From: Jan Beulich <jbeulich@suse.com>
Subject: x86/PV: check GDT/LDT limits during emulation

Accesses beyond the LDT limit originating from emulation would trigger
the ASSERT() in pv_map_ldt_shadow_page(). On production builds such
accesses would cause an attempt to promote the touched page (offset from
the present LDT base address) to a segment descriptor one. If this
happens to succeed, guest user mode would be able to elevate its
privileges to that of the guest kernel. This is particularly easy when
there's no LDT at all, in which case the LDT base stored internally to
Xen is simply zero.

Also adjust the ASSERT() that was triggering: It was off by one to
begin with, and for production builds we also better use
ASSERT_UNREACHABLE() instead with suitable recovery code afterwards.

This is XSA-298.

Reported-by: Andrew Cooper <andrew.cooper3@citrix.com>
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>
---
v2: Correct 64-bit-only limit check (by folding into the common one).

--- a/xen/arch/x86/pv/emul-gate-op.c
+++ b/xen/arch/x86/pv/emul-gate-op.c
@@ -51,7 +51,13 @@ static int read_gate_descriptor(unsigned
     const seg_desc_t *pdesc = gdt_ldt_desc_ptr(gate_sel);
 
     if ( (gate_sel < 4) ||
-         ((gate_sel >= FIRST_RESERVED_GDT_BYTE) && !(gate_sel & 4)) ||
+         /*
+          * We're interested in call gates only, which occupy a single
+          * seg_desc_t for 32-bit and a consecutive pair of them for 64-bit.
+          */
+         ((gate_sel >> 3) + !is_pv_32bit_vcpu(v) >=
+          (gate_sel & 4 ? v->arch.pv.ldt_ents
+                        : v->arch.pv.gdt_ents)) ||
          __get_user(desc, pdesc) )
         return 0;
 
@@ -70,7 +76,7 @@ static int read_gate_descriptor(unsigned
     if ( !is_pv_32bit_vcpu(v) )
     {
         if ( (*ar & 0x1f00) != 0x0c00 ||
-             (gate_sel >= FIRST_RESERVED_GDT_BYTE - 8 && !(gate_sel & 4)) ||
+             /* Limit check done above already. */
              __get_user(desc, pdesc + 1) ||
              (desc.b & 0x1f00) )
             return 0;
--- a/xen/arch/x86/pv/emulate.c
+++ b/xen/arch/x86/pv/emulate.c
@@ -31,7 +31,14 @@ int pv_emul_read_descriptor(unsigned int
 {
     seg_desc_t desc;
 
-    if ( sel < 4)
+    if ( sel < 4 ||
+         /*
+          * Don't apply the GDT limit here, as the selector may be a Xen
+          * provided one. __get_user() will fail (without taking further
+          * action) for ones falling in the gap between guest populated
+          * and Xen ones.
+          */
+         ((sel & 4) && (sel >> 3) >= v->arch.pv.ldt_ents) )
         desc.b = desc.a = 0;
     else if ( __get_user(desc, gdt_ldt_desc_ptr(sel)) )
         return 0;
--- a/xen/arch/x86/pv/mm.c
+++ b/xen/arch/x86/pv/mm.c
@@ -92,12 +92,16 @@ bool pv_map_ldt_shadow_page(unsigned int
     BUG_ON(unlikely(in_irq()));
 
     /*
-     * Hardware limit checking should guarantee this property.  NB. This is
+     * Prior limit checking should guarantee this property.  NB. This is
      * safe as updates to the LDT can only be made by MMUEXT_SET_LDT to the
      * current vcpu, and vcpu_reset() will block until this vcpu has been
      * descheduled before continuing.
      */
-    ASSERT((offset >> 3) <= curr->arch.pv.ldt_ents);
+    if ( unlikely((offset >> 3) >= curr->arch.pv.ldt_ents) )
+    {
+        ASSERT_UNREACHABLE();
+        return false;
+    }
 
     if ( is_pv_32bit_domain(currd) )
         linear = (uint32_t)linear;