Messages in this thread | | | Date | Mon, 31 Aug 2015 08:02:08 +0200 | From | Ingo Molnar <> | Subject | Re: [PATCH] task_work: remove fifo ordering guarantee |
| |
* Oleg Nesterov <oleg@redhat.com> wrote:
> On 08/29, Ingo Molnar wrote: > > > > So I'm wondering, is there any strong reason why we couldn't use a double linked > > list and still do FIFO and remove that silly linear list walking hack? > > This will obviously enlarge callback_head, and it is often embedded. > But this is minor. > > If we use a double linked list we can't do task_work_add() lockless. > So we will need another spinlock_t in task_struct. We can't use pi_lock.
The fact that the O(N) overhead was measured in real apps to be in the milliseconds IMHO weakens cycle-level concerns about also having a spinlock next to the list head. (There's no additional cacheline bouncing concerns with the spinlock: the head of a LIFO list is essentially a bouncing cacheline.)
If there's some other solution, sure, but LIFO queues tend to be trouble down the line.
Thanks,
Ingo
| |