From: Boqun Feng <boqun.feng@gmail.com>
To: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>,
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, fweisbec@gmail.com,
oleg@redhat.com, will.deacon@arm.com
Subject: Re: [PATCH tip/core/rcu 4/5] sys_membarrier: Add expedited option
Date: Thu, 27 Jul 2017 21:55:51 +0800 [thread overview]
Message-ID: <20170727135551.tg7cduw22krjzbvt@tardis> (raw)
In-Reply-To: <20170727130816.GP3730@linux.vnet.ibm.com>
[-- Attachment #1: Type: text/plain, Size: 1981 bytes --]
Hi Paul,
I have a side question out of curiosity:
How does synchronize_sched() work properly for sys_membarrier()?
sys_membarrier() requires every other CPU does a smp_mb() before it
returns, and I know synchronize_sched() will wait until all CPUs running
a kernel thread do a context-switch, which has a smp_mb(). However, I
believe sched flavor RCU treat CPU running a user thread as a quiesent
state, so synchronize_sched() could return without that CPU does a
context switch.
So why could we use synchronize_sched() for sys_membarrier()?
In particular, could the following happens?
CPU 0: CPU 1:
========================= ==========================
<in user space> <in user space>
{read Y}(reordered) <------------------------------+
store Y; |
read X; --------------------------------------+ |
sys_membarrier(): <timer interrupt> | |
synchronize_sched(); update_process_times(user): //user == true | |
rcu_check_callbacks(usr): | |
if (user || ..) { | |
rcu_sched_qs() | |
... | |
<report quesient state in softirq> | |
<return to user space> | |
read Y; --------------------------------------+----+
store X; |
{read X}(reordered) <-------------------------+
I assume the timer interrupt handler, which interrupts a user space and
reports a quiesent state for sched flavor RCU, may not have a smp_mb()
in some code path.
I may miss something subtle, but it just not very obvious how
synchronize_sched() will guarantee a remote CPU running in userspace to
do a smp_mb() before it returns, this is at least not in RCU
requirements, right?
Regards,
Boqun
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
next prev parent reply other threads:[~2017-07-27 13:55 UTC|newest]
Thread overview: 68+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-07-24 21:57 [PATCH tip/core/rcu 0/5] Related non-RCU updates Paul E. McKenney
2017-07-24 21:58 ` [PATCH tip/core/rcu 1/5] module: Fix pr_fmt() bug for header use of printk Paul E. McKenney
2017-07-24 21:58 ` [PATCH tip/core/rcu 2/5] init_task: Remove redundant INIT_TASK_RCU_TREE_PREEMPT() macro Paul E. McKenney
2017-07-24 21:58 ` [PATCH tip/core/rcu 3/5] EXPERIMENTAL sched: Allow migrating kthreads into online but inactive CPUs Paul E. McKenney
2017-07-24 21:58 ` [PATCH tip/core/rcu 4/5] sys_membarrier: Add expedited option Paul E. McKenney
2017-07-25 4:27 ` Boqun Feng
2017-07-25 16:24 ` Paul E. McKenney
2017-07-25 13:21 ` Mathieu Desnoyers
2017-07-25 16:48 ` Paul E. McKenney
2017-07-25 16:33 ` Peter Zijlstra
2017-07-25 16:49 ` Paul E. McKenney
2017-07-25 16:59 ` Peter Zijlstra
2017-07-25 17:17 ` Paul E. McKenney
2017-07-25 18:53 ` Peter Zijlstra
2017-07-25 19:36 ` Paul E. McKenney
2017-07-25 20:24 ` Peter Zijlstra
2017-07-25 21:19 ` Paul E. McKenney
2017-07-25 21:55 ` Peter Zijlstra
2017-07-25 22:39 ` Mathieu Desnoyers
2017-07-25 22:50 ` Mathieu Desnoyers
2017-07-26 0:01 ` Paul E. McKenney
2017-07-26 7:46 ` Peter Zijlstra
2017-07-26 15:42 ` Paul E. McKenney
2017-07-26 18:01 ` Mathieu Desnoyers
2017-07-26 18:30 ` Paul E. McKenney
2017-07-26 20:37 ` Mathieu Desnoyers
2017-07-26 21:11 ` Paul E. McKenney
2017-07-27 1:45 ` Paul E. McKenney
2017-07-27 12:39 ` Mathieu Desnoyers
2017-07-27 14:44 ` Paul E. McKenney
2017-07-27 10:24 ` Peter Zijlstra
2017-07-27 14:52 ` Paul E. McKenney
2017-07-27 8:53 ` Peter Zijlstra
2017-07-27 10:09 ` Peter Zijlstra
2017-07-27 10:22 ` Will Deacon
2017-07-27 13:14 ` Paul E. McKenney
2017-07-25 23:59 ` Paul E. McKenney
2017-07-26 7:41 ` Peter Zijlstra
2017-07-26 15:41 ` Paul E. McKenney
2017-07-27 8:30 ` Peter Zijlstra
2017-07-27 13:08 ` Paul E. McKenney
2017-07-27 13:49 ` Peter Zijlstra
2017-07-27 14:32 ` Paul E. McKenney
2017-07-27 14:36 ` Peter Zijlstra
2017-07-27 14:46 ` Paul E. McKenney
2017-07-27 13:55 ` Boqun Feng [this message]
2017-07-27 14:16 ` Paul E. McKenney
2017-07-27 14:29 ` Boqun Feng
2017-07-27 14:36 ` Paul E. McKenney
2017-07-27 14:41 ` Will Deacon
2017-07-27 14:47 ` Boqun Feng
2017-07-27 14:55 ` Paul E. McKenney
2017-07-27 13:56 ` Peter Zijlstra
2017-07-27 15:19 ` Peter Zijlstra
2017-07-26 9:36 ` Will Deacon
2017-07-26 15:46 ` Paul E. McKenney
2017-07-27 10:14 ` Peter Zijlstra
2017-07-27 12:56 ` Paul E. McKenney
2017-07-27 13:37 ` Peter Zijlstra
2017-07-27 14:33 ` Paul E. McKenney
2017-07-24 21:58 ` [PATCH tip/core/rcu 5/5] EXP: sched/cputime: Fix using smp_processor_id() in preemptible Paul E. McKenney
2017-07-24 22:01 ` Wanpeng Li
2017-07-24 22:29 ` Paul E. McKenney
2017-07-31 22:51 ` [PATCH tip/core/rcu 0/5] Related non-RCU updates Paul E. McKenney
2017-07-31 22:53 ` [PATCH v2 tip/core/rcu 1/4] module: Fix pr_fmt() bug for header use of printk Paul E. McKenney
2017-07-31 22:53 ` [PATCH v2 tip/core/rcu 2/4] init_task: Remove redundant INIT_TASK_RCU_TREE_PREEMPT() macro Paul E. McKenney
2017-07-31 22:53 ` [PATCH v2 tip/core/rcu 3/4] sched: Allow migrating kthreads into online but inactive CPUs Paul E. McKenney
2017-07-31 22:53 ` [PATCH v2 tip/core/rcu 4/4] membarrier: Expedited private command Paul E. McKenney
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=20170727135551.tg7cduw22krjzbvt@tardis \
--to=boqun.feng@gmail.com \
--cc=akpm@linux-foundation.org \
--cc=dhowells@redhat.com \
--cc=dipankar@in.ibm.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=paulmck@linux.vnet.ibm.com \
--cc=peterz@infradead.org \
--cc=rostedt@goodmis.org \
--cc=tglx@linutronix.de \
--cc=will.deacon@arm.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).