linux-kernel - Re: [PATCH] sysfs: device-core: sysfs open close notify

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:	Thu, 4 Nov 2010 11:27:58 -0700
From:	Greg KH <gregkh@...e.de>
To:	Onkalo Samu <samu.p.onkalo@...ia.com>
Cc:	"hmh@....eng.br" <hmh@....eng.br>,
	"alan@...ux.intel.com" <alan@...ux.intel.com>,
	"akpm@...ux-foundation.org" <akpm@...ux-foundation.org>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] sysfs: device-core: sysfs open close notify

On Thu, Nov 04, 2010 at 06:37:36PM +0200, Onkalo Samu wrote:
> On Thu, 2010-11-04 at 17:03 +0100, ext Greg KH wrote:
> > On Thu, Nov 04, 2010 at 02:32:15PM +0100, samu.p.onkalo@...ia.com wrote:
> > > It is easy to get rid of if the mode parameter is used to pass the information
> > > that this entry uses open_close_notify.  What do you think, is it ok to use
> > > mode also to that purpose?
> > 
> > Don't try to overload a parameter that has been used for the past 40+
> > years in one way, to try to add additional side-band data that has
> > nothing to do with it.
> > 
> > That way lies madness.
> > 
> 
> And that is why I didn't even tried to do that in the first place - even
> if it would have been the simple way.
> 
> Is the implementation ok otherwise?
> 
> I'll add sysfs_create_file_notify which sets the control bit save way.
> I think it is enough if these entries can be done attribute by
> attribute. It is still possible delete them using normal sysfs
> operations.

No, don't do a separate file create because no one should ever create a
file on its own.  It should only be done by bus through a default
attribute, or, in extreem cases, done by purposefully controlling the
kobject uevents where you know exactly what you are doing.

Odds are, any user of this call wouldn't know exactly what they are
doing, so don't give them that opportunity to mess it up.

thanks,

greg k-h
--
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/