From: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
To: Peter Zijlstra <peterz@infradead.org>
Cc: linux-kernel@vger.kernel.org, mingo@kernel.org,
jiangshanlai@gmail.com, dipankar@in.ibm.com,
akpm@linux-foundation.org, mathieu.desnoyers@efficios.com,
josh@joshtriplett.org, tglx@linutronix.de, rostedt@goodmis.org,
dhowells@redhat.com, edumazet@google.com, dvhart@linux.intel.com,
fweisbec@gmail.com, oleg@redhat.com, bobby.prani@gmail.com,
dave@stgolabs.net, waiman.long@hp.com
Subject: Re: [PATCH tip/core/rcu 19/19] rcu: Add fastpath bypassing funnel locking
Date: Fri, 31 Jul 2015 08:57:13 -0700 [thread overview]
Message-ID: <20150731155713.GQ27280@linux.vnet.ibm.com> (raw)
In-Reply-To: <20150730163427.GO25159@twins.programming.kicks-ass.net>
On Thu, Jul 30, 2015 at 06:34:27PM +0200, Peter Zijlstra wrote:
> On Thu, Jul 30, 2015 at 08:34:52AM -0700, Paul E. McKenney wrote:
> > On Thu, Jul 30, 2015 at 04:44:55PM +0200, Peter Zijlstra wrote:
>
> > > If the extra read before the cmpxchg() does not hurt, we should do the
> > > same for mutex and make the above redundant.
> >
> > I am pretty sure that different hardware wants it done differently. :-/
>
> I think that most archs won't notice since any RmW includes a load of
> that variable anyhow. The only case where it can matter is if the RmW is
> done outside of the normal cache hierarchy -- like on Power, where the
> ll/sc bypasses the L1.
Some years back, AMD and Intel variants of x86 had different preferences
on this matter. Timings indicated that one or the other of them (I
cannot recall which) would get the cacheline shared, then have to get
it exclusive, while the other would get it exclusive to begin with.
I honestly do not know what the preferences of current Power hardware
might be.
Thanx, Paul
next prev parent reply other threads:[~2015-07-31 15:57 UTC|newest]
Thread overview: 36+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-07-17 23:29 [PATCH tip/core/rcu 0/19] Expedited grace period changes for 4.3 Paul E. McKenney
2015-07-17 23:29 ` [PATCH tip/core/rcu 01/19] rcu: Stop disabling CPU hotplug in synchronize_rcu_expedited() Paul E. McKenney
2015-07-17 23:29 ` [PATCH tip/core/rcu 02/19] rcu: Remove CONFIG_RCU_CPU_STALL_INFO Paul E. McKenney
2015-07-30 12:49 ` Peter Zijlstra
2015-07-30 15:13 ` Paul E. McKenney
2015-07-30 15:31 ` Peter Zijlstra
2015-07-30 15:45 ` Josh Triplett
2015-07-17 23:29 ` [PATCH tip/core/rcu 03/19] rcu: Switch synchronize_sched_expedited() to stop_one_cpu() Paul E. McKenney
2015-07-17 23:29 ` [PATCH tip/core/rcu 04/19] rcu: Rework synchronize_sched_expedited() counter handling Paul E. McKenney
2015-07-17 23:29 ` [PATCH tip/core/rcu 05/19] rcu: Get rid of synchronize_sched_expedited()'s polling loop Paul E. McKenney
2015-07-17 23:29 ` [PATCH tip/core/rcu 06/19] rcu: Make expedited GP CPU stoppage asynchronous Paul E. McKenney
2015-07-17 23:29 ` [PATCH tip/core/rcu 07/19] rcu: Abstract sequence counting from synchronize_sched_expedited() Paul E. McKenney
2015-07-17 23:29 ` [PATCH tip/core/rcu 08/19] rcu: Make synchronize_rcu_expedited() use sequence-counter scheme Paul E. McKenney
2015-07-17 23:29 ` [PATCH tip/core/rcu 09/19] rcu: Abstract funnel locking from synchronize_sched_expedited() Paul E. McKenney
2015-07-17 23:29 ` [PATCH tip/core/rcu 10/19] rcu: Fix synchronize_sched_expedited() type error for "s" Paul E. McKenney
2015-07-17 23:29 ` [PATCH tip/core/rcu 11/19] rcu: Use funnel locking for synchronize_rcu_expedited()'s polling loop Paul E. McKenney
2015-07-17 23:29 ` [PATCH tip/core/rcu 12/19] rcu: Apply rcu_seq operations to _rcu_barrier() Paul E. McKenney
2015-07-17 23:29 ` [PATCH tip/core/rcu 13/19] rcu: Consolidate last open-coded expedited memory barrier Paul E. McKenney
2015-07-17 23:29 ` [PATCH tip/core/rcu 14/19] rcu: Extend expedited funnel locking to rcu_data structure Paul E. McKenney
2015-09-20 14:58 ` Sasha Levin
2015-09-21 4:12 ` Paul E. McKenney
2015-09-21 22:04 ` Sasha Levin
2015-09-22 15:10 ` Paul E. McKenney
2015-07-17 23:29 ` [PATCH tip/core/rcu 15/19] rcu: Add stall warnings to synchronize_sched_expedited() Paul E. McKenney
2015-07-17 23:29 ` [PATCH tip/core/rcu 16/19] documentation: Describe new expedited stall warnings Paul E. McKenney
2015-07-17 23:29 ` [PATCH tip/core/rcu 17/19] rcu: Pull out wait_event*() condition into helper function Paul E. McKenney
2015-07-17 23:29 ` [PATCH tip/core/rcu 18/19] rcu: Rename RCU_GP_DONE_FQS to RCU_GP_DOING_FQS Paul E. McKenney
2015-07-17 23:29 ` [PATCH tip/core/rcu 19/19] rcu: Add fastpath bypassing funnel locking Paul E. McKenney
2015-07-30 14:44 ` Peter Zijlstra
2015-07-30 15:34 ` Paul E. McKenney
2015-07-30 15:40 ` Peter Zijlstra
2015-08-03 20:05 ` Steven Rostedt
2015-08-03 20:06 ` Peter Zijlstra
2015-07-30 16:34 ` Peter Zijlstra
2015-07-31 15:57 ` Paul E. McKenney [this message]
2015-07-31 2:03 ` Waiman Long
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20150731155713.GQ27280@linux.vnet.ibm.com \
--to=paulmck@linux.vnet.ibm.com \
--cc=akpm@linux-foundation.org \
--cc=bobby.prani@gmail.com \
--cc=dave@stgolabs.net \
--cc=dhowells@redhat.com \
--cc=dipankar@in.ibm.com \
--cc=dvhart@linux.intel.com \
--cc=edumazet@google.com \
--cc=fweisbec@gmail.com \
--cc=jiangshanlai@gmail.com \
--cc=josh@joshtriplett.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mathieu.desnoyers@efficios.com \
--cc=mingo@kernel.org \
--cc=oleg@redhat.com \
--cc=peterz@infradead.org \
--cc=rostedt@goodmis.org \
--cc=tglx@linutronix.de \
--cc=waiman.long@hp.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).