Re: [Xen-devel] [xtf test] 100874: all pass - PUSHED

[Konrad Rzeszutek Wilk <konrad.wilk@xxxxxxxxxx>]

On Mon, Sep 19, 2016 at 02:51:03PM +0100, Wei Liu wrote:
> On Mon, Sep 19, 2016 at 02:46:11PM +0100, Ian Jackson wrote:
> > Wei Liu writes ("Re: [Xen-devel] [xtf test] 100874: all pass - PUSHED"):
> > > On Mon, Sep 19, 2016 at 04:56:01AM -0600, Jan Beulich wrote:
> > > > I guess the respective command line option ("allowsuperpage") was
> > > > not given?
> > > 
> > > Correct, that option is not given in osstest.
> > 
> > It should be, in the XTF tests at least, perhaps ?
> > 
> 
> Adding that in would be trivial. But the value of testing that ...
> 
> > Unless we're proposing to rip the feature out in stable branches.
> > 
> > Konrad Rzeszutek Wilk writes ("Re: [Xen-devel] [xtf test] 100874: all pass 
> > - PUSHED"):
> > > We don't use it anymore either. I had a patch to rip it out but .. I need
> > > to dust it off.
> > 
> > That would be great, thanks.
> > 
> > Failing that, we could clarify the security support status.
> > 
> 
> ... depends on the security support status of that particular
> feature.

To enable it you had to provide the 'allowsuperpage' parameter so by default
nobody would use it. Which means no security support (like the #UD traping
for cross-vendor migration).

I somehow assumed that any non-default enabled parameters by default would not
receive any security support.

Like running an 'debug=y' kernel or such?



> 
> Wei.
> 
> > Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel