linux-kernel - Re: [PATCH v2] srcu/tiny: Remove preempt_disable/enable() in srcu_gp_start_if

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]

Message-ID: <430d5714-4a00-4a75-b763-34918f8017c3@paulmck-laptop>
Date: Thu, 11 Sep 2025 04:37:39 -0700
From: "Paul E. McKenney" <paulmck@...nel.org>
To: Zqiang <qiang.zhang@...ux.dev>
Cc: Joel Fernandes <joelagnelf@...dia.com>, frederic@...nel.org,
	neeraj.upadhyay@...nel.org, boqun.feng@...il.com, urezki@...il.com,
	rcu@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] srcu/tiny: Remove preempt_disable/enable() in
 srcu_gp_start_if_needed()

On Thu, Sep 11, 2025 at 12:36:45AM +0000, Zqiang wrote:
> > 
> > On Wed, Sep 10, 2025 at 10:36:20AM -0400, Joel Fernandes wrote:
> > 
> > > 
> > > [..]
> > >  > kernel/rcu/srcutiny.c | 4 +---
> > >  > 1 file changed, 1 insertion(+), 3 deletions(-)
> > >  > 
> > >  > diff --git a/kernel/rcu/srcutiny.c b/kernel/rcu/srcutiny.c
> > >  > index b52ec45698e8..b2da188133fc 100644
> > >  > --- a/kernel/rcu/srcutiny.c
> > >  > +++ b/kernel/rcu/srcutiny.c
> > >  > @@ -181,10 +181,9 @@ static void srcu_gp_start_if_needed(struct srcu_struct *ssp)
> > >  > {
> > >  > unsigned long cookie;
> > >  > 
> > >  > - preempt_disable(); // Needed for PREEMPT_LAZY
> > >  > + lockdep_assert_preemption_disabled();
> > >  
> > >  nit: Do we still want to keep the comment that the expectation of preemption
> > >  being disabled is for the LAZY case?
> > > 
> > Good point, and I do believe that we do. Zqiang, any reason not to
> > add this comment back in?
> 
> in rcu-tree, this commit:
> 
> (935147775c977 "EXP srcu: Enable Tiny SRCU On all CONFIG_SMP=n kernels")
> 
> make preempt disable needed for CONFIG_PREEMPT=y or CONFIG_PREEMPT_LAZY=y
> when the CONFIG_SMP=n. do we need to replace "Needed for PREEMPT_LAZY"
> comments with "Needed for PREEMPT or PREEMPT_LAZY"?

Good point as well, thank you!  And I need to decide whether I should
send that patch upstream.  Its original purpose was to test PREEMPT_LAZY=y
better than could be tested with PREEMPT_LAZY.

Thoughts?

							Thanx, Paul

> Thanks
> Zqiang
> 
> 
> > 
> >  Thanx, Paul
> > 
> > > 
> > > thanks,
> > >  
> > >  - Joel
> > >  
> > >  
> > >  > cookie = get_state_synchronize_srcu(ssp);
> > >  > if (ULONG_CMP_GE(READ_ONCE(ssp->srcu_idx_max), cookie)) {
> > >  > - preempt_enable();
> > >  > return;
> > >  > }
> > >  > WRITE_ONCE(ssp->srcu_idx_max, cookie);
> > >  > @@ -194,7 +193,6 @@ static void srcu_gp_start_if_needed(struct srcu_struct *ssp)
> > >  > else if (list_empty(&ssp->srcu_work.entry))
> > >  > list_add(&ssp->srcu_work.entry, &srcu_boot_list);
> > >  > }
> > >  > - preempt_enable();
> > >  > }
> > >  > 
> > >  > /*
> > >  > -- 
> > >  > 2.48.1
> > >  >
> > >
> >