**** BEGIN LOGGING AT Fri Mar 13 03:00:43 2020 Mar 13 03:56:49 hello world!!! Mar 13 04:00:23 could any one point me in the right direction? my logs under systemD and supervisorD have symbols i assume it a lang mismatch but what is it that causes python(supervisorD) and systemD(assuming C and bash) both have symbols in logs Mar 13 05:11:51 anotherNewb: what do you mean by symbols? do you mean special chars Mar 13 07:00:36 New news from stackoverflow: Bootargs to Mount Rootfs on SAMA5D27-SOM1-EK board using TFTP & QSPI Mar 13 07:30:42 New news from stackoverflow: Restrict the size of Root Partition for hddimg Mar 13 08:57:43 "switching DISTRO_FEATURES on an existing build directory is not supported". Is that really true ? Mar 13 08:57:58 RP, ^^^ Mar 13 09:01:01 New news from stackoverflow: The postinstall intercept hook 'update_gio_module_cache' failed || Where yocto assign default kernel? Mar 13 09:12:36 kroon: technically. It works much better than it ever used to and likely works most of the time now Mar 13 09:15:54 hi all, I've some client on Ubuntu 16.04 and building thud. They have some python function called in a variable assignment which fails because it's python2 valid but not python3. /usr/bin/python points to python2.7... Does someone have a clue why Yocto/bitbake is trying to run that python function with apparently python3? Mar 13 09:16:51 qschulz: wouldn't bitbake be running under py3? Mar 13 09:17:09 RP, ok. I just got that long error message about a recipe trying to install files into a shared area when the files already exsists, in sstate.bbclass (error on my part). But I guess we keep it the way it is then ? Mar 13 09:19:44 RP: that what makes no sense to me! That puzzles me a lot. I'm pretty sure Yocto/bitbake wasn't ready for python3 until zeus right? Mar 13 09:22:20 but the error is pretty explicit, it's a python2 function which does not exist in python3 anymore... (range vs xrange and another string function). I don't see an explanation for that /o\ Mar 13 09:24:29 qschulz: explicit? parental advisory? Mar 13 09:25:06 qschulz: no, our metadata was not. Bitbake switched ages ago Mar 13 09:25:32 kroon: that can probably be fixed or maybe was fixed in master. Which version? Mar 13 09:26:07 qschulz: http://git.yoctoproject.org/cgit.cgi/poky/tree/bitbake/bin/bitbake?h=sumo - sumo has bitbake using python3 Mar 13 09:29:15 RP, this is master. The message mentions as an example that switching from "systemd" to "udev" in DISTRO_FEATURES is not supported Mar 13 09:31:22 RP, https://git.openembedded.org/openembedded-core/tree/meta/classes/sstate.bbclass#n267 Mar 13 09:33:29 RP: I'm confused now.... since bitbake is running on python3, should all python task be executed with python3? Mar 13 09:38:42 qschulz: yes Mar 13 09:39:19 kroon: its never been officially supported, it does actually mostly work now Mar 13 09:39:43 kroon: there are still some issues by the sounds of it. I would be interested in test case Mar 13 09:44:00 RP: aha! krogoth was still executing bitbake with python2. That's code we didn't migrate! Mar 13 09:44:08 RP: everything's clear! thanks a bunch! Mar 13 09:50:36 Good morning everybody. I hope you have a great corona free day. So, I'm failing at injection my own wpa_supplicant.conf into my image :( Do you have a hint on how to do that or which line of the mega manual I should reread? Mar 13 09:50:45 I also phrased my question here. https://stackoverflow.com/questions/60668192/patch-bitbake-to-use-custom-wpa-supplicant-conf Mar 13 09:50:51 So feel free to add your thoughts there Mar 13 09:51:26 emrius: _prepend not _append Mar 13 09:51:51 ah oh! yeah? hmm... Mar 13 09:52:05 emrius: check the paths given in $WORKDIR/temp/log.do_fetch, you see in which order they are traversed for files Mar 13 09:52:07 RP, well.. dunno about test cases.. the fault I had was unrelated to DISTRO_FEATURES, I've never had any problems with switching distro features Mar 13 09:52:27 or, bitbake -e wpa-supplicant | grep -e "^FILESPATH" Mar 13 09:52:31 Oh great! thanks! debugging hints are always the most lasting! I'll dive right into it Mar 13 10:01:11 New news from stackoverflow: Patch bitbake to use custom `wpa_supplicant.conf` Mar 13 10:11:01 Hi, Could somebody verify if I have followed the correct procedure to send a patch. Mar 13 10:11:03 https://lists.yoctoproject.org/g/poky/message/11987 Mar 13 10:11:40 Will this be somewhere added to a pachwork website? Mar 13 10:12:10 And because it's a patch on a poky script I'm not 100% sure it's the correct mailinglist I have send it. Mar 13 10:13:39 emrius: also, wpa-supplicant and not wpa_supplicant in the path, but I'm gonna answer on SO Mar 13 10:14:23 Oh, ok. Thanks. I basically copied the layout from underlying poky stencil. At least I thought I had ... Mar 13 10:14:37 Mmmh… is there a way to remove a "inherit" from a recipe with a .bbappend ? Mar 13 10:17:48 I'm trying to re-introduce a do_install task in a recipe which inherits nopackages.bbclass Mar 13 10:19:31 meego: why? Mar 13 10:20:43 meego: nopackages doesn't remove the install task? Mar 13 10:24:25 rburton: I want a copy of the BCM2835 bootfiles to be installed somewhere outside of the boot partition https://github.com/agherzan/meta-raspberrypi/blob/warrior/recipes-bsp/bootfiles/bcm2835-bootfiles.bb Mar 13 10:25:39 meego: state your problem clearly.. nopackages indeed does not remove the install task so your issue isn't very clear to us Mar 13 10:26:43 meego: it looks like do_deploy already puts it in the deploy directory, can't you use it from there? Mar 13 10:27:27 qschulz: Thanks! `bitbake-layers show-appends wpa-supplicant` alone was super helpful! Mar 13 10:28:50 qschulz: i'll try. Please forgive me if I don't provide the most relevant context. I understand there's such thing as properly asking for help Mar 13 10:28:54 emrius: pleasure, let us know if it works Mar 13 10:29:11 I will Mar 13 10:30:28 meego: it's fine, we're listening :) Think, what's the original issue? What are you trying to do, what's the error (and forget all you've done so far, tinkering the recipe trying to fix the original problem). Honestly, it's hard to ask proper questions, don't worry Mar 13 10:37:20 qschulz `bitbake-layers show-appends` indicates that the bbappend is respected. However, The image doesn't seem to be rebuilding. When I fire up the machine after a rebuild and check history it shows the previous couple of commands. I haven't included `wpa-suppplicant` as a dependency explicitly because it's getting installed anyway. Do I still Mar 13 10:37:20 have to add this dependency explicitly in my image confiugration? Mar 13 10:38:22 Hang on Mar 13 10:39:06 emrius: I' Mar 13 10:39:28 the file that is acually used from the `poky` - `meta` is called `wpa_supplicant.conf-sane` https://git.congatec.com/yocto/poky/blob/f6917ad47b0425f9ccb3965aa9220ed3ba0e49ec/meta/recipes-connectivity/wpa-supplicant/wpa-supplicant-0.6.10/wpa_supplicant.conf-sane Mar 13 10:39:31 've put more info in your SO, there were more small mistakes than I saw in the first read, please check Mar 13 10:39:37 ok Mar 13 10:40:54 wait... those recipes are 9yo /o\ Mar 13 10:41:30 oh hmm... Mar 13 10:41:42 no wait Mar 13 10:41:49 this is not from poky but some other repo Mar 13 10:41:57 sorry Mar 13 10:43:18 sorry got kicked out Mar 13 10:44:12 emrius: ok so, what's the version shown by `bitbake-layers show-appends wpa-supplicant`? Mar 13 10:44:40 === Matched appended recipes ===wpa-supplicant_2.6.bb: /home/marius/mender-qemu/sources/meta-quakesaver/recipes-system/wpa-supplicant/wpa-supplicant_2.6.bbappend Mar 13 10:44:51 2.6 Mar 13 10:45:02 So, actually it must be thud, not warrior. Mar 13 10:45:48 emrius: I'm confused, why did you send us the link for this very old poky version from congatec? Mar 13 10:46:20 cause I'm a lazy googler, copy-pasteing the first hit I find O:3 Mar 13 10:46:29 sorry for the confusion Mar 13 10:46:40 Not using that Mar 13 10:47:06 ok :) so what's the issue now? have you done everything explained in my SO answer? Mar 13 10:47:23 yes Mar 13 10:48:32 But hang on. I just changed `wpa_supplicant.conf` to `wpa_supplicant.conf-sane` which is the nameing used for the file that is acually used here. Mar 13 10:49:03 I assume it's somewhere `mv`ed along the road. At least the image is clearly now rebuilding as wpa-supplicant just got rebuild... Hang tight Mar 13 10:49:59 emrius: is there any wpa_supplicant.conf-sane in the base recipe of wpa-supplicant? Mar 13 10:50:07 yes Mar 13 10:50:48 The wpa_supplicant.conf has tons of comments and documentation. Thus, I assumed they included the sane version to be actually used in the image (confirmed that already) Mar 13 10:51:37 yes, wpa_supplicant.conf-sane seems to be installed as wpa_supplicant.conf in the image according to the recipe I've in thud Mar 13 10:51:51 Heureka! There it is!!! All in it's place Mar 13 10:51:52 and wpa_supplicant.conf is just put in the doc directory Mar 13 10:51:54 <3 Mar 13 10:52:11 let me update the answer Mar 13 10:52:26 ok great! I owe you a beer Mar 13 10:52:38 beer? i hear beer? Mar 13 10:52:59 LetoThe2nd: do you really have a trigger on the word beer :'D ? Mar 13 10:53:11 or better, a bot answering for you :'D? Mar 13 10:53:22 qschulz was thinking exactly the same Mar 13 10:54:10 If I have an urgent question for you, LetoThe2nd, I know now how to trigger your attention. Might become useful one day Mar 13 10:54:46 * LetoThe2nd is mightily disappointed now. Mar 13 10:55:54 emrius: updated the SO answer Mar 13 10:56:01 so people can have the correct answer :) Mar 13 10:56:23 Accepted! Thanks a lot! Mar 13 10:57:17 emrius: nice discussion, learned something new as well :) Mar 13 10:57:40 great to hear that!! Win-win :) Mar 13 11:01:05 With multiple .bbappends in different layers _append:ing patches to SRC_URI, is there a way to control the patch apply order, besides using BBFILE_PRIORITY_ ? Because I guess filename doesn't matter ? Mar 13 11:03:50 kroon: if all bbappends are using _append and you want to be before, SRC_URI += in your bbappend will work because _append are resolved after all = .= =. += =+ ?= ?== Mar 13 11:04:12 that's what came to my mind, but hackish :/ Mar 13 11:06:08 qschulz: thanks for the encouragement :) Mar 13 11:06:16 The high-level problem I'm trying to solve is improving the over-the-air update logic for Raspberry-Pi based devices. Right now I use Mender for OTA updates. It works well, with a 4-partition layout system (boot, active system, inactive system, persistent data) but there's a catch. It never updates the files located in the boot partition. Mender update artefacts only contain an image of the system partition. So I need to f Mar 13 11:06:16 ind a way to copy the bootfiles to the system partition, so they can be deployed later with a script to the boot partition. Mar 13 11:06:29 My idea up to now has been this: #1 figure out which recipes provide the files in the boot partition, #2 write a bbappend for each recipe that copies the bootfiles to a location where I'm sure they'll end up on the system partition during image/artefact building. Mar 13 11:07:27 The first obstacle I hit, and the reason for my last question, was with the bcm2835-bootfiles recipe linked above. It do_deploy's files in a path which is not picked up during Mender artefact building. So my objective at this point is to re-introduce a "regular" do_install/do_package/do_populate_sysroot flow to keep the files in the system partition. Mar 13 11:08:36 meego: or, since this looks very image specific, maybe something to do from the image recipe? before it creates the partitions? apparently the files are available in the deploy_dir, so that's a way to get your file into the partition you want Mar 13 11:09:32 qschulz: sounds much saner, yes Mar 13 11:11:49 meego: I don't know what Mender is using for creating the partitions, maybe wic? maybe you need to modify the .wks to have the files/directory you want in there. Hopefully I'm not giving you wrong hints Mar 13 11:14:08 Hi. I just tried to update Zeus from 3.0.1 to 3.0.2. I'm now unable to start the bitbake server. It fails with an error about failure to expand AVAILABLE_LICENSES. I took a quick look through the mailing list and found no mention. Before I start diving in, I'm hoping someone can give me a quick fix. Mar 13 11:18:18 mabnhdev: give the full error :) Mar 13 11:18:19 qschulz: that's useful, thanks ! Mender indeed uses wic. I'll follow that lead. I don't know much about wic & what happens during do_image TBH. I'll start w/ the manual… Mar 13 11:19:15 IMAGE_CMD_wic ;) Mar 13 11:19:28 in classes/image_types_wic.bbclass if you want to read code Mar 13 11:20:43 How come raising the layer priority causes a corresponding .bbappend to be applied *later* ? Mar 13 11:20:46 @qschulz Sorry - thought it may be an obvious one. https://gist.github.com/mabnhdev/521c37882d42091b9ffd993480f2b7c4 Mar 13 11:21:27 kroon: probably because *later* means it can override *more*. so it takes higher precedence, which corresponds to a higher priority. Mar 13 11:22:48 LetoThe2nd, yeah.. makes sense now that you say it Mar 13 11:24:36 :) Mar 13 11:25:07 mabnhdev: where's the licenses dir in meta-exos? Mar 13 11:27:55 @qschulz I've never had one. Mar 13 11:27:56 mabnhdev: try to find if there's any incorrect LICENSE_PATH in some conf/layer.conf. The error is that it's looking for /data/mberger/exos-yocto-307-update/layers/meta-exos/licenses directory but couldn't find it Mar 13 11:31:35 @qschulz Aha. My layers.conf had LICENSE_PATH += "${LAYERDIR}/licenses" (since day 1). I just commented it out and all is good. I should probably put together a Licenses dir for my layer. Mar 13 11:32:24 mabnhdev: depends, is there a custom license for any of the recipes in your layer? no, no need. yes, please do so :) (you'll be reminded by a warning I think anyway) Mar 13 11:33:14 @qschulz Thanks for pointing me in the right direction! Cheers. Mar 13 11:33:44 mabnhdev: my pleasure, learned something as well :) Mar 13 11:46:17 qschulz: I think i get it now. The bootfiles are listed under IMAGE_BOOT_FILES. And I've found the image bbclass where they're added to the wks. But now I realize: this Mender logic is indeed image-specific, but what i'm trying to accomplish isn't. Since I'm aiming to have the files in the system partition, the do_install/do_package/do_populate_sysroot path seems relevant in retrospect Mar 13 11:52:09 meego: I fail to see why it's not image-specific? Mar 13 12:07:38 Hi Guys! Ive a problem with a bbapend for my kernel. I have two different machine that uses the same kernel but I need different DTBs and kernel configs for them. How do I solve this in my linux-ti-staging_%.bbappend file? Or should I solve this elsewhere? Mar 13 12:08:09 Linus_SWE: put it into the machine.conf Mar 13 12:08:32 Linus_SWE: you can use machine specific overrides in bbappend too, if machine.conf comes from a third party Mar 13 12:09:26 tahat another option Mar 13 12:09:26 LetoThe2nd, erbo Thanks! Im the creator for the machine.conf so no problem. Mar 13 12:09:56 but usually you'd just stick this into the MACHINE: KERNEL_DEVICETREE = "mymachine.dtb" Mar 13 12:10:31 Yeah, sorry I wanst clear. I need to build my own DTBs Mar 13 12:11:23 so in my bbappend i do SRC_URI += "file://am335x-xxxx.dts" today Mar 13 12:12:03 Can I do that from machine.conf? Mar 13 12:12:19 no Mar 13 12:12:31 in that case you need a kernel recipe append. Mar 13 12:13:26 ok, thats what I have today. But should I always build all DTBs and just select them in machine.conf? Mar 13 12:16:55 Thats probably the easiest way, yes. Mar 13 12:17:05 ok, thanks! Mar 13 12:17:36 How about kernel configs? Today I add do like this: Mar 13 12:17:41 #Enable RNDIS in kernelSRC_URI += "file://config-usb-net-rndis.cfg"KERNEL_CONFIG_FRAGMENTS_append = " ${WORKDIR}/config-usb-net-rndis.cfg" Mar 13 12:17:49 in my bbappend Mar 13 12:17:58 How do I move this to machine.conf? Mar 13 12:18:32 sorry about the bad format, there should be some new lines in there.. :) Mar 13 12:19:35 Can I just do the same and have the files in a folder where the machine.conf is? Mar 13 12:20:32 nope Mar 13 12:20:51 if you need to reference files then it has to go into a recipe Mar 13 12:21:57 ok, thanks. Can I enable specific kernel configs without files? Mar 13 12:24:45 if the kernel is yocto-ized, then i think so. but thats pretty much beyond my expertise. Mar 13 12:27:27 LetoThe2nd, Ok. Thanks for your help. Ill do some research :). Mar 13 12:28:47 Linus_SWE: the kernel manual certainly knows more about it. Mar 13 12:31:52 Could somebody check if I have followed the correct procdure to submit my first patch? Mar 13 12:31:53 https://lists.yoctoproject.org/g/poky/message/11987 Mar 13 12:32:32 Is this the correct mailing list for a patch in poky? And will it appear somewhere in a patchwork site? Mar 13 12:34:33 Hi y'all! Mar 13 12:35:18 I'm working with a bbclass and I wonder if I can somehow add to FILES_${PN} from the install task? Mar 13 12:35:39 I tried but it didn't seem to work... Mar 13 12:38:09 no Mar 13 12:38:51 the variables are supposed to be static after recipe parsing/evaluation, and not to be dynamically tinkered with in functions. Mar 13 12:39:09 Bummer... Mar 13 12:39:30 or bad recipe architecture, depends on the point of view :) Mar 13 12:39:51 I find out the names during compile of some config files so I guess I must parse from annonymus then? Mar 13 12:40:00 reason is, if they are static changes can be detected to trigger tasks and also find problems. Mar 13 12:40:17 hmm just wondering if FILES_${PN}_append may work or ${PN} is just unevaluted in class context ? Mar 13 12:40:57 dev1990: the append should work fine, IMHO Mar 13 12:41:38 I know the PN from before and I'm trying d.appendVar('FILES_' + pkg, ...) Mar 13 12:41:40 karrman7: ok then mayby try with _append Mar 13 12:42:59 But _append seems like something that IS handled during parsing... Mar 13 12:44:29 karrman7: maybe just do the easy way out and add the directory that your unknown config files are ending up to files? Mar 13 12:44:51 (essentially globbing them without knowing the exact names) Mar 13 12:46:09 Hmm... there are a bunch of files there that should end up in different packages Mar 13 12:46:45 was just an idea. Mar 13 12:47:06 I am thinking maybe I could add some restriction to the naming, like a prefix, but I rather not. Mar 13 12:55:34 karrman7: "I find out the names during compile of some config files"? I'm wondering if the filename matching logic you have in do_install couldn't be used in FILES_${PN} directly. I mean, what's the crazy thing about those files? Also, FILES_ can have more paths or glob without actually matching... Mar 13 12:57:18 The config file contain the names of the files I generate Mar 13 13:03:44 OK, thanks for suggestions! It will be either naming restrictions or parsing in __anonymous(). Bye! Mar 13 13:11:33 Doesn't wiping TMPDIR and rebuilding from sstate cache regenerate the task signature files ? Mar 13 13:14:35 from experience, not in thud at least Mar 13 13:15:12 kroon: I would use bitbake -S none like scripts/sstate-diff-machines.sh does Mar 13 13:28:21 JaMa, thanks. seems to be just what I needed Mar 13 13:29:45 Are changes from openembedded-core are backporeted to poky overtime or patches should be created and submitted to both projects? Mar 13 13:30:36 dev1990: submit to openembedded-core, poky is a generated repo Mar 13 13:30:39 they are integrated with combo-layer tool Mar 13 13:30:42 JaMa, so I can now: 1. wipe TMPDIR 2. bitbake 3. bitbake -S none 4. Update layers 5. bitbake 6. finally enjoy a working bitbake-diffsigs ? Mar 13 13:31:19 RP: ok, thanks Mar 13 13:34:04 kroon: nice! learned something new. That's a good Friday so far :) Mar 13 13:34:07 JaMa: ^ Mar 13 13:44:45 JaMa, qschulz, yeah, that has been bugging me for quite a while Mar 13 14:02:08 New news from stackoverflow: Is there a way for a custom aarhc64 cross compiler to search for headers and libraries when it is built --with-sysroot=/not/exist Mar 13 15:09:45 not really yocto-specific, but I know a lot of people are suddenly WFH more than they used to be: Mar 13 15:09:46 https://benediktlehnert.github.io/ Mar 13 15:31:19 for u-boot do I have to explicitly enable i2c? I get Unknown command 'i2c' - try 'help' Mar 13 15:31:19 in u-boot console Mar 13 15:31:58 Ad0: https://elixir.bootlin.com/u-boot/latest/source/cmd/Kconfig#L997 Mar 13 15:32:10 I'm pretty sure it should be this one Mar 13 15:32:28 when typing help, if you don't see it in the list, it's not there, no magic Mar 13 15:33:46 thanks. it looks a lot like kernel config Mar 13 15:34:34 it's in 2019.01 too so I should be good Mar 13 15:34:42 but it's simply not there Mar 13 15:35:01 is this a part of the kernel config? Mar 13 15:37:23 I was looking for recipe magic to set it Mar 13 15:37:47 Ad0: kernel config is for the kernel, u-boot config is for u-boot. There are defconfigs for U-Boot yes Mar 13 15:38:11 There's no magic process. Modify the defconfig used by U-Boot, put it in the recipe, build your image, flash u-boot Mar 13 16:10:16 qschulz, thanks Mar 13 16:18:26 modify the defconfig using menuconfig please. Then make savedefconfig, then take the defconfig in the root of U-Boot :) Mar 13 16:18:58 Ad0: ^ Mar 13 16:20:15 thanks :) well I use mender, and mender is patching and doing a lot of stuff with the config it seems, but there's no "defconfig" filename to be found Mar 13 16:20:17 really confusing Mar 13 16:21:37 it's called .config it seems Mar 13 16:23:47 ah it runs kconfig with --defconfig Mar 13 16:25:26 when doing make defconfig, it creates a .config which is a list of all Kconfig options available and their value. Defconfig is just a stripped down version of .config. .config is the file used when compiling. That's he difference between both Mar 13 16:26:35 ok Mar 13 16:26:56 https://github.com/mendersoftware/meta-mender/blob/master/meta-mender-raspberrypi/recipes-bsp/u-boot/patches/0001-CONFIGS-rpi-enable-mender-requirements.patch Mar 13 16:27:01 when they do stuff like that Mar 13 16:27:14 I was hoping for a config fragment system :) Mar 13 17:10:30 Hi guys, do you know when does linux-intel-4.19 go EOL? I tried searching on the internet, but I couldn't find any info Mar 13 17:13:10 i *think* fairly soon, but you'll want to ask anuj, the maintainer of meta-intel Mar 13 17:13:12 kriive: official 4.19 EOL is end of this year so... more or less same? Mar 13 19:06:50 RP: nice reduction in master-next :) Mar 13 19:07:07 I have root rights on my build box now, what's still missing is it's on the wrong network Mar 13 19:07:19 big company stuff as usual, takes time Mar 13 19:07:55 but it is a 2990 32-core AMD TR Mar 13 22:18:13 Is there a way to set SRC_URI if it's not already set? SRC_URI?= doesn't work, since it appear Poky always sets it to "". Mar 13 22:24:27 xyzzy42, some anonymous python could probably do it Mar 13 22:26:04 xyzzy42, but I'd question why its even set in bitbake.conf to begin with, judging by the git history Mar 13 22:31:43 why would you want to set it only if its not already set? curious what the use case is Mar 13 22:33:40 why it is set to "" would be a even more interesting to know to me .. Mar 13 22:36:32 but apparently lots of python code assumes SRC_URI != None Mar 13 23:01:33 kanavin_home: sounds like progress at least! :) Mar 13 23:01:56 kanavin_home: we're actually getting green master-next builds Mar 13 23:02:12 The history of https://autobuilder.yoctoproject.org/typhoon/#/builders/83 is "interesting" Mar 13 23:10:58 RP: right, but master-next will fail more often than not, more interesting would be to see a history of plain master builds :) Mar 13 23:15:01 kanavin_home: those are master and master-next. We've not had green builds working for ages :( Mar 13 23:15:36 kanavin_home: https://autobuilder.yoctoproject.org/typhoon/#/builders/85 is quick builds which are master or stable work and it doesn't look much better :/ Mar 13 23:15:54 anyhow, builds have improved Mar 13 23:18:45 RP: right, hopefully there'll be less firefighting, this time it seemed particularly protracted Mar 14 00:51:24 RP: I was looking at meta/lib/oeqa/runtime/cases/parselogs.py and it lists errors by machine types Mar 14 00:52:04 I wanted to add some exceptions for rpi preferably Mar 14 00:55:10 in the SOC layer, but I did not find an easy way Mar 14 01:10:03 see https://git.openembedded.org/openembedded-core-contrib/commit/?h=yoe/mut&id=2e9acabea250ecb7c0cb06834373cea8490f5b2b **** ENDING LOGGING AT Sat Mar 14 03:02:45 2020