clean up comment. prepare cgroup return 0 or -ENOMEN, others return -EAGAIN.
avoid conflict meanings.
Signed-off-by: Figo.zhang <[email protected]>
---
mm/migrate.c | 3 +--
1 files changed, 1 insertions(+), 2 deletions(-)
diff --git a/mm/migrate.c b/mm/migrate.c
index 34132f8..d65b351 100644
--- a/mm/migrate.c
+++ b/mm/migrate.c
@@ -647,7 +647,6 @@ static int unmap_and_move(new_page_t get_new_page, unsigned long private,
if (unlikely(split_huge_page(page)))
goto move_newpage;
- /* prepare cgroup just returns 0 or -ENOMEM */
rc = -EAGAIN;
if (!trylock_page(page)) {
@@ -687,7 +686,7 @@ static int unmap_and_move(new_page_t get_new_page, unsigned long private,
goto unlock;
}
- /* charge against new page */
+ /* charge against new page, return 0 or -ENOMEM */
charge = mem_cgroup_prepare_migration(page, newpage, &mem, GFP_KERNEL);
if (charge == -ENOMEM) {
rc = -ENOMEM;
On Sat, 07 May 2011 00:03:11 +0800
"Figo.zhang" <[email protected]> wrote:
>
> clean up comment. prepare cgroup return 0 or -ENOMEN, others return -EAGAIN.
> avoid conflict meanings.
>
> Signed-off-by: Figo.zhang <[email protected]>
> ---
> mm/migrate.c | 3 +--
> 1 files changed, 1 insertions(+), 2 deletions(-)
>
> diff --git a/mm/migrate.c b/mm/migrate.c
> index 34132f8..d65b351 100644
> --- a/mm/migrate.c
> +++ b/mm/migrate.c
> @@ -647,7 +647,6 @@ static int unmap_and_move(new_page_t get_new_page, unsigned long private,
> if (unlikely(split_huge_page(page)))
> goto move_newpage;
>
> - /* prepare cgroup just returns 0 or -ENOMEM */
> rc = -EAGAIN;
>
> if (!trylock_page(page)) {
> @@ -687,7 +686,7 @@ static int unmap_and_move(new_page_t get_new_page, unsigned long private,
> goto unlock;
> }
>
> - /* charge against new page */
> + /* charge against new page, return 0 or -ENOMEM */
> charge = mem_cgroup_prepare_migration(page, newpage, &mem, GFP_KERNEL);
> if (charge == -ENOMEM) {
> rc = -ENOMEM;
Well it's still pretty confusing - the function can also return -EAGAIN
and -EBUSY, at least.
It would be better to remove this random sprinkle of commentlets and to
properly document unmap_and_move()'s interface in the usual fashion, in
its leading comment block.