linux-kernel - Re: [PATCH] Export GPIO control symbols from mfp-pxa2xx.c

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] [day] [month] [year] [list]

Date:	Mon, 11 May 2009 16:08:44 +0100 (BST)
From:	Michael Abbott <michael@...neidae.co.uk>
To:	Eric Miao <eric.y.miao@...il.com>
cc:	Christoph Hellwig <hch@...radead.org>,
	linux-arm-kernel@...ts.arm.linux.org.uk,
	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] Export GPIO control symbols from mfp-pxa2xx.c

On Mon, 11 May 2009, Eric Miao wrote:
> On Mon, May 11, 2009 at 3:28 PM, Michael Abbott <michael@...neidae.co.uk> wrote:
> > On Mon, 11 May 2009, Christoph Hellwig wrote:
> >> On Mon, May 11, 2009 at 07:43:29AM +0100, Michael Abbott wrote:
> >> > With very lightweight board support providing a number of 
> >> > uncommitted GPIO pins I found that my out of kernel driver needs to 
> >> > call symbols that aren't actually exported.  This patch adds these 
> >> > exports.
> >> As usual we don't export symbols for out of tree drivers.
> >
> > I understand that's the default condition, but presumably there's no 
> > compelling reason for the mfp configuration functions to be withheld 
> > (I was simply trying to obey the obsolescence instructions in 
> > pxa2xx-gpio.h!)
> >
> > As for bringing this driver in tree?  Not a great deal of point, in 
> > this particular case, but I have no problem with doing that.
> 
> I don't see many real requirements here to export them for the moment. 
> By keeping these functions/symbols not-exported, I'd expect 
> board-specific code (usually manipulation of these pins are 
> board-specific) to be kept in the "board.c" as well, which I do think is 
> a good practice.

That's kind of fair, but in the case where the board is a CPU module 
rather than an entire system (which applies both for my XCEP module, the 
Colibri modules, and others I'm sure) this isn't necessarily so 
straightforward.

> Since your driver is out of the tree, you can certainly keep this change 
> out of the tree as well :-)

Heh.  There should be an emoticon for sticking my tongue out ;^).  Ok, I 
can live with that for the time being.