Can mtd partition span multiple devices

Can mtd partition span multiple devices

[John Burch]

I'm wondering if a single mtd partition can span more than one physical
flash device.  For example, if two flash devices span the following
address ranges, 0x0-0x3FFFFF and 0x400000-0x7FFFFF, can an mtd partition
be defined as follows?

Name: spanning_partition
Size: 0x200000
Offset: 0x300000

So this partition would span physical addresses 0x300000 - 0x3FFFFF
(device 1) and 0x400000 - 0x4FFFFF (device 2).


John

Re: Can mtd partition span multiple devices

[Thomas Gleixner]

On Thursday 03 April 2003 00:56, John Burch wrote:
> I'm wondering if a single mtd partition can span more than one physical
> flash device.  For example, if two flash devices span the following
> address ranges, 0x0-0x3FFFFF and 0x400000-0x7FFFFF, can an mtd partition
> be defined as follows?
>
> Name: spanning_partition
> Size: 0x200000
> Offset: 0x300000
>
> So this partition would span physical addresses 0x300000 - 0x3FFFFF
> (device 1) and 0x400000 - 0x4FFFFF (device 2).

Enable MTD concatenating support.

-- 
Thomas
________________________________________________________________________
linutronix - competence in embedded & realtime linux
http://www.linutronix.de
mail: tglx@linutronix.de

Re: Can mtd partition span multiple devices

[Charles Manning]

I believe mtdconcat will do what you want.

On Thu, 03 Apr 2003 10:56, John Burch wrote:
> I'm wondering if a single mtd partition can span more than one physical
> flash device.  For example, if two flash devices span the following
> address ranges, 0x0-0x3FFFFF and 0x400000-0x7FFFFF, can an mtd partition
> be defined as follows?
>
> Name: spanning_partition
> Size: 0x200000
> Offset: 0x300000
>
> So this partition would span physical addresses 0x300000 - 0x3FFFFF
> (device 1) and 0x400000 - 0x4FFFFF (device 2).
>
>
> John
>
>
> ______________________________________________________
> Linux MTD discussion mailing list
> http://lists.infradead.org/mailman/listinfo/linux-mtd/

Re: Can mtd partition span multiple devices

[Kenneth Johansson]

On Thu, 2003-04-03 at 02:58, Thomas Gleixner wrote:
> On Thursday 03 April 2003 00:56, John Burch wrote:
> > I'm wondering if a single mtd partition can span more than one physical
> > flash device.  For example, if two flash devices span the following
> > address ranges, 0x0-0x3FFFFF and 0x400000-0x7FFFFF, can an mtd partition
> > be defined as follows?
> >
> > Name: spanning_partition
> > Size: 0x200000
> > Offset: 0x300000
> >
> > So this partition would span physical addresses 0x300000 - 0x3FFFFF
> > (device 1) and 0x400000 - 0x4FFFFF (device 2).
> 
> Enable MTD concatenating support.
> 

You mean MTD partitioning no need for concatenating with this layout. 

-- 
Kenneth Johansson	
Ericsson AB                       Tel: +46 8 719 70 20
Tellusborgsvägen  90              Fax: +46 8 719 29 45
126 25 Stockholm                  ken@switchboard.ericsson.se

Re: Can mtd partition span multiple devices

[David Woodhouse]

On Thu, 2003-04-03 at 12:15, Kenneth Johansson wrote:
> On Thu, 2003-04-03 at 02:58, Thomas Gleixner wrote:
> > On Thursday 03 April 2003 00:56, John Burch wrote:
> > > I'm wondering if a single mtd partition can span more than one physical
> > > flash device.  For example, if two flash devices span the following
> > > address ranges, 0x0-0x3FFFFF and 0x400000-0x7FFFFF, can an mtd partition
> > > be defined as follows?
> > >
> > > Name: spanning_partition
> > > Size: 0x200000
> > > Offset: 0x300000
> > >
> > > So this partition would span physical addresses 0x300000 - 0x3FFFFF
> > > (device 1) and 0x400000 - 0x4FFFFF (device 2).
> > 
> > Enable MTD concatenating support.
> > 
> 
> You mean MTD partitioning no need for concatenating with this layout. 

Probably true -- if the two physical devices are laid out such that a
single chip probe will find them together, they can be used as a single
MTD device even without mtdconcat just by setting your map to cover the
whole range 0->8M. You only really need mtdconcat for more complicated
setups.

-- 
dwmw2

RE: Can mtd partition span multiple devices

[John Burch]

> 
> On Thu, 2003-04-03 at 12:15, Kenneth Johansson wrote:
> > On Thu, 2003-04-03 at 02:58, Thomas Gleixner wrote:
> > > On Thursday 03 April 2003 00:56, John Burch wrote:
> > > > I'm wondering if a single mtd partition can span more than one 
> > > > physical flash device.  For example, if two flash 
> devices span the 
> > > > following address ranges, 0x0-0x3FFFFF and 
> 0x400000-0x7FFFFF, can 
> > > > an mtd partition be defined as follows?
> > > >
> > > > Name: spanning_partition
> > > > Size: 0x200000
> > > > Offset: 0x300000
> > > >
> > > > So this partition would span physical addresses 0x300000 - 
> > > > 0x3FFFFF (device 1) and 0x400000 - 0x4FFFFF (device 2).
> > > 
> > > Enable MTD concatenating support.
> > > 
> > 
> > You mean MTD partitioning no need for concatenating with 
> this layout.
> 
> Probably true -- if the two physical devices are laid out 
> such that a single chip probe will find them together, they 
> can be used as a single MTD device even without mtdconcat 
> just by setting your map to cover the whole range 0->8M. You 
> only really need mtdconcat for more complicated setups.
> 

I'm using 2.4.17 and I don't think mtdconcat is part of that kernel
(though I may not need it anyway based on dwmw2's response).  Are the
changes since then substantial, and can 2.4.17 be patched to support the
latest mtd/jffs2 code?

Second, you imply that a single chip probe will handle multiple devices
(assuming contiguous addressing?).  Is this supported in 2.4.17 and is
it a common approach to use for multiple devices?

John

Re: Can mtd partition span multiple devices

[Jörn Engel]

On Thu, 3 April 2003 10:36:54 -0500, John Burch wrote:
> 
> I'm using 2.4.17 and I don't think mtdconcat is part of that kernel
> (though I may not need it anyway based on dwmw2's response).  Are the
> changes since then substantial, and can 2.4.17 be patched to support the
> latest mtd/jffs2 code?

I wouldn't patch the mtd/jffs2 part, unless I had to. The better
strategy is to update the whole kernel, but that is always a bit
tricky inside some companies.

> Second, you imply that a single chip probe will handle multiple devices
> (assuming contiguous addressing?).  Is this supported in 2.4.17 and is
> it a common approach to use for multiple devices?

That support has been in the official kernel since 2.4.2 or even
longer.

Jörn

-- 
Measure. Don't tune for speed until you've measured, and even then
don't unless one part of the code overwhelms the rest.
-- Rob Pike