lk/linux-kernel-test
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

memcg: simple migration handling

Browse Source 
Now, management of "charge" under page migration is done under following
manner. (Assume migrate page contents from oldpage to newpage)

 before
  - "newpage" is charged before migration.
 at success.
  - "oldpage" is uncharged at somewhere(unmap, radix-tree-replace)
 at failure
  - "newpage" is uncharged.
  - "oldpage" is charged if necessary (*1)

But (*1) is not reliable....because of GFP_ATOMIC.

This patch tries to change behavior as following by charge/commit/cancel ops.

 before
  - charge PAGE_SIZE (no target page)
 success
  - commit charge against "newpage".
 failure
  - commit charge against "oldpage".
    (PCG_USED bit works effectively to avoid double-counting)
  - if "oldpage" is obsolete, cancel charge of PAGE_SIZE.

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Reviewed-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
Cc: Balbir Singh <balbir@in.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This commit is contained in:
KAMEZAWA Hiroyuki
2009-01-07 18:07:50 -08:00
committed by Linus Torvalds
parent bced0520fe
commit 01b1ae63c2
3 changed files with 73 additions and 96 deletions
Download Patch File Download Diff File Expand all files Collapse all files

42
mm/migrate.c
View File

@@ -121,20 +121,6 @@ static void remove_migration_pte(struct vm_area_struct *vma,
if (!is_migration_entry(entry) || migration_entry_to_page(entry) != old)
goto out;
/*
* Yes, ignore the return value from a GFP_ATOMIC mem_cgroup_charge.
* Failure is not an option here: we're now expected to remove every
* migration pte, and will cause crashes otherwise. Normally this
* is not an issue: mem_cgroup_prepare_migration bumped up the old
* page_cgroup count for safety, that's now attached to the new page,
* so this charge should just be another incrementation of the count,
* to keep in balance with rmap.c's mem_cgroup_uncharging. But if
* there's been a force_empty, those reference counts may no longer
* be reliable, and this charge can actually fail: oh well, we don't
* make the situation any worse by proceeding as if it had succeeded.
*/
mem_cgroup_charge_migrate_fixup(new, mm, GFP_ATOMIC);
get_page(new);
pte = pte_mkold(mk_pte(new, vma->vm_page_prot));
if (is_write_migration_entry(entry))
@@ -378,9 +364,6 @@ static void migrate_page_copy(struct page *newpage, struct page *page)
anon = PageAnon(page);
page->mapping = NULL;
if (!anon) /* This page was removed from radix-tree. */
mem_cgroup_uncharge_cache_page(page);
/*
* If any waiters have accumulated on the new page then
* wake them up.
@@ -614,6 +597,7 @@ static int unmap_and_move(new_page_t get_new_page, unsigned long private,
struct page *newpage = get_new_page(page, private, &result);
int rcu_locked = 0;
int charge = 0;
struct mem_cgroup *mem;
if (!newpage)
return -ENOMEM;
@@ -623,24 +607,26 @@ static int unmap_and_move(new_page_t get_new_page, unsigned long private,
goto move_newpage;
}
charge = mem_cgroup_prepare_migration(page, newpage);
if (charge == -ENOMEM) {
rc = -ENOMEM;
goto move_newpage;
}
/* prepare cgroup just returns 0 or -ENOMEM */
BUG_ON(charge);
rc = -EAGAIN;
if (!trylock_page(page)) {
if (!force)
goto move_newpage;
lock_page(page);
}
/* charge against new page */
charge = mem_cgroup_prepare_migration(page, &mem);
if (charge == -ENOMEM) {
rc = -ENOMEM;
goto unlock;
}
BUG_ON(charge);
if (PageWriteback(page)) {
if (!force)
goto unlock;
goto uncharge;
wait_on_page_writeback(page);
}
/*
@@ -693,7 +679,9 @@ static int unmap_and_move(new_page_t get_new_page, unsigned long private,
rcu_unlock:
if (rcu_locked)
rcu_read_unlock();
uncharge:
if (!charge)
mem_cgroup_end_migration(mem, page, newpage);
unlock:
unlock_page(page);
@@ -709,8 +697,6 @@ unlock:
}
move_newpage:
if (!charge)
mem_cgroup_end_migration(newpage);
/*
* Move the new page to the LRU. If migration was not successful