**** BEGIN LOGGING AT Wed Mar 21 02:59:58 2012 Mar 21 03:09:00 without a doubt, yes. Mar 21 03:09:25 but there are too few devices available where one can play directly with the metal... Mar 21 03:09:46 i should qualify that with 'fun'. Mar 21 03:09:49 too few fun devices. Mar 21 05:04:20 does building qt4-embedded give me qmake? Mar 21 05:05:54 oh yeah, i see it Mar 21 05:08:27 it's not actually a usable install though... Mar 21 05:08:36 JViz, yeah you are supposed to use this version of qmake over the one that might be available for the native platform Mar 21 05:09:43 i'm trying to point qt-creator to the one angstrom used to build it, so i can use it to build my project against Mar 21 05:10:34 if i build qt4-embedded-tools, it'll put it in the sysroot? Mar 21 05:12:54 weird Mar 21 05:15:55 sundar: what do you mean by "this" version of qmake? Mar 21 05:17:08 JViz, I assumed that you were compiling qt4-embedded for the beagleboard. when you do so, you get qmake which will generate Makefile taking the cross-compiler tolls into account. Mar 21 05:17:25 tools* Mar 21 05:17:59 so if you do make followed by a qmake, you compile for the target Mar 21 05:19:10 suppose if you had installed an x86/x86_64 version of Qt4 in your development machine, and if you use this qmake instead with the wrong specs then it may not generate binary that can be run on beagleboard Mar 21 05:19:22 i ended up with a few different qmakes Mar 21 05:19:29 all in work directories Mar 21 05:20:13 i think you should use the qmake that was generated after you compile Qt4 embedded. Do not mix with the native ones. Mar 21 05:20:28 where would that be? in a sysroot? Mar 21 05:21:00 while configuring Qt you would give the location in -prefix right? Mar 21 05:21:21 i just ran bitbake qt4-embedded Mar 21 05:22:21 i am not familiar on bitbake, but should be configured automatically in this case Mar 21 05:22:38 oh i know what you're saying now Mar 21 05:24:01 yeah, so i should try to find the config that angstrom built the local packages for, and build it on my native box seperately, with a slightly modified config? Mar 21 05:25:08 i guess modifications should not be necessary. just check where the built package is available. Mar 21 05:25:24 there should not be more than one qmake Mar 21 05:26:30 if you compile Qt4 development package instead, i think it might generate packages that runs on ARM Mar 21 05:26:49 so as long as i get the mkspecs right for my local version of qt4-embedded, then it should run right on the beagle? Mar 21 05:26:59 it's like having the beagleboard as both the development and target machone Mar 21 05:27:08 yes, it should Mar 21 05:27:18 sweet Mar 21 05:27:24 gonna give that a shot now Mar 21 05:27:26 wish me luck Mar 21 05:27:29 thank you Mar 21 05:27:38 sure, good luck :) Mar 21 07:04:40 Hi all Mar 21 07:51:48 great, cpu misdetection in linux Mar 21 07:52:31 hi Mar 21 07:58:14 hi Mar 21 08:05:20 can any body help me out about usb boot Mar 21 08:13:46 Dipto_: you might want to explain a bit more detailed your problem. (don't bother to ask me though, I can't help with it) Mar 21 08:13:58 hi Mar 21 08:14:25 I want to Boot the board with a pendrive. is that possible? Mar 21 08:14:31 i'm getting pretty decent downloads on my 3G hookup to the beagle board Mar 21 08:14:40 using laptop as a router Mar 21 08:15:11 satellite connection only had 2K byte /sec lol Mar 21 08:19:57 Dipto why not boot from the SD? that is the standard way Mar 21 08:21:45 yes, but I want to do that way. The mmc will not be present, I have to boot only using usb. Mar 21 08:22:38 is this for a class assignment or something like that? Mar 21 08:23:10 yes.. something like that.. Mar 21 08:31:36 Dipto_: talking about which board? Mar 21 08:31:51 BeagleBoard Rev C4 Mar 21 08:32:07 you cannot boot from a usb stick Mar 21 08:32:14 you can boot from usb via a host PC Mar 21 08:33:04 so there is no way to boot from usb stic Mar 21 08:33:25 Dipto_: well the beagleboard has nand, it should be possible to have the bootloader in nand, eventually also the kernel and have the rootfs on usb. Mar 21 08:33:32 even by modifying u-boot or x-loader? Mar 21 08:35:14 where are you going to put xloader? Mar 21 08:35:26 if it's not on the usb stick, you aren't booting from it, no? Mar 21 08:35:35 i can flash the x loader to the nand Mar 21 08:36:09 but that's booting from nand, no? Mar 21 08:36:38 ya.. but the u-boot will be in the usb stick Mar 21 08:36:45 no. Mar 21 08:37:23 sure, add a usb stack to x-loader and load uboot from usb stick Mar 21 08:37:52 Dipto_: you can have the kernel and rootfs on the stick, eg. start a linux from the stick. you can _not_ have the bootloader on usb. Mar 21 08:38:11 How to add usb stack to xloader? i am new to these Mar 21 08:38:36 (except for arcane coding skillz applied, which are rare to find and you obviously don't possess ) Mar 21 08:39:02 Dipto_: if you have to ask, you can't Mar 21 08:39:06 If it is possible to start linux from usb stick, with out the sd card, please tell me the way Mar 21 08:39:09 Dipto_: the usb stack was a joke. you are not able to do this. Mar 21 08:41:12 i've got 226MB updates to ubuntu downloading Mar 21 08:49:06 i'm using my novatel 2200 MiFi box (supports 5 wifi devices), the laptop is using the wifi and connection sharing over a cat5 to the beagle board Mar 21 09:02:18 Dipto_: no queries, please. Mar 21 09:02:51 ok, thanks Mar 21 09:04:27 Dipto_: and its been said all already - ahve xloader and uboot in nand, have them boot to usb. modify the scripts accordingly, there should be enough examples, its very similar to booting from mmc after all. Mar 21 09:04:45 or just boot from sdcard Mar 21 09:04:53 and be done Mar 21 09:04:54 Thanks Mar 21 09:05:13 (you said its something lika class assignment, and i'm not gonna do your homework. go read the docs yourself, its part of the learning process :P) Mar 21 09:05:33 or: http://26-26-54.hardwarebug.org/43 Mar 21 09:06:36 <3 http://26-26-54.hardwarebug.org/110 Mar 21 09:07:40 av500: fun problem: find out if the kernel is wrong, TI programmed the efuses wrong or that the disty gave us the wrong chips Mar 21 09:08:04 [ 0.000000] OMAP3630 ES1.0 (l2cache iva sgx neon isp 192mhz_clk ) Mar 21 09:08:04 vs Mar 21 09:08:08 [ 0.000000] OMAP3630 ES1.2 (l2cache neon isp 192mhz_clk ) Mar 21 09:09:33 so 1.2 has no iva and sgx :) Mar 21 09:09:44 nice upgrade Mar 21 09:09:56 it looks like an AM370x Mar 21 09:10:06 instead of a DM3730 Mar 21 09:11:31 the software just parses the register Mar 21 09:11:52 I'm 90% sure the kernel isn't wrong Mar 21 09:12:04 you are holding it wrong Mar 21 09:12:24 so it's either the efuses that set the register bits or some physical mixup Mar 21 09:12:49 koen: that when you buy "factor floor sweeps" Mar 21 09:12:59 that's Mar 21 09:15:03 yeah Mar 21 09:43:39 well, at least it does have l2cache Mar 21 09:44:01 so it's still > rpi :p Mar 21 09:44:28 ah, no sgx, bummer then, you're doomed Mar 21 09:44:58 *DOOM* Mar 21 09:45:03 piratebay won't consider your kitty for their drones Mar 21 09:50:58 http://www.humanbirdwings.net/ - how udderly ubercool is this? :D Mar 21 09:55:12 lol Mar 21 09:56:53 what a summer this will be - sky full of flying people! :) Mar 21 09:57:26 and add some alcohol || dope... Mar 21 09:57:31 he even takes off from level ground. Mar 21 09:57:58 imagine he had the pirate bay database and a wifi hotspot with him.... Mar 21 09:58:15 now, for a way to make this bb-related... Mar 21 10:00:43 * av500 considers buying: http://www.pollin.de/shop/dt/MjE5OTgxOTk-/Bausaetze_Module/Bausaetze/Bausatz_SUB_D_Anschlussplatine.html Mar 21 10:02:08 no soldering, that's boring Mar 21 11:05:40 is the arago site down? Mar 21 11:38:49 jipi, http://www.downforeveryoneorjustme.com/arago-project.org (appears down) Mar 21 11:40:18 bradfa: thx.. Mar 21 12:27:51 Hi Mar 21 12:28:48 I am having some problems reading from the analog input ports of the beaglebone Mar 21 12:29:25 has someone experienced buffered problems from ain reading ?? Mar 21 12:44:10 how does one buffer problems? Mar 21 12:55:32 av500, unresolved ones can be considered to be buffered Mar 21 12:56:29 I use a EIFO type for my problem buffer Mar 21 12:56:41 EIFO? Mar 21 12:56:53 Easy In First Out, of course Mar 21 12:57:06 :) Mar 21 12:57:50 help! during the execution of "bitbake virtual/kernel -c menuconfig" i return this error "NOTE: Running task 190 of 395 (ID: 204, virtual:native:/home/gia/Desktop/oe/openembedded/recipes/gmp/gmp_5.0.1.bb, do_qa_configure) ERROR: fork failed: 12 (Cannot allocate memory)" Mar 21 12:58:05 for me, CINGO is the type then. Complex In Never Gets Out Mar 21 12:58:20 I use PWP Mar 21 12:58:34 problem? what problem? Mar 21 12:58:39 lol Mar 21 12:58:53 i don't see any problem in using that :) Mar 21 12:59:18 there's also the Not My [Fine] Problem buffer Mar 21 12:59:32 I like to mux in an appropriate buffer algorithm as necessary though Mar 21 13:00:13 just set up an SEP field and be happy Mar 21 13:01:10 Guest1376: well it says out of memory. Mar 21 13:01:45 always install as much memory as the motherboard will take Mar 21 13:02:30 .. and having some swap ready for emergencies also helps sometimes. Mar 21 13:02:50 swap is for wimps Mar 21 13:03:04 mru: ram only is for posers. Mar 21 13:03:23 *sing* wimps ans posers, leave the hall! Mar 21 13:04:05 LetoThe2nd: sung like a true brother! Mar 21 13:04:16 hello Mar 21 13:04:25 cruel world Mar 21 13:04:32 tnx Mar 21 13:04:38 I'm leaving you today Mar 21 13:12:51 hey. i am having a problem with loading my beagle board. the Nand is probably screwed up as it gives Error: reading boot sector, ECC failed, page 0x00080a00 Mar 21 13:13:08 and if i press user button while reset is says 40W Mar 21 13:13:31 then it hangs at Loading u-boot.bin from mmc Mar 21 13:13:52 can someone pls help me out Mar 21 13:16:00 what bb revision? Mar 21 13:16:27 revc4 i guess :) Mar 21 13:16:50 yea Mar 21 13:17:15 can you enter the uboot command line? Mar 21 13:17:23 if yes, do a nand erase Mar 21 13:17:23 no Mar 21 13:17:44 you have an sdcard inside? Mar 21 13:17:50 yea Mar 21 13:18:21 well, holding the user button should make it boot from SD Mar 21 13:18:43 well as i said it hangs on "Loading u-boot.bin from mmc" Mar 21 13:19:03 please pastebin a full boot log Mar 21 13:20:58 just a sec Mar 21 13:21:29 and nothing else? Mar 21 13:21:33 oops Mar 21 13:21:36 undo Mar 21 13:24:19 how can I connect to the beaglebone having a direct ethernet connection with it (no router or switch)? Mar 21 13:25:28 rucketho@lear:~# ps aux|grep wget Mar 21 13:25:28 rucketho 3396 0.4 0.7 67488 28940 pts/2 S+ 14:58 0:06 wget -e robots=off -k --mirror http://us0XXXXX.eu.tieto.com:8181/ Mar 21 13:25:31 rucketho@lear:~/foobar$ find .|wc -l Mar 21 13:25:32 ohs: make sure that the other device has specification-compliant gbit ethernet AND/OR use cross-patchcable Mar 21 13:25:33 grrrr Mar 21 13:25:35 4855 Mar 21 13:25:35 pastefail Mar 21 13:26:38 http://pastebin.com/tqSxNaeu Mar 21 13:26:56 yeah it has necessary connection and cables. but I dont know what to do next Mar 21 13:27:04 I can not find its IP Mar 21 13:27:12 is there a way to see its IP? Mar 21 13:27:43 ohs: how would it obtain one? either have the other device provide dhcp, or set one manually. Mar 21 13:28:15 I tried to set static IP but it did not work. It stills get IP from DHCP server Mar 21 13:28:30 revc4: ok, so xloader is loaded from the card, that is good Mar 21 13:28:39 do you know is there any small program that acts as a DHCP server to assign an IP to the bone? Mar 21 13:28:52 revc4: now, do you have u-boot.bin on the card? Mar 21 13:29:27 ohs: there certainly are, but i don't feel like googling for you. Mar 21 13:29:51 yes Mar 21 13:31:54 I googled it but I couldnt find it, I wonder if anyone knows the solution Mar 21 13:33:43 is it possible to find a toolchain for angstroms distribution for windows environment? Mar 21 13:34:03 ohs: no. Mar 21 13:34:38 av500: do u have any idea what i could do to solve this problem Mar 21 13:36:50 revc4: could you please try MLO and u-bott.img from here: http://www.angstrom-distribution.org/demo/beagleboard/ Mar 21 13:39:04 dammit I have willie nelson / julio iglesias "to all the girls I've loved before" stuck in my head. Friggin 93.3. Mar 21 13:39:38 * LetoThe2nd hands over a Meshuggah cd to SilicaGel Mar 21 13:40:05 do i just directly copy onto the card. or do i have to partition and set up the card Mar 21 13:40:40 just copy Mar 21 13:40:46 overwrite Mar 21 13:52:38 i got this when booting from mmc(on mmc there is only MLO, uboot.bin and uImage) http://pastebin.com/grcPAUXj Mar 21 13:54:39 hmm Mar 21 13:59:08 does anybody know if bitbake can have a SRC_URI that just pulls directly out of github? Mar 21 13:59:23 It awnts the URI to include ;md5sum= Mar 21 13:59:26 so I'm not sure how to do that :( Mar 21 14:00:07 SilicaGel: just make a tgz from github :) Mar 21 14:00:15 and add patches Mar 21 14:00:23 and re-upload it as a 'download' on github? Mar 21 14:00:29 \o/ Mar 21 14:00:43 call it *_psp_* something Mar 21 14:01:11 upside-download Mar 21 14:02:53 av500: any ideas?? Mar 21 14:03:28 no Mar 21 14:03:32 how do you power the board? Mar 21 14:04:05 usb mini Mar 21 14:04:20 hmm Mar 21 14:04:43 could you power it from 5v dc-in? Mar 21 14:05:15 will try Mar 21 14:05:29 just to rule out stuff Mar 21 14:05:42 koen: ^^^ Mar 21 14:05:57 i have a wierd problem with dropbear Mar 21 14:06:25 I've had many weird problems with that Mar 21 14:06:28 i can start it up when logged into serial, but then it stops when i disconnect from serial Mar 21 14:06:41 how are you starting it? Mar 21 14:06:57 maybe it's still attached to the tty and dies when it goes away Mar 21 14:07:18 etc/init.d/dropbear start Mar 21 14:07:44 also etc/init.d/dropbear start & Mar 21 14:08:15 why would the init script attach it to the tty? Mar 21 14:08:21 no reason Mar 21 14:08:28 but you didn't say how you were starting it Mar 21 14:08:58 maybe because of the '&'? it should background itself automagivcally. Mar 21 14:09:16 s/background/daemonize/ Mar 21 14:09:19 i tried it both ways Mar 21 14:09:25 same behavior Mar 21 14:09:27 LetoThe2nd: if it does, the & should make no difference Mar 21 14:09:36 ok Mar 21 14:11:17 av500: My problem yesterday with SCHED_FIFO and TIImgenc1 was the CONFIG_RT_GROUP_SCHED option. Turning that off fixed it. Mar 21 14:13:17 chrisw957: ic Mar 21 14:14:25 av500: same error with external supply as well Mar 21 14:24:27 i replaced dropbear with sshd, now all is well Mar 21 14:24:31 thank you Mar 21 14:24:33 bbl Mar 21 14:24:34 SilicaGel: koen@dominion:/OE/tentacle/sources/meta-ti$ git grep github.com | grep SRC Mar 21 14:24:37 recipes-kernel/linux/linux-ti33x-psp_3.1.bb:SRC_URI += "git://github.com/beagleboard/linux.git;branch=${BRANCH} \ Mar 21 14:24:40 recipes-misc/payload/beaglebone-getting-started.bb:SRC_URI = "git://github.com/jadonk/beaglebone-getting-started.git" Mar 21 14:24:43 recipes-ti/beagleboard/beaglebone-tester.bb:SRC_URI = "git://github.com/koenkooi/validation-scripts.git;protocol=git \ Mar 21 14:25:14 koen: any idea: http://pastebin.com/grcPAUXj Mar 21 14:26:00 koen: holy crap Mar 21 14:26:21 koen: thanks! Mar 21 14:27:34 koen: I'm trying to make my PRU debugger build as bitbake so taht I can set a build dependency of ncurses and possibly the panels library Mar 21 14:27:42 koen: this sentence no tgz! Mar 21 14:33:26 does watching Angstrom compile count as "working" ? Mar 21 14:33:34 * SilicaGel stares off into space Mar 21 14:33:38 303 Mar 21 14:33:51 depends are you filling out timesheets :P Mar 21 14:35:34 * mranostay yawns Mar 21 14:35:50 s/yawns/compiles/ Mar 21 14:36:02 * av500 compiles a lot Mar 21 14:36:33 av500: http://xkcd.com/303/ Mar 21 14:37:04 [15:33:37] 303 Mar 21 14:38:03 is av500 a bot today? Mar 21 14:40:11 today? Mar 21 14:40:48 the av500 bot is just not communicating in "most interesting man" memes today…switched to xkcd mode Mar 21 14:43:30 koen: it shouldn't matter that my project has no 'configure' script right? it's just a makefile Mar 21 14:43:47 (well, and a pile of source) Mar 21 14:44:19 SilicaGel: what, you dont test for fortram compile? Mar 21 14:44:26 or that 1+1 == 2? Mar 21 14:44:27 precisely Mar 21 14:44:30 heh Mar 21 14:44:46 you SW will never run on an unpowered PDP11 Mar 21 14:44:54 maybe i should have a configure script that tests for ncurses and all that, but I dno't know how to make one Mar 21 14:45:05 i'll whatever av500 is drinking this morning :P Mar 21 14:45:15 what if it's arsenic Mar 21 14:45:21 SilicaGel: and what are you going to do if it's not there? Mar 21 14:45:32 if you're just going to fail, you might as well do that when compiling Mar 21 14:45:34 mru: cry? rm -rf / ? Mar 21 14:45:46 mru: yeah. ok. sounds reasonable to me. Mar 21 14:46:02 checking for stuff only makes sense if you can do something useful with or without it Mar 21 14:46:09 yeah Mar 21 14:46:44 well the bitbake way I say DEPENDS= and add panel and ncurses and I think it should take care of thing sfor me .. I hope Mar 21 14:47:02 mru: in that respect checking for fortran compiler is ok Mar 21 14:47:06 if not, then I don't really need bitbake I suppose. I can just say "modify the makefile to point to your cross compiler and type make" Mar 21 14:47:33 av500: fine, checking makes sense if you need to do something differently depending on what you find Mar 21 14:47:45 * mranostay goes back to digging around for a smoker detector to disassemble Mar 21 14:47:49 I am trying to convince myself that this is all worth it. Setting up the bitbake/oe environment takes a long as stime and a lot of disk space Mar 21 14:48:24 SilicaGel: disk space is a concern? Mar 21 14:48:34 heh is this 1999? :) Mar 21 14:48:47 must be Mar 21 14:48:50 not so much. It's more that if someone else wants to build this they have to go through all this too Mar 21 14:48:54 do you know what I mean Mar 21 14:48:58 mranostay: smokers can usually be found in the back gardens of pubs Mar 21 14:49:13 SilicaGel: yocto is easier… http://vimeo.com/24743626 Mar 21 14:49:35 like "here's a 300 line long C program. But to build it, first you have to download a gigabyte worth of crap. And then compile it all, which takes 4 hours. And uses up 24G of your disk." Mar 21 14:49:57 jwinnebeck keeps telling me this is worth it. I'm just not sure. Mar 21 14:49:59 SilicaGel: that all depend on who your users are Mar 21 14:50:05 why would you make it _require_ OE? Mar 21 14:50:11 just write a damn makefile Mar 21 14:50:30 yeah. that's what I have now. A damned makefile. :) Mar 21 14:51:21 and a readme that says "Modify CC= and MACHINE= to match your target" Mar 21 14:51:28 well, that's paraphrased, but you get the idea Mar 21 14:51:45 mru: the radioactive ones still legal in the .eu? Mar 21 14:51:51 Illegal!?!!? Mar 21 14:51:58 radioactive smokers? Mar 21 14:51:59 mranostay: radioactive smokers? Mar 21 14:52:02 How do european smoke detectors work without a souce of ionization? Mar 21 14:52:06 mru: u lag! Mar 21 14:52:20 SilicaGel: optical Mar 21 14:52:23 Are they optical only? or rate of rise? Mar 21 14:52:25 huh. Mar 21 14:52:36 SilicaGel: they have ones that don't use an ion chamber Mar 21 14:52:49 Do they work as well? Mar 21 14:52:55 namely people that view anything nuclear as bad :) Mar 21 14:53:03 and no i've heard they don't Mar 21 14:53:45 d Mar 21 14:53:45 NOTE: Running task 477 of 1174 Mar 21 14:53:48 haha I love it. Mar 21 14:56:16 mranostay: the one in my flat claims it's radioactive Mar 21 14:56:23 run! Mar 21 14:57:17 i like Berkeley's Nuclear Free Zone city ordinance Mar 21 14:57:51 yeah and i was looking forward to testing nuclear weapons Mar 21 14:58:04 koen/av500: http://pastebin.com/YWJqWAtT (i took a previously working sd card and replaced the boot files(mlo,uboot,uimage) from http://www.angstrom-distribution.org/demo/beagleboard/) Mar 21 14:58:37 no Mar 21 14:58:40 that boots from nand Mar 21 14:58:46 Texas Instruments X-Loader 1.4.2 (Feb 19 2009 - 12:01:24) Mar 21 14:58:51 2009 -> NAND Mar 21 14:59:16 the xloader on SD crashes/hangs for some reason Mar 21 15:05:49 av500: but the first time it tried to load from sd card Mar 21 15:06:07 yes, and that crahses Mar 21 15:08:19 yeah Mar 21 15:09:42 also if i remember correctly both the lights(USR0 USR1) blink when booting..but now only pwr light turns on. Mar 21 15:09:55 nothing blinks until the kernel is up Mar 21 15:10:07 sometime I think these leds do more bad than good Mar 21 15:10:23 only when they are lit Mar 21 15:10:36 so when linux is running, it's evil only part of the time Mar 21 15:11:01 it makes all the people report meaningless LED patterns Mar 21 15:11:19 beep codes ftw Mar 21 15:11:38 yeah, add them to a bug report :) Mar 21 15:12:49 av500: LOL Mar 21 15:15:56 led patterns only if they are morse code! :) Mar 21 15:16:28 or use 6 leds and display messages in braille Mar 21 15:17:08 blind people have optical sensors in fingertips these days? Mar 21 15:18:24 I'm sure some kind of implant would be possible Mar 21 15:18:40 in that case implant RS232 Mar 21 15:18:50 can talk to a beagle directly Mar 21 15:19:35 helloooo ppl...i have a problem here. i would appreciate some help.. my major project is stuck cuz of this Mar 21 15:20:01 revc4: did you switch personality? Mar 21 15:20:22 what makes u say that? Mar 21 15:20:38 I dunno, feels like you just reset Mar 21 15:20:59 and had blond hairs before Mar 21 15:20:59 since we talk about your "problem" already.... Mar 21 15:21:36 i reset the board so many times.. maybe i reset myself by \mistake Mar 21 15:21:48 * mdp prepares to determine if the problem matches his Easy filter Mar 21 15:22:03 42! Mar 21 15:22:11 revc4: [15:58:03] koen/av500: http://pastebin.com/YWJqWAtT (i took a previously working sd card and replaced the boot files(mlo,uboot,uimage) from http://www.angstrom-distribution.org/demo/beagleboard/) Mar 21 15:22:24 what do you mean with "previously working"? Mar 21 15:23:08 oh..i meant that before it used to work(like last month)..but today it stopped working suddenly Mar 21 15:23:38 but that was booting from nand I guess Mar 21 15:24:00 and since that stopped suddenly to work Mar 21 15:24:08 I guess something "struck" your board Mar 21 15:24:24 i hadnt used the board for quite a while Mar 21 15:24:44 do you've cat? Mar 21 15:25:05 no. Mar 21 15:25:14 dog? Mar 21 15:25:38 hm, then it's not cat urine Mar 21 15:25:48 Wookie spit? Mar 21 15:25:49 * ynezz moves to the next point in his problem solving diagram Mar 21 15:26:16 i don't think poor wokkie has got something to do with it. Mar 21 15:26:25 revc4: do you have another card to try? Mar 21 16:33:24 Hi all. When I give bitbake virtual/kernel -c menuconfig, new terminal flashes infront of me and dissapears. Why?? Mar 21 16:34:03 it hates you Mar 21 16:34:16 well Mar 21 16:34:35 * mranostay been around av500 too long Mar 21 16:34:39 *has Mar 21 16:36:14 learned his trolling from the master. Mar 21 16:37:21 tess: in gnome terminal or something? Mar 21 16:39:24 gnome terminal Mar 21 16:40:46 or xterm, i tried both Mar 21 16:41:07 no the menuconfig window that pops up Mar 21 16:41:32 install a popup blocker Mar 21 16:43:04 I tried to modify bitbake.conf for use xterm and gnome-terminal Mar 21 16:44:27 * florian dislikes this feature badly since you are usually running bitbake in your preferred terminal anyway Mar 21 16:44:51 ... not speaking its "usefulness" for remote connections Mar 21 16:51:04 when terminal return: NOTE: package linux-omap-2.6.39-r102o: task do_menuconfig: Started -> run menuconfig gnome-terminal . Mar 21 16:51:41 when terminal return: NOTE: package linux-omap-2.6.39-r102o: task do_menuconfig: Succeeded the terminal closed Mar 21 16:53:25 bitbake forks a new terminal for menuconfig? Mar 21 16:53:35 yes Mar 21 16:53:44 how silly Mar 21 16:53:58 what happens if you run it in a text console? Mar 21 16:54:09 it forks a screen window? Mar 21 16:54:39 maybe it orders a new monitor from amazon Mar 21 16:55:01 or it gets the hose again. Mar 21 16:55:04 http://upload.wikimedia.org/wikipedia/en/4/49/DESQview_2.8_screenshot.jpg Mar 21 16:55:08 maybe it does that Mar 21 17:00:55 i don't know, i'm newbie Mar 21 17:03:29 * koen always runs menuconfig by hand Mar 21 17:03:39 e.g. http://www.slimlogic.co.uk/2011/05/openembeddedangstrom-kernel-workflow/ Mar 21 18:16:26 good grief, koen ... but thanks for documenting that Mar 21 18:16:50 koen: ping Mar 21 18:24:05 i just built the toolchain Mar 21 18:24:11 bitbaked I mean Mar 21 18:24:30 now i'm trying to figure out how the hell to make it run my new .bb Mar 21 18:26:31 i am having difficulty concentrating, thanks to drinking way too much caffeine today, and the fact that we broke a record high temperature here and it's the most beautiful day in YEARS Mar 21 18:28:06 sounds like you need some beer to counteract the caffeine Mar 21 18:28:49 I'm not sure that works? Mar 21 18:28:58 but you can try and find out Mar 21 18:29:12 I must have to put my .bb some magical place Mar 21 18:29:20 do I need to make a meta-chrisp ? Mar 21 18:34:16 <_av500_> get real crisps Mar 21 18:35:27 * mranostay smacks _av500_ Mar 21 18:35:55 with a trout? Mar 21 18:36:16 sure why not Mar 21 18:36:33 then he can have fish & chips Mar 21 18:37:21 mmmm the one english dish i like Mar 21 18:39:43 <_av500_> mru: fish & crisps? Mar 21 18:39:55 * mranostay steal a Guinness from mru's fridge Mar 21 18:40:22 what? you don't like mushy peas and blood pudding? :P Mar 21 18:40:23 impossible, I didn't have any guiness Mar 21 18:41:04 we now have that exquisite dish of pink slime Mar 21 18:41:25 that seems to be falling out of fashion now Mar 21 18:41:31 maybe if you eat at fast food joints Mar 21 18:41:32 or school cafeteria Mar 21 18:41:33 there's a reasonable selection of other beer though Mar 21 18:41:49 <_av500_> not falling out of meat machines? Mar 21 18:41:49 think of it as soylent pink Mar 21 18:41:55 ew Mar 21 18:42:01 hmm, lets guess the beer in mru's fridge Mar 21 18:42:13 St poly girl? Mar 21 18:42:33 MGD 64bits Mar 21 18:42:45 good one Mar 21 18:42:59 css dry Mar 21 18:43:45 or CSS Lite :) Mar 21 18:44:31 india pale arm? Mar 21 18:44:59 this is even causing *me* pain. Mar 21 18:45:17 there are a couple of IPAs there, indeed Mar 21 18:45:25 HEX code brew Mar 21 18:45:36 i think we all need to get out more Mar 21 18:45:54 omg, a social life,, horrors Mar 21 18:46:09 if we all go out together it won't be so bad Mar 21 18:46:16 lol Mar 21 18:46:19 we'll just take #beagle to the pub Mar 21 18:46:48 and toss around a bone Mar 21 18:46:52 yeah i'm sure we'll be a hit with the ladies :)... Mar 21 18:47:12 * mranostay pictures all the macbooks on the bar Mar 21 18:47:22 they'll be fine with it onc we buy them a boneweiser Mar 21 18:47:47 well, there was that blonde in that bar in SF... Mar 21 18:47:53 * djlewis trying to not laugh out load here in the office Mar 21 18:47:53 so don't count it out Mar 21 18:48:01 Hey, don't talk about merlyn lik ethat Mar 21 18:48:04 mru: story? Mar 21 18:48:41 in sf a good looking blonde might be either sex :( Mar 21 18:50:13 heh Mar 21 18:50:25 there was this rather drunk swedish blonde there Mar 21 18:50:34 do tell :) Mar 21 18:50:45 mru: lie about what you do for a living? :) Mar 21 18:51:14 mru: fake your vocabulary? Mar 21 18:51:28 why bother? Mar 21 18:56:22 i find I have little social BS to share with people. Mar 21 18:56:56 well, except you guys :) Mar 21 19:02:57 djlewis: no you just share BS :P Mar 21 19:03:00 * mranostay ducks Mar 21 19:04:21 speaking of ducks, its been raining hard and constant here for two days Mar 21 19:04:34 <_av500_> raining ducks? Mar 21 19:04:45 <_av500_> or you were ducking rain? Mar 21 19:05:11 raining nuff to drown a duck and choke a bull frog Mar 21 19:05:12 <_av500_> or your ducker got rained on? Mar 21 19:06:08 duck is damn tasty Mar 21 19:06:22 _av500_: i like your G+ trolling :) Mar 21 19:20:10 <_av500_> mranostay: i'd never do that Mar 21 19:22:57 _av500_: should i make a reference to how i love lamb and veal? :) Mar 21 19:23:30 <_av500_> the ethic thread? Mar 21 19:24:15 yep Mar 21 19:24:28 <_av500_> that would be cheap Mar 21 19:29:04 what did I miss? Mar 21 19:30:24 hey hi Mar 21 19:31:27 <_av500_> mru: Jon Masters on the ethics of eating yummy meat Mar 21 19:32:15 I don't see how ethics and flavour are related Mar 21 19:32:52 <_av500_> its the usual vegan scare that we might eat them after we eat all the meat Mar 21 19:33:12 I will..I love pasture and grain fed beef alike Mar 21 19:33:31 I will cannibalize veg*'s before I'll ever eat other meat-eaters Mar 21 19:37:33 _av500_: nice trolling indeed Mar 21 19:37:49 anyone having idea of BB-xm Mar 21 19:41:39 ygsh: i think one or two people here may own one Mar 21 19:42:10 yes, gerald had that idea Mar 21 19:42:39 @mranostay : who are those? i have doubts to be cleared. Mar 21 19:43:01 * _av500_ has no idea Mar 21 19:43:12 just ask your question please:) Mar 21 19:43:16 * _av500_ checks his plan file Mar 21 19:44:36 ah my favorite command to explain to an non-*nix guy :) Mar 21 19:44:58 "did you say finger myself"? Mar 21 19:45:10 i am trying to boot my BB-xm using MMC in putty (windows) Mar 21 19:45:25 but what i am getting on the terminal is 60 Mar 21 19:45:35 60 what? Mar 21 19:45:37 <_av500_> most people get only 40 Mar 21 19:45:43 <_av500_> you have 20 more Mar 21 19:45:50 give it back! Mar 21 19:45:53 yeah Mar 21 19:46:01 <_av500_> but I would guess your card is not formatted correctly Mar 21 19:46:05 thats what i get Mar 21 19:46:13 <_av500_> use a known good image for a start Mar 21 19:46:18 can you helpme in that Mar 21 19:46:40 _av500_'s card formatting service? Mar 21 19:47:02 yes in windows Mar 21 19:47:21 hmm, perhaps there is money to be had distributing all variations of SD builds Mar 21 19:48:20 @djlewis : ?? Mar 21 19:48:37 <_av500_> the only thing i know in windows is solitaire.exe and canyon.mid Mar 21 19:49:18 * djlewis prefers command prompt Mar 21 19:49:53 the demo card is actually working Mar 21 19:49:55 ygsh_: your question comes up most every day. Mar 21 19:50:45 @djlewis: so is there any solution? Mar 21 19:51:14 hey! dont pick on me. I dont have a XM :) Mar 21 19:51:49 ok!! but any idea that could help Mar 21 19:51:55 <_av500_> http://www.angstrom-distribution.org/demo/beagleboard/ Mar 21 19:52:03 <_av500_> there you will find all files that you need Mar 21 19:52:09 and instructions Mar 21 19:52:20 <_av500_> since you are on windows, you will fail to use the mkcard script Mar 21 19:52:32 <_av500_> so make a 1:1 image of the demo card Mar 21 19:52:36 yes Mar 21 19:52:45 <_av500_> or copy one of the 2GB demo image to "format" the card Mar 21 19:53:06 <_av500_> how to copy such an image under windows I leave up to you to find out Mar 21 19:53:32 won't simple copy / paste work? Mar 21 19:53:39 uhoh Mar 21 19:54:00 <_av500_> no Mar 21 19:54:06 <_av500_> you need to write an image Mar 21 19:54:20 <_av500_> but any life distro that uses a usb stick has to do the same Mar 21 19:54:24 <_av500_> so there must be ways Mar 21 19:55:51 will writing the image onto the card make the card bootable? Mar 21 19:59:49 <_av500_> yes Mar 21 19:59:53 I cannot seem to find anything regarding pin mux configuration when using GPMC in AD multiplexed mode in the TRM (spruh73c) any idea where I could obtain that information Mar 21 20:00:00 <_av500_> since the image is of a bootable card Mar 21 20:00:16 <_av500_> jsabeaudry: it should be in the trm Mar 21 20:00:53 _av500_: when I make an image of a card, is it important that it be placed on a wooden table? Mar 21 20:03:14 @av500: and am i supposed to create partitions on sd card? Mar 21 20:03:19 mru: its important not to scale it up or down or it wont fit properly Mar 21 20:03:36 <_av500_> ygsh_: no Mar 21 20:03:50 <_av500_> since you write an image, it already has the partitions Mar 21 20:04:14 <_av500_> of course you also try to partition the card correctly Mar 21 20:04:33 <_av500_> and then copy MLO and uboot etc... Mar 21 20:04:38 <_av500_> but i guess you tried that Mar 21 20:04:45 <_av500_> andended up here Mar 21 20:05:05 i didn't write the image Mar 21 20:05:16 just copy/pasted tem Mar 21 20:05:26 so thats how to get a 60 Mar 21 20:06:35 so do i have to write the remaining files also? Mar 21 20:06:57 <_av500_> ygsh_: if you write an image no Mar 21 20:07:01 <_av500_> then it has all the files Mar 21 20:07:09 <_av500_> if your format and partition, yes Mar 21 20:07:48 ok that means ii should try writing an angstrom image onto the card? Mar 21 20:08:12 <_av500_> yes Mar 21 20:08:19 <_av500_> using some windows image writer tool Mar 21 20:08:23 <_av500_> or a proper linux pc Mar 21 20:08:37 prpplague: pong Mar 21 20:08:49 are u sure then it won't require MLO and other files? Mar 21 20:09:09 <_av500_> ygsh_: well, you will see if it has MLO or not, no? Mar 21 20:09:21 <_av500_> the image wont be write-only Mar 21 20:10:43 any recommended s/w for writing the image? Mar 21 20:10:50 <_av500_> yes, google Mar 21 20:11:24 too many results Mar 21 20:11:29 can you add filters? Mar 21 20:11:46 <_av500_> google google Mar 21 20:12:01 <_av500_> or call the google help line Mar 21 20:12:32 ygsh_: I recommend beagleboard.org and some reading Mar 21 20:13:26 ygsh_: under heading "Getting Started" Mar 21 20:17:14 i have a ubuntu raw file. can that be burnt? Mar 21 20:17:24 ygsh: no, because it's x86. Mar 21 20:17:28 (probably) Mar 21 20:17:45 <_av500_> ygsh: read the backlog and read beagleboard.org Mar 21 20:20:33 i am trying to write angstrom image onto the SD card using flashboot. Mar 21 20:20:48 lets see whether i get 60 or not Mar 21 20:20:59 koen: sorry was out getting cables Mar 21 20:21:06 koen: still around? Mar 21 20:23:42 yes Mar 21 20:24:26 koen: do you have time to add in some support for the mini-lvds to angstrom builds for beagle? or are you too swamped atm? Mar 21 20:25:45 not yet, preparing for ESC Mar 21 20:26:25 prpplague: I'll be in dallas next week Mar 21 20:26:38 not sure on the timing, since we are preparing for esc Mar 21 20:26:51 koen: ahh ok Mar 21 20:26:59 This LIC_FILE_CHKSUM thing is making me very angry Mar 21 20:27:10 koen: ping me when you get here Mar 21 20:28:27 why would you lick a checksum Mar 21 20:29:40 XMPPwocky: to see if it tastes salty? Mar 21 20:30:40 If you don't lick things you're wasting 20% of your senses Mar 21 20:30:52 16.66% if you believe in ESP Mar 21 20:31:38 prpplague: somewhere sunday evening, the cco folks are picking me up from the airport Mar 21 20:31:49 what? Mar 21 20:31:57 * prpplague 's feelings are hurt Mar 21 20:32:06 prpplague: it sounds like we'll fly to CA early on monday Mar 21 20:32:32 I don't need to book my flights and hotel myself, so I don't know the schedule :) Mar 21 20:33:03 koen: not like i would have picked you up from the airport, but i would have liked the opportunity to tell you "hell no!" Mar 21 20:33:16 koen could this be right? http://pastebin.com/nrycXqVP Mar 21 20:33:26 I don't think I should have to be using ${WORKDIR}/git Mar 21 20:33:35 koen: coming back through dallas on the way home? Mar 21 20:33:38 that's the only way I could figure out how to make it work Mar 21 20:33:45 prpplague: no, dulles Mar 21 20:34:06 SilicaGel: S = "${WORKDIR}/git" Mar 21 20:34:12 koen: ahh Mar 21 20:34:20 I thought I tried that. Let me try it again Mar 21 20:34:39 Greetings. Mar 21 20:34:41 S being already defined right? you'r enot telling me to add that definition of S are you? Mar 21 20:34:55 I have a problem to run xorg on BB rev C4. I think that has to do something to kernel parameters passed by uEnv.txt script because when I start the board along with pressing "USER" button - I can start X server normally. I'm almost sure that the MLO and u-boot.img placed on my mmc are the same files I used to flash the nand with. Mar 21 20:35:41 SilicaGel: http://pastebin.com/3nJXf1VZ Mar 21 20:35:45 I just had it echo ${S} and it told me | Your source dir is /mnt/bulk/cepasp/setup-scripts/build/tmp-angstrom_2010_x-eglibc/work/armv7a-angstrom-linux-gnueabi/prude-1.0-r0/prude-1.0 Mar 21 20:35:47 I've set up my uEnv.txt according to http://elinux.org/BeagleBoardUbuntu#Upgrade_X-loader_and_U-boot Mar 21 20:36:21 SilicaGel: S defaults to WORKDIR/PN-PV, but git stuff gets put in 'git', so S needs to get replaced Mar 21 20:38:10 oh Mar 21 20:38:24 interesing Mar 21 20:39:05 hrm that sort of worked, it apparently didn't do -lncurses on the link step Mar 21 20:39:10 * SilicaGel reads manual on how to add libraries Mar 21 20:39:13 SilicaGel: I already rearranged you recipe a bit Mar 21 20:39:26 SilicaGel: RDEPENDS on libs isn't needed, OE figures that out on its own Mar 21 20:39:41 oh ok Mar 21 20:40:43 is it actually running my makefile? I wonder what's wrong ... my makefile does LDFLAGS+=-lncurses which maybe is wrong'ish Mar 21 20:42:59 My failure with startx ends with "omapfb(0): Failed to open /dev/fb1: No such file or directory Mar 21 20:43:15 but I have no /dev/fb1 in fact Mar 21 20:44:03 SilicaGel: yes, CFLAGS and LDFLAGS are for the 'user', not the 'buildsystem' Mar 21 20:44:30 SilicaGel: so you'd to something like LFLAGS = -lncurses $(LDFLAGS) Mar 21 20:44:41 and then use LFLAGS in your linker cmd Mar 21 20:44:46 oh do that in the recipe Mar 21 20:52:25 Maybe someone could tell if during the "USER"-pressed-boot on BBrevC4 u-boot reads uEnv.txt in the same way as it reads this script file when booting from nand? Mar 21 20:52:50 SilicaGel: no, in your makefile :) Mar 21 20:52:51 I don't get why the behaviour is different depending on the booting method. Mar 21 20:53:14 SilicaGel: e.g. http://git.profusion.mobi/cgit.cgi/kmod.git/commit/?id=e48f37657dc03aee9aace60bf14ef9af489bf47c Mar 21 20:53:30 my makefile already does I'll show you. https://github.com/wz2b/prude/blob/master/Makefile Mar 21 20:53:55 I think if user is pressed it looks for a different file... I didn't spend much time on it, just noted that it was not reading uEnv.txt Mar 21 20:54:09 I set LDFLAGS, then in the actual link like I invoke the linker using that. I just modified my makefile to rpint out what it's actually running to link it though and it's not that! :( Mar 21 20:54:18 SilicaGel: you use LDFLAGS directly Mar 21 20:54:38 thurbad: thanks for the tip. Mar 21 20:54:40 SilicaGel: note the lack of 'D' in LFLAGS in my example above :) Mar 21 20:54:42 and what, oe_runmake overrides it? Mar 21 20:55:24 it calls make -e Mar 21 20:56:06 so CC, CFLAGS, LDLFLAGS, etc get nuked Mar 21 20:57:25 ahhhhhhhhhh Mar 21 20:57:27 i got you Mar 21 20:58:27 SilicaGel: arguably it should do 'make CC=foo CFLAGS=bar' instead of make -e, but historical reasons and all Mar 21 20:58:47 question .. should a 2.6.34.7 be able to boot on a beagleboard-xm + i get an error on uart3 douring bootup and anotherone saying cannot open initial console (not that they need to be connected9 booting with console=ttyS2,115200n8 as default bootargs Mar 21 20:58:48 yeah. Mar 21 20:59:13 s/9/) Mar 21 20:59:33 it's FREAKING OUT on me Mar 21 20:59:41 Do I have to tell it to use g++ somehow Mar 21 20:59:47 post 32, use ttyO2 Mar 21 21:00:18 thurbad: ahh yea that was my fuckup .. thanks Mar 21 21:00:28 i bet it needs to be CXX or something, not CC Mar 21 21:00:40 i guess the uart3 is the uart4 beeing not properly handled or something Mar 21 21:00:50 thuban: indeed, removing uEnv.txt solved my problem - I'll try to find what exactly was disturbing my Xorg. Mar 21 21:01:19 yep. now it built. Mar 21 21:01:22 WOW that was brutal. Mar 21 21:01:30 * unsolo_ is glad thurbad had an answer. now i dont have to work untill 1am .. Mar 21 21:02:14 you have to start working at 1am? Mar 21 21:02:35 mru: no i start at 9am but tend to end after midnight these days Mar 21 21:03:01 which is way to painful Mar 21 21:04:53 SilicaGel: I keep making the CPP=g++ thinko Mar 21 21:05:19 k;k' Mar 21 21:05:20 ah so Mar 21 21:05:25 I jsut changed my makefile to use CXX and it worked Mar 21 21:05:28 I hope more people read CPP as C Plus Plus instead of C Pre Processor Mar 21 21:05:34 However, it made two .ipk files Mar 21 21:05:36 not good Mar 21 21:05:53 one is an empty debian archive, the other is a 0 byte file (i.e. not even a valid archive) Mar 21 21:05:59 SilicaGel: no install target Mar 21 21:06:19 ok Mar 21 21:06:24 cool that's fixable Mar 21 21:06:31 do I need to somehow tell it what goes in -dev and what goes in Mar 21 21:06:31 koen: i would think so Mar 21 21:06:36 why it's -dbg I don't get either Mar 21 21:06:48 gdb you mean + Mar 21 21:06:49 SilicaGel: http://pastebin.com/6vkDNbwm Mar 21 21:06:58 SilicaGel: add that to the recipe Mar 21 21:07:12 SilicaGel: -dbg holds all the debug symbols that get stripped out Mar 21 21:07:30 SilicaGel: so if you want to gdb somthing, you just need to install foo-dbg instead of rebulding it Mar 21 21:08:21 can someone knows a way to override USB otg power limiation? I've connected power lines directly to 5VDC supply but I've got dmesg "usb 3-1: rejected 1 configuration due to insufficient avaliable bus power". Mar 21 21:08:47 maybe I can switch off power checking on OTG port? Mar 21 21:08:57 oh ok the dbg one actually hrm Mar 21 21:09:14 ixtec: you can probably mess with the driver to achieve that Mar 21 21:09:19 it looks like it put all my sources in ./usr/src/debug/prude-1.0-r0/git/ Mar 21 21:09:19 Mar 21 21:09:22 so taht's expected? Mar 21 21:09:23 at least it is detecting the r-id Mar 21 21:09:48 and the debug symbols I bet are this ./usr/bin/.debug/prude Mar 21 21:13:57 ok i git pushed. now it's time to go home. Mar 21 21:14:20 It's going to be annoying that I have to modify the .bb to specify a revision hash every time. In fact it's a chicken and egg situation I think Mar 21 21:14:26 if I keep the recipe in the same project anyway Mar 21 21:14:31 so I must have that somehow super screwed up Mar 21 21:14:34 SilicaGel: yes, -dbg includes sources to do line number matching Mar 21 21:15:12 SilicaGel: you can set SRCREV = "${AUTOREV}", but that's not recommended for obvious reasons Mar 21 21:15:51 yeah. Mar 21 21:16:39 What am I supposed to do with the .bb file really though? I've got it in my project itself, right next to the amkefile, but to make it actually build I had to run bitbake from ./setup-scripts so what I did was made ./setup-scripts/chrisp and put my prude.bb in there Mar 21 21:16:46 then did: bitbake -b chrisp/prude.bb Mar 21 21:18:19 hi Mar 21 21:18:26 * mranostay gets popcorn Mar 21 21:18:49 * mranostay bets $20 this is a SD card question Mar 21 21:20:12 @av500: still not getting the output. Mar 21 21:20:38 i have build the AM335XPSP_04.06.00.06 from git repos but under /sys/devices/platform/tsc isnt any ainX Mar 21 21:21:30 but maybe it would be possible to change the power requirement behaviour by some parameter? Mar 21 21:21:34 have anyone an idea Mar 21 21:22:28 ? parameters of ? Mar 21 21:22:51 I considering OTG power limitation. Mar 21 21:23:18 mranostay: he threw you for one ;) Mar 21 21:23:42 ixtec: is this xm or bone ? Mar 21 21:23:50 BBrevC4 Mar 21 21:23:56 oh Mar 21 21:23:58 its an bone Mar 21 21:24:04 A4 Mar 21 21:24:34 ixtec: iirc ive seen somewhere that you can specify the amount of amps for the tps Mar 21 21:24:56 I remember such parameter in the kernel Mar 21 21:24:58 unsure if that is only used by negotiation protocol or actually in use Mar 21 21:25:03 but in fact it's default value is 2 Mar 21 21:25:16 2mA doesnt get you far hehe Mar 21 21:25:17 and this seems not to be related to any real value Mar 21 21:25:29 since on OTG there is 100mA Mar 21 21:25:30 ixtec: and it is not used by the protocol + Mar 21 21:26:36 ixtec: but will the vbus detect work if you supply the chip outside of it _ Mar 21 21:26:52 w/o actually doing anything on the vbus Mar 21 21:27:24 VBUS seems to see its own voltage - not influenced by the load (?) Mar 21 21:28:27 ixtec: you said you wanted to supply the device outside of vbus right ? Mar 21 21:28:51 unsolo_: yes Mar 21 21:29:20 usolo_: just to have stable 5V @ 500mA Mar 21 21:29:27 ok nobody has an idea for my "no ainX is in sysfs" problem Mar 21 21:30:36 unsolo_: using even dumbest hub solves the problem Mar 21 21:31:02 unsolo_: but I have to get it working without one :( Mar 21 21:31:16 gn8 Mar 21 21:31:26 Failing to find information about pin mux for gpmc I resorted to using the TI pinmux utility, one thing is curious though, it puts the gpmc_clk in "input enable", does that make sense? Mar 21 21:31:39 Hexxter: 26+26 = 54! Mar 21 21:32:21 jsabeaudry: i have been warned the pinmux might not match the new boards Mar 21 21:32:30 with the utility Mar 21 21:32:54 djlewis, new chips? Mar 21 21:33:06 i didnt dig into it yet Mar 21 21:33:24 this is on A3 anyways Mar 21 21:33:34 still it is a cool tool to get a sense of the mux'ing Mar 21 21:34:09 but I'm far from a GPMC expert so I have trouble validating that IEN is a correct setting for the gpmc clock which I would have put in output only mode Mar 21 21:34:32 djlewis, definitely, especially too see quickly if you will be able to fit all your peripherals Mar 21 21:34:56 jsabeaudry: that sounds familiar. i remember seeing something similar in SDRC Mar 21 21:35:29 they wanted some memory clocks in input enable contrary to intuition Mar 21 21:35:50 @mranostay im meant it is an known bug, error, config mismatch ... Mar 21 21:36:47 jay6981, interesting, I'll leave it as suggested by the pinmux utility then, I just hope I won't break a bone or the peripheral Mar 21 21:37:16 i was working on the omap35xx but i bet your soc TRM calls it out Mar 21 21:38:09 jay6981, I've being going through the TRM but I think I don't have the right document because I fail to find guidelines regarding pin mux Mar 21 21:38:28 did you check the section on gpmc clocking? Mar 21 21:38:57 jsabeaudry: imho that does not make sense Mar 21 21:38:59 yes, the GPMC section I've gone through it a couple of times to understand everything Mar 21 21:39:16 gpmc is fun Mar 21 21:39:26 jsabeaudry: the input pin thing.. Mar 21 21:39:43 wait pins can also be used as irq pins so thats all inputs.. Mar 21 21:39:51 cs are all outputs Mar 21 21:39:59 address are all outputs Mar 21 21:40:04 data are all bidirectional Mar 21 21:40:20 read and write is output Mar 21 21:40:32 be is output Mar 21 21:40:45 what did i forget.. Mar 21 21:41:04 clk is output but only if synchronous mode is active (at least where it would matter) Mar 21 21:42:20 unsolo_, the pinmux utility agrees with you on everything but the clk Mar 21 21:42:39 could it be to support asynch ? Mar 21 21:42:50 well it would still be an output Mar 21 21:43:12 but it changes from cs to cs.. which is a bit crazy Mar 21 21:44:42 jsabeaudry: this is what I remembered re sdrc_clk Mar 21 21:44:44 "This output signal is also used as retiming input (the CONTROL_PADCONF_x.INPUTENABLE bit must be set to 1" Mar 21 21:45:29 oO Mar 21 21:46:00 btw how come the pinmux output from the ti tool is GPl ? Mar 21 21:46:04 table 11-95 of spruf98v (omap35xx trm) Mar 21 21:46:21 unsolo_ why not? Mar 21 21:46:34 they call out the same thing for gpmc_clk Mar 21 21:46:46 jsabeaudry: interesting Mar 21 21:46:51 err jay6981 :) Mar 21 21:47:08 woglinde: is the xl spreadsheat its generated from gpl ? Mar 21 21:47:43 woglinde: if it did something and not just list a bunch of settings i would agree it could be gpl Mar 21 21:47:55 but isnt it all just the way to do it ? Mar 21 21:49:32 woglinde: i do not want to argue that it needs to be gpl compatible to be linked into u-boot and the kernel. but BSD and mit would make it re usable in stuff like proprietary booters Mar 21 21:49:38 same goes for the x-loader imho Mar 21 21:59:19 jay6981, unsolo_, thanks for the help guys! Can't wait to receive the hw to test that gpmc bus Mar 21 22:01:42 jsabeaudry: np.. i might be able to fix up a kernel module that uses it too .. (kernel needs a few missing exports) Mar 21 22:01:52 but i have it somewhere. Mar 21 22:02:12 jsabeaudry: what are you hanging off gpmc? Mar 21 22:02:34 Spartan 6 Mar 21 22:02:35 32MB of sram :) Mar 21 22:02:41 jsabeaudry: nice Mar 21 22:02:44 cool Mar 21 22:41:10 hey there# Mar 21 22:41:31 hEy ThErE! Mar 21 22:43:06 has someone made some experience with MATLAB/Simulink and the beagleboard? Mar 21 22:44:13 i would like to know, what exactly is on the SD card written by Simulink for the beagleboard Mar 21 22:44:16 there is an arm version of matlab? Mar 21 22:44:22 intressting Mar 21 22:44:42 woglinde no.. but matlab has got the ability to install so called targets Mar 21 22:44:42 hmm I would think that would be some interface thing for matlab. Mar 21 22:44:57 yes sry for my bad terminology Mar 21 22:45:42 you can export C code from matlab and import it to the board Mar 21 22:45:58 and it seems to run via a linux distribution Mar 21 22:46:02 on the dsp then i recon ? Mar 21 22:46:15 that simply runs the compiled cide Mar 21 22:46:24 unsolo_: i dunno right now Mar 21 22:46:29 if it exports C code, I guess you can compile that for whatever you please Mar 21 22:46:29 i hope so Mar 21 22:46:56 the c code is noen optimized? Mar 21 22:46:57 yes but there is an image for the beagleboard Mar 21 22:47:01 I rather guess not Mar 21 22:47:09 ups neon even Mar 21 22:47:10 not yez Mar 21 22:47:30 I'd rather not run generated neon code Mar 21 22:47:36 it can't possibly be any good Mar 21 22:47:51 gm mru ;) Mar 21 22:47:55 haha Mar 21 22:48:02 is the DSP a floating point processor? Mar 21 22:48:10 no Mar 21 22:48:11 m0erk: no its fixed Mar 21 22:48:20 k, thx Mar 21 22:48:25 fixed-point or broken-point? Mar 21 22:48:27 ARM Cortex also? Mar 21 22:48:35 the arm has floating-point Mar 21 22:48:43 oh really? Mar 21 22:49:36 well, double precision and strict ieee single precision are a bit slow Mar 21 22:50:14 neon is fast though Mar 21 22:50:31 what simd is slow ? Mar 21 22:50:42 no, neon is simd Mar 21 22:50:44 neon is fast Mar 21 22:50:47 thus simd is fast Mar 21 22:50:53 iam right at the beginning, no hardware yet but i am planning to do some audio/video application with Matlab/Simulink with the beagle board Mar 21 22:51:02 mru: what i meant is that most simd's are fast ;) Mar 21 22:51:19 well, that's kind of the point Mar 21 22:51:59 mru: if a cpu had "multi" issue simd lets say like 5 neon pipelines.. it would be good fun Mar 21 22:52:14 * mru looks at cortex-a15 Mar 21 22:53:33 yes thats multiple simd's Mar 21 22:53:41 well, two Mar 21 22:54:27 where does it say that ? Mar 21 22:54:40 two lines up from here :) Mar 21 22:54:52 haha Mar 21 22:56:15 doesnt say so here http://www.arm.com/images/Eagle_New_Look_Chip-600.jpg Mar 21 22:56:39 that doesn't mean it's not true Mar 21 22:56:43 indeed Mar 21 22:56:59 so will it do a dual issue thing ? Mar 21 22:57:18 <_av500_> or at least dual boot Mar 21 22:57:21 or simply "schedule" it automagically for you.. Mar 21 22:57:34 _av500_: i thought the a15 didnt do AMP.. Mar 21 22:58:19 hmm is the SCU capable of doing anything useful ? Mar 21 22:58:58 <_av500_> snooping Mar 21 23:02:19 i need to get some hacking going soon on the neon.. Mar 21 23:02:39 just reading up on the A15 makes me miss the cell. Mar 21 23:03:02 in a good way for the a15 Mar 21 23:06:28 I can't find the image on arm.com, but here's a simplified diagram of the a15: http://www.wired.com/cloudline/wp-content/uploads/2011/10/arma15-blockdiagram.png Mar 21 23:06:38 stuff thats fun and relatively easy to do on a simd.. FFT iDCT DCT MATRIX operation (especially the ones that are used for calculating stuff in 3d space where the identity matrix is always a 4x4 matrix) Mar 21 23:06:43 to mention some Mar 21 23:07:34 so a neon instruction takes 10 cycles to complete ? Mar 21 23:07:58 nobody said that Mar 21 23:08:06 but a floating-point divide probably does Mar 21 23:08:20 i guess its a worst case scenario Mar 21 23:08:41 but even so if you issue one every cycle you execute one every cycle right ? Mar 21 23:09:00 ah, found it: http://www.arm.com/files/downloads/big.LITTLE_Final.pdf Mar 21 23:09:20 that's the pdf that image came from Mar 21 23:10:48 latencies for neon and vfp should be roughly the same as for a8 and a9 Mar 21 23:11:01 k Mar 21 23:11:13 but better throughput Mar 21 23:11:17 with two pipelines Mar 21 23:11:21 and out of order execution Mar 21 23:11:57 indeed Mar 21 23:12:00 hmm Mar 21 23:12:38 i notice that it mentions that A7 provides higher performance than current A8 implementations for a fraction of the power. Mar 21 23:12:49 page 3 last section Mar 21 23:13:02 that's not entirely true Mar 21 23:13:11 a7 neon performance is crap Mar 21 23:13:27 plain integer is probably about the same as a9 Mar 21 23:13:33 k Mar 21 23:15:31 so the neon should on the A15 in theory perfom insanely well Mar 21 23:15:37 +r Mar 21 23:15:57 it should be pretty good, yes Mar 21 23:16:42 VFPv3 is available? Mar 21 23:16:44 * unsolo_ hopes for beagleboard-xxm Mar 21 23:17:04 vfpv3 is in a8 and a9 Mar 21 23:17:11 <_av500_> will be more likely a pandaboard-ESX Mar 21 23:17:25 _av500_: oO Mar 21 23:17:57 <_av500_> or eagleboard.org Mar 21 23:18:00 sry i was talking about beagleboard Mar 21 23:18:17 beagle has vfpv3 Mar 21 23:18:18 m0erk: beagle is an A8.. Mar 21 23:18:21 so yes Mar 21 23:18:25 yes # Mar 21 23:21:54 _av500_: looks like something happened in eagleboard in march oO Mar 21 23:22:43 guess i have to pay some attention in the weeks/months to come Mar 21 23:51:02 hi Mar 21 23:55:37 * mranostay hides Mar 22 00:00:27 finally got sound out of my beagle Mar 22 00:00:50 the audio jacks were really stiff I didn't think a speaker cable would fit it Mar 22 00:02:04 now I'll try transferring an avi movie and see how video decoding dos Mar 22 00:06:43 pimping Mar 22 00:06:55 hi Mar 22 00:08:01 deja vu Mar 22 00:08:20 and now for something completely different Mar 22 00:08:26 monty python Mar 22 00:08:56 this is an ex-parrot.... Mar 22 00:09:13 It's pining for the fjords! Mar 22 00:12:27 seems to be able to play audio and ftp a movie at the same time Mar 22 00:47:20 avi movies play ok at a small window size Mar 22 01:21:35 I have just acquired a Beagle Bone rev A4. Documentation suggests that 5V adapter & connection to LAN should get me going, but I'm not seeing a DHCP lease being given out. Ideas? Mar 22 01:22:16 LAN cable is showing as OK, have checked router config (also OK.) Mar 22 01:50:43 hi **** ENDING LOGGING AT Thu Mar 22 02:59:58 2012