linux-kernel - Re: [PATCH] rcu: Add rnp->expmask output to show_rcu_gp

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Wed, 25 May 2022 07:03:28 -0700
From:   "Paul E. McKenney" <paulmck@...nel.org>
To:     Mukesh Ojha <quic_mojha@...cinc.com>
Cc:     Zqiang <qiang1.zhang@...el.com>, frederic@...nel.org,
        rcu@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] rcu: Add rnp->expmask output to show_rcu_gp_kthreads()

On Wed, May 25, 2022 at 10:29:23AM +0530, Mukesh Ojha wrote:
> 
> 
> On 5/25/2022 10:15 AM, Zqiang wrote:
> > Currently, for expedited rcu we only check 'rnp->exp_tasks', however
> > the 'rnp->exp_tasks' are not always exist for expedited grace period
> > that not end. this commit add 'rnp->expmask' condition.
> > 
> > Signed-off-by: Zqiang <qiang1.zhang@...el.com>
> 
> Thanks !! for the patch.
> Commit log could require wordsmith from Paul :-)

For Paul to wordsmith this commit log, he must first understand why this
change is useful, keeping in mind that show_rcu_gp_kthreads() is invoked
for normal grace periods, but not for expedited ones.

The point of ->exp_mask for debugging forward-progress issues for normal
grace periods is to understand why tasks blocking the current normal
grace period are not yet being boosted (because the expedited grace
period's boosting takes precedence).  How do these additions help you?

(Not saying that they cannot possibly help you.  I am instead saying
that you must tell me how they have helped you guys.)

							Thanx, Paul

> Reviewed-by: quic_mojha@...cinc.com
> 
> -Mukesh
> 
> > ---
> >   kernel/rcu/tree_stall.h | 6 ++++--
> >   1 file changed, 4 insertions(+), 2 deletions(-)
> > 
> > diff --git a/kernel/rcu/tree_stall.h b/kernel/rcu/tree_stall.h
> > index 2464b0eccfd0..5c4df855f2ac 100644
> > --- a/kernel/rcu/tree_stall.h
> > +++ b/kernel/rcu/tree_stall.h
> > @@ -865,13 +865,15 @@ void show_rcu_gp_kthreads(void)
> >   	rcu_for_each_node_breadth_first(rnp) {
> >   		if (ULONG_CMP_GE(READ_ONCE(rcu_state.gp_seq), READ_ONCE(rnp->gp_seq_needed)) &&
> >   		    !data_race(READ_ONCE(rnp->qsmask)) && !data_race(READ_ONCE(rnp->boost_tasks)) &&
> > -		    !data_race(READ_ONCE(rnp->exp_tasks)) && !data_race(READ_ONCE(rnp->gp_tasks)))
> > +		    !data_race(READ_ONCE(rnp->expmask)) && !data_race(READ_ONCE(rnp->exp_tasks)) &&
> > +			!data_race(READ_ONCE(rnp->gp_tasks)))
> >   			continue;
> > -		pr_info("\trcu_node %d:%d ->gp_seq %ld ->gp_seq_needed %ld ->qsmask %#lx %c%c%c%c ->n_boosts %ld\n",
> > +		pr_info("\trcu_node %d:%d ->gp_seq %ld ->gp_seq_needed %ld ->qsmask %#lx ->expmask %#lx %c%c%c%c ->n_boosts %ld\n",
> >   			rnp->grplo, rnp->grphi,
> >   			(long)data_race(READ_ONCE(rnp->gp_seq)),
> >   			(long)data_race(READ_ONCE(rnp->gp_seq_needed)),
> >   			data_race(READ_ONCE(rnp->qsmask)),
> > +			data_race(READ_ONCE(rnp->expmask)),
> >   			".b"[!!data_race(READ_ONCE(rnp->boost_kthread_task))],
> >   			".B"[!!data_race(READ_ONCE(rnp->boost_tasks))],
> >   			".E"[!!data_race(READ_ONCE(rnp->exp_tasks))],