linux-kernel - Re: [PATCH v8 11/17] zsmalloc: make zspage lock preemptible

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: <nfkblmczqmfizey3fmd5566xthbrbg3dpsdu7ldsoddnwlqnmv@ef7lqqpa66bg>
Date: Tue, 25 Feb 2025 13:28:07 +0900
From: Sergey Senozhatsky <senozhatsky@...omium.org>
To: Sebastian Andrzej Siewior <bigeasy@...utronix.de>
Cc: Sergey Senozhatsky <senozhatsky@...omium.org>, 
	Andrew Morton <akpm@...ux-foundation.org>, Yosry Ahmed <yosry.ahmed@...ux.dev>, 
	Hillf Danton <hdanton@...a.com>, Kairui Song <ryncsn@...il.com>, Minchan Kim <minchan@...nel.org>, 
	linux-mm@...ck.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v8 11/17] zsmalloc: make zspage lock preemptible

On (25/02/24 09:59), Sebastian Andrzej Siewior wrote:
> > +/*
> > + * The zspage lock can be held from atomic contexts, but it needs to remain
> > + * preemptible when held for reading because it remains held outside of those
> > + * atomic contexts, otherwise we unnecessarily lose preemptibility.
> > + *
> > + * To achieve this, the following rules are enforced on readers and writers:
> > + *
> > + * - Writers are blocked by both writers and readers, while readers are only
> > + *   blocked by writers (i.e. normal rwlock semantics).
> > + *
> > + * - Writers are always atomic (to allow readers to spin waiting for them).
> > + *
> > + * - Writers always use trylock (as the lock may be held be sleeping readers).
> > + *
> > + * - Readers may spin on the lock (as they can only wait for atomic writers).
> > + *
> > + * - Readers may sleep while holding the lock (as writes only use trylock).
> > + */
> > +static void zspage_read_lock(struct zspage *zspage)
> > +{
> > +	struct zspage_lock *zsl = &zspage->zsl;
> > +
> > +	rwsem_acquire_read(zsl_dep_map(zsl), 0, 0, _RET_IP_);
> > +
> > +	spin_lock(&zsl->lock);
> > +	zsl->cnt++;
> > +	spin_unlock(&zsl->lock);
> 
> How is this working given that a read_lock always increments the
> counter? If it is write_locked then a read_lock makes it UNLOCKED.

If zspage is write-locked then zsl->lock is also locked, because
write-lock returns with zsl->lock acquired and releases it in
write-unlock.