Catastrophic Power Off

SamK · posted: 2015-07-07 11:11

This laptop has run antiX-13.2 (kernel 3.7.10 and above) since November 2013 without experiencing a catastrophic shutdown.

Immediately after installing antiX-15 to disk, the uncontrolled shutdowns began.

Cold components
Booting with the CPU and hard disk temperatures cold (e.g. having been powered down overnight) the system fan operates until the kernel starts to load. At this point it switches off. The boot continues and the desktop is displayed as expected. The internal temperature machine builds steadily over a period of time, but the fan is not restarted. Eventually, the system experiences a catastrophic shutdown due to overheating.

Hot components
Booting immediately following a catastrophic shutdown, the system fan operates throughout the entire boot process. The boot completes normally and the desktop is displayed as expected. The system fan manages the temperature of the machine and a shutdown does not occur even during extended periods of use.

Even though it was not needed in 13.2, the boot parameter acpi=force has been tried in 15 but the above symptoms persist.

cat /proc/cmdline

Code: Select all

BOOT_IMAGE=/boot/vmlinuz-4.0.5-antix.1-486-smp root=UUID=blurb... ro acpi=force vga=791 quiet

inxi -c 0 -F

Code: Select all

System:    Host: clevo1 Kernel: 4.0.5-antix.1-486-smp i686 (32 bit) Desktop: JWM sn-579
           Distro: antiX-15-V_386-full Killah P 30 June 2015
Machine:   Mobo: KAPOK model: SiS 630S Bios: SystemSoft v: Version 1.00 date: 10/22/97
CPU:       Single core Intel Celeron (-UP-) cache: 256 KB speed: 1295 MHz (max)
Graphics:  Card: Silicon Integrated Systems [SiS] 630/730 PCI/AGP VGA Display Adapter
           Display Server: X.Org 1.16.4 driver: N/A Resolution: 1024x768@60.00hz
           GLX Renderer: Gallium 0.4 on llvmpipe (LLVM 3.5, 128 bits) GLX Version: 3.0 Mesa 10.3.2
Audio:     Card Silicon Integrated Systems [SiS] SiS PCI Audio Accelerator driver: snd_trident
           Sound: Advanced Linux Sound Architecture v: k4.0.5-antix.1-486-smp
Network:   Card: Silicon Integrated Systems [SiS] SiS900 PCI Fast Ethernet driver: sis900
           IF: eth0 state: up speed: 100 Mbps duplex: full mac: 00:90:f5:13:a7:df
Drives:    HDD Total Size: 20.0GB (14.6% used) ID-1: /dev/sda model: FUJITSU_MHR2020A size: 20.0GB
Partition: ID-1: / size: 6.2G used: 2.6G (45%) fs: ext4 dev: /dev/sda1
           ID-2: /home size: 12G used: 101M (1%) fs: ext4 dev: /dev/sda2
           ID-3: swap-1 size: 0.05GB used: 0.00GB (0%) fs: swap dev: /dev/zram0
Sensors:   None detected - is lm-sensors installed and configured?
Info:      Processes: 109 Uptime: 3:35 Memory: 91.8/308.0MB Client: Shell (bash) inxi: 2.2.25

rokytnji · posted: 2015-07-07 12:38

Cold components
Booting with the CPU and hard disk temperatures cold (e.g. having been powered down overnight) the system fan operates until the kernel starts to load. At this point it switches off. The boot continues and the desktop is displayed as expected. The internal temperature machine builds steadily over a period of time, but the fan is not restarted. Eventually, the system experiences a catastrophic shutdown due to overheating.

Just wondering if the compressed old dmesg logs in /var/log may provide info on why.

========= SCRAPER REMOVED AN EMBEDDED LINK HERE ===========
url was:"http://i44.tinypic.com/1t3fk7.jpg"
linktext was:"My Kapok."
====================================

Still runs also. But. It is a dust collector now on the shelf.
Good luck with it.

rokytnji · posted: 2015-07-07 12:41

When in doubt. I roll with a older Debian Kernel using smxi on older gear also.

SamK · posted: 2015-07-08 08:05

rokytnji wrote:My Kapok...Still runs also. But. It is a dust collector now on the shelf.
Good luck with it.

Given light tasks using lightweight apps the one here still performs remarkably well. It can even play video via Streamlight at less than 50% CPU usage, which is not too shabby for a machine of this vintage. There is still a useful lifespan for this system.

I was reporting because something has changed in the shipped antiX-15 (kernel?) that adversely affects this machine compared to antiX-13.2. It might be worth investigating in view of antiX supporting older kit.

My next step is to install the Debian kernel for older 486 systems and possibly try and find an older antiX kernel.

SamK · posted: 2015-07-10 17:43

Installed the 13.2 kernel 3.7.10-antix.7-486 in antiX-15 (thanks Dave)

System fan now operates as expected.
Seems to strongly implicate the shipped antix-15 kernel as the cause.

skidoo · posted: 2015-07-10 19:54

reiterating something that I've previously posted elsewhere:

beyond 3.7, several/various linux kernels I've tested (not specific to the antiX kernel builds) yield a poor result on certain hardware.

rokytnji · posted: 2015-07-10 23:35

This old IBM T23 with a Pentium 3 floats the new kernel OK

Code: Select all

harry@fud:~
$ acpi -t
Thermal 0: ok, 64.0 degrees C
harry@fud:~
$ uname -a
Linux fud 4.0.5-antix.1-486-smp #1 SMP Mon Jun 8 00:21:57 EEST 2015 i686 GNU/Linux
harry@fud:~
$ free
             total       used       free     shared    buffers     cached
Mem:       1028984     895192     133792       2332      49052     524792
-/+ buffers/cache:     321348     707636
Swap:      2046972       2576    2044396

rokytnji · posted: 2015-07-10 23:37

My graphics is weird like your Sis also with these newer kernels

Code: Select all

harry@fud:~
$ inxi -G
Graphics:  Card: S3 Graphics SuperSavage IX/C SDR
           Display Server: X.Org 1.16.4 driver: vesa
           Resolution: 1024x768@61.00hz
           GLX Renderer: Gallium 0.4 on llvmpipe (LLVM 3.5, 128 bits)
           GLX Version: 3.0 Mesa 10.3.2

SamK · posted: 2015-07-13 08:37

For the sake of completeness...

The Debian 8 kernel aimed at older kit does not control the system fan so catastrophic shutdown due to overheating is produced.

Kernels tested which fail to control the system fan and temperature

antiX 4.0.5-antix.1-486-smp
Debian 3.16.0-4-586

Kernel tested which succeeds in controlling the system fan and temperature

antiX 3.7.10-antix.7-486-smp

Perhaps there is an opportunity for antiX to either incorporate the configuration of the earlier kernel into the the newer versions (if possible), or produce a kernel specifically aimed at older kit.

Related topic
error-building-kernel-3-7-10-antix-7-486-in-antix-15-t5774.html

jdmeaux1952 · posted: 2015-07-13 10:53

SamK wrote: Perhaps there is an opportunity for antiX to either incorporate the configuration of the earlier kernel into the the newer versions (if possible), or produce a kernel specifically aimed at older kit.

Now that's an idea worth trying to accomsplish.

anticapitalista · posted: 2015-09-28 20:55

It might have just happened!