Drivers: hv: balloon: account for gaps in hot add regions

I'm observing the following hot add requests from the WS2012 host:

hot_add_req: start_pfn = 0x108200 count = 330752
hot_add_req: start_pfn = 0x158e00 count = 193536
hot_add_req: start_pfn = 0x188400 count = 239616

As the host doesn't specify hot add regions we're trying to create
128Mb-aligned region covering the first request, we create the 0x108000 -
0x160000 region and we add 0x108000 - 0x158e00 memory. The second request
passes the pfn_covered() check, we enlarge the region to 0x108000 -
0x190000 and add 0x158e00 - 0x188200 memory. The problem emerges with the
third request as it starts at 0x188400 so there is a 0x200 gap which is
not covered. As the end of our region is 0x190000 now it again passes the
pfn_covered() check were we just adjust the covered_end_pfn and make it
0x188400 instead of 0x188200 which means that we'll try to online
0x188200-0x188400 pages but these pages were never assigned to us and we
crash.

We can't react to such requests by creating new hot add regions as it may
happen that the whole suggested range falls into the previously identified
128Mb-aligned area so we'll end up adding nothing or create intersecting
regions and our current logic doesn't allow that. Instead, create a list of
such 'gaps' and check for them in the page online callback.

Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

1 files changed, 94 insertions, 37 deletions

diff --git a/drivers/hv/hv_balloon.c b/drivers/hv/hv_balloon.c
index 4ae26d60565f..18766f616b8d 100644
--- a/drivers/hv/hv_balloon.c
+++ b/drivers/hv/hv_balloon.c
@@ -441,6 +441,16 @@ struct hv_hotadd_state {
 	unsigned long covered_end_pfn;
 	unsigned long ha_end_pfn;
 	unsigned long end_pfn;
+	/*
+	 * A list of gaps.
+	 */
+	struct list_head gap_list;
+};
+
+struct hv_hotadd_gap {
+	struct list_head list;
+	unsigned long start_pfn;
+	unsigned long end_pfn;
 };
 
 struct balloon_state {
@@ -596,18 +606,46 @@ static struct notifier_block hv_memory_nb = {
 	.priority = 0
 };
 
+/* Check if the particular page is backed and can be onlined and online it. */
+static void hv_page_online_one(struct hv_hotadd_state *has, struct page *pg)
+{
+	unsigned long cur_start_pgp;
+	unsigned long cur_end_pgp;
+	struct hv_hotadd_gap *gap;
+
+	cur_start_pgp = (unsigned long)pfn_to_page(has->covered_start_pfn);
+	cur_end_pgp = (unsigned long)pfn_to_page(has->covered_end_pfn);
 
-static void hv_bring_pgs_online(unsigned long start_pfn, unsigned long size)
+	/* The page is not backed. */
+	if (((unsigned long)pg < cur_start_pgp) ||
+	    ((unsigned long)pg >= cur_end_pgp))
+		return;
+
+	/* Check for gaps. */
+	list_for_each_entry(gap, &has->gap_list, list) {
+		cur_start_pgp = (unsigned long)
+			pfn_to_page(gap->start_pfn);
+		cur_end_pgp = (unsigned long)
+			pfn_to_page(gap->end_pfn);
+		if (((unsigned long)pg >= cur_start_pgp) &&
+		    ((unsigned long)pg < cur_end_pgp)) {
+			return;
+		}
+	}
+
+	/* This frame is currently backed; online the page. */
+	__online_page_set_limits(pg);
+	__online_page_increment_counters(pg);
+	__online_page_free(pg);
+}
+
+static void hv_bring_pgs_online(struct hv_hotadd_state *has,
+				unsigned long start_pfn, unsigned long size)
 {
 	int i;
 
-	for (i = 0; i < size; i++) {
-		struct page *pg;
-		pg = pfn_to_page(start_pfn + i);
-		__online_page_set_limits(pg);
-		__online_page_increment_counters(pg);
-		__online_page_free(pg);
-	}
+	for (i = 0; i < size; i++)
+		hv_page_online_one(has, pfn_to_page(start_pfn + i));
 }
 
 static void hv_mem_hot_add(unsigned long start, unsigned long size,
@@ -684,26 +722,24 @@ static void hv_online_page(struct page *pg)
 	list_for_each(cur, &dm_device.ha_region_list) {
 		has = list_entry(cur, struct hv_hotadd_state, list);
 		cur_start_pgp = (unsigned long)
-			pfn_to_page(has->covered_start_pfn);
-		cur_end_pgp = (unsigned long)pfn_to_page(has->covered_end_pfn);
+			pfn_to_page(has->start_pfn);
+		cur_end_pgp = (unsigned long)pfn_to_page(has->end_pfn);
 
-		if (((unsigned long)pg >= cur_start_pgp) &&
-			((unsigned long)pg < cur_end_pgp)) {
-			/*
-			 * This frame is currently backed; online the
-			 * page.
-			 */
-			__online_page_set_limits(pg);
-			__online_page_increment_counters(pg);
-			__online_page_free(pg);
-		}
+		/* The page belongs to a different HAS. */
+		if (((unsigned long)pg < cur_start_pgp) ||
+		    ((unsigned long)pg >= cur_end_pgp))
+			continue;
+
+		hv_page_online_one(has, pg);
+		break;
 	}
 }
 
-static bool pfn_covered(unsigned long start_pfn, unsigned long pfn_cnt)
+static int pfn_covered(unsigned long start_pfn, unsigned long pfn_cnt)
 {
 	struct list_head *cur;
 	struct hv_hotadd_state *has;
+	struct hv_hotadd_gap *gap;
 	unsigned long residual, new_inc;
 
 	if (list_empty(&dm_device.ha_region_list))
@@ -718,6 +754,24 @@ static bool pfn_covered(unsigned long start_pfn, unsigned long pfn_cnt)
 		 */
 		if (start_pfn < has->start_pfn || start_pfn >= has->end_pfn)
 			continue;
+
+		/*
+		 * If the current start pfn is not where the covered_end
+		 * is, create a gap and update covered_end_pfn.
+		 */
+		if (has->covered_end_pfn != start_pfn) {
+			gap = kzalloc(sizeof(struct hv_hotadd_gap), GFP_ATOMIC);
+			if (!gap)
+				return -ENOMEM;
+
+			INIT_LIST_HEAD(&gap->list);
+			gap->start_pfn = has->covered_end_pfn;
+			gap->end_pfn = start_pfn;
+			list_add_tail(&gap->list, &has->gap_list);
+
+			has->covered_end_pfn = start_pfn;
+		}
+
 		/*
 		 * If the current hot add-request extends beyond
 		 * our current limit; extend it.
@@ -734,19 +788,10 @@ static bool pfn_covered(unsigned long start_pfn, unsigned long pfn_cnt)
 			has->end_pfn += new_inc;
 		}
 
-		/*
-		 * If the current start pfn is not where the covered_end
-		 * is, update it.
-		 */
-
-		if (has->covered_end_pfn != start_pfn)
-			has->covered_end_pfn = start_pfn;
-
-		return true;
-
+		return 1;
 	}
 
-	return false;
+	return 0;
 }
 
 static unsigned long handle_pg_range(unsigned long pg_start,
@@ -785,6 +830,8 @@ static unsigned long handle_pg_range(unsigned long pg_start,
 			if (pgs_ol > pfn_cnt)
 				pgs_ol = pfn_cnt;
 
+			has->covered_end_pfn +=  pgs_ol;
+			pfn_cnt -= pgs_ol;
 			/*
 			 * Check if the corresponding memory block is already
 			 * online by checking its last previously backed page.
@@ -793,10 +840,8 @@ static unsigned long handle_pg_range(unsigned long pg_start,
 			 */
 			if (start_pfn > has->start_pfn &&
 			    !PageReserved(pfn_to_page(start_pfn - 1)))
-				hv_bring_pgs_online(start_pfn, pgs_ol);
+				hv_bring_pgs_online(has, start_pfn, pgs_ol);
 
-			has->covered_end_pfn +=  pgs_ol;
-			pfn_cnt -= pgs_ol;
 		}
 
 		if ((has->ha_end_pfn < has->end_pfn) && (pfn_cnt > 0)) {
@@ -834,13 +879,19 @@ static unsigned long process_hot_add(unsigned long pg_start,
 					unsigned long rg_size)
 {
 	struct hv_hotadd_state *ha_region = NULL;
+	int covered;
 
 	if (pfn_cnt == 0)
 		return 0;
 
-	if (!dm_device.host_specified_ha_region)
-		if (pfn_covered(pg_start, pfn_cnt))
+	if (!dm_device.host_specified_ha_region) {
+		covered = pfn_covered(pg_start, pfn_cnt);
+		if (covered < 0)
+			return 0;
+
+		if (covered)
 			goto do_pg_range;
+	}
 
 	/*
 	 * If the host has specified a hot-add range; deal with it first.
@@ -852,6 +903,7 @@ static unsigned long process_hot_add(unsigned long pg_start,
 			return 0;
 
 		INIT_LIST_HEAD(&ha_region->list);
+		INIT_LIST_HEAD(&ha_region->gap_list);
 
 		list_add_tail(&ha_region->list, &dm_device.ha_region_list);
 		ha_region->start_pfn = rg_start;
@@ -1585,6 +1637,7 @@ static int balloon_remove(struct hv_device *dev)
 	struct hv_dynmem_device *dm = hv_get_drvdata(dev);
 	struct list_head *cur, *tmp;
 	struct hv_hotadd_state *has;
+	struct hv_hotadd_gap *gap, *tmp_gap;
 
 	if (dm->num_pages_ballooned != 0)
 		pr_warn("Ballooned pages: %d\n", dm->num_pages_ballooned);
@@ -1601,6 +1654,10 @@ static int balloon_remove(struct hv_device *dev)
 #endif
 	list_for_each_safe(cur, tmp, &dm->ha_region_list) {
 		has = list_entry(cur, struct hv_hotadd_state, list);
+		list_for_each_entry_safe(gap, tmp_gap, &has->gap_list, list) {
+			list_del(&gap->list);
+			kfree(gap);
+		}
 		list_del(&has->list);
 		kfree(has);
 	}