[PATCH v4 03/11] mtd: spi-nor: core: Use auto-detection only once

[Tudor.Ambarus at microchip.com]

On 4/27/22 08:20, Pratyush Yadav wrote:
> EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
> 
> On 21/04/22 01:41PM, Tudor.Ambarus at microchip.com wrote:
>> On 4/21/22 16:16, Pratyush Yadav wrote:
>>> EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
>>>
>>> On 21/04/22 07:18AM, Tudor.Ambarus at microchip.com wrote:
>>>> Hi, Pratyush,
>>>>
>>>> I forgot to remove few checks, would you please remove them when applying?
>>>> See below.
>>>>
>>>> On 4/20/22 13:34, Tudor Ambarus wrote:
>>>>> In case spi_nor_match_name() returned NULL, the auto detection was
>>>>> issued twice. There's no reason to try to detect the same chip twice,
>>>>> do the auto detection only once.
>>>>>
>>>>> Signed-off-by: Tudor Ambarus <tudor.ambarus at microchip.com>
>>>>> Reviewed-by: Michael Walle <michael at walle.cc>
>>>>> ---
>>>>>  drivers/mtd/spi-nor/core.c | 13 ++++++++-----
>>>>>  1 file changed, 8 insertions(+), 5 deletions(-)
>>>>>
>>>>> diff --git a/drivers/mtd/spi-nor/core.c b/drivers/mtd/spi-nor/core.c
>>>>> index b9cc8bbf1f62..b55d922d46dd 100644
>>>>> --- a/drivers/mtd/spi-nor/core.c
>>>>> +++ b/drivers/mtd/spi-nor/core.c
>>>>> @@ -2896,13 +2896,14 @@ static const struct flash_info *spi_nor_get_flash_info(struct spi_nor *nor,
>>>>>  {
>>>>>     const struct flash_info *info = NULL;
>>>>>
>>>>> -   if (name)
>>>>> +   if (name) {
>>>>>             info = spi_nor_match_name(nor, name);
>>>>> +           if (IS_ERR(info))
>>>>> +                   return info;
>>>>
>>>> As Michael suggested spi_nor_match_name() returns NULL or valid entry, so this
>>>> check is not necessary, let's remove them.
>>>>
>>>>> +   }
>>>>>     /* Try to auto-detect if chip name wasn't specified or not found */
>>>>>     if (!info)
>>>>> -           info = spi_nor_read_id(nor);
>>>>> -   if (IS_ERR_OR_NULL(info))
>>>>> -           return ERR_PTR(-ENOENT);
>>>>> +           return spi_nor_read_id(nor);
>>>>>
>>>>>     /*
>>>>>      * If caller has specified name of flash model that can normally be
>>>>> @@ -2994,7 +2995,9 @@ int spi_nor_scan(struct spi_nor *nor, const char *name,
>>>>>             return -ENOMEM;
>>>>>
>>>>>     info = spi_nor_get_flash_info(nor, name);
>>>>> -   if (IS_ERR(info))
>>>>> +   if (!info)
>>>>> +           return -ENOENT;
>>>>
>>>> also according to Michael, this change is not needed as spi_nor_get_flash_info() can't
>>>> return NULL. Here we can keep the code as it was. Let me know if you want me to respin.
>>>
>>> TBH I don't think a NULL check here hurts much since the behaviour might
>>> change later, and error paths don't get exercised as often. But I have
>>
>> I agree, but at the same time we're introducing checks gratuitously. Since
>> Michael cared about it, it's fine that we removed it. I don't care too much
>> about it.
>>
>>> made both changes when applying. You can double-check at [0] if you
>>> want.>
>>> [0] https://github.com/prati0100/linux-0day/commit/67d913746833ee54bf4c661040f3ef13657dffd8
>>
>> looks good.
>>
>> btw: I think this patch
>> https://github.com/prati0100/linux-0day/commit/b45bbff85d49529f8daff83c341a292f6c6492ca
>> may introduce a regression on some atmel chips. Let me try it please.
> 
> Did you get a chance to try this out? If it works fine, I would like to
> apply it.

I tried it on a at25df321a. When calling unlock, everything seems fine. However
I haven't tried a lock/unlock cycle as lock is not supported and I couldn't
allocate more time. Even if there will be regressions we can handle them
afterwards, so let's apply it.

Cheers,
ta