Date: Thu, 3 Jun 2010 12:05:33 +0100
From: Russell King - ARM Linux <linux@arm.linux.org.uk>
To: Jeremy Kerr <jeremy.kerr@canonical.com>
Cc: Ben Dooks <ben-linux@fluff.org>,
       Ben Herrenchmidt <benh@kernel.crashing.org>,
       linux-kernel@vger.kernel.org, linux-arm-kernel@lists.infradead.org
Subject: Re: [RFC,PATCH 1/2] Add a common struct clk
Message-ID: <20100603110533.GB7127@n2100.arm.linux.org.uk>
References: <1275479804.137633.565764505843.0.gpush@pororo> <201006031121.21896.jeremy.kerr@canonical.com> <20100603081354.GA4720@trinity.fluff.org> <201006031824.53832.jeremy.kerr@canonical.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <201006031824.53832.jeremy.kerr@canonical.com>
User-Agent: Mutt/1.5.19 (2009-01-05)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1140
Lines: 33

On Thu, Jun 03, 2010 at 06:24:50PM +0800, Jeremy Kerr wrote:
> OK, this would mean adding parent to struct clk:
> 
> struct clk {
> 	struct clk_operations	*ops;
> 	atomic_t				enable_count;

I don't think it makes sense for this to be an atomic type.

> static inline int clk_enable(struct clk *clk)
> {
> 	int ret;
> 
> 	if (atomic_inc(clk->enable_count) != 1))
> 		return 0;

Okay, so what if we sleep at this point, and someone else comes along
and calls clk_enable(), which'll merely increment the count and return.

Unfortunately, the clock is still disabled at that point - which is a
violation of what's supposed to happen.

Or to put it another way, the above method results in clk_enable()
sometimes returning with the clock enabled and sometimes with the
clock still disabled.

That's not nice behaviour for drivers which may need the clock enabled
to read/write the device registers.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/