linux-kernel - Re: [PATCH 1/3] ARM: Introduce atomic MMIO clear/set

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: <20130812182942.GA28695@mudshark.cambridge.arm.com>
Date:	Mon, 12 Aug 2013 19:29:42 +0100
From:	Will Deacon <will.deacon@....com>
To:	Ezequiel Garcia <ezequiel.garcia@...e-electrons.com>
Cc:	"linux-arm-kernel@...ts.infradead.org" 
	<linux-arm-kernel@...ts.infradead.org>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	Lior Amsalem <alior@...vell.com>,
	Thomas Petazzoni <thomas.petazzoni@...e-electrons.com>,
	Russell King <linux@....linux.org.uk>,
	Jason Cooper <jason@...edaemon.net>,
	Andrew Lunn <andrew@...n.ch>,
	Gregory Clement <gregory.clement@...e-electrons.com>,
	Sebastian Hesselbarth <sebastian.hesselbarth@...il.com>
Subject: Re: [PATCH 1/3] ARM: Introduce atomic MMIO clear/set

On Sat, Aug 10, 2013 at 01:43:00PM +0100, Ezequiel Garcia wrote:
> Some SoC have MMIO regions that are shared across orthogonal
> subsystems. This commit implements a possible solution for the
> thread-safe access of such regions through a spinlock-protected API
> with clear-set semantics.
> 
> Concurrent access is protected with a single spinlock for the
> entire MMIO address space. While this protects shared-registers,
> it also serializes access to unrelated/unshared registers.
> 
> Signed-off-by: Ezequiel Garcia <ezequiel.garcia@...e-electrons.com>

[...]

> +void atomic_io_clear_set(void __iomem *reg, u32 clear, u32 set)
> +{
> +	spin_lock(&__io_lock);
> +	writel((readl(reg) & ~clear) | set, reg);
> +	spin_unlock(&__io_lock);
> +}

I appreciate that you've lifted this code from a previous driver, but this
doesn't really make any sense to me. The spin_unlock operation is
essentially a store to normal, cacheable memory, whilst the writel is an
__iowmb followed by a store to device memory.

This means that you don't have ordering guarantees between the two accesses
outside of the CPU, potentially giving you:

	spin_lock(&__io_lock);
	spin_unlock(&__io_lock);
	writel((readl(reg) & ~clear) | set, reg);

which is probably not what you want.

I suggest adding an iowmb after the writel if you really need this ordering
to be enforced (but this may have a significant performance impact,
depending on your SoC).

Will
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/