**** BEGIN LOGGING AT Thu Dec 11 02:59:58 2014 Dec 11 08:04:44 good morning Dec 11 09:20:29 Is there a task to only run device-tree generation? Dec 11 11:02:25 morning all Dec 11 11:02:47 hi Dec 11 11:04:50 I know it's not yocto specific, but I recently switched over from ash to bash in yocto and I noticed that my line wrap does not work properly anymore when connected via serial Dec 11 11:06:13 it seems to be the same behavior as described in http://www.linuxquestions.org/questions/linux-general-1/bash-does-not-wrap-long-lines-correctly-121061/#post626174 Dec 11 11:06:24 but I did not modify PS1 Dec 11 11:06:37 and it works when using ash instead of bash Dec 11 11:07:39 hmm, I don't know what would cause that but it sounds like a bug Dec 11 11:34:00 bluelightning: mh, okay Dec 11 12:23:30 can any one please point me out what UBOOT_MACHINES ?= in recipe file Dec 11 12:26:31 ramose__, could you rephrase the question? Dec 11 12:27:50 There is variable named UBOOT_MACHINES in .bb file ,I wanted to know what it is for? Dec 11 12:30:34 UBOOT_MACHINE is used to select which u-boot configuration to build. Thus it's normally set in your .conf Dec 11 12:35:41 AndersD: Thanks,so I need to check in u-boot source what is my u-boot configuration file and place it in machine.conf? Dec 11 12:38:09 AndersD :Also wanted to know when I need to provide JFFS2_ERASEBLOCK varibale in machine.conf file? Dec 11 12:41:15 Well, if you're going to use jffs2, then you should specify the erase block size, which is done by JFFS2_ERASEBLOCK Dec 11 12:48:06 hi Dec 11 12:48:31 I want to add mets-efl layer to my yocto dizzy Dec 11 12:48:35 meta-efl Dec 11 12:48:52 Can anyone suggest me the changes require Dec 11 12:49:39 It has two extra version LAYERVERSIOn and LAYERDEPENDS Dec 11 12:59:58 "It has two extra version LAYERVERSIOn and LAYERDEPENDS" ??? Dec 11 13:00:32 and you're probably looking at master branch instead of dizzy of meta-openembedded Dec 11 13:15:29 AndersD: Thanks :) Dec 11 13:59:58 Hello, trying to use systemd.bbclass. It is working fine of most of my custom recipes, but on one (that uses the same kind of bbappend) the daemon will not start on boot (when the package is in the image). But, what I find funny, is that if I install the rpm manually, it will work. The only thing I add in my recipe is : 1. inherit sytemd, 2. SYSTEMD_SERVICE_${PN} = "my.service". 3. FILES_${PN} += "/lib/" 4. do_install_append() { } which copy the service file Dec 11 13:59:59 in the systemd directory. How can I debug this ?!!! Thank you ! Dec 11 14:04:57 Can anyone point out how can we run echo command from .bb file? Dec 11 14:06:15 remose_ : maybe this could work : python some_python_function () { Dec 11 14:06:16 d.setVar("TEXT", "Hello World") Dec 11 14:06:16 print d.getVar("TEXT", True) Dec 11 14:06:16 } Dec 11 14:09:23 try bb.warn() instead of print() Dec 11 14:12:34 weeb0: (a) have a look at packages-split/ under the workdir for the recipe to verify the service file is being installed and packaged correctly; then (b) you can check the postinstall script for the recipe by looking at bitbake -e recipename | less and searching for ^pkg_postinst Dec 11 14:16:22 thank you bluelightning Dec 11 14:17:05 This package is the only one that asks me to add FILES_${PN}-dbg ... don't know why... Dec 11 15:31:56 denix: you sound a bit like a dalek Dec 11 15:32:14 rburton: I'm redialing now Dec 11 15:44:04 <_qwerty_> Hi all, I have to add two udev rules, how I can set from yocto? Dec 11 15:51:05 _qwerty_: Have your recipe install the rules into /lib/udev/rules.d/ Dec 11 15:52:05 <_qwerty_> gabredd: where I have to put it? Dec 11 15:54:01 bluelightning: A) in packages-split, the service file is there. for B) I cannot see anything relevant to the systemd in postinstall Dec 11 16:01:38 _qwerty_: Usually, udev rules are tied to some specific feature that already has a bitbake recipe for drivers and whatnot. Dec 11 16:01:46 _qwerty_: So you usually add it to those recipes. Dec 11 16:02:08 _qwerty_: otherwise, you could create a recipe that has the sole purpose of installing the udev rule. Dec 11 16:02:40 _qwerty_: or you could possibly do a .bbappend to some other package (like udev) and have the rule installed that way. Dec 11 16:04:39 <_qwerty_> gabrbedd: Yes ok Dec 11 16:43:12 weeb0: as far as I can tell looking at systemd.bbclass if your recipe inherits systemd, you have systemd in DISTRO_FEATURES, you haven't touched SYSTEMD_PACKAGES, and you have set SYSTEMD_SERVICE_${PN}, then it should work Dec 11 16:43:17 weeb0: are all of those true? Dec 11 16:46:22 bluelightning: all of those are true. and when I install manually the rpm it works great. it is only when I load the package with an image recipe it will not start on boot. but the service file is copied in the system. Dec 11 16:47:05 weeb0: ah, I see... so the postinstall script must be created, or that wouldn't work Dec 11 16:47:29 it's just when the package is installed as part of the image that isn't working then Dec 11 16:47:49 bluelightning: exact. it is very strange for me. Dec 11 16:48:27 weeb0: maybe have a look at log.do_rootfs to see if there is anything suspicious in there relating to this Dec 11 16:48:59 bluelightning: thank you, I will have a look Dec 11 17:16:38 RP: did you see the image generation fixes I sent? Dec 11 17:24:59 otavio: see, yes, digest and review, no Dec 11 17:25:15 otavio: will likely be next week before I get to it to be honest, I have some vacation coming up Dec 11 17:48:51 RP: right; it is not urgent but it will allow for a great cleanup in our fs handling routines internally Dec 11 17:49:01 RP: and also in fsl-arm Dec 11 18:20:18 I noticed that "read-only-rootfs" in IMAGE_FEATURES was not sufficient for it to work. Dec 11 18:20:31 Had to also change the rcS-default in sysvinit Dec 11 18:20:38 File a documentation bug? Dec 11 18:24:23 gabrbedd: I don't think any extra steps should really be necessary for core stuff like that, so that would be more of a general bug I would think Dec 11 18:34:13 bluelightning: ok, thanks. Dec 11 19:57:09 Is there a layer that includes the latest linux release? Dec 11 20:09:05 Hi everyone. Probably not completely problem of yocto but still. Using mainline 3.14.16 kernel., beaglebone black am3358. In kernel there are two drivers: omap_sham for sha and md5 crypto hw and omap_aes for aes crypto hw. However in /proc/interrupts i can only see omap_sham and while using it I see that it is used. However I cannot bring omap_aes, anyone had this problem? Dec 11 20:15:46 JustasMika: you'll probably get more response for that on #beagle Dec 11 20:17:03 thanks, will try :) Dec 11 22:03:49 I recently moved to 1.6 (don't ask how far behind I was), and I'm having trouble figuring out the equivalent of doing "guilt/quilt applied" in a kernel devshell. **** ENDING LOGGING AT Fri Dec 12 02:59:59 2014