linux-kernel - Re: [PATCH] locking/percpu_rwsem: Rewrite to not use rwsem

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: <20190805154328.GJ2332@hirez.programming.kicks-ass.net>
Date:   Mon, 5 Aug 2019 17:43:28 +0200
From:   Peter Zijlstra <peterz@...radead.org>
To:     Boqun Feng <boqun.feng@...il.com>
Cc:     Oleg Nesterov <oleg@...hat.com>, Will Deacon <will@...nel.org>,
        Ingo Molnar <mingo@...nel.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        linux-kernel@...r.kernel.org, bigeasy@...utronix.de,
        juri.lelli@...hat.com, williams@...hat.com, bristot@...hat.com,
        longman@...hat.com, dave@...olabs.net, jack@...e.com
Subject: Re: [PATCH] locking/percpu_rwsem: Rewrite to not use rwsem

On Mon, Aug 05, 2019 at 10:58:13PM +0800, Boqun Feng wrote:
> On Mon, Aug 05, 2019 at 10:43:18PM +0800, Boqun Feng wrote:
> > On Mon, Aug 05, 2019 at 04:02:41PM +0200, Peter Zijlstra wrote:
> > [...]
> > >  
> > >  static inline void percpu_up_read(struct percpu_rw_semaphore *sem)
> > >  {
> > > +	rwsem_release(&sem->dep_map, 1, _RET_IP_);
> > > +
> > >  	preempt_disable();
> > >  	/*
> > >  	 * Same as in percpu_down_read().
> > >  	 */
> > > -	if (likely(rcu_sync_is_idle(&sem->rss)))
> > > +	if (likely(rcu_sync_is_idle(&sem->rss))) {
> > >  		__this_cpu_dec(*sem->read_count);
> > > -	else
> > > -		__percpu_up_read(sem); /* Unconditional memory barrier */
> > > -	preempt_enable();
> > > +		preempt_enable();
> > > +		return;
> > > +	}
> > >  
> > > -	rwsem_release(&sem->rw_sem.dep_map, 1, _RET_IP_);
> > 
> > Missing a preempt_enable() here?
> > 
> 
> Ah.. you modified the semantics of __percpu_up_read() to imply a
> preempt_enable(), sorry for the noise...

Yes indeed; I suppose I should've noted that in the Changlog. The reason
is that waitqueues use spin_lock() which change into a sleepable lock on
RT and thus cannot be used with preeption disabled. We also cannot
(easily) switch to swait because we use both exclusive and !exclusive
waits.