linux-kernel - Re: [PATCH v2 4/8] mtd: spi-nor: core: Introduce method for RDID op

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Mon, 21 Mar 2022 13:18:48 +0000
From:   <Tudor.Ambarus@...rochip.com>
To:     <p.yadav@...com>
CC:     <michael@...le.cc>, <miquel.raynal@...tlin.com>, <richard@....at>,
        <vigneshr@...com>, <linux-mtd@...ts.infradead.org>,
        <linux-kernel@...r.kernel.org>, <Nicolas.Ferre@...rochip.com>
Subject: Re: [PATCH v2 4/8] mtd: spi-nor: core: Introduce method for RDID op

On 3/21/22 14:21, Pratyush Yadav wrote:
> EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
> 
> On 28/02/22 01:17PM, Tudor Ambarus wrote:
>> RDID is used in the core to auto detect the flash, but also by some
>> manufacturer drivers that contain flashes that support Octal DTR mode,
>> so that they can read the flash ID after the switch to Octal DTR was made
>> to test if the switch was successful. Introduce a core method for RDID op
>> to avoid code duplication.
>>
>> Signed-off-by: Tudor Ambarus <tudor.ambarus@...rochip.com>
>> ---
>>  drivers/mtd/spi-nor/core.c | 58 ++++++++++++++++++++++++++------------
>>  drivers/mtd/spi-nor/core.h |  9 ++++++
>>  2 files changed, 49 insertions(+), 18 deletions(-)
>>
>> diff --git a/drivers/mtd/spi-nor/core.c b/drivers/mtd/spi-nor/core.c
>> index b1d6fa65417d..281e3d25f74c 100644
>> --- a/drivers/mtd/spi-nor/core.c
>> +++ b/drivers/mtd/spi-nor/core.c
>> @@ -369,6 +369,41 @@ int spi_nor_write_disable(struct spi_nor *nor)
>>       return ret;
>>  }
>>
>> +/**
>> + * spi_nor_read_id() - Read the JEDEC ID.
>> + * @nor:     pointer to 'struct spi_nor'.
>> + * @naddr:   number of address bytes to send. Can be zero if the operation
>> + *           does not need to send an address.
>> + * @ndummy:  number of dummy bytes to send after an opcode or address. Can
>> + *           be zero if the operation does not require dummy bytes.
>> + * @id:              pointer to a DMA-able buffer where the value of the JEDEC ID
>> + *           will be written.
>> + * @reg_proto:       the SPI protocol for register operation.
>> + *
>> + * Return: 0 on success, -errno otherwise.
>> + */
>> +int spi_nor_read_id(struct spi_nor *nor, u8 naddr, u8 ndummy, u8 *id,
>> +                 enum spi_nor_protocol reg_proto)
> 
> Nitpick: Could just call it 'proto'.

sure, will update

> 
>> +{
>> +     int ret;
>> +
>> +     if (nor->spimem) {
>> +             struct spi_mem_op op =
>> +                     SPI_NOR_READID_OP(naddr, ndummy, id, SPI_NOR_MAX_ID_LEN);
>> +
>> +             spi_nor_spimem_setup_op(nor, &op, reg_proto);
>> +             ret = spi_mem_exec_op(nor->spimem, &op);
>> +     } else {
>> +             ret = nor->controller_ops->read_reg(nor, SPINOR_OP_RDID, id,
>> +                                                 SPI_NOR_MAX_ID_LEN);
>> +     }
>> +
>> +     if (ret)
>> +             dev_dbg(nor->dev, "error %d reading JEDEC ID\n", ret);
> 
> I think this message should be in spi_nor_detect(). Let octal DTR enable

As of now every SPI NOR operation that return an error also prints a dbg
message. I like this because it offers a smaller granularity on the error
cause.

> methods print their own, more specific error messages.

How about duplicating the error in the octal dtr enable methods if you
feel it is worth it?

> 
>> +
>> +     return ret;
>> +}
>> +
>>  /**
>>   * spi_nor_read_sr() - Read the Status Register.
>>   * @nor:     pointer to 'struct spi_nor'.
>> @@ -1649,28 +1684,15 @@ static const struct flash_info *spi_nor_match_id(struct spi_nor *nor,
>>       return NULL;
>>  }
>>
>> -static const struct flash_info *spi_nor_read_id(struct spi_nor *nor)
>> +static const struct flash_info *spi_nor_detect(struct spi_nor *nor)
>>  {
>>       const struct flash_info *info;
>>       u8 *id = nor->bouncebuf;
>>       int ret;
>>
>> -     if (nor->spimem) {
>> -             struct spi_mem_op op =
>> -                     SPI_MEM_OP(SPI_MEM_OP_CMD(SPINOR_OP_RDID, 1),
>> -                                SPI_MEM_OP_NO_ADDR,
>> -                                SPI_MEM_OP_NO_DUMMY,
>> -                                SPI_MEM_OP_DATA_IN(SPI_NOR_MAX_ID_LEN, id, 1));
>> -
>> -             ret = spi_mem_exec_op(nor->spimem, &op);
>> -     } else {
>> -             ret = nor->controller_ops->read_reg(nor, SPINOR_OP_RDID, id,
>> -                                                 SPI_NOR_MAX_ID_LEN);
>> -     }
>> -     if (ret) {
>> -             dev_dbg(nor->dev, "error %d reading JEDEC ID\n", ret);
>> +     ret = spi_nor_read_id(nor, 0, 0, id, nor->reg_proto);
> 
> Hmm, I wonder if it is better to explicitly use SNOR_PROTO_1_1_1 so
> clearly signify that this is intended to use 1S-1S-1S only. What do you
> think?

I would keep it as it is for now, because it offers flexibility.
If we ever gonna determine the protocol at runtime this will come in handy
because it will work without touching the code. JESD216 suggests an algorithm
that tries to determine the mode depending on the SFDP signature.

Cheers,
ta