From: Mike Rapoport <rppt@kernel.org>
To: Baoquan He <bhe@redhat.com>
Cc: linux-kernel@vger.kernel.org, linux-mm@kvack.org,
akpm@linux-foundation.org, gopakumarr@vmware.com,
david@redhat.com
Subject: Re: [PATCH v2 1/5] mm: memmap defer init dosn't work as expected
Date: Mon, 21 Dec 2020 08:32:52 +0200 [thread overview]
Message-ID: <20201221063252.GC392325@kernel.org> (raw)
In-Reply-To: <20201220082754.6900-2-bhe@redhat.com>
On Sun, Dec 20, 2020 at 04:27:50PM +0800, Baoquan He wrote:
> VMware observed a performance regression during memmap init on their platform,
> and bisected to commit 73a6e474cb376 ("mm: memmap_init: iterate over memblock
> regions rather that check each PFN") causing it.
>
> Before the commit:
>
> [0.033176] Normal zone: 1445888 pages used for memmap
> [0.033176] Normal zone: 89391104 pages, LIFO batch:63
> [0.035851] ACPI: PM-Timer IO Port: 0x448
>
> With commit
>
> [0.026874] Normal zone: 1445888 pages used for memmap
> [0.026875] Normal zone: 89391104 pages, LIFO batch:63
> [2.028450] ACPI: PM-Timer IO Port: 0x448
>
> The root cause is the current memmap defer init doesn't work as expected.
> Before, memmap_init_zone() was used to do memmap init of one whole zone, to
> initialize all low zones of one numa node, but defer memmap init of the
> last zone in that numa node. However, since commit 73a6e474cb376, function
> memmap_init() is adapted to iterater over memblock regions inside one zone,
> then call memmap_init_zone() to do memmap init for each region.
>
> E.g, on VMware's system, the memory layout is as below, there are two memory
> regions in node 2. The current code will mistakenly initialize the whole 1st
> region [mem 0xab00000000-0xfcffffffff], then do memmap defer to iniatialize
> only one memmory section on the 2nd region [mem 0x10000000000-0x1033fffffff].
> In fact, we only expect to see that there's only one memory section's memmap
> initialized. That's why more time is costed at the time.
>
> [ 0.008842] ACPI: SRAT: Node 0 PXM 0 [mem 0x00000000-0x0009ffff]
> [ 0.008842] ACPI: SRAT: Node 0 PXM 0 [mem 0x00100000-0xbfffffff]
> [ 0.008843] ACPI: SRAT: Node 0 PXM 0 [mem 0x100000000-0x55ffffffff]
> [ 0.008844] ACPI: SRAT: Node 1 PXM 1 [mem 0x5600000000-0xaaffffffff]
> [ 0.008844] ACPI: SRAT: Node 2 PXM 2 [mem 0xab00000000-0xfcffffffff]
> [ 0.008845] ACPI: SRAT: Node 2 PXM 2 [mem 0x10000000000-0x1033fffffff]
>
> Now, let's add a parameter 'zone_end_pfn' to memmap_init_zone() to pass
> down the real zone end pfn so that defer_init() can use it to judge whether
> defer need be taken in zone wide.
>
> Fixes: commit 73a6e474cb376 ("mm: memmap_init: iterate over memblock regions rather that check each PFN")
> Reported-by: Rahul Gopakumar <gopakumarr@vmware.com>
> Signed-off-by: Baoquan He <bhe@redhat.com>
> Cc: stable@vger.kernel.org
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com>
> ---
> arch/ia64/mm/init.c | 4 ++--
> include/linux/mm.h | 5 +++--
> mm/memory_hotplug.c | 2 +-
> mm/page_alloc.c | 8 +++++---
> 4 files changed, 11 insertions(+), 8 deletions(-)
>
> diff --git a/arch/ia64/mm/init.c b/arch/ia64/mm/init.c
> index 9b5acf8fb092..e76386a3479e 100644
> --- a/arch/ia64/mm/init.c
> +++ b/arch/ia64/mm/init.c
> @@ -536,7 +536,7 @@ virtual_memmap_init(u64 start, u64 end, void *arg)
>
> if (map_start < map_end)
> memmap_init_zone((unsigned long)(map_end - map_start),
> - args->nid, args->zone, page_to_pfn(map_start),
> + args->nid, args->zone, page_to_pfn(map_start), page_to_pfn(map_end),
> MEMINIT_EARLY, NULL, MIGRATE_MOVABLE);
> return 0;
> }
> @@ -546,7 +546,7 @@ memmap_init (unsigned long size, int nid, unsigned long zone,
> unsigned long start_pfn)
> {
> if (!vmem_map) {
> - memmap_init_zone(size, nid, zone, start_pfn,
> + memmap_init_zone(size, nid, zone, start_pfn, start_pfn + size,
> MEMINIT_EARLY, NULL, MIGRATE_MOVABLE);
> } else {
> struct page *start;
> diff --git a/include/linux/mm.h b/include/linux/mm.h
> index e4e5be20b0c2..92e06ea053f4 100644
> --- a/include/linux/mm.h
> +++ b/include/linux/mm.h
> @@ -2432,8 +2432,9 @@ extern int __meminit early_pfn_to_nid(unsigned long pfn);
> #endif
>
> extern void set_dma_reserve(unsigned long new_dma_reserve);
> -extern void memmap_init_zone(unsigned long, int, unsigned long, unsigned long,
> - enum meminit_context, struct vmem_altmap *, int migratetype);
> +extern void memmap_init_zone(unsigned long, int, unsigned long,
> + unsigned long, unsigned long, enum meminit_context,
> + struct vmem_altmap *, int migratetype);
> extern void setup_per_zone_wmarks(void);
> extern int __meminit init_per_zone_wmark_min(void);
> extern void mem_init(void);
> diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
> index af41fb990820..f9d57b9be8c7 100644
> --- a/mm/memory_hotplug.c
> +++ b/mm/memory_hotplug.c
> @@ -713,7 +713,7 @@ void __ref move_pfn_range_to_zone(struct zone *zone, unsigned long start_pfn,
> * expects the zone spans the pfn range. All the pages in the range
> * are reserved so nobody should be touching them so we should be safe
> */
> - memmap_init_zone(nr_pages, nid, zone_idx(zone), start_pfn,
> + memmap_init_zone(nr_pages, nid, zone_idx(zone), start_pfn, 0,
> MEMINIT_HOTPLUG, altmap, migratetype);
>
> set_zone_contiguous(zone);
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 8cea0823b70e..32645f2e7b96 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -423,6 +423,8 @@ defer_init(int nid, unsigned long pfn, unsigned long end_pfn)
> if (end_pfn < pgdat_end_pfn(NODE_DATA(nid)))
> return false;
>
> + if (NODE_DATA(nid)->first_deferred_pfn != ULONG_MAX)
> + return true;
> /*
> * We start only with one section of pages, more pages are added as
> * needed until the rest of deferred pages are initialized.
> @@ -6116,7 +6118,7 @@ overlap_memmap_init(unsigned long zone, unsigned long *pfn)
> * zone stats (e.g., nr_isolate_pageblock) are touched.
> */
> void __meminit memmap_init_zone(unsigned long size, int nid, unsigned long zone,
> - unsigned long start_pfn,
> + unsigned long start_pfn, unsigned long zone_end_pfn,
> enum meminit_context context,
> struct vmem_altmap *altmap, int migratetype)
> {
> @@ -6152,7 +6154,7 @@ void __meminit memmap_init_zone(unsigned long size, int nid, unsigned long zone,
> if (context == MEMINIT_EARLY) {
> if (overlap_memmap_init(zone, &pfn))
> continue;
> - if (defer_init(nid, pfn, end_pfn))
> + if (defer_init(nid, pfn, zone_end_pfn))
> break;
> }
>
> @@ -6307,7 +6309,7 @@ void __init __weak memmap_init(unsigned long size, int nid,
>
> if (end_pfn > start_pfn) {
> size = end_pfn - start_pfn;
> - memmap_init_zone(size, nid, zone, start_pfn,
> + memmap_init_zone(size, nid, zone, start_pfn, range_end_pfn,
> MEMINIT_EARLY, NULL, MIGRATE_MOVABLE);
> }
>
> --
> 2.17.2
>
--
Sincerely yours,
Mike.
next prev parent reply other threads:[~2020-12-21 6:33 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-12-20 8:27 [PATCH v2 0/5] Fix the incorrect memmep defer init handling and do some cleanup Baoquan He
2020-12-20 8:27 ` [PATCH v2 1/5] mm: memmap defer init dosn't work as expected Baoquan He
2020-12-21 6:32 ` Mike Rapoport [this message]
2020-12-20 8:27 ` [PATCH v2 2/5] mm: rename memmap_init() and memmap_init_zone() Baoquan He
2020-12-21 6:33 ` Mike Rapoport
2020-12-20 8:27 ` [PATCH v2 3/5] mm: simplify parater of function memmap_init_zone() Baoquan He
2020-12-21 6:34 ` Mike Rapoport
2020-12-20 8:27 ` [PATCH v2 4/5] mm: simplify parameter of setup_usemap() Baoquan He
2020-12-21 6:34 ` Mike Rapoport
2020-12-20 8:27 ` [PATCH v2 5/5] mm: remove unneeded local variable in free_area_init_core Baoquan He
2020-12-21 6:35 ` Mike Rapoport
2020-12-23 1:46 ` [PATCH v2 0/5] Fix the incorrect memmep defer init handling and do some cleanup Andrew Morton
2020-12-23 2:05 ` Baoquan He
2020-12-23 8:12 ` Baoquan He
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20201221063252.GC392325@kernel.org \
--to=rppt@kernel.org \
--cc=akpm@linux-foundation.org \
--cc=bhe@redhat.com \
--cc=david@redhat.com \
--cc=gopakumarr@vmware.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).