**** BEGIN LOGGING AT Wed Sep 29 02:59:57 2010 Sep 29 05:52:08 gm Sep 29 07:24:28 morning Sep 29 07:43:52 good morning Sep 29 07:56:07 hello Sep 29 07:56:23 i m new to oe Sep 29 07:56:53 m trying to build one recipe Sep 29 07:57:17 nd i want to call my Makefile from my recipe Sep 29 07:57:31 how can i do that Sep 29 07:57:35 ? Sep 29 08:07:46 what a patience :p Sep 29 08:08:06 :-) Sep 29 08:31:30 hmm... stupid rpath Sep 29 08:50:52 good morning Sep 29 08:53:38 florian: hi Sep 29 09:15:54 03Thilo Fromm  07org.openembedded.dev * rd318d0667d 10openembedded.git/recipes/wvstreams/wvstreams_4.6.bb: (log message trimmed) Sep 29 09:15:54 wvstreams-4.6: Remove dbus support (breaks build) Sep 29 09:15:54 If wvstreams is configured with DBus support enabled (by not providing Sep 29 09:15:54 --without-dbus) the the build fails at link time as libdbus cannot be Sep 29 09:15:54 found: Sep 29 09:15:54 | ./CXX -c dbus/wvdbusconn Sep 29 09:15:54 | ./CXX -c dbus/wvdbusmarshal Sep 29 10:01:40 hm.. mupdf build failed.. Sep 29 10:03:21 03Dmitry Eremin-Solenikov  07org.openembedded.dev * ra0fcaa97de 10openembedded.git/recipes/t1lib/ (4 files in 2 dirs): Sep 29 10:03:22 t1lib: update to 5.1.2 Sep 29 10:03:22 t1lib 5.0.2 is so old, it was removed from main Debian mirrors. Update Sep 29 10:03:22 to current t1lib. Sep 29 10:03:22 Signed-off-by: Dmitry Eremin-Solenikov Sep 29 10:03:22 03Dmitry Eremin-Solenikov  07org.openembedded.dev * rcfb5385859 10openembedded.git/: Merge branch 'master' of git://git.openembedded.org/openembedded into org.openembedded.dev Sep 29 10:03:26 03Dmitry Eremin-Solenikov  07org.openembedded.dev * r99797a4769 10openembedded.git/recipes/fortune-mod/fortune-mod_1.99.1.bb: Sep 29 10:03:26 fortune-mod: update to 1.99.1-4 and drop host-only tools Sep 29 10:03:26 Old debian patch (1.99.1-2) was removed from Debian mirrors. Update to Sep 29 10:03:26 current patch (1.99.1-4). Also fix QA errors by removing host tools Sep 29 10:03:34 from target pakages (strfile/unstr). Sep 29 10:03:34 Signed-off-by: Dmitry Eremin-Solenikov Sep 29 10:03:34 03Dmitry Eremin-Solenikov  07org.openembedded.dev * rfe3c6c50f7 10openembedded.git/recipes/libqpe/ (libqpe-opie/double_name.patch libqpe-opie_1.2.4.bb): Sep 29 10:03:34 libqpe-opie: copy double_name.patch from opie-mediaplayer1 Sep 29 10:03:35 Signed-off-by: Dmitry Eremin-Solenikov Sep 29 10:03:35 03Dmitry Eremin-Solenikov  07org.openembedded.dev * r9819d93bbb 10openembedded.git/recipes/opie-taskbar/ (5 files): Sep 29 10:03:36 opie-taskbar-images: add prerm script removing alternatives Sep 29 10:03:36 Signed-off-by: Dmitry Eremin-Solenikov Sep 29 10:03:37 03Dmitry Eremin-Solenikov  07org.openembedded.dev * rc9391dfd9f 10openembedded.git/recipes/recode/recode_3.6.bb: Sep 29 10:03:37 recode: use debian orig.tar.gz instead of gnu one Sep 29 10:03:38 recode-3.6.tar.gz was removed from gnu mirrors (moved to old-gnu). Sep 29 10:03:38 Use recode_3.6.orig.tar.gz from Debian which is the same file. Sep 29 10:03:54 Signed-off-by: Dmitry Eremin-Solenikov Sep 29 10:27:01 hi, what kind of instruction is that: push {r4, r5, r6, r7, r8, lr} ? neon? Sep 29 10:27:11 it's not stock arm instructions Sep 29 10:28:53 can't find -lfreetype.. Sep 29 10:29:05 looks like lost dependency Sep 29 10:31:03 no.. freetype is there.. Sep 29 10:31:13 well, ppl, mupdf fails to build Sep 29 10:31:51 http://tinderbox.openembedded.net/public/logs/task/8205889.txt Sep 29 11:12:05 Hi, I'm trying to bitbake a new recipe. It works fine when I hardcode something like "-I /opt/include" in one of the App's Makefile.am. Is there a way I can get rid of this hardcoding by exporting a variable "INC=/opt/include" before bitbaking and changing the Makefile.am entry to "-I $(INC)" ? Sep 29 11:12:44 hi , my wifi always closing itself when not used long , and closing my ssh connection , when i return back , i see my ssh closed and wifi power down , then i need to open wifi back, how can i prevent this happen ? Sep 29 11:12:56 screwgoth1 CFLAGS Sep 29 11:13:18 screwgoth1 read the automake manual Sep 29 11:13:36 hi , my wifi always closing itself when not used long , and closing my ssh connection , when i return back , i see my ssh closed and wifi power down , then i need to open wifi back, how can i prevent this happen ? i tried "iwconfig wlan0 power off" but its not helping me too , any more help ? please help Sep 29 11:14:56 woglinde_: ok. thanks. I'll look it up. But from what I gather , passing CFLAGS option to EXTRA_OECONF variable in the recipe, should work, right ? Sep 29 11:15:18 no Sep 29 11:15:32 EXTRA_OECONF is for configure options Sep 29 11:18:41 woglinde_: yes. I know that. I don't want to touch the application's source files, which is where I would usually define CFLAGS. If I pass it as a configure option, I can ask users to change only the recipe Sep 29 11:19:07 o.O? Sep 29 11:22:30 screwgoth1: so you have a binary only library that you need to link against? the easiest way would be to 'stage'/install that library as a recipe, and then have the app DEPEND on this recipe Sep 29 11:22:53 screwgoth1: it will also help with deploying your application (e.g. when the app gets installed, the library will be installed automatically) Sep 29 11:23:31 woglinde_: :-) . Ok I know I might not have made much sense there. But basically, my whole aim is to make a recipe which users can configure with their specific include paths without touching the patch Sep 29 11:24:01 zecke: I'm not really doing that. this is slightly more convoluted in terms of the way I want it to be deployed Sep 29 11:24:06 what users? Sep 29 11:24:15 users installing .ipk Sep 29 11:25:34 screwgoth1: i think you need to provide more context. Right now the closest thing I can think of is the PowerVR library for OMAP3. Users will need to download the tarball themselves and place it in the DL_DIR, the rest will then work Sep 29 11:25:52 zecke nope Sep 29 11:25:55 :) Sep 29 11:26:06 sgx has now a freely downloadable url Sep 29 11:26:54 zecke: I know... my mistake for not providing the entire context. But woglinde_'s suggestion of CFLAGS actually might solve my problem. Maybe I can post a write-up about my issue when I resolve it Sep 29 11:52:42 when i bitbake "bitbake libxml-parser-perl" Sep 29 11:52:50 then it gives error: Sep 29 11:53:10 ERROR: Required build target 'libxml-parser-perl' has no buildable providers. Sep 29 11:53:10 Missing or unbuildable dependency chain was: ['libxml-parser-perl', 'expat-native'] Sep 29 11:54:37 hm expat-native should be there Sep 29 11:54:46 are you using latest dev? Sep 29 11:54:58 yes Sep 29 12:01:25 Hi Sep 29 12:01:56 Say I want some MACHINE-specific recipes which don't fit in any particular category, where do I put those? Sep 29 12:03:29 Custom init scripts, stuff like that Sep 29 12:04:24 B_Lizzard: some zaurus stuff is laying in recipes/zaurus-utils Sep 29 12:04:38 but not sure about proper way Sep 29 12:05:13 I hate having stuff outside of OE, and I'd hate to clutter OE with stuff which is irrelevant Sep 29 12:08:14 i want to call makefile of my extracted source code from my recipe Sep 29 12:08:32 how can i do tht in oe? Sep 29 12:11:26 thanks Sep 29 12:13:56 arti call? Sep 29 12:14:06 arti autotools? Sep 29 12:14:28 means in my source code i have makefile Sep 29 12:14:29 arti you read the manual? Sep 29 12:14:54 not whole Sep 29 12:15:17 wht shold i write in my makefile Sep 29 12:15:54 sorry in my recipe Sep 29 12:59:41 morning Sep 29 13:01:51 gm Sep 29 13:25:56 hi , is wvdial buggy ? i cant use wvdial to make 3g gprs connection , vdial: utils/wvtask.cc:202: WvTaskMan::WvTaskMan(): Assertion `getcontext(&get_stack_return) == 0' failed. how can i fix this bug ? Sep 29 13:30:04 it is Sep 29 13:30:28 I have seen this error before... seems to be specific for ARM architecture. Sep 29 13:30:40 An older one worked for me... Sep 29 13:33:29 hey ka6sox, any news from the freenx front? Sep 29 13:35:12 zecke: I can reproduce the iwmmxt problem with 2.6.34.6; but only with './a | ./a' pipes. I can not reproduce it by forking and calling ./a from a C program although a similar sequence of syscalls is done there :( Sep 29 13:35:56 eFfeM_work, got it to build for i686 with oe now fixing things for ARM. Sep 29 13:36:07 cool Sep 29 13:37:29 btw my building of testing-next for my platforms was successful except for neek, this one will probably work next week when denix' patch for SRC_URI is in Sep 29 13:44:20 zecke: https://www.cvg.de/people/ensc/x1.s is the ./a program Sep 29 13:47:11 hi , is wvdial buggy ? i cant use wvdial to make 3g gprs connection , vdial: utils/wvtask.cc:202: WvTaskMan::WvTaskMan(): Assertion `getcontext(&get_stack_return) == 0' failed. how can i fix this bug ? Sep 29 13:49:00 JDuke128: there is not really a point in repeating your Q (and florian already gave you an answer) Sep 29 13:50:48 btw the way to fix a bug is: debug the program, analyse the root cause, fix the problem, test it, submit patch to mailing list :-) Sep 29 13:54:59 03Dmitry Eremin-Solenikov  07org.openembedded.dev * r015e393510 10openembedded.git/recipes/gpm/ (gpm-1.99.7/processcreds.patch gpm_1.99.7.bb): Sep 29 13:54:59 gpm: fix compilation due to missing struct ucred Sep 29 13:54:59 Signed-off-by: Dmitry Eremin-Solenikov Sep 29 14:06:46 florian , how did you used 3g modem without wvdial ? Sep 29 14:07:07 JDuke128: I never did this... Sep 29 14:07:53 how can i make gprs/3g usb connection without wvdial ? someone know ? Sep 29 14:07:55 I cursed at a customer I didn't manage to convinve no to use this.. ;) Sep 29 14:08:21 shpuld be pretty easy with a simple chat script Sep 29 14:08:34 ppp ? Sep 29 14:08:41 is there any tutorial for that ? Sep 29 14:10:27 JDuke128: do you have bloated stuff like NetworkManger in your image, that can wrap the dial side of 3G and works well enough. Sep 29 14:11:52 DJWillis , yes i ve it Sep 29 14:12:05 i can see 3g modem on /dev/ttyUSB0 Sep 29 14:15:03 03Dmitry Eremin-Solenikov  07org.openembedded.dev * rad701be682 10openembedded.git/recipes/gdb/gdb-cross-sdk.inc: Sep 29 14:15:03 gdb-cross-sdk: fix compilation on non-arm platforms Sep 29 14:15:03 Not all platforms are arm. Not all platforms have sim/arm/Makefile for Sep 29 14:15:03 sed. Grrr. Sep 29 14:15:03 Signed-off-by: Dmitry Eremin-Solenikov Sep 29 14:15:04 Cc: Eric BĂ©nard Sep 29 14:15:04 Cc: Tom Rini Sep 29 14:56:07 ensc|w: wonderful, from a quick look at the iwmmxt.S it seems they try hard to avoid saving/restoring context if it is not needed Sep 29 15:30:06 hi i am trying to bitbake but am getting this error Sep 29 15:30:08 ERROR: Please set the 'PERSISTENT_DIR' or 'CACHE' variable. Sep 29 15:30:16 i gooogled Sep 29 15:30:21 but no answer Sep 29 15:30:22 your BBPATH isn't set correctly or isn't set at all Sep 29 15:30:22 \?? Sep 29 15:30:25 or you forgot to export it Sep 29 15:30:38 it's using the bitbake conf/bitbake.conf rather than the OE one Sep 29 15:30:56 i have edited the .bashrc file Sep 29 15:31:25 r0ck_: nice, but as kergoth said. your BBPATH is wrong or not exported. Sep 29 15:31:26 which tells me nothing Sep 29 15:32:02 # end oe setup # # # Environment setup for oe # export BB_ENV_EXTRAWHITE="MACHINE DISTRO ANGSTROM_MODE OVEROTOP OEBRANCH USERBRANCH TITOOLSDIR" export OVEROTOP="${HOME}/overo-oe" export OEBRANCH="${OVEROTOP}/org.openembedded.dev" export USERBRANCH="${OVEROTOP}/user.collection" export PATH="${OVEROTOP}/bitbake/bin:$PATH" export BBPATH="${OVEROTOP}/build:${USERBRANCH}:${OEBRANCH}" export TITOOLSDIR="${OVEROTOP}/ti" umask 0002 # # end Sep 29 15:32:14 the bashrc file Sep 29 15:32:46 r0ck_: well, you didn't export or don't have the correct path in it. Sep 29 15:32:58 hmmm Sep 29 15:33:21 r0ck_: and i think your actual bashrc content is different, so you might have added something which is not there (humans tend to do these things). Sep 29 15:33:41 r0ck_: echo $BBPATH, what does it look like? Sep 29 15:33:54 i have appended the above PATHs to my bashrc file Sep 29 15:34:18 warcraft@warcraft-desktop:~/overo-oe$ echo $BBPATH /home/warcraft/overo-oe/build:/home/warcraft/overo-oe/user.collection:/home/warcraft/overo-oe/org.openembedded.dev Sep 29 15:35:08 is this fine? Sep 29 15:35:56 r0ck_: well, tell me (you have access to your machine). do the paths you list exist? Sep 29 15:36:05 r0ck_: which of them have a conf/bitbake.conf subdirectory? Sep 29 15:37:09 thats in overo-oe/bitbake/bitbake.conf Sep 29 15:37:48 r0ck_: I said conf/bitbake.conf, so either /home/warcraft/overo-oe/org.openembedded.dev/conf/bitbake.conf or somewhere else Sep 29 15:38:45 ya its there also Sep 29 15:39:18 @zecke its there Sep 29 15:39:19 r0ck_: well, my question was. Which of the directories in your BBPATH have a conf/bitbake.conf file? Sep 29 15:41:53 when i typed echo $BBPATH Sep 29 15:42:30 i got thid Sep 29 15:42:33 this* Sep 29 15:43:42 none Sep 29 15:43:44 .. Sep 29 15:44:09 export BBPATH="${OVEROTOP}/build:${USERBRANCH}:${OEBRANCH}" Sep 29 15:44:18 this is my BBTH Sep 29 15:44:30 *BBPATH Sep 29 15:44:37 @zecke Sep 29 15:44:41 u there Sep 29 15:46:14 r0ck_: I have asked a simple question. Your BBPATH contains three paths, I ask you.. which of these three have conf/bitbake.conf subpath? Sep 29 15:46:34 and of course, i assume you've actually spawned a new shell to get the updated bashrc, or sourced it manually? Sep 29 15:47:51 none of them contain conf/bitbake Sep 29 15:48:12 OE contains conf/bitbake.conf. if your oe doesn't, you're doing something wrong. Sep 29 15:48:43 r0ck_: if it is only the org.openembedded.dev/conf/bitbake.conf, you either don't have spawned a shell with the BBPATH set or we will have something else Sep 29 15:50:57 but the same setting was working on another machine Sep 29 15:52:02 computers aren't magic. Sep 29 15:52:09 wow now its working Sep 29 15:52:12 something in your setup is different, or it would be working Sep 29 15:52:17 how its possible Sep 29 15:52:24 you changed something. Sep 29 15:52:30 r0ck_: do you still have the original shell open? Sep 29 15:52:55 08:45 < kergoth_> and of course, i assume you've actually spawned a new shell to get the updated bashrc, or sourced it manually? Sep 29 15:53:13 do we have an FAQ on the wiki? the PERSISTENT_DIR/CACHE one seems pretty common Sep 29 15:53:37 do u have the link Sep 29 15:53:40 for that wiki Sep 29 15:53:51 see the topic of this channel Sep 29 15:54:25 one more question ?? Sep 29 15:54:53 if you have a question, ask it Sep 29 15:54:55 what is spawned a new shell means?? Sep 29 15:54:58 r0ck_: sure, just ask Sep 29 15:55:09 run a new terminal.. Sep 29 15:55:21 r0ck_: executed bash, e.g by opening a new terminal window Sep 29 15:55:25 the one you're in doesn't magically get the changes from the .bashrc unless you source it manually or run a new terminal Sep 29 15:56:07 oh man Sep 29 15:56:08 zecke: http://github.com/kergoth/bitbake/commit/0c87dae7c4aefe8375ca9f397f98daf4626758ff seem reasonable to you? Sep 29 15:56:17 ya i got it Sep 29 15:56:31 i just ran the whole thing on that terminal only Sep 29 15:56:53 but it was written there in that docs to restart the terminal Sep 29 15:56:56 kergoth_: oh, I have never worked with that code. Sep 29 15:57:38 its just a change in API Sep 29 15:58:01 @zecke and @ kergoth thnaks for helping me out Sep 29 15:58:36 np Sep 29 15:59:12 bye Sep 29 15:59:49 kergoth_: looks good Sep 29 16:29:34 re Sep 29 16:34:45 hi Sep 29 16:40:42 hi ericben Sep 29 16:41:53 hi woglinde Sep 29 16:42:16 preparing an answer to a guy which believes OE is not a serious solution for industrial solutions and that Debian is the way to go :-) Sep 29 16:42:34 in summary the answer is "go for debian and forget me " :-D Sep 29 16:42:43 sure Sep 29 16:43:12 I always like people like this which explain your job when their activity goes from web design to what they call industrial solutions Sep 29 16:43:25 *g* Sep 29 16:43:29 webdesign Sep 29 16:43:32 php and mysql Sep 29 16:43:46 real industrial peoples understand the interest of OE because they have real industrial problems Sep 29 16:44:04 dont tell him webos is using oe Sep 29 16:44:07 sometimes they can adopt it because of management's choices or things like that Sep 29 16:44:18 and Atmel, Ti and others and using it ;-) Sep 29 16:44:25 but that's not industrial peoples Sep 29 16:44:35 they don't design their website themself :-D Sep 29 17:18:44 ericben: what sort of industrial solution do they want Sep 29 17:18:48 btw gm all Sep 29 17:19:54 Hi, I could use some clarification on *.bbappend files Sep 29 17:20:26 dickelbeck: read my blog http:/sakrah.homelinux.org Sep 29 17:20:32 Can I put them in one overlay and have them use relative references to another overlay's recipes? Sep 29 17:20:40 blog, yeah Sep 29 17:21:02 info has been very scarce. Sep 29 17:21:27 dickelbeck: feel free to contribute to wiki Sep 29 17:21:55 oe wiki? Sep 29 17:22:13 yes Sep 29 17:22:17 or documentation Sep 29 17:22:20 whichever Sep 29 17:22:43 hi khem: well not many things a few apps. But debian has stable in one of it's release and there is 1GB of flash on the board so that's a good solution :-) Sep 29 17:24:19 to be fair, there is one reason to prefer debian over oe, and that's reproducibility. since they don't crosscompile, their builds are less prone to failures caused by the build machine environment. nature of the beast Sep 29 17:25:35 not false but one you setup a build server (or a virtual machine for this) you can keep a stable configuration for the build of a release Sep 29 17:26:01 yeah, that's true. separate build server or chroot will get you there Sep 29 17:26:52 kergoth_: hmmm do they use qemu for armel builds? Sep 29 17:26:57 I keep thinking about doing builds in the cloud Sep 29 17:27:03 the problem is RAM & Flash are to cheap today, that prevent people from thinking to optimization ;-) Sep 29 17:27:40 ericben: thats a shallow thought IMO better code does always better Sep 29 17:28:21 at Cisco when the size of flash outgrew 32M it was a big expense for new production routers Sep 29 17:29:09 and in these days margins are low :) Sep 29 17:29:10 yes for large companies that's always a problem, but for peoples doing 100-1k products, adding 1 or 2 $ to get 1GB instead of 256MB of NAND is not always a problem Sep 29 17:29:15 GNUtoo|laptop: around ? Sep 29 17:29:27 yes Sep 29 17:29:32 mapping keyboard but arround Sep 29 17:29:58 Did you have a chance to try the gcc with the fix I pushed yesterday which fixed the mplayer issue ? Sep 29 17:30:07 not yet Sep 29 17:30:27 GNUtoo|laptop: hmmm I am waiting on you before I merge it into master Sep 29 17:30:35 you had some niggles Sep 29 17:30:47 ok Sep 29 17:30:54 I'll try very soon then Sep 29 17:30:56 as soon as I can Sep 29 17:31:03 the only issue is this one: Sep 29 17:31:11 we use a branch on SHR for the n900 Sep 29 17:31:18 in order to prevent irreversible damage to oe Sep 29 17:31:19 GNUtoo|laptop: my images has been building and booting well on efikamx Sep 29 17:31:30 like going backward kenrel versions Sep 29 17:31:32 so.... Sep 29 17:31:43 I don't remember how to rebase current oe against that branch Sep 29 17:31:46 how can I do that? Sep 29 17:31:46 GNUtoo|laptop: ok cherry pick the gcc fixed Sep 29 17:31:48 fixes Sep 29 17:31:50 ok Sep 29 17:31:55 I'll do that then Sep 29 17:32:08 so Sep 29 17:32:12 there are 2 patches you need Sep 29 17:32:18 I checrry pick(which patches?) Sep 29 17:32:23 then I rebuild from scratch? Sep 29 17:32:30 or rebuild only mplayer or specific packages? Sep 29 17:33:08 http://git.openembedded.org/cgit.cgi/openembedded/commit/?h=khem/import-linaro-gcc-4.5&id=b42902fe8ab2160cb9b01c468a7670fd88ce9238 Sep 29 17:33:18 http://git.openembedded.org/cgit.cgi/openembedded/commit/?h=khem/import-linaro-gcc-4.5&id=fa11285ca5b6fcdd9f91ac548ae275ea076201bd Sep 29 17:33:44 GNUtoo|laptop: I would suggest to build mplayer only and ffmpeg Sep 29 17:33:46 first Sep 29 17:33:49 ok Sep 29 17:34:10 and for better if you have cycles then do a build from scratch that will be very helpful too Sep 29 17:36:44 it's too late to rebuild from scratch now Sep 29 17:36:47 but maybe tomorrow Sep 29 17:39:11 khem, what flags should I use? Sep 29 17:39:12 -mvectorize-with-neon-quad ? Sep 29 17:39:18 -ftree-vectorize ? Sep 29 17:39:28 GNUtoo|laptop: no workarounds needed Sep 29 17:39:28 -f-no-tree-vectorize Sep 29 17:39:38 ah ok so the default-in-oe one Sep 29 17:39:39 remove all workarounds that are there Sep 29 17:39:42 yep Sep 29 17:39:45 thanks Sep 29 18:07:39 hi i have my recipe in which i want to compile my source code with makefile. so wht should i write in my recipe? Sep 29 18:09:54 knock knock ... wake up! Sep 29 18:10:11 there are 7,000+ recipes to read Sep 29 18:11:17 kergoth, i dint get you Sep 29 18:11:35 what is there to get? there are over 7,000 recipes you can read to see how things are done. Sep 29 18:20:02 kergoth, i knw tht nd m following opkg recipe nd want to build the image but in my case things are not working Sep 29 18:20:48 if you have a specific problem, then ask about it, and preferably link to the exact error / log in a pastebin Sep 29 18:20:59 re Sep 29 18:23:14 ok will ask tomorrow Sep 29 18:33:11 Aarti: how familiar are you with OE Sep 29 18:33:26 m new to oe Sep 29 18:33:52 ok then you should first try to build an exisitng image Sep 29 18:34:24 learn OE nomenculature Sep 29 18:35:52 then you can look into existing recipes e.g. recipes/helloworld/helloworld_1.0.0.bb Sep 29 18:35:59 is a good example to start with Sep 29 18:36:25 and spend time reading documentation, wiki and mailing list archives Sep 29 18:36:47 khem, i knw tht nd i have done tht example Sep 29 18:37:43 Aarti: if you have done all what I said you will be asking more specific questions Sep 29 18:38:59 khem, my problem is I m fetching the source code from local svn repository nd unpacking in tmp/work/armv7.../ Sep 29 18:39:19 nd thr i have Makefile in sourcecode Sep 29 18:40:24 I want to compile my sourcecode with that makefile from my recipe,so wht i have to write in my recipe Sep 29 18:40:44 thats a common case Sep 29 18:40:56 look into other recipes they all do that Sep 29 18:41:39 ya m following opkg recipe but i m not getting my image Sep 29 18:42:27 opkg uses autotools Sep 29 18:42:34 does your app used autools ? Sep 29 18:42:41 if not then its not right sample Sep 29 18:44:55 Aarti: look into recipes/zeroconf/zeroconf_0.9.bb e.g. Sep 29 18:44:55 no my app is not using autotools Sep 29 18:45:03 that might be more useful Sep 29 18:46:09 khem, thnx Sep 29 18:55:56 khem: to use *.bbappend files, is there a minimum version of bitbake required? I am currently using version 1.8.18 from stock Ubuntu Lucid. Sep 29 18:57:25 can anyone else say? Sep 29 19:00:28 dickelbeck: master. Sep 29 19:00:32 dickelbeck: you need git/master version Sep 29 19:00:35 neither 1.8 nor 1.10 support it Sep 29 19:01:16 dickelbeck: btw dont use stock bitbake always use it from git even if you use a release IMO Sep 29 19:03:16 more convenient to try updates, test fixes, etc that way Sep 29 19:07:05 thanks re: bitbake Sep 29 19:18:07 hi Sep 29 20:08:22 khem, whats the state of your gcc 4.5/liniaro work? Sep 29 20:08:51 Crofton: I have pushed it to branch Sep 29 20:08:55 Its ready for merge Sep 29 20:09:37 when do you think you will merge it? Sep 29 20:09:46 Crofton: I am waiting for word from GNUtoo|laptop and JaMa who are using it to generate complete rebuilt of SHR for N900 Sep 29 20:09:54 ok Sep 29 20:10:08 Crofton: may be at the end of this week Sep 29 20:10:10 ah sorry I went to eat and forgott about it Sep 29 20:10:28 khem, I remember now Sep 29 20:10:38 GNUtoo|laptop: no worries Sep 29 20:10:45 about ffmpeg Sep 29 20:10:59 I'll rebuild it and it'll break my other programs Sep 29 20:11:06 but mplayer is more important Sep 29 20:11:12 and gnash is broken anyway Sep 29 20:11:14 so... Sep 29 20:11:19 s/my/the Sep 29 20:12:03 Crofton: on my end I have tested it quite a bit and it has withstood all Sep 29 20:12:20 I dont have multimedia sort of things though Sep 29 20:13:43 03Frans Meulenbroeks  07org.openembedded.dev * r4a44cef2da 10openembedded.git/recipes/udev/files/ (fix-alignment.patch tmpfs.patch): Sep 29 20:13:43 udev : moved unused files to obsolete dir Sep 29 20:13:43 Signed-off-by: Frans Meulenbroeks Sep 29 20:13:43 03Frans Meulenbroeks  07org.openembedded.dev * r17c21c62b4 10openembedded.git/recipes/unionfs/ (3 files in 3 dirs): Sep 29 20:13:44 unionfs : moved unused files to obsolete dir Sep 29 20:13:44 Signed-off-by: Frans Meulenbroeks Sep 29 20:15:08 Crofton: primarily there are two patches to 4.5 Sep 29 20:15:17 if you wanna try Sep 29 20:15:32 if it gets merge I will try Sep 29 20:15:36 | /tmp/ccxKaFpV.s:377: Error: VFP/Neon double precision register expected -- `vmovl.u8 q0,s16' Sep 29 20:15:41 if I get bored, I will try the branch :) Sep 29 20:15:56 GNUtoo|laptop: I think you do not have my second patch Sep 29 20:16:18 GNUtoo|laptop: that problem should be fixed with the second patch to gcc Sep 29 20:16:22 ok Sep 29 20:16:25 I'll look Sep 29 20:17:03 thats the top patch on http://git.openembedded.org/cgit.cgi/openembedded/commit/?h=khem/import-linaro-gcc-4.5&id=fa11285ca5b6fcdd9f91ac548ae275ea076201bd Sep 29 20:17:15 you need to cherry pick that Sep 29 20:17:19 and then rebuild gcc-cross Sep 29 20:18:24 ahh Sep 29 20:18:27 rebuild gcc-cross Sep 29 20:18:33 that was the thing that I didn't do Sep 29 20:20:17 re Sep 29 20:24:56 GNUtoo|laptop: leave me a message here how it goes Sep 29 20:25:07 if I am not here I will read it later Sep 29 20:26:57 ok Sep 29 20:27:10 for now it's cleaning gcc-cross Sep 29 20:35:11 hm, http://lists.linuxtogo.org/pipermail/openembedded-devel/ seems non responding Sep 29 20:35:59 hi I have a strange thing in glibc-initial for a arm926 target using angstrom : do install fails with : linux_embarque/cpuimx25/tmp_angstrom/sysroots/x86_64-linux/usr/bin/install: cannot create regular file `/usr/local/include/limits.h': Permission denied Sep 29 20:36:04 make[1]: *** [/usr/local/include/limits.h] Error 1 Sep 29 20:36:23 the PC seems to be a 64bits slackware Sep 29 20:36:39 same build works fine on x86 & amd64 ubuntu Sep 29 20:37:11 what could be the reason of glibc-initial trying to install in /usr/local/include ? Sep 29 20:37:39 recipe version is : glibc-initial - 2.9 - r36.3 Sep 29 20:47:11 sounds like it is trying to create a file on the build machine ... Sep 29 20:49:45 yes Sep 29 20:50:01 I don't understand what make it trying to do this Sep 29 20:50:32 the configuration is script generated so it's the same on several build hosts. What differs is that the host PC is a slack Sep 29 20:50:49 03Frans Meulenbroeks  07org.openembedded.dev * r9a5382aad4 10openembedded.git/ (2 files in 2 dirs): Sep 29 20:50:50 wpa-supplicant : moved unused files to obsolete dir Sep 29 20:50:50 Signed-off-by: Frans Meulenbroeks Sep 29 20:50:53 03Frans Meulenbroeks  07org.openembedded.dev * rb723fda80f 10openembedded.git/recipes/wireless-tools/wireless-tools/fixheaders.patch: Sep 29 20:50:53 wireless-tools : moved unused files to obsolete dir Sep 29 20:50:53 Signed-off-by: Frans Meulenbroeks Sep 29 20:50:54 03Frans Meulenbroeks  07org.openembedded.dev * r1b6ffe25f9 10openembedded.git/recipes/wt/wt3/cheader.patch: Sep 29 20:50:54 wt : moved unused files to obsolete dir Sep 29 20:50:54 Signed-off-by: Frans Meulenbroeks Sep 29 20:50:58 03Frans Meulenbroeks  07org.openembedded.dev * r30098f1915 10openembedded.git/recipes/wifi-radar/files/ (wifi-radar.patch wifi-radar_destktop.patch): Sep 29 20:50:58 wifi-radar: moved files dir to more specific wifi-radar Sep 29 20:50:58 Signed-off-by: Frans Meulenbroeks Sep 29 20:51:23 yes Frans, you are at w, cleanup is nearly finished good job :-) Sep 29 21:04:17 ericben: it is not, I didn't really stick with alphabetical order Sep 29 21:05:35 still 98 packages to go, but some of these do not fetch so they need some additional verification Sep 29 21:05:35 (i normally do a bitbake -cpatch of all recipes in the dir after removing the file and inspect the results) Sep 29 21:06:17 guess it'll take 2-3 more weeks, trying to do a few every day (max 10, but preferably 5 or 6 or so) Sep 29 21:06:59 btw formal patch for kernel note is posted, if there is no objection I will push tomorrow evening; already added the acks that were given before Sep 29 21:07:06 calling it a day, cya tomorrow Sep 29 21:07:44 bye eFfeM_work Sep 29 21:14:48 hmm Sep 29 21:14:55 I hope he saw Tartarus 's remark Sep 29 21:18:34 03Martin Jansa  07master * r7a54b54c10 10openembedded.git/recipes/xorg-lib/ (13 files in 6 dirs): Sep 29 21:18:34 libx11: add 1.3.6 and 1.3.99.902, remove 1.3.5, change keysymdef_include_patch Sep 29 21:18:34 * point to dir instead of keysymdef.h directly, because newer versions Sep 29 21:18:34 also check for vendor keysyms, see Sep 29 21:18:34 http://cgit.freedesktop.org/xorg/lib/libX11/commit/?h=libX11-1.3-branch&id=a32370c0854c8ad6f526dd997f14dbe3f466d4db Sep 29 21:18:34 Signed-off-by: Martin Jansa Sep 29 21:46:33 03Martin Jansa  07master * r3870681ff6 10openembedded.git/recipes/linux/ (4 files in 3 dirs): Sep 29 21:46:34 linux(-kexecboot)_git: bump SRCREV for 2.6.36-rc6 Sep 29 21:46:34 Signed-off-by: Martin Jansa Sep 29 21:46:44 ericben: whats make version on your box Sep 29 21:47:30 moin. i always get this error while running bitbake: http://pastebin.com/8ZATyXYe Sep 29 21:56:02 playya__: what rev of OE are you at Sep 29 21:58:40 khem, ok it wokrs Sep 29 21:58:48 I'm unsure about performances but it works Sep 29 21:58:50 let me check where's the last merge with org.oe.dev Sep 29 21:58:52 I must go to sleep now Sep 29 21:59:50 good night GNUtoo|laptop Sep 29 22:00:10 i just checked the recipe and the patch, same as HEAD in org.oe.dev Sep 29 22:00:17 I'm unsure because of 500/600 Mhz scaling maximum Sep 29 22:00:29 GNUtoo|laptop: cool Sep 29 22:00:34 how is experience :) Sep 29 22:00:38 GNUtoo|laptop: is it better Sep 29 22:00:58 playya__: whats your build/os Sep 29 22:01:12 I don't know Sep 29 22:01:27 GNUtoo|laptop: ok. You did not run it yet ? Sep 29 22:01:34 I just remembered 500/600 Mhz Sep 29 22:01:41 I'm running it right now Sep 29 22:01:42 xv works Sep 29 22:01:49 GNUtoo|laptop: are you able to play HD in mplayer Sep 29 22:02:13 if you mean the h264 movie, no Sep 29 22:02:21 let me disable some stuff Sep 29 22:02:23 khem, ubuntu 10.04.1 Sep 29 22:02:55 GNUtoo|laptop: ok, I am interested to know if you see some visible performance gain Sep 29 22:03:09 playya__: ok and you are using latest oe Sep 29 22:03:16 basically I remember now that the huge performances were because I switched from 500Mhz to 600Mhz Sep 29 22:03:17 playya__: and bitbake also from git/master Sep 29 22:03:34 ok Sep 29 22:03:35 not huge Sep 29 22:03:40 I'm using 1.10 Sep 29 22:03:41 I'm unsure Sep 29 22:03:45 I'll test tomorrow Sep 29 22:03:52 GNUtoo|laptop: ok thx Sep 29 22:03:57 good night Sep 29 22:04:08 playya__: and whats your shell Sep 29 22:04:12 is is dash or bash Sep 29 22:04:14 thanks Sep 29 22:04:14 bye Sep 29 22:04:28 bash Sep 29 22:04:41 playya__: hmm interesting Sep 29 22:04:49 iirc bb complains if you use sth different Sep 29 22:04:49 playya__: what distro/machine Sep 29 22:05:03 64 bit ubuntu lucid Sep 29 22:05:09 I mean target Sep 29 22:06:26 good nite Sep 29 22:06:43 ah. palmpre Sep 29 22:07:03 just compiling sth from fso Sep 29 22:07:37 hmm. the patch is already applied Sep 29 22:07:48 let me clean my tmp/ Sep 29 22:09:17 btw. why oe still uses pkgconfig 0.23? compabilty? Sep 29 22:11:31 Hi Sep 29 22:11:46 anybody can help-me with this: http://pastebin.com/uGQse4F4 ? Sep 29 22:15:24 angelox_123: eina-0 Sep 29 22:15:28 is missing Sep 29 22:15:33 i'm using a created by me distro based on angstrom... Sep 29 22:15:37 and... Sep 29 22:15:40 no Sep 29 22:15:42 how to fix? Sep 29 22:15:43 eina-0 is now eina-1 Sep 29 22:16:05 either esmart should be fixed to use the latest eina in oe Sep 29 22:16:13 or provide compat package for older version Sep 29 22:18:06 i don't understand(newbie)... Sep 29 22:19:28 ? Sep 29 22:26:24 khem, do_patch works now. thanks Sep 29 22:26:41 maybe just ctrl. +c at the wrong time Sep 29 22:29:14 angelox_123, what are you building? Sep 29 22:29:37 esmart is obsolete and nobody seems to update it Sep 29 22:35:44 sorry by delay... task-e-x11-core Sep 29 22:42:01 khem: I'll check make version with the user of the box thanks I didn't think to verify this Sep 29 22:50:10 angelox_123, you should set a different EFLSRCREV Sep 29 22:50:35 how? Sep 29 22:53:38 are you using HEAD of org.openembedded.dev? Sep 29 22:54:43 and which distro? Sep 29 22:54:50 it first look into angstrom repo,if don't found,look at org.openembedded.dev Sep 29 22:55:07 it's a based angstrom distro(i only changed the name) Sep 29 23:00:50 angstrom-2008.1.conf ? Sep 29 23:01:13 yes,based in it Sep 29 23:03:29 so you should have the line require conf/distro/include/sane-srcrevs.inc Sep 29 23:03:52 and sane-srcrevs.inc sets EFL_SRCREV to 52678 Sep 29 23:06:16 this line where(i mean of require....)? Sep 30 00:12:59 ah,now i understand sorry Sep 30 00:13:01 i'll try Sep 30 01:14:20 Hi back Sep 30 01:14:26 now i have this error: Sep 30 01:14:30 http://pastebin.com/0JKmsb64 Sep 30 01:28:58 angelox_123, my logs from 2 days ago shows that patch applies... Sep 30 01:29:36 hum..i'm using an angstrom-based distro(only with the name changed) Sep 30 01:29:50 i had do this: Sep 30 01:29:52 so you should have the line require conf/distro/include/sane-srcrevs.inc Sep 30 01:29:52 and sane-srcrevs.inc sets EFL_SRCREV to 52678 Sep 30 01:31:29 http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?id=d33ca8e16602a72d99f47a465bb56a613273cbc6 Sep 30 01:32:16 what i do? Sep 30 01:32:22 they voluntarily went from setuptools to autotools? ... not the direction i'd have gone.. heh Sep 30 01:32:29 sorry please i'm newbye Sep 30 01:33:01 i don't know setuptools, but i'm no fan of autotools Sep 30 01:33:14 angelox_123, make sure your tree is up to date i suppose Sep 30 01:33:40 i'll update now! (i forget,but have 1 month i don't updated =O ) Sep 30 01:33:42 lol Sep 30 01:33:58 well that could explain a lot Sep 30 01:34:01 it's git pull yes? Sep 30 01:34:05 yep Sep 30 01:34:18 git pull --rebase please Sep 30 01:34:23 hi grg Sep 30 01:34:25 and then merge everything that's broken in the last month Sep 30 01:34:29 morning khem Sep 30 01:34:36 grg: australia is playing india next month Sep 30 01:34:45 * khem thinks of cricket Sep 30 01:34:48 :) Sep 30 01:34:50 thanks Sep 30 01:35:05 what time is on your country grg ? Sep 30 01:35:13 11:04 am Sep 30 01:35:19 lol Sep 30 01:35:26 here is 22:34 pm Sep 30 01:35:26 tomorrow Sep 30 01:35:33 30th sept Sep 30 01:35:45 29 here =D Sep 30 01:36:08 what is your real name? (my name is Angelo) Sep 30 01:36:33 Graham Sep 30 01:36:47 thanks Sep 30 01:37:14 you can probably find that out from your irc client, i'm not hiding Sep 30 01:37:59 ah ok sorry i'm some newbie Sep 30 01:39:31 khem, a two test series? They're asking for a drawn series, right? Sep 30 01:40:19 bye (i'll to bath! =P ) Sep 30 01:40:23 i come soon Sep 30 01:41:28 grg: yeah but I think these days they cant play full five days so I am expecting some results Sep 30 02:04:21 i back Sep 30 02:04:28 i do git pull but the same error Sep 30 02:04:35 did* Sep 30 02:06:03 i did a mistake Sep 30 02:07:41 Hi there. Just wondering if anyone has made an OE distro for a PC/104 stack before? Sep 30 02:11:51 as i had say here,i maked a distro,anybody can place it on oe tree(oficial) **** ENDING LOGGING AT Thu Sep 30 02:59:57 2010