KVM: arm64: Clarify the ordering requirements for vcpu/RD creation
authorMarc Zyngier <maz@kernel.org>
Wed, 27 Sep 2023 09:09:11 +0000 (10:09 +0100)
committerOliver Upton <oliver.upton@linux.dev>
Sat, 30 Sep 2023 18:15:44 +0000 (18:15 +0000)
commitf9940416f193d148dd7ad7498ce4f40c83d12e7a
tree5afc552517fcfec5f4fd03a68cc2f16ab6dd6d1d
parentb5daffb120bb60f974ae1a5589160b05c98e00e5
KVM: arm64: Clarify the ordering requirements for vcpu/RD creation

It goes without saying, but it is probably better to spell it out:

If userspace tries to restore and VM, but creates vcpus and/or RDs
in a different order, the vcpu/RD mapping will be different. Yes,
our API is an ugly piece of crap and I can't believe that we missed
this.

If we want to relax the above, we'll need to define a new userspace
API that allows the mapping to be specified, rather than relying
on the kernel to perform the mapping on its own.

Reviewed-by: Zenghui Yu <yuzenghui@huawei.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20230927090911.3355209-12-maz@kernel.org
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Documentation/virt/kvm/devices/arm-vgic-v3.rst