$NetBSD: patch-CVE-2014-7154,v 1.1 2014/09/26 10:45:00 bouyer Exp $

x86/shadow: fix race condition sampling the dirty vram state

d->arch.hvm_domain.dirty_vram must be read with the domain's paging lock held.

If not, two concurrent hypercalls could both end up attempting to free
dirty_vram (the second of which will free a wild pointer), or both end up
allocating a new dirty_vram structure (the first of which will be leaked).

This is XSA-104.

Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
Reviewed-by: Tim Deegan <tim@xen.org>

--- xen/arch/x86/mm/shadow/common.c.orig	2013-09-10 08:42:18.000000000 +0200
+++ xen/arch/x86/mm/shadow/common.c	2014-09-26 12:21:33.000000000 +0200
@@ -3640,7 +3640,7 @@
     int flush_tlb = 0;
     unsigned long i;
     p2m_type_t t;
-    struct sh_dirty_vram *dirty_vram = d->arch.hvm_domain.dirty_vram;
+    struct sh_dirty_vram *dirty_vram;
     struct p2m_domain *p2m = p2m_get_hostp2m(d);
 
     if (end_pfn < begin_pfn
@@ -3649,6 +3649,7 @@
         return -EINVAL;
 
     shadow_lock(d);
+    dirty_vram = d->arch.hvm_domain.dirty_vram;
 
     if ( dirty_vram && (!nr ||
              ( begin_pfn != dirty_vram->begin_pfn