**** BEGIN LOGGING AT Tue Jan 22 02:59:58 2013 Jan 22 04:45:49 Can anyone help me setup open embedded? Jan 22 04:46:09 I have the core running and i want to get some of the stuff in meta-openembedded repo built Jan 22 05:22:48 when i run bitbake, it doesnt scroll through all the tasks it has to do but it just shows how many threads and what its doing on those threads.. is this a more recent addition to bitbake? Jan 22 05:23:08 i checked out oe & bb from october and it doesnt have this feature Jan 22 05:23:35 or is this just feature i need to enable? Jan 22 05:23:46 rsync --recursive gw.xora.org.uk:.irssi . Jan 22 05:23:55 WTF Jan 22 05:24:05 XorA: wrong window :b Jan 22 05:24:18 not even sure how I did that Jan 22 05:24:29 palm on touchpad error I guess Jan 22 05:24:46 pretty dead in here hey Jan 22 05:30:36 :( Jan 22 05:37:45 * XorA is watching UFC Jan 22 05:37:51 not much coding Jan 22 05:41:00 Xora: help me out! Jan 22 05:41:17 Go Vitor Jan 22 05:41:49 XorA: :S Jan 22 05:41:56 Im still on FX 5 Jan 22 05:41:59 catching up Jan 22 05:42:25 trung: I have never seen the affect you see Jan 22 05:42:40 trung: I checked out and built yesterday and everything was normal Jan 22 05:43:10 XorA: i checked out today and it was sweet.. no scrolling through 100s of pages of output.. just 4 lines (concurrent processes) Jan 22 05:43:14 but I am using Angstrom Jan 22 05:43:25 XorA: gumstix?? Jan 22 05:43:31 no, beagles Jan 22 05:43:47 XorA: tell me one thing.. how do u get uImage and u-boot on the beagle? Jan 22 05:44:06 cp MLO u-boot.bin uImage /media/sd/ Jan 22 05:44:26 you mean just copy it to the root dir? Jan 22 05:44:39 use omap3-mkcard.sh Jan 22 05:44:53 it will make 2 partitions, one fat for those files, one ext2 for your rootfs Jan 22 05:44:58 ok. where do you get MLO, u-boot.bin and uImage? Do you build it using bitbake? Jan 22 05:45:05 yes, built them Jan 22 05:45:21 how? do you have any docs i can read? Jan 22 05:45:52 yocto project has many docs for beginners Jan 22 05:45:55 AFAIK Jan 22 05:46:03 i cant find omap3-mkcard.sh ive seen it before Jan 22 05:46:24 www.xora.org.uk :-D Jan 22 05:46:39 ive got the console image working fine.. and NFS network booting.. but i think my uImage is ancient andi need to update it.. been too long.. Jan 22 05:46:41 but latest version is always in meta-angstrom Jan 22 05:47:06 i see let me check Jan 22 05:47:36 oh yes ive found it Jan 22 05:48:11 do you put meta-angstrom in ~/yocto/poky ? even though it's a git repo? Jan 22 05:48:23 see the docs Jan 22 05:48:32 layers are too complex to explain here Jan 22 05:48:46 my head hurts. ok Jan 22 05:49:40 hehe Jan 22 05:49:50 coule be worse, could have some large dude punching it Jan 22 05:50:57 lol.. ufc, insane.. Jan 22 05:52:04 XorA: so tell me where you put YOUR meta-angstrom.. i know i can configure it to be anywhere in bblayers.conf but hell i just want to follow best pracs Jan 22 05:52:30 trung: I used the angstong setup scripts because Im lazy Jan 22 05:55:34 ill check em out Jan 22 05:56:27 XorA: should i buy xM or beagle bone? Jan 22 06:02:13 trung: depends on what you want to do Jan 22 06:04:47 XorA: just play around.. hobby.. compare it to the gumstix Jan 22 06:04:52 use oe.. whatever Jan 22 06:19:19 bone is cheaper then Jan 22 06:21:58 XorA: which bitbake targets did u use to build uImage, u-boot? Jan 22 06:24:14 console image so I get the filesystem as well Jan 22 06:25:09 oh really? that gives u MLO too? Jan 22 06:25:15 yup Jan 22 06:25:24 OE is smart that wat Jan 22 06:25:25 way Jan 22 06:26:52 XorA: i am not seeing that.. i build core-image-minimal and only got the bz2 file and some other things Jan 22 06:27:01 in ~/oe-core/build/tmp-eglibc/deploy/images Jan 22 06:28:28 but we already established we are using totally different stuff Jan 22 06:29:00 mm so its a behaviour of angstrom recipe? Jan 22 06:29:26 ok, that should be fine then, i am currently building the gumstix receipe, i should expect MLO then Jan 22 06:29:28 angstrom distro has full meta-ti layer, I dont know what you have Jan 22 06:29:31 iuts taking ages.... Jan 22 06:30:36 even on a 4 core machine! Jan 22 06:30:50 yup Jan 22 06:42:40 XorA: i figured out my previous bitbake question Jan 22 06:43:19 use `bitbake -ui knotty2 console-image' Jan 22 06:43:31 it shows a condensed overview of what is building.. quite nice Jan 22 06:50:12 cool Jan 22 07:59:08 trung: knotty2 is default for last few months atleast Jan 22 08:13:47 hrw: i am using gumstix and it uses an older version.. in fact it keeps displaying 4 of 4 .. over and over again.. i think it might be a bug.. but HEAD works as intended Jan 22 08:28:30 good morning Jan 22 08:41:13 Hello all! Jan 22 08:53:13 hi Jan 22 08:56:49 trung: hello Jan 22 09:09:17 mckoan: hows it going? so quiet in here Jan 22 09:11:33 trung: usually more people come later or at US daytime Jan 22 09:12:11 mckoan: been here all day.. no one talks! Jan 22 09:12:18 last 8 hours!! Jan 22 09:12:28 trung: no news good news :-D Jan 22 09:12:34 #gumstix is even worse!!! Jan 22 09:13:02 mckoan: not when you are struggling with trying to build images… :( Jan 22 09:13:24 trung: you could try offering a pint of beer to everyone posting in the next 5 min :-D Jan 22 09:13:45 ha! i'd buy a few beers to anyone who could help me! Jan 22 09:13:55 ~ask Jan 22 09:13:56 Questions in the channel should be specific, informative, complete, concise, and on-topic. Don't ask if you can ask a question first. Don't ask if a person is there; just ask what you intended to ask them. Better questions more frequently yield better answers. We are all here voluntarily or against our will. Jan 22 09:15:51 I have successfully built my gumstix image recipe. I would like to now create a bootable microsd card. Is there any current documentation for gumstix which works? The one at http://www.gumstix.org/create-a-bootable-microsd-card.html is not working for me. Jan 22 09:16:29 gm all Jan 22 09:20:52 trung: have you tried asking to gumstix people? Jan 22 09:21:37 trung: http://www.gumstix.org/community/help.html Jan 22 09:21:39 mckoan: #gumstix only has 10 people. Jan 22 09:21:58 trung: Sign up for the Community Mailing List too Jan 22 09:22:12 mckoan: thanks i have. unfortunately time is not on my side Jan 22 09:24:54 trung: first of all, have you been able to use pre built images? http://www.gumstix.org/software-development/pre-built-images.html Jan 22 09:25:09 mckoan: yes the pre built ones are fine Jan 22 09:25:30 mckoan: i am trying to move to the next stage now which is build my own images… Jan 22 09:27:32 hmmm angstrom-uclibc seem broken Jan 22 09:27:35 my questions are pretty basic.. i can't believe i can't achieve this.. the gumstix docs are all over the place Jan 22 09:30:52 trung: usually (very often) howtos from board producers are not updated to latest OE modifications Jan 22 09:31:12 mckoan: found out the hard way today. you are correct Jan 22 09:31:25 mckoan: its like they don't want us to succeed! Jan 22 09:31:50 trung: switch to another producer then! ;-) Jan 22 09:32:03 mckoan: haha not a current option Jan 22 09:34:45 trung: anyway, IMHO you should explain in detail (in ML or pastebin here) what problem you are facing to Jan 22 09:38:30 mckoan: thanks Jan 22 09:39:02 Bitbake scripts have a lot of declarations in them. Is there some documentation on what they all are and what the all do? Jan 22 09:39:42 sorry bitbake recipes ^^ Jan 22 09:41:53 yes, on yocto website Jan 22 10:02:09 trung: have a look at the reference manual: http://www.yoctoproject.org/docs/current/poky-ref-manual/poky-ref-manual.html Jan 22 10:02:43 particularly chapter 10 (variable glossary) Jan 22 10:03:38 bluelightning: exactly what i was looking for Jan 22 10:03:40 thanks! Jan 22 10:04:13 np Jan 22 11:07:16 Does more than 8gig ram assist OE compile and compute faster? Jan 22 11:07:47 trung: faster processor and faster harddrives are the main factor I believe Jan 22 11:07:58 trung: although more ram never hurts Jan 22 11:08:22 jackmitchell: i've got a i5-2500k 3.3ghz CPU and sata 3 SSD.. still takes time.. Jan 22 11:08:29 trung: how much time? Jan 22 11:08:49 trung: less than 4 hours is a good run for a fresh build Jan 22 11:08:52 i've set the parallelism to 4.. i think a basic console image took about 45 mins still Jan 22 11:09:40 trung: that's fast, you'll be struggling to get any faster unless you upgrade to a better cpu, i.e. a 3 series i7 Jan 22 11:10:16 jackmitchell: thanks. just for comparisons what sort of hardware and timings do you have? Jan 22 11:11:31 trung: it's hard to say, I have a dog slow hdd but an i7-2600 and I also build my own custom images. If I were to do a fresh build of core-image-minimal, I would be pleased if it completed in less than 2 hours Jan 22 11:12:05 jackmitchell: interesting! i highly recommend the SSD Jan 22 11:13:09 trung: yes, I have been considoring it - but bear in mind the strain compiling puts on it. I know Intel were using SSD's for thier OpenEmbedded testing and they would apparently usually kill them within 3 months Jan 22 11:13:16 with a slow disk, memory helps a lot Jan 22 11:13:41 jackmitchell: it depends on the quality of the SSD Jan 22 11:13:44 RP: is that because of the caching linux does? Jan 22 11:14:02 jackmitchell: certainly some of the first/cheap ones were pretty unreliable for heavy use like that Jan 22 11:14:06 RP: are Intel ones not the reliable ones? They wouldn't be using cheaper alternatives now, would they ;) Jan 22 11:14:23 trung: it means it can build up plenty of disk cache Jan 22 11:14:36 jackmitchell: I don't think they were using Intel ones at that time Jan 22 11:14:42 but I agree, SSD's have come a long way recently (unless you're buying OCZ) Jan 22 12:31:44 hello Jan 22 12:32:12 I saw Wolfgang Denx sending patches to OE. had to check again ;D Jan 22 12:33:22 re Jan 22 12:33:27 hrw: heh Jan 22 12:35:43 I just was surpriced Jan 22 12:56:18 is recent oe-core with systemd integration already usable or are there conflicts with oe-meta? Jan 22 13:00:03 ensc|w: systemd recipe and bbclass from meta-systemd still works after oe-core changes Jan 22 13:01:34 thx Jan 22 13:02:23 ensc|w: don't forget to add systemd to DISTRO_FEATURES Jan 22 13:38:51 JaMa, remember my patch to stop gnuradio from building if neon is not in tune features? Jan 22 13:39:11 and I do not suppose you could add an x86 machine to your build :) Jan 22 13:43:16 is "?? =" different from "??=" Jan 22 13:44:00 heh Jan 22 13:44:05 stupid commit message ..... Jan 22 13:44:07 Crofton|work: I'm building x86 locally (but not world) and on jenkins there isn't enough space to build world for more then 1 machine Jan 22 13:44:13 yeah Jan 22 13:44:23 anyway, the code in gnuradio breaks x86 :( Jan 22 13:58:51 so if I set PACKAGECONFIG in a recipe, how do I cahnage it, say from local.conf? Jan 22 14:10:28 Crofton|work: typically PACKAGECONFIG would be set in a bbappend Jan 22 14:10:58 Crofton|work: however with some help from overrides like any other variable it can be set per-package from local.conf or distro config Jan 22 14:11:18 i.e. PACKAGECONFIG_pn-yourrecipe = "some value" Jan 22 14:11:22 it seems like PACKAGECONFIG_gnuradio = "qtgui grc uhd" Jan 22 14:11:22 Jan 22 14:11:24 ah thansk Jan 22 14:11:29 pn-gnuradio Jan 22 14:11:29 * Crofton|work slaps himself Jan 22 14:12:03 I am trying test the cases so I know what the options are for people Jan 22 14:12:16 yet another senior moment Jan 22 14:12:21 ok, np Jan 22 14:12:22 :) Jan 22 14:12:36 btw, I have OE stickers! Jan 22 14:12:43 cool! Jan 22 14:12:47 don't forget to put yourself down on the FOSDEM wiki page btw Jan 22 14:12:52 ah yes Jan 22 14:12:53 http://www.openembedded.org/wiki/FOSDEM_2013 Jan 22 14:15:50 bluelightning, I should put myself in too Jan 22 14:15:59 bluelightning, but I don't know what devices to take Jan 22 14:16:03 denisATeukrea: please do Jan 22 14:16:04 :) Jan 22 14:16:13 bug20 doesn't work anymore because it's stuck at systemd Jan 22 14:16:27 then rest some phones Jan 22 14:16:33 4 are still supported by SHR Jan 22 14:16:48 nokia900,gta04,gta002,crespo Jan 22 14:17:04 well, anything interesting really Jan 22 14:17:26 I think we would probably only want to have one phone on there... gta04 would be good I think Jan 22 14:17:30 somone asked for commercial finished products Jan 22 14:17:40 yes, if possible that would be ideal Jan 22 14:18:06 last time I asked help for gta02, no one stepped in Jan 22 14:18:10 oops Jan 22 14:18:13 bug20 I mean Jan 22 14:18:43 you'll be near the gta04 boost? Jan 22 14:18:46 and the coreboot boost? Jan 22 14:18:53 *booth Jan 22 14:20:50 like the last year would be ideal for me Jan 22 14:21:00 because I've to bring my x60 for coreboot booth Jan 22 14:24:29 denisATeukrea, something that works and was built with OE :) Jan 22 14:24:38 and prefeably fairly self contained Jan 22 14:24:43 ok Jan 22 14:24:50 so GTA04 and GTA02 Jan 22 14:25:00 I can bring at91sam9m10-ekes running wince and very old angstrom Jan 22 14:25:24 same as we had in 2011 in OE raffle organized by Ulf Jan 22 14:25:29 I'll bring stable images tough Jan 22 14:25:30 can they make calls? Jan 22 14:25:37 yes Jan 22 14:25:41 :) Jan 22 14:25:46 but if I bring it then I do not want to take it back home Jan 22 14:25:48 touchscreen is a bit problematic tough Jan 22 14:25:58 debouncing was desactivated to solve a bug Jan 22 14:26:11 hrw: didn't you have vortex486? :P Jan 22 14:26:32 suihkulokki: I have. but it is too crappy to show I think Jan 22 14:27:07 suihkulokki: not to mention fact that I do not remember when I have CF cards to get it booted Jan 22 14:27:11 hrw: I was thinking if you wanted to bring hw you didn't want to take back... Jan 22 14:27:30 hmmm Jan 22 14:27:42 I wonder if I should give away my bugs or not Jan 22 14:27:48 I don't use them that much anymore Jan 22 14:27:52 suihkulokki: you mean: at91sam9m10-ekes, bug11, bug12, sim.one, nhk-15 and other boards? Jan 22 14:28:04 and I'm fedup of trying to fix booting each time Jan 22 14:28:20 denisATeukrea: does someone use them still? Jan 22 14:28:36 no idea Jan 22 14:28:54 I've the 2.0 version Jan 22 14:28:57 complete one Jan 22 14:29:20 lucky bastard Jan 22 14:29:26 I only had prototype Jan 22 14:29:43 ok Jan 22 14:29:51 and had to send it back Jan 22 14:29:55 yes I know Jan 22 14:30:24 ~curse acer for breaking microphone in 722 Jan 22 14:30:25 May you be reincarnated as a Windows XP administrator, acer for breaking microphone in 722 ! Jan 22 14:30:47 I guess you do not have much interrest in it anymore either Jan 22 14:31:19 ok suggestions needed Jan 22 14:31:20 http://cgit.openembedded.org/meta-openembedded/tree/meta-oe/recipes-connectivity/gnuradio/gnuradio_git.bb Jan 22 14:31:28 denisATeukrea: not at all. it is cortex-a8 (bug20) and arm11 (bug1x) and I try to get rid of all the python at the bottom needs to only test if we are doing an arm build Jan 22 14:31:37 ok Jan 22 14:31:59 and for fosdem, working examples of what we can do with OE/Yocto :) Jan 22 14:32:19 Crofton|work, you will bring the E100? Jan 22 14:32:31 maybe Jan 22 14:32:34 ok Jan 22 14:32:35 wow Jan 22 14:32:47 * denisATeukrea only has an rtl_sdr Jan 22 14:32:47 If I can think of a good way to show what I just said :) Jan 22 14:32:58 Crofton|work: so gnuradio is armv7a-neon only? Jan 22 14:33:03 rtl_sdr is very good bang for the buck :) Jan 22 14:33:06 yes Jan 22 14:33:13 there us a section of code that is Jan 22 14:33:26 and the test for neon in cmake is broke atm Jan 22 14:33:31 hopefully I've bought it and saw that it was very dependant on: Jan 22 14:33:36 ->the antenna kind Jan 22 14:33:54 -> the area where you are Jan 22 14:34:08 I wanted to catch ham radios Jan 22 14:34:11 but that failed Jan 22 14:34:17 Crofton|work: use COMPATIBLE_HOST? Jan 22 14:34:26 but it's still usefull Jan 22 14:34:32 where HOST is machine? Jan 22 14:34:39 like for GSM stuff (osmocombb port on top of nuttx) Jan 22 14:34:50 rtl_sdr is a lot like R-Pi Jan 22 14:34:56 good for some things and not others Jan 22 14:35:02 COMPATIBLE_HOST_arm = "broken due neon" + COMPATIBLE_HOST_armv7ahf-neon = "armv7ahf-neon" etc Jan 22 14:35:50 my favorite use of that var is: COMPATIBLE_HOST_aarch64 = "1" which means 'it is broken for aarch64, do not even try' Jan 22 14:35:55 Crofton|work, what can't I do with rtl_sdr that I can do with the newer standalone USRP Jan 22 14:36:01 I mean like the N210 Jan 22 14:36:03 transmit :) Jan 22 14:36:14 ok but I need licenses for that anyway Jan 22 14:36:22 and I've no licenses Jan 22 14:36:23 and the receiver + ADC is better Jan 22 14:36:30 there are ISM bands Jan 22 14:36:43 and you can transmit into a dummy load Jan 22 14:36:46 yes, I know but practically speaking it get more stuff? Jan 22 14:36:47 ok Jan 22 14:36:53 the leakage is enough for short range tests Jan 22 14:36:59 yes Jan 22 14:37:00 wifi would be interesting with N210 Jan 22 14:37:05 dynamic range is helpful Jan 22 14:37:07 ok Jan 22 14:37:18 but, for learning the rtl stuff is great Jan 22 14:37:27 like a not-cheap spectrum analyzer for wifi perturbations Jan 22 14:37:49 basically, when you figure out what the rtl can not do for you, it is time to upgrade Jan 22 14:37:58 ok Jan 22 14:38:15 if I fail to get radio hams with rtl_sdr will the USRP get them? Jan 22 14:38:20 unless you have extra money :) Jan 22 14:38:41 It cuold, but there is no daughter board that I think would be good for hams Jan 22 14:38:45 ok Jan 22 14:38:49 most of their signals are narrow band Jan 22 14:38:54 yes Jan 22 14:38:57 and weak Jan 22 14:38:59 ok Jan 22 14:39:11 so the USRP can digitize 30 MHz of spectrum Jan 22 14:39:20 which is the entire hf allocation :) Jan 22 14:39:23 wow Jan 22 14:39:26 at once? Jan 22 14:39:27 but Jan 22 14:39:32 basically yes Jan 22 14:39:35 wow Jan 22 14:40:10 but there are very strong signals in that range Jan 22 14:40:17 difficult problem Jan 22 14:40:24 ah ok Jan 22 14:40:41 also with rtl_sdr I ear some beeps with the multimode.py Jan 22 14:40:46 any ideas what they are? Jan 22 14:41:06 not sure Jan 22 14:41:08 like continuous beeps Jan 22 14:41:20 ask in #gnuradio, or the rtlsdr channel Jan 22 14:41:29 ok Jan 22 14:41:54 also I wonder how limited I am with USRP if: Jan 22 14:42:06 -> I cannot transmit due to regulations and lack of license Jan 22 14:42:12 -> I want to stay legal Jan 22 14:43:40 basically I really fear that if I buy one it would be very useless Jan 22 14:43:45 so I didn't buy one Jan 22 14:43:56 but rtl_sdr was cheap Jan 22 14:44:03 so it was not that risky Jan 22 14:44:24 The problem with sdr is there are lots of things to learn and it is not magical Jan 22 14:44:35 yes I realized that Jan 22 14:44:45 so real world things like dynamic range and bandwidth are really important :) Jan 22 14:44:51 yes ninquist Jan 22 14:45:16 it is sort of like me and mountain bikes Jan 22 14:45:26 I just bought one with a front shock Jan 22 14:45:40 before the bike limited me, now I am the limiting factor :) Jan 22 14:46:22 ok Jan 22 14:46:24 hopefully that makes some sense :) Jan 22 14:47:18 ok, I need to submit some pacthes Jan 22 14:47:25 and look at the gnuradio neon issue a little Jan 22 14:47:56 ok Jan 22 15:40:01 what class is handling device tree stuff in oe-core? Jan 22 15:43:46 gah linux-yocto.inc Jan 22 15:47:42 do you know of musl support for oe? Jan 22 16:44:37 i noticed ntp is not in oe-core, are there alternatives? Jan 22 16:48:15 OWayne: it's available in meta-networking Jan 22 16:50:03 so it is, was looking in wrong layer, ta. Jan 22 16:50:52 no worries Jan 22 16:51:02 in the not too distant future we'll have a searchable index of recipes Jan 22 17:11:51 kergoth: is there some util function in oe or bb by which I can replace a substring inside a bb variable's RHS ? Jan 22 17:13:45 not that I know of, just getVar() + re.sub() or str.replace() + setVar(), afaik Jan 22 17:14:12 ok Jan 22 17:15:51 it'd be so much more convenient if the __getitem__ returned an expanded value Jan 22 17:16:04 then appendvar == d["foo"] += "bar" Jan 22 17:16:16 or whatever Jan 22 17:16:37 though, the particular case of append, that would actually work.. Jan 22 17:16:53 but obviously taht doesn't help this case Jan 22 17:16:54 heh Jan 22 18:09:06 hi, I'm getting an error from bitbake failing to fetch 'git://git.yoctoproject.org/prelink-cross.git;protocol=git'. cloning it myself worked fine. where should I move it? should I tar it? touch something? to make bitbake happy Jan 22 18:16:19 mnemoc: are you behind any kind of firewall/proxy? Jan 22 18:18:30 bluelightning: no, nothing Jan 22 18:19:10 bluelightning: and `git clone git://git.yoctoproject.org/prelink-cross.git` worked just fine.... just not within bitbake Jan 22 18:19:28 mnemoc: anything in the fetch log to indicate why it failed? Jan 22 18:19:29 so I wanted to know how to "fake" bitbake's download Jan 22 18:21:16 bluelightning: http://sprunge.us/jURc Jan 22 18:22:49 hmm... nothing indicative Jan 22 18:27:37 how can I tell it to skip the mirror and clone directly? Jan 22 18:34:25 it is attempting to clone directly, it's in the first few lines in the log Jan 22 18:38:37 When using mini-x-session and booting into runlevel 5, how do I customize so that the initscripts will start an xsession for particular user with a particular desktop environment ? Jan 22 18:41:22 mnemoc: it seems to me to be a deficiency in the error reporting; something is going wrong but it's not reporting it Jan 22 18:42:32 bluelightning: anything I can export or set to make it verbose? Jan 22 18:42:43 mnemoc: just checking that now Jan 22 18:43:11 mnemoc: which version of git is installed on your system ? Jan 22 18:45:29 1.8.0.3 Jan 22 18:45:54 bluelightning: ---^ Jan 22 18:48:20 mnemoc: would you be able to run the commands at the top of the file and see what happens? Jan 22 18:48:45 (after "DEBUG: Trying Upstream") Jan 22 18:49:02 probably be good to do it in a temporary directory somewhere Jan 22 18:50:44 bluelightning: i made this using angstrom scripts... don't know how to make them use a different dir. but one the cause of the problem is found I can just rm -rf and start over Jan 22 18:51:26 I'm interested in solving the underlying problem Jan 22 18:51:46 sources/bitbake/lib/bb/fetch2/__init__.py:1002: logger.debug(1, "Trying Upstream") Jan 22 18:51:55 so I figured re-running the exact commands it is running would determine where the issue lies Jan 22 18:53:05 you mean the (incomplete) commands in the log or sources/bitbake/lib/bb/fetch2/__init__.py ? Jan 22 18:53:54 the incomplete commands yes: http://pastebin.com/GBYiTw7H Jan 22 18:53:57 mnemoc: ^ Jan 22 18:54:19 hmm actually you'd probably need to cd into the directory extracted from the tarball Jan 22 18:55:29 fatal: bad object 7b47f2f8a15ed13b7905bc120bb2586f3e164f7d Jan 22 18:55:45 err Jan 22 18:55:46 1m Jan 22 18:56:21 f* ... the tar is a bare git Jan 22 18:56:36 without any dir on top :< Jan 22 18:57:40 bluelightning: fails where redefining origin Jan 22 18:57:45 when* Jan 22 18:58:18 it probably shouldb be remove, not prune Jan 22 18:58:52 it doesn't like git remote remote origin either Jan 22 18:59:19 remove* Jan 22 19:01:10 I think the error from the latter is ignored Jan 22 19:03:41 what's the point of that line? Jan 22 19:03:48 origin is already what it should be Jan 22 19:04:05 don't know I'm afraid Jan 22 19:04:13 what happens at the last command? Jan 22 19:04:30 fatal: remote origin already exists. Jan 22 19:04:54 remote.origin.url=git://git.yoctoproject.org/prelink-cross.git Jan 22 19:04:55 remote.origin.fetch=+refs/*:refs/* Jan 22 19:06:07 if bitbake wants to be sure "my" url is used instead of the one used to make the mirror it should use git config, not a duplicated git remote add.... Jan 22 19:06:30 git remote remove leaves tons of orphan branches Jan 22 19:11:00 mnemoc: ah, I see what has happened Jan 22 19:11:59 mnemoc: which bitbake version are you using? Jan 22 19:13:59 got master:b1dbc24ebcc4e5100c32568c2c41fd982fb4bcce Jan 22 19:14:23 there is already a fix for this problem: http://cgit.openembedded.org/bitbake/commit/?id=2ba23df5fad4b94d38a6aed97f7822226d72eb89 Jan 22 19:15:28 that revision is quite old... Jan 22 19:16:39 bbl Jan 22 20:48:05 hello Jan 22 20:50:38 yes Jan 22 20:54:59 We have two hardware platforms for our product, the two use very similar processors on boards from different vendors. Is it better to create one bps layer for both, or two separate layers? Jan 22 20:55:11 bps… damn apple auto-correct Jan 22 20:55:16 BSP Jan 22 20:55:48 hm Jan 22 20:55:54 does not matter Jan 22 20:55:59 use the simpler for your case Jan 22 20:56:27 I would choose 2 bsp's so customer dont get unwanted sru Jan 22 20:56:33 ups stuff Jan 22 20:56:57 but two bsp's might duplicate some work Jan 22 20:57:11 also, one of the designs is a custom main board with a gumstix over attached. I'm trying to figure out how to make my own meta-bsp layer for our board that just slightly modifies anything found in meta-gumstix Jan 22 20:57:39 overo Jan 22 20:58:37 I'm looking at the Yocto docs right now for guidance as well. Jan 22 21:01:26 Can I recursively bbappend a recipe that was bbappend'ed in meta-gumstix? Jan 22 21:01:53 In other words, I want the recipe with their changes, and then I want to make a few changes on top of theirs. Jan 22 21:01:56 I did not check who made contrib/vim for bitbake but great thanks! Jan 22 21:04:53 How is the easier way to resolve conflicts in a patch of a recipe? Jan 22 21:05:24 otavio? Jan 22 21:05:33 woglinde: hey Jan 22 21:05:34 :) Jan 22 21:06:17 hm icu needs fixing for uclibc Jan 22 21:06:35 otavio: if PATCHRESOLVE is user, it'll open a devshell to resolve any conflicts. i don't often use it, but it should still work Jan 22 21:06:54 kergoth: but can I open a devshell byhand? Jan 22 21:06:58 what the hell is __timezone Jan 22 21:06:58 kergoth: grr Jan 22 21:07:06 let me rephrase Jan 22 21:07:18 kergoth: can I use quilt inside the devshell for it? Jan 22 21:08:17 devshell is a task Jan 22 21:08:20 htat task depends on do_patch Jan 22 21:08:32 which means a failed patch application will prevent do_devshell from running Jan 22 21:09:04 could add a new similar task which spawns one after do_unpack and before do_patch, but such a thing doesn't currently exist Jan 22 21:10:23 kergoth: I setted PATCHRESOLVE Jan 22 21:10:26 kergoth: thx Jan 22 21:10:54 np Jan 22 21:19:35 how to get just compiler name? $CC is now 'aarch64....-gcc --sysroot=/path/to/somewhere/' Jan 22 21:20:40 probably want ${HOST_PREFIX}gcc or so Jan 22 21:20:49 * kergoth shrugs Jan 22 21:21:26 thx kergoth Jan 22 23:47:17 do bbappend files stack when used in multiple layers or os is only one used based on layer priority or layer ordering? Jan 22 23:48:05 i.e. can I bbappend a recipe that is already bbappended in another layer? Jan 22 23:48:34 they all apply Jan 22 23:48:41 in an order controlled by layer priority Jan 22 23:48:49 so yes, you can Jan 23 00:21:47 kergoth: thanks for the reply and sorry for the late thanks. Had a small toddler emergency here. Jan 23 00:21:57 np Jan 23 00:36:07 isn't that a daily occurrence with toddlers? Jan 23 00:37:16 and while i'm here, any doc pointers on building for or creating a read-only rootfs? Jan 23 00:37:42 we have a wonky one now, and i got elected to fix it... Jan 23 00:44:05 should probably look at the read-only-rootfs threads on the oe core list recently, someone submitted a patch series to improve the situation **** ENDING LOGGING AT Wed Jan 23 02:59:59 2013