Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Doug Berger <opendmb@gmail.com>
To:     Andrew Morton <akpm@linux-foundation.org>
Cc:     Jonathan Corbet <corbet@lwn.net>, Mike Rapoport <rppt@kernel.org>,
        Borislav Petkov <bp@suse.de>,
        "Paul E. McKenney" <paulmck@kernel.org>,
        Neeraj Upadhyay <quic_neeraju@quicinc.com>,
        Randy Dunlap <rdunlap@infradead.org>,
        Damien Le Moal <damien.lemoal@opensource.wdc.com>,
        Muchun Song <songmuchun@bytedance.com>,
        Vlastimil Babka <vbabka@suse.cz>,
        Johannes Weiner <hannes@cmpxchg.org>,
        Michal Hocko <mhocko@suse.com>,
        KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>,
        Mel Gorman <mgorman@suse.de>,
        Mike Kravetz <mike.kravetz@oracle.com>,
        Florian Fainelli <f.fainelli@gmail.com>,
        David Hildenbrand <david@redhat.com>,
        Oscar Salvador <osalvador@suse.de>,
        Joonsoo Kim <iamjoonsoo.kim@lge.com>,
        linux-doc@vger.kernel.org, linux-kernel@vger.kernel.org,
        linux-mm@kvack.org, Doug Berger <opendmb@gmail.com>
Subject: [PATCH v3 3/9] mm/page_alloc: prevent creation of empty zones
Date:   Thu, 20 Oct 2022 14:53:12 -0700
Message-Id: <20221020215318.4193269-4-opendmb@gmail.com>
In-Reply-To: <20221020215318.4193269-1-opendmb@gmail.com>
References: <20221020215318.4193269-1-opendmb@gmail.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

If none of the pages a zone spans are present then its start pfn
and span should be zeroed to prevent initialization.

This prevents the creation of an empty zone if all of its pages
are moved to a zone that would overlap it.

The real_size name is reverted to just size to reduce the burden
of supporting multiple realities.

Signed-off-by: Doug Berger <opendmb@gmail.com>
---
 mm/page_alloc.c | 20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)

diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 92908c51f1c3..2f7b88d78bc2 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -7548,8 +7548,7 @@ static void __init calculate_node_totalpages(struct pglist_data *pgdat,
 	for (i = 0; i < MAX_NR_ZONES; i++) {
 		struct zone *zone = pgdat->node_zones + i;
 		unsigned long zone_start_pfn, zone_end_pfn;
-		unsigned long spanned, absent;
-		unsigned long size, real_size;
+		unsigned long spanned, absent, size;
 
 		spanned = zone_spanned_pages_in_node(pgdat->node_id, i,
 						     node_start_pfn,
@@ -7560,20 +7559,21 @@ static void __init calculate_node_totalpages(struct pglist_data *pgdat,
 						   node_start_pfn,
 						   node_end_pfn);
 
-		size = spanned;
-		real_size = size - absent;
+		size = spanned - absent;
 
-		if (size)
+		if (size) {
 			zone->zone_start_pfn = zone_start_pfn;
-		else
+		} else {
+			spanned = 0;
 			zone->zone_start_pfn = 0;
-		zone->spanned_pages = size;
-		zone->present_pages = real_size;
+		}
+		zone->spanned_pages = spanned;
+		zone->present_pages = size;
 #if defined(CONFIG_MEMORY_HOTPLUG)
-		zone->present_early_pages = real_size;
+		zone->present_early_pages = size;
 #endif
 
-		totalpages += real_size;
+		totalpages += size;
 	}
 
 	pgdat->node_spanned_pages = node_end_pfn - node_start_pfn;
-- 
2.25.1