**** BEGIN LOGGING AT Mon Mar 11 02:59:57 2019
Mar 11 07:41:42 <Tofe> Morning !
Mar 11 07:57:01 <Tofe> JaMa: ping
Mar 11 08:00:23 <Tofe> bshah: I wonder, do you setup a usb network in initrd, for debug purposes ? Where would be that part of the initrd ? I'm struggling a bit with mine, tbh :)
Mar 11 08:00:47 <Tofe> (I'm talking about mainline hammerhead, again )
Mar 11 08:00:57 <bshah> Tofe: it uses normal functionfs
Mar 11 08:01:08 <Tofe> oh, ok
Mar 11 08:01:32 <Tofe> I tried to use the same way as flto, but it's a neverending story with the qmi utilities
Mar 11 08:11:33 <Tofe> bshah: it's smart, the way it's done: always start telnetd, and if initrd goes well kill telnetd before starting systemd
Mar 11 08:11:43 <Tofe> i'll copy that :p
Mar 11 08:31:30 <JaMa> Tofe: pong
Mar 11 08:31:46 <JaMa> Tofe: separate MACHINE looks reasonable, but it might be easier to just test it in branch
Mar 11 08:32:00 <JaMa> and when it works, merge it and forget about the libhybris version
Mar 11 08:32:27 <Tofe> JaMa: ah, thanks, that was exactly the topic I wanted to talk about
Mar 11 08:32:45 <JaMa> otherwise hammerhead-mainline will end in sstate, some bbappends etc and we'll probably want to clean that and rename back to hammerhead once it's the preferred version of it
Mar 11 08:33:31 <Tofe> JaMa: ok, let's try the "branch" way; that's what I'm doing right now anyway
Mar 11 08:34:21 <Tofe> and it's true that 3.4 kernels are going to be EOL very soon, so, let's prepare the future
Mar 11 08:34:40 <Tofe> (well they've been EOL for several years already :p )
Mar 11 08:35:20 <Tofe> https://gitlab.com/postmarketOS/pmaports/blob/7df2c93459424af257d673f055e5d86c6e6db46c/main/postmarketos-mkinitfs/init_functions.sh#L204 <-- I'll just drop that here, I'll try to integrate that into our init later
Mar 11 08:45:51 <Tofe> Herrie|Laptop: o/
Mar 11 10:11:40 <Herrie|Laptop> Morning!
Mar 11 10:11:45 <Herrie|Laptop> Tofe: Ah nice stuff :)
Mar 11 10:15:15 <Tofe> Herrie|Laptop: yes, I find their script nice and well organized
Mar 11 10:17:22 <Herrie|Laptop> There are more useful things in there it seems :)
Mar 11 10:23:10 <Herrie|Laptop> bshah: Can you merge this? It's mainly cosmetics and been outstanding for long time: https://github.com/Halium/android/pull/30/files
Mar 11 10:26:00 <bshah> I will nominate you for fixing any merge conflicts you create from this :P
Mar 11 10:29:18 <Herrie|Laptop> bshah: There's only 1 PR pending, I think I can deal with that
Mar 11 16:03:08 <saidinesh5> ookay this is interesting.. Today I found out that the efi binary from intel-meta is definitely broken and isn't even booting. So i went back to create a franken image with debian netinstaller and now it is stuck in the initrd trying to loop mount the root
Mar 11 17:19:41 <saidinesh5> WebOS on cherrytrail https://imgur.com/a/Fr7KDTz
Mar 11 17:19:55 <saidinesh5> Not much but hey, it's a start!
Mar 11 17:21:10 <Herrie|Laptop> saidinesh5: That's not very helpful about meta-intel ;)
Mar 11 17:21:41 <saidinesh5> Herrie|Laptop: how much of the backlog do you have?
Mar 11 17:22:07 <saidinesh5> so basically the efi stuff from meta-intel seems to be broken and just hangs the device.. and i am not even sure how to get about debugging THAT
Mar 11 17:22:21 <Herrie|Laptop> Well frankenstuff works ;)
Mar 11 17:22:34 <Herrie|Laptop> saidinesh5: Shouldn't be too long LOL
Mar 11 17:22:36 <saidinesh5> so I created a franken USB stick with efi + grub from debian net installer and try to chroot / switchroot from the shell
Mar 11 17:22:41 <Herrie|Laptop> 1 line in log LOL
Mar 11 17:22:45 <saidinesh5> ahh
Mar 11 17:22:52 <saidinesh5> so you've seen the pictures?
Mar 11 17:22:54 <Herrie|Laptop> Logged in and you posted something new ;)
Mar 11 17:22:57 <Herrie|Laptop> Yes
Mar 11 17:23:00 <saidinesh5> yep
Mar 11 17:23:03 <Herrie|Laptop> Volatile is suspicious?
Mar 11 17:23:14 <Herrie|Laptop> Could be defconfig related, not sure
Mar 11 17:23:54 <saidinesh5> one of the biggest things i hate about systemd - binary logs
Mar 11 17:24:08 <saidinesh5> and yeah not sure how to get these logs .. let me look
Mar 11 17:25:01 <Herrie|Laptop> SNTP = normal, don't worry about that one
Mar 11 17:25:26 <saidinesh5> okay logs are gone ... because /var/log -> /run/log
Mar 11 17:26:35 <saidinesh5> is it worth trying rocko or some new branch from meta-intel?
Mar 11 17:26:54 <saidinesh5> maybe that would hopefully spit out working efi images...
Mar 11 17:27:00 <Herrie|Laptop> We switched to Sumo last week
Mar 11 17:27:04 <saidinesh5> Ahh
Mar 11 17:27:16 <Herrie|Laptop> After fixing the 3.4 kernels with a single line patch
Mar 11 17:27:23 <saidinesh5> Oooo
Mar 11 17:27:30 <Herrie|Laptop> To fix systemd 233+ to behave ;)
Mar 11 17:27:43 <saidinesh5> do you have any idea about efi stuff?
Mar 11 17:27:57 <Herrie|Laptop> Not really sorry :(
Mar 11 17:28:00 <saidinesh5> Ahh
Mar 11 17:28:07 <Herrie|Laptop> Trial & error I'd say
Mar 11 17:28:16 <saidinesh5> yep. that's te current path lol
Mar 11 17:28:17 <Herrie|Laptop> Intel is a big sponsor of Yocto
Mar 11 17:28:33 <Herrie|Laptop> Something crucial like this should've been back ported
Mar 11 17:28:42 <Herrie|Laptop> Or maybe it was and you need to set some flags etc
Mar 11 17:28:46 <Herrie|Laptop> Hard to say
Mar 11 17:29:08 <saidinesh5> hmm.. Yep. but i tried booting that iso image even on my desktop and it got stuck the same way
Mar 11 17:29:16 <saidinesh5> so i am guessing the efi stuff may be broken...
Mar 11 17:29:47 <saidinesh5> i know that efi expects a fat32  partition with the folder structure EFI/<osname>/filename.efi
Mar 11 17:30:03 <saidinesh5> but not sure what those binaries do..or how they do it
Mar 11 17:30:05 <saidinesh5> to load grub
Mar 11 17:30:55 <Herrie|Laptop> Not sure if this is related or of any use, just came across it when Googling something: https://github.com/intel/kernelflinger/issues/4
Mar 11 17:33:34 <saidinesh5> unlikely.. i mean i don't even have any android installation on this one
Mar 11 17:34:23 <saidinesh5> will try out the sumo build i think
Mar 11 17:37:42 <saidinesh5> Out of curiosity, how come you guys went for this Makefile + update xyz approach instead of just using a git repo for everything?
Mar 11 17:39:05 <saidinesh5> Aand how do i change the branch?
Mar 11 17:39:07 <saidinesh5> to sumo?
Mar 11 17:40:16 <Herrie|Laptop> saidinesh5: Well historically. I wasn't there when this was decided
Mar 11 17:40:23 <Herrie|Laptop> JaMa might tell you more ;)
Mar 11 17:40:26 <saidinesh5> Ahh
Mar 11 17:40:36 <Herrie|Laptop> He's been around for long, also from within HP/LG ;)
Mar 11 17:40:42 <Herrie|Laptop> Even back to Palm days I think
Mar 11 17:40:51 <saidinesh5> O_O
Mar 11 17:41:47 <Herrie|Laptop> This now points to sumo: https://github.com/webOS-ports/webos-ports-setup/blob/testing/conf/layers.txt
Mar 11 17:42:15 <Herrie|Laptop> Simply doing: make update-conffiles && make update
Mar 11 17:42:21 <Herrie|Laptop> Should help to update your layers
Mar 11 17:42:30 <Herrie|Laptop> Expect a long rebuilding time ;)
Mar 11 17:43:06 <Herrie|Laptop> Since systemd and glibc were updated pretty much everything will be rebuild ;)
Mar 11 17:44:25 <saidinesh5> can't update-conffiles because my bblayers and layers had some intel stuff
Mar 11 17:44:27 <saidinesh5> but yeah
Mar 11 17:44:29 <saidinesh5> updated
Mar 11 17:45:54 <saidinesh5> would be nice to just do a git clone webos-ports-env and git checkout <pyro,sumo,etc..>
Mar 11 17:46:05 <saidinesh5> that way git diff would also nicely show the differences in conffiles
Mar 11 17:47:10 <saidinesh5> https://bpaste.net/show/9a20d3ab13c9
Mar 11 17:47:13 <saidinesh5> looks decent
Mar 11 17:52:34 <saidinesh5> --- Starting bitbake server pid 7808 at 2019-03-11 17:51:15.921353 ---
Mar 11 17:52:34 <saidinesh5> ERROR: Layer directory '/src/LuneOS/webos-ports-env/webos-ports/meta-openembedded/meta-systemd' does not exist! Please check BBLAYERS in /src/LuneOS/webos-ports-env/webos-ports/conf/bblayers.conf
Mar 11 17:52:37 <saidinesh5> hmm...
Mar 11 17:52:40 <saidinesh5> make
Mar 11 17:56:17 <saidinesh5> Herrie|Laptop: ^ any idea?
Mar 11 18:16:40 <JaMa> saidinesh5: remove it from bblayers.conf
Mar 11 18:17:17 <JaMa> saidinesh5: everything _is_ git repo, the makefile stuff is in "common" directory
Mar 11 18:17:43 <JaMa> you can go there and checkout different branch and do make update to checkout right revisions of all included layers
Mar 11 18:17:56 <saidinesh5> JaMa oh? But systemd seems important.... So are you guys changing init systems?
Mar 11 18:18:00 <JaMa> if you're using e.g. testing, then it will be switched from pyro to sumo automatically
Mar 11 18:18:19 <saidinesh5> Ah.  I was using testing but it didn't change
Mar 11 18:18:45 <JaMa> saidinesh5: meta-systemd doesn't exist anymore, because first systemd itself was merged to oe-core
Mar 11 18:18:45 <JaMa> then meta-systemd had some bbappends for backward compatibility
Mar 11 18:18:45 <saidinesh5> Oh interesting
Mar 11 18:18:48 <JaMa> and now it wasn't needed anymore, so it was removed completely
Mar 11 18:19:03 <saidinesh5> Aye 1 sec.. removing it and starting
Mar 11 18:22:40 <saidinesh5> JaMa: WARNING: Layer intel-luneos should set LAYERSERIES_COMPAT_intel-luneos in its conf/layer.conf file to list the core layer names it is compatible with.
Mar 11 18:22:44 <saidinesh5> anything i should do about that?
Mar 11 18:24:19 <saidinesh5> Ahh nevermind
Mar 11 18:24:25 <saidinesh5> i wasn't building intel-luneos
Mar 11 18:29:18 <JaMa> you'll need to update conf/layer.conf
Mar 11 18:29:58 <JaMa> see "meta-*: add LAYERSERIES_COMPAT to layer.conf files" commit in meta-webos-ports
Mar 11 18:32:18 <saidinesh5> Yeah i wasn't using that one at all..  was just midway through creating a meta-intel-luneos to customise things
Mar 11 18:32:39 <saidinesh5> But for now just using a MACHINE directly from meta-intel
Mar 11 19:18:05 <Tofe> saidinesh5: ooh great work here :)
Mar 11 19:18:17 <Tofe> I see we have almost the same screenshots :D
Mar 11 19:18:26 <Tofe> (me with hammerhead mainline)
Mar 11 19:18:27 <saidinesh5> O_O same screenshots?
Mar 11 19:18:29 <saidinesh5> Ahh
Mar 11 19:18:30 <saidinesh5> lol
Mar 11 19:18:41 <saidinesh5> i think you have a tougher battle ahead of you
Mar 11 19:19:27 <saidinesh5> WARNING: pkgconfig-native-0.29.2+gitAUTOINC+edf8e6f0ea-r0 do_fetch: Failed to fetch URL git://anongit.freedesktop.org/pkg-config, attempting MIRRORS if available
Mar 11 19:19:29 <saidinesh5> hmm..
Mar 11 19:20:43 <Tofe> saidinesh5: we'll see :p
Mar 11 19:20:58 <saidinesh5> you're using freedreno drivers?
Mar 11 19:21:16 <Tofe> no, just regular linux 4.17 + mesa
Mar 11 19:21:25 <Tofe> so it's bare framebuffer right now
Mar 11 19:21:32 <saidinesh5> Ahh..
Mar 11 19:21:41 <saidinesh5> didn't know it even worked that way!
Mar 11 19:21:59 <Tofe> it's not optimized at all, but at least it'll show something
Mar 11 19:22:58 <Tofe> for me what's interesting is to have one of our target devices using a near-mainline kernel, and a full regular glibc stack
Mar 11 19:23:14 <saidinesh5> Ahhh.. so it is a competition between you and me i see .. lol
Mar 11 19:23:36 <Tofe> :)
Mar 11 19:23:54 <saidinesh5> ihave no idea why the efi binary from meta-intel is so broke though.. no way to even debug it
Mar 11 19:23:58 <Tofe> do you have a remote access to your booting device ? ssh, adb, telnet ... ?
Mar 11 19:24:12 <saidinesh5> well i have keyboard access.. but no remote access
Mar 11 19:24:21 <saidinesh5> i think i can trick it into some usb networking
Mar 11 19:24:34 <Tofe> ok, you're one step ahead -- but I'm working on it
Mar 11 19:24:57 <saidinesh5> compiling sumo from scratch .. so my build will take time again
Mar 11 19:25:17 <saidinesh5> btw. the number of tasks reduced for sumo?
Mar 11 19:25:32 <saidinesh5> 434 out of 6294 ..
Mar 11 19:25:41 <saidinesh5> for pyro iirc it was 7000 +
Mar 11 19:26:24 <Tofe> mmmh maybe, I didn't look carefully
Mar 11 19:27:45 <saidinesh5> I think i will have to create  meta-intel-luneos anyway though.. and remove half the stuff this tablet doesn't need/can't use
Mar 11 19:27:58 <saidinesh5> like the telephony for eg.
Mar 11 19:30:08 <Tofe> But if you define the image yourself, you get to choose what to include
Mar 11 19:30:32 <saidinesh5> Yep and that has to go into my own meta-intel-luneos or something right?
Mar 11 19:30:37 <saidinesh5> like meta-intel is straight from upstream
Mar 11 19:31:13 <saidinesh5> and this is what i am using: MACHINE=intel-corei7-64 bb luneos-dev-package
Mar 11 23:29:37 <elvispre> So am I the only one left with a mako as well now?
Mar 11 23:31:06 <elvispre> Flashed mako, onyx and tissot with local builds from the current (sumo) layers just now.
Mar 11 23:31:35 <elvispre> mako boots but behaves "strangely" :)
Mar 11 23:32:01 <elvispre> onyx does not appear to boot, but I haven't made that patch to the kernel yet so that is probably expected.
Mar 11 23:32:40 <elvispre> tissot boots quite nicely. No wifi though even after a second boot.
Mar 11 23:32:58 <elvispre> Good to be on sumo!
**** ENDING LOGGING AT Tue Mar 12 02:59:57 2019