Sunday, 2019-05-19

pseudodevMy bluetooth and wifi wont work04:36
pseudodevI already did what was mentioned in faq. As far as bluetooth is concerned, "CONFIG_BT_HCISMD" doesnt exist. Making those sepcific changes mentioned in ( gives a compilation error.04:37
pseudodevFor WiFi, I did add those files in my droid-configs/sparse. The modules mentioned in the faq dont exist in my defconf04:38
pseudodevNeed some help here.04:40
T4<austenite> Does your inbuilt WiFi work04:52
pseudodevIt works on andrid04:53
T4<austenite> Pseudodev Processor?04:55
T4<austenite> Try this if other one gives compilation error , worked for me... 8937
T4<austenite> Pseudodev
T4<austenite> wlan works after : echo sta > /sys/module/wlan/parameters/fwpath05:01
T4<austenite> Build wlan as module if giving troubles as module then to use in built try this
pseudodevaustenite: I will try and tell you in 10mins05:07
rinigusmal: thank you very much!05:12
pseudodevaustenite: THnak you so much! My WLAN is fixed05:25
pseudodevany ideas how do i make "echo sta > /sys/module/wlan/parameters/fwpath" change permanent?05:29
T4<austenite> Add systemd service as in link/build as module05:31
pseudodevNow I dont have ril05:36
pseudodevI mean no sim card detected05:37
pseudodevCannot compile the kernel after applying the fix05:49
pseudodevCan anyone tell me what is wrong with my ril?06:06
pseudodevps ax | grep rild
T4<austenite> Pseudodev, make clean and build kernel06:07
pseudodev"/usr/libexec/droid-hybris/system/bin/logcat -b radio"
pseudodevI have attached my ril logs06:08
pseudodevwhat is wrong there?06:08
*** OhYash1 is now known as ohyash07:17
malpseudodev: in 15.1 build and install bluebinder for bluetooth09:38
maland ofono-ril-binder-plugin for modem09:38
malthose should be built automatically, just add those to patterns09:39
malrinigus: do you still need libhybris on onyx? what was the problem with .31?10:27
pseudodevmal: what do I do about the wlan?10:32
pseudodevRunning "echo sta > /sys/module/wlan/parameters/fwpath" makes it work. How can I make it permanent?10:32
malpseudodev: do you have wlan as module?10:40
malpseudodev: if not then something like this if as a module then just modify the wlan-module-load.service and add it as ExecStartPre=10:43
rinigusmal: as far as i remember, there was no gui with .31. should test it again later11:07
rinigusmal: Symptoms were very similar to one other port (bacon?).11:08
*** OhYash1 is now known as ohyash11:19
pseudodevmal: Bluetooth wont work even after building bluebinder11:38
pseudodevcant see ril either11:38
pseudodevI see this in journalctl "/usr/bin/droid/ line 16: /sys/module/hci_smd/parameters/hcismd_set: No such file or directory"11:39
pseudodevi'm building for if that seems relevant11:46
pseudodevin the faq, it shows building for, but I am building for Will the fix mentioned in faq work?11:51
malwhat fix11:52
malalso bluebinder and the old bt instruction are not to be mixed11:52
pseudodevL#703 in faq11:52
pseudodevmal: for bluebinder, I just added it to my hw_daptation pattern and built the package using build_packages11:53
pseudodevDid the same with ofono11:53
malthat 3.0.1 ofono plugin build is not relevant11:54
pseudodevAlright. I just added ofono-ril-binder-plugin to the patterns11:55
pseudodevit built the package on it's own11:55
malpseudodev: those bt and ofono plugins are automatically built when you run without parameters11:55
malofono plugin needs config also
pseudodevYeah. I noticed they got built. But after creating a new rootfs flash file, bluetooh and ril wont work11:56
pseudodevLet me add that .conf you just sent. What about bluetooth?11:57
malsome parts of those might not be relevant, you need at least the ril_0 section, the stuff in settings parts you can comment out11:57
malI cannot help with bluetooth if you don't give logs11:57
pseudodevmal: Do you nearly boot logs or any time log is fine?11:59
malall as usual12:00
pseudodevhot cup of logs coming up...12:01
pseudodevmal: dmesg -
pseudodevjournalctl -
pseudodevlogcat -
rinigusmal: after update to 3.0.3, I get lipstick failure message in journal: lipstick[3976]: /usr/bin/lipstick: symbol lookup error: /usr/lib/ undefined symbol: _ZN8QUsbMode4Mode16ChargingFallbackE12:16
rinigusmaybe should update to libhybris .31 and see if that helps...12:16
malrinigus: do you have any custom packages12:18
malrinigus: that error is not related to libhybris12:18
rinigusmal: probably... depends on how you define custom.12:18
malusb-moded or lipstick12:19
pseudodevmal: did you check my logs?12:19
malpseudodev: only a quick look, is bluebinder running?12:20
rinigusmal: I don't think lipstick is custom, but I'll see how to check them12:22
malrinigus: verify that lipstick and all usb-moded related packages are correct, pkcon search name lipstick for example nicely gives available versions12:24
pseudodevmal: running top, doesnt look like it is running12:24
mal"ps aux | grep blue" is better12:26
rinigusjust a note before I forget it: /lib/systemd/system/systemd-tmpfiles-setup.service required addition of --exclude-prefix=/sys to its exec line12:29
rinigusmal: nothing really strikes as out of order:
rinigusmal: I'm using devel build and made an ota from earlier image (one of the latest 2.x series)12:33
T4<elros34> I think you have too old usb-moded: in it should be 0.86.0+mer3312:35
riniguselros34: thanks. Let me see how to update that12:36
malrinigus: oh, I think there were some issues if you don't do the normal update steps12:42
malcan't remember which ones were the stop releases12:43
pseudodevmal: did you check logs?12:43
malpseudodev: bluebinder is not running, is it really installed, check zypper if bluebinder12:43
malrinigus: did you run pkcon refresh before that search12:44
malrinigus: 3.0.3 should have libusb-moded-qt5 1.812:45
malrinigus: probably something went wrong with the update12:46
malrinigus: also usb-moded should be 0.86.0+mer3312:47
rinigusmal: I did zypper ref -f, as always. there were no error messages after update, as far as I could see12:49
pseudodevmal: it returns nothing12:50
maltry pkcon refresh now and check the versions again, also check ssu lr12:50
malpseudodev: so you don't have it installed, did you add it correctly to patterns, and did you update patterns12:50
pseudodevmal: I added it to patterns. After adding, I ran build_packages. It asked me if i wanted to build bluebinder. I built it. Then i processed patterns and built a new image.12:52
malhow did you add it12:53
rinigusmal: ok, probably caused by jump in versions (2.something-> zypper up refused to installed newer version since its in conflict with buteo-mtp.12:55
rinigusError message: usb-moded-0.86.0+mer33-1.3.2.jolla.armv7hl conflicts with buteo-mtp-qt5-sync-plugin provided by buteo-mtp-qt5-sync-plugin-0.7.0-1.6.1.jolla.armv7hl12:55
rinigusit suggests to uninstall buteo-mtp as a solution. should I go for that?12:55
malrinigus: try "zypper in usb-moded -buteo-mtp-qt5-sync-plugin"12:55
pseudodevmal: I added "- bluebinder" to "/home/pseudodev/hadk/hybris/droid-configs/patterns/jolla-hw-adaptation-z2_plus.yaml"12:57
pseudodevAdded it at the very end before "Summary: Jolla HW Adaptation z2_plus"12:57
rinigusmal: and I have all nicely booted up. will continue testing12:58
rinigusmal and elros34: thank you for help!12:58
malrinigus: I suggest you try zypper dup or version --dup again to make sure all packages are updated13:03
rinigusmal: looks like everything else is fine - nothing to update13:21
pseudodevmal: what do i do?14:10
pseudodevAlso, how to fix the blackscreen on images bug? I already added a new file in /var/lib/environment/nemo putting "Define QT_OPENGL_NO_BGRA=1" and added the lines as specified in faq14:11
pseudodevbrowser works btw14:13
malpseudodev: can't remember if that black images was that or the adreno quirk, check droid-hal-sony-nile in mer-hybris, the common inc there14:27
pseudodevand bluetooth?14:27
T4<elros34> pseudodev: it's "QT_OPENGL_NO_BGRA=1" not "Define QT_OPENGL_NO_BGRA=1"14:29
pseudodevI want to build my wlan as module14:31
pseudodevhow do i know which config to set as m?14:31
rinigusfolks on do you have systemd-tmpfiles-setup.service running correctly? without adding --exclude-prefix=/sys to the end of the command, it was failing on onyx. that file is a part of systemd-225+git13-1.4.2.jolla.armv7hl14:48
malpseudodev: read the faq about wlan modules, it has some examples15:00
malrinigus: what did it cause when it failed?15:00
rinigusmal: no idea. when lipstick was not showing up, I checked which services were down. that was the one. does it work on fp2?15:01
malrinigus: seems to be working15:04
malpseudodev: for bluetooth I already what you need to do, not sure which step you did wrong15:04
rinigusmal: thanks. When I run the command without excluding sys, I am getting 'Failed to create directory or subvolume "/sys/fs/cgroup/memory/noswapping": No such file or directory'. I wonder if I should submit bugreport somewhere regarding it...15:06
pseudodevmal: I basically made all configs with "WLAN" in it as m15:08
pseudodevNow I get this:
riniguskimmoli: so far I had no issues with I just submitted PR and let's build it on testing. this will allow me to test it for some time as a daily driver. would you mind to do your magic and promote hal packages from devel to testing and start the build after merging PR and tagging config? (all other repos stayed the same)15:09
Mister_Magistermal: did you have time to look at my log?15:10
malMister_Magister: I had a quick look, nothing obvious so far15:16
malrinigus: maybe you are missing some cgroup kernel config?15:16
Mister_Magistermal: will try journal15:30
Mister_Magisteri mean logcayt15:30
malrinigus: can you check on device if some configs are enabled: "zcat /proc/config.gz | grep MEM_RES_CTRL"15:38
malrinigus: why do you have two commits in the config PR, maybe you could squash those to one?15:52
masya_Hi! I get this error during make hybris-hal15:57
masya_arch/arm64/Makefile:47: *** CROSS_COMPILE_ARM32 not defined or empty, the compat vDSO will not be built.  Stop.15:57
masya_How can I fix it?15:57
malmasya_: pseudodev had the same error I think16:03
pseudodevmayasa: clone this:
pseudodevmasya: Then run: export CROSS_COMPILE_ARM32=$(pwd)/bin/arm-linux-androideabi-  #( Note: This command must be run in "$ANDROID_ROOT/prebuilts/gcc/linux-x86/arm/arm-linux-androideabi-4.9")16:06
pseudodevmasya_: ping ^16:06
pseudodevSorry got your name wrong twice :P16:07
pseudodevoh. He quit16:07
pseudodevmal: what do i do about the bluetooth16:09
pseudodevI tried it again. Same thing16:09
masya_pseudodev: I'm here! There are some troubles with internet connection.16:14
pseudodevmasya_: Which device btw?16:14
masya_Xiaomi Mi 6x (wayne)16:15
T4<elros34> rinigus: I have the same systemd-tmpfiles-setup error but I don't have all cgroup recommended settings in kernel16:15
rinigusmal: re number of commits: those are easy to squash on merge. I have two since i updated to the last dcd first, failed to build packages, checked logs and found that but older version is needed.16:20
rinigusRe cgroup: I'll check, thanks16:20
Mister_Magisteralso mal on g2 i seem to don't have internet connection when i connect to internet16:51
Mister_Magisteroh journal might not be useful i forgot i should take logcat16:52
Mister_Magisteragain mal
Mister_Magisteralso abranson if you could please look at it maybe something will catch your eye16:57
rinigusmal: CONFIG_CGROUP_MEM_RES_CTLR is not set.17:00
Mister_Magisterrinigus: if that's newer then it's okay iirc17:01
Mister_Magisterthere was change in name at some point17:01
rinigusI wonder if there is any reason for it.17:01
rinigusMister_Magister: I wouldn't count on our onyx kernel to be "newer". if I understood you correctly17:02
Mister_Magisterwhat is kernel version of it then17:02
Mister_Magisterthen it uses old names17:05
rinigus(uname -a)17:05
Mister_Magisterso CONFIG_CGROUP_MEM_RES_CTLR should be set17:05
rinigusso, is it recent when this setting was recommended? recent is within 1-2 years?17:06
Mister_Magisterit was there  since i can remember17:06
rinigusok, then that's strange...17:07
riniguskimmoli: any side effects for enabling CONFIG_CGROUP_MEM_RES_CTLR that you remember about?17:07
Mister_Magisteroh wait no i'm confusing things17:08
Mister_Magistermy 3.4.X moto g2 doesn't have it17:08
Mister_Magistersorry for causing confusion17:08
Mister_Magisterit has to be recent then17:08
Mister_Magisterno i'm confusing people again17:08
Mister_Magisternow this is correct and this was there since always17:09
Mister_Magistersorry for AGAIN causing confusion17:09
Mister_Magisteri just read your message incorrectly17:09
Mister_Magisterthird commit in this repo adds this so it was since forever17:13
malrinigus: not older version, you should use the upgrade-3.0.3 branch17:17
kimmolirinigus: not that i remember, found old issue referencing it
rinigusmal: it's using the same commit as in fp2 (git hard reset), should be the same branch - detached state for git17:23
riniguskimmoli: i guess we should enable it then.17:23
malrinigus: ok17:26
malrinigus: I think you should enable that CONFIG_CGROUP_MEM_RES_CTLR, kernel checker says that is recommended17:27
rinigusmal: will do17:44
T4<elros34> can we have a custom 999-android-system.rules ? systemd-udevd spams a lot in 3.0.317:57
piggzmal: for build of qtbase, i think I want eglfs_kms, but do I also want eglfs_kms_egldevice ?18:02
malpiggz: not sure what the difference is18:04
mal@elros34 what kind of errors?18:04
T4<elros34> mal: bunch of "Process '/bin/chown system:system /sys//devices/system/cpu/cpu1/cpufreq/*' failed with exit code 1." when one cpu core is offline18:07
piggzmal: edicated myself, i think the answer is no
piggzsecond is for tegra devices18:08
mal@elros34 hmm, I seem to get the same18:09
T4<elros34> maybe solution would be to add ACTION=="online" to udev rule generator18:10
mal@elros34 it seems the paths don't exist18:10
mal@elros34 hmm, it seems the path disappear if the cpu is inactive18:12
T4<elros34> yes, it exist when action is online18:13
mal@elros34 maybe because of this
malassuming you rebuilt droid-hal with latest dhd submodule18:19
T4<elros34> on 3.0.2 looks like these cpu1/cpufreq/ rules are incorrect so they are never executed and never fail:)18:29
mal@elros34 yes, I think so, wondering why it still tries to run that if the path isn't there18:53
mal@elros34 you can always downgrade the dhd submodule19:09
riniguskimmoli: new kernel/hal is uploaded to obs devel. I'll be able to test it tomorrow night. if all is ok, we can promote it to testing.20:08
malrinigus: do you get those chmod failed errors in journal?20:10
rinigusmal: I'm back on 3.0.2 for now (its my daily driver, has to work), will check it tomorrow. on 3.0.2 - don't have them.20:15
malrinigus: I did my fp2 3.0.3 testing with my daily driver :)20:15
malrinigus: it's only if you updated dhd submodule within last 2 weeks and rebuilt droid-ahl20:16
malrinigus: it's of course a bit risky to test like this but I'm used to it20:17
maland I rarely break it20:17
rinigusmal: I intend to do the same way with onyx testing. but for that, it will have to be build using obs testing. right now we are on devel stage - will hopefully move on to testing tomorrow if all is ok with recompiled kernel.20:17
rinigusmal: no, I kept dhd the same as we had for 3.0.220:18
rinigusmal: ^ that decision was based on fp2 changes20:19
rinigusgn! that's it for me for today, will catch up tomorrow.20:19
malrinigus: hmm, so you didn't update system/core?20:25

Generated by 2.17.1 by Marius Gedminas - find it at!