[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Fri, 7 Jul 2017 19:47:58 +0200
From: Manfred Spraul <manfred@...orfullife.com>
To: Ingo Molnar <mingo@...nel.org>,
Peter Zijlstra <peterz@...radead.org>
Cc: "Paul E. McKenney" <paulmck@...ux.vnet.ibm.com>,
David Laight <David.Laight@...LAB.COM>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"netfilter-devel@...r.kernel.org" <netfilter-devel@...r.kernel.org>,
"netdev@...r.kernel.org" <netdev@...r.kernel.org>,
"oleg@...hat.com" <oleg@...hat.com>,
"akpm@...ux-foundation.org" <akpm@...ux-foundation.org>,
"mingo@...hat.com" <mingo@...hat.com>,
"dave@...olabs.net" <dave@...olabs.net>,
"tj@...nel.org" <tj@...nel.org>, "arnd@...db.de" <arnd@...db.de>,
"linux-arch@...r.kernel.org" <linux-arch@...r.kernel.org>,
"will.deacon@....com" <will.deacon@....com>,
"stern@...land.harvard.edu" <stern@...land.harvard.edu>,
"parri.andrea@...il.com" <parri.andrea@...il.com>,
"torvalds@...ux-foundation.org" <torvalds@...ux-foundation.org>
Subject: Re: [PATCH v2 0/9] Remove spin_unlock_wait()
Hi Ingo,
On 07/07/2017 10:31 AM, Ingo Molnar wrote:
>
> There's another, probably just as significant advantage: queued_spin_unlock_wait()
> is 'read-only', while spin_lock()+spin_unlock() dirties the lock cache line. On
> any bigger system this should make a very measurable difference - if
> spin_unlock_wait() is ever used in a performance critical code path.
At least for ipc/sem:
Dirtying the cacheline (in the slow path) allows to remove a smp_mb() in
the hot path.
So for sem_lock(), I either need a primitive that dirties the cacheline
or sem_lock() must continue to use spin_lock()/spin_unlock().
--
Manfred
Powered by blists - more mailing lists