[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <20241002145738.38226-1-frederic@kernel.org>
Date: Wed, 2 Oct 2024 16:57:35 +0200
From: Frederic Weisbecker <frederic@...nel.org>
To: LKML <linux-kernel@...r.kernel.org>
Cc: Frederic Weisbecker <frederic@...nel.org>,
Boqun Feng <boqun.feng@...il.com>,
Joel Fernandes <joel@...lfernandes.org>,
Neeraj Upadhyay <neeraj.upadhyay@....com>,
"Paul E . McKenney" <paulmck@...nel.org>,
Uladzislau Rezki <urezki@...il.com>,
Zqiang <qiang.zhang1211@...il.com>,
rcu <rcu@...r.kernel.org>
Subject: [PATCH 0/3] rcu: Fix yet another wake up from offline related issue
Hi,
A new warning has been reported due to swake_up_one_online() use
from an offline IRQ:
https://lore.kernel.org/oe-lkp/202409231644.4c55582d-lkp@intel.com
Here is a tentative fix.
Similar issue can happen with exp kthread and GP kthread if offline tick
fires and there is a pending deferred quiescent state to report on
PREEMPT_RCU. Currently only oneshot ticks are disabled while the CPU is
offlining. I have yet to take care of the periodic tick implementation.
Work in progress...
Frederic Weisbecker (3):
rcu/nocb: Use switch/case on NOCB timer state machine
rcu/nocb: Fix rcuog wake-up from offline softirq
rcu: Report callbacks enqueued on offline CPU blind spot
kernel/rcu/tree.c | 3 +++
kernel/rcu/tree.h | 1 +
kernel/rcu/tree_nocb.h | 47 +++++++++++++++++++++++++++++++-----------
3 files changed, 39 insertions(+), 12 deletions(-)
--
2.46.0
Powered by blists - more mailing lists