On Thu, Feb 27, 2020 at 10:51:46AM -0500, Peter Xu wrote: > Right. Another thing I just noticed is that we must _not_ attach the > resamplefd to the KVM_IRQFD ioctl when we registered this in the > userspace. One thing is it's for sure not needed at all. More > importantly, on new kernels (after commit 654f1f13ea56) the KVM_IRQFD > could fail (because it sees both split irqchip and resamplefd) and > QEMU will fallback again to the slow path even if the fast path can > work.
I've just posted a v1.1 version of patch 4 only, which should fix this problem on new kernels, so fast path will always be preferred. The only changed function is kvm_irqchip_assign_irqfd(). -- Peter Xu
