for each cape combination Nov 01 20:21:13 which means it doesn't work for users Nov 01 20:21:24 it's "suboptimal" Nov 01 20:21:41 to the point of said user RMAing the bone & the cape Nov 01 20:22:04 we cannot rely on shipping dts in the code images.... Nov 01 20:22:22 rabryn: check that all that stuff gets called..sprinkle printks in your board file and see what happens with spi1 Nov 01 20:22:29 the relevant information *must* come from the cape EEPROMs. Nov 01 20:22:34 so first goal is to ship a working kernel+distro that can support any capes the user has Nov 01 20:22:37 jkridner, right Nov 01 20:23:14 now, a major sticking point is the use of omap_platform_device after the bone has booted Nov 01 20:23:21 mdp, thanks! will keep plugging away... Nov 01 20:23:36 which still, is something about not having DT bindings for two drivers Nov 01 20:23:47 one of them is having them done, so we're down to one Nov 01 20:24:05 I'm just not sure if that is understood upstream yet or if mdp's comment comprehended that. it isn't "suboptimal", it doesn't work, if that is the statement. Nov 01 20:24:36 tell me how it doesn't work Nov 01 20:24:40 it works statically Nov 01 20:24:46 it only works if you manually edit the dts for a given cape definition Nov 01 20:24:52 which is suboptimal Nov 01 20:24:56 yeah Nov 01 20:25:01 and the EEPROM is not read at all Nov 01 20:25:06 yep Nov 01 20:25:11 but it works Nov 01 20:25:13 or not used in that decision Nov 01 20:25:22 include dvi_audio_cape.dtsi Nov 01 20:25:23 it works for me and you, not for the user Nov 01 20:25:36 include bone_6502_cape.dtsi Nov 01 20:25:46 that's not an end user solution Nov 01 20:25:53 I didn't say it was Nov 01 20:25:56 I said it works Nov 01 20:26:01 and I even said it's suboptimal Nov 01 20:26:11 that's what I said too, that it works for us Nov 01 20:26:21 and going out and telling people in a thread that "nothing works" just bogs down the discussion Nov 01 20:26:22 not for the customer Nov 01 20:26:32 wait Nov 01 20:26:38 so I'm playing devil's advocate here Nov 01 20:26:41 it only works for trivial capes Nov 01 20:26:42 it does work, but it sucks Nov 01 20:26:46 just need to split the issues.... Nov 01 20:26:48 i.e. generic capes Nov 01 20:26:54 it is not sufficient. Nov 01 20:26:59 for complex capes that doesn't work at all Nov 01 20:27:09 "suboptimial" isn't accurate. Nov 01 20:27:11 gogo go speedracer Nov 01 20:27:40 the whole question is whether we need capebus at all Nov 01 20:27:44 panto, another problem...you keep saying trivial versus complex..but not once have I heard an example Nov 01 20:27:58 we'll never be able to have 3rd parties make 100s of capes if every one has to merge their .dtsi into the mainline kernel. Nov 01 20:28:27 complex cape == needs a driver to work Nov 01 20:28:27 should have just put pci-e on the part Nov 01 20:28:50 panto, so by definition, the st7735 cape is a complex cape Nov 01 20:29:00 no, it's not Nov 01 20:29:05 whoa, wait! Nov 01 20:29:11 it *needs* a driver Nov 01 20:29:14 it's simple because it's only a single SPI device Nov 01 20:29:14 ok, try again then Nov 01 20:29:17 ok, rewording Nov 01 20:29:30 needs coordination of multiple drivers to perform the work Nov 01 20:29:30 ok, try my 6502 cape then Nov 01 20:30:13 you 6502 would fall into that category if you were to make a more complex interface that the one offered by the standard platform devices Nov 01 20:30:36 so just 6502 where you're hacking about, not a complex cape Nov 01 20:30:40 ok, multiple drivers is it then! Nov 01 20:30:42 thank you Nov 01 20:30:58 cause this one just uses lots of pins Nov 01 20:31:25 6502 where you are offering a complete solution for driving some kind of machinery Nov 01 20:31:28 complex Nov 01 20:31:30 oh, when I get done with the pruss remoteproc stuff, it'll use two drivers...but actually it will take care of calling them Nov 01 20:31:50 well, that would be done via virtio drivers which instantiate from the firmware resource table anyway Nov 01 20:31:56 which is another discussion Nov 01 20:31:56 right, the way to think about it is like this: Nov 01 20:32:14 simple - something that once you plug it into the Linux device model it just works Nov 01 20:32:17 they don't even have any relation to platform drivers and DT at all Nov 01 20:32:38 wait, is a dvi + audio cape I have sitting here a complex cape? Nov 01 20:32:41 complex - something that requires a coordinating driver presenting a more comprehensive interface for the task that this cape is built for Nov 01 20:32:53 because it needs da8xx-fb and davinci-pcm? Nov 01 20:33:01 nope, simple Nov 01 20:33:05 damn! Nov 01 20:33:23 ok, weathercape? ;) Nov 01 20:33:24 once you register the video/audio device needs no more work Nov 01 20:33:35 simple too (at least how koen is using it) Nov 01 20:34:00 let's take your 6502 cape Nov 01 20:34:01 ok, tell me the one existing cape that represents the complex case Nov 01 20:34:07 so I have something solid to consider Nov 01 20:34:13 assume I'm an idiot, ok? ;) Nov 01 20:34:21 the only one at the moment is the geiger cape Nov 01 20:34:50 because you have a number of peripherals connected in a way that all the details are abstracted Nov 01 20:35:17 the user interface presented back to the user, is a sysfs interface, with 3 files Nov 01 20:35:20 geiger cape doesn't have an open design yet so I can't really understand it Nov 01 20:35:37 it's totally open, only you weren't in the loop when designing it Nov 01 20:35:48 care to check it out? Nov 01 20:35:53 this is the geiger cape driver you are describing? Nov 01 20:35:57 yes Nov 01 20:36:04 hw Nov 01 20:36:07 ok, go on Nov 01 20:36:28 let me find the schematics and sent it over Nov 01 20:36:37 so there's a geiger device described in DT then Nov 01 20:37:27 and so tell me what pin resources / hwmods it uses Nov 01 20:37:31 gpio? pwm? Nov 01 20:37:48 spi, I think, I saw the awesome display Nov 01 20:37:52 leds, pwm, gpio, and a analog vsense input Nov 01 20:38:09 spi, no, that's an override cape Nov 01 20:38:16 ok, sounds like my 6502 cape so far Nov 01 20:38:19 it's not part of the single cape Nov 01 20:38:25 right Nov 01 20:38:51 a driver that uses various resources (gpio/pwm/prupins) Nov 01 20:39:15 so what it does, is use a pwm to boost voltage of the geiger tube to about 500+ V and then count the number of events (interrupts) detected Nov 01 20:39:23 it's not complicated Nov 01 20:39:40 but it *is* complex Nov 01 20:39:40 but you can't do this by just creating platform devices Nov 01 20:39:53 yes, complex by the cape bus definition Nov 01 20:40:28 I'm not sure what your platform device omment is about Nov 01 20:41:09 the simple capes only need to create the devices that are described in their cape definition in the DTS Nov 01 20:41:18 then they just work Nov 01 20:41:42 you don't need to do anything else with them Nov 01 20:41:47 no need to count events Nov 01 20:42:01 * alan_o shakes his fist at radio shack standoffs that are too big Nov 01 20:42:28 panto, using ecap? Nov 01 20:42:30 nor check anything like the analog input vsense for proper operation Nov 01 20:42:39 mdp, not yet, there's no ecap driver Nov 01 20:42:51 counting events via gpio interrupts Nov 01 20:42:54 ok, saw that on the schematic and wondered for that very reason Nov 01 20:43:08 ok, yes, saw the gpios on 08 Nov 01 20:43:28 eventually if ecap driver comes, then, yes, switch to using that, although the event rate is very low for making it matter Nov 01 20:43:42 but I don't understand how this is different from the st7735 cape Nov 01 20:43:58 it requires a driver that then uses gpio/pwms/spi1 Nov 01 20:44:21 the st7735 is implementing a standard linux video driver Nov 01 20:44:39 and b6502 implements a standard remoteproc driver..I'm with you Nov 01 20:44:50 what does geiger cape do? why would it implement a standard driver? Nov 01 20:45:01 it doesn't Nov 01 20:45:10 s/would it/would it not/ Nov 01 20:45:12 that's the point, it's an one-off Nov 01 20:45:25 so your driver s/w is broken :) Nov 01 20:45:35 what? Nov 01 20:45:52 a driver is a driver Nov 01 20:46:03 yes, but it's a driver for something arbitrary Nov 01 20:46:08 no it's not Nov 01 20:46:21 it's a driver specifically for this bone+geiger cape Nov 01 20:46:33 right Nov 01 20:47:08 which doesn't fit in any kind of mold Nov 01 20:47:08 and I believe you said static dtsi does not work for these complex cases Nov 01 20:47:12 which is not true Nov 01 20:47:21 sure it does Nov 01 20:47:35 it fits into drivers/misc just like the sram driver and everything else there Nov 01 20:48:09 well, true Nov 01 20:48:21 I fail to see how a unique peripheral is a special case...beyond requireing a unique driver Nov 01 20:48:38 a single peripheral no Nov 01 20:48:47 that is a peripheral Nov 01 20:49:02 I see your point Nov 01 20:49:12 but it's just semantics Nov 01 20:49:53 whether you put the driver in a box, you will still need to solve the problem that capebus does Nov 01 20:49:55 so back to the static thing...it does work if one distributes a driver patch + .dtsi for a cape Nov 01 20:50:00 the include it into the dts Nov 01 20:50:20 and conflicts will be horrible and whined about by pinctrl if the drivers are well done Nov 01 20:50:36 it sucks and is suboptimal by my definition but it does work Nov 01 20:50:48 it works for developers only Nov 01 20:51:23 well, for one, you've got distribute the drivers Nov 01 20:51:54 so let's not forget that nobody gets away with just shipping a cape the requires a driver and avoids the wrath of the kernel learning curve overhead Nov 01 20:52:13 unless the normal users wait for another angstrom build that includes the driver and grab that image Nov 01 20:52:30 you're simplifying things Nov 01 20:53:07 it's damn near trivia to build a stopgap userspace tool that glues a dtb from a dtsi cape database on the target and you reboot to use that dtb that has your stack of capes working properly Nov 01 20:53:31 and that's even been suggested in past discussion from !me as a stopgap to work around DT's staticisms Nov 01 20:53:56 I don't think it's the long term solution, just pointing out that it's a better user experience Nov 01 20:54:26 it's only a better user experience for the mess that is currently the normal case Nov 01 20:54:51 you don't need anything like that with something (let's not call it) like capebus Nov 01 20:54:59 it's a better experience that works against stock mainline..incremental improvement Nov 01 20:55:11 * woglinde chuckles Nov 01 20:55:18 yep, and that's why I support the concept embodied in capebus Nov 01 20:55:32 it's only a horrible kludge so that a few egos don't get bruised Nov 01 20:56:07 I'm just wondering why nobody is willing to roll out an incremental improvement in angstrom while the ideal solution is being worked on Nov 01 20:56:12 mdp, figured it out... in board-am335xevm.c the spi1 pin mux was being setup then the next line uses bone_spidev2_info in the spi_register_board_info call (not spi1's). Nov 01 20:56:14 so we have a real fix, from a user experience perspective, and we have to compromise, for what exactly Nov 01 20:56:19 koen tried Nov 01 20:56:35 rabryn: nice catch! yeah, printks are your friend ;) Nov 01 20:56:36 it's not going to work just from user-space Nov 01 20:56:58 ask koen about it, he had written up the concept Nov 01 20:57:21 I know, and he chucked it away when using the alternative Nov 01 20:57:22 it can work..it simply running dtc based on user selection Nov 01 20:57:28 instead of probing Nov 01 20:57:32 *sigh* Nov 01 20:57:37 I guess you don't even see the point Nov 01 20:57:43 I see the point Nov 01 20:57:54 it's something that might be accepted more easily by mainline people Nov 01 20:58:16 it's still sub-optimal though Nov 01 20:58:37 well, anyway, for the ideal approach, tony has suggest a path to deal with the one big issue Nov 01 20:58:40 so go forth ;) Nov 01 20:59:04 userspace dtc walk yiikes.. Nov 01 20:59:06 hint: buy rob herring some beers Nov 01 20:59:24 and fun at the same time.. Nov 01 20:59:36 unsolo: yeah, I think he even prototyped it in frustration with the cape situation Nov 01 21:00:05 probe the eeproms in the stack...then it's trivial to construct a dts including the right cape dtsis..run dtc.. Nov 01 21:00:06 mdp, I think you're talking about the DT object file method Nov 01 21:00:11 cp the dtb to the sd card and reboot Nov 01 21:00:31 it's fugly as can be Nov 01 21:00:39 or mess the dtb and fail to boot at all Nov 01 21:00:45 details! Nov 01 21:01:02 it's where the devil lies Nov 01 21:01:30 my point is that I'm not married to capebus Nov 01 21:01:50 panto, we'll spin some KISS records and have a beer next week while talking about this ;) Nov 01 21:02:14 but I do want, is that the user doesn't have to suffer Nov 01 21:02:16 mdp, deal! Nov 01 21:02:33 I live by one motto: Don't break your customer's balls Nov 01 21:02:54 and linux people sure like kicking them in the nuts Nov 01 21:03:18 panto: grumpy old man Nov 01 21:03:30 panto, ouch Nov 01 21:03:38 panto linuxer user are expecting breaking stuff Nov 01 21:03:53 well, we'll do beer assuming newark is running ok again saturday Nov 01 21:03:58 not acceptable IMO Nov 01 21:04:06 it's a failure of the community Nov 01 21:04:13 you perfectionists Nov 01 21:04:18 what did I tell you about customers? Nov 01 21:04:19 ;) Nov 01 21:04:37 hehe Nov 01 21:05:01 I'm not defending them, they sure do their own share of dumb things Nov 01 21:05:05 but we should know better Nov 01 21:05:14 we _do_ know better Nov 01 21:05:31 they are always wrong Nov 01 21:05:54 irrelevant, they still pay for your product/services Nov 01 21:06:31 actually, I once had to defend a customer here from my boss...told him, "look, this is great! they are doing us a favor detailing above our heads how bad our kernel sucks" Nov 01 21:06:51 panto, but I still hated them for pulling string to get me on a plane with 4 hours notice for a week Nov 01 21:07:07 I never said not to hate your customers :) Nov 01 21:07:16 I'll be sure to bring the t-shirt they gave me for my trouble Nov 01 21:07:20 good troll shirt Nov 01 21:07:40 but if we've stopped kicking them in the balls, perhaps they would be in a better mood Nov 01 21:07:56 keep kicking 'em and keep speechless Nov 01 21:07:59 that'll work too Nov 01 21:08:32 ok, let me reverse the question about the way to go Nov 01 21:09:35 panto, you are now into my 6502 hacking time...please don't disturb my beloved accumulator ;) Nov 01 21:11:01 mdp: i don't se why a cape couldnt be "activated" post boot.. Nov 01 21:11:09 panto, would you mind if I add a slide entitled, "Would CapeBus Help Here" Nov 01 21:11:25 unsolo: just current mach-omap2 hwmod assumptions prevent it Nov 01 21:11:42 panto's original patch was controversial, but tony has softened his stance Nov 01 21:11:49 hehe Nov 01 21:12:31 panto, testing my raw virtual uart speed...important stuff Nov 01 21:12:32 to bad there is no hypervisor laying around in the a8 Nov 01 21:14:06 then you could move all this stuff out of linux Nov 01 21:14:27 i suppose one could theoretically run a small manager in the sram or something but still.. Nov 01 21:15:01 would be cool to have a hot pluggable cape to right :) Nov 01 21:16:20 * unsolo is only trolling but still interesting way of seeing things Nov 01 21:39:36 mdp, mranostay: are you guys going to say "you can download all the code for this demo from...." ? Nov 01 21:39:46 on my last slide Nov 01 21:40:00 because I'm not sure I can do that. It's such a mess, and I don't have time to commit it, and it's a lot of pieces. Nov 01 21:40:06 I usually reference links throughout as necessary Nov 01 21:40:07 <_av500_> mdp: i had an idea, could a vendor tree help? Nov 01 21:40:18 now that's just trolling Nov 01 21:40:25 _av500_, only an *evil* vendor tree Nov 01 21:40:49 congrats :) Nov 01 21:41:00 alan_o, consider that people will be looking through your slideware 2 years from now Nov 01 21:41:19 mdp: maybe I'll see what I can put together... The beaglebone part is really a bodge. Maybe I can put up a link and say "I'll put stuff up here later." Nov 01 21:41:42 alan_o, tbh, I mark stuff that's in the state of what mine is in as "WIP" Nov 01 21:41:47 buyer beware Nov 01 21:41:49 half the point of my presentation is that this that 802.15.4 on Linux is currently a bodge Nov 01 21:42:02 not everything that's published has to be medical mission critical to be useful to people Nov 01 21:42:06 people like to learn Nov 01 21:42:13 yeah, true Nov 01 21:42:19 I know I greatly appreciate all the fragments that google turns up for me Nov 01 21:42:33 g'night Nov 01 21:42:44 and I have a lot ugly code in my b6502 related trees Nov 01 21:43:51 I wish it was just ugly code Nov 01 21:43:57 the stuff I polish up on TI time is just so uninteresting that I could never talk about it Nov 01 21:44:05 I would fall asleep in front of everybody ;) Nov 01 21:44:42 it's ugly hacks to the bb kernel and dts, a patchset from another guy for 802.15.4... Then there's two platforms (PC and bone) with two different wireless modules and drivers. Nov 01 21:44:55 yeah, mainlining sucks Nov 01 21:44:59 I know I would find it interesting Nov 01 21:45:07 alan_o, it has ups and downs Nov 01 21:45:11 that's true Nov 01 21:45:21 seeing the email from DaveM in the inbox.... Nov 01 21:45:28 scrolling down.... Nov 01 21:45:30 I'm leveraging a lot of things that I did on TI time Nov 01 21:45:51 like reviewing and patching the generic sram driver that's hopefully going in for 3.8 Nov 01 21:46:14 I needed a feature for this...but it's also needed in davinci land so... Nov 01 21:46:36 alan_o, it's hard to find something that isn't useful in multiple places Nov 01 21:46:43 That's true Nov 01 21:46:45 same with knowledge Nov 01 21:46:47 oh, that's in my slides too Nov 01 21:47:08 I should count the original code in this project Nov 01 21:47:17 it's got to be less than 200 hundred lines or so Nov 01 21:47:50 like right now... since I pulled my hair out over something months ago and then detailed it in an email on the mc1322x mailing list, I can go back and reference it now (since I've long forgotten, but I now need it ) Nov 01 21:48:05 mdp: oh yeah, I don't feel like I do any acutal coding anymre Nov 01 21:48:38 Embedded Linux: 2% art, 5% science, 93% archeology Nov 01 21:49:11 * alan_o picks up a skull.... what the.....? Nov 01 21:49:22 run! Nov 01 21:49:33 embedded isn't what it used to be Nov 01 21:49:46 jacekowski: it still is that way for a _lot_ of people Nov 01 21:49:56 people who use micros Nov 01 21:50:12 you want to bitbang? Here's the register, and here's a hammer Nov 01 21:50:29 nothing wrong with bitbanging Nov 01 21:50:36 memory controller? there _isn't_ one Nov 01 21:50:49 boot up? execution starts at 0x0 Nov 01 21:50:52 it's that word! Nov 01 21:51:03 hello. Cubieboard user here. im am using a linaro-ubuntu rootfs, but im not sutisfied. Im looking for a pure debian rootfs to install on cubieboard. i find noone but http://elinux.org/BeagleBoardDebian. Is this installation porcess usable also on cubieboard? Nov 01 21:51:22 bootloader? That's when you make a program that reflashes the program from USB Nov 01 21:51:22 someone has one it seems Nov 01 21:51:31 debianin: wrong channel! Nov 01 21:51:53 mdp, ops! sorry. what the right one? (if you know) Nov 01 21:52:07 nope, look for a #cubieboard Nov 01 21:52:16 we train beagles here Nov 01 21:52:37 and sell them for their fur Nov 01 21:52:49 oh wait, wrong channel..... Nov 01 21:52:59 on #cubieboard the say to search "debian rootfs harmhf", and google bring me here :) Nov 01 21:53:09 debian rootfs harmful? Nov 01 21:53:18 haven't read that one yet Nov 01 21:53:48 * alan_o realizes he's turned into av500 Nov 01 21:53:50 cubieboard run A10 S.o.C. Nov 01 21:54:27 so it is ARMv7, the cpu that debina 7 is made for. Nov 01 21:55:17 alan_o, there can only be one Nov 01 21:55:23 if also BBoard run A10, so i suppose and hope debina for BBoard can be user als ofor Cboard. Nov 01 21:57:29 debianin: looks like you are in the game early. Nov 01 21:57:43 product just hit the market Nov 01 21:58:20 trailblazing Nov 01 21:58:45 yeah, they got it ready to ship a few but no support Nov 01 21:58:59 debianin: out of curiousity, do they ship register-level documentation for the A10 SoC with the Cubieboard? Nov 01 22:00:30 mdp, i duno. anyway documentation is poor, the just start and have many things to do. Nov 01 22:00:33 download section of their site only has the autocad dimensions Nov 01 22:00:40 no software Nov 01 22:00:54 yes, poos, site, few infos. Nov 01 22:01:00 poor site. Nov 01 22:01:02 in my searches it appears there is no documentation on the A10 itself Nov 01 22:01:10 wait Nov 01 22:01:15 so I can't purchase that as a developer board Nov 01 22:01:52 debianin: did you get the pre-installed buildroot with it? Nov 01 22:02:57 most useful info here: http://linux-sunxi.org/Cubieboard Nov 01 22:19:02 <_av500_> debianin: the beagle does not use the A10 Nov 01 22:19:11 <_av500_> and this is not the place to get cubieboard support Nov 01 22:19:26 hehe Nov 01 22:19:27 * _av500_ realizes he has turned into av500 Nov 01 22:19:43 ok, no A10 so no help fo me. thank the same and sorry to have flood the channel. Nov 01 22:19:44 bye Nov 01 22:20:14 _av500_: you are good :) Nov 01 22:20:15 _av500_: hehe it seems all these boards with no support suddenly are swamping the beagle/panda channels for support, hehe Nov 01 22:20:39 <_av500_> we should send them to rpi Nov 01 22:20:44 <_av500_> they dont have a TRM either :) Nov 01 22:21:19 <_av500_> I asked cubie for the TRM today on g+, not reply so far Nov 01 22:21:23 Hello dear friends: My beagleboard XM rev C just arrive. I would like to install UBUNTU and OpenCV on it, please could you give more information about how to do it? Nov 01 22:21:52 <_av500_> why not run angstrom? that comes with opencv? Nov 01 22:21:58 <_av500_> -? Nov 01 22:23:03 I recall something about a beagleboard.org .. Nov 01 22:23:13 I tried to run a program in OPENCV and it didn't run. How can I know if the OpenCV is installed? Nov 01 22:23:29 that sounds like mru Nov 01 22:23:40 <_av500_> JAIME: did you install it? Nov 01 22:23:45 <_av500_> if not, probably not Nov 01 22:23:52 <_av500_> angstrom has opencv packages Nov 01 22:24:10 <_av500_> something like opkg update Nov 01 22:24:21 <_av500_> opkg list |grep opencv Nov 01 22:26:36 <_av500_> JAIME: if you are running ubunt, no idea about opencv Nov 01 22:26:50 <_av500_> but I guess usual ubuntu sw installation rules apply Nov 01 22:26:54 <_av500_> apg-get something Nov 01 22:26:55 The beagle came with Amstrong, I didn't install opencv. Do I have to install the openCV manually? Nov 01 22:28:11 <_av500_> could be Nov 01 22:28:21 <_av500_> see if you have opencv libraries installed Nov 01 22:28:34 <_av500_> //usr/lib/libopencv or so Nov 01 22:29:35 I don't mind if is ubuntu or amstrong. I just want openCV run. Nov 01 22:30:52 How can I see the libraries installed on Beagle? The beagleboard XM rev C is new. Nov 01 22:42:28 Dear _av500_ can we talk by skype ? Nov 01 22:47:37 <_av500_> lol, no Nov 01 22:47:46 <_av500_> why? you cant read what I write here? Nov 01 22:48:06 <_av500_> where did you use opencv before? Nov 01 22:48:12 <_av500_> on your windows pc? Nov 01 22:48:24 <_av500_> then you might need to invest some time into basic linux Nov 01 22:55:34 I need help with beagleboard and Opencv Nov 01 22:56:00 can anybody help me? Nov 01 23:21:22 _av500_: what is Windows? Nov 01 23:22:59 a porthole Nov 01 23:23:16 an undesireable opening that leaks energy, and other things Nov 01 23:23:27 something to be dispose of and sealed off ASAP. Nov 01 23:24:17 plays games really well though :) Nov 01 23:26:24 ugh... 3d on windows is crap compared to Linux Nov 01 23:26:29 not individual games I'm talking about Nov 01 23:26:43 just in general, the realtime performance is better in Linux Nov 01 23:26:56 why would one need 3D on windows? Nov 01 23:27:15 go for the real thing instead of looking through a port hole Nov 01 23:27:27 well, for my particular project, it was in Windows when I got there, and was cross-platform when I left :) Nov 01 23:28:18 We had a timing graph thing... looked really nice and smooth in Linux, jagged and awful in Windows Nov 01 23:28:19 OpenGL Nov 01 23:28:31 (of course...) Nov 01 23:28:51 must be bad glass Nov 01 23:29:05 could be Nov 01 23:29:24 hey, quick question, for my 7805 issue, I got an 8-way AA holder Nov 01 23:29:43 Was wondering if there's any benefit to putting 8 rather than going with 6 and just soldering a wire across two of them Nov 01 23:30:11 mechanical Nov 01 23:30:25 :) Nov 01 23:30:29 I'd put a dowel w/a wire Nov 01 23:30:39 coiled around the dowel Nov 01 23:30:40 inductor Nov 01 23:30:42 mechanical stuff is surprisingly important Nov 01 23:30:46 yes Nov 01 23:31:09 why not just get a USB battery? Nov 01 23:31:10 why would you use opengl in windows unless you were concerned with cross platform Nov 01 23:31:18 they were $40 :) Nov 01 23:31:24 $40?! Nov 01 23:31:30 Well, in best buy Nov 01 23:31:32 they should be around $20-$25 Nov 01 23:31:41 and that was only 1000mAh Nov 01 23:31:56 heh "1000mAh" Nov 01 23:32:00 ?? Nov 01 23:32:06 I mean that's what it said, who knows. Nov 01 23:32:16 think that is 1000mAH battery, not 1000mAH available Nov 01 23:32:31 yeah, after its regulator burns up enough, etc Nov 01 23:32:42 another thing that made me hesitate Nov 01 23:32:53 i.e you get (3.7/5.0)*0.8 Nov 01 23:33:05 those regulators should be more efficient Nov 01 23:33:18 interesting, according to Felipe, I²C is totally plug-n-play Nov 01 23:33:21 you have to use a switcher of sorts ot boost from the nominal 3.7V of the lithium cell Nov 01 23:33:32 Russ: which Felipe? Nov 01 23:33:39 Balbi Nov 01 23:33:41 gah these linkedin endorsement were cool... for about the first hour... Nov 01 23:33:58 the MUSB felipe? Nov 01 23:34:05 or the DSP Bridge Felipe? Nov 01 23:34:14 I'm not sure that he's actually looked at ->detect() methods for I²C devices Nov 01 23:34:28 TI Felipe Nov 01 23:34:35 think they are both TI felipes Nov 01 23:34:47 balbi@ti.com Nov 01 23:34:53 has he read the doc on I2C detection and how it can kill a device? Nov 01 23:35:05 Russ: nice proper squared sign Nov 01 23:35:23 or how many devices look identical, except they are different Nov 01 23:37:35 by that line of reasoning, SoC's are also plug and play Nov 01 23:37:36 :) Nov 02 00:07:02 how is in ds2-world? Nov 02 00:11:55 mranostay: is that a theme park? Nov 02 00:15:09 when you guys boot up your bone, how long until you can connect to the serial port? Nov 02 00:15:30 For me it's arbitrary, I get "device busy" or some such for a while Nov 02 00:15:39 maybe for 30 seconds or so Nov 02 00:15:49 sometimes once I can connect, it's already booting Nov 02 00:15:53 like the FTDI startup is slow or something Nov 02 00:15:57 but it's not Nov 02 00:16:07 for all that time, I can see /dev/ttyUSB* Nov 02 00:16:14 I didn't think it was alwys like that. Nov 02 00:17:39 it is great...everything "just works" :D Nov 02 00:17:56 hmm so yours doesn't do that? Nov 02 00:18:06 i take care not to get it into that case Nov 02 00:18:11 I assume you are power cycling? Nov 02 00:18:17 no, just when I plug it in Nov 02 00:18:23 following the exact steps on the website Nov 02 00:18:29 but your "plug" is the power, isn't it? Nov 02 00:18:33 yeah Nov 02 00:18:34 oh Nov 02 00:18:35 right Nov 02 00:18:36 but Nov 02 00:18:54 I tried using external power and that didn't help (added other problems iirc) Nov 02 00:19:04 this is what i do - Nov 02 00:19:11 hold down reset, plug in USB, let it settle Nov 02 00:19:13 start kermit Nov 02 00:19:16 release reset Nov 02 00:19:32 that's not a bad idea Nov 02 00:19:37 I'll give that a go next time. thanks Nov 02 00:19:45 if the AM33x wasn't brain damaged from the start, I'd argue for doing a clone w/o the FTDI chip Nov 02 00:19:58 not a fan of am33x? Nov 02 00:19:58 but given that it is brain damaged, I'll wait Nov 02 00:20:03 PM is still broq Nov 02 00:20:15 I like my chips to have good working PM Nov 02 00:21:14 FTDI works mostly Nov 02 00:21:43 that's not my complain about it Nov 02 00:21:54 adds too many complications Nov 02 00:22:08 i like a raw serial port or at least jumpers to disable the FTDI Nov 02 00:22:26 yeah, I'd like 2 headers where I could connect TTL serial Nov 02 00:22:30 I mean that said.... Nov 02 00:22:41 u-boot can be built to use the other serial port Nov 02 00:22:57 u-boot can, but the ROM is a bit less flexible Nov 02 00:23:06 hmm Nov 02 00:23:09 the ROM writes to uart? Nov 02 00:23:19 I suppose I can use the JTAG interface to hold the thing in reset... Nov 02 00:23:24 ROM UART boot Nov 02 00:23:34 i.e. magic with pserial and friends Nov 02 00:56:11 are you talking busy on the host side? Nov 02 00:58:01 ds2, since I don't care about PM, I'm happy with am33x...though if I did.... Nov 02 00:58:57 ds2, re: FTDI maybe the worst kept sekrit in beagle-land will address what you want Nov 02 01:00:25 alan_o, btw, I boot my bone with console on ttyO2 with real RS-232..so it is rocking immediately Nov 02 01:16:30 Hey guys. I have two BeagleBoard xMs. Why can't I share one MicroSD amongst them? e.g. I can pop one MicroSD into BeagleBoard xM #1 and get an OS running off of it, yet when I take that same MicroSD and place it into BeagleBoard xM #2, the device fails to boot. Nov 02 01:18:24 Geminizer: one of the beagles was a pound puppy and was abused, so it fears strange sd cards Nov 02 01:18:39 you're ejecting the rootfs.... linux doesn't like that Nov 02 01:19:23 or are you shutting down between the swaps? Nov 02 01:19:39 I'm shutting down between swaps Nov 02 01:20:10 does the 2nd board ever boot? Nov 02 01:20:20 nope Nov 02 01:20:35 probably not the card then Nov 02 01:21:06 I would expect it to... though could it be the case where that first boot associates the OS with that particular hardware set (if that makes any sense)? Nov 02 01:21:48 no, if they're both xM's the hardware is essentially the same Nov 02 01:22:12 are they both the same rev? Nov 02 01:22:18 that was my thought as well. Nov 02 01:22:24 Yes, Rev C Nov 02 01:37:27 mdp: I understand. but w/o PM, there are cheaper alternatives Nov 02 02:21:27 Anyone here familiar with the netinstall script from RCNelson? Nov 02 02:23:30 ah what is that? Nov 02 02:24:25 It's a nice set of scripts this guy wrote for building various sdcards for the xM Nov 02 02:26:22 The problem I'm having is getting debian and/or ubuntu up and going on my board, shortly after boot the usb and ethernet on the board drop out and become completely non functional until I reset the board **** ENDING LOGGING AT Fri Nov 02 03:00:00 2012