free_pages_bulk(struct zone *zone, int count,
struct list_head *list, unsigned int order)
{
- unsigned long flags;
struct page *page = NULL;
int ret = 0;
- spin_lock_irqsave(&zone->lock, flags);
+ spin_lock(&zone->lock);
zone->all_unreclaimable = 0;
zone->pages_scanned = 0;
while (!list_empty(list) && count--) {
__free_pages_bulk(page, zone, order);
ret++;
}
- spin_unlock_irqrestore(&zone->lock, flags);
+ spin_unlock(&zone->lock);
return ret;
}
void __free_pages_ok(struct page *page, unsigned int order)
{
+ unsigned long flags;
LIST_HEAD(list);
int i;
int reserved = 0;
list_add(&page->lru, &list);
mod_page_state(pgfree, 1 << order);
kernel_map_pages(page, 1<<order, 0);
+ local_irq_save(flags);
free_pages_bulk(page_zone(page), 1, &list, order);
+ local_irq_restore(flags);
}
static int rmqueue_bulk(struct zone *zone, unsigned int order,
unsigned long count, struct list_head *list)
{
- unsigned long flags;
int i;
int allocated = 0;
struct page *page;
- spin_lock_irqsave(&zone->lock, flags);
+ spin_lock(&zone->lock);
for (i = 0; i < count; ++i) {
page = __rmqueue(zone, order);
if (page == NULL)
allocated++;
list_add_tail(&page->lru, list);
}
- spin_unlock_irqrestore(&zone->lock, flags);
+ spin_unlock(&zone->lock);
return allocated;
}
#if defined(CONFIG_PM) || defined(CONFIG_HOTPLUG_CPU)
static void __drain_pages(unsigned int cpu)
{
+ unsigned long flags;
struct zone *zone;
int i;
struct per_cpu_pages *pcp;
pcp = &pset->pcp[i];
+ local_irq_save(flags);
pcp->count -= free_pages_bulk(zone, pcp->count,
&pcp->list, 0);
+ local_irq_restore(flags);
}
}
}
if (pcp->count <= pcp->low)
pcp->count += rmqueue_bulk(zone, 0,
pcp->batch, &pcp->list);
- if (pcp->count) {
+ if (likely(pcp->count)) {
page = list_entry(pcp->list.next, struct page, lru);
list_del(&page->lru);
pcp->count--;