[PATCH 0/2] mtd: spi-nor: Fix Quad Enable method for MX25L12835F

[Michael Walle]

Am 2022-03-01 13:46, schrieb Tudor.Ambarus at microchip.com:
> On 3/1/22 14:36, Michael Walle wrote:
>> EXTERNAL EMAIL: Do not click links or open attachments unless you know 
>> the content is safe
>> 
>> Am 2022-03-01 11:52, schrieb Tudor.Ambarus at microchip.com:
>>> On 3/1/22 12:38, Michael Walle wrote:
>>>> EXTERNAL EMAIL: Do not click links or open attachments unless you 
>>>> know
>>>> the content is safe
>>>> 
>>>> Am 2022-03-01 11:29, schrieb Heiko Thiery:
>>>>> Hi Tudor,
>>>>> 
>>>>> Am Di., 1. März 2022 um 10:56 Uhr schrieb Tudor Ambarus
>>>>> <tudor.ambarus at microchip.com>:
>>>>>> 
>>>>>> Hi, Heiko,
>>>>>> 
>>>>>> Would you please apply this patch set on top of:
>>>>>> https://lore.kernel.org/linux-mtd/20220228134505.203270-1-tudor.ambarus@microchip.com/
>>>>> 
>>>>> When applying that series the flash will be properly detected.
>>>>> 
>>>>> [    1.322655] spi-nor spi0.0: mx25l12835f (16384 Kbytes)
>>>>> 
>>>>> Thanks
>>>> 
>>>> But doesn't that mean that the previous series introduces
>>>> possible regressions for other flashes, too?
>>>> 
>>> no, because the other flashes are using the deprecated way of
>>> initializing
>>> parameters which calls default_init() hooks.
>> 
>> ahh right!
>> 
>> But we should clear the quad_enable in the case the SFDP doesn't
>> specify it. Right now, we are falling back to a function
>> which doesn't make sense, and might even be harmful.
>> 
> 
> do you mean the one set in the default_init() hook for macronix? we 
> should
> get rid of the default_init(), yes. We should use SFDP where possible, 
> where
> not possible we should use the late_init() hook to set the Quad Enable 
> method.

The one we set in spi_nor_init_default_params(), is that planned to be 
removed,
too?

I was thinking about something along the following patch:
https://lore.kernel.org/linux-mtd/20220301124935.2893622-1-michael@walle.cc/

-michael