**** BEGIN LOGGING AT Fri Apr 17 02:59:57 2020 Apr 17 04:34:09 Crofton|road, we try Apr 17 13:07:23 hmm, it seems that I fail to understand how to use the NO_GENERIC_LICENSE variable, I tried something like this (also various combinations with $S, without it, just the .c file) but I still get the "No generic license" warning when building Apr 17 13:07:28 LICENSE = "SLA0044" Apr 17 13:07:30 LIC_FILES_CHKSUM = "file://${S}/src/main.c;beginline=7;endline=16;md5=22739fe427a325307a9b52ae6ac61197" Apr 17 13:07:32 NO_GENERIC_LICENSE[SLA0044] = "${S}/src/main.c" Apr 17 13:07:34 what am I am missing? Apr 17 14:11:00 why don't you just add SLA0044 to licenses in your layer? Apr 17 14:19:59 well, it's just a comment in a source file... Apr 17 14:20:07 I could copy it out I guess, but seemed like too much hassle? Apr 17 14:21:34 I package an example program from ST, no extra license file was supplied, just a header comment Apr 17 15:00:13 I'm still running Zeus, but I'm finding that some things (nativesdk is where I'm focused) are not producing archiver sources.. anyone know why (or if there might be a fix in a newer version? Apr 17 15:14:08 Ohh I think it's copyleft_should_include.. Apr 17 15:14:18 but I see no way to say, no really I want everything.. Apr 17 15:16:47 COPYLEFT_LICENSE_INCLUDE = "*" ..... Apr 17 18:15:09 JPEW: Jin^eLD: it's friday, so it's rockchip day ;-) Apr 17 18:21:37 tlwoerner: you focus on rockchips exclusively on fridays or how come? :) Apr 17 18:22:22 $DAYJOB and $WORK-LIFE-BLANACE occupy the rest of the week ;-) Apr 17 18:23:06 i'm just kidding, i don't explicitly set aside friday afternoons for rockchip, but it was a decent week so i'm taking a little break Apr 17 18:23:26 ah :) Apr 17 18:23:43 Jin^eLD: and i'm curious to look at what you sent regarding the tinker and setting the clock Apr 17 18:24:48 that was really some weird stuff, but since I am really a userspace habitant the lowlevel stuff was somewhat over my head Apr 17 18:25:21 I was happy I got it to work at all without knowing which patch it was... Apr 17 18:26:07 I tried taking them out, but some build on top of others and there were too many combinations, so I just gave up and took all of them Apr 17 18:26:44 it'd be great to figure out what makes the difference and mainline it... Apr 17 18:29:06 Jin^eLD: i'm curious to know how you're getting on with that A/B update stuff you were working on, any chance you've blogged about it, or have a repo somewhere? Apr 17 18:30:58 I did it as a job for friends of mine, I do not think they finished anything or published anything yet, but I finished the basic thing, i.e. you had a script to flash a downloaded image into A or B (depending what was currently booted) Apr 17 18:31:17 rest would be application stuff, download, decide when, blah Apr 17 18:31:27 but I tell you what I did Apr 17 18:31:42 I also replicated that on other hardware later on Apr 17 18:32:10 only thing you really need is working u-boot with environment that can be saved and also accessed/modified from user space Apr 17 18:32:30 and a few simple scripts, most workload is done by swupdate anyway Apr 17 18:33:09 my only blog is about my lathe I think :) oh and one about my car restoration, but I do not have an IT blog, sorry Apr 17 18:34:40 *I can tell you what i did (is what I wanted to write above, missed the "can" doh) Apr 17 18:35:30 where is the u-boot environment saved? flash? filesystem? Apr 17 18:35:52 one thing that's still worth adding is some u-boot script for detecting when a partition died without having been reflashed; i.e. if you have not been updating but your rootA blew up, you could in theory still boot rootB, but you'd need some additional u-boot logic to notice that "current root" does not boot Apr 17 18:36:24 tlwoerner: for the rockchip? flash... and I did submit a patch for fw_env config :> you did not merge it ;( Apr 17 18:36:57 oops, good thing it's rockchip-friday Apr 17 18:37:00 I consider savint u-boot env in the fs unsafe and if you wanted to do it like that you'd still need an own partition for it, because you are reflashing rootA/rootB but you want the env to survive Apr 17 18:37:59 on the tinker specifically u-boot env is in the emmc directly in some of their custom boot partitions or whatever, I did not have to change anything there Apr 17 18:38:14 I only had to add an fw_env.config file to tell userpsace where the environmen tis Apr 17 18:38:22 which is what my patch was all about... Apr 17 18:39:20 on which hw would you want to try the A/B thing? Apr 17 18:40:10 an ancient piece of armv5 hardware, but if you have it working on tinker i could use that to experiment on Apr 17 18:40:52 I think as long as you have enough space and a working u-boot environment you could probably use any hardware Apr 17 18:42:51 also check out: https://github.com/sbabic/meta-swupdate-boards/blob/master/recipes-extended/images/update-image/raspberrypi3/sw-description Apr 17 18:43:04 basically thats an example of the swupdate sw-description file for the image Apr 17 18:43:38 I did not use that lua script though Apr 17 18:43:59 lol, i was just rolling my eyes when i saw that Apr 17 18:44:10 but once you get the idea it's generally quite straightforward Apr 17 18:45:08 I did use an embedded lua script to parse /proc/cmddline Apr 17 18:45:17 that's how I figure out what root partition is booted Apr 17 18:45:29 I parse out the root=/dev/mmc... parameter Apr 17 18:46:05 and then you can check if the script that called swupdate and told it where to flash the stuff did not mess up i.e. you can make sure you're not trying to flash the currently booted partition Apr 17 18:46:56 but thats just an additional safeguard, if you trust your shellscript that tells swupate which image to flash, you can avoid lua completely I Think Apr 17 20:02:22 tlwoerner: Cool :) Apr 17 20:05:28 JPEW: the SRC_URI_append_rk3288 casts too wide a net, since it also includes firefly and vyasa, but there's no harm in patching a dts file for another machine, so i'll apply without worrying about it Apr 17 20:07:44 JPEW: i'll add another patch to add compatible machines for the other yocto kernel recipes too: linux-yocto-dev, linux-yocto-rt, linux-yocto-tiny Apr 17 20:14:00 but other than that, i like it :-) i think it's better than what we had before **** ENDING LOGGING AT Sat Apr 18 02:59:57 2020