Battery gauge issue (mainline kernel) - Printable Version +- PINE64 (https://forum.pine64.org) +-- Forum: Pinebook Pro (https://forum.pine64.org/forumdisplay.php?fid=111) +--- Forum: Linux on Pinebook Pro (https://forum.pine64.org/forumdisplay.php?fid=114) +--- Thread: Battery gauge issue (mainline kernel) (/showthread.php?tid=13546) |
Battery gauge issue (mainline kernel) - lfourquaux - 04-03-2021 Hello, I'm trying to use Debian on a Pinebook Pro using the official bullseye installer. (I'm aware of the unofficial installer!) Mostly, it works quite well! For the record: * the installer created an additional partition (dummy partition to protect the bootloader, maybe?) that confused u-boot; deleting the partition fixed the problem (I plan to reproduce the unofficial installer's nice partition scheme later); * the screen remains blank (I installed using a serial cable) because the kernel gets confused by the video + usb-c port; this can be worked around by removing Code: &cdn_dp { * the battery gauge is not working (this is the reason for this post); * there is something wrong with the audio output, which I have not investigated (the uart/audio switch is set to uart, for serial output, it would make sense to try setting it to audio first). From dmesg, I see a few messages related to the battery gauge: Code: [ 0.000000] Booting Linux on physical CPU 0x0000000000 [0x410fd034] My understanding is that cw2015 support has been mainlined and should work with a mainline kernel. I don't really understand what the "Not all required supplies found" message is about, but dmesg does not seem to have errors relating to the DC input not the USB-C input. This is the only occurrence of "power supply" in these messages. The same errors appear with the original .dtb file, so this is not a side effect of the above change. Do you know what is causing the cw2015 driver to fail? Any suggestion for a fix or workaround would be highly appreciated! Best regards, Lionel RE: Battery gauge issue (mainline kernel) - lfourquaux - 05-22-2021 Answering my own post: this was caused by missing drivers in the Debian kernel: enabling CONFIG_TYPEC_FUSB302 (requires CONFIG_TYPEC and CONFIG_TYPEC_TCPM) and CONFIG_SND_SOC_ES8316 makes the type-c port, battery gauge and audio work. This is currently fixed in Debian sid (from version 5.10.38-1 of linux-image-amd64). One remaining oddness is that the usb-c port does not always work in all orientations: reversing the usb connector can make a usb device work. Does anyone understand why this is happening? RE: Battery gauge issue (mainline kernel) - moonwalkers - 05-22-2021 (05-22-2021, 03:58 AM)lfourquaux Wrote: Answering my own post: this was caused by missing drivers in the Debian kernel: enabling CONFIG_TYPEC_FUSB302 (requires CONFIG_TYPEC and CONFIG_TYPEC_TCPM) and CONFIG_SND_SOC_ES8316 makes the type-c port, battery gauge and audio work.I believe you meant `linux-image-arm64`. What's your experience with it so far? I've seen some instability around WiFi - device likes to disappear out of the blue, haven't noticed any pattern yet. With xmixahx's 5.8.x I had it disappear only occasionally after deep sleep, but with 5.10.0-7 of Debian it can sometimes disappear in the middle of my typing up a message in IM. (05-22-2021, 03:58 AM)lfourquaux Wrote: One remaining oddness is that the usb-c port does not always work in all orientations: reversing the usb connector can make a usb device work. Does anyone understand why this is happening?AFAIK Type-C is not perfectly reversible, and that was brought up before. Somebody on these forums even referred to the USB-C spec not guaranteeing 100% reversible connection. Which seems to me to defeat the purpose, but oh well. I haven't spent enough time reading the spec to tell whether it is an issue with the hardware spec or if it is a software (driver) issue. RE: Battery gauge issue (mainline kernel) - lfourquaux - 05-22-2021 (05-22-2021, 09:53 AM)moonwalkers Wrote: I believe you meant `linux-image-arm64`. Indeed! (05-22-2021, 09:53 AM)moonwalkers Wrote: What's your experience with it so far? I've seen some instability around WiFi - device likes to disappear out of the blue, haven't noticed any pattern yet. With xmixahx's 5.8.x I had it disappear only occasionally after deep sleep, but with 5.10.0-7 of Debian it can sometimes disappear in the middle of my typing up a message in IM. Very good experience overall. Regarding WiFi, it has been very stable for me, though a bit slow (the Pinebook Pro has 1 stream only; even my phone has 2). I have read on the Pine64 wiki that 2.4GHz WiFi may be unstable. I have mostly, but not exclusively, used 5GHz networks. (FWIW, I used iwd to configure the WiFi adapter.) The remaining pain point is audio input. The builtin microphone seems very poor, but I also had poor sound quality with a usb mic that worked fine on another computer. I'm still experimenting and it may be only a software misconfiguration. (05-22-2021, 09:53 AM)moonwalkers Wrote: AFAIK Type-C is not perfectly reversible, and that was brought up before. Somebody on these forums even referred to the USB-C spec not guaranteeing 100% reversible connection. Which seems to me to defeat the purpose, but oh well. I haven't spent enough time reading the spec to tell whether it is an issue with the hardware spec or if it is a software (driver) issue. Thanks for these informations! RE: Battery gauge issue (mainline kernel) - jiyong - 05-22-2021 What software are you using to record audio? I use a RØDE NT-USB Mini with SimpleScreenRecorder and I it works like a charm. I haven't tried to use it with video conferencing, as you need to keep the RØDE very close to your mouth. This is an example of how it sounds. https://youtu.be/XN7ZEmgVBzA |