linux-kernel - Re: [PATCH] clk.h: Fix shim ifdef guard (HAVE_CLK -> COMMON

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:	Mon, 27 Aug 2012 16:53:44 -0400
From:	Jeff Mahoney <jeffm@...e.com>
To:	Russell King <rmk@....linux.org.uk>
Cc:	Linux Kernel Maling List <linux-kernel@...r.kernel.org>,
	Viresh Kumar <viresh.kumar@...com>,
	Wolfram Sang <w.sang@...gutronix.de>,
	Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
	Jeff Garzik <jgarzik@...hat.com>, Andrew Lunn <andrew@...n.ch>,
	Bhupesh Sharma <bhupesh.sharma@...com>,
	Giuseppe Cavallaro <peppe.cavallaro@...com>,
	Mike Turquette <mturquette@...aro.org>,
	Sergei Shtylyov <sshtylyov@...mvista.com>
Subject: Re: [PATCH] clk.h: Fix shim ifdef guard (HAVE_CLK -> COMMON_CLK)

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 8/27/12 2:52 PM, Russell King wrote:
> On Mon, Aug 27, 2012 at 11:28:15AM -0400, Jeff Mahoney wrote:
>> Commit 93abe8e4 (clk: add non HAVE_CLK routines) added shims for 
>> the clk code but HAVE_CLK isn't enough. It's possible to have
>> the clk support but not enable it. We end up with full prototypes
>> for code that is never built - causing module linking to fail
>> later.
>> 
>> This patch changes the guard to use COMMON_CLK, which actually
>> guards the code.
> 
> This is wrong.  COMMON_CLK is an _implementation_ of the CLK API.
> It is not the only implementation in the kernel.  Conditionalizing
> like this breaks existing users.
> 
> HAVE_CLK is the right thing here - if you define HAVE_CLK then you
> _are_ providing an implementation of clk_get() et.al.  If you're
> not, then you do not define HAVE_CLK.  Simples.
> 

Ok. Thanks for the feedback.

It would seem, then, that the powerpc clock implementation doesn't
implement the full API and that's where we're running into problems.

Specifically, the use of clk_devm_get in
drivers/usb/chipidea/ci13xxx_imx.c and
drivers/char/hw_random/exynos-rng.ko is causing module link errors on
powerpc configs for the EFIKA hardware. The EXYNOS RNG driver should
probably have a dependency on ARCH_EXYNOS. The chipidea driver looks
like it was developed on ARM hardware - does it exist anywhere else?

Should the clk_devm stuff be moved someplace generic? I don't have a
horse in this race. My only interest here is to get these configs to
build again. :)

- -Jeff

- -- 
Jeff Mahoney
SUSE Labs
-----BEGIN PGP SIGNATURE-----
Version: GnuPG/MacGPG2 v2.0.18 (Darwin)
Comment: GPGTools - http://gpgtools.org
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
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=/UyR
-----END PGP SIGNATURE-----
--
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/