Subject: [to-be-updated] mm-create-a-separate-slab-for-page-ptl-allocation.patch removed from -mm tree
To: kirill.shutemov@xxxxxxxxxxxxxxx,fengguang.wu@xxxxxxxxx,mingo@xxxxxxxxxx,peterz@xxxxxxxxxxxxx,mm-commits@xxxxxxxxxxxxxxx
From: akpm@xxxxxxxxxxxxxxxxxxxx
Date: Wed, 06 Nov 2013 13:38:26 -0800
The patch titled
Subject: mm: create a separate slab for page->ptl allocation
has been removed from the -mm tree. Its filename was
mm-create-a-separate-slab-for-page-ptl-allocation.patch
This patch was dropped because an updated version will be merged
------------------------------------------------------
From: "Kirill A. Shutemov" <kirill.shutemov@xxxxxxxxxxxxxxx>
Subject: mm: create a separate slab for page->ptl allocation
If DEBUG_SPINLOCK and DEBUG_LOCK_ALLOC are enabled spinlock_t on x86_64 is
72 bytes. For page->ptl they will be allocated from kmalloc-96 slab, so
we loose 24 on each. An average system can easily allocate few tens
thousands of page->ptl and overhead is significant.
Let's create a separate slab for page->ptl allocation to solve this.
Fengguang said:
: In a 4p server, we noticed up to +469.1% increase in will-it-scale page_fault3
: test case and +199.8% in vm-scalability case-shm-pread-seq-mt.
:
: 5c02216ce3110aab070d 5a58baaa0a1af0a43d7c
: ------------------------ ------------------------
: 300409.00 +440.2% 1622770.80 TOTAL will-it-scale.page_fault3.90.threads
:
: 5c02216ce3110aab070d 5a58baaa0a1af0a43d7c
: ------------------------ ------------------------
: 291257.80 +469.1% 1657582.20 TOTAL will-it-scale.page_fault3.120.threads
:
: ...
:
: 5c02216ce3110aab070d 5a58baaa0a1af0a43d7c
: ------------------------ ------------------------
: 4034831.40 +199.8% 12095649.80 TOTAL vm-scalability.throughput
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@xxxxxxxxxxxxxxx>
Tested-by: Fengguang Wu <fengguang.wu@xxxxxxxxx>
Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Cc: Ingo Molnar <mingo@xxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
include/linux/mm.h | 8 ++++++++
init/main.c | 2 +-
mm/memory.c | 12 ++++++++++--
3 files changed, 19 insertions(+), 3 deletions(-)
diff -puN include/linux/mm.h~mm-create-a-separate-slab-for-page-ptl-allocation include/linux/mm.h
--- a/include/linux/mm.h~mm-create-a-separate-slab-for-page-ptl-allocation
+++ a/include/linux/mm.h
@@ -1317,6 +1317,7 @@ static inline pmd_t *pmd_alloc(struct mm
#endif /* CONFIG_MMU && !__ARCH_HAS_4LEVEL_HACK */
#if USE_SPLIT_PTE_PTLOCKS
+void __init ptlock_cache_init(void);
bool __ptlock_alloc(struct page *page);
void __ptlock_free(struct page *page);
static inline bool ptlock_alloc(struct page *page)
@@ -1369,6 +1370,7 @@ static inline void pte_lock_deinit(struc
}
#else /* !USE_SPLIT_PTE_PTLOCKS */
+static inline void ptlock_cache_init(void) {}
/*
* We use mm->page_table_lock to guard all pagetable pages of the mm.
*/
@@ -1380,6 +1382,12 @@ static inline bool ptlock_init(struct pa
static inline void pte_lock_deinit(struct page *page) {}
#endif /* USE_SPLIT_PTE_PTLOCKS */
+static inline void pgtable_init(void)
+{
+ ptlock_cache_init();
+ pgtable_cache_init();
+}
+
static inline bool pgtable_page_ctor(struct page *page)
{
inc_zone_page_state(page, NR_PAGETABLE);
diff -puN init/main.c~mm-create-a-separate-slab-for-page-ptl-allocation init/main.c
--- a/init/main.c~mm-create-a-separate-slab-for-page-ptl-allocation
+++ a/init/main.c
@@ -476,7 +476,7 @@ static void __init mm_init(void)
mem_init();
kmem_cache_init();
percpu_init_late();
- pgtable_cache_init();
+ pgtable_init();
vmalloc_init();
}
diff -puN mm/memory.c~mm-create-a-separate-slab-for-page-ptl-allocation mm/memory.c
--- a/mm/memory.c~mm-create-a-separate-slab-for-page-ptl-allocation
+++ a/mm/memory.c
@@ -4272,11 +4272,19 @@ void copy_user_huge_page(struct page *ds
#endif /* CONFIG_TRANSPARENT_HUGEPAGE || CONFIG_HUGETLBFS */
#if USE_SPLIT_PTE_PTLOCKS
+struct kmem_cache *page_ptl_cachep;
+void __init ptlock_cache_init(void)
+{
+ if (sizeof(spinlock_t) > sizeof(long))
+ page_ptl_cachep = kmem_cache_create("page->ptl",
+ sizeof(spinlock_t), 0, SLAB_PANIC, NULL);
+}
+
bool __ptlock_alloc(struct page *page)
{
spinlock_t *ptl;
- ptl = kmalloc(sizeof(spinlock_t), GFP_KERNEL);
+ ptl = kmem_cache_alloc(page_ptl_cachep, GFP_KERNEL);
if (!ptl)
return false;
page->ptl = (unsigned long)ptl;
@@ -4286,6 +4294,6 @@ bool __ptlock_alloc(struct page *page)
void __ptlock_free(struct page *page)
{
if (sizeof(spinlock_t) > sizeof(page->ptl))
- kfree((spinlock_t *)page->ptl);
+ kmem_cache_free(page_ptl_cachep, (spinlock_t *)page->ptl);
}
#endif
_
Patches currently in -mm which might be from kirill.shutemov@xxxxxxxxxxxxxxx are
x86-mm-get-aslr-work-for-hugetlb-mappings.patch
mm-huge_memoryc-fix-stale-comments-of-transparent_hugepage_flags.patch
mm-thp-cleanup-mv-alloc_hugepage-to-better-place.patch
mm-thp-khugepaged-add-policy-for-finding-target-node.patch
mm-thp-khugepaged-add-policy-for-finding-target-node-fix.patch
mm-drop-actor-argument-of-do_generic_file_read.patch
mm-drop-actor-argument-of-do_generic_file_read-fix.patch
mm-avoid-increase-sizeofstruct-page-due-to-split-page-table-lock.patch
mm-rename-use_split_ptlocks-to-use_split_pte_ptlocks.patch
mm-convert-mm-nr_ptes-to-atomic_long_t.patch
mm-introduce-api-for-split-page-table-lock-for-pmd-level.patch
mm-thp-change-pmd_trans_huge_lock-to-return-taken-lock.patch
mm-thp-move-ptl-taking-inside-page_check_address_pmd.patch
mm-thp-do-not-access-mm-pmd_huge_pte-directly.patch
mm-hugetlb-convert-hugetlbfs-to-use-split-pmd-lock.patch
mm-hugetlb-convert-hugetlbfs-to-use-split-pmd-lock-checkpatch-fixes.patch
mm-convert-the-rest-to-new-page-table-lock-api.patch
mm-implement-split-page-table-lock-for-pmd-level.patch
x86-mm-enable-split-page-table-lock-for-pmd-level.patch
x86-mm-enable-split-page-table-lock-for-pmd-level-checkpatch-fixes.patch
x86-add-missed-pgtable_pmd_page_ctor-dtor-calls-for-preallocated-pmds.patch
cris-fix-potential-null-pointer-dereference.patch
m32r-fix-potential-null-pointer-dereference.patch
xtensa-fix-potential-null-pointer-dereference.patch
mm-allow-pgtable_page_ctor-to-fail.patch
microblaze-add-missing-pgtable_page_ctor-dtor-calls.patch
mn10300-add-missing-pgtable_page_ctor-dtor-calls.patch
openrisc-add-missing-pgtable_page_ctor-dtor-calls.patch
alpha-handle-pgtable_page_ctor-fail.patch
arc-handle-pgtable_page_ctor-fail.patch
arm-handle-pgtable_page_ctor-fail.patch
arm64-handle-pgtable_page_ctor-fail.patch
avr32-handle-pgtable_page_ctor-fail.patch
cris-handle-pgtable_page_ctor-fail.patch
frv-handle-pgtable_page_ctor-fail.patch
hexagon-handle-pgtable_page_ctor-fail.patch
ia64-handle-pgtable_page_ctor-fail.patch
m32r-handle-pgtable_page_ctor-fail.patch
m68k-handle-pgtable_page_ctor-fail.patch
m68k-handle-pgtable_page_ctor-fail-fix.patch
m68k-handle-pgtable_page_ctor-fail-fix-fix.patch
metag-handle-pgtable_page_ctor-fail.patch
mips-handle-pgtable_page_ctor-fail.patch
parisc-handle-pgtable_page_ctor-fail.patch
powerpc-handle-pgtable_page_ctor-fail.patch
s390-handle-pgtable_page_ctor-fail.patch
score-handle-pgtable_page_ctor-fail.patch
sh-handle-pgtable_page_ctor-fail.patch
sparc-handle-pgtable_page_ctor-fail.patch
tile-handle-pgtable_page_ctor-fail.patch
um-handle-pgtable_page_ctor-fail.patch
unicore32-handle-pgtable_page_ctor-fail.patch
x86-handle-pgtable_page_ctor-fail.patch
xtensa-handle-pgtable_page_ctor-fail.patch
iommu-arm-smmu-handle-pgtable_page_ctor-fail.patch
xtensa-use-buddy-allocator-for-pte-table.patch
mm-dynamically-allocate-page-ptl-if-it-cannot-be-embedded-to-struct-page.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
