| spiiroin | seems that piggz is already gone, but: pong | 06:02 |
|---|---|---|
| T42 | <adampigg> Spiiroin: give me 50 mins :) | 07:24 |
| T42 | <adampigg> Spiiroin: how does usbmoded on the pinephonepeo doesn't react to connect\disconnect, so wondering how that is handled. Mce does see correct charge state, but thats using the power devices | 08:03 |
| T42 | <adampigg> Also, im able to configure gadget modes in configfs manually ok | 08:05 |
| Thaodan | @elros34: Yes I plan to do so, last weekend the obs had some issues, now everything is build. | 08:10 |
| Thaodan | I didn't wrote anything down but the update should work for existing installations. | 08:11 |
| T42 | <elros34> @thx for info. @Mat60000 ^ | 08:17 |
| spiiroin | @adampigg IIRC the problem with pinephone was (still is?) that there just are no change notifications coming from kernel on cable connect/disconnect | 08:42 |
| T42 | <adampigg> This is the pro...ill have to check the original pp | 08:43 |
| T42 | <adampigg> What device should emit those events? | 08:43 |
| spiiroin | In case of mce, there was some charging related event that could be used as a trigger to re-read all power supply device properties -> such poll catches also cable state change | 08:43 |
| spiiroin | and polling also on system heart beat -> eventually catches cable disconnect (at up to 24 second delay) | 08:44 |
| T42 | <adampigg> On the pro, mce is quick, and the kernel spits ous some usb message, but im at work atm and cant check | 08:45 |
| T42 | <adampigg> Seperate, mcetool -U doesn't wake the screen, only power button press | 08:46 |
| spiiroin | basically whenever there are significant changes in what one gets via "grep . /sys/class/power_supply/*/uevent" -> a notification should already have been sent | 08:46 |
| spiiroin | where significant would be state changes like connect/disconnec, charging/not etc | 08:47 |
| T42 | <adampigg> Where does usb moded look for those events and i can check later | 08:47 |
| spiiroin | (whereas things like voltage/current changes are probably too volatile for such reporting) | 08:47 |
| spiiroin | @adampigg this is where events are processed: https://github.com/sailfishos/usb-moded/blob/master/src/usb_moded-udev.c#L321 | 08:50 |
| spiiroin | for tracing what happens: usb-moded is interested in subset of uevents mce handles -> mce verbose logging can be used for following also things usb-moded is interested in | 08:52 |
| T42 | <adampigg> Thx, ive got moded debugging on, will add mce, as i say, mce seems ok | 08:53 |
| spiiroin | ... or mce sort of works because of all the workarounds, some of which are pinephone specific | 09:00 |
| spiiroin | @adampigg about that "mcetool -U" vs. powerkey. one thing that comes to mind is that by default: powerkey unblanks regardless of proximity sensor state, whereas everything else needs proximity sensor reporting "not covered" | 09:01 |
| spiiroin | ... and about the same applies for lid/cover/hall effect sensor too | 09:04 |
| T42 | <adampigg> Spiiroin thx will check those | 09:57 |
| *** amccarthy is now known as Guest1076 | 13:53 | |
| *** amccarthy_ is now known as amccarthy | 13:53 | |
| T42 | <Mat60000> Just wanted dual sim ans waydroid on my Xperia 5;) | 16:56 |
| piggz | spiiroin: just checked with udevadmin monitor, pluggin a cable and get the following events | 18:57 |
| piggz | KERNEL[59.938341] change /devices/platform/typec-extcon/extcon/extcon2 (extcon) | 18:57 |
| piggz | UDEV [59.942207] change /devices/platform/typec-extcon/extcon/extcon2 (extcon) | 18:57 |
| piggz | full log https://paste.mozilla.org/nk9AjV6w | 18:59 |
| piggz | so, seems to me like there are event that usb-moded could monitor? | 18:59 |
| spiiroin | piggz: can you take traces of connect and disconnect with "udevadm monitor -p -s power_supply" | 19:16 |
| piggz | spiiroin: sure.. | 19:17 |
| piggz | spiiroin: https://paste.mozilla.org/WJNkvguB | 19:18 |
| spiiroin | interesting, there is sane lookin POWER_SUPPLY_ONLINE=0/1 changes | 19:30 |
| spiiroin | piggz: can you do "grep . /sys/class/power_supply/*/uevent" too, in connected and disconnected states | 19:31 |
| piggz | sure | 19:31 |
| piggz | spiiroin: https://paste.mozilla.org/fJ4z61es | 19:33 |
| piggz | spiiroin: i dont want to confuse devices, but this is the same for the original pinephone https://paste.mozilla.org/USre3OYM | 19:35 |
| piggz | other device is pinephone-pro | 19:36 |
| piggz | any suggestions that work on both is nice | 19:38 |
| piggz | spiiroin: also, on the PPP, im using your new PR for usb-moded for cached network | 19:38 |
| spiiroin | hmm.. there are rk818-usb and tcpm-source-psy-4-0022, both of which look like "charger", but only the latter one showed up in that udevadm event log | 19:57 |
| spiiroin | might be that usb-moded happens to choose wrong device to track | 19:57 |
| spiiroin | I'll take a closer look at this later, but if you want to investigate ^ would be something to check | 19:58 |
| T42 | <adampigg> spiiroin: im off out for the night now, but i know now where to add debugging if you dont get to look at it, thx | 20:11 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!