**** BEGIN LOGGING AT Mon Nov 09 02:59:59 2015 Nov 09 08:10:17 morning all Nov 09 08:42:52 good morning Nov 09 08:46:14 how do I switch the rootfs for my yocto image from ext4 to btrfs? Nov 09 08:46:21 is there a standard way? Nov 09 08:46:42 or is this something which needs to be done "manually"? Nov 09 09:10:44 bambule__: if you do IMAGE_FSTYPES += "btrfs" in your configuration that should be enough Nov 09 09:26:43 bluelightning: will try that. thanks! Nov 09 12:37:11 kbingham-: am just trying to reproduce your issue with devtool; though linux-imx seems to be taking quite some time to fetch unfortunately... Nov 09 12:37:34 kbingham-: I don't suppose you can shed any further light on what triggers it? Nov 09 12:41:17 Hi everybody Nov 09 12:41:22 I'm trying to integrate RTAI as a yocto-project recipe and I have some problems with that: Nov 09 12:41:29 To compile RTAI, I need a prepared kernel source tree including the kernel scripts and symbol version dump etc. Nov 09 12:41:35 But the KERNEL_STAGING_DIR points to "work_shared" where i only find a source tree that doesn't seem to pe prepared. Nov 09 12:41:40 How can I compile RTAI or anything else against a prepared kernel source with yocto-project, or do I run in a completely wrong direction? Nov 09 12:41:43 Hello everyone! I'm grasping at straws with a custom i.mx6 board. I've been using the 3.10.53 kernel very happily for a while now, but want to upgrade to the 3.14.28 GA kernel. However, USB is not working. USB OTG, or either USB hub. The only warnings/errors I see on boot for the USB devices are messages about using dummy regulators for vbus. I've been poring over the device tree of similar boards, like the SabreLite and SabreSD, bu Nov 09 12:42:34 I'm not sure what I might be missing, still. I'm reasonably sure it is a device tree requirement that changed, as I doubt the drivers all of a sudden broke... Nov 09 12:42:56 Any help would be appreciated! Even which direction to look, as I'm hitting brick walls now... Nov 09 13:30:22 kbingham-: ah, found the cause; it is indeed fixed in 2.0 (and we ought to backport for 1.8.2) Nov 09 13:52:39 * joshuagl fido senses tingling Nov 09 13:55:06 joshuagl: yes... I should really have done another set of backports before 1.8.1, oh well Nov 09 13:55:37 I have the important fixes on a branch but need to run oe-selftest over them before sending Nov 09 13:59:51 joshuagl: please include the fontcache.bbclass change as well Nov 09 14:00:28 "Oct 27 To openembedded- [PATCH][fido] fontcache: allow to pass extra parameters and environment to fc-cache" Nov 09 14:01:46 JaMa: oh yeah, I wanted to spend some more time looking at that to try and understand the impact Nov 09 15:18:27 can anyone else whose using FIDO branch compile binutils for the target? for me I'm getting a compiler error for libiberty LONG_MIN not defined. Nov 09 15:18:42 looking at config.h HAVE_LIMITS is not defined to 1 Nov 09 15:55:01 mtownsend1973: seemed to build OK here FWIW Nov 09 15:55:27 that was for imx6sxsabresd (just what I happened to be testing at the moment) Nov 09 16:02:19 bluelightning: thanks for trying out, out of interest was HAVE_LIMITS defined and set to 1 in build.arm-oe-linux-gnueabi.arm-oe-linux-gnueabi/libiberty/config.h? Nov 09 16:08:16 mtownsend1973: HAVE_LIMITS_H was defined to 1, I assume that's what you meant... Nov 09 16:11:34 bluelightning: thanks, I just need to work out why it's not on my system now :) Nov 09 16:33:57 Hi. Anyone here who has worked with gpgme recipes before? Nov 09 16:34:43 gatisp: probably, why? Nov 09 16:35:25 there seem to be some issues with BBCLASSEXTEND = "native nativesdk" Nov 09 16:35:37 after i install the toolchain Nov 09 16:35:54 it still seems to search for things in the default prefix Nov 09 16:36:27 GPGME 2015-11-09 16:28:00 <0x0d8f> gpgme_dinfo: gpg='/opt/b2qt/1.7/sysroots/x86_64-pokysdk-linux/usr/bin/gpg' Nov 09 16:36:56 instead of gpg='/${CUSTOM_PREFIX}/sysroots/x86_64-pokysdk-linux/usr/bin/gpg' Nov 09 16:37:51 if i set symlink in the place where it search to the same path in my toolchain gpgme works Nov 09 16:42:01 gatisp: is that on master branch? anyway if it's oe-core-1.7 or later open a defect in https://bugzilla.yoctoproject.org/ -- seems like an easy fix. Nov 09 16:43:13 vmeson, do you have a fix in mind already? Nov 09 16:44:26 looks like I have the latest version of that recipe Nov 09 16:45:44 or not really, will have to try more recent version Nov 09 16:50:14 gatisp: nothing specific, if you explain how to reproduce the bug in the defect, we'll figure it out. Let me know if it still happens on master. Nov 09 16:59:47 vmeson, yeah, the issue is still there, only i am unsure at the moment how to make this into a simple test case for bug report Nov 09 17:01:41 talked to fast, looking at gpgme debug output, path seems to be correct, probably the rest i will figure out my self :) Nov 09 17:02:30 gatisp: a relocation script runs at install time of the sdk which replaces those paths with the new install prefix Nov 09 17:02:43 its expected they'll be hardcoded, but it should be substituted at install time Nov 09 17:16:04 kergoth, not everything seems to be replaced with the install prefix Nov 09 17:16:06 gp@gp:~/b2qt-4.0/Boot2Qt-4.x/iMX6-eLinux/toolchain/sysroots/x86_64-pokysdk-linux/usr/bin$ ./gpgconf Nov 09 17:16:06 gpg:GPG for OpenPGP:/opt/b2qt/1.7/sysroots/x86_64-pokysdk-linux/usr/bin/gpg2 Nov 09 17:16:06 gpg-agent:GPG Agent:/opt/b2qt/1.7/sysroots/x86_64-pokysdk-linux/usr/bin/gpg-agent Nov 09 17:16:06 scdaemon:Smartcard Daemon:/opt/b2qt/1.7/sysroots/x86_64-pokysdk-linux/usr/lib/gnupg/scdaemon Nov 09 17:16:07 gpgsm:GPG for S/MIME:/opt/b2qt/1.7/sysroots/x86_64-pokysdk-linux/usr/bin/gpgsm Nov 09 17:16:09 dirmngr:Directory Manager:/opt/b2qt/1.7/sysroots/x86_64-pokysdk-linux/usr/bin/dirmngr Nov 09 17:16:31 it still have that "/opt/b2qt/1.7" Nov 09 17:16:57 afaik it does a sed search/replace, so only bits in text files / scripts will be adjusted, not binaries. but i'm not sure as to the exact logic, see relocate_sdk.py in the sdk if you're curious Nov 09 17:18:18 i think the issue probably is in the recipe Nov 09 17:18:25 not sdk install script Nov 09 20:37:55 <[Sno]> is there a bug tracked regarding the race condition described in https://lists.yoctoproject.org/pipermail/yocto/2015-May/024750.html Nov 09 20:38:34 <[Sno]> I run into incomplete Module.symvers when building external wifi driver depending on cfg80211 - configured as module Nov 09 20:39:17 <[Sno]> I figured out, doing staging_workdir for kernel should be prepared after do_compile_kernelmodules - not after do_compile Nov 09 22:05:08 otavio: heads up: mut is running tonight, if the only failures are in fsl's gstreamer patching then it's going into master tomorrow **** ENDING LOGGING AT Tue Nov 10 02:59:59 2015