2008-10-19 10:28:16 +07:00
|
|
|
#include <linux/mm.h>
|
|
|
|
#include <linux/mmzone.h>
|
|
|
|
#include <linux/bootmem.h>
|
|
|
|
#include <linux/bit_spinlock.h>
|
|
|
|
#include <linux/page_cgroup.h>
|
|
|
|
#include <linux/hash.h>
|
2008-10-23 04:15:05 +07:00
|
|
|
#include <linux/slab.h>
|
2008-10-19 10:28:16 +07:00
|
|
|
#include <linux/memory.h>
|
2008-10-23 04:14:58 +07:00
|
|
|
#include <linux/vmalloc.h>
|
2008-10-23 04:15:05 +07:00
|
|
|
#include <linux/cgroup.h>
|
2009-01-08 09:07:58 +07:00
|
|
|
#include <linux/swapops.h>
|
2010-07-19 17:54:14 +07:00
|
|
|
#include <linux/kmemleak.h>
|
2008-10-19 10:28:16 +07:00
|
|
|
|
|
|
|
static unsigned long total_usage;
|
|
|
|
|
|
|
|
#if !defined(CONFIG_SPARSEMEM)
|
|
|
|
|
|
|
|
|
2008-11-23 00:33:24 +07:00
|
|
|
void __meminit pgdat_page_cgroup_init(struct pglist_data *pgdat)
|
2008-10-19 10:28:16 +07:00
|
|
|
{
|
|
|
|
pgdat->node_page_cgroup = NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
struct page_cgroup *lookup_page_cgroup(struct page *page)
|
|
|
|
{
|
|
|
|
unsigned long pfn = page_to_pfn(page);
|
|
|
|
unsigned long offset;
|
|
|
|
struct page_cgroup *base;
|
|
|
|
|
|
|
|
base = NODE_DATA(page_to_nid(page))->node_page_cgroup;
|
2012-01-13 08:18:40 +07:00
|
|
|
#ifdef CONFIG_DEBUG_VM
|
|
|
|
/*
|
|
|
|
* The sanity checks the page allocator does upon freeing a
|
|
|
|
* page can reach here before the page_cgroup arrays are
|
|
|
|
* allocated when feeding a range of pages to the allocator
|
|
|
|
* for the first time during bootup or memory hotplug.
|
|
|
|
*/
|
2008-10-19 10:28:16 +07:00
|
|
|
if (unlikely(!base))
|
|
|
|
return NULL;
|
2012-01-13 08:18:40 +07:00
|
|
|
#endif
|
2008-10-19 10:28:16 +07:00
|
|
|
offset = pfn - NODE_DATA(page_to_nid(page))->node_start_pfn;
|
|
|
|
return base + offset;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int __init alloc_node_page_cgroup(int nid)
|
|
|
|
{
|
2012-01-13 08:18:18 +07:00
|
|
|
struct page_cgroup *base;
|
2008-10-19 10:28:16 +07:00
|
|
|
unsigned long table_size;
|
2012-01-13 08:18:18 +07:00
|
|
|
unsigned long nr_pages;
|
2008-10-19 10:28:16 +07:00
|
|
|
|
|
|
|
nr_pages = NODE_DATA(nid)->node_spanned_pages;
|
2008-12-10 04:14:20 +07:00
|
|
|
if (!nr_pages)
|
|
|
|
return 0;
|
|
|
|
|
2008-10-19 10:28:16 +07:00
|
|
|
table_size = sizeof(struct page_cgroup) * nr_pages;
|
2009-06-12 14:33:53 +07:00
|
|
|
|
|
|
|
base = __alloc_bootmem_node_nopanic(NODE_DATA(nid),
|
|
|
|
table_size, PAGE_SIZE, __pa(MAX_DMA_ADDRESS));
|
|
|
|
if (!base)
|
2008-10-19 10:28:16 +07:00
|
|
|
return -ENOMEM;
|
|
|
|
NODE_DATA(nid)->node_page_cgroup = base;
|
|
|
|
total_usage += table_size;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2009-06-12 14:33:53 +07:00
|
|
|
void __init page_cgroup_init_flatmem(void)
|
2008-10-19 10:28:16 +07:00
|
|
|
{
|
|
|
|
|
|
|
|
int nid, fail;
|
|
|
|
|
2009-01-08 09:08:02 +07:00
|
|
|
if (mem_cgroup_disabled())
|
2008-10-23 04:15:05 +07:00
|
|
|
return;
|
|
|
|
|
2008-10-19 10:28:16 +07:00
|
|
|
for_each_online_node(nid) {
|
|
|
|
fail = alloc_node_page_cgroup(nid);
|
|
|
|
if (fail)
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
printk(KERN_INFO "allocated %ld bytes of page_cgroup\n", total_usage);
|
2009-06-18 06:26:32 +07:00
|
|
|
printk(KERN_INFO "please try 'cgroup_disable=memory' option if you"
|
|
|
|
" don't want memory cgroups\n");
|
2008-10-19 10:28:16 +07:00
|
|
|
return;
|
|
|
|
fail:
|
2009-06-18 06:26:32 +07:00
|
|
|
printk(KERN_CRIT "allocation of page_cgroup failed.\n");
|
|
|
|
printk(KERN_CRIT "please try 'cgroup_disable=memory' boot option\n");
|
2008-10-19 10:28:16 +07:00
|
|
|
panic("Out of memory");
|
|
|
|
}
|
|
|
|
|
|
|
|
#else /* CONFIG_FLAT_NODE_MEM_MAP */
|
|
|
|
|
|
|
|
struct page_cgroup *lookup_page_cgroup(struct page *page)
|
|
|
|
{
|
|
|
|
unsigned long pfn = page_to_pfn(page);
|
|
|
|
struct mem_section *section = __pfn_to_section(pfn);
|
2012-01-13 08:18:40 +07:00
|
|
|
#ifdef CONFIG_DEBUG_VM
|
|
|
|
/*
|
|
|
|
* The sanity checks the page allocator does upon freeing a
|
|
|
|
* page can reach here before the page_cgroup arrays are
|
|
|
|
* allocated when feeding a range of pages to the allocator
|
|
|
|
* for the first time during bootup or memory hotplug.
|
|
|
|
*/
|
2009-06-18 06:26:34 +07:00
|
|
|
if (!section->page_cgroup)
|
|
|
|
return NULL;
|
2012-01-13 08:18:40 +07:00
|
|
|
#endif
|
2008-10-19 10:28:16 +07:00
|
|
|
return section->page_cgroup + pfn;
|
|
|
|
}
|
|
|
|
|
2011-05-27 06:25:29 +07:00
|
|
|
static void *__meminit alloc_page_cgroup(size_t size, int nid)
|
2011-03-24 06:42:40 +07:00
|
|
|
{
|
2012-01-13 08:18:18 +07:00
|
|
|
gfp_t flags = GFP_KERNEL | __GFP_ZERO | __GFP_NOWARN;
|
2011-03-24 06:42:40 +07:00
|
|
|
void *addr = NULL;
|
|
|
|
|
2011-11-03 03:38:11 +07:00
|
|
|
addr = alloc_pages_exact_nid(nid, size, flags);
|
|
|
|
if (addr) {
|
|
|
|
kmemleak_alloc(addr, size, 1, flags);
|
2011-03-24 06:42:40 +07:00
|
|
|
return addr;
|
2011-11-03 03:38:11 +07:00
|
|
|
}
|
2011-03-24 06:42:40 +07:00
|
|
|
|
|
|
|
if (node_state(nid, N_HIGH_MEMORY))
|
2012-01-13 08:18:18 +07:00
|
|
|
addr = vzalloc_node(size, nid);
|
2011-03-24 06:42:40 +07:00
|
|
|
else
|
2012-01-13 08:18:18 +07:00
|
|
|
addr = vzalloc(size);
|
2011-03-24 06:42:40 +07:00
|
|
|
|
|
|
|
return addr;
|
|
|
|
}
|
|
|
|
|
2011-06-16 05:08:42 +07:00
|
|
|
static int __meminit init_section_page_cgroup(unsigned long pfn, int nid)
|
2008-10-19 10:28:16 +07:00
|
|
|
{
|
2011-03-24 06:42:30 +07:00
|
|
|
struct mem_section *section;
|
2012-01-13 08:18:18 +07:00
|
|
|
struct page_cgroup *base;
|
2008-10-19 10:28:16 +07:00
|
|
|
unsigned long table_size;
|
|
|
|
|
2012-01-13 08:18:18 +07:00
|
|
|
section = __pfn_to_section(pfn);
|
2011-03-24 06:42:30 +07:00
|
|
|
|
|
|
|
if (section->page_cgroup)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
table_size = sizeof(struct page_cgroup) * PAGES_PER_SECTION;
|
2011-03-24 06:42:40 +07:00
|
|
|
base = alloc_page_cgroup(table_size, nid);
|
|
|
|
|
2011-03-24 06:42:30 +07:00
|
|
|
/*
|
|
|
|
* The value stored in section->page_cgroup is (base - pfn)
|
|
|
|
* and it does not point to the memory block allocated above,
|
|
|
|
* causing kmemleak false positives.
|
|
|
|
*/
|
|
|
|
kmemleak_not_leak(base);
|
2008-10-19 10:28:16 +07:00
|
|
|
|
|
|
|
if (!base) {
|
|
|
|
printk(KERN_ERR "page cgroup allocation failure\n");
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
|
2011-06-16 05:08:42 +07:00
|
|
|
/*
|
|
|
|
* The passed "pfn" may not be aligned to SECTION. For the calculation
|
|
|
|
* we need to apply a mask.
|
|
|
|
*/
|
|
|
|
pfn &= PAGE_SECTION_MASK;
|
2008-10-19 10:28:16 +07:00
|
|
|
section->page_cgroup = base - pfn;
|
|
|
|
total_usage += table_size;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
#ifdef CONFIG_MEMORY_HOTPLUG
|
2012-01-13 08:19:08 +07:00
|
|
|
static void free_page_cgroup(void *addr)
|
|
|
|
{
|
|
|
|
if (is_vmalloc_addr(addr)) {
|
|
|
|
vfree(addr);
|
|
|
|
} else {
|
|
|
|
struct page *page = virt_to_page(addr);
|
|
|
|
size_t table_size =
|
|
|
|
sizeof(struct page_cgroup) * PAGES_PER_SECTION;
|
|
|
|
|
|
|
|
BUG_ON(PageReserved(page));
|
|
|
|
free_pages_exact(addr, table_size);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2008-10-19 10:28:16 +07:00
|
|
|
void __free_page_cgroup(unsigned long pfn)
|
|
|
|
{
|
|
|
|
struct mem_section *ms;
|
|
|
|
struct page_cgroup *base;
|
|
|
|
|
|
|
|
ms = __pfn_to_section(pfn);
|
|
|
|
if (!ms || !ms->page_cgroup)
|
|
|
|
return;
|
|
|
|
base = ms->page_cgroup + pfn;
|
2011-03-24 06:42:40 +07:00
|
|
|
free_page_cgroup(base);
|
|
|
|
ms->page_cgroup = NULL;
|
2008-10-19 10:28:16 +07:00
|
|
|
}
|
|
|
|
|
2008-11-23 00:33:24 +07:00
|
|
|
int __meminit online_page_cgroup(unsigned long start_pfn,
|
2008-10-19 10:28:16 +07:00
|
|
|
unsigned long nr_pages,
|
|
|
|
int nid)
|
|
|
|
{
|
|
|
|
unsigned long start, end, pfn;
|
|
|
|
int fail = 0;
|
|
|
|
|
2011-07-26 07:12:13 +07:00
|
|
|
start = SECTION_ALIGN_DOWN(start_pfn);
|
|
|
|
end = SECTION_ALIGN_UP(start_pfn + nr_pages);
|
2008-10-19 10:28:16 +07:00
|
|
|
|
2011-06-16 05:08:42 +07:00
|
|
|
if (nid == -1) {
|
|
|
|
/*
|
|
|
|
* In this case, "nid" already exists and contains valid memory.
|
|
|
|
* "start_pfn" passed to us is a pfn which is an arg for
|
|
|
|
* online__pages(), and start_pfn should exist.
|
|
|
|
*/
|
|
|
|
nid = pfn_to_nid(start_pfn);
|
|
|
|
VM_BUG_ON(!node_state(nid, N_ONLINE));
|
|
|
|
}
|
|
|
|
|
2008-10-19 10:28:16 +07:00
|
|
|
for (pfn = start; !fail && pfn < end; pfn += PAGES_PER_SECTION) {
|
|
|
|
if (!pfn_present(pfn))
|
|
|
|
continue;
|
2011-06-16 05:08:42 +07:00
|
|
|
fail = init_section_page_cgroup(pfn, nid);
|
2008-10-19 10:28:16 +07:00
|
|
|
}
|
|
|
|
if (!fail)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
/* rollback */
|
|
|
|
for (pfn = start; pfn < end; pfn += PAGES_PER_SECTION)
|
|
|
|
__free_page_cgroup(pfn);
|
|
|
|
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
|
2008-11-23 00:33:24 +07:00
|
|
|
int __meminit offline_page_cgroup(unsigned long start_pfn,
|
2008-10-19 10:28:16 +07:00
|
|
|
unsigned long nr_pages, int nid)
|
|
|
|
{
|
|
|
|
unsigned long start, end, pfn;
|
|
|
|
|
2011-07-26 07:12:13 +07:00
|
|
|
start = SECTION_ALIGN_DOWN(start_pfn);
|
|
|
|
end = SECTION_ALIGN_UP(start_pfn + nr_pages);
|
2008-10-19 10:28:16 +07:00
|
|
|
|
|
|
|
for (pfn = start; pfn < end; pfn += PAGES_PER_SECTION)
|
|
|
|
__free_page_cgroup(pfn);
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
}
|
|
|
|
|
2008-11-23 00:33:24 +07:00
|
|
|
static int __meminit page_cgroup_callback(struct notifier_block *self,
|
2008-10-19 10:28:16 +07:00
|
|
|
unsigned long action, void *arg)
|
|
|
|
{
|
|
|
|
struct memory_notify *mn = arg;
|
|
|
|
int ret = 0;
|
|
|
|
switch (action) {
|
|
|
|
case MEM_GOING_ONLINE:
|
|
|
|
ret = online_page_cgroup(mn->start_pfn,
|
|
|
|
mn->nr_pages, mn->status_change_nid);
|
|
|
|
break;
|
|
|
|
case MEM_OFFLINE:
|
|
|
|
offline_page_cgroup(mn->start_pfn,
|
|
|
|
mn->nr_pages, mn->status_change_nid);
|
|
|
|
break;
|
2008-12-02 04:13:48 +07:00
|
|
|
case MEM_CANCEL_ONLINE:
|
2008-10-19 10:28:16 +07:00
|
|
|
case MEM_GOING_OFFLINE:
|
|
|
|
break;
|
|
|
|
case MEM_ONLINE:
|
|
|
|
case MEM_CANCEL_OFFLINE:
|
|
|
|
break;
|
|
|
|
}
|
2008-12-02 04:13:48 +07:00
|
|
|
|
2011-03-23 06:30:49 +07:00
|
|
|
return notifier_from_errno(ret);
|
2008-10-19 10:28:16 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
void __init page_cgroup_init(void)
|
|
|
|
{
|
|
|
|
unsigned long pfn;
|
2011-06-16 05:08:42 +07:00
|
|
|
int nid;
|
2008-10-19 10:28:16 +07:00
|
|
|
|
2009-01-08 09:08:02 +07:00
|
|
|
if (mem_cgroup_disabled())
|
2008-10-23 04:15:05 +07:00
|
|
|
return;
|
|
|
|
|
2011-06-16 05:08:42 +07:00
|
|
|
for_each_node_state(nid, N_HIGH_MEMORY) {
|
|
|
|
unsigned long start_pfn, end_pfn;
|
|
|
|
|
|
|
|
start_pfn = node_start_pfn(nid);
|
|
|
|
end_pfn = node_end_pfn(nid);
|
|
|
|
/*
|
|
|
|
* start_pfn and end_pfn may not be aligned to SECTION and the
|
|
|
|
* page->flags of out of node pages are not initialized. So we
|
|
|
|
* scan [start_pfn, the biggest section's pfn < end_pfn) here.
|
|
|
|
*/
|
|
|
|
for (pfn = start_pfn;
|
|
|
|
pfn < end_pfn;
|
|
|
|
pfn = ALIGN(pfn + 1, PAGES_PER_SECTION)) {
|
|
|
|
|
|
|
|
if (!pfn_valid(pfn))
|
|
|
|
continue;
|
|
|
|
/*
|
|
|
|
* Nodes's pfns can be overlapping.
|
|
|
|
* We know some arch can have a nodes layout such as
|
|
|
|
* -------------pfn-------------->
|
|
|
|
* N0 | N1 | N2 | N0 | N1 | N2|....
|
|
|
|
*/
|
|
|
|
if (pfn_to_nid(pfn) != nid)
|
|
|
|
continue;
|
|
|
|
if (init_section_page_cgroup(pfn, nid))
|
|
|
|
goto oom;
|
|
|
|
}
|
2008-10-19 10:28:16 +07:00
|
|
|
}
|
2011-06-16 05:08:42 +07:00
|
|
|
hotplug_memory_notifier(page_cgroup_callback, 0);
|
2008-10-19 10:28:16 +07:00
|
|
|
printk(KERN_INFO "allocated %ld bytes of page_cgroup\n", total_usage);
|
2011-06-16 05:08:42 +07:00
|
|
|
printk(KERN_INFO "please try 'cgroup_disable=memory' option if you "
|
|
|
|
"don't want memory cgroups\n");
|
|
|
|
return;
|
|
|
|
oom:
|
|
|
|
printk(KERN_CRIT "try 'cgroup_disable=memory' boot option\n");
|
|
|
|
panic("Out of memory");
|
2008-10-19 10:28:16 +07:00
|
|
|
}
|
|
|
|
|
2008-11-23 00:33:24 +07:00
|
|
|
void __meminit pgdat_page_cgroup_init(struct pglist_data *pgdat)
|
2008-10-19 10:28:16 +07:00
|
|
|
{
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
#endif
|
2009-01-08 09:07:58 +07:00
|
|
|
|
|
|
|
|
|
|
|
#ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
|
|
|
|
|
|
|
|
static DEFINE_MUTEX(swap_cgroup_mutex);
|
|
|
|
struct swap_cgroup_ctrl {
|
|
|
|
struct page **map;
|
|
|
|
unsigned long length;
|
2010-03-15 11:34:57 +07:00
|
|
|
spinlock_t lock;
|
2009-01-08 09:07:58 +07:00
|
|
|
};
|
|
|
|
|
2011-11-03 03:38:36 +07:00
|
|
|
static struct swap_cgroup_ctrl swap_cgroup_ctrl[MAX_SWAPFILES];
|
2009-01-08 09:07:58 +07:00
|
|
|
|
|
|
|
struct swap_cgroup {
|
2009-04-03 06:57:45 +07:00
|
|
|
unsigned short id;
|
2009-01-08 09:07:58 +07:00
|
|
|
};
|
|
|
|
#define SC_PER_PAGE (PAGE_SIZE/sizeof(struct swap_cgroup))
|
|
|
|
|
|
|
|
/*
|
|
|
|
* SwapCgroup implements "lookup" and "exchange" operations.
|
|
|
|
* In typical usage, this swap_cgroup is accessed via memcg's charge/uncharge
|
|
|
|
* against SwapCache. At swap_free(), this is accessed directly from swap.
|
|
|
|
*
|
|
|
|
* This means,
|
|
|
|
* - we have no race in "exchange" when we're accessed via SwapCache because
|
|
|
|
* SwapCache(and its swp_entry) is under lock.
|
|
|
|
* - When called via swap_free(), there is no user of this entry and no race.
|
|
|
|
* Then, we don't need lock around "exchange".
|
|
|
|
*
|
|
|
|
* TODO: we can push these buffers out to HIGHMEM.
|
|
|
|
*/
|
|
|
|
|
|
|
|
/*
|
|
|
|
* allocate buffer for swap_cgroup.
|
|
|
|
*/
|
|
|
|
static int swap_cgroup_prepare(int type)
|
|
|
|
{
|
|
|
|
struct page *page;
|
|
|
|
struct swap_cgroup_ctrl *ctrl;
|
|
|
|
unsigned long idx, max;
|
|
|
|
|
|
|
|
ctrl = &swap_cgroup_ctrl[type];
|
|
|
|
|
|
|
|
for (idx = 0; idx < ctrl->length; idx++) {
|
|
|
|
page = alloc_page(GFP_KERNEL | __GFP_ZERO);
|
|
|
|
if (!page)
|
|
|
|
goto not_enough_page;
|
|
|
|
ctrl->map[idx] = page;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
not_enough_page:
|
|
|
|
max = idx;
|
|
|
|
for (idx = 0; idx < max; idx++)
|
|
|
|
__free_page(ctrl->map[idx]);
|
|
|
|
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
|
2012-01-13 08:18:48 +07:00
|
|
|
static struct swap_cgroup *lookup_swap_cgroup(swp_entry_t ent,
|
|
|
|
struct swap_cgroup_ctrl **ctrlp)
|
|
|
|
{
|
|
|
|
pgoff_t offset = swp_offset(ent);
|
|
|
|
struct swap_cgroup_ctrl *ctrl;
|
|
|
|
struct page *mappage;
|
2012-03-06 11:52:55 +07:00
|
|
|
struct swap_cgroup *sc;
|
2012-01-13 08:18:48 +07:00
|
|
|
|
|
|
|
ctrl = &swap_cgroup_ctrl[swp_type(ent)];
|
|
|
|
if (ctrlp)
|
|
|
|
*ctrlp = ctrl;
|
|
|
|
|
|
|
|
mappage = ctrl->map[offset / SC_PER_PAGE];
|
2012-03-06 11:52:55 +07:00
|
|
|
sc = page_address(mappage);
|
|
|
|
return sc + offset % SC_PER_PAGE;
|
2012-01-13 08:18:48 +07:00
|
|
|
}
|
|
|
|
|
2010-03-11 06:22:17 +07:00
|
|
|
/**
|
|
|
|
* swap_cgroup_cmpxchg - cmpxchg mem_cgroup's id for this swp_entry.
|
2012-06-21 02:53:01 +07:00
|
|
|
* @ent: swap entry to be cmpxchged
|
2010-03-11 06:22:17 +07:00
|
|
|
* @old: old id
|
|
|
|
* @new: new id
|
|
|
|
*
|
|
|
|
* Returns old id at success, 0 at failure.
|
2011-03-31 08:57:33 +07:00
|
|
|
* (There is no mem_cgroup using 0 as its id)
|
2010-03-11 06:22:17 +07:00
|
|
|
*/
|
|
|
|
unsigned short swap_cgroup_cmpxchg(swp_entry_t ent,
|
|
|
|
unsigned short old, unsigned short new)
|
|
|
|
{
|
|
|
|
struct swap_cgroup_ctrl *ctrl;
|
|
|
|
struct swap_cgroup *sc;
|
2010-03-15 11:34:57 +07:00
|
|
|
unsigned long flags;
|
|
|
|
unsigned short retval;
|
2010-03-11 06:22:17 +07:00
|
|
|
|
2012-01-13 08:18:48 +07:00
|
|
|
sc = lookup_swap_cgroup(ent, &ctrl);
|
2010-03-11 06:22:17 +07:00
|
|
|
|
2010-03-15 11:34:57 +07:00
|
|
|
spin_lock_irqsave(&ctrl->lock, flags);
|
|
|
|
retval = sc->id;
|
|
|
|
if (retval == old)
|
|
|
|
sc->id = new;
|
2010-03-11 06:22:17 +07:00
|
|
|
else
|
2010-03-15 11:34:57 +07:00
|
|
|
retval = 0;
|
|
|
|
spin_unlock_irqrestore(&ctrl->lock, flags);
|
|
|
|
return retval;
|
2010-03-11 06:22:17 +07:00
|
|
|
}
|
|
|
|
|
2009-01-08 09:07:58 +07:00
|
|
|
/**
|
|
|
|
* swap_cgroup_record - record mem_cgroup for this swp_entry.
|
|
|
|
* @ent: swap entry to be recorded into
|
2012-06-21 02:53:01 +07:00
|
|
|
* @id: mem_cgroup to be recorded
|
2009-01-08 09:07:58 +07:00
|
|
|
*
|
2009-04-03 06:57:45 +07:00
|
|
|
* Returns old value at success, 0 at failure.
|
|
|
|
* (Of course, old value can be 0.)
|
2009-01-08 09:07:58 +07:00
|
|
|
*/
|
2009-04-03 06:57:45 +07:00
|
|
|
unsigned short swap_cgroup_record(swp_entry_t ent, unsigned short id)
|
2009-01-08 09:07:58 +07:00
|
|
|
{
|
|
|
|
struct swap_cgroup_ctrl *ctrl;
|
|
|
|
struct swap_cgroup *sc;
|
2009-04-03 06:57:45 +07:00
|
|
|
unsigned short old;
|
2010-03-15 11:34:57 +07:00
|
|
|
unsigned long flags;
|
2009-01-08 09:07:58 +07:00
|
|
|
|
2012-01-13 08:18:48 +07:00
|
|
|
sc = lookup_swap_cgroup(ent, &ctrl);
|
2009-01-08 09:07:58 +07:00
|
|
|
|
2010-03-15 11:34:57 +07:00
|
|
|
spin_lock_irqsave(&ctrl->lock, flags);
|
|
|
|
old = sc->id;
|
|
|
|
sc->id = id;
|
|
|
|
spin_unlock_irqrestore(&ctrl->lock, flags);
|
2009-01-08 09:07:58 +07:00
|
|
|
|
|
|
|
return old;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2012-01-13 08:18:48 +07:00
|
|
|
* lookup_swap_cgroup_id - lookup mem_cgroup id tied to swap entry
|
2009-01-08 09:07:58 +07:00
|
|
|
* @ent: swap entry to be looked up.
|
|
|
|
*
|
2009-04-03 06:57:45 +07:00
|
|
|
* Returns CSS ID of mem_cgroup at success. 0 at failure. (0 is invalid ID)
|
2009-01-08 09:07:58 +07:00
|
|
|
*/
|
2012-01-13 08:18:48 +07:00
|
|
|
unsigned short lookup_swap_cgroup_id(swp_entry_t ent)
|
2009-01-08 09:07:58 +07:00
|
|
|
{
|
2012-01-13 08:18:48 +07:00
|
|
|
return lookup_swap_cgroup(ent, NULL)->id;
|
2009-01-08 09:07:58 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
int swap_cgroup_swapon(int type, unsigned long max_pages)
|
|
|
|
{
|
|
|
|
void *array;
|
|
|
|
unsigned long array_size;
|
|
|
|
unsigned long length;
|
|
|
|
struct swap_cgroup_ctrl *ctrl;
|
|
|
|
|
|
|
|
if (!do_swap_account)
|
|
|
|
return 0;
|
|
|
|
|
2011-05-27 06:25:30 +07:00
|
|
|
length = DIV_ROUND_UP(max_pages, SC_PER_PAGE);
|
2009-01-08 09:07:58 +07:00
|
|
|
array_size = length * sizeof(void *);
|
|
|
|
|
2011-05-29 00:36:34 +07:00
|
|
|
array = vzalloc(array_size);
|
2009-01-08 09:07:58 +07:00
|
|
|
if (!array)
|
|
|
|
goto nomem;
|
|
|
|
|
|
|
|
ctrl = &swap_cgroup_ctrl[type];
|
|
|
|
mutex_lock(&swap_cgroup_mutex);
|
|
|
|
ctrl->length = length;
|
|
|
|
ctrl->map = array;
|
2010-03-15 11:34:57 +07:00
|
|
|
spin_lock_init(&ctrl->lock);
|
2009-01-08 09:07:58 +07:00
|
|
|
if (swap_cgroup_prepare(type)) {
|
|
|
|
/* memory shortage */
|
|
|
|
ctrl->map = NULL;
|
|
|
|
ctrl->length = 0;
|
|
|
|
mutex_unlock(&swap_cgroup_mutex);
|
2011-05-27 06:25:31 +07:00
|
|
|
vfree(array);
|
2009-01-08 09:07:58 +07:00
|
|
|
goto nomem;
|
|
|
|
}
|
|
|
|
mutex_unlock(&swap_cgroup_mutex);
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
nomem:
|
|
|
|
printk(KERN_INFO "couldn't allocate enough memory for swap_cgroup.\n");
|
|
|
|
printk(KERN_INFO
|
2011-07-26 07:12:12 +07:00
|
|
|
"swap_cgroup can be disabled by swapaccount=0 boot option\n");
|
2009-01-08 09:07:58 +07:00
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
|
|
|
|
void swap_cgroup_swapoff(int type)
|
|
|
|
{
|
2011-05-27 06:25:31 +07:00
|
|
|
struct page **map;
|
|
|
|
unsigned long i, length;
|
2009-01-08 09:07:58 +07:00
|
|
|
struct swap_cgroup_ctrl *ctrl;
|
|
|
|
|
|
|
|
if (!do_swap_account)
|
|
|
|
return;
|
|
|
|
|
|
|
|
mutex_lock(&swap_cgroup_mutex);
|
|
|
|
ctrl = &swap_cgroup_ctrl[type];
|
2011-05-27 06:25:31 +07:00
|
|
|
map = ctrl->map;
|
|
|
|
length = ctrl->length;
|
|
|
|
ctrl->map = NULL;
|
|
|
|
ctrl->length = 0;
|
|
|
|
mutex_unlock(&swap_cgroup_mutex);
|
|
|
|
|
|
|
|
if (map) {
|
|
|
|
for (i = 0; i < length; i++) {
|
|
|
|
struct page *page = map[i];
|
2009-01-08 09:07:58 +07:00
|
|
|
if (page)
|
|
|
|
__free_page(page);
|
|
|
|
}
|
2011-05-27 06:25:31 +07:00
|
|
|
vfree(map);
|
2009-01-08 09:07:58 +07:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
#endif
|