linux-kernel - Re: [PATCH RFC] PPC-BRIQ_PANEL: Remove BKL and replace with atomic variable.

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:	Tue, 20 Oct 2009 00:01:31 +0200
From:	Frederic Weisbecker <fweisbec@...il.com>
To:	Thomas Gleixner <tglx@...utronix.de>
Cc:	John Kacur <jkacur@...hat.com>, linux-kernel@...r.kernel.org,
	Alan Cox <alan@...rguk.ukuu.org.uk>,
	Arnd Bergmann <arndbergmann@...glemail.com>,
	Ingo Molnar <mingo@...e.hu>
Subject: Re: [PATCH RFC] PPC-BRIQ_PANEL: Remove BKL and replace with atomic
	variable.

On Mon, Oct 19, 2009 at 07:04:02AM +0200, Thomas Gleixner wrote:
> B1;2005;0cOn Sun, 18 Oct 2009, John Kacur wrote:
> 
> > >From b64c7d0f11eab96cb253b23c7264c999746116c0 Mon Sep 17 00:00:00 2001
> > From: John Kacur <jkacur@...hat.com>
> > Date: Sun, 18 Oct 2009 21:29:21 +0200
> > Subject: [PATCH] PPC-BRIQ_PANEL: Remove BKL and replace with atomic variable.
> > 
> > There are no locks here except the bkl in briq_panel_open. It's only
> > purpose is to ensure single access. Remove the bkl and ensure single access
> > by making vfd_is_open an atomic_variable.
> 
> And again, can you please look more carefully at the init
> vs. read/write functions ?
> 
> The BKL is not only protecting the single user variable it's also
> serializing write against the access to the display in init.
> 
> Thanks,
> 
> 	tglx


That could be solved by statically initializing vfd_is_open to -1
and then set it to 0 once briq_panel_init has finished initializing
the device.

Another thing, I really don't see the point in this check in
briq_panel_read() and briq_panel_write():

	if (!vfd_is_open)
		return -ENODEV;


You can't read/write if vfd_is_open hasn't been set to 1 (open set)
and you're not racing against the release callback since it is called
after the file is closed.

I guess this check can disappear from read/write callbacks.

--
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/