**** BEGIN LOGGING AT Tue Feb 02 02:59:58 2016 Feb 02 05:14:06 RP: I am going to upgrade ltp to latest release which should fix it Feb 02 07:31:31 hi folks Feb 02 07:50:08 commit 8d310b24927d0f348fb4 kernel/kernel-arch: Explicitly mapping between i386/x86_64 and x86 for kernel ARCH in oe-core breaks my build because it cannot find my KBUILD_DEFCONFIG anymore Feb 02 07:51:05 I have my defconfig inside the kernel source under the arch/x86 dir and that commit maps OE to search for x86_64 instead of x86 how it was before Feb 02 07:51:38 the only way I can get it to work as before that commit with that change is to set TARGET_ARCH = athlon which will map to x86 Feb 02 08:21:35 good morning Feb 02 09:03:36 wow cool Feb 02 09:04:11 any one here,I am trying to build yocto for raspberry pi, Feb 02 09:05:39 hi Feb 02 09:14:42 Hi everyone Feb 02 09:15:24 It seems I'm stucked on a classical error: I get "glxx_client_GetRenderbufferParameteriv" while compiling meta-qt on Raspberry. Feb 02 09:17:25 I added a .bbappend rule for qtbase to select gles2. PACKAGECONFIG_GL = "gles2" Feb 02 09:18:04 And then this error comes out. Can anyone tell me where I should have a look? Feb 02 09:19:48 http://comments.gmane.org/gmane.linux.embedded.yocto.general/19025 Feb 02 09:22:05 In the meta raspberry there is: PROVIDES = "virtual/libgles2 \ Feb 02 09:22:07 virtual/egl" Feb 02 09:45:52 hi Feb 02 09:46:56 I'm developing an image and I would like to include boost library in it. I know there's a recipe for it, but I would like to use a different version (1.55 instead of current 1.58). How can I do that? Feb 02 09:48:46 Spulit: you'd have to update the recipe :) Feb 02 09:51:14 CTtpollard, by looking at the recipes repository, I see that 1.55 version was once supported in a previous version of yocto/poky. Maybe I can use that recipe, right? Feb 02 09:52:28 Spulit: yes, quite of the only thing that has changed is the src uri and src rev Feb 02 09:52:35 *quite often Feb 02 10:18:32 Anyone playing with meta-qt arround? Feb 02 14:00:47 hi. how can I add a new variable to BB_ENV_WHITELIST/BB_ENV_EXTRAWHITE? Feb 02 14:18:21 Hi. I want to add systemd service files for recipes present in other layers from my layes. How do I do it? Feb 02 14:18:59 rtr_: create a bbappend to the recipe for your layer, this will allow to to add extra files, such a .service Feb 02 14:19:46 @CTtpollard Can I keep the service files in my layes Feb 02 14:19:49 layer*? Feb 02 14:20:15 rtr_: yes, you can have them where you please Feb 02 14:20:56 Ok. Can you please point me to an example (if you know of any)? Feb 02 14:21:12 meta-systemd probably still has a few in Feb 02 14:52:55 since I rebased to recent CVE fixes in poky/jethro I get: Feb 02 14:52:57 sh -c "export LANG=de_DE; ls -l" Feb 02 14:52:57 sh: loadlocale.c:130: _nl_intern_locale_data: Assertion `cnt < (sizeof (_nl_value_type_LC_COLLATE) / sizeof (_nl_value_type_LC_COLLATE[0]))' failed. Feb 02 14:52:57 Aborted Feb 02 14:53:12 any idea? Feb 02 14:54:34 did that include " glibc: Backported a patch to fix glibc's bug(18589)" Feb 02 14:54:40 i *think* that was implicated in that Feb 02 14:54:50 let me check ... Feb 02 14:55:40 commit 96b1b5c1 glibc: Backported a patch to fix glibc's bug(18589) from 25th Nov 2015 Feb 02 14:56:08 * rburton checks with colleague Feb 02 15:01:06 Hi! I've done https://wiki.yoctoproject.org/wiki/How_do_I#Q:_How_do_I_create_my_own_source_download_mirror_.3F, and I can see the git repo in my folder. However when I enable BB_NO_NETWORK, I still get "ERROR: Function failed: Network access disabled through BB_NO_NETWORK ... but access requested with command git -c core.fsyncobjectfiles=0 ls-remote ssh://example.server.com:29418/linux-2.6-imx.git (for url None)". In the receipe it is like SRC_URI = ... Feb 02 15:01:12 ... "git://example.server.com:29418/linux-2.6-imx.git;protocol=ssh};branch=${SRCBRANCH}" SRCREV = "${AUTOREV}". Any ideas? Feb 02 15:01:31 AUTOREV Feb 02 15:01:40 that means "go and see what the latest rev is Feb 02 15:01:51 so fundamentally incompatible with BB_NO_NETWORK Feb 02 15:02:00 I see. thanks. Feb 02 15:02:26 recommendation is not to put autorev in recipes for this and other reasons Feb 02 15:02:46 there is an alternative to BB_NO_NETWORK (on newer Yocto Project versions) Feb 02 15:02:52 it's BB_ALLOWED_NETWORK (I think).. Feb 02 15:03:07 that permits you to put a simple globbed lists of networks you ARE allowed to access.. everything else is forbidden Feb 02 15:03:20 that will allow your stuff to access a local mirror, but prevent external access Feb 02 15:04:24 rburton: any check result from colleague? Feb 02 15:05:32 any reasonable change to image language settings I can do to avoid that (IMAGE_LINGUAS="de-de en-us " is the current setting) Feb 02 15:08:35 rdmsno: not yet, but if you have the time i recommend reverting and seeing Feb 02 15:10:14 that would be a "quick shot" - sooner or later the problem must be solved ^^ Feb 02 15:10:19 but will do Feb 02 15:11:13 Before I start diving into the code and write it myself: Is there a way to measure download cache hit rate? Feb 02 15:22:19 hi folks Feb 02 15:22:41 How do edit + rebuilt source, without going through all the hassle of creating a patch. Feb 02 15:23:05 I just want to play around with the code a little, I don't want to make a patch everytime I make a change. Feb 02 15:23:59 ignore me - I just figured it out. Feb 02 15:24:01 raykinsella78: you can try using devtool Feb 02 15:24:17 bboozzoo: yeah, just rememberd Feb 02 15:24:21 bboozzoo: thanks Feb 02 15:35:25 does anyone have any idea how http://pastie.org/10705670 might affect build output even if the ad-hoc QA check never triggers? somehow it's causing a system to not boot properly. Feb 02 15:38:14 only thing i can think of is that bitbake might be automatically inferring that the task (package_do_shlibs) that's patched depends on GITPKGV, GITPKGVTAG, and/or PKGV, and that somehow breaking things Feb 02 15:52:34 YPTM: armin is on Feb 02 15:56:31 YPTM: Ready-Access Number: 8007302996 Access Code: 2705751 Feb 02 15:56:43 YPTM: Stephen Joined Feb 02 15:59:56 YPTM: Joshua joined Feb 02 16:00:37 YPTM: Belen joined Feb 02 16:01:26 YPTM: sona joined Feb 02 16:01:59 YPTM ross joined Feb 02 16:05:42 gah, i hate it when a 'no buildable providers' situation is preventing me from using bitbake -e to see why there's a no buildable providers situation. i thought i prevented those things from coming in via dependencies, but apparently not.. Feb 02 16:16:04 rburton: gnerally https://sourceware.org/git/gitweb.cgi?p=glibc.git;a=blobdiff;f=locale/loadlocale.c;h=dcbb833a16d7845059816cd17ff132d928feca04;hp=fdba6e93c4cd3e400062f57f31b1633df78902c9;hb=a73a62562a198072ab417e28d4c0315a78c49cfe;hpb=7ad78f869815a73dc26c93b72d156812d93eb026 looks as a candidate to "solve" the issue, too Feb 02 16:19:53 kergoth: I thought we stopped bitbake -e being blocked by that? :/ Feb 02 17:02:10 I got it to W/ork Feb 02 17:02:14 HELL YEA Feb 02 17:22:22 fray: thoughts on a sdk postprocess command for sdkmingw32 which removes case duplicates and warns the user it's doing it? on case insensitive filesystems, there are conflicts for soe files in the kernel source tree for example.. rather than letting tar just overwrite one with the other, better to explicitly remove all but one dupe and warn at do_populate_sdk-time Feb 02 17:22:31 we were doing that for one of our sdk-like things Feb 02 17:22:48 i have a function to do this now, just wondering if it'd be something worth submitting to meta-mingws Feb 02 17:31:18 kergoth, we've got something similar -- but it's never been in a form I could send it back.. Feb 02 17:31:43 the way we implemented was adding another "format" thing (like .zip or whatever) so we could just put the case folding in-line with the SDK generation Feb 02 17:32:05 we also output the SDK in a .zip (but it needs 7zip to extract because the length of the paths are larger then winzip can handle) Feb 02 17:32:11 (I've determined winzip is crap BTW) Feb 02 17:36:23 yeah, we've played a bit with zip construction and the like too, but in our case it's specifically for use by eclipse equinox p2 for distribution into codebench Feb 02 17:39:53 ya.. our usecase is developer builds an SDK, and distributes it to their app developers.. SDK in either win (.zip) or Linux (.sh) format... Feb 02 17:40:02 no connection to eclipse or anything else is implied.. Feb 02 17:40:23 (our internal eclipse stuff knows how to import and use an SDK on both Linux and Windows.. but they are seperate entities) Feb 02 17:40:42 * kergoth nods Feb 02 17:41:15 that reminds me, the fact that sdk tar'ing is implicit rather than eitehr part of SDK_PACKAGING_FUNC or another hook is slightly irritating for this use case Feb 02 17:42:07 ya.. we "fixed that" and sent up a patch to the oe-core list a year or more ago.. but it was ignored.. Feb 02 17:42:10 we've just been carrying it forward.. Feb 02 17:42:17 feel free to take up the cause on that one and get it fixed Feb 02 19:01:39 fray: thanks again, mgh/meta-mingw is quite helpful Feb 02 22:24:35 rburton: see https://github.com/rehsack/poky/commit/46ac36953e2e76ba716c7ba42ddfe41c9a19a11d wrt. "glibc: Backported a patch to fix glibc's bug(18589)" Feb 02 22:25:27 sno: please post that to oe-core \o/ Feb 02 22:25:48 rburton: sure, but it's close to midnight Feb 02 22:25:56 I sleep over it and post tomorrow Feb 02 22:26:00 heh, okay :) Feb 02 22:26:11 its half ten here so i'm going shortly too Feb 02 22:27:36 I think there should be a common inc for glibc and cross-localedef-native for the used glibc-base Feb 02 22:27:52 but that's a patch for master ... Feb 02 22:28:57 yeah, agreed Feb 02 22:34:57 kergoth -- a day late, but the meta-mingw I pushed to the poky-contrib.. it works.. I just got my Win 7 VM re-installed. Feb 02 22:36:48 nice Feb 02 22:40:28 waiting on our IT guys to get me the right Win 10 key and I'll try it there as well. (I've never run any OE stuff on Win 10 before) Feb 02 22:41:08 (I got a new laptop, so everything is starting from scatch.. VMs and such.. so it's taking longer then normal to install/test as I figure this thing out..) Feb 02 22:42:36 (and yes, it's already getting covered with YP and related stickers) Feb 02 22:44:37 you put stickers on your new laptop? Feb 02 22:44:56 good man Feb 02 22:45:09 did you peel off the windows stickers though Feb 02 22:45:14 really stick it to The Man Feb 02 22:45:44 peanut butter will take off that OEM sticker glue Feb 02 22:49:01 rburton, new MacBook Pro.. no windows stickers.. ;) Feb 02 22:49:29 which is why I had to get IT's help to get a windows VM so I could test YP/OE meta-mingw SDKs.. ;P Feb 02 22:50:45 I really need to get the cash together to upgrade my machine, I'm using a Mid 2011 Macbook Air that wasn't even max spec'd when I bought it in 2011 as my daily driver Feb 02 22:50:53 thankfully i do most of my stuff on remote linux VMs, but still Feb 02 22:51:05 120gb ssd, 4gb of ram = ow Feb 02 22:53:14 * kergoth bangs head against wall Feb 02 23:42:29 Hi, I'm having some issues trying to set SRC_URI using a python task. In my bitbake recipe I inherit some tasks which run and are supposed to modify the SRC_URI, but the changes don't seem to stick after the tasks run. Here's an example using the hello world recipe http://paste2.org/LYLDhPzF . This is on top of yocto 2.0.1 with BitBake 1.28.0. Any ideas? Feb 02 23:43:25 tasks get isolated copies of metadata Feb 02 23:43:29 otherwise builds would be non-deterministic Feb 02 23:43:45 if you need to use python to alter src_uri, you'd need to use an anonymous python funciton, which would run at the end of the parsing Feb 02 23:44:06 ok, so once recipes are parsed I can't change variables? Feb 02 23:44:12 between tasks that is Feb 02 23:46:19 correct, as i said, each task gets its own isolated copy of the metadata Feb 02 23:46:29 otherwise the results would change completely depending on the order of task execution Feb 02 23:47:55 okay, thanks! I think I can work around this then. Feb 03 01:24:23 how could i add to my recipe something like do_configure_append(){ sed -e foo -i bar} if a distro feature "systemd" is enabled? Feb 03 01:54:46 hi Feb 03 01:55:27 I have questions about build yocto with meta-raspberrypi layer Feb 03 01:56:06 I follow the README file in the meta-raspberrypi,and bitbke Feb 03 01:57:03 but it comes out with several errors Feb 03 01:58:45 must of the people that I know that answer are usuallyy more Feb 03 02:43:09 hi Feb 03 02:43:14 anyone here Feb 03 02:46:31 jackzhang1992: hi Feb 03 02:52:39 jackzhang1992: what are the errors? Feb 03 02:53:54 jackzhang1992: are you using a release? or master? Feb 03 02:59:48 ok i will print the erros here **** ENDING LOGGING AT Wed Feb 03 02:59:58 2016