Overcoming 8GB BIOS limit

  • Thread starter Thread starter Michael
  • Start date Start date
Joep said:
Yes thank you for confirming your inteventions would have been abolutely
meaningless. And yet you still suggest someone should start a partition
table editor without giving correct advice. That in itself is stupid.

Calm down, Joepie, you don't make any sense. Watch the bloodpressure.
Yes, indeed, however my remark on the creating an extended (0F) instead of a
'free space partition' was for you.

Posture all you want, Joepie, I don't need to partition free space,
he did, so your comment was obviously addressed to him, not me.
No he doesn't because he ASKS!

Read again carefully, Joepie, the comment was to him and the "he" in that
comment, that is *you*.
Count the quoting characters if you can't figure it out by the comment alone.
I quote OP to give some indication of his experience with using partition table
editors: "I have no idea how to edit the master boot record".

Unless he wanted to say that he doesn't know how to use a keyboard or doesn't
understand english, that made totally no sense after the instructions I gave.
No he didn't.

You did, Joepie, you did.
If you will use Fdisk from Windows the whole partition table editing execise
is pointless.

Not from the standpoint that Fdisk would see no free space and that that
free space was to be made visible so that Fdisk would then be able to use it.
You suggest to someone, incomplete and icorrect modifications to the
partition table.

In my then understanding of the problem they were totally complete.
As I pointed out my then understanding of the problem was incorrect.
That you took it a step further and proposed to actually edit it into
an extended partition without setting up the EMBR (talking about
being incomplete) is your problem.
 
Arno Wagner said:
The limitation in LILO is the same as for all boot-managers:

Not necessarily. All the bios has to do is load block 0.
What comes next can be anything you choose, as boot overlays prove.
Unless bios independent code can be stored in a single block it only
takes a bios call extra to load a few blocks more to load all of
it and use that code in the further process to startup the system.
They need the BIOS to load the system.

Only one call and that is by the bios itself.
Hence for Linux the kernel has to reside in a BIOS-accessable area.

Bummer.
Here we always hear how Linux is bios independent and now we hear
that it's loader isn't.
 
Joep said:
Where exactly? Free space (I prefer unallocated space) isn't recorded,
that's why it's free space.

Free space is the space that can be included in partitions. Once the maximum
allowed number of partitions is defined, unused space stops being free space.
Free space is what remains outside of areas that are claimed by a partition.

In general. Not with Fdisk.
In practice it depends on how many (primary) partitions are allowed by
the OS that you are using and whether that includes an extended one.
Once a primary and an extended partition are defined and not using the
whole of the drive's available space, that unused space is not free space
in the eyes of Fdisk. That is what I meant earlier with "finalized".
To free-up the unused space you will have to resize the extended par-
tition and probably any other extended one further up in the chain.
That's not an Fdisk thing, it's a BIOS thing.

No, since we are obviously talking about Fdisk here, it is a Microsoft
thing.
Other apps that accept more than one primary may see it different.
Before you ask; yes 'thing' is an existing technical term.

If you say so.
It isn't, partitions are defined at 'partitioning time'.

And so is the usable space once the Extended partition is defined.
Is this the same Folkert who always tells people to get their facts straight
by referring to the various ATA and whatever specs he can come up with?

Yup, and I already admitted that I'm probably wrong.
Having an off day?

As a matter of fact, Yes. Can't you tell?
And as it appears, so have you. Watch your bloodpressure.
I quote OP: "The BIOS limits the system to seeing only 8GB".
We're talking about a BIOS limitation here.

Yes, and I was talking about what that limitation might have caused to free
space being available or not. Once you define an extended partition during
the 8GB limitation the unused space will still be unavailable once you go into
an environment that hasn't got that limitation. That is what I overlooked and
what made me think that Fdisk was recording the seen capacity somewhere.

And in sense that makes me right afterall even though I didn't understand
it properly before.
Also, how DOS sees a disk or Windows are entirely different matters.

When your bloodpressure drops your eyes will clear and you'll notice
that I wasn't talking about the 8GB bios limitation at all, but about
what Fdisk sees once 2 partitions in the MBR have been defined.
And you don't wonder why they do?


Read back from where exactly? Are you referring to Ptedit? If so, you humor
me and you read it back to us. I bet without even looking that Ptedit will
not mention a 'free space partition' anywhere.

Fraid so, although by implication, not exact words.
Are you Joep@diydatarecovery saying that you have never used PTEdit?
Shame on you, Joepie.

In the right place. Are you Joep@diydatarecovery saying
that you never used Fdisk and don't know where to look?
If that works, then I consider that an Fdisk bug.

What, you didn't check?
It implies that I could edit the partition table to add numbers (start and
end values) to an unused entry that exceed the physical capacity of the
physical disk,

You didn't check, did ya.
and that after that Fdisk would happily
create partitions in non existing space.

Don't know about that (yes, I did not check) but it happily ac-
cepts it to be bigger than that part of the drive in reality is.
When Fdisk is run in Windows, the BIOS limitation doesn't apply so there
would be no need for your Ptedit voodoo.

There still may have been some "need" if he had an extended partition defined.
But yes, he doesn't have one so the free space will probably be available for
partitioning then, as I already mentioned.
(you just repeated what I already spoonfed to you, below)
Yes, even without all your ptedit hocus pocus.

And now you are repeating yourself.
However, the issue that the BIOS does not support disks > 8 Gb remains.

Well, as long as they are incompatible with use under some
Low Level OS that is relying on the BIOS, he can't care less.
 
Michael said:
I thought that Windows 3.1/95/98/ME all used the BIOS and only 2000/XP
couldn't have this problem.

Everything uses the bios in the boot process to a certain extend, even 2000/XP.
It depends on that "extend" how much problems you will have with the 8GB
limit and whether you intend to boot anything beyond that 8GB limit and/or
what it is that is booting to it (MBR boot code, bootmanager or otherwise).
Apparently, so they say, 2000/XP uses it's own code in an earlier stage so
some of the problems don't exist. Some say you don't need a bios overlay
with 2000/XP because it has it builtin but unless this is in the 2000/XP
associated bootmanager I fail to see where it would make a difference.
And when the bios limitation fails to recognize and/or accept a
drive properly at POST, then that does't help in any way.

Even 2000/XP will have problems with applications (or rather the applica-
tions that don't use it in some stage e.g. Imaging/backup etc) that shutdown
the system to run in a Dos or other environment that relies on the BIOS.
Also, I might want to create a Linux partition somewhere in the unalloca-
ted space, but I think there's some kind of LILO problem with 8GB too.

Anything that builds on the BIOS and tries to address beyond the 8GB limit
(with a 8GB limited bios, in case Joepie tries to misunderstand that too) will
have a problem.
 
Michael said:
I meant, could I have a, say, 7GB Windows partition and then have a
Linux partition which overlapped the 8GB mark, not having the two
partitions overlap.
You could have 7GB for Windows and 1GB for the Linux root, and rest could be
swap, /usr, ... in extended.
 
Previously Michael said:
I meant, could I have a, say, 7GB Windows partition and then have a
Linux partition which overlapped the 8GB mark, not having the two
partitions overlap.

I see. Yes, no problem on that. Here is what I use (XP and Linux):

Disk /dev/hda: 160.0 GB, 160041885696 bytes
255 heads, 63 sectors/track, 19457 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System
/dev/hda1 1 10 80324+ 83 Linux
/dev/hda2 * 11 764 6056505 b W95 FAT32
/dev/hda3 765 19457 150151522+ f W95 Ext'd (LBA)
.... more partitions below, Linux partitions at end ....


As you can see you can have a small Linux partition even before
the Windows primary partition. The above set-up was created
with Linux fdisk, don't know whether it can be created with
Windows.

Arno
 
You could have 7GB for Windows and 1GB for the Linux root, and rest could be
swap, /usr, ... in extended.
I'm still not clear here. This is an ASCII diagram of what I am asking
about:

Windows C: Windows D:
| 8GB Linux / |
| line | |
|-|-----|--------|-----------|--------|
|@@@##################@@@@@@@@@@@@@@@@|
|-----|-------------------------------|
|
Linux kernel

So in this diagram *one* Linux partition, the / partition, is
overlapping the troublesome 8GB line with the kernel being before that
line. Would it just be easier to put the kernel on the windows
partition?
 
:>No, definitely not!!! Never, ever overlap partitions!
:I meant, could I have a, say, 7GB Windows partition and then have a
:Linux partition which overlapped the 8GB mark, not having the two
:partitions overlap.

The problem is that you can't guarantee where any particular file will
be placed in the partition. You might get lucky and have all the
critical files entirely below the 8GB boundary, and then someday update
your kernel and find that it's outside the BIOS-accessible region. You
have to have a partition that is entirely below the 8GB mark.
 
Anyone can see how that is an obvious lie, Joepie.

As it is only me who can say anything about my own intentions you'll have to
take my word for it. Since you prefer to call me Joepie, which is very
childish, I will lower myself to your level and start addressing you as
F'NUT again.
 
Folkert Rienstra said:
Calm down, Joepie, you don't make any sense. Watch the bloodpressure.

Well, the thing is, I *am* calm. But even if I weren't, yes you could ask me
to calm down. Unfortunaly it would be unfair to ask from you to smarten up a
little and use some common sense.

Posture all you want, Joepie, I don't need to partition free space,
he did, so your comment was obviously addressed to him, not me.

Look F'NUT, look up the original message where I comment on your suggestion,
and quot your suggestions. It is obvious my answer is for YOU, F'NUT and
no-one else.
Read again carefully, Joepie, the comment was to him and the "he" in that
comment, that is *you*.
Count the quoting characters if you can't figure it out by the comment
alone.

Oh, now you got me! Ouch, I miscount a quote.
Unless he wanted to say that he doesn't know how to use a keyboard or doesn't
understand english, that made totally no sense after the instructions I gave.

You did, Joepie, you did.

You're even more stupid than I thought. I explained YOU how defining a 'free
space partition' was completely useless, and that defining an extended
partition would make more sense. I explained *you* that there is no thind as
a 'free space partition.

BTW. You still didn't tell us how you can create a 'free space partition'
with Ptedit.
Not from the standpoint that Fdisk would see no free space and that that
free space was to be made visible so that Fdisk would then be able to use it.

Fdisk didn't free space because the BIOS didn't show it to Fdisk. Fdisk in
Windows will show the entire disk.
In my then understanding of the problem they were totally complete.

F'NUT, you have to accept the fact that your capabilities to understand
anything are very limited. I am not sure you can even understand this.
 
Folkert Rienstra said:
Free space is the space that can be included in partitions. Once the maximum
allowed number of partitions is defined, unused space stops being free space. partition.

In general. Not with Fdisk.
In practice it depends on how many (primary) partitions are allowed by
the OS that you are using and whether that includes an extended one.
Once a primary and an extended partition are defined and not using the
whole of the drive's available space, that unused space is not free space
in the eyes of Fdisk.

You can ot define partitions there, but not because there's no free space
left.
No, since we are obviously talking about Fdisk here, it is a Microsoft
thing.
Other apps that accept more than one primary may see it different.

Since we're talking a P166 PC it's a BIOS thing. I do not care what you do
at home, I answer and comment taking OP's situation into account. All noise
you decide to introduce trying to talk your way out is irrelevant.
And so is the usable space once the Extended partition is defined.

So now you decide to call it 'usable' space. Let me remind you how this
discussion started; you were going to tell us about 'free space partitions'.
Yup, and I already admitted that I'm probably wrong.

It took you a long time to figure it out though.
As a matter of fact, Yes. Can't you tell?
And as it appears, so have you. Watch your bloodpressure.

Blood preseure is okido, thank you.
Yes, and I was talking about what that limitation might have caused to free
space being available or not.

And why not take the obvious into account ... A P166 ... BIOS limitation?
Once you define an extended partition during
the 8GB limitation the unused space will still be unavailable once you go into
an environment that hasn't got that limitation. That is what I overlooked and
what made me think that Fdisk was recording the seen capacity somewhere.

And in sense that makes me right afterall even though I didn't understand
it properly before.

No F'Nut, it doesn't.
When your bloodpressure drops your eyes will clear and you'll notice
that I wasn't talking about the 8GB bios limitation at all, but about
what Fdisk sees once 2 partitions in the MBR have been defined.

No, you were talking about 'free space partitions' and how to define them
using Ptedit. While the real issue was a 8 Gb BIOS limitation.
Fraid so, although by implication, not exact words.

Please tell us what it says there F'Nut.
Are you Joep@diydatarecovery saying that you have never used PTEdit?
Shame on you, Joepie.

Thanks F'Nut at (e-mail address removed)
In the right place. Are you Joep@diydatarecovery saying
that you never used Fdisk and don't know where to look?

I didn't say that F'Nut at (e-mail address removed)
What, you didn't check?


You didn't check, did ya.

Nope. Doesn't matter, if Fdisk does that then that's a bug. It's a simple IF
... THEN.
Don't know about that (yes, I did not check) but it happily ac-
cepts it to be bigger than that part of the drive in reality is.

So? Thazabug
There still may have been some "need" if he had an extended partition defined.
But yes, he doesn't have one so the free space will probably be available for
partitioning then, as I already mentioned.
(you just repeated what I already spoonfed to you, below)


And now you are repeating yourself.

Yes, so? Maybe if I repeat myself often enough you will understand.
Well, as long as they are incompatible with use under some
Low Level OS that is relying on the BIOS, he can't care less.

Indeed. But without knowing yourself, you suggested him to 'work around
that'. And that's what my comments were all about: Warn OP that your
proposal was incorrect and potentially dangerous.
 
Fdisk didn't free space because the BIOS didn't show it to Fdisk.
Fdisk in
Windows will show the entire disk.
No it won't. I tried. And in case anybody's wondering, I have *one
primary partition*, not any extended ones.
Notwithstanding the 8GB problem, can you have a DOS primary partition,
then an extended partition with however many logical drives in it, and
then two Linux partitions?
 
That might just be fdisk tho. Try partitioning it with something
else.
Yeah, have a look at what it puts on the CD and just download one of
them.

The only ones under 'Partition' are MBRWiz and TestDisk, neither of
which *create* partitions.
 
Joep said:
The statement I replied to was incorrect, so my answer is correct.

Nope, he ripped it out of context and then you snipped it even more
and that made it "semi-casual, incorrect and incomplete" and therefor
wrong, according to your own values.
 
Joep said:
Well, the thing is, I *am* calm.

Nonsense, anyone can see by the number of typos that you are red hot and steaming.
But even if I weren't, yes you could ask me to calm down.
Unfortunaly it would be unfair to ask from you to smarten up a little
and use some common sense.

Which apparently has left you.
Look F'NUT, look up the original message where I comment on your suggestion,
and quot your suggestions. It is obvious my answer is for YOU, F'NUT and
no-one else.


Oh, now you got me! Ouch, I miscount a quote.

It's that red daze in front of your eyes, Joepie, you need to calm down.
You're even more stupid than I thought. I explained YOU how defining a 'free
space partition' was completely useless, and that defining an extended
partition would make more sense.

But it didn't make more sense. That was at least equally stupid, if not more.
I explained *you* that there is no thind as a 'free space partition.

Yes, you keep repeating yourself.
BTW. You still didn't tell us how you can create a 'free space partition'
with Ptedit.

Yes, I did. Read my first post. That is, after you have calmed down.
Fdisk didn't free space because the BIOS didn't show it to Fdisk. Fdisk in
Windows will show the entire disk.

I believe OP just proved you wrong.
(It's that red daze again, isn't it, Joepie. You really need to calm down).
I won't be going so far as to say that he proved me right though.
In my then understanding of the problem they were totally complete.

[idiot's rant snipped]
 
Remember the 'Fdisk bug' technique mentioned by Folkert? If I want to
try that, how do I find out *exactly* what the number of sectors is?
 
Michael said:
No it won't. I tried. And in case anybody's wondering, I have *one
primary partition*, not any extended ones.
Notwithstanding the 8GB problem, can you have a DOS primary partition,
then an extended partition with however many logical drives in it, and
then two Linux partitions?
You can create 3 primaries and an extended with Linux fdisk in any order.

However, all FAT volumes must all be below 8GB, also the Linux boot (which is
usually root).
 
Joep said:
You can ot define partitions there, but not because there's no free space
left.

Wow, for once you actually understood what I said. What was the magic word?
Since we're talking a P166 PC it's a BIOS thing. I do not care what you do
at home, I answer and comment taking OP's situation into account.

That's an obvious lie, Joepie.
All noise you decide to introduce trying to talk your way out is irrelevant.

Anyone can see who it is that is making all the noise, Joepie.
So now you decide to call it 'usable' space. Let me remind you how this
discussion started; you were going to tell us about 'free space partitions'.


It took you a long time to figure it out though.

And since OP can't add space even in Windows it may take yet a bit longer, Joepie.
Blood preseure is okido, thank you.


And why not take the obvious into account ... A P166 ... BIOS limitation?

And if he had an extended defined under that limitation, that limitation stays,
even under Windows. Similar to my case where I sector cloned a smaller drive
to a bigger one. Different cause but same limitation.
No F'Nut, it doesn't.

Have it your way.
No, you were talking about 'free space partitions' and how to define them
using Ptedit. While the real issue was a 8 Gb BIOS limitation.

The real issue was how to partition the unused space even with that 8 Gb BIOS
limitation.
Please tell us what it says there F'Nut.


Thanks F'Nut at (e-mail address removed)>

I didn't say that F'Nut at (e-mail address removed)

Then "please" prove it to us, Joepie.
Nope. Doesn't matter, if Fdisk does that then that's a bug. It's a simple IF
.. THEN.


So? Thazabug.

Yeah and you Joep@diydatasomething obviously didn't know about it.
Shame on you, Joepie.
Yes, so? Maybe if I repeat myself often enough you will understand.

Maybe you should beat it into me, right Joepie?
Indeed. But without knowing yourself, you suggested him to 'work around
that'.

I don't think that he is as dense as you make him out to be, Joepie.
And that's what my comments were all about: Warn OP that your
proposal was incorrect and potentially dangerous.

Sorry Joepie, but I don't think that you fooled anyone.
 
Back
Top