Re: [PATCH -next 1/3] genirq/affinity: replace managed with is_managed in irq_affinity_desc

[Marc Zyngier <maz@kernel.org>]

On Sat, 27 Aug 2022 02:13:49 +0100,
Xu Qiang <xuqiang36@huawei.com> wrote:
> 
> This submission is based on the following two considerations:
> 
> 1. The definition of is_managed field is misleading to assume
>    that it only uses 1 bit of memory, which is not the case;

You realise that a bitfield is not about the memory used, but the
number of significant bits, right? The memory it uses is the
compiler's business.

> 2. from the actual use of is_managed, it should be a Boolean type;

Why? What is wrong with the existing bitfield? Why renaming it?

>
> Fixes: c410abbbacb9 (“genirq/affinity: Add is_managed to struct irq_affinity_desc”)

I don't see any fix here, only some seemingly pointless bike-shedding.
If you have identified an actual issue, please spell it out for me,
because I cannot see it.

Thanks,

	M.


-- 
Without deviation from the norm, progress is not possible.