| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
When an ipvlan device is created on a bond device, the link state
of the ipvlan device may be abnormal. This is because bonding device
allows to add physical network card device in the down state and so
NETDEV_CHANGE event will not be notified to other listeners, so ipvlan
has no chance to update its link status.
The following steps can cause such problems:
1) bond0 is down
2) ip link add link bond0 name ipvlan type ipvlan mode l2
3) echo +enp2s7 >/sys/class/net/bond0/bonding/slaves
4) ip link set bond0 up
After these steps, use ip link command, we found ipvlan has NO-CARRIER:
ipvlan@bond0: <NO-CARRIER, BROADCAST,MULTICAST,UP,M-DOWN> mtu ...>
We can deal with this problem like VLAN: Add handling of NETDEV_UP
events. If we receive NETDEV_UP event, we will update the link status
of the ipvlan.
Signed-off-by: Di Zhu <zhudi21@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
currently, only 'ingress' and 'clsact ingress' qdiscs store the tc 'chain
id' in the skb extension. However, userspace programs (like ovs) are able
to setup egress rules, and datapath gets confused in case it doesn't find
the 'chain id' for a packet that's "recirculated" by tc.
Change tcf_classify() to have the same semantic as tcf_classify_ingress()
so that a single function can be called in ingress / egress, using the tc
ingress / egress block respectively.
Suggested-by: Alaa Hleilel <alaa@nvidia.com>
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Reviewed-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Ioana Ciornei says:
====================
dpaa2-switch: add mirroring support
This patch set adds per port and per VLAN mirroring in dpaa2-switch.
The first 4 patches are just cosmetic changes. We renamed the
dpaa2_switch_acl_tbl structure into dpaa2_switch_filter_block so that we
can reuse it for filters that do not use the ACL table and reorganized
the addition of trap, redirect and drop filters into a separate
function. All this just to make for a more streamlined addition of the
support for mirroring.
The next 4 patches are actually adding the advertised support. Mirroring
rules can be added in shared blocks, the driver will replicate the same
configuration on all the switch ports part of the same block.
The last patch documents the feature, presents its behavior and
limitations and gives a couple of examples.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Document the mirroring capabilities of the dpaa2-switch driver,
any restrictions that are imposed and some example commands.
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
When mirroring rules are added in shared filter blocks, the same
mirroring rule has to be configured on all the switch ports that are
part of the same block.
In case a switch port joins a shared block after mirroring filters have
been already added to it, then all the mirror rules should be offloaded
to the port. The reverse, removal of mirroring rules, has to be done at
block unbind.
For this purpose, the dpaa2_switch_block_offload_mirror() and
dpaa2_switch_block_unoffload_mirror() functions are added and called
upon binding and unbinding a switch port to/from a block.
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Using the infrastructure added in the previous patch, extend tc-flower
support with FLOW_ACTION_MIRRED based on VLAN.
Tested with:
tc qdisc add dev eth8 ingress_block 1 clsact
tc filter add block 1 ingress protocol 802.1q flower skip_sw \
vlan_id 100 action mirred egress mirror dev eth6
tc filter del block 1 ingress pref 49152
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add support for per port mirroring for the DPAA2 switch. We support
only single mirror port, therefore we allow mirroring rules only as long
as the destination port is always the same.
Unlike all the actions (drop, redirect, trap) already supported by the
dpaa2-switch driver, adding mirroring filters in shared blocks is not
achieved by a singular ACL entry added in a table shared by the ports.
This is why, when a new mirror filter is added in a block we have to got
through all the switch ports sharing it and configure the filter
individually on all.
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add the necessary MC API for setting up and configuring the mirroring
feature on the DPSW DPAA2 object.
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Extract the necessary steps to offload a filter by using the ACL table
in a separate function - dpaa2_switch_cls_matchall_replace_acl().
This is intended to help with the code readability when the mirroring
support is added.
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Extract the necessary steps to offload a filter by using the ACL table
in a separate function - dpaa2_switch_cls_flower_replace_acl().
This is intended to help with the code readability when the mirroring
support is added.
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Until now, shared filter blocks were implemented only by ACL tables
shared between ports. Going forward, when the mirroring support will be
added, this will not be true anymore.
Rename the dpaa2_switch_acl_tbl into dpaa2_switch_filter_block so that
we make it clear that the structure is used not only for filters that
use the ACL table but will be used for all the filters that are added in
a block.
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Until now, the dpaa2_switch_tc_parse_action() function was used for all
the supported tc actions since all of them were implemented by adding
ACL table entries. In the next commits, the dpaa2-switch driver will
gain mirroring support which is not using the same HW feature.
Make sure that we specify the ACL in the function name so that we make
it clear that it's only used for specific actions.
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Removing the qede module version which is not needed and not allowed
with inbox drivers.
Signed-off-by: Prabhakar Kushwaha <pkushwaha@marvell.com>
Signed-off-by: Ariel Elior <aelior@marvell.com>
Signed-off-by: Shai Malin <smalin@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Removing the qed module version which is not needed and not allowed
with inbox drivers.
Signed-off-by: Prabhakar Kushwaha <pkushwaha@marvell.com>
Signed-off-by: Ariel Elior <aelior@marvell.com>
Signed-off-by: Shai Malin <smalin@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Vladimir Oltean says:
====================
NXP SJA1105 VLAN regressions
These are 3 patches to fix issues seen with some more varied testing
done after the changes in the "Traffic termination for sja1105 ports
under VLAN-aware bridge" series were made:
https://patchwork.kernel.org/project/netdevbpf/cover/20210726165536.1338471-1-vladimir.oltean@nxp.com/
Issue 1: traffic no longer works on a port after leaving a VLAN-aware bridge
Issue 2: untagged traffic not dropped if pvid is absent from a VLAN-aware port
Issue 3: PTP and STP broken on ports under a VLAN-aware bridge
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
On RX, a control packet with SJA1110 will have:
- an in-band control extension (DSA tag) composed of a header and an
optional trailer (if it is a timestamp frame). We can (and do) deduce
the source port and switch id from this.
- a VLAN header, which can either be the tag_8021q RX VLAN (pvid) or the
bridge VLAN. The sja1105_vlan_rcv() function attempts to deduce the
source port and switch id a second time from this.
The basic idea is that even though we don't need the source port
information from the tag_8021q header if it's a control packet, we do
need to strip that header before we pass it on to the network stack.
The problem is that we call sja1105_vlan_rcv for ports under VLAN-aware
bridges, and that function tells us it couldn't identify a tag_8021q
header, so we need to perform imprecise RX by VID. Well, we don't,
because we already know the source port and switch ID.
This patch drops the return value from sja1105_vlan_rcv and we just look
at the source_port and switch_id values from sja1105_rcv and sja1110_rcv
which were initialized to -1. If they are still -1 it means we need to
perform imprecise RX.
Fixes: 884be12f8566 ("net: dsa: sja1105: add support for imprecise RX")
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Surprisingly, this configuration:
ip link add br0 type bridge vlan_filtering 1
ip link set swp2 master br0
bridge vlan del dev swp2 vid 1
still has the sja1105 switch sending untagged packets to the CPU (and
failing to decode them, since dsa_find_designated_bridge_port_by_vid
searches by VID 1 and rightfully finds no bridge VLAN 1 on a port).
Dumping the switch configuration, the VLANs are managed properly:
- the pvid of swp2 is 1 in the MAC Configuration Table, but
- only the CPU port is in the port membership of VLANID 1 in the VLAN
Lookup Table
When the ingress packets are tagged with VID 1, they are properly
dropped. But when they are untagged, they are able to reach the CPU
port. Also, when the pvid in the MAC Configuration Table is changed to
e.g. 55 (an unused VLAN), the untagged packets are also dropped.
So it looks like:
- the switch bypasses ingress VLAN membership checks for untagged traffic
- the reason why the untagged traffic is dropped when I make the pvid 55
is due to the lack of valid destination ports in VLAN 55, rather than
an ingress membership violation
- the ingress VLAN membership cheks are only done for VLAN-tagged traffic
Interesting. It looks like there is an explicit bit to drop untagged
traffic, so we should probably be using that to preserve user expectations.
Note that only VLAN-aware ports should drop untagged packets due to no
pvid - when VLAN-unaware, the software bridge doesn't do this even if
there is no pvid on any bridge port and on the bridge itself. So the new
sja1105_drop_untagged() function cannot simply be called with "false"
from sja1105_bridge_vlan_add() and with "true" from sja1105_bridge_vlan_del.
Instead, we need to also consider the VLAN awareness state. That means
we need to hook the "drop untagged" setting in all the same places where
the "commit pvid" logic is, and it needs to factor in all the state when
flipping the "drop untagged" bit: is our current pvid in the VLAN Lookup
Table, and is the current port in that VLAN's port membership list?
VLAN-unaware ports will never drop untagged frames because these checks
always succeed by construction, and the tag_8021q VLANs cannot be changed
by the user.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Now that we no longer have the ultra-central sja1105_build_vlan_table(),
we need to be more careful about checking all corner cases manually.
For example, when a port leaves a VLAN-aware bridge, it becomes
standalone so its pvid should become a tag_8021q RX VLAN again. However,
sja1105_commit_pvid() only gets called from sja1105_bridge_vlan_add()
and from sja1105_vlan_filtering(), and no VLAN awareness change takes
place (VLAN filtering is a global setting for sja1105, so the switch
remains VLAN-aware overall).
This means that we need to put another sja1105_commit_pvid() call in
sja1105_bridge_member().
Fixes: 6dfd23d35e75 ("net: dsa: sja1105: delete vlan delta save/restore logic")
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Jeremy Kerr says:
====================
Add Management Component Transport Protocol support
This series adds core MCTP support to the kernel. From the Kconfig
description:
Management Component Transport Protocol (MCTP) is an in-system
protocol for communicating between management controllers and
their managed devices (peripherals, host processors, etc.). The
protocol is defined by DMTF specification DSP0236.
This option enables core MCTP support. For communicating with other
devices, you'll want to enable a driver for a specific hardware
channel.
This implementation allows a sockets-based API for sending and receiving
MCTP messages via sendmsg/recvmsg on SOCK_DGRAM sockets. Kernel stack
control is all via netlink, using existing RTM_* messages. The userspace
ABI change is fairly small; just the necessary AF_/ETH_P_/ARPHDR_
constants, a new sockaddr, and a new netlink attribute.
For MAINTAINERS, I've just included netdev@ as the list entry. I'm happy
to alter this based on preferences here - an alternative would be the
OpenBMC list (the main user of the MCTP interface), or we can create a
new list entirely.
We have a couple of interface drivers almost ready to go at the moment,
but those can wait until the core code has some review.
This is v4 of the series; v1 and v2 were both RFC.
selinux folks: CCing 01/15 due to the new PF_MCTP protocol family.
linux-doc folks: CCing 15/15 for the new MCTP overview document.
Review, comments, questions etc. are most welcome.
Cheers,
Jeremy
v2:
- change to match spec terminology: controller -> component
- require specific capabilities for bind() & sendmsg()
- add address and tag defintions to uapi
- add selinux AF_MCTP table definitions
- remove strict cflags; warnings are present in common headers
v3:
- require caps for MCTP bind() & send()
- comment typo fixes
- switch to an array for local EIDs
- fix addrinfo dump iteration & error path
- add RTM_DELADDR
- remove GENMASK() and BIT() from uapi
v4:
- drop tun patch; that can be submitted separately
- keep nipa happy: add maintainer CCs, including doc and selinux
- net-next rebase
- Include AF_MCTP in af_family_slock_keys and pf_family_names
- Introduce MODULE_ definitions earlier
- upstream change: set_link_af no longer called with RTNL held
- add kdoc for net_device.mctp_ptr
- don't inline mctp_rt_match_eid
- require rtm_type == RTN_UNICAST in route management handlers
- remove unused RTAX policy table
- fix mctp_sock->keys rcu annotations
- fix spurious rcu_read_unlock in route input
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This change adds a brief document about the sockets API provided for
sending and receiving MCTP messages from userspace.
This is roughly based on the OpenBMC design document, at:
https://github.com/openbmc/docs/blob/master/designs/mctp/mctp-kernel.md
Signed-off-by: Jeremy Kerr <jk@codeconstruct.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Currently we have a compile-time default network
(MCTP_INITIAL_DEFAULT_NET). This change introduces a default_net field
on the net namespace, allowing future configuration for new interfaces.
Signed-off-by: Matt Johnston <matt@codeconstruct.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Now that we have a neighbour implementation, hook it up to the output
path to set the dest hardware address for outgoing packets.
Signed-off-by: Matt Johnston <matt@codeconstruct.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This change implements MCTP fragmentation (based on route & device MTU),
and corresponding reassembly.
The MCTP specification only allows for fragmentation on the originating
message endpoint, and reassembly on the destination endpoint -
intermediate nodes do not need to reassemble/refragment. Consequently,
we only fragment in the local transmit path, and reassemble
locally-bound packets. Messages are required to be in-order, so we
simply cancel reassembly on out-of-order or missing packets.
In the fragmentation path, we just break up the message into MTU-sized
fragments; the skb structure is a simple copy for now, which we can later
improve with a shared data implementation.
For reassembly, we keep track of incoming message fragments using the
existing tag infrastructure, allocating a key on the (src,dest,tag)
tuple, and reassembles matching fragments into a skb->frag_list.
Signed-off-by: Jeremy Kerr <jk@codeconstruct.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Start filling-out the socket syscalls: bind, sendmsg & recvmsg.
This requires an input route implementation, so we add to
mctp_route_input, allowing lookups on binds & message tags. This just
handles single-packet messages at present, we will add fragmentation in
a future change.
Signed-off-by: Jeremy Kerr <jk@codeconstruct.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This change adds the netlink interfaces for manipulating the MCTP
neighbour table.
Signed-off-by: Matt Johnston <matt@codeconstruct.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add an initial neighbour table implementation, to be used in the route
output path.
Signed-off-by: Matt Johnston <matt@codeconstruct.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This change adds RTM_GETROUTE, RTM_NEWROUTE & RTM_DELROUTE handlers,
allowing management of the MCTP route table.
Includes changes from Jeremy Kerr <jk@codeconstruct.com.au>.
Signed-off-by: Matt Johnston <matt@codeconstruct.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add a simple routing table, and a couple of route output handlers, and
the mctp packet_type & handler.
Includes changes from Matt Johnston <matt@codeconstruct.com.au>.
Signed-off-by: Jeremy Kerr <jk@codeconstruct.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This change adds the infrastructure for managing MCTP netdevices; we add
a pointer to the AF_MCTP-specific data to struct netdevice, and hook up
the rtnetlink operations for adding and removing addresses.
Includes changes from Matt Johnston <matt@codeconstruct.com.au>.
Signed-off-by: Jeremy Kerr <jk@codeconstruct.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add an empty drivers/net/mctp/, for future interface drivers.
Signed-off-by: Jeremy Kerr <jk@codeconstruct.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This change introduces the user-visible MCTP header, containing the
protocol-specific addressing definitions.
Signed-off-by: Jeremy Kerr <jk@codeconstruct.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Simple packet header format as defined by DMTF DSP0236.
Signed-off-by: Jeremy Kerr <jk@codeconstruct.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add an empty socket implementation, plus initialisation/destruction
handlers.
Signed-off-by: Jeremy Kerr <jk@codeconstruct.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add basic Kconfig, an initial (empty) af_mctp source object, and
{AF,PF}_MCTP definitions, and the required definitions for a new
protocol type.
Signed-off-by: Jeremy Kerr <jk@codeconstruct.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Krzysztof Kozlowski says:
====================
nfc: constify, continued (part 2)
On top of:
nfc: constify pointed data
https://lore.kernel.org/lkml/20210726145224.146006-1-krzysztof.kozlowski@canonical.com/
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
File-scope struct nfcmrvl_if_ops is not modified so can be made const.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Several functions do not modify pointed data so arguments and local
variables can be const for correctness and safety.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Several functions do not modify pointed data so arguments and local
variables can be const for correctness and safety.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Several functions do not modify pointed data so arguments and local
variables can be const for correctness and safety. This allows also
making file-scope nci_core_get_config_otp_ram_version array const.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Loop iterators are simple integers, no point to optimize the size and
use u8. It only raises the question whether the variable is used in
some other context.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Size of firmware is a type of size_t, so print it directly instead of
casting to int.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
nfcsim_abort_cmd() does not modify struct nfcsim, so local variable
can be a pointer to const.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
virtual_ncidev_ioctl() does not modify struct nfc_dev, so local variable
can be a pointer to const.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Several functions do not modify pointed data so arguments and local
variables can be const for correctness and safety.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Several functions do not modify pointed data so arguments and local
variables can be const for correctness and safety.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The buffer passed to mei_nfc_send() can be const for correctness and
safety.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The struct nfc_dev is not modified by nfc_get_drvdata() and
nfc_device_name() so it can be made a const.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Paolo Abeni says:
====================
sk_buff: optimize GRO for the common case
This is a trimmed down revision of "sk_buff: optimize layout for GRO",
specifically dropping the changes to the sk_buff layout[1].
This series tries to accomplish 2 goals:
- optimize the GRO stage for the most common scenario, avoiding a bunch
of conditional and some more code
- let owned skbs entering the GRO engine, allowing backpressure in the
veth GRO forward path.
A new sk_buff flag (!!!) is introduced and maintained for GRO's sake.
Such field uses an existing hole, so there is no change to the sk_buff
size.
[1] two main reasons:
- move skb->inner_ field requires some extra care, as some in kernel
users access and the fields regardless of skb->encapsulation.
- extending secmark size clash with ct and nft uAPIs
address the all above is possible, I think, but for sure not in a single
series.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Leveraging the previous patch we can now avoid orphaning the
skb in the veth gro path, allowing correct backpressure.
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This change leverages the infrastructure introduced by the previous
patches to allow soft devices passing to the GRO engine owned skbs
without impacting the fast-path.
It's up to the GRO caller ensuring the slow_gro bit validity before
invoking the GRO engine. The new helper skb_prepare_for_gro() is
introduced for that goal.
On slow_gro, skbs are aggregated only with equal sk.
Additionally, skb truesize on GRO recycle and free is correctly
updated so that sk wmem is not changed by the GRO processing.
rfc-> v1:
- fixed bad truesize on dev_gro_receive NAPI_FREE
- use the existing state bit
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
Di Zhu
Davide Caratti
David S. Miller
Ioana Ciornei
Shai Malin
Vladimir Oltean
Jeremy Kerr
Matt Johnston
Krzysztof Kozlowski
Paolo Abeni