[PATCH] ethtool: Use kcalloc instead of kmalloc for ethtool_get_strings
It seems that kernel memory can leak into userspace by a
kmalloc, ethtool_get_strings, then copy_to_user sequence.
Avoid this by using kcalloc to zero fill the copied buffer.
Signed-off-by: Joe Perches
---
stable too...
On Tue, 2015-10-13 at 23:59 -0700, Jeff Kirsher wrote:
> From: Jacob Keller
[]
> diff --git a/drivers/net/ethernet/intel/fm10k/fm10k_ethtool.c
> b/drivers/net/ethernet/intel/fm10k/fm10k_ethtool.c
[]
> @@ -206,13 +206,13 @@ static void fm10k_get_stat_strings(struct net_device
> *dev, u8 *data)
> }
>
> for (i = 0; i < interface->hw.mac.max_queues; i++) {
> - sprintf(p, "tx_queue_%u_packets", i);
> + snprintf(p, ETH_GSTRING_LEN, "tx_queue_%u_packets", i);
It seems these need a memset after the snprintf to zero fill
bytes after the string terminating \0 to avoid leaking
contents of any unset bytes.
It'd probably be better to allocate a zeroed buffer instead.
> p += ETH_GSTRING_LEN;
> - sprintf(p, "tx_queue_%u_bytes", i);
> + snprintf(p, ETH_GSTRING_LEN, "tx_queue_%u_bytes", i);
so...
net/core/ethtool.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/core/ethtool.c b/net/core/ethtool.c
index b495ab1..29edf74 100644
--- a/net/core/ethtool.c
+++ b/net/core/ethtool.c
@@ -1284,7 +1284,7 @@ static int ethtool_get_strings(struct net_device *dev,
void __user *useraddr)
gstrings.len = ret;
- data = kmalloc(gstrings.len * ETH_GSTRING_LEN, GFP_USER);
+ data = kcalloc(gstrings.len, ETH_GSTRING_LEN, GFP_USER);
if (!data)
return -ENOMEM;
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] ethtool: Use kcalloc instead of kmalloc for ethtool_get_strings
On Wed, 2015-10-14 at 01:09 -0700, Joe Perches wrote:
> It seems that kernel memory can leak into userspace by a
> kmalloc, ethtool_get_strings, then copy_to_user sequence.
>
> Avoid this by using kcalloc to zero fill the copied buffer.
>
> Signed-off-by: Joe Perches
> ---
>
> stable too...
>
> On Tue, 2015-10-13 at 23:59 -0700, Jeff Kirsher wrote:
> > From: Jacob Keller
> []
> > diff --git a/drivers/net/ethernet/intel/fm10k/fm10k_ethtool.c
> > b/drivers/net/ethernet/intel/fm10k/fm10k_ethtool.c
> []
> > @@ -206,13 +206,13 @@ static void fm10k_get_stat_strings(struct net_device
> > *dev, u8 *data)
> > > >> > }
> >
> > > >> > for (i = 0; i < interface->hw.mac.max_queues; i++) {
> > -> >> > > > sprintf(p, "tx_queue_%u_packets", i);
> > +> >> > > > snprintf(p, ETH_GSTRING_LEN, "tx_queue_%u_packets",
> > i);
>
> It seems these need a memset after the snprintf to zero fill
> bytes after the string terminating \0 to avoid leaking
> contents of any unset bytes.
Right. It used to be that all drivers were memcpy()ing from a static
array which had all the necessary zero bytes, but now there are a bunch
of them using s{,n}printf() or otherwise dynamically generating names
for statistics or tests. And I don't think there's any snprintf()-
alike function that will fix that.
At least these drivers aren't zero-padding all strings: bnx2x, bnad,
i40e, i40evf, igb, ixgbe, liquidio, mlx4_en, mlx5e, nicvf, qlcnic, sfc,
vxge.
Acked-by: Ben Hutchings
Ben.
> It'd probably be better to allocate a zeroed buffer instead.
>
> > p += ETH_GSTRING_LEN;
> > - sprintf(p, "tx_queue_%u_bytes", i);
> > +> >> > > > snprintf(p, ETH_GSTRING_LEN, "tx_queue_%u_bytes",
> > i);
>
> so...
>
> net/core/ethtool.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/net/core/ethtool.c b/net/core/ethtool.c
> index b495ab1..29edf74 100644
> --- a/net/core/ethtool.c
> +++ b/net/core/ethtool.c
> @@ -1284,7 +1284,7 @@ static int ethtool_get_strings(struct net_device *dev,
> void __user *useraddr)
>
> gstrings.len = ret;
>
> - data = kmalloc(gstrings.len * ETH_GSTRING_LEN, GFP_USER);
> + data = kcalloc(gstrings.len, ETH_GSTRING_LEN, GFP_USER);
> if (!data)
> return -ENOMEM;
>
>
>
--
Ben Hutchings
[W]e found...that it wasn't as easy to get programs right as we had thought.
... I realized that a large part of my life from then on was going to be spent
in finding mistakes in my own programs. - Maurice Wilkes, 1949
signature.asc
Description: This is a digitally signed message part
Re: [PATCH] ethtool: Use kcalloc instead of kmalloc for ethtool_get_strings
From: Joe Perches Date: Wed, 14 Oct 2015 01:09:40 -0700 > It seems that kernel memory can leak into userspace by a > kmalloc, ethtool_get_strings, then copy_to_user sequence. > > Avoid this by using kcalloc to zero fill the copied buffer. > > Signed-off-by: Joe Perches Applied and queued up for -stable, thanks Joe. -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
