Re: [PATCH 1/3] mtd: nand: Add on-die ECC support

[Richard Weinberger <richard@nod.at>]

Am 08.05.2015 um 23:39 schrieb Brian Norris:
> On Fri, May 08, 2015 at 04:26:32PM -0500, Ben Shelton wrote:
>> On 04/27, Brian Norris wrote:
>>> On Tue, Apr 28, 2015 at 08:18:12AM +0530, punnaiah choudary kalluri wrote:
>>>> On Tue, Apr 28, 2015 at 4:53 AM, Brian Norris
>>>> <computersforpeace@gmail.com> wrote:
>>>>> On Tue, Apr 28, 2015 at 12:19:16AM +0200, Richard Weinberger wrote:
>>>>>> Oh, I thought every driver has to implement that function. ;-\
>>>>>
>>>>> Nope.
>>>>>
>>>>>> But you're right there is a corner case.
>>>>>
>>>>> And it's not the only one! Right now, there's no guarantee even that
>>>>> read_buf() returns raw data, unmodified by the SoC's controller. Plenty
>>>>> of drivers actually have HW-enabled ECC turned on by default, and so
>>>>> they override the chip->ecc.read_page() (and sometimes
>>>>> chip->ecc.read_page_raw() functions, if we're lucky) with something
>>>>> that pokes the appropriate hardware instead. I expect anything
>>>>> comprehensive here is probably going to have to utilize
>>>>> chip->ecc.read_page_raw(), at least if it's provided by the hardware
>>>>> driver.
>>>>
>>>> Yes, overriding the chip->ecc.read_page_raw would solve this.
>>>
>>> I'm actually suggesting that (in this patch set, for on-die ECC
>>> support), maybe we *shouldn't* override chip->ecc.read_page_raw() and
>>> leave that to be defined by the driver, and then on-die ECC support
>>> should be added in a way that just calls chip->ecc.read_page_raw(). This
>>> should work for any driver that already properly supports the raw
>>> callbacks.
>>>
>>
>> Hi Richard et al,
>>
>> I'm guessing it's probably too late for the on-die ECC stuff to land in
>> 4.2 at this point.
> 
> Not technically. We've got several weeks (approx 5 to 6?) before 4.1 is
> released. 4.2 material should be getting finalized by a week or so
> before the merge window (i.e., 4 to 5 weeks from now).
> 
>> Is there anything I can do to help this along
>> (testing, etc.)?
> 
> This is going to need to get rewritten. I'm not sure if Richard is going
> to tackle this again, as he hasn't responded to the points I brought up.
> (Note that Richard is not the first to have tried to implement this,
> without initial success.)

I'm definitely willing to take the challenge.
But as I'm currently very busy with non-MTD stuff I had no time
to address your comments.

Thanks,
//richard