| T42 | <edp_17> hi all, is there a trick to make sensors work in sfos4.1? | 07:17 |
|---|---|---|
| T42 | <edp_17> For my last port I needed to downgrade sensorfw to 0.11.9 to make it work. But when I built 4.1, neither 0.11.9 nor 0.12.0 works. | 07:18 |
| T42 | <edp_17> Here is the log: https://paste.ubuntu.com/p/34ZHtYM2rX/ | 07:44 |
| T42 | <edp_17> Is this message can be the reason why none of the sensors work?: | 07:44 |
| T42 | <edp_17> SensorManagerError: "requested sensor id 'alssensor' not registered" | 07:44 |
| T42 | <edp_17> For sensorfwd.service, I've changed --log-level to test. This gives more details in the log: https://paste.ubuntu.com/p/VW9fQrWxjT/ | 09:23 |
| piggz | better do a kernel check, 4.2 coming with a new class of security errors | 10:17 |
| piggz | Jul 22 11:14:44 PinePhone invoker[4053]: Invoking execution: '/usr/bin/fingerterm' | 10:17 |
| piggz | Jul 22 11:14:44 PinePhone mapplauncherd[4049]: Connection: getMountNamespace: stat failed for pid 4053: Permission denied | 10:17 |
| piggz | Jul 22 11:14:44 PinePhone mapplauncherd[4049]: Connection: getExecutablePath: readlink failed for pid 4053: Permission denied | 10:17 |
| mal | @edp_17 does test_sensors work? | 10:39 |
| piggz | mal: re above ... my kernel config only has one error for a netfilter rule, which i doubt is related to the above.... | 10:42 |
| piggz | any idea what changed between 4.1 and 4.2? probably sailjail? | 10:43 |
| T42 | <edp_17> mal: yes, it lists the sensors | 10:43 |
| T42 | <edp_17> mal: Now I am back on 0.11.9 version | 10:44 |
| mal | @edp_17 and restarting sensorfw doesn't help? | 10:45 |
| T42 | <edp_17> I've restarted the service. | 10:46 |
| mal | do you have your repos somewhere? | 10:46 |
| T42 | <edp_17> mal: https://github.com/edp17/droid-config-treltexx | 10:47 |
| T42 | <edp_17> I only have one thing that is not pushed. | 10:49 |
| T42 | <edp_17> A symlink in sparse/usr/lib/systemd/system/actdead-pre.target.wants/droid-hal-init.service | 10:49 |
| T42 | <edp_17> I think this is fixig somethig related to when the device is off and hooked up on charger. | 10:51 |
| T42 | <edp_17> mal: I rebooted and sensorfw outputs the same error into log | 10:52 |
| T42 | <edp_17> mal: I have checked on device. I still have sensorfw-qt5-0.12.2 installed. | 10:57 |
| T42 | <edp_17> When I went back to 0.11.9., I used this command to build: rpm/dhd/helpers/build_packages.sh -D -o -b hybris/mw/sensorfw -s rpm/sensorfw-qt5-hybris.spec | 11:02 |
| T42 | <edp_17> And this built two files | 11:02 |
| T42 | <edp_17> hybris-libsensorfw-qt5-hal-0.11.9-0.armv7hl.rpm | 11:02 |
| T42 | <edp_17> hybris-libsensorfw-qt5-hal-devel-0.11.9-0.armv7hl.rpm | 11:02 |
| T42 | <edp_17> I guess I would need a sensorfw-qt5-0.11.9-xxx.rpm too. Right? | 11:03 |
| T42 | <edp_17> Or should I go back to version 0.12.2? (I found a comment from you that 0.11.9 only needed for sfos 3.4) | 11:05 |
| mal | @edp_17 you should have both at 0.12.x | 11:06 |
| mal | sensorfw-qt5 and hybris-libsensorfw-qt5-hal, which android base is that? | 11:06 |
| T42 | <edp_17> mal: It is working now. | 11:23 |
| T42 | <edp_17> I've deleted the sensorfw folder, cloned it from fresh, rebuilt, installed the hybris-libsensorfw-qt5-hal-0.12.2-0.armv7hl.rpm and then restarted the service. | 11:25 |
| mal | good | 11:25 |
| T42 | <edp_17> This is the second package where the git fetch xxx and then git pull xxx master didn't update properly. I needed to do the same with dummy_netd package too. | 11:26 |
| T42 | <edp_17> mal: Another question regarding bluetooth. It seems the built broadcom-bluetooth and bluetooth-rfkill-event packages are not installed into the build. After installed the build on device, I needed to install those .rpms manually to make the BT work. | 11:30 |
| T42 | <edp_17> What should I check/change to get them included into the build? | 11:30 |
| mal | patterns | 12:14 |
| mal | or what used to be patterns, the patterns folder in config repo which contains the meta-packages | 12:15 |
| mal | @edp_17 ^ | 12:15 |
| mal | it seems highlight is not working well anymore for me, need debug at some point, anyone pinging me from telegram bridge is not showing as highlight | 12:16 |
| T42 | <edp_17> mal: It is already in there (line 74, 75): https://paste.ubuntu.com/p/zsPghs46zT/ | 12:58 |
| T42 | <elros34> compare patterns-sailfish-device-adaptation-treltexx package version installed in image with that onein your local repo. If they are same then use rpm -qp --requires path/to/rpm to confirm dpendencies | 13:09 |
| T42 | <edp_17> @elros34 : It looks like the package in local repo and what is installed on device is the same: https://paste.ubuntu.com/p/Pb8gtQ4rT6/ | 13:16 |
| T42 | <edp_17> "rpm -qp --requires path/to/rpm" Should I issue this command on device or host? What is the rpm you are referring to at the "path/to/rpm"? patterns-sailfish-device-adaptation-treltexx-1-202107211221.armv7hl.rpm or the roadcom-bluetooth and bluetooth-rfkill-event packages? | 13:18 |
| T42 | <elros34> patterns package in your local repo, this one supposed to pull your bluetooth rpm packages so it must have dependencies you set in patterns | 13:20 |
| T42 | <elros34> on device you can use rpm -q --requires patterns-sailfish-device-adaptation-treltexx | 13:20 |
| T42 | <edp_17> Both packages are listed: https://paste.ubuntu.com/p/89k38GHYsz/ | 13:32 |
| T42 | <elros34> so they should be installed, are you sure they are not? zypper se -s bluetooth | 13:35 |
| T42 | <elros34> if dependencies for patterns package in local repo are also fine then bluetooth packages must also be in your Jolla-4.1.0.24-treltexx-armv7hl.packages | 13:36 |
| T42 | <edp_17> I am not suer they were not installed but BT didn't work until I manually instaled them. Now, of course, both are installed. | 13:38 |
| T42 | <edp_17> Jolla-4.1.0.24-treltexx-armv7hl.packages, this is where I checked whether these were included. | 13:39 |
| T42 | <elros34> so according to *.packages, there were installed or not? | 13:42 |
| T42 | <edp_17> Based on the .packages, a different version were installed: https://paste.ubuntu.com/p/6BQds42tZq/ | 13:44 |
| T42 | <edp_17> This is what I have on the local repo: https://paste.ubuntu.com/p/K8Q6s4G9d3/ | 13:46 |
| T42 | <elros34> so most obvious reason, you just rebuild bluetooth-rfkill-event-hciattach after building image:) | 13:49 |
| T42 | <elros34> ah you have old version 1.0.8 vs 1.0.9 | 13:51 |
| T42 | <edp_17> @elros34 : Yeah, that could be a reason but I copied those two .rpm packages to the device right after the image is created. (I had this issue with my previous builds, so wanted to be prepared.) So, on the device I still have the same rpms which were generated by the command 'rpm/dhd/helpers/build_packages.sh -o' | 14:14 |
| T42 | <edp_17> And those packages are the same that I currently have in local repo. | 14:14 |
| T42 | <edp_17> I see. So, the newer version (1.0.9) was installed rather than the one that I built (1.0.8). | 14:15 |
| T42 | <edp_17> I guess now I need to update those modules in mw then apply the necessary changes, then rebuild. That way I will have a newer that 1.0.9 version, and that will (at least expected) be installed. Right? | 14:17 |
| T42 | <elros34> I think git pull in bluetooth-rfkill-event source and rebuild should be enough, latest comit is about spec file. I guess this would be done automatically and you would not even notice it if you would not use offline option. | 14:35 |
| T42 | <elros34> zypper or whatever is used internally by mic "always' install latest available version. In your case latest 1.0.9 was in common devel repo | 14:37 |
| T42 | <edp_17> Thanks. I'll try this. Glad now I know why BT didn't work and what the solution would be. 👍 | 15:11 |
| piggz | mal: in good news, the default browser install in 4.2 does render on the pinephone now | 17:03 |
| mal | piggz: nice | 17:04 |
| piggz | mal: just need to dig into this mapplauncherd problem | 17:04 |
| piggz | mal: its fun ... launching calculator from lipstick doesnt work, but running the full sailjail command on terminal does work! | 17:06 |
| T42 | <b100dian> Well thats probably the mapplauncherd | 17:08 |
| T42 | <b100dian> Isnt that the booster from the desktop file? | 17:08 |
| piggz | @b100dian definitely mapplauncherd is logging some error | 17:24 |
| piggz | when icon is clicked | 17:24 |
| T42 | <b100dian> Does it have | 17:28 |
| T42 | <b100dian> X-Nemo-Application-Type=silica-qt5 | 17:28 |
| piggz | @b100dian no, neither jolla-calculator or notes have that | 17:38 |
| T42 | <webdotcom> Where you able to port sailfish for rosy? (re @SailfishFreenodeIRCBridgeBot: <Tofe> mal: I also h...) | 18:12 |
| piggz | mal: @b11dian: atleast found the offending code https://git.sailfishos.org/mer-core/mapplauncherd/commit/3cb9c45c4159e5079dd94f000c6cd6464b8ce490 | 18:22 |
| T42 | <webdotcom> Lmfao i replied to a bot | 18:25 |
| mal | piggz: btw, use github.com/sailfishos for the repos, those have the latest versions | 18:38 |
| T42 | <b100dian> piggz: use the exploit in the comment:) | 18:39 |
| piggz | mal: true ... thats just what google found for me when googling errors | 18:39 |
| mal | piggz: which part of that fails | 18:40 |
| T42 | <b100dian> so this error is not volla+firejails as before, is pinephone and 4.2? I guess we all wait to see that error for ourselves | 18:40 |
| piggz | yes .... | 18:41 |
| mal | piggz: have you checked if that /proc/*/ns/mnt exists and which permissions it has | 18:41 |
| piggz | mal: on the pinephone, after 4.2 upgrade, apps dont launch from lipstick | 18:41 |
| T42 | <webdotcom> are you a bot? (re @SailfishFreenodeIRCBridgeBot: <piggz>mal: on the p...) | 18:43 |
| T42 | <webdotcom> or you are just replying with a bot? | 18:43 |
| piggz | wtf | 18:44 |
| T42 | <b100dian> @webdotcom the discusson is on the IRC channel, T42 helps us connect the two groups | 18:44 |
| piggz | god damn TG users :D | 18:44 |
| mal | hehe | 18:44 |
| T42 | <webdotcom> what channel is that? (re @b100dian: @webdotcom the discu...) | 18:44 |
| piggz | mal: perms are 777 defaultuser:defaultuser | 18:44 |
| mal | piggz: those are actually links | 18:45 |
| T42 | <b100dian> #sailfishos-porters on irc.oftc.net | 18:45 |
| piggz | mal: yes... | 18:45 |
| Nico | "are you a bot" <- That's great :D | 18:45 |
| T42 | <b100dian> piggz talks so much to be passed as a bot. 'sup? | 18:45 |
| T42 | <webdotcom> ok then. where you able to port sailfish for rosy? (re @SailfishFreenodeIRCBridgeBot: <Nico>"are you a bot...) | 18:45 |
| T42 | <webdotcom> i can not load that website (re @b100dian: #sailfishos-porters ...) | 18:46 |
| T42 | <b100dian> @webdotcom please see the telegram pinned message, for compatiblity with IRC don't use the Reply feature in Telegram | 18:46 |
| piggz | @webdotcom no, ive not ported that device | 18:46 |
| T42 | <b100dian> @webdotcom notice the rosy message is written by Tofe | 18:47 |
| T42 | <webdotcom> ok.ok. i did not know for what that rule is | 18:47 |
| piggz | mal: for a random process ownder by defaultuser: https://paste.mozilla.org/H9UgS10p | 18:47 |
| mal | piggz: which namespace options do you have enabled in kernel config? | 18:48 |
| piggz | s/owned | 18:48 |
| piggz | i actuall have the open right now! ... | 18:48 |
| piggz | CONFIG_NAMESPACES=y | 18:48 |
| piggz | CONFIG_UTS_NS=y | 18:48 |
| piggz | CONFIG_TIME_NS=y | 18:48 |
| piggz | CONFIG_IPC_NS=y | 18:48 |
| piggz | CONFIG_USER_NS=y | 18:48 |
| piggz | CONFIG_PID_NS=y | 18:48 |
| piggz | CONFIG_NET_NS=y | 18:48 |
| T42 | <webdotcom> /notes | 18:49 |
| T42 | <webdotcom> no notes in here? | 18:49 |
| deathmist | @webdotcom what exactly are you looking for? does the pinned message not satisfy your needs? | 20:18 |
| T42 | <webdotcom> I didn't know this group isn't like most of the groups. I thought there are notes here as well. I didn't look for something in particular. | 20:22 |
| T42 | <b100dian> mal: one last point about the ultrasound proximity sensor: I switched the SIM card and the workaround (disabling mapplauncherd-booster-silica-qt5-media) doesn't work anymore. A race is a race.. need to understand how pulse could help me with that if I am not wrong | 20:36 |
| mal | does the sensor really work via sensorfw if you have those audio service running? | 20:43 |
| T42 | <b100dian> the sensor seems two have two parts: a control device /dev/elliptic0 and some dsp devices | 20:47 |
| T42 | <b100dian> the last part is unknown to me, so only enabling the android audioservice and audio-hal-2.0 does the control part work | 20:48 |
| T42 | <b100dian> short answer: yes, sensorfw works if I enable the audio services | 20:50 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!