**** BEGIN LOGGING AT Thu Sep 18 03:00:00 2014 Sep 18 03:35:14 I build an os image, but udev daemon does not detect rules changes automatically, Please help Sep 18 03:39:09 znode: what are you doing? Sep 18 03:40:09 znode: you can review the current rules with udevadm Sep 18 03:40:30 after reload, it works. Sep 18 03:40:39 ok Sep 18 03:41:02 znode: what is the scenario? Sep 18 03:41:22 But not detect changes automatically Sep 18 03:42:35 need doing this cmmand line "udevadm control --reload-rules" Sep 18 03:44:09 znode: i don't know if udev reload changes located at conf file auto Sep 18 03:44:20 i remember that you need to sent singal to it Sep 18 03:44:29 or execute this command Sep 18 03:44:31 I write my own rules, and after I replug my modem, rules not applied, after reload manually, then replug modem, it worked Sep 18 03:44:40 yeah Sep 18 03:44:48 i think that's ok Sep 18 03:45:38 Udev automatically detects changes to rules files, so changes take effect immediately without requiring udev to be restarted. Sep 18 03:45:52 https://wiki.archlinux.org/index.php/udev#Testing_rules_before_loading Sep 18 03:46:31 I tested on ubuntu pc, rules changes detected automatically Sep 18 03:48:00 \alimon Sep 18 03:48:08 znode: what kind of modem do you have? Sep 18 03:49:07 A longcheer wcdma modem Sep 18 03:50:19 If i need do some settings to make udev daemon to detect rules changes? Sep 18 03:52:52 Have inotify supported in kernel already. Sep 18 03:53:06 ok Sep 18 03:53:58 it should be enough Sep 18 03:58:42 alimon any suggestion? Sep 18 04:05:33 znode: nope Sep 18 04:06:19 can you publish the log of udev running in debug? Sep 18 04:27:07 alimon: OK Sep 18 04:28:10 znode: i'm review the code from udev/systemd Sep 18 04:28:31 and if inotify detects the change should print Sep 18 04:29:03 how to get the output? Sep 18 04:29:14 just udevd --debug? Sep 18 04:29:43 but on ubuntu pc, everything is ok, and no print too. Sep 18 04:29:52 yes Sep 18 04:30:02 uhhmm Sep 18 04:30:24 what version of udev you use? Sep 18 04:30:52 211 Sep 18 04:31:10 systemd-udevd Sep 18 06:30:45 is there a way to exclude some repositories from BB_GENERATE_MIRROR_TARBALLS? Sep 18 06:51:37 vincent-: ping Sep 18 07:32:43 ant_work: pong. Sep 18 07:32:53 hello Sep 18 07:33:18 Hello, good morning. Sep 18 07:34:02 about the MXU patch for mips, it seems necessary for XBurst Sep 18 07:34:37 I did a bit of reading and it is a SIMD instruction set Sep 18 07:34:50 like MMX, SSE, for x86 Sep 18 07:36:23 can you please check your logs and/or verify whether the OE/Yocto crosscompiler is using MXU for XBurst targets? Sep 18 07:36:44 good morning Sep 18 07:36:52 (I don't think so, I guess we need that Buildroot patch) Sep 18 07:37:40 ant_work: I will check it when I arrive to the office. Give me one hour or so... Sep 18 08:32:04 morning all Sep 18 08:36:21 morning Sep 18 08:58:25 (UMT) Sep 18 09:04:21 hi, is there a way to get an SDK where python is not broken in its core? Sep 18 09:04:32 currently, it does not have standard modules, like json, etc.... Sep 18 09:06:06 in fact, I do not understand why Yocto would care about python in the SDK. Sep 18 09:06:17 this has nothing to do with cross-stuff. Sep 18 09:06:51 but even if it duplicates python that is already available on the host, it should be possible to get a standard python, rather than a vastly truncated. Sep 18 11:40:46 Hello everybody! I've got a question about meta-qt; maybe someone can point me to the right direction: Sep 18 11:42:29 I cross-compiled for ARMv7, created the SDK and everything works basically. Where I'm stuck now, is how to add some additional packages to the sysroot which is installed by the script in deploy/sdk/xxx.sh? Sep 18 11:43:18 I tried to have two different sysroot (one containing Qt) and one where I have my other libs (e.g. gstreamer) which I need Sep 18 11:43:41 lfa: there is some sdk task package thing that should be mean for that Sep 18 11:43:48 but I didn't get that to work and I'm not sure if that's the right way Sep 18 11:43:59 sorry don't know more precisely, just remember the buzzword right now Sep 18 11:46:03 lfa: like, https://gitorious.org/arago-bitbake/arago-oe-dev/source/09aca7d418b893be7f668dc3e2374637f289ae6b:packages/tasks/task-native-sdk.bb Sep 18 11:46:11 (just a quick google) Sep 18 11:46:17 there is "populate_sdk_qt5" which i tried to inherit from and add my own packages but i got a lot of missing dependencies Sep 18 11:46:34 which is strange as the exact same image without the "inherit populate_sdk_qt5" works fine Sep 18 11:48:42 ok, thanks Sep 18 11:48:55 googling I found http://lists.openembedded.org/pipermail/openembedded-devel/2014-February/094113.html Sep 18 11:49:18 just wanted to know if that's the way how it's supposed to work Sep 18 11:49:27 then i can try to figure out why it's not working for me :) Sep 18 12:32:36 how to tell bitbake to ignore all download and sstate caches? I'm testing fixes to pre mirrors... Sep 18 12:33:14 hello Sep 18 12:34:07 trying to start with hob, but it says that layer yocto depends on layer core. it is openembedded-core or something else? Sep 18 12:36:11 xata: the layer "meta" which should be enabled by default should already provide that; something has gone wrong there Sep 18 12:37:17 mcfrisk: you can't really do that but you can clean those out for one recipe by doing bitbake -c cleanall recipename Sep 18 12:37:48 (well, of course, another alternative would be to change DL_DIR and SSTATE_CACHE to some other directory) Sep 18 12:37:55 er, SSTATE_DIR I meant Sep 18 13:16:44 thanks bluelightning, used different SSTATE_DIR and got the mirror stuff working Sep 18 13:16:56 great :) np Sep 18 13:18:23 next question, trying to set CCACHE_DIR to ~/.ccache but that fails with (No such file or directory), so bb doesn't do any standard shell variable expansions on variables/directory names? Sep 18 13:19:02 use ${HOME} Sep 18 13:19:11 ok, thanks rburton Sep 18 13:37:54 bluelightning: thanks. How do i fix such conflict? http://pastebin.com/y8TsFSut Sep 18 13:45:50 xata: I'm not sure... rburton ^ Sep 18 13:47:12 rburton: hello. can you help me somehow? Sep 18 13:49:08 unless this has come about due to some misconfiguration, I'm guessing that ultimately the maintainer of the BSP (otavio?) should fix it though Sep 18 13:55:29 bluelightning: i made no configuration, so this can be Sep 18 14:03:57 xata: ok, so it's something that the maintainer ought to fix Sep 18 14:03:59 otavio: ^ Sep 18 14:07:41 otavio: help? Sep 18 14:08:21 please. or maybe i can roll to daisy instead of daisy-next Sep 18 14:08:32 how to shutdown bitbake properly? It seems ctrl-c is too often leaking bitbake-worker, cc and other processes in my case. Sep 18 14:08:50 mcfrisk: remind me which branch are you using? Sep 18 14:09:21 xata: probably best to use daisy rather than daisy-next Sep 18 14:10:14 bluelightning: daisy Sep 18 14:12:32 ok, a brute workaround of 'killall python' seemed to do the trick, now lock is clear and I can start bitbake with ionice -n 3 instead (to not kill my Firefox responsiveness...) Sep 18 14:13:18 I'm just curious, but, can someone tell me what is the `d' parameter on OE python functions like in this call, `${@base_contains("MACHINE_FEATURES", "systemd", "${SYSTEMD_INSTALL}", "${SYSV_INSTALL}", d)}' Sep 18 14:15:46 mcfrisk: ok, I guess we need to backport some fixes from master for that then Sep 18 14:16:11 geckos: it's bitbake's datastore, it's the means by which you can access bitbake variable values from python basically Sep 18 14:21:54 bluelightning: same thing on daisy instead of daisy-next Sep 18 14:22:04 strange, it should be stable Sep 18 14:22:39 xata: I agree, it should be Sep 18 14:22:59 otavio: halp :C Sep 18 14:22:59 is that the branch of meta-fsl-arm you're referring to? or poky? Sep 18 14:23:08 both Sep 18 14:23:19 ok Sep 18 14:26:55 anyone working to provide an unified /bin x /usr/bin, /lib x /usr/lib, in the lines of systemd recommendation? Sep 18 14:35:24 k-s: it should be mostly doable already, simply by setting the the appropriate variables in your custom distro config Sep 18 14:49:34 made everything fresh and blank reclone, from daisy branch of both. nothing changes, still that trouble Sep 18 14:49:36 fray: ping. Sep 18 14:50:22 joeythesaint hey Sep 18 14:51:17 Sweet, you're here. :-) Sep 18 14:51:33 for a bit.. have to head into the office soon.. (but I'll be back online after that..) Sep 18 14:51:51 I wanted to check with you on master-next in meta-selinux. Are you or Pascal using it for anything a.t.m.? Sep 18 14:52:15 I was going to push some trail refpolicy updates there, but I didn't want to stomp on anything you guys were doing. Sep 18 14:52:20 AFAIK, it's NOT being used right now.. Sep 18 14:52:27 It doesn't look like it. Sep 18 14:52:38 HEAD commit is something from Wenzong from January. Sep 18 14:52:48 I had intended to start using it again after M3.. but frankly with my time being split a million differnet ways, I doubt I will.. Sep 18 14:53:18 my original plan was post M3 release, we're start staging in master-next.. and try to keep master up to dtate w/ Poky, within a few days or so.. Sep 18 14:53:22 Well, this being a policy update, it goes directly to stuff we probably want in the 1.7 meta-selinux branch anyway. So it's relevant. :-) Sep 18 14:53:27 but again, I think -I- won't be able to do that.. Sep 18 14:53:40 do you still have push access to meta-selinux? Sep 18 14:53:44 I do. Sep 18 14:53:51 hi guys Sep 18 14:54:07 I have a strange problem with a variable in a do_install_append() Sep 18 14:54:08 I'd drop a note to pascal.. but I'm fine (and think it's a good idea) if you want to reset master-next and push updates for testing there.. Sep 18 14:54:12 I define a variable in there Sep 18 14:54:21 TEST=${sysconfdir} Sep 18 14:54:24 Let's coordinate our plans for post M3 stuff later today or tomorrow if you've got time. Be good to nail that down soon, I think. Sep 18 14:54:25 we should sync with the Yocto Project list and Phillip and others who have been contributing.. but otherwise it seems like a good idea Sep 18 14:54:32 yes.. I think so Sep 18 14:54:34 and then try to do a mkdir -p ${D}${TEST} Sep 18 14:54:43 Done deal. I'll send out an email soon. Sep 18 14:54:44 my goal is near or soon after 1.7 release meta-selinux is "working" ;) Sep 18 14:54:48 thanks! Sep 18 14:54:55 np Sep 18 14:54:55 but strangely, in run.do_install I see that my $TEST is not expanded Sep 18 14:55:06 any idea why not? Sep 18 14:55:33 if there is no bitbake variable, you won't see it expand.. as to why in one area or another, I'm not sure.. Sep 18 14:55:36 typo is the usual cause.. Sep 18 14:56:44 so you are saying it will not show up expanded in the logs but the expansion will tka eplace actually? Sep 18 15:02:03 fray, could you please rephrase that? Sep 18 15:03:17 bitbake variables, when the 'run' files are expanded will be expanded -- if defined.. Sep 18 15:05:37 okay, thanks Sep 18 15:41:21 morning all. I imagine the 1.7 release is coming out soon. just wondering when that might be? Sep 18 15:41:36 I'm planning to do a refresh of my old intro screencast and I think I'm going to wait for the new release Sep 18 15:53:38 zenlinux: end of October I think Sep 18 15:53:44 (hi btw!) Sep 18 15:56:30 gotcha Sep 18 17:07:32 ANy thoughts on how to properly fix this? http://pastie.org/9572899 Sep 18 17:39:54 Hi to all. i add in local.conf FFTW (in IMAGE_INSTALL_append) but on sysroot(my image) i can't find it. Also i generate toolchain (for host machine) and i can't find libfftw. Can anybody help? Sep 18 19:24:05 I have (r)provides for kernel-modules-rtl872be to have rtlwifi .. yet the system is still trying to install kernel-modules-rtlwifi _after_ Sep 18 19:24:22 and this is only happening on one of my build trees, so it isn't making sense. Sep 18 19:53:18 Hi Sep 18 19:56:03 Hello otavio **** ENDING LOGGING AT Fri Sep 19 03:00:00 2014