Xen devel - Mar 2013 - [GIT PULL] (xen) stable/for-linus-3.9-rc3-tag

Hey Linus,

Please git pull the following tag:

 git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen.git
stable/for-linus-3.9-rc3-tag

which is full of fixes. The signed tag has the details (as always) but
this is mostly just the last stragglers of the regression bugs that this
merge window had. There are also two normal bug-fixes in the lot.

P.S.
We are working out the kinks in an early warning #linux-next testing framework
so that it can give us heads up in the future - so the when the next merge
window opens we shouldn''t have surprises.

Please pull!

 arch/x86/include/asm/xen/hypercall.h |  4 +--
 arch/x86/xen/mmu.c                   |  3 +-
 drivers/xen/Kconfig                  |  2 +-
 drivers/xen/events.c                 | 10 +++---
 drivers/xen/fallback.c               |  3 +-
 drivers/xen/xen-acpi-processor.c     |  3 ++
 drivers/xen/xen-pciback/pci_stub.c   | 59 +++++++++++++++++++++++++++---------
 include/xen/interface/physdev.h      |  6 ++++
 8 files changed, 64 insertions(+), 26 deletions(-)

David Vrabel (1):
      xen/events: avoid race with raising an event in unmask_evtchn()

Jan Beulich (1):
      xen-pciback: notify hypervisor about devices intended to be assigned to
guests

Konrad Rzeszutek Wilk (3):
      xen/acpi-processor: Don''t dereference struct acpi_processor on
all CPUs.
      xen/acpi-stub: Disable it b/c the acpi_processor_add is no longer called.
      xen/mmu: Move the setting of pvops.write_cr3 to later phase in bootup.

On Fri, 22 Mar 2013, Konrad Rzeszutek Wilk wrote:
> Hey Linus,
> 
> Please git pull the following tag:
> 
>  git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen.git
stable/for-linus-3.9-rc3-tag
> 
> which is full of fixes. The signed tag has the details (as always) but
> this is mostly just the last stragglers of the regression bugs that this
> merge window had. There are also two normal bug-fixes in the lot.
> 
> P.S.
> We are working out the kinks in an early warning #linux-next testing
framework
> so that it can give us heads up in the future - so the when the next merge
> window opens we shouldn''t have surprises.
> 
> Please pull!
> 
>  arch/x86/include/asm/xen/hypercall.h |  4 +--
>  arch/x86/xen/mmu.c                   |  3 +-
>  drivers/xen/Kconfig                  |  2 +-
>  drivers/xen/events.c                 | 10 +++---
>  drivers/xen/fallback.c               |  3 +-
>  drivers/xen/xen-acpi-processor.c     |  3 ++
>  drivers/xen/xen-pciback/pci_stub.c   | 59
+++++++++++++++++++++++++++---------
>  include/xen/interface/physdev.h      |  6 ++++
>  8 files changed, 64 insertions(+), 26 deletions(-)
> 
> David Vrabel (1):
>       xen/events: avoid race with raising an event in unmask_evtchn()
Unless I am looking in the wrong place, this patch is still the same one
that breaks PV on HVM and Xen ARM guests, see:

http://marc.info/?l=xen-devel&m=136377731913034

David is aware of the issue and I think is working on an alternative
fix.

On Sat, Mar 23, 2013 at 10:08 AM, Stefano Stabellini
<stefano.stabellini@eu.citrix.com> wrote:
>
> Unless I am looking in the wrong place, this patch is still the same one
> that breaks PV on HVM and Xen ARM guests, see:
Konrad?

I''ll hold off on the pull request for now.

             Linus

> David Vrabel (1): xen/events: avoid race with raising an event in 
unmask_evtchn()
> Unless I am looking in the wrong place, this patch is still the same one
> that breaks PV on HVM and Xen ARM guests, see:
>
> http://marc.info/?l=xen-devel&m=136377731913034
>
> David is aware of the issue and I think is working on an alternative
> fix.
Thank you for noticing that. Linus does not look to have pulled it - so 
we still have some time
to fix this right.

On 23/03/13 23:13, konrad wilk wrote:
>  > David Vrabel (1): xen/events: avoid race with raising an event in 
> unmask_evtchn()
>> Unless I am looking in the wrong place, this patch is still the same
one
>> that breaks PV on HVM and Xen ARM guests, see:
>>
>> http://marc.info/?l=xen-devel&m=136377731913034
>>
>> David is aware of the issue and I think is working on an alternative
>> fix.
> 
> Thank you for noticing that. Linus does not look to have pulled it - so 
> we still have some time
> to fix this right.
Konrad,

I''ve just posted an updated version of this patch.  I''d
suggest waiting
for an ack from Stefano before applying it.

David