Goodbye, Éowyn

Goodbye

The last few months have been rough on Éowyn (Dell XPS 15 9570). I've had her since June 2019, right up till February 2022. It hasn't even been 3 years. That's the computer where I first messed around with my setups - I customized and debugged them until I was happy with where they were. I got to try out different window managers and Linux distros, and this was the first computer that I took apart to replace parts.

What happened?

Rather than divide this section into "what did I try that didn't work" and a solutions section, I'm choosing to write this more as a story.

December 11, 2021

After keeping my laptop running for about two weeks (not doing particularly CPU-intensive tasks, just editing, PDF-viewing, and music), on December 11, 2021, I decided to reboot it. It wouldn't wake up.

I tried some combination of

Most of the time, the keyboard would light up, but I'd get nothing more. A couple times, I got the Dell logo, but it'd immediately turn off, not even showing the boot messages. I can't remember which combination of things got the logo to appear. I'd often get an amber blinking pattern of 3 amber flashes followed by 5 white flashes, indicating a SBIOS corruption. I didn't have another computer with which I could burn a BIOS, but after a bit of waiting and hitting the power button, the issue somehow resolved itself. I was warned to update my BIOS soon, but I was too lazy to do that.

January 23, 2022

Between Dec 11 and Jan 23, reboots would take a few attempts. I tried to keep these to a minimum, since I was scared that each time would be the last. Sometimes I'd get nothing, but typically within 5-10 minutes of trying, I'd be able to boot up and log in, and everything was fine-ish. On Jan 23, after almost an hour of trying, I couldn't get it to boot. The keyboard would occasionally light up, but I wasn't even able to reach the Dell logo. I maybe got the Dell logo twice, but I wasn't able to log in at all. Finally, disconnecting the battery let me log in. This time, I got the 2, 1 blinking pattern that indicates CPU failure. However, I'd also occasionally get the 3, 5 pattern, which indicated I had BIOS issues.

I wasn't able to re-flash the BIOS immediately, since that needed both battery and AC power. I swapped the battery and reflashed the BIOS. It would let me log in provided I wasn't connected to AC power. I tried disabling the adapter power warning, but that didn't seem to make much of an impact. I was able to boot after shutdowns and reboots, and everything seemed to be okay for a while.

February 20, 2022

The WiFi stopped working. I can't remember what error message iwd gave me, but it was one I'd never seen before. iwctl wouldn't even show me its menu. I tried running sudo systemctl restart iwd, which is what I typically do when I run into any sort of issue with iwd. Restarting iwd normally takes around 5 seconds, but this time it froze up. I Ctrl + C-d that, and tried to restart. That also froze up, so I hit the power button instead. After it powered off (this took a bit, since it had to kill iwd properly and stuff), it wouldn't wake up.

Based on the 2, 1 blinking pattern, it looks to be a CPU failure or a board corruption, both of which are difficult to fix. I suspected this had something to do with the hardware components related to WiFi modules, since the iwd error was one I hadn't seen before. I haven't gotten a successful boot up in days. Either way, I needed a more reliable computer (Zoom classes from a phone are a pain), so I got Mithrandir, a MacBook Pro. I will probably have a write up about that set up at some point. I might wind up making that a dual boot with Arch; I miss Arch. Alternatively, I might keep this is as a reliable back up, and get another computer for my main stuff. I wound up choosing the latter option; that computer is now Faramir.

Current status

March 3, 2022

It was a bit too hard to say goodbye, so I went ahead and tried to boot again. I was finally able to reach the Dell page, but I wasn't able to boot up properly. It would give me an error about not being able to find the root partition, and dropped me into an emergency shell (rootfs). There was no sda or disk related device in the /dev directory, so it wasn't detecting the hard drive. This is similar to what happened in February, where it wasn't recognizing peripherals. I tried rebooting normally a few times, which did nothing (the exit code was 0, and I was able to run commands in the shell perfectly normally). reboot -f, however, worked. This was oddly reminiscent of the previous issue I had with rebooting, where it wasn't correctly processing/executing the command. On rebooting, it booted me back into the emergency shell. I tried taking a look at the BIOS, but there wasn't anything there of use. I haven't even gotten the Dell logo since.

If it really is a chip failure, there's not much I can do at this stage. Goodbye; I loved you.