Re: [PATCH v2] x86: assorted array_index_nospec() insertions

[Paul Durrant <Paul.Durrant@citrix.com>]

> -----Original Message-----
> From: Jan Beulich [mailto:JBeulich@suse.com]
> Sent: 26 July 2018 14:07
> To: xen-devel <xen-devel@lists.xenproject.org>
> Cc: Razvan Cojocaru <rcojocaru@bitdefender.com>; Andrew Cooper
> <Andrew.Cooper3@citrix.com>; Paul Durrant <Paul.Durrant@citrix.com>;
> Tamas K Lengyel <tamas@tklengyel.com>
> Subject: [PATCH v2] x86: assorted array_index_nospec() insertions
> 
> Don't chance having Spectre v1 (including BCBS) gadgets. In some of the
> cases the insertions are more of precautionary nature rather than there
> provably being a gadget, but I think we should err on the safe (secure)
> side here.
> 
> Signed-off-by: Jan Beulich <jbeulich@suse.com>
> ---
> v2: Re-base. Drop guest_cpuid() changes. Fix off-by-1 in
>     {do,compat}_dm_op().
> 
> --- a/xen/arch/x86/domctl.c
> +++ b/xen/arch/x86/domctl.c
> @@ -28,6 +28,7 @@
>  #include <xen/hypercall.h> /* for arch_do_domctl */
>  #include <xsm/xsm.h>
>  #include <xen/iommu.h>
> +#include <xen/nospec.h>
>  #include <xen/vm_event.h>
>  #include <public/vm_event.h>
>  #include <asm/mem_sharing.h>
> @@ -93,27 +94,34 @@ static int update_domain_cpuid_info(stru
>      /* Insert ctl data into cpuid_policy. */
>      switch ( ctl->input[0] )
>      {
> +        unsigned int idx;
> +
>      case 0x00000000 ... ARRAY_SIZE(p->basic.raw) - 1:
>          switch ( ctl->input[0] )
>          {
>          case 4:
> -            p->cache.raw[ctl->input[1]] = leaf;
> +            idx = array_index_nospec(ctl->input[1], ARRAY_SIZE(p->cache.raw));
> +            p->cache.raw[idx] = leaf;
>              break;
> 
>          case 7:
> -            p->feat.raw[ctl->input[1]] = leaf;
> +            idx = array_index_nospec(ctl->input[1], ARRAY_SIZE(p->feat.raw));
> +            p->feat.raw[idx] = leaf;
>              break;
> 
>          case 0xb:
> -            p->topo.raw[ctl->input[1]] = leaf;
> +            idx = array_index_nospec(ctl->input[1], ARRAY_SIZE(p->topo.raw));
> +            p->topo.raw[idx] = leaf;
>              break;
> 
>          case XSTATE_CPUID:
> -            p->xstate.raw[ctl->input[1]] = leaf;
> +            idx = array_index_nospec(ctl->input[1], ARRAY_SIZE(p->xstate.raw));
> +            p->xstate.raw[idx] = leaf;
>              break;
> 
>          default:
> -            p->basic.raw[ctl->input[0]] = leaf;
> +            idx = array_index_nospec(ctl->input[0], ARRAY_SIZE(p->basic.raw));
> +            p->basic.raw[idx] = leaf;
>              break;
>          }
>          break;
> @@ -127,7 +135,9 @@ static int update_domain_cpuid_info(stru
>          break;
> 
>      case 0x80000000 ... 0x80000000 + ARRAY_SIZE(p->extd.raw) - 1:
> -        p->extd.raw[ctl->input[0] - 0x80000000] = leaf;
> +        idx = array_index_nospec(ctl->input[0] & 0xffff,
> +                                 ARRAY_SIZE(p->extd.raw));
> +        p->extd.raw[idx] = leaf;
>          break;
>      }
> 
> --- a/xen/arch/x86/hvm/dm.c
> +++ b/xen/arch/x86/hvm/dm.c
> @@ -17,6 +17,7 @@
>  #include <xen/event.h>
>  #include <xen/guest_access.h>
>  #include <xen/hypercall.h>
> +#include <xen/nospec.h>
>  #include <xen/sched.h>
> 
>  #include <asm/hap.h>
> @@ -232,7 +233,7 @@ static int set_mem_type(struct domain *d
>                          struct xen_dm_op_set_mem_type *data)
>  {
>      xen_pfn_t last_pfn = data->first_pfn + data->nr - 1;
> -    unsigned int iter = 0;
> +    unsigned int iter = 0, mem_type;
>      int rc = 0;
> 
>      /* Interface types to internal p2m types */
> @@ -252,7 +253,9 @@ static int set_mem_type(struct domain *d
>           unlikely(data->mem_type == HVMMEM_unused) )
>          return -EINVAL;
> 
> -    if ( data->mem_type  == HVMMEM_ioreq_server )
> +    mem_type = array_index_nospec(data->mem_type,
> ARRAY_SIZE(memtype));
> +
> +    if ( mem_type == HVMMEM_ioreq_server )
>      {
>          unsigned int flags;
> 
> @@ -279,10 +282,10 @@ static int set_mem_type(struct domain *d
> 
>          if ( p2m_is_shared(t) )
>              rc = -EAGAIN;
> -        else if ( !allow_p2m_type_change(t, memtype[data->mem_type]) )
> +        else if ( !allow_p2m_type_change(t, memtype[mem_type]) )
>              rc = -EINVAL;
>          else
> -            rc = p2m_change_type_one(d, pfn, t, memtype[data->mem_type]);
> +            rc = p2m_change_type_one(d, pfn, t, memtype[mem_type]);
> 
>          put_gfn(d, pfn);
> 
> @@ -387,6 +390,8 @@ static int dm_op(const struct dmop_args
>          goto out;
>      }
> 
> +    op.op = array_index_nospec(op.op, ARRAY_SIZE(op_size));
> +
>      if ( op_args->buf[0].size < offset + op_size[op.op] )
>          goto out;
> 
> @@ -739,7 +744,7 @@ int compat_dm_op(domid_t domid,
>          return -E2BIG;
> 
>      args.domid = domid;
> -    args.nr_bufs = nr_bufs;
> +    args.nr_bufs = array_index_nospec(nr_bufs, ARRAY_SIZE(args.buf) + 1);

Using something called 'array_index_nospec()' for an array size and having to adjust by 1 is kind of a bit ugly but it looks correct.

Reviewed-by: Paul Durrant <paul.durrant@citrix.com> 

> 
>      for ( i = 0; i < args.nr_bufs; i++ )
>      {
> @@ -776,7 +781,7 @@ long do_dm_op(domid_t domid,
>          return -E2BIG;
> 
>      args.domid = domid;
> -    args.nr_bufs = nr_bufs;
> +    args.nr_bufs = array_index_nospec(nr_bufs, ARRAY_SIZE(args.buf) + 1);
> 
>      if ( copy_from_guest_offset(&args.buf[0], bufs, 0, args.nr_bufs) )
>          return -EFAULT;
> --- a/xen/arch/x86/hvm/hypercall.c
> +++ b/xen/arch/x86/hvm/hypercall.c
> @@ -20,6 +20,7 @@
>   */
>  #include <xen/lib.h>
>  #include <xen/hypercall.h>
> +#include <xen/nospec.h>
> 
>  #include <asm/hvm/support.h>
> 
> @@ -181,8 +182,15 @@ int hvm_hypercall(struct cpu_user_regs *
>      BUILD_BUG_ON(ARRAY_SIZE(hvm_hypercall_table) >
>                   ARRAY_SIZE(hypercall_args_table));
> 
> -    if ( (eax >= ARRAY_SIZE(hvm_hypercall_table)) ||
> -         !hvm_hypercall_table[eax].native )
> +    if ( eax >= ARRAY_SIZE(hvm_hypercall_table) )
> +    {
> +        regs->rax = -ENOSYS;
> +        return HVM_HCALL_completed;
> +    }
> +
> +    eax = array_index_nospec(eax, ARRAY_SIZE(hvm_hypercall_table));
> +
> +    if ( !hvm_hypercall_table[eax].native )
>      {
>          regs->rax = -ENOSYS;
>          return HVM_HCALL_completed;
> --- a/xen/arch/x86/mm/mem_access.c
> +++ b/xen/arch/x86/mm/mem_access.c
> @@ -23,6 +23,7 @@
> 
>  #include <xen/guest_access.h> /* copy_from_guest() */
>  #include <xen/mem_access.h>
> +#include <xen/nospec.h>
>  #include <xen/vm_event.h>
>  #include <xen/event.h>
>  #include <public/vm_event.h>
> @@ -334,6 +335,7 @@ static bool xenmem_access_to_p2m_access(
>      switch ( xaccess )
>      {
>      case 0 ... ARRAY_SIZE(memaccess) - 1:
> +        xaccess = array_index_nospec(xaccess, ARRAY_SIZE(memaccess));
>          *paccess = memaccess[xaccess];
>          break;
>      case XENMEM_access_default:
> --- a/xen/arch/x86/pv/hypercall.c
> +++ b/xen/arch/x86/pv/hypercall.c
> @@ -21,6 +21,7 @@
> 
>  #include <xen/compiler.h>
>  #include <xen/hypercall.h>
> +#include <xen/nospec.h>
>  #include <xen/trace.h>
> 
>  #define HYPERCALL(x)                                                \
> @@ -99,8 +100,15 @@ void pv_hypercall(struct cpu_user_regs *
>      BUILD_BUG_ON(ARRAY_SIZE(pv_hypercall_table) >
>                   ARRAY_SIZE(hypercall_args_table));
> 
> -    if ( (eax >= ARRAY_SIZE(pv_hypercall_table)) ||
> -         !pv_hypercall_table[eax].native )
> +    if ( eax >= ARRAY_SIZE(pv_hypercall_table) )
> +    {
> +        regs->rax = -ENOSYS;
> +        return;
> +    }
> +
> +    eax = array_index_nospec(eax, ARRAY_SIZE(pv_hypercall_table));
> +
> +    if ( !pv_hypercall_table[eax].native )
>      {
>          regs->rax = -ENOSYS;
>          return;
> 
> 


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel