**** BEGIN LOGGING AT Tue Dec 06 02:59:56 2022 Dec 06 14:07:28 WARNING: Your host glibc version (2.36) is newer than that in uninative (2.34). Disabling uninative so that sstate is not corrupted. Dec 06 14:07:31 this is annoying Dec 06 14:09:46 Crofton: change to a newer uninative version, is easy and safe Dec 06 14:10:00 hmm Dec 06 14:10:02 OK Dec 06 14:10:31 at least it clearly told you what the problem was rather than giving you obtuse errors Dec 06 15:27:19 it could mention updating the uninative :) Dec 06 16:29:33 vmeson: so, I tried to follow your hints for the reust thing, but I think I am missing one last important step for understanding how to complete it; so I am on master, I ran bitbake -c update_crates recipe-name which generated me that .inc ile with the SRC_URI cargo entries Dec 06 16:29:58 what now? I I supposed to remove the cargo-update-recipe-crates inherit? Dec 06 16:30:41 and thats it? or do I need some other rust related inherit? or does the fetcher know how to handle cargo already? especially in regard to moving the generated recipeto Kirkstone Dec 06 16:31:27 its also not entirely clear to me if I do need meta-rust or not at this point, i.e. if cargo-update-recipe-crates sort of replaces it or if these are independend things? Dec 06 16:32:45 Jin^eLD: you're one step ahead of me! I confirmed that the 'cargo bitbake' approach still worked but haven't tried the new approach yet. I may get to that this afternoon... Dec 06 16:33:29 but is the new approach documented somewhere except for that one line in the bbclass header on how to create the .inc file? Dec 06 16:33:50 I am lacking the global picture a bit I guess Dec 06 16:34:48 I am now naively simply bitbaking the recipe (still on master) and it seems not to complain so far, so curious on what comes out Dec 06 18:21:22 vmeson: well, it "did something", but the result of the build are two ipk's, a -dev and a -dbg one, which are essentially empty Dec 06 18:21:44 I need to break into devshell to see if any binaries were produced which may have needed manual installation Dec 06 18:22:19 but the "generated" package copied over to kirkstone (removed inherit of cargo-update-recipe-crates) did build on kirkstone as well, with the same result though Dec 06 18:30:06 vmeson: as far as I can tell the compile step does not build anything at all.. I think I should have inherited "cargo" Dec 06 18:39:01 Jin^eLD: best to follow-up on the thread where the new method was introduced. I have some other things going on today and lilely this week. Dec 06 18:39:45 I am not sure I can join a thread without being able to reply to an existing mail from that thread? Dec 06 18:40:02 I wasn't subscribed lately Dec 06 20:33:50 ANy have an answer to this question?https://www.yoctoproject.org/pipermail/yocto/2019-October/047112.html Dec 06 20:34:00 When do you change LCONF_VERSION? Dec 06 20:41:49 Jin^eLD I'll reply and loop you in. Please DM your email address. Dec 06 20:42:17 wait, I have to subscribe first Dec 06 20:47:14 aha, seems I was subscribed after all but my account was bouncingn for some reason Dec 06 20:47:36 Jin^eLD: let me know if you need help fixing it Dec 06 20:48:25 Crofton: thank you! I "unbounced" it via the provided link, lets see Dec 06 20:48:32 ok Dec 06 20:49:35 there is no "forgot password" link? I am here https://lists.yoctoproject.org/login Dec 06 20:50:07 hmm Dec 06 20:50:26 I ahd it send me a link to login via email Dec 06 20:50:36 groups can be funny to login to Dec 06 20:50:49 ok thats weird, it says it does not know my name, but it still offered to unboubce me? Dec 06 20:50:51 We may need halstead to help if you can't get back in Dec 06 20:50:52 Jin^eLD: I took the time to add ripgrep using : bitbake -c update_crates . It failed silently at first but once I added S = "${WORKDIR}/git" it worked fine. Dec 06 20:51:21 Identical dependencies to what 'cargo bitbake' produced as you'd expect. Dec 06 20:51:32 Crofton: I'll just register anew Dec 06 20:51:44 I can help if needed. Dec 06 20:52:59 vmeson: it first kept failing for me with a libudev related error, I figured out that I had to add DEPENDS = "eudev", still failed, figured out had to also inherit pkgconfig, then it kind of worked - the proggy I am trying to compile failed with a rust error, the guy who wrote it says it works fine for him when he does "cargo build --target=armv7-unknown-linux-musleabihf release" locally in a non Yocto Dec 06 20:53:01 env Dec 06 20:54:57 Jin^eLD: are you both using the same version of rust? What's the package / error? Dec 06 20:55:10 vmeson: seems he is building for musl and I am using glibc Dec 06 20:55:14 I wonder why this is a problem though Dec 06 20:55:23 the rust version should be fine, we compared Dec 06 20:56:00 Jin^eLD: I'm not sure and can't help without seeing what the error is about. Dec 06 20:56:40 the error is `rustc --explain E0507` Dec 06 20:56:48 but I let him debug it now as I have no idea of rust Dec 06 20:56:55 kk, checking. Dec 06 20:57:59 but it seems to be toolchain related as the same code works for him, just on a different toolchain Dec 06 20:58:02 and not within Yocto Dec 06 20:58:09 Jin^eLD: A borrowed value was moved out. -- I haven't had the pleasure of wrestling with the borrow checker yet ! ;-) Dec 06 20:59:29 Jin^eLD: if you plan to share the recipe, and we think this is a problem in the rust / yocto-gcc toolhchain, then I *may* be able to have someone take a look. Dec 06 20:59:38 I have no clue what is happening there as the only rust I know is the one I am fighting on my car, hail WD40 ;) Dec 06 21:00:30 vmeson: thank you, let me have my colleague debug it first before you spend time on it, at least to a point where we have more clues if its Yocto related at all Dec 06 21:01:14 Jin^eLD: I see a link labeled, "Forgot your password, or don't have one yet?" onhttps://lists.yoctoproject.org/login. Can you see it? Dec 06 21:01:28 Jin^eLD: sure, that makes sense. If you decide you need help, please open a defect in bugzilla if the issue is with oe-core. A reproducer would be ideal even if that's not yet merged. Dec 06 21:01:46 halstead: it seems to know my e-mail address, saying I am already registered when I tried to register anew, at the same time it does not let me in with my saved credentials Dec 06 21:01:56 halstead: I mst be blind... let me disable the adblocker Dec 06 21:03:27 halstead: https://snipboard.io/vqfQ20.jpg Dec 06 21:04:03 Jin^eLD: That's after you visit https://lists.yoctoproject.org/sendloginlink and enter your email? Dec 06 21:04:57 ehm no, I thought "send link" is one of many options to log in, I assumed I can just log in using my mail and password? or is the "sendloginlink" the "forgot password thing"? Dec 06 21:05:01 it has a telegram icon Dec 06 21:05:14 so I was assuming it was TG, alongside facebook and google Dec 06 21:05:36 vmeson: ack! Dec 06 21:06:23 Jin^eLD: When I visit https://lists.yoctoproject.org/login I see a link labeled ""Forgot your password, or don't have one yet?"" that leads to https://lists.yoctoproject.org/sendloginlink which has a form to send a login link to my email. I can use that to log in and set a password. Dec 06 21:06:56 halstead: oooh, that is visible _before_ entering anything Dec 06 21:06:58 not after Dec 06 21:07:16 which imho makes less sense :) because I know that I forgot the password only after getting the error response that my pw is bad ;) Dec 06 21:07:19 Jin^eLD: Ah... perhaps I can make a suggestion to their UI team. Dec 06 21:07:45 so now that I repasted https://lists.yoctoproject.org/login into the browser to get into the "entry" state to that page, then I saw the link Dec 06 21:08:42 not sure why they decided to hide the "forgot pw" link if your login failed... so that'd be worth improving Dec 06 21:14:32 vmeson: I managed to reactivate my account, my e-mail is jin@mediatomb.cc Dec 06 21:14:49 so if you could loop me in on the thread, that'd be great Dec 06 21:18:29 Jin^eLD: sure, I'll get to that in a bit. Dec 06 21:19:01 thx Dec 06 22:25:43 vmeson: was some problem in the code, not a Yocto issue, so overall it worked (including the backport of the recipe into kirkstone) Dec 06 22:51:01 Jin^eLD: ah, good to know. Dec 06 22:51:33 I'll reply to the thread just to let people know that the new approach works and is being used. Dec 06 22:52:14 sure, I'll post my steps and info on the problems I had while trying it out **** ENDING LOGGING AT Wed Dec 07 02:59:56 2022