This issue occurred at a very low probability, If we inject delays in
address_space_write_continue like this, the issue occurred inevitably：
###
diff --git a/exec.c b/exec.c
index e8d7b33..b9779e0 100644
--- a/exec
On 02/10/2018 03:29 AM, Wang King wrote:
> Empty IDE CD-ROM configured on the VM:
>
>
>
>
>
>
> Make migration for this VM, then qemu aborted in ide_restart_bh. IDEState
> expect
> end_transfer_func equal to ide_atapi_cmd, but it refer to
> ide_dummy_transf
Empty IDE CD-ROM configured on the VM:
Make migration for this VM, then qemu aborted in ide_restart_bh. IDEState expect
end_transfer_func equal to ide_atapi_cmd, but it refer to
ide_dummy_transfer_stop.
I have no idea about this, can anyone help me?
qemu vers
