From: Frederic Weisbecker <frederic@kernel.org>
To: paulmck@kernel.org
Cc: rcu@vger.kernel.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, peterz@infradead.org,
rostedt@goodmis.org, dhowells@redhat.com, edumazet@google.com,
fweisbec@gmail.com, oleg@redhat.com, joel@joelfernandes.org,
"Paul E. McKenney" <paulmck@linux.ibm.com>
Subject: Re: [PATCH tip/core/rcu 08/12] rcu: Force tick on for nohz_full CPUs not reaching quiescent states
Date: Thu, 3 Oct 2019 16:50:55 +0200 [thread overview]
Message-ID: <20191003145054.GC27555@lenoir> (raw)
In-Reply-To: <20191003013903.13079-8-paulmck@kernel.org>
On Wed, Oct 02, 2019 at 06:38:59PM -0700, paulmck@kernel.org wrote:
> From: "Paul E. McKenney" <paulmck@linux.ibm.com>
>
> CPUs running for long time periods in the kernel in nohz_full mode
> might leave the scheduling-clock interrupt disabled for then full
> duration of their in-kernel execution. This can (among other things)
> delay grace periods. This commit therefore forces the tick back on
> for any nohz_full CPU that is failing to pass through a quiescent state
> upon return from interrupt, which the resched_cpu() will induce.
>
> Reported-by: Joel Fernandes <joel@joelfernandes.org>
> [ paulmck: Clear ->rcu_forced_tick as reported by Joel Fernandes testing. ]
> [ paulmck: Apply Joel Fernandes TICK_DEP_MASK_RCU->TICK_DEP_BIT_RCU fix. ]
> Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
> ---
> kernel/rcu/tree.c | 38 +++++++++++++++++++++++++++++++-------
> kernel/rcu/tree.h | 1 +
> 2 files changed, 32 insertions(+), 7 deletions(-)
>
> diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c
> index 74bf5c65..621cc06 100644
> --- a/kernel/rcu/tree.c
> +++ b/kernel/rcu/tree.c
> @@ -650,6 +650,12 @@ static __always_inline void rcu_nmi_exit_common(bool irq)
> */
> if (rdp->dynticks_nmi_nesting != 1) {
> trace_rcu_dyntick(TPS("--="), rdp->dynticks_nmi_nesting, rdp->dynticks_nmi_nesting - 2, rdp->dynticks);
> + if (tick_nohz_full_cpu(rdp->cpu) &&
> + rdp->dynticks_nmi_nesting == 2 &&
> + rdp->rcu_urgent_qs && !rdp->rcu_forced_tick) {
> + rdp->rcu_forced_tick = true;
> + tick_dep_set_cpu(rdp->cpu, TICK_DEP_MASK_RCU);
I understand rdp->cpu is always smp_processor_id() here, right? Because calling
tick_dep_set_cpu() to a remote CPU while in NMI wouldn't be safe. It would warn anyway.
next prev parent reply other threads:[~2019-10-03 14:50 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-10-03 1:38 [PATCH tip/core/rcu 0/12] NO_HZ fixes for v5.5 Paul E. McKenney
2019-10-03 1:38 ` [PATCH tip/core/rcu 01/12] nohz: Add TICK_DEP_BIT_RCU paulmck
2019-10-03 1:38 ` [PATCH tip/core/rcu 02/12] time: Export tick start/stop functions for rcutorture paulmck
2019-10-03 1:38 ` [PATCH tip/core/rcu 03/12] rcu: Force on tick when invoking lots of callbacks paulmck
2019-10-03 14:10 ` Frederic Weisbecker
2019-10-05 16:42 ` Paul E. McKenney
2019-10-03 1:38 ` [PATCH tip/core/rcu 04/12] rcutorture: Force on tick for readers and callback flooders paulmck
2019-10-03 14:14 ` Frederic Weisbecker
2019-10-05 16:52 ` Paul E. McKenney
2019-10-03 1:38 ` [PATCH tip/core/rcu 05/12] stop_machine: EXP Provide RCU quiescent state in multi_cpu_stop() paulmck
2019-10-03 1:38 ` [PATCH tip/core/rcu 06/12] rcu: Make CPU-hotplug removal operations enable tick paulmck
2019-10-03 14:34 ` Frederic Weisbecker
2019-10-05 17:17 ` Paul E. McKenney
2019-10-03 1:38 ` [PATCH tip/core/rcu 07/12] stop_machine: Use {READ,WRITE)_ONCE() for multi_cpu_stop() ->state paulmck
2019-10-03 1:38 ` [PATCH tip/core/rcu 08/12] rcu: Force tick on for nohz_full CPUs not reaching quiescent states paulmck
2019-10-03 14:50 ` Frederic Weisbecker [this message]
2019-10-05 17:21 ` Paul E. McKenney
2019-10-03 1:39 ` [PATCH tip/core/rcu 09/12] rcu: Force nohz_full tick on upon irq enter instead of exit paulmck
2019-10-03 1:39 ` [PATCH tip/core/rcu 10/12] rcu: Reset CPU hints when reporting a quiescent state paulmck
2019-10-03 1:39 ` [PATCH tip/core/rcu 11/12] rcu: Confine ->core_needs_qs accesses to the corresponding CPU paulmck
2019-10-03 1:39 ` [PATCH tip/core/rcu 12/12] rcu: Make kernel-mode nohz_full CPUs invoke the RCU core processing paulmck
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=20191003145054.GC27555@lenoir \
--to=frederic@kernel.org \
--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=joel@joelfernandes.org \
--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@kernel.org \
--cc=paulmck@linux.ibm.com \
--cc=peterz@infradead.org \
--cc=rcu@vger.kernel.org \
--cc=rostedt@goodmis.org \
--cc=tglx@linutronix.de \
/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).