| *** ubuntu is now known as Guest2206 | 01:31 | |
| *** niven.freenode.net sets mode: +v T4 | 01:44 | |
| *** ced117 is now known as Guest31599 | 01:48 | |
| *** juiceme is now known as Guest14302 | 01:50 | |
| *** ahjolinna_ is now known as ahjolinna | 05:32 | |
| *** ChanServ sets mode: +v T4 | 05:39 | |
| Mister_Magister | hmmm im wondering who can i ask about statefs problem | 06:03 |
|---|---|---|
| Mister_Magister | monich: not only signal strength changes to plane mode but also battery indicator freezes and charging is not detected but its detected by AIDA64 and after i restarted statefs it works | 06:04 |
| T4 | <adampigg> mal, qgsutils in qtmm helps...shows how to do type conversion from GValue | 06:26 |
| T4 | <adampigg> i think i have the gst-droid pat done, just need the qtmm side finished | 06:26 |
| spiiroin | Mister_Magister: simple version: statefs notifications via fuse just do not work reliably. in any device. | 06:28 |
| Mister_Magister | spiiroin: but no other device have problems like that | 06:29 |
| Mister_Magister | i would like to fix that | 06:29 |
| Mister_Magister | @adampigg hi whatcha doin? | 06:29 |
| T4 | <adampigg> secret ;) | 06:30 |
| Mister_Magister | echhh… | 06:30 |
| spiiroin | Mister_Magister: yes they do, just when and how and how much and do you/users in general notice varies. and the fix involves rewriting battery tracking. | 06:31 |
| T4 | <adampigg> patching qtmm to read camera supported features | 06:31 |
| Mister_Magister | spiiroin: but i havent seen any device facing that issue :O | 06:31 |
| Mister_Magister | spiiroin: every two days is a bit too often | 06:32 |
| spiiroin | Mister_Magister: it does not work properly even in j1. i.e. it has never worked reliably for any device. some are more affected than others. | 06:33 |
| Mister_Magister | spiiroin: i havent seen any problem with it on jolla phone or any other | 06:33 |
| Mister_Magister | like freezing battery indicator (in aida its working) and changing signal bar to plane icon | 06:34 |
| Mister_Magister | spiiroin: if you say so sure | 06:35 |
| Mister_Magister | its just weird how much my device is affected | 06:35 |
| Mister_Magister | spiiroin: so not possible to make it any better? | 06:37 |
| * Mister_Magister wants to make phone more fully working just like mal | 06:37 | |
| spiiroin | Mister_Magister: fyi: just opened bug about this. at least xperia x is affected seriously enough. probably causes things like https://together.jolla.com/question/169321/xperia-x-white-led-stuck-on-after-charging/ | 06:38 |
| Mister_Magister | spiiroin: that doesnt sound like my problem tho :/ | 06:39 |
| Mister_Magister | led sometimes stopped turning on tho for no reason | 06:39 |
| Mister_Magister | spiiroin: on xperia x only led freezes and after 10 days its not that bad | 06:44 |
| spiiroin | Mister_Magister: the problem is "stuff in /run/state/namespaces/Battery changes, mce does not get woken up to process it" -> battery level, charging state, younameit can be affected | 06:44 |
| Mister_Magister | on my device battery and signal indicator freezes after 2 days | 06:44 |
| Mister_Magister | that a bit a lot more serious which i would like to fix it didnt happen before 2.1.4 | 06:45 |
| spiiroin | as said, details vary, root cause is the same | 06:45 |
| Mister_Magister | sure i get it now | 06:45 |
| Mister_Magister | spiiroin: it didnt happen before 2.1.4 and since 2.1.4 i had random reboots which i fixed but now i get this and this is kinda really serious "oh i have 50% battery i have nothing to worry about" meanwhile it will die in 5 minutes | 06:47 |
| Mister_Magister | i really want to fix that | 06:47 |
| Mister_Magister | spiiroin: is there any way to fix it? or make it last longer at least | 06:53 |
| * Mister_Magister make service that restarts statefs every 5 minutes | 06:53 | |
| spiiroin | Mister_Magister: quick and dirty band aid = partially ignore change notifications, force prop re-read on heartbeat & cable state change | 06:55 |
| Mister_Magister | spiiroin: sorry what | 06:56 |
| Mister_Magister | spiiroin: what about longer more cleaner band aid | 06:56 |
| spiiroin | Mister_Magister: that would be the least bad "any way to fix it" without actually fixing it / not resorting to blindly restarting stuff periodically | 06:57 |
| spiiroin | the real fix is to eliminate needing/relying on notifications from fuse fs changes altogether | 06:58 |
| Mister_Magister | spiiroin: i see so how can i achieve that | 06:58 |
| spiiroin | patch mce | 06:58 |
| Mister_Magister | seems like hard thing to do | 06:59 |
| Mister_Magister | ill try | 06:59 |
| Mister_Magister | i will ask you on the way of fixing kay? | 06:59 |
| * spiiroin ponders about doing it, would take less time than explaining what needs to be done ;-/ | 07:00 | |
| Mister_Magister | spiiroin: sorry man | 07:00 |
| spiiroin | Mister_Magister: np, just bad timing - I'm trying to justify using time for proper fix, and having bandaid makes it less convincing | 07:02 |
| Mister_Magister | spiiroin: no need to hurry nobody uses this port except me | 07:02 |
| Mister_Magister | i can restart statefs from time to time | 07:02 |
| Mister_Magister | i mean i wish somebody would use it :/ | 07:03 |
| spiiroin | Mister_Magister: that is just so ugly solution that I'd rather not make you need such a thing ... | 07:05 |
| Mister_Magister | spiiroin: waiting for real fix then | 07:05 |
| Mister_Magister | and for fix for moto x play led that you said you are taking half year ago :) | 07:06 |
| Mister_Magister | or maybe year ago | 07:06 |
| Mister_Magister | spiiroin: i bet you forgot didnt ya? :P | 07:06 |
| Mister_Magister | spiiroin: https://bugs.merproject.org/show_bug.cgi?id=1813 yep almost year ago | 07:08 |
| merbot | Mer bug 1813 in mce "Add option to set On and Off value in Binary led configuration file" [Enhancement,New] | 07:08 |
| Mister_Magister | spiiroin: ? | 07:22 |
| Mister_Magister | monich: in my other phone after some time ofono goes into loop aka no signal -> signal -> no signal -> signal and it changes in 1 second loop. logs in a second | 07:23 |
| * Mister_Magister wonders where to put logs | 07:31 | |
| Mister_Magister | monich: https://cloud.verdanditeam.com/index.php/s/stLfXS7Sa4QNAK8 | 07:33 |
| spiiroin | testing/review appreciated: https://git.merproject.org/mer-core/mce/merge_requests/92 | 08:05 |
| spiiroin | ^ Mister_Magister | 08:05 |
| Mister_Magister | spiiroin: how can i test that | 08:06 |
| spiiroin | Mister_Magister: you have sdk? check out the branch, build mce, install on device, see if you still see battery prop freezes | 08:07 |
| spiiroin | Mister_Magister: that led thing is not forgotten, but higher priority stuff keeps popping up | 08:07 |
| Mister_Magister | sure :P | 08:07 |
| Mister_Magister | spiiroin: would like to make that on obs it would be easier | 08:07 |
| spiiroin | Mister_Magister: well, that should work too. but I have zero experience with custom obs builds, so I can't help with that. | 08:12 |
| Mister_Magister | spiiroin: i will manage | 08:17 |
| Mister_Magister | spiiroin: https://build.merproject.org/package/show/nemo:devel:hw:motorola:kinzie/mce werks | 08:20 |
| Mister_Magister | and it failed xd | 08:21 |
| *** Piece_Maker is now known as Acou_Bass | 08:21 | |
| Mister_Magister | spiiroin: https://build.merproject.org/package/live_build_log/nemo:devel:hw:motorola:kinzie/mce/sailfish_latest_armv7hl/armv8el | 08:22 |
| Mister_Magister | spiiroin: its problem with code | 08:22 |
| Mister_Magister | and its not even in your changes wow | 08:24 |
| Mister_Magister | spiiroin: also you thought about that im charging phone using wireless charger so usb cable wont be connected right? | 08:25 |
| mgrover | krnlyng, why do you never fork the libhybris with the packaging :P am guessing want the latest upstream? | 08:31 |
| krnlyng | mgrover, when i moved it from my github to mer-hybris i rebased it | 08:33 |
| krnlyng | or do you mean something else? | 08:33 |
| mgrover | krnlyng, ah so the mer-hybris is the most upt ot date one? | 08:35 |
| krnlyng | mgrover, yep | 08:35 |
| krnlyng | mgrover, the android8-initial | 08:35 |
| spiiroin | Mister_Magister: well, that is reported by statefs -> no workaround. but heartbeat/resume checks /should/ help with that too | 08:35 |
| Mister_Magister | spiiroin: sure. would be great if mce would like to compile tho | 08:36 |
| spiiroin | Mister_Magister: you need up to date usb-moded headers | 08:37 |
| Mister_Magister | where can i get those | 08:37 |
| Mister_Magister | update droid-hal? | 08:37 |
| spiiroin | Mister_Magister: usb-moded-devel rpm | 08:37 |
| mgrover | krnlyng, ahh fair enough, time to rock n roll | 08:37 |
| Mister_Magister | spiiroin: lol where is it | 08:38 |
| krnlyng | mgrover, had some success with the kernel? | 08:38 |
| Mister_Magister | spiiroin: if its needed why isnt it in spec file? | 08:39 |
| spiiroin | Mister_Magister: git is here https://git.merproject.org/mer-core/usb-moded | 08:39 |
| Mister_Magister | spiiroin: ^ | 08:40 |
| spiiroin | Mister_Magister: because I forgot to update version deps? ;-) | 08:40 |
| Mister_Magister | :P | 08:40 |
| Mister_Magister | i need to add it to spec then | 08:40 |
| Mister_Magister | which means i need to fork it. so much hassle :P | 08:40 |
| mgrover | krnlyng, yes and no, the problem is definitley CONFIG_VT. But now am migrating bullhead to android 8 in hopes of fixing the bluetooth, then the onyl thing left is fingerprint scanner :P | 08:42 |
| Mister_Magister | spiiroin: welp it didnt help https://build.merproject.org/package/live_build_log/home:mister/mce/latest_armv7hl/armv8el | 08:48 |
| Mister_Magister | and i added BuildRequires: usb-moded-devel | 08:49 |
| Mister_Magister | spiiroin: or do i need newer version | 08:50 |
| spiiroin | Mister_Magister: you need usb-moded >= 0.86.0+mer25. in that spec file it is that "BuildRequires: pkgconfig(usb_moded)", but ... | 08:55 |
| spiiroin | ... now I recall. those pkg config files are stuck at 0.86.0 and I'm not sure if those +merXX things can even be used | 08:56 |
| spiiroin | Mister_Magister: local mce build takes ~ 20 seconds ... | 08:56 |
| Mister_Magister | spiiroin: it built when i built new usb-moded | 09:05 |
| Mister_Magister | spiiroin: i cant install new usb-modded | 09:20 |
| Mister_Magister | it uninstalls tons of packages | 09:20 |
| Mister_Magister | spiiroin: kay i updated only mce. i will let you know in 2 days if it works or nah | 09:22 |
| Mister_Magister | for now everything seems to be working | 09:23 |
| spiiroin | Mister_Magister: ok, thanks (there are changes to mtp handling -> those could trigger overzealous removals unless bunch of other stuff & patterns too are updated - doing forced nodep install of mce should be ok in this situation) | 09:25 |
| Mister_Magister | spiiroin: nah i just made zypper in mce and it didnt even want to update anything else | 09:28 |
| Mister_Magister | oh maybe because of old deps in spec file :) | 09:28 |
| spiiroin | Mister_Magister: mce works with or without usb-moded -> no runtime dependencies -> zypper in just installs mce [but compiling requires up to date headers to be available] | 09:33 |
| Mister_Magister | spiiroin: sure | 09:40 |
| Mister_Magister | spiiroin: what am i testing btw? | 09:41 |
| elros1 | I have bunch of sensorfwd[997]: virtual bool HybrisAdaptor::setInterval(unsigned int, int) setInterval not ok | 09:46 |
| elros1 | in 2.2.1.18 and looks like issues with random orientation/proximity sensor fail returns. | 09:46 |
| elros1 | I switched to proximity-evdev adaptor but can't do this for orientation sensor | 09:47 |
| elros1 | spiiroin: is this setInterval errors critical? | 09:48 |
| spiiroin | elros1: should not be - anymore, knock knock | 10:02 |
| * spiiroin goes check something | 10:02 | |
| spiiroin | elros1: valid range varies from one device to another, sensorfwd applies strict range checks -> client interval request can end up ignored. but since some time ago: every sensor is programmed to have some valid-from-hal-pov interval on startup -> the reporting might happen at different rate from what app requests, but it should still report something | 10:24 |
| spiiroin | so unlikely to be critical, but can still cause some sort of hiccups | 10:25 |
| spiiroin | probably that range check should be changed to "use closest thing supported by hal" rather than "use as-is or ignore" | 10:26 |
| spiiroin | but even then clients would need to be able to handle situations where they do not get exactly what they asked for | 10:26 |
| elros1 | ok, thx I will need to investigate it further | 10:30 |
| ghosalmartin | r0kk3rz, that pipeline is becoming more and more handy by the minute :P | 13:24 |
| r0kk3rz | glad to be of service | 13:25 |
| *** ChanServ sets mode: +v T4 | 14:28 | |
| *** mgrover is now known as ghosalmartin | 15:50 | |
| ghosalmartin | krnlyng, mal finally booted into android 8...everythings broken for now but thats fine :P | 15:50 |
| krnlyng | ghosalmartin, nice, what was the problem? | 15:51 |
| ghosalmartin | krnlyng, atm no ui so ill see whats going on there first | 15:51 |
| krnlyng | ghosalmartin, you mean sfos on android 8? | 15:51 |
| krnlyng | ah ok:) | 15:51 |
| ghosalmartin | yeah sfos on android 8 | 15:52 |
| krnlyng | neat :) | 15:52 |
| krnlyng | ghosalmartin, don't forget surfaceflinger stuff and so on | 15:52 |
| krnlyng | ghosalmartin, take a look at nile repo | 15:52 |
| ghosalmartin | ive nabbed that, might need to steal some ore | 15:52 |
| ghosalmartin | more | 15:52 |
| krnlyng | ghosalmartin, go ahead | 15:52 |
| ghosalmartin | krnlyng, what happens fi arm64-v8a is the product abi? what rc file should it hit then? | 16:06 |
| krnlyng | ghosalmartin, the init.environ.rc stuff? | 16:06 |
| ghosalmartin | krnlyng, yeahh | 16:06 |
| krnlyng | ghosalmartin, none ;) | 16:07 |
| ghosalmartin | ah so we dont need to mess around with ld_library path? | 16:07 |
| krnlyng | we may not | 16:07 |
| krnlyng | otherwise things break | 16:07 |
| krnlyng | also we don't need :) | 16:07 |
| krnlyng | android 8 has long moved away from the old logd kernel stuff | 16:07 |
| ghosalmartin | ahh I see, hopefully after i get servicemanager working it should sort itself out :P | 16:08 |
| ghosalmartin | seen "property_set("ro.boottime.servicemanager", "254293855008") failed: property already set | 16:08 |
| ghosalmartin | " that before? | 16:08 |
| krnlyng | means it crashes :) | 16:08 |
| ghosalmartin | ahh okaystime to see why :P | 16:10 |
| krnlyng | ghosalmartin, https://github.com/mer-hybris/droid-config-sony-nile/blob/master/sparse/usr/libexec/droid-hybris/system/etc/init/vndservicemanager.rc#L2 | 16:10 |
| krnlyng | i don't recall but i think normal servicemanager.rc also requires that | 16:10 |
| ghosalmartin | krnlyng, that seems to of done the trick, now everything has stopped so am guessing its waiting for its boot done signal | 16:15 |
| mal | ghosalmartin: did you copy all rc files from link krnlyng gave | 16:15 |
| ghosalmartin | mal, i did | 16:15 |
| krnlyng | yeah true it's best to copy all | 16:16 |
| ghosalmartin | so whats vndservicemanager compared to the normal one? | 16:16 |
| krnlyng | mal, maybe we should move those to some generic folder starting from android 8 | 16:17 |
| krnlyng | ghosalmartin, there are three binder devices now | 16:17 |
| krnlyng | ghosalmartin, /dev/binder for framework stuff, /dev/vndbinder for stuff that HAL modules previously created and /dev/hwbinder for hardware services (i.e. the replacement for the HAL apis) | 16:17 |
| ghosalmartin | krnlyng, ahh okays :) | 16:24 |
| mal | krnlyng: what kind of folder? | 16:25 |
| krnlyng | mal, not sure | 16:25 |
| krnlyng | need to think myself | 16:25 |
| krnlyng | but basically those are required on all android 8 devices for now | 16:26 |
| ghosalmartin | hmm for whatever reason droid-bootctl is failing. when i ran it manually with LD_LIBRARY_PATH as /system/lib it kicked me noff | 16:28 |
| ghosalmartin | well ti is a 64bit exec :P | 16:29 |
| ghosalmartin | seen this before? Aug 30 21:31:39 Sailfish sh[1239]: CANNOT LINK EXECUTABLE "/usr/libexec/droid-hybris/system/bin/bootctl": library "android.hardware.boot@1.0.so" not found | 16:30 |
| ghosalmartin | i cant actualyl find that file | 16:30 |
| mal | ghosalmartin: maybe your device doesn't need bootctl? | 16:30 |
| mal | ghosalmartin: it's only needed if your device has validation of boot in use | 16:31 |
| ghosalmartin | ahh i see, well nothing else happens then :P | 16:31 |
| mal | ghosalmartin: you will notice if that is the case, device fails to boot after about 5 reboots | 16:31 |
| ghosalmartin | thats always super handy :P | 16:32 |
| ghosalmartin | well atleast test_gps still owrks :P | 16:34 |
| ghosalmartin | well i get to the language picker, then it all goe sbadly after that :P | 16:42 |
| *** ChanServ sets mode: +v T4 | 17:20 | |
| ghosalmartin | mal, krnlyng made it to ui, albeit with a few crashes and reboots in between but UI :D | 17:26 |
| ghosalmartin | lipstick likes ot restart a lot :P | 17:26 |
| krnlyng | heh | 17:27 |
| mal | ghosalmartin: anything about oom in logs? | 17:40 |
| ghosalmartin | mal: don tthink so? https://pastebin.com/ttR3LZeK | 20:06 |
| ghosalmartin | i already have the kernel patch that prevents that | 20:06 |
| mal | ok | 20:14 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!