hi.
for partitions not in the first 8gb of a harddisk, what
should the c/h/s start and end value be ?
most fdisks seem to set start and end to 255/63/1023.
but partition magic creates partitions with start set to
0/1/1023 and end set to 255/63/1023, and detects a problem
if start is set to 255/63/1023.
so, how should a partition table look like ?
regards, andreas
> for partitions not in the first 8gb of a harddisk, what
> should the c/h/s start and end value be ?
http://www.linuxdoc.org/HOWTO/Large-Disk-HOWTO.html
--
On Sun, Feb 25, 2001 at 04:35:34PM +0100, Andreas Jellinghaus wrote:
> for partitions not in the first 8gb of a harddisk, what
> should the c/h/s start and end value be ?
>
> most fdisks seem to set start and end to 255/63/1023.
> but partition magic creates partitions with start set to
> 0/1/1023 and end set to 255/63/1023, and detects a problem
> if start is set to 255/63/1023.
>
> so, how should a partition table look like ?
Since these values will never be used you can imagine that they
are not of great interest. Do whatever you like.
Please also check vger.timpanogas.org/nwfs/nwfs.tar.gz:disk.c for NetWare
specific calculations of the CHS values, a different method is used for
NetWare partitions vs. everything else (Novell just had to be different).
If you do not use their methods on NetWare partitions, NetWare will not
recognize the partition entries correctly, and will attempt to reinitialize
the entire partition table on a system if they are wrong (Ouch!). NetWare
does a different calculation for conversion of cylinder values above 1024.
Jeff
----- Original Message -----
From: "Guest section DW" <[email protected]>
To: "Andreas Jellinghaus" <[email protected]>;
<[email protected]>
Sent: Sunday, February 25, 2001 2:47 PM
Subject: Re: partition table: chs question
> On Sun, Feb 25, 2001 at 04:35:34PM +0100, Andreas Jellinghaus wrote:
>
> > for partitions not in the first 8gb of a harddisk, what
> > should the c/h/s start and end value be ?
> >
> > most fdisks seem to set start and end to 255/63/1023.
> > but partition magic creates partitions with start set to
> > 0/1/1023 and end set to 255/63/1023, and detects a problem
> > if start is set to 255/63/1023.
> >
> > so, how should a partition table look like ?
>
> Since these values will never be used you can imagine that they
> are not of great interest. Do whatever you like.
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
----- Original Message -----
From: "Jeff V. Merkey" <[email protected]>
To: "Guest section DW" <[email protected]>; "Andreas Jellinghaus"
<[email protected]>; <[email protected]>
Sent: Sunday, February 25, 2001 5:02 PM
Subject: Re: partition table: chs question
>
> Please also check vger.timpanogas.org/nwfs/nwfs.tar.gz:disk.c for NetWare
> specific calculations of the CHS values, a different method is used for
> NetWare partitions vs. everything else (Novell just had to be different).
> If you do not use their methods on NetWare partitions, NetWare will not
> recognize the partition entries correctly, and will attempt to
reinitialize
> the entire partition table on a system if they are wrong (Ouch!). NetWare
> does a different calculation for conversion of cylinder values above 1024.
>
> Jeff
>
>
> ----- Original Message -----
> From: "Guest section DW" <[email protected]>
> To: "Andreas Jellinghaus" <[email protected]>;
> <[email protected]>
> Sent: Sunday, February 25, 2001 2:47 PM
> Subject: Re: partition table: chs question
>
>
> > On Sun, Feb 25, 2001 at 04:35:34PM +0100, Andreas Jellinghaus wrote:
> >
> > > for partitions not in the first 8gb of a harddisk, what
> > > should the c/h/s start and end value be ?
> > >
> > > most fdisks seem to set start and end to 255/63/1023.
> > > but partition magic creates partitions with start set to
> > > 0/1/1023 and end set to 255/63/1023, and detects a problem
> > > if start is set to 255/63/1023.
> > >
> > > so, how should a partition table look like ?
> >
> > Since these values will never be used you can imagine that they
> > are not of great interest. Do whatever you like.
This statement is not true of NetWare partitions. They are used to validate
whether or not the partition table has been corrupted.
Jeff
> > -
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel"
in
> > the body of a message to [email protected]
> > More majordomo info at http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at http://www.tux.org/lkml/
>
It does not matter because the usage of CHS will dies soon because it was
voted to death in Austin last week. There will only be LBA addressing
from now on out.
On Sun, 25 Feb 2001, Jeff V. Merkey wrote:
>
> Please also check vger.timpanogas.org/nwfs/nwfs.tar.gz:disk.c for NetWare
> specific calculations of the CHS values, a different method is used for
> NetWare partitions vs. everything else (Novell just had to be different).
> If you do not use their methods on NetWare partitions, NetWare will not
> recognize the partition entries correctly, and will attempt to reinitialize
> the entire partition table on a system if they are wrong (Ouch!). NetWare
> does a different calculation for conversion of cylinder values above 1024.
>
> Jeff
>
>
> ----- Original Message -----
> From: "Guest section DW" <[email protected]>
> To: "Andreas Jellinghaus" <[email protected]>;
> <[email protected]>
> Sent: Sunday, February 25, 2001 2:47 PM
> Subject: Re: partition table: chs question
>
>
> > On Sun, Feb 25, 2001 at 04:35:34PM +0100, Andreas Jellinghaus wrote:
> >
> > > for partitions not in the first 8gb of a harddisk, what
> > > should the c/h/s start and end value be ?
> > >
> > > most fdisks seem to set start and end to 255/63/1023.
> > > but partition magic creates partitions with start set to
> > > 0/1/1023 and end set to 255/63/1023, and detects a problem
> > > if start is set to 255/63/1023.
> > >
> > > so, how should a partition table look like ?
> >
> > Since these values will never be used you can imagine that they
> > are not of great interest. Do whatever you like.
> > -
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to [email protected]
> > More majordomo info at http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at http://www.tux.org/lkml/
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>
Andre Hedrick
Linux ATA Development
ASL Kernel Development
-----------------------------------------------------------------------------
ASL, Inc. Toll free: 1-877-ASL-3535
1757 Houret Court Fax: 1-408-941-2071
Milpitas, CA 95035 Web: http://www.aslab.com
On Sun, Feb 25, 2001 at 05:02:09PM -0700, Jeff V. Merkey wrote:
> Please also check vger.timpanogas.org/nwfs/nwfs.tar.gz:disk.c for NetWare
> specific calculations of the CHS values, a different method is used for
> NetWare partitions vs. everything else (Novell just had to be different).
> > On Sun, Feb 25, 2001 at 04:35:34PM +0100, Andreas Jellinghaus wrote:
> >
> > > for partitions not in the first 8gb of a harddisk, what
> > > should the c/h/s start and end value be ?
> > >
> > > most fdisks seem to set start and end to 255/63/1023.
> > > but partition magic creates partitions with start set to
> > > 0/1/1023 and end set to 255/63/1023, and detects a problem
> > > if start is set to 255/63/1023.
Good. I added this to
http://www.win.tue.nl/~aeb/partitions/partition_types-2.html#above1024chs
Now that I looked at this disk.c anyway: it has a table of
partition types, and it seems I collect these.
(See http://www.win.tue.nl/~aeb/partitions/partition_types-1.html )
Are types 57 and 77, labeled "VNDI Partition", actually in use?
Andries
On Mon, 26 Feb 2001, Jeff V. Merkey wrote:
> On Sun, Feb 25, 2001 at 05:59:33PM -0800, Andre Hedrick wrote:
> >
> >
> > It does not matter because the usage of CHS will dies soon because it was
> > voted to death in Austin last week. There will only be LBA addressing
> > from now on out.
>
> If someone has Linux and NetWare dual booted on a system, and does not
> fill out the CHS fields properly for NetWare partitions, When NetWare
> boots, it will wipe the partition table (it will ask you first) and
> will not recognize any of the partitions. It does this because if it
> sees CHS values it does not expect, it assumes the partition table
> has been corrupted.
Then Netware is a bad HOST-Driver and people should expect to be hurt by
using a HOST that is not compliant. It is the responsiblity of the user
to tell the HOST-OS what it needs to do. Especially if one of the OSes
can not be intelligent enough to adpat to the changes.
Regards,
Andre Hedrick
Linux ATA Development
ASL Kernel Development
-----------------------------------------------------------------------------
ASL, Inc. Toll free: 1-877-ASL-3535
1757 Houret Court Fax: 1-408-941-2071
Milpitas, CA 95035 Web: http://www.aslab.com
On Mon, Feb 26, 2001 at 04:11:56AM +0100, Guest section DW wrote:
> On Sun, Feb 25, 2001 at 05:02:09PM -0700, Jeff V. Merkey wrote:
>
> > Please also check vger.timpanogas.org/nwfs/nwfs.tar.gz:disk.c for NetWare
> > specific calculations of the CHS values, a different method is used for
> > NetWare partitions vs. everything else (Novell just had to be different).
>
> > > On Sun, Feb 25, 2001 at 04:35:34PM +0100, Andreas Jellinghaus wrote:
> > >
> > > > for partitions not in the first 8gb of a harddisk, what
> > > > should the c/h/s start and end value be ?
> > > >
> > > > most fdisks seem to set start and end to 255/63/1023.
> > > > but partition magic creates partitions with start set to
> > > > 0/1/1023 and end set to 255/63/1023, and detects a problem
> > > > if start is set to 255/63/1023.
>
> Good. I added this to
> http://www.win.tue.nl/~aeb/partitions/partition_types-2.html#above1024chs
>
> Now that I looked at this disk.c anyway: it has a table of
> partition types, and it seems I collect these.
> (See http://www.win.tue.nl/~aeb/partitions/partition_types-1.html )
> Are types 57 and 77, labeled "VNDI Partition", actually in use?
No. They are not. 65, and 77 are the ones in use. Novell was using
67 for Wolf Mountain, but for NSS, they are exclusively using
"69" (I know who at Novell picked this number, and I'll give you
three guesses what is always on his mind).
Jeff
>
> Andries
On Sun, Feb 25, 2001 at 05:59:33PM -0800, Andre Hedrick wrote:
>
>
> It does not matter because the usage of CHS will dies soon because it was
> voted to death in Austin last week. There will only be LBA addressing
> from now on out.
If someone has Linux and NetWare dual booted on a system, and does not
fill out the CHS fields properly for NetWare partitions, When NetWare
boots, it will wipe the partition table (it will ask you first) and
will not recognize any of the partitions. It does this because if it
sees CHS values it does not expect, it assumes the partition table
has been corrupted.
Jeff
>
> On Sun, 25 Feb 2001, Jeff V. Merkey wrote:
>
> >
> > Please also check vger.timpanogas.org/nwfs/nwfs.tar.gz:disk.c for NetWare
> > specific calculations of the CHS values, a different method is used for
> > NetWare partitions vs. everything else (Novell just had to be different).
> > If you do not use their methods on NetWare partitions, NetWare will not
> > recognize the partition entries correctly, and will attempt to reinitialize
> > the entire partition table on a system if they are wrong (Ouch!). NetWare
> > does a different calculation for conversion of cylinder values above 1024.
> >
> > Jeff
> >
> >
> > ----- Original Message -----
> > From: "Guest section DW" <[email protected]>
> > To: "Andreas Jellinghaus" <[email protected]>;
> > <[email protected]>
> > Sent: Sunday, February 25, 2001 2:47 PM
> > Subject: Re: partition table: chs question
> >
> >
> > > On Sun, Feb 25, 2001 at 04:35:34PM +0100, Andreas Jellinghaus wrote:
> > >
> > > > for partitions not in the first 8gb of a harddisk, what
> > > > should the c/h/s start and end value be ?
> > > >
> > > > most fdisks seem to set start and end to 255/63/1023.
> > > > but partition magic creates partitions with start set to
> > > > 0/1/1023 and end set to 255/63/1023, and detects a problem
> > > > if start is set to 255/63/1023.
> > > >
> > > > so, how should a partition table look like ?
> > >
> > > Since these values will never be used you can imagine that they
> > > are not of great interest. Do whatever you like.
> > > -
> > > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > > the body of a message to [email protected]
> > > More majordomo info at http://vger.kernel.org/majordomo-info.html
> > > Please read the FAQ at http://www.tux.org/lkml/
> >
> > -
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to [email protected]
> > More majordomo info at http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at http://www.tux.org/lkml/
> >
>
> Andre Hedrick
> Linux ATA Development
> ASL Kernel Development
> -----------------------------------------------------------------------------
> ASL, Inc. Toll free: 1-877-ASL-3535
> 1757 Houret Court Fax: 1-408-941-2071
> Milpitas, CA 95035 Web: http://www.aslab.com
----- Original Message -----
From: "Andre Hedrick" <[email protected]>
To: "Jeff V. Merkey" <[email protected]>
Cc: "Jeff V. Merkey" <[email protected]>; "Guest section DW"
<[email protected]>; "Andreas Jellinghaus" <[email protected]>;
<[email protected]>
Sent: Monday, February 26, 2001 10:32 AM
Subject: Re: partition table: chs question
> On Mon, 26 Feb 2001, Jeff V. Merkey wrote:
>
> > On Sun, Feb 25, 2001 at 05:59:33PM -0800, Andre Hedrick wrote:
> > >
> > >
> > > It does not matter because the usage of CHS will dies soon because it
was
> > > voted to death in Austin last week. There will only be LBA addressing
> > > from now on out.
> >
> > If someone has Linux and NetWare dual booted on a system, and does not
> > fill out the CHS fields properly for NetWare partitions, When NetWare
> > boots, it will wipe the partition table (it will ask you first) and
> > will not recognize any of the partitions. It does this because if it
> > sees CHS values it does not expect, it assumes the partition table
> > has been corrupted.
>
> Then Netware is a bad HOST-Driver and people should expect to be hurt by
> using a HOST that is not compliant. It is the responsiblity of the user
> to tell the HOST-OS what it needs to do. Especially if one of the OSes
> can not be intelligent enough to adpat to the changes.
Andre,
I am not going to debate whether what they do is good or bad, there's
arguments either
way for the value in being able to detect a corrupted boot sector.
Unfortunately, this is how all the installed NetWare versions on planet
Earth behave. If someone is rearranging a partition table and fills out
these values wrong, it will make the drive unreadable to NetWare, and make
it look like Linux just corrupted the drive (since Linux will get blamed).
The disk
tools should observe each OS's weirdness if it may potentially change values
in this
table and cause data loss.
Jeff
>
> Regards,
>
> Andre Hedrick
> Linux ATA Development
> ASL Kernel Development
> --------------------------------------------------------------------------
---
> ASL, Inc. Toll free: 1-877-ASL-3535
> 1757 Houret Court Fax: 1-408-941-2071
> Milpitas, CA 95035 Web: http://www.aslab.com
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
On Mon, Feb 26, 2001 at 11:24:07AM -0700, Jeff V. Merkey wrote:
> On Mon, Feb 26, 2001 at 04:11:56AM +0100, Guest section DW wrote:
> > (See http://www.win.tue.nl/~aeb/partitions/partition_types-1.html )
> > Are types 57 and 77, labeled "VNDI Partition", actually in use?
>
> No. They are not. 65, and 77 are the ones in use. Novell was using
> 67 for Wolf Mountain, but for NSS, they are exclusively using 69.
Wait! 57 and 77 are not in use, but 65 and 77 are?
(Is the second 77 a typo for 66 or 67?)
A lot of partition IDs are attributed to Novell.
64 (Netware 286) and 65 (Netware 386) are well established, and
you tell me 66 is Netware SMS, 67 is Wolf Mountain and 69 is Netware NSS.
But also 51 and 68 occur in reports. Do you know anything about those?
Andries
----- Original Message -----
From: "Guest section DW" <[email protected]>
To: "Jeff V. Merkey" <[email protected]>
Cc: "Jeff V. Merkey" <[email protected]>; "Andreas Jellinghaus"
<[email protected]>; <[email protected]>; <[email protected]>
Sent: Monday, February 26, 2001 1:11 PM
Subject: Re: partition table: chs question
> On Mon, Feb 26, 2001 at 11:24:07AM -0700, Jeff V. Merkey wrote:
> > On Mon, Feb 26, 2001 at 04:11:56AM +0100, Guest section DW wrote:
>
> > > (See http://www.win.tue.nl/~aeb/partitions/partition_types-1.html )
> > > Are types 57 and 77, labeled "VNDI Partition", actually in use?
> >
> > No. They are not. 65, and 77 are the ones in use. Novell was using
> > 67 for Wolf Mountain, but for NSS, they are exclusively using 69.
>
> Wait! 57 and 77 are not in use, but 65 and 77 are?
> (Is the second 77 a typo for 66 or 67?)
>
> A lot of partition IDs are attributed to Novell.
> 64 (Netware 286) and 65 (Netware 386) are well established, and
> you tell me 66 is Netware SMS, 67 is Wolf Mountain and 69 is Netware NSS.
> But also 51 and 68 occur in reports. Do you know anything about those?
Novell is assigned block 60-70 by some committee way back in the mid-1980's.
51 is not legal for them, I would like to look at this one. Any numbers in
range 60-70 they can and do use, though most are experimental. 66 and 64
are no longer used, but older stuff may use them.
77 is one we are using internally for M2FS/M2CS partitions. VNDI can be
dropped.
:-)
Jeff
>
> Andries