xen/mm: don't unconditionally clear PGC_need_scrub in alloc_heap_pages()

alloc_heap_pages() will unconditionally clear PGC_need_scrub, even when
MEMF_no_scrub is requested.  This is kind of expected as otherwise some
callers will assert on seeing non-expected flags set on the count_info
field.

Introduce a new MEMF bit to signal to alloc_heap_pages() that non-scrubbed
pages should keep the PGC_need_scrub bit set. This fixes returning dirty
pages from alloc_domheap_pages() without the PGC_need_scrub bit set for
populate_physmap() to consume.

With the above change alloc_domheap_pages() needs an adjustment to cope
with allocated pages possibly having the PGC_need_scrub set.

Fixes: 83a784a15b ("xen/mm: allow deferred scrub of physmap populate allocated pages")
Reported-by: Ayden Bottos <aydenbottos12@gmail.com>
Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>

xen/common/memory.c

@@ -345,7 +345,8 @@ static void populate_physmap(struct memop_args *a)
                 unsigned int scrub_start = 0;
                 unsigned int memflags =
                     a->memflags | (d->creation_finished ? 0
-                                                        : MEMF_no_scrub);
+                                                        : (MEMF_no_scrub |
+                                                           MEMF_keep_scrub));
                 nodeid_t node =
                     (a->memflags & MEMF_exact_node) ? MEMF_get_node(a->memflags)
                                                     : NUMA_NO_NODE;

xen/common/page_alloc.c

@@ -989,6 +989,8 @@ static struct page_info *alloc_heap_pages(
     ASSERT(zone_lo <= zone_hi);
     ASSERT(zone_hi < NR_ZONES);
 
+    ASSERT(!(memflags & MEMF_keep_scrub) || (memflags & MEMF_no_scrub));
+
     if ( unlikely(order > MAX_ORDER) )
         return NULL;
 
@@ -1110,17 +1112,26 @@ static struct page_info *alloc_heap_pages(
     {
         bool cold = d && d != current->domain;
 
-        for ( i = 0; i < (1U << order); i++ )
+        if ( !(memflags & MEMF_no_scrub) )
         {
-            if ( test_and_clear_bit(_PGC_need_scrub, &pg[i].count_info) )
+            for ( i = 0; i < (1U << order); i++ )
             {
-                if ( !(memflags & MEMF_no_scrub) )
+                if ( test_and_clear_bit(_PGC_need_scrub, &pg[i].count_info) )
+                {
                     scrub_one_page(&pg[i], cold);
-
-                dirty_cnt++;
+                    dirty_cnt++;
+                }
+                else
+                    check_one_page(&pg[i]);
             }
-            else if ( !(memflags & MEMF_no_scrub) )
-                check_one_page(&pg[i]);
+        }
+        else
+        {
+            for ( i = 0; i < (1U << order); i++ )
+                if ( (memflags & MEMF_keep_scrub)
+                     ? test_bit(_PGC_need_scrub, &pg[i].count_info)
+                     : test_and_clear_bit(_PGC_need_scrub, &pg[i].count_info) )
+                    dirty_cnt++;
         }
 
         if ( dirty_cnt )
@@ -2696,8 +2707,10 @@ struct page_info *alloc_domheap_pages(
 
             for ( i = 0; i < (1UL << order); i++ )
             {
-                ASSERT(!pg[i].count_info);
-                pg[i].count_info = PGC_extra;
+                ASSERT(!(pg[i].count_info &
+                         ~((memflags & MEMF_keep_scrub) ? PGC_need_scrub
+                                                        : 0UL)));
+                pg[i].count_info |= PGC_extra;
             }
         }
         if ( assign_page(pg, order, d, memflags) )

xen/include/xen/mm.h

@@ -208,6 +208,9 @@ struct npfec {
 #define  MEMF_no_refcount (1U<<_MEMF_no_refcount)
 #define _MEMF_populate_on_demand 1
 #define  MEMF_populate_on_demand (1U<<_MEMF_populate_on_demand)
+/* MEMF_keep_scrub is only valid when specified together with MEMF_no_scrub. */
+#define _MEMF_keep_scrub  2
+#define  MEMF_keep_scrub  (1U << _MEMF_keep_scrub)
 #define _MEMF_no_dma      3
 #define  MEMF_no_dma      (1U<<_MEMF_no_dma)
 #define _MEMF_exact_node  4