Laptop broke down after installing kernel 3.0

Question

My laptop died little after installing kernel 3.0 (the "linux" package). Details:
- Laptop model: ASUS K50AB-SX029L, it was exactly 2 years minus 10 days old (based on the purchase date); specifictions here.
- using cpufreq frequency scaling (ondemand) and the catalyst drivers for video - if this is useful.
- the laptop used to get heated  when the ventilation became obstructed by dust; however, the sensors worked, it used to instantly shut down when it was intensely used and too hot in the room.
- after I installed linux-3.0, I noticed some error message early at boot time, something about "soft restart" or "soft reboot" and something about "device" (not configured, maybe) - I don't remember well.
How it happened: I installed kernel 3.0 yesterday, rebooted, then built/installed the AUR ATI drivers (catalyst-utils, catalyst, lib32-catalyst-utils), rebooted again and then I left it building an AUR package. When I came back to it, the desktop was frozen, I forced a shutdown using the power button and since then, every time I turn it on, it gets powered, the fan is spinning and the USB external drive starts, but nothing else happen - no display, no boot, no spinning HDD. I removed its CMOS battery, but the same.
Now, since too many possible factors coincide - I installed kernel 3.0 for the first time, laptop was 2 y.o., it was used somehow intensely along the time, its warranty was just expiring - it is of concern to me what happened, if possible to find out. I mention that it was 2 years old because another laptop left me down before, after the exactly same period of using Linux (Arch for like 4 months on that one), different model and different brand (Acer). What is of interest:
- someone else's laptop breaking down coincided with updating to kernel 3.0?
- someone else's laptop broke down after 2 years or similar of using Linux/Arch daily? Any difference in reliability compared to using other OSs?
- any insight in the reliability/durability of the notebook hardware in general (things that may not depend on the OS)? I mean if it was supposed to die anyway after 2 years of usage - which sounds too short to me, btw - maybe it was just a big coincidence that it happened right after linux 3.0 was installed. But there is a need for some arguments supporting this hypothesis, while searching the web comes with a lot of irrelevant or contradictory information.
- if the case, how could the malfunctioning/misconfiguration of the OSs in general, Linux in particular, damage the hardware? I think overheating is out of question when the sensors work, but I'm open to information on this subject.
- what data (logs, etc) could be relevant and how should I analyze it, after retrieving it off the dead laptop HDD?

Answer 1

MoonSwan wrote:Funny thing, my netbook's battery (an asus eeepc) pretty much died at exactly 1 year and 362 days old.  Coincidence?  I don't know.
Inxsible wrote:In my opinion, if you buy a laptop from the "Home and Home Office" section --- no matter what company, you most often get shitty computers that last all of 1 -2 years if you are lucky. My HP dv9000t mobo got fried exact 2 weeks after the warranty expired. Then fixing that, the wireless broke down, then the lcd got dead pixels. I think this is not a coincidence... Lenovo Thinkpad T series is what I went with after I gave up on the HP.
+1 on both. I had a kernel panic in my head this AM, too, but if you ask me it sounds suspiciously like a planned 2-year lifecycle on the products. Sack of middle fingers to the OEM retail products divisions.
Last night I had started compiling a kernel and was working on my thesis (which has nothing to do with computers besides using a word processor). Did a $ sudo pm-suspend and went to sleep when the machine finished compiling. This morning I'm at my desk with a multimeter and trying to find a local place that carries SATA to USB assembly.
BIOS won't come up, screen won't turn on, I'm definitely not getting to GRUB--pressing enter repeatedly and waiting does not lead to the hard disk thrashing at all. Yanked the battery, tested the adapter, cycled the power switch b/c I can't find the CMOS jumper and there's no repair manual for this thing. I realize I did leave a single USB fan and a USB light (turned off) plugged in during the suspend, but am having a hard time thinking that drawing power at USB spec (5V/100mA max) from USB could have done this especially with the A/C power source available. I'm not familiar with the actual suspend process, but a 5V rail is a 5V rail when it's got power, right?
I'm not going to start yanking individual components just yet, I actually want to get sh!t done today. The machine in question has a manufacture date of January 2009 stamped on the bottom.
Mfg Date: 2009/01/07
So two years, seven months and about three weeks of life. Even with a backup of mission-critical files from last week, this sucks--losing 60 hours of work is FTL. Fingers crossed that whatever zapped it didn't get to the disk. I'm still kinda shocked about all this. Get it? Shocked? Har har X-(
I'm not going to blame the 3.0 kernel though I was tempted. "But I didn't even install the new kernel... how can this be"? I thought maybe I blew up my processor or something by compiling, running libreoffice, VLC and firefox all at once... hardly. Bad timing on a new kernel, maybe, or I managed to piss off some DelHewlitSUSway god, or something. I've had battery/power issues on this machine almost since day one, but never any indication of impending doom.
I agree completely with the advice to get laptops from a "small business" or institutionally-oriented line, like Lenovo/IBM's Thinkpads, and even thought about Mac this morning, but don't have a trust fund. I'm writing this on a second-hand Thinkpad R series almost six years old. The LCD screen and/or cable are damaged, the disk spins at an underwhelming 4300 RPM or whatever and I have to keep the fan at max levels all the time but at least it f-cking boots up.
Sorry if this is a thread hijack, but if you can't get to BIOS you can't get to BIOS. After confirming it's not a video issue the only options are to yank all parts not necessary to booting up, clear the CMOS by finding the jumper or battery or trying to cycle power via the switch (for, like, 5 minutes), and reinstall parts one-by-one until replicating the issue. If you can't get that far, the mobo, BIOS or processor is probably nuked.