**** BEGIN LOGGING AT Sat Jan 25 02:59:59 2020 Jan 25 05:41:37 New news from stackoverflow: How to load image in sama5d27-som1-ek kit using SDcard Jan 25 09:47:24 hmmm, I ahve a vendor how sets defconfig in the linux-ycoto bbappend Jan 25 09:47:44 Does that mean it ignore all attempts to modify behavior with config fragments? Jan 25 10:03:33 Crofton|road: i think so, yes Jan 25 10:05:03 hmm mega manual is suggesting otherwise, but imperical test suggest sdefconfig rules Jan 25 10:07:18 imperial test? Jan 25 10:07:26 * LetoThe2nd cues the Imperial March Jan 25 10:14:59 weird Jan 25 10:16:43 do not understand what I am seeing, maybe order in SRC_URI depenednet Jan 25 10:25:56 welcome to my life Jan 25 10:46:47 OK so it does use the fragments, but it is wise to enable the required CONFIG DEps Jan 25 10:46:58 even though they end up on in th efinal .config Jan 25 10:47:35 CONFIG_USB_SUPPORT=y Jan 25 10:47:35 CONFIG_USB=y Jan 25 10:47:35 CONFIG_TTY=y Jan 25 10:47:35 CONFIG_USB_SERIAL=y Jan 25 10:47:49 Just listing CONFIG_USB_SERIAL failed Jan 25 10:48:00 adding the three listed depends fixed it Jan 25 11:14:12 bluelightning: thanks for the hint yesterday, it worked :) Jan 25 11:14:28 crazy_imp: no worries :) Jan 25 11:14:39 glad to be able to help Jan 25 11:49:17 @Crofton|road: I would assume that first defconfig is used than the defconfig bbappend is applied over it and then the config fragments kick in Jan 25 11:49:46 it is Jan 25 11:50:03 was just confusing until I added options needed to turn on USB_SERIAL Jan 25 11:50:36 Crofton|road: but thats always true, that fragments have to take care of configureation dependencies. Jan 25 11:50:44 yeah Jan 25 11:51:01 just annoying when a file that worked once, doesn't work the next time :) Jan 25 11:51:10 due to different intil conditions Jan 25 11:51:26 Crofton|road: you can now take an elaborated guess why i avoid config fragments. Jan 25 11:51:33 @Crofton|road: Ah I see. Did you generate the fragment via menuconfig/diffconfig? Then additional dependencies should be in the fragment automagically. Jan 25 11:51:53 that sound like something I need to try Jan 25 11:52:11 is there a good guide to working with frgaments? Jan 25 11:52:32 Crofton|road: the kernel manual Jan 25 11:52:39 (the YP one) Jan 25 11:52:46 who reads those? Jan 25 11:52:58 Crofton|road: and once you've figured out, get a hair cut and in front of a webcam! Jan 25 11:53:19 I jsut had a haircut, but onyl the NSA sees my webcam Jan 25 11:53:37 meh Jan 25 11:53:54 we need more streamers! Jan 25 11:54:20 incentive: the uploaded episode of tuesday already has 120+ views on youtube. Jan 25 11:57:52 @LetoThe2nd 120+ views due to all those buzzwords. Just add Iot, AI as well and it will go crazy ;) Jan 25 11:58:10 RobertBerger: go ahead. Jan 25 11:59:15 @LetoThe2nd Once they reach 120K ;) Jan 25 11:59:42 RobertBerger: if you got something to show on those buzzwords, i'll happily assist in getting it out to the public. Jan 25 12:01:11 @LetoThe2nd: I might have at some point in time, but until the Embedded World I am 1 week in the office, where i'll prep equipment for EW. Otherwise I'm on the road training companies. Lack of time. Jan 25 12:02:53 just do it in the evening when you would be drinking anyways. thats what i do. Jan 25 12:07:20 @LetoThe2nd: As a self employed either you drink with customers or you don't drink a lot in the evening. Also a lot of prep time is required so examples actually work and slides are prepared, which should not be under the influence of "substances" like alc. Jan 25 12:08:25 * LetoThe2nd take due note. Jan 25 12:13:24 @LetoThe2nd: oci, Kubernetes, Iot, Grafana, Influx, Python3,... maybe AI,... the question is more when I will have some prototype with my current schedule (and I was not much different the last couple of years) Jan 25 12:13:42 and it was Jan 25 12:17:12 @LeoThe2nd: Ah and instead of drinking in the evening also I need to do some kind of training maintenance. Only on the YP training the last couple of weeks: 853 files changed, 60455 insertions(+), 2894 deletions(-) Jan 25 12:17:58 * LetoThe2nd takes more due notes. :) Jan 25 14:34:19 Hi, I try to up and init can0 after boot automatically with systemd. Althought eth0,eth1 up, can0 and can1 not up after boot. It is my systemd bbappend and can0.service files : https://pastebin.pl/view/238c14a5 Can you help me? Jan 25 14:42:49 I have multiconfig. Why do I see "Layer version information not found; Check if the bitbake server was configured to inherit toaster.bbclass" message in Toaster ? Jan 25 14:54:00 armpit: the recipes-kernel/linux/linux-%_5.%.bbappend in meta-security seems problematic, I believe bitbake is treating that as linux-%.bbappend, as I'm seeing it picked when parsing linux-raspberrypi_4.19.bb (breaking compilation)... Jan 25 14:54:42 smurray: security is problematic, generally.... Jan 25 14:54:47 Jan 25 14:54:51 LetoThe2nd: heh Jan 25 14:57:54 armpit: also recipes-kernel/linux/linux-yocto-dev.bbappend has a line starting with a '+', seems like a patch cut-and-paste Jan 25 15:05:19 hmmm Jan 25 18:07:07 smurray, thanks. I will look into it Jan 25 18:08:28 armpit: thanks Jan 25 18:18:06 RP: I have another glibc upgrade three things 1. Bump to latest 2. Drop long standing ppc cacheline patch 3_ Drop fortify patch Jan 25 18:19:00 RP: I will post it today and also update it on kraj/pu on poky-contrib final glibc releaase then will be a lot smaller Jan 25 18:33:43 khem, need that built ? Jan 25 18:36:11 armpit: sure give me a minute to update it Jan 25 18:45:12 armpit: its ready now https://git.yoctoproject.org/clean/cgit.cgi/poky-contrib/log/?h=kraj/pu Jan 25 18:57:37 khem: I'll get it into the queue, thanks Jan 25 19:13:15 thank you Jan 25 19:14:28 RP: I have got most of gcc10 issues reported and fixed in upstream gcc as well, right now gcc trunk can build world cleanly and core-image-sato/testimage works too Jan 25 19:14:51 but gcc10 wont release in time for 3.1 I think Jan 25 19:15:05 it aligns with fedora release Jan 25 19:15:32 not surprisingly, thats where most of gcc dev work Jan 25 19:19:18 real kicker will be -fcommon defaults change which I have disabled but I will let other distros chime in we have around 130 apps which needs fixing with -fno-common Jan 25 19:21:25 Hi. Can somebody help me? https://stackoverflow.com/questions/59912651/yocto-systemd-can-service-not-start-after-boot Jan 25 19:23:50 khem: sounds like good work, thanks Jan 25 19:23:57 khem: the fun just keeps coming... Jan 25 19:38:53 yeah it never ends does it Jan 25 19:40:43 beratiks: you have a typo FILES_${PN} += "{sysconfdir}/systemd/system/* " Jan 25 19:40:56 {sysconfdir} should be ${sysconfdir} Jan 25 19:41:19 khem: if someone opens a defect on missing RDEPENDS for a package and they're using zeus, is there anything that'd block me from submitting a patch to fix it? Specifically https://bugzilla.yoctoproject.org/show_bug.cgi?id=13741, which is actually a problem with python-twisted being missing RDEPENDS Jan 25 19:41:20 Bug 13741: normal, Undecided, ---, unassigned, NEW , python-twisted declares PyHamcrest>=1.9.0 in requirements.txt but it is not in RDEPENDS (not even needed?) Jan 25 19:42:18 I've already submitted the patch for master, but the user's backtrace suggests they're using zeus Jan 25 19:42:19 tgamblin: it flows via master, so if its fixed in master indirectly then mention that and it can be proposed for zeus Jan 25 19:42:48 khem: Is mistakes at "systemd-machine-units_1.0.bbappend" ? Jan 25 19:42:52 khem: alright, thanks Jan 25 19:43:15 beratiks: yes Jan 25 19:44:01 actually systemd_%.bbappend Jan 25 19:44:12 New news from stackoverflow: yocto systemd can service not start after boot Jan 25 19:46:01 actually it should be FILES_${PN} += "${systemd_unitdir}/system/" Jan 25 19:52:33 khem: I added to "systemd-machine-units_1.0.bbappend" FILES_${PN} += "${sysconfdir}/systemd/system/* ". But nothing change. Jan 25 19:53:15 khem: I have seen now. I add and try Jan 25 19:59:26 khem: It seems nothing change. There is not can0.service at target's /lib/systemd/system Jan 25 20:00:17 is same bbappend adding can0.service too ? Jan 25 20:06:29 If you mean , added "FILES_${PN} += "${systemd_unitdir}/system/" to can0.service, yes same Jan 25 20:08:24 Also for be sure recipe is work, I check : Jan 25 20:08:38 bitbake-layers show-appends|grep systemd-machinesystemd-machine-units_1.0.bb: /home/berat/meta-tq/ci-meta-tq/sources/meta-tq/recipes-core/systemd/systemd-machine-units_1.0.bbappend Jan 25 20:42:26 do you see can0.service anywhere in rootfs ? Jan 25 20:42:49 and ensure that systemd-machine-units is added to image perhaps via IMAGE_INSTALL Jan 25 20:51:14 khem: I havent seen in rootfs. (I search with find -name "*can0*" at / ) Jan 25 20:51:59 hmm then its not adding to your rootfd Jan 25 20:52:03 rootfs Jan 25 20:52:16 khem: Now i build with systemd_%.bbappend. Now I think can0.service will be add. Jan 25 20:52:18 so please ensure the relevant package is added via IMAGE_INSTALL for testing Jan 25 21:57:16 khem: I finished the build. systemd_%.append add can0.service to /lib/systemd/system as I expected but : $systemctl status can0 says that : Jan 25 21:57:23 ��● can0.service - can0 interface setup Loaded: loaded (/lib/systemd/system/can0.service; disabled; vendor preset: enabled) Active: inactive (dead)Jan 25 21:56:55 tqma8xqp-mba8xx systemd[1]: [[0;1;39m[[0;1;31m[[0;1;39m/lib/systemd/system/can0.service:16: Unknown lvalue 'FILES_${PN} +' in section 'Install'[[0m Jan 26 00:12:13 hmm can you post /lib/systemd/system/can0.service somewhere Jan 26 02:17:16 khem: also this could be useful https://gerrit.automotivelinux.org/gerrit/c/AGL/meta-agl/+/20147 Jan 26 02:18:08 uses systemd .network stuff.. then again has to register a real can netdev device, and slcan Jan 26 02:18:13 *not slcan **** ENDING LOGGING AT Sun Jan 26 02:59:58 2020