**** BEGIN LOGGING AT Mon Jun 10 02:59:58 2013 Jun 10 07:35:34 Morning everybody. Suppose I have a license which is not in this list https://wiki.yoctoproject.org/wiki/License_Infrastructure_Interest_Group#Licenses , in my case this one: https://github.com/rafalmiel/glmark2-wl/blob/master/COPYING.SGI Jun 10 07:35:47 what should I put in the LICENSE field in my recipe? Jun 10 08:31:47 morning all Jun 10 08:35:45 good morning Jun 10 08:54:08 EsbenH_: Jun 10 08:54:14 EsbenH_: sorry Jun 10 08:54:20 Hello all. Jun 10 08:54:56 hi agherzan Jun 10 08:55:07 I spotted a bashism (or something that behaves wrongly in dash). Is there anybody willing to test and give me a hint fix? Jun 10 08:55:19 agherzan: if I can, sure Jun 10 08:55:25 bluelightning: great Jun 10 08:56:02 just bake an image where you include the systemd package udev-hwdb having dash as sh. Jun 10 08:56:10 Tell me if it fails. Jun 10 09:10:07 bluelightning: that was the build? #162? Jun 10 09:10:58 bluelightning: (realized the answer: no - checked timing) Jun 10 10:50:46 Hi. Anyone here working on qmx6 ? Jun 10 11:16:04 ChristianBS: imx6q? Jun 10 12:46:17 ChristianBS: Hello Jun 10 12:46:28 ChristianBS: from Congatec? Jun 10 12:53:03 Nope Jun 10 12:53:26 otavio: Jun 10 12:57:52 ChristianBS: so which som is it? Jun 10 12:57:54 hihi Jun 10 12:58:11 ChristianBS: The only qmx6 SOM I know is from Congatec IIRC Jun 10 12:58:28 ChristianBS: and it is supported by meta-fsl-arm-extra Jun 10 13:00:22 can anyone tel me basics of creating an SDK for some project in yocto? Jun 10 13:03:11 rao_: easiest method is: bitbake -c populate_sdk imagename Jun 10 13:05:16 thank you . i tried the same command and installed the sdk in some target directory. it contains one sysroot and some script files. Now i am totally confused what to do with those files and how i can basically use it? pls help me Jun 10 13:05:53 is there any possible way that i can emulate it Using QEMU? Jun 10 13:14:25 otavio: yup its indeed the conga som. Have you tried using the *-x11 libs from the vivante blob and gotten them working on x11? Jun 10 13:15:06 ChristianBS: I don't have the Congatec SOM here; they never sent it to me so I cannot help on it Jun 10 13:15:23 got another board that you used the vivante libs on? Jun 10 13:15:29 ChristianBS: it works in other i.MX6 boards (with known issues) Jun 10 13:15:56 otavia: known issues = undefined symbols in the vivante_dri causing it fallback to mesa? Jun 10 13:16:12 ChristianBS: yes; for windowed mode, yes Jun 10 13:16:22 ChristianBS: but fullscreen mode works Jun 10 13:17:08 ChristianBS: It'd be great if you could e-mail the mailing list so others can help. Jun 10 13:17:52 ChristianBS: and I also would like to backport the U-Boot support for it, so it'd be great if you could test it as well. Jun 10 13:18:32 otavio: i'll try to post it on the mailing list aswell Jun 10 13:19:32 otavio: how to ensure I run it in fullscreen? Jun 10 13:22:53 ChristianBS: I am not a GPU guy; panda84kde is a better guy to help here. Jun 10 14:06:27 ChristianBS: hi Jun 10 14:07:19 panda84kde: hi Jun 10 14:09:37 I don't have problems with software fallback; I have problems with wrong 3D rendering Jun 10 14:10:25 okay. Are you able to display e.g. the vivante demos using the -x11 libs? Jun 10 14:11:21 they start and run without errors on the command line, but they display only white windows Jun 10 14:11:45 and you absolutely certain thats it not mesa your using? Jun 10 14:14:25 when X starts, it attempts to start the vivante x11 driver, but (atleast for me) fails. If mesa is present, it fallbacks to using software rendering. Jun 10 14:16:02 es2info reports Vivante Driver is being used Jun 10 14:17:01 yes, I'm sure. Freescale's Ubuntu has the same problem as Yocto Jun 10 14:17:07 Could I ask you to post a dump of the /var/log/Xorg.log ? Jun 10 14:18:37 ChristianBS: ok, just a minute Jun 10 15:01:37 Good morning, #yocto! Jun 10 15:03:26 davest: good morning! Jun 10 15:08:21 anyone else getting mail bounces: 5.1.2 - Bad destination host "DNS Hard Error looking up yoctoproject.org' (MX): NXDomain" Jun 10 15:11:45 dig works for me Jun 10 15:12:09 davest, we gave up and all started using buildroot while you were gone! Jun 10 15:13:28 and RP took up professional motorcycle racing Jun 10 15:41:31 Crofton|work: :) Jun 10 15:59:54 I'm experiencing something odd with building an image. I have added systemd to DISTRO_FEATURES, but have not added it to IMAGE_INSTALL. Now if I build my image it includes udev 182 (i.e., the non-systemd version). However, if I explicitly add udev to IMAGE_INSTALL it installs udev 199 (i.e., the systemd version). Any explanation? Jun 10 16:03:33 Saur: have you set VIRTUAL-RUNTIME_init_manager ? Jun 10 16:05:38 bluelightning: Nope (as I said systemd is not yet installed, still working on that). I was just confused why explicitly adding udev to IMAGE_INSTALL would make a difference as a udev package is installed in both cases. Jun 10 16:06:37 Saur: I'm not entirely sure how that would come about either; it may have to do with udev being brought in through some other dependency Jun 10 16:07:45 bluelightning: Yeah, but PREFERRED_PROVIDER_udev == "systemd" (duse to systemd being in DISTRO_FEATURES), which is why I don't get what's happening... Jun 10 16:07:58 due to* Jun 10 16:09:37 Saur: what package management backend are you using? (ipk/rpm/deb) Jun 10 16:09:48 bluelightning: rpm Jun 10 16:11:09 Saur: was it a clean build from scratch after changing DISTRO_FEATURES or an existing TMPDIR? Jun 10 16:12:30 bluelightning: An existing... Jun 10 16:16:05 Saur: typically we recommend clearing out your TMPDIR after changing DISTRO_FEATURES Jun 10 16:16:16 Now you tell me ;) Jun 10 16:16:53 bluelightning: Wouldn't it just recreate from the sstate cache though? Jun 10 16:20:52 Saur: the parts that don't include DISTRO_FEATURES in their signature, yes Jun 10 16:23:18 Saur: the safest option is not to keep that either I suspect Jun 10 16:25:39 morning... Jun 10 16:25:54 bluelightning: Guess it's finally time to create a distro configuration and stop relying on poky.conf... Jun 10 16:26:30 Saur: if you're wanting to do customisation at that level I would recommend that yes **** BEGIN LOGGING AT Mon Jun 10 16:51:49 2013 Jun 10 18:43:26 RP: could use your thoughts on https://bugzilla.yoctoproject.org/show_bug.cgi?id=4634 whenever you have a chance. Jun 10 18:43:27 Bug 4634: normal, Undecided, ---, richard.purdie, NEW , BitBake default 'dirs' in exec_func (${B}) causes build failures Jun 10 19:27:12 what do I have to do to update the rpm metadata in tmp/deploy/rpm? Jun 10 19:27:52 because I'm building a recipe, it installs packages into tmp/deploy/rpm but when I do smart update, I don't get any new packages, which means the metadata inside the repodata/ is not been updated Jun 10 20:15:29 Hi Everyone. Definitely a new guy with yocto, and have everything running, but now trying to install chromium. Having some issues, here's what I get: Jun 10 20:15:34 ryan@Ryan:~/fsl-community-bsp/build$ bitbake chromium Jun 10 20:15:34 Loading cache: 100% |###############################################################################################################################################################################| ETA: 00:00:00 Jun 10 20:15:34 Loaded 1711 entries from dependency cache. Jun 10 20:15:34 ERROR: No recipes available for: Jun 10 20:15:34 /home/ryan/fsl-community-bsp/sources/meta-browser/recipes-gnome/gnome-settings-daemon/gnome-settings-daemon_2.32.1.bbappend Jun 10 20:15:37 /home/ryan/fsl-community-bsp/sources/meta-browser/recipes-support/nspr/nspr_4.9.6.bbappend Jun 10 20:15:39 /home/ryan/fsl-community-bsp/sources/meta-browser/recipes-graphics/libxi/libxi_1.7.1.bbappend Jun 10 20:15:41 ERROR: Command execution failed: Exited with 1 Jun 10 20:15:43 Summary: There were 2 ERROR messages shown, returning a non-zero exit code. Jun 10 20:25:47 Ryan777: sounds like a branch mismatch (e.g. using master meta-browser with dylan oe-core/poky) or meta-browser is just out of date wrt upstream. Jun 10 20:36:37 pastebin! Jun 10 20:41:03 meta-browser should stop bbappending things... Jun 10 20:43:52 that does seem odd, indeed Jun 10 20:56:20 hi folks. I'm trying to use the 'archive-patched-source' class to get an srpm of the sources for all my packages (including the patches). However, when I build my target, I don't get srpms, I get only xxx.diff.gz and xxx.showdata.dump. Is there some knob that I should be adjusting and am not? Jun 10 20:57:07 I've set ARCHIVER_MODE = "patched", ARCHIVER_MODE[type] = "srpm" Jun 10 20:57:10 in local.conf Jun 10 20:57:28 Also SOURCE_ARCHIVE_PACKAGE_TYPE = "srpm" Jun 10 20:58:05 SOURCE_ARCHIVE_PACKAGE_TYPE is set basd on ARCHIVER_MODE[type] by default, both should be redundant, but not harmful Jun 10 20:58:34 hmm Jun 10 20:58:43 * kergoth shrugs, never tried to use the srpm stuff personally, hopefully someone else has Jun 10 20:59:39 maybe I'll give the tar variety a go, but the srpm would be nice —we have some automated tool that accepts srpms Jun 10 21:09:31 Garibaldi|work: I think you should be setting ARCHIVE_TYPE and not ARCHIVER_MODE[type] Jun 10 21:10:32 * kergoth still rather dislikes the archiver.bbclass use semantics, confusing Jun 10 21:10:38 yes :/ Jun 10 21:10:47 actually it's not ARCHIVE_TYPE Jun 10 21:11:22 SOURCE_ARCHIVE_PACKAGE_TYPE, as you said Jun 10 21:11:27 *grumble* Jun 10 21:12:10 it's a weird hodgepodge of INHERIT-based choices, flags, and actual variables Jun 10 21:12:12 heh Jun 10 21:12:40 honestly I think we need to clean it up, this is too complicated Jun 10 21:13:53 there's a fair bit of non-idiomatic python in the implementation that makes it harder to read/grasp than necessary, too Jun 10 21:13:56 heh Jun 10 21:14:00 Jun 10 21:20:34 so the 'tar' type worked where the 'srpm' didn't Jun 10 21:21:00 I'm now getting xxx-patched.tar.gz Jun 10 21:22:49 but then, I removed SOURCE_ARCHIVE_PACKAGE_TYPE; maybe me having had set it caused some problems for srpm Jun 10 21:37:11 yeah, ARCHIVER_MODE[type] = "tar" works, = "srpm" doesn't :-( Jun 10 21:38:53 with srpm, I get only xxx.showdata.dump and xxx.diff.gz Jun 10 21:39:37 and some -script-logs.tar.gz Jun 10 21:43:49 Garibaldi|work: looking at the implementation to be honest I can't see how it could create srpms despite claiming to do so... Jun 10 21:44:31 bluelightning: yeah, I'm looking at the same Jun 10 21:46:02 tar it is then :-) **** ENDING LOGGING AT Tue Jun 11 02:59:58 2013