**** BEGIN LOGGING AT Tue Apr 06 02:59:56 2010 Apr 06 04:21:21 03Mike Westerhof  07org.openembedded.dev * rc906fd7faa 10openembedded.git/recipes/initscripts/initscripts-slugos_1.0.bb: SlugOS: initscripts-slugos - add dependency on update-rc.d-native Apr 06 04:45:04 * mwester kicks somebody's *** for screwing with a perfectly fine way SRCREV and PREFERRED_VERSIONs *used* to work. :( Apr 06 04:46:03 $10 says that change was yet another "let's make it pretty" sort of change, having little to do with actually fixing any bugs. Bah. Apr 06 05:05:30 and all of util-linux-ng and e2fsprogs is being pulled in now -- no space left in the rootfs. Looks like someone added something to one of those two that changed that behavior recently. :( Apr 06 05:06:07 * mwester starts searching for the commit that broke it... :( Apr 06 07:08:39 good morning Apr 06 07:12:15 mornign Apr 06 07:25:19 good morning Apr 06 07:25:51 , ,,Gm Apr 06 07:56:51 morning Apr 06 07:57:24 hrw: hi Apr 06 08:00:21 morning all Apr 06 08:15:38 ~curse karmic kernel update Apr 06 08:15:39 May you be reincarnated as a Windows XP administrator, karmic kernel update ! Apr 06 08:39:39 03Marcin Juszkiewicz  07org.openembedded.dev * r0165ce1850 10openembedded.git/recipes/vim/ (gvim_7.2.bb vim-7.2/001-394.diff vim_7.2.bb): vim: updated to upstream patch 411 Apr 06 08:39:50 03Marcin Juszkiewicz  07org.openembedded.dev * rd9326807a2 10openembedded.git/recipes/giflib/giflib_4.1.6.bb: Apr 06 08:39:50 giflib: added native, moved to new staging, improved Apr 06 08:39:50 Signed-off-by: Marcin Juszkiewicz Apr 06 08:43:06 hi guys - do you know about of any written literature about OpenEmbedded? Apr 06 08:43:59 hi mckoan Apr 06 08:44:21 today the big pc , i.e. picione, arrives Apr 06 08:45:03 recalcati: good news, PC-one aka pidgeon Apr 06 08:45:26 nice joke Apr 06 08:45:37 recalcati: remember to fasten you seat belt Apr 06 08:45:42 right Apr 06 08:45:49 new build machine? Apr 06 08:45:55 and to have virtualbox running. Apr 06 08:46:01 hrw: yes. Apr 06 08:46:27 which o.s. ? debian or ubuntu? I need to have virtualbox running and compiling OE Apr 06 08:46:56 I use debian to run vbox Apr 06 08:47:12 ubuntu 9.04/9.10, xp, maemosdk in vbox Apr 06 08:48:04 so the base is debian (which version? I'm not expert) and than to do special things you use virtualbox Apr 06 08:48:54 I use Debian 'sid' (aka unstable) for my desktop + Personal edition of vbox (not opensource one) Apr 06 08:49:14 vbox runs x86 and x86-64 versions of ubuntu for my tests Apr 06 08:50:39 vbox not opensource means that you needed to buy it ? Apr 06 08:50:50 there are 3 licenses of vbox Apr 06 08:51:12 ose (opensource one), personal, 3rd one Apr 06 08:51:29 I try. thx Apr 06 08:51:34 personal adds sata emulation, usb support and few other things compared to opensource one Apr 06 08:51:38 check on website Apr 06 08:52:02 http://www.virtualbox.org/wiki/Editions Apr 06 08:54:12 "Personal use is when you install the product on one or more PCs yourself and you make use of it (or even your friend, sister and grandmother). It doesn't matter whether you just use it for fun or run your multi-million euro business with it. " Apr 06 08:59:32 thank you very much Apr 06 09:15:27 np Apr 06 12:18:41 woglinde: ping Apr 06 12:20:38 mickey|office: alive? Apr 06 12:22:10 hoi ant Apr 06 12:22:21 jo hrw Apr 06 12:22:43 hrw is qemu support for n800/810 currently borken in debian unstable? Apr 06 12:23:05 no idea Apr 06 12:23:16 last time I used it in September 2009 Apr 06 12:23:24 okay Apr 06 12:23:27 if not September 2008 even Apr 06 12:23:35 at Berlin Maemo Summit Apr 06 12:23:39 so rather 2008 Apr 06 12:23:41 jo Apr 06 12:23:47 okay Apr 06 12:23:52 woglinde: was about http://cgit.openembedded.org/cgit.cgi/openembedded/tree/recipes/kexecboot/linux-kexecboot_2.6.33+2.6.34-rc3.bb. But something is fishy Apr 06 12:23:53 woglinde: better grab n8x0 then use qemu for it Apr 06 12:24:16 hrw I liked to tested kexecboot Apr 06 12:24:23 before flashing it Apr 06 12:24:52 woglinde: flasher has option to boot kernel without flashing Apr 06 12:24:56 woglinde: I guess koen is fixing SRC_URI right now Apr 06 12:25:15 ant sure will test it next week Apr 06 12:25:21 woglinde: and testing kexecboot is safe on n8x0 - you have separate kernel partition in flash Apr 06 12:25:46 hrw I like to put my kernels on the mmc Apr 06 12:26:02 woglinde: sure Apr 06 12:26:19 woglinde: you want to have kexecboot kernel in flash and use it with kernels on mmc Apr 06 12:26:33 hrw yes thats the plan Apr 06 12:26:54 woglinde: and as such testing kexecboot is safe on n8x0 - you have separate kernel partition in flash which you use for kexecboot and if something is wrong then you just flash new kernel with flasher without touching maemo rootfs Apr 06 12:27:26 right Apr 06 12:27:41 but thanks for the info I can use flasher for load kernels Apr 06 12:27:47 wasnt aware of it Apr 06 12:28:15 so I can test it without hassle Apr 06 12:28:22 -l, --load Only load all supplied images Apr 06 12:32:30 woglinde, btw I see you're mtouching icons. We aimed at http://gnome-look.org/content/show.php/Black-Gnome?content=110418 Apr 06 12:32:45 but these are 48x native Apr 06 12:32:56 ant???? Apr 06 12:33:06 in #navit? Apr 06 12:33:45 the question was: do you know of any tool doing some fine risizing 48->40 or 32x ? Apr 06 12:33:57 imagemagick and gimp ...no Apr 06 12:34:00 for which images? Apr 06 12:34:05 icons 48x Apr 06 12:34:22 pdf bmp png gif tiff? Apr 06 12:34:45 xpm Apr 06 12:34:56 for kexecboot Apr 06 12:35:20 we have 'slots' of 40 pixel, divisor of 320, 480, 640,.. Apr 06 12:35:33 hm why is xpm used? Apr 06 12:35:39 is all in c code Apr 06 12:35:46 ??? Apr 06 12:35:57 xpm support is smaller as png? Apr 06 12:36:15 iirc yes Apr 06 12:36:34 sorry for xpm I dont know Apr 06 12:36:44 maybee redesign it with gimp Apr 06 12:37:46 I fear I'll have to do by hand... Apr 06 12:39:06 svg is the way to go and than convert it to the format you want Apr 06 12:40:27 woglinde: so excuse us for the ugly icons at present ;) Apr 06 12:49:38 Having reconfigured the kernel to compile a driver as a module and rebuilding the kernel, I did 'bitbake package-index' and 'bitbake console-image' and expected my .ko file to be in the rootfs .tar, but it's not. What am I missing? Apr 06 13:29:08 hrw: ping Apr 06 13:30:08 pog Apr 06 13:30:41 have you got any problem with nfs server with debian? Apr 06 13:30:53 with kernel 2.6.32 should have this pb Apr 06 13:31:02 somebody told me Apr 06 13:31:47 I use 2.6.34-rc2 now and exported nfs to laptop running 2.6.33/debian, ubuntu 9.10 running in vm Apr 06 13:32:03 my desktop does not use debian kernels Apr 06 13:32:32 ok, thx Apr 06 13:35:58 recalcati: what is your actual problem? Apr 06 13:36:25 check the best distro for my new compilation box Apr 06 13:37:28 is better no raid or raid 1 ? Apr 06 13:38:46 raid 1 Apr 06 13:39:06 ok, my technician is telling me this answer! Apr 06 15:12:01 libtiff fails? Apr 06 15:13:35 All sorts of packages fail lately :( But libtiff isn't listed in my set of failed packages... Apr 06 15:14:10 (which might just mean that a dependency of libtiff failed, so OE didn't even try to build it.) Apr 06 15:23:15 morning Apr 06 15:24:20 oe reminds me a lot of gentoo, lots of time and headaches Apr 06 15:27:05 patches welcome. Apr 06 15:27:36 it's like any other distro, only in some ways more complex, since multiple distros share metadata Apr 06 15:29:17 though of course, our reliance on upstream locations for sources does add some headaches to the mix, but mirrors usually take care of that Apr 06 15:30:36 hmm Apr 06 15:37:28 Actually, OE (and bitbake) are fine. It's the human element, and no amount of patching can fix instability caused by changes, especially when there are a lot of unrelated changes from many contributors. OE was very stable over the winter, but now that its spring (in this hemisphere), OE seems to be suffering. Apr 06 15:38:28 mwester: Islands of stablity - a well documented feature of software development cycles :/ Apr 06 15:38:30 I just wish people would take a little more time to test things before they commit. And I really dislike the "commit the change, and let other people fix what breaks" Apr 06 15:38:56 mwester: re: your SRCREV problem, there was an email from me predicting it and with a documented fix Apr 06 15:39:00 I generally test my changes, but that doesn't mean I test them by building every recipe in the repository, or for every distro Apr 06 15:39:13 I do build test most of time Apr 06 15:39:18 mwester: summary, use PREFERRRED_VERSION = "x.y.z+svn%", the % being the key part Apr 06 15:39:30 oh good, my memory didn't fail me this time :) Apr 06 15:39:39 mwester: I've not got the reply written yet :/ Apr 06 15:39:42 RP: Ok, I'll search for it. I can't keep up with the emails anymore since I've been imprisoned behind firewalls that prohit that sort of thing... Apr 06 15:40:22 (I've escaped for a few days this week, and am trying to get the feeds to build again) Apr 06 15:43:03 kergoth_: re testing: I think there are different types of commits with differing testing requirements. Those with limited impact (i.e. one recipe) don't need widespread testing. Those that rearrange core class files, or single commits that change dozens of different recipes probably need to have a standard set of distros that get built before they get commited. SlugOS takes 45 minutes on my quad-core to build the core image -- that's not too much to Apr 06 15:43:03 ask. And that would have uncovered at least 5 different broken commits in the past weeks (and I'm not yet done fixing) Apr 06 15:44:32 I suspect that the recent set of changes is suffering once again from "only-tested-with-angstrom-itis" :D Apr 06 15:44:35 Then I'd suggest proposing a recommended set of distros and recipes and machine combinations to be tested for every major change, get the TSC to ratify it, and contribute a script to build it Apr 06 15:44:51 * kergoth doesn't build angstrom, personally Apr 06 15:45:48 mwester: oddly when I make changes and carefully mke they Angstrom only I get shouted at by minimal/micro/kaelios people as they cant be bothered to redo testing for their own distro Apr 06 15:47:04 please help: I just made meta-toolchain and I've got a nice *.bz2 file. But it does not have all my needed development libraries in it, such as libxml2. Apr 06 15:47:04 Ok. Well I'm not shouting. I'm just grumbling. :) Apr 06 15:47:16 Well; buildbot might solve it Apr 06 15:47:23 mwester: I dont like breaking SlugOS :-D Apr 06 15:47:35 One buildbot for each of those distros does the trick Apr 06 15:47:38 * XorA has seen mad max Apr 06 15:47:53 :D Apr 06 15:48:41 otavio, SlugOS (le and be) are in tinderbox (or whatever that is called). Apr 06 15:48:54 Is there a template for how to add libraries to the sdk? And is this done post installation on the development machine or done in such a way as to have it show up in the bz2 file? Apr 06 15:51:02 03Koen Kooi  07org.openembedded.dev * r326ddfffc1 10openembedded.git/recipes/gnome/cheese_2.30.0.bb: cheese: add 2.30.0 Apr 06 15:55:07 Hmm, who admins the git server again? my memory is failing me again Apr 06 15:55:20 cbrake? Apr 06 15:55:56 hrw: hello Apr 06 15:56:02 cbrake, mickey, RP I think Apr 06 15:56:26 oh no we woke the cbrake RUUUUUN! Apr 06 15:56:52 cbrake: think we can get the CIA notification hook added to the bitbake repository? update, I assume Apr 06 15:57:27 * cbrake emerges drowsly from his cave Apr 06 15:57:34 hehe Apr 06 15:57:35 and looks for something to eat Apr 06 16:00:11 kergoth_: looking ... Apr 06 16:00:34 cbrake: We can probably copy the hook from the OE repo Apr 06 16:00:56 yeah. might need to set something in .git/config, depends on the hook you're using Apr 06 16:01:37 * XorA imagines cbrake is kinda like Smaug Apr 06 16:07:26 * mwester points cbrake in the direction of whomever commited the SRCREV reorganization changes. Apr 06 16:20:12 After the SDK is installed on the development system, what defines the set of packages that can be installed into this cross compiling environment? Apr 06 16:21:43 dickelbeck: define SDK Apr 06 16:22:06 on-host-sdk or on-target-sdk? Apr 06 16:22:28 hrw, thanks it was getting lonely in a room of 153 people. This is the host sdk, "meta-toolchain"s bz2 file. Apr 06 16:24:00 There is a opkg configuration file for this host. I'm wondering how I prepare and add packages to the set. Apr 06 16:27:43 dickelbeck: http://marcin.juszkiewicz.com.pl/2008/07/17/using-poky-sdk-to-build-software/ Apr 06 16:27:55 dickelbeck: just do mental s/poky/openembedded/ Apr 06 16:28:48 byebye Apr 06 16:30:42 hrw, thanks! I will read through it. I am building a new SDK for a custom target, but you probably assumed that. Apr 06 16:30:55 code is same Apr 06 16:30:59 * ka6sox waits till cbrake gets done with what he is doing to ask him to check something else. Apr 06 16:32:09 03Steve Sakoman  07org.openembedded.dev * r1748d1ea1e 10openembedded.git/recipes/gnome/gnome-games-extra-data_2.30.0.bb: gnome-games-extra-data: add 2.30.0 Apr 06 16:32:09 03Steve Sakoman  07org.openembedded.dev * ra751044225 10openembedded.git/recipes/gnome/gnome-mount_0.8.bb: gnome-mount: add 0.8 Apr 06 16:32:10 03Steve Sakoman  07org.openembedded.dev * rd819930c04 10openembedded.git/recipes/gnome/gnome-backgrounds_2.30.0.bb: gnome-backgrounds: add 2.30.0 Apr 06 16:32:11 03Steve Sakoman  07org.openembedded.dev * radd4efda83 10openembedded.git/recipes/gnome/epiphany-extensions_2.30.0.bb: epiphany-extensions: add 2.30.0 Apr 06 16:32:12 03Steve Sakoman  07org.openembedded.dev * re0c3741345 10openembedded.git/recipes/gcalctool/gcalctool_5.30.0.bb: gcalctool: add 5.30.0 Apr 06 16:32:13 03Steve Sakoman  07org.openembedded.dev * r9997569390 10openembedded.git/recipes/gnome/gvfs_1.6.0.bb: gvfs: add 1.6.0 Apr 06 16:32:16 03Steve Sakoman  07org.openembedded.dev * rd8313a7cf1 10openembedded.git/recipes/gftp/gftp_2.0.19.bb: gftp: add 2.0.19 Apr 06 16:32:17 03Steve Sakoman  07org.openembedded.dev * r82edc61018 10openembedded.git/recipes/evince/evince_2.30.0.bb: evince: add 2.30.0 Apr 06 16:32:18 03Steve Sakoman  07org.openembedded.dev * r33e0b8e4ae 10openembedded.git/recipes/gnome/gnome-nettool_2.30.0.bb: gnome-nettool: add 2.30.0 Apr 06 16:32:22 03Steve Sakoman  07org.openembedded.dev * r0fef31550a 10openembedded.git/recipes/gnome/gnome-games_2.30.0.bb: gnome-games: add 2.30.0 Apr 06 16:32:45 * mwester says "hello" to CIA-2 and wonders where CIA has been for so long :-D Apr 06 16:33:11 hrw, what I will look forward to learning is are the tricks in building a package (*.ipk) for installation on the host vs. target. Apr 06 16:34:14 dickelbeck, IMO the biggest "trick" is to ensure that the package installs into the correct location (which is usually /usr/local/... rather than /usr). Apr 06 16:34:53 hrw, awesomely helpful. Got to run now, but many many thanks! Apr 06 16:35:50 mwester, thank you too. Apr 06 16:36:25 mwester: SDK contains aliases which takes care Apr 06 16:38:00 hi,I've a strange issue with STAGING_KERNEL_DIR: http://pastebin.com/q9y5E2f4 Apr 06 16:38:06 it varies from recipes to recipes Apr 06 16:40:14 mickey|zzZZzz, do you have an idea? Apr 06 16:42:38 i guess it's only correct on kernel recipes Apr 06 16:43:04 probably due to the way it is defined, it might defend on the package arch of a recipe Apr 06 16:43:07 mickey|zzZZzz: /query? Apr 06 16:45:39 mickey|zzZZzz, so what should I do? Apr 06 16:46:09 mickey|zzZZzz, we already talked about and if I remember well you told me to include theses kernel header Apr 06 16:46:17 I can't find the logs altough Apr 06 16:46:22 maybe they are in another computer Apr 06 16:48:12 if the msm package needs these headers, bring them in with the package Apr 06 16:48:35 mickey|zzZZzz, how? I can't access kenrel staging dir Apr 06 16:48:36 and link to workdir Apr 06 16:48:45 doesn't need to be in kernel staging dir does it? Apr 06 16:49:28 no so I copy manually theses headers into the xf86-video-msm dir where the patches and other files lies and shipt it from there with install? Apr 06 16:50:18 that would be a start, yeah Apr 06 16:50:20 my xf86-video-msm I mean recipes/xorg-driver/xf86-video-msm Apr 06 16:50:21 ok Apr 06 16:50:29 I think any recipe using KERNEL_STAGING_DIR must be PACKAGE_ARCH="${MACHINE_ARCH}" Apr 06 16:50:30 that will be a very ugly recipe Apr 06 16:50:36 neon disabled for everybody Apr 06 16:50:38 from memory Apr 06 16:50:47 ah ok nice Apr 06 16:50:55 that would solve a lot of issues Apr 06 16:50:57 that should hopefully fix your issue Apr 06 16:51:07 also I could patch neon only for htcdream Apr 06 16:51:14 thanks a lot Apr 06 16:51:26 yes, obviously the palmpre wants to use the neon instructions Apr 06 16:51:40 pre has neon doesnt it? Apr 06 16:51:43 yep Apr 06 16:51:57 ah read those statements with inverse logic :-D Apr 06 16:52:09 qualcom too cheap to pay for neon? Apr 06 16:52:19 mickey|zzZZzz, you mean palm pixie Apr 06 16:52:35 no, the pre uses neon instructions Apr 06 16:52:40 XorA, armv6-novfp Apr 06 16:52:59 * XorA is so looking forward to omap4 phones Apr 06 16:53:00 mickey|zzZZzz, the pre user use omapfb not xf86-video-msm no? Apr 06 16:53:05 no upgrade for me until then Apr 06 16:53:33 not all qualcomm processors have neon Apr 06 16:53:46 GNUtoo: ah right, i was thinking pulseaudio for some reason Apr 06 16:53:57 ok Apr 06 16:54:03 but xf86-video-msm was only ever intended for processors that *do* have neon - patches welcome Apr 06 16:54:12 hi CosmicPenguin Apr 06 16:54:57 CosmicPenguin, what are the advantages/disavantaegs of using xf86-video-msm on htcdream(armv6-novfp) Apr 06 16:55:15 also: what about rotation,would it be easier to implement with xf86-video-msm? Apr 06 16:55:17 no real advantages Apr 06 16:55:43 because some people insist on the fact that the output looks beter on xf86-video-msm Apr 06 16:55:53 Than with surfaceflinger? Apr 06 16:56:04 than with xf86-video-msm Apr 06 16:56:10 oops Apr 06 16:56:14 than with xf86-video-fbdev Apr 06 16:56:22 thats odd Apr 06 16:56:26 that is odd Apr 06 16:56:27 one of them is busted then Apr 06 16:56:30 i would except video-msm to use all the shiny acceleration ioctls Apr 06 16:56:46 s/except/expect/ Apr 06 16:56:57 any two X drivers should produce the same pixel results on framebuffer4 Apr 06 16:57:02 mickey|zzZZzz, it's when you move you see a wave or something like that,I didn't dig into it but leviathan reported that Apr 06 16:57:15 ah, the "tearing" i told you Apr 06 16:57:18 GNUtoo: ah fbdev is tearing Apr 06 16:57:19 that sounds like vsync Apr 06 16:57:27 that is to be expected Apr 06 16:57:38 indeed, due to that strange msm fbrefresh thread Apr 06 16:57:57 we do not use doubble buffering on xf86-video-fbdev Apr 06 16:58:37 that reminds me... Apr 06 16:58:49 i should patch evas to use the msm ioctls Apr 06 16:58:58 so we can stop using the refresh thread Apr 06 16:59:25 mickey|zzZZzz, ah the non-implemented ioctls? Apr 06 16:59:37 no, those which are implemented :) Apr 06 16:59:43 mickey|zzZZzz, they are not Apr 06 16:59:44 i had a quick look at msm_fb Apr 06 17:00:01 I'll look too Apr 06 17:00:10 basically what the refresh thread does is unconditionally updating the whole display at a fixed frequency Apr 06 17:00:12 ahh theses ones Apr 06 17:00:19 the 2 unused ones Apr 06 17:00:24 GNUtoo: PACKAGE_ARCH = "${MACHINE_ARCH}" will help you a lot Apr 06 17:00:26 that kill and restore the refresh thread Apr 06 17:00:30 hrw, thanks a lot Apr 06 17:00:31 we might as well update only the affected areas and only at the time we actually draw something new Apr 06 17:00:53 mickey|zzZZzz, what you are describing exist but is not in our kenrel Apr 06 17:02:12 mickey|zzZZzz, that's all we have: http://pastebin.com/dwQzp0aJ Apr 06 17:02:29 hmm Apr 06 17:02:31 that's lacking Apr 06 17:02:43 we need the same level of control as *droid Apr 06 17:03:04 mickey|zzZZzz, also some commit logs are wrong,sometime when leviathan says implement he means print something when the ioctl is called Apr 06 17:03:09 ok Apr 06 17:03:13 LATER Apr 06 17:03:14 then Apr 06 17:03:24 (caps for beeing like bug status) Apr 06 17:03:34 :D Apr 06 17:03:35 righto Apr 06 17:03:54 i'll have a go at the firmware uploader later tonite Apr 06 17:04:02 so I fix xf86-video-msm and keep it or revert to xf86-video-fbdev Apr 06 17:04:12 can you run expedite? Apr 06 17:04:17 lets see some benchmarks Apr 06 17:04:24 mickey|zzZZzz, maybe it's because udev 151 is broken....we must try udev 141 Apr 06 17:04:25 before deciding Apr 06 17:04:28 ok Apr 06 17:05:00 mickey|zzZZzz, btw xf86-video-msm has no more issues with gtk Apr 06 17:05:27 pretty good Apr 06 17:05:58 thanks dcordes for pointing me that they did some update in another branch Apr 06 17:05:59 The fbrefresh stuff is strange Apr 06 17:06:13 it probably doesn't have much value unless you use a specific type of panel Apr 06 17:11:46 i think it was just for testing Apr 06 17:11:56 while they didn't have the "real" gfx driver done Apr 06 17:12:09 it's disabled on a stock kernel Apr 06 17:12:17 http://marcin.juszkiewicz.com.pl/2010/04/06/another-job-change/ Apr 06 17:13:13 hmm Apr 06 17:13:24 do they need some more of those foundation engineers? :D Apr 06 17:14:27 I know something of that build system... you will miss OE Apr 06 17:15:09 CosmicPenguin: I know Apr 06 17:15:56 Now, if you can teach OE to use stock .debs for dependencies and use it as the infrastructure for building rootfsen Apr 06 17:16:17 then you will be a hero to many downtrodden developers Apr 06 17:17:15 including me Apr 06 17:21:45 hrw: message noted Apr 06 17:28:40 thx Apr 06 17:29:03 * hrw -> out Apr 06 17:31:06 hi, is qte the qt/embedded sdk in openembedded recipes?? Apr 06 17:31:27 in other words will bitbake qte install qt/embedded sdk for me? Apr 06 18:23:20 03Roman I Khimov  07org.openembedded.dev * r51310ad412 10openembedded.git/recipes/poptop/poptop_1.3.4.bb: Apr 06 18:23:20 poptop: fix broken sed script Apr 06 18:23:20 Signed-off-by: Roman I Khimov Apr 06 18:29:51 03Roman I Khimov  07org.openembedded.dev * ra980755dd7 10openembedded.git/recipes/ragel/ragel-native_6.6.bb: (log message trimmed) Apr 06 18:29:51 ragel-native: new recipe Apr 06 18:29:51 Ragel compiles executable finite state machines from regular languages. Apr 06 18:29:51 Ragel targets C, C++, Objective-C, D, Java and Ruby. Ragel state machines Apr 06 18:29:51 can not only recognize byte sequences as regular expression machines do, Apr 06 18:29:51 but can also execute code at arbitrary points in the recognition of a Apr 06 18:29:52 regular language. Code embedding is done using inline operators that do Apr 06 18:31:07 03Roman I Khimov  07org.openembedded.dev * raa7fa11e61 10openembedded.git/recipes/util-linux-ng/util-linux-ng.inc: Apr 06 18:31:07 util-linux-ng: ensure that LDFLAGS are clean on native build Apr 06 18:31:07 uclibc-based configurations fail to build util-linux-ng-native because Apr 06 18:31:07 of "-lintl". Fix LDFLAGS for native builds. Apr 06 18:31:07 Signed-off-by: Roman I Khimov Apr 06 19:45:40 hi,is there a _neon override? Apr 06 19:46:26 I bet no so I'll override with my machine Apr 06 19:46:35 anyway xf86-video-msm is not very used Apr 06 19:46:52 if someone wants to use it he'll shurely modify the recipe and find a better way no? Apr 06 19:47:10 Thats what open source is all about Apr 06 19:47:20 or maybe I do some @python_function with TARGET_CC_ARCH Apr 06 20:02:02 re Apr 06 20:22:07 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * rf7e26d55ae 10openembedded.git/conf/machine/htcdream.conf: htcdream.conf: added mesa-dri as PREFERRED_PROVIDER for virtual/libgl under shr Apr 06 20:22:08 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * rc233ed019d 10openembedded.git/recipes/xorg-driver/ (4 files in 2 dirs): (log message trimmed) Apr 06 20:22:08 xf86-video-msm : bump git version(fixes gtk issue),rebased patch,also made it compile automatically Apr 06 20:22:08 *We had an issue with gtk+ programs resulting in screen corruption,this is now fixed Apr 06 20:22:08 with this version Apr 06 20:22:09 *rebased the patches to make it work with the new version Apr 06 20:22:09 *switched to machine arch(thanks XorA and hrw),that permit us to copy a kernel header in the sources, Apr 06 20:22:10 unfortunately it also depends on another non-staged kernel header(kgsl_drm.h) that Apr 06 20:22:12 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * rcd60ddc923 10openembedded.git/recipes/mesa/mesa-dri_7.8.bb: (log message trimmed) Apr 06 20:22:12 mesa-dri 7.8: disable egl for htcdream(this recipe is machine arch) Apr 06 20:22:12 If we don't disable egl,we have some undefined functions at linking: Apr 06 20:22:12 | xeglgears.o: In function `main': Apr 06 20:22:12 | xeglgears.c:(.text+0x1afc): undefined reference to `eglCreateImageKHR' Apr 06 20:22:13 | xeglgears.c:(.text+0x1c4c): undefined reference to `eglCreateImageKHR' Apr 06 20:22:13 | xeglgears.c:(.text+0x1cf8): undefined reference to `glEGLImageTargetTexture2DOES' Apr 06 20:31:22 03Andrea Adami  07org.openembedded.dev * r379a4b8cc4 10openembedded.git/recipes/kexecboot/linux-kexecboot_2.6.33+2.6.34-rc3.bb: linux-kexecboot: remove bogus omap3 refs in 2.6.33+2.6.34-rc3 recipe Apr 06 20:52:05 i'm building a metatoolchain... is there a way to not make the host cross-compiler dependent on the host's glibc? Apr 06 21:01:45 hi JaMa Apr 06 21:05:20 * RP can't quite seem to find his way out from underneath his todo list today :( Apr 06 21:06:45 ant_home: Hi Apr 06 21:12:08 GNUtoo: stage that kernel headers in do_stage_append then Apr 06 21:12:20 ok Apr 06 21:28:54 03Chris Larson  07org.openembedded.dev * r03983dc2b0 10openembedded.git/classes/clean.bbclass: Apr 06 21:28:54 clean.bbclass: add pstage bits, for use with srctree Apr 06 21:28:54 Signed-off-by: Chris Larson Apr 06 21:29:03 03Chris Larson  07org.openembedded.dev * r17604d29b9 10openembedded.git/site/common: Apr 06 21:29:04 Add ac_cv_sys_symbol_underscore to site/ for gnupg Apr 06 21:29:04 Signed-off-by: Chris Larson Apr 06 21:42:09 hmm Apr 06 21:42:42 * kergoth_ can't seem to focus very well today Apr 06 21:42:49 * kergoth_ hates those days Apr 06 21:51:02 RP: thoughts on searching up the tree to root for 'conf/bblayers.conf', so you can run commands in subdirs of topdir rather than just in topdir? Apr 06 22:08:45 kergoth: I've done a fresh git pull. Still having issues with base_version_less_or_equal Apr 06 22:08:47 http://www.pastebin.ca/1854693 Apr 06 22:10:07 kergoth_: bitbake_1.8.18 can live with this Apr 06 22:13:07 dunno, builds just fine here Apr 06 22:13:35 I'm looking ehre this snippet comes from... Apr 06 22:13:39 *where Apr 06 22:14:34 ah..kernel.bbclass Apr 06 22:15:32 ah, i see Apr 06 22:17:15 may I point you to kexecboot/linux-kexecboot.inc for a substitute function working with linux_2.6.33+2.6.34-rc3 alike strings? Apr 06 22:17:35 err..just happening to work in some mysterious way :D Apr 06 22:20:02 yeah, I see what happened. When I moved things from the bb package into bb.utils, I inadvertantly removed the second vercmp. We had two functions that did different things with the same name.. bad Apr 06 22:20:20 oh..great you remember the issue Apr 06 22:20:27 no Apr 06 22:20:34 the two vercmp Apr 06 22:20:49 i just saw the error, read the vercmp definition, and checked the bitbake log for refs to vercmp :P Apr 06 22:21:07 03Michael 'Mickey' Lauer  07org.openembedded.dev * r4c29054b77 10openembedded.git/MAINTAINERS: MAINTAINERS: fix my entry Apr 06 22:21:08 03Michael 'Mickey' Lauer  07org.openembedded.dev * r840cae7425 10openembedded.git/recipes/linux/ (3 files in 2 dirs): linux-leviathan: bump; remove patch that has been applied upstream Apr 06 22:21:08 03Michael 'Mickey' Lauer  07org.openembedded.dev * rddba3186ec 10openembedded.git/recipes/images/fso2-demo-image.bb: fso2-demo-image: install iptables and some alsa utils Apr 06 22:21:10 03Michael 'Mickey' Lauer  07org.openembedded.dev * rce5330e6ed 10openembedded.git/recipes/linux/ (linux-leviathan/defconfig linux-leviathan_git.bb): linux-leviathan: bump to latest version; enable mac80211 and cfg80211 Apr 06 22:21:11 03Michael 'Mickey' Lauer  07org.openembedded.dev * r420695af7a 10openembedded.git/conf/machine/htcdream.conf: conf/machine/htcdream.conf: add wifi to MACHINE_FEATURES and add required kernel modules Apr 06 22:22:52 kergoth_: Sounds like a nice improvement Apr 06 22:25:43 kergoth_: We should make those two version comparisions use the same function Apr 06 22:26:19 * RP has been trying to get us using once function (in bitbake) for a while rather than the odd assortment we were using :/ Apr 06 22:30:50 for now i resurrected the old one under another name, so we don't break the metadata. can consolidate them in the future Apr 06 22:31:03 ant_home: should be fixed now, pull Apr 06 22:31:09 ant_home: (in bitbake, master) Apr 06 22:31:12 thx Apr 06 22:31:33 RP: still planning on branching 1.10 / releasing 1.10.[01] this week? Apr 06 22:32:26 man portage had some ugly, ugly code back when we started this :) Apr 06 22:32:42 hm.. doing -c clean I see it deletes the modules one-by-one... Apr 06 22:33:14 well, one note per module Apr 06 22:33:41 this is slow Apr 06 22:34:08 it lists each header removed :/ Apr 06 22:34:47 OMG, still cleaning Apr 06 22:36:46 I mean , these were last lines http://gist.github.com/358211 Apr 06 22:39:33 kergoth_: I'd like to think so, just snowed under atm :( Apr 06 22:39:57 kergoth_: Thats what a bank holiday weekend followed by a business trip does to you Apr 06 22:40:10 kergoth_: btw, are you or Tom going to be at ELC next week? Apr 06 22:42:54 kergoth: packaged just fine after git pull Apr 06 22:43:06 RP: I'm not, not sure about tom Apr 06 22:43:08 ant_home: great Apr 06 22:56:10 anyone going to ESC San Jose Apr 06 23:00:19 'nite Apr 06 23:03:38 RP: it still happens, just again Apr 06 23:03:41 KeyError: 'virtual:native:/data/pkg/oe/openembedded/recipes/util-linux-ng/util-linux-ng_2.17.bb' Apr 06 23:03:54 very often, when one recipe fails to build Apr 06 23:04:14 then trying again leads to that **** ENDING LOGGING AT Wed Apr 07 02:59:57 2010