**** BEGIN LOGGING AT Tue Oct 10 03:00:01 2017
Oct 10 06:44:55 <Dvorkin> rburton, thank you!
Oct 10 07:06:27 <Dvorkin> may I use OEROOT in my local.conf as the base path?
Oct 10 07:41:53 <knip> hello
Oct 10 07:42:20 <knip> I have builded an image for raspberry pi 3 using meta-rpi and meta-raspberrypi
Oct 10 07:42:23 <knip> it boots
Oct 10 07:42:36 <knip> but I don't see the login prompt
Oct 10 07:42:44 <knip> any idea ?
Oct 10 07:43:10 <LetoThe2nd> where do you look fopr the login prompt?
Oct 10 07:45:01 <knip> I'm using an FTDI cable
Oct 10 07:45:22 <knip> connected to the pins of the rpi3
Oct 10 07:45:36 <knip> I see all the boot messages
Oct 10 07:45:42 <LetoThe2nd> ok, so on the serial port. thats right, usually there should be one.
Oct 10 07:46:09 <knip> the last message I see is "INIT: Id "S0" respawning too fast: disabled for 5 minutes"
Oct 10 07:47:08 <LetoThe2nd> well thats the reason.
Oct 10 07:47:22 <LetoThe2nd> are you using a custom kernel config?
Oct 10 07:47:35 <LetoThe2nd> this can happen for example, if system lacks some kernel features
Oct 10 07:47:47 <knip> no
Oct 10 07:48:07 <knip> I'm using the original recipe inside meta-rpi
Oct 10 07:48:20 <LetoThe2nd> ok, and why meta-rpi AND meta-raspberrypi?
Oct 10 07:48:39 <knip> do you mean why I'm using both of them ?
Oct 10 07:48:50 <LetoThe2nd> yes
Oct 10 07:49:11 <LetoThe2nd> i'm no rpi user, but mixing BSP usually is not a good idea. and it sounds a bit like it
Oct 10 07:49:50 <knip> I was following an online tutorial
Oct 10 07:50:14 <knip> maybe I have just to try with one and I have to use the normal core-image-minimal
Oct 10 07:51:03 <LetoThe2nd> thats a good idea!
Oct 10 08:32:07 <open-nandra> hi
Oct 10 08:32:36 <open-nandra> I'm bumping poky from krogoth to rocko and getting errors like during expanding variables like: which triggered exception ImportError: No module named 'features'
Oct 10 08:32:43 <open-nandra> am I missing something on my host PC?
Oct 10 09:09:09 <nayfe> open-nandra: Hi, hard to tell, can you provide more informations? do you have python3 or python2.7 only ?
Oct 10 09:09:45 <open-nandra> nayfe: I found a reason it was some internal features.py file which was missing
Oct 10 09:10:14 <open-nandra> btw I'm seeing also error in lines like:       try:
Oct 10 09:10:15 <open-nandra>             ver = gitrev_run("git rev-parse --short HEAD", gitdir)
Oct 10 09:10:15 <open-nandra>         except Exception, e:
Oct 10 09:10:15 <open-nandra>             bb.fatal(str(e))
Oct 10 09:10:35 <rburton> any particular bit of that?
Oct 10 09:10:39 <open-nandra> poky complains about except Exception, e: like: SyntaxError: invalid syntax
Oct 10 09:10:47 <open-nandra> no idea
Oct 10 09:10:49 <open-nandra> why
Oct 10 09:10:51 <rburton> thats becayse we use python3 now
Oct 10 09:10:55 <rburton> and that isn't valid python3
Oct 10 09:11:06 <open-nandra> oh
Oct 10 09:11:07 <open-nandra> ok
Oct 10 09:11:20 <rburton> if you're porting between releases 1) read the release notes 2) you'll need to port all your custom classes to python3
Oct 10 09:11:49 <LetoThe2nd> rburton: 2) is obvious, but hey, c'mon, 1)?
Oct 10 09:11:58 <LetoThe2nd> </SCNR>
Oct 10 09:12:08 <rburton> i know it's controversial  :)
Oct 10 09:12:13 <open-nandra> rburton: I did look at : http://www.yoctoproject.org/docs/2.3/ref-manual/ref-manual.html#moving-to-the-yocto-project-2.3-release
Oct 10 09:12:20 <open-nandra> maybe I overlooked it
Oct 10 09:12:27 <open-nandra> sorry about spamming :)
Oct 10 09:12:40 * LetoThe2nd feels totally left out alone and starts a sadsad sob story about being antialphabetic!
Oct 10 09:13:04 <rburton> open-nandra: it was in the 2.2 release: http://www.yoctoproject.org/docs/2.3/ref-manual/ref-manual.html#migration-2.2-metadata-now-must-use-python-3-syntax
Oct 10 09:14:03 <open-nandra> rburton: thanks a lot
Oct 10 12:39:59 <diego_r> fray: I know it's old, but thanks for the explanation here: https://lists.yoctoproject.org/pipermail/yocto/2011-November/003542.html I had same question yesterday, and you had the perfect answer!
Oct 10 12:41:47 <peacememories> hmm, how do i know which version of bitbake to use with openembedded-core pyro?
Oct 10 12:44:58 <diego_r> peacememories: have you already selected a distribution? e.g. poky has bitbake included
Oct 10 12:45:17 <peacememories> i'm trying to setup my own minimal distribution
Oct 10 12:46:08 <diego_r> poky is for sure more minimal than say, angstrom.
Oct 10 12:48:59 <LetoThe2nd> peacememories: i second the position to start with poky, but if you absolutely think you need to do the matching by hand, see: https://www.openembedded.org/wiki/OE-Core_Standalone_Setup
Oct 10 13:00:16 <rburton> peacememories: https://wiki.yoctoproject.org/wiki/Releases tells you
Oct 10 13:00:28 <peacememories> thank you
Oct 10 13:12:06 <kanavin> rburton: WARNING: atk-2.26.0-r0 do_configure: QA Issue: atk: configure was passed unrecognised options: --disable-glibtest [unknown-configure-option]
Oct 10 13:12:12 <kanavin> rburton: WARNING: at-spi2-core-2.26.0-r0 do_configure: QA Issue: at-spi2-core: configure was passed unrecognised options: --disable-xevie [unknown-configure-option]
Oct 10 13:12:20 <rburton> kanavin: thanks :)
Oct 10 13:12:25 <kanavin> rburton: I'll send my own version of those patches if you don't mind :)
Oct 10 13:12:33 <rburton> please
Oct 10 13:12:45 <kanavin> rburton: do fujitsu even pay attention to these things? :/
Oct 10 13:13:25 <rburton> distros can turn those off (or more likely their distro has never turned it on)
Oct 10 13:14:06 <kanavin> rburton: to be honest another reason I'm not dropping my patches is that it makes the patch count a nice round 50 :-)
Oct 10 13:14:13 <rburton> haha
Oct 10 13:14:35 <kanavin> just running world on it, followed by sanity checking core-image-sato
Oct 10 13:16:45 <open-nandra> hi again, another issue when bumping to rocko: subprocess.CalledProcessError: Command '['rm', '-rf', '/path/to/recipe/../'] returned non-zero exit status 1
Oct 10 13:17:22 <rburton> open-nandra: full stack?
Oct 10 13:19:03 <open-nandra> rburton: ERROR: myrecipe-3120-g2a9c579-r14 do_unpack: Build of do_unpack failed
Oct 10 13:19:03 <open-nandra> ERROR: myrecipe-3120-g2a9c579-r14 do_unpack: Traceback (most recent call last):
Oct 10 13:19:04 <open-nandra>   File "/home/marek/projects/yocto/build/poky/bitbake/lib/bb/build.py", line 646, in exec_task
Oct 10 13:19:04 <open-nandra>     return _exec_task(fn, task, d, quieterr)
Oct 10 13:19:04 <open-nandra>   File "/home/marek/projects/yocto/build/poky/bitbake/lib/bb/build.py", line 584, in _exec_task
Oct 10 13:19:04 <open-nandra>     exec_func(task, localdata)
Oct 10 13:19:06 <open-nandra>   File "/home/marek/projects/yocto/build/poky/bitbake/lib/bb/build.py", line 203, in exec_func
Oct 10 13:19:08 <open-nandra>     bb.utils.remove(cdir, True)
Oct 10 13:19:10 <open-nandra>   File "/home/marek/projects/yocto/build/poky/bitbake/lib/bb/utils.py", line 683, in remove
Oct 10 13:19:12 <open-nandra>     subprocess.check_call(['rm', '-rf'] + glob.glob(path))
Oct 10 13:19:14 <open-nandra>   File "/usr/lib/python3.5/subprocess.py", line 581, in check_call
Oct 10 13:19:16 <open-nandra>     raise CalledProcessError(retcode, cmd)
Oct 10 13:22:47 <open-nandra> rburton: I've added smalla debug to build.py so lines are shifted by 1
Oct 10 13:31:24 <peacememories> do i need to put nonremovable layers into both BBLAYERS and BBLAYERS_nonremovable?
Oct 10 13:42:04 <open-nandra> rburton: maybe I need to remove do_unpack because it probably refuse to delete path when project is
Oct 10 13:42:35 <open-nandra> I tried to set do_unpack[noexec] = "1"
Oct 10 13:42:40 <open-nandra> but it doesn't help
Oct 10 17:10:13 <khem> https://www.ostechnix.com/convert-deb-packages-arch-linux-packages/
Oct 10 17:12:54 <kergoth> huh
Oct 10 19:17:24 <khem> Dont know how arch community will like it
Oct 10 19:17:35 <khem> since arch's principle is KISS
Oct 10 20:52:05 <Tartarus> bitbake question
Oct 10 20:52:37 <Tartarus> Is there a better way than bb.utils.contains_any('VAR', ['foo', 'foo.xz', 'foo.gz'], ...) ?
Oct 10 20:53:10 <Tartarus> ie I just want to know if foo.* is in VAR
Oct 10 21:12:21 <kergoth> Tartarus: it'd be easy to do programmatically, but you'd be doing so manually, and bypassing use of contains* would result in the checksum depending on the entire value of the variable, rather than just the presence/absence of those values..
Oct 10 21:12:26 <kergoth> so there's a tradeoff there
Oct 10 21:12:51 <Tartarus> kergoth: So there's not a util function that does what I'm thinking of already?
Oct 10 21:14:00 <kergoth> not that i know of. even if there was, codeparser only knows about getvar/appendvar/prependvar, getvarflag/appendvarflag/prependvarflag, contains/contains_any, and exec_func/exec_task
Oct 10 21:14:16 <kergoth> if you're not opposed to depending on the entire variable, any(i.startswith('foo.') for i in d.getVar('VAR').split()) would do
Oct 10 21:14:34 <kergoth> of course, use of vardepvalue could help with that, depending on wha tyou need to do
Oct 10 21:17:01 <Tartarus> kergoth: OK, thanks
Oct 10 21:17:55 <Tartarus> AGL does additional things for vmdk images vs non-vmdk images.  Today it's just checking for vmdk and vmdk.xz (I added the latter to pyro only for them), but with rocko and later it's wic.vmdk.<anything>
Oct 10 21:29:13 <kergoth> checking for the existence of particular image tpyes and/or compression types seems like a pretty common thing to do, i had to do it with wic to handle the dependency on the wks file checksum
Oct 10 21:29:21 <kergoth> maybe we should write a helper for it that properly parses the fstypes out
Oct 10 21:29:34 <kergoth> rather than the semi-naïve text processing we're doing
Oct 10 22:20:21 <tyler-baker> In one of my layers I'm trying to clone a git repo via https which is private
Oct 10 22:20:42 <tyler-baker> my .netrc works when using git or repo directly, but fails when I run bitbake
Oct 10 22:21:14 <tyler-baker> I see on the mailing list most people suggest to use ssh instead of https, is this because private https clones are not supported by the git fetcher?
Oct 10 22:23:48 <rburton> the default protocol if you do a git: clone is git:// did you tell it to do http?
Oct 10 22:24:53 <tyler-baker> rburton, here is my example: SRC_URI = "git://github.com/private/linux.git;protocol=https;branch=linux-v4.11.y-dev;name=kernel
Oct 10 22:25:39 <tyler-baker> I've also tried SRC_URI = "git://github.com/private/linux.git;protocol=https;user=username:password;branch=linux-v4.11.y-dev;name=kernel hoping that username:password would come from the .netrc
Oct 10 22:25:52 <tyler-baker> but it is not evaluated like that at runtime
Oct 10 22:27:03 <rburton> i'd say look at the fetch log to see what its running and if the error is helpful
Oct 10 22:27:14 <tyler-baker> k let take a peek
Oct 10 22:37:54 <Crofton|work> I'm pretty sure .netrc works for me with github
Oct 11 01:53:15 <kergoth> tyler-baker: i use .netrc for https fetching with bitbake on a regular basis, for what it's worth
Oct 11 01:55:54 <tyler-baker> kergoth, thanks, I was able to get it working. It was an error in our build env, the .netrc was being deployed for the incorrect user.
Oct 11 01:56:06 <tyler-baker> all is good now
Oct 11 01:56:13 <tyler-baker> thanks for the help!
Oct 11 01:58:12 <kergoth> ah, good to hear, np
**** ENDING LOGGING AT Wed Oct 11 03:00:02 2017