Re: RFC/patch portability: split kvm_vm_ioctl v3

[Carsten Otte <cotte-tA70FqPdS9bQT0dZR+AlfA@public.gmane.org>]

Dong, Eddie wrote:
> BTW, how S390 user/kernel interrupt signal is communicated?
Our s90host prototype does inject it from userspace, so there is no 
need to have a call for that.

Nevertheless, I really love the lightweight exits that kvm has 
invented and I want to use it on s390 with kvm too. In that case, 
we'll have an interrupt facility in the kernel that is in terms of 
functionality close to the pic/lapic work for x86.

For that, I think we could encode an interrupt number in an __u64 that 
has a platform specific helper function in userspace to generate that 
interrupt number: for x86 that would just equal the irq line, for s390 
that would be more complex. The common kvm code can then take that 
__u64 number and call an architecture specific callback that injects 
the interrupt.

In addition, I would love to be able to specify which target CPUs may 
receive that interrupt because our IPI equivalent comes out just like 
a regular interrupt on just one target CPU.

That boils down to something like this:
struct kvm_interrupt_data {
	__u64	 interrupt_number;
	cpuset_t possible_target_cpus;
}
and an KVM_INJECT_INTERRUPT common ioctl for the vm to provide this.

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/