Meaningful error messages

morbiuswilters

@Severity One said:

So what were you doing at the time that computers were too weak to do much of value? Cleaning toilets at a railway station?

What? How did you arrive at this conclusion? I used computers then (and even enjoyed myself) but that doesn't make me nostalgic for those times.

@Severity One said:

The point of putting device drivers (and libraries) on the expansion card itself, like was the case with Amiga AutoConfig devices, was that the device driver was indeed part of the hardware: it got installed automagically as soon as you plugged in the card into your Amiga. True plug-and-play.

Yes, I understand why it was done. I also understand that although it may sound nice it's a terrible idea for the reasons you've cited. So why are we even talking about it?

morbiuswilters

@roelforg said:

they bought the pc before Dell sold crap hw

Dell sells crap hardware? That's news to me.

@roelforg said:

Besides, when windows has boot issues, you only have stopcodes with terse/cryptic descriptions of the problem to go by.

Right, because Linux kernel panics are so very useful.

serguey123

@morbiuswilters said:

I used computers then (and even enjoyed myself)

Porn?

@morbiuswilters said:

that doesn't make me nostalgic for those times.

Ohhh, porn!

roelforg

@morbiuswilters said:

@roelforg said:
they bought the pc before Dell sold crap hw

Dell sells crap hardware? That's news to me.

@roelforg said:
Besides, when windows has boot issues, you only have stopcodes with terse/cryptic descriptions of the problem to go by.

Right, because Linux kernel panics are so very useful.

Was thinking about hp consumer hw, oops... (i still think vendors prefer stylish thin and flimsy case over an solid, can't break case)

And yes, kernelpanics can be usefull!

Proof: One install disk kept crashing when it tried to boot the installer, the stacktrace resulting from the panic pointed me to faulty mem. (the initram in mem didn't match the one on cd, the cd was fine so it had to be mem. Replaced ram and it ran fine)

morbiuswilters

@roelforg said:

And yes, kernelpanics can be usefull!

Yes, about as useful as BSODs...

@roelforg said:

Proof: One install disk kept crashing when it tried to boot the installer, the stacktrace resulting from the panic pointed me to faulty mem. (the initram in mem didn't match the one on cd, the cd was fine so it had to be mem. Replaced ram and it ran fine)

I can barely follow what you are saying here.. how does the stack trace point to faulty mem? How did you compare the initram in memory to the one on disk? Especially since the kernel died?

roelforg

@morbiuswilters said:

@roelforg said:
And yes, kernelpanics can be usefull!

Yes, about as useful as BSODs...

@roelforg said:
Proof: One install disk kept crashing when it tried to boot the installer, the stacktrace resulting from the panic pointed me to faulty mem. (the initram in mem didn't match the one on cd, the cd was fine so it had to be mem. Replaced ram and it ran fine)

I can barely follow what you are saying here.. how does the stack trace point to faulty mem? How did you compare the initram in memory to the one on disk? Especially since the kernel died?

Let me explain my thoughtrain from when i needed to fix the mem:(it was yesterday)

When going through the trace i recorgnised something:

in the middle it mentioned several funcs from init/initramfs.c (http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git;a=blob;f=init/initramfs.c;h=8216c303b0821b15f1a353a2fea7af84abb8f4bf;hb=HEAD).

having build a few linux initramfs from scratch (to do some partititioning (doing that to several identical pc's was worth the 3houres free time)) i examined that source file once to see why one of my initramfs' wouldn't boot.

Then in a "gut"-moment i swapped out the ram and problem solved! (a memtest showed the ram was >75% broken.

Turned out that the memory manager doesn't like broken ram (initramfs->decompressing rootfs->lot's of kmalloc'ing and free'ing->memory manager assigns bad block->decompress yields broken rootfs->failure to run corrupted init->long stacktrace).

blakeyrat

@roelforg said:

Let me explain my thoughtrain from when i needed to fix the mem:(it was yesterday)

When going through the trace i recorgnised something:

in the middle it mentioned several funcs from init/initramfs.c (http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git;a=blob;f=init/initramfs.c;h=8216c303b0821b15f1a353a2fea7af84abb8f4bf;hb=HEAD).

having build a few linux initramfs from scratch (to do some partititioning (doing that to several identical pc's was worth the 3houres free time)) i examined that source file once to see why one of my initramfs' wouldn't boot.

Then in a "gut"-moment i swapped out the ram and problem solved! (a memtest showed the ram was >75% broken.

Turned out that the memory manager doesn't like broken ram (initramfs->decompressing rootfs->lot's of kmalloc'ing and free'ing->memory manager assigns bad block->decompress yields broken rootfs->failure to run corrupted init->long stacktrace).

Jesus.

morbiuswilters

@blakeyrat said:

Jesus.

lol

Severity_One

@TheCPUWizard said:

Just a FWIW: I had a number of the Seagate 1.5TB disks, and had a number of probems, many of which were very subtle. Since many of my systems are near-realtime systems, I am very sensitive to "freeze" even when it is below the threshold of being noticable at the user level. On a whim, I replaced the 1.5TB drives in one system, and no longer experienced that problem. Since then I have removed all of them from my systems, andmany of the issues have al gone away. Nothing conclusive (i.e. a disk test did not fail), but there is other antedoctal evidenve on the internet to support my "feeling" that the drives were indeed the root cause.

Well, waddayaknow. Went to the Seagate site yesterday, located a firmware upgrade for the 1.5 TB disc, installed it, and the whole system feels a lot snappier. I very much suspect that the freezing problem is a thing of the past as well.

However, what with a Barracuda dying on me before, this problem, and the Seagate diagnostics utilities being somewhat crap, I doubt I'll buy one of their products ever again.

TheCPUWizard

@Severity One said:

@TheCPUWizard said:
Just a FWIW: I had a number of the Seagate 1.5TB disks, and had a number of probems, many of which were very subtle. Since many of my systems are near-realtime systems, I am very sensitive to "freeze" even when it is below the threshold of being noticable at the user level. On a whim, I replaced the 1.5TB drives in one system, and no longer experienced that problem. Since then I have removed all of them from my systems, andmany of the issues have al gone away. Nothing conclusive (i.e. a disk test did not fail), but there is other antedoctal evidenve on the internet to support my "feeling" that the drives were indeed the root cause.

Well, waddayaknow. Went to the Seagate site yesterday, located a firmware upgrade for the 1.5 TB disc, installed it, and the whole system feels a lot snappier. I very much suspect that the freezing problem is a thing of the past as well.
However, what with a Barracuda dying on me before, this problem, and the Seagate diagnostics utilities being somewhat crap, I doubt I'll buy one of their products ever again.

Any chance you have a date on that??? [I am too swamped right now to try to find it...but definately interested..as some of the drives stil in use (but not in my work systems)]

sh_code

@pauly said:

@boomzilla said:
@SEMI-HYBRID code said:
it doesn't when this is all you know, but it actually does, because most of the "my program doesn't work after OS upgrade" is caused by the program exploiting a bug/undocumented ( = unofficial, internal, not to be used by public) feature to work, and that feature/bug getting changed/corrected, so most of these are really program's faults, not operating system's.

And you still haven't explained why it would be illogical to dislike the OS for this. Note that I haven't disagreed with anything you've said.
He should be blaming the games for not being future-proof and being written with Vista compatiblity in mind.

it would be completely sufficient for the game/program be present-proof and not use features that were never supposed to be used by any third party in the first place.

Severity_One

@TheCPUWizard said:

@Severity One said:
Well, waddayaknow. Went to the Seagate site yesterday, located a firmware upgrade for the 1.5 TB disc, installed it, and the whole system feels a lot snappier. I very much suspect that the freezing problem is a thing of the past as well.
However, what with a Barracuda dying on me before, this problem, and the Seagate diagnostics utilities being somewhat crap, I doubt I'll buy one of their products ever again.
Any chance you have a date on that??? [I am too swamped right now to try to find it...but definately interested..as some of the drives stil in use (but not in my work systems)]

Um... that must be a bit of idiom going right past me. What kind of date are you looking for?