Re: [Xen-devel] [PATCH] xl: suppress suspend/resume functions on platforms which do not support it.

[Ian Campbell <Ian.Campbell@xxxxxxxxxx>]

On Wed, 2014-02-12 at 16:54 +0000, Ian Campbell wrote:
> On Wed, 2014-02-12 at 16:49 +0000, George Dunlap wrote:
> > On 02/12/2014 03:57 PM, Ian Campbell wrote:
> > > On Wed, 2014-02-12 at 16:21 +0100, Olaf Hering wrote:
> > >> On Wed, Feb 12, Ian Campbell wrote:
> > >>
> > >>>   #define LIBXL_HAVE_SIGCHLD_SHARING 1
> > >>>   
> > >>> +/*
> > >>> + * LIBXL_HAVE_NO_SUSPEND_RESUME
> > >> Think positive?
> > >> Make that HAVE_FEATURE and define it on x86.
> > > Could do -- anyone got any strong feeling one way or the other?
> > 
> > I'm a fan of consistency, so "HAVE_SUSPEND" sounds a little bit better 
> > to me.  But either way is fine.
> 
> Right. I think I'll flip it then, it does make more logical sense that
> way.

Actually, no I wont.

Anyone who uses the positive version of this will find that migration
support suddenly disappears on x86 when they build against Xen 4.3 or
earlier, unless they are careful to use
        #if defined(LIBL_HAVE_NO_SUSPEND_RESUME) || defined(__i386__) || 
defined(__x86_64__)
which is worse than a -ve feature flag IMHO.

I will stick with LIBXL_HAVE_NO_SUSPEND_RESUME then.

Ian.


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