[PATCH] mm: page_state opt
Optimise page_state manipulations by introducing interrupt unsafe accessors
to page_state fields. Callers must provide their own locking (either
disable interrupts or not update from interrupt context).
Switch over the hot callsites that can easily be moved under interrupts off
sections.
Signed-off-by: Nick Piggin <npiggin@suse.de>
Cc: Hugh Dickins <hugh@veritas.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
diff --git a/include/linux/page-flags.h b/include/linux/page-flags.h
index 32d09c8..dede8d4 100644
--- a/include/linux/page-flags.h
+++ b/include/linux/page-flags.h
@@ -144,22 +144,33 @@
extern void get_page_state(struct page_state *ret);
extern void get_page_state_node(struct page_state *ret, int node);
extern void get_full_page_state(struct page_state *ret);
-extern unsigned long __read_page_state(unsigned long offset);
-extern void __mod_page_state(unsigned long offset, unsigned long delta);
+extern unsigned long read_page_state_offset(unsigned long offset);
+extern void mod_page_state_offset(unsigned long offset, unsigned long delta);
+extern void __mod_page_state_offset(unsigned long offset, unsigned long delta);
#define read_page_state(member) \
- __read_page_state(offsetof(struct page_state, member))
+ read_page_state_offset(offsetof(struct page_state, member))
#define mod_page_state(member, delta) \
- __mod_page_state(offsetof(struct page_state, member), (delta))
+ mod_page_state_offset(offsetof(struct page_state, member), (delta))
-#define inc_page_state(member) mod_page_state(member, 1UL)
-#define dec_page_state(member) mod_page_state(member, 0UL - 1)
-#define add_page_state(member,delta) mod_page_state(member, (delta))
-#define sub_page_state(member,delta) mod_page_state(member, 0UL - (delta))
+#define __mod_page_state(member, delta) \
+ __mod_page_state_offset(offsetof(struct page_state, member), (delta))
-#define mod_page_state_zone(zone, member, delta) \
- do { \
+#define inc_page_state(member) mod_page_state(member, 1UL)
+#define dec_page_state(member) mod_page_state(member, 0UL - 1)
+#define add_page_state(member,delta) mod_page_state(member, (delta))
+#define sub_page_state(member,delta) mod_page_state(member, 0UL - (delta))
+
+#define __inc_page_state(member) __mod_page_state(member, 1UL)
+#define __dec_page_state(member) __mod_page_state(member, 0UL - 1)
+#define __add_page_state(member,delta) __mod_page_state(member, (delta))
+#define __sub_page_state(member,delta) __mod_page_state(member, 0UL - (delta))
+
+#define page_state(member) (*__page_state(offsetof(struct page_state, member)))
+
+#define state_zone_offset(zone, member) \
+({ \
unsigned offset; \
if (is_highmem(zone)) \
offset = offsetof(struct page_state, member##_high); \
@@ -169,7 +180,17 @@
offset = offsetof(struct page_state, member##_dma32); \
else \
offset = offsetof(struct page_state, member##_dma); \
- __mod_page_state(offset, (delta)); \
+ offset; \
+})
+
+#define __mod_page_state_zone(zone, member, delta) \
+ do { \
+ __mod_page_state_offset(state_zone_offset(zone, member), (delta)); \
+ } while (0)
+
+#define mod_page_state_zone(zone, member, delta) \
+ do { \
+ mod_page_state_offset(state_zone_offset(zone, member), (delta)); \
} while (0)
/*
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 7f58077..fd47494 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -424,9 +424,9 @@
return;
list_add(&page->lru, &list);
- mod_page_state(pgfree, 1 << order);
kernel_map_pages(page, 1<<order, 0);
local_irq_save(flags);
+ __mod_page_state(pgfree, 1 << order);
free_pages_bulk(page_zone(page), 1, &list, order);
local_irq_restore(flags);
}
@@ -674,18 +674,14 @@
}
#endif /* CONFIG_PM */
-static void zone_statistics(struct zonelist *zonelist, struct zone *z)
+static void zone_statistics(struct zonelist *zonelist, struct zone *z, int cpu)
{
#ifdef CONFIG_NUMA
- unsigned long flags;
- int cpu;
pg_data_t *pg = z->zone_pgdat;
pg_data_t *orig = zonelist->zones[0]->zone_pgdat;
struct per_cpu_pageset *p;
- local_irq_save(flags);
- cpu = smp_processor_id();
- p = zone_pcp(z,cpu);
+ p = zone_pcp(z, cpu);
if (pg == orig) {
p->numa_hit++;
} else {
@@ -696,7 +692,6 @@
p->local_node++;
else
p->other_node++;
- local_irq_restore(flags);
#endif
}
@@ -716,11 +711,11 @@
if (free_pages_check(page))
return;
- inc_page_state(pgfree);
kernel_map_pages(page, 1, 0);
pcp = &zone_pcp(zone, get_cpu())->pcp[cold];
local_irq_save(flags);
+ __inc_page_state(pgfree);
list_add(&page->lru, &pcp->list);
pcp->count++;
if (pcp->count >= pcp->high)
@@ -753,49 +748,58 @@
* we cheat by calling it from here, in the order > 0 path. Saves a branch
* or two.
*/
-static struct page *
-buffered_rmqueue(struct zone *zone, int order, gfp_t gfp_flags)
+static struct page *buffered_rmqueue(struct zonelist *zonelist,
+ struct zone *zone, int order, gfp_t gfp_flags)
{
unsigned long flags;
struct page *page;
int cold = !!(gfp_flags & __GFP_COLD);
+ int cpu;
again:
+ cpu = get_cpu();
if (order == 0) {
struct per_cpu_pages *pcp;
- page = NULL;
- pcp = &zone_pcp(zone, get_cpu())->pcp[cold];
+ pcp = &zone_pcp(zone, cpu)->pcp[cold];
local_irq_save(flags);
- if (!pcp->count)
+ if (!pcp->count) {
pcp->count += rmqueue_bulk(zone, 0,
pcp->batch, &pcp->list);
- if (likely(pcp->count)) {
- page = list_entry(pcp->list.next, struct page, lru);
- list_del(&page->lru);
- pcp->count--;
+ if (unlikely(!pcp->count))
+ goto failed;
}
- local_irq_restore(flags);
- put_cpu();
+ page = list_entry(pcp->list.next, struct page, lru);
+ list_del(&page->lru);
+ pcp->count--;
} else {
spin_lock_irqsave(&zone->lock, flags);
page = __rmqueue(zone, order);
- spin_unlock_irqrestore(&zone->lock, flags);
+ spin_unlock(&zone->lock);
+ if (!page)
+ goto failed;
}
- if (page != NULL) {
- BUG_ON(bad_range(zone, page));
- mod_page_state_zone(zone, pgalloc, 1 << order);
- if (prep_new_page(page, order))
- goto again;
+ __mod_page_state_zone(zone, pgalloc, 1 << order);
+ zone_statistics(zonelist, zone, cpu);
+ local_irq_restore(flags);
+ put_cpu();
- if (gfp_flags & __GFP_ZERO)
- prep_zero_page(page, order, gfp_flags);
+ BUG_ON(bad_range(zone, page));
+ if (prep_new_page(page, order))
+ goto again;
- if (order && (gfp_flags & __GFP_COMP))
- prep_compound_page(page, order);
- }
+ if (gfp_flags & __GFP_ZERO)
+ prep_zero_page(page, order, gfp_flags);
+
+ if (order && (gfp_flags & __GFP_COMP))
+ prep_compound_page(page, order);
return page;
+
+failed:
+ local_irq_restore(flags);
+ put_cpu();
+ return NULL;
}
#define ALLOC_NO_WATERMARKS 0x01 /* don't check watermarks at all */
@@ -871,9 +875,8 @@
continue;
}
- page = buffered_rmqueue(*z, order, gfp_mask);
+ page = buffered_rmqueue(zonelist, *z, order, gfp_mask);
if (page) {
- zone_statistics(zonelist, *z);
break;
}
} while (*(++z) != NULL);
@@ -1248,7 +1251,7 @@
__get_page_state(ret, sizeof(*ret) / sizeof(unsigned long), &mask);
}
-unsigned long __read_page_state(unsigned long offset)
+unsigned long read_page_state_offset(unsigned long offset)
{
unsigned long ret = 0;
int cpu;
@@ -1262,18 +1265,26 @@
return ret;
}
-void __mod_page_state(unsigned long offset, unsigned long delta)
+void __mod_page_state_offset(unsigned long offset, unsigned long delta)
+{
+ void *ptr;
+
+ ptr = &__get_cpu_var(page_states);
+ *(unsigned long *)(ptr + offset) += delta;
+}
+EXPORT_SYMBOL(__mod_page_state_offset);
+
+void mod_page_state_offset(unsigned long offset, unsigned long delta)
{
unsigned long flags;
- void* ptr;
+ void *ptr;
local_irq_save(flags);
ptr = &__get_cpu_var(page_states);
- *(unsigned long*)(ptr + offset) += delta;
+ *(unsigned long *)(ptr + offset) += delta;
local_irq_restore(flags);
}
-
-EXPORT_SYMBOL(__mod_page_state);
+EXPORT_SYMBOL(mod_page_state_offset);
void __get_zone_counts(unsigned long *active, unsigned long *inactive,
unsigned long *free, struct pglist_data *pgdat)
diff --git a/mm/rmap.c b/mm/rmap.c
index 4107f64..6f3f7db 100644
--- a/mm/rmap.c
+++ b/mm/rmap.c
@@ -451,7 +451,11 @@
page->index = linear_page_index(vma, address);
- inc_page_state(nr_mapped);
+ /*
+ * nr_mapped state can be updated without turning off
+ * interrupts because it is not modified via interrupt.
+ */
+ __inc_page_state(nr_mapped);
}
/**
@@ -498,7 +502,7 @@
BUG_ON(!pfn_valid(page_to_pfn(page)));
if (atomic_inc_and_test(&page->_mapcount))
- inc_page_state(nr_mapped);
+ __inc_page_state(nr_mapped);
}
/**
@@ -522,7 +526,7 @@
*/
if (page_test_and_clear_dirty(page))
set_page_dirty(page);
- dec_page_state(nr_mapped);
+ __dec_page_state(nr_mapped);
}
}
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 7681d8e..be8235f 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -645,16 +645,17 @@
goto done;
max_scan -= nr_scan;
- if (current_is_kswapd())
- mod_page_state_zone(zone, pgscan_kswapd, nr_scan);
- else
- mod_page_state_zone(zone, pgscan_direct, nr_scan);
nr_freed = shrink_list(&page_list, sc);
- if (current_is_kswapd())
- mod_page_state(kswapd_steal, nr_freed);
- mod_page_state_zone(zone, pgsteal, nr_freed);
- spin_lock_irq(&zone->lru_lock);
+ local_irq_disable();
+ if (current_is_kswapd()) {
+ __mod_page_state_zone(zone, pgscan_kswapd, nr_scan);
+ __mod_page_state(kswapd_steal, nr_freed);
+ } else
+ __mod_page_state_zone(zone, pgscan_direct, nr_scan);
+ __mod_page_state_zone(zone, pgsteal, nr_freed);
+
+ spin_lock(&zone->lru_lock);
/*
* Put back any unfreeable pages.
*/
@@ -816,11 +817,13 @@
}
}
zone->nr_active += pgmoved;
- spin_unlock_irq(&zone->lru_lock);
- pagevec_release(&pvec);
+ spin_unlock(&zone->lru_lock);
- mod_page_state_zone(zone, pgrefill, pgscanned);
- mod_page_state(pgdeactivate, pgdeactivate);
+ __mod_page_state_zone(zone, pgrefill, pgscanned);
+ __mod_page_state(pgdeactivate, pgdeactivate);
+ local_irq_enable();
+
+ pagevec_release(&pvec);
}
/*