[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <1351070872.2237.122.camel@zakaz.uk.xensource.com>
Date: Wed, 24 Oct 2012 10:27:52 +0100
From: Ian Campbell <Ian.Campbell@...rix.com>
To: Konrad Rzeszutek Wilk <konrad.wilk@...cle.com>
CC: "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"xen-devel@...ts.xensource.com" <xen-devel@...ts.xensource.com>,
"mukesh.rathor@...cle.com" <mukesh.rathor@...cle.com>,
Stefano Stabellini <Stefano.Stabellini@...citrix.com>
Subject: Re: [PATCH 1/6] xen/pvh: Support ParaVirtualized Hardware
extensions.
On Tue, 2012-10-23 at 15:46 +0100, Konrad Rzeszutek Wilk wrote:
> > diff --git a/include/xen/interface/memory.h b/include/xen/interface/memory.h
> > index b66d04c..8beebdb 100644
> > --- a/include/xen/interface/memory.h
> > +++ b/include/xen/interface/memory.h
> > @@ -169,7 +169,13 @@ struct xen_add_to_physmap {
> > /* Source mapping space. */
> > #define XENMAPSPACE_shared_info 0 /* shared info page */
> > #define XENMAPSPACE_grant_table 1 /* grant table page */
> > - unsigned int space;
> > +#define XENMAPSPACE_gmfn 2 /* GMFN */
> > +#define XENMAPSPACE_gmfn_range 3 /* GMFN range */
> > +#define XENMAPSPACE_gmfn_foreign 4 /* GMFN from another guest */
> > + uint16_t space;
> > + domid_t foreign_domid; /* IFF XENMAPSPACE_gmfn_foreign */
> > +
> > +#define XENMAPIDX_grant_table_status 0x80000000
> >
> > /* Index into source mapping space. */
> > xen_ulong_t idx;
>
> And this breaks 32-bit PVHVM :-(
>
> The reason being that in xen_hvm_init_shared_info we allocate the
> structure on the stack, and do not clear the foreign_domid to zero.
>
> This means we can (and do get) for the argument space, something
> like this:
> xatp.space == 0xffff0001;
> instead of
> xatp.space = 0x1;
We should be reverting this patch and using XENMEM_add_to_physmap range
instead, see my "xen: x86 pvh: use XENMEM_add_to_physmap_range for
foreign gmfn mappings" from last week.
Mukesh wanted to go with this original version in the interim so
probably fixing it as you have makes sense anyhow.
>
> This fixes it:
>
> diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
> index e3497f2..b679f86 100644
> --- a/arch/x86/xen/enlighten.c
> +++ b/arch/x86/xen/enlighten.c
> @@ -1509,6 +1509,7 @@ void __ref xen_hvm_init_shared_info(void)
> xatp.domid = DOMID_SELF;
> xatp.idx = 0;
> xatp.space = XENMAPSPACE_shared_info;
> + xatp.foreign_domid = 0;
> xatp.gpfn = __pa(shared_info_page) >> PAGE_SHIFT;
> if (HYPERVISOR_memory_op(XENMEM_add_to_physmap, &xatp))
> BUG();
> diff --git a/drivers/xen/grant-table.c b/drivers/xen/grant-table.c
> index b2b0a37..cbfd929 100644
> --- a/drivers/xen/grant-table.c
> +++ b/drivers/xen/grant-table.c
> @@ -1044,6 +1044,7 @@ static int gnttab_map(unsigned int start_idx, unsigned int end_idx)
> do {
> xatp.domid = DOMID_SELF;
> xatp.idx = i;
> + xatp.foreign_domid = 0;
> xatp.space = XENMAPSPACE_grant_table;
> xatp.gpfn = (xen_hvm_resume_frames >> PAGE_SHIFT) + i;
> rc = HYPERVISOR_memory_op(XENMEM_add_to_physmap, &xatp);
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists