**** BEGIN LOGGING AT Thu May 14 02:59:57 2020
May 14 05:54:20 <davidlt> build tools use poky sdk, but that's missing openmp, e.g. rpm-native: parseSpec.c:10:10: fatal error: omp.h: No such file or directory
May 14 06:47:36 <khem> davidlt: are you using clang ?
May 14 06:48:04 <davidlt> khem, no
May 14 06:48:27 <davidlt> khem, for now I just disable openmp for rpm-native
May 14 06:48:53 <khem> it needs libgomp in gcc-runtime which should have worked 
May 14 06:48:57 <davidlt> khem, but I am surprised that opemp is not part of poky sdk (i.e. buildtools) // I create Yocto ticket for it
May 14 06:49:09 <khem> if you are using clang then you need libomp from llvm 
May 14 06:49:10 <davidlt> khem, that's not in poky sdk
May 14 06:50:32 <khem> maybe you need https://git.openembedded.org/openembedded-core/commit/?id=319ae973740bc291cb3e02e3cb2b4df2109df53e
May 14 06:51:56 <davidlt> khem, yes, when new build tools with it should be available?
May 14 06:55:13 <khem> its already in dunfell so perhaps when there is point release for 3.1 
May 14 06:55:42 <khem> you can build your own too if need is quick 
May 14 06:56:00 <khem> checkout dunfell branch and bitbake buildtools-extended-tarball
May 14 06:56:47 <davidlt> khem, that's what I am considering now
May 14 07:04:26 <khem> sounds good 
May 14 15:26:14 <Plinnie> What is the correct way to specify a source URI for self-osted gitlab using a private key? Last attempt: SRC_URI = "gitsm://gitlab.nikhef.nl/spidr4/controlboard_control.git;protocol=ssh;branch=vvb_plx_integ;user=git" According to gitlab: git@gitlab.nikhef.nl:spidr4/controlboard_control.git 
May 14 15:28:35 <Plinnie> it says "ERROR: control-1.0-r0 do_fetch: The URL: 'gitsm:spidr4/spidr4_shared.git;protocol=git@gitlab.nikhef.nl;name=shared;bareclone=1;nocheckout=1;nobranch=1' is invalid and cannot be interpreted " which is not at all the URI I used
May 14 15:29:44 <Jin^eLD> Plinnie: just add the public key from your building machine account to github, i.e. outside of Yocto/OE?
May 14 15:30:12 <Jin^eLD> oh you mean the format
May 14 15:30:25 <Plinnie> Yeah, I have no issues cloning the repo outside of the build
May 14 15:30:44 <Plinnie> all keys are Ok, but somehow it gets mangled 
May 14 15:30:54 <Jin^eLD> I think you are missing git@github
May 14 15:31:06 <Jin^eLD> gitsm://git@gitlab.. ?
May 14 15:31:20 <Plinnie> Yeah, tried that too... 
May 14 15:32:13 <Plinnie> bitbake manual isn't very specific on these things
May 14 15:32:15 <Jin^eLD> SRC_URI = "gitsm://git@git.somegitserver.org:port/project/repo.git;protocol=ssh; is what I have here
May 14 15:32:51 <Jin^eLD> why bareclone?
May 14 15:33:05 <Jin^eLD> oh thats from the log
May 14 15:33:08 <Jin^eLD> sorry :)
May 14 15:33:14 <Jin^eLD> I am confused today hehe
May 14 15:33:42 <Plinnie> yeah, the src URI seems to be Ok, but somehow it gets weird
May 14 15:33:54 <Plinnie> WARNING: control-1.0-r0 do_fetch: Failed to fetch URL gitsm://git@gitlab.nikhef.nl/spidr4/controlboard_control.git;protocol=ssh;branch=vvb_plx_integ, attempting MIRRORS if availableERROR: control-1.0-r0 do_fetch: The URL: 'gitsm:spidr4/spidr4_shared.git;protocol=git@gitlab.nikhef.nl;name=shared;bareclone=1;nocheckout=1;nobranch=1' is invalid and
May 14 15:33:54 <Plinnie> cannot be interpreted
May 14 15:34:19 <Plinnie> I mean, why does the protocol part become the user+host?
May 14 15:34:26 <Plinnie> The top is the actual URL I'm using
May 14 15:35:12 <Plinnie> maybe its the submodule fetcher which causes this?
May 14 15:36:19 <Jin^eLD> really don't know, the line I pasted works fine for mit with gitsm... are you sure you have submodules in that repo at all, maybe it gets confused if it does not see any?
May 14 15:37:12 <Plinnie> yeah, I'm sure. Let me try it w/o the gitsm
May 14 15:38:12 <Plinnie> (using plain git fetcher)
May 14 15:40:24 <Plinnie> ok, that works fine
May 14 15:40:35 <Plinnie> except that the build then fails, of course :-S
May 14 15:41:18 <Jin^eLD> :)
May 14 15:42:23 <Plinnie> Its strange that bitbake has so much trouble with submodules. It would simply be a matter of initializing them after the fetch. 
May 14 15:42:52 <Jin^eLD> who doesnt have troubles with submodules? ;))) 
May 14 15:43:08 <Plinnie> :*-)
May 14 15:43:54 <Plinnie> Well, if there are no quick solutions I could always use CMake's fetchproject function, but I was kinda hoping it would work out of the box
May 14 15:44:23 <Jin^eLD> but its indeed wierd, imho SRC_URI="gitsm://git@gitlab.nikhef.nl:spidr4/controlboard_control.git;protocol=ssh;branch=vvb_plx_integ" should have worked
May 14 15:44:37 <Jin^eLD> and that branch does exist?
May 14 15:45:07 <Jin^eLD> wait
May 14 15:45:11 <Plinnie> yes. it just seems to rewrite it
May 14 15:45:12 <Jin^eLD> I asted your line
May 14 15:45:18 <Jin^eLD> pasted your line
May 14 15:45:20 <paulbarker> git submodule URIs get converted and handed to the git fetcher, bitbake doesn't run `git submodule update` directly
May 14 15:45:22 <Jin^eLD> and then noticed a mistake
May 14 15:45:35 <Jin^eLD> gitsm://git@gitlab.nikhef.nl:spidr4/controlboard_control.git see the colon?
May 14 15:45:38 <Jin^eLD> thats where port would be expected
May 14 15:45:47 <Jin^eLD> replace the colon by /
May 14 15:45:50 <Jin^eLD> if you are on a standard port
May 14 15:45:53 <Plinnie> ah, yes, it is the gitlab url format
May 14 15:46:01 <Plinnie> yeah, plain simple ssh
May 14 15:46:01 <Jin^eLD> i.e.: gitsm://git@gitlab.nikhef.nl/spidr4....
May 14 15:46:27 <Jin^eLD> was that it?
May 14 15:46:29 <Plinnie> wait, that is actually not what is in my recipe
May 14 15:46:57 <paulbarker> Plinnie: What version of OE & bitbake are you using?
May 14 15:46:57 <Plinnie> This is in my recipe: SRC_URI = "gitsm://git@gitlab.nikhef.nl/spidr4/controlboard_control.git;protocol=ssh;branch=vvb_plx_integ"
May 14 15:47:07 <Jin^eLD> dman, that would have been easy
May 14 15:47:10 <Jin^eLD> *damn
May 14 15:48:54 <Plinnie> I'm actually using Petalinux 2019.1 , which uses Yocto 2.6.1 Thud . Not sure which versions of OE and bitbake are used
May 14 15:49:35 <paulbarker> Plinnie: Ok, that's not too old I guess. What's the submodule URL in .gitmodules?
May 14 15:50:53 <Plinnie> I have multiple
May 14 15:51:11 <Plinnie> [submodule "thirdparty/easyloggingpp"]        path = thirdparty/easyloggingpp        url = https://github.com/amrayn/easyloggingpp.git[submodule "thirdparty/simpleini"]        path = thirdparty/simpleini        url = https://github.com/brofield/simpleini[submodule "thirdparty/csv-parser"]        path = thirdparty/csv-parser        url =
May 14 15:51:12 <Plinnie> https://github.com/ben-strasser/fast-cpp-csv-parser[submodule "shared"]        path = shared        url = git@gitlab.nikhef.nl:spidr4/spidr4_shared.git
May 14 15:51:16 <Plinnie> Oh, that is messed up
May 14 15:51:37 <Jin^eLD> paulbarker: gitsm worked for us even with 2.0.1, so 2.6 should work I think...
May 14 15:54:22 <Plinnie> ok, I'm actually getting further now. I don't understand, because afaik I didn't change anything. Now it cplains about the last submodule
May 14 15:55:00 <Plinnie> Oh wait, no never mind. I forgot to clean the recipe
May 14 15:55:14 <paulbarker> Plinnie: Check if your bitbake repo has this commit: https://git.openembedded.org/bitbake/commit/lib/bb/fetch2/gitsm.py?id=dcac05e7dc6b0c5f8e63d36ad105b7eab41f0016
May 14 15:56:07 <Plinnie> one moment, need to locate bb in the install
May 14 15:56:43 <paulbarker> Plinnie: Not sure how PetaLinux pulls in the various Yocto components so not sure where to look
May 14 15:57:34 <Plinnie> No, and it is quite difficult to actually find the location of things, as its all tucked away in higher level scripts. Even if you source petalinux, bitbake is actually not on the path
May 14 15:57:38 <paulbarker> But that commit adds support for the SSH format `git@host:path/to/repo`. It wasn't in the initial thud release but looks like it was added to the stable branch after release to fix this exact issue
May 14 15:58:04 <Plinnie> ls
May 14 15:58:13 <Plinnie> sorry, wrong screen :-)
May 14 15:58:28 <paulbarker> You probably need to update bitbake to the latest commit in the 1.40 branch (or update poky to the latest commit on the thud branch if it's using the combined poky repo)
May 14 16:00:55 <Plinnie> hmm.. darn those vendor packed distributions!
May 14 16:03:04 <paulbarker> Plinnie: I usually rip them apart, determine what commits of what repositories they're pulling and try to check it all out locally myself. If that doesn't work I shout at the vendor
May 14 16:03:32 <Plinnie> I found bitbake
May 14 16:05:11 <Plinnie> No, it doesn't have that commit
May 14 16:06:06 <Plinnie> *sigh* I mean the petalinux tools are essential for getting the right devicetree and all. I'm afraid I'd open a can of worms when I try to pull other things
May 14 16:07:13 <Plinnie> I'll think about what to do. I could also upgrade to  a newer petalinux, or not use submodules. Thanks for your assistence paulbarker and Jin^eLD 
May 14 16:07:40 <Plinnie> assistance*
May 14 21:19:18 <Crofton|cloud> bluelightning: ping
May 14 21:35:47 <bluelightning> hi Crofton|cloud
May 14 21:36:15 <Crofton[m]> hey, working on scheduling an OE happy hour
May 14 21:36:51 <Crofton[m]> do any mornings not work for you?
May 14 21:37:04 <Crofton[m]> suggested time is 0900 for you
May 14 21:37:28 <bluelightning> let me double-check
May 14 21:38:22 <bluelightning> generally any day is good except US Friday which is my Saturday
May 14 21:38:36 <bluelightning> and my Monday/Tuesday are out next week, but that's unusual
May 14 21:39:28 <Crofton[m]> THinking about Wed 27, which would be Thu am for you
May 14 21:40:29 <bluelightning> that should be fine
May 14 21:40:47 <Crofton[m]> ok thanks
May 14 21:41:26 <Crofton[m]> Be good to have at least one person from that part of the world to validate not being .eu/.us centric :)
May 14 21:41:36 <bluelightning> thank you for scheduling it at a convenient time for me :)
May 14 21:42:09 <Crofton[m]> Also good for Steve in Hawaii
May 14 21:42:50 <bluelightning> ah yes of course
May 14 21:45:40 <Crofton[m]> and armpit 
May 14 22:41:20 <khem> Friday should be out of question 
May 14 22:41:22 <armpit> I am not in Hawaii.
May 14 22:42:57 <armpit> when did volunteers have  any issue with days-of-the-week ?
May 14 22:44:24 <Crofton[m]> Heh, I just figured you would appreciate something not early in the morning
May 14 22:46:35 <kergoth> I'll have to see if i can make the next one, missed the happy hour last time
May 14 22:46:49 <armpit> I was waiting for the grumpy hour
May 14 22:47:53 <khem> armpit: people might have something else planned I am usually OK oon Friday
May 14 22:50:21 <khem> 7am works for me, I can make it green tea hour
May 14 22:50:38 <armpit> planned like drinking ; )
May 14 22:51:56 <khem> or 5pm again second green tea hr
May 14 22:59:03 <armpit> Green tea Martinis 
May 14 23:00:27 <khem> :)
**** ENDING LOGGING AT Fri May 15 02:59:57 2020