linux-kernel - Re: [PATCH 0/5] Pushdown bkl from v4l ioctls

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]

Message-Id: <201004290910.43412.laurent.pinchart@ideasonboard.com>
Date:	Thu, 29 Apr 2010 09:10:42 +0200
From:	Laurent Pinchart <laurent.pinchart@...asonboard.com>
To:	Hans Verkuil <hverkuil@...all.nl>
Cc:	Frederic Weisbecker <fweisbec@...il.com>,
	LKML <linux-kernel@...r.kernel.org>,
	Arnd Bergmann <arnd@...db.de>, John Kacur <jkacur@...hat.com>,
	Linus Torvalds <torvalds@...ux-foundation.org>,
	Jan Blunck <jblunck@...il.com>,
	Thomas Gleixner <tglx@...utronix.de>,
	Mauro Carvalho Chehab <mchehab@...radead.org>,
	Greg KH <gregkh@...e.de>,
	Linux Media Mailing List <linux-media@...r.kernel.org>
Subject: Re: [PATCH 0/5] Pushdown bkl from v4l ioctls

Hi Hans,

On Thursday 29 April 2010 08:44:29 Hans Verkuil wrote:
> On Thursday 29 April 2010 05:42:39 Frederic Weisbecker wrote:
> > Hi,
> > 
> > Linus suggested to rename struct v4l2_file_operations::ioctl
> > into bkl_ioctl to eventually get something greppable and make
> > its background explicit.
> > 
> > While at it I thought it could be a good idea to just pushdown
> > the bkl to every v4l drivers that have an .ioctl, so that we
> > actually remove struct v4l2_file_operations::ioctl for good.
> > 
> > It passed make allyesconfig on sparc.
> > Please tell me what you think.
> 
> I much prefer to keep the bkl inside the v4l2 core. One reason is that I
> think that we can replace the bkl in the core with a mutex. Still not
> ideal of course, so the next step will be to implement proper locking in
> each driver. For this some additional v4l infrastructure work needs to be
> done. I couldn't proceed with that until the v4l events API patches went
> in, and that happened yesterday.
> 
> So from my point of view the timeline is this:
> 
> 1) I do the infrastructure work this weekend. This will make it much easier
> to convert drivers to do proper locking. And it will also simplify
> v4l2_priority handling, so I'm killing two birds with one stone :-)
> 
> 2) Wait until Arnd's patch gets merged that pushes the bkl down to
> v4l2-dev.c
> 
> 3) Investigate what needs to be done to replace the bkl with a v4l2-dev.c
> global mutex. Those drivers that call the bkl themselves should probably be
> converted to do proper locking, but there are only about 14 drivers that do
> this. The other 60 or so drivers should work fine if a v4l2-dev global lock
> is used. At this point the bkl is effectively removed from the v4l
> subsystem.
> 
> 4) Work on the remaining 60 drivers to do proper locking and get rid of the
> v4l2-dev global lock. This is probably less work than it sounds.
> 
> Since your patch moves everything down to the driver level it will actually
> make this work harder rather than easier. And it touches almost all drivers
> as well.

Every driver will need to be carefully checked to make sure the BKL can be 
replaced by a v4l2-dev global mutex. Why would it be more difficult to do so 
if the BKL is pushed down to the drivers ?

-- 
Regards,

Laurent Pinchart
--
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/