Re: [PATCH] locking/rtmutex: remove unused cmpxchg_relaxed

From: Davidlohr Bueso
Date: Wed Feb 12 2020 - 10:18:58 EST


On Tue, 04 Feb 2020, Alex Shi wrote:

Thanks Thomas and David!
Is this following patch ok?

So if anything, this really wants to be two patches.

---
From 4cf9e38a73c67c6894f3addb2ddca26bb51b1a28 Mon Sep 17 00:00:00 2001
From: Alex Shi <alex.shi@xxxxxxxxxxxxxxxxx>
Date: Tue, 21 Jan 2020 15:03:33 +0800
Subject: [PATCH v2] locking/rtmutex: optimize rt_mutex_cmpxchg_xxx series func

rt_mutex_cmpxchg_relexed isn't interested by anyone, so remove it.
And Davidlohr Bueso suggests check l->owner before cmpxchg to reduce
lock contention.

Signed-off-by: Alex Shi <alex.shi@xxxxxxxxxxxxxxxxx>
Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Cc: Davidlohr Bueso <dave@xxxxxxxxxxxx>
Cc: Ingo Molnar <mingo@xxxxxxxxxx>
Cc: Will Deacon <will@xxxxxxxxxx>
Cc: linux-kernel@xxxxxxxxxxxxxxx
---
kernel/locking/rtmutex.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/kernel/locking/rtmutex.c b/kernel/locking/rtmutex.c
index 851bbb10819d..eb26f4e57ce4 100644
--- a/kernel/locking/rtmutex.c
+++ b/kernel/locking/rtmutex.c
@@ -141,9 +141,10 @@ static void fixup_rt_mutex_waiters(struct rt_mutex *lock)
* set up.
*/
#ifndef CONFIG_DEBUG_RT_MUTEXES
-# define rt_mutex_cmpxchg_relaxed(l,c,n) (cmpxchg_relaxed(&l->owner, c, n) == c)
-# define rt_mutex_cmpxchg_acquire(l,c,n) (cmpxchg_acquire(&l->owner, c, n) == c)
-# define rt_mutex_cmpxchg_release(l,c,n) (cmpxchg_release(&l->owner, c, n) == c)
+# define rt_mutex_cmpxchg_acquire(l,c,n) \
+ (l->owner == c && cmpxchg_acquire(&l->owner, c, n) == c)
+# define rt_mutex_cmpxchg_release(l,c,n) \
+ (l->owner == c && cmpxchg_release(&l->owner, c, n) == c)

Thomas, should I resend the top-waiter spin series? Otherwise yeah,
I see little point to the CCAS fastpath thing.

Thanks,
Davidlohr