2019-11-19 13:21:40

by Andrey Ryabinin

Subject: [PATCH] mm/ksm: Don't WARN if page is still mapped in remove_stable_node()

It's possible to hit the WARN_ON_ONCE(page_mapped(page)) in
remove_stable_node() when it races with __mmput() and squeezes
in between ksm_exit() and exit_mmap().

WARNING: CPU: 0 PID: 3295 at mm/ksm.c:888 remove_stable_node+0x10c/0x150

Call Trace:
remove_all_stable_nodes+0x12b/0x330
run_store+0x4ef/0x7b0
kernfs_fop_write+0x200/0x420
vfs_write+0x154/0x450
ksys_write+0xf9/0x1d0
do_syscall_64+0x99/0x510
entry_SYSCALL_64_after_hwframe+0x49/0xbe

Remove the warning as there is nothing scary going on.

Fixes: cbf86cfe04a6 ("ksm: remove old stable nodes more thoroughly")
Signed-off-by: Andrey Ryabinin <[email protected]>
Cc: <[email protected]>
---
mm/ksm.c | 14 +++++++-------
1 file changed, 7 insertions(+), 7 deletions(-)

diff --git a/mm/ksm.c b/mm/ksm.c
index dbee2eb4dd05..7905934cd3ad 100644
--- a/mm/ksm.c
+++ b/mm/ksm.c
@@ -885,13 +885,13 @@ static int remove_stable_node(struct stable_node *stable_node)
return 0;
}

- if (WARN_ON_ONCE(page_mapped(page))) {
- /*
- * This should not happen: but if it does, just refuse to let
- * merge_across_nodes be switched - there is no need to panic.
- */
- err = -EBUSY;
- } else {
+ /*
+ * Page could be still mapped if this races with __mmput() running in
+ * between ksm_exit() and exit_mmap(). Just refuse to let
+ * merge_across_nodes/max_page_sharing be switched.
+ */
+ err = -EBUSY;
+ if (!page_mapped(page)) {
/*
* The stable node did not yet appear stale to get_ksm_page(),
* since that allows for an unmapped ksm page to be recognized
--
2.23.0

2019-11-21 04:16:08

by Hugh Dickins

Subject: Re: [PATCH] mm/ksm: Don't WARN if page is still mapped in remove_stable_node()

On Tue, 19 Nov 2019, Andrey Ryabinin wrote:

> It's possible to hit the WARN_ON_ONCE(page_mapped(page)) in
> remove_stable_node() when it races with __mmput() and squeezes
> in between ksm_exit() and exit_mmap().
>
> WARNING: CPU: 0 PID: 3295 at mm/ksm.c:888 remove_stable_node+0x10c/0x150
>
> Call Trace:
> remove_all_stable_nodes+0x12b/0x330
> run_store+0x4ef/0x7b0
> kernfs_fop_write+0x200/0x420
> vfs_write+0x154/0x450
> ksys_write+0xf9/0x1d0
> do_syscall_64+0x99/0x510
> entry_SYSCALL_64_after_hwframe+0x49/0xbe
>
> Remove the warning as there is nothing scary going on.
>
> Fixes: cbf86cfe04a6 ("ksm: remove old stable nodes more thoroughly")
> Signed-off-by: Andrey Ryabinin <[email protected]>

Thanks -
Acked-by: Hugh Dickins <[email protected]>

> Cc: <[email protected]>

For removing a WARN_ON_ONCE? No, I don't think this needs to go to
stable; but I won't resist if Andrew or autobot insist otherwise.

> ---
> mm/ksm.c | 14 +++++++-------
> 1 file changed, 7 insertions(+), 7 deletions(-)
>
> diff --git a/mm/ksm.c b/mm/ksm.c
> index dbee2eb4dd05..7905934cd3ad 100644
> --- a/mm/ksm.c
> +++ b/mm/ksm.c
> @@ -885,13 +885,13 @@ static int remove_stable_node(struct stable_node *stable_node)
> return 0;
> }
>
> - if (WARN_ON_ONCE(page_mapped(page))) {
> - /*
> - * This should not happen: but if it does, just refuse to let
> - * merge_across_nodes be switched - there is no need to panic.
> - */
> - err = -EBUSY;
> - } else {
> + /*
> + * Page could be still mapped if this races with __mmput() running in
> + * between ksm_exit() and exit_mmap(). Just refuse to let
> + * merge_across_nodes/max_page_sharing be switched.

Good comment, thank you.

> + */
> + err = -EBUSY;
> + if (!page_mapped(page)) {
> /*
> * The stable node did not yet appear stale to get_ksm_page(),
> * since that allows for an unmapped ksm page to be recognized
> --
> 2.23.0