[REPORT] Lenovo IdeaPad Slim 3 15Q8X10 83TW (not 83N3)

[esveiga-antimatter]

Hardware: LENOVO 83TW / LNVNB161216, BIOS RVCN25WW (11/26/2025)
CPU (SMBIOS): Snapdragon(R) X - X125100
Kernel: 7.2.5-jg-0-qcom-x1e with x126100-lenovo-ideapad-slim3.dtb, installed via Ubuntu 26.04 ISO

Working: boot from NVMe, internal keyboard (048D:83DB on 1-003a, TLMM 67), battery (ADSP attached
from UEFI), Bluetooth (ROM firmware only, see below), Wi-Fi (with workaround, see below), display.

Issues found:

1. Bluetooth node declared as WCN7850, chip is WCN6855 (SOC version 0x400c1211). The driver asks for
   qca/hmtbtfw21.tlv and runs ROM-only; hpbtfw21.tlv is installed. Probably affects 83N3 too.
2. Wi-Fi: no board data for subsystem 17aa:e12a in linux-firmware (variant LE_S390).
   Works temporarily with the 17aa:9309 qmi-chip-id=18 entry as board.bin.
3. USB controller a800000 fails: "dwc3-qcom: failed to register DWC3 Core"
   (PTN3222 eUSB2 repeater not found). One USB port likely dead on 83TW.
4. Audio: LPASS pinctrl and codecs stuck in deferred probe.
5. GPU: Zap shader missing on Ubuntu 26.04 → software rendering. Fixed by extracting Qualcomm's driver installer exe
6. eDP panel BOE 0x0cea unknown to panel-edp (warning, display works).

Logs attached: dmesg, lspci, deferred devices, hwids.
Happy to test patches or DTB changes.
deferred.txt
dmesg-full.txt
dmesg-wireless.txt
firmware-files.txt
lspci.txt
rfkill.txt
uname.txt

[jglathe]

83TW is actually one of the better supported devices IMO, just not by me. I'll ask @polkorny if he can help out.

[jglathe]

My own ips3x adventures are a bit on the backburner, sadly. I know the path to take (EC support to unfuck the touchpad on my SKU), but no time atm.

[polkorny]

Just a heads-up: I’m on vacation at the moment. When I’m back at work I’ll pick the Ideapad Slim 3 (83TW) up again and continue from there.

In the meantime, I rebased the commits onto the current branch, in case it’s useful for testing: https://github.com/polkorny/linux_ms_dev_kit/commits/jg/ubuntu-qcom-x1e-7.2.y/

1. Bluetooth: My model also uses the WCN6855, so you can use the commit that changes the DTS for that chip.

2. Wi-Fi: Yes, we are in contact with Lenovo to add this vendor to linux-firmware. Mine is also the LE_S390 variant, so it should be added soon. For now, using a different entry as board.bin is the best option.

3. USB: I worked on this before and I remember fixing it, but I don’t remember which commit it is in. I will look for it and tell you.

4. Audio: I made some commits for this machine. I also opened a PR on alsa-ucm-conf, because this machine needs its own UCM config.

[esveiga-antimatter]

minor bump, to report some updates as I tinker away with this model and after testing @polkorny 's current branch.

I figured out how to make the touchpad on my unit work. This specific unit has a Goodix GXTP5100 (27c6:01e0) at i2c bus 0 (88c000), addr 0x5d**. Nothing answers at 0x38. The interrupt for this is in TLMM gpio 19, level low, with external pull-up. So in this specific case, there's no need for an IRQ mux and both kb and touchpad work.

Regarding audio, seems like my habit of nuking the hard drive when first booting the computer bit me, because now I dont have the windows factory driver store to extract the audio drivers.

[jglathe]

Now thats news. TLMM19 is what the "formula" actually gives as result, but it never worked. So its not really a shared interrupt? That pdc patch appears to work wonders then. My ips3x also has a Goodix touchpad. I will test that again, of course.
Audio drivers? You mean the firmwares for adsp and cdsp I guess. If required, just send me a private email.

[esveiga-antimatter]

This is the diff against this repo's current DTS:

@@ touchpad@5d {
 		reg = <0x5d>;

 		hid-descr-addr = <0x1>;
+		interrupts-extended = <&tlmm 19 IRQ_TYPE_LEVEL_LOW>;
+
+		pinctrl-0 = <&tpad_default>;
+		pinctrl-names = "default";

 		vdd-supply = <&vreg_misc_3p3>;
 		vddl-supply = <&vreg_l12b_1p2>;
@@ &tlmm {
+	tpad_default: tpad-default-state {
+		pins = "gpio19";
+		function = "gpio";
+		bias-disable;
+	};
+
 	kybd_default: kybd-default-state {

I found it by sampling the TLMM levels in debugfs while touching the pad, then confirmed it with gpiomon -c gpiochip0 -e both 19. That shows nothing at rest, a falling edge every ~200 ms while a finger is down (unread reports), and a rising edge on release. With the mux I also confirmed that touching the pad never raises GPIO 67.

Yes i meant audio firmware. I'll shoot you an email.