2024-02-16 23:31:44

by Kees Cook

[permalink] [raw]
Subject: [PATCH] netfilter: x_tables: Use unsafe_memcpy() for 0-sized destination

The struct xt_entry_target fake flexible array has not be converted to a
true flexible array, which is mainly blocked by it being both UAPI and
used in the middle of other structures. In order to properly check for
0-sized destinations in memcpy(), an exception must be made for the one
place where it is still a destination. Since memcpy() was already
skipping checks for 0-sized destinations, using unsafe_memcpy() is no
change in behavior.

Signed-off-by: Kees Cook <[email protected]>
---
Cc: Pablo Neira Ayuso <[email protected]>
Cc: Jozsef Kadlecsik <[email protected]>
Cc: Florian Westphal <[email protected]>
Cc: "David S. Miller" <[email protected]>
Cc: Eric Dumazet <[email protected]>
Cc: Jakub Kicinski <[email protected]>
Cc: Paolo Abeni <[email protected]>
Cc: Gustavo A. R. Silva <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
---
net/netfilter/x_tables.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/net/netfilter/x_tables.c b/net/netfilter/x_tables.c
index 21624d68314f..da5d929c7c85 100644
--- a/net/netfilter/x_tables.c
+++ b/net/netfilter/x_tables.c
@@ -1142,7 +1142,8 @@ void xt_compat_target_from_user(struct xt_entry_target *t, void **dstptr,
if (target->compat_from_user)
target->compat_from_user(t->data, ct->data);
else
- memcpy(t->data, ct->data, tsize - sizeof(*ct));
+ unsafe_memcpy(t->data, ct->data, tsize - sizeof(*ct),
+ /* UAPI 0-sized destination */);

tsize += off;
t->u.user.target_size = tsize;
--
2.34.1



2024-02-19 17:19:36

by Simon Horman

[permalink] [raw]
Subject: Re: [PATCH] netfilter: x_tables: Use unsafe_memcpy() for 0-sized destination

On Fri, Feb 16, 2024 at 03:31:32PM -0800, Kees Cook wrote:
> The struct xt_entry_target fake flexible array has not be converted to a
> true flexible array, which is mainly blocked by it being both UAPI and
> used in the middle of other structures. In order to properly check for
> 0-sized destinations in memcpy(), an exception must be made for the one
> place where it is still a destination. Since memcpy() was already
> skipping checks for 0-sized destinations, using unsafe_memcpy() is no
> change in behavior.
>
> Signed-off-by: Kees Cook <[email protected]>

Reviewed-by: Simon Horman <[email protected]>