bpf: Add warning to detect memory leak in bpf_selem_unlink_nofail()
PATCH 1/1 bpf: Add warning to detect memory leak in bpf_selem_unlink_nofail()
While very unlikely, local storage theoretically may leak memory of the size of "struct bpf_local_storage" when destroy() fails to grab local_storage->lock and initializes selem->local_storage before other racing map_free() see it. Warn the user to allow debugging the issue instead of leaking the memory silently.
Note that test_maps in bpf selftests already stress tested bpf_selem_unlink_nofail() by creating 4096 sockets and then immediately destroying them in multiple threads. With 64 threads, 64 x 4096 socket local storages were created and destroyed during the test and no warning in the function were triggered.
--- kernel/bpf/bpf_local_storage.c | 8 ++++++++ 1 file changed, 8 insertions(+)
kernel/bpf/bpf_local_storage.c +8 -0
@@ -494,6 +494,14 @@ static void bpf_selem_unlink_nofail(struct bpf_local_storage_elem *selem, }raw_res_spin_unlock_irqrestore (&local_storage->lock, flags); } +/* +* Highly unlikely scenario: memory leak +* +* When destroy() fails to acqurire local_storage->lock and initializes +* selem->local_storage to NULL before any racing map_free() sees the same +* selem, no one will free the local storage. +*/ +WARN_ON_ONCE (err && !in_map_free);if (!err || !in_map_free)RCU_INIT_POINTER (selem->local_storage,NULL ); }
Signature · 1 line
2.52.0
Quoted 7 lines · Amery Hung
While very unlikely, local storage theoretically may leak memory of the size of "struct
bpf_local_storage" whendestroy()fails to grablocal_storage->lockand initializesselem->local_storagebefore other racingmap_free()see it. Warn the user to allow debugging the issue instead of leaking the memory silently.
Quoted 8 lines
Note that
test_mapsin bpf selftests already stress testedbpf_selem_unlink_nofail()by creating 4096 sockets and then immediately destroying them in multiple threads. With 64 threads, 64 x 4096 socket local storages were created and destroyed during the test and no warning in the function were triggered.Signed-off-by Amery Hung <[email protected]>---
Quoted 25 lines
kernel/bpf/bpf_local_storage.c | 8 ++++++++ 1 file changed, 8 insertions(+)kernel/bpf/bpf_local_storage.c +8
-1@@ -494,6 +494,14 @@ static void bpf_selem_unlink_nofail(struct bpf_local_storage_elem *selem, }raw_res_spin_unlock_irqrestore (&local_storage->lock, flags); } +/* +* Highly unlikely scenario: memory leak +* +* When destroy() fails to acqurire local_storage->lock and initializes +* selem->local_storage to NULL before any racing map_free() sees the same +* selem, no one will free the local storage. +*/ +WARN_ON_ONCE (err && !in_map_free);if (!err || !in_map_free)RCU_INIT_POINTER (selem->local_storage,NULL ); } --2.52.0
This patch was applied to bpf/bpf-next.git (master) by Martin KaFai Lau <[email protected]>:
Hello:
This patch was applied to bpf/bpf-next.git (master) by Martin KaFai Lau <[email protected]>:
Quoted 6 lines · On Wed, 18 Mar 2026 15:42:19 -0700 you wrote:
While very unlikely, local storage theoretically may leak memory of the size of "struct
bpf_local_storage" whendestroy()fails to grablocal_storage->lockand initializesselem->local_storagebefore other racingmap_free()see it. Warn the user to allow debugging the issue instead of leaking the memory silently.
Quoted 7 lines
Note that
test_mapsin bpf selftests already stress testedbpf_selem_unlink_nofail()by creating 4096 sockets and then immediately destroying them in multiple threads. With 64 threads, 64 x 4096 socket local storages were created and destroyed during the test and no warning in the function were triggered.[...]
Here is the summary with links:
- [bpf-next,v1,1/1] bpf: Add warning to detect memory leak in bpf_selem_unlink_nofail()
https://git.kernel.org/bpf/bpf-next/c/4b21ea502483
You are awesome, thank you!
Signature · 2 lines
Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html