Replies: 34 comments 141 replies
-
|
Update, I tried your 6.14-rc1 kernel packages and swallowed my pride with qemu and got my laptop to boot. It even started X11, but then after a while of showing a rotating circle it rebooted and the UEFI wanted to "repair the system." |
Beta Was this translation helpful? Give feedback.
-
|
Hi @axu-trex, yeah that is a nice budget laptop with loads of RAM and nice peripherals. I have mine up with Ubuntu, didn't publish the image yet. Let me update that one with the latest kernel and dtb, will be worth it. I recommend booting either from type-a SSD enclosure (fastest), stick, or SD-card. SD-card is odd, you can't use the fast UHS-II ones. Haven't tested UHS-I yet. Re installing on x86 box in chrooted mount: yes that should be possible, but there are easier tasks. I would recommend to wait until I publish the bootable image. Today. |
Beta Was this translation helpful? Give feedback.
-
|
Having used the laptop for a while now, I'm growing frustrated with the touchpad. It's so big that it's difficult to use the keyboard in an ergonomic posture without accidentally touching the touchpad. I also haven't figured out how to generate a middle mouse button click, have you? Do you know if there are any tools on Linux to configure this touchpad? I've never disliked any touchpad before. |
Beta Was this translation helpful? Give feedback.
-
|
The first reboot is intentional, btw. You need to be fast with F12 |
Beta Was this translation helpful? Give feedback.
-
|
Hi @axu-trex , the image is ready for download and test. You can find it here. It should work ootb, including local keyboard and touchpad. Depends on the actual SKU. There is no gpu support yet, its still software-rendered. WiFi, BT work. Card reader works, but not with UHS-II. type-c ports and type-a ports work. DP Altmode on type-c works, with 2 lanes. HDMI not yet, Sound/speakers not yet. Sound via USB / BT works fine. |
Beta Was this translation helpful? Give feedback.
-
|
Thanks, writing it to a USB stick right now. |
Beta Was this translation helpful? Give feedback.
-
|
(this stick is so slow) |
Beta Was this translation helpful? Give feedback.
-
|
My SKU is |
Beta Was this translation helpful? Give feedback.
-
|
I've got the new image installed now and snap repaired. I've gotten one weird hang so far, and on the next boot it didn't seem to find the SSD which it found on the first boot. About the hang: |
Beta Was this translation helpful? Give feedback.
-
|
A slow USB stick is no fun with this. My recommendation would be to try with an SSD enclosure as boot device. For reasons (adsp firmware booting and VBUS woes) I recommend using a USB-A port... directly, not behind a hub. I have a UGreen SSD Enclosure with RTL9210 chip, they are good (as in no crippling bugs) and support type-c and type-a cables. |
Beta Was this translation helpful? Give feedback.
-
|
I've had several enclosures and SATA adapters over the years, and they've all broken down just by sitting unused on a shelf. Computer no longer recognizes that anything is connected when I try to plug them in. 🙁 I'll give UGreen a go, if I can find them in some shop around here tho. |
Beta Was this translation helpful? Give feedback.
-
|
Otherwise, seen only minor issues on the Thinkbook so far. It changes magically with fast I/O available. |
Beta Was this translation helpful? Give feedback.
-
|
What hurdles do we still have to jump over before it can boot off of the built-in SSD? How about the built-in camera and audio subsystem? |
Beta Was this translation helpful? Give feedback.
-
|
Run it on the nvme: just some courage, and maybe a recovery stick if it goes south (you can order the image from Lenovo support online, need to register). Make sure you have a backup of all your data on this laptop.
We're in the new system now and need to modify
Camera, sound, fingerprint sensor - not even looked at yet, tinkering to get it working on HP Omnibook X14, on the list |
Beta Was this translation helpful? Give feedback.
-
|
After some banging of the head, I've now got Debian/sid installed and booting from the laptop internal SSD. |
Beta Was this translation helpful? Give feedback.
-
|
The Debian grub.d scripts don't handle devicetrees. I'm thinking of submitting a bug report to get them included, but I'd like some input from you:
|
Beta Was this translation helpful? Give feedback.
-
|
Actually, this is quite the topic. Ubuntu carries the patch, but its also incomplete (booting on other drives has it missing). Grub didn't include them yet. Is your fix for the other disks? |
Beta Was this translation helpful? Give feedback.
-
|
No I wouldn't have been submitting a patch, just a bug... 😇 I found these two while looking around:
I would be basically saying "hi Debian, copy this patch from Ubuntu!" |
Beta Was this translation helpful? Give feedback.
-
|
The devicetrees are needed mostly only on arm64, right? Or do a lot of armel and armhf platforms need it too? My raspberries running Debian don't need it, but that may be a special case, what do I know... While looking around, I also found some cases where it was needed on an obscure x86 platform as well... I suppose before grub-upstream will accept a patch, they want it to be perfect and to cover all corner cases? You mentioned that os-prober sourced grub menu entries are missing the devicetree setting in Ubuntu. Yes, I ran into that while setting up Debian on the SSD... But I fear that might be a tough one: correct me if I'm wrong, but you can also have the device tree staticly compiled into the kernel in some operating systems, can't you? And that's not really possible to detect. Might be easier, at least in the case of Linux, if each OS had a grub.cfg clip for themselves in some standard location that the os-prober looks at... On second thought, that's a serious security issue. |
Beta Was this translation helpful? Give feedback.
-
|
I tried the kernel 6.14.0-17-jg-2 on the ThinkBook and couldn't get it to work. The screen just stayed black forever, after it switched console, so I couldn't even see if it panicked or what. Not sure if it was even meant to work tho. |
Beta Was this translation helpful? Give feedback.
-
|
Ouch. Well, no it was not supposed to, not very surprising it didn't. From 6.14-rc3 on there are patches for the retimers (Parade PS8830), and one property is that it really goes south when the chip is not there but assumed to be, ending in a blank screen and probably sort of crashed, as it was on my HP Omnibook X 14, and on the Devkit (or so), but not yet on the Thinkbook. The latest that should work is 6.14-rc3-0. |
Beta Was this translation helpful? Give feedback.
-
|
I was traveling when you released it, and I tried it at an airport, but once I got home I got sick and forgot about it. Anyway, it works! I'm still getting a bunch of kernel thread crashes(?) every boot. I don't know if they're related to Snapdragon or something else, but here are some of the kernel logs in case they help you: These are all from the same boot-up, but ntp adjusted the clock in the middle of the log... |
Beta Was this translation helpful? Give feedback.
-
|
Well those splats are Iris based. Which is a pain, as it comes from failing to load the firmware vpu30_p4.bin. To avoid this we need to point to the "right" irmware, although this might (and surely will) kill video playback in firefox. So... I disabled it by creating an empty file: The newer Ubuntu Concept X1E based kernels have iris disabled AFAIK. |
Beta Was this translation helpful? Give feedback.
-
|
There is a new build, 6.14.0-17-jg-3. The repository branch is jg/ubuntu-concept-6.14 but it is not tagged yet, expecting further changes. I am using this corrected dtb, looking good so far. |
Beta Was this translation helpful? Give feedback.
-
|
I ran a tests again on the ThinkBook 16 G7 QOY variant 21NH with the with the Unencrypted Windows Partition Here are the findings out of the box: Tested Laptop: ThinkBook 16 G7 QOY (Variant: 21NH) Test Results Thanks for all the work if a testing is needed, I’m happy to help! |
Beta Was this translation helpful? Give feedback.
-
|
Hi, congrats! Results looking good / as expected. You could try a newer kernel if you want. I'm running 6.14.0-25-jg-0 currently. Works nice, unless you need to change the WLAN network (will crash it). The software rendering is due to missing GPU support, nothing published yet. |
Beta Was this translation helpful? Give feedback.
-
|
Hmm ... what's the deal with RTC? Recently my laptop always boots up with a date a week or two in the past, and I think it's the date when I switched from one kernel version to another. So I'm thinking maybe there's some bug with RTC in the more recent versions of the kernel? Searching for "rtc" from the beginning of my kern.log: (where kernel version is 6.14.0-5-qcom-x1e) And from the end of the log: (where kernel version is 6.14.0-17-jg-3-qcom-x1e) |
Beta Was this translation helpful? Give feedback.
-
|
There was a bug in my tree, yes. Fixed with 6.14-rc6 and 6.14.0-25-jg-0. |
Beta Was this translation helpful? Give feedback.
-
|
I understand that this laptop has the Adreno GPU, and that Linux kernel already supports Adreno. So what's missing? Is the Adreno on this laptop a different variant than the two(?) chips that already have Linux support? Or do we just need to find the right device tree entries for the GPU on this laptop to get the driver (and presumably also HDMI) working? |
Beta Was this translation helpful? Give feedback.
-
|
Hmm a bit more complicated. At least Qualcomm is aware of this. That said its really a matter of correct node description on the SoC. My attempts ended with correct identification of the chip and trying to power it up. Probably it is mainly that. The adreno drivers are for many GPU generations now and sort of generic. There is a catalog record for a new variant, which I filled in by guessing. X1-45 and X1-85 are functionally the same (?), only different sizes. I guess we will have to wait. |
Beta Was this translation helpful? Give feedback.
-
|
Hi, smee again. Can you give me pointers with tracking battery status? Upower seems to identify the battery, but can't get any real numbers for it. BTW, I'm testing 6.14.0-31-jg-1. Haven't noticed any new issues. |
Beta Was this translation helpful? Give feedback.
-
|
Hi, i'm kopernikustheog from UBUNTU thread. i did decrypted windows partition and used USB type-a I also checked this^ discussion and haven't found any type of problem that would be similar to the one i'm facing |
Beta Was this translation helpful? Give feedback.
-
|
Ok, i'll try that tomorrow it's 11 pm in here. Thx again |
Beta Was this translation helpful? Give feedback.
-
|
it didn't work, I'm starting to think it's because of the processor in the laptop. i'll try other images from the link you gave me |
Beta Was this translation helpful? Give feedback.
-
|
Do you have some more debug info? I have the same laptop, so lets assume the thing can work. Can you try a different USB stick, perchance? The other type-a connector? No other peripherals connected (except for maybe a mouse)? I've seen some odd things there. |
Beta Was this translation helpful? Give feedback.
-
|
|
Beta Was this translation helpful? Give feedback.
-
|
Sorry for not giving more specific informations, i'm new into linux |
Beta Was this translation helpful? Give feedback.
-
|
WOW, great work on the camera! |
Beta Was this translation helpful? Give feedback.
-
|
🙏 Well for the tb16 it was sort of easy (no it wasnt) because it is a usb camera connected via the usb_2 bus. For the T14s there should be camera already, it probably isn't upstream yet. Its an Omnivision camera on the mipi (?) bus. Maybe @kuruczgy or @alexVinarskis can help with that. @alexVinarskis was a mayor force behind getting that tb16 cam online, too. |
Beta Was this translation helpful? Give feedback.
-
|
Mipi camera on Lenovo t14s, Lenovo Yoga Slim7x, Dell XPS 9345, Dell Inspiron 14 Plus 7441 is supported on this branch from Bryan from Linaro. By extension, I added support for Asus Zenbook A14, but it still requires changes for camms from this branch since not everything made it upstream just yet. |
Beta Was this translation helpful? Give feedback.
-
|
Here are my cherry-pick commands for the camera stuff: |
Beta Was this translation helpful? Give feedback.
-
|
@jglathe I just wanted to say thank you! Yesterday, I finally removed all the Windows stuff from my machine, and now with your ISO and Kernel v5, my laptop is running Ubuntu directly from the hard drive without partition. Out of the box, the sound isn’t working, but I’ll try to fix that today. By the way, amazing work with the camera! 👍 If you need any help or testing, don’t hesitate to ask—I'm more than willing to help. Alejandro 
Here is a small guide if someone is interested on try it:
1. Download the system imageDownload the compressed image ( 2. Uncompress the
|
Beta Was this translation helpful? Give feedback.
-
|
Nice! Sound is there, but separate installation. Thank you for documenting your procedure. Please keep in mind that the source of new firmwares is still Windows, so... could be more work later. What are these different shades on the keyboard, key stickers? Looks like labeled for US ASCII. |
Beta Was this translation helpful? Give feedback.
-
|
Tip to avoid step 10: you can run the same fsck at step 8 after the dd. |
Beta Was this translation helpful? Give feedback.
-
|
@jglathe Thanks! Yes, I’m aware that removing the Windows partition could bring extra challenges down the line—especially with firmware updates—but I’m willing to take that tradeoff. For now, everything works great. I’ll continue working on the audio and will share any progress. Regarding the keyboard: good eye! The machine originally came with a QWERTZ layout, which I found really uncomfortable. I managed to find a ThinkBook 16 G6 keyboard with a US layout and swapped the physical keys. It’s not a perfect match—two keys are still mismapped—but with a simple remap, the typing experience is now smooth, and the backlight works properly too. |
Beta Was this translation helpful? Give feedback.
-
|
I tried |
Beta Was this translation helpful? Give feedback.
-
|
Hi @axu-trex, yes, I tried the version RC2 and also had some problems. For the moment, I have a direct installation on the hard drive, and I have been using the |
Beta Was this translation helpful? Give feedback.
-
|
Apropos, Debian's initramfs builder has given the following warning for all these kernels: I wonder if it's correct or mistaken? |
Beta Was this translation helpful? Give feedback.
-
|
This is correct, this firmware seems to be missing. I get the same message, although without negative effect since we're not this Apple device. |
Beta Was this translation helpful? Give feedback.
-
|
Hi @jglathe, I want to ask you if you have been able to enable the GPU? Is it the only part that is missing for me to use the laptop more frequently? |
Beta Was this translation helpful? Give feedback.
-
|
The short answer is yes. Current Mesa main will do. It's not in packaged versions yet. Will add a Wiki on how to - based on work from @alexVinarskis . |
Beta Was this translation helpful? Give feedback.
-
|
I've tried a few 6.16-rc kernels on my machine, but I keep getting a black screen on all of them. Here's a trace from my latest attempt: Do I need to have the correct Mesa installed for the kernel module to work? |
Beta Was this translation helpful? Give feedback.
-
|
I'm afraid this is tricky. You need the firmware files gen71500_gmu.bin and gen71500_sqe.fw, and you need the updated mesa. I have it running again on my TB16, but the way to get there seems a bit fuzzy. When the firmware files are not there, it will spam the dmesg with firmware load attempts. |
Beta Was this translation helpful? Give feedback.
-
|
Are those firmware files in the Windows install, or only available via some SDK with an NDA? It's not even trying to load those for me tho. |
Beta Was this translation helpful? Give feedback.
-
|
Those firmware files are on the kernel mailing list, and on newer dedicated install images. I'm a bit ambiguous re those, enabling of the GPU currently still destabilizes the snap functionality (which is sort of crucial for ubuntu). |
Beta Was this translation helpful? Give feedback.
-
|
Hi @jglathe, today I finally managed to enable the GPU, with your last image with 6.16, and the first glance looks everything fine. The battery lasts long, the CPU name is not recognized, but the GPU just works fine. Thanks a lot, I am happy that now I can use it for real things. On the other hand, I saw that the kernel 6.16 was released, but I browsed the file and I could not find the DTS for the TB16. Do you think that they approved it for kernel 6.17? |
Beta Was this translation helpful? Give feedback.
-
|
Thanks. That looks like my panel, and I guess also my touchpad. |
Beta Was this translation helpful? Give feedback.
-
|
Hi @jglathe, here is my info. From dmsg: Form libinput listdevices: |
Beta Was this translation helpful? Give feedback.
-
|
Oh interesting. The touchpad appears to be the same. |
Beta Was this translation helpful? Give feedback.
-
|
Hi @jglathe sorry for the late answer I was not around, here is the info: |
Beta Was this translation helpful? Give feedback.
-
|
Thank you. One with a nicer screen :) I'll add an entry in panel-edp.c for it. |
Beta Was this translation helpful? Give feedback.
-
|
Hi @jglathe, I see that the dts for this Thinkbook is on linux-next. I believe it will be available on kernel 6.18. I’m currently waiting to test another distribution on my laptop, Thanks a lot. |
Beta Was this translation helpful? Give feedback.
-
|
by the way I see the code now in the mainline 6.18-rc1 Now looks official in 6.18. |
Beta Was this translation helpful? Give feedback.
-
|
It‘s good that its finally upstream. Need to add a few more nodes:
So, still something to chart out 😀 |
Beta Was this translation helpful? Give feedback.
-
|
I installed 6.17.0-jg-1-qcom-x1e a few weeks back, and I've been able to use it. (All versions of 6.16 ended up in black screen for me.) It's still a bit unstable, and DHCP somehow always fails on shutdown, but on the positive side the battery remaining meter now works more often than not. It still doesn't detect if the battery is charging or not tho. HDMI output still doesn't work, but now I randomly get a bunch of realistic (albeit wrong) resolution options on the output. Here's some errors when the resolutions look more realistic: I also get a lot of cma spam on boot: |
Beta Was this translation helpful? Give feedback.
-
|
Here's some errors with 6.18: |
Beta Was this translation helpful? Give feedback.
-
|
Hmm there's multiple things: cma errors: add cma=128M in the boot command line, should fix it (IMO). Actually, there was only one small time window when I had a version without the required patch for X1-45 up (6.18). Didn't announce it either. 6.18.2-jg-2 should be quite good, including HDMI support with the "right" gpio. |
Beta Was this translation helpful? Give feedback.
-
|
Thanks for the CMA trick. As for 6.18.2-jg-2: still black screen. Currently the only versions I've found that don't blackscreen are 6.15.0-jg-5-qcom-x1e and 6.17.0-jg-1-qcom-x1e. And at least the latter also doesn't have the iris errors. I've tried to find kernel command line parameters that would stop the console from switching emulation modes so many times during boot, so that I could observe what's happening. Perhaps in my poking around, I've added something that is causing this problem? Or maybe I'm missing something essential? My current command line is |
Beta Was this translation helpful? Give feedback.
-
|
Another thought that occurred to me is that at some point you said I needed to disable certain firmwares and modules from loading. Perhaps something has changed? Which modules and firmwares should I allow to be loaded now? |
Beta Was this translation helpful? Give feedback.
-
|
All, actually. As long as you boot from nvme, no limitation. Do you have a dmesg for when it does come up? |
Beta Was this translation helpful? Give feedback.
-
|
Hi @jglathe, how are you? Happy 2026 I’ve picked up again the idea of using the ThinkBook as my main Linux laptop. I tried to find a distro already shipping kernel 6.18 with proper support, but so far without success (I guess I was a bit too optimistic thinking that once 6.18 landed, more distros would be readily usable). I also tried building my own setup based on Alpine Linux 3.23 with kernel 6.18. It does boot and the DT is correctly loaded, the firmware looks ok, but a few seconds after reaching the login prompt the screen goes black (something very similar reported by @axu-trex), so for now I’ve paused that attempt. If Alpine is of any interest, I’d be happy to share the project/ISO with you. I wanted to ask whether you’ve created any new ISO recently to retry with Ubuntu, and once again thank you very much for all your contributions to make this hardware usable. |
Beta Was this translation helpful? Give feedback.
-
|
Hi, @alejandroqh , thanks and also a happy 2026. I'm using the Thinkbook with 6.18.3, working mostly well. I didn't come up with a new ISO, but I played around with the 26.04 daily build from Canonical and this should actually do fine-ish. 6.18 kernels on x1/x1p suffered a bug with GPU, wouldn't come up with GUI. But this should be fixed...? It is on my tree. With that in it makes sense to try a new ISO as I only add some grub entries, actually. The original ISO content, like booting kernel or live fs stays untouched. So it should better work with x1/x1p. |
Beta Was this translation helpful? Give feedback.
-
|
I noticed some Device Tree related bullets in the changelogs of the recent linux-image packages in Debian/sid, so that side is looking brighter. However, I don't see any movement in the grub config generator for DTBs. I feel Debian is waiting for "upstream" to do something, and upstream is clueless... |
Beta Was this translation helpful? Give feedback.
-
|
Hi @jglathe, I restarted the process of installing Ubuntu and wanted to check if you’ve encountered a similar issue with the device tree not being installed correctly. What I did
ProblemAfter rebooting into the internal SSD:
Workaround (manual)
What I tried next
QuestionHave you encountered a similar issue where the device tree is not automatically installed or referenced after installation? Is there a known correct way to make GRUB persistently load the DTB on this setup? Thanks in advance for the help |
Beta Was this translation helpful? Give feedback.
-
|
In short: yes 😁 That's one if the reasons I looked into stubble, to get a kernel that can detect/assign the right dtb. Which is a lofty goal, especially with all the vendors being uncooperative in general. The best method to force a certain device tree I found was to use flash-kernel, I wrote about it a few times. You can create a /etc/flash-kernel/machine file and define the model string that is set for your box. And /etc/flash-kernel/db with the same model string and the dtb name you want used. This fixes grub shenanigans almost completely. When selecting that route, the -nostubble kernels are easier to handle. Otherwise you need |
Beta Was this translation helpful? Give feedback.
-
|
Now it works, It was funny. I restarted the complete process to check the HWIDs, but for some reason this time the devicetree was autoloader without intervention in the grub config as expected. Then I updated to 6.18.3 and was fine. I enabled the GPU smoothly, and the audio took me a while, but at the end everything is working again. Thanks a lot for the help. Now I will try to switch my stuff and let you know, but for the moment everything runs smoothly. |
Beta Was this translation helpful? Give feedback.
-
|
What are the -nostubble kernels? |
Beta Was this translation helpful? Give feedback.
-
|
"Old style" Kernel without the stubble boot loader which tries to match the dtb by hwid. You need this if the UEFI has problems with stubble, f.e. when booting in EL2. I need it on the Windows Dev Kit 2023 with EL2. Some people need it for the Vivobook S15 x1p4-variant. |
Beta Was this translation helpful? Give feedback.
-
|
Hi @jglathe, if I try with the stubble kernel, maybe this is why I need to put the "devicetree" manually because the stubble HWD is failing in my machine. How can I double-check the hwid? |
Beta Was this translation helpful? Give feedback.
-
|
|
Beta Was this translation helpful? Give feedback.
-
|
Thanks I checked it and all are the same, thanks for the info. |
Beta Was this translation helpful? Give feedback.
-
|
I tried 6.18.3 from Debian and it ended up in black screen, but I was happy to see that it still mostly worked. It also seems that the working kernels (6.15 and 6.17) are loading firmware for adreno, while the failing ones are not even trying. Here are the same lines as above from 6.17.0-jg-1: Why are firmware loads for qccdsp8380 and qcadsp8380 failing? |
Beta Was this translation helpful? Give feedback.
-
|
Ok this is my dmesg on Thinkbook 16 with 6.18.3: jglathe@tb16-jg:~$ sudo dmesg|grep -iE "drm|gpu|adreno" That sync_state message is not so good, but the thing works with GPU acceleration. 🤷♂️ I know that the mainline kernels miss a patch that you need for X1-45 to work, this gave me some headaches when I rebased my tree to 6.18. So... this looks like you might not use my kernel tree? |
Beta Was this translation helpful? Give feedback.
-
|
I'm downloading kernel packages from https://drive.google.com/drive/folders/1Lps5o3FXroAJFDiKj18vutJbC1uld49s which I assume are all your kernel tree. There's not a lot of documentation on what they are and there are clearly some completely different kernels there as well. I ran the your grep command above on vmlinuz-6.17.0-jg-1-qcom-x1e: And I ran it again on vmlinuz-6.18.3-jg-0-qcom-x1e: And for comparison, here's the same grep on the Debian packaged vmlinuz-6.18.3+deb14-rt-arm64: |
Beta Was this translation helpful? Give feedback.
-
|
I have fix 2 small issues that I want to share, hope this helps others with the same hardware! System Info
Fix 1: ath12k WiFi Stats Timeout ErrorsProblemath12k_pci 0004:01:00.0: failed to pull fw stats: -71 Solution: Update ath12k Firmware# Clone latest linux-firmware
cd /tmp
git clone --depth 1 https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git
# Backup and update
sudo cp -r /lib/firmware/ath12k /lib/firmware/ath12k.bak
sudo cp -r linux-firmware/ath12k/* /lib/firmware/ath12k/
# Reboot
sudo rebootResult Fixed! Firmware version after update: 0x110cffff (2025-06-25) Fix 2: Fingerprint Reader UnreliableProblem fpc_cmd_ssm_done error: Operation was cancelled Solution: Udev Rule Create udev rule to disable autosuspendReload and restartsudo udevadm control --reload-rules Re-enroll fingerprintfprintd-delete $USER Enable for sudo/terminal sudo pam-auth-update
Result Fixed! Fingerprint now works reliably. |
Beta Was this translation helpful? Give feedback.
-
|
I've been running Ubuntu 25.10 (Questing) on the Lenovo ThinkBook 16 Gen 7 QOY (Snapdragon X1P42100) as a daily driver. Here are some fixes and configs that I have been using. Hopefully this saves others some time. Hardware
Software
1. Audio (Speakers, Headphones, Mic)Audio is completely dead out of the box. The root cause is a firmware path mismatch: the kernel identifies this machine as The fix requires machine-specific ADSP/CDSP firmware, a custom audioreach topology, and ALSA UCM configuration. All credit to jglathe for figuring this out. Key gotchas:
Copy ThinkBook 16-specific firmware (ADSP/CDSP) Using sudo mkdir -p /lib/firmware/updates/qcom/x1p42100/LENOVO/21NH
SRC=firmware-x1p42100-21NH # or wherever the machine-specific files are
# ADSP firmware (note: source file has .disabled suffix — rename to .mbn)
sudo cp $SRC/qcadsp8380.mbn.disabled /lib/firmware/updates/qcom/x1p42100/LENOVO/21NH/qcadsp8380.mbn
sudo cp $SRC/adsp_dtbs.elf /lib/firmware/updates/qcom/x1p42100/LENOVO/21NH/
sudo cp $SRC/adspr.jsn /lib/firmware/updates/qcom/x1p42100/LENOVO/21NH/
sudo cp $SRC/adsps.jsn /lib/firmware/updates/qcom/x1p42100/LENOVO/21NH/
sudo cp $SRC/adspua.jsn /lib/firmware/updates/qcom/x1p42100/LENOVO/21NH/
# CDSP firmware
sudo cp $SRC/qccdsp8380.mbn /lib/firmware/updates/qcom/x1p42100/LENOVO/21NH/
sudo cp $SRC/cdsp_dtbs.elf /lib/firmware/updates/qcom/x1p42100/LENOVO/21NH/
sudo cp $SRC/cdspr.jsn /lib/firmware/updates/qcom/x1p42100/LENOVO/21NH/
# Battery manager
sudo cp $SRC/battmgr.jsn /lib/firmware/updates/qcom/x1p42100/LENOVO/21NH/Steps summary:
2. Fingerprint Reader — identity_type MismatchThe fingertprint works out of the box but after some boots start to missmatch,then I notice that enrollment works fine, but verification always fails with "verify-no-match" after some time. The FPC MOC driver in libfprint expects Fix: Patch # Build from patched source
apt source libfprint-2-2
cd libfprint-1.94.9+tod1
patch -p1 < fpc-identity-type-workaround.patch
dpkg-buildpackage -us -uc -b
sudo dpkg -i ../libfprint-2-2_*.deb ../libfprint-2-tod1_*.deb
sudo systemctl restart fprintdYou can download the files already compiled here 3. GPU Firmware in InitramfsAt early boot, the kernel fails to load Fix: Create #!/bin/sh
PREREQ=""
prereqs() { echo "$PREREQ"; }
case "$1" in prereqs) prereqs; exit 0;; esac
. /usr/share/initramfs-tools/hook-functions
for fw in qcom/gen71500_sqe.fw qcom/gen71500_gmu.bin; do
for ext in "" ".zst"; do
src="/lib/firmware/${fw}${ext}"
if [ -e "$src" ]; then
dest="${DESTDIR}/usr/lib/firmware/${fw}${ext}"
mkdir -p "$(dirname "$dest")"
cp "$src" "$dest"
break
fi
done
doneThen: Note: Ubuntu's Performance OptimizationsThese are optional but made a noticeable difference on this hardware. Boot Time: 34s to ~21s
Memory & I/OReplace the swap file with zram (compressed swap in RAM): sudo apt install zram-tools
# Default config: PERCENT=50, ALGO=lz4 → ~15 GB compressed swap on 32 GB
sudo swapoff /swap.img && sudo rm /swap.img
# Comment out /swap.img in /etc/fstabSysctl tuning ( vm.swappiness=10
vm.vfs_cache_pressure=50
vm.dirty_ratio=15
vm.dirty_background_ratio=5
vm.dirty_writeback_centisecs=1500
net.core.default_qdisc=fq
net.ipv4.tcp_congestion_control=bbr
net.ipv4.tcp_fastopen=3
fs.inotify.max_user_watches=524288Add Power
Running x86-64 Apps via Box64Some apps don't have aarch64 builds. Box64 with the Oryon DynaRec ( Currently running via Box64:
Setup: git clone https://github.com/ptitSeb/box64 && cd box64
mkdir build && cd build
cmake .. -DSDORYON1=1 -DCMAKE_BUILD_TYPE=RelWithDebInfo
make -j8 && sudo make install
sudo systemctl restart systemd-binfmt # enables transparent x86-64 executionPower LED Investigation (Unsolved)The power LED stays solid during Linux suspend instead of pulsing like in Windows. After extensive investigation:
Next step: Capture the EC I2C traffic in Windows during sleep/wake transitions to reverse-engineer the command protocol. If anyone has done this or written a ThinkBook EC driver, please reach out. |
Beta Was this translation helpful? Give feedback.
-
|
I found 2 thinks that maybe are could be the problem:
Have you tried ubuntu-concept/x1e Kernel? |
Beta Was this translation helpful? Give feedback.
-
|
The log says some scmi failure, cascading to gpu-smmu and lots of clocks. Race condition in scmi, IMO I've seen this, too, but maybe once or twice. |
Beta Was this translation helpful? Give feedback.
-
|
Unfortunately, no clue on how to handle this yet. |
Beta Was this translation helpful? Give feedback.
-
|
@alejandroqh I checked out of curiosity with my FP reader on the TB16 as it degraded as well (and on the X13s, too): Somebody has fixed this, more or less back to normal as before.
|
Beta Was this translation helpful? Give feedback.
-
|
@jglathe surprising that also is happening to X13s, Now I have 2 week with this patch and has been working, but if there is a deep solution I will try it thanks for let me know. This days I will test the suspension with this kernel, thanks have a grate week. |
Beta Was this translation helpful? Give feedback.
-
|
Hey @alejandroqh , I liked your initramfs hook for firmware. I found some useful hook functions in Debian/sid (which I'm running) and adapted your script a bit. My script looks like this now: The last file |
Beta Was this translation helpful? Give feedback.
-
|
Thanks for the feedback, the file tplg is an audio topology, follow the guide of @jglathe to generate it, also I have a version here how to prepare the aduio for the machine. |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
Hi jg!
I bought the Snapdragon laptop mentioned in the title this January after having read a lot of good stuff about Snapdragon X Elite. Turns out it's Snapdragon X Plus (x1p42100) instead, and I was misled by the reseller, but I've noticed you've been furiously working on support for this laptop here too.
I'd like to test drive your work. Have you made
.debpackages of the kernel and the required.dtbavailable somewhere? (I did try to look through your googledrive folders mentioned elsewhere, but couldn't find this.)And as a followup question: Is it possible to install these into your modified desktop images while chrooted on an x86 machine? Without qemu?
Beta Was this translation helpful? Give feedback.
All reactions