Re: Please revert / review 077fc1c04d70ef1748ac2daa6622b3320a1a004c

["Zhang, Yang Z" <yang.z.zhang@intel.com>]

Matthias wrote on 2014-06-20:
> Just for me for clarification: So is it reasonable that my server
> crashed with th
> #3 patch used, or is this something I should investigate furthor /
> maybe do another xen build.?

As Jan pointed out that the patch should not work on AMD platform and cause the Xen crash. I will work out a NPT patch for you to have a try.

> 
> 
> Also, I don't have a serial port, but i got a netconsole setup
> running. Is there any way to pipe the xl dmesg output into the normal
> dmesg output so it would be caught by netconsole?
> 
> 
> 
> 2014-06-20 10:22 GMT+02:00 Jan Beulich <JBeulich@suse.com>:
> 
> 
> 	>>> On 19.06.14 at 18:22, <matthias.kannenberg@googlemail.com> wrote:
> 	> Test 2: 	> 	> Xen-Version: Compiled a normal 077 commit version
> without any of your 	> patches 	> Xen cmd line: added immo=no-sharept as
> suggested 	> Passed-through devices: vga, 2 usb controller (ohci+ehci),
> sound card 	> Result: works fine. domU boots, no BSOD, no other
> problems, server seems to 	> be stable. So not sharing the VT-d page
> tables seems to do the trick / the 	> problem is in the page table
> sharing.
> 
> 
> 	Just to clarify - page table sharing is intended to get disabled in
> 	-unstable on the AMD side ( 	
> http://lists.xenproject.org/archives/html/xen-devel/2014-06/msg00416.h t
> ml, 	still awaiting an AMD IOMMU maintainer ack). But as that's not
> 	immediately a backporting candidate, understanding the problem 	to find
> a solution also for 4.4.x would still be necessary.
> 
> 	Jan
> 
> 
>


Best regards,
Yang