[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH 2/4] xen/x86: Use memblock_reserve for sensitive areas.



On Fri, Jul 27, 2012 at 11:49:02AM +0100, Stefano Stabellini wrote:
> On Thu, 26 Jul 2012, Konrad Rzeszutek Wilk wrote:
> > instead of a big memblock_reserve. This way we can be more
> > selective in freeing regions (and it also makes it easier
> > to understand where is what).
> > 
> > [v1: Move the auto_translate_physmap to proper line]
> > Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@xxxxxxxxxx>
> > ---
> >  arch/x86/xen/enlighten.c |   38 ++++++++++++++++++++++++++++++++++++++
> >  arch/x86/xen/p2m.c       |    5 +++++
> >  arch/x86/xen/setup.c     |    9 ---------
> >  3 files changed, 43 insertions(+), 9 deletions(-)
> > 
> > diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
> > index ff962d4..9b1afa4 100644
> > --- a/arch/x86/xen/enlighten.c
> > +++ b/arch/x86/xen/enlighten.c
> > @@ -998,7 +998,44 @@ static int xen_write_msr_safe(unsigned int msr, 
> > unsigned low, unsigned high)
> >  
> >     return ret;
> >  }
> > +static void __init xen_reserve_mfn(unsigned long mfn)
> > +{
> > +   unsigned long pfn;
> > +
> > +   if (!mfn)
> > +           return;
> > +   pfn = mfn_to_pfn(mfn);
> > +   if (phys_to_machine_mapping_valid(pfn))
> > +           memblock_reserve(PFN_PHYS(pfn), PAGE_SIZE);
> > +}
> 
> If the mfn is not in the m2p xen_reserve_mfn won't do anything. It is
> worth writing it down in a comment.

Meaning in a printk?
> 
> 
> > +static void __init xen_reserve_internals(void)
> > +{
> > +   unsigned long size;
> > +
> > +   if (!xen_pv_domain())
> > +           return;
> > +
> > +   memblock_reserve(__pa(xen_start_info), PAGE_SIZE);
> 
> xen_start_info is not in the m2p, so you cannot use xen_reserve_mfn

It seems to work for me. For both the toolstack created guests
and dom0. Let me double check thought.
> 
> 
> > +   xen_reserve_mfn(PFN_DOWN(xen_start_info->shared_info));
> > +   xen_reserve_mfn(xen_start_info->store_mfn);
> 
> Are we sure that shared_info points to an mfn that is in the m2p (rather
> than a special mfn not present in the list)?
> 
> 
> > +   if (!xen_initial_domain())
> > +           xen_reserve_mfn(xen_start_info->console.domU.mfn);
> > +
> > +   if (xen_feature(XENFEAT_auto_translated_physmap))
> > +           return;
> > +
> > +   /*
> > +    * ALIGN up to compensate for the p2m_page pointing to an array that
> > +    * can partially filled (look in xen_build_dynamic_phys_to_machine).
> > +    */
> > +
> > +   size = PAGE_ALIGN(xen_start_info->nr_pages * sizeof(unsigned long));
> > +   memblock_reserve(__pa(xen_start_info->mfn_list), size);
> 
> I take that here you are using memblock_reserve again, rather than
> xen_reserve_mfn, because the corresponding mfn is not in the m2p?

<nods> Well, they are - but they are 55555555..
> 
> 
> > +   /* The pagetables are reserved in mmu.c */
> > +}
> >  void xen_setup_shared_info(void)
> >  {
> >     if (!xen_feature(XENFEAT_auto_translated_physmap)) {
> > @@ -1362,6 +1399,7 @@ asmlinkage void __init xen_start_kernel(void)
> >     xen_raw_console_write("mapping kernel into physical memory\n");
> >     pgd = xen_setup_kernel_pagetable(pgd, xen_start_info->nr_pages);
> >  
> > +   xen_reserve_internals();
> >     /* Allocate and initialize top and mid mfn levels for p2m structure */
> >     xen_build_mfn_list_list();
> >  
> > diff --git a/arch/x86/xen/p2m.c b/arch/x86/xen/p2m.c
> > index e4adbfb..6a2bfa4 100644
> > --- a/arch/x86/xen/p2m.c
> > +++ b/arch/x86/xen/p2m.c
> > @@ -388,6 +388,11 @@ void __init xen_build_dynamic_phys_to_machine(void)
> >     }
> >  
> >     m2p_override_init();
> > +
> > +   /* NOTE: We cannot call memblock_reserve here for the mfn_list as there
> > +    * isn't enough pieces to make it work (for one - we are still using the
> > +    * Xen provided pagetable). Do it later in xen_reserve_internals.
> > +    */
> >  }
> >  
> >  unsigned long get_phys_to_machine(unsigned long pfn)
> > diff --git a/arch/x86/xen/setup.c b/arch/x86/xen/setup.c
> > index a4790bf..9efca75 100644
> > --- a/arch/x86/xen/setup.c
> > +++ b/arch/x86/xen/setup.c
> > @@ -424,15 +424,6 @@ char * __init xen_memory_setup(void)
> >     e820_add_region(ISA_START_ADDRESS, ISA_END_ADDRESS - ISA_START_ADDRESS,
> >                     E820_RESERVED);
> >  
> > -   /*
> > -    * Reserve Xen bits:
> > -    *  - mfn_list
> > -    *  - xen_start_info
> > -    * See comment above "struct start_info" in <xen/interface/xen.h>
> > -    */
> > -   memblock_reserve(__pa(xen_start_info->mfn_list),
> > -                    xen_start_info->pt_base - xen_start_info->mfn_list);
> > -
> >     sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &e820.nr_map);
> >  
> >     return "Xen";
> > -- 
> > 1.7.7.6
> > 
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to majordomo@xxxxxxxxxxxxxxx
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at  http://www.tux.org/lkml/
> > 
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@xxxxxxxxxxxxx
> http://lists.xen.org/xen-devel

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


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.