Re: [PATCH v2 01/15] pci: Introduce parse_pci_sbdf{_seg}()

[Jan Beulich <jbeulich@xxxxxxxx>]

On 19.06.2026 11:17, Andrew Cooper wrote:
> On 19/06/2026 7:14 am, Jan Beulich wrote:
>> On 18.06.2026 16:50, Teddy Astie wrote:
>>> In many places, we're parsing a PCI string into individual
>>> parts (seg, bus, dev, fn) and then transform it into a pci_sbdf_t
>>> using PCI_SBDF macro. Rather than converting from parts to pci_sbdf_t
>>> and vice versa, introduce a new function that parses a PCI string
>>> into a pci_sbdf_t structure directly.
>>>
>>> Signed-off-by: Teddy Astie <teddy.astie@xxxxxxxxxx>
>> You introduce two functions without callers, hence two new violations of
>> Misra rule 2.1. I understand there will be callers by the end of the series,
>> but I think we want to remain Misra-unregressed at every patch boundary.
>> Even for rules which aren't marked as clean (and hence blocking) yet.
> 
> I'm going to firmly object to this request.
> 
> 2.1 is not checked and has no viable plan to be.

Okay, then we need to discuss this on, perhaps, the next committers' call.
Yes, the rule isn't among the clean ones. But it is among the ones checked
(because it is among the ones we accepted). As per the most recent staging
pipeline, at least in the -amd jobs it also doesn't have many remaining
violations.

Jan

> This patch series is nicely structured for easy review, and I'm not
> happy forcing it to be worse just to satisfy a check that is firmly off.
> 
> ~Andrew