Implementing OTP

Implementing OTP

[Tom Tatakis]

Hello,

I have implemented OTP functionality in the NAND driver for the
Macronix MX30LF1G18AC device.  I would like to send it on up in a
patch but I'd like to confirm that what I did makes sense.  I modified
 nand_base.c file to include the requisite functions for reading and
writing the OTP.  I also modified the nand_scan_tail function to
assign the names of the functions in the mtd structure.  The potential
issue is that the ONFI command used to read and write the OTP uses a
feature address of 0x90 which is defined as part of the 'Vendor
Specific' area if I understand the ONFI spec (it's entirely possible
that I do not).  Specifically, section 5.30 in Version 4.0 of the ONFI
spec.

If my modification of nand_base.c to include this makes sense, then I
can put together a patch.

I did try and figure out a way to get this potentially device
dependent code into its own file, but I do not see how to do that
other than creating a whole other driver for it complete with a
configuration setting.  If that's what I need to do then I will try
and put it together.  If there is a better way and you can point me
int the right direction, that would be helpful as well.

Thanks for you time
Tom

Re: Implementing OTP

[Boris Brezillon]

Hi Tom,

Sorry for the late reply (remember to also Cc the NAND and MTD
maintainers next time).

On Thu, 5 May 2016 13:35:32 -0400
Tom Tatakis <ttatakis@appliedlogix.com> wrote:

> Hello,
> 
> I have implemented OTP functionality in the NAND driver for the
> Macronix MX30LF1G18AC device.  I would like to send it on up in a
> patch but I'd like to confirm that what I did makes sense.  I modified
>  nand_base.c file to include the requisite functions for reading and
> writing the OTP.  I also modified the nand_scan_tail function to
> assign the names of the functions in the mtd structure.  The potential
> issue is that the ONFI command used to read and write the OTP uses a
> feature address of 0x90 which is defined as part of the 'Vendor
> Specific' area if I understand the ONFI spec (it's entirely possible
> that I do not).  Specifically, section 5.30 in Version 4.0 of the ONFI
> spec.
> 
> If my modification of nand_base.c to include this makes sense, then I
> can put together a patch.
> 
> I did try and figure out a way to get this potentially device
> dependent code into its own file, but I do not see how to do that
> other than creating a whole other driver for it complete with a
> configuration setting.  If that's what I need to do then I will try
> and put it together.  If there is a better way and you can point me
> int the right direction, that would be helpful as well.

You might want to have a look at [1] ;).

Regards,

Boris

[1]https://lkml.org/lkml/2016/5/27/264

-- 
Boris Brezillon, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com