On Thu, 2016-01-14 at 17:32 +0000, Stefano Stabellini wrote:
> 
> > > +    LOG(DEBUG, "QEMU needs to be run as root for PCI passthrough to
> > > work");
> > 
> > Shouldn't there be an if here, and/or an error return?
> 
> Unfortunately we cannot get the user used to run QEMU with from here.
> However, even without this change, there is already plenty of
> information printed out:
> 
> - xl prints:
> libxl: error: libxl_qmp.c:287:qmp_handle_error_response: received an
> error message from QMP server: Device initialization failed
> 
> - qemu prints:
> [00:03.0] xen_pt_initfn: Error: Failed to "open" the real pci device. rc:
> -13

All that is under normal (i.e. successful) usage? The presence of such spew
under those circumstances should probably be considered a bug. Certainly
adding more noise doesn't seem desirable.

libxl should be able to get the user, either from the json stash of the
config or by writing it somewhere on create so it can get it back when it
needs it.

That reminds me -- how does the qemu user selection work over migrate? Does
it remember the specific user or does it try and pick it again on the other
end? Do we require all hosts in a pool to have been setup in the same way
i.e. either xen-qemuuser-domid$i or xen-qemuuser-shared but not a mixture?

Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

Reply via email to