linux-kernel - Re: [PATCH v8 16/19] locking/rwsem: Guard against making count negative

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: <20190611131131.GG3402@hirez.programming.kicks-ass.net>
Date:   Tue, 11 Jun 2019 15:11:31 +0200
From:   Peter Zijlstra <peterz@...radead.org>
To:     Waiman Long <longman@...hat.com>
Cc:     Ingo Molnar <mingo@...hat.com>, Will Deacon <will.deacon@....com>,
        Thomas Gleixner <tglx@...utronix.de>,
        Borislav Petkov <bp@...en8.de>,
        "H. Peter Anvin" <hpa@...or.com>, linux-kernel@...r.kernel.org,
        x86@...nel.org, Davidlohr Bueso <dave@...olabs.net>,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        Tim Chen <tim.c.chen@...ux.intel.com>,
        huang ying <huang.ying.caritas@...il.com>
Subject: Re: [PATCH v8 16/19] locking/rwsem: Guard against making count
 negative

On Mon, May 20, 2019 at 04:59:15PM -0400, Waiman Long wrote:

> +static inline long rwsem_read_trylock(struct rw_semaphore *sem, long *cnt)
> +{
> +	long adjustment = -RWSEM_READER_BIAS;
> +
> +	*cnt = atomic_long_fetch_add_acquire(RWSEM_READER_BIAS, &sem->count);

I'm thinking we'd actually want add_return_acquire() here.

> +	if (unlikely(*cnt < 0)) {
> +		atomic_long_add(-RWSEM_READER_BIAS, &sem->count);
> +		adjustment = 0;
> +	}
> +	return adjustment;
> +}

> @@ -1271,9 +1332,10 @@ static struct rw_semaphore *rwsem_downgrade_wake(struct rw_semaphore *sem)
>   */
>  inline void __down_read(struct rw_semaphore *sem)
>  {
> +	long tmp, adjustment = rwsem_read_trylock(sem, &tmp);
> +
> +	if (unlikely(tmp & RWSEM_READ_FAILED_MASK)) {
> +		rwsem_down_read_slowpath(sem, TASK_UNINTERRUPTIBLE, adjustment);
>  		DEBUG_RWSEMS_WARN_ON(!is_rwsem_reader_owned(sem), sem);
>  	} else {
>  		rwsem_set_reader_owned(sem);
> @@ -1282,9 +1344,11 @@ inline void __down_read(struct rw_semaphore *sem)
>  
>  static inline int __down_read_killable(struct rw_semaphore *sem)
>  {
> +	long tmp, adjustment = rwsem_read_trylock(sem, &tmp);
> +
> +	if (unlikely(tmp & RWSEM_READ_FAILED_MASK)) {
> +		if (IS_ERR(rwsem_down_read_slowpath(sem, TASK_KILLABLE,
> +						    adjustment)))
>  			return -EINTR;
>  		DEBUG_RWSEMS_WARN_ON(!is_rwsem_reader_owned(sem), sem);
>  	} else {

I'm confused by the need for @tmp; isn't that returning the exact same
state !adjustment is?

Also; half the patch seems to do cnt<0, while the other half (above)
does &READ_FAILED, what gives?