Saturday, 2019-01-26

*** ChanServ sets mode: +v T402:55
T4<BusterBg_18> So I got maintainership and already uploaded the packages stated on the faq (droid hal, droidmedia, audioflingerglue) Im kinda lost, whats next?06:50
T4<abhishek_0> @BusterBg_18 look at other devices as an example07:29
T4<abhishek_0> Replicate what they have done07:29
T4<BusterBg_18> It already have things in there so I just literally reuploaded the packages and triggered the rebuild07:32
*** ChanServ sets mode: +v T408:48
piggzjusa: mornin....09:05
piggzxpolicy.conf errors09:05
piggzi think starting this line https://paste.gnome.org/pcnz1knil#line-34209:07
piggzim guessing the variable is correct?09:10
piggzseems the variable only doesnt work in the device section09:19
HelloPotter_hey there while building i am getting this error build/core/product_config.mk:238: *** Can not locate config makefile for product "potter".  Stop.09:54
HelloPotter_anyone10:37
T4<elros34> maybe show your local manifest10:45
HelloPotter_potter.xml or manifest.xml10:58
T4<abhishek_0> potter.xml11:59
HelloPotter_https://pastebin.com/GfE9AM0012:29
jusapiggz: variable works in device section as well, just that I forgot equals:$foo from every line :) updated the file12:56
lbtPSA: 3.0.1.11 is available on OBS - it took a while due to some massive rebuilds happening13:17
* lbt wonders if NeoChapay needs all the mer:core copies to rebuild all the time on all the architectures including MIPS...? :)13:19
T4<vorombe> btw, https://git.merproject.org is down13:21
T4<NotKit> lbt, I think he has been porting for MIPS13:27
lbtOK, fair enough. It does use a lot of resources :)13:28
lbtI'll look at git now13:28
kimmolirinigus: FYI.. https://build.merproject.org/project/monitor/nemo:testing:hw:oneplus:onyx13:31
lbtgit is fixed13:34
r0kk3rzlbt: bump sailfishos:latest?13:35
lbtr0kk3rz: you need to get mal to OK that13:35
r0kk3rzmal says its ok ;)13:36
kimmolichange nick to ma1 and say ok ?13:37
* kimmoli hides13:37
lbtyou know I'm right here?13:37
r0kk3rzhehe13:38
kimmoli1bt: yes13:38
T4<vorombe> :D13:38
r0kk3rzim sure mal will be around soon enough13:38
riniguskimmoli: and I was wondering when these updates are coming? any idea what broke gst-droid? and what else should we do to get it all green again?13:41
kimmolino ideas atm13:41
kimmolijust added target and looked monitor13:41
rinigusgst-droid is probably missing some updates from upstream. I'll make an update and see if it helps13:46
kimmoliyeah notices13:50
r0kk3rzwhen was the last time you updated droidmedia?13:51
rinigus4 month ago13:51
r0kk3rzhmm, maybe you're missing some piggz-o-vision stuff13:52
T4<elros34> 3.0.1.11 target ready, let kernel panic begin!13:53
r0kk3rzyeah looks like13:53
rinigussurely we do.13:53
rinigusthen we will have to schedule droidmedia update as well. ... and that requires local build, unfortunately.13:55
r0kk3rzyeah13:55
r0kk3rzor drop a few commits off gst-droid13:56
r0kk3rzbut updating droidmedia is a good idea13:56
rinigusnaaah, if we stay behind then it will all accumulate. better deal with this as it comes13:56
rinigus... but maybe not today13:57
malHelloPotter_: I think you should be using this as device repo https://github.com/boulzordev/android_device_motorola_potter and branch lineage-15.1-6414:09
malrinigus: kimmoli geoclue-providers-hybris cannot be latest, for 3.0.1.11 you should use commit 38ef0f1d8a56b56db5bfe99fdb88b9a0e684ca6d14:16
piggzr0kk3rz: shhh, the first rule of piggz-o-vision, is, dont talk about piggz-o-vision :)14:17
malpiggz: :)14:17
piggzjusa: thx14:17
r0kk3rzgood thing im a serial breaker of rules14:17
malpiggz: is it ok if I remove the 3.0.0.8 devel target from mido OBS?14:17
piggzmal: sure14:18
piggzr0kk3rz: rule 2, you can talk about piggz-o-vision after your first PR14:18
malpiggz: normally only testing should have versioned targets, unless EA target is used14:19
piggzi may do some more icons and polish this weekend...tho, atm im plumbing and fixing my heating14:19
piggzmal: i know, was needed until latest !== latest14:19
piggzs/while14:20
mal@BusterBg_18 I'll do some maintanance in kenzo devel OBS, it has a useless target I'm going to remove, also read what I said a few lines back about geoclue version14:20
malpiggz: I just wanted to remove the obsolete 3.0.0.8 devel common target14:21
malnow it's gone14:21
UmeaboyI just for the fun of it downloaded and ran the sfa-mer script and I thought that this error was fixed already: Directory /var/run/dbus is missing in SDK root - please report this bug14:22
malUmeaboy: is that really an error, probably just a warning14:26
Umeaboymal: It stops the script so YES I'd say it's an error.14:27
rinigusmal: does it mean that we have to drop webhook?14:27
malrinigus: well it will be broken until next release probably14:28
malbut I doubt there will be much more changes before that14:28
Umeaboymal: !! command failed at Sat Jan 26 14:16:48 UTC 2019, dying...14:29
malrinigus: there are no open PRs to geoclue at the monent so just changing the commit hash is enough for now14:29
malUmeaboy: show the whole build log14:29
rinigusmal: testing it - then we should just release update before next commit will come :)14:30
malrinigus: and remember to use testing for release and disable build once it's ready14:32
rinigusmal: will do (or kimmoli will)14:32
mallbt: I just did a test build and looks like 3.0.1.11 target works well on Mer OBS so it can be made latest now14:33
rinigusmal: thanks, geoclue is fine now - would just have to get piggz brilliance incorporated and we'll be all set.14:34
HelloPotter_@mal yes, i am using that branch,14:34
Umeaboymal: Whole build log: https://pastebin.com/6NBTciNJ14:34
piggzrinigus: i expect PRs :P there is much to do14:34
malpiggz: about qtmultimedia, wondering if we should add that to common14:35
kimmolionyx has 3.0.1.11 target active under testing, i'll disable it when ready to trigger CI build14:35
malUmeaboy: is the platform sdk and target you use new enough?14:36
malUmeaboy: the real error could be "Package 'android-tools-hadk' not found."14:37
malwhich suggest too old platform sdk maybe?14:37
riniguspiggz: please don't. with the porting I just follow the lead.14:37
malHelloPotter_: your local manifest used <project name="boulzordev/device_motorola_potter" path="device/motorola/potter" remote="github" revision="oreo-mr1" />14:38
malHelloPotter_: which is different from what I said14:38
malHelloPotter_: also there are other errors in your local manifest, some branches are wrong14:40
HelloPotter_@mal  yes because  they have no branch called lineage-15.1-64, but i fixed that prob , by grepping the device tree from official lineage repository14:40
kimmolir0kk3rz: how does the CI get @release etc?14:40
malHelloPotter_: https://github.com/boulzordev/android_device_motorola_potter is not the same as https://github.com/boulzordev/device_motorola_potter14:41
T4<MarcoDS_bit> noob question, is the data/ directory we see in hybris-boot with telnet the same we get using adb shell in /data?14:41
r0kk3rzkimmoli: magic14:41
*** ChanServ sets mode: +v T414:41
r0kk3rztheres a global variable i set in the whole ci group14:42
malHelloPotter_: ah, sorry, the incorrect other lines in manifest were commented out, so only that device repo is wrong14:42
r0kk3rzothers are set in the .env in each project14:42
kimmolir0kk3rz: yep, those i found14:42
malHelloPotter_: do you now see the mistake in device repo?14:43
HelloPotter_@mal yup but iam stucked in an other prob build/core/base_rules.mk:238: error: vendor/codeaurora/telephony/internal: MODULE.TARGET.ETC.bootanimation.zip already defined by test/vts/utils/python/coverage.14:43
HelloPotter_@mal yes14:45
malHelloPotter_: very odd error14:50
HelloPotter_@mal i think  i need to undefine one of them, am i correct14:51
mal@It_sMike how often did you have the reboot issue with 3.0.1.11?14:51
malHelloPotter_: I didn't find that in the latter path14:52
UmeaboyIs there a chance that maple could have its own DHD repo and MW repo so that the sfa-mer can continue?14:52
malHelloPotter_: so I don't know why it finds it14:52
malUmeaboy: ?14:52
malwhat does that have to do with anything14:52
r0kk3rzUmeaboy: cant you just do things in the usual way?14:53
malr0kk3rz: I have always wondered the same thing, probably the script doesn't have the latest changes14:56
Umeaboymal: Nope. It doesn't.14:56
UmeaboyIt still uses the curlfix.14:56
malthere wasn't that many changes but some14:57
HelloPotter_@mal any solution15:00
UmeaboyIt still fails to unpack the tar.gz even though I double checked the device directory for errors.15:00
Umeaboyhttps://pastebin.com/GHSu1nkz15:00
T4<MarcoDS_bit> mal: we can connect with telnet on port 2323 but all freeze after some seconds15:01
mal@MarcoDS_bit get logs15:05
UmeaboyI meant tar.bz215:05
maldmesg, journalctl and /usr/libexec/droid-hybris/system/bin/logcat15:05
malUmeaboy: why can't you just follow the hadk pdf15:06
Umeaboymal: I wanted to try the sfa-mer script to see if it worked.15:06
malUmeaboy: for example the target location on server changed15:06
UmeaboyI know it's outdated and I solved some of the problems.15:07
malmaybe you use the old url15:07
T4<elros34> Umeaboy: you should already know that, add set -x to scripts15:07
Umeaboyelros34: Yeah. I'll check.15:07
malUmeaboy: so the proper way would be to look at hadk when you have issues, not ask here first15:08
malHelloPotter_: well one option is to comment out one of those repos in manifest, like vendor/codeaurora in snippets manifest, it should be i .repo/manifests/snippets/lineage.xml and then doing repo sync15:19
HelloPotter_ok i will try that15:20
HelloPotter_hey there i go another problem BOARD_KERNEL_IMAGE_NAME no defined17:05
malHelloPotter_: show your current local manifest17:08
HelloPotter_https://paste.ubuntu.com/p/SgV4YPkbVQ/17:10
ghosalmartinso hows sailfish land doing17:13
malHelloPotter_: why didn't you change the device repo like I said17:15
malHelloPotter_: it should be <project name="boulzordev/android_device_motorola_potter" path="device/motorola/potter" remote="github" revision="lineage-15.1-64" />17:16
HelloPotter_i changed the dvice repo in roomservice17:16
malHelloPotter_: remove roomservice17:17
HelloPotter_ok17:17
malHelloPotter_: and add that line to local manifest17:17
HelloPotter_ok17:17
wdehoogHi. testing 3.0.1.11 on bacon. unfortunately it reboots after a few minutes18:04
T4<elros34> third device with such a issue18:04
r0kk3rzdun dun dunnnn18:05
r0kk3rzthe plot thickens18:05
wdehoogall 12.1?18:06
wdehoognothing to be seen in journalctl. is there anywhere else I should look?18:07
T4<elros34> no, get dmesg from that reboot from last_kmsg or ramops18:07
mal@elros34 is your device also 12.1 based?18:08
T4<elros34> mal: no, it's 1118:08
malok18:09
wdehoogawful lot of:  mdss_fb_ioctl: unsupported ioctl (40b86da4)18:09
malwe really need to figure out what is causing that, it has to be some kernel bug18:09
malwdehoog: doesn't seem related, that looks like display related message18:10
wdehoogmal: ok18:12
wdehoogthis is in my last_kmsg: https://pastebin.com/dxGzGYNv18:12
HelloPotter_@mal hey it worked, but got another error ninja: error: 'libsf_compat_layer', needed by 'hybris-hal', missing and no known rule to make it18:12
wdehooghow can one see it is from a previous boot?18:12
malwdehoog: could you try this patch for your kernel https://github.com/mlehtima/android_kernel_fairphone_fp2/commit/d725094e30538b3f5c85bf338e251714965e0b96.patch18:14
malwdehoog: although one other person tried that but not 100% is was properly tested18:14
wdehoogmal: thanks I'll try18:14
malHelloPotter_: go to $ANDROID_ROOT/external/ and run git clone --recurse-submodules https://github.com/mer-hybris/libhybris.git -b android8-initial18:15
malwdehoog: just because fp2 and onyx has that patch and these don't have reboot issues18:16
HelloPotter_ok18:17
malit could be something else also18:17
malwdehoog: actually it seems you have that patch in hybris-12.1 branch18:20
wdehoogmal: indeed it is already there.18:23
wdehoogtoo bad18:23
T4<BusterBg_18> mal: the commit for geoclue is locally? Im still a little bit confused about the whole OBS thing18:24
T4<elros34> wdehoog: remove /etc/connman/firewall* and reboot to confirm18:25
wdehoogelros34: I'll do. reboot appears (to my mind) related to heavy network18:26
mal@BusterBg_18 what do you mean? just change the revision in the geoclue _service file on OBS18:26
wdehoogif I do nothing, nothing happens18:26
wdehoogwhen I try to start using an app sailfish IDE kaboom18:27
HelloPotter_@mal hey there  i got an error, but i didn't understand what is the error, please lokk it up https://paste.ubuntu.com/p/mXw7XkM4jt/18:29
T4<BusterBg_18> Ahhh, now I get it, thanks18:29
malHelloPotter_: your fixup-mountpoints has some bug, pastebin the whole file18:30
HelloPotter_wait18:30
HelloPotter_@mal https://paste.ubuntu.com/p/nHY846wm4M/18:32
T4<BusterBg_18> Does sensorfw need to be also at the previous commit at faq? Or was that a 3.0.0.8 only thing?18:32
T4<elros34> iirc latest works ok now18:38
mal@BusterBg_18 sensorfw needs old commit only on 3.0.0.818:38
malHelloPotter_: indentation is incorrect, you have mixture of spaces and tabs, also there is empty line after "$@"18:41
r0kk3rzhmm, my sensors stopped working19:04
r0kk3rzi wonder why19:04
HelloPotter_@hey there it seems to be ok, i can't able to figure out where is the problem with my fixup moutpoints(which you told), can you please guide me with using an image19:20
HelloPotter_@mal ok i go it , i just solved the problem20:07
wdehoogelros34: no crash since mv firewall.* elsewhere20:11
T4<elros34> wdehoog: ok, do you have pstore enabled? Maybe there you will have more network related kernel-panic.20:15
wdehoogsorry, no pstore20:16
wdehoogdoes pstore only save 1 boot or multiple? do you know?20:16
*** ChanServ sets mode: +v T420:20
T4<elros34> never heard of multiple boots logs20:20
wdehoogelros34: I am using multirom so another kernel starts the sf kernel. I guess the error logs are gone then20:24
T4<elros34> oh so maybe you could try: while true; do dmesg -c; done. Anyway I'm not sure it will be fast enough to catch error20:26
wdehoogelros34: not fast enough https://pastebin.com/JZhvCfZN21:32
T4<MarcoDS_bit> mal: dmesg https://paste.ubuntu.com/p/n92Kj6MvJj/21:33
T4<elros34> wdehoog: ok thanks21:34
T4<MarcoDS_bit> Still it freezes after like 20 seconds21:35
wdehoogelros34: https://pastebin.com/EgfvJZfp . I sacrificed my ubports phone so now I have a last_kmsg without multirom.21:38
T4<elros34> oh great so it's really the same issue. For comparision https://gist.github.com/elros34/be59c64e29fcc04bbfce227c86e3c48f21:42
T4<It_sMike> mal : once I run browser or download any app from store, phone gets reboot :D21:48
T4<It_sMike> Hmmm.... Elros did you find out anything?21:49
T4<It_sMike> mal : first crash :/ then reboots for twice ...21:50
T4<elros34> no, I didn't search yet. Only found another victim21:51
T4<It_sMike> lol , whom?21:51
T4<It_sMike> in what device?21:51
T4<It_sMike> which* :D21:51
mal@It_sMike did disabling firewall help you?21:52
T4<elros34> wdehoog's bacon21:52
mal@MarcoDS_bit freezes or reboots?21:52
T4<It_sMike> mal : CONFIG_NETFILTER_XT_MATCH_CONNTRACK you mean unset this config? or removing firewall folder?21:53
T4<MarcoDS_bit> mal: freezes, stays on the xiaomi logo but telnet times out21:53
T4<It_sMike> unseting*21:53
T4<It_sMike> @MarcoDS_bit Same here !!!!21:54
T4<It_sMike> so 3 victims till now :D21:54
T4<MarcoDS_bit> oh sad ahah21:54
mal@It_sMike removing the folder is the current workaround21:55
T4<It_sMike> people are just killing me at xda :D should roll them back to 3.0 and wait for any bugfixs ..21:55
T4<It_sMike> mal : yeah, removing the folder makes everything fine again...21:55
mal@It_sMike the bug is in your device kernel, not in sailfish21:55
mal@MarcoDS_bit if you have time before the issue comment out any usb related imports in /init*rc files21:56
wdehoogcan we try to disable ipv6? and if so how?21:56
T4<It_sMike> O_o what kind of bug mal? why it wasn't there before this update ? a bit confused !21:57
mal@It_sMike somewhere in the netfilter related code probably, 3.0.1 has some firewall stuff in use21:57
T4<It_sMike> Currently building with 3.0.1 sdk and targets.... will report back if it helps...21:58
T4<It_sMike> Ah.... understood...21:58
T4<It_sMike> mal, is everything ok on fp2 with 3.0.1 ?21:59
mal@It_sMike so far yes, I have used it a few hours now without any problems21:59
T4<It_sMike> Hmmm....22:00
mal@It_sMike how often do you have those crashes?22:00
T4<It_sMike> Just tried everything that I could :D but still no luck :/ now, it's over a week22:00
T4<It_sMike> too many times.22:01
T4<It_sMike> when I turn on wlan and start browsing, it would be crashed after a while22:02
T4<It_sMike> same for cellular too22:02
T4<It_sMike> Oh... I've find out another bug too... when I use pully menu in ambiences section, settings will be stucked at that point...22:04
T4<It_sMike> Untill I reboot the device again.22:04
malhow many minutes did you have to browse?22:05
T4<It_sMike> less then a minute !22:05
malI'll leave the device to play some stuff from youtube and see what happens22:05
T4<It_sMike> Nice :)22:06
T4<It_sMike> Could you also try browsing the xda?22:06
T4<It_sMike> seems it has a serious problem with xda :D , when I open xda, it just crashed and then reboots...22:08
malhmm, xda seems to be very slow or not working22:09
T4<It_sMike> Do you have some kind of crashs too?22:09
malno22:09
malthe page just doesn't load22:10
T4<It_sMike> Mine will be crash after trying to load xda...22:10
T4<It_sMike> Really confusing :( !22:10
malok, it just went to a non-working place for some reason22:11
malnow with proper url it seems to work22:11
T4<It_sMike> Nice ! so no problem so far...22:12
malno, how long did you have to browse in xda?22:12
T4<It_sMike> Could you also try the pully menu in ambiences ? ( tell me if you don't get my meaning )22:12
malxda causes quite much memory usage22:13
T4<It_sMike> less then a minute.... could not browse it due the crashing...22:13
T4<It_sMike> Yeah much heavy ...22:13
malno crashing yet22:16
malI have scrolling through some threads22:17
T4<It_sMike> I can't even do this :D22:17
T4<It_sMike> maybe I should clone your netfilter to my kernel ( a very noob solution lol )22:19
malthe commit history looked the same22:20
malexpect that one commit which didn't help22:20
T4<It_sMike> Yep :( tried it last night....22:21
maleven speed test websites don't cause a crash and those should cause quite heavy network usage22:23
T4<It_sMike> so yeah.... the problem is from our kernels ( me and other 2 victims :D )22:24
malhmm, my fp2 gave quite bad results from network tests22:25
malwondering what is causing those22:25
T4<It_sMike> o-o22:26
T4<It_sMike> Just for sure ... are you on 3.10 ? ( kernel I mean )22:27
mal3.4 kernel22:27
T4<It_sMike> oh ... same as me...22:27
T4<It_sMike> now it's more confusing ...22:28
malalthough not sure how fast wifi fp2 even should have22:28
malmy laptop nicely shows 100/100 MB/s but fp2 only showed 30 to 50 and upload was even worse22:29
T4<It_sMike> Hmm.... is ther any diffrence between the results from 3.0 and 3.0.1 ?22:32
T4<It_sMike> there*22:32
T4<It_sMike> Still building.... can't wait anymore...22:36
wdehoogI disabled ipv6 (connmanctl) and so far so good22:36
malno idea, but testing on another sailfish phone and upload was only 15 MB/s in that also22:36
T4<It_sMike> O_O seems something's really wrong with the new update22:37
T4<It_sMike> wdehoog : Should try it too....22:37
mal@It_sMike the wifi speed issue on fp2 is not because of 3.0.1, I tested with 3.0.0.8 and the results are the same, 50/5 MB/s22:42
malso that in 3.0.0.8 fp222:42
T4<It_sMike> So seems everything ok, maybe it's a hardware related thing...22:43
malmaybe phones don't have very good upload speeds22:44
T4<It_sMike> Yep :)22:45
T4<It_sMike> And still building :/22:46
T4<It_sMike> ( system I mean )22:46
T4<It_sMike> Finished finally ( lol ) .... let's get in22:55
T4<It_sMike> Booting....23:00
T4<It_sMike> lol23:03
T4<It_sMike> no cellular data or network :D23:03
malhow did you manage to break those23:04
T4<It_sMike> Thanks SailfishOS :)))) still have reboots, so no luck with new targets too23:04
T4<It_sMike> Deleted ofono folder from etc :D , thats the reason ...23:05
T4<It_sMike> Had some problem while building droid-hal-version23:06
T4<It_sMike> seems it's a bit better now, mal23:09
T4<It_sMike> no crashing from the first reboot23:12
T4<It_sMike> xda looks fine23:12
T4<It_sMike> let's open about 100 tabs :D , will it survive ?23:14
T4<It_sMike> scroling doesn't caused reboot right now...23:14
T4<It_sMike> Working fine....23:16
T4<It_sMike> Till now...23:16
T4<It_sMike> So the problem was too old target and tooling all the time :/ ?23:17
T4<It_sMike> Any idea mal :D ?23:17
T4<It_sMike> I should make ofono thing back and test it again.... as I removed it right now23:18
T4<It_sMike> Ok... crashed again...23:19
T4<It_sMike> But took longer to ran into crash23:19
r0kk3rzdebugging random issues is always fun23:19
T4<It_sMike> lol23:20
r0kk3rzits like, did you fix it? is it better? or is it just chance23:20
T4<It_sMike> Not fixed yet. but much better then the last time !23:20
T4<It_sMike> As I said took longer for crashing....23:21
T4<It_sMike> enough for tonight ... Continue tomorrow.... :)23:23
T4<It_sMike> Goodnight sailors ( here is 3:00 am )23:24
mal@It_sMike in theory the browser crashing can also be OOM related23:57
malcheck journalctl for anything related to oom23:57
mal@It_sMike what is that ofono issue?23:58
malyou mean a conflict between two packages?23:58
malif you have something ofono related in droid-config sparse you need this in droid-config spec https://github.com/mlehtima/droid-config-fp2-sibon/blob/master/rpm/droid-config-fp2-sibon.spec#L2223:59

Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!