Porting to Intel P33 flash

Porting to Intel P33 flash

[jkimble]

Does anyone know at what kernel release this particular flash (Intel
256P33B StrataFlash) was supported (if any)?

I can't go to the latest kernel and back porting the MTD section form
2.6.24 to 2.6.10 turned out to be impractical. There's just too many basic
structures (like devices Vs platform_devices) that I'm not prepared to
change in 2.6.10.

I just need to know when/if P33 was supported. It would be good to have a
version I know works to work from.

Thanks,

James Kimble

Re: Porting to Intel P33 flash

[Alexey Korolev]

Hi,

> Does anyone know at what kernel release this particular flash (Intel
> 256P33B StrataFlash) was supported (if any)?
> 
> I can't go to the latest kernel and back porting the MTD section form
> 2.6.24 to 2.6.10 turned out to be impractical. There's just too many basic
> structures (like devices Vs platform_devices) that I'm not prepared to
> change in 2.6.10.
> 
> I just need to know when/if P33 was supported. It would be good to have a
> version I know works to work from.
>


There are two versions of P33 one has version of CFI 1.4 another version
1.5 

P33 with CFI 1.4 works Ok starting 2.6.17 
P33 with CFI 1.5 works Ok starting 2.6.24 

Since the difference between P33 with CFI 1.4 and P33 with CFI 1.5 are
minor applying the following very simple patch:
http://git.infradead.org/mtd-2.6.git?a=commitdiff;h=b1c9c9be6da010510459aca93f5754efb19695ff 
should solve the problem for linux >= 2.6.17.

Thanks,
Alexey

Re: Porting to Intel P33 flash

[jkimble]

Thank you very, very much! At least that gives me a place to start from
with my 2.6.10 kernel.

James

>
>
> There are two versions of P33 one has version of CFI 1.4 another version
> 1.5
>
> P33 with CFI 1.4 works Ok starting 2.6.17
> P33 with CFI 1.5 works Ok starting 2.6.24
>
> Since the difference between P33 with CFI 1.4 and P33 with CFI 1.5 are
> minor applying the following very simple patch:
> http://git.infradead.org/mtd-2.6.git?a=commitdiff;h=b1c9c9be6da010510459aca93f5754efb19695ff
> should solve the problem for linux >= 2.6.17.
>
> Thanks,
> Alexey
>
> ______________________________________________________
> Linux MTD discussion mailing list
> http://lists.infradead.org/mailman/listinfo/linux-mtd/
>
>

Re: Porting to Intel P33 flash

[Hamish Moffatt]

On Wed, May 14, 2008 at 06:06:35PM +0100, Alexey Korolev wrote:
> There are two versions of P33 one has version of CFI 1.4 another version
> 1.5 
> 
> P33 with CFI 1.4 works Ok starting 2.6.17 
> P33 with CFI 1.5 works Ok starting 2.6.24 
> 
> Since the difference between P33 with CFI 1.4 and P33 with CFI 1.5 are
> minor applying the following very simple patch:
> http://git.infradead.org/mtd-2.6.git?a=commitdiff;h=b1c9c9be6da010510459aca93f5754efb19695ff 
> should solve the problem for linux >= 2.6.17.

I've been running 2.6.15 with P33 CFI 1.5 without issue. Almost
completely read-only operation though in my application.

Hamish
-- 
Hamish Moffatt VK3SB <hamish@debian.org> <hamish@cloud.net.au>

Re: Porting to Intel P33 flash

[jkimble]

Funny you say that...  I just modified the code where it checks CFI
version  and it seems to be working without further modification. I've not
tested much though so I may have more to do. I do have to write to this
but not much.

James

>
> I've been running 2.6.15 with P33 CFI 1.5 without issue. Almost
> completely read-only operation though in my application.
>
> Hamish
> --
> Hamish Moffatt VK3SB <hamish@debian.org> <hamish@cloud.net.au>
>
> ______________________________________________________
> Linux MTD discussion mailing list
> http://lists.infradead.org/mailman/listinfo/linux-mtd/
>
>

Re: Porting to Intel P33 flash

[Alexey Korolev]

Hi

> 
> Funny you say that...  I just modified the code where it checks CFI
> version  and it seems to be working without further modification. I've not
> tested much though so I may have more to do. I do have to write to this
> but not much.
> 
> James
It is good to hear this. Please check number of RWW partitons
(numvirtchips variable) and 9th bit of extp->FeatureSupport. 
Everything is Ok if numvirtchips == 1 or extp->FeatureSupport bit 9 is 0.

As I remember some versions of MTD prior to 2.6.17 had
problems with CFI offset calculation. It led chip to be identified as
multiple partitioned device when it should have 1 partiton. 

Thanks,
Alexey

Re: Porting to Intel P33 flash

[Hamish Moffatt]

On Fri, May 16, 2008 at 01:59:43PM +0100, Alexey Korolev wrote:
> Hi
> 
> > 
> > Funny you say that...  I just modified the code where it checks CFI
> > version  and it seems to be working without further modification. I've not
> > tested much though so I may have more to do. I do have to write to this
> > but not much.
> > 
> > James
> It is good to hear this. Please check number of RWW partitons
> (numvirtchips variable) and 9th bit of extp->FeatureSupport. 
> Everything is Ok if numvirtchips == 1 or extp->FeatureSupport bit 9 is 0.
> 
> As I remember some versions of MTD prior to 2.6.17 had
> problems with CFI offset calculation. It led chip to be identified as
> multiple partitioned device when it should have 1 partiton. 

How do you mean multiple partitions?

Here is what I see with the 8Mb P33 NOR flash with 2.6.15:

IXP4XX-Flash.0: Found 1 x16 devices at 0x0 in 16-bit bank
 Intel/Sharp Extended Query Table at 0x010A
 Intel/Sharp Extended Query Table at 0x010A
 Intel/Sharp Extended Query Table at 0x010A
 Intel/Sharp Extended Query Table at 0x010A
 Intel/Sharp Extended Query Table at 0x010A
Using buffer write method
cfi_cmdset_0001: Erase suspend on write enabled
erase region 0: offset=0x0,size=0x20000,blocks=63
erase region 1: offset=0x7e0000,size=0x8000,blocks=4

The output with 2.6.24 is the same. I only modified 2.6.15 to recognise
the new minor revision of the P33 flash.

Anyway 2.6.15 is history for me now (we ran it for 1 year but just
upgraded to 2.6.24), I only mention it for James's information.

Hamish
-- 
Hamish Moffatt VK3SB <hamish@debian.org> <hamish@cloud.net.au>