Re: Fwd: [PATCH] x86_64: typo in __assign_irq_vector when update pos for vector and offset

[ebiederm@xmission.com (Eric W. Biederman)]

"Lu, Yinghai" <yinghai.lu@amd.com> writes:

>>So to get things going making TARGET_CPUS cpu_online_map looks like
>>the right thing to do.
>
> Yes. but need to other reference to TARGET_CPUS to verify...it doesn't
> break sth.

I just looked and tested and we are fine.

>>My question is are your io_apics pci devices?  Not does the kernel
>>have them.
>
> Yes, I'm testing with 32 amd8132 in the simulator. Or forget about about
> ioapic, and use MSI, and HT-irq directly...?

Ok.  So if want a pci device we can have one :)
Usually what I have seen is that all io_apics except the
first one show up as pci devices.

>>There are a lot of ways we can approach assigning irqs to cpus and
> there
>>is a lot of work there.  I think Adrian Bunk has been doing some work
>>with the user space irq balancer, and should probably be involved.
>
> Right. We need only do needed in kernel space, and leave most to irq
> balancer.

Actually I was just being pragmatic.  Make it work now.  Make it optimal
later :)

Eric