Re: [PATCH] Revert "mtd: spinand: esmt: fix id code for F50D1G41LB"

[Ziyang Huang <hzyitc@outlook.com>]

在 2025/12/8 16:33, Miquel Raynal 写道:
> On 02/12/2025 at 23:05:56 +08, Ziyang Huang <hzyitc@outlook.com> wrote:
> 
>> This reverts commit dd26402642a0899fde59ea6b0852fad3d799b4cc.
>>
>> The issue George met is due to the limit of QPIC, not the issue of the
>> flash chip.
>>
>> QPIC only supports 4 bytes ID. So the fifth byte is always 0.
>>
>> If we use spi-gpio, the fifth byte can be read correctly.
>>
>> Signed-off-by: Ziyang Huang <hzyitc@outlook.com>
>> ---
> 
> So how do we solve that?
> 
> Is there anything than can be done on the qpic side?
> 
> Thanks,
> Miquèl

Unfortunately, There is not a good way to solve it now.

In Openwrt, I made a hack to force the max compare bytes to 4.

I'm trying to find a way to let QPIC do a raw SPI op. But without 
documents, it's hard.

For router devices, an other typical way is to use parameter page. This 
feature is supported in almost all of the high-speed QSPI NANDs they used.

But for now, 4 bytes is enough for the router devices since only a few 
flash models will be used. More the that, some vendor SDKs use 2 bytes.