Messages in this thread | | | Date | Tue, 25 Jul 2023 09:54:52 +0200 | Subject | Re: [PATCH V3 1/2] xen: Update dm_op.h from Xen public header | From | Juergen Gross <> |
| |
On 25.07.23 09:49, Jan Beulich wrote: > On 25.07.2023 09:42, Viresh Kumar wrote: >> On 25-07-23, 09:18, Jan Beulich wrote: >>> I question that use, btw, but it is not up to me to decide whether to >>> accept such a layering violation in Linux. dm-op is, as its name says, >>> for device models to use. Your intended use doesn't fall in that >>> category, aiui. Imo the present contents of dm_op.h in Linux is indeed >>> all a kernel is supposed to know about, unless it was to gain in-kernel >>> device models. >> >> Is there any other way by which an interrupt can be raised for the >> guest VM ? I was only aware of this method and so implemented it like >> this. >> >> I am open to suggestions on this. > > Well. I don't know your requirements. Generally I would suggest using > event channels, not interrupts, when talking about injecting events > into guests. If it strictly needs to be an interrupt, then I guess a > non-dm-op means would need introducing if none already exists.
I think the best way would be to let the user mode device model (i.e. the backend) construct the dm-op parameters like qemu is doing it and pass it via the ioctl to the privcmd driver as part of struct privcmd_irqfd. Then it would be opaque to the kernel like every other dm-op.
Juergen
[unhandled content-type:application/pgp-keys][unhandled content-type:application/pgp-signature] | |