**** BEGIN LOGGING AT Thu Mar 12 02:59:57 2020 Mar 12 07:55:59 New news from stackoverflow: Fatal error: cstddef: No such file or directory when trying to cross compile c++ libraries using a custom yocto sdk aarch64 Mar 12 08:31:13 Hi, Mar 12 08:31:37 Could it be the poky/scripts/tiny/ksize.py is, although converted to python3, not fully working with python3? Mar 12 08:32:04 The tables is more in detail when running the script with python2. Mar 12 08:32:17 BTW: working on master of a week ago Mar 12 08:46:53 vermaete: certainly possible, i don't think these parts get much testing/exercise these days. Mar 12 08:52:50 fine, so if I can, I can make a patch and send it to the mailinglist. Correct? Mar 12 08:53:09 vermaete: thats would be the preferred conclusion, yes :) Mar 12 08:53:43 vermaete: (see also https://git.yoctoproject.org/cgit/cgit.cgi/poky/log/scripts/tiny/ksize.py) Mar 12 08:53:48 BTW: any reason why these are not much used? Is there something new I've missed? Mar 12 08:54:34 vermaete: probably nobody needs them :) Mar 12 09:10:50 hi :)) Mar 12 09:11:31 i want to build image for zynq7000 Mar 12 09:11:36 lo :(( Mar 12 09:12:08 oh Mar 12 09:12:59 i think i watch youre videos on utube Mar 12 09:13:13 thanks man Mar 12 09:13:43 can i ask a question? Mar 12 09:13:58 i think i watch youre videos on utube Mar 12 09:14:15 you can of course ask any question. :) Mar 12 09:14:20 you're welcome Mar 12 09:14:45 yet no guarantee and *any* question will also yield an answer :) Mar 12 09:14:58 thanks :] Mar 12 09:18:44 i want to build an image with fsbl and etc thing for zynq 7000 , i now sth is wrong with device tree that i the device doesnt work , the best branch for xilinx tools is thud i think . thats true? and can you help me for more enformation about device tree and ... Mar 12 09:20:53 first, i have never used a xilinx board and have no idea what "fsbl" and "etc thing" are. Mar 12 09:21:24 so the first thing would be, to start out with their documentation and build a well-known image. have you already done that? Mar 12 09:23:01 yea i do that but i with petalinux Mar 12 09:23:39 petalinux use yocto Mar 12 09:24:06 i also have no idea about petalinux. Mar 12 09:24:42 anyways. so you have a working build and a resulting, bootable image that works, right? Mar 12 09:24:55 yea Mar 12 09:25:32 do you know about sources for device tree stuffs? Mar 12 09:25:47 *sigh* Mar 12 09:25:57 what answer do you expect to that question? Mar 12 09:26:11 a book or refrence :) Mar 12 09:26:18 sorry Mar 12 09:26:19 New news from stackoverflow: yocto offline build fails cause fetches from recipes Mar 12 09:26:37 so your problem is in understanding how the device tree in itself is structured? Mar 12 09:27:22 yea Mar 12 09:27:49 md_micheal: https://elinux.org/images/f/f9/Petazzoni-device-tree-dummies_0.pdf Mar 12 09:28:06 https://www.youtube.com/watch?v=m_NyYEBxfn8 Mar 12 09:28:58 the last time i build image for zynq in zeus branch with the xilinx layers Mar 12 09:29:17 i have some errore about initramdisk i think Mar 12 09:29:55 and i think this related to dt Mar 12 09:30:13 md_micheal: I'm sorry but you're just throwing "it does not work" or "how does that work" without giving us any context, any error, nothing. How are we supposed to help you? Mar 12 09:31:02 At the same time, this is most likely NOT related to yocto, so I'd suggest to start contact your vendor and ask them those questions. Especially since Yocto BSP layers are usually not really following best practices (I'm saying this as nicely as I can, it's often total garbage) Mar 12 09:32:41 youre true , im building image again and send you the error again . excuse me for insufficient data Mar 12 09:32:51 :] Mar 12 09:36:52 i actually thinkt the xilinx bsps should be ok, but no idea about their petalinux stuff. Mar 12 09:43:18 md_micheal: bsp is for evaluate boards. Mar 12 10:22:15 hello, is there any body here using yocto project for Microchip SAMA5D27-SOM1-EK1 bord ? Mar 12 10:22:45 userr8: maybe. Mar 12 10:22:56 userr8: just ask your *ACTUAL* question. Mar 12 10:23:42 well I insist on this board because, first, I have a hardware question Mar 12 10:24:10 userr8: then we will insist on you going to #at91, if you prefer to start out like that, ok? Mar 12 10:24:41 even moreso if it is a hardware question :) Mar 12 10:25:09 okey Mar 12 10:43:58 halstead: Hi, hope everything's alright by you. FF74.0 refuses to connect me to wiki.yoctoproject.org due to TLS1.2 not being used. Just to let you know :) Mar 12 13:13:26 Hey, is there a certain QC procedure I have to follow before I submit a pull request for a recipe that I would like to be added in `meta-openembedded/meta-python` or can I simply send a pull request? Mar 12 13:15:59 emrius: i don't think they accept pull requests, better make it a patch onto the oe-devel mailing list. Mar 12 13:16:12 Ah alright. Will do so Mar 12 13:16:16 thanks Mar 12 13:29:18 RP: Are you missing the autobuilder.json update I sent, or did you move it elsewhere? Mar 12 13:42:33 JPEW: have another look at the branches? I'm having problems getting the patches applied to the repos correctly Mar 12 13:42:44 JPEW: I think its right now Mar 12 13:48:57 JPEW: the logging isn't quite doing what I'd expect in testing, or I have it setup incorrectly :/ Mar 12 13:52:14 RP: Ok, looks correct now Mar 12 13:52:57 RP: What the unexpected behavior? Mar 12 13:56:51 JPEW: I think I'm misunderstanding. We've already turned off some logging I think we should be putting in log files Mar 12 13:57:39 RP: Hash equivalence related logging? Mar 12 13:58:01 JPEW: yes. Just running some tests to confirm but my tests aren't working :( Mar 12 13:58:16 RP: Running on the AB or locally? Mar 12 13:58:26 JPEW: locally Mar 12 14:00:31 RP: Hmm, I think all of the hash equivalence logs that used to be displayed on the console should be recorded to the console-latest.log file by default (e.g. w/o setting BB_LOGCONFIG) Mar 12 14:01:21 RP: And setting BB_LOGCONFIG to the autobuilder.json should make them appear on the console also Mar 12 14:01:21 JPEW: what I'm not seeing are the "Task X unihash changed to Y" messages Mar 12 14:03:29 RP: Are they in the log file? Mar 12 14:03:49 JPEW: its as if equivlance in my build has just stopped entirely :/ Mar 12 14:04:23 RP: Huh, weird Mar 12 14:04:39 JPEW: I just don't understand it Mar 12 14:08:45 RP: I'll try again with just master-next and no local patches Mar 12 14:10:35 JPEW: I'm fairly sure this is me Mar 12 14:19:52 RP: Ok. Well, I just tried it on master-next and it all looks like it's working as expected to me Mar 12 14:20:20 Hmm, a lot of bitbake utilities fail if BBTARGETS is set.... Mar 12 14:20:29 JPEW: there is something wrong with my build I think Mar 12 14:41:31 hi question, can i change SRC_URI of a kernel recipe to another repo, from a .bbappend ? testing this, seems not. Should i create a new .bb ? Mar 12 14:43:14 angelo__: you can if you remove the original entry in SRC_URI Mar 12 14:43:53 two ways. you redefine SRC_URI. Easy if there's only one entry, more complex if there are some patches or files to be included. Mar 12 14:44:37 Or you do SRC_URI_remove = "git://foo-bar.git" of the repo you want to remove and then SRC_URI += (or _append) with your repo Mar 12 15:11:28 qschulz, thanks, trying Mar 12 15:39:35 qschulz, seems only way worked is total re-assign of SRC_URI Mar 12 15:55:33 JPEW: the issue was my timestamps change to hashequiv sigs Mar 12 15:55:40 JPEW: it meant I had no matches ever Mar 12 15:57:16 angelo__: SRC_URI_remove will work 100% of the time. Maybe a typo? it has to be the exact same line Mar 12 15:57:42 JPEW: sadly the knotty output gets broken with your change :( Mar 12 15:58:07 JPEW: The trouble is its not disabling the header before printing and hence the consule gets confused if its interactive Mar 12 16:41:42 qschulz, mm looks like i was getting errors not in the _remove but in the += part, readding the repo uri Mar 12 16:41:53 btw, thanks anyway, will retry in case Mar 12 17:05:50 RP: Which header? Mar 12 17:06:59 JPEW: in the knotty code, the current task execution status Mar 12 17:11:16 RP: The filtering done in InteractiveConsoleLogFilter? Mar 12 17:11:55 JPEW: yes Mar 12 17:12:19 JPEW: if I run with the autobuilder config and it has hash equiv output, the console is corrupted Mar 12 17:12:37 Ah, OK. I'll give that a try quick Mar 12 17:14:29 RP: OK, I think I see it. It leaves partially completed progress bars? Mar 12 17:17:06 JPEW: yes Mar 12 17:17:25 JPEW: exactly what happens depends on the current messages but I think the output is bypassing one of the handlers Mar 12 17:17:29 RP: Ok, got it. Fix is pretty easy Mar 12 17:17:32 (probably the wrong terms) Mar 12 17:20:11 JPEW: sorry, poking at things around meetings hence the lagged responses Mar 12 17:20:26 RP: It's fine, I was gone for lunch Mar 12 17:22:02 RP: Hah, you probably meant right now. It's fine :) Mar 12 17:23:17 Hi, I would like to use wpa_supplicant to connect to wifi. There is `poky/meta/recipes-connectivity/wpa-supplicant/wpa-supplicant/wpa_supplicant.conf-sane` which is placed in to `/etc/`. How can I inject my own `wpa_supplicant.conf` right from bb to connect to a known host? Mar 12 17:23:37 JPEW: both :) Mar 12 17:23:59 I could of course modify that `*-sane` file but there must be a better way to overwrite. Mar 12 17:24:14 JPEW: good news is logging is basically working, ignoring that corruption which I agree is fixable one way or another :) Mar 12 17:24:27 JPEW: I think I'll get on and merge things and then we go from there Mar 12 17:27:34 for anyone considering to buy AMD 3990X for OE builds: https://www.youtube.com/watch?v=VQa6r6Ci1jg but he didn't say much about 3990x nor sent me the pull request with detailed results yet Mar 12 17:28:10 I think I have it Mar 12 17:33:15 JaMa: have a tl;dw for the video? Currently exploring options for an OE build machine. Mar 12 17:42:10 jrdn: 3970X is great for OE builds for reasonable price (much faster than older dual socket Xeons), 3990X is significantly faster but for most people might not be worth double the price Mar 12 17:43:37 jrdn: I'm even more interested in comparing 3990X with single socket 7702P which is only slightly faster, but isn't limited to 256G ram and with 8 memory channels and ECC might be faster and more reliable for long-term build machine Mar 12 18:30:31 RP: Not that I care, but all these hot patches make me look a little scatter brained ;) Mar 12 18:30:47 Anyway, the latest set should fix the console corruption Mar 12 19:12:15 hm, is gcc 9.3 too late for m3? Mar 12 19:18:10 rburton: heh, I had that same thought earlier Mar 12 19:50:28 dooooo it Mar 12 19:50:36 I'll roll a new kernel as well! :P Mar 12 21:21:01 rburton: its perhaps not Mar 12 21:26:57 given LTS Mar 12 21:27:23 and my branch is always at top of the branch Mar 12 21:30:46 is that King branch ? Mar 12 21:32:02 khem, send patches Mar 12 21:35:46 yes I am preparing, it will be in tonight Mar 12 22:11:46 JPEW: we can squash some? Mar 12 22:11:56 (thanks, will take a look) Mar 12 22:12:31 gah. M3, the build that we never build Mar 12 22:26:47 Hello. I'm trying to do 'bitbake core-image-minimal' using the default .conf's (except with the 3.0.2 SState mirror defined), but it crashed out with an "error: Failed to initialize NSS library" error Mar 12 22:27:35 This happened after "NOTE: Setscene tasks completed", as the result of a dnf command Mar 12 22:52:58 JPEW: btw, if you want to feel better, look at autobuilder-helper :/ Mar 12 22:54:24 khem: are we aiming for gcc 9.3? Mar 12 22:54:34 khem: i.e. should I wait on M3? Mar 12 23:03:01 yes Mar 12 23:03:06 sending patch in few mins Mar 12 23:09:47 RP: patches on ML now Mar 12 23:10:11 khem: thanks, will take a look and likely run another test build Mar 12 23:11:28 RP: I have done SDK builds and run core-image-minimal tests so far on qemux86 looks ok Mar 12 23:11:39 I will also run world build tonight Mar 12 23:11:45 so will know more tomorrow morning Mar 12 23:11:50 but it seems to hold Mar 12 23:12:50 khem: ok, its running on the autobuilder Mar 12 23:14:47 RP: I think we also need to backport https://sourceware.org/git/?p=binutils-gdb.git;a=commit;h=ae531041c7c5956672342f89c486a011c84f027f Mar 12 23:14:55 for binutils Mar 12 23:19:30 also systemd 245 released but I see that ubuntu 20.04 is still on 244 so perhaps we should stick to 244 Mar 12 23:20:03 khem: 245 has reported issues so we're not doing that Mar 12 23:20:35 RP: at this time, I am also looking at aligning versions with ubuntu 20.04 since that will be LTS too Mar 12 23:20:57 so we can share some maintainance pain with wider distro community Mar 12 23:21:01 khem: fair enough, just saying I know 245 will break things Mar 12 23:21:10 right Mar 12 23:22:33 I want to spend some time on ptesting here on Mar 12 23:22:54 but folks keep breaking musl :) Mar 12 23:24:34 khem: its not the only thing... Mar 12 23:25:09 RP: yeah I am sure you see a lot more Mar 12 23:26:19 khem, get bigger musl's Mar 12 23:29:24 RP: Squashing is unnecessary :) Mar 12 23:31:57 JPEW: I did have a look and I think the history is good to explain the changes Mar 12 23:42:06 armpit: good idea Mar 12 23:55:03 any chance that libarchive-3.4.2.tar.gz could be added to the mirrors? I can't download it from the primary source Mar 13 00:02:53 hmm, its the latest. it should be on the yp sources Mar 13 00:05:29 HunterZ, this one ? http://downloads.yoctoproject.org/mirror/sources/libarchive-3.4.2.tar.gz Mar 13 00:06:30 the yp mirrors should be part of the poky* distros Mar 13 00:20:53 when i try to bitbake from master it only tries to download that from libarchive.org, which is blocked by my firewall (apparently some old version had a keylogger vulnerability, so out goes the baby with the bathwater) Mar 13 00:23:03 to get past that, can you wget from YP archives and put that in your download dir? Mar 13 00:37:57 armpit: Thanks, I'll try that next week if I run into the issue again Mar 13 01:23:43 oe sources mirror should also have things like this **** ENDING LOGGING AT Fri Mar 13 03:00:43 2020