**** BEGIN LOGGING AT Tue Mar 05 02:59:57 2019
Mar 05 07:13:59 <Herrie|Laptop> Morning!
Mar 05 07:17:21 <Herrie|Laptop> Tofe: Seems I have some issues with exiv2 on Sumo. exiv2.org is down. I can get 0.26 from GitHub, however the do_install for man then fails. I can remove it, since it seems to be documentation mainly, it might be good to compare with 0.25 archive to see what changed and maybe we can keep man?
Mar 05 07:17:48 <Herrie|Laptop> https://github.com/webOS-ports/meta-webos-ports/blob/pyro/meta-luneos/recipes-multimedia/exiv2/exiv2_0.25.bb#L15
Mar 05 07:21:40 <Herrie|Laptop> Ah seems 0.25 is available on http://exiv2.dyndns.org/releases/exiv2-0.25.tar.gz
Mar 05 07:25:02 <Tofe> Morning!
Mar 05 07:28:02 <Tofe> Herrie|Laptop: I'm building sumo right now, so I guess I'll soon have the same issue... Though I seem to already have 0.25 in my downloads
Mar 05 07:28:35 <Herrie|Laptop> Tofe: Well I created a new folder for sumo to start from scratch
Mar 05 07:28:42 <Herrie|Laptop> So pyro and it's downloads are in another folder
Mar 05 07:28:51 <Herrie|Laptop> Download speed isn't the issue here :P
Mar 05 07:28:59 <Herrie|Laptop> So don't care too much about reusing it ;)
Mar 05 07:29:02 <Herrie|Laptop> Except for disk space
Mar 05 07:31:58 <Herrie|Laptop> Ooh they have good and detailed changelogs, that's helpful :) http://exiv2.dyndns.org/changelog.html#v0.26
Mar 05 07:32:13 <Tofe> Herrie|Laptop: I'm always on the same SSD, Halium + LuneOS, so I try to be careful :)
Mar 05 07:32:48 <Herrie|Laptop> Tofe: Well here too, 1TB so it's OK for a while ;)
Mar 05 07:32:48 <Tofe> wow, yes, impressive changelog quality
Mar 05 08:37:33 <Herrie|Laptop> OK flashing sumo to Hammerhead now
Mar 05 08:38:02 <Tofe> ah, interesting :)
Mar 05 08:39:33 <Herrie|Laptop> Seems bootloop
Mar 05 08:40:15 <Herrie|Laptop> I didn't merge your PR in my build
Mar 05 08:40:21 <Herrie|Laptop> But for booting that shouldn't matter normally
Mar 05 08:41:11 <Herrie|Laptop> https://bpaste.net/show/8185e8352120
Mar 05 08:41:17 <Herrie|Laptop> [    1.777557] initrd: PANIC for reason: Couldn't find data partition.
Mar 05 08:41:45 <Tofe> ok, you'll need gcc 7 patch from JaMa
Mar 05 08:42:33 <Tofe> which is currently only applied on our tissot kernel
Mar 05 08:44:19 <Herrie|Laptop> Tofe: OK
Mar 05 08:44:45 <Tofe> https://github.com/Tofee/tissot/commit/a8412322dc34a81517cfdb9f4ef1c75bcf263ad0
Mar 05 08:44:51 <Herrie|Laptop> Well original boeffla patch has a lot more flags, but they're on 3.0 kernel. https://github.com/zanezam/boeffla-kernel-cm-n51x0/commit/228cc26
Mar 05 08:44:56 <Herrie|Laptop> I can try with this one first
Mar 05 08:46:27 <Tofe> seems like our commit is a further filtering from boeffla's patch, and it proved working wekk on tissot
Mar 05 08:46:29 <Tofe> well*
Mar 05 08:48:45 <Herrie|Laptop> Well Tissot is on 3.18 ;)
Mar 05 08:49:49 <Tofe> ah, yes, true
Mar 05 08:50:34 <Herrie|Laptop> Tofe: Seems we're missing https://github.com/Tofee/tissot/blob/a8412322dc34a81517cfdb9f4ef1c75bcf263ad0/Makefile#L624 as well
Mar 05 08:50:38 <Herrie|Laptop> On Hammerhead
Mar 05 08:50:41 <Herrie|Laptop> Should I add it too?
Mar 05 08:52:23 <Herrie|Laptop> Details on that change here: https://github.com/Tofee/tissot/commit/69102311a57d1fd65cdc4002c55c5d551c799044
Mar 05 08:55:14 <Herrie|Laptop> I guess it doesn't hurt to have
Mar 05 08:57:36 <Tofe> yes, let's have it too
Mar 05 08:59:01 <Herrie|Laptop> While I'm at it, I'll add https://github.com/Tofee/tissot/commit/1873e870fd63ee4b87dbe0125ca373e420fb4987 as well
Mar 05 08:59:13 <Herrie|Laptop> Doesn;t hurt to have for debugging which we might need with these old kernels :P
Mar 05 09:02:06 <Tofe> ok; though I hope we'll naver come to need that...
Mar 05 09:03:23 <Herrie|Laptop> Tofe: Yeah ASM is never fun LOL
Mar 05 09:05:52 <Herrie|Laptop> I can just flash the boot.img with fastboot flash boot boot.img right?
Mar 05 09:05:58 <Herrie|Laptop> That should suffice?
Mar 05 09:06:19 <Tofe> right
Mar 05 09:06:38 <Herrie|Laptop> OK building now :)
Mar 05 09:17:03 <Herrie|Laptop> Tofe: Seems same issue
Mar 05 09:17:30 <Herrie|Laptop> https://bpaste.net/show/6726ab7ea5ce
Mar 05 09:17:44 <Herrie|Laptop> I can try to downgrade systemd to pyro version just to see if that solves it
Mar 05 09:17:52 <Herrie|Laptop> Then we have some idea where to look maybe
Mar 05 09:20:18 <Tofe> Herrie|Laptop: spare the hassle, systemd isn't even started at this point
Mar 05 09:20:37 <Herrie|Laptop> Tofe: OK
Mar 05 09:20:41 <Herrie|Laptop> Any other suggestions?
Mar 05 09:21:41 <Tofe> Getting panic adb to work would help, of course; otherwise it'll be printf in halium script
Mar 05 09:21:58 <Tofe> like the output of find /dev -name "userdata"
Mar 05 09:24:10 <Tofe> One of the only things that has changed here are: 1) compiler, and 2) busybox
Mar 05 09:24:18 <Tofe> and udev, maybe
Mar 05 09:24:53 <Tofe> ok udev is now part of systemd, you got a point :)
Mar 05 09:32:18 <Herrie|Laptop> So should try older systemd for now
Mar 05 09:32:19 <Herrie|Laptop> ?
Mar 05 09:33:44 <Tofe> mmmh... well, either that, or solve the "mounting adb on /dev/usb-ffs/adb failed: No such device" issue
Mar 05 09:35:41 <Tofe> functionfs and adb are enabled in hammerhead kernel, right?
Mar 05 09:38:17 <Herrie|Laptop> Yes they should be
Mar 05 09:40:18 <Tofe> I wonder why the adb functionfs mount fails, then
Mar 05 09:46:33 <Herrie|Laptop> I suspect udev issues somehow
Mar 05 09:48:17 <Tofe> possibly
Mar 05 09:57:41 <Tofe> https://groups.google.com/forum/#!topic/android-kernel/EDPxbSze-6Q do we put this "g_ffs.idVendor=0x18d1" in the boot params ?
Mar 05 10:05:50 <Herrie|Laptop> Seems to half work. ADB seems to do somethign but cannot shell in: https://bpaste.net/show/5237d78a9c01
Mar 05 11:40:50 <saidinesh5> Herrie|Laptop: btw. the namespace patches to 3.4 were a headache and you are better off not going down that path
Mar 05 11:41:32 <saidinesh5> like even if you found a namespace patch you like.. some filesystem you use (like overlayfs or squashfs etc..) may not support the namespaces and you see weird issues
Mar 05 11:41:53 <saidinesh5> and whenyou add a patch for that,.. it requires another patch.. and breaks without even a compilation error at times
Mar 05 11:42:16 <saidinesh5> you are better off porting the 3.10 kernel for your device (that'swhat they did for my cancro) ..
Mar 05 11:44:07 <saidinesh5> iirc the hammerhead even boots and has graphics with upstream kernel these days right?
Mar 05 11:46:03 <Tofe> saidinesh5: but that won't support Halium, which means a whole lot of things not working... including rild
Mar 05 11:46:51 <saidinesh5> Ahh I see.. the 3.10 kernel maybe  your best bet. that guy took like a weekend to get 3.10 kernel up and running on cancro.. iirc there was a 3.10 kernel even for hammerhead
Mar 05 11:48:44 <saidinesh5> https://github.com/vasishath/kernel_cancro
Mar 05 11:48:52 <saidinesh5> this is for cancro btw.
Mar 05 11:49:13 <Tofe> https://masneyb.github.io/nexus-5-upstream/ this seems to be the current status of mainline hammerhead support
Mar 05 11:50:47 <saidinesh5> yeah iirc that's probably the one that showed up in my rss feed a couple of days ago
Mar 05 11:50:56 <saidinesh5> about how 5.1 kernel may be able to boot hammerhead
Mar 05 11:51:44 <Tofe> even 4.9 might boot it
Mar 05 11:52:08 <saidinesh5> it did.. but without graphics iirc
Mar 05 11:52:55 <saidinesh5> but backporting the namespace patches was a nightmare
Mar 05 11:53:13 <saidinesh5> for sfdroid we needed squashfs and overlayfs with namespace support
Mar 05 11:53:17 <Tofe> :) we'll try to avoid that
Mar 05 11:53:45 <saidinesh5> and each patch kept giving more compiler errors and each compiler error kind of demanded another patch
Mar 05 11:53:49 <saidinesh5> yeah
Mar 05 11:53:57 <Tofe> but here it might just be a udev issue with gcc 7.3
Mar 05 11:54:33 <saidinesh5> what's the issue?
Mar 05 11:55:04 <Tofe> it doesn't find the "userdata" partition
Mar 05 11:55:09 <Tofe> in initrd
Mar 05 11:55:35 <Tofe> but dmesg shows that partition n°28 is there
Mar 05 11:55:57 <Tofe> so the name mapping might be incomplete somehow
Mar 05 11:56:04 <saidinesh5> Ahh I see
Mar 05 11:56:14 <saidinesh5> you run udev in initrd?
Mar 05 11:56:18 <Tofe> yes
Mar 05 11:56:39 <saidinesh5> Ahh.. interesting.. for sailfish we used just hardcoded partition names
Mar 05 11:57:04 <Tofe> we did that before, but Halium brought a generalized boot script
Mar 05 11:57:21 <saidinesh5> ahh.. wow.. one year changes a lot of things...
Mar 05 11:57:28 <Tofe> hehe :)
Mar 05 11:58:17 <saidinesh5> so booting halium = just makign the kernel config changes for systemd these days?
Mar 05 11:58:22 <Tofe> still, one thing that didn't change, is that our kernels are far too old for the current userspace :p
Mar 05 11:58:44 <saidinesh5> amen.. even for systemd update on sailfish 2.x i had to update the 3.4 kernel with2 more patches
Mar 05 11:59:12 <Tofe> halium also means using LXC
Mar 05 11:59:45 <Tofe> and iirc, Halium doesn't (yet?) require systemd itself, though everyone does that
Mar 05 12:00:07 <saidinesh5> it was using LXC even before no? iirc the usernamespace was broken on 3.4 devices
Mar 05 12:00:46 <Tofe> yes, LXC has always been there
Mar 05 12:01:33 <saidinesh5> the namespace support on 3.4 kernels was very meh .. like some configs show that they are supported but when you enable them it behaves weird.. like the devpts iirc
Mar 05 12:03:37 <Tofe> yes, the implementations were quite young and only partially tested
Mar 05 12:07:38 <saidinesh5> i am surprised your hammerhead phones still function though
Mar 05 12:07:50 <saidinesh5> my cancro battery had puffed and broke the screen
Mar 05 12:08:06 <saidinesh5> well it bent the screen and when pulling out the battery i broke the screen ..
Mar 05 12:08:53 <Tofe> Well I replaced the battery once
Mar 05 12:09:05 <saidinesh5> what about the wearing out emmc?
Mar 05 12:09:30 <Tofe> I didn't see the effects so far
Mar 05 12:09:41 <saidinesh5> Ahh
Mar 05 12:10:04 <Tofe> I'm ususally lucky with my phones :) I still have a fully working N900, usb connector unhacked and all
Mar 05 12:11:04 <saidinesh5> Ahh i lent my beloved N900 to a friend and he dropped it and broked it
Mar 05 12:11:09 <saidinesh5> *broke it
Mar 05 12:11:12 <Tofe> tss
Mar 05 12:11:19 <saidinesh5> the N9 battery puffed up too
Mar 05 12:11:31 <saidinesh5> the N950 got bricked during a firmware update iirc
Mar 05 12:11:39 <saidinesh5> N9 was probably my fav.
Mar 05 12:12:03 <saidinesh5> N900  close second.. but it's 256MB ram always showed ..
Mar 05 12:12:16 <Tofe> both were very nice, though N950 never got polished enough software-wise
Mar 05 12:12:26 <saidinesh5> oh but it did!
Mar 05 12:12:34 <saidinesh5> N950 got the PR 1.3 leaked
Mar 05 12:12:45 <Tofe> oh, I missed that news :)
Mar 05 12:12:56 <saidinesh5> yeah .. i still have it if you want
Mar 05 12:13:06 <saidinesh5> although it would be pretty useless at this point i think
Mar 05 12:13:24 <Tofe> I don't have any N950, just tried a couple of times with a friend
Mar 05 12:13:29 <saidinesh5> Ahh
Mar 05 12:13:47 <saidinesh5> i think my N9 was a loan device and N950 was a prize
Mar 05 12:13:49 <saidinesh5> or vice versa
Mar 05 12:19:44 <Tofe> https://talk.maemo.org/showthread.php?t=100554 <-- with the right software.....
Mar 05 12:22:22 <saidinesh5> tbh i don't think i am fond of hardware keyboard anymore..
Mar 05 12:22:29 <saidinesh5> like honestly swiping is wayy wayy faster i think
Mar 05 12:22:49 <saidinesh5> and when doing tinkery things.. hacker's keyboard with the right onscreen keys.. is oookayish
Mar 05 12:22:56 <saidinesh5> plus.. ahem ahem.. Price; $649
Mar 05 12:23:29 <saidinesh5> And no way i am going back to a Battery: 3,200 mAh ..
Mar 05 12:24:19 <Tofe> well, yes, it's just plain N9x0 nostalgia here :)
Mar 05 12:27:11 <saidinesh5> yep
Mar 05 12:28:57 <saidinesh5> i mean back in the good old days when wifi acceess points were still WEPand you can run airshark on your N9x0 .. things were different lol
Mar 05 12:29:09 <saidinesh5> *aircrack
Mar 05 12:44:45 <Tofe> yup
Mar 05 16:02:22 <Herrie|Laptop> Tofe: Seems that there are no results to https://github.com/Herrie82/initramfs-tools-halium/blob/tofe/ab-scheme/scripts/halium#L402
Mar 05 16:14:18 <Herrie|Laptop> Tofe: When I do: fastboot -c "datapart=mmcblk0p28" boot "boot.img" I get the following: https://bpaste.net/show/b7aa7897f3dd
Mar 05 16:14:41 <Herrie|Laptop> [    1.995235] mount: mounting mmcblk0p28 on /tmpmnt failed: No such file or directory
Mar 05 16:52:30 <JaMa> Tofe: sorry for leaving quickly yesterday, will look into that PR shortly
Mar 05 17:29:54 <Tofe> Herrie: that's most peculiar
Mar 05 17:31:50 <Tofe> I feel I didn't pick the right build target to debug things on sumo
Mar 05 17:32:04 <Tofe> (I built tissot, I'm now flashing it to test it)
Mar 05 17:33:09 <Tofe> JaMa: thanks
Mar 05 17:35:06 <Tofe> Tissot if as good as on pyro. Nothing to report -- except that wifi doesn't show up during firstuse, but it was the same with pyro
Mar 05 17:35:45 <Tofe> I'll just build hammerhead then, so that we fix it
Mar 05 17:38:42 <Tofe> I'll just build the bootimg... JaMa: would it be possible to build a new hammerhead sumo luneos-dev-image ?
Mar 05 17:39:17 <Tofe> though it'll probably take the same time -- maybe I could steal Herrie's image ;)
Mar 05 17:48:02 <JaMa> lets see if nyx invalidates qtwebengine, but still will take longer than herrie's :)
Mar 05 17:48:34 <JaMa> you want sumo? not rocko?
Mar 05 17:48:44 <Tofe> sumo
Mar 05 17:48:58 <JaMa> ok, switching unstable branch and workspace now
Mar 05 17:49:02 <Tofe> works as well as rocko, since the pulseaudio fix
Mar 05 17:49:28 <Tofe> JaMa: rocko is on gcc 7.3.0 too ?
Mar 05 17:50:41 <JaMa> pulseaudio fix should be needed only for sumo
Mar 05 17:50:43 <JaMa> IIRC
Mar 05 17:51:01 <Tofe> yes, that's what I meant, in a circonvoluted way
Mar 05 17:53:22 <JaMa> and yes, rocko is on 7.3 as well
Mar 05 17:54:07 <JaMa> rocko, sumo, thud all using 7.3 now
Mar 05 17:56:43 <Tofe> ok; so yes, right now there's no point in using rocko I think
Mar 05 17:59:07 <JaMa> it's still 1/3 of the changes included in thud
Mar 05 18:00:39 <Tofe> yes but thud crashes on tissot, whereas sumo is almost ready, so it's a good in-between step
Mar 05 18:02:07 <JaMa> have you tried tissot image built with gcc-7?
Mar 05 18:02:19 <Tofe> the sumo one, yes
Mar 05 18:02:28 <Tofe> works like a charm
Mar 05 18:02:47 <Tofe> Is there a gcc-7 build of thud ?
Mar 05 18:02:49 <JaMa> I've switched thud to gcc-7, maybe it works like a charm as well :)
Mar 05 18:02:57 <JaMa> yes, last one on jenkins is with gcc-7
Mar 05 18:03:01 <Tofe> oh! I didn't know :)
Mar 05 18:03:11 <Tofe> let me try that asap
Mar 05 18:03:36 <JaMa> https://github.com/webOS-ports/meta-webos-ports/commit/c558cccf65019b41eb2c40173fd527f50081afe7
Mar 05 18:03:49 <JaMa> that unblocks also the other MACHINEs with very old kernels
Mar 05 18:05:48 <Tofe> ok good
Mar 05 18:17:42 <Herrie|Phone> Tofe: www.herrie.org/ regular hammerhead filename I think
Mar 05 18:18:54 <Herrie|Phone> I uploaded it today and then downloaded and flashed
Mar 05 18:18:58 <Herrie|Phone> So should be OK
Mar 05 18:30:11 <Tofe> Herrie|Phone: thanks, I'm fetching it
Mar 05 18:34:09 <Tofe> JaMa: I still get Qt crashes during its scenegraph rendering on tissot with gcc 7.3 thud
Mar 05 18:35:02 <Tofe> "tissot kernel: Linux version 3.18.31 (oe-user@oe-host) (gcc version 7.3.0 (GCC) ) #1 SMP PREEMPT Sat Mar 2 16:54:10"
Mar 05 18:35:46 <Tofe> (also it would be good to import my pulseaudio fixes, as thud is now on PA 12.2)
Mar 05 18:38:04 <Herrie|Phone> Tofe: kernel is in herrie82 GitHub
Mar 05 18:38:39 <Tofe> Herrie|Phone: ok !
Mar 05 18:51:31 <JaMa> Tofe: ok, interesting datapoint, qt should be the same and gcc as well
Mar 05 18:52:04 <Tofe> yes; I didn't really investigate at that point, I must say. It could still be a TLS issue, fwiw
Mar 05 18:53:47 <JaMa> looks like buildhistory pushing was broken since ubuntu upgrade
Mar 05 18:53:49 <JaMa> NOTE: Writing buildhistory
Mar 05 18:53:49 <JaMa> The authenticity of host 'github.com (192.30.253.113)' can't be established.
Mar 05 18:53:52 <JaMa> RSA key fingerprint is SHA256:nThbg6kXUpJWGl7E1IGOCspRomTxdCARLviKw6E5SY8.
Mar 05 18:53:55 <JaMa> Are you sure you want to continue connecting (yes/no)?
Mar 05 18:55:28 <Tofe> oh, it lost its authorized hosts list...
Mar 05 18:59:35 <JaMa> looks like someone was doing something on 25th Jan
Mar 05 18:59:38 <JaMa> known_hosts               known_hosts.bak-20190125
Mar 05 18:59:44 <JaMa> novaldex|away: ^ any idea?
Mar 05 19:00:57 <JaMa> I'm checking why testing build failed for mido (with unrelated sstate issue)
Mar 05 19:03:20 <Herrie|Phone> JaMa: Novaldex was solving some key issues for Preware feeds
Mar 05 19:03:31 <Herrie|Phone> He broke another and fixed that last week
Mar 05 19:52:18 <Tofe> Herrie|Phone: I get the impression that udev doesn't actually do anything...
Mar 05 19:55:19 <Tofe> mmh no it's worse, /dev/block should be there even without udev....
Mar 05 21:33:01 <Herrie|Laptop> Tofe: Well yes that's not good...
Mar 05 21:33:12 <Herrie|Laptop> Seems it's likely systemd somehow, but not sure how to fix it yet
Mar 05 21:33:25 <Herrie|Laptop> https://github.com/systemd/systemd/issues/10041 and https://github.com/systemd/systemd/issues/9924 subscribe similar problems
**** ENDING LOGGING AT Wed Mar 06 02:59:57 2019