| T42 | <edp_17> @elros34: Thanks, I have seen this too but I am not sure how I can add the KEY_HOMEPAGE into it. I'll try to find out. | 00:05 |
|---|---|---|
| T42 | <edp_17> @ankaos: Are you sure the wifi is not built in? | 00:05 |
| T42 | <ankaos> now i am trying to fix | 00:06 |
| T42 | <edp_17> @ankaos: This is not answer to the question. Are you sure that the wifi driver is compiled as module? | 00:11 |
| T42 | <ankaos> I guess, no | 00:14 |
| T42 | <edp_17> Check you defconfig, that will tell you for sure. | 00:15 |
| r0kk3rz | check your out folder | 00:24 |
| T42 | <elros34> for this device wlan is wil6210 by default build as a module but he set it to builtin and enabled different wlan drivers which probably cause issues. So after reverting all wlan related changes modprobe wil6210 should be enough to get wifi working | 00:25 |
| T42 | <elros34> I cannot communicate with him properly but maybe you will explain it to him | 00:26 |
| T42 | <edp_17> @elros34: evdev.ini helped! Now the home button wakes the screen up! Thanks. | 00:29 |
| T42 | <elros34> hmm or maybe this wilocity driver is only for 60Mhz | 00:41 |
| T42 | <elros34> GHz* | 00:41 |
| T42 | <edp_17> Now the S-Pen also wakes the screen up! 😊 Now I would like to be able to swipe with it. I may need to contact spiiroin as he done that evdev.ini. | 00:42 |
| T42 | <edp_17> *he authored | 00:43 |
| r0kk3rz | the spen is some kind of bluetooth device? | 00:51 |
| T42 | <edp_17> Well, I am not sure. It works without bt on, so I guess it is not a bt device. | 00:55 |
| r0kk3rz | if its just a touchscreen stylus it should work already | 00:56 |
| T42 | <edp_17> No, it is more than that. | 00:56 |
| T42 | <edp_17> And it doesn't work. | 00:57 |
| T42 | <edp_17> The evdev_trace shows its events when I put it close to the screen. It has multiple pressure levels, and the evdev_trace shows when I pull it out and put it back. | 00:58 |
| r0kk3rz | "S Pen remote function subject to Bluetooth connection quality." | 00:59 |
| r0kk3rz | says samsung | 00:59 |
| r0kk3rz | so its some kind of bluetooth device | 00:59 |
| T42 | <edp_17> It might be, but it doesn't at all require to turn the bt on. | 01:00 |
| T42 | <edp_17> It works with bt off on android. | 01:00 |
| r0kk3rz | 'it works' to do what? | 01:01 |
| T42 | <edp_17> Many things. Write onto the screen, select something on the screen then cut/copy/paste it. It is like a stylus but has its own menu options in a menu which appears if the S-pen is pulled out. | 01:03 |
| r0kk3rz | sailfish wont be able to do things like that | 01:04 |
| T42 | <edp_17> I would be happy if I could use it as a normal stylus. I don't need the other functionalities like copy/paste. | 01:04 |
| r0kk3rz | sounds like a fair amount of software integration work | 01:04 |
| r0kk3rz | im surprised it doesnt work as a normal stylus | 01:04 |
| kalube | r0kk3rz: @edp_17: it probably wouldn't be too bad to modify the kernel driver to emulate a mouse\ | 01:05 |
| T42 | <edp_17> Yeah, I don't want to develop for sailfish what Samsung has developed for years for android. 😊 | 01:05 |
| T42 | <edp_17> @kalube: That sounds very good to me. My original intention was to be able to use it as a mouse. | 01:06 |
| T42 | <edp_17> However, I have no idea how and where to start, so if you could help me, that would be fantastic. | 01:06 |
| r0kk3rz | digging into how the thing works is a start | 01:07 |
| r0kk3rz | it must be using some kind of wireless protocol | 01:07 |
| kalube | @edp_17 I have no idea to be honest, I'd start by tracking down the downstream driver (grep your kernel source for the name given in evdev_trace), then I'd find docs on mouse events and how to send them and just add function calls to the existing driver | 01:08 |
| r0kk3rz | if its touches are showing up in evdev then it shouldnt be too bad | 01:09 |
| r0kk3rz | print logs of evdev trace for normal touching and stylus touching | 01:10 |
| T42 | <edp_17> Yes, evdev_trace shows all events when I pull out, put back, take it close or touches the screen. | 01:10 |
| T42 | <elros34> Have you tried removing hardcoded event node from droid-hal-device.conf? Maybe qt input will take care of it and listen on both events | 01:10 |
| kalube | yeah, it probably has some function that takes an x/y value and the pressure or whatever, you could just call mouse function when it's touching etc, and probably implement something nice with the unused functionality | 01:10 |
| T42 | <edp_17> How can I see evdev_trace in human readable format? | 01:10 |
| T42 | <edp_17> Yes it has X/Y axis and pressure level. | 01:11 |
| T42 | <edp_17> Like: ABS_X=0 [0,7056] ABS_Y=0 [0,12544] ABS_PRESSURE=0 [0,2047] ABS_DISTANCE=0 [0,1024] ABS_TILT_X=0 [-63,63] | 01:11 |
| kalube | @edp_17 EASY | 01:11 |
| T42 | <edp_17> @kalube: Is it? Might be for you. :P | 01:12 |
| T42 | <edp_17> However I am happy to learn. | 01:12 |
| T42 | <edp_17> @elros34: No, I have not tried that. | 01:13 |
| kalube | @edp_17 what's the device called in evdev_trace? and link your kernel sources | 01:13 |
| kalube | Definitely try what @elros34 said first though | 01:13 |
| T42 | <elros34> based on qt doc, maybe adding evdevtablet plugin | 01:14 |
| T42 | <edp_17> @kalube: https://paste.ubuntu.com/p/y5vQ9zwXP3/ | 01:15 |
| kalube | Would be cool if it showed up as a drawing tablet or something like that | 01:15 |
| kalube | oh it does | 01:15 |
| kalube | wacom | 01:15 |
| kalube | You can probably get it working without messing with kernel then | 01:16 |
| T42 | <edp_17> Okay, I have removed the /dev/input/event2 from /var/lib/environment/compositor/droid-hal-device.conf. | 01:17 |
| T42 | <edp_17> Should I reboot? | 01:17 |
| T42 | <edp_17> Or just restart lipstick? | 01:17 |
| kalube | sure | 01:17 |
| T42 | <elros34> add evdevtablet also | 01:17 |
| kalube | restarting lipstick isn't consistent is it? | 01:17 |
| T42 | <edp_17> evdevtablet? How? | 01:18 |
| T42 | <elros34> should be pretty obvious if you look in that file | 01:18 |
| T42 | <edp_17> Okay, so as a plugin like : -plugin: evdevtablet? | 01:20 |
| T42 | <edp_17> or just simply add to the end of: -plugin evdevkeyboard:keymap=/usr/share/qt5/keymaps/droid.qmap | 01:20 |
| T42 | <edp_17> like: -plugin evdevkeyboard:keymap=/usr/share/qt5/keymaps/droid.qmap:evdevtablet | 01:20 |
| T42 | <edp_17> Okay, to the end like -plugin evdevtablet | 01:21 |
| T42 | <edp_17> Is there any parameters/arguments/whatever? | 01:21 |
| T42 | <elros34> https://doc.qt.io/archives/qt-5.6/embedded-linux.html | 01:24 |
| T42 | <edp_17> I have added and rebooted but doesn't work. | 01:24 |
| T42 | <edp_17> I have tried these: https://paste.ubuntu.com/p/fNVhqFt7s7/ | 01:37 |
| T42 | <ankaos> @elros34 [for this device wlan is wil6210 by default bui …], Not work wil6210 | 01:38 |
| T42 | <elros34> so modprobe wlan | 01:41 |
| T42 | <elros34> @edp_17 without colon | 01:42 |
| T42 | <edp_17> Only -plugin evdevtablet/dev/input/event3 ? | 01:43 |
| T42 | <elros34> -plugin evdevtouch | 01:43 |
| T42 | <edp_17> Okay, I am doing it. | 01:44 |
| T42 | <elros34> QT_QPA_EVDEV_DEBUG could also be useful or/and QT_LOGGING_RULES="qt.qpa.input=true" | 01:45 |
| T42 | <edp_17> I have realized that the s-pen also react for cat /dev/input/mice and /dev/input/mouse0 | 01:46 |
| T42 | <edp_17> @elros34: where should I set these? | 01:47 |
| T42 | <edp_17> Got that. In the same file. | 01:48 |
| T42 | <edp_17> Okay, the droid-hal-device.conf on the device looks like: https://paste.ubuntu.com/p/cFZBmqdjkj/ | 01:49 |
| T42 | <edp_17> Okay, I missed that I need to remove the : from the -plugin evdevtouch. Sorry. | 01:50 |
| T42 | <edp_17> Okay, now the droid-hal-device.conf looks like: https://paste.ubuntu.com/p/YNdWBp26ZQ/ | 01:55 |
| T42 | <elros34> QT_QPA_EVDEV_DEBUG is missing value | 01:56 |
| T42 | <edp_17> And what should that be? | 01:56 |
| T42 | <elros34> 1? | 01:57 |
| T42 | <edp_17> So like: QT_QPA_EVDEV_DEBUG="1" ? | 01:58 |
| T42 | <edp_17> Or just: QT_QPA_EVDEV_DEBUG=1? | 01:58 |
| T42 | <elros34> id doesnt matter, its env varable | 01:59 |
| T42 | <edp_17> Okay, I set QT_QPA_EVDEV_DEBUG=1. | 02:00 |
| T42 | <edp_17> And rebooted. | 02:00 |
| T42 | <edp_17> How can I get those logs? | 02:00 |
| T42 | <elros34> same as always: journalctl | 02:01 |
| T42 | <edp_17> Here it is: https://paste.ubuntu.com/p/3jG6N7BdnN/ | 02:03 |
| T42 | <edp_17> At one point I pulled out the s-pen and touched the screen a few times. | 02:03 |
| T42 | <edp_17> Sorry guys, I need some sleep now. Can we continue tomorrow? | 02:05 |
| T42 | <edp_17> Good night! | 02:06 |
| T42 | <elros34> at least evdevtablet plugin is loaded. BTW you should disable TvoutService_C as it fails anyway | 02:14 |
| *** ahjolinna_ is now known as ahjolinna | 02:45 | |
| T42 | <Verevka86> @adampigg Hello, its missing- | 04:21 |
| T42 | <Verevka86> https://github.com/piggz/sailfish-fpd-community/blame/master/README.md#L24 | 04:21 |
| T42 | <Verevka86> And missing include- | 05:12 |
| T42 | <Verevka86> https://github.com/SailfishOS-sagit/android_device_xiaomi_msm8998-common/commit/739b4affcca5c52590bcbe6a915282864d9aa239 | 05:12 |
| spiiroin | rinigus: fpd Verify = start dummy enroll, 1st progress report from below -> cancel enroll + report success upwards | 05:20 |
| spiiroin | rinigus: it is used only from csd app | 05:20 |
| rinigus | Verevka86: PR with Android packaging scripts is not need yet. Wait a bit and we should get it ready | 05:26 |
| rinigus | spiiroin: thank you very much! Will have to implement it | 05:27 |
| rinigus | spiiroin: from current API, it's not clear how to prevent manipulation with fingerprints without some Auth layer. Unless it's enforced by app somehow | 05:29 |
| spiiroin | rinigus: can you elaborate what you mean? (atm downstairs = user can add/remove fingerprints, if that is what you meant) | 05:33 |
| spiiroin | @edp_17: from mce pov that s-pen probably looks like single-touch device.. to verify do: systemctl mce stop; mce -Tq -levent-input.c:evin_iomon_device_add --auto-exit=0 | 05:40 |
| spiiroin | ... lists evdev probing results, stte: /dev/input/event1: name='synaptics_dsx' type=TOUCHSCREEN | 05:40 |
| spiiroin | that type is used for mice etc also -> double click == double tap etc | 05:41 |
| spiiroin | for actual interaction with ui also lipstick needs to grok what is going on | 05:41 |
| rinigus | spiiroin: sure. we can add/remove fingerprints without asking for PIN code or any other way of identification if the phone was unlocked. in settings app, you do ask for PIN/auth before operations on fingerprints, but that is not somehow enforced on dbus. | 05:43 |
| rinigus | unless we miss some challenge/response in FP DBUS api | 05:44 |
| spiiroin | rinigus: devicelock manages options that can be used for auth. fpd identify is used only when devicelock thinks it is possible and ok. | 05:47 |
| spiiroin | earlier fp use depended only on past attempt count. | 05:47 |
| spiiroin | currently devel head state = lock code required after reboot / user switch | 05:48 |
| spiiroin | what fpd needs to do: when stateful op is started, lock on to client -> cancel etc is accepted only from client who initiated the op | 05:49 |
| spiiroin | the latter also allows "shared fp use" between mce (wakeup) and devicelock (auth) | 05:52 |
| spiiroin | mce parts are here: https://git.sailfishos.org/mer-core/mce/blob/master/modules/fingerprint.c | 05:52 |
| spiiroin | behind the scene fpd also needs to deal with user session tracking, fp set switching, etc in a manner that is "invisible" to apps using/tracking dbus stuff | 05:54 |
| rinigus | spiiroin: interesting Todo, will be critical when we get to multiuser support. | 06:01 |
| T42 | <adampigg> @Verevka86 [@adampigg Hello, its missing- | 06:21 |
| T42 | <adampigg> https://github.c …], Its in a PR atm | 06:21 |
| rinigus | spiiroin: I filed the issues based on your comments, thank you! (cc @adampigg) | 06:29 |
| spiiroin | rinigus: I guess the important bits atm: that "lock-onto-client" + every op that starts should end with: aborted, failed, or appropriate success signal 1st, then fpd going idle | 06:47 |
| spiiroin | so that signaling that is related to operation happens while fpd is in state that indicates that operation | 06:47 |
| T42 | <adampigg> Yeah, i found that emitting idle too soon broke things | 06:49 |
| spiiroin | ... clients still need to deal also with sun-spots like fpd restarts / out of state signaling | 06:49 |
| T42 | <adampigg> So chanfed that a.few days ago | 06:49 |
| rinigus | spiiroin: that is not as it is now, as far as I know. I pushed the signaling for earlier as some ops also lead to recheck of the list of fingerprints to ensure android<->sfos maps consistency. will have to revise that | 06:50 |
| spiiroin | rinigus: getting list of fingerprints is exception, it is not stateful | 06:51 |
| rinigus | spiiroin: on dbus side it is not, so we have to keep a map ready | 06:51 |
| spiiroin | rinigus: yes. in some/many devices you can't even enumerate templates android side has -> have to have cache and hope it stays in sync | 06:52 |
| rinigus | ... I probably overdid it as I did not trust recorded state. maybe should just init it once on start and then trust the changes as reported | 06:52 |
| rinigus | (cross-posting) | 06:53 |
| T42 | <adampigg> Rinigus, will merge all prs this morning, got some meetings first tho | 06:54 |
| rinigus | @adampigg: I can work on that in the evening only, so that sounds fine | 06:54 |
| rinigus | spiiroin: we then will revise accordingly and make signaling work as expected | 06:55 |
| T42 | <adampigg> We got any more porters trying yet? | 06:55 |
| * spiiroin tries to recall if there has been any device where fp hal would have worked as-expected ... | 06:55 | |
| T42 | <birdzhang> @adampigg i tried on davinci, in-screen fp, when enrolling, Acquisition: FPACQUIRED_UNRECOGNIZED, and no progress, no light up on screen | 07:38 |
| T42 | <birdzhang> dmesg and logcat logs when enroll https://pastebin.com/kpK43YRH | 07:42 |
| T42 | <edp_17> spiiroin and @elros34, thanks for the advice. I'll try that you suggested however I am not sure about the grok you mentioned for lipstick. | 07:44 |
| rinigus | @birdzhang: just in case, try to use -test app https://github.com/piggz/sailfish-fpd-community-test | 07:52 |
| rinigus | ... assuming that you used Settings | 07:52 |
| T42 | <birdzhang> no, i use this test app | 07:52 |
| rinigus | then please add journal log as well | 07:53 |
| T42 | <birdzhang> no journal log at all | 07:53 |
| T42 | <birdzhang> btw, it's not working on ubports too, @erfanoabdi | 07:54 |
| rinigus | strange, the daemon should be rather chatty in the journal | 07:54 |
| T42 | <birdzhang> https://imgur.com/gallery/YjPKQuN | 07:54 |
| rinigus | birdzhang: no idea what's piggz position on it, but I would prefer us to get to the first release and then start testing on other devices. last night people were trying to follow README that was referring to things which are not merged. we will have to reorganize storage => mess with databases and there is no wish to start data migration support | 08:06 |
| rinigus | there are probably more issues to be solved. | 08:06 |
| rinigus | we do have a decent speed in development, so it should be ready soon. just wait a bit. | 08:07 |
| T42 | <birdzhang> Yeah, no hurry, and good job ;) | 08:08 |
| rinigus | birdzhang: as you seem to have issues with UT as well, try to get android logcats corresponding to FP operations meanwhile. then you can later debug and see what is failing when you have good log to compare to | 08:08 |
| T42 | <birdzhang> i will do it | 08:09 |
| T42 | <edp_17> spiiroin: The result of 'mce -Tq -levent-input.c:evin_iomon_device_add --auto-exit=0': https://paste.ubuntu.com/p/9MQ327B6Qm/ | 08:16 |
| T42 | <edp_17> event-input.c: evin_iomon_device_add(): /dev/input/event3: name='sec_e-pen' type=TOUCHSCREEN | 08:16 |
| T42 | <edp_17> How do you mean: "lists evdev probing results, stte: /dev/input/event1: name='synaptics_dsx' type=TOUCHSCREEN" | 08:18 |
| T42 | <edp_17> If you meant evdev_trace output, that is here: https://paste.ubuntu.com/p/Zdrmbms8z6/ | 08:19 |
| T42 | archi_dagachs was added by: archi_dagachs | 08:21 |
| spiiroin | @edp_17: I meant that line with "name='sec_e-pen' type=TOUCHSCREEN" -> from mce pov the pen is mouse/touch kind device = good (in theory it could have been treated as sensor / whatnot) | 08:34 |
| T42 | <edp_17> spiiroin: I see, thanks for confirming. Thought you advised me to run a command or something and it was not clear. 😊 | 08:49 |
| T42 | <edp_17> Last night I played with cat /dev/input/mice and cat /dev/input/mouse0 and cat /dev/input/event3. All three did the same when I touched the screen with the s-pen, so any of them can be used to interact with the s-pen. However I don't know how. 😁 | 08:51 |
| T42 | <KernelPanix> build/make/core/droiddoc.mk:188: error: FindEmulator: find: `tools/metalava/manual': No such file or directory | 08:55 |
| T42 | <KernelPanix> @KernelPanix [build/make/core/droiddoc.mk:188: error: FindEm …], please help me this errors | 08:56 |
| spiiroin | @edp_17 "evdev_trace -t 3" -> prints events from /dev/input/event3 (in more human readable from than plain cat does) | 09:01 |
| rinigus | mal: ping regarding compass-related MRs to qtsensors and sensorfwd review. | 09:03 |
| T42 | <edp_17> spiiroin: Thank you! It looks much better. I used the evdev_trace -t (without the event number) before but I thought there is something else as well. Anyway, for referece the output is here: https://paste.ubuntu.com/p/txZZc7sK8n/ | 09:08 |
| spiiroin | @edp_17 looks like mix betweem single-touch device (ABS_X/Y + BTN_TOUCH) and mouse (reports also when not pressed down) with some extras | 09:20 |
| spiiroin | does it report also when display is off? | 09:20 |
| T42 | <edp_17> Yes it does report when the screen is off. All what you see in the last log recorded with screen off. | 09:23 |
| zinstack625 | @KernelPanix hybris-16.0? If yes, then hybris-patches/apply-patches.sh --mb | 09:24 |
| T42 | <KernelPanix> @zinstack625 [@KernelPanix hybris-16.0? If yes, then hybris- …], okay thanks u | 09:24 |
| T42 | <KernelPanix> (Sticker, 512x507) https://irc.thaodan.de/.imgstore/R2qcePaLob.webp | 09:24 |
| spiiroin | @edp_17 that sw_13 is interesting... 0x13 > 0x0f = SW_MAX when it possibly could/should be SW_PEN_INSERTED=0x0f? | 09:29 |
| zinstack625 | Building rootfs or whatnot. rpm/dhd/helpers/build_packages.sh dies like this: https://pastebin.com/UMDUwGmd. Log has this: https://pastebin.com/eXwUuT4A. Tried reinstalling target to no avail. I did interrupt the building as it was stuck on something I can't recall. I could try deleting that file, but I'm not sure where it is. Any ideas? | 09:30 |
| T42 | <edp_17> zinstack625: I got this before. You need to install droid-config in Platform. use this command: sb2 -t $VENDOR-$DEVICE-$PORT_ARCH -m sdk-install -R zypper install droid-config | 09:35 |
| T42 | <edp_17> Then re-run the rpm/dhd/helpers/build_packages.sh | 09:35 |
| T42 | <edp_17> spiiroin: Tbh, I don't know. The 0x013/SW_0013 does change its value to 1 or 0 depending on the s-pen is out or not. | 09:37 |
| T42 | <edp_17> The other thing I do not understand is the home button. On the S2 port, the home button just works out of the box, while on this N4 port it doesn't. | 09:39 |
| zinstack625 | @edp_17 Still fails. From what I see is it's trying to unpack initramfs to somewhere, but there this file already exists. In Platform, /boot is empty, so, perhaps there is rootfs getting ready somewhere and there it should be? | 09:52 |
| T42 | <edp_17> zinstack625: what's the new error message? | 09:55 |
| T42 | <edp_17> zinstack625: have you seen this? http://collabedit.com/kr9xx | 09:57 |
| zinstack625 | @edp_17 I have, but I forgot the url. Thanks | 09:59 |
| T42 | <edp_17> zinstack625: No problem, your solution is on that link or here https://pastebin.com/nC9Dxh1q | 10:02 |
| T42 | <edp_17> For me on one of my ports it was enough to install droid-config, but on some others I needed to apply this fix. | 10:02 |
| T42 | <adampigg> rinigus: @birdzhang Ive no issue with others trying it out, so long as they understand the rate of flux | 10:04 |
| piggz | spiiroin: rinigus: so, for Verify, state changes to enrolling, emits a progress update, csd calls Abort and state goes back to idle? sounds like that can just use a timer to emit the progress? | 10:07 |
| T42 | <adampigg> or, do you actually use the finger. | 10:10 |
| T42 | <adampigg> just ran csd and thats what it looks like | 10:10 |
| T42 | <adampigg> thats better actually | 10:10 |
| T42 | <erfanoabdi> @birdzhang [btw, it's not working on ubports too, @erfanoa …], Sorry can't remember your device model | 10:32 |
| T42 | <erfanoabdi> Is it fingerprint on screen? | 10:32 |
| T42 | <birdzhang> Yeah, Redmi K20 | 10:37 |
| T42 | <erfanoabdi> @birdzhang [Yeah, Redmi K20], Fod still not supported | 10:39 |
| T42 | <birdzhang> Okay | 10:40 |
| rinigus | @adampigg: I guess we need to file the issue to support Verify as well | 10:47 |
| rinigus | thnx for merging it all | 10:47 |
| piggz | rinigus: struggling with Verify ... its getting called, im emitting the progress, but its not Aborting like spiiroin suggested | 11:09 |
| piggz | maybe it expects the daemon to abort after the progress? | 11:10 |
| spiiroin | piggz: nope... a sec ;-) | 11:10 |
| spiiroin | piggz: fpd does hal->enroll() ... hal->cancel; csd does fpd->verify() and waits for success/failure | 11:11 |
| spiiroin | i.e. verify is implemented as enroll operation that is terminated when the info that is needed has been received | 11:11 |
| spiiroin | which is finger-on-sensor is detected = the 1st progress report fpd gets from hal | 11:12 |
| piggz | spiiroin: so, like i said, the fpd needs to terminate :#0 | 11:12 |
| piggz | :) | 11:12 |
| piggz | one sec, lets try that | 11:12 |
| piggz | spiiroin: in normal enroll, success is indicated by the Added signal which contains the name supplied by the Enroll call..... what signal is expected for Verify, which doesnt take a string name parameter? | 11:18 |
| T42 | <edp_17> @elros34: There are a few problems with the bt. 1,. After 3 minutes, it turns off itself (even if it is connected to another device). 2,. Not always can be paired (sometimes I need to turn the bt on and reboot) 3,. However it shows the device connected to a bluetooth loudspeaker, the speaker still blinking (waiting for the connection) | 11:43 |
| T42 | and the sound is not routed. | 11:43 |
| T42 | <edp_17> Is number 3 can be connected the incorrect scopcm value? | 11:44 |
| T42 | <elros34> I have no idea but adding some random values for for hciattach can cause random issues. | 11:46 |
| T42 | <elros34> BTW I am not bluetooth expert:) | 11:47 |
| T42 | <edp_17> @elros34: Okay, I see. Now it does make sense what you told me last time. 😊 | 12:02 |
| T42 | <edp_17> This: http://www.merproject.org/logs/%23sailfishos-porters/%23sailfishos-porters.2020-05-29.log.html#t2020-05-29T14:30:23 | 12:05 |
| T42 | <edp_17> 😊 | 12:05 |
| piggz | spiiroin: rinigus: Verify() implemented and passes csd test. | 12:48 |
| piggz | pushed | 12:49 |
| rinigus | piggz: great! I am still watching / will be able to join later tonight | 13:07 |
| zinstack | Rootfs 3.3.0.16 is built, it was not happy at hwservicemanager being unable to start (probably due to selinux being dead). I copied a sparce from sony-nile to deal with it and now it's not giving out ips over usb and bootloops after about a minute. From recovery I did ln -s /dev/null /data/.stowaways/sailfishos/etc/systemd/system/ofono.service, but it still bootloops. Here I am again, somebody knows what to do? | 13:43 |
| zinstack | Also, I'd like to package rootfs as gzip, as twrp doesn't support bzip2 | 13:43 |
| T42 | <edp_17> zinstack: Next to the bzip2, you should have got a normal zip image too. I am always installing that with twrp. | 14:05 |
| T42 | <edp_17> If you don't have the zip image, the build is not complete. I think. | 14:05 |
| zinstack | @edp_17 There is a zip archive, but installing it is impossible, because inside it there is a rootfs tarball packed in bzip2. Installing it will copy that to /data, create a stowaway and extract it to the stowaway. That last part is bothersome, because there is no support for bzip2 in my twrp image (it's official btw) | 14:11 |
| T42 | <edp_17> I am always installing the zip. It need to be flashed the android image first, then the sailfish image on top of it. Then, reboot. I have never heard of any other ways. It works the same way even on the officially supported Sony XA2. | 14:14 |
| T42 | <edp_17> If your twrp doesn't support this, you should find an older one that does support. | 14:15 |
| zinstack | @edp_17 The eldest one is 3.3.1-0 and it uses toybox as well. Don't know much, but it still can't bunzip anything | 14:26 |
| T42 | <edp_17> Oh, is that something very new device? | 14:27 |
| zinstack | Perhaps it's related to it being new, but I can remember sony amami had that same issue. Can't remember the twrp version though | 14:28 |
| zinstack | @edp_17 Found it, it had twrp 3.2.3-1 and no bzip2 | 14:36 |
| T42 | <edp_17> What is this device? | 14:43 |
| zinstack | The one that had twrp 3.2.3-1 was sony xperia z1 compact (amami), mine now is motorola moto g7 (river) | 14:44 |
| T42 | <edp_17> I see, that really is a new one. :😄 So basically, you manually need to install the bzip2 from the image? | 14:49 |
| rinigus | piggz: pulling in your changes and starting to test | 14:49 |
| T42 | <adampigg> rinigus: gr8 | 14:49 |
| T42 | <edp_17> zinstack: And what about if you extract the zip on your pc and manually gzip it? | 14:50 |
| zinstack | @edp_17 I unpack bzip2 on pc and send over the unpacked tarball. It's manual but it works. At least to make it say init-debug in real rootfs for a minute | 14:51 |
| rinigus | @adampigg: meanwhile you could think what should I start working on. #17 or #18? | 14:52 |
| piggz | rinigus: 18 is proabably a quick win? mainyl code removal? | 14:53 |
| rinigus | piggz: none of them is quick. in 18 we have to keep track on connected client | 14:54 |
| zinstack | Installing is the least of my problems now. It started bootlooping and now I'm unable to telnet after I merged in hwservicemanager.rc from sony nile. That's a dead end for me, unless somebody knows what to do | 14:54 |
| rinigus | but I can start there, no problem | 14:54 |
| piggz | rinigus: how do we do that, we dont get any info for that from qtdbus? | 14:54 |
| T42 | <edp_17> zinstack: How quickly does it reboot? | 14:54 |
| rinigus | piggz: we should. just have to read the docs. | 14:55 |
| zinstack | @edp_17 it does take a minute, during which it shows up as init-debug in real rootfs. I can't telnet to it though, because it doesn't give out ips for some reason | 14:56 |
| piggz | zinstack: try a fixed ip on your host | 14:56 |
| T42 | <elros34> @zinstack: first remove any changes you copied from different device then paste full journalctl. About gzip, maybe it's enough to amend pack_package-droid-updater in droid config, updater-unpack.sh in hybris-boot, and use gzip when creating image with mic | 14:57 |
| T42 | <edp_17> zinstack: try to assign an ip and bring up the usb manually. (Write a script that does it for you and at the end you can go in with telnet.) Some info is here: http://docs.halium.org/en/latest/porting/debug-build/early-init.html | 14:59 |
| T42 | <edp_17> Then once you are in: to stop the continuesly reboot: touch /init_enter_debug2 | 14:59 |
| zinstack | Well, bootloop stopped, but for what I can see, it's a regression. Systemd no longer starts after that merge, out it goes. After building, i'll report what it says. Thanks all | 15:07 |
| zinstack | So, here's my journalctl: https://pastebin.com/v7ns9ZHV. And dmesg: https://pastebin.com/wybWJ2sk. All I can see is that I need to fix hwservicemanage somehow. Message points to some selinux code, which may be a hint. In the code it obtains an selinux handle, successful, then checks the context and flops. I don't think I should touch that code, as that will be a pain to build and maintain, and, most likely, will | 15:56 |
| zinstack | be extremely unstable | 15:56 |
| rinigus | piggz: how do you specify spec file when you build in sdk? or you use sb2 from terminal? | 15:59 |
| rinigus | mb2 | 16:00 |
| rinigus | piggz: I am on #18 | 16:08 |
| T42 | <adampigg> rinigus: i use terminal | 16:23 |
| T42 | <adampigg> rinigus: gr8, i am going for icecream and chips while u work :D | 16:23 |
| rinigus | :) | 16:23 |
| T42 | <kkk_sw> zinstack: https://github.com/WinKarbik/droid-config-ido/commit/0b0b9b8cfc07f760fcfee6fbc4c3622eabbd0095 | 16:44 |
| T42 | <elros34> you are missing hybris 16 specific changes from github.com/mer-hybris/hadk-faq. Also vendor-dsp.mount is incorrect. Do you have /usr/sbin/ohmd? | 16:48 |
| T42 | <elros34> zinstack: did you even run apply-patches.sh? | 16:53 |
| zinstack | @elros yes, I did, it wouldn't compile without them, would it? | 16:53 |
| zinstack | @kkk_sw I knew it had something to do with that, but only thing I could find referencing that was an illiterate (both in language and in computers) gentelman being sworn at by another gentelman on this irc with broken links from ages ago. You are my reassurance and saviour | 16:55 |
| zinstack | gentleman* Wow, that's embarrasing | 16:56 |
| T42 | <ot_kurva> "Could not run transaction" in buildpackages.sh | 16:59 |
| T42 | <ot_kurva> https://pastebin.com/f5EYJFz2 | 16:59 |
| T42 | <elros34> looks like too old submodules | 17:00 |
| T42 | <ot_kurva> it hepls. thanks | 17:15 |
| T42 | OwL_Is_NooB was added by: OwL_Is_NooB | 19:04 |
| rinigus | piggz: fix for #18 submitted. also PR for gui after I was looking while failed signal and error were not propagated... | 19:39 |
| rinigus | piggz: I think we have to fix #17 before release as well. at least decide how to store FPs and implement it. | 19:40 |
| rinigus | as we don't have multiuser yet, the rest of #17 will be hard to debug | 19:41 |
| T42 | <adampigg> rinigus: yeah, im thinking how to even track user sessions .... does systemd emit such things? | 19:49 |
| rinigus | @adampigg: let's not bother with tracking until we get the functionality. its probably somewhere in SFOS beta/alpha codes and can maybe change as well | 19:50 |
| rinigus | but we have to store FPs correctly already so we will not loose them when the update comes | 19:50 |
| T42 | <adampigg> Ok, that should be easy enough | 19:51 |
| rinigus | @adampigg: then #17 is on you? :) | 19:56 |
| rinigus | piggz: I am off for tonight. gn | 20:25 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
