On 2022/4/29 17:54, David Hildenbrand wrote:
> On 25.04.22 15:27, Miaohe Lin wrote:
>> rcu_read_lock is required by grabbing the task refcount but it's not
>> needed for ptrace_may_access. So we could release the rcu lock after
>> task refcount is successfully grabbed to reduce the rcu holding time.
>>
>> Reviewed-by: Muchun Song <songmuchun@xxxxxxxxxxxxx>
>> Reviewed-by: Christoph Hellwig <hch@xxxxxx>
>> Signed-off-by: Miaohe Lin <linmiaohe@xxxxxxxxxx>
>> Cc: Huang Ying <ying.huang@xxxxxxxxx>
>> Cc: David Howells <dhowells@xxxxxxxxxx>
>> Cc: Christoph Lameter <cl@xxxxxxxxx>
>> ---
>> mm/migrate.c | 3 +--
>> 1 file changed, 1 insertion(+), 2 deletions(-)
>>
>> diff --git a/mm/migrate.c b/mm/migrate.c
>> index b2678279eb43..b779646665fe 100644
>> --- a/mm/migrate.c
>> +++ b/mm/migrate.c
>> @@ -1902,17 +1902,16 @@ static struct mm_struct *find_mm_struct(pid_t pid, nodemask_t *mem_nodes)
>> return ERR_PTR(-ESRCH);
>> }
>> get_task_struct(task);
>> + rcu_read_unlock();
>>
>> /*
>> * Check if this process has the right to modify the specified
>> * process. Use the regular "ptrace_may_access()" checks.
>> */
>> if (!ptrace_may_access(task, PTRACE_MODE_READ_REALCREDS)) {
>> - rcu_read_unlock();
>> mm = ERR_PTR(-EPERM);
>> goto out;
>> }
>> - rcu_read_unlock();
>>
>> mm = ERR_PTR(security_task_movememory(task));
>> if (IS_ERR(mm))
>
> Similar pattern in:
>
> mm/mempolicy.c:kernel_migrate_pages()
> kernel/futex/syscalls.c:get_robust_list()
> kernel/nsproxy.c:validate_nsset()
>
> Exception:
>
> sched/core_sched.c:sched_core_share_pid()
>
>
> Should we unify -- i.e., adjust the remaining 3 as well?
>
Sorry for late respond. I think it's fine to do all of this together. But this patch is indeed
under verifying now. I will try to do that after verified. Thanks!
