linux-kernel - Re: [PATCH v5 1/5] mtd: spi-nor: core: add manufacturer flags

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]

Message-ID: <906df2b1-2827-444f-9de2-9b59f4d49b5c@linaro.org>
Date: Mon, 23 Sep 2024 19:04:21 +0100
From: Tudor Ambarus <tudor.ambarus@...aro.org>
To: Erez <erezgeva2@...il.com>, Michael Walle <mwalle@...nel.org>
Cc: Erez Geva <erezgeva@...ime.org>, linux-mtd@...ts.infradead.org,
 Pratyush Yadav <pratyush@...nel.org>, linux-kernel@...r.kernel.org,
 Miquel Raynal <miquel.raynal@...tlin.com>,
 Richard Weinberger <richard@....at>, Vignesh Raghavendra <vigneshr@...com>,
 devicetree@...r.kernel.org, Rob Herring <robh@...nel.org>,
 Krzysztof Kozlowski <krzk+dt@...nel.org>, Conor Dooley
 <conor+dt@...nel.org>, Esben Haabendal <esben@...nix.com>
Subject: Re: [PATCH v5 1/5] mtd: spi-nor: core: add manufacturer flags



On 9/23/24 5:31 PM, Erez wrote:
> On Mon, 23 Sept 2024 at 18:19, Michael Walle <mwalle@...nel.org> wrote:
>>
>> Hi,
>>
>>>>> I would gladly remove the obsolete mx25l12805d.
>>>> Why? I don't see any need for that.
>>> Maybe because we do not want compatibility table?
>>
>> I don't get this? Anyway, we do not remove support for older
>> flashes for no reason.
> 
> I did not insist, you asked.
> Macronix stopped selling these chips 15 year ago.
> How long do you want to support old chips?
> 

as long as there's at least a user in the kernel.
>>
>>> No, reading the SFDP is great.
>>> Except for OTP parameters/configuration.
>>> As there is not way to find OTP parameters using JEDEC ID/SFDP
>>> So as I understand there are only 2 ways to set the OTP parameters:
>>> * Use a compatibility.
>>> * Use dynamic OTP configuration, through DT, sysfs,
>>
>> * Use the in-kernel database to look up the parameters as it is done
>>   with any other flash device. If the id is reused, look for
> 
> All IDs in table are reused.
> The change I suggest is to read SFDP to all Macronix chips.
> As today, we skip the RDSFDP. "to avoid no-op".
> 
>>   differences in the SFDP to figure out the correct flash device,
> 
> You can not do that unless you actually read the SFDP.
> 
>>   then add some .fixups to manually add the OTP flags and
>>   parameters.
> 
> I am in favour.
> As we can not use JEDEC ID or SFDP for OTP parameters.
> How do you suggest manually OTP parameters?

Michael literally just said how to do it. Here it is again:
```
* Use the in-kernel database to look up the parameters as it is done
  with any other flash device. If the id is reused, look for
  differences in the SFDP to figure out the correct flash device,
  then add some .fixups to manually add the OTP flags and
  parameters.
```
> Tudor Ambarus rejected my proposal.
> I am open to implementing your suggestion.
> 
> I will split the two patches to avoid confusion.
> So I will submit 2 patches:
> * Always read Macronix chips SFDP, as Macronix replaced all old chips
> in the Manufacture table.

I'll NACK it unless you prove that the old flavors of flashes are not
used anymore in the kernel.
> * Support Macronix OTP.

Hopefully this time you describe your problem in the commit message and
convince us it is worth fixing and that it makes sense for us to read
past the first paragraph.