Tuesday, 2018-09-04

*** rburton <rburton!~textual@> has quit IRC00:11
*** rburton <rburton!~textual@> has joined #yocto00:12
*** rburton <rburton!~textual@> has quit IRC00:14
*** rburton <rburton!~textual@> has joined #yocto00:15
*** yizhao <yizhao!~zhaoyi@> has quit IRC00:37
*** rburton <rburton!~textual@> has quit IRC00:37
*** rburton <rburton!~textual@> has joined #yocto00:38
*** khem <khem!~khem@unaffiliated/khem> has quit IRC00:40
*** yizhao <yizhao!~zhaoyi@> has joined #yocto00:42
*** khem <khem!~khem@unaffiliated/khem> has joined #yocto00:52
*** khem <khem!~khem@unaffiliated/khem> has quit IRC00:56
*** khem <khem!~khem@unaffiliated/khem> has joined #yocto00:56
*** jkridner|pd <jkridner|pd!~jkridner@pdpc/supporter/active/jkridner> has joined #yocto01:12
*** jkridner_ <jkridner_!~jkridner@pdpc/supporter/active/jkridner> has joined #yocto01:14
*** jkridner <jkridner!~jkridner@pdpc/supporter/active/jkridner> has quit IRC01:15
*** jkridner|pd <jkridner|pd!~jkridner@pdpc/supporter/active/jkridner> has quit IRC01:17
*** sgw <sgw!~sgw@> has quit IRC02:08
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has quit IRC02:19
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has joined #yocto02:19
*** kaspter <kaspter!~Instantbi@> has quit IRC02:32
*** kaspter <kaspter!~Instantbi@> has joined #yocto02:33
*** arielmr <arielmr!~quassel@187-163-217-93.static.axtel.net> has quit IRC03:04
*** aratiu <aratiu!~adi@> has quit IRC03:09
*** aratiu <aratiu!~adi@> has joined #yocto03:17
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has quit IRC03:22
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has joined #yocto03:42
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has quit IRC03:51
*** rburton <rburton!~textual@> has quit IRC04:13
*** rburton <rburton!~textual@> has joined #yocto04:14
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has joined #yocto04:19
*** seebs <seebs!~seebs@> has quit IRC04:20
*** seebs <seebs!~seebs@> has joined #yocto04:36
*** sgw <sgw!sgw@nat/intel/x-jipuqwaigrzhagvu> has joined #yocto04:48
*** kanavin_ <kanavin_!~kanavin@> has quit IRC04:58
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has quit IRC05:01
*** agust <agust!~agust@p508862E3.dip0.t-ipconnect.de> has joined #yocto05:40
*** ka6sox <ka6sox!ka6sox@nasadmin/ka6sox> has quit IRC05:48
*** rperier <rperier!~rperier@unaffiliated/bambee> has quit IRC05:49
*** shauno <shauno!~soneil@pdpc/supporter/professional/shauno> has quit IRC05:50
*** shauno <shauno!~soneil@pdpc/supporter/professional/shauno> has joined #yocto05:50
*** ka6sox <ka6sox!ka6sox@nasadmin/ka6sox> has joined #yocto05:52
*** olani <olani!~user@proxy02.se.axis.com> has joined #yocto05:59
*** pohly <pohly!~pohly@p54BD55A3.dip0.t-ipconnect.de> has joined #yocto06:03
*** frsc <frsc!~frsc@2003:a:e7a:6200:246c:2a8b:f45a:a33d> has joined #yocto06:06
*** rob_w <rob_w!~bob@unaffiliated/rob-w/x-1112029> has joined #yocto06:11
*** rperier <rperier!~rperier@unaffiliated/bambee> has joined #yocto06:46
*** fl0v0 <fl0v0!~fvo@i577B9FB1.versanet.de> has joined #yocto06:50
*** runde <runde!sid228344@gateway/web/irccloud.com/x-cszwfqlosahbqvgt> has joined #yocto06:50
*** T_UNIX <T_UNIX!uid218288@gateway/web/irccloud.com/x-qrtvqrmgdfjikdnp> has joined #yocto06:52
*** varjag <varjag!~user@122.62-97-226.bkkb.no> has joined #yocto07:12
*** xtron <xtron!~mentor@> has quit IRC07:18
mcfriskhmm, any ideas what could be wrong in my locale/libc config: nothing provides glibc-gconv-cp125207:37
mcfriskfor nativesdk it is compiled but not for target07:37
ak77why would cmake build ninja make files? I inherit cmake, and I can see it generates ninja files07:38
ak77Ok, i was looking at older cmake.bbclass.07:40
ak77there's a OECMAKE_GENERATOR variable now07:40
mcfriskak77: it's the default07:44
ak77mcfrisk: ok, but why is then bitbake trying to run make after configure07:44
*** xtron <xtron!~mentor@> has joined #yocto07:48
mcfriskak77: cmake.bbclass doesn't do that, maybe it's the recipe trying to call oemake07:48
ak77mcfrisk: can you take a look https://pastebin.com/zxErUDuZ07:50
ak77mcfrisk: this project provides CMakeLists.txt and Makefile (maybe that's the culprit) (PN=lua-cjson)07:53
*** joaocfernandes <joaocfernandes!~Joao@> has joined #yocto08:01
*** Bunio_FH <Bunio_FH!~bunio@clj-165.netdrive.pl> has joined #yocto08:03
*** boucman_work <boucman_work!~jrosen@wesnoth/developer/boucman> has quit IRC08:04
*** Bunio_FH <Bunio_FH!~bunio@clj-165.netdrive.pl> has quit IRC08:08
*** mns__ <mns__!4fab95ac@gateway/web/freenode/ip.> has joined #yocto08:09
*** mns__ <mns__!4fab95ac@gateway/web/freenode/ip.> has left #yocto08:10
*** mns__ <mns__!4fab95ac@gateway/web/freenode/ip.> has joined #yocto08:18
*** Bunio_FH <Bunio_FH!~bunio@81-18-201-214.static.chello.pl> has joined #yocto08:19
derRichardwhen i have meta-freescale and meta-browser in my project, bitbake is unhappy because of:08:20
derRichardfor now i set BB_DANGLINGAPPENDS_WARNONLY to 1, but is this really the right way?08:20
derRichardto me it looks like these two layers depend on each other but are out of sync08:21
derRichardup, the error/warning is:08:21
derRichardNo recipes available for:08:21
*** boucman_work <boucman_work!~jrosen@wesnoth/developer/boucman> has joined #yocto08:22
*** florian <florian!~florian_k@Maemo/community/contributor/florian> has joined #yocto08:25
*** ndec <ndec!sid219321@linaro/ndec> has joined #yocto08:35
*** dv_ <dv_!~dv@62-178-50-190.cable.dynamic.surfer.at> has quit IRC08:36
*** boucman_work <boucman_work!~jrosen@wesnoth/developer/boucman> has quit IRC08:40
*** boucman_work <boucman_work!~jrosen@wesnoth/developer/boucman> has joined #yocto08:40
*** kaspter <kaspter!~Instantbi@> has quit IRC08:44
*** kaspter <kaspter!~Instantbi@> has joined #yocto08:45
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has joined #yocto08:45
*** dv_ <dv_!~dv@62-178-50-190.cable.dynamic.surfer.at> has joined #yocto08:51
*** gtristan <gtristan!~tristanva@> has joined #yocto08:57
*** awe00 <awe00!~awe00@unaffiliated/awe00> has joined #yocto09:00
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has quit IRC09:02
*** boucman_work <boucman_work!~jrosen@wesnoth/developer/boucman> has quit IRC09:04
*** boucman_work <boucman_work!~jrosen@wesnoth/developer/boucman> has joined #yocto09:19
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has joined #yocto10:10
*** morphis <morphis!~morphis@p200300CCFBEB6E006D44A5210733229F.dip0.t-ipconnect.de> has joined #yocto10:15
*** kanavin <kanavin!~kanavin@> has joined #yocto10:20
*** rob_w <rob_w!~bob@unaffiliated/rob-w/x-1112029> has quit IRC10:41
*** mns__ <mns__!4fab95ac@gateway/web/freenode/ip.> has quit IRC10:54
*** bluelightning_ <bluelightning_!~paul@pdpc/supporter/professional/bluelightning> has joined #yocto11:01
*** bluelightning <bluelightning!~paul@pdpc/supporter/professional/bluelightning> has quit IRC11:04
*** bluelightning <bluelightning!~paul@pdpc/supporter/professional/bluelightning> has joined #yocto11:05
*** bluelightning_ <bluelightning_!~paul@pdpc/supporter/professional/bluelightning> has quit IRC11:06
*** RyanMeulenkamp <RyanMeulenkamp!~ryan.meul@lorentz.bad-bit.net> has joined #yocto11:19
*** rob_w <rob_w!~rob@unaffiliated/rob-w/x-1112029> has joined #yocto11:37
*** eduardas_m <eduardas_m!~eduardas@> has joined #yocto11:42
*** BCMM <BCMM!~BCMM@unaffiliated/bcmm> has joined #yocto11:42
eduardas_mhello, I get a urllib.error.URLError: <urlopen error [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed (_ssl.c:719) when trying to do_install ntpath for python311:43
eduardas_mcan certificate checking be disabled for a recipe that inherits both pypi and setuptools3?11:44
rburtona better question is why is do_install trying to fetch?11:58
derRichardand before disabling a ssl check, figure why is it failing. nsa? ;)12:04
kanavinRP: openssl 1.1.1 merging might have been a bit premature, as we now have a 1.1.1-pre9 recipe in master, and 1.1.1 final will have PV that's smaller12:08
kanavinthat said, I am glad you took it in :)12:09
RPkanavin: hmm, I hadn't realised the version issue. We might want to tweak that to 1.1.1~pre912:09
kanavinRP: will that sort to less than 1.1.1?12:10
RPkanavin: its supposed to12:10
kanavinoh - then we definitely should12:10
kanavinI can prepare the commit?12:10
RPkanavin: please12:10
*** agust <agust!~agust@p508862E3.dip0.t-ipconnect.de> has quit IRC12:41
nayfeHi, I'm going crazy with cmake, is there any good reference for a cmake executable depending on other cmake lib? When I set "target_include_directories(${TARGET_NAME} PUBLIC $<INSTALL_INTERFACE:include/${TARGET_NAME}>)" I get an absolute dir in INTERFACE_INCLUDE_DIRECTORIES then, with sysroot by recipe in yocto, my executable won't compile ...12:44
*** BCMM <BCMM!~BCMM@unaffiliated/bcmm> has quit IRC12:50
*** bluelightning <bluelightning!~paul@pdpc/supporter/professional/bluelightning> has quit IRC13:15
mcfriskqa check for FILE_bla defined but does not exists in PACKAGES would be nice :)13:24
eduardas_meven though the pathlib module is available in Python since 3.4, it does not seem to be present in Yocto 2.5 Sumo (Python 3.5.5). Why is that so?13:35
eduardas_mI check for this using help('modules') in Python prompt13:36
eduardas_mhow do I make the pathlib module available on my system? is this some kind of python3-core build option?13:38
*** BCMM <BCMM!~BCMM@unaffiliated/bcmm> has joined #yocto13:42
*** JPEW <JPEW!cc4da337@gateway/web/freenode/ip.> has joined #yocto13:42
yoctiNew news from stackoverflow: Yocto - "Nothing provides websocketpp = 0.7.0-r0 needed by websocketpp-dev-0.7.0-r0" <https://stackoverflow.com/questions/52167777/yocto-nothing-provides-websocketpp-0-7-0-r0-needed-by-websocketpp-dev-0-7-0>13:48
RPJPEW: looks like I'm chasing a bug similar to you fd on fork problem :/13:50
JPEWRP: Oh? As part of what?13:50
RPJPEW: I was poking around the issue on the bitbake list, "process: Flush server stdout/stderr before reporting failure"13:51
RPJPEW: There is this idea of the "readypipe" in server/process.py. Closing the pipe isn't enough to trigger the client to realise the server failed13:52
RPJPEW: my theory is the write pipe is open for writing in multiple processes in the server :(13:52
JPEWRP: Ya, that would make sense.... are they all just fork() or fork() + exec() ?13:53
JPEWRP: If fork() + exec() O_CLOEXEC should fix it13:53
RPJPEW: I don't know which process specifically as yet13:54
JPEWHmm, it might not be bad practice for the server to mark all the non-stdio FD's as O_CLOEXEC anyway. I can't imagine anything else would need to be passed to a child process....13:57
*** agust <agust!~agust@p508862E3.dip0.t-ipconnect.de> has joined #yocto14:02
*** ant_work <ant_work!~ant__@host63-251-dynamic.54-79-r.retail.telecomitalia.it> has joined #yocto14:07
RPJPEW: O_CLOEXEC didn't help so its probably some fork call :/14:07
*** ant_work <ant_work!~ant__@host63-251-dynamic.54-79-r.retail.telecomitalia.it> has quit IRC14:08
*** ant_work <ant_work!~ant__@host63-251-dynamic.54-79-r.retail.telecomitalia.it> has joined #yocto14:09
*** BCMM <BCMM!~BCMM@unaffiliated/bcmm> has quit IRC14:11
RPJPEW: I think there may be a different set of exception handling code catching and retrying the server start without me noticing :/14:12
*** marka <marka!~masselst@> has joined #yocto14:18
*** [RLA]Eske <[RLA]Eske!c0e1ba62@gateway/web/freenode/ip.> has joined #yocto14:18
*** agust <agust!~agust@p508862E3.dip0.t-ipconnect.de> has quit IRC14:20
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has quit IRC14:25
RPJPEW: There is a problem with http://git.yoctoproject.org/cgit.cgi/poky/commit/?id=cbe2d3cb0af6c7a5cd368e7dc489960a13648bd0 -  bb.event.print_ui_queue() trashes stdout/stderr14:25
JPEWRP: And it was never seen before because the buffered stdout/stderr was lost?14:27
RPJPEW: I don't think we've realised how much data was getting lost, we've patched things up now and again but not realised how many ways this is breaking :/14:28
RPI think a step back and rethink may be needed for that function in server context14:28
RPJPEW: certainly that function needs to flush out/err before it changes them14:28
*** armpit <armpit!~armpit@2601:202:4180:c33:ed90:f8dd:8dd6:4edd> has quit IRC14:29
*** maxin <maxin!~maxin@> has quit IRC14:30
JPEWRP: Ya, that makes sense. There is also another caveat I discovered: stdio and stderr in the server are now fully buffered instead of line buffered (as one would normally expect).14:31
JPEWRP: Actually, they were always fully buffered. I have an (unsubmitted) patch to make the line buffered so that they act like you would expect14:32
*** BCMM <BCMM!~BCMM@unaffiliated/bcmm> has joined #yocto14:32
*** rburton <rburton!~textual@> has quit IRC14:37
mcfrisksigh, got bitten several times today by _append vs += differences..14:37
*** rburton <rburton!~textual@> has joined #yocto14:38
mcfriskeven in https://www.yoctoproject.org/docs/latest/mega-manual/mega-manual.html#adding-ptest-to-your-build it's not clear why using DISTRO_FEATURES_append = " ptest" and EXTRA_IMAGE_FEATURES += "ptest-pkgs"14:39
*** armpit <armpit!~armpit@2601:202:4180:c33:45cc:4239:c235:fa42> has joined #yocto14:41
*** stephano <stephano!~stephano@> has joined #yocto14:49
RPJPEW: I've also just realised we have separate buffers for stderr and stdout which makes no sense14:52
*** varjag <varjag!~user@122.62-97-226.bkkb.no> has quit IRC14:52
*** gtristan <gtristan!~tristanva@> has quit IRC14:53
JPEWRP: In print_ui_queue()?14:53
armpitYPTM: Armin in on14:54
*** sjolley_ <sjolley_!86868b4c@gateway/web/freenode/ip.> has joined #yocto14:55
sjolley_YPTM: Join at Zoom https://zoom.us/j/99089271214:56
sjolley_YPTM: Stephen Joined14:56
armpitno you haven't14:56
*** maxin <maxin!~maxin@> has joined #yocto14:57
sjolley_So Zoom is slow14:57
RPJPEW: I think we need something like http://git.yoctoproject.org/cgit.cgi/poky-contrib/commit/?h=rpurdie/t222&id=82f4e1dfd4bdc5869ca5a04649d4871d4db681d314:58
*** AbleBacon <AbleBacon!~AbleBacon@unaffiliated/ablebacon> has joined #yocto14:59
*** JPEW_ <JPEW_!cc4da369@gateway/web/freenode/ip.> has joined #yocto14:59
RPYPTM: Richard joined14:59
*** RyanMeulenkamp <RyanMeulenkamp!~ryan.meul@lorentz.bad-bit.net> has quit IRC15:00
*** smurray <smurray!sid98062@gateway/web/irccloud.com/x-pydldanqypldsczt> has joined #yocto15:00
zeddiiYPTM: Bruce has joined15:00
rburtonYPTM ross joined15:00
tgoodwinOther than patching it, is there any support for extending WIC's Partition class to support other file system types or modify its fstab behavior?15:00
JPEW_YPTM: Joshua Watt has joined15:01
tlwoernerYPTM: trevor has joined15:01
vmesonYPTM: Randy joined15:05
denixYPTM: Denys joined15:05
*** arielmr <arielmr!~quassel@187-163-217-93.static.axtel.net> has joined #yocto15:06
*** dc13ff <dc13ff!uid190567@gateway/web/irccloud.com/x-xqtlhbzbwuipgnua> has joined #yocto15:06
*** frsc <frsc!~frsc@2003:a:e7a:6200:246c:2a8b:f45a:a33d> has quit IRC15:07
vmesonRP:  your audio has dropped15:08
*** thaytan <thaytan!~thaytan@121-200-23-18.cust.aussiebb.net> has quit IRC15:10
*** gtristan <gtristan!~tristanva@> has joined #yocto15:16
*** BCMM <BCMM!~BCMM@unaffiliated/bcmm> has quit IRC15:17
vmesontopics: python profile optimized build: OFF by default unless there's proof of improved performance or used canned data.15:19
*** BCMM <BCMM!~BCMM@unaffiliated/bcmm> has joined #yocto15:20
*** eduardas_m <eduardas_m!~eduardas@> has quit IRC15:23
*** jacques <jacques!~jacques@nslu2-linux/jacques> has joined #yocto15:26
*** jacques is now known as linuxjacques15:26
*** thaytan <thaytan!~thaytan@121-200-23-18.cust.aussiebb.net> has joined #yocto15:28
*** tgoodwin <tgoodwin!~tgoodwin@static-108-40-78-74.bltmmd.fios.verizon.net> has quit IRC15:31
*** chandana73 <chandana73!~ckalluri@> has joined #yocto15:34
*** sgw <sgw!sgw@nat/intel/x-jipuqwaigrzhagvu> has quit IRC15:39
*** chandana73 <chandana73!~ckalluri@> has quit IRC15:39
vmesonfor those not on the call: the hope is that we can leave the python pgo optimization on. There are 10 runs used to generate the pgo data so we can reduce that to 1 or 2 and do a couple other things and lower the overhead from ~ 20 minutes to ~ 2-4 minutes15:40
*** JPEW_ <JPEW_!cc4da369@gateway/web/freenode/ip.> has quit IRC15:40
*** kuzulis <kuzulis!~Denis@> has joined #yocto15:41
kuzulisHi guys. What is a best way to customize the layers to have a several images? E.g. I want to have a possibility to create an own images like: bitbake image1, bitbake imageN. But it is uncleas for me how to add or remove required features from this images?15:43
kuzulisShould I change the IMAGE_FEATURES in each image.bb file?15:44
kuzulisE.g. to add or remove the X, wayland Qt4/Qt5 and so on15:45
rob_wkuzulis, how about taking the default console-minimal-image and add your whished on top each time ?15:46
*** chandana73 <chandana73!~ckalluri@> has joined #yocto15:46
markakuzulis: there are plenty of examples, take a look at the images we define in meta-overc for example https://github.com/OverC/meta-overc/tree/master-oci/meta-cube/recipes-core/images15:46
markaoverall it is easiest to create your own image files instead of attempting to find ways to manipulate your local.conf to define them (in case that is what you were thinking about doing)15:48
*** sjolley_ <sjolley_!86868b4c@gateway/web/freenode/ip.> has quit IRC15:48
*** varjag <varjag!~user@ti0040a400-6639.bb.online.no> has joined #yocto15:49
kuzulisI don't want to change the local.conf file. I want to have a ready pre-sets of image1.bb, image2.bb, imageN.bb files. And in this files I want to change required features. Is it right way?15:51
markayou can use bitbake and bitbake-layers to create your layer15:51
markathen you are off to the races to define your images and add any recipes which you might need that don't exist15:52
kuzulisSo, is just one inheritable like: "inherit core-image" means that will be assembled the my custom core-image-minimal?15:52
markafor the most part the images are defining the high level contents of the image, the lower level stuff will come in via dependencies15:54
markacore-image will get you base-files and a few other bits that most rootfs will need/want15:55
markaif you get more complex you will want to define a distro as well, but usually that can be avoided15:55
markastart with creating and adding your layer, define your images and go from there15:57
kuzulisis this content of resulting rootfs depends on DISTRO_FEATURES_DEFAULT property?15:57
JPEWRP: That change looks good in general... I would have some general cleanup coments if you don't change them before submitting to the mailing list15:58
*** sgw <sgw!~sgw@> has joined #yocto15:58
markaDISTRO_FEATURES can be used to do things like configure packages differently and stuff such as this (think about Ubuntu, there is one repo for debs, regardless if you are building a small footprint rootfs or a rootfs for big iron the emacs will be configured and built with the same configurations)16:00
JPEWRP: I *think* you want to keep the dup2 of stderr though? Otherwise child processes won't see the FD 2 as the log file.... unless python does some magic when you do sys.stderr sys.stderr?16:00
JPEWRP: Argh, can16:00
JPEW*when you do sys.stderr = sys.stdout16:01
markaDISTRO_FEATURES let you change how emacs is configured, for example, so you can define a set of DISTRO_FEATURES for a small footprint distro and another set of configs for large footprint16:01
markaagain, only worries about tuning these nobs if needed16:02
RPJPEW: with the other dup2, I was seeing logline duplication16:02
markaif you only want packages configured in one way for all your images you can included a bbappend in your layer and adjust the PACKAGECONFIG and be done with it16:02
* marka is glad fray is most likely not watching as he would school me on my abuses of DISTRO_FEATURES16:06
*** florian <florian!~florian_k@Maemo/community/contributor/florian> has quit IRC16:08
kuzulishmm.. I don't understand a bit... e.g. I have created an own my-layer... And then I want to produce the different images: e.g. image-with-X11 and image-without-X11... Should I in this case just add the REQUIRED_DISTRO_FEATURES = "x11" to image-with-X11? And to do not add nothing to image-without-X11? It is not clear to me a tit..16:09
JPEWRP: That..... doens't make sense. It seems like not redirecting stderr with dup2 is only fixing that because the duplicates to stderr would be getting lost....16:11
kuzulisHow I will know that a minimal image will be compiled without of X11?16:11
*** stephano <stephano!~stephano@> has quit IRC16:11
RPJPEW: it does make sense in that there are two buffers, stdout and stderr and if you flush them at different times the lines are reordered16:11
RPJPEW: as you commented earlier, they're not line buffered16:12
markakuzulis: the x11 distro features won't cause any X stuff to be in your image16:15
*** robbawebba <robbawebba!~rob@> has joined #yocto16:17
markait may turn on some PACKAGECONFIGs to enable the graphical component of packages16:18
JPEWRP: Agreed, I've seen the reordering issues (bitbake can time-travel!), I hadn't seen any duplication of log messages. My comment was more that if you don't dup stderr to the output file, anything a fork+exec'd child process prints to that FD might get lost (or end up who knows where).16:18
markaso for example you can see16:18
markameta-oe/recipes-extended/hwloc/hwloc_1.11.10.bb:PACKAGECONFIG ?= "pci libxml2 ${@bb.utils.contains('DISTRO_FEATURES', 'x11', 'x11', '', d)}"16:18
markameta-oe/recipes-extended/hwloc/hwloc_1.11.10.bb:PACKAGECONFIG[x11] = "--with-x,--without-x,virtual/libx11 cairo,cairo"16:19
markaso hwloc will be configured with x support16:20
*** BCMM <BCMM!~BCMM@unaffiliated/bcmm> has quit IRC16:20
markaso again start by defining the image contents, then branch out to other bits as needed16:22
RPJPEW: I guess we can dup, then trample over it :)16:23
*** rubdos <rubdos!~rubdos@ptr-1uzevqefjgxdm5wv65h.18120a2.ip6.access.telenet.be> has joined #yocto16:27
kuzulismarka: Many thanks anyway... But it is soo hard for me as newbie :)16:27
JPEWRP: That would work. I'm still struggling with the fix though.... someone must be outputting logs to both stdout and stderr, so why does making them the same Python object (sys.stderr = sys.stdout) fix it? I feel I'm missing something :)16:28
*** fl0v0 <fl0v0!~fvo@i577B9FB1.versanet.de> has quit IRC16:31
RPJPEW: normal prints to stdout, traceback to strerr was my test case16:35
RPJPEW: the traceback appeared before stdout output16:36
*** tgoodwin <tgoodwin!~tgoodwin@static-108-40-78-74.bltmmd.fios.verizon.net> has joined #yocto16:36
kuzulismarka: I have compiled core-image-minimal, but it contains libX11 library.. Why?16:40
JPEWRP: Ok, I think thats what I was missing. It's not "duplication", just out of order? In that case, yes, that fix make sense.16:42
*** joaocfernandes <joaocfernandes!~Joao@> has quit IRC16:44
RPJPEW: yes, out of order, not duplication16:45
RPJPEW: sorry, I did say duplication earlier as well as reordering :/16:46
markakuzulis: you can do things like 'bitbake <your-image-name> -g -u depex' or just 'bitbake <your-image-name> -g' to examine the dependencies16:58
markafigure out what is depending on libX11 then work from there, in the easy case you might be able to adjust the PACKAGECONFIG for the recipe that is resulting in this, in the extreme case you might have to look at manipulating distro features or creating distros to fork your x11 vs non-x11 images17:00
*** BCMM <BCMM!~BCMM@unaffiliated/bcmm> has joined #yocto17:01
*** seebs <seebs!~seebs@> has quit IRC17:02
kuzulisok, many thanks17:03
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has joined #yocto17:03
*** dc13ff <dc13ff!uid190567@gateway/web/irccloud.com/x-xqtlhbzbwuipgnua> has quit IRC17:15
JPEWRP: No worries :)17:16
*** kuzulis <kuzulis!~Denis@> has quit IRC17:28
*** linuxjacques <linuxjacques!~jacques@nslu2-linux/jacques> has quit IRC17:28
*** awe00 <awe00!~awe00@unaffiliated/awe00> has quit IRC17:29
*** seebs <seebs!~seebs@> has joined #yocto17:37
*** T_UNIX <T_UNIX!uid218288@gateway/web/irccloud.com/x-qrtvqrmgdfjikdnp> has quit IRC17:40
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has quit IRC17:48
*** stephano <stephano!~stephano@> has joined #yocto17:54
*** BCMM <BCMM!~BCMM@unaffiliated/bcmm> has quit IRC17:54
*** BCMM <BCMM!~BCMM@unaffiliated/bcmm> has joined #yocto17:57
*** kaspter <kaspter!~Instantbi@> has quit IRC17:57
*** kaspter <kaspter!~Instantbi@> has joined #yocto17:58
*** Bunio_FH <Bunio_FH!~bunio@81-18-201-214.static.chello.pl> has quit IRC17:58
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has joined #yocto18:06
*** arielmr <arielmr!~quassel@187-163-217-93.static.axtel.net> has quit IRC18:10
*** likewise <likewise!~leon@ip56501413.direct-adsl.nl> has joined #yocto18:15
likewisegood morning18:15
likewiseI have a systemd service that Requires=systemd-sysusers.service. But how do I indicate to Yocto that the latter should also be enabled? It doesn't seem to parse "Requires=", while it does parse "Also=".18:17
likewise'It' in this context is systemd.bbclass.18:18
kergothsounds like you need to enable the packageconfig for sysusers18:19
likewisekergoth: "My" package is pulseaudio, but running in system mode on a read-only filesystem. I wasn't aware that packageconfig works cross-package. So it can influence the configuration of all dependants?18:22
kergothsystemd.bbclass just calls systemctl. if systemctl enable doesn't do what you want, your service needs fixing18:24
kergothand no, i mentioned packageconfig just to ensure sysusers support is enabled and its service file is enabled by default as a baseline before you go depending on it18:25
likewisekergoth: thanks. Further debugging to find -EUSER.18:28
likewisesystemd's PACKAGECONFIG_CONFARGS contains --enable-sysusers. As I am on a RO FS, I should be looking at systemctl-native being called within the pulseaudio build. I expect it to setup a symlink as a target dependency due to the Requires=18:32
*** jacques <jacques!~jacques@nslu2-linux/jacques> has joined #yocto18:38
*** arielmr <arielmr!~quassel@187-163-217-93.static.axtel.net> has joined #yocto18:45
rburtonvmeson:   python3    do_compile            -940.5s   -74.6%   1260.5s -> 319.9s18:53
rburtonvmeson: thats master vs my changes to use a different benchmark for pgo18:53
*** awe00 <awe00!~awe00@unaffiliated/awe00> has joined #yocto18:57
*** ant_work <ant_work!~ant__@host63-251-dynamic.54-79-r.retail.telecomitalia.it> has quit IRC18:58
khemvmeson: how good is benchmark ?19:03
khemPGO is very subjective and its as good as your test matrix can cover19:03
khemrburton: I think we should offload the profile data creation19:04
khemand use that data until we bump up python verson say 3.7 to 3.8 and so on19:04
khemkanavin: seen this http://errors.yoctoproject.org/Errors/Details/190427/19:06
khemthis is when some recipes are changed to use openssl10 in meta-oe19:07
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has quit IRC19:10
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has joined #yocto19:17
rburtonkhem: its the same benchmark upstream py use19:20
tgoodwinDoes anyone know how to specify an alternate imager plugin in WIC?  I don't see any mention of "DirectPlugin", the default, but it's clearly used.19:27
*** [RLA]Eske <[RLA]Eske!c0e1ba62@gateway/web/freenode/ip.> has quit IRC19:28
tgoodwinNever mind.  It seems hard-coded in rocko via the engine to only ever pull in the direct file.19:33
khemrburton: that sounds promising19:33
kergothtgoodwin: lines 194 to 202 in engine.py19:33
kergothnot ideal19:34
kergothtgoodwin: i'm a little surprised direct is hardcoded like that, not typical for how we usually do things (i.e. maximize flexibility)19:39
tgoodwinkergoth: yep...not really a hard patch by any stretch.  All I'm trying to do is swap in cryptsetup instead of parted so that the built partition table for the direct file can have some LUKs partitions.19:39
*** georgem_home <georgem_home!uid210681@gateway/web/irccloud.com/x-payouljunsecdthq> has joined #yocto19:42
khemsend a patch to make it param I guess19:43
*** georgem <georgem!~georgem@> has quit IRC19:51
*** georgem <georgem!~georgem@> has joined #yocto19:54
*** bluelightning <bluelightning!~paul@pdpc/supporter/professional/bluelightning> has joined #yocto19:59
*** arielmr <arielmr!~quassel@187-163-217-93.static.axtel.net> has quit IRC20:01
tgoodwinkhem: debating it over on this end since even if we had this as an option, we would end up replicating most of DirectPlugin by copy-paste for our own nuanced need (since it's not a module that can be imported).  May just be better to define a different image class and go at it that way.20:01
*** dc13ff <dc13ff!uid190567@gateway/web/irccloud.com/x-jzyujketccgvkbyf> has joined #yocto20:07
*** marka <marka!~masselst@> has quit IRC20:16
FrostEyes_P1Evening folks. I am working on a project using a Xilinx ZynqMP, where there is a custom build setup for the "boot.bin". Though is the rootfs a mess currently located as binaries in a git repository. So would like to either use petalinux from Xilinx (Consist of yocto) or yocto directly.20:16
FrostEyes_P1To build the rootfs part of the project. Being new to Yoco, I am puzzled on how you "version control" your setup..20:18
*** joaocfernandes <joaocfernandes!~Joao@a213-22-65-25.cpe.netcabo.pt> has joined #yocto20:19
FrostEyes_P1The first impression is to clone the poky repo, and just hack the code insight this, though seems a bit off to me.20:25
kergothyou should be creating your own layer or layers, not modifying upstream ones20:30
*** ant_home <ant_home!~ant__@host205-252-dynamic.245-95-r.retail.telecomitalia.it> has joined #yocto20:30
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has quit IRC20:30
*** ant_home <ant_home!~ant__@host205-252-dynamic.245-95-r.retail.telecomitalia.it> has joined #yocto20:30
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has joined #yocto20:33
*** tgoodwin <tgoodwin!~tgoodwin@static-108-40-78-74.bltmmd.fios.verizon.net> has quit IRC20:37
*** pohly <pohly!~pohly@p54BD55A3.dip0.t-ipconnect.de> has quit IRC20:39
*** arielmr <arielmr!~quassel@187-163-217-93.static.axtel.net> has joined #yocto20:44
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has quit IRC20:49
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has joined #yocto20:51
*** likewise <likewise!~leon@ip56501413.direct-adsl.nl> has quit IRC21:00
RPhmm, bitbake parsing of oe-core took 49 minutes under qemu on i586, single threaded21:07
RP(before it crashed)21:07
RPand python ctypes.util needs gcc :(21:08
mcfriskI'm impressed. finally got around to properly derive our disto from poky.conf and the qa checks are finding a lot issues. should have done this a lot earlier.. thanks for the really useful QA check in poky!21:13
*** morphis <morphis!~morphis@p200300CCFBEB6E006D44A5210733229F.dip0.t-ipconnect.de> has quit IRC21:14
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has quit IRC21:24
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has joined #yocto21:27
*** joaocfernandes <joaocfernandes!~Joao@a213-22-65-25.cpe.netcabo.pt> has quit IRC21:28
RPmcfrisk: nice :). Makes me wonder if we should update defaultdistro...21:33
*** kaspter <kaspter!~Instantbi@> has quit IRC21:37
*** rob_w <rob_w!~rob@unaffiliated/rob-w/x-1112029> has quit IRC21:37
*** kaspter <kaspter!~Instantbi@> has joined #yocto21:37
*** joaocfernandes <joaocfernandes!~Joao@a213-22-65-25.cpe.netcabo.pt> has joined #yocto21:42
*** varjag <varjag!~user@ti0040a400-6639.bb.online.no> has quit IRC21:52
*** gtristan <gtristan!~tristanva@> has quit IRC21:52
otavioRP: indeed. It is better than defaultdistro21:55
*** georgem_home <georgem_home!uid210681@gateway/web/irccloud.com/x-payouljunsecdthq> has quit IRC22:02
*** rburton <rburton!~textual@> has quit IRC22:04
khemRP: I encourage to make all poky policies go into default distro especially QA checks22:13
RPkhem: the differences were mainly things OE didn't seem to want by default iirc22:14
khemI am not against code that makes things good quality22:15
*** dc13ff <dc13ff!uid190567@gateway/web/irccloud.com/x-jzyujketccgvkbyf> has quit IRC22:16
neverpanicmcfrisk: Get some sleep ;-)22:18
mcfriskneverpanic: you too!22:24
*** joaocfernandes <joaocfernandes!~Joao@a213-22-65-25.cpe.netcabo.pt> has quit IRC22:36
*** nighty- <nighty-!~nighty@s229123.ppp.asahi-net.or.jp> has quit IRC22:51
kergothheh, https://codewithoutrules.com/2018/09/04/python-multiprocessing/22:54
RPkergoth: we seem to have been there in a number of different ways :)23:01
kergothi didn't know about contexts though, that's interesting23:01
khemRP: http://cgit.openembedded.org/openembedded-core-contrib/log/?h=kraj/misc23:04
khemyou did not pick top 2 patches. any issues with them ?23:04
RPkhem: don't think I saw them for some reason23:06
* RP wonders if we can remove the gcc 4/8/4.9 mess from the esdk installer23:06
RPanswer, no, centos723:07
khemRP: centos7 is latest release23:08
*** demonimin <demonimin!~demonimin@unaffiliated/demonimin> has joined #yocto23:08
*** awe00 <awe00!~awe00@unaffiliated/awe00> has quit IRC23:09
khemubuntu 14.04 is also gcc 4.823:09
khemRP: think containers for next release23:10
RPkhem: right, I just for a second hoped we might be able to avoid that23:11
khemI jusr use containers with OE now a days except master where I use arch23:12
khemso we can fix the upcoming issues23:12
RPkhem: makes sense23:12
*** tgraydon <tgraydon!~textual@> has joined #yocto23:12
RPkhem: I don't think forcing people to containers will help our usability/image though23:12
khemlike py723:12
khempython 3.7 I mean23:13
khemRP: who would care of container was not coming in the way23:13
khemsee https://github.com/cbrake/oe-build/blob/master/envsetup.sh#L37723:14
khemhere I launch container when building and directories are exactly same23:14
khemon host and container23:15
*** BCMM <BCMM!~BCMM@unaffiliated/bcmm> has quit IRC23:15
khemso I use my faviourite editor to make changes no issues23:15
khemand when it comes to build it uses a debian stretch container  or debian jessie depending upon which version of oe I am building23:15
khemjessie still works btw with master which is impressive23:16
*** georgem_home <georgem_home!uid210681@gateway/web/irccloud.com/x-ivkcfrungrhqxslf> has joined #yocto23:17
RPkhem: I'm not convinced we can hide the containers from the user, we swap one set of problems for another23:17
khemwe can download the container during setup or clearly instruct how to do that23:18
khemand ask for docker to be installed23:18
khemI might even be able to build on mac osx now a days should probably try23:19
RPkhem: something to think about but I remain unsure we'd want to go this route at the expense of general distro support23:19
khemcontainer will still run a supported distro23:20
khembut it will be just 123:20
khemfolks have receieved it well, they like the fact that they can have a well contained reproducible environment23:21
RPkhem: would be interesting to see what the OSVs make of it23:21
khemeven in data centers they dont have single distro installs23:21
khemCant speak for OSVs but it reduces there work manifolds23:21
RPkhem: yes and no, it forces their users onto one kind of container and given what I've heard about user workflows, that likely won't be acceptable23:22
RPkhem: the older distros don't tend to cause problems anyway, its the newer ones...23:23
khemAgreed there might be such usecases but distros are also in same boat23:23
khemWe can also do the same with lxc and even systemd-nspawn23:24
khemso it doesnt have to be docker only solution23:24
RPkhem: ah, now two things to test ;-)23:24
RPkhem: seriously, I think we should explore it but its very unlikely to switch for 2.723:25
RPit would be hard enough to get that working as one of the workflows when you consider the dev work and docs needed23:25
khemthats fine as long as long we keep making our mind, I am sure we will do something23:25
khemto my mind incentives are high23:26
khembtw. resuming the distro policy discussions, if we can get poky policies ( all sane ones) into nodistro that will be great23:27
RPkhem: probably a case of proposing patches23:27
khemand I was planning to harden toolchains for 2.723:27
* RP is drowning just trying to sort build failures and keep patches moving. Not getting to do much development :(23:28
khemRP: I am getting some of that heat since I am trying to keep up meta-openembedded with oe-core23:28
RPkhem: right :(23:29
khemalmost got to 3 failures and openssl 1.1 came in :(23:29
RPkhem: now I get torn between a) just take 1.1 back out b) try and help sort the failures c) accept meta-oe will sort things out eventually23:30
khemI am progressively minded23:32
khemso I will hope to fix it along the way23:33
RPkhem: I also think we do have to try and move forward even if that is slightly painful at times23:33
khemand I am hoping that meta-oe CI will bring these issues ahead of time so sometimes we can decide23:34
RPkergoth: any neat ideas on how to stop ${@oe.utils.host_gcc_version(d)} expanding in create_shar() unless its running in task context? :/23:34
RPkergoth: I have ideas but I don't like most of them23:35
*** jacques is now known as linuxjacques23:39
RPhttp://git.yoctoproject.org/cgit.cgi/poky-contrib/commit/?h=rpurdie/t222&id=ce3b35f7739e1ccbe00503cc482a2e4dc2ff96ed is the least evil version I can think of right now. I suspect that may cause checksum issues but we'll see23:43
RPFileExistsError: [Errno 17] Link /usr/lib/libcrypto.so already exists to a different location? (libcrypto.so.1.0.2 vs libcrypto.so.1.1)23:47
*** Son_Goku <Son_Goku!~King_InuY@fedora/ngompa> has quit IRC23:47
RPso now I get to debug a new problem before the other one :(23:47
*** grahamgelding <grahamgelding!ca86f31a@gateway/web/freenode/ip.> has joined #yocto23:49
RPhmm, its from a combination of things the autobuilder doesn't test23:52
*** tgraydon <tgraydon!~textual@> has quit IRC23:55

Generated by irclog2html.py 2.11.0 by Marius Gedminas - find it at mg.pov.lt!