On Wed, 30 Jan 2008 17:43:44 +0100, Arun Dev wrote:

> Hi all
>
> I just tried my luck on an old P-II, 233 MHz, 64 MB laptop.
>
> Booting huge.s I could finish the installation. But I can't start
> the installation. It gives
>
> no file system could mount root, tried:
> Kernel panic - not syncing: VFS: Unable to mount root fs on
> unknown-block(8,2)
>
> Am I correct in assuming that the initrd is too big for the 64 MB?
>
> The root device is accoding to lilo.conf /dev/hda2
>
> Kernel was 2.6.22 "generic".
>
> Arun
>
Caveat: I do not know the specific system requiremens for Slackware 12.0.
My guess is that 64M is enough to be able to /*run*/ Slack 12 (probably
via command line only). However, the Slackware /*setup*/ environment may
require more memory than that. If you could get around that catch-22, then
you could use that computer with Slack 12.

I have used various methods to setup Slack 12, including cloning a preset
image with the baseline set of packages. If you elect to go this route,
be aware that Slackware's recent kernels require an initrd and that can be
a stumbling block when setting up using a non-traditional method.

It might just be the easiest course to break down and use the "screw
driver" method. You could swapout the disk to a computer with the
resources necessary, then swap it back into the older computer
when done.

--
Douglas Mayne

Arun Dev wrote:
> Hi all
>
> I just tried my luck on an old P-II, 233 MHz, 64 MB laptop.
>
> Booting huge.s I could finish the installation. But I can't start
> the installation. It gives
>
> no file system could mount root, tried:
> Kernel panic - not syncing: VFS: Unable to mount root fs on
> unknown-block(8,2)
>
> Am I correct in assuming that the initrd is too big for the 64 MB?

No the initrd is quite small and you won't have this kind of message if
it were the case. The problem is (most probably); you do not include the
right modules for your root filesystem or do not have specified it
correctly. See man mkinitrd. If you use ext3 on /dev/hda2 you should
generate a good initrd by

mkinitrd -c -r /dev/hda2 -f ext3 -m ext3 -o initrd.img

(boot first with the huge kernel and make the initrd once the system is
booted; then you can boot with another kernel).

Olive

Arun Dev wrote:
> Hi all
>
> I just tried my luck on an old P-II, 233 MHz, 64 MB laptop.
>
> Booting huge.s I could finish the installation. But I can't start
> the installation. It gives
>
> no file system could mount root, tried:
> Kernel panic - not syncing: VFS: Unable to mount root fs on
> unknown-block(8,2)
>
> Am I correct in assuming that the initrd is too big for the 64 MB?

No the initrd is quite small and you won't have this kind of message if
it were the case. The problem is (most probably); you do not include the
right modules for your root filesystem or do not have specified it
correctly. See man mkinitrd. If you use ext3 on /dev/hda2 you should
generate a good initrd by

mkinitrd -c -r /dev/hda2 -f ext3 -m ext3 -o initrd.img

(boot first with the huge kernel and make the initrd once the system is
booted; then you can boot with another kernel).

Olive

Hallo, Arun,

Du meintest am 30.01.08:


> I just tried my luck on an old P-II, 233 MHz, 64 MB laptop.

> Booting huge.s I could finish the installation. But I can't start
> the installation. It gives

> no file system could mount root, tried:
> Kernel panic - not syncing: VFS: Unable to mount root fs on
> unknown-block(8,2)

May not help you: I find this message on 1 of my machines with many
kernels (Compaq 500, 498 MB RAM), I don't find it on my (many) other
machines. Some kernels have no problems - I haven't found why.

I'd found this message under Slackware 11.x too.

Viele Gruesse
Helmut

"Ubuntu" - an African word, meaning "Slackware is too hard for me".

On Wed, 30 Jan 2008 17:43:44 +0100
Arun Dev wrote:
> I just tried my luck on an old P-II, 233 MHz, 64 MB laptop.
>
> Booting huge.s I could finish the installation. But I can't start
> the installation. It gives
>
> no file system could mount root, tried:
> Kernel panic - not syncing: VFS: Unable to mount root fs on
> unknown-block(8,2)
>
> Am I correct in assuming that the initrd is too big for the 64 MB?

IIRC, huge.s doesn't require initrd (providing /boot partition
is of the type compiled in the kernel, e.g., ext3). Are you sure
you have properly configured lilo.conf?

--
Mikhail

> "Ubuntu" - an African word, meaning "Slackware is too hard for me".
>

You did complain about troll... If it is really what you think; why
don't you go to Ubuntu? What do you do on this list?

Olive

i think i'b better try this slack-distro

http://www.pcbypaul.com/absolute/

It works well on old pcs

On Wed, 30 Jan 2008 17:43:44 +0100
Arun Dev wrote:

Hello all

Am 30.01.2008 17:43, Arun Dev schrieb:
>
> I just tried my luck on an old P-II, 233 MHz, 64 MB laptop.
>
> Booting huge.s I could finish the installation. But I can't start
> the installation. It gives
>
> no file system could mount root, tried:
> Kernel panic - not syncing: VFS: Unable to mount root fs on
> unknown-block(8,2)

It could be missing modules for the file system then? The root
partition was formatted ext3. Is it possible that Slack "generic"
kernel has no ext3 built-in?

Anyway I don't have the computer right now, will check tomorrow.

BTW, I like the idea behind "Absolute Linux". In fact I was
thinking of trying Puppy Linux if there is a fundamental
problem with Slack 12.

greetings
Arun

On Wed, 30 Jan 2008 21:35:25 +0100, Arun Dev wrote:

> Hello all
>
> Am 30.01.2008 17:43, Arun Dev schrieb:
>>
>> I just tried my luck on an old P-II, 233 MHz, 64 MB laptop.
>>
>> Booting huge.s I could finish the installation. But I can't start the
>> installation. It gives
>>
>> no file system could mount root, tried: Kernel panic - not syncing:
>> VFS: Unable to mount root fs on unknown-block(8,2)
>
> It could be missing modules for the file system then? The root partition
> was formatted ext3. Is it possible that Slack "generic" kernel has no
> ext3 built-in?

No, it definitely has ext3 supported.

> Anyway I don't have the computer right now, will check tomorrow.
>
> BTW, I like the idea behind "Absolute Linux". In fact I was thinking of
> trying Puppy Linux if there is a fundamental problem with Slack 12.

From my experience, Absolute is a *lot* faster on the same hardware than
Puppy. Paul put a lot of effort into minimising the footprint of
Absolute and it shows.

Arun Dev wrote:
> Hi all
>
> I just tried my luck on an old P-II, 233 MHz, 64 MB laptop.
>
> Booting huge.s I could finish the installation. But I can't start
> the installation. It gives
>
> no file system could mount root, tried:
> Kernel panic - not syncing: VFS: Unable to mount root fs on
> unknown-block(8,2)
>
> Am I correct in assuming that the initrd is too big for the 64 MB?
>
> The root device is accoding to lilo.conf /dev/hda2
>
> Kernel was 2.6.22 "generic".
>
> Arun
I had similar problem which turned out to be bad ram , even though the bios
ram test passed.
Originally thought it was the multiprocessor kernel, on my first attempt.

Am 31.01.2008 02:24, Sambo schrieb:
> Arun Dev wrote:
>>
>> I just tried my luck on an old P-II, 233 MHz, 64 MB laptop.
>>
>> Booting huge.s I could finish the installation. But I can't start
>> the installation. It gives
>>
>> no file system could mount root, tried:
>> Kernel panic - not syncing: VFS: Unable to mount root fs on
>> unknown-block(8,2)
>>
>> Am I correct in assuming that the initrd is too big for the 64 MB?
>>
> I had similar problem which turned out to be bad ram , even though the
> bios ram test passed. Originally thought it was the multiprocessor
> kernel, on my first attempt.

Could it be a hardware problem after all?

I reinstalled with ext2. During boot-up the filesystem check fails.
I booted from the cd and tried "e2fsck -v -y /dev/hda2" and get
this story (manually typed):
---
BUG: unable to handle kernel paging request at virtual address
00005e40
printing eip:
*pde = 00000000
Oops: 0002 [#1]
CPU: 0
EIP: 00c0:[<00003c65>] Not tainted VLI
EFLAGS: 00010003 (2.6.21.5 #2)
..... stack trace follows ...
Call Trace:
=========================
Code: Bad EIP value
EIP: [<00003c65>] 0x3c65 SS:ESP 0068:c3b45e3e
Segmentation fault
---

This laptop was not used for years. The CMOS battery had "feathers"
around the edges :-( We replaced it.

Arun

Arun Dev wrote in news:47a1f939$0$3421
$5402220f@news.sunrise.ch:

> Am 31.01.2008 02:24, Sambo schrieb:
>> Arun Dev wrote:
>>>
>>> I just tried my luck on an old P-II, 233 MHz, 64 MB laptop.
>>>
>>> Booting huge.s I could finish the installation. But I can't start
>>> the installation. It gives
>>>
>>> no file system could mount root, tried:
>>> Kernel panic - not syncing: VFS: Unable to mount root fs on
>>> unknown-block(8,2)
>>>
>>> Am I correct in assuming that the initrd is too big for the 64 MB?
>>>
>> I had similar problem which turned out to be bad ram , even though the
>> bios ram test passed. Originally thought it was the multiprocessor
>> kernel, on my first attempt.
>
> Could it be a hardware problem after all?
>

[snip]

>
> This laptop was not used for years. The CMOS battery had "feathers"
> around the edges :-( We replaced it.
>
> Arun
>

Yes, it could. Depending upon the type of battery it might have leaked a
caustic base (alkali for alkaline batteries) onto the motherboard. This
is a not-uncommon occurrence among older equipment. Sometimes damaged
traces can be repaired, sometimes not.

It is interesting, though, that you seem to be able to boot from CD and
do enough to install an OS. Perhaps you might try a less ambitious
install, or try installing a small variety of different small distros to
see if one likes your hardware better than slackware?

I suggest that because I've access to one computer (a Shuttle brand) that
simply refuses to boot any linux distro if you leave Linux's USB enabled.
Booting simply hangs. Perhaps some laptop-specific drive-interface
electronics is simply failing to respond well to Linux?

--
The email address, above, is most certainly munged. Perhaps you
might reply to the newsgroup, instead? Thanks!

On 2008-01-31, Arun Dev wrote:
> I reinstalled with ext2. During boot-up the filesystem check fails.
> I booted from the cd and tried "e2fsck -v -y /dev/hda2" and get
> this story (manually typed):
> ---
> BUG: unable to handle kernel paging request at virtual address
> 00005e40
> printing eip:
> *pde = 00000000
> Oops: 0002 [#1]
> CPU: 0
> EIP: 00c0:[<00003c65>] Not tainted VLI
> EFLAGS: 00010003 (2.6.21.5 #2)
> .... stack trace follows ...
> Call Trace:
> =========================
> Code: Bad EIP value
> EIP: [<00003c65>] 0x3c65 SS:ESP 0068:c3b45e3e
> Segmentation fault
> ---
>
> This laptop was not used for years. The CMOS battery had "feathers"
> around the edges :-( We replaced it.


Sounds like a memory problem is the most likely culprit.
I'll admit to not having read anything before this post, so if you've
already ruled that out, mea culpa. There are quite a few other things
that could potentially cause this, but faulty memory is the most common
IME. Check to be sure it's seated properly in the sockets.

I had a very similar problem on a brand new box a few weeks ago.
All of the memory tested fine *most* of the time. It would usually
boot up just fine, but occasionally init would even segfault. When
it booted fine, it would sometimes run for days with no problem;
other times, I'd come back after leaving it idle and all I'd see on
the console was a "panic party" that started with a kernel OOPS.
I *thought* I had it traced back to a bad driver (snd_hda_intel),
and after blacklisting it, we went for several days with no problem
until suddenly the gremlin returned. Anyway, it turned out to be...
Yeah, something I should have investigated very early - the BIOS.
I flashed it with a newer BIOS and all is well.

Now, with all that said, I doubt a newer BIOS is what you need, and
since it's an older laptop, I seriously doubt you'll even be able to
find one. However, maybe that gives you some ideas to consider.
I strongly suspect hardware problems, or possibly even cooling issues.

Good luck!

-RW

Arun Dev wrote:
> Could it be a hardware problem after all?
>

try downloading and running memtest, and letting it run for 72 hours.

My old fileserver ran fine for years, and then popped a cap on the mobo
AND blew a stick of ram at the same time, and went from 100% uptime to
crashing every 15 minutes, which made getting the data off a real pain.

(and before you ask why I didn't pull the disks, I only had one other
machine that could read disks that big and that was the NEW server,
which is full.)

Ray

Hi Federico and Mark

Am 30.01.2008 21:40, Mark South schrieb:
> On Wed, 30 Jan 2008 21:35:25 +0100, Arun Dev wrote:
>>
>> BTW, I like the idea behind "Absolute Linux". In fact I was thinking of
>> trying Puppy Linux if there is a fundamental problem with Slack 12.
>
> From my experience, Absolute is a *lot* faster on the same hardware than
> Puppy. Paul put a lot of effort into minimising the footprint of
> Absolute and it shows.

I gave "abosolute" a try. Hey, the install script is almost the same
as Slack. But there was one difference, and that one killed my
effort. I didn't find a package selection menu. It is either you
want all or nothing. The "all" is according to a message 3 GB. So
as expected the installation bogged down after my 2 GB got full :-(

Pretty bad, on this lap top it took hours to get those 2 GB copied.

As I suspect hardware anyway, I'll do some checks before dumping the
whole exercise.

Thanks for all the pointers!

Arun

Am 31.01.2008 18:38, Ray schrieb:
> Arun Dev wrote:
>> Could it be a hardware problem after all?
>
> try downloading and running memtest, and letting it run for 72 hours.

I booted from a CD which has memtest (Memtest86+ v1.65) on it. I've
started the test a couple of minutes ago.


Arun

On Thu, 31 Jan 2008 19:20:22 +0100, Arun Dev wrote:

> Hi Federico and Mark
>
> Am 30.01.2008 21:40, Mark South schrieb:
>> On Wed, 30 Jan 2008 21:35:25 +0100, Arun Dev wrote:
>>>
>>> BTW, I like the idea behind "Absolute Linux". In fact I was thinking
>>> of trying Puppy Linux if there is a fundamental problem with Slack 12.
>>
>> From my experience, Absolute is a *lot* faster on the same hardware
>> than Puppy. Paul put a lot of effort into minimising the footprint of
>> Absolute and it shows.
>
> I gave "abosolute" a try. Hey, the install script is almost the same as
> Slack. But there was one difference, and that one killed my effort. I
> didn't find a package selection menu. It is either you want all or
> nothing. The "all" is according to a message 3 GB. So as expected the
> installation bogged down after my 2 GB got full :-(

Yeah, it is all or nothing, no choice of kernel either. Sorry to have
not noticed your disk constraints.

> Pretty bad, on this lap top it took hours to get those 2 GB copied.

See, this is why you need several different systems under your desk :-)

> As I suspect hardware anyway, I'll do some checks before dumping the
> whole exercise.

If your memory and disk space are so severely limited, I could only
suggest that you try NetBSD, which will definitely not be resource
limited in 64MB and 2GB. Read the NetBSD FAQ before doing anything.

> Thanks for all the pointers!

Hope it's been of some help, at least.

Hi all

Am 31.01.2008 20:03, Arun Dev schrieb:
> Am 31.01.2008 18:38, Ray schrieb:
>>
>> try downloading and running memtest, and letting it run for 72 hours.
>
> I booted from a CD which has memtest (Memtest86+ v1.65) on it. I've
> started the test a couple of minutes ago.

If I knew what the expected outcome is!
;-[

Now (about 14h later) the screen stands still
---
Memtest86+ Pass 17% #####
Pentium-MMX 266.7 MHz Test 47% #################
L1 Cache: 16K 127 MB/s Test #4 [Moving inversions, random pattern]
L2 Cache: Unknown Testing: 108K - 64M 64M
Memory: 64M 110MB/s Pattern: 75394bd2
Chipset:

WallTime 2:26:26
Cached 64M
RsvdMem 56K
MemMap e820-Std
Cache on
ECC off
Test Std
Pass 4
Error 0
Ecc Errs
---

What does that horoscope mean?

Arun

On Fri, 01 Feb 2008 13:07:08 +0100, Arun Dev wrote:

> Memtest86+ Pass 17% #####
....
> WallTime 2:26:26
> Cached 64M
> RsvdMem 56K
> MemMap e820-Std
> Cache on
> ECC off
> Test Std
> Pass 4
> Error 0
> Ecc Errs
> ---
>
> What does that horoscope mean?

It ran through the test cycle 3 times without showing an error, and
that's good, but 17% of the way through the 4th pass the machine locked
up. I see from the clock that it ran tests for 2 hours and 26 minutes
before locking. Lockups are not unknown with memtest, since it is
messing with the memory after all. The fact that no errors were detected
is a good thing, but it may be that the lockup was caused by a memory
error due to the stress of the testing.

I'd recommend runnng memtest again and checking that the walltime clock
is still running every hour or so.