| rinigus | piggz: that would be handy! | 05:23 |
|---|---|---|
| rinigus | But looking into the list of PRs sitting and waiting there, I wouldn't be that hopeful in timely manner of the merge. Unless you submit it as sledges (see merged PRs) | 05:26 |
| spiiroin | piggz: the usual suspects: misidentified power supply devices / problems with change notifications | 06:38 |
| spiiroin | what mce version you have there? 1.106.0 has fixes / enablers for some cases | 06:39 |
| piggz | spiiroin: 1.105.3 | 06:40 |
| spiiroin | piggz: for example what happened in pinephone on cable disconnect: charger device notified about connected->disconnect, but battery device did not notify about charging->discharging => state evaluation saw "battery is getting charged" -> heuristics misfire -> "there has to be a charger connected" -> similar "it thinks its plugged in all the time" situation | 06:52 |
| piggz | spiiroin: ill update mce then get back to you if there are still issues | 06:53 |
| spiiroin | workaround: make mce poll all power supply devices when one of them reports change -> https://git.sailfishos.org/mer-core/mce/commit/a01dacd6caa88aafc1810761c7a3cc8476c98a65 | 06:53 |
| spiiroin | needs to be enabled via config file | 06:54 |
| spiiroin | (the defaults for 1.106.0 = polls info periodically -> helps with things like not missing percentage changes, but not with "hard edges" such as what happens with cable disconnect) | 06:55 |
| piggz | spiiroin: usb plug in > developer mode in use, usb unplug > charging :D | 07:20 |
| spiiroin | piggz: what would be expected in case of: connect -> usb.online=1 + battery.state=charging ; disconnect -> usb.online=0 + battery.state left in charging -> "some sort of charger that is not pc connected" | 07:22 |
| piggz | spiiroin: with mcetool, and cable disconnected, i see charger cable: disconnected, charger state: on | 07:40 |
| piggz | there are 5 devices in /sys/class/power | 07:40 |
| piggz | https://paste.mozilla.org/ZhDdE9rB | 07:41 |
| piggz | spiiroin: when unplugged, /sys/class/power/*/online are all 0 | 07:58 |
| spiiroin | piggz: the suspected problem comes from: tracking state via udev notifications produces different result vs reading that data directly | 09:13 |
| piggz | spiiroin: i wonder what causes it to think its charging when unplugged tho | 09:15 |
| piggz | i tried that config you sent, but that seems to be for something different, and didnt work anyway | 09:15 |
| spiiroin | piggz: basically, if mce ends up in situation where every charger.online==0 but battery.state==charging -> disconnected but charging | 09:16 |
| spiiroin | what you could do is: enable verbose logging from battery-udev plugin -> follow change notifications mce sees & compare vs what you see by reading directly from sysfs | 09:18 |
| spiiroin | assumed to be the most relevant bits being online props for charger devices and status prop for the (hopefully only one) battery device | 09:20 |
| piggz | spiiroin: i have logs....looks like it could be an unknown charget type.... | 09:21 |
| piggz | spiiroin: the gap at line 30 is where i unplug https://paste.mozilla.org/VYOe34Tj | 09:22 |
| spiiroin | piggz: ok, this was a novel gotcha. seems battery is treated as charger ... | 09:23 |
| spiiroin | can you paste uevent content from that cw-bat | 09:24 |
| piggz | spiiroin: wdym uvent content? | 09:25 |
| piggz | spiiroin: in /sys its type is "Battery" | 09:26 |
| spiiroin | cat /sys/class/power_supply/cw-bat/uevent | 09:26 |
| piggz | POWER_SUPPLY_NAME=cw-bat | 09:26 |
| piggz | POWER_SUPPLY_CAPACITY=91 | 09:26 |
| piggz | POWER_SUPPLY_HEALTH=Good | 09:26 |
| piggz | POWER_SUPPLY_PRESENT=1 | 09:26 |
| piggz | POWER_SUPPLY_VOLTAGE_NOW=4318000 | 09:26 |
| piggz | POWER_SUPPLY_TIME_TO_EMPTY_NOW=8191 | 09:26 |
| piggz | POWER_SUPPLY_TECHNOLOGY=Li-ion | 09:26 |
| spiiroin | it does not have status property. does /sys/class/power_supply/battery/uevent have one? | 09:27 |
| piggz | spiiroin: yes | 09:28 |
| piggz | [root@Phone nemo]# cat /sys/class/power_supply/battery/uevent | 09:28 |
| piggz | POWER_SUPPLY_NAME=battery | 09:28 |
| piggz | POWER_SUPPLY_STATUS=Charging | 09:28 |
| spiiroin | piggz: ok, that is probably it: lack of status prop -> false "this must be charger" -> to avoid it, blacklist that device | 09:29 |
| spiiroin | printf >> /etc/mce/99-hackhack.ini "[BatteryUDevDeviceBlacklist]\ncw-bat=true\n" | 09:30 |
| spiiroin | or so, and then restart mce | 09:30 |
| piggz | thx | 09:30 |
| piggz | btw, need to move away from term blacklist :) | 09:31 |
| * spiiroin waits for somebody to find red-black trees ... | 09:32 | |
| piggz | spiiroin: working :) | 09:32 |
| piggz | spiiroin: oddity, when i unplug, it says "charging complete" is tht normal? its at 92% | 09:33 |
| spiiroin | piggz: hopefully with charger only? i.e. not when connecting/disconnecting from pc | 09:35 |
| piggz | spiiroin: no,its on my pc atm | 09:36 |
| spiiroin | urgh. some sort of false battery full perhaps | 09:37 |
| spiiroin | there were problems with some mtk devices and battery full reporting -> more heuristics that can misfire | 09:38 |
| spiiroin | do you still have that ini-file in place that was discussed earlier, the one for polling stuff? | 09:38 |
| piggz | spiiroin: no, i changed that one for the new hack | 09:38 |
| spiiroin | ok, good. then the 1st thought is: battery=discharging gets acted on before charger=disconnected -> former might tickle battery full heurisitics (charger connected, battery from charging to discharging @ capacity > N == full) | 09:41 |
| spiiroin | so, it could be caused by order/timing of change notifications in kernel->udev->mvce, or mce->dbus->battery status tracker in lipstick | 09:43 |
| spiiroin | to verify, you could run "mcetool --set-battery-level=42 -B" in one shell and leave it there, then attach/detach cable and see if it still happens | 09:47 |
| spiiroin | requires that you have "devel flavor" mce though, battery simulator api is not made available normally | 09:47 |
| T42 | Kovid1984 was added by: Kovid1984 | 13:31 |
| *** MagneFire_ is now known as MagneFire | 13:32 | |
| piggz | mal: remember u promised to look at dhd pr ;) | 16:57 |
| mal | piggz: oops | 17:28 |
| piggz | mal: yes, oops! :D | 17:36 |
| mal | piggz: maybe fix the typos in the commit message | 17:38 |
| piggz | :/ | 17:39 |
| piggz | mal: done | 17:41 |
| mal | looking good now | 17:53 |
| T42 | <ItsMeShouko> Hello, I need some help. I am porting 3.3.0 to mi a2 and I got error while trying to run: rpm/dhd/helpers/build_packages.sh --mw | 19:21 |
| T42 | <ItsMeShouko> Cloning into 'libnciplugin'... | 19:21 |
| T42 | <ItsMeShouko> Already up to date. | 19:21 |
| T42 | <ItsMeShouko> No provider of 'pkgconfig(nfcd-plugin) >= 1.0.27' found. | 19:21 |
| T42 | <ItsMeShouko> Setting version: 1.0.6 | 19:21 |
| T42 | <ItsMeShouko> error: Failed build dependencies: | 19:21 |
| T42 | <ItsMeShouko> pkgconfig(libncicore) is needed by libnciplugin-1.0.6-0.armv7hl | 19:21 |
| T42 | <ItsMeShouko> pkgconfig(nfcd-plugin) >= 1.0.27 is needed by libnciplugin-1.0.6-0.armv7hl | 19:21 |
| T42 | <ItsMeShouko> Building target platforms: armv7hl-meego-linux | 19:21 |
| T42 | <ItsMeShouko> Building for target armv7hl-meego-linux | 19:21 |
| T42 | <ItsMeShouko> How can I fix this? | 19:21 |
| T42 | <ItsMeShouko> Here are my repos: https://github.com/SailfishOS-MI-A2 | 19:23 |
| mal | try running "rpm/dhd/helpers/build_packages.sh --mw=https://git.sailfishos.org/mer-core/nfcd.git" | 20:07 |
| T42 | <ItsMeShouko> It worked, thank you. However I got one error after running | 20:41 |
| T42 | <ItsMeShouko> rpm/dhd/helpers/build_packages.sh --version | 20:41 |
| T42 | <ItsMeShouko> https://paste.linux.pizza/?cdd5d971fb609ac0#AzQka36LsY6u6GbmEMPgoWJqQ4DdibFDoWcYCuCrRVZU | 20:41 |
| mal | sb2 -t $VENDOR-$DEVICE-$PORT_ARCH -m sdk-install -R chmod 755 /boot | 20:46 |
| mal | if that doesn't help then you need to provide the droid-hal-version-jasmine_sprout.log which is mentioned in the paste you showed | 20:47 |
| T42 | <ItsMeShouko> Yes, it didn't work | 20:48 |
| T42 | <ItsMeShouko> (Document) https://irc.thaodan.de/.imgstore/P1naKz6wKS.null | 20:49 |
| mal | sb2 -t $VENDOR-$DEVICE-$PORT_ARCH -R -msdk-install zypper in droid-config-$DEVICE -ofono-configs-mer | 20:58 |
| xmn | Hey Sailfish peeps. Question, would this be a good device to try a. run SF on? https://forum.xda-developers.com/lg-g-pad-83 | 21:55 |
| xmn | *and | 21:57 |
| T42 | <ItsMeShouko> Thanks mal, it worked. After solving some more issues like these I am getting package issues with --mic command, will post about it later | 22:09 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!