**** BEGIN LOGGING AT Tue Feb 26 02:59:58 2013 Feb 26 03:11:36 thanks dvhart, I just try to build 2.6 and 3.4 , to check if our own PCIe driver works same on both Feb 26 03:14:19 we have 1 PCIe hardware works on 2.6 desktop Linux with Intel server, but can't get it work on Yocto with new designed jasperforest board, just try to verify if it's driver's problem or hardware's problem Feb 26 03:26:09 another question , can I force all kernel modules build into target binray image by add line : IMAGE_INSTALL += "kernel-modules" to conf/local.conf? Feb 26 03:27:02 tried, but still nothing in the /lib/modules/3.4.xxx/ Feb 26 04:17:55 EddyDev, hrm, I don't remember the package name off the top of my head. But the variable you want to set is CORE_IMAGE_EXTRA_INSTALL (not IMAGE_INSTALL) Feb 26 04:18:21 even better would be to use the MACHINE_ESSENTIAL* variables, see the kernel-dev manual for details (or the reference manual glossary) Feb 26 08:45:23 Hello. Is somebody able to help with an rpath-qa-error in "meta/recipes-support/db/db_5.3.15.bb"? occurs in poky-danny-8.0 and a recent poky.git upon adding the configure-flag "--enable-cxx" to the "DB5_CONFIG" option. Feb 26 09:13:27 good morning Feb 26 09:14:50 good morning~ Feb 26 09:42:03 morning all Feb 26 13:22:16 Can anyone tel me how to add prebuilt toolchain in yocto 1.3? Feb 26 13:35:39 Keshav: is this because you want to use a vendor toolchain or because you are interested in speeding up the build? Feb 26 15:59:45 YPTM: Darren has joined Feb 26 15:59:51 YPTM: welcome to the technical team meeting. If you are on the bridge, please let me know. Thanks! Feb 26 15:59:52 YPTM: Paul Eggleton is here Feb 26 15:59:57 YPTM: belen has joined Feb 26 16:00:09 YPTM: Tom Z here Feb 26 16:00:23 YPTM: Mark Hatle is here.. Feb 26 16:00:37 YPTM: Laurentiu Palcu joined Feb 26 16:01:21 YPTM: Scott Rifenbark joined call Feb 26 16:01:27 YPTM: Beth is here Feb 26 16:01:33 YPTM: Jeff Polk is here Feb 26 16:01:42 YPTM: Bruce Ashfield on the call, but only on for 15 mins. Feb 26 16:02:39 YPTM: jzhang's on line will drop early Feb 26 16:02:53 YPTM: Cristian is online Feb 26 16:03:16 Song_Liu: YPTM: Sean just joined. Feb 26 16:03:22 YPTM: Any opens for the meeting? Feb 26 16:03:25 YPTM: nitin joined the bridge Feb 26 16:03:51 Song_Liu: Yocto BoF notes collection? Feb 26 16:03:57 YTPM: Saul is here Feb 26 16:04:01 https://wiki.yoctoproject.org/wiki/1.4_QA_Status Feb 26 16:04:11 * zeddii can't hear anything on the call. but I think it is my line. Feb 26 16:04:30 zeddii: I can hear fine. Do you have your phone upside down? ;) Feb 26 16:04:45 Song_Liu: that was an open item^^^ Feb 26 16:04:57 hah. not this time. I could barely hear the bridge prompts, so it's my line. Feb 26 16:05:16 Yocto BoF notes collection Feb 26 16:05:26 "Birds of a Feather" Feb 26 16:05:40 tig Feb 26 16:05:41 think of it as a "user's group" Feb 26 16:05:54 YPTM: davest on the call Feb 26 16:05:54 'er.. no "'" Feb 26 16:08:19 :) Feb 26 16:09:05 oops, usually do Feb 26 16:09:26 blloyd_: thank Feb 26 16:13:45 darn double click... Feb 26 16:13:55 Song_Liu: link for the WDD? Feb 26 16:14:43 aw crap. I heard my name and nothing more. Feb 26 16:16:18 https://wiki.yoctoproject.org/wiki/Yocto_Bug_Trend Feb 26 16:16:22 WDD - Weighted Defect Density Feb 26 16:16:31 * zeddii has to drop. Feb 26 16:16:45 Song_Liu: thx Feb 26 16:18:20 * darknighte is still looking at the bug trends Feb 26 16:21:21 fray: do you have some time to review the RO from Qi.Chen? Feb 26 16:21:31 I should... Feb 26 16:21:59 fray: I am going to another build with it enabled today also while waiting for 1.3.1 to build out. Feb 26 16:22:17 has it been put into a single set of commits? I just need to know what to review Feb 26 16:23:16 dvhart: what's a minnow? ;) Feb 26 16:24:07 minnowboard.org Feb 26 16:25:10 darknighte, well placed questions, I'll pay you later ;-) Feb 26 16:25:10 hello, i have a warning it says: "Files/directories were installed but not shipped" why could this warning appear Feb 26 16:25:28 dvhart: send me a minnowboard and we'll call it even! ;) Feb 26 16:25:51 wfailla, it happens with the install step installs files (make install) but those files are not covered by any of the FILES_* variables in the recipe Feb 26 16:27:12 dvhart: btw, the datasheet link for the minnow doesn't really have any info when I look at it. (http://www.minnowboard.org/wp-content/themes/responsive/images/minnowflyer_thumb.png) Feb 26 16:27:18 dvhart, but they are Feb 26 16:27:32 dvhart, well it is Feb 26 16:28:08 darknighte, are you right clicking or something? Feb 26 16:28:19 darknighte, the link actually goes to a pdf... Feb 26 16:28:31 ya.. it's basic information, but there is info there.. Feb 26 16:28:35 http://www.minnowboard.org/wp-content/themes/responsive/files/MinnowFlyer-ELC2013_new.pdf Feb 26 16:28:42 dvhart: right. the pdf has just an image, no info for me. Feb 26 16:28:59 there should be two pages.. second page is the info.. first page the image(s) Feb 26 16:29:02 darknighte, you sent a link to a png above Feb 26 16:29:21 darknighte, are you look for something more like a technical datasheet of a processor? Feb 26 16:29:22 weird. it just updated. Feb 26 16:29:25 * darknighte shrugs Feb 26 16:29:57 ok :-) Feb 26 16:30:02 dvhart: actually yes, I was, but the blank page seemed fishy. Feb 26 16:30:50 dvhart: fwiw, the 2nd page is still incomplete. probably it's just my browser, or something Feb 26 16:33:00 I see both pages there Feb 26 16:33:08 way younger :) Feb 26 16:33:25 davest is in far better shape though ;-) Feb 26 16:34:01 My slides are here: http://elinux.org/images/9/96/2013-ELC-Building-a-distro-with-YP.pdf Feb 26 16:35:00 * darknighte does a little dance Feb 26 16:35:29 * davest heads for the showers Feb 26 16:35:30 tomz2, zeddii, nitink : we could meet now for 25m if you prefer Feb 26 16:35:40 dvhart: ok Feb 26 16:35:42 YPTM: thank you all for joining the meeting, have a nice day/evening! Feb 26 16:35:54 * zeddii is still on his all hands call. sorry. Feb 26 16:36:04 but go ahead without me, I won't spoil the fun :P Feb 26 16:52:59 dvhart: are you guys dialed in - just noticed the 25min meeting thing above Feb 26 16:54:37 hey tomz2 we just dropped Feb 26 16:54:45 I have a 9, no worries, email is fine Feb 26 16:54:52 sorry for causing confusion Feb 26 17:18:39 Hey y'all, I appear to be having trouble with libtool and have seen some threads related to that on the list, but I haven't found a real resolution to my problem Feb 26 17:19:16 -L/usr/lib is winding up in my path and it's blowing up packages seemingly randomly. This is coming in from .la files, from what I can tell. What can I do to stop this from happening? Feb 26 17:19:48 peachj, get your distribution or OS to delete .la files from /usr/lib Feb 26 17:19:57 peachj, e.g. http://wiki.debian.org/ReleaseGoals/LAFileRemoval Feb 26 17:20:08 Oooh, okay, so that's where it's coming from Feb 26 17:20:09 Thanks Feb 26 17:20:14 use the libtoolize within the YP build system.. it will fix many of those issues.. Feb 26 17:20:18 we patch libtool to avoid unexpected occurrances of that, but also make sure you don't have a Makefile.am passing it explicitly in LDFLAGS/CFLAGS. libtool won't remove explicitly added bits Feb 26 17:20:32 well, we used to, i expect we rely on its sysroot support omstly now Feb 26 17:20:34 if your recipes don't use the libtool within YP/OE environment, you will get host contamination Feb 26 17:26:47 kergoth: I just tried your latest patch 9ca8e513dc142dd7aefadeb6db2ccdb00f38b39f (cooker: now that we use a Pool, raise the exceptions) Feb 26 17:26:53 and it loops indefinitely or something. Feb 26 17:27:07 I reverted it and am back to the TypeError: object of type 'ExpansionError' has no len() Feb 26 17:27:21 NOTE, I have a project that definitely has errors in it intentially. Feb 26 17:27:25 intentionally that is. Feb 26 17:27:31 like: NOTE: Error during finalise of /space/jw/5/small/iso_small/layers/oe-core/meta/recipes-extended/images/core-image-lsb.bb Feb 26 17:28:05 But with your latest patch it never emits any of that it just sits there indefinitely. Any ideas how to debug it? Feb 26 17:28:37 hmm, interesting, that makes no sense Feb 26 17:29:02 multiprocessing's Pool has built in exception handling, if it detects one, it passes it along and reraises in the other process, in the next() method of IMapResults Feb 26 17:29:13 I agree. I think there is some other underlying problem. Of course if I also revert 8af519a49a3374bd9004864ef31ca8aa328e9f34, then I get the origin error. Feb 26 17:30:22 kergoth: So the way I am getting this error is by making changes that don't compile with in a bbclass. Feb 26 17:31:15 I am assuming if my "totally doesn't work for oe because it is broken" patch is applied you might see the same behavior, but I am not sure. Next I'll try the same thing with bitbake latest / oe-latest. Feb 26 17:31:55 exactly what do you mean by 'changes that don't compile'? Feb 26 17:31:59 a python SyntaxError? Feb 26 17:32:02 Yes. Feb 26 17:32:11 in what, a task, event handler, or anonymous python? Feb 26 17:32:27 Changes that don't parse. I am not exactly sure, where I broke it. Feb 26 17:32:43 I'll have a more concrete test case in a few minutes. Feb 26 17:32:45 just pastebin what you're doing to force the failure Feb 26 17:32:58 there are a lot of slightly different codepaths for the many ways in which a parse can fail Feb 26 17:33:00 Yes, I am trying to isolate that right now against poky master. Feb 26 17:33:07 we really need test case metadata for each Feb 26 17:33:30 k Feb 26 17:35:46 there's recipe syntax, config file syntax, ExpansionError (e.g. a failure in a := + ${@}), SyntaxError, ordinary Exception, and BaseException (SystemExit/os.exit/bb.fatal), and much of those need testing in various contexts (recipe vs config, event handlers). i tested nearly all combinations i could come up with when doing the orignial parallel parsing implementation, but don't have that around anymore Feb 26 17:36:45 http://pastebin.com/f1rm9AN9 Feb 26 17:37:58 the code is near the bottom of bitbake/lib/bb/cooker.py if you want to investigate any further Feb 26 17:38:07 otherwise i'll see if i can look into it sometime in the next few hours Feb 26 17:38:43 kergoth: poky master doesn't have your latest bitbake patch yet. Feb 26 17:38:58 I applied it and used my "oops I broke syslinux patch", and it hangs indefinitely. Feb 26 17:39:27 kergoth: So if you use poky master as the reference your latest patch has to be applied first Feb 26 17:40:12 I am not exactly sure why it hangs / loops or what ever it is doing. I just bisected it to the to commits I mentioned ealier, both for the pool code. Feb 26 17:41:33 k, thanks Feb 26 17:42:02 On the libtool stuff: I don't appear to have any .la files in /usr/lib. There's a chance they may be in some of the packages for which we've downloaded tarballs. If those packages are using autotools, do those .la need to be manually removed from the tarballs? Feb 26 17:44:28 FWIW The error in the pastebin is the result of referencing a variable that didn't exist. I named the variable incorrectly in one place vs another. Feb 26 17:48:26 actually I was wrong. The failure is: cfgfile.write('SERIAL %s', % syslinux_serial) vs cfgfile.write('SERIAL %s' % syslinux_serial) Feb 26 17:48:34 note the comma is the difference. Feb 26 19:34:33 I see live image still uses syslinux for booting. Anyone know if there is any work to move this to grub 2? Feb 26 20:19:20 blloyd_: i think the plans are to remove grub entirely Feb 26 20:19:48 blloyd_: did you know that grub is responsible approximately ~10 minutes of your from-scratch build time? Feb 26 20:19:54 * rburton has a pet peve that makes him angry Feb 26 20:40:52 quelle surprise Feb 26 20:41:28 ERROR: No recipes available for: Feb 26 20:41:30 /srv/home/pokybuild/yocto-autobuilder/yocto-slave/nightly-fsl-arm/build/yocto/meta-fsl-arm/recipes-core/udev/udev_173.bbappend Feb 26 20:41:32 grrr Feb 26 21:09:52 otavio: you submitted some patches for Boundary Devices sabrelite u-boot recently. where did those end up? Feb 26 21:10:42 I don't see the 6x_bootscript stuff anywhere in meta-fsl-arm Feb 26 21:12:15 oh hm, meta-fsl-arm-extra is something else? Feb 26 21:21:22 hollisb: it is for support to non FSL boards Feb 26 21:21:44 hollisb: but you're looking to sabrelite or nitrogen? Feb 26 21:22:23 otavio: sabrelite Feb 26 21:22:39 hollisb: sabrelite is at fsl-arm Feb 26 21:23:12 but stock sabrelite uboot uses /6q_bootscript (or 6x*), and I want to build an image that provides those Feb 26 21:23:28 IOW works without requiring people to flash a new uboot Feb 26 21:25:05 otavio: so naturally I came across your uboot recipe in -extra Feb 26 21:40:45 hollisb: the sabrelite is using those IIRC Feb 26 21:40:55 otavio: exactly Feb 26 21:41:14 otavio: I want a package that provides those scripts Feb 26 21:44:52 hollisb: u-boot-bondary-scripts IIRC Feb 26 21:45:55 otavio: do you recommend just adding the -extra layer to my build? Feb 26 21:46:08 hollisb: yes; seems easier Feb 26 21:46:18 otavio: btw, the README for that repo doesn't say anything about support for non-Freescale boards. might be helpful :-) Feb 26 21:46:25 hollisb: you can check the nitrogen Feb 26 21:46:32 hollisb: please send a patch ;-) Feb 26 21:46:40 :-) Feb 26 22:05:36 halstead: ping Feb 26 22:26:09 I am in hell Feb 26 22:26:29 do we have a list of "supported" distros for running builds on? Feb 26 22:27:11 Crofton, We have a list of tested distros. Bitbake will warn you if you aren't using one. Feb 26 22:27:55 yeah, do we have the list where I can refer to it Feb 26 22:32:03 Crofton, In http://git.yoctoproject.org/cgit/cgit.cgi/poky/tree/meta-yocto/conf/distro/poky.conf?h=danny at about line 71 is what I found. Feb 26 22:40:33 we should probably make that more obvious :) Feb 26 22:41:31 http://lists.linuxtogo.org/pipermail/openembedded-core/2012-December/033982.html Feb 26 22:41:46 and this seems to contradict centos 5.6 working out of the box? Feb 27 00:18:20 Okay, idle curiousity: We're seeing some strangeness on systems (I think only CentOS 5.9 systems). It looks suspiciously as though, long after our prelinker is done running, the sizes of some prelinked files occasionally change on disk. Feb 27 00:19:07 This is sort of familiar because we had a problem in the past where du would show a bunch of files as 0 bytes until some unspecified later time, but the size changes (we think) are being reported by tar, which would be checking buf.st_size. And that should, I think, NOT be changing after a file has been closed... Feb 27 00:19:35 And the obvious reason for this to come up would be that pseudo 1.5 suppresses fsyncs. (Our branch of prelink has fsyncs in it precisely because of the earlier problem.) Feb 27 00:19:55 Anyone seen cases where, after a file has been closed, buf.st_size changes at unspecified later times? Feb 27 02:07:05 is it possible to buy a Fish River Island 2 board? Perhaps the Kontron dev kit? I seem to recall some being shipped recently. Feb 27 02:12:28 wmat: didn't go to ELC last year? Feb 27 02:12:46 mranostay: no, i couldn't make it Feb 27 02:13:26 * mranostay notes the indefinite lease terms in box for his sample :) Feb 27 02:13:42 aka "if we catch you ebay'ing these" :) Feb 27 02:13:53 mranostay: prpplague needs one Feb 27 02:15:06 er he wasn't at Yocto Day last year? Feb 27 02:15:20 i guess not Feb 27 02:15:28 mine is collecting dust atm Feb 27 02:16:38 i've let him know Feb 27 02:33:47 bluelightning : Yes i want to use a vendor toolchain **** ENDING LOGGING AT Wed Feb 27 02:59:58 2013