Messages in this thread | | | Date | Sat, 27 Aug 2022 16:18:48 +0100 | From | Marc Zyngier <> | Subject | Re: [PATCH -next 1/3] genirq/affinity: replace managed with is_managed in irq_affinity_desc |
| |
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.
| |