[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <87zg2hcufn.ffs@tglx>
Date: Wed, 23 Aug 2023 23:36:44 +0200
From: Thomas Gleixner <tglx@...utronix.de>
To: Z qiang <qiang.zhang1211@...il.com>,
Huacai Chen <chenhuacai@...nel.org>
Cc: paulmck@...nel.org, Huacai Chen <chenhuacai@...ngson.cn>,
Frederic Weisbecker <frederic@...nel.org>,
Neeraj Upadhyay <quic_neeraju@...cinc.com>,
Joel Fernandes <joel@...lfernandes.org>,
Josh Triplett <josh@...htriplett.org>,
Boqun Feng <boqun.feng@...il.com>,
Ingo Molnar <mingo@...nel.org>,
John Stultz <jstultz@...gle.com>,
Stephen Boyd <sboyd@...nel.org>,
Steven Rostedt <rostedt@...dmis.org>,
Mathieu Desnoyers <mathieu.desnoyers@...icios.com>,
Lai Jiangshan <jiangshanlai@...il.com>,
Sergey Senozhatsky <senozhatsky@...omium.org>,
rcu@...r.kernel.org, linux-kernel@...r.kernel.org,
stable@...r.kernel.org, Binbin Zhou <zhoubinbin@...ngson.cn>
Subject: Re: [PATCH V4 2/2] rcu: Update jiffies in rcu_cpu_stall_reset()
On Wed, Aug 16 2023 at 18:06, Z. qiang wrote:
>> On Wed, Aug 16, 2023 at 1:09 PM Z qiang <qiang.zhang1211@...il.com> wrote:
>> > You can replace rcu_state.jiffies_stall update by setting rcu_cpu_stall_suppress
>> > in rcu_cpu_stall_reset(), and reset rcu_cpu_stall_suppress in rcu_gp_init() and
>> > rcu_gp_cleanup().
>> What's the advantage compared with updating jiffies? Updating jiffies
>> seems more straight forward.
>>
>
> In do_update_jiffies_64(), need to acquire jiffies_lock raw spinlock,
> like you said, kgdb is not the only caller of rcu_cpu_stall_reset(),
> the rcu_cpu_stall_reset() maybe invoke in NMI (arch/x86/platform/uv/uv_nmi.c)
What's worse is that KGDB can set breakpoints pretty much everywhere and
there is no guarantee that the jiffies lock is not held when a
breakpoint hits.
Thanks,
tglx
Powered by blists - more mailing lists