From nobody Wed Apr 1 22:20:02 2026 Received: from mgamail.intel.com (mgamail.intel.com [192.198.163.11]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 8C9D92D1914 for ; Wed, 1 Apr 2026 07:01:52 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=192.198.163.11 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1775026914; cv=none; b=h8xKkSQwOvnQ6JUwwRy/U05nYOcoFw3Aoh2J4bnfWzfEED2JMeJcw6iq5V2kee5ZKqLn2errLCv3rHuMCQF5qbt0lGusnC9vgUE9VVntmXcSUvZL7bVQpwzVx2YT4uZpQ1QJvy2MbLAWtq/m2ad5+pqbelkeaAKCBgz1XLw18DQ= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1775026914; c=relaxed/simple; bh=cO2HAq7Tbla5DCgH2yGoOYpWj/xxLKqY/nmwAcNmCmE=; h=From:To:Cc:Subject:Date:Message-ID:MIME-Version; b=qN+pDgrnDRSdtHiwQJRMUmw0bX0RnYPAKgLnc9s9y/dQcq9FoYywb5LQy2mMO/WP2XLz1vkIRfTcqf7Zy3Fuv0rMbdrFR8f4qIiSr14zgyucORKkwURgNVsxluKW2gBecAroT5DRLdw2xkL+EnLFW7q4TtYi7B8+TM3BhzdZZ1o= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=intel.com; spf=pass smtp.mailfrom=intel.com; dkim=pass (2048-bit key) header.d=intel.com header.i=@intel.com header.b=h8Kc2RLT; arc=none smtp.client-ip=192.198.163.11 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=intel.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=intel.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=intel.com header.i=@intel.com header.b="h8Kc2RLT" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1775026912; x=1806562912; h=from:to:cc:subject:date:message-id:mime-version: content-transfer-encoding; bh=cO2HAq7Tbla5DCgH2yGoOYpWj/xxLKqY/nmwAcNmCmE=; b=h8Kc2RLTAY1mnfCE/J6eVYdz8wYBj2gvYskPA8qpH/2SUOFtOZavpVlp SivHeX0CobNu8oeqcAzcrJUjxw11iLSqBCsIP2zJsPjnwxFEx4yPBF4/F uZoPmXCIJwcRb4grNotvo+FoK9OYse51r7KXiKWDXJ1mU1OachpLspHoh 5UC8uTea6Y7Pc8c0v0G4ousZm2K88bTk55ILDo0uKz6Pkf7OEA4vUNm/M I+76lXV+cblTbVdSK55pmgqG/jvuGf/oE/byw4SIo8DMi+/5msIYqXf/c +O7wBDDBblYdYJXp2Bcxas+PKtX41TEr++4cNEYAgTkq2RScGvqCF5smk Q==; X-CSE-ConnectionGUID: xw9T3uerQmGrq+iJw+mctQ== X-CSE-MsgGUID: 1D7jcqpTRLOItpRzucnf6Q== X-IronPort-AV: E=McAfee;i="6800,10657,11745"; a="86670940" X-IronPort-AV: E=Sophos;i="6.23,153,1770624000"; d="scan'208";a="86670940" Received: from orviesa002.jf.intel.com ([10.64.159.142]) by fmvoesa105.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 01 Apr 2026 00:01:39 -0700 X-CSE-ConnectionGUID: xzw0AiRbSv61DOsFc5UM6w== X-CSE-MsgGUID: vskgqkMkQ3uNPB953sBlkg== X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="6.23,153,1770624000"; d="scan'208";a="257092471" Received: from spr10.sh.intel.com (HELO localhost) ([10.239.23.75]) by orviesa002.jf.intel.com with ESMTP; 01 Apr 2026 00:01:36 -0700 From: Yuan Liu To: David Hildenbrand , Oscar Salvador , Mike Rapoport , Wei Yang Cc: linux-mm@kvack.org, Yong Hu , Nanhai Zou , Yuan Liu , Tim Chen , Qiuxu Zhuo , Yu C Chen , Pan Deng , Tianyou Li , Chen Zhang , linux-kernel@vger.kernel.org Subject: [PATCH v2] mm/memory hotplug/unplug: Optimize zone contiguous check when changing pfn range Date: Wed, 1 Apr 2026 03:01:55 -0400 Message-ID: <20260401070155.1420929-1-yuan1.liu@intel.com> X-Mailer: git-send-email 2.47.3 Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" When move_pfn_range_to_zone() or remove_pfn_range_from_zone() updates a zone, set_zone_contiguous() rescans the entire zone pageblock-by-pageblock to rebuild zone->contiguous. For large zones this is a significant cost during memory hotplug and hot-unplug. Add a new zone member pages_with_online_memmap that tracks the number of pages within the zone span that have an online memmap (including present pages and memory holes whose memmap has been initialized). When spanned_pages =3D=3D pages_with_online_memmap the zone is contiguous and pfn_to_page() can be called on any PFN in the zone span without further pfn_valid() checks. Only pages that fall within the current zone span are accounted towards pages_with_online_memmap. A "too small" value is safe, it merely prevents detecting a contiguous zone. The following test cases of memory hotplug for a VM [1], tested in the environment [2], show that this optimization can significantly reduce the memory hotplug time [3]. +----------------+------+---------------+--------------+----------------+ | | Size | Time (before) | Time (after) | Time Reduction | | +------+---------------+--------------+----------------+ | Plug Memory | 256G | 10s | 3s | 70% | | +------+---------------+--------------+----------------+ | | 512G | 36s | 7s | 81% | +----------------+------+---------------+--------------+----------------+ +----------------+------+---------------+--------------+----------------+ | | Size | Time (before) | Time (after) | Time Reduction | | +------+---------------+--------------+----------------+ | Unplug Memory | 256G | 11s | 4s | 64% | | +------+---------------+--------------+----------------+ | | 512G | 36s | 9s | 75% | +----------------+------+---------------+--------------+----------------+ [1] Qemu commands to hotplug 256G/512G memory for a VM: object_add memory-backend-ram,id=3Dhotmem0,size=3D256G/512G,share=3Don device_add virtio-mem-pci,id=3Dvmem1,memdev=3Dhotmem0,bus=3Dport1 qom-set vmem1 requested-size 256G/512G (Plug Memory) qom-set vmem1 requested-size 0G (Unplug Memory) [2] Hardware : Intel Icelake server Guest Kernel : v7.0-rc4 Qemu : v9.0.0 Launch VM : qemu-system-x86_64 -accel kvm -cpu host \ -drive file=3D./Centos10_cloud.qcow2,format=3Dqcow2,if=3Dvirtio \ -drive file=3D./seed.img,format=3Draw,if=3Dvirtio \ -smp 3,cores=3D3,threads=3D1,sockets=3D1,maxcpus=3D3 \ -m 2G,slots=3D10,maxmem=3D2052472M \ -device pcie-root-port,id=3Dport1,bus=3Dpcie.0,slot=3D1,multifunction= =3Don \ -device pcie-root-port,id=3Dport2,bus=3Dpcie.0,slot=3D2 \ -nographic -machine q35 \ -nic user,hostfwd=3Dtcp::3000-:22 Guest kernel auto-onlines newly added memory blocks: echo online > /sys/devices/system/memory/auto_online_blocks [3] The time from typing the QEMU commands in [1] to when the output of 'grep MemTotal /proc/meminfo' on Guest reflects that all hotplugged memory is recognized. Reported-by: Nanhai Zou Reported-by: Chen Zhang Tested-by: Yuan Liu Reviewed-by: Tim Chen Reviewed-by: Qiuxu Zhuo Reviewed-by: Yu C Chen Reviewed-by: Pan Deng Reviewed-by: Nanhai Zou Co-developed-by: Tianyou Li Signed-off-by: Tianyou Li Signed-off-by: Yuan Liu --- Documentation/mm/physical_memory.rst | 11 +++++ drivers/base/memory.c | 6 +++ include/linux/mmzone.h | 44 +++++++++++++++++++ mm/internal.h | 8 +--- mm/memory_hotplug.c | 12 +----- mm/mm_init.c | 64 +++++++++++++++++----------- 6 files changed, 102 insertions(+), 43 deletions(-) diff --git a/Documentation/mm/physical_memory.rst b/Documentation/mm/physic= al_memory.rst index b76183545e5b..e47e96ef6a6d 100644 --- a/Documentation/mm/physical_memory.rst +++ b/Documentation/mm/physical_memory.rst @@ -483,6 +483,17 @@ General ``present_pages`` should use ``get_online_mems()`` to get a stable value= . It is initialized by ``calculate_node_totalpages()``. =20 +``pages_with_online_memmap`` + Tracks pages within the zone that have an online memmap (present pages a= nd + memory holes whose memmap has been initialized). When ``spanned_pages`` = =3D=3D + ``pages_with_online_memmap``, ``pfn_to_page()`` can be performed without + further checks on any PFN within the zone span. + + Note: this counter may temporarily undercount when pages with an online + memmap exist outside the current zone span. Growing the zone to cover su= ch + pages and later shrinking it back may result in a "too small" value. Thi= s is + safe: it merely prevents detecting a contiguous zone. + ``present_early_pages`` The present pages existing within the zone located on memory available s= ince early boot, excluding hotplugged memory. Defined only when diff --git a/drivers/base/memory.c b/drivers/base/memory.c index a3091924918b..2b6b4e5508af 100644 --- a/drivers/base/memory.c +++ b/drivers/base/memory.c @@ -246,6 +246,7 @@ static int memory_block_online(struct memory_block *mem) nr_vmemmap_pages =3D mem->altmap->free; =20 mem_hotplug_begin(); + clear_zone_contiguous(zone); if (nr_vmemmap_pages) { ret =3D mhp_init_memmap_on_memory(start_pfn, nr_vmemmap_pages, zone); if (ret) @@ -270,6 +271,7 @@ static int memory_block_online(struct memory_block *mem) =20 mem->zone =3D zone; out: + set_zone_contiguous(zone); mem_hotplug_done(); return ret; } @@ -282,6 +284,7 @@ static int memory_block_offline(struct memory_block *me= m) unsigned long start_pfn =3D section_nr_to_pfn(mem->start_section_nr); unsigned long nr_pages =3D PAGES_PER_SECTION * sections_per_block; unsigned long nr_vmemmap_pages =3D 0; + struct zone *zone; int ret; =20 if (!mem->zone) @@ -294,7 +297,9 @@ static int memory_block_offline(struct memory_block *me= m) if (mem->altmap) nr_vmemmap_pages =3D mem->altmap->free; =20 + zone =3D mem->zone; mem_hotplug_begin(); + clear_zone_contiguous(zone); if (nr_vmemmap_pages) adjust_present_page_count(pfn_to_page(start_pfn), mem->group, -nr_vmemmap_pages); @@ -314,6 +319,7 @@ static int memory_block_offline(struct memory_block *me= m) =20 mem->zone =3D NULL; out: + set_zone_contiguous(zone); mem_hotplug_done(); return ret; } diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h index 3e51190a55e4..011df76a03b6 100644 --- a/include/linux/mmzone.h +++ b/include/linux/mmzone.h @@ -943,6 +943,17 @@ struct zone { * cma pages is present pages that are assigned for CMA use * (MIGRATE_CMA). * + * pages_with_online_memmap tracks pages in the zone that have an + * online memmap (present pages and holes whose memmap was initialized). + * When spanned_pages =3D=3D pages_with_online_memmap, pfn_to_page() can + * be performed without further checks on any PFN in the zone span. + * + * Note: pages_with_online_memmap may temporarily undercount when pages + * with an online memmap exist outside the current zone span (e.g., from + * init_unavailable_range() during boot). Growing the zone to cover such + * pages and later shrinking it back may result in a "too small" value. + * This is safe: it merely prevents detecting a contiguous zone. + * * So present_pages may be used by memory hotplug or memory power * management logic to figure out unmanaged pages by checking * (present_pages - managed_pages). And managed_pages should be used @@ -967,6 +978,7 @@ struct zone { atomic_long_t managed_pages; unsigned long spanned_pages; unsigned long present_pages; + unsigned long pages_with_online_memmap; #if defined(CONFIG_MEMORY_HOTPLUG) unsigned long present_early_pages; #endif @@ -1601,6 +1613,38 @@ static inline bool zone_is_zone_device(const struct = zone *zone) } #endif =20 +/** + * zone_is_contiguous - test whether a zone is contiguous + * @zone: the zone to test. + * + * In a contiguous zone, it is valid to call pfn_to_page() on any PFN in t= he + * spanned zone without requiring pfn_valid() or pfn_to_online_page() chec= ks. + * + * Note that missing synchronization with memory offlining makes any PFN + * traversal prone to races. + * + * ZONE_DEVICE zones are always marked non-contiguous. + * + * Return: true if contiguous, otherwise false. + */ +static inline bool zone_is_contiguous(const struct zone *zone) +{ + return zone->contiguous; +} + +static inline void set_zone_contiguous(struct zone *zone) +{ + if (zone_is_zone_device(zone)) + return; + if (zone->spanned_pages =3D=3D zone->pages_with_online_memmap) + zone->contiguous =3D true; +} + +static inline void clear_zone_contiguous(struct zone *zone) +{ + zone->contiguous =3D false; +} + /* * Returns true if a zone has pages managed by the buddy allocator. * All the reclaim decisions have to use this function rather than diff --git a/mm/internal.h b/mm/internal.h index cb0af847d7d9..92fee035c3f2 100644 --- a/mm/internal.h +++ b/mm/internal.h @@ -793,21 +793,15 @@ extern struct page *__pageblock_pfn_to_page(unsigned = long start_pfn, static inline struct page *pageblock_pfn_to_page(unsigned long start_pfn, unsigned long end_pfn, struct zone *zone) { - if (zone->contiguous) + if (zone_is_contiguous(zone)) return pfn_to_page(start_pfn); =20 return __pageblock_pfn_to_page(start_pfn, end_pfn, zone); } =20 -void set_zone_contiguous(struct zone *zone); bool pfn_range_intersects_zones(int nid, unsigned long start_pfn, unsigned long nr_pages); =20 -static inline void clear_zone_contiguous(struct zone *zone) -{ - zone->contiguous =3D false; -} - extern int __isolate_free_page(struct page *page, unsigned int order); extern void __putback_isolated_page(struct page *page, unsigned int order, int mt); diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c index bc805029da51..cd9c89de6ed2 100644 --- a/mm/memory_hotplug.c +++ b/mm/memory_hotplug.c @@ -565,18 +565,13 @@ void remove_pfn_range_from_zone(struct zone *zone, =20 /* * Zone shrinking code cannot properly deal with ZONE_DEVICE. So - * we will not try to shrink the zones - which is okay as - * set_zone_contiguous() cannot deal with ZONE_DEVICE either way. + * we will not try to shrink the zones. */ if (zone_is_zone_device(zone)) return; =20 - clear_zone_contiguous(zone); - shrink_zone_span(zone, start_pfn, start_pfn + nr_pages); update_pgdat_span(pgdat); - - set_zone_contiguous(zone); } =20 /** @@ -753,8 +748,6 @@ void move_pfn_range_to_zone(struct zone *zone, unsigned= long start_pfn, struct pglist_data *pgdat =3D zone->zone_pgdat; int nid =3D pgdat->node_id; =20 - clear_zone_contiguous(zone); - if (zone_is_empty(zone)) init_currently_empty_zone(zone, start_pfn, nr_pages); resize_zone_range(zone, start_pfn, nr_pages); @@ -782,8 +775,6 @@ void move_pfn_range_to_zone(struct zone *zone, unsigned= long start_pfn, memmap_init_range(nr_pages, nid, zone_idx(zone), start_pfn, 0, MEMINIT_HOTPLUG, altmap, migratetype, isolate_pageblock); - - set_zone_contiguous(zone); } =20 struct auto_movable_stats { @@ -1079,6 +1070,7 @@ void adjust_present_page_count(struct page *page, str= uct memory_group *group, if (early_section(__pfn_to_section(page_to_pfn(page)))) zone->present_early_pages +=3D nr_pages; zone->present_pages +=3D nr_pages; + zone->pages_with_online_memmap +=3D nr_pages; zone->zone_pgdat->node_present_pages +=3D nr_pages; =20 if (group && movable) diff --git a/mm/mm_init.c b/mm/mm_init.c index df34797691bd..b8187a22e90e 100644 --- a/mm/mm_init.c +++ b/mm/mm_init.c @@ -842,7 +842,7 @@ overlap_memmap_init(unsigned long zone, unsigned long *= pfn) * zone/node above the hole except for the trailing pages in the last * section that will be appended to the zone/node below. */ -static void __init init_unavailable_range(unsigned long spfn, +static unsigned long __init init_unavailable_range(unsigned long spfn, unsigned long epfn, int zone, int node) { @@ -858,6 +858,36 @@ static void __init init_unavailable_range(unsigned lon= g spfn, if (pgcnt) pr_info("On node %d, zone %s: %lld pages in unavailable ranges\n", node, zone_names[zone], pgcnt); + return pgcnt; +} + +/* + * Initialize unavailable range [spfn, epfn) while accounting only the pag= es + * that fall within the zone span towards pages_with_online_memmap. Pages + * outside the zone span are still initialized but not accounted. + */ +static void __init init_unavailable_range_for_zone(struct zone *zone, + unsigned long spfn, + unsigned long epfn) +{ + int nid =3D zone_to_nid(zone); + int zid =3D zone_idx(zone); + unsigned long in_zone_start; + unsigned long in_zone_end; + + in_zone_start =3D clamp(spfn, zone->zone_start_pfn, zone_end_pfn(zone)); + in_zone_end =3D clamp(epfn, zone->zone_start_pfn, zone_end_pfn(zone)); + + if (spfn < in_zone_start) + init_unavailable_range(spfn, in_zone_start, zid, nid); + + if (in_zone_start < in_zone_end) + zone->pages_with_online_memmap +=3D + init_unavailable_range(in_zone_start, in_zone_end, + zid, nid); + + if (in_zone_end < epfn) + init_unavailable_range(in_zone_end, epfn, zid, nid); } =20 /* @@ -956,9 +986,10 @@ static void __init memmap_init_zone_range(struct zone = *zone, memmap_init_range(end_pfn - start_pfn, nid, zone_id, start_pfn, zone_end_pfn, MEMINIT_EARLY, NULL, MIGRATE_MOVABLE, false); + zone->pages_with_online_memmap +=3D end_pfn - start_pfn; =20 if (*hole_pfn < start_pfn) - init_unavailable_range(*hole_pfn, start_pfn, zone_id, nid); + init_unavailable_range_for_zone(zone, *hole_pfn, start_pfn); =20 *hole_pfn =3D end_pfn; } @@ -996,8 +1027,11 @@ static void __init memmap_init(void) #else end_pfn =3D round_up(end_pfn, MAX_ORDER_NR_PAGES); #endif - if (hole_pfn < end_pfn) - init_unavailable_range(hole_pfn, end_pfn, zone_id, nid); + if (hole_pfn < end_pfn) { + struct zone *zone =3D &NODE_DATA(nid)->node_zones[zone_id]; + + init_unavailable_range_for_zone(zone, hole_pfn, end_pfn); + } } =20 #ifdef CONFIG_ZONE_DEVICE @@ -2261,28 +2295,6 @@ void __init init_cma_pageblock(struct page *page) } #endif =20 -void set_zone_contiguous(struct zone *zone) -{ - unsigned long block_start_pfn =3D zone->zone_start_pfn; - unsigned long block_end_pfn; - - block_end_pfn =3D pageblock_end_pfn(block_start_pfn); - for (; block_start_pfn < zone_end_pfn(zone); - block_start_pfn =3D block_end_pfn, - block_end_pfn +=3D pageblock_nr_pages) { - - block_end_pfn =3D min(block_end_pfn, zone_end_pfn(zone)); - - if (!__pageblock_pfn_to_page(block_start_pfn, - block_end_pfn, zone)) - return; - cond_resched(); - } - - /* We confirm that there is no hole */ - zone->contiguous =3D true; -} - /* * Check if a PFN range intersects multiple zones on one or more * NUMA nodes. Specify the @nid argument if it is known that this --=20 2.47.3