[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <871psb1zte.fsf@bootlin.com>
Date: Mon, 26 May 2025 11:33:33 +0200
From: Miquel Raynal <miquel.raynal@...tlin.com>
To: Álvaro Fernández Rojas <noltari@...il.com>
Cc: linux-mtd@...ts.infradead.org, dregan@...adcom.com,
bcm-kernel-feedback-list@...adcom.com, florian.fainelli@...adcom.com,
rafal@...ecki.pl, computersforpeace@...il.com, kamal.dasu@...adcom.com,
dan.beygelman@...adcom.com, william.zhang@...adcom.com,
frieder.schrempf@...tron.de, linux-kernel@...r.kernel.org,
vigneshr@...com, richard@....at, bbrezillon@...nel.org,
kdasu.kdev@...il.com, jaimeliao.tw@...il.com, kilobyte@...band.pl,
jonas.gorski@...il.com, dgcbueu@...il.com
Subject: Re: [PATCH v5] mtd: rawnand: brcmnand: legacy exec_op implementation
Hello,
>> AFAIK, the legacy functions were only using it for
>> NAND_CMD_SET_FEATURES, which we don't support:
>> https://github.com/torvalds/linux/blob/c86b63b82fde4f96ee94dde827a5f28ff5adeb57/drivers/mtd/nand/raw/brcmnand/brcmnand.c#L1922-L1938
>>
>> The other uses I could find are already covered by our
>> chip->ecc.read/write functions.
>>
>> In any case I've tested the patch for reading, erasing and writing the
>> NAND and so far I haven't found any unsupported error apart from
>> NAND_CMD_GET_FEATURES with a Macronix NAND in the Sercom H500-s
>> (BCM63268).
>> I believe it's used for unlocking the NAND, which isn't needed in that
>> device.
>
> Well, you are restoring an old behavior so I won't ask for a better
> support, but you should normally allow software ECC engines (and even no
> engine at all) and in this case the core will require a write path. I
> honestly think it is not very complex to implement but if someone is
> lacking this feature it can be added later.
>
> Please just fix the braces in the for loop that was reported, but no
> hurry, I'll only take this after -rc1.
Nevermind, I'm applying now. You can send a follow-up patch for -rc1 if
you want.
Powered by blists - more mailing lists