linux-kernel - Re: [PATCH] regulator: vctrl: Fix out of bounds array access for vctrl->vtable

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:   Fri, 14 Apr 2017 12:01:06 -0700
From:   Matthias Kaehlcke <mka@...omium.org>
To:     Axel Lin <axel.lin@...ics.com>
Cc:     Mark Brown <broonie@...nel.org>, Rob Herring <robh@...nel.org>,
        Liam Girdwood <lgirdwood@...il.com>,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH] regulator: vctrl: Fix out of bounds array access for
 vctrl->vtable

El Fri, Apr 14, 2017 at 10:11:37AM -0700 Matthias Kaehlcke ha dit:

> Hi Axel,
> 
> El Fri, Apr 14, 2017 at 10:50:43AM +0800 Axel Lin ha dit:
> 
> > Current code only allocates rdesc->n_voltages entries for vctrl->vtable.
> > Thus use rdesc->n_voltages instead of n_voltages in the for loop.
> 
> This is intended. n_voltages is the number of voltages of the control
> regulator. In the first loop of the function we determine how many of
> these voltages are usable by the vctrl regulator (=> rdesc->n_voltages).
> The loop that populates vctrl->vtable iterates over n_voltages,
> however it skips those that are outside of the voltage range for the
> control regulator.

Sorry, I should have gotten coffee first. Your patch changes the
pre-calc and not the population loop, the use of n_voltages there
is indeed incorrect. Thanks for catching this!

Reviewed-by: Matthias Kaehlcke <mka@...omium.org>

> > While at it, also switch to use devm_kcalloc instead of devm_kmalloc_array
> > + __GFP_ZERO flag and fix the argument order.
> 
> This looks good to me.
> 
> Thanks
> 
> Matthias
> 
> > Signed-off-by: Axel Lin <axel.lin@...ics.com>
> > ---
> >  drivers/regulator/vctrl-regulator.c | 8 ++++----
> >  1 file changed, 4 insertions(+), 4 deletions(-)
> > 
> > diff --git a/drivers/regulator/vctrl-regulator.c b/drivers/regulator/vctrl-regulator.c
> > index 6baadef..78de002 100644
> > --- a/drivers/regulator/vctrl-regulator.c
> > +++ b/drivers/regulator/vctrl-regulator.c
> > @@ -345,9 +345,9 @@ static int vctrl_init_vtable(struct platform_device *pdev)
> >  		return -EINVAL;
> >  	}
> >  
> > -	vctrl->vtable = devm_kmalloc_array(
> > -		&pdev->dev, sizeof(struct vctrl_voltage_table),
> > -		rdesc->n_voltages, GFP_KERNEL | __GFP_ZERO);
> > +	vctrl->vtable = devm_kcalloc(&pdev->dev, rdesc->n_voltages,
> > +				     sizeof(struct vctrl_voltage_table),
> > +				     GFP_KERNEL);
> >  	if (!vctrl->vtable)
> >  		return -ENOMEM;
> >  
> > @@ -371,7 +371,7 @@ static int vctrl_init_vtable(struct platform_device *pdev)
> >  	     NULL);
> >  
> >  	/* pre-calculate OVP-safe downward transitions */
> > -	for (i = n_voltages - 1; i > 0; i--) {
> > +	for (i = rdesc->n_voltages - 1; i > 0; i--) {
> >  		int j;
> >  		int ovp_min_uV = (vctrl->vtable[i].out *
> >  				  (100 - vctrl->ovp_threshold)) / 100;