| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Thu, 4 Feb 2016 16:49:29 -0800
From: Andrew Morton <akpm@...ux-foundation.org>
To: Joonsoo Kim <js1304@...il.com>
Cc: Vlastimil Babka <vbabka@...e.cz>, Aaron Lu <aaron.lu@...el.com>,
Mel Gorman <mgorman@...e.de>, Rik van Riel <riel@...hat.com>,
David Rientjes <rientjes@...gle.com>,
linux-kernel@...r.kernel.org, linux-mm@...ck.org,
Joonsoo Kim <iamjoonsoo.kim@....com>
Subject: Re: [PATCH v2 3/3] mm/compaction: speed up pageblock_pfn_to_page()
when zone is contiguous
On Thu, 4 Feb 2016 15:19:35 +0900 Joonsoo Kim <js1304@...il.com> wrote:
> There is a performance drop report due to hugepage allocation and in there
> half of cpu time are spent on pageblock_pfn_to_page() in compaction [1].
> In that workload, compaction is triggered to make hugepage but most of
> pageblocks are un-available for compaction due to pageblock type and
> skip bit so compaction usually fails. Most costly operations in this case
> is to find valid pageblock while scanning whole zone range. To check
> if pageblock is valid to compact, valid pfn within pageblock is required
> and we can obtain it by calling pageblock_pfn_to_page(). This function
> checks whether pageblock is in a single zone and return valid pfn
> if possible. Problem is that we need to check it every time before
> scanning pageblock even if we re-visit it and this turns out to
> be very expensive in this workload.
>
> Although we have no way to skip this pageblock check in the system
> where hole exists at arbitrary position, we can use cached value for
> zone continuity and just do pfn_to_page() in the system where hole doesn't
> exist. This optimization considerably speeds up in above workload.
>
> Before vs After
> Max: 1096 MB/s vs 1325 MB/s
> Min: 635 MB/s 1015 MB/s
> Avg: 899 MB/s 1194 MB/s
>
> Avg is improved by roughly 30% [2].
>
> [1]: http://www.spinics.net/lists/linux-mm/msg97378.html
> [2]: https://lkml.org/lkml/2015/12/9/23
>
> ...
>
> --- a/include/linux/memory_hotplug.h
> +++ b/include/linux/memory_hotplug.h
> @@ -196,6 +196,9 @@ void put_online_mems(void);
> void mem_hotplug_begin(void);
> void mem_hotplug_done(void);
>
> +extern void set_zone_contiguous(struct zone *zone);
> +extern void clear_zone_contiguous(struct zone *zone);
> +
> #else /* ! CONFIG_MEMORY_HOTPLUG */
> /*
> * Stub functions for when hotplug is off
Was it really intended that these declarations only exist if
CONFIG_MEMORY_HOTPLUG? Seems unrelated.
The i386 allnocofnig build fails in preditable ways so I fixed that up
as below, but it seems wrong.
From: Andrew Morton <akpm@...ux-foundation.org>
Move CONFIG_MEMORY_HOTPLUG code into memory_hotplug.c, fix
CONFIG_MEMORY_HOTPLUG=n build.
Cc: Joonsoo Kim <iamjoonsoo.kim@....com>
Cc: Aaron Lu <aaron.lu@...el.com>
Cc: Aaron Lu <aaron.lu@...el.com>
Cc: Mel Gorman <mgorman@...e.de>
Cc: Rik van Riel <riel@...hat.com>
Cc: David Rientjes <rientjes@...gle.com>
Cc: Vlastimil Babka <vbabka@...e.cz>
Signed-off-by: Andrew Morton <akpm@...ux-foundation.org>
---
include/linux/memory_hotplug.h | 12 +++++++++++-
include/linux/mmzone.h | 2 ++
mm/memory_hotplug.c | 27 +++++++++++++++++++++++++++
mm/page_alloc.c | 27 ---------------------------
4 files changed, 40 insertions(+), 28 deletions(-)
diff -puN include/linux/memory_hotplug.h~mm-compaction-speed-up-pageblock_pfn_to_page-when-zone-is-contiguous-fix include/linux/memory_hotplug.h
--- a/include/linux/memory_hotplug.h~mm-compaction-speed-up-pageblock_pfn_to_page-when-zone-is-contiguous-fix
+++ a/include/linux/memory_hotplug.h
@@ -200,7 +200,10 @@ void mem_hotplug_done(void);
extern void set_zone_contiguous(struct zone *zone);
extern void clear_zone_contiguous(struct zone *zone);
-
+static inline bool zone_contiguous(struct zone *zone)
+{
+ return zone->contiguous;
+}
#else /* ! CONFIG_MEMORY_HOTPLUG */
/*
* Stub functions for when hotplug is off
@@ -243,6 +246,13 @@ static inline void put_online_mems(void)
static inline void mem_hotplug_begin(void) {}
static inline void mem_hotplug_done(void) {}
+static inline void set_zone_contiguous(struct zone *zone) {}
+static inline void clear_zone_contiguous(struct zone *zone) {}
+static inline bool zone_contiguous(struct zone *zone)
+{
+ return false;
+}
+
#endif /* ! CONFIG_MEMORY_HOTPLUG */
#ifdef CONFIG_MEMORY_HOTREMOVE
diff -puN include/linux/mmzone.h~mm-compaction-speed-up-pageblock_pfn_to_page-when-zone-is-contiguous-fix include/linux/mmzone.h
--- a/include/linux/mmzone.h~mm-compaction-speed-up-pageblock_pfn_to_page-when-zone-is-contiguous-fix
+++ a/include/linux/mmzone.h
@@ -522,7 +522,9 @@ struct zone {
bool compact_blockskip_flush;
#endif
+#ifdef CONFIG_MEMORY_HOTPLUG
bool contiguous;
+#endif
ZONE_PADDING(_pad3_)
/* Zone statistics */
diff -puN mm/memory_hotplug.c~mm-compaction-speed-up-pageblock_pfn_to_page-when-zone-is-contiguous-fix mm/memory_hotplug.c
--- a/mm/memory_hotplug.c~mm-compaction-speed-up-pageblock_pfn_to_page-when-zone-is-contiguous-fix
+++ a/mm/memory_hotplug.c
@@ -130,6 +130,33 @@ void mem_hotplug_done(void)
memhp_lock_release();
}
+void set_zone_contiguous(struct zone *zone)
+{
+ unsigned long block_start_pfn = zone->zone_start_pfn;
+ unsigned long block_end_pfn;
+ unsigned long pfn;
+
+ block_end_pfn = ALIGN(block_start_pfn + 1, pageblock_nr_pages);
+ for (; block_start_pfn < zone_end_pfn(zone);
+ block_start_pfn = block_end_pfn,
+ block_end_pfn += pageblock_nr_pages) {
+
+ block_end_pfn = min(block_end_pfn, zone_end_pfn(zone));
+
+ if (!__pageblock_pfn_to_page(block_start_pfn,
+ block_end_pfn, zone))
+ return;
+ }
+
+ /* We confirm that there is no hole */
+ zone->contiguous = true;
+}
+
+void clear_zone_contiguous(struct zone *zone)
+{
+ zone->contiguous = false;
+}
+
/* add this memory to iomem resource */
static struct resource *register_memory_resource(u64 start, u64 size)
{
diff -puN mm/page_alloc.c~mm-compaction-speed-up-pageblock_pfn_to_page-when-zone-is-contiguous-fix mm/page_alloc.c
--- a/mm/page_alloc.c~mm-compaction-speed-up-pageblock_pfn_to_page-when-zone-is-contiguous-fix
+++ a/mm/page_alloc.c
@@ -1347,33 +1347,6 @@ struct page *__pageblock_pfn_to_page(uns
return start_page;
}
-void set_zone_contiguous(struct zone *zone)
-{
- unsigned long block_start_pfn = zone->zone_start_pfn;
- unsigned long block_end_pfn;
- unsigned long pfn;
-
- block_end_pfn = ALIGN(block_start_pfn + 1, pageblock_nr_pages);
- for (; block_start_pfn < zone_end_pfn(zone);
- block_start_pfn = block_end_pfn,
- block_end_pfn += pageblock_nr_pages) {
-
- block_end_pfn = min(block_end_pfn, zone_end_pfn(zone));
-
- if (!__pageblock_pfn_to_page(block_start_pfn,
- block_end_pfn, zone))
- return;
- }
-
- /* We confirm that there is no hole */
- zone->contiguous = true;
-}
-
-void clear_zone_contiguous(struct zone *zone)
-{
- zone->contiguous = false;
-}
-
#ifdef CONFIG_CMA
/* Free whole pageblock and set its migration type to MIGRATE_CMA. */
void __init init_cma_reserved_pageblock(struct page *page)
_
Powered by blists - more mailing lists