Re: [Xen-devel] PATCH 0/10: Merge PV framebuffer & console into QEMU

[Alex Williamson <alex.williamson@xxxxxx>]

On Thu, 2007-10-25 at 22:23 +0100, Daniel P. Berrange wrote:
> > Seems like the async hookup isn't happening.  Anything obvious to check?
> > I confirmed I do still have vfb on ia64 with a tools build from the cset
> > before this series.  Thanks,
> 
> Another thing worth testing, is to try a build at 16225:695871933840
> vs 16226:695871933840.  16226 contains a refactoring of the front/back
> handshake code, so if it still works on 16225 then it points to a bug
> in the state machine changes I did.

   Yes, that is the case.  16225 works fine.  On 16226, the domu hangs,
then gives me a vkbd error:

Error: Device 0 (vkbd) could not be connected. Hotplug scripts not
working.

Here's was I see in the qmeu log file:

xs_read(): vncpasswd get
error. /vm/c766b61d-58e1-8efe-b853-40addbd451ae/vncpasswd.
shift keysym 003e keycode 86
shift keysym 00a6 keycode 86
keypad keysym ffae keycode 83
keypad keysym ffac keycode 83
...
shift keysym 0021 keycode 2
shift keysym 0040 keycode 3
shift keysym 0023 keycode 4
...
shift keysym 003f keycode 53
FB: Waiting for KBD backend creation
Doing backend watch on 
Watch for dev failed

The same failure happens in the qmeu log with tip.  It seems odd to me
that xenfb_new() can return failure, but we don't make use of the return
value from xenfb_wait_for_backend().  I also attached the xenstore-ls
output.  This was domid 5 for this latest attempt on cset 16226.
Thanks,

        Alex

-- 
Alex Williamson                             HP Open Source & Linux Org.

xenstore-ls.log.bz2
Description: application/bzip

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel