search

QubesOS / qubes-issues

The Qubes OS Project issue tracker
https://www.qubes-os.org/doc/issue-tracking/
541 stars 48 forks source link

RTL8821CE WIFI Controller crashes sys-net #8126

Open hadmut opened 1 year ago

hadmut commented 1 year ago

Hi, some time ago I have installed QubesOS 4.1 on a HP notebook, but didn't do much with that particular machine. Now I went through upgrading the machine, dom0, templates, VMs and downloaded the fedora-37 image. After setting sys-net to use fedora-37 instead of fedora-34 (not sure, whether this or upgrading dom0 was the reason), sys-net didn't run anymore. When starting, it ran for some seconds, and then shut itself down again, error message that the manager can't connect to the VM.

For unknown (i.e. not obvious) reasons, sys-net repeatedly crashed some seconds after booting, while the underlying fedora-37 template itself boots well. The console log did not reveal any reason. Even switching back to fedora-34 did not help. But when using debian-11 instead, it showed that there is a kernel crash right after loading the rtl88 driver. A Null pointer exception.

After removing the RTL8821CE Wifi Controller PCIe device from the device list, sys-net worked normally again with both, debian-11 and fedora-37 as underlying templates (but of course, ethernet only, no Wifi).

So there is currently a problem with this Wifi chip causing sys-net to not boot, and thus not having network access to download patches.

regards

pasajeros commented 1 year ago

I have this problem as well on my Lenovo IdeaPad laptop.

ProGamer2508-0 commented 4 months ago

HI. I too have the same problem, but this issue is a copy of another older issue which also concerns 4.2 (basically the same hardware and the same poorly made driver that doesn't work. It's not a hardware problem in truth, but of the driver that QubesOS still keeps in its database)

ProGamer2508-0 commented 4 months ago

8472

ProGamer2508-0 commented 4 months ago

before creating a new post on this issue you should always check if there is a post similar to the problem encountered

ProGamer2508-0 commented 4 months ago

anyway... this problem hasn't been solved for years. see the post I mentioned to see how much is ignored

ProGamer2508-0 commented 4 months ago

see the post I mentioned to see how much is ignored

from the dev I mean

ProGamer2508-0 commented 4 months ago

in the post there are also problems that arise from this problem (for example that at the time of installation, sys-usb is not created... and that it cannot be created in normal ways in post-installation)

andrewdavidwong commented 4 months ago

@ProGamer2508-0: Please refrain from making multiple separate comments on an issue in quick succession. Instead, you can use the edit function to add to a comment that was made recently.

As for whether this issue is a duplicate of https://github.com/QubesOS/qubes-issues/issues/8472, I don't know. They mention the same networking component model, but https://github.com/QubesOS/qubes-issues/issues/8472 says sys-net "does not start," whereas this says sys-net "crashes," so I'm not sure.

ProGamer2508-0 commented 4 months ago

@ProGamer2508-0: Please refrain from making multiple separate comments on an issue in quick succession. Instead, you can use the edit function to add to a comment that was made recently.

Sorry. I got used to writing messages like this. from this moment on I will write in the issues in the way you said

@andrewdavidwong, crash and start are the same thing, but it's better to say crash as an error occurs when it tries to start. it doesn't start, but this is due to a crash

see the issue I mentioned before. preferably focus on the photos sent by me and a user with whom I discussed

sorry again: crash and start are not the same thing, but in this issue it is

(I only realized now that it is from 2023 and not 2024 😑)

I'm sorry if I bothered you. I had seen this post on the first page of the issue (although I don't know why) and I thought it was recent... but the problem remains the same: this hardware is literally the worst of the worst for sys-net, it prevents the creation of sys-usb, many posts on the community forum have been created to try to solve this problem, now the repository that once fixed this problem has practically made it more serious (now it destroys the packages for the daemon with dom0) and, As the cherry on the cake (I don't like cherries but that's an irrelevant statement), it has been going on for years