| T42 | <edp_17> @elros34 : Do you have idea why there is no IMEI number? | 00:07 |
|---|---|---|
| T42 | <elros34> sorry, no | 00:08 |
| T42 | <edp_17> Ok, no worries. | 00:08 |
| T42 | <edp_17> The other one. You may remember how we fixed the mount points. (We removed the block from the path.) | 00:09 |
| T42 | <edp_17> I have updated the fixup-mountpoints script but when I created a new image recently, the block still was in the path. I thought if I change the fixup-mountpoints script that would be enough. | 00:10 |
| T42 | <elros34> mount units are provided by droid-hal so you must rebuild it | 00:13 |
| T42 | <edp_17> Because I needed to add some stuff into the kernel I did run both 'make hybris-hal' and 'rpm/dhd/helpers/build_packages.sh'. | 00:18 |
| T42 | <edp_17> I will try again and will see what happens. | 00:18 |
| T42 | <edp_17> My other question is how can I add the wifi driver .ko file by default into the package? In the last image that was also an old one and I needed to manually copy over it from the out/target/product/i9100/obj/KERNEL_OBJ/drivers/net/wireless/bcmdhd folder. | 00:21 |
| T42 | <elros34> it should be packaged, check droid-hal-Device.log if it's not inluded in droid-hal*rpm | 00:24 |
| T42 | <edp_17> I have searched for dhd.ko and dhd in that log and there is nothing. | 00:26 |
| T42 | <edp_17> Does it help if I manually copy the dhd.ko from 'out/target/product/...' into 'mer/android/droid/out/target/product/i9100/system/lib/modules/' or this will be overwritten when I run the build_packages or make droid_hal? | 00:30 |
| T42 | <elros34> yes, it should help. You are second person with such a issue, maybe something is broken | 00:32 |
| T42 | <edp_17> Thanks. Yeah, I remember that I read a few days back. | 00:33 |
| T42 | <edp_17> Maybe the mount points also didn't take effect because of a similar reason? | 00:37 |
| T42 | <edp_17> Maybe you know something about this issue: Only two statuses of sound settings: silent or sound. The volume rocker moves the sound bar up and down at the top of the screen accordingly but doesn't increase/decrease the sound level. (Once the bar reaches 0 the sound disappears. Once the bar reaches anything more than 0 the sound comes ba | 00:45 |
| T42 | ck and its level remain is the same, seemingly the volume rocker doesn't make any effect. | 00:45 |
| T42 | <elros34> Pure guess. You can set various quirks for pulseaudio droid module, maybe check no_hw_volume: https://raw.githubusercontent.com/mer-hybris/pulseaudio-modules-droid/master/src/common/droid-util.c | 00:54 |
| T42 | <elros34> you set it like here: https://github.com/elros34/droid-config-moto_msm8960_jbbl/blob/master/sparse/etc/pulse/arm_droid_card_custom.pa | 00:55 |
| T42 | <edp_17> Thank you. | 00:58 |
| T42 | <edp_17> You may also can suggest something about the strange "screen refresh delay" how I call it issue. | 01:05 |
| T42 | <elros34> how about udevadm monitor when you touch screen? | 01:09 |
| T42 | <edp_17> It looks like the screen only gets refreshed when I touch it. (Not always but in most cases.) Example: I am in fingerterm and typing characters. The character that I tap pops up and then remains visible until I tap on another one. When I switch to numbers I see the numbers but when I switch back to characters I see the numbers until I | 01:09 |
| T42 | press something. As I don't see the character keyboard I can only tap on a number but at that moment the number keyboard switches to show the characters and the pressed character gets into the command line. | 01:09 |
| T42 | <edp_17> Another example: still fingerterm, if I rotate the screen it is not rotated until I touch the screen. Well, it is rotated just the screen doesn't get refreshed until I touch it. | 01:10 |
| T42 | <edp_17> Have you came accros with this or similar? | 01:10 |
| T42 | <elros34> You use old hwcomposer, probably nobody else us it anymore. How about you finally tell whether you have udev spam when you touch screen? | 01:14 |
| T42 | <elros34> looks like missing vsync | 01:14 |
| r0kk3rz | didnt you say you've set it to fbdev? | 01:16 |
| T42 | <edp_17> Well, I ran the udevadm monitor but it shows only two lines: UDEV and KERNEL. And there is no flood. even when I touch the screen. I didn't say this because I wasn't sure what I supposed to see on the screen. | 01:16 |
| T42 | <elros34> ok | 01:17 |
| T42 | <edp_17> I only see these 4 lines and time to time it is growing with 2: https://pastebin.com/riZp7X97 | 01:18 |
| r0kk3rz | you can also turn on graphics monitoring in the developer settings | 01:19 |
| T42 | <edp_17> I have realized a nice surprise on the device. Originally it has two soft keys next to the home button at the bottom middle. The one on the right works as a back button. 😄 | 01:19 |
| T42 | <edp_17> @r0kk3rz: I have tried 4.4 from the hadk-faq, but that made the whole experience worse, so removed them. | 01:21 |
| r0kk3rz | but what does the framerate monitor look like? | 01:22 |
| r0kk3rz | super red? | 01:22 |
| r0kk3rz | which isnt that surprising if you set it to fbdev | 01:23 |
| T42 | <edp_17> Not at all. When I start an app it gives me red but then 98% green and sometimes red. | 01:23 |
| r0kk3rz | thats ok then | 01:23 |
| T42 | <edp_17> Yeah, as I said I tried both from that chapter but made it worse so I removed them. | 01:24 |
| r0kk3rz | but getting the hwcomposer backend running would probably improve it a lot | 01:24 |
| T42 | <edp_17> How can I do that? | 01:25 |
| T42 | <edp_17> Should I compile hwcomposer for this device? | 01:25 |
| r0kk3rz | it doesnt have it already? | 01:26 |
| T42 | <edp_17> Yes, you are right. 😄 I meant an older version. | 01:26 |
| r0kk3rz | run EGL_PLATFORM=hwcomposer test_hwcomposer | 01:28 |
| T42 | <edp_17> I have just changed the EGL_PLATFORM to hwcomposer and rebooted. 😄 | 01:28 |
| r0kk3rz | ok, but if lipstick doesnt work then do that | 01:29 |
| T42 | <edp_17> Ok. | 01:29 |
| T42 | <edp_17> No, it didn't work so I changed back to fbdev. | 01:32 |
| T42 | <edp_17> The 'EGL_PLATFORM=hwcomposer test_hwcomposer' gives me: 'test_hwcomposer: test_hwcomposer.cpp:185: int main(int, char**): Assertion `err == 0' failed. Aborted' | 01:33 |
| T42 | <elros34> you need to stop/mask user@100000.service first | 01:34 |
| T42 | <edp_17> In droid-hal-device.conf, should I change the QT_PA_PLATFORM=hwcomposer to fbdev? | 01:34 |
| T42 | <elros34> no, there is no qpa fbdev platform only: https://github.com/mer-hybris/qt5-qpa-hwcomposer-plugin | 01:35 |
| T42 | <edp_17> @elros34 : ok, thanks. | 01:37 |
| T42 | <edp_17> I stopped the user@100000.service and rerun the 'EGL_PLATFORM=hwcomposer test_hwcomposer'. It gave me: 'Segmentation fault' | 01:38 |
| T42 | <edp_17> Tried to strace but I got: 'strace: Can't stat 'EGL_PLATFORM=hwcomposer': No such file or directory' | 01:38 |
| r0kk3rz | wrong command | 01:41 |
| r0kk3rz | variables first, then programs | 01:41 |
| T42 | <edp_17> Thanks, it worked that way: https://pastebin.com/GvSYzkMN | 01:43 |
| r0kk3rz | not quite... | 01:43 |
| T42 | <edp_17> I meant, I was able to grab the strace. | 01:44 |
| r0kk3rz | which isnt really that useful if you dont get the command right :P | 01:44 |
| T42 | <edp_17> I thought it was right. 😄 What's the correct command then? | 01:46 |
| r0kk3rz | i told you, variables first | 01:47 |
| r0kk3rz | which of those three terms do you think is the variable? | 01:47 |
| T42 | <edp_17> The program is the test_hwcomposer, right? | 01:49 |
| r0kk3rz | well, theres two programs in that command | 01:49 |
| T42 | <edp_17> yes the strace is the second. | 01:49 |
| r0kk3rz | and the command reads left to right :P | 01:50 |
| T42 | <edp_17> There you go: https://pastebin.com/wiNmpnfa | 01:51 |
| T42 | <edp_17> :P | 01:51 |
| T42 | <edp_17> 😄 | 01:51 |
| T42 | <edp_17> Is it the right one? | 01:52 |
| r0kk3rz | yes | 01:54 |
| T42 | <edp_17> Damn, I am so good. 😄 | 01:55 |
| T42 | <edp_17> However to me this log an the previous looks identical. | 01:55 |
| r0kk3rz | thats beside the point | 01:58 |
| T42 | <edp_17> Sorry guys, I am half dead tired again. I'll continue later. Thank you for the help so far! Good night. | 02:02 |
| T42 | <RealDanct12> is there anyway to build a base los 15.1 image from hybris-hal? | 08:55 |
| T42 | <RealDanct12> is there anyway to skip the lineage flashing step? | 09:32 |
| T42 | <RealDanct12> or is there anyway to build a lineageos image from sailfishos build? | 09:32 |
| T42 | <Akatsu %lastname%> you have outputed basic OS image | 10:40 |
| T42 | <Akatsu %lastname%> like with halium | 10:40 |
| T42 | <Akatsu %lastname%> it will be cool to have system.img like with halium so we can flash it onto /system | 10:41 |
| r0kk3rz | productised devices like the xperias do that | 11:05 |
| r0kk3rz | so yes, you can | 11:05 |
| T42 | <ot_kurva> Hi! Does the device tree which I use to build SFOS must be the same as the tree, which was used to build LOS zip? I found zip for Xiaomi Wayne (los-16.0). But it is based on UnparallelSky's tree. And this tree is unavailable now. I found actual trees on xiaomi-sdm600 GitHub. But I can't find zip, which is based on this tree. | 11:31 |
| T42 | <ot_kurva> Do I have to build LOS on this tree before building LOS? Or I can use the existing zip. | 11:34 |
| deathmist | @ot_kurva zip should not matter, but you do also need some LOS device sources (+ vendor blobs) for building the kernel & HAL during SFOS development | 11:44 |
| deathmist | btw read the pinned message (especially regarding FAQ) as that is crucial for 16.0 building | 11:46 |
| T42 | <ot_kurva> I have to get vendor blobs from xiaomi-sdm660 and flash vendor.img from out dir. And then - flash SFOS zip. Is it right? | 11:49 |
| deathmist | you should not need to flash anything from out/ dir other than hybris-boot.img to boot partition. blobs in /vendor should be supplied by the existing LOS zip you're going to flash before SFOS. you usually need them in your local manifest because they are required for building HAL to succeed | 11:51 |
| T42 | Raj %lastname% was added by: Raj %lastname% | 12:31 |
| T42 | <Akatsu %lastname%> How can i remove nfcd-binder-plugin from building? my device doesnt have nfc and it throws me error during packaging | 16:18 |
| T42 | <elros34> Say no when it ask you whether you want to build it. You are building middleware only once, later just droid-config and droid-hal. | 16:29 |
| T42 | <Akatsu %lastname%> can i get into this configurator once again? i build it long ago and now it just goes straight to building | 16:38 |
| T42 | <elros34> rpm/dhd/helpers/build_packages.sh --mw will ask you about every package [Y/n/all] | 16:43 |
| T42 | <Akatsu %lastname%> ok ty. | 16:44 |
| T42 | <edp_17> Hi guys. | 19:33 |
| T42 | <edp_17> Has anyone experienced issue with the wireless charging on a xperia xa2? | 19:34 |
| mal | what? xa2 hardware doesn't support wireless charging | 19:44 |
| T42 | <edp_17> Oh my, it's my bad. I mixed up when I read it supports quick charge 3.0 and fast battery charging 18W. | 19:54 |
| T42 | <edp_17> Thank you. | 19:54 |
| T42 | <Akatsu %lastname%> Any idea why i have permissions errors on rpm/dhd/helpers/build_packages.sh and after chmod +x i have more errors on accessing HADK ./out? | 19:57 |
| mal | are you sure the partition you have the sources is mounted with correct permissions and that you ran the commands with correct user? | 19:59 |
| T42 | <Akatsu %lastname%> (Photo, 1280x465) https://irc.thaodan.de/.imgstore/o2D2kuVH0q.png | 20:06 |
| T42 | <Akatsu %lastname%> i have build droid-hal under ubuntu chroot just fine then exited and error, i did chmod and it cannot access files | 20:07 |
| mal | well that claims the issue is in rpm/dhd submodule, which cannot be unless something is really wrong, show output "ls -la $ANDROID_ROOT" | 20:17 |
| T42 | <Akatsu %lastname%> for now i just tried most idiotic way ("sudo chmod +x rpm/dhd/helpers/*") and it seems like somehow got some progress | 20:18 |
| piggz | mal: spiiroin: who knows mce ? | 20:25 |
| mal | what is the problem? | 20:31 |
| piggz | mal: volume keys are not being acted on | 20:31 |
| piggz | evdev_trace shows them working ok | 20:31 |
| piggz | mce seems to detect the keyboard ok according to the logs | 20:32 |
| piggz | Feb 02 20:29:16 PinePhone mce[3612]: event-input.c: evin_iomon_device_add(): /dev/input/event3: name='1c21800.lradc' type=VOLUME KEYS | 20:32 |
| piggz | when key is press, mce says: | 20:32 |
| piggz | Feb 02 20:30:43 PinePhone mce[3612]: tklock.c: tklock_notif_slot_set(): notification 'mce_volume_key' added | 20:32 |
| piggz | Feb 02 20:30:45 PinePhone mce[3612]: tklock.c: tklock_notif_slot_free(): notification 'mce_volume_key' removed | 20:32 |
| mal | hmm | 20:32 |
| mal | piggz: could there be some configuration missing? | 20:33 |
| mal | from other parts | 20:33 |
| piggz | mal: maybe, not sure | 20:34 |
| T42 | <Ege %lastname%> Hello when i try to fetch halium-16.0 i get a gpg error,gpg: Can't check signature: public key not found, how do i fix this? | 20:44 |
| T42 | <Ege %lastname%> the command is repo init -u git://github.com/mer-hybris/android.git -b hybris-16.0 | 20:45 |
| mal | piggz: which pulse configs did you take from dhc submodule for that device? | 21:35 |
| mal | piggz: there are many things related to audio in there, including volume steps | 21:35 |
| piggz | mal: that could be it ... we were missing the dbus module being loaded until recently.... | 21:37 |
| mal | piggz: which droid-config packages do you have installed? | 21:38 |
| piggz | https://github.com/sailfish-on-dontbeevil/droid-config-pinephone/blob/master/sparse/etc/pulse/arm_native_default.pa | 21:38 |
| piggz | mal: i see some modules not being loaded | 21:39 |
| mal | piggz: do you have the keepalive installed? | 21:40 |
| T42 | <Ege %lastname%> how can i fix this? repo sync --fetch-submodules | 21:40 |
| T42 | <Ege %lastname%> error: in `sync --fetch-submodules`: revision refs/tags/android-9.0.0_r46 in platform/external/guice not found | 21:40 |
| piggz | mal: no | 21:40 |
| mal | try without --fetch-submodules | 21:40 |
| mal | piggz: try installing it, I assume it's needed | 21:40 |
| mal | piggz: you should show some logs | 21:41 |
| T42 | <Ege %lastname%> wont i need those submodules later on? | 21:41 |
| piggz | mal: let me have a go first....we are missing some meego modules for volume | 21:41 |
| r0kk3rz | @Ege yea but you do it in two stages because its a big checkout | 22:18 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
