[PATCH net-next V2] net/vxlan: Avoid unaligned access in vxlan_build_skb()

The vxlan header may not be aligned to 4 bytes in
vxlan_build_skb (e.g., for MLD packets). This patch
avoids unaligned access traps from vxlan_build_skb
(in platforms like sparc) by making struct vxlanhdr __packed.

Signed-off-by: Sowmini Varadhan <redacted>
---
 include/net/vxlan.h |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/include/net/vxlan.h b/include/net/vxlan.h
index 0255613..1ec56f4 100644
--- a/include/net/vxlan.h
+++ b/include/net/vxlan.h

@@ -18,7 +18,7 @@
 struct vxlanhdr {
 	__be32 vx_flags;
 	__be32 vx_vni;
-};
+} __packed;
 
 /* VXLAN header flags. */
 #define VXLAN_HF_VNI	cpu_to_be32(BIT(27))

-- 
1.7.1

On Tue, 20 Sep 2016 14:57:37 -0400, Sowmini Varadhan wrote:

The vxlan header may not be aligned to 4 bytes in
vxlan_build_skb (e.g., for MLD packets). This patch
avoids unaligned access traps from vxlan_build_skb
(in platforms like sparc) by making struct vxlanhdr __packed.

Signed-off-by: Sowmini Varadhan <redacted>

Acked-by: Jiri Benc <redacted>

On 20.09.2016 20:57, Sowmini Varadhan wrote:

The vxlan header may not be aligned to 4 bytes in
vxlan_build_skb (e.g., for MLD packets). This patch
avoids unaligned access traps from vxlan_build_skb
(in platforms like sparc) by making struct vxlanhdr __packed.

Signed-off-by: Sowmini Varadhan <redacted>

Performance wise this should only affect code generation for archs where
it matters anyway.

Acked-by: Hannes Frederic Sowa <redacted>

Thanks,
Hannes

On Wed, Sep 21, 2016 at 12:10:55PM +0200, Hannes Frederic Sowa wrote:

On 20.09.2016 20:57, Sowmini Varadhan wrote:

quoted

The vxlan header may not be aligned to 4 bytes in
vxlan_build_skb (e.g., for MLD packets). This patch
avoids unaligned access traps from vxlan_build_skb
(in platforms like sparc) by making struct vxlanhdr __packed.

Signed-off-by: Sowmini Varadhan <redacted>

Performance wise this should only affect code generation for archs where
it matters anyway.

I think it's the opposite. Even on x86 compiler will use byte loads.

On 21.09.2016 18:14, Alexei Starovoitov wrote:

On Wed, Sep 21, 2016 at 12:10:55PM +0200, Hannes Frederic Sowa wrote:

quoted

On 20.09.2016 20:57, Sowmini Varadhan wrote:

quoted

The vxlan header may not be aligned to 4 bytes in
vxlan_build_skb (e.g., for MLD packets). This patch
avoids unaligned access traps from vxlan_build_skb
(in platforms like sparc) by making struct vxlanhdr __packed.

Signed-off-by: Sowmini Varadhan <redacted>

Performance wise this should only affect code generation for archs where
it matters anyway.

I think it's the opposite. Even on x86 compiler will use byte loads.

I checked that on x86-64 actually. Also clearly visible here:

https://godbolt.org/g/xsW2P1

Bye,
Hannes

On Wed, 2016-09-21 at 09:14 -0700, Alexei Starovoitov wrote:

I think it's the opposite. Even on x86 compiler will use byte loads.

Unless you tweaked gcc, it should still use word loads on x86.

On Wed, Sep 21, 2016 at 09:53:31AM -0700, Eric Dumazet wrote:

On Wed, 2016-09-21 at 09:14 -0700, Alexei Starovoitov wrote:

quoted

I think it's the opposite. Even on x86 compiler will use byte loads.

Unless you tweaked gcc, it should still use word loads on x86.

checked that on x86-64 actually. Also clearly visible here:

ahh. ok. good to know. thanks guys!