On 3/8/23 22:46, Richard W.M. Jones wrote:
The patch series looks generally fine, but it really needs a test ...
We find that having tests prevents unnoticed regressions from
happening later.
Rich.
Could you please elaborate on how do you imagine such a test case?
Basically what I've been doing while making this series is installing a
guest Win on the VM with an IDE controller only, then plugging
virtio-scsi controller instead of IDE and running in-place conversion.
VM should be able to boot as a result.
I guess in the test we should play such a trick with both virtio-blk and
virtio-scsi. But can we do that using phony windows.img from the
test-data? And how do we actually check whether the guest is bootable
after our manipulations?