[PATCH] mm/damon/ops-common: ignore migration request to invalid nodes

[PATCH] mm/damon/ops-common: ignore migration request to invalid nodes

[SeongJae Park]

damon_migrate_pages() try migration even if the target node is invalid.
If users mistakenly make such invalid requests via
DAMOS_MIGRATE_{HOT,COLD} action, below kernel BUG can happen.

    [ 7831.883495] BUG: unable to handle page fault for address: 0000000000001f48
    [ 7831.884160] #PF: supervisor read access in kernel mode
    [ 7831.884681] #PF: error_code(0x0000) - not-present page
    [ 7831.885203] PGD 0 P4D 0
    [ 7831.885468] Oops: Oops: 0000 [#1] SMP PTI
    [ 7831.885852] CPU: 31 UID: 0 PID: 94202 Comm: kdamond.0 Not tainted 6.16.0-rc5-mm-new-damon+ #93 PREEMPT(voluntary)
    [ 7831.886913] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.3-4.el9 04/01/2014
    [ 7831.887777] RIP: 0010:__alloc_frozen_pages_noprof (include/linux/mmzone.h:1724 include/linux/mmzone.h:1750 mm/page_alloc.c:4936 mm/page_alloc.c:5137)
    [...]
    [ 7831.895953] Call Trace:
    [ 7831.896195]  <TASK>
    [ 7831.896397] __folio_alloc_noprof (mm/page_alloc.c:5183 mm/page_alloc.c:5192)
    [ 7831.896787] migrate_pages_batch (mm/migrate.c:1189 mm/migrate.c:1851)
    [ 7831.897228] ? __pfx_alloc_migration_target (mm/migrate.c:2137)
    [ 7831.897735] migrate_pages (mm/migrate.c:2078)
    [ 7831.898141] ? __pfx_alloc_migration_target (mm/migrate.c:2137)
    [ 7831.898664] damon_migrate_folio_list (mm/damon/ops-common.c:321 mm/damon/ops-common.c:354)
    [ 7831.899140] damon_migrate_pages (mm/damon/ops-common.c:405)
    [...]

Add a target node validity check in damon_migrate_pages().  The validity
check is stolen from that of do_pages_move(), which is being used for
move_pages() system call.

Fixes: b51820ebea65 ("mm/damon/paddr: introduce DAMOS_MIGRATE_COLD action for demotion") # 6.11.x
Cc: stable@vger.kernel.org
Cc: Honggyu Kim <honggyu.kim@sk.com>
Signed-off-by: SeongJae Park <sj@kernel.org>
---
 mm/damon/ops-common.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/mm/damon/ops-common.c b/mm/damon/ops-common.c
index 6a9797d1d7ff..99321ff5cb92 100644
--- a/mm/damon/ops-common.c
+++ b/mm/damon/ops-common.c
@@ -383,6 +383,10 @@ unsigned long damon_migrate_pages(struct list_head *folio_list, int target_nid)
 	if (list_empty(folio_list))
 		return nr_migrated;
 
+	if (target_nid < 0 || target_nid >= MAX_NUMNODES ||
+			!node_state(target_nid, N_MEMORY))
+		return nr_migrated;
+
 	noreclaim_flag = memalloc_noreclaim_save();
 
 	nid = folio_nid(lru_to_folio(folio_list));

base-commit: e2c90d41402c324ea81fa3d9c2c1d0f61906c161
-- 
2.39.5

Re: [PATCH] mm/damon/ops-common: ignore migration request to invalid nodes

[Joshua Hahn]

On Sun, 20 Jul 2025 11:58:22 -0700 SeongJae Park <sj@kernel.org> wrote:

> damon_migrate_pages() try migration even if the target node is invalid.
> If users mistakenly make such invalid requests via
> DAMOS_MIGRATE_{HOT,COLD} action, below kernel BUG can happen.
> 
>     [ 7831.883495] BUG: unable to handle page fault for address: 0000000000001f48
>     [ 7831.884160] #PF: supervisor read access in kernel mode
>     [ 7831.884681] #PF: error_code(0x0000) - not-present page
>     [ 7831.885203] PGD 0 P4D 0
>     [ 7831.885468] Oops: Oops: 0000 [#1] SMP PTI
>     [ 7831.885852] CPU: 31 UID: 0 PID: 94202 Comm: kdamond.0 Not tainted 6.16.0-rc5-mm-new-damon+ #93 PREEMPT(voluntary)
>     [ 7831.886913] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.3-4.el9 04/01/2014
>     [ 7831.887777] RIP: 0010:__alloc_frozen_pages_noprof (include/linux/mmzone.h:1724 include/linux/mmzone.h:1750 mm/page_alloc.c:4936 mm/page_alloc.c:5137)
>     [...]
>     [ 7831.895953] Call Trace:
>     [ 7831.896195]  <TASK>
>     [ 7831.896397] __folio_alloc_noprof (mm/page_alloc.c:5183 mm/page_alloc.c:5192)
>     [ 7831.896787] migrate_pages_batch (mm/migrate.c:1189 mm/migrate.c:1851)
>     [ 7831.897228] ? __pfx_alloc_migration_target (mm/migrate.c:2137)
>     [ 7831.897735] migrate_pages (mm/migrate.c:2078)
>     [ 7831.898141] ? __pfx_alloc_migration_target (mm/migrate.c:2137)
>     [ 7831.898664] damon_migrate_folio_list (mm/damon/ops-common.c:321 mm/damon/ops-common.c:354)
>     [ 7831.899140] damon_migrate_pages (mm/damon/ops-common.c:405)
>     [...]
> 
> Add a target node validity check in damon_migrate_pages().  The validity
> check is stolen from that of do_pages_move(), which is being used for
> move_pages() system call.
> 
> Fixes: b51820ebea65 ("mm/damon/paddr: introduce DAMOS_MIGRATE_COLD action for demotion") # 6.11.x
> Cc: stable@vger.kernel.org
> Cc: Honggyu Kim <honggyu.kim@sk.com>
> Signed-off-by: SeongJae Park <sj@kernel.org>
> ---

LGTM, thank you SJ!

On a side note... This seems like it would be a common check. However, doing a
(quick) search seems to return no function that checks whether a node is valid.
Perhaps it would make sense to look deeper and see how many other functions
make this check, and export this as a function? I can try spinning something
if it makes sense to you : -)

Reviewed-by: Joshua Hahn <joshua.hahnjy@gmail.com>

Sent using hkml (https://github.com/sjp38/hackermail)

Re: [PATCH] mm/damon/ops-common: ignore migration request to invalid nodes

[SeongJae Park]

On Mon, 21 Jul 2025 08:28:26 -0700 Joshua Hahn <joshua.hahnjy@gmail.com> wrote:

> On Sun, 20 Jul 2025 11:58:22 -0700 SeongJae Park <sj@kernel.org> wrote:
> 
> > damon_migrate_pages() try migration even if the target node is invalid.
> > If users mistakenly make such invalid requests via
> > DAMOS_MIGRATE_{HOT,COLD} action, below kernel BUG can happen.
[...]
> > Add a target node validity check in damon_migrate_pages().  The validity
> > check is stolen from that of do_pages_move(), which is being used for
> > move_pages() system call.
> > 
> > Fixes: b51820ebea65 ("mm/damon/paddr: introduce DAMOS_MIGRATE_COLD action for demotion") # 6.11.x
> > Cc: stable@vger.kernel.org
> > Cc: Honggyu Kim <honggyu.kim@sk.com>
> > Signed-off-by: SeongJae Park <sj@kernel.org>
> > ---
> 
> LGTM, thank you SJ!
> 
> On a side note... This seems like it would be a common check. However, doing a
> (quick) search seems to return no function that checks whether a node is valid.
> Perhaps it would make sense to look deeper and see how many other functions
> make this check, and export this as a function? I can try spinning something
> if it makes sense to you : -)

My humble impression was that this check is short enough to be ok to be
open-coded, but please don't be blocked on my opinion :)

> 
> Reviewed-by: Joshua Hahn <joshua.hahnjy@gmail.com>

Thank you!

> 
> Sent using hkml (https://github.com/sjp38/hackermail)

Thanks,
SJ

Re: [PATCH] mm/damon/ops-common: ignore migration request to invalid nodes

[Andrew Morton]

On Sun, 20 Jul 2025 11:58:22 -0700 SeongJae Park <sj@kernel.org> wrote:

> damon_migrate_pages() try migration even if the target node is invalid.
> If users mistakenly make such invalid requests via
> DAMOS_MIGRATE_{HOT,COLD} action, below kernel BUG can happen.
> 
>     [ 7831.883495] BUG: unable to handle page fault for address: 0000000000001f48
>     [ 7831.884160] #PF: supervisor read access in kernel mode
>     [ 7831.884681] #PF: error_code(0x0000) - not-present page
>     [ 7831.885203] PGD 0 P4D 0
>     [ 7831.885468] Oops: Oops: 0000 [#1] SMP PTI
>     [ 7831.885852] CPU: 31 UID: 0 PID: 94202 Comm: kdamond.0 Not tainted 6.16.0-rc5-mm-new-damon+ #93 PREEMPT(voluntary)
>     [ 7831.886913] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.3-4.el9 04/01/2014
>     [ 7831.887777] RIP: 0010:__alloc_frozen_pages_noprof (include/linux/mmzone.h:1724 include/linux/mmzone.h:1750 mm/page_alloc.c:4936 mm/page_alloc.c:5137)
>     [...]
>     [ 7831.895953] Call Trace:
>     [ 7831.896195]  <TASK>
>     [ 7831.896397] __folio_alloc_noprof (mm/page_alloc.c:5183 mm/page_alloc.c:5192)
>     [ 7831.896787] migrate_pages_batch (mm/migrate.c:1189 mm/migrate.c:1851)
>     [ 7831.897228] ? __pfx_alloc_migration_target (mm/migrate.c:2137)
>     [ 7831.897735] migrate_pages (mm/migrate.c:2078)
>     [ 7831.898141] ? __pfx_alloc_migration_target (mm/migrate.c:2137)
>     [ 7831.898664] damon_migrate_folio_list (mm/damon/ops-common.c:321 mm/damon/ops-common.c:354)
>     [ 7831.899140] damon_migrate_pages (mm/damon/ops-common.c:405)
>     [...]
> 
> Add a target node validity check in damon_migrate_pages().  The validity
> check is stolen from that of do_pages_move(), which is being used for
> move_pages() system call.
> 
> Fixes: b51820ebea65 ("mm/damon/paddr: introduce DAMOS_MIGRATE_COLD action for demotion") # 6.11.x
> Cc: stable@vger.kernel.org
>
> ...
>
> --- a/mm/damon/ops-common.c
> +++ b/mm/damon/ops-common.c
> @@ -383,6 +383,10 @@ unsigned long damon_migrate_pages(struct list_head *folio_list, int target_nid)
>  	if (list_empty(folio_list))
>  		return nr_migrated;
>  
> +	if (target_nid < 0 || target_nid >= MAX_NUMNODES ||
> +			!node_state(target_nid, N_MEMORY))
> +		return nr_migrated;
> +
>  	noreclaim_flag = memalloc_noreclaim_save();
>  
>  	nid = folio_nid(lru_to_folio(folio_list));
> 

OK.  damon_migrate_pages() exists only in mm.git thanks to 13dde31db71f
("mm/damon: move migration helpers from paddr to ops-common").  I
assume that you'll send the -stable people a patch which adds this check into
damon_pa_migrate_pages() when called upon to do so.

Re: [PATCH] mm/damon/ops-common: ignore migration request to invalid nodes

[SeongJae Park]

On Mon, 21 Jul 2025 19:56:58 -0700 Andrew Morton <akpm@linux-foundation.org> wrote:

> On Sun, 20 Jul 2025 11:58:22 -0700 SeongJae Park <sj@kernel.org> wrote:
> 
[...]
> > Add a target node validity check in damon_migrate_pages().  The validity
> > check is stolen from that of do_pages_move(), which is being used for
> > move_pages() system call.
> > 
> > Fixes: b51820ebea65 ("mm/damon/paddr: introduce DAMOS_MIGRATE_COLD action for demotion") # 6.11.x
> > Cc: stable@vger.kernel.org
> >
> > ...
> >
> > --- a/mm/damon/ops-common.c
> > +++ b/mm/damon/ops-common.c
> > @@ -383,6 +383,10 @@ unsigned long damon_migrate_pages(struct list_head *folio_list, int target_nid)
> >  	if (list_empty(folio_list))
> >  		return nr_migrated;
> >  
> > +	if (target_nid < 0 || target_nid >= MAX_NUMNODES ||
> > +			!node_state(target_nid, N_MEMORY))
> > +		return nr_migrated;
> > +
> >  	noreclaim_flag = memalloc_noreclaim_save();
> >  
> >  	nid = folio_nid(lru_to_folio(folio_list));
> > 
> 
> OK.  damon_migrate_pages() exists only in mm.git thanks to 13dde31db71f
> ("mm/damon: move migration helpers from paddr to ops-common").  I
> assume that you'll send the -stable people a patch which adds this check into
> damon_pa_migrate_pages() when called upon to do so.

That's very correct, Andrew.  I am planning to do so as soon as this is merged
into the mainline :)


Thanks,
SJ