**** BEGIN LOGGING AT Fri Jan 11 02:59:58 2013 Jan 11 05:22:48 ericben: otavio: [meta-browser] we noticed one more problem with chromium browser that the address bar text does not render properly on imx6-sabrelite target. Disabling compiler optimizations for cairo solves this problem. should I send a patch for this as well Jan 11 06:23:01 https://gist.github.com/4508410 Jan 11 07:31:32 fusman: which option did you disable Jan 11 07:32:26 khem: EXTRA_OECONF_append_imx6qsabrelite += " CFLAGS='-O0' " Jan 11 07:32:38 hmmm thats not good Jan 11 07:32:52 yeah Jan 11 07:32:55 disabling it completely may not be best Jan 11 07:33:01 and cairo is core lib Jan 11 07:33:05 no, it was only for cairo Jan 11 07:33:49 do you have a complete compiler cmdline Jan 11 07:33:52 to show Jan 11 07:34:16 let me check Jan 11 07:37:41 khem: arm-none-linux-gnueabi-gcc -march=armv7-a -mthumb-interwork -mfloat-abi=softfp -mfpu=neon -mno-unaligned-access --sysroot=/media/sda6/home/fahad/stuff/mel6/build_sabrelite_chromium_jan10/tmp/sysroots/imx6qsabrelite -DHAVE_CONFIG_H -I. -I.. -I. -D_REENTRANT -I/media/sda6/home/fahad/stuff/mel6/build_sabrelite_chromium_jan10/tmp/sysroots/imx6qsabrelite/usr/include/pixman-1 -I/media/sda6/home/fahad/stuff/mel6/build_sabrelite_chromium_jan10/tmp/sysroots/imx6qs Jan 11 07:37:59 khem: is this what you wanted? Jan 11 07:38:33 yes Jan 11 07:40:00 fusman: leaving all -I option out Jan 11 07:40:15 is that all compiler options that are in play ? Jan 11 07:42:11 yeah, I can pastebin log.do_compile if you want Jan 11 07:45:31 khem: !!! Jan 11 07:45:35 how's it going ;) Jan 11 07:48:49 msm: hey Jan 11 07:48:53 its not bad Jan 11 07:49:24 finally have some time to do some upstream work Jan 11 07:49:43 fusman: ok pastebin it Jan 11 07:50:30 khem: i saw new activity! Jan 11 07:52:13 there is more I have in queue like python3 Jan 11 07:54:10 http://pastebin.mozilla.org/2051032 Jan 11 07:54:19 khem: ^ Jan 11 07:54:47 pastebin.com seem to be blocked in my country/organization Jan 11 07:55:22 fusman: use pastey.org Jan 11 07:55:53 http://pastie.org I mean Jan 11 07:56:12 msm: I have new eglibc 2.17 patches Jan 11 07:56:29 msm: It would be intresting if you can try those on various fsl boards Jan 11 07:56:35 khem: i informed fsl Jan 11 07:56:37 see if all boots fine Jan 11 07:56:41 k cool Jan 11 07:56:52 khem: we are hiring a fsl toolchain person for yocto btw Jan 11 07:56:54 if you know anyone Jan 11 07:57:03 msm: ah cool Jan 11 07:57:15 ive not been on master in a while Jan 11 07:57:21 im about to start testing it Jan 11 07:57:23 should include 2.17 Jan 11 07:57:29 I see ok Jan 11 07:58:15 I know few folks in russia good toolchain engineers Jan 11 07:58:27 let me see if someone is inbetween jobs Jan 11 08:02:16 what is ADT ? Jan 11 08:02:28 Application Dev Kit Jan 11 08:03:52 khem: won't this pastebin.mozilla work for you, pastie.org does not allow more than 64kb per post, and the log file is ~2mb Jan 11 08:04:15 yeah it works fine Jan 11 08:04:36 should I use ADT if I want to build a qt app for the target device ? Jan 11 08:04:51 fusman: which version of gcc are you using Jan 11 08:05:06 is it OE internal toolchain ? or something external Jan 11 08:05:43 qt-x: you need yocto SDK for QT Jan 11 08:07:59 khem, do you know of some documents that explain how do i setup yocto SDK for QT? Jan 11 08:08:34 or some pointers in the right direction Jan 11 08:10:11 khem: gcc version 4.7.2 Jan 11 08:10:19 Sourcery CodeBench 2012.09-93 Jan 11 08:10:32 fusman: ok so thats using internal toolchain Jan 11 08:10:49 no, external Jan 11 08:11:51 fusman: hmmm external ? who is the vendor ? Jan 11 08:11:54 codesourcery Jan 11 08:12:17 yeah, codesourcery Jan 11 08:12:45 we noticed the same problem with gcc version 4.6.3 (Sourcery CodeBench 2012.03-47) as well Jan 11 08:13:01 fusman: hmmm can you reproduce this problem iwth internal yocto toolchain Jan 11 08:13:27 I haven't tried, but I can check Jan 11 08:13:34 fusman: if you could it would be interesting if not then you should let CSL know of the bug in their gcc Jan 11 08:14:13 and meta-browser should not work around for one compiler IMO Jan 11 08:15:16 khem, I believe this problem has already been communicated to CSL. I will check again though. thanks Jan 11 08:17:03 k Jan 11 08:17:15 qt-x: bitbake mno-unaligned-access Jan 11 08:17:18 errrrr Jan 11 08:17:32 qt-x: bitbake meta-toolchain-qte Jan 11 08:18:01 qt-x: this will build the whole SDK for you then follow https://wiki.yoctoproject.org/wiki/SDK_Generator Jan 11 08:18:05 to install and use it Jan 11 08:22:54 khem: should I first issue bitbake mno-unaligned-access or the order it is not important (I issued yesterday bitbake meta-toolchain-qte ) ? Jan 11 08:23:08 fusman: in case you try internal toolchain please report back on your findings I will be interested to hear back Jan 11 08:23:21 qt-x: that was typo Jan 11 08:23:40 you only need bitbake meta-toolchain-qte Jan 11 08:43:23 hi khem I see you talk about meta-toolchain-qte, I'm actually testing a meta-toolchain-qt to provide the toolchain for Qt/X11 patches should reach ML today Jan 11 10:10:00 hi Jan 11 10:11:06 does IMAGE_FEATURES +="dev-pkgs" means that all dev-pks (if exists) of all package of the image will be installed also ? Jan 11 10:11:30 b737800: yes Jan 11 10:13:36 also packages installed via IMAGE_INSTALL+="..." Jan 11 10:13:38 ? Jan 11 10:13:56 yes Jan 11 10:14:01 afaik Jan 11 10:14:19 ok thanks that's great Jan 11 10:24:49 i want to check the installed packages on a dev-image is there a way to get a list of packages effectively installed on an image ? Jan 11 10:26:49 b737800: on the machine, use the package manager. or in the workdir, machine/image/version/installed_pkgs.txt Jan 11 10:30:08 thanks installed_pkgs.txt worked Jan 11 10:46:39 hello again, is there a variable (accessible via bitbake -e) which holds a list of packages that will be installed on an image (before build) ? Jan 11 10:55:13 good morning Jan 11 11:35:27 hi everybody! I've got a strange issue on booting and i.MX6 board. I have bootloader, kernel and filesystem, but i'm not able to get a serial console. this is boot log: http://pastebin.com/CcsdCHEf Jan 11 11:35:29 (the hwclock error is not fatal) Jan 11 11:35:30 [12:33] I've checked my /etc/inittab and set checked the presence of this line: "S:2345:respawn:/sbin/getty 115200 ttymxc1" Jan 11 11:35:32 [12:33] it is actually present but I still fail to get a serial console login. Anything that anyone can suggest? Jan 11 11:36:30 damnit musicbrainz failing again Jan 11 11:36:58 rburton: musicbrainz, the deadly beast :) Jan 11 11:44:26 rburton: why is musicbrainz even in core...? Jan 11 11:44:43 a sata throwback? Jan 11 11:44:56 s/sata/sato Jan 11 11:45:32 jackmitchell: pretty much. when it broke i made a note to move it to meta-multimedia Jan 11 11:45:48 its for a gstreamer plugin that is rarely used Jan 11 11:46:09 the telepathy stack needs to move out too - nothing exercises it in oe-core Jan 11 11:47:11 is there a variable (accessible via bitbake -e) which holds a list of packages that will be installed on an image (before build) ? Jan 11 15:23:26 i don't suppose anyone cares about gtk+ on directfb? Jan 11 15:58:26 RP: https://github.com/kergoth/oe-test-scripts#readme - rewrote these in python, verified both function properly. not ideal, but better than nothing, I think Jan 11 15:59:39 kergoth: yes, I've wanted to test things like this for a while. Sounds good Jan 11 16:00:07 kergoth: worth pulling into scripts or scripts/contrib in the main tree? Jan 11 16:00:42 RP: Hi Jan 11 16:00:59 I want to ask one thing regarding bitbake patches Jan 11 16:01:49 RP: Peter Seebach submitted some patches which handles history of variable when -e option is used Jan 11 16:02:06 any plans to merge those patches in main bitbake Jan 11 16:02:40 RP: Hmm, didn't think about that, but it's a good idea. I'll do a little clean up and submit a patch series for consideration Jan 11 16:06:06 Noor: Its planned, I just want to sit down and check a few things Jan 11 16:06:23 Noor: Sadly, I don't get many such chances :( Jan 11 16:06:25 RP: OK ... I really like to see those Jan 11 16:06:30 * RP too Jan 11 16:06:46 NP ... at least its in your plan :) Jan 11 16:06:53 that's the +ve sign Jan 11 16:07:19 Noor: yes, its just been there for a while Jan 11 16:07:45 Taking vacation at the end of last year hasn't helped :( Jan 11 16:11:12 i still don't really like how it sprinkles bits to a pile of getVar calls around the code, but I can't really come up with much better Jan 11 16:11:21 heh Jan 11 16:28:39 rburton: heh gst123 now fails with different error (not curses-32.h) | /home/jenkins/oe/shr-core-branches/shr-core/tmp-eglibc/sysroots/x86_64-linux/usr/lib/libgobject-2.0.so: file not recognized: File format not recognized Jan 11 16:29:10 rburton: this is qemuarm build (notice native sysroot) Jan 11 16:29:22 | arm-oe-linux-gnueabi-g++ -march=armv5te -marm -mthumb-interwork --sysroot=/home/jenkins/oe/shr-core-branches/shr-core/tmp-eglibc/sysroots/qemuarm -pthread -I/home/jenkins/oe/shr-core-branches/shr-core/tmp-eglibc/sysroots/qemuarm/usr/include/gstreamer-0.10 -I/home/jenkins/oe/shr-core-branches/shr-core/tmp-eglibc/sysroots/qemuarm/usr/include/glib-2.0 -I/home/jenkins/oe/shr-core-branches/shr-core/tmp-eglibc/sysroots/qemuarm/usr/lib/glib-2. Jan 11 16:29:32 sorry for long oneliner Jan 11 16:31:10 wtf Jan 11 16:37:31 Hopefully someone can help clarify something for me. I come OE-Classic using Angstrom. I'm using Angstrom now on OE-Core. Am I correct to assert that poky is the Yocto Project equivalent of OE-Core (specifically git://git.openembedded.org/openembedded-core vs git://git.yoctoproject.org/poky)? Jan 11 16:37:51 jawilson: no Jan 11 16:37:55 * jawilson sighs Jan 11 16:38:07 jawilson: oe-core comes under the yocto project Jan 11 16:38:24 poky is a reference distro using oe-core and meta-yocto (bad choice of name) Jan 11 16:38:38 so poky is bitbake+oe-core+meta-yocto Jan 11 16:38:44 in one convenient bundle Jan 11 16:38:51 ohhhhh, ok Jan 11 16:39:27 so if I indepenently check out bitbake, oe-core and meta-yocto, I'm using poky Jan 11 16:39:35 effectively. Jan 11 16:39:55 just like right now I have bitbake, oe-core (and meta-oe among others), and meta-angstrom and I have Angstrom Jan 11 16:41:08 but does that mean I'm using OE or am I using Yocto, or is the differentiation pointless? Jan 11 16:41:37 well, yocto contains oe Jan 11 16:42:18 if I'm just using oe-core, am I still "using" yocto, or is that all just semantics? Jan 11 16:42:47 jawilson: semantics :) as oe-core is part of yocto, you're using yocto. Jan 11 16:43:09 jawilson: basically, if you're using oe-core you're welcome here :) Jan 11 16:43:22 * rburton -> dinner Jan 11 16:43:47 rburton, gotchya, thanks for the clarification. I'm just trying to figure out what exactly to tell my corporate overloards Jan 11 16:44:53 there is a blog post about it somewhere... Jan 11 16:45:14 jawilson: http://blogs.mentor.com/chrishallinan/blog/2012/04/13/yocto-versus-poky-versus-angstrom-etc/ Jan 11 16:45:24 that is a brief introduction to it Jan 11 16:45:38 jackmitchell, YES! Thank you. Jan 11 16:47:07 jawilson: don't take it for gospel but use it to help garner a better understanding :) Jan 11 16:49:55 jackmitchell, will do Jan 11 17:06:01 ok, so I get how oe-core, and poky is part of the Yocto Project and Angstrom depends upon it. How does bitbake fit in? Is bitbake part of poky or a seperate Yocto Project project of its own? Jan 11 17:54:46 bitbake has always been seperate to OE, but I am not certain anyone has used it outside OE Jan 11 17:56:34 jawilson: poky is an integration of bitbake, oe-core, meta-yocto and other associated bits. Jan 11 17:56:48 jawilson: poky is also a distro (held in meta-yocto) Jan 11 17:58:50 yeah, I got it Jan 11 17:59:05 my coworker and I also just enjoyed my discovery of meta-jarvis Jan 11 17:59:42 heh, always good to find someone else has already done some of the work you need to do :) Jan 11 18:01:02 so true Jan 11 18:01:52 hmm, need to handle sigint/sigterm better in the oe-core bitbake wrapper Jan 11 18:34:49 I've got some questions about "sdcard_image.bbclass" in meta-ti, and image types in general. Proper place to ask? Jan 11 18:36:02 #oe or the lists might be best Jan 11 18:36:06 * kergoth shrugs Jan 11 18:37:17 I'll try Jan 11 18:37:32 mostly the same people anyway :) Jan 11 18:38:24 yep Jan 11 18:50:09 what's up Jan 11 18:50:16 I just rebooted, 20G of ram now Jan 11 18:50:39 nice Jan 11 18:50:44 yeah Jan 11 18:50:53 tired of driving machine way into swap Jan 11 18:51:49 jawilson, I have the gnuradio 3.6.3 and freinds updates ready Jan 11 18:51:55 well almost ready Jan 11 18:53:41 Crofton|work, well I hope that's enough Jan 11 18:53:53 Crofton|work, you going to push that into denzil for me? Jan 11 18:55:38 not into the release, only master Jan 11 18:55:50 you can ask it be put in denzil though Jan 11 18:56:02 oh... I will Jan 11 18:56:10 thanks Jan 11 18:56:23 Crofton|work, than you for doing the work! Jan 11 18:57:33 np Jan 11 18:57:38 you might alos ask about fftw Jan 11 18:57:52 bother, display on otehr computer flipped out after I put the cover on Jan 11 18:58:23 Crofton, huh, I thought fftw was already on there, haven't really been paying attention Jan 11 18:59:39 I updated to the latest Jan 11 18:59:45 gotchya Jan 11 18:59:45 might be a little better at neon Jan 11 18:59:52 that'd certainly be nice Jan 11 19:00:40 if you are going to ask gnuradio get updated, it would be good to also do fftw Jan 11 19:01:54 understood Jan 11 19:10:08 anyone see this on danny: https://gist.github.com/4513124 Jan 11 19:31:07 someone remind me what beth's irc nick is? i'm drawing a blank for some reason Jan 11 19:31:20 pidge? Jan 11 19:33:57 yep Jan 11 19:39:51 ah, right, thanks Jan 11 19:47:03 guys Jan 11 19:47:06 somethging wired Jan 11 19:47:36 im compiling xserver-xorg (specific version.. so im using the xorg-xserver.inc + my recipe) Jan 11 19:47:48 which is the same as the default one but with different version Jan 11 19:48:12 but, bitbake is creating a work directory like this: xserver-xorg-2_1.10.4-r3 Jan 11 19:48:16 xserver-xorg-2 ?! Jan 11 19:48:24 and xserver is not been installed in my image Jan 11 19:51:15 that just means PE is set to 2 Jan 11 19:51:22 the package name is still xserver-xorg Jan 11 19:53:55 kergoth: but why PE is 2? Jan 11 19:54:20 ok, i saw on the .inc Jan 11 19:54:39 ftonello: package epoch Jan 11 19:55:49 most significant part of PKGV to prevent PKGV going backwards when PV has to go backwards Jan 11 21:03:31 Anotehr use of Poky where oe-core hsould be used: https://www.yoctoproject.org/tools-resources/projects/pseudo Jan 11 23:44:57 <_Lucretia_> hi, anyone building on debian 7.0? I'm failing on autoconf, for some reason it's building it's own 2.68 version and it's failing on it also Jan 11 23:50:09 * kergoth ponders Jan 11 23:51:02 <_Lucretia_> I have autoconf2.64 installed on my debian 7 Jan 11 23:51:13 <_Lucretia_> surely, it's not picking it up from there? Jan 11 23:51:20 <_Lucretia_> even though it's building it's own Jan 11 23:51:58 <_Lucretia_> nope Jan 11 23:52:00 <_Lucretia_> cant be Jan 11 23:52:05 <_Lucretia_> tis the only version in yocto Jan 11 23:52:16 <_Lucretia_> and it was working before my upgrade to 7 Jan 11 23:52:26 <_Lucretia_> kergoth: can paste the full log if ya like? Jan 11 23:55:09 <_Lucretia_> ignore that... Jan 11 23:56:16 <_Lucretia_> might be me Jan 12 00:10:58 i dont know whats happening.. but xserver-xorg is not been installed in my image... does ant one had this before? Jan 12 00:11:38 im manually adding the package to the CORE_IMAGE_EXTRA_INSTALL and still not installing Jan 12 00:11:51 i checked all the packages were created correctly Jan 12 00:58:54 FWIW, I decided to make one more quality pass over the tracking fixes and clean up a couple of loose ends. Jan 12 01:41:30 Sent out updated version. I found an actual bug, too! There were circumstances under which the variable listing (bitbake -e) would list things as functions that weren't. **** ENDING LOGGING AT Sat Jan 12 02:59:59 2013