**** BEGIN LOGGING AT Thu Aug 09 02:59:59 2012 Aug 09 03:02:30 I found some program to control the gpio output. when I try to cross compile it with eclipse in ubuntu, some errors occur-------cannot find some header files Aug 09 03:04:14 my question is what's the directory of header files such as , Aug 09 03:06:10 most likely in the header source Aug 09 03:06:41 plat/mux.h definitely there Aug 09 03:07:33 where? Aug 09 03:08:24 what do you mean about the header source? where is it? Aug 09 03:08:25 you're going to need to compile it as a module probably Aug 09 03:08:52 I meant to type kernel source Aug 09 03:08:57 sorry about that Aug 09 03:11:24 It's ok. thank you! Could you tell me where I can down load the kernel source file? Aug 09 03:11:27 which distro are you using? Aug 09 03:12:21 sorry, I don't know what distro means. Aug 09 03:12:54 which linux distribution, such as angstrom/ubuntu/etc Aug 09 03:16:43 Oh. I forgot its distribution. I can check it tomorrow. If I know the distribution, how can I get the kernel source code? Aug 09 03:20:00 if you can type "uname -a" at the command line it should give some hint Aug 09 03:20:17 I feel so puzzle that I don't know which website has the application source code of Beaglebone Aug 09 03:20:57 ok, I will try tomorrow Aug 09 03:21:00 I don't use the bone, but angstrom is generally the default beagle distro Aug 09 03:23:59 if you didn't go out of your way to switch it's probably what you're using Aug 09 03:24:10 Yes, I'm sure that my beaglebone is now using the angstrom Aug 09 03:25:11 mranostay: hehe no Aug 09 03:25:19 mranostay: like that post? hehe Aug 09 03:28:16 you'll probably want to get the setup-scripts toolchain, if you can live without eclipse Aug 09 03:28:59 I assume that's what the bone uses still... someone correct me if I'm wrong Aug 09 03:33:26 I've never tried cross compiling the kernel in eclipse... Aug 09 03:38:18 how does the am355 compare speed wise to the with the c4 beagleboard anyway? Aug 09 03:44:59 I realize they're relatively the same as far as MHz Aug 09 03:53:59 so where do you cross compile the kernel? Aug 09 03:55:22 I have installed the toolchain. Now it can cross compile the user application project in eclipse. Aug 09 03:57:18 I use eclipse for regular app development, but oebb setup-scripts for kernel stuff Aug 09 04:00:15 ok, thank you so much for your help, thurbad. Aug 09 04:01:46 I don't know enough about using eclipse in a non managed makefile setting to set up the kernel Aug 09 04:02:25 so I lean on angstrom Aug 09 05:13:04 <_av500_> gm Aug 09 05:17:37 gm Aug 09 05:32:17 <_av500_> prpplague: rpi trainer: http://www.cnx-software.com/2012/08/09/gertboard-expansion-board-for-raspberry-pi-is-now-available-for-60/ Aug 09 05:38:50 yo Aug 09 05:39:06 anyone know about xm serial port boot issue Aug 09 05:39:18 i got my xm board to want to boot from serial port Aug 09 05:39:32 the problem is i dont want it to do that Aug 09 05:41:28 er Aug 09 05:42:28 _av500_: the fuse is nice touch Aug 09 05:42:34 *a Aug 09 05:46:10 <_av500_> edmen: insert a good sdcard and it wil boot from that Aug 09 05:46:44 av500 Aug 09 05:46:46 hey Aug 09 05:46:51 hows it going Aug 09 05:46:52 yeah Aug 09 05:46:58 i know sd card should do it Aug 09 05:47:00 but Aug 09 05:47:10 but what? Aug 09 05:47:16 i have a beagle board Aug 09 05:47:18 too Aug 09 05:47:21 ok? Aug 09 05:47:37 and i just tried it in the xm not sure if thats a dumb thing to do our not Aug 09 05:47:40 but i did it Aug 09 05:47:47 and surprise Aug 09 05:47:50 it didnt work Aug 09 05:47:51 :) Aug 09 05:48:11 my goal was to get the xm board to have ubuntu Aug 09 05:48:24 and was able to Aug 09 05:48:24 and so i followed a tutorial Aug 09 05:48:27 the first time Aug 09 05:48:30 ubuntu worked Aug 09 05:48:39 and everything was great Aug 09 05:48:43 then i turned off the board Aug 09 05:48:48 xm board Aug 09 05:48:50 and Aug 09 05:48:57 turned back on , and it was dead Aug 09 05:49:06 well not dead but , not alive Aug 09 05:49:14 i get some jiberish Aug 09 05:49:20 on the uart side Aug 09 05:49:21 holy spamming edmen Aug 09 05:49:39 lol Aug 09 05:49:41 nananananana edmen! ;) Aug 09 05:50:46 here is the current theory : Aug 09 05:50:48 The SD-card is the only media from which the BeagleBoard xM boots on its own: unlike the Beagleboard, it does not have any non-volatile RAM chips installed (BeagleBoard has NAND flash memory). The xM boot sequence is quite elaborate. There is a little bootstrap loader located in the CPU internal ROM. This loader boots MLO. In turn, MLO starts u-Boot. Finally,u-boot sets the table up for Linux (or any other operating systems) and start i Aug 09 05:51:12 this is what i know have come to belive is happening only that i dont know how to fix it Aug 09 05:51:36 well thats no theory so far, but common knowledge Aug 09 05:51:43 lol Aug 09 05:51:54 the theory would include an idea which part is broken ;) Aug 09 05:51:59 i am a beginner so everything is a theory at this point :) Aug 09 05:52:37 the part that is broken is the part where my xm board doesnt boot from sd card Aug 09 05:53:26 My board is a beagleboard xm rev c Aug 09 05:54:19 I tried all sorts of things , only to keep getting jiberish on the uart Aug 09 05:55:30 if at least *something* is coming out of the uart, i'd guess that it sucessfully hits MLO Aug 09 05:55:39 which is probably the xm board trying to boot from uart.. Aug 09 05:55:45 i am thinking Aug 09 05:56:15 sorry if you have already heard this, but please, do really triple check that your terminal is set to 115200-8N1, flow control off Aug 09 05:56:42 and some terminals as well as usb usarts are easily confused, so please cold start both Aug 09 05:56:44 thats a good check, I have checked that Aug 09 05:56:59 i even tried all standard bauds Aug 09 05:56:59 lol Aug 09 05:57:22 only to find out that 115200 8N1 flow control off make the jiberish look more correct Aug 09 05:57:25 howdy ilyanok Aug 09 05:58:05 the jiberish is an attempt from the MLO to boot from the serial port Aug 09 05:58:15 from what i am understanding Aug 09 05:58:27 if i am understanding anything correctly Aug 09 05:58:31 edmen: my next step would then be to put the sd card aside, and redo the procedure with a second one. Aug 09 05:59:04 if that card also boots only once, there is some software part that breaks MLO/uboot Aug 09 06:00:13 LetoThe2nd : i tried the same procedure on the same card and i was able to reproduce but then after turning off the board it had the same problem and I have not been able to recover from that can the boot strap get damaged ? Aug 09 06:01:05 edmen: ok, then redo the card, and before the first boot, save the contents of the first (FAT) partition. Aug 09 06:01:19 to isolate the part tha gets broken. Aug 09 06:01:30 i see, Aug 09 06:01:48 edmen: the MLO and uboot and most probably anlso kernel should reside there. Aug 09 06:01:56 yes Aug 09 06:02:33 so for me now it would be time to do a comparison [before first boot]:[after first boot] Aug 09 06:02:55 one thing i did which could of screwed it up , was use the beaglebone standard sd card on the xm ( out of desperation ) it booted amazingly Aug 09 06:03:10 but stopped working after two power cycles Aug 09 06:03:12 don't think so. Aug 09 06:03:37 the card still works on the beaglebone , so what ever happen it only screwed up the xm board Aug 09 06:04:00 i'm no pro on the beagles, but are you sure it is sufficiently powered? Aug 09 06:04:24 I did check power, using a 5 v 1.5 amp Aug 09 06:04:26 thats a problem we keep on seeing constantly on the pandas. Aug 09 06:04:31 power supply adaptor Aug 09 06:04:39 maybe it needs more current ? Aug 09 06:05:03 (where 1.5A would certainly be too little for running ubuntu stable. but like i said, no idea about the beagles consumption) Aug 09 06:05:32 hmm i can try a more beefy adaptor Aug 09 06:05:43 5v 3am or so Aug 09 06:05:48 mranostay: hi Aug 09 06:05:59 edmen: yeah thats what we recommend. Aug 09 06:06:16 but i do know if the current is not enough the kernel will panic Aug 09 06:06:26 when the board was working thats what i would see Aug 09 06:06:30 so i use adaptor Aug 09 06:06:38 to boot angstrom Aug 09 06:06:45 nope, it *might* panic. Aug 09 06:07:14 do you guys think i would need something like this http://www.liquidware.com/shop/show/GL000604/BeagleBoard+Recovery+Disk Aug 09 06:07:16 it does not *have* *to* Aug 09 06:07:17 to fix it ? Aug 09 06:07:36 it claims to be able to fix hard core issues Aug 09 06:07:43 what a rip off Aug 09 06:07:50 yeah thats what i thought Aug 09 06:07:53 <_av500_> hard core? Aug 09 06:08:03 yeah lol Aug 09 06:08:09 i dono what it does Aug 09 06:08:20 that i wouldnt be able to do manually using ubuntu terminal Aug 09 06:08:39 probably reflash the uboot. it's meant for the standard beagle, where you have a nand that can be bricked. Aug 09 06:08:46 totally pointless for XM Aug 09 06:09:02 cool thanks Aug 09 06:09:41 for the info on that card Aug 09 06:09:56 * LetoThe2nd bbiab, gotta get caffeine Aug 09 06:11:03 doh... but i still dont know what i should do exactly , isnt flashing the card to the demo image enough ? beacuase when i have tried that it seems it not becuase it still tried to boot from serial port Aug 09 06:11:08 and shows giberish Aug 09 06:13:22 <_av500_> mranostay: :) Aug 09 06:14:35 ok can anyone explain something, the xm boot order Aug 09 06:14:39 how does that work Aug 09 06:15:21 because the idea was that the jiberish from the serialport is actually the board trying to boot from the uart Aug 09 06:15:40 why is it choosing to boot from the uart to begin with ? Aug 09 06:15:53 does that mean that the sd card is not bootable ? Aug 09 06:21:33 <_av500_> yes Aug 09 06:21:34 yes Aug 09 06:21:43 see TRM. Aug 09 06:21:44 <_av500_> and that is not a theory Aug 09 06:22:07 since we don't have sources for the ROM, it is just a theory ;) Aug 09 06:22:34 <_av500_> its highly advanced, thus magic! Aug 09 06:22:58 hehe Aug 09 06:23:11 i see Aug 09 06:23:17 http://elinux.org/BeagleBoardUbuntu Aug 09 06:23:41 would this help me get out my situation ? Aug 09 06:24:50 my dream is to run ubuntu on the xm board ;) Aug 09 06:25:07 but your abilities is limiting you to digging ditches. Aug 09 06:25:31 lol Aug 09 06:25:42 yes it would appear so Aug 09 06:25:53 * _av500_ will not publicly state that ubuntu gets one out of anything Aug 09 06:26:00 <_av500_> except out of a sane mind Aug 09 06:26:01 lol Aug 09 06:26:11 thats true Aug 09 06:26:22 my reason for wanting ubuntu Aug 09 06:26:28 of all the things that i have lost, i miss my mind the most. Aug 09 06:26:34 (citation: ozzy osbourne) Aug 09 06:26:38 is uvc camera support Aug 09 06:27:07 if angstrom was nice to uvc camera i would use that Aug 09 06:27:23 angstrom can be nice to everything Aug 09 06:28:20 after recompiling lots of stuff :) right ? Aug 09 06:28:40 my force is not strong yet Aug 09 06:28:42 oO(isn't the install opkg install uvc-magic) Aug 09 06:28:45 IIRC, no. Aug 09 06:29:15 uvc-magic ? Aug 09 06:29:17 really Aug 09 06:29:19 lol Aug 09 06:29:36 somthing like that. just check whats in the repos. Aug 09 06:29:44 kool Aug 09 06:29:58 ubuntu doesnt run at full speed in xm through Aug 09 06:30:03 from what i have read Aug 09 06:30:09 which is not something i like Aug 09 06:30:20 <_av500_> yes, its limited to quarter speed Aug 09 06:30:28 that blows. Aug 09 06:30:38 <_av500_> in order to make it running an a quad core look nice... Aug 09 06:30:39 in a negative preasure kind way Aug 09 06:30:42 <_av500_> polotics... Aug 09 06:30:44 <_av500_> politics Aug 09 06:30:56 _av500_: an if i pay you a nickel? Aug 09 06:31:08 <_av500_> id be rich Aug 09 06:31:21 will you make it run faster then? Aug 09 06:31:29 <_av500_> run it downhill Aug 09 06:31:37 i want my nickelback! Aug 09 06:31:42 *SCNR* Aug 09 06:32:11 you mean physically make it change position faster, simple just throw it Aug 09 06:33:00 * LetoThe2nd does in fact not own a beagleboard of any kind and also does not plan to. Aug 09 06:33:23 i am starting to not want to own one too Aug 09 06:33:28 limited to quarter speed? what? Aug 09 06:33:30 lol Aug 09 06:34:09 <_av500_> Kevin`: yes, long story, go ask Canonical Aug 09 06:34:27 from 1ghz to about 700mhz Aug 09 06:34:45 isn't it based on that opensource linux thing, and running on a chip without any annoying config locks related to clocking? Aug 09 06:34:49 * Kevin` confused Aug 09 06:36:04 people should learn not to take _av500_, myself and a few other too seriously. Aug 09 06:36:18 * LetoThe2nd admits he is on metal right now, but caffeine depraved. Aug 09 06:36:27 http://www.gigamegablog.com/2010/12/09/beagleboard-xm-and-angstrom-getting-the-big-dog-to-run-at-full-speed/ Aug 09 06:36:58 its kinda old info Aug 09 06:37:03 kevin Aug 09 06:37:23 I would expect boot time to be influenced more by a bad sd card choice (or using an sd card at all vs using nand on the old models) Aug 09 06:38:07 does anyone here played around with a xm board before ? Aug 09 06:38:45 <_av500_> lots Aug 09 06:39:29 did you ever get it to a state where it wouldnt boot from anything, such as that jiberish show up on the uart only Aug 09 06:40:19 i think i am going to try the suggestion of loading the demo image , but i am not clear about one thing, which is the following: Aug 09 06:40:28 there's one rather nice thing about the omap chips, they can be recovered from any software fault rather easily because of the romstage loaded Aug 09 06:40:37 loader* Aug 09 06:40:47 ha you read my mind Aug 09 06:40:48 and on the xm even more so because it ONLY boots from external storage Aug 09 06:40:52 i was about to say that Aug 09 06:41:16 so i was a bit confused about that part Aug 09 06:41:28 so i have tried to trigger the rom boot Aug 09 06:41:39 or so say romstage boot Aug 09 06:41:45 by user button Aug 09 06:41:51 on the board its self Aug 09 06:41:56 but it does not do anything Aug 09 06:42:03 or i dont know how to do it Aug 09 06:42:08 which probablly that Aug 09 06:42:13 what would you expect it to do? the default is to boot from sd card, right? Aug 09 06:42:22 yeah Aug 09 06:42:29 which it does not Aug 09 06:42:31 do Aug 09 06:42:35 if one of the boot config lines is connected to a button (I didn't check, for xm), what it would likely do is boot from usb or rs232 Aug 09 06:42:51 yes!! Aug 09 06:43:02 thats exactly what i been postulating it does Aug 09 06:43:03 you won't see any serial output from the rom, the first thing you'll see is x-loader or whatever they use now Aug 09 06:43:20 well i do see something on the uart Aug 09 06:43:27 which is a few garbage charcters Aug 09 06:43:44 * LetoThe2nd is so happy we are finally reaching the repetition phase. Aug 09 06:43:46 but i think you maybe right there Aug 09 06:44:10 i love repetition ! Aug 09 06:44:44 one of those therapeutic programming syndrome Aug 09 06:44:52 :) Aug 09 06:45:07 so kevin what should i do Aug 09 06:45:11 exactly Aug 09 06:45:32 (sorry I am still new to the beagle board science) Aug 09 06:45:55 use a known working image, verify it after writing to the card (sd cards are stupid about breaking). boot as normal. Aug 09 06:46:57 i see, and how do i force it to boot to the sd card ? Aug 09 06:47:14 i mean how do i make it not try to boot from other sources Aug 09 06:47:58 <_av500_> it trie sd first Aug 09 06:48:11 I suspect that's the default first boot device. check the configuration of sys_boot to be sure though Aug 09 06:49:22 isee Aug 09 06:49:31 <_av500_> LetoThe2nd: +1 Aug 09 06:49:52 will sys_boot be a text file ? Aug 09 06:50:31 <_av500_> no Aug 09 06:50:38 _av500_: well i've already said everything an hour ago. Aug 09 06:50:51 yeah i know Aug 09 06:50:54 sorry guys Aug 09 06:51:08 i am like truly struggling here Aug 09 06:52:03 you guys have been great by the way Aug 09 06:52:32 i am truly greatful here i would be totally lost if it werent for you guys Aug 09 06:54:08 edmen: http://beagle.s3.amazonaws.com/design/xM-A3/BB-xM_Schematic_REVA3.pdf - note the "DNI" for parts that are NOT populated Aug 09 06:55:21 ok Aug 09 06:55:30 i see Aug 09 06:56:51 is the omap 3530 trm still the document to use for the 3730? Aug 09 06:57:27 ok I am going to try out some of these suggestions thanks a ton av500 and LetoThe2nd and kevin . Aug 09 06:58:49 edmen: assuming the above is true, you should also have this, it's informative: http://www.ti.com/litv/pdf/spruf98x Aug 09 07:00:31 oh thanks! Aug 09 07:02:38 * KeatonT waves Aug 09 07:05:10 the chip is probably dm3730 (as referenced elsewhere) and the file is just derped, since that part actually exists. grab the document for that instead. Aug 09 10:36:53 Hi. Does anyone know if BeagleBoard-xM is capable of doing H.264 decoding and encoding simultaneously? Aug 09 10:38:11 if then you'd need to run one in the 'DSP' and the other on the CPU Aug 09 10:39:35 freen: h.264 is a non free codec and i advise using ogg theora instead Aug 09 10:44:07 dm8tbr: ok, thanks Aug 09 10:57:06 jkridner: I'm going over the test* script to adapt them to the new world order Aug 09 10:57:34 jkridner: any objections to including the .wav file in the repo? that would avoid the sox dep Aug 09 10:57:42 I've switched to aplay already Aug 09 11:26:19 koen: I like having sox, but if you have some reason to remove it, I don't have any objections to shipping a .wav Aug 09 11:27:25 I'll still ship sox, but I'm running into some problems with it after updating it Aug 09 11:27:33 we were using a version from 2007 Aug 09 11:27:39 I updated to a 2012 one :) Aug 09 11:27:53 (this is to test audio capes on bone) Aug 09 12:24:18 does anyone have any pointers to 4/8-bit NAND ECC support in u-boot? I'm looking in the denx tree, and not coming up with much yet Aug 09 12:32:27 cbrake: http://elinux.org/BeagleBoardNAND might provide some pointers Aug 09 12:37:11 jkridner: is it safe to update bonescript to master? Aug 09 13:15:44 jkridner: https://github.com/koenkooi/validation-scripts/commit/168a9604e170931d9de228f0d774f4fbda58aa37 Aug 09 13:37:55 wmat: thanks for the link -- its not clear how to set up for 4/8 bit hw ECC vs 1 bit Aug 09 13:58:47 cbrake, it's one of the most convoluted issues on "omap" IMHO Aug 09 14:01:20 cbrake, there's at least some foundational work now being done on the kernel side to address some of the gpmc infrastructure...it'll be nice when nand driver works for all cases. Aug 09 14:53:21 mdp: I'm having pretty good results with the 3.5 kernel and 8-bit ECC Aug 09 14:53:46 cool...using h/w ecc there? Aug 09 14:54:03 mdp: as best I can tell -- hw detection, sw correction Aug 09 14:54:44 mdp: I instrumented a few of the calls like omap3_calculate_ecc_bch8, and they are getting called Aug 09 14:55:13 mdp: nandtest /dev/mtdX runs about the same speed as with the micron internal ECC engine Aug 09 14:55:31 mdp: we've used the micron ECC engine on everything that needed 4-bit ECC to date Aug 09 14:56:08 mdp: but ran into a problem with the AM35xx where it will power on fine, but won't boot from nand if we warm/cold reset after the system has been running Aug 09 14:56:41 mdp: it appears the NAND flash is stuck in the internal Micron ECC mode, and the AM35xx does not execute a ONFI RESET to the NAND in this case Aug 09 14:56:42 cbrake, yeah, I was thankful to see the Ivan's version finally get cleaned and go in...the TI BCH solution that has shipped for a long time is buggy :( Aug 09 14:57:12 mdp: yeah, it looks pretty reasonable compared to the code I was looking at several years ago Aug 09 14:57:19 cbrake, I've never seen the value of using the internal engine on these parts Aug 09 14:58:04 mdp: the internal engine is very valuable when the 4-bit ECC mode is broke in the OMAP3530, and you can't figure out the kernel support for hw ECC detect :-) Aug 09 14:58:32 mdp: from what I can tell, there is no real cost to doing 8-bit ECC over 4-bit with the OMAP/AM ECC engine Aug 09 14:58:38 cbrake, yeah, I ran some comparisons on a rough version of it and their s/w bch solution after hacking too many production stop 911 fixes on the shipping stuff Aug 09 14:59:21 cbrake, right, the correction part is the same instructions...the costly part is done in h/w Aug 09 14:59:37 cbrake, ok, so that's why I didn't see the value...omap3530 doesn't exist to me :) Aug 09 14:59:42 mdp: IC, neat Aug 09 14:59:44 mdp: :-) Aug 09 14:59:59 mdp: now if I could just find some MLO/u-boot bits ... Aug 09 15:00:22 that Tartarus guy could be beaten up for those bits Aug 09 15:01:42 cbrake, you might want to check and see what, if anything, has been worked on for that...depending on which part you need those for.. Aug 09 15:01:51 he's working am335x stuff for mainline u-boot now Aug 09 15:02:09 * Tartarus reads Aug 09 15:03:11 I didn't know how the ecc bits are in mainline u-boot these days.. Aug 09 15:03:15 yeah, nand isn't up on the mainline list yet Aug 09 15:04:44 Tartarus: ahh, so that explains it -- are there any repos/patches available? Aug 09 15:05:25 Tartarus: am I correct that the 3.5 omap-linux kernel is in pretty good shape with 4/8 bit hw ecc? Aug 09 15:08:19 I'm just a poo-flinging monkey, I know nothing of the kernel Aug 09 15:08:48 wrt U-Boot, the arago-project u-boot-am33x tree supports 8bit hw ecc, with a caveat that the env isn't covered right Aug 09 15:08:52 which leads to silliness Aug 09 15:09:00 that needs to be fixed when it goes mainlin Aug 09 15:09:00 e Aug 09 16:03:16 koen: you get a chance to look at your man stuff? The changes you pushed got me the packages, but it looks like gtbl has been omitted from groff. Aug 09 16:20:53 spacecolonyone: not yet, sorry Aug 09 16:21:19 * koen root@beaglebone:~# opkg update ; opkg install man [ OK ] Aug 09 16:23:28 same here. I also did install less to get rid of the other. man man gives: /usr/bin/gtbl no such file Aug 09 16:29:48 spacecolonyone: I see what happened Aug 09 16:29:55 spacecolonyone: man worked in the old, non-yocto OE tree Aug 09 16:30:08 but in the new yocto OE tree it's missing some fixes Aug 09 16:30:17 like dependencies on less and groff Aug 09 16:30:20 spacecolonyone: wget http://dominion.thruhere.net/koen/angstrom/man.conf -O /etc/man.conf Aug 09 16:31:10 ehm Aug 09 16:31:15 man.config of course Aug 09 16:31:20 stupid copy/pase Aug 09 16:31:23 paste Aug 09 16:32:25 sidenote: i catted my current man.conf and it refrences /home/koen Aug 09 16:32:39 yeah Aug 09 16:32:54 that's one of the bugs than was fixed in the old tree Aug 09 16:33:21 woot! Aug 09 16:34:42 works, ok, so did the change in the config remove the dependency on gtbl? for that matter, if gtbl is part of groff and groff is installed why didn't i have it? soo many (mostly unnecessary) questions! :) Aug 09 16:34:52 yes Aug 09 16:34:57 groff installs tbl Aug 09 16:35:43 but not gtbl Aug 09 16:35:45 spacecolonyone: http://pastebin.com/WUXiY8wS Aug 09 16:36:28 same here. ty! Aug 09 16:37:22 koen: any pointers as to how I could have debugged this on my own? remembering to always learn about and check the config file is one thing. Aug 09 16:41:27 spacecolonyone: well... Aug 09 16:41:42 spacecolonyone: I realized that when I said "worked months ago" it might have been the old OE tree Aug 09 16:42:02 and that old recipe said: Aug 09 16:42:02 # Note: The default man.conf uses wrong names for GNU eqn and troff, Aug 09 16:42:02 # so we install our own Aug 09 16:43:39 k Aug 09 16:44:18 the long way around would have been strace and grep Aug 09 16:45:25 the funny thing is that someone in yocto went to a lot of trouble of adding patches to fix things and add feature Aug 09 16:45:39 but we now know it had never actually been runtime tested :) Aug 09 16:52:54 spacecolonyone: patches sent, so hopefully it will be fixed once and for all Aug 09 16:53:01 spacecolonyone: updates binaries on their way Aug 09 17:43:59 koen, angstrom-distribution is all borked up Aug 09 17:44:09 dotster screwed up roayllay Aug 09 17:44:24 I am trying to figure out how to sort it out, but will take time Aug 09 17:44:42 after the website is back, if we are missing other dns entries, can you let me know Aug 09 17:51:00 Crofton|work: sure Aug 09 17:51:18 I am all pissed off Aug 09 17:51:33 so is twitter Aug 09 17:51:51 * koen received his new makerbot today and doesn't really care Aug 09 17:52:07 I am annoyed at the US west coast and their pesky timezone Aug 09 17:52:17 10pm conf calls are no fun Aug 09 17:52:23 but I'm taking one for the team :) Aug 09 17:53:04 heh Aug 09 17:53:16 I will be on the west coast for a couple of weeks Aug 09 17:57:36 koen: hey! Aug 09 18:00:10 * koen blames mranostay Aug 09 18:45:41 koen, sanity seems to be coming back, can you let me know if things like lists are broken etc Aug 09 18:57:57 Crofton|work: sure Aug 09 18:58:00 Crofton|work: feeds are back up Aug 09 18:58:46 sanity is still MIA Aug 09 19:00:33 heh Aug 09 19:01:15 what is the name of the feeds? Aug 09 19:01:51 insanity? Aug 09 19:15:47 Crofton|work: feeds.angstrom and eu.feeds.angstrom, us.feeds.angstrom Aug 09 19:18:16 can you email me what a records should be defined Aug 09 19:18:22 I fear things will fall apart Aug 09 19:29:01 denix: sanity never existed Aug 09 19:56:26 i recently purchased the flyswatter JTAG debugger for the beagle. I am trying to debug the kernel using gdb and JTAG, but I am unable to load vmlinux using the "load vmlinux" command in gdb. Has anyone encountered this error? And if so, how do you fix it? Aug 09 20:53:04 cboyd: what's the error? Aug 09 20:53:20 cboyd: I assume you mean Flyswatter2? Aug 09 20:55:03 wmat: i bought the flyswatter because it was cheaper Aug 09 20:55:50 wmat: when I try to load vmlinux ... the error says "load failed" Aug 09 20:55:56 cboyd: are you using gdb via openocd? Aug 09 20:56:11 wmat: yes Aug 09 20:57:03 i think the problem is the linker stamps the image with the wrong address Aug 09 20:57:23 cboyd: have you gotten anything simple to work, like LEDblink? Aug 09 20:58:13 wmat: yes, LEDblink works fine Aug 09 20:58:43 cboyd: via gdb i meant? Aug 09 21:00:07 yes i use gdb and openocd Aug 09 21:00:25 wmat: yes i use gdb and openocd Aug 09 21:01:43 cboyd: try the 'Loading a kernel in memory' instructions here http://elinux.org/Debugging_The_Linux_Kernel_Using_Gdb Aug 09 21:02:14 wmat: the error I get is "Loading section .head.text, size 0x180 lma 0xc0008000 Aug 09 21:03:02 wmat: the error I get is "Loading section .head.text, size 0x180 lma 0xc0008000 Aug 09 21:03:03 Load failed" Aug 09 21:03:37 cboyd: I recall that error showing up on the list before Aug 09 21:03:52 cboyd: https://groups.google.com/forum/?fromgroups#!topic/beagleboard/h_k5Fn9rLQI%5B1-25%5D Aug 09 21:04:34 wmat: I am familiar with these instructions Aug 09 21:05:34 cboyd: so have you tried using the address param of the load command to pass the correct value? Aug 09 21:05:44 wmat: I believe so Aug 09 21:06:12 wmat: I used the following command "load vmlinux 0xbfffffff" Aug 09 21:06:30 wmat: is that the correct way to execute the command? Aug 09 21:08:33 cboyd: did you calculate the 0xbfffffff? Aug 09 21:09:02 cboyd: the value in the post is 0xBFFFFFFFFF Aug 09 21:09:28 wmat: no I didn't calculate the address Aug 09 21:09:50 wmat: I tried the address from the post and it didn't work Aug 09 21:10:14 wmat: the address appeared to be to long Aug 09 21:10:27 wmat: 32 bit addresses right? Aug 09 21:13:24 cboyd: yeah Aug 09 21:13:54 cboyd: try load vmlinux address 0xBFFFFFFFFF Aug 09 21:14:58 wmat: verbatim Aug 09 21:15:27 wmat "load vmlinux address 0xBFFFFFFFFF" Aug 09 21:15:31 cboyd: you can try with your address first if you like Aug 09 21:16:01 wmat: i tried the 0xBFFFFFFFFF address and got an error Aug 09 21:16:06 cboyd: I'm guessing the parameter addition command to be address Aug 09 21:16:20 that looks like one too many F's for a 32 bit address Aug 09 21:16:27 wmat: i am work right now ... so I can't see the actual error message Aug 09 21:16:36 agreed Aug 09 21:16:43 wmat: agreed Aug 09 21:17:00 thurbad: agreed Aug 09 21:17:18 cboyd: well, I have to go to dinner now. But I'll set up my board when I get home. May be as late as >9pm EDT Aug 09 21:17:22 actually 2 too many Aug 09 21:17:30 cboyd: what BB specifically are you using? Aug 09 21:17:50 wmat: BB xm rev A I believe Aug 09 21:18:03