Description of problem

When using virtio-fs with shimv2, if qemu or virtiofsd exits abnormally, e.g. got killed, virtiofsd becomes a zombie process, whose parent is the shimv2 process.

root 23624 1 87 14:10 ? 00:00:33 /usr/local/bin/containerd-shim-kata-v2 -namespace default -address /var/run/containerd.sock -publish-binary /usr/local/bin/containerd -id 967adcbd236c5a7aab2d713e784947c5cb352aa8fd5a4c3e2fae48ec878ebcf9 -debug
root 23780 23624 0 14:10 ? 00:00:00 [virtiofsd]

Expected result

virtiofsd should be waited to release the process resource.

Actual result

virtiofsd becomes zombie.

Additional info

shimv2 process should exit as well, and it reaps virtiofsd automatically, but I think that's another issue.