On 05/11/2017 07:23 PM, David Miller wrote:
From: Daniel Borkmann
Date: Thu, 11 May 2017 18:49:45 +0200
This actually won't work, see also commit 79adffcd6489 ("bpf,
verifier:
fix rejection of unaligned access checks for map_value_adj") with some
longer explanation. In case of map_value_adj, r
From: Daniel Borkmann
Date: Thu, 11 May 2017 18:49:45 +0200
> This actually won't work, see also commit 79adffcd6489 ("bpf,
> verifier:
> fix rejection of unaligned access checks for map_value_adj") with some
> longer explanation. In case of map_value_adj, reg->id is always 0.
I see
Ok if I
On 05/11/2017 06:06 PM, David Miller wrote:
We can validate PTR_TO_MAP_VALUE_ADJ accesses in the same way that we
do for PTR_TO_PACKET. The only difference is that we don't plug
NET_IP_ALIGN into the equation.
Signed-off-by: David S. Miller
---
kernel/bpf/verifier.c | 25 +++
We can validate PTR_TO_MAP_VALUE_ADJ accesses in the same way that we
do for PTR_TO_PACKET. The only difference is that we don't plug
NET_IP_ALIGN into the equation.
Signed-off-by: David S. Miller
---
kernel/bpf/verifier.c | 25 +
1 file changed, 21 insertions(+), 4 del