Re: [PATCH v4 6/9] mtd: rawnand: ingenic: Separate top-level and SoC specific code

From: Paul Cercueil <paul@crapouillou.net>
To: Miquel Raynal <miquel.raynal@bootlin.com>
Cc: Mark Rutland <mark.rutland@arm.com>,
	devicetree@vger.kernel.org,
	Boris Brezillon <bbrezillon@kernel.org>,
	Richard Weinberger <richard@nod.at>,
	linux-kernel@vger.kernel.org, Marek Vasut <marek.vasut@gmail.com>,
	Rob Herring <robh+dt@kernel.org>,
	linux-mtd@lists.infradead.org,
	Harvey Hunt <harveyhuntnexus@gmail.com>,
	Brian Norris <computersforpeace@gmail.com>,
	David Woodhouse <dwmw2@infradead.org>
Subject: Re: [PATCH v4 6/9] mtd: rawnand: ingenic: Separate top-level and SoC specific code
Date: Mon, 04 Mar 2019 19:26:08 +0100	[thread overview]
Message-ID: <1551723968.4932.4@crapouillou.net> (raw)
In-Reply-To: <20190304112058.67cd5194@xps13>

On Mon, Mar 4, 2019 at 11:20 AM, Miquel Raynal 
<miquel.raynal@bootlin.com> wrote:
> Hi Paul,
> 
> Paul Cercueil <paul@crapouillou.net <mailto:paul@crapouillou.net>> 
> wrote on Sat,  9 Feb 2019 16:23:02
> -0300:
> 
>>  The ingenic-nand driver uses an API provided by the jz4780-bch 
>> driver.
>>  This makes it difficult to support other SoCs in the jz4780-bch 
>> driver.
>>  To work around this, we separate the API functions from the 
>> SoC-specific
>>  code, so that these API functions are SoC-agnostic.
>> 
> 
> I like the idea, actually I am working on this separation (see
> [1]) and I would really appreciate that you try to implement the
> interface when it will be available (v2 is coming this week, I think 
> v3
> will be the one to test when raw NAND devices will be properly
> supported). I will add you in Cc: if you want to follow/review.
> 
> [1] 
> <http://lists.infradead.org/pipermail/linux-mtd/2019-February/087815.html>

Do you think this will be ready for 5.2?

You can add me in Cc:.

>>  Signed-off-by: Paul Cercueil <paul@crapouillou.net 
>> <mailto:paul@crapouillou.net>>
>>  ---
>> 
>>  v2: Add an optional .probe() callback. It is used for instance to 
>> set
>>      the clock rate in the JZ4780 backend.
>> 
>>  v3: The common code is now inside the ingenic-ecc module. Each
>>      SoC-specific ECC code is now in its own module, which leaves to 
>> the
>>      user the choice of which (if any) ECC code should be supported.
>> 
>>  v4: No change
>> 
>>   drivers/mtd/nand/raw/ingenic/Kconfig        |  17 +++
>>   drivers/mtd/nand/raw/ingenic/Makefile       |   5 +-
>>   drivers/mtd/nand/raw/ingenic/ingenic_ecc.c  | 157 
>> +++++++++++++++++++++++++
>>   drivers/mtd/nand/raw/ingenic/ingenic_ecc.h  |  84 ++++++++++++++
>>   drivers/mtd/nand/raw/ingenic/ingenic_nand.c |  38 +++----
>>   drivers/mtd/nand/raw/ingenic/jz4780_bch.c   | 170 
>> +++++-----------------------
>>   drivers/mtd/nand/raw/ingenic/jz4780_bch.h   |  40 -------
>>   7 files changed, 308 insertions(+), 203 deletions(-)
>>   create mode 100644 drivers/mtd/nand/raw/ingenic/ingenic_ecc.c
>>   create mode 100644 drivers/mtd/nand/raw/ingenic/ingenic_ecc.h
>>   delete mode 100644 drivers/mtd/nand/raw/ingenic/jz4780_bch.h
>> 
> 
> [...]
> 
>>  diff --git a/drivers/mtd/nand/raw/ingenic/ingenic_nand.c 
>> b/drivers/mtd/nand/raw/ingenic/ingenic_nand.c
>>  index 8c73f7c5be9a..0f51fd15fe79 100644
>>  --- a/drivers/mtd/nand/raw/ingenic/ingenic_nand.c
>>  +++ b/drivers/mtd/nand/raw/ingenic/ingenic_nand.c
>>  @@ -22,7 +22,7 @@
>> 
>>   #include <linux/jz4780-nemc.h>
>> 
>>  -#include "jz4780_bch.h"
>>  +#include "ingenic_ecc.h"
>> 
>>   #define DRV_NAME	"ingenic-nand"
>> 
>>  @@ -40,7 +40,7 @@ struct ingenic_nand_cs {
>> 
>>   struct ingenic_nfc {
>>   	struct device *dev;
>>  -	struct jz4780_bch *bch;
>>  +	struct ingenic_ecc *ecc;
>>   	struct nand_controller controller;
>>   	unsigned int num_banks;
>>   	struct list_head chips;
>>  @@ -124,10 +124,10 @@ static int ingenic_nand_ecc_calculate(struct 
>> nand_chip *chip, const u8 *dat,
>>   {
>>   	struct ingenic_nand *nand = to_ingenic_nand(nand_to_mtd(chip));
>>   	struct ingenic_nfc *nfc = to_ingenic_nfc(nand->chip.controller);
>>  -	struct jz4780_bch_params params;
>>  +	struct ingenic_ecc_params params;
>> 
>>   	/*
>>  -	 * Don't need to generate the ECC when reading, BCH does it for 
>> us as
>>  +	 * Don't need to generate the ECC when reading, ECC does it for 
>> us as
> 
> "the ECC engine does it for us" would be more meaningful.

OK.

>>   	 * part of decoding/correction.
>>   	 */
>>   	if (nand->reading)
>>  @@ -137,7 +137,7 @@ static int ingenic_nand_ecc_calculate(struct 
>> nand_chip *chip, const u8 *dat,
>>   	params.bytes = nand->chip.ecc.bytes;
>>   	params.strength = nand->chip.ecc.strength;
>> 
>>  -	return jz4780_bch_calculate(nfc->bch, &params, dat, ecc_code);
>>  +	return ingenic_ecc_calculate(nfc->ecc, &params, dat, ecc_code);
>>   }
>> 
> 
> Thanks,
> Miquèl

______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/