Re: [PATCH -tip 0/2] locking/rwbase: Two reader optimizations

[Waiman Long]

On 9/20/21 1:20 AM, Davidlohr Bueso wrote:
> Hi,
>
> Patch 1 is a barrier optimization that came up from the reader
> fastpath ordering auditing.
>
> Patch 2 is a resend of the previous broken patch that attempts
> to use wake_q for read_unlock() slowpath.
>
> Tested on v5.15.y-rt. Applies against tip/urgent.
>
> Thanks!
>
> Davidlohr Bueso (2):
>    locking/rwbase: Optimize rwbase_read_trylock
>    locking/rwbase: Lockless reader waking up a writer
>
>   kernel/locking/rtmutex.c   | 19 +++++++++++++------
>   kernel/locking/rwbase_rt.c | 11 +++++++----
>   2 files changed, 20 insertions(+), 10 deletions(-)
>
> --
> 2.26.2
Your patches look good to me.

Acked-by: Waiman Long <longman@xxxxxxxxxx>