linux-kernel - Re: [PATCH v4] clk: allow reentrant calls into the clk framework

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]

Date:	Wed, 27 Mar 2013 22:56:10 +0000
From:	Russell King - ARM Linux <linux@....linux.org.uk>
To:	Mike Turquette <mturquette@...aro.org>
Cc:	linux-kernel@...r.kernel.org, Ulf Hansson <ulf.hansson@...aro.org>,
	linaro-kernel@...ts.linaro.org, patches@...aro.org,
	Laurent Pinchart <laurent.pinchart@...asonboard.com>,
	Rajagopal Venkat <rajagopal.venkat@...aro.org>,
	David Brown <davidb@...eaurora.org>,
	linux-arm-kernel@...ts.infradead.org
Subject: Re: [PATCH v4] clk: allow reentrant calls into the clk framework

On Wed, Mar 27, 2013 at 12:09:43AM -0700, Mike Turquette wrote:
> +	/* set context for any reentrant calls */
> +	atomic_set(&prepare_context, (int) get_current());
...
> +	if (mutex_is_locked(&prepare_lock))
> +		if ((void *) atomic_read(&prepare_context) == get_current())
> +			return true;

If you really want to do it like this, then the _correct_ way to do it
is:

		if (atomic_read(&prepare_context) == (int)get_current())

So that any effects from the cast are the same in both parts.  Otherwise,
you will be running into the possibility that a cast could do something
like truncate the returned value, resulting in the test condition using
pointers always being false.

It's not that much of a problem on ARM, but it's a matter of good
programming discpline that such issues are avoided.

Even better would be to cast it to unsigned long - this is the value
which the atomic types use, and that is less likely to be truncated.
It also helps to avoid the possibility of compilers complaining about
a narrowing cast too - especially as the entire Linux kernel assumes
that pointers can be cast to unsigned long and back again with no loss.
--
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/