Friday, 2018-07-06

UmeaboyI'll type arm64 instead.00:03
UmeaboyI managed to get into the PlatformSDK, but as I re entered it to solve the error with the symbolic links I saw this:00:19
*** ubuntu is now known as Guest8001501:50
*** Guest80015 is now known as zhxt02:08
branek_Hi all... still not able to see any gui on my device... lipstick is failing with08:56
branek_[F] unknown:0 - QPA-SF: sf_client in create returned 008:56
branek_Problem choosing egl config on creation08:56
malbranek_: you expect us to be able to help if you don't provide full logs09:37
branek_mal: this is output of EGL_PLATFORM=null QT_QPA_PLATFORM=surfaceflinger strace -o lipstick.txt lipstick10:00
malbranek_: I'm pretty sure I asked you before to provide the following logs: dmesg, journalctl, systemctl and /usr/libexec/droid-hybris/system/bin/logcat10:04
branek_mal: dmesg: journalctl: systemctl: logcat:
malcould you give a link to your adaptation repos, and make sure those are updated10:21
branek_mal: ok, need some time to update everything10:30
NeoChapay_happy new HADK
*** NeoChapay_ is now known as NeoChapay10:59
locusfsledges: wot11:10
branek_mal: dhd -
branek_hybris-boot -
branek_droid-config-dumpling -
branek_dhc -
krnlyngbranek_, hybris-boot has been merged already11:15
branek_krnlyng: i'm still do not understand whole build tree and build cycle, so there is things like that... fill clean up once I got some progress...11:18
malbranek_: the /var/lib/environment/compositor/droid-hal-device.conf looks wrong, you have added something to LIPSTICK_OPTIONS line, not sure why you have added "-plugin surfaceflinger" there, that is not in the reference repo11:19
branek_mal: saw that in some discussion here i think... I tried also without that on first try...11:22
malbut was that when you had other problems?11:22
mallike the missing package11:22
malbranek_: also this is wrong
malthat should not be a file, it should be a symlink11:25
malbranek_: so you go to that folder and remove the surfaceflinger.service and then run "ln -s ../surfaceflinger.service surfaceflinger.service"11:26
malbranek_: I think also the other symlinks are wrong11:27
malthe firmware symlinks in two folders11:28
malso in here and
branek_aha... I see... when I implemented changes, i made files insted of linkd...11:29
malbranek_: you probably misunderstood because github web interface shows the symlinks in a not so clear way11:30
branek_Is it ok to fix it on device now and try?11:30
riniguskimmoli: anything on eta for pr's and moving devel to testing 2.2.0.x? or the weather is too good to waste the time on that?11:30
Mister_Magistermal: after updating to 2.2 (without changing dhd) udev rules failed and so there is no /dev/block/platform. any idea what could cause that?11:32
malMister_Magister: what do you have there?11:33
kskarthik[m]whats the easier way to posrt sailfish, halium or hadk?11:46
Mister_Magistermal: what?11:46
malMister_Magister: ls -lR /dev/block/11:47
malMister_Magister: dhd has nothing to do with udev rules, at least not those device symlinks we add11:48
malMister_Magister: those are completely handled in dhc11:48
malMister_Magister: you are sure you update dhc submodule to latest?11:49
Mister_Magistermal: yeah12:06
branek_mal: I fixed linking... no change. on nile there is /etc/firmware -> /odm/firmware/... but I do not have that odm dir..12:06
malbranek_: then you don't need that12:06
Mister_Magistermal: i just said i dont have /dev/block/platform in /dev/block/12:06
malMister_Magister: so does that cause boot failure?12:07
Mister_Magistermal: no but modem failiture12:07
Mister_Magisterand modem causes reboot12:08
branek_mal: same result... lipstick does not run on init... fails with previous error: [F] unknown:0 - QPA-SF: sf_client in create returned 0 - Problem choosing egl config on creation12:09
malMister_Magister: check the date of the submodule update here and compare to changelog here
malMister_Magister: you might notice something12:09
malbranek_: did you fix the LIPSTICK_OPTIONS ?12:10
Keijo__mal: /dev/block/platform and /dev/block/bootdevice disappeared after on lux and now the kernel panics12:11
Mister_Magistermal: that its outdated?12:11
malMister_Magister: yes12:11
Mister_MagisterKeijo__: we are talking about it12:11
Mister_Magistermal: well its new enough for 2.212:11
Mister_Magisteri updated is along with others the same day12:11
Keijo__K, went to do other things in the meantime12:12
malMister_Magister: I don't understand what you have done anymore12:12
Mister_Magistersorry in the same day i updated this submodule i updated dcd in other devices too12:12
malMister_Magister: you were talking about removing the udev rules and didn't update submodule or what?12:12
Mister_Magisterand other devices work12:12
Mister_Magisteri mean updating submodules could help but thats weird that this submodule doesnt work only on one device12:14
Mister_Magisterand i dont see any commits udev related12:15
malhave you tried the new udev rules on that device or not?=12:15
malif the rules don't work then fix those12:15
Mister_Magisterwhat new udev rules?12:16
malalthough that was older than I remembered12:16
malwhat rules are you using now?12:16
malMister_Magister: I'm still lost, are you using these rules or not
maland if yes how is it possible it broke12:17
Mister_Magistermal: im lost too12:18
Mister_Magisteryes yes i do12:18
Mister_Magistereverything that is in that repo is in use12:18
malso you are saying the rule just randomly stopped working? that doesn't make any sense12:19
Mister_Magisterafter updating to 2.212:19
Mister_Magisteri know it doesnt make sense12:19
Mister_Magisterbut 2.1.4 works 2.2 nope12:20
Mister_Magistershould i remove that custom file from my sparse and  it will use dcd sparse?12:20
malI suggest you try first on device12:21
Mister_Magisterthe one from sparse? gotcha12:22
Mister_Magisteryeah as i look now other devices dont thave that rule12:22
maljust replace the old rules on device with the new ones and see how that works12:22
Mister_Magisteryeah i got you12:22
malso you copy this to your device
Mister_Magistermal: i seriously seriously sincerely truely got you12:23
Mister_Magistermal: seriously :) 3 messages ago :)12:24
Mister_Magisteri know you want to make sure12:24
Mister_Magisterim not suprised at all…12:24
malI never trust anyone here anymore12:24
Mister_Magister[14:24] <Mister_Magister> im not suprised at all…12:24
Mister_Magisterthe fact that you still want to help is amaizng12:25
Mister_Magistermal: you still cant sleep?12:26
Mister_Magisternope means as you cant or you can12:27
malI can't12:28
branek_mal: when I start, I only get OnePlus logo, and it stays there forever... it looks like display is never initialised...12:38
malwell that's obvious if lipstick fails12:38
malbranek_: update your config repo in github again12:38
Mister_Magistermal: yep these new udev rules work12:41
malthought so12:41
Mister_Magisterso just remove old one from my sparse and they will be get from dcd sparse right?12:41
Mister_Magistergooootcha :P12:42
malMister_Magister: also if you update to latest submodule this file is not needed anymore
malbranek_: you didn't need straggler_files in droid-hal spec?12:46
malbranek_: this one
Mister_Magistermal: but i still need to add one rule12:48
Mister_Magisterotherwise modem will fail12:48
branek_mal: check now, it's updated...12:50
Mister_Magistermal: i guess add updated udev and your rule into it12:50
malbranek_: you are missing droid-hal-dumpling-detritus from patterns (add it somewhere close to here
malMister_Magister: what?12:53
Mister_Magistermal: i need this rule to have modem working12:53
malMister_Magister: the new rule should handle all cases, I thought you just tried the new rule?12:53
Mister_Magisterit doesnt handle this case12:54
maldid you replace the old rule or not?12:54
Mister_Magisterand it helped12:54
Mister_Magisterbut not exactly12:54
malyou are confusing me again12:54
Mister_Magisterill put it simple12:54
malyou first said the new rule worked but now you say it doesn't work12:54
Mister_Magisteri got this file which you gave me and put it on device replaced old12:54
malmake up your mind12:55
Mister_Magisterit created /dev/block/platform/soc.0/ 7824900.sdhci/ by-name/12:55
Mister_Magisterso its working becuse before i didnt have /dev/block/platform folder12:55
Mister_Magisterbut i need /dev/block/platform/soc.0/by-name/12:56
Mister_Magisterand  this creates it12:56
malso why did you say the new rule worked if it didn't12:57
Mister_Magisterit created folder but wrong one :P12:57
Mister_Magisterso it worked but didnt helped with my problem :P12:57
malhow can you count that as working?12:57
Mister_Magisterprevious didnt create folder at all12:57
malkrnlyng: some issues the device udev rules again ^12:58
branek_mal: added and building again...12:58
malbranek_: you could have copied the rpm to the device and installed it manually13:02
krnlyngmal, Mister_Magister ? why do you need  /dev/block/platform/soc.0/by-name/ ?13:05
malI have no idea13:05
malhe says it's needed, not 100% sure about that13:06
krnlyngis it needed by some script that you created yourself Mister_Magister ? because in android it shouldn't be needed unless i overlooked something13:06
demonaxshHello. I switched to a newer version (2.2.0) and when building rootfs:13:07 - returning FAIL to postpone oneshot to first boot13:07
demonaxshoneshot: /etc/oneshot.d/100000/ - FAIL13:07
demonaxshqemu: uncaught target signal 11 (Segmentation fault) - core dumped13:07
demonaxsh/etc/oneshot.d/100000/zz-initialize-tracker-index: line 11:  7171 Segmentation fault      (core dumped) tracker daemon --start13:07
demonaxshqemu: uncaught target signal 11 (Segmentation fault) - core dumped13:07
demonaxshqemu: uncaught target signal 11 (Segmentation fault) - core dumped13:07
demonaxshTracker miners busy, waiting...13:07
demonaxshqemu: uncaught target signal 11 (Segmentation fault) - core dumped13:07
demonaxshTracker miners busy, waiting...13:07
demonaxshI'm bulding on Ubuntu 16.04 LTS13:07
Mister_Magisterkrnlyng: sure its needed by ril13:07
krnlyngalso if you have the proper rules, then it should create /dev/block/platform/soc.0 as a fallback. even though it shouldn't be necessary13:08
Mister_Magister>load_utags ERR opening (/dev/block/platform/soc.0/by-name/utags) errno13:08
Mister_Magisterril needs this path and it doesnt exist13:08
krnlyngMister_Magister, check why this rule fails then:
Mister_Magisterkrnlyng: but how?13:10
branek_mal: done that way, install rpm, restarted, nothing changed...13:13
malbranek_: make sure you see the files listed here on your device
branek_everything is there13:17
branek_mal: ^^13:17
Mister_Magisterkrnlyng: how can i check why it fails?13:20
malbranek_: get new logcat13:30
malbranek_: there seems to be some permission issue13:30
*** Nokius_ is now known as Nokius13:32
malkrnlyng: do you remember this error "qdgralloc: Init: Failed to open ion device - Permission denied"13:35
bshah /dev/ion permissions not good?13:36
malyes, but why is the question13:36
krnlyngmal, most likely some ueventd rule processing issue13:37
malbranek_: what do you get when you run "ls -l /dev/ion" on your device13:37
branek_mal: as root: crw------- 1 root root 10, 126 May 29 20:56 /dev/ion13:38
branek_mal: sam as nemo...13:39
malbranek_: go to your build environment and in there to $ANDROID_ROOT, then run "find out/ -name ueventd.rc"13:42
branek_mal: out/target/product/dumpling/root/ueventd.rc13:44
malbranek_: show the file /lib/udev/rules.d/999-android-system.rules from your device13:46
malbranek_: you need to check if any .rc file under out/target/product/dumpling/ has something about device called /dev/ion13:50
malin your build environment13:50
branek_ok.. doing that now...13:50
branek_mal: nothing in *.rc13:57
branek_only one line in out/target/product/dumpling/root/plat_file_contexts13:57
malbranek_: hmm, very odd, need to invetigate a bit more13:58
malbranek_: can you find that anywhere in device repos, maybe just run this in your build enviroment: grep -RI "dev/ion" device/14:02
branek_mal: device/oneplus/msm8998-common/init/vendor/ueventd.rc:/dev/ion                  0664   system     system14:07
branek_mal: do you need whole  file?14:07
malthe question is why is that content not in out/14:07
malbranek_: try running grep -RI "dev/ion" out/target/product/dumpling/14:08
malfor some odd reason the ueventd.rc is not copied to out/ the build rules tell it should have been copied14:22
branek_it could be that I made some mistake...14:24
maldemonaxsh: it's normal to get various errors or warnings from mic build, as long as you get a zip at the end it should be fine14:33
demonaxshYes, I see. Rebuilding on the new version, even launched a user session, although now the screen is black, but already progress.14:36
demonaxshStrange, but now password for ssh not nemo, I thought that this is always  by default.14:36
malbranek_: I think you should change the dhd submodule and hybris-boot to use upstream again, just remember to add the fixup-mountpoints changes back to hybris-boot after changing the branch14:36
maldemonaxsh: no, password needs to be set after boot, for example via telnet or from settings if gui is up14:37
demonaxshok, where can I change it?14:38
demonaxshfrom telent14:38
branek_mal: ok... will do... will tell you results little bit later...14:39
maldemonaxsh: "passwd nemo"14:40
demonaxshyeees, it's linux, i fogot14:42
malbranek_: remember to rerun make hybris-hal after changing hybris-boot repo15:01
sailor97Finally camera is working on kenzo, cm-14.1 base16:26
sailor97but, i still have an issue, after stopping video recorder I can no longer use camera without rebooting16:26
T4<adampigg> What did you do?16:28
sailor97Cam sensor defaults to HAL3 and wouldn't connect. Forcing it to HAL1 worked16:29
sailor97FM is another story16:29
sailor97Just one error, FM RECV -1916:29
sailor97No clue what it means16:29
malsailor97: do you see anything in logcat after video recording?16:31
sailor97mal: af_map_input_dciaf_info:620  input is invalid16:32
sailor97will give you a pastebin in a minute16:32
malsailor97_: are photos working without problem?16:42
malsailor97_: did the video even record really?16:43
sailor97_mal: photos are working fine but won't open in gallery17:07
sailor97_Video is partly recorded17:07
malsailor97_: for gallery you probably need this
malsailor97_: what do you mean part recorded?17:09
sailor97_mal: Suppose i've recorded 20 sec video, only 10 seconds are actually recorded17:11
sailor97_typo, i meant to say partially recorded17:11
malsailor97_: but you can still watch the videos?17:12
sailor97_Do you mean in the gallery?17:12
malyes, or on your computer if you copy the video file there17:14
sailor97_yes, i can watch it17:14
T4<adampigg> sailor97_: for fm, you could try my patched version if the iris plugin, its a xiaomie device so might work the same way as mido17:15
sailor97_adampigg: sure, can you point me to the source?17:18
mal@adampigg or patch the kernel :P17:18
malsailor97_: just curious, did you do anything with camera.fifo.disable ?17:19
sailor97_mal: yes, I did17:21
malsailor97_: ok, good to know17:22
malsailor97_: that didn't help with hal3?17:22
sailor97_mal: no, hal3 didn't even give any error. It just says exit code 0 for pid some no.17:23
sailor97_mal: here is a video captured using SFOS
T4<adampigg> mal: the ficticious kernel patch ;)17:34
mal@adampigg even I had to patch kernel to get fm radio working on fp2, it also had that kind of thing that it needed to write to /sys17:38
piggzmal: is that the extent of patches ... quite small i agree17:41
malpiggz: that's what I needed17:41
sailor97_piggz: still not working but that error -19 is gone17:42
malsailor97_: do you see any errors? is your headphone detection working?17:43
piggzsailor97_: did you define the variable?17:43
sailor97_mal: did you see that video? Actual duration was around 1 min but it has recorded only a few sconds17:44
sailor97_Yes headphone detection is working17:44
malsailor97_: I couldn't see the video17:44
sailor97_piggz: Yes, I've defined the variab;e17:44
malsailor97_: maybe some permission issue in youtube?17:44
malalso maybe this is needed ?17:45
sailor97_mal: Sorry, I've uploaded it as private. Please check now17:46
malsailor97_: yes, now it worked17:46
Holmes466Hi, when i try make a command:17:55
Holmes466I get this error:17:55
Holmes466* Building rpm/droid-hal-pioneer.spec17:55
Holmes466Fatal: sony-pioneer-armv8-a is an invalid sb2 target17:55
Holmes466* Check /home/hadk/droid-hal-pioneer.log for full log.17:55
Holmes466!! building of package failed17:55
merbotHolmes466: Error: "!" is not a valid command.17:55
sailor97_mal, piggz: they didn't work as well. I haven't tried the kernel patch though. Actually the frequency is always at 0.0MHz17:55
piggzsailor97_: have a look in /system/etc for any scripts that may configure the radio17:56
piggzome moment....17:56
malsailor97_: do you have built iris-transport as module?17:57
malpiggz: did you need to do that btw17:57
piggzmal: i cant remmeber ... but it rings a bell17:57
sailor97_mal: i've built it as a module. default is y in kernel config17:58
piggzmal: sailor97_:
demonaxsh library "" not found18:00
demonaxshMay 29 15:02:10 Sailfish jolla-startupwizard-pre-user-session[1426]: libminisf is incompatible or missing. Can not possibly start the SurfaceFlinger service. If you're experiencing troubles with media try updating droidmedia (and/or this plugin).== hwcomposer module ==18:00
demonaxshwithout gui not worked? now, screen is black18:00
maldemonaxsh: which android base was that?18:01
demonaxshandroid 6.018:01
maldemonaxsh: what exactly is your problem?18:02
demonaxshVery simply, the session of the user started, but the screen after booting, just black18:05
malwhich device was that?18:06
maldemonaxsh: libminisf is built with droidmedia18:06
demonaxshI see, and I can build it's, I just do not know if it's necessary, and it's strange that it was not in the rutfs.18:09
malbecause it's optional, some devices need it but not all18:10
demonaxshit's mtk device18:11
maldemonaxsh: could you provide some logs after boot attempt, dmesg, journalctl, /usr/libexec/droid-hybris/system/bin/logcat18:11
demonaxshyes, but jornalctl very looong18:12
maldemonaxsh: for some reason there is no sign of lipstick process in journal log18:22
malhmm, but actually that might not be a problem yet18:22
maldemonaxsh: is there any startupwizard process running, jolla-startupwizard-pre-user-session or something like that?18:25
demonaxshjolla-startupwizard-pre-user-session? this is in the log18:28
demonaxshMay 29 15:02:10 Sailfish jolla-startupwizard-pre-user-session[1426]: libminisf is incompatible or missing. Can not possibly start the SurfaceFlinger service. If you're experiencing troubles with media try updating droidmedia (and/or this plugin).== hwcomposer module ==18:28
maldemonaxsh: I asked if the process is running18:29
maldemonaxsh: ps aux | grep startupwizard or something like htta18:29
demonaxshaah, now I'll take a look18:30
Mister_Magistermal: krnlyng: how can i debug udev rule?18:31
maludevadm or something that18:31
demonaxshnemo      1426  1.4  3.8 250828 73848 ?        Ssl  15:02   3:33 /usr/bin/jolla-startupwizard-pre-user-session -plugin evdevtouch:/dev/input/event0 -plugin evdevkeyboard:keymap=/usr/share/qt5/keymaps/droid.qmap18:33
merbotdemonaxsh: Error: Error getting Nemo bug #1426: InvalidBugId18:33
maldemonaxsh: so now get the /usr/libexec/droid-hybris/system/bin/logcat output I asked18:34
maldemonaxsh: maybe you should test to see if building libminisf helps, do chapter 13.2 in hadk pdf, note that the is a new version released of the pdf
malmake sure you have new enough external/droidmedia (run repo sync external/droidmedia)18:37
demonaxshyes, I will try18:37
maldemonaxsh: the process seems to be running but nothing is shown on display, but without logcat output I cannot tell if there are any errors in the android side18:39
sailor97_mal, piggz: fm is working, needed all of those fixes. Thank you...18:53
piggzsailor97_: \o/18:53
demonaxsh logcat output18:53
sailor97_kenzo build is almost stable now!!18:54
sailor97_piggz, mal: additionally i had to make radio_iris as a module too18:55
branek_mal: I finished make clean, than make hybris-hal, and than build packages and rootfs.. regarding ueventd.rc and /dev/ion, nothing changed on host, it looks like .rc is not copied again19:07
branek_mal: same on device - no changes19:07
malbranek_: maybe then manully edit the /lib/udev/rules.d/999-android-system.rules to set the correct permissions on device until we figure out why the file is not copied19:15
*** sailor97_ is now known as akhil_surabhi19:15
malbranek_: I have there KERNEL=="ion", MODE="0664", GROUP="system", OWNER="system"19:16
malbranek_: you can check the file here for reference
malbranek_: it should be simple thing to determine what rules you might need19:17
malbranek_: or other option is to manually append the content of that file to the ueventd.rc in out19:17
malafter building hybris-hal and before build_packages.sh19:18
branek_cool... will try that...19:26
branek_mal: and, thanks for your patience...19:27
steeevelbt: obs repo plz19:52
malsteeeve: you should give all needed information when asking19:54
mallike which repo and username if it's not the same as your IRC nick19:54
steeeveUsrnm: deadsquirrel, repo nemo:devel:hw:samsung:a5ultexx19:59
* branek_ is so happy! GUI is on on OnePlus 5T (dumpling). And WiFi working out of the box! hahaha! \o/22:05
UmeaboyWhy does this line show an error about changing to double quotes when the HADK doesn't use it?22:25
Umeaboyecho 'alias sfossdk=$PLATFORM_SDK_ROOT/sdks/sfossdk/mer-sdk-chroot' >> ~/.bashrc22:25
UmeaboyAnd this: echo 'PS1="PlatformSDK $PS1"' > ~/.mersdk.profile22:26
UmeaboyI tried switching ' for " on both sides, but it's still wrong.22:26
malbranek_: good to hear, we still need to solve the build issue at some point22:35

Generated by 2.17.1 by Marius Gedminas - find it at!