[PATCH] macvtap: fix TUNSETSNDBUF values > 64k

[PATCH] macvtap: fix TUNSETSNDBUF values > 64k

[Michael S. Tsirkin]

Upon TUNSETSNDBUF,  macvtap reads the requested sndbuf size into
a local variable u.
commit 39ec7de7092b ("macvtap: fix uninitialized access on
TUNSETIFF") changed its type to u16 (which is the right thing to
do for all other macvtap ioctls), breaking all values > 64k.

The value of TUNSETSNDBUF is actually a signed 32 bit integer, so
the right thing to do is to read it into an int.

Cc: David S. Miller <davem@davemloft.net>
Fixes: 39ec7de7092b ("macvtap: fix uninitialized access on TUNSETIFF")
Reported-by: Mark A. Peloquin
Bisected-by: Matthew Rosato <mjrosato@linux.vnet.ibm.com>
Reported-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---

This patch probably makes sense on stable.

 drivers/net/macvtap.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
index edd7734..248478c 100644
--- a/drivers/net/macvtap.c
+++ b/drivers/net/macvtap.c
@@ -1111,10 +1111,10 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
 		return 0;
 
 	case TUNSETSNDBUF:
-		if (get_user(u, up))
+		if (get_user(s, sp))
 			return -EFAULT;
 
-		q->sk.sk_sndbuf = u;
+		q->sk.sk_sndbuf = s;
 		return 0;
 
 	case TUNGETVNETHDRSZ:
-- 
MST

Re: [PATCH] macvtap: fix TUNSETSNDBUF values > 64k

[Christian Borntraeger]

Am 18.09.2015 um 12:41 schrieb Michael S. Tsirkin:
> Upon TUNSETSNDBUF,  macvtap reads the requested sndbuf size into
> a local variable u.
> commit 39ec7de7092b ("macvtap: fix uninitialized access on
> TUNSETIFF") changed its type to u16 (which is the right thing to
> do for all other macvtap ioctls), breaking all values > 64k.
> 
> The value of TUNSETSNDBUF is actually a signed 32 bit integer, so
> the right thing to do is to read it into an int.
> 
> Cc: David S. Miller <davem@davemloft.net>
> Fixes: 39ec7de7092b ("macvtap: fix uninitialized access on TUNSETIFF")
> Reported-by: Mark A. Peloquin
> Bisected-by: Matthew Rosato <mjrosato@linux.vnet.ibm.com>
> Reported-by: Christian Borntraeger <borntraeger@de.ibm.com>
> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

You can add
Tested-by:  Matthew Rosato <mjrosato@linux.vnet.ibm.com>
as this looks identical to an early version of my patch which was tested,
by Matt. (I send you the other version that changes back u as I felt that
u and up are named to identify unsigned)

and please add
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>

what about
CC: stable@vger.kernel.org

Christian

> ---
> 
> This patch probably makes sense on stable.
> 
>  drivers/net/macvtap.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
> index edd7734..248478c 100644
> --- a/drivers/net/macvtap.c
> +++ b/drivers/net/macvtap.c
> @@ -1111,10 +1111,10 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
>  		return 0;
> 
>  	case TUNSETSNDBUF:
> -		if (get_user(u, up))
> +		if (get_user(s, sp))
>  			return -EFAULT;
> 
> -		q->sk.sk_sndbuf = u;
> +		q->sk.sk_sndbuf = s;
>  		return 0;
> 
>  	case TUNGETVNETHDRSZ:
> 

Re: [PATCH] macvtap: fix TUNSETSNDBUF values > 64k

[Michael S. Tsirkin]

On Fri, Sep 18, 2015 at 12:49:07PM +0200, Christian Borntraeger wrote:
> Am 18.09.2015 um 12:41 schrieb Michael S. Tsirkin:
> > Upon TUNSETSNDBUF,  macvtap reads the requested sndbuf size into
> > a local variable u.
> > commit 39ec7de7092b ("macvtap: fix uninitialized access on
> > TUNSETIFF") changed its type to u16 (which is the right thing to
> > do for all other macvtap ioctls), breaking all values > 64k.
> > 
> > The value of TUNSETSNDBUF is actually a signed 32 bit integer, so
> > the right thing to do is to read it into an int.
> > 
> > Cc: David S. Miller <davem@davemloft.net>
> > Fixes: 39ec7de7092b ("macvtap: fix uninitialized access on TUNSETIFF")
> > Reported-by: Mark A. Peloquin
> > Bisected-by: Matthew Rosato <mjrosato@linux.vnet.ibm.com>
> > Reported-by: Christian Borntraeger <borntraeger@de.ibm.com>
> > Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> 
> You can add
> Tested-by:  Matthew Rosato <mjrosato@linux.vnet.ibm.com>
> as this looks identical to an early version of my patch which was tested,
> by Matt. (I send you the other version that changes back u as I felt that
> u and up are named to identify unsigned)
> 
> and please add
> Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
> 
> what about
> CC: stable@vger.kernel.org
> 
> Christian

David Miller adds that himself for netdev patches.

> > ---
> > 
> > This patch probably makes sense on stable.
> > 
> >  drivers/net/macvtap.c | 4 ++--
> >  1 file changed, 2 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
> > index edd7734..248478c 100644
> > --- a/drivers/net/macvtap.c
> > +++ b/drivers/net/macvtap.c
> > @@ -1111,10 +1111,10 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
> >  		return 0;
> > 
> >  	case TUNSETSNDBUF:
> > -		if (get_user(u, up))
> > +		if (get_user(s, sp))
> >  			return -EFAULT;
> > 
> > -		q->sk.sk_sndbuf = u;
> > +		q->sk.sk_sndbuf = s;
> >  		return 0;
> > 
> >  	case TUNGETVNETHDRSZ:
> > 

Re: [PATCH] macvtap: fix TUNSETSNDBUF values > 64k

[David Miller]

From: "Michael S. Tsirkin" <mst@redhat.com>
Date: Fri, 18 Sep 2015 13:41:09 +0300

> Upon TUNSETSNDBUF,  macvtap reads the requested sndbuf size into
> a local variable u.
> commit 39ec7de7092b ("macvtap: fix uninitialized access on
> TUNSETIFF") changed its type to u16 (which is the right thing to
> do for all other macvtap ioctls), breaking all values > 64k.
> 
> The value of TUNSETSNDBUF is actually a signed 32 bit integer, so
> the right thing to do is to read it into an int.
> 
> Cc: David S. Miller <davem@davemloft.net>
> Fixes: 39ec7de7092b ("macvtap: fix uninitialized access on TUNSETIFF")
> Reported-by: Mark A. Peloquin
> Bisected-by: Matthew Rosato <mjrosato@linux.vnet.ibm.com>
> Reported-by: Christian Borntraeger <borntraeger@de.ibm.com>
> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

Applied and queued up for -stable, thanks.