Xen4.2-rc1 test result

Xen4.2-rc1 test result

[Ren, Yongjie]

Hi All,
We did a round testing for Xen 4.2 RC1 (CS#25692) with Linux 3.4.7 dom0.
We covered VT-d, SR-IOV, Power Management, TXT, IVB new features, HSW new features.
We covered many cases for HVM guests (both Redhat Linux and MS Windows guest).
We tested on Westmere-EP, SandyBridge-EP, IvyBridge desktop, and Haswell hardware platforms.
We found no new issues, and verified 1 fixed bug.

Fixed bug (1):
1. parameter 'maxvcpus' causes hvm guest boots up with wrong vcpu number
http://bugzilla.xen.org/bugzilla/show_bug.cgi?id=1825
-- Fixed by Yang from Intel.

The following are some of the old issues which we guess are something important.
Some of the old issues:
1. long stop during the guest boot process (too slow bootup)
  http://bugzilla.xen.org/bugzilla/show_bug.cgi?id=1821
2. Poor performance when do guest save/restore and migration with linux 3.x dom0
  http://bugzilla.xen.org/bugzilla/show_bug.cgi?id=1784
3. 'xl vcpu-set' can't decrease the vCPU number of a HVM guest
  http://bugzilla.xen.org/bugzilla/show_bug.cgi?id=1822
4. after detaching a VF from a guest, shutdown the guest is very slow
  http://bugzilla.xen.org/bugzilla/show_bug.cgi?id=1812
5. Dom0 cannot be shutdown before PCI detachment from guest and when pci assignment conflicts
  (xl allows same PCI device to be assigned to multiple guests)
  http://bugzilla.xen.org/bugzilla/show_bug.cgi?id=1826
6. Guest hang after resuming from S3
  http://bugzilla.xen.org/bugzilla/show_bug.cgi?id=1828


Best Regards,
     Yongjie (Jay)

Re: Xen4.2-rc1 test result

[Konrad Rzeszutek Wilk]

On Mon, Aug 06, 2012 at 02:34:35PM +0000, Ren, Yongjie wrote:
> Hi All,
> We did a round testing for Xen 4.2 RC1 (CS#25692) with Linux 3.4.7 dom0.
> We covered VT-d, SR-IOV, Power Management, TXT, IVB new features, HSW new features.
> We covered many cases for HVM guests (both Redhat Linux and MS Windows guest).
> We tested on Westmere-EP, SandyBridge-EP, IvyBridge desktop, and Haswell hardware platforms.
> We found no new issues, and verified 1 fixed bug.

Cool!
> 
> Fixed bug (1):
> 1. parameter 'maxvcpus' causes hvm guest boots up with wrong vcpu number
> http://bugzilla.xen.org/bugzilla/show_bug.cgi?id=1825
> -- Fixed by Yang from Intel.
> 
> The following are some of the old issues which we guess are something important.
> Some of the old issues:
> 1. long stop during the guest boot process (too slow bootup)
>   http://bugzilla.xen.org/bugzilla/show_bug.cgi?id=1821
> 2. Poor performance when do guest save/restore and migration with linux 3.x dom0
>   http://bugzilla.xen.org/bugzilla/show_bug.cgi?id=1784
> 3. 'xl vcpu-set' can't decrease the vCPU number of a HVM guest
>   http://bugzilla.xen.org/bugzilla/show_bug.cgi?id=1822
> 4. after detaching a VF from a guest, shutdown the guest is very slow
>   http://bugzilla.xen.org/bugzilla/show_bug.cgi?id=1812
> 5. Dom0 cannot be shutdown before PCI detachment from guest and when pci assignment conflicts
>   (xl allows same PCI device to be assigned to multiple guests)
>   http://bugzilla.xen.org/bugzilla/show_bug.cgi?id=1826
> 6. Guest hang after resuming from S3
>   http://bugzilla.xen.org/bugzilla/show_bug.cgi?id=1828


So those that are Linux related are going to have to wait. I am ..swamped.
The merge window just closed and its time to deal with regressions. Is there
way for Intel engineers in helping to come up with a patch for some of
these issues?