**** BEGIN LOGGING AT Thu Oct 20 02:59:57 2011 Oct 20 05:58:19 ei Oct 20 07:10:37 good morning Oct 20 08:42:31 mornin Oct 20 08:42:51 Jin^eLD: hi Oct 20 08:53:47 pb_, you about? Oct 20 08:59:03 hello, does anyone on here know how I can get bitbake to tell me if it's actually reading the layer.conf files from my specified directories or not? Oct 20 08:59:36 bitbake -e | grep BBFILES just gives me the BBFILES line from local.conf Oct 20 09:03:31 ka6sox: hi, just got to my desk Oct 20 09:05:50 hmmm, bitbake-layers show_layers just spouts a load of deprecation warnings and then prints "INFO: None", wich suggests that no layers are being loaded Oct 20 09:05:58 morning pb_ can you get into opal? Oct 20 09:23:14 let me have a go Oct 20 09:23:38 ... no :-( Oct 20 09:25:21 hello Oct 20 09:25:51 while compiling xserver-kdrive package in oe..... I m getting this msg... http://pastebin.mozilla.org/1360287 Oct 20 09:27:45 snkt: devinde PREFERRED_PROVIDER_virtual/xserver = "xserver-kdrive" Oct 20 09:28:33 JaMa, actually I want to enable Xephyr.... Oct 20 09:28:37 s/vi/fi/g Oct 20 09:40:06 anyone know if there's a way of getting bitbake to tell you which recipe file it will use for a particular target? so if I were to run "bitbake virtual/kernel" is there any way to get it to tell me which recipe it would use? Oct 20 09:45:13 pb_, okay let me instrument up to see what going...give me a minute Oct 20 09:47:41 sebaz: ja :P Oct 20 09:47:44 oops wrong window Oct 20 09:48:51 ? Oct 20 09:51:41 ah, no, that's not the one I was using Oct 20 09:52:15 thats what I have Oct 20 09:52:17 on melo I had three keys installed, and the second one (pb@hetzner) is for the system I am using right now Oct 20 09:52:25 okay let me copy from there. Oct 20 09:52:29 thanks Oct 20 09:53:40 now? Oct 20 09:53:57 works now, awesome. thanks. Oct 20 09:54:49 try getting into garnet Oct 20 09:54:53 fixed that one too. Oct 20 09:55:04 yup, that works too Oct 20 09:56:26 How can I specify my own mirror that should be used as first server where bitbake should look for the sources? Oct 20 09:56:52 pb_, okay when you have time we need to move the mirroring system to opal. Oct 20 09:57:00 righto Oct 20 09:57:04 I can do that later today Oct 20 09:57:12 I should have a spare hour this afternoon Oct 20 09:58:18 tyvm Oct 20 10:00:59 How can I specify my own mirror that should be used as first server where bitbake should look for the sources? -> in OpenEmbedded classic? Oct 20 10:03:20 kobe_: you can use PREMIRRORS in OE classic Oct 20 10:03:38 kobe_: look at classes/angstrom-mirrors.bbclass Oct 20 10:04:38 jamesba: I see you got an answer of sorts on #angstrom... you may find the -s and -v bitbake command line options useful also Oct 20 10:05:45 do I have to change that .bbclass file or can I set the PREMIRRORS somewhere in the configuration files? Oct 20 10:25:00 which package provide libGL.so.1 in oe... Oct 20 10:32:33 kobe_: you can either make a copy of it and then INHERIT += "my-mirrors", or you can just copy its contents into your local.conf if you prefer Oct 20 10:51:53 snkt: mesa Oct 20 10:52:25 snkt: see virtual/libgl Oct 20 10:55:14 JaMa, I got it.... Oct 20 10:56:38 JaMa, can you please help me regarding changing screen resolution xorg.conf file ..... I had builded a x11-image for mini2440...... Oct 20 10:59:53 you can limit supported resolutions in xorg.conf by Section "Screen" -> SubSection "Display" -> Modes "640x480" and then change them with xrandr (if avalable) Oct 20 11:17:25 JaMa, I had tried a lot..... will you provide me a sample xorg.conf Oct 20 11:18:19 there are sample xorg.conf files in oe Oct 20 11:25:41 JaMa: any news about xserver-nodm-init? one side has rootless x and the other has systemd improvements :/ Oct 20 11:26:20 I pointed that today to koen's attenction Oct 20 11:30:08 ant_work: I didn't have time to work on it, so nothing from me Oct 20 11:31:21 np, thx Oct 20 11:31:52 we already did a big step toward unification with your work Oct 20 11:32:42 there is need for cleanings now Oct 20 11:33:27 yes, but I have also few more patches from meta-oe xserver-xorg to new one in oe-core and that's higher on my todo Oct 20 11:38:35 in meta-oe there is much stuff gta-specific while in oe-core I found occurrences of Omap1, Nokia 770/800 and compulab-x270 Oct 20 11:40:08 for the latter I'll ask bluelightning Oct 20 11:40:47 now you're talking about xserver-common, right? because xserver-nodm-init is quite generic Oct 20 11:41:05 yes, looking now at that Oct 20 11:41:45 yes this needs to be moved to bsp layers in some clever way Oct 20 11:42:10 in theory with a formfactor you're done Oct 20 11:42:49 now about tslib...wasn't dumped? Oct 20 11:43:21 iirc the latest kdrive doesn't use tslib Oct 20 11:43:59 I don't care about kdrive.. Oct 20 11:44:14 nor I, why is tslib still there then? Oct 20 11:44:16 but I've fixed xf86-input-tslib to be compatible with new xserver Oct 20 11:44:22 hm... Oct 20 11:44:40 isn't xinput-calibrator the way to go? Oct 20 11:44:49 formfactor doesn't say ie that we don't want cursor Oct 20 11:46:12 ant_work: evdev+xinput-calibrator is way to go if you have kernel driver without jitter Oct 20 11:46:31 ant_work: and without good driver you need evfilter Oct 20 11:46:33 I tested yesterday on poodle and is working very well Oct 20 11:46:37 which is still not in oe Oct 20 11:46:42 it isn't ads7846 though Oct 20 11:47:18 ie jitter on om-gta02 is visible only in some places Oct 20 11:47:29 ie in some configuration dialog in e17 Oct 20 11:47:45 I see, driver... Oct 20 11:48:01 but works fine to launch apps/use virt. keyboard (when it's not scrollable) Oct 20 11:48:16 but now we have patched kernel driver to be jiterless Oct 20 11:48:24 and evdev works great Oct 20 11:49:19 ok, what is the standard for calibration in oe-core then? iirc there was xtscal but not xinput-calibrator Oct 20 11:49:28 I added it by hand Oct 20 11:49:40 is there any magic var? Oct 20 11:50:01 I'm using xinput-calibrator everywhere Oct 20 11:50:13 oe-core doesn't have that Oct 20 11:50:59 and also x11-common in oe-core doesn't use it.. so moving xinput-calibrator to oe-core makes sense only with xserver-common changes.. Oct 20 11:51:17 why? you can just run it Oct 20 11:51:30 or provide a pointercal (with a small patch) Oct 20 11:51:51 xserver-common calls it on xserver start Oct 20 11:52:07 to use stored pointercal-xinput files Oct 20 11:52:46 and? we can give the pointercal in BSP. Anyway you can reach the calibration utility using cursor keys on first boot Oct 20 11:52:54 without that you need to run it after flashing Oct 20 11:53:04 sure Oct 20 11:53:09 and then execute those xinput calls after every xserver start Oct 20 11:53:26 I don't have any cursor keys on om-gta* Oct 20 11:53:35 ok ok .. I did read the scripts and did that... Oct 20 11:53:37 ;) Oct 20 11:53:46 BSPs are providing pointercal-xinput alredy.. Oct 20 11:53:54 those layers which care :) Oct 20 11:54:14 I'd like sane defaults + BSP responsability Oct 20 11:54:54 the Xserver scripts are terribly comlicated by the presence of kdrive compatibility Oct 20 11:57:31 that's why we had xserver-kdrive-common in OE-classic Oct 20 11:59:32 imho the right way is using the formfactor and x11-common/etc/X11/Xserver as default Oct 20 12:00:04 so you see where $ARGS are defined Oct 20 12:00:44 if necessary, override that in your BSP Oct 20 12:01:39 it's bettter but not good enough for me Oct 20 13:09:56 JaMa, which package provides hwinfo Oct 20 13:17:42 what you mean by hwinfo? Oct 20 13:40:16 JaMa, hardware info Oct 20 15:07:16 what is the procedure for generating a SDK using oe-core? Oct 20 15:08:06 i think in theory meta-toolchain should work, but never actually tried any of that stuff myself :\ Oct 20 15:08:46 * Crofton|work is compiling PyQt onan omap3 atm :) Oct 20 15:09:07 * cbrake tries meta-toolchain ... Oct 20 16:11:55 florian: have you seen the multitouch tslib plugin posted to its mailing list? you might be a better choice to review it than me, but it looks pretty promising, and a relatively straightforward modification of the regular input layer event interface plugin Oct 20 16:17:41 kergoth: hi - thanks for your mail :) Oct 20 16:17:52 kergoth: not yet, but sounds promising... Oct 20 16:21:44 cbrake: oe or oe-core? Oct 20 16:28:11 florian: looks quite straightforward, seems to send basically a context switch between touches, so it retains an array of the current x/y/p for each of those contexts. The hard part, naturally, is how for the user calling ts_read to know which one they're dealing with.. Oct 20 16:31:27 otavio: oe-core Oct 20 16:31:35 otavio: meta-toolchain built fine :-) Oct 20 16:32:11 cbrake: yes; be prepared to some problems regarding path Oct 20 16:32:47 cbrake: heh Oct 20 16:38:13 otavio: stuff like this: cpp-static-vars.cpp:1:19: fatal error: stdio.h: No such file or directory Oct 20 16:38:32 cbrake: and more heh; Oct 20 16:38:42 cbrake: has it generated two paths? Oct 20 16:39:35 otavio: as far as I know, only /usr/local/angstrom-eglibc-x86_64-armv5te/ Oct 20 16:39:57 cbrake: in your case it is better then mine Oct 20 16:40:08 otavio: has this been reported? Oct 20 16:40:14 cbrake: ah, angstrom Oct 20 16:40:20 cbrake: that explains much Oct 20 16:40:34 cbrake: yes; long time ago Oct 20 16:40:43 cbrake: but nothing has agreed Oct 20 16:42:15 cbrake: [PATCH 5/8] fix SDK building due TARGET_ARCH use in installation path Oct 20 16:42:26 cbrake: in my case it ends with two paths Oct 20 16:48:38 in trying to debug a problem with bitbake if I use -DD or -DDD bitbake hangs indefinitely, though -D works any advice? Oct 20 16:56:22 Haven't used OE in a while, bitbake 1.12.0 or git? Oct 20 17:02:36 oe-core Oct 20 17:02:46 with bitbake from git Oct 20 17:19:35 what the hell Oct 20 17:19:41 just got this trying to do a send-email: "5.5.2 Syntax error. q57sm14766684yhi.22" Oct 20 17:19:45 what the fuck is that supposed to be Oct 20 17:20:11 5.5.2 Oct 20 17:20:17 hehe Oct 20 17:20:25 read the rfc Oct 20 17:20:27 * woglinde runs Oct 20 17:21:02 kergoth: maybe "Syntax error"? *duck* Oct 20 17:21:24 * kergoth twitches Oct 20 17:22:32 lol Oct 20 18:53:17 oe-core? Oct 20 18:53:20 WHAT IS THIS Oct 20 18:53:55 where have you been in last few months? :) Oct 20 18:55:44 :) Oct 20 18:58:10 B_Lizzard: it seems you woke up from slumber Oct 20 18:58:12 :) Oct 20 18:59:36 Ooooh, stability! Oct 20 19:02:55 * kergoth chuckles Oct 20 20:22:33 interesting Oct 20 23:24:23 hi! could it be that iwmmxt support is somehow broken? Oct 20 23:24:56 I haven't been able to compile gcc-cross when using iwmmxt :-/ Oct 20 23:25:22 OE-classic? Oct 20 23:25:34 if I switched to gcc 4.3.4, I could, but then gtk+ broke Oct 20 23:25:51 JaMa: angstrom, but I'm asking here to know if it is a general problem Oct 20 23:26:05 hmm Oct 20 23:26:50 well, now that I remember, I think I had the gcc-cross problem on org.openembedded.dev Oct 20 23:26:56 (too) Oct 20 23:38:01 Hi, I am trying to build OE for Beagleboard and I'm getting the following errors in my build, does anyone know what may be causing these? Parsing recipes: 100% |###############################################################################| Time: 00:01:16 Oct 20 23:38:01 Parsing of 7333 .bb files complete (6576 cached, 757 parsed). 7980 targets, 315 skipped, 0 masked, 0 errors. Oct 20 23:38:01 ERROR: Openembedded's config sanity checker detected a potential misconfiguration. Oct 20 23:38:01 Either fix the cause of this error or at your own risk disable the checker (see sanity.conf). Oct 20 23:38:01 Following is the list of potential problems / advisories: Oct 20 23:38:03 Please install following missing utilities: makeinfo,svn Oct 20 23:38:05 Using dash as /bin/sh causes various subtle build problems, please use bash instead. Oct 20 23:38:07 ERROR: Command execution failed: Exited with 1 Oct 20 23:38:53 dmitryp: install makeinfo and svn (also called subversion) Oct 20 23:39:09 dmitryp: also dpkg-reconfigure dash and set bash as your default sh Oct 20 23:39:28 you will need root/sudo access Oct 20 23:40:29 i can just get these with apt-get install subversion (then makeinfo) right? Oct 20 23:40:56 it can't find makeinfo or svn but it found subversion Oct 20 23:41:34 subversion *is* svn Oct 20 23:41:50 what about makeinfo? Oct 20 23:41:57 checking Oct 20 23:42:52 dmitryp: install texinfo http://packages.debian.org/search?searchon=contents&keywords=makeinfo&mode=path&suite=stable&arch=any Oct 20 23:44:13 dmitryp: by the way, you can install them in a single aptitude call: aptitude install package1 package2 package3 Oct 20 23:44:24 /quit Oct 20 23:44:46 sorry, is that the command Oct 20 23:45:52 or do you mean package1 is texinfo package2 is subversion.. etc Oct 20 23:47:59 dmitryp: exactly, aptitude install subversion texinfo Oct 20 23:48:54 sweet it didn't crash this time, so will I find the uimage of the kernel in the build/tmp folder? Oct 20 23:49:13 well, you should :) Oct 20 23:49:55 thank you very much Oct 20 23:50:08 you're welcomed Oct 20 23:50:29 does this build include the rootfs? Oct 20 23:51:07 dmitryp: you haven't said exactly what you are building Oct 20 23:51:14 ie, the bitbake call Oct 20 23:54:20 oh sorry it was the MACHINE=beagleboard ./oebb.sh bitbake virtual/kernel Oct 20 23:54:38 i'm guessing it's just going to build me a uImage.bin Oct 21 00:07:44 dmitryp: indeed, just a kernel Oct 21 00:08:09 dmitryp: check openembedded/recipes/images for images Oct 21 00:08:32 * lisandro doesn't knows wich is a suitable beagleboard image Oct 21 00:17:01 use console-image or systemd-gnome-image Oct 21 00:30:32 how do I use those? Oct 21 00:30:40 or where do I unput those? Oct 21 00:32:04 I have tons more errors from the build Oct 21 00:32:13 ERROR: Function 'Fetch failed: ftp://ftp.gnu.org/gnu/binutils/binutils-2.20.1.tar.bz2;name=tarball checksum mismatch.' failed Oct 21 00:32:13 ERROR: Task 501 (/home/dmitry/dev-beagle/sources/openembedded/recipes/binutils/binutils-cross_2.20.1.bb, do_fetch) failed with exit code '1' Oct 21 00:32:24 ERROR: '/home/dmitry/dev-beagle/sources/openembedded/recipes/binutils/binutils-cross_2.20.1.bb' failed Oct 21 00:34:22 can't find a uimage in the build folder either Oct 21 00:35:25 I'm trying to get this camera to run on the beagleboard xm, and use some openCV validation program to test it and make my own, maybe I'm doing this the wrong way? http://blog.galemin.com/2011/04/li-5m03-camera-on-beagleboard-xm/#comment-634 Oct 21 00:35:36 thanks, Oct 21 00:44:18 dmitryp: I've hitted the binutils stuff too Oct 21 00:44:26 the checksum of the tarball has changed Oct 21 00:44:36 you will have to change them in the recipe Oct 21 00:45:28 oh ok, well I already have binutils installed, is that all the command does? Oct 21 00:45:36 wha tdo I change the checksum to? Oct 21 00:45:55 dmitryp: I think you will need to read openembedded's doc Oct 21 00:49:21 hmm, I can't find anything about the checksum in the recipe Oct 21 00:49:41 can I ask you if this is the right direction to go for getting that image sensor working with the beagleboard? Oct 21 00:49:42 dmitryp: binutil's recipe Oct 21 00:49:54 dmitryp: well, you seem quite confused Oct 21 00:50:14 yes, the directions on that link are very vague Oct 21 00:50:14 maybe this is not the best choice Oct 21 00:51:21 dmitryp: as I see it, you will need to invest some time into understanding how oe works Oct 21 00:51:32 I don't know another solution Oct 21 00:52:59 dmitryp: does the camera work on a standard desktop system? Oct 21 00:53:15 mm, seems not Oct 21 00:54:45 yeah it's MiPi Oct 21 00:57:33 the other kernel, the angstrom validation came with make instead of bitbake and I had no problem compiling it, is there a source distribution of OE that uses make? Oct 21 00:59:57 dmitryp: try #angstrom Oct 21 01:18:24 dmitryp: angstrom doesn't use make instead of bitbake. angstrom has a makefile wrapper around execution of bitbake to make it more convenient. that's all Oct 21 01:19:14 dmitryp: further, most of hte things built out of oe use make themselves. bitbake *runs* make to build individual recipes. it just handles the level above that. **** ENDING LOGGING AT Fri Oct 21 02:59:57 2011