[PATCH 1/6] x86/p2m: tidy p2m_add_foreign() a little

[Jan Beulich <jbeulich@xxxxxxxx>]

Drop a bogus ASSERT() - we don't typically assert incoming domain
pointers to be non-NULL, and there's no particular reason to do so here.

Replace the open-coded DOMID_SELF check by use of
rcu_lock_remote_domain_by_id(), at the same time covering the request
being made with the current domain's actual ID.

Move the "both domains same" check into just the path where it really
is meaningful.

Swap the order of the two puts, such that
- the p2m lock isn't needlessly held across put_page(),
- a separate put_page() on an error path can be avoided,
- they're inverse to the order of the respective gets.

Signed-off-by: Jan Beulich <jbeulich@xxxxxxxx>
---
The DOMID_SELF check being converted also suggests to me that there's an
implication of tdom == current->domain, which would in turn appear to
mean the "both domains same" check could as well be dropped altogether.

--- a/xen/arch/x86/mm/p2m.c
+++ b/xen/arch/x86/mm/p2m.c
@@ -2560,9 +2560,6 @@ int p2m_add_foreign(struct domain *tdom,
     int rc;
     struct domain *fdom;
 
-    ASSERT(tdom);
-    if ( foreigndom == DOMID_SELF )
-        return -EINVAL;
     /*
      * hvm fixme: until support is added to p2m teardown code to cleanup any
      * foreign entries, limit this to hardware domain only.
@@ -2573,13 +2570,15 @@ int p2m_add_foreign(struct domain *tdom,
     if ( foreigndom == DOMID_XEN )
         fdom = rcu_lock_domain(dom_xen);
     else
-        fdom = rcu_lock_domain_by_id(foreigndom);
-    if ( fdom == NULL )
-        return -ESRCH;
+    {
+        rc = rcu_lock_remote_domain_by_id(foreigndom, &fdom);
+        if ( rc )
+            return rc;
 
-    rc = -EINVAL;
-    if ( tdom == fdom )
-        goto out;
+        rc = -EINVAL;
+        if ( tdom == fdom )
+            goto out;
+    }
 
     rc = xsm_map_gmfn_foreign(XSM_TARGET, tdom, fdom);
     if ( rc )
@@ -2593,10 +2592,8 @@ int p2m_add_foreign(struct domain *tdom,
     if ( !page ||
          !p2m_is_ram(p2mt) || p2m_is_shared(p2mt) || p2m_is_hole(p2mt) )
     {
-        if ( page )
-            put_page(page);
         rc = -EINVAL;
-        goto out;
+        goto put_one;
     }
     mfn = page_to_mfn(page);
 
@@ -2625,8 +2622,6 @@ int p2m_add_foreign(struct domain *tdom,
                  gpfn, mfn_x(mfn), fgfn, tdom->domain_id, fdom->domain_id);
 
  put_both:
-    put_page(page);
-
     /*
      * This put_gfn for the above get_gfn for prev_mfn.  We must do this
      * after set_foreign_p2m_entry so another cpu doesn't populate the gpfn
@@ -2634,9 +2629,13 @@ int p2m_add_foreign(struct domain *tdom,
      */
     put_gfn(tdom, gpfn);
 
-out:
+ put_one:
+    put_page(page);
+
+ out:
     if ( fdom )
         rcu_unlock_domain(fdom);
+
     return rc;
 }