**** BEGIN LOGGING AT Wed Apr 18 02:59:59 2012 Apr 18 03:32:04 so I built a new kernel with cdc acm configured, how do I get the nmea data from the /dev/ttyACM0 to where it needs to go for google maps and things to work? Apr 18 04:53:06 <_av500_> ds2: nonsense, compared to the r-pi, beagle is rocket surgery Apr 18 06:17:11 Microsoft just grew a WART Apr 18 06:18:47 ? Apr 18 06:23:12 goodie for the asm-freaks: https://github.com/jmechner/Prince-of-Persia-Apple-II Apr 18 06:30:11 what is the rt in "ARM RT"? Apr 18 06:30:52 _av500_: can you say pogoplug? Apr 18 06:31:35 I have 4 of them ;) Apr 18 06:31:52 aholler: Really a Tablet Apr 18 06:32:01 and are they less useful then a pogostick? :D Apr 18 06:32:24 Linux eiche 3.3.2-dockstar-00007-g78e0c06 #40 PREEMPT Sun Apr 15 11:08:49 CEST 2012 armv5tel Feroceon 88FR131 rev 1 (v5l) Seagate FreeAgent DockStar GNU/Linux Apr 18 06:32:35 oh the fake pogoplugs ;) Apr 18 06:32:41 one is my 24/7 workhorse Apr 18 06:32:50 aholler: is that a pxa168? Apr 18 06:33:11 or is the 88FR131 truely a diff processor? Apr 18 06:33:23 don't know what the pxa168 is ;) Apr 18 06:33:37 :) Apr 18 06:41:15 looking at the pxa168, the kirkwood seems to lack just the graphics accelerator Apr 18 06:41:34 oh Apr 18 06:41:34 and the ac97 Apr 18 06:41:44 like the 3530 vs 3503 Apr 18 06:43:32 hmm, otg is missing too Apr 18 06:47:04 the pxa seems to be for things with an gui while the kirkwood is more for nas-boxes Apr 18 06:50:22 anyway, both are outdated hw ;) Apr 18 06:51:43 except the gbit-if and working usb ;) Apr 18 06:53:50 ah Apr 18 06:53:59 usb is overrated Apr 18 06:54:37 yes Apr 18 06:54:50 which OTG driver does the pxa use anyways? Apr 18 06:55:35 the EHCI one appropriated by the other A9 folks or is it another monster USB implementation? Apr 18 06:58:01 the otg on the pxa must be different than the one on the kirkwood. the orion (kirkwood) doesn't support otg, just device-mode (udc) Apr 18 06:58:53 lokks like it's ARC Apr 18 06:59:45 never heard that before, but I didn't know the m in musb before too ;) Apr 18 07:04:34 oh I see Apr 18 07:04:51 the real question of course is - does OTG really work or is it like the "OTG" on the PXA270 Apr 18 07:06:04 seems to be different: http://lists.infradead.org/pipermail/linux-arm-kernel/2010-June/017655.html Apr 18 07:08:13 are you searching for an usb-ip? Apr 18 07:09:36 I wouldn't buy the ones from intel and m* Apr 18 07:10:19 got burned by both Apr 18 07:16:29 aholler: write your own with an LS138 and a few GPIOs... Apr 18 07:25:10 and the I replace the faulty ip in my core-i7 with that Apr 18 07:26:18 I'm unsure if I really want to buy a laptop with intels new ivy-bridge, based on the experience I made with my first-gen core i7 :/ Apr 18 07:27:56 I'm still unable to copy from one usb-hd to another. Apr 18 07:29:47 even using just one is like playing lotto Apr 18 08:29:16 hmm, I think I would get rich when I sell all my ac-adaptors as scrap metal Apr 18 08:29:19 most beatiful patch ever: http://code.google.com/p/android-x86/issues/attachmentText?id=474&aid=4740002000&name=new+file&token=vrh4WfqRk_i03pDeqhVqRyz15co%3A1334358364102 Apr 18 08:29:31 aholler: the old non-switching ones? Apr 18 08:29:50 all sorts Apr 18 08:30:10 the switching one have little metal inside Apr 18 08:31:32 just got my pollin package, including two small switching adaptors with changeable voltage and multiple connectors. Now I can get rid of a hundreds of others ;) Apr 18 08:35:13 who writes such patches? Apr 18 08:39:49 apparently somebody smoking too much Apr 18 08:40:00 (or not enough) Apr 18 08:42:16 adroid people Apr 18 08:42:41 cool patch indeed. Apr 18 08:43:19 we should comment that numbering is uncool, and the controllers should be renamed "super" and "fantastic" or such Apr 18 09:08:13 what are differences between the ARM A8 and the OMAP, except for power? Apr 18 09:08:42 lpi: one is an ip core, the other is an actual incarnation. Apr 18 09:08:43 a8 is a processor, omap a soc Apr 18 09:09:32 A10 is a SoC Apr 18 09:09:45 ir even has an A8 inside Apr 18 09:09:48 it Apr 18 09:17:12 thanks Apr 18 09:34:58 koen: I'm playing with kernels and instead of creating a new one (duplicate) I'm including a .bb from another recipe. Known drawbacks? Apr 18 09:38:16 huge explosions? Apr 18 09:38:52 almost, unexpected flames Apr 18 09:39:14 add curves and you have a michael bay movie Apr 18 09:49:32 av500: where can i sign up? Apr 18 09:50:22 LetoThe2nd: you have the curves? Apr 18 09:50:52 nope, i want to see the movie afterwards. Apr 18 10:06:35 hello guys Apr 18 10:07:33 my BB-xM cannot start Apr 18 10:07:58 I just recieved "60" via console Apr 18 10:07:58 :( Apr 18 10:08:07 ant_work: no, I do that a lot locally as well Apr 18 10:08:35 nice, I'll send you a patch for this soon http://paste.debian.net/163654/ Apr 18 10:13:51 my BB-xM cannot start , I just recieved "60" via console Apr 18 10:13:59 How can I fix it? Apr 18 10:14:06 thank you so much Apr 18 10:14:12 sdcard inside? Apr 18 10:14:20 correctly prepared? Apr 18 10:14:35 yes Apr 18 10:14:53 just led D5 and D14 On Apr 18 10:15:15 well, then the card is wrong/bad/not inserted Apr 18 10:15:54 oh, I formated Apr 18 10:16:04 and reInstall Apr 18 10:16:17 but nothing changes Apr 18 10:16:28 formated the wrong way Apr 18 10:17:34 I used Disk Utility to format Apr 18 10:18:31 ? Apr 18 10:18:58 and I use Imager to write file Angstrom-Beagleboard-demo-image-glibc-ipk-2010.3-beagleboard.sd-image-2GiB.img Apr 18 10:19:05 on SD card Apr 18 10:19:14 if you used an image, there is nothing to format Apr 18 10:19:33 the "image" is formatted already Apr 18 10:20:00 yes, I mean I format SDcard and write Image on it Apr 18 10:20:23 how did you "write" it? Apr 18 10:20:29 please xplain the exact step Apr 18 10:20:31 steps Apr 18 10:20:32 it sames if I write Image without format it Apr 18 10:21:16 1st, I format SD card type FAT Apr 18 10:21:39 OMG a Pi just sold for 205 quid Apr 18 10:21:51 2nd, I use ImageWriter on Ubuntu to write Image file on SD card Apr 18 10:21:59 XorA: link? Apr 18 10:22:06 http://www.ebay.co.uk/itm/Raspberry-Pi-Model-B-Brand-New-Box-/180864377586?pt=LH_DefaultDomain_3&hash=item2a1c5b62f2&autorefresh=true Apr 18 10:22:34 that's all. Plz help. Thank you so much. Apr 18 10:22:34 XorA: I think I saw one for 240 even Apr 18 10:22:44 NUTTERS!!!!! Apr 18 10:22:45 XorA: Ok, I'll start my business Apr 18 10:22:48 want to join? :D Apr 18 10:23:08 http://www.ebay.co.uk/itm/Raspberry-Pi-Model-B-ModelB-Brand-New-Box-/140739793135?pt=UK_Computing_DesktopComponents_RL&hash=item20c4bed0ef Apr 18 10:23:10 240 Apr 18 10:23:43 heh, I would by two pandas and a pint for that Apr 18 10:23:46 av500 you still here? Apr 18 10:23:50 Cuongle: "use ImageWriter on Ubuntu" how? Apr 18 10:24:20 XorA: I thought the pis were cheap Apr 18 10:24:26 on irc, always imagine that we cannot see your screen Apr 18 10:24:48 koen: not after you add the idiotax Apr 18 10:24:55 av500: yes, it just a tool to write Image ON SDcard, like win32diskimager on Windows Apr 18 10:25:06 looks like over 50% of the first batch are on ebay and all approaching 100 quid or more Apr 18 10:25:07 yes, I know Apr 18 10:25:25 Cuongle: "use ImageWriter on Ubuntu" how? Apr 18 10:25:32 av500: but they said is was CHEAP Apr 18 10:25:46 koen: get some bones on there listed as Pi compatible Apr 18 10:25:48 remember, "they" control world politics Apr 18 10:26:52 koen: I guess you should have actioned away the 1st bone batch too..... Apr 18 10:26:57 auctioned Apr 18 10:27:13 av500: I mean I using Imagewriter to write Angstrom Image to sdcard Apr 18 10:27:26 ok, I give up Apr 18 10:27:28 we have it to TI to include the box and software damage Apr 18 10:27:31 gave* Apr 18 10:28:34 Pi is the new Openmoko Apr 18 10:29:17 av500: Im the freshman with Linux...please be patient..thank you so much Apr 18 10:29:51 [12:24:26] on irc, always imagine that we cannot see your screen Apr 18 10:30:35 .. Apr 18 10:34:23 linux guys dont have patience, otherwise they would run windows :-D Apr 18 10:37:24 * av500 runs windows to play a Patience Apr 18 10:40:22 anyone here? Apr 18 10:41:29 everybody left for lunch Apr 18 10:41:44 [12:23:50] Cuongle: "use ImageWriter on Ubuntu" how? Apr 18 10:42:41 wish you good food Apr 18 10:44:46 av500: sorry, I not really understand your Q: "use ImageWriter on Ubuntu" how? Apr 18 10:44:54 :( Apr 18 10:45:46 use dd and be carfull to select the correct /dev/sdN Apr 18 10:46:00 as every tutorial suggests Apr 18 10:47:36 its ok, writing /dev/sda is ok until you reboot :_D Apr 18 10:50:03 XorA: yes, all are OK Apr 18 10:50:23 Cuongle: PLEASE DONE WRITE TO sda!!!!! Apr 18 10:50:44 DONT Apr 18 10:50:54 yes, what should I do? Apr 18 10:51:15 do as aholler says Apr 18 10:52:19 Ok, thank you all my buddies. Apr 18 11:51:10 hmm, a cortex-m0 emu for avr: http://dmitry.co/index.php?p=./04.Thoughts/08.%20uM0 Apr 18 11:52:36 so, I can run my Cortex A15 emulator written for M0 on it? Apr 18 11:53:07 why not Apr 18 11:53:17 does that avr have enough memory to store the internal state of the a15? Apr 18 11:54:12 * av500 does handwaving Apr 18 11:54:41 chain enough i2c-rams Apr 18 11:55:23 use a long shift register Apr 18 11:56:19 * panto reads and boggles Apr 18 11:56:56 oh, lab126? Apr 18 11:57:15 panto, :) Apr 18 11:57:50 somebody is making too much money on these arm cores if we can just emulate them on an avr Apr 18 11:57:55 revolution! Apr 18 11:58:48 emulate A15 in r-pi Apr 18 11:58:50 win-win Apr 18 11:58:54 200kHz should be enough to run systemd right? :) Apr 18 11:59:01 not with dbus Apr 18 11:59:44 and without dbus it's udev, what was your question? Apr 18 11:59:45 * mdp notes the beginning of an e2e post: "I have been stucked at a Point with confusion." Apr 18 12:00:09 epic Apr 18 12:00:48 * av500 must not write "I have been stucked to a cross with nails" Apr 18 12:01:21 yes Apr 18 12:01:27 because that's bad grammar Apr 18 12:01:38 right, it should be "at a" Apr 18 12:02:07 shouldn't that just be "I have been nailed to a cross"? Apr 18 12:02:21 'to stick' is a strong verb Apr 18 12:02:46 so 'stucked' is one past tense too many Apr 18 12:02:54 depends on the glue Apr 18 12:03:06 "I have been bluetacked to a cross" Apr 18 12:03:23 koen: yes, it happened in a previous life Apr 18 12:03:25 (and yes, it's present perfect, not past tense) Apr 18 12:03:35 "Jesus has been blue-tacked for our sins" doesn't have a nice ring to it Apr 18 12:05:21 it also violates causality Apr 18 12:05:55 I doubt causality is the only violation of christianity :) Apr 18 12:06:00 at one point I tried to answer all the e2e questions..but quickly found that it would consume 3x of a typical work week Apr 18 12:06:34 mdp: you joined after the big ph0rum push Apr 18 12:06:40 mdp: per day Apr 18 12:07:05 mdp: with the big push every post needed an answer within 48 hours Apr 18 12:07:15 koen, oh…well…I was told we were responsible for answering them…what I later found out is that it's only expected if you know the answer off the top of your head Apr 18 12:07:22 mdp: so AMBU apps had some people fulltime on it to avoid the wrath of upper mgmt Apr 18 12:07:29 koen, we still have that..expect it's just a reply within 24 hrs Apr 18 12:07:47 koen, do you know how easy it is to fulfill the reply metric? :) Apr 18 12:07:51 I do Apr 18 12:07:57 oh yeah :) Apr 18 12:08:05 it used to be 'answer' as in 'solution' Apr 18 12:08:13 yeah, good luck there Apr 18 12:08:15 that lasted a month or 2 Apr 18 12:08:33 I've been down a rat hole on a couple trying to be helpful..then quit replying Apr 18 12:08:54 then get a note from some people in TII begging me to reengage with the crazy person Apr 18 12:10:05 just write a m0-emu for the msp, and afterwards an avr-emu too Apr 18 12:11:15 it's enough with an avr-emu Apr 18 12:11:19 it can run the m0-emu Apr 18 12:12:36 well, but the MSP itself should be written in javascript to begin with Apr 18 12:13:00 and that javascript should run in a browser in an emulated cpu Apr 18 12:13:11 emulators all the way down Apr 18 12:13:16 msp on qemu ;) Apr 18 12:13:51 or "msp430 for android" Apr 18 12:15:17 hmpf, two bt-headsets are dead, f* lipos Apr 18 12:15:55 don't sit on them Apr 18 12:16:05 leave them alone and they will die Apr 18 12:16:11 yes Apr 18 12:16:23 it's neat Apr 18 12:16:30 once gets to sell new hardware every 2 years Apr 18 12:16:39 exactly Apr 18 12:16:40 but you can run yours from NiCDs Apr 18 12:16:45 I still don't get the value of an msp430 if it can't run android Apr 18 12:16:46 100 fps! Apr 18 12:17:04 and no easy way to interface a 4.5" super amoled display Apr 18 12:18:17 mdp: why not? all of the MSPs GPIOs together could maybe power the backlight Apr 18 12:19:12 excellent! Apr 18 12:19:18 jkridner: Oh cmon, are you serious? Apr 18 12:19:25 about? Apr 18 12:19:37 G+ stuff? Apr 18 12:19:44 a blog post about a github that has a PDF that has "opkg install task-native-sdk" inside? Apr 18 12:19:50 hehe. Apr 18 12:19:59 yeah, I was less than impressed. Apr 18 12:20:01 av500: I was just going "WTF?" about the same Apr 18 12:20:03 sorry for the noise.... Apr 18 12:20:06 yeah, so why post it? Apr 18 12:20:10 this is not facebook Apr 18 12:20:26 av500, I'm starting to think the e2e people and the #beagle people prpplague references are the same Apr 18 12:20:33 see, Crofton left in disgust... Apr 18 12:20:49 k. feedback noted. I'll be a bit more discretionary and find another way to log stuff in a more exhaustive manner. Apr 18 12:20:53 heh Apr 18 12:20:55 no Apr 18 12:21:07 found out that machine had been tethereed via phone all night Apr 18 12:21:26 Are the Tenet folks here on the channel? Apr 18 12:21:27 and now you are bankrupt? Apr 18 12:21:43 perhaps they could comment on the usefulness of this new repo.... Apr 18 12:22:01 these might be useful tidbits for the SRM updates Koen is working on.... Apr 18 12:22:06 only 12.72 clicks to reach this useful pdf Apr 18 12:22:08 and the usefulness of the PDFs inside Apr 18 12:22:12 which was part of what I was thinking when I shared it. Apr 18 12:22:22 it shows the types of things newbies get hung up on . Apr 18 12:22:34 jkridner: well then share: "opkg install task-native-sdk" Apr 18 12:22:37 it is the sort of thing Koen really needs to document in his SRM contributions. Apr 18 12:22:47 +1 is so much easier Apr 18 12:22:52 +1 Apr 18 12:23:10 gotcha.... don't make you dig for it.... let you know how banal it is in the title. Apr 18 12:23:22 I want a +2 option so my vote counts…Chicago style Apr 18 12:23:31 jkridner: actually not Apr 18 12:23:34 +2 is patented Apr 18 12:23:38 jkridner: task-native-sdk is already included Apr 18 12:23:41 jkridner, the name of the game is attention span Apr 18 12:24:35 I can of course make a section titled "how to install stuff that is already installed, for the people who think everything on the internet is true" Apr 18 12:24:38 just preinstall gcc Apr 18 12:24:46 jkridner: actually I am waiting for a "how to move the mouse cursor on a beagleboard XM" post Apr 18 12:24:49 and the kernel as git-repo Apr 18 12:25:25 av500: I've top-right corner badge Apr 18 12:25:27 git boot kernel? Apr 18 12:25:29 well, guess you could mention that it is installed in the SRM then. Apr 18 12:25:35 I'm considering a post on how "getting started" on osx or winders is an epic fail Apr 18 12:26:09 and don't forget to preinstall opencv too Apr 18 12:26:10 koen: "Sir, I have a doubt, how do I install opkg install task-native-sdk in the SRM? Apr 18 12:26:17 "Don't listen to the official beagleboard.org folks when they try to convince you that hosting on those platforms is at all useful" Apr 18 12:26:24 av500: I can show you the exact steps Apr 18 12:26:44 no document handy? Apr 18 12:26:52 * koen also adds a section on how to liberate funds from a exiled nigerian prince Apr 18 12:27:01 maybe a pdf or ppt, nothing fancy, a few animations shoud do.... Apr 18 12:27:17 beagleboard as spam-bot Apr 18 12:27:25 koen, I want animations like the one elc session I went to Apr 18 12:27:27 new gsoc project Apr 18 12:27:37 koen, animation that are only possible in powerpoint Apr 18 12:27:56 mdp: did it get you animated? Apr 18 12:28:08 I was bedazzled, I must admit Apr 18 12:28:15 nfc what the session was about Apr 18 12:28:52 guys it would be a whole lot easier if you would just design the boards to climb back into the box and mail themselves back to distributer if user failed an IQ test Apr 18 12:29:11 koen: +1 for removing the ancient 2GB demo images!!! Apr 18 12:29:20 XorA: marketing rejected that technology Apr 18 12:29:35 pack marketing back into a box and mail to ..... Apr 18 12:29:39 XorA: they projected a 90% drop in sales Apr 18 12:29:54 s/sales/recognized revenue/ Apr 18 12:30:43 \o/ Apr 18 12:30:44 https://github.com/xbmc/xbmc/commit/667fd9ff9c6a45be2b1e10782573ca5cb02c7763 Apr 18 12:30:55 free swap real hardware for foam version? Apr 18 12:32:25 koen: wow, even with drama Apr 18 12:34:00 av500: yes, both debian and ffmpeg drama Apr 18 12:34:10 drama ^ 2 Apr 18 12:34:14 ddrama Apr 18 12:34:40 I bet debian optmizes ffmpeg for 286/hurd Apr 18 12:36:15 * koen waits for richardson, TX to wake up and drive to work Apr 18 12:37:28 press fast forward on the world :-D Apr 18 12:37:52 * mdp eagerly awaits part 2 of the getting started with osx lion….compiling the kernel Apr 18 12:38:06 well, be careful ... if all of richardson reacts at the same time that will cause traffic jams without end Apr 18 12:40:06 that osx link isn't that bad actually Apr 18 12:41:53 yep Apr 18 12:42:07 ogra_: add jitter Apr 18 12:42:17 heh Apr 18 12:42:36 RED on highways would be awesome Apr 18 12:42:47 http://highscalability.com/blog/2012/4/17/youtube-strategy-adding-jitter-isnt-a-bug.html Apr 18 12:43:24 ttp://en.wikipedia.org/wiki/Random_early_detection Apr 18 12:44:20 nice, simulink adds dsp support to Arduino.... Apr 18 12:53:10 I also like how people confuse the matlab "dsp toolbox" with "matlabs uses the dsp on beagle" Apr 18 12:54:17 I also like that people think matlab is used outside academia Apr 18 12:54:34 in companies I saw, they use excel Apr 18 12:54:54 as a db Apr 18 12:55:03 as a scientific tool Apr 18 12:55:16 because the nice thing about excel, it always outputs a result Apr 18 12:59:08 xD Apr 18 13:01:04 koen: I actually have a guy sitting next to me with a big MatLab sticker on the back of his corporate issue StinkPad. Apr 18 13:01:46 does he actually use it at work? Apr 18 13:02:05 yes Apr 18 13:02:19 what's bad with matlab, Apr 18 13:02:21 ? Apr 18 13:02:31 but he's also the algorithm brain for a highly advanced self learning system Apr 18 13:02:53 koen, I only take issue with that path because it's a productivity dead-end Apr 18 13:02:55 matlab is great for prototyping and solving one-off problems Apr 18 13:14:23 Anyone managed to get ethernet working with a CONFIG_PREEMPT=y kernel? Apr 18 13:15:02 define "working" Apr 18 13:15:19 can ping something Apr 18 13:15:32 (pretty low standard) Apr 18 13:15:33 jsabeaudry, ubuntu uses it on all its builds, given the omap3 beagle images work i would say yes Apr 18 13:15:40 that works, but more than that, no Apr 18 13:15:46 ogra_: funny man Apr 18 13:15:50 why ? Apr 18 13:16:17 ogra_: you're saying beaglebone ethernet works with PREEMPT=y Apr 18 13:16:23 i have working ethernet on all my beagles with the ubuntu kernel which defaults to CONFIG_PREEMPT=y Apr 18 13:16:40 (as all ubuntu kernels do) Apr 18 13:16:40 even on the bone? Apr 18 13:16:43 no Apr 18 13:16:43 I didn't know ubuntu supported bone at all Apr 18 13:17:01 * ogra_ missed that the question was about bone, sorry Apr 18 13:17:07 koen, so one has the choice between ethernet and decent irq latency ? Apr 18 13:17:16 jsabeaudry: correct Apr 18 13:17:41 run an rtos if irq latency matters Apr 18 13:17:50 qnx has a bsp for the bone Apr 18 13:19:01 mru, well "matters", i can live with 1-2 ms wich is pretty awful by an rtos point of view, but the 50ms of a non preemptible kernel really break what an irq is supposed to be imho Apr 18 13:19:30 on x86 you can spend 4 seconds on an SMI according to intel folks Apr 18 13:22:12 koen: but you have more than one core :) Apr 18 13:22:24 jsabeaudry: usa a beaglebone Duo Apr 18 13:22:27 use* Apr 18 13:24:47 In that case i'd rather buy a fatter fpga that will be able to buffer 50ms of data Apr 18 13:25:01 50ms? Apr 18 13:25:09 where do you get 50ms latency? Apr 18 13:25:20 gpio irq Apr 18 13:25:32 and you measured that? Apr 18 13:25:38 worst case is around 50ms Apr 18 13:25:40 yes with a scope Apr 18 13:25:52 best case is around 500us Apr 18 13:26:30 I've been running user space stuff with a 1ms scheduler for years and never saw 50ms latency Apr 18 13:26:51 av500, how do you configure a 1ms scheduler? Apr 18 13:26:59 in the kernel config Apr 18 13:27:09 disable dyntick and set HZ to 1000? Apr 18 13:27:11 unless you have NOHZ Apr 18 13:27:24 that was from times before dyntick Apr 18 13:27:48 definitely worth a try Apr 18 13:27:57 and since I never saw user space block for 50ms, I wonder how kernel code could be Apr 18 13:28:20 in fact, 1ms scheduler and SCHED_FIFO in user space was just fine Apr 18 13:28:47 rcn-ee: I hear that the http://elinux.org/BeagleBoardUbuntu image is running at only 350MHz.... do you know if that is correct? Apr 18 13:29:09 jkridner: it's to keep the "ubuntu is sluggish" image.... Apr 18 13:29:10 could it be that it is simply running with an on-demand governor and people are interpreting the clock frequency incorrectly? Apr 18 13:29:26 jkridner, is it under load? because omap-cpufreq is enabled.. Apr 18 13:29:49 so it'll move between 350 -> 800... Apr 18 13:30:13 "sudo cpufreq-info" will print out the stats Apr 18 13:32:36 k. have you looked at 1GHz support? Apr 18 13:33:12 used to work fine Apr 18 13:33:14 no more Apr 18 13:33:28 is that what we call progress these days? Apr 18 13:33:35 must be obama's fault Apr 18 13:33:47 as soon as sr1.5 patches appear again... ;) based on his email, that's after the new avs (smartreflex) gets moved to the new dir.. Apr 18 13:34:14 why can't TI finish something just one time? Apr 18 13:34:16 it will all be working with DT Apr 18 13:34:21 mru: unpossible Apr 18 13:34:40 mru: if the actually "finish", they lost their linux patch contrib ranking Apr 18 13:34:44 lose* Apr 18 13:34:45 they are moving it from the mach-omap2 -> a drivers... Apr 18 13:34:52 and MGT does not like that Apr 18 13:35:18 well, they can always create st1.5++ .. and change everything again.. Apr 18 13:35:19 rewriting PM 3x gives you high linux patch stats Apr 18 13:35:30 so it seems Apr 18 13:36:02 but that's not the best part of those stats Apr 18 13:36:06 the other explanation is that they finish something and then look at at it shame and rewrite it - other just push it in mainline Apr 18 13:36:18 microsoft contributing more than canonical, that's funny Apr 18 13:36:28 ogra_: ^^ what can you say about that? Apr 18 13:36:44 the other explanation is that they finish something and then look at it in shame and rewrite it - others just push it in mainline Apr 18 13:36:50 there, with sense Apr 18 13:37:59 mru, that microsoft is like 1000x bigger than canonical ? Apr 18 13:38:25 and that ubuntu is a pure downstream distro as well :) Apr 18 13:38:41 ubuntu is a pure patch-hoarding distro is what Apr 18 13:39:04 jkridner, just watching/reading http://www.spinics.net/lists/linux-omap/msg67659.html hopefully Jean is able to implement all that.. Apr 18 13:39:26 well, ubuntu is focused on adding glue, nothing else Apr 18 13:39:50 * Crofton|work makes a note about that next time someone asks him why not ubuntu Apr 18 13:39:53 if you want upstream stuff, go to linaro ;) Apr 18 13:39:59 so what are those ~1000 patches in ubuntu kernels? Apr 18 13:40:10 ask the ubuntu kernel team ? Apr 18 13:40:31 there are two kinds of patches: ones that should go upstream and useless ones Apr 18 13:40:48 * ogra_ doesnt work on kernels usually ... Apr 18 13:40:57 that applies to all patches :) Apr 18 13:41:02 and i dont know where that number comes from either Apr 18 13:41:11 its the one after 999 Apr 18 13:41:18 heh Apr 18 13:42:14 * ogra_ also doesnt get why everyone looks at kernel only ... take a look at gcc and the toolchain ... Apr 18 13:42:23 distros are more than just a kernel Apr 18 13:42:43 as well as are contributions ... Apr 18 13:43:05 like Unity Apr 18 13:43:30 well unity is a canonical upstream project ... ubuntu is only a consumer here Apr 18 13:52:23 ogra_: I didn't count them, but there are quite a few ubuntu kernel patches: http://kernel.ubuntu.com/git?p=ubuntu%2Fubuntu-precise.git&a=search&h=HEAD&st=commit&s=^ubuntu&sr=1 Apr 18 13:52:45 might be Apr 18 13:52:56 as i said, i try to stay away from kernels Apr 18 13:53:18 (beyond making them run on target HW at least) Apr 18 13:53:40 the ubuntu kernel team can surely elaborate Apr 18 13:54:17 but you can be sure they wont carry any patches (and add maintenance penalty for 5 year supported releases) without very valid reasons Apr 18 13:54:24 ogra_, understand we bitch at everyone with out of mainline patches, or poor performance upstreamimg stuff Apr 18 13:54:56 we bitch at everyone. period Apr 18 13:55:00 Crofton|work, i dont feel attacked (especially not if mru is involved in such a discussion ;) ), no worries Apr 18 13:55:31 how much RAM is needed to build Angstrom, say the cloud9 beaglebone image? Apr 18 13:55:34 hi Guest21555 Apr 18 13:56:01 rcn-ee was just mentioning http://www.spinics.net/lists/linux-omap/msg67659.html, but I think that is a long-term fix of some kind. Apr 18 13:56:59 rcn-ee was saying that http://elinux.org/BeagleBoardUbuntu kernel runs 350-800MHz using omap-cpufreq, but does not have whatever it takes to get to 1GHz. Apr 18 13:57:31 I believe that adaptive body biasing is required to enable 1GHz, but I'm not aware of that being around for the newer kernels. Apr 18 13:58:19 rcn-ee: are you still around to confirm? Apr 18 13:59:01 wasnt that for 1.2? Apr 18 13:59:29 no, 800Mhz+ needed the voltage increase for reliably... so it needs that stuff. Apr 18 13:59:48 but yeah, jkridner is correct... Apr 18 14:00:42 I thought the reason the voltage increase wasn't there was because without ABB, you might overly shorten the life of your device (running a hot device at a higher voltage than needed). Apr 18 14:02:32 i keep getting out of memory errors when building Angstrom. how much RAM do you guys recommend? my VM is currently set to 1GB Apr 18 14:02:46 plenty Apr 18 14:02:48 jkridner, that too... two of my xM (early A and B versions) will hardlock when forcing 1Ghz on bootup.. Angstrom and also the mmc voltage regulator hack would make it reliable.. Apr 18 14:02:50 in 2001 Apr 18 14:03:28 av500: lol, yea. so what's the 2012 number? 8GB? Apr 18 14:03:39 jkridner, so maybe, they just work with later silicon as the process got refined.. Apr 18 14:04:57 Guest21555: is that useful info? Apr 18 14:07:18 1GHz is available from PSP release. Is that feasible to be used? Apr 18 14:29:37 Guest21555: I'm not sure exactly why not, but apparently it isn't easy to pull the patch and apply it to a newer kernel. Apr 18 14:31:52 jkridner, because there is a lot of churn in omap Apr 18 14:32:02 particularly in that area Apr 18 14:32:10 no, omap is totally fucked Apr 18 14:32:14 that too Apr 18 14:32:25 mru, thank you..I couldn't quite find the words Apr 18 14:32:26 wasnt all of arm fucked? Apr 18 14:32:34 no, just omap Apr 18 14:32:37 ah Apr 18 14:32:53 I'm sure rmk's old arm9tdmi is doing just fine Apr 18 14:33:38 jkridner, so anyway, it requires somebody to 100% understand the changes and refactor for the current kernel…it's not a "trained patch monkey" exercise. Apr 18 14:34:03 anybody qualified working on it? Apr 18 14:34:40 I have no visibility into people working on stuff at TI beyond what hits linux-omap Apr 18 14:34:58 so from my pov, I know of nobody Apr 18 14:35:17 on linux-omap, it looks like jean.pihet it looking on it with his "AVS" driver.. Apr 18 14:36:15 mdp: I guess you are working on other subsystems? Apr 18 14:36:27 jkridner, what do you mean? Apr 18 14:36:38 koen, av500: Thanks for the HZ=1000 and no dynamic ticks, the worst case latency now seems to be around 10ms instead of 50ms, any other sweet tricks like this that could get me closer to 1ms ? Apr 18 14:36:44 aren't you working on TI ARM Cortex-A8 devices? Apr 18 14:37:11 jkridner, me personally? I'm specifically looking at am37x-evm ehci host breakage Apr 18 14:37:26 k Apr 18 14:37:54 jsabeaudry: what exactly are you measuring? Apr 18 14:37:59 for those looking at beagleboard xm, 1ghz isn't a requirement...fwiw Apr 18 14:38:12 it's in the "nice to have" class Apr 18 14:38:27 phew Apr 18 14:38:53 mdp: ehci host breakage? that rings a few bell towers Apr 18 14:39:21 av500, it's a seasonal thing Apr 18 14:39:35 and now is hurting season? Apr 18 14:39:36 av500, the time between the rise of a gpio and my first access to the gpmc (in the irq handler) Apr 18 14:39:46 all in kernel? Apr 18 14:40:42 jkridner, there's enough more fundamental things broken to consume a team 5x the size right now :) Apr 18 14:40:48 av500, the code is in a kernel module yes Apr 18 14:41:06 jkridner, dunno if you ever heard of this thing called musb? Apr 18 14:41:15 never! Apr 18 14:41:20 its like GPTIMER12 Apr 18 14:41:24 count yourself lucky Apr 18 14:41:32 :) Apr 18 14:41:40 mdp: I have people in my team that dream musb Apr 18 14:41:48 and I have people that dream EHCI Apr 18 14:41:56 unpleasant dreams Apr 18 14:42:04 it's common to mix those with unicorns and kittens Apr 18 14:42:09 and prancing ponies Apr 18 14:42:17 we had a conf call with TI on a future SoC Apr 18 14:42:29 we asked them "does EHCI work finally" Apr 18 14:42:30 Has anyone tried 1GHz with Ubuntu? Apr 18 14:42:33 they laughed Apr 18 14:42:38 then we read them the eratum Apr 18 14:42:43 then they fell silent Apr 18 14:43:08 Guest21555, linaor has i belive Apr 18 14:43:11 *linaro Apr 18 14:43:15 Lenor Apr 18 14:43:17 ... ask in #linaro Apr 18 14:43:28 av500, heh, that didnt come to mind yet Apr 18 14:43:30 Guest21555, just dump the uImage/modules from angstrom's 3.0.X kernel on top of the ubuntu rootfs.. Apr 18 14:43:48 * ogra_ always thinks of a sanitary napkin hearing "linaro" though Apr 18 14:43:55 av500, the trick is having our h/w apps guys hand hold your h/w designer through the twisted maze and you can get it working on am37x, at least Apr 18 14:44:16 mdp: am37xx is omap37xx? Apr 18 14:44:25 so the bugs should be the same? Apr 18 14:44:43 Thanks Robert for the suggestion Apr 18 14:45:05 what rcn-ee mean to say "Linux runs Linux" Apr 18 14:45:08 meant* Apr 18 14:45:25 (its easy to test wit5h the ubuntu kernel though, its just bumping a cmdline option to enable 1GHz) Apr 18 14:46:13 yeah, but from what i've read, it sounds like he's using my image.. by forcing mpurate=1000, it should hardlock on bootup. ;) Apr 18 14:47:34 rcn-ee: I'm trying to get over my extreme dislike for git submodules and see if I can fix the script instead :) Apr 18 14:48:25 koen, me too.. i almost got something working, then after a push, and then a pull on another machine the submodule wasn't pull in.. Apr 18 14:48:49 want to just .gitignore a 'dl' directory, and have a script pull in linux-stable? ;) Apr 18 14:48:51 av500, I still don't understand the gobbldygoop of the history of these parts, quite frankly Apr 18 14:48:59 What is the highest safe value for HZ on the bone? Can I go much higher than 1000 you think? Apr 18 14:49:17 jsabeaudry: I would try to understand where that latency comes from Apr 18 14:49:24 av500, iirc, it's a 3630 derivative Apr 18 14:49:25 rcn-ee: cp -a patches something ; git-foo bar ; sh something/script Apr 18 14:49:28 not increase HZ all the time Apr 18 14:49:32 mdp: right Apr 18 14:49:43 so your EHCI bugs should be well known here :) Apr 18 14:50:22 well, there's the known stuff I've learned about just supporting the antique blessed kernels for customers... Apr 18 14:50:34 then there's the new bugs inserted due to bitrot upstream Apr 18 14:51:24 due to all the churn, and lack of real testing, there's been a lot of quirky breakage Apr 18 14:51:54 mdp: av500 also counts silicon bugs Apr 18 14:51:59 mag found a lot of PM stuff just hammering on bogus registers on am35x... Apr 18 14:52:01 mostly Apr 18 14:52:04 silicon bugs are features Apr 18 14:52:05 SW we can fix :) Apr 18 14:52:34 think of them as challenges to conquer like Everest…they are "there" to be conquered Apr 18 14:52:49 mdp: yes, but why year after year? Apr 18 14:52:58 or if it's too painful, you choose a less difficult mountain Apr 18 14:53:06 we conquered EHCI for omap3430, 3440, 3630, 4430 and 4460 Apr 18 14:53:09 av500, process fail Apr 18 14:53:14 it gets plain boring Apr 18 14:53:23 yep Apr 18 14:53:48 what I guees is happening that despite "one silicon every 6month", we are still seeing the old 3y dev cycle Apr 18 14:54:00 so all these SoCs are WIP snapshots Apr 18 14:54:24 I was little annoyed to be put on a plane with 4 hours notice only to have the fix be a backport of a handful of patches from mainline to 2.6.32….after a long week of debug Apr 18 14:54:34 hmmm Apr 18 14:54:35 http://cgit.freedesktop.org/libminidump/tree/src/read-coredump.c#n173 Apr 18 14:54:51 wtf? Apr 18 14:54:55 a variant of the HOME = /home/crofton && break Apr 18 14:55:09 av500: it's on fd.o, no futher explanation necessary Apr 18 14:55:37 av500, we've since been awarded a huge team of 2.5 people to fix all the issues in the kernel from the past umpteen years Apr 18 14:55:54 mdp: you should have just asked us :) Apr 18 14:56:07 not those issues :) Apr 18 14:56:10 mru: its a tool to read Lennart Apr 18 14:56:16 mru: its a tool to read Lennart's core dumps? Apr 18 15:07:14 is it Jan Axelson that is constantly breaking threads on the ML? Apr 18 15:10:08 koen, this will pull in a good kernel sha commit, that patch apply function needs a little tweak stil... https://github.com/RobertCNelson/kernel-try3/blob/master/patches/patch.sh Apr 18 15:15:40 hmm, just had a nip and it seems I've missed a good time to spray some blame too ;) Apr 18 15:16:06 btw. 3.3.x is unusable ;) Apr 18 15:16:56 and I like git submodules Apr 18 15:18:19 aholler, seeing same here, 3.3.x just can't handle a load.. Apr 18 15:20:45 rcn-ee: I like the way you are heading with that script! Apr 18 15:26:26 thanks, koen.. getting stuck right here in the patch section.. http://pastebin.com/D7ZBhgC3 Apr 18 15:27:10 rcn-ee: yeah, that's why I wanted to cp patches/ to something in .gitignore when doing the actual patching Apr 18 15:28:22 ah i see.. but we should be able to access them from the patches dir.. Apr 18 15:32:43 rcn-ee: at least for one problem Russ found/got a patch: http://marc.info/?l=linux-omap&m=133406536900864&w=2 Apr 18 15:33:32 rcn-ee: patches/ is under git control, so it goes away with the reset --hard Apr 18 15:33:45 rcn-ee: maybe a git checkout patches OLDSHA would work for that Apr 18 15:33:48 but I still have to bisect for at least on other thing as more seems to be broken Apr 18 15:34:19 koen, but git reset should only be in the .gitignored directory holding the linux.git repo.. Apr 18 15:36:47 Crofton|work: what's your opinion on the tag name? Apr 18 15:37:02 as in the default Apr 18 15:37:04 ? Apr 18 15:37:12 Crofton|work: no, as in the name I used Apr 18 15:37:29 Crofton|work: "angstrom-v2012.05-yocto1.2" Apr 18 15:37:35 Crofton|work: as opposed to "denzil" Apr 18 15:37:45 no code names Apr 18 15:37:50 denzil? Apr 18 15:38:02 also, see if the other tags will have yocto in the front or the back Apr 18 15:38:07 and follow the same format Apr 18 15:42:19 koen, it just didn't like $PATCHPATH so just hardcoded it to ../patches and now it's patching just fine.. Apr 18 15:45:54 opps, it didn't like patching the defconfig, better just do *.patch ;) Apr 18 15:46:33 I ran into some bugs with git, hence the patch-by-patch git am , instead of git-am $list Apr 18 15:47:22 yeah, it's a must for the am335 v3.2 tree, you found the only way that worked relabily for merging in those stable patches.. Apr 18 15:48:38 it has a conflicting merge inside in net/ Apr 18 15:49:00 * koen mumbles something about a millionn typewriters Apr 18 15:52:43 av500: as you like erratas about usb: http://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/5-and-3400-chipset-specification-update.pdf Apr 18 15:52:54 humm, it now makes it thru all the patches... now to fix the src-uri.txt file location.. Apr 18 15:54:52 aholler: yes maybe Apr 18 15:55:20 aholler: but with Intel stuff you have a million people working around the same HW Apr 18 15:55:29 doesn't help me Apr 18 15:55:37 s/help/helped/ Apr 18 15:56:17 they just stopped accepting usb-bugs at bugs.kernel.org and are forwarding to the ml where you've got lost Apr 18 15:56:27 and intel doesn't support linux Apr 18 15:56:55 servers dont need usb Apr 18 15:57:07 and desktops dont need linux :) Apr 18 15:57:20 solved Apr 18 15:57:23 so it looks like Apr 18 15:58:25 therefor my next desktop wilk be amd (again) ;) Apr 18 15:59:30 but for my laptop I'm lost in intel-land again :/ Apr 18 16:00:18 amd, lol Apr 18 16:00:53 I never had a problem with my amd64 Apr 18 16:01:27 well, you are right, I have more dead P4s than dead Athlons Apr 18 16:01:36 but the core i* is just a usb-nightmare Apr 18 16:01:45 but at least the P4s died gracefully Apr 18 16:03:16 e.g. one 'fix' they did in the linux kernel, is just to restart usb in case of an error, without throwing out an error in dmesg Apr 18 16:04:33 aholler: yes, it can be ascribed to "The user does not care!" Apr 18 16:05:32 hmm, such a restart needs seconds. I've just stopped using usb-storage on my intel-box Apr 18 16:07:42 so I now have 6 idle half-cores Apr 18 16:08:37 i see no issue on my I7 Apr 18 16:08:52 but its an old I7 Apr 18 16:09:22 hmm Apr 18 16:09:31 Hmm, I was an early adopter, that stupid chipset just appeared when I bought it. Apr 18 16:09:40 which one? Apr 18 16:09:52 p55 Apr 18 16:09:59 uart4 on the beaglebone p9 pins 11 and 13 should work just fine shoudln't it Apr 18 16:10:11 speaking of usb bug... i'm running 3.2.0-psp7 on a beaglebone, and i've got a usb camera that randomly timesout (always before getting a single picture) Apr 18 16:10:20 I wonder if the default pinmux for uart4_rx is somewhere else, so that I can't change that pin to mode 6 Apr 18 16:10:23 is it known/is there some workaround/fixes available somewhere ? Apr 18 16:10:32 This document says uart4_rx_mux2 Apr 18 16:10:39 that mux2 is probably important Apr 18 16:11:32 aholler: x58 chipset is working fine over here Apr 18 16:11:40 phh, that version had the latest workaround i pulled from http://arago-project.org/git/projects/?p=linux-am33x.git;a=shortlog;h=refs/heads/v3.2-staging (the top 3 where musb "fixes") Apr 18 16:13:14 mru: I assume they fixed some of the stuff in later revisions. Apr 18 16:13:35 Therfor I'm unsure if I really should by an ivy-bridge laptop when it appears Apr 18 16:14:25 anyone manged to vnc into angstrom 04/09 build on the beaglebone? Apr 18 16:14:36 I've never had any serious problems with intel chipsets Apr 18 16:15:05 my 440bx is still running great. ;) Apr 18 16:15:43 I still have a pentium with that float-bug lying around ;) Apr 18 16:18:08 I have plenty of cpus with lots of bugs Apr 18 16:18:13 most of them quite new too Apr 18 16:18:41 yeah, hw is a pain Apr 18 16:19:10 I miss 68k Apr 18 16:19:31 that one worked like a charm ;) Apr 18 16:19:37 my personal favourite cpu bug is the avr32 branch folding bug I found Apr 18 16:19:37 flawless Apr 18 16:19:43 so utterly stupid Apr 18 16:19:52 we should never have left 6502 Apr 18 16:20:06 hmm, that had bugs too Apr 18 16:20:37 but the sw was small enough to care about them Apr 18 16:39:37 anyway, back to bisecting 3.3 on the omap ;) Apr 18 16:43:02 aholler, everybody's doing it…regressions give you something to work on at least Apr 18 16:43:26 it doesn't give me something to work, but my beagle ;) Apr 18 16:45:02 hehe Apr 18 16:45:30 does that mean you're doing the ethical thing and building the kernel on the beagle? Apr 18 16:47:54 exactly, I've time and bisecting doesn't cost me much ;) Apr 18 16:49:42 aholler, which breakage are you tracking down on the beagle? Apr 18 16:50:21 can't say as I don't know. but it doesn't boot from usb with all kind of wired errors Apr 18 16:50:47 so ehci broken? Apr 18 16:51:12 not really, that seemed to be fixed by the patch above Apr 18 16:51:33 i think the clock-stuff is somewhat broken Apr 18 16:51:41 "weird errors"? Apr 18 16:52:18 aholler, you have: https://github.com/RobertCNelson/stable-kernel/blob/v3.3.x/patches/beagle/0001-omap3_beagle-init-uart2-for-beagle-rev-AX-BX-only.patch Apr 18 16:52:34 yeah, getty bails out, or not more than just "starting kernel" and such things Apr 18 16:53:13 Hmm, How can I set CONFIG_HZ=1000, it does not seem to be in the menuconfig Apr 18 16:53:15 hrm, I wonder if Russ and panto also see that…they are working on 3.3 beagle stuffs Apr 18 16:54:21 rcn-ee: I have http://www.spinics.net/lists/linux-omap/msg68227.html Apr 18 16:54:51 that fixed booting from ehci Apr 18 16:55:26 humm, i ignored that patch, ehci works here with mine, must do the same thing in the end.. Apr 18 16:56:43 koen, really close.. src-uri.txt is getting the full path, but otherwise everything else now works.. gota run to a lunch/meeting, but will see if i can fix that too.. Apr 18 16:56:55 rcn-ee, there were a few iterations of the fix Apr 18 16:59:07 mdp, ? Apr 18 16:59:34 mdp, yes, I have this in to boot Apr 18 16:59:42 I still don't like the way it's handled Apr 18 16:59:43 panto, aholler was describing other breakages Apr 18 17:00:08 panto, it is arguably better than fixing it for just one board Apr 18 17:00:45 hmm, I missed those breakages Apr 18 17:00:50 aholler, can you elaborate? Apr 18 17:01:11 * mdp goes back to bisecting other breakages Apr 18 17:01:26 * mdp detects a pattern Apr 18 17:07:21 panto: with 3.3 sometimes it doesn't boot at all. disabling resetting clocks helps (at least somewhat). Apr 18 17:07:48 gm Apr 18 17:08:18 aholler, hmm, let me try to boot mine Apr 18 17:23:20 | rasqal_internal.h:39:25: fatal error: git-version.h: No such file or directory Apr 18 17:23:23 I hate kde Apr 18 17:25:47 panto, first kindly make a port of linux to your beagle Apr 18 17:26:13 and show us exact steps Apr 18 17:26:18 then tell use the codes Apr 18 17:26:23 z Apr 18 17:26:23 because we have doubts Apr 18 17:26:31 s/doubts/doubt/ Apr 18 17:27:55 doing some research in open source hardware (master, sigh) Apr 18 17:28:09 if i wanted to talk with someone at beagleboard about the project, who to contact? Apr 18 17:28:18 my beagle is at the vet Apr 18 17:28:31 my beagles are bored Apr 18 17:28:43 <_av500_> invresearch: ask, dont ask to ask Apr 18 17:28:48 http://i282.photobucket.com/albums/kk254/9cleo/Puppies/sad-little-beagle.jpg Apr 18 17:28:53 _av500_: link? Apr 18 17:29:01 sad beagle at the vet Apr 18 17:29:48 well, i guess my "problem" is that i assume that there is no real management for the project? Apr 18 17:30:02 or, looking at this page it seems that way at least Apr 18 17:30:10 <_av500_> ask, dont ask to ask Apr 18 17:30:12 <_av500_> just ask Apr 18 17:31:07 right now i am listing up which projects i want to look at, the questions are still being prepared Apr 18 17:31:15 I just got my GTA04 in the mail. Not sure what I'll do with it---but plan to put it on the WiFi network to start. Apr 18 17:32:15 here is a question for you: Who is the contact person for this project- Apr 18 17:32:41 jkridner: you already given up on the gsm part? Apr 18 17:32:49 invresearch: from the web page, off of the "about" link: http://beagleboard.org/brief Apr 18 17:33:22 I am not invested enough to put my only working SIM into it. Apr 18 17:33:31 I cheated and didn't use google..I went to the url myself and clicked on the link Apr 18 17:33:40 I'll get it on the web, then decide if I want to actually use it as a phone and get a SIM for it. Apr 18 17:34:05 invresearch: I'm someone good to start with. Apr 18 17:34:33 ok, i saw your name on that page, i will send you an email at a later time if that is ok Apr 18 17:34:38 jkridner, coincidentally, your name is even at the bottom of the brief. Apr 18 17:34:53 wow, what a coincidence! ;-) Apr 18 17:35:36 shocking, I'd say Apr 18 17:35:38 e-mail is indeed welcome at any time. Apr 18 17:35:54 thank you very much Apr 18 17:36:06 have a nice day, thanks for the help Apr 18 17:36:14 <_av500_> now we are all curious Apr 18 17:36:25 <_av500_> and only jkridner__ will get an email Apr 18 17:36:30 urgh. Apr 18 17:36:30 not sure what is up with my network. Apr 18 17:36:34 <_av500_> HOW UNFAIR! Apr 18 17:36:44 i will post my findings, if i not flunk ;) Apr 18 17:37:01 invresearch: I do always prefer queries going to the discussion group (http://beagleboard.org/discuss). Apr 18 17:37:01 <_av500_> post all the codes NOW! Apr 18 17:37:08 far more likely to get a complete answer there. Apr 18 17:37:10 * mdp gathers the shell casings for forensics Apr 18 17:37:16 ok, will do Apr 18 17:37:22 aholler, root@beagleboard:~# uname -a Apr 18 17:37:22 Linux beagleboard 3.4.0-rc3-00019-gbece2fb #7 SMP Thu Apr 19 17:24:35 UTC 2012 armv7l GNU/Linux Apr 18 17:37:23 jkridner__: speaking of email, I forgot to copy you on the roadmap one Apr 18 17:37:29 seems to boot over here Apr 18 17:37:36 jkridner__: but most info is in the google doc I sent you Apr 18 17:37:50 panto, did you mean to phrase that differently? "works for me!" Apr 18 17:37:51 panto: nice. Apr 18 17:38:10 hope it does more than just boot! Apr 18 17:38:17 panto: 3.4 is too new for me Apr 18 17:38:35 "έργα για μένα" Apr 18 17:38:41 jkridner__, it blinks lots of blinkenlights! Apr 18 17:38:45 jkridner__, there's something beyond the boot? Apr 18 17:38:49 I only care about kernels named stable Apr 18 17:39:07 aholler, there's no such thing ;) - but I get your point Apr 18 17:39:21 mdp, no I don't do that :) Apr 18 17:39:44 panto: "named" Apr 18 17:39:59 s/named/called/ Apr 18 17:40:00 panto, nfsroot? Apr 18 17:40:17 no, mmc Apr 18 17:40:34 of course Apr 18 17:40:59 mdp, for some reason nfsroot doesn't work for me Apr 18 17:41:02 and I boot from ehci ;) Apr 18 17:41:13 and I hasn't bothered fixing it yet Apr 18 17:41:31 panto, it was a fail for me recently on xM w/ the uart pad fix applied Apr 18 17:41:47 mdp, the uart fix seems to be applied in mainline now Apr 18 17:41:50 but lemme check Apr 18 17:42:00 not in 3.3.2 Apr 18 17:42:16 yes, it is Apr 18 17:42:58 aholler, mdp, care to send me your kernel command lines & put your rootfses someplace? Apr 18 17:43:09 this was prior to it being merged and is out of date info Apr 18 17:43:39 I see lots of scary messages on boot Apr 18 17:43:55 panto: my rootfs is a gentoo full with stuff like kde and such ;) Apr 18 17:44:08 panto, that's "normal" Apr 18 17:44:24 aholler, tell me you're not emerge'ing on the bbxm :P Apr 18 17:44:35 no, I don't Apr 18 17:44:37 mdp, for a very weird definition of 'normal' Apr 18 17:44:44 I'm doing it on a beagle c4 ;) Apr 18 17:44:48 lol Apr 18 17:45:01 is that part of your monk patience training? :) Apr 18 17:45:03 panto, in omap-land it's welcomed to have broken clocks, failing hwmod messages etc…as long as they are known to be benign Apr 18 17:45:35 panto: it helps the beagle not to sleep. Apr 18 17:45:58 I'm usually don't look at it while it compiles ;) Apr 18 17:46:18 the poor mutt is going to be buffed from all that exercise :) Apr 18 17:46:40 mdp, you want to have a chat about that? :) Apr 18 17:47:15 panto, no, rmk established the root cause Apr 18 17:48:24 aholler, define "boot from ehci" please Apr 18 17:48:48 gee five with same time stamp Apr 18 17:49:30 mdp, nfsroot over gadget ethernet? Apr 18 17:49:47 he said ehci Apr 18 17:50:01 oh? Apr 18 17:50:04 and C4 Apr 18 17:50:15 the plot thickens Apr 18 17:50:46 in my work "boot" is either loading any part of the MLO/u-boot/uImage chain Apr 18 17:50:47 panto: my cmdline is console=tty1 console=ttyO2,115200n8 mpurate=720 musb_hdrc.fifo_mode=5 debug vram=1572864 omapfb.vram=0:1572864 omapfb.mode=dvi:1024x768-16@60 omapfb.debug=y omapdss.def_disp=dvi root=/dev/sda5 rw rootfstype=ext4 rootwait Apr 18 17:50:57 and the config is http://ahsoftware.de/config.beagle.3.2+.txt Apr 18 17:51:19 aholler, where do you load your uImage from? Apr 18 17:51:25 ehci Apr 18 17:51:34 mass storage in u-boot? Apr 18 17:51:39 yes Apr 18 17:51:49 ok, cool, thanks Apr 18 17:51:58 MLO/u-boot from mmc? Apr 18 17:52:02 or nand? Apr 18 17:52:10 I think nand Apr 18 17:52:19 but works both Apr 18 17:52:25 have an sd there too Apr 18 17:52:29 sure, just curious Apr 18 17:53:23 and you aren't using musb at all , right? Apr 18 17:53:42 no, I have a bt-dongle there for networking Apr 18 17:54:16 and keyboard/audio, if I want such Apr 18 17:54:26 ok Apr 18 17:54:38 but usally just for ssh Apr 18 17:55:00 desktop replacement Apr 18 17:55:05 playground Apr 18 17:55:22 yeah, I guess it can't be a desktop without Unity :) Apr 18 17:55:30 kde works Apr 18 17:55:36 just a bit slow Apr 18 17:55:44 but usable Apr 18 17:56:51 I even have some meego-stuff compiled there ;) Apr 18 17:56:57 nooo!!! Apr 18 17:57:03 too bad that died Apr 18 17:57:09 ittthatshallnotbenamed Apr 18 17:57:22 ok, so "sometimes it doesn't boot" Apr 18 17:57:23 it got renamed ;) Apr 18 17:57:49 Hi, from the beaglebone angstrom build, how do I get only the lib-dev files? I want to keep these dev files separately on which I can build my app development. Apr 18 17:57:52 you load from usb msc device, it uncompressed and hang, no output, etc? Apr 18 17:58:00 sometimes Apr 18 17:58:40 but no output is the symptom…sometimes Apr 18 17:58:42 with 3.3.2 I never got it to boot, with 3.3.0 other things happen Apr 18 17:58:56 but disabling restting clocks helps Apr 18 17:59:20 resetting clocks..or resetting the hwmods? Apr 18 17:59:44 resetting unused clocks is it named, not sure Apr 18 17:59:56 would have to look at menuconfig ;) Apr 18 18:00:32 ahh, that one, ok Apr 18 18:01:29 you may have simply perturbed a heisenbug Apr 18 18:02:08 btw, emerge --sync over bt over musb kills musb ;) Apr 18 18:03:32 aholler, my Keurig heating up kills musb Apr 18 18:05:45 but that might be a bug in bluez too Apr 18 18:06:12 didn't found the reason, fifo=5 didn't help Apr 18 18:07:08 "We've added an entirely new system of Keurig® Brewers and Packs called Vue™." Apr 18 18:08:16 they need cookies, how awful ;) Apr 18 18:11:32 Hi, from the beaglebone angstrom build, how do I get only the lib-dev files? I want to keep these dev files separately using which I can create my app development. Apr 18 18:16:19 aholler, you haven't lived until you've had the opportunity to hand tune (hack the driver) the fifo modes to make a device work at the expense of others. Apr 18 18:19:42 aholler, we use our own coffee in the keurig, their stuff is overpriced crap Apr 18 18:19:43 <_av500_> prasant: its the "sdk" part of the build Apr 18 18:20:51 only the -dev packages are useless Apr 18 18:21:03 since you'll need the actual libraries to link Apr 18 18:22:03 koen, i think i got all the bugs worked out, just going to recreate your beagleboard-3.0 branch to verfiy it... Once that's done, i'll create a new git tree, any preference on the name? Apr 18 18:22:07 <_av500_> mdp: keurig cannot do espresso -> useless Apr 18 18:22:23 rcn-ee: tree or branch? Apr 18 18:22:33 <_av500_> rcn-ee: zendil Apr 18 18:22:35 rcn-ee: feel free to wipeout the current beagle/kernel tree Apr 18 18:22:45 _av500_: I do have to go out for my espresso Apr 18 18:23:01 tree.. https://github.com/beagleboard/kernel do you want to reuse "kernel" or something, else.. Apr 18 18:23:04 <_av500_> get a nespresso and a spending account :) Apr 18 18:23:15 rcn-ee: kernel is ok for me Apr 18 18:23:20 okay will do.. Apr 18 18:24:18 koen: yes that as well. Basically I do not want to integrate my development environment with the Angstrom build but maintain a separate one. So is there a way to extract the relevant files from angstrom build. Apr 18 18:25:31 <_av500_> what for? Apr 18 18:25:42 <_av500_> just point to the angstrom stuff Apr 18 18:27:04 _av500_: that separate build is easier to move across systems (in the team) rather than the entire angstrom build.. Apr 18 18:27:21 _av500_: we looked at those and decided it's not any cheaper than going out for it :) Apr 18 18:27:38 <_av500_> mdp: well, its 0.4€ per shot Apr 18 18:27:50 <_av500_> and it tastes really good Apr 18 18:28:13 yeah, I've heard good things Apr 18 18:31:46 I use powder to make my "wiener melange" ;) Apr 18 18:32:07 <_av500_> white powder like Mr. Wecker? Apr 18 18:32:45 its brown, but that might just be some color Apr 18 18:33:09 <_av500_> i stopped drinking coffee with milk after i drank a liter of milk in the end Apr 18 18:35:06 if I need coffein, I have penguins ;) Apr 18 18:36:14 I wonder what the .us shipping cost is on a typical 50-100 capsule order from the club Apr 18 18:37:47 sorry this is more of a linux kernel question, but say i was writing some module to do interrupt handling on two GPIO pins. i want to gather data from the interrupts for 100ms or so after the first event and after the 100ms to dump it all to char device in /dev/. Apr 18 18:38:30 to speed up interrupt handling, i'd rather not to all that checking and dumping in the IRQ handler, right? i imagine setting some flag to do it later but that would mean some kind of event loop in the kernel module. is such a thing common to do? Apr 18 18:39:06 use a time Apr 18 18:39:08 r Apr 18 18:39:22 and let it start a thread Apr 18 18:39:46 ah, and initialize it on the first event? ok thanks! i'll look into using timers Apr 18 18:40:38 btw, beaglebone can do kernel-level IRQ from GPIO pins, right? Apr 18 18:41:38 yes, gpio-keys is an example Apr 18 18:42:04 awesome. thank you very much aholler Apr 18 18:45:09 <_av500_> mdp: I order 200 with free shipping :) Apr 18 18:45:46 ok, anything free there has got to be better here :P Apr 18 18:46:21 back when I first looked at them, I was surprised to find we had retail outlets here in "Real America" Apr 18 18:46:48 I originally assumed you had to be in NYC, LA, or SF :) Apr 18 18:54:06 can we enable the pullup resistor on the gpio pins user /sys fs Apr 18 18:57:44 thanks jkridner, apparently my eyes weren't working to locate that logo Apr 18 18:58:19 nobody scrolls. :) Apr 18 19:07:30 <_av500_> only the elders Apr 18 19:07:37 example of a failed boot: http://fpaste.org/e1MD/ Apr 18 19:08:40 sorry, more has come: http://fpaste.org/iWRK/ Apr 18 19:09:13 got a fast question : written up there :P with the beaglebone : can we enable the pullup resistor on the gpio pins Apr 18 19:11:14 yes Apr 18 19:11:30 using /sys? Apr 18 19:11:34 every pin has a pull up and pull down. Apr 18 19:11:55 well, it is available via debugfs right now, not sysfs. Apr 18 19:12:10 ahh ok , i gotta learn debugfs too Apr 18 19:12:12 thanks Apr 18 19:12:23 under /sys/kernel/debug ... Apr 18 19:12:26 omapmux or something. Apr 18 19:12:49 it isn't obvious as the register values are shared a bit raw. Apr 18 19:12:52 ahhh thanks ! must be done on boot Apr 18 19:12:58 ? Apr 18 19:13:09 in the mode, some of the bits are there for pull-up/pull-down. Apr 18 19:13:53 the place I've put it into code is in bonescript/eeprom.js Apr 18 19:14:06 the fields in the EEPROM match the locations in the register. Apr 18 19:14:24 jkridner: ohh thanks! i now know one more place to look in Apr 18 19:14:43 of course, the TRM is your friend! :) Apr 18 19:14:55 jkridner__, imagine, I looked through the logos you provide on the one page…and figured all the content was on that page :) Apr 18 19:15:00 TRM ? Apr 18 19:16:35 damn me that did a cape yesterday with resistors for everything Apr 18 19:16:50 see http://beagleboard.org/bone and search for Technical Reference Manual Apr 18 19:16:56 koen, looks like i can't create the repo, any chance can you delete it and recreate it.. then i'll be able to push it.. Apr 18 19:17:03 urgh.... Apr 18 19:17:05 morons. Apr 18 19:17:08 freakin' morons. Apr 18 19:17:42 jkridner: thanks Apr 18 19:17:51 jkridner: http://beagleboard.org/media#logos Apr 18 19:18:08 jkridner: in my defense Apr 18 19:18:31 jkridner: the link is broken XD but found it on ti site Apr 18 19:19:34 where can i get the rubber cases? Apr 18 19:19:40 aholler, turn off SUSPEND? Apr 18 19:20:54 thus my "morons" comment in regards to the broken link to the TRM. Apr 18 19:21:03 I'll go update the /media page. Apr 18 19:22:22 jkridner: ahhh tought was about me but i was taking it well Apr 18 19:22:25 ok, great, that will help the next person Apr 18 19:22:34 new idea, connect the reset and user-button to some gpios of the ftdi and write some sw to reset/press button by remote-usb ;) Apr 18 19:22:35 http://www.ti.com/lit/ug/spruh73d/spruh73d.pdf Apr 18 19:23:43 mdp: is that a new requirement? Apr 18 19:25:34 actually my hope is that there will be vanilla kernel sometime which I can boot when the beagle is powered by usb only ;) Apr 18 19:25:48 and not .32 ;) Apr 18 19:25:54 koen: where can i get this case? http://www.flickr.com/photos/koenkooi/4816086934/ Apr 18 19:27:59 koen builds them himself Apr 18 19:28:10 with a cnc machine? Apr 18 19:28:37 http://www.flickr.com/photos/koenkooi/4678613966/in/photostream/ Apr 18 19:29:14 ahh sweet Apr 18 19:29:49 any details on the cnc machine? I would build one just to get that case Apr 18 19:31:14 aholler, something to try is all Apr 18 19:31:51 smplman: it's a modified makerbot Apr 18 19:32:50 You can get BeagleBone cases from AdaFruit Apr 18 19:32:54 at least he once had one, that thing on the photo above looks like something else Apr 18 19:33:22 aholler, sysv init or systemd? Apr 18 19:34:43 that is openrc, but thats just one of the errors. as said, there are various boot errors. Apr 18 19:36:47 but I'll bisecting, just give me 2 days: Bisecting: 3194 revisions left to test after this (roughly 12 steps) Apr 18 19:37:28 using ktest.pl? :) Apr 18 19:44:04 no, I've just decided to stop bisecting. that uart-stuff makes it to cumbersome Apr 18 19:47:02 the needed patch doesn't apply to every commit git bisect chooses. Apr 18 19:47:17 yeah, bisecting is tough on unstable stuff like omap Apr 18 19:47:45 too many changes in the omap-tree between 3.2 and 3.3 Apr 18 19:47:53 yeah Apr 18 19:48:14 ok, off to bed Apr 18 19:53:21 mdp: steven paying you to pimp his tool? :) Apr 18 19:54:25 just another huge perl fan Apr 18 19:55:19 mdp: you use perl? Apr 18 19:59:37 I must if I use ktest.pl, right? Apr 18 19:59:52 srostedt made me blink Apr 18 20:00:03 fell off the wagon Apr 18 20:01:22 mdp: i don't know you at all anymore :P Apr 18 20:04:40 * mdp welcomes the intervention Apr 18 20:05:05 mranostay, intervention in barcelona :) Apr 18 20:08:29 heh Apr 18 20:15:20 where do I find the omap3->pmic mapping stuff for smart reflex? Apr 18 20:15:35 in mainline kernels Apr 18 20:15:55 git grep Apr 18 20:16:14 tried, but because I dont know what it looks like Apr 18 20:17:32 git grep smartreflex.h Apr 18 20:17:36 mdp: isn't in on the wagon? Apr 18 20:17:48 * mranostay hopes mdp gets the reference Apr 18 20:19:20 aholler: I dont think you understand Apr 18 20:20:03 aholler: there should be a mapping somewhere between the omap smartreflex IP block and the PMIC SMPS IP block, I know where it is in TI kernels but cant find it in kernel.org ones Apr 18 20:21:10 omap_twl.c, not the most obvious name :-( Apr 18 20:42:24 mdp: http://elinux.org/Running_OpenOCD_on_Linux_with_the_Beagleboard Apr 18 20:42:38 mdp: might compare that with your notes Apr 18 20:43:54 is there a follow up on using OpenOCD to trace the ROM on HS devices? ;) Apr 18 20:44:15 ds2: sure, i'll get that posted asap Apr 18 20:44:20 :) Apr 18 20:50:01 when I try "opkg update" on a Craneboard, i get the following error: * Failed to download http://feeds.arago-project.org/feeds/live/ipk/armv5te/Packages.gz. any tips on what a better source repo might be? Apr 18 20:51:46 thos37: try angstrom feeds, arago no longer provides binary feeds due to legal issues Apr 18 20:53:51 heh Apr 18 20:55:52 any idea about how to read 2 byte of data from I2C of beagleboard, I am reading a sensors data which are in 2's complement? and I want to read them without any dely in between Apr 18 20:56:36 hiwa, i2cget 3 0x78 0 w Apr 18 20:57:37 I know that, but the sensors data are in different register addresses, and I want to do it in C language Apr 18 20:59:04 hiwa, I'm afraid I dont know how to read two registers from an i2c sensor without any delay Apr 18 20:59:09 best of luck Apr 18 20:59:11 write a proper kernel driver Apr 18 20:59:21 and make sure the sensor is sane Apr 18 20:59:48 most of them have hw interlocks to insure sanity but it requires the reads to be done in a specific way. check datasheet Apr 18 21:02:03 I have a gyroscope sensor connected to beagleboard via i2c, the gyro values (high byte and low byte) should be read on time, otherwise the value will be wrong Apr 18 21:02:17 Any idea how to troubleshoot a situation where you get high latency for 1 in 5000 to 1 in 500 interrupts? (I'm afraid this is normal and should move to a rtos if I can't live with that Apr 18 21:02:46 that is why I want to read them together Apr 18 21:03:27 did you understand my point? Apr 18 21:04:17 yup, see what ds2 said, I think it's your best bet Apr 18 21:08:18 which gyro? Apr 18 21:08:25 what is the proper way to read from any sensor through beagleboard i2c ?? Apr 18 21:08:38 the proper way is the Linux way Apr 18 21:08:50 most gyro's have drivers already or are trivial to write a driver for Apr 18 21:08:53 any idea which machine repo, if any, I should use for the craneboard from this selection: http://feeds.angstrom-distribution.org/feeds/core/ipk/eglibc/armv7a/machine/ ? craneboard? am3517_evm? Apr 18 21:08:56 MPU6050 Apr 18 21:09:22 strictly speaking that is not a plain old gyro Apr 18 21:09:31 yes Apr 18 21:09:34 why are you not using the official sw? Apr 18 21:09:38 that is the problem Apr 18 21:10:04 it is not only a gyro Apr 18 21:10:22 you need more then just i2c access to make that part work as a gyro w/o the official sw Apr 18 21:10:40 avoid that part if you are just looking for gyro and want to not go bald Apr 18 21:10:42 all this talk about gyros is making me hungry Apr 18 21:11:34 I am doing master thesis, and I can't simply give it up Apr 18 21:11:40 there are lots of gyros around Apr 18 21:12:01 but we need that IMU Apr 18 21:12:15 since you are in academia... please define what is an IMU Apr 18 21:12:43 sorry , that is Inertial Measurement Unit Apr 18 21:12:56 i googled faster than that Apr 18 21:12:59 no. not what the letters stand for... what are YOU considering an IMU Apr 18 21:13:20 cuz saying you need an IMU and insisting you need the MPU6050 seems a bit absurd Apr 18 21:13:28 now if you are working for the guys that make it... Apr 18 21:13:52 no, I am not Apr 18 21:14:07 I am working on north finding system Apr 18 21:14:12 then define what makes an IMU an IMU Apr 18 21:15:44 could you please tell how to read the register values in a correct way and on time Apr 18 21:15:58 w/o the magic sauce that won't help Apr 18 21:15:59 which function to use? Apr 18 21:18:22 I have 4 of them each 2 on i2c of beagleboard Apr 18 21:18:47 since I can have 2 addresses for each sensor Apr 18 21:21:22 hiwa - read about linux iio, it has several gyro/accelerometer/imu drivers, you can look at how they work Apr 18 21:21:40 what is iio? Apr 18 21:21:44 google it Apr 18 21:22:38 I've added some angstrom repos to the craneboard, and used the beagleboard machine repo, but when I try to install avahi-daemon, I get an error that one of the packages is not a compatible architecture ... http://pastebin.com/bW4i01wi which begs the question what IS the architecture of the craneboard or how do I find out? uname -a : Linux am3517-crane 2.6.32 #13 Thu Mar 15 21:18:03 IST 2012 armv7l unknown Apr 18 21:25:26 Is it possible to disable the gp timer? Apr 18 21:26:42 hiwa: this might help: http://elinux.org/BeagleBoard_Trainer_Nunchuk Apr 18 21:27:04 thos, from wha tyou postedI would guess you need arm7l, not arm7a Apr 18 21:30:08 I already read this page before.. I have a diffrent case Apr 18 21:31:13 but my question is how to read data from beagleboard i2c ? which function to use in C??? Apr 18 21:32:49 the craneboard has an AM3517 chip Apr 18 21:33:06 as far as the cpu goes, it's the same as the beagle Apr 18 21:41:20 why is my xm hanging at ehci-omap.0 supply hsusb1 not found, using dummy regulator ? Apr 18 22:31:34 koen: https://github.com/prpplague/fb-test-app Apr 18 22:31:40 koen: new and improved Apr 18 22:35:28 so when i want to use gpio1_16 i need to export it as 48 because that's the hex number, right? so if i was trying to access that GPIO from a kernel module, would it also be 48 there? Apr 18 22:40:12 prpplague, help lol Apr 18 22:41:25 prpplague: how can something be both new and improved Apr 18 22:43:19 mru can't you have a v2 but it not be better than the first :P ? Apr 18 22:52:33 new means it's something that didn't exist before Apr 18 22:52:50 improve is something you do to existing things Apr 18 22:54:31 mru, it combines new and improved things? Apr 18 22:54:45 perhaps it was new, and then it was improved before they told anyone about the new Apr 19 00:04:23 huh, there is no linux/module.h in the kernel-headers package? Apr 19 00:34:49 nemik: why would there be? Apr 19 00:36:56 koen: ping Apr 19 00:49:36 neato http://i.imgur.com/vw1SE.png Apr 19 01:14:15 hmmm..I have a really strong desire to type 'root' into my mplayer window Apr 19 01:14:20 maybe there is a way to make that happen Apr 19 01:24:56 ok, so plugging in a USB keyboard does it...Is there a host<->gadget usb thingy Apr 19 01:34:28 is it safe/possible to do a dist-upgrade from 11.10 to 12.04? Apr 19 01:35:05 Russ: host to gadget? Apr 19 01:35:42 yes, I don't see drivers for any Apr 19 01:35:45 but it'd be handly Apr 19 01:35:47 er, handy Apr 19 01:36:06 I have a netchip pci card I can use, but that'd only be good for one device Apr 19 01:36:22 Russ: sorry i don't understand what you are asking for Apr 19 01:36:48 I want to plug in a virtual usb keyboard to my beagle board and also other devices Apr 19 01:38:27 a device that is a microcontroller with two usb gadget interface would be a really cool way of doing that, from one side, it'd be connected to a USB host port of a linux system that would drive the other gadget interface with the linux gadget api Apr 19 01:38:40 Russ: ahh like have your desktop pc pretend to be a usb keyboard plugged into the usb host port of the other linux device like the beagle Apr 19 01:38:58 right, and it could change based on what you need, USB mass storage, network, etc Apr 19 01:39:24 Russ: we've had that working in the past on the nail board Apr 19 01:39:45 http://elinux.org/Nail_Board Apr 19 01:40:42 it would need two usb slave ports Apr 19 01:41:21 ah, a leftover from the tuxscreen days makes a cameo in your images section Apr 19 01:41:27 Russ: yea we had a variation like that Apr 19 01:42:00 Russ: oh which item? Apr 19 01:42:32 the compact flash wireless card Apr 19 01:42:49 Russ: ahh that was from the old AML M7100 Apr 19 01:43:15 ah, I had bought the same one for use with my tuxscreen Apr 19 01:43:51 yea they were pretty popular for a long time Apr 19 01:44:31 Russ: we had a version of the nail with two usb slave ports Apr 19 01:45:03 maybe I can find a controllable USB switch for using my single usb gadget controller with multiple devices Apr 19 01:45:09 Russ: there are few bridge chips that are made for doing data syncs Apr 19 01:45:19 prpplague, wonder how hackable they are Apr 19 02:06:15 wonder if a 743253 would function as a good usb multiplexer (dual 1-of-4- fet multiplexer) Apr 19 02:07:19 LS or FS or HS? Apr 19 02:13:44 probably cbt Apr 19 02:17:39 why is my xm hanging at ehci-omap.0 supply hsusb1 not found, using dummy regulator ? Apr 19 02:21:19 we don't know Apr 19 02:21:40 post your full log, to pastebin, maybe someone can look at it for you Apr 19 02:27:50 there is no log lol thats the issue Apr 19 02:29:13 is there something I can copy off the sdcard thats automatically getting created? Apr 19 02:31:01 capture the output of the on the serial port Apr 19 02:31:23 kk Apr 19 02:31:34 have minicom or whatever you're using set to log it Apr 19 02:44:11 thurbad http://pastebin.com/FetbB40v Apr 19 02:45:56 is there supposed to be more detail than that default Apr 19 02:46:03 the xm has no nand Apr 19 02:46:28 this kernel seems to be trying to use nand Apr 19 02:46:55 odd Apr 19 02:47:12 kk, Ill download another image and try again Apr 19 02:49:53 not sure which android port has the best xm support, but you may be able to ask around and save yourself some time Apr 19 02:57:58 thurbad, this is the one that TI supplies on their site for it Apr 19 02:58:30 the rowboat one? Apr 19 02:59:22 this rev A2-01 does have onboard nand "which cant boot from NAND but has NAND present", yes rowboat Apr 19 02:59:46 I will try and see if its erasable tho, maybe something got in there some how ? **** ENDING LOGGING AT Thu Apr 19 02:59:59 2012