Re: [patch 33/37] iommu/arm-smmu-v3: Use msi_get_virq()

[Thomas Gleixner]

On Tue, Nov 30 2021 at 09:36, Will Deacon wrote:
> On Mon, Nov 29, 2021 at 02:54:18PM +0000, Robin Murphy wrote:
>> On 2021-11-29 14:42, Thomas Gleixner wrote:
>> > It's actually 0 when the vector cannot be found.
>> 
>> Oh, -1 for my reading comprehension but +1 for my confidence in the patch
>> then :)
>> 
>> I'll let Will have the final say over how cautious we really want to be
>> here, but as far as I'm concerned it's a welcome cleanup as-is. Ditto for
>> patch #32 based on the same reasoning, although I don't have a suitable test
>> platform on-hand to sanity-check that one.
>
> If, as it appears, msi_get_virq() isn't going to fail meaningfully after
> we've successfully called platform_msi_domain_alloc_irqs() then it sounds
> like the patch is fine. Just wanted to check though, as Spring cleaning at
> the end of November raised an eyebrow over here :)

Fair enough. Next time I'll name it 'Cleaning the Augean stables' when
it's the wrong season.

Thanks,

        tglx