On 3/3/25 12:38, Rafael J. Wysocki wrote:

On Mon, Mar 3, 2025 at 4:43 AM Haoxiang Li haoxiang_li2024@163.com wrote:

...

Replace kfree() with em_table_free() to free the memory allocated by em_table_alloc().

Ostensibly, this is fixing a problem, but there's no problem described above. Please describe it.

Thank Rafael for adding me on CC.

...

Fixes: 24e9fb635df2 ("PM: EM: Remove old table") Cc: stable@vger.kernel.org Signed-off-by: Haoxiang Li haoxiang_li2024@163.com

---

kernel/power/energy_model.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/kernel/power/energy_model.c b/kernel/power/energy_model.c index 3874f0e97651..71b60aa20227 100644 --- a/kernel/power/energy_model.c +++ b/kernel/power/energy_model.c @@ -447,7 +447,7 @@ static int em_create_pd(struct device *dev, int nr_states, return 0;

free_pd_table:

•   kfree(em_table);
  

•   em_table_free(em_table);
  

  free_pd: kfree(pd); return -EINVAL;

-- 2.25.1

IMO there is no need to use RCU freeing mechanism, since this table is not used yet. We failed in the initialization steps, so we can simply call kfree() on that memory.

That 'free_pd_table' goto label is triggered before the call to:

rcu_assign_pointer(pd->em_table, em_table);

IMO this is even dangerous in the patch to abuse RCU free for such case.

Regards, Lukasz