In some situations, like when error happens in block attach for
a guest in dom0, backend send us XenbusStateClosing notification.
However, as frontend were never properly initialized, it fails
to change its own state to XenbusStateClosed, leaving the system
in a dead-end state.
Steps to reproduce:
1 - pass a nonexistant file, like:
xm block-attach <domId> file://nonexistant xvdb w
2 - xm block-list <domId> <== Will still be there
3 - xm block-detach <domId> <== Will still be there.
Created attachment 142570 [details]
patch already merged upstream
The only ACK missing here is the QA ack, but since this bug is exceedingly easy
to test I'll apply the patch anyway :)
1. configure an HVM guest's fda as /dev/fd0
2. try to boot the HVM guest
Without this patch, the Xen utilities refuse to create the guest.
With this patch, the Xen utilities should create the guest just fine.
Oops, that was the wrong bugzilla yesterday!
QE ack for RHEL5.
2.6.18-7.el5 included in 20070125.0.