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

Re: [Xen-devel] [PATCH] x86: allow Dom0 read-only access to IO-APICs



On 02/05/2013 07:31, "Jan Beulich" <JBeulich@xxxxxxxx> wrote:

> There are BIOSes that want to map the IO-APIC MMIO region from some
> ACPI method(s), and there is at least one BIOS flavor that wants to
> use this mapping to clear an RTE's mask bit. While we can't allow the
> latter, we can permit reads and simply drop write attempts, leveraging
> the already existing infrastructure introduced for dealing with AMD
> IOMMUs' representation as PCI devices.
> 
> This fixes an interrupt setup problem on a system where _CRS evaluation
> involved the above described BIOS/ACPI behavior, and is expected to
> also deal with a boot time crash of pv-ops Linux upon encountering the
> same kind of system.
> 
> Signed-off-by: Jan Beulich <jbeulich@xxxxxxxx>

mmio_ro_ranges is not specific to ioapic handling, so its initialisation
should probably not go to init_ioapic_mappings()? If it is x86 specific (for
now at least) then it could just go in __start_xen()?

 -- Keir

> --- a/xen/arch/x86/domain_build.c
> +++ b/xen/arch/x86/domain_build.c
> @@ -1114,7 +1114,7 @@ int __init construct_dom0(
>      for ( i = 0; i < nr_ioapics; i++ )
>      {
>          mfn = paddr_to_pfn(mp_ioapics[i].mpc_apicaddr);
> -        if ( smp_found_config )
> +        if ( !rangeset_contains_singleton(mmio_ro_ranges, mfn) )
>              rc |= iomem_deny_access(dom0, mfn, mfn);
>      }
>  
> --- a/xen/arch/x86/io_apic.c
> +++ b/xen/arch/x86/io_apic.c
> @@ -2468,6 +2468,9 @@ void __init init_ioapic_mappings(void)
>      unsigned int i, idx = FIX_IO_APIC_BASE_0;
>      union IO_APIC_reg_01 reg_01;
>  
> +    mmio_ro_ranges = rangeset_new(NULL, "r/o mmio ranges",
> +                                  RANGESETF_prettyprint_hex);
> +
>      if ( smp_found_config )
>          nr_irqs_gsi = 0;
>      for ( i = 0; i < nr_ioapics; i++ )
> @@ -2507,6 +2510,11 @@ void __init init_ioapic_mappings(void)
>              reg_01.raw = io_apic_read(i, 1);
>              nr_ioapic_entries[i] = reg_01.bits.entries + 1;
>              nr_irqs_gsi += nr_ioapic_entries[i];
> +
> +            if ( rangeset_add_singleton(mmio_ro_ranges,
> +                                        ioapic_phys >> PAGE_SHIFT) )
> +                printk(KERN_ERR "Failed to mark IO-APIC page %lx
> read-only\n",
> +                       ioapic_phys);
>          }
>      }
>  
> --- a/xen/drivers/passthrough/pci.c
> +++ b/xen/drivers/passthrough/pci.c
> @@ -105,8 +105,6 @@ void __init pt_pci_init(void)
>      radix_tree_init(&pci_segments);
>      if ( !alloc_pseg(0) )
>          panic("Could not initialize PCI segment 0\n");
> -    mmio_ro_ranges = rangeset_new(NULL, "r/o mmio ranges",
> -                                  RANGESETF_prettyprint_hex);
>  }
>  
>  int __init pci_add_segment(u16 seg)
> 
> 
> 



_______________________________________________
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®.