Nikunj A Dadhania nik...@linux.vnet.ibm.com writes:
The following commit caused the regression in qemu-system-ppc64
7effdaa3: spapr: Fix return value of vga initialization
d44229c5: Fix vga_interface_type for command line argument '-device VGA'
Even when -nodefaults was provided, USB
Markus Armbruster arm...@redhat.com writes:
Nikunj A Dadhania nik...@linux.vnet.ibm.com writes:
Have you considered extending QEMUMachineInitArgs instead of adding this
function?
Did not think of this option earlier. You mean doing something like
this?
diff --git a/hw/ppc/spapr.c
Nikunj A Dadhania nik...@linux.vnet.ibm.com writes:
Markus Armbruster arm...@redhat.com writes:
Nikunj A Dadhania nik...@linux.vnet.ibm.com writes:
Have you considered extending QEMUMachineInitArgs instead of adding this
function?
Did not think of this option earlier. You mean doing
Il 04/04/2014 12:58, Markus Armbruster ha scritto:
Have you considered extending QEMUMachineInitArgs instead of adding this
function?
Did not think of this option earlier. You mean doing something like
this?
Yes. Looks nicer, doesn't it?
I still think it's a libvirt bug. Mixing
Il 04/04/2014 07:28, Nikunj A Dadhania ha scritto:
And -usb is translated to adding pci-ohci controller for spapr
Yeah, but with -nodefaults it's better to use -device directly.
I think there is special handling for this in vl.c
bool usb_enabled(bool default_usb)
{
return
Paolo Bonzini pbonz...@redhat.com writes:
Il 04/04/2014 12:58, Markus Armbruster ha scritto:
Have you considered extending QEMUMachineInitArgs instead of adding this
function?
Did not think of this option earlier. You mean doing something like
this?
Yes. Looks nicer, doesn't it?
Paolo Bonzini pbonz...@redhat.com writes:
Il 04/04/2014 07:28, Nikunj A Dadhania ha scritto:
And -usb is translated to adding pci-ohci controller for spapr
Yeah, but with -nodefaults it's better to use -device directly.
I think there is special handling for this in vl.c
bool
Il 04/04/2014 13:40, Nikunj A Dadhania ha scritto:
Sure. However, I'm saying that it's fine for spapr to make -usb mean
OHCI, and also keyboard mouse if there is a VGA card in the system.
If libvirt used -device pci-ohci unconditionally, it would fix the bug
*and* it would ensure that the
Am 04.04.2014 10:28, schrieb Nikunj A Dadhania:
diff --git a/vl.c b/vl.c
index 017f92d..0d6c36c 100644
--- a/vl.c
+++ b/vl.c
@@ -4348,7 +4348,8 @@ int main(int argc, char **argv, char **envp)
.kernel_filename = kernel_filename,
[adding libvir-list]
On 04/04/2014 05:23 AM, Markus Armbruster wrote:
Paolo Bonzini pbonz...@redhat.com writes:
Il 04/04/2014 12:58, Markus Armbruster ha scritto:
Have you considered extending QEMUMachineInitArgs instead of adding this
function?
Did not think of this option earlier. You
Am 03.04.2014 18:56, schrieb Nikunj A Dadhania:
The following commit caused the regression in qemu-system-ppc64
7effdaa3: spapr: Fix return value of vga initialization
d44229c5: Fix vga_interface_type for command line argument '-device VGA'
Even when -nodefaults was provided, USB Keyboard
The following commit caused the regression in qemu-system-ppc64
7effdaa3: spapr: Fix return value of vga initialization
d44229c5: Fix vga_interface_type for command line argument '-device VGA'
Even when -nodefaults was provided, USB Keyboard and Mouse was added
to the machine. This breaks
Andreas Färber afaer...@suse.de writes:
Am 03.04.2014 18:56, schrieb Nikunj A Dadhania:
The following commit caused the regression in qemu-system-ppc64
7effdaa3: spapr: Fix return value of vga initialization
d44229c5: Fix vga_interface_type for command line argument '-device VGA'
Even
Am 03.04.2014 19:06, schrieb Nikunj A Dadhania:
The following commit caused the regression in qemu-system-ppc64
7effdaa3: spapr: Fix return value of vga initialization
d44229c5: Fix vga_interface_type for command line argument '-device VGA'
Even when -nodefaults was provided, USB
Il 03/04/2014 18:56, Nikunj A Dadhania ha scritto:
The following commit caused the regression in qemu-system-ppc64
7effdaa3: spapr: Fix return value of vga initialization
d44229c5: Fix vga_interface_type for command line argument '-device VGA'
Even when -nodefaults was provided, USB Keyboard
On 04/03/2014 12:01 PM, Paolo Bonzini wrote:
Il 03/04/2014 18:56, Nikunj A Dadhania ha scritto:
The following commit caused the regression in qemu-system-ppc64
7effdaa3: spapr: Fix return value of vga initialization
d44229c5: Fix vga_interface_type for command line argument '-device VGA'
Paolo Bonzini pbonz...@redhat.com writes:
Il 03/04/2014 18:56, Nikunj A Dadhania ha scritto:
The following commit caused the regression in qemu-system-ppc64
7effdaa3: spapr: Fix return value of vga initialization
d44229c5: Fix vga_interface_type for command line argument '-device VGA'
Even
Il 03/04/2014 21:24, Nikunj A Dadhania ha scritto:
Does libvirt use -nodefaults -machine usb=true? It should create the
OHCI controller separately instead of using -machine.
I see it creating:
-nodefaults -usb -device usb-kbd,id=input0 -device usb-mouse,id=input1
And -usb is translated to
Paolo Bonzini pbonz...@redhat.com writes:
Il 03/04/2014 21:24, Nikunj A Dadhania ha scritto:
Does libvirt use -nodefaults -machine usb=true? It should create the
OHCI controller separately instead of using -machine.
I see it creating:
-nodefaults -usb -device usb-kbd,id=input0 -device
19 matches
Mail list logo
