path: root/0013-IOMMU-x86-disallow-device-assignment-to-PoD-guests.patch

From 838f6c211f7f05f107e1acdfb0977ab61ec0bf2e Mon Sep 17 00:00:00 2001
From: Jan Beulich <jbeulich@suse.com>
Date: Tue, 7 Jun 2022 14:03:20 +0200
Subject: [PATCH 13/51] IOMMU/x86: disallow device assignment to PoD guests
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

While it is okay for IOMMU page tables to be set up for guests starting
in PoD mode, actual device assignment may only occur once all PoD
entries have been removed from the P2M. So far this was enforced only
for boot-time assignment, and only in the tool stack.

Also use the new function to replace p2m_pod_entry_count(): Its unlocked
access to p2m->pod.entry_count wasn't really okay (irrespective of the
result being stale by the time the caller gets to see it). Nor was the
use of that function in line with the immediately preceding comment: A
PoD guest isn't just one with a non-zero entry count, but also one with
a non-empty cache (e.g. prior to actually launching the guest).

To allow the tool stack to see a consistent snapshot of PoD state, move
the tail of XENMEM_{get,set}_pod_target handling into a function, adding
proper locking there.

In libxl take the liberty to use the new local variable r also for a
pre-existing call into libxc.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
master commit: ad4312d764e8b40a1e45b64aac6d840a60c59f13
master date: 2022-05-02 08:48:02 +0200
---
 xen/arch/x86/mm.c                   |  6 +---
 xen/arch/x86/mm/p2m-pod.c           | 43 ++++++++++++++++++++++++++++-
 xen/common/vm_event.c               |  2 +-
 xen/drivers/passthrough/x86/iommu.c |  3 +-
 xen/include/asm-x86/p2m.h           | 21 +++++++-------
 5 files changed, 57 insertions(+), 18 deletions(-)

diff --git a/xen/arch/x86/mm.c b/xen/arch/x86/mm.c
index e222d9aa98ee..4ee2de11051d 100644
--- a/xen/arch/x86/mm.c
+++ b/xen/arch/x86/mm.c
@@ -4777,7 +4777,6 @@ long arch_memory_op(unsigned long cmd, XEN_GUEST_HANDLE_PARAM(void) arg)
     {
         xen_pod_target_t target;
         struct domain *d;
-        struct p2m_domain *p2m;
 
         if ( copy_from_guest(&target, arg, 1) )
             return -EFAULT;
@@ -4812,10 +4811,7 @@ long arch_memory_op(unsigned long cmd, XEN_GUEST_HANDLE_PARAM(void) arg)
         }
         else if ( rc >= 0 )
         {
-            p2m = p2m_get_hostp2m(d);
-            target.tot_pages       = domain_tot_pages(d);
-            target.pod_cache_pages = p2m->pod.count;
-            target.pod_entries     = p2m->pod.entry_count;
+            p2m_pod_get_mem_target(d, &target);
 
             if ( __copy_to_guest(arg, &target, 1) )
             {
diff --git a/xen/arch/x86/mm/p2m-pod.c b/xen/arch/x86/mm/p2m-pod.c
index d8d1a0ce7ed7..a3c9d8a97423 100644
--- a/xen/arch/x86/mm/p2m-pod.c
+++ b/xen/arch/x86/mm/p2m-pod.c
@@ -20,6 +20,7 @@
  */
 
 #include <xen/event.h>
+#include <xen/iocap.h>
 #include <xen/ioreq.h>
 #include <xen/mm.h>
 #include <xen/sched.h>
@@ -362,7 +363,10 @@ p2m_pod_set_mem_target(struct domain *d, unsigned long target)
 
     ASSERT( pod_target >= p2m->pod.count );
 
-    ret = p2m_pod_set_cache_target(p2m, pod_target, 1/*preemptible*/);
+    if ( has_arch_pdevs(d) || cache_flush_permitted(d) )
+        ret = -ENOTEMPTY;
+    else
+        ret = p2m_pod_set_cache_target(p2m, pod_target, 1/*preemptible*/);
 
 out:
     pod_unlock(p2m);
@@ -370,6 +374,23 @@ out:
     return ret;
 }
 
+void p2m_pod_get_mem_target(const struct domain *d, xen_pod_target_t *target)
+{
+    struct p2m_domain *p2m = p2m_get_hostp2m(d);
+
+    ASSERT(is_hvm_domain(d));
+
+    pod_lock(p2m);
+    lock_page_alloc(p2m);
+
+    target->tot_pages       = domain_tot_pages(d);
+    target->pod_cache_pages = p2m->pod.count;
+    target->pod_entries     = p2m->pod.entry_count;
+
+    unlock_page_alloc(p2m);
+    pod_unlock(p2m);
+}
+
 int p2m_pod_empty_cache(struct domain *d)
 {
     struct p2m_domain *p2m = p2m_get_hostp2m(d);
@@ -1387,6 +1408,9 @@ guest_physmap_mark_populate_on_demand(struct domain *d, unsigned long gfn,
     if ( !paging_mode_translate(d) )
         return -EINVAL;
 
+    if ( has_arch_pdevs(d) || cache_flush_permitted(d) )
+        return -ENOTEMPTY;
+
     do {
         rc = mark_populate_on_demand(d, gfn, chunk_order);
 
@@ -1408,3 +1432,20 @@ void p2m_pod_init(struct p2m_domain *p2m)
     for ( i = 0; i < ARRAY_SIZE(p2m->pod.mrp.list); ++i )
         p2m->pod.mrp.list[i] = gfn_x(INVALID_GFN);
 }
+
+bool p2m_pod_active(const struct domain *d)
+{
+    struct p2m_domain *p2m;
+    bool res;
+
+    if ( !is_hvm_domain(d) )
+        return false;
+
+    p2m = p2m_get_hostp2m(d);
+
+    pod_lock(p2m);
+    res = p2m->pod.entry_count | p2m->pod.count;
+    pod_unlock(p2m);
+
+    return res;
+}
diff --git a/xen/common/vm_event.c b/xen/common/vm_event.c
index 70ab3ba406ff..21d2f0edf727 100644
--- a/xen/common/vm_event.c
+++ b/xen/common/vm_event.c
@@ -639,7 +639,7 @@ int vm_event_domctl(struct domain *d, struct xen_domctl_vm_event_op *vec)
 
             rc = -EXDEV;
             /* Disallow paging in a PoD guest */
-            if ( p2m_pod_entry_count(p2m_get_hostp2m(d)) )
+            if ( p2m_pod_active(d) )
                 break;
 
             /* domain_pause() not required here, see XSA-99 */
diff --git a/xen/drivers/passthrough/x86/iommu.c b/xen/drivers/passthrough/x86/iommu.c
index a36a6bd4b249..dc9936e16930 100644
--- a/xen/drivers/passthrough/x86/iommu.c
+++ b/xen/drivers/passthrough/x86/iommu.c
@@ -502,11 +502,12 @@ bool arch_iommu_use_permitted(const struct domain *d)
 {
     /*
      * Prevent device assign if mem paging, mem sharing or log-dirty
-     * have been enabled for this domain.
+     * have been enabled for this domain, or if PoD is still in active use.
      */
     return d == dom_io ||
            (likely(!mem_sharing_enabled(d)) &&
             likely(!mem_paging_enabled(d)) &&
+            likely(!p2m_pod_active(d)) &&
             likely(!p2m_get_hostp2m(d)->global_logdirty));
 }
 
diff --git a/xen/include/asm-x86/p2m.h b/xen/include/asm-x86/p2m.h
index 357a8087481e..f2af7a746ced 100644
--- a/xen/include/asm-x86/p2m.h
+++ b/xen/include/asm-x86/p2m.h
@@ -661,6 +661,12 @@ int p2m_pod_empty_cache(struct domain *d);
  * domain matches target */
 int p2m_pod_set_mem_target(struct domain *d, unsigned long target);
 
+/* Obtain a consistent snapshot of PoD related domain state. */
+void p2m_pod_get_mem_target(const struct domain *d, xen_pod_target_t *target);
+
+/* Check whether PoD is (still) active in a domain. */
+bool p2m_pod_active(const struct domain *d);
+
 /* Scan pod cache when offline/broken page triggered */
 int
 p2m_pod_offline_or_broken_hit(struct page_info *p);
@@ -669,11 +675,6 @@ p2m_pod_offline_or_broken_hit(struct page_info *p);
 void
 p2m_pod_offline_or_broken_replace(struct page_info *p);
 
-static inline long p2m_pod_entry_count(const struct p2m_domain *p2m)
-{
-    return p2m->pod.entry_count;
-}
-
 void p2m_pod_init(struct p2m_domain *p2m);
 
 #else
@@ -689,6 +690,11 @@ static inline int p2m_pod_empty_cache(struct domain *d)
     return 0;
 }
 
+static inline bool p2m_pod_active(const struct domain *d)
+{
+    return false;
+}
+
 static inline int p2m_pod_offline_or_broken_hit(struct page_info *p)
 {
     return 0;
@@ -699,11 +705,6 @@ static inline void p2m_pod_offline_or_broken_replace(struct page_info *p)
     ASSERT_UNREACHABLE();
 }
 
-static inline long p2m_pod_entry_count(const struct p2m_domain *p2m)
-{
-    return 0;
-}
-
 static inline void p2m_pod_init(struct p2m_domain *p2m) {}
 
 #endif
-- 
2.35.1