Friday, 2019-01-25

SimdolDo I have to use LineageOS as a base? LineageOS for OnePlus 6 Oreo is very primitive and I am leaning towards Omnirom.00:34
SimdolAssuming that Android Pie is not supported, of course.00:34
SimdolAlso where can I get SDK for aarch64?00:45
SimdolI don't know why it won't let me chroot into sdk. What can I do to figure this error out? unshare: cannot change root filesystem propagation: Invalid argument00:56
SimdolManually chrooting to the chroot seems to work.01:01
SimdolOkay, so is the reason why I have to use LineageOS as a base is because I have to use hybris-15.1 branch for Android?01:07
*** ChanServ sets mode: +v T402:09
*** ChanServ sets mode: +v T408:15
r0kk3rzyeah our hybris trees are usually based on lineageos08:34
malwhat sources does Omnirom use?10:08
T4<NotKit> Omnirom is AOSP + CAF based, different tree from LOS10:10
T4<adampigg> jusa: ping .... on our new port, we have audio working/ atlast (its intel based and we hadto write an alsa config) .....12:50
T4<adampigg> video// ,media works ok12:51
T4<adampigg> but BT headphones dont, the audio doent route12:51
T4<adampigg> headphones pair/connect ok12:51
T4<adampigg> would our PA config be wrong, or what else could it be?12:51
HelloPotter_hey there i am porting sailfish for my device potter, i wrote fixup mountpoints(which i get from my device), can any one tell me is this correct??
malHelloPotter_: that is not correct, see the link above how to define the paths, those are not absolute paths13:07
HelloPotter_@mal now?13:09
malremove the empty lines near the end, otherwise looks ok13:16
T4<adampigg> jusa, our sound config
T4<adampigg> which is based on the emu;lator config iirc13:17
T4<eugenio_g7> I have some fixes on that that I need to push13:27
T4<adampigg> ooooo13:28
T4<adampigg> go on13:28
T4<adampigg> i/was going o13:28
T4<adampigg> along the ines of comaring the pa config with the droid one...13:28
T4<adampigg> yes13:43
T4<adampigg> an x86 device13:44
T4<adampigg> we use mesa for gfx ... its almost like a pure port ... hoping to use vaapi for video accel13:44
mal@adampigg which parts use droid stuff currently?13:45
T4<adampigg> none i think13:45
T4<adampigg> sensors are native, as are gfx13:46
T4<adampigg> sound is alsa13:46
T4<adampigg> there is no camera working13:46
T4<adampigg> and it doesnt work in los13:46
T4<adampigg> not sure how the camera works or how its connected yet13:47
T4<adampigg> it works in the/offiiial android and windows 1013:48
T4<eugenio_g7> Yup no droid bits13:52
T4<eugenio_g7> I have dhi masked too on mine13:52
T4<adampigg> i can/probably/mask/mine too nowwd is fixed13:53
T4<adampigg> sorry about / .... it is next to spacebar!13:53
T4<abhishek_0> If it is x86 , it should be able to run mainline fine, have you tried @adampigg ?14:00
*** ChanServ sets mode: +v T414:01
T4<adampigg> yes, and we had modesetting issues14:02
jusaare you using bluez4 with the device?14:09
T4<adampigg> jusa, 514:12
mal@adampigg which kernel version does it have?14:14
jusaok, change line 43 on to load-module module-bluez5-discover14:14
T4<adampigg> mal, 3.1414:14
T4<adampigg> jusa, i might have done that locally already, lemme check14:15
T4<adampigg> jusa, actually on that line i have the bluez-discover load, with bluez4 and bluez5 args14:16
T4<adampigg> will change14:16
jusathat should be fine as well.. you'll need to add flags1 = disable_notify to bta2dp card section, line 165 in the xpolicy.conf, though I think that's not enough to get everything ok..14:19
T4<adampigg> @eugenio_g7 what changes do youhave?14:20
jusacould you take verbose pulseaudio logs from connecting the bt device?14:20
T4<adampigg> sure, probbaly this evenig tho if thats ok14:20
jusayep, I'm not sure how I have time but I'll try look at that during the weekend14:21
T4<adampigg> ta14:22
T4<eugenio_g7> @adampigg [@eugenio_g7 what changes do youhave?], Some changes to the xpolicy.conf regarding the correct switch of sinks (which is only one, but due to a configuration error it tries to switch to an unexistant switch) and outputs... Managed to make PA switch the output correctly in case an headphone is plugged (now I need to get the al14:49
T4sa mixers right)14:49
T4<MarcoDS_bit> mal so now we have no errors in kernel defconfig, I have just reinstalled hybris-boot and sailfish chroot, but the strange thing is that I can telnet into it with port 23 as sailfish was not installed (I followed the manual installation)16:04
mal@MarcoDS_bit do you see something in /diagnosis.log16:25
T4<MarcoDS_bit> mal "CONFIG_AUTOFS4_FS=y not found in /proc/config.gz"16:27
mal@MarcoDS_bit so you need to fix that16:28
T4<MarcoDS_bit> In kernel defconfig right?16:29
T4<MarcoDS_bit> OK thank you16:29
maland make sure you find it in .config found in out/16:29
T4<MarcoDS_bit> Sure :P16:30
malafter make hybris-boot of course16:31
T4<adampigg> @eugenio_g7 looking forward to those fixes!16:53
T4<It_sMike> Ah... Tired of debugging .... could anyone help me please ??????17:49
mal@It_sMike what is the problem?17:50
T4<It_sMike> Phones get reboot while browsing :(17:51
T4<It_sMike> or installing apps17:51
r0kk3rzlike instant boot loop?17:56
r0kk3rzturn on pstore and check for kernel panics17:56
T4<It_sMike> Everything was fine on , but when I've updated to 3.0.1 it's starts happening...17:57
mal@It_sMike just to be sure, did you update target correctly before building
T4<It_sMike> to which version?17:58
malto obviously17:58
T4<It_sMike> Hmmm... I guess so. but let me check it for sure! just target ?17:59
malin platform sdk check sb2 -t SailfishOS-latest-armv7hl -R -msdk-install ssu re17:59
maloops, replace the target name with the correct one in that18:01
T4<It_sMike> yeah I know :) no problem !18:01
T4<It_sMike> It's on :D18:04
T4<It_sMike> SO that was the problem all the time :/// ?18:04
r0kk3rznah its gonna be kernel panics18:06
r0kk3rzi think elros was having a similar issue18:06
T4<It_sMike> Yeah I see in the logs...18:06
T4<abhishek_0> 3.0.1 on obs yet?18:07
T4<It_sMike> @abhishek_0 What do you mean?18:08
T4<abhishek_0> @It_sMike not asking you :P18:08
T4<It_sMike> Sorry :D18:08
mallbt: reminder of adding target to Mer OBS, when you have time :)18:08
T4<It_sMike> noob question :D , how do you work with pstore?18:15
r0kk3rzits a kernel option you enable18:30
T4<It_sMike> Already enabled :D , then what should I do? ( sorry I'm a bit noob with the whole kernel )18:31
r0kk3rzwhich pstore options you have enabled18:34
T4<It_sMike> Let me check ...18:34
T4<BusterBg_18> Hey mal I fixed my droidmedia issues and dropped all  anbox patches, and got my obs account,  can I get maintainership for xiaomi:kenzo?:D18:35
T4<It_sMike> r0kk3rz : CONFIG_PSTORE=y18:36
r0kk3rzany others?18:37
T4<It_sMike> nope!18:37
r0kk3rzcheck in /sys/fs/pstore for a ramoops file18:37
r0kk3rzif not, you might need to enable more18:37
T4<It_sMike> OK !18:38
mal@BusterBg_18 which was the original maintainer?18:40
T4<BusterBg_18> Akhil Kalwakurthy18:41
malhave you had any contact with him?18:44
T4<BusterBg_18> Yeah, he helped me to fix droidmedia actually18:46
malcan't he give you permissions? he has the permissions to change
T4<BusterBg_18> I will contact him18:49
T4<BusterBg_18> I hope he have internet access18:49
malit's just in a way cleaner if the original porter gives the permissions to new one, if that is not possible then it can be handled by me18:50
r0kk3rzespecially since that ports not that old18:52
piggz@eugenio_g7 .... that new kernel is acting much better for me regarding USB connection .... i no longer get lags and errors regarding charging voltage18:53
piggzmal: you think 3.14 kernel will be ok?18:54
piggzjusa: getting logs now...18:54
T4<BusterBg_18> Ok, I will contact you in case he doesn't respond in some days18:55
malpiggz: ok for what?18:55
T4<BusterBg_18> It's not that old but he can't continue supporting it anymore18:56
piggzmal: you asked what kernel version we are using18:56
piggzso, i wondered why18:56
malpiggz: just curious, it's fine, just asked because of bluez518:56
piggzjusa: there was _nothing_ printed in journal when the headphones connected .... i am running PA with -vvv18:57
r0kk3rz@BusterBg_18 theres nothing stopping you from making an update though18:58
T4<BusterBg_18> That's why I started the whole OTA thing, I also solved some of his bugs, I really want to keep this device update for as long as I can19:00
r0kk3rzoh so you have updated the repos already?19:03
T4<BusterBg_18> The ones on github yes19:04
r0kk3rzbtw dont do things like this -
r0kk3rztheres really no need, the out-of-box setup prompts to download apps19:05
T4<BusterBg_18> Oh ok, that's  just for when I was trying builds and didn't want to set up the device19:09
maldoes that device really still need ofono delay hack?19:11
mal@BusterBg_18 installroot and tmp should not be in git19:11
T4<BusterBg_18> Already put them on gitignore19:14
maldocumentation.list is also useless in git19:16
r0kk3rzyou could probably squash those two commits to fix the history a bit19:17
r0kk3rzbut in general i have no major issues adding you to the obs if you have an account already19:18
r0kk3rzi mean, if you put your changes on master and tagged it, the webhooks should pick it up anyway19:21
malindeed, that would do it19:22
mal@BusterBg_18 could you squash the last two commits to one commit19:22
malr0kk3rz: although droid-hal needs updating anyway19:23
malso it wouldn't fix everything19:23
r0kk3rzyeah true19:27
piggzr0kk3rz: argh, make me a newer libva19:29
r0kk3rzcan do19:29
r0kk3rzshould just be a case of bumping the submodule19:30
r0kk3rzbut it wants new wayland, so probably build it in your repo19:30
T4<It_sMike> r0kk3rz : There's such no directory :D , /sys/fs/pstore19:30
r0kk3rzyou know, the linux kernel is quite a popular thing, you can find out all kinds of things on your favourite internet search engine19:32
T4<It_sMike> Right :) sorry! may I ask another question?19:33
piggzr0kk3rz: ok.... im hoping it defines I915_EXEC_VEBOX19:33
piggzis the intel driver fails to build...19:33
piggzbut, im just searching for that atm19:33
r0kk3rzis this for the driver thing?19:34
piggzr0kk3rz: yes ... bugger, looks like it comes fromthe kernel19:35
piggzr0kk3rz: though, accoridng to it should be in my kernel19:36
piggzso, some build problem19:36
mal@It_sMike does the device have /proc/last_kmsg ?19:37
T4<It_sMike> let me see mal !19:38
r0kk3rzpiggz: where did you get the spec from19:38
piggzr0kk3rz: fedora19:38
r0kk3rztanks lbt :)19:40
T4<It_sMike> Yes mal. it does, so what's next!?19:40
mal@It_sMike depends on whether your device actually uses that, it will contain kernel logs if there is a crash19:41
T4<It_sMike> Thanks! I should take a look on it...19:42
r0kk3rzpiggz: must be that gen4asm thing?19:42
piggzr0kk3rz: what is that?19:44
r0kk3rznot sure19:45
r0kk3rzsays intel on it19:46
r0kk3rzmaybe you'll have to build that on device...19:47
r0kk3rzyou do need it19:52
r0kk3rzpiggz: 'The intel-vaapi-driver will not be actively maintained. From BDW and SKL+, intel-media-driver for VAAPI is recommended. '19:56
r0kk3rzhmm, nope, cant use that one, gcc too olde19:58
piggzr0kk3rz: so, stick with intel-vaapi-driver?20:04
r0kk3rzi think so20:05
r0kk3rzidk, this stuff is confusing20:05
piggzr0kk3rz: so, on obs, do I just need to click 'branch package'20:21
piggzan stick it in one of my repos20:21
*** ChanServ sets mode: +v T420:22
r0kk3rzyou could do that20:23
r0kk3rzor just copy the _service file20:23
r0kk3rzand probably fork the repo on gmo20:24
piggzactually, yes, i need a newer version anyway maybe20:26
piggzr0kk3rz: thoughts on build problem?20:26
r0kk3rzwhich one?20:27
r0kk3rznot especially20:28
r0kk3rzyou must have to build that thing against the right kernel20:28
r0kk3rzso whether you can package kernel headers and use that for build?20:29
piggzr0kk3rz: it seems to be in there....20:31
piggz[SB2 sdk-build xiaomi-latte-i486] piggz@linux-f1uu / $ cat ./usr/lib/droid-devel/include/libdrm/i915_drm.h | grep I915_EXEC_VEBOX20:31
piggz#define I915_EXEC_VEBOX                  (4<<0)20:31
T4<MarcoDS_bit> mal all done but not working anyway, but, I tried extracting the chroot from my pc and I actually got way more files than extracting it from the phone (and from the phone it gave me "tar: short read"), so I checked the partitions size and I see this and all seems alright (?). now I'm try copying the extra20:31
T4cted files from the pc to the phone manually20:31
piggzr0kk3rz: ah20:32
piggzit must be building against ./usr/include/libdrm/i915_drm.h20:32
piggzwhich doesnt have it20:32
T4<BusterBg_18> Once I get back home I will squash those commits, but still have to update droid-hal and droidmedia to obs right?20:33
T4<It_sMike> mal could you help me with something please? this is the last_kmsg after crashing : , Any idea about what happend?20:37
piggzr0kk3rz: i guess i need to modify the configure checks to look for droid-devel instead of libdrm-devel20:38
T4<It_sMike> I'm gonna make some changes to the kernel, but would be happy to know your opinion too, mal :)20:40
T4<It_sMike> Uh.... still crashing :( got really tired.... took about a week and still no fix :(20:46
r0kk3rzwhat did you do?20:49
r0kk3rzthat last_kmsg doesnt really say anything useful20:49
T4<It_sMike> Yeah I know :(20:49
T4<It_sMike> Disabled some of the useless kernel configs...20:49
mal@It_sMike sailfish release shouldn't really break such things20:50
malso it's odd how didn't have the issue20:50
T4<It_sMike> Yeah! much odd :(20:50
mal@It_sMike how exactly did you build the new version?20:50
r0kk3rzits not that odd, since theres new kernel functionality being used now20:51
T4<It_sMike> mal : Like everytime I build images ....20:52
malr0kk3rz: but odd how that kind of "basic" feature in kernel would be broken20:52
T4<It_sMike> r0kk3rz : Applied those too.... but no luck :(20:52
r0kk3rzmal: yeah i know, but im reasonably sure elros had a similar issue20:53
r0kk3rzid disable CONFIG_NETFILTER_XT_MATCH_CONNTRACK as a test20:54
mal@It_sMike what does "zcat /proc/config.gz | grep CONFIG_NETFILTER_XT_MATCH_CONNTRACK" return?20:55
T4<It_sMike> let me run it ....20:55
T4<It_sMike> Uh crashed again :D while running the command20:55
malthat command cannot cause a crash, it's just random probably20:56
T4<It_sMike> Yeah... random reboots :( Are everywhere now :/20:56
T4<elros34> not very similar, I had kernel panic after loading wlan module20:57
T4<elros34> Removing /etc/connmand/firewall* "fix" it20:57
T4<It_sMike> @elros34 same here ...20:58
T4<It_sMike> I should test that too :) thanks Elros :)20:58
mal@It_sMike can you give a link to your kernel source repo20:59
T4<It_sMike> Sure !20:59
T4<It_sMike> Here you are mal :)21:00
mal@It_sMike you use that hybris-12.1 branch?21:00
T4<It_sMike> mal it returns this : CONFIG_NETFILTER_XT_MATCH_CONNTRACK=y21:00
T4<It_sMike> Yes.21:00
mal@It_sMike a very random idea, apply this patch to your kernel
T4<It_sMike> WoW :))))21:11
T4<It_sMike> Thanks a lot mal. ok I'll do it and report back !21:11
T4<It_sMike> Hmmmm... seems everything's working fine after removing /etc/connman/firewall* , let me test everything again ....21:20
T4<It_sMike> Yeah, Everything's OK now ! ( with Elros solution ) . mal , let me apply your patch too....21:30
mal@It_sMike the method by elros is not a proper fix, hopefully the kernel patch works, just remember to add back the firewall configs21:31
malfor testing21:31
T4<elros34> certainly it's not solution but workaround21:31
T4<It_sMike> Yeah I know mal... I hope too :D21:31
T4<It_sMike> @elros34 :) Thanks anyway...21:32
mal@It_sMike that was the only relevant looking difference in my working kernel and your kernel, at least in the netfilter code21:32
T4<It_sMike> Applying right now mal ! Thanks so much for looking around my kernel :)21:33
r0kk3rzso it was that netfilter thing, weird21:49
T4<It_sMike> Building new image right now....21:55
piggzr0kk3rz: our libdrm is _reall_ old21:59
jusapiggz: hm..22:03
piggzjusa: hit me with cmds to run :D22:05
jusapiggz: it would be nice to see ohmd logs of that but the command line for enabling logs for ohmd is not something one remembers by the heart.. :)22:05
piggzr0kk3rz: i was able to build ver 1.5 of the driver22:06
jusapiggz: as root:22:07
jusasystemctl stop ohmd22:08
jusaohmd --no-daemon --verbose=all --trace="* enabled;* target stdout;*.*=all"22:08
piggzthat line is approx when i turned on headphones22:12
piggzi actually turned off/on22:13
T4<WaseemAlkurdi> Hi guys22:15
T4<WaseemAlkurdi> Excuse me for the question22:15
T4<WaseemAlkurdi> How do I add a stub to libhybris?22:15
jusapiggz: ok so ohmd side looks ok, I'll re-check the policy-enforcement configs..22:19
jusapiggz: could you take full pulseaudio logs of starting & connecting headset, even if connecting headset wouldn't log anything extra22:23
piggzjusa: seems i was wrong ... earlier, journald had stopped logging!22:25
jusaah, if you take logs from journal, make sure rate limiter is off, otherwise journal will drop log lines from pulse22:26
jusaRateLimitBurst and RateLimitInterval set to 0 in /etc/systemd/journal.conf22:27
jusaand restart22:27
piggzjusa: line 1418 maybe ?22:31
T4<It_sMike> Uh... mal, just applied your patch, then clean installed, but still have random reboots :(22:33
T4<It_sMike> Should we give up at this point :D ?22:39
T4<elros34> mike: could you try: while true;do dmesg -c; done. You should have some logs if the issue is same like mine22:41
T4<It_sMike> Sure Elros, but what did you do for sloving this problem? ( instead of removing those files )22:44
T4<elros34> I did not solve it. I was hoping more user will have same issue when target will be ready. Also I tested only in 1 device for now22:47
T4<It_sMike> Understood... had the same idea too :)22:48
jusapiggz: I'll update your config a bit, takes some minutes22:48
piggzjusa: once again, gr8 service!22:49
jusapiggz: could you still paste output of "pactl list" when pulse is running22:51
T4<elros34> It's like lack of audio after watching video bug reported in tjc. Not everyone have it22:52
T4<It_sMike> :( Maybe I should downgrade users to 3.0 and wait for a bugfix from jolla ....22:53
T4<It_sMike> Anyway... tomorrow, I'll try building over 3.0.1 tooling and target ... ( I know it would not solved it... but better than nothing )22:55
*** Benclark006_ is now known as Benclark00623:01
T4<It_sMike> Goodnight sailors :)23:03
T4<It_sMike> Thanks everyone :)23:03
piggzr0kk3rz: gst-vaapi builds without gst-doc, it just skips it23:07
jusapiggz: - try those for pulseaudio config, they are not optimal (or even complete) but should be more in line how things need to be done23:47
jusait's nice to figure out alsa-based configs, easier to see where things would need to be more generic, as currently the config has mostly gone "droid things first"23:48
jusanow some sleep :)23:48
jusaanyways, as that most likely blows up pulse logs would be nice :)23:49

Generated by 2.17.1 by Marius Gedminas - find it at!