On Thu, 21 Oct 2021 14:30:50 +0100, Pavel Begunkov wrote:
> Kill extra rcu_read_lock/unlock() pair in blk_try_enter_queue().
> Testing with io_uring (high batching) with nullblk:
>
> Before:
> 3.20% io_uring [kernel.vmlinux] [k] __rcu_read_unlock
> 3.05% io_uring [kernel.vmlinux] [k] __rcu_read_lock
>
> [...]
Applied, thanks!
[1/2] percpu_ref: percpu_ref_tryget_live() version holding RCU
commit: 3b13c168186c115501ee7d194460ba2f8c825155
[2/2] block: kill extra rcu lock/unlock in queue enter
commit: e94f68527a35271131cdf9d3fb4eb3c2513dc3d0
Best regards,
--
Jens Axboe
