- 09 Mar, 2022 23 commits
-
-
Vladimir Oltean authored
The Felix driver declares FDB isolation but puts all standalone ports in VID 0. This is mostly problem-free as discussed with Alvin here: https://patchwork.kernel.org/project/netdevbpf/cover/20220302191417.1288145-1-vladimir.oltean@nxp.com/#24763870 however there is one catch. DSA still thinks that FDB entries are installed on the CPU port as many times as there are user ports, and this is problematic when multiple user ports share the same MAC address. Consider the default case where all user ports inherit their MAC address from the DSA master, and then the user runs: ip link set swp0 address 00:01:02:03:04:05 The above will make dsa_slave_set_mac_address() call dsa_port_standalone_host_fdb_add() for 00:01:02:03:04:05 in port 0's standalone database, and dsa_port_standalone_host_fdb_del() for the old address of swp0, again in swp0's standalone database. Both the ->port_fdb_add() and ->port_fdb_del() will be propagated down to the felix driver, which will end up deleting the old MAC address from the CPU port. But this is still in use by other user ports, so we end up breaking unicast termination for them. There isn't a problem in the fact that DSA keeps track of host standalone addresses in the individual database of each user port: some drivers like sja1105 need this. There also isn't a problem in the fact that some drivers choose the same VID/FID for all standalone ports. It is just that the deletion of these host addresses must be delayed until they are known to not be in use any longer, and only the driver has this knowledge. Since DSA keeps these addresses in &cpu_dp->fdbs and &cpu_db->mdbs, it is just a matter of walking over those lists and see whether the same MAC address is present on the CPU port in the port db of another user port. I have considered reusing the generic dsa_port_walk_fdbs() and dsa_port_walk_mdbs() schemes for this, but locking makes it difficult. In the ->port_fdb_add() method and co, &dp->addr_lists_lock is held, but dsa_port_walk_fdbs() also acquires that lock. Also, even assuming that we introduce an unlocked variant of the address iterator, we'd still need some relatively complex data structures, and a void *ctx in the dsa_fdb_walk_cb_t which we don't currently pass, such that drivers are able to figure out, after iterating, whether the same MAC address is or isn't present in the port db of another port. All the above, plus the fact that I expect other drivers to follow the same model as felix where all standalone ports use the same FID, made me conclude that a generic method provided by DSA is necessary: dsa_fdb_present_in_other_db() and the mdb equivalent. Felix calls this from the ->port_fdb_del() handler for the CPU port, when the database was classified to either a port db, or a LAG db. For symmetry, we also call this from ->port_fdb_add(), because if the address was installed once, then installing it a second time serves no purpose: it's already in hardware in VID 0 and it affects all standalone ports. This change moves dsa_db_equal() from switch.c to dsa.c, since it now has one more caller. Fixes: 54c31984 ("net: mscc: ocelot: enforce FDB isolation when VLAN-unaware") Signed-off-by:
Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Vladimir Oltean authored
The two blamed commits were written/tested individually but not together. When put together, commit 90897569 ("net: dsa: felix: start off with flooding disabled on the CPU port"), which deletes a reinitialization of PGID_UC/PGID_MC/PGID_BC, is no longer sufficient to ensure that these port masks don't contain the CPU port module. This is because commit b903a6bd ("net: dsa: felix: migrate flood settings from NPI to tag_8021q CPU port") overwrites the hardware default settings towards the CPU port module with the settings that used to be present on the NPI port treated as a regular port. There, flooding is enabled, so flooding would get enabled on the CPU port module too. Adding conditional logic somewhere within felix_setup_tag_npi() to configure either the default no-flood policy or the flood policy inherited from the tag_8021q CPU port from a previous call to dsa_port_manage_cpu_flood() is getting complicated. So just let the migration logic do its thing during initial setup (which will temporarily turn on flooding), then turn flooding off for the NPI port after felix_set_tag_protocol() finishes. Here we are in felix_setup(), so the DSA slave interfaces are not yet created, and this doesn't affect traffic in any way. Fixes: 90897569 ("net: dsa: felix: start off with flooding disabled on the CPU port") Fixes: b903a6bd ("net: dsa: felix: migrate flood settings from NPI to tag_8021q CPU port") Signed-off-by:
-
Vladimir Oltean authored
Since the slave unicast address is synced to hardware and to the DSA master during dsa_slave_open(), this means that a call to dsa_slave_set_mac_address() while the slave interface is down will result to a call to dsa_port_standalone_host_fdb_del() and to dev_uc_del() for the MAC address while there was no previous dsa_port_standalone_host_fdb_add() or dev_uc_add(). This is a partial revert of the blamed commit below, which was too aggressive. Fixes: 35aae5ab ("net: dsa: remove workarounds for changing master promisc/allmulti only while up") Signed-off-by:
-
Vladimir Oltean authored
We no longer need the workaround in the felix driver to avoid calling dsa_port_walk_fdbs() when &dp->fdbs is an uninitialized list, because that list is now initialized from all call paths of felix_set_tag_protocol(). Signed-off-by:
-
Vladimir Oltean authored
&cpu_db->fdbs and &cpu_db->mdbs may be uninitialized lists during some call paths of felix_set_tag_protocol(). There was an attempt to avoid calling dsa_port_walk_fdbs() during setup by using a "bool change" in the felix driver, but this doesn't work when the tagging protocol is defined in the device tree, and a change is triggered by DSA at pseudo-runtime: dsa_tree_setup_switches -> dsa_switch_setup -> dsa_switch_setup_tag_protocol -> ds->ops->change_tag_protocol dsa_tree_setup_ports -> dsa_port_setup -> &dp->fdbs and &db->mdbs only get initialized here So it seems like the only way to fix this is to move the initialization of these lists earlier. dsa_port_touch() is called from dsa_switch_touch_ports() which is called from dsa_switch_parse_of(), and this runs completely before dsa_tree_setup(). Similarly, dsa_switch_release_ports() runs after dsa_tree_teardown(). Fixes: f9cef64f ("net: dsa: felix: migrate host FDB and MDB entries when changing tag proto") Signed-off-by: -
Vladimir Oltean authored
There has been recent work towards matching each switchdev object addition with a corresponding deletion. Therefore, having elements in the fdbs, mdbs, vlans lists at the time of a shared (DSA, CPU) port's teardown is indicative of a bug somewhere else, and not something that is to be expected. We shouldn't try to silently paper over that. Instead, print a warning and a stack trace. This change is a prerequisite for moving the initialization/teardown of these lists. Make it clear that clearing the lists isn't needed. Signed-off-by:
-
David S. Miller authored
Jonathan Lemon says: ==================== ptp: ocp: update devlink information Both of these patches update the information displayed via devlink. v1 -> v2: remove board.manufacture information ==================== Signed-off-by: -
Jonathan Lemon authored
Cache the firmware version when the card is initialized, and use this field to populate the devlink firmware information. The cached firmware version will be used for feature gating in upcoming patches. Signed-off-by:
Jonathan Lemon <jonathan.lemon@gmail.com> Signed-off-by:
-
Jonathan Lemon authored
Add the at24 drivers for the eeprom, and use the accessors via the nvmem API instead of direct i2c accesses. This makes things cleaner. Add an eeprom map table which specifies where the pre-defined information is located. Retrieve the information and and export it via the devlink interface. Signed-off-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/nextDavid S. Miller authored
-queue Tony Nguyen says: ==================== 10GbE Intel Wired LAN Driver Updates 2022-03-08 This series contains updates to ixgbe and ixgbevf drivers. Slawomir adds an implementation for ndo_set_vf_link_state() to allow for disabling of VF link state as well a mailbox implementation so the VF can query the state. Additionally, for 82599, the option to disable a VF after receiving several malicious driver detection (MDD) events are encountered is added. For ixgbevf, the corresponding implementation to query and report a disabled state is added. ==================== Signed-off-by: -
Colin Ian King authored
Don't populate the read-only array client_map on the stack but instead make it static const. Also makes the object code a little smaller. Signed-off-by:
Colin Ian King <colin.i.king@gmail.com> Link: https://lore.kernel.org/r/20220307221349.164585-1-colin.i.king@gmail.comSigned-off-by:
Jakub Kicinski <kuba@kernel.org>
-
Jonathan Lemon authored
When devlink_register() was removed from the error path, the corresponding label was not updated. Rename the label for readability puposes, no functional change. Signed-off-by:
-
Samuel Thibault authored
ENOTSUPP is documented as "should never be seen by user programs", and thus not exposed in <errno.h>, and thus applications cannot safely check against it (they get "Unknown error 524" as strerror). We should rather return the well-known -EOPNOTSUPP. This is similar to 2230a7ef ("drop_monitor: Use correct error code") and 4a5cdc60 ("net/tls: Fix return values to avoid ENOTSUPP"), which did not seem to cause problems. Signed-off-by:
Samuel Thibault <samuel.thibault@labri.fr> Acked-by:
Willem de Bruijn <willemb@google.com> Link: https://lore.kernel.org/r/20220307223126.djzvg44v2o2jkjsx@beginSigned-off-by:
-
Jakub Kicinski authored
Mat Martineau says: ==================== mptcp: Advertisement reliability improvement and misc. updates Patch 1 adds a helpful debug tracepoint for outgoing MPTCP packets. Patch 2 is a small "magic number" refactor. Patches 3 & 4 refactor parts of the mptcp_join.sh selftest. No change in test coverage. Patch 5 ensures only advertised address IDs are un-advertised. Patches 6-8 improve handling of an edge case where endpoint IDs need to be created on-the-fly when adding subflows. Includes selftest coverage. Patch 9 adds validation of the fullmesh flag in a MPTCP netlink command, which was overlooked when this flag was introduced for 5.18. ==================== Link: https://lore.kernel.org/r/20220307204439.65164-1-mathew.j.martineau@linux.intel.comSigned-off-by:
-
Geliang Tang authored
The fullmesh flag mustn't be used with the signal flag when adding an address. This patch added the necessary flags check for this case. Fixes: 73c762c1 ("mptcp: set fullmesh flag in pm_netlink") Signed-off-by:
Geliang Tang <geliang.tang@suse.com> Signed-off-by:
Mat Martineau <mathew.j.martineau@linux.intel.com> Signed-off-by:
-
Paolo Abeni authored
Ensure implicit endpoint are created when expected and that the user-space can update them Reviewed-by:
Matthieu Baerts <matthieu.baerts@tessares.net> Co-developed-by:
Paolo Abeni <pabeni@redhat.com> Signed-off-by:
-
Paolo Abeni authored
The address ID selection for MPJ subflows created in response to incoming ADD_ADDR option is currently unreliable: it happens at MPJ socket creation time, when the local address could be unknown. Additionally, if the no local endpoint is available for the local address, a new dummy endpoint is created, confusing the user-land. This change refactor the code to move the address ID selection inside the rebuild_header() helper, when the local address eventually selected by the route lookup is finally known. If the address used is not mapped by any endpoint - and thus can't be advertised/removed pick the id 0 instead of allocate a new endpoint. Signed-off-by:
-
Paolo Abeni authored
In some edge scenarios, an MPTCP subflows can use a local address mapped by a "implicit" endpoint created by the in-kernel path manager. Such endpoints presence can be confusing, as it's creation is hard to track and will prevent the later endpoint creation from the user-space using the same address. Define a new endpoint flag to mark implicit endpoints and allow the user-space to replace implicit them with user-provided data at endpoint creation time. Signed-off-by:
-
Paolo Abeni authored
The in-kernel MPTCP path manager, when processing the MPTCP_PM_CMD_FLUSH_ADDR command, generates RM_ADDR events for each known local address. While that is allowed by the RFC, it makes unpredictable the exact number of RM_ADDR generated when both ends flush the PM addresses. This change restricts the RM_ADDR generation to previously explicitly announced addresses, and adjust the expected results in a bunch of related self-tests. Signed-off-by:
-
Mat Martineau authored
The "selftests: mptcp: improve 'fair usage on close' stability" commit changed that self test to check the TcpAttemptFails MIB instead of looking for TW sockets. The associated bash function wasn't renamed in that commit because of the merge conflicts it would cause, so this commit updates the function name as Paolo originally intended. Cc: Paolo Abeni <pabeni@redhat.com> Reviewed-by:
-
Matthieu Baerts authored
Without this patch, no tests would be ran when launching: mptcp_join.sh -cCi In any order or a combination with 2 of these letters. The recommended way with getopt is first parse all options and then act. This allows to do some actions in priority, e.g. display the help menu and stop. But also some global variables changing the behaviour of this selftests -- like the ones behind -cCi options -- can be set before running the different tests. By doing that, we can also avoid long and unreadable regex. Signed-off-by:
-
Geliang Tang authored
Set subflow->data_avail with the enum value MPTCP_SUBFLOW_NODATA, instead of using 0 directly. Reviewed-by:
-
Geliang Tang authored
The tracepoint in get_mapping_status() only dumped the incoming mpext fields. This patch added a new tracepoint in mptcp_sendmsg_frag() to dump the outgoing mpext too. Signed-off-by:
-
- 08 Mar, 2022 12 commits
-
-
Slawomir Mrozowicz authored
Add possibility to disable link state if it is administratively disabled in PF. It is part of the general functionality that allows the PF driver to control the state of the virtual link VF devices. Signed-off-by:
Slawomir Mrozowicz <slawomirx.mrozowicz@intel.com> Tested-by:
Konrad Jankowski <konrad0.jankowski@intel.com> Signed-off-by:
Tony Nguyen <anthony.l.nguyen@intel.com>
-
Slawomir Mrozowicz authored
The 82599 PF driver disable VF driver after a special MDD event occurs. Adds the option for administrators to control whether VFs are automatically disabled after several MDD events. The automatically disabling is now the default mode for 82599 PF driver, as it is more reliable. This addresses CVE-2021-33061. Signed-off-by:
-
Slawomir Mrozowicz authored
Add support for ndo_set_vf_link_state the Network Device Option that allows the PF driver to control the virtual link state of the VF devices. Without this change a VF cannot be disabled/enabled by the administrator. In the implementation the auto state takes over PF link state to VF link setting, the enable state is not supported, the disable state shut off the VF link regardless of the PF setting. Signed-off-by:
-
Paolo Abeni authored
Arun Ramadoss says: ==================== net: phy: lan87xx: use genphy_read_master_slave function LAN87xx T1 Phy has the same register field as gigabit phy for reading the master slave configuration. But the genphy_read_master_slave function has a check of gigabit phy. So refactored the function in such a way, moved the speed check to the genphy_read_status function. Analyzed the nxp-tja11xx function for refactoring, but the register for configuring master/slave is nxp specific which is not extended phy register. And analyzed the reusing genphy_setup_master_slave, but for LAN87xx MASTER_ENABLE is always 1 and Preferred state is always 0. So, I didn't try to change it. ==================== Link: https://lore.kernel.org/r/20220307161515.14970-1-arun.ramadoss@microchip.comSigned-off-by:
-
Arun Ramadoss authored
To read the master slave configuration of the LAN87xx T1 phy, used the generic phy driver genphy_read_master_slave function. Removed the local lan87xx_read_master_slave function. Signed-off-by:
Arun Ramadoss <arun.ramadoss@microchip.com> Reviewed-by:
Andrew Lunn <andrew@lunn.ch> Signed-off-by:
-
Arun Ramadoss authored
genphy_read_master_slave function allows to configure the master/slave for gigabit phys only. In order to use this function irrespective of speed, moved the speed check to the genphy_read_status call. Signed-off-by:
-
Tobias Waldekranz authored
In this situation (VLAN filtering disabled on br0): br0.10 / br0 / \ swp0 swp1 When a frame is transmitted from the VLAN upper, the bridge will send it down to one of the switch ports with forward offloading enabled. This will cause tag_dsa to generate a FORWARD tag. Before this change, that tag would have it's VID set to 10, even though VID 10 is not loaded in the VTU. Before the blamed commit, the frame would trigger a VTU miss and be forwarded according to the PVT configuration. Now that all fabric ports are in 802.1Q secure mode, the frame is dropped instead. Therefore, restrict the condition under which we rewrite an 802.1Q tag to a DSA tag. On standalone port's, reuse is always safe since we will always generate FROM_CPU tags in that case. For bridged ports though, we must ensure that VLAN filtering is enabled, which in turn guarantees that the VID in question is loaded into the VTU. Fixes: d352b20f ("net: dsa: mv88e6xxx: Improve multichip isolation of standalone ports") Signed-off-by:Tobias Waldekranz <tobias@waldekranz.com> Tested-by:
Vladimir Oltean <olteanv@gmail.com> Link: https://lore.kernel.org/r/20220307110548.812455-1-tobias@waldekranz.comSigned-off-by:
-
Horatiu Vultur authored
In case the MAC is using 'netif_rx()' to deliver the skb up the network stack, it needs to check whether 'skb_defer_rx_timestmap()' is necessary or not. In case is needed then don't call 'netif_rx()' Signed-off-by:
Horatiu Vultur <horatiu.vultur@microchip.com> Link: https://lore.kernel.org/r/20220307094632.3764266-1-horatiu.vultur@microchip.comSigned-off-by:
-
Heiner Kallweit authored
In case of DT-configured systems it may be hard to identify the PHY interrupt in the /proc/interrupts output. Therefore add the name to the id to make clearer that it's about a device on a muxed mdio bus. In my case: Now: mdio_mux-0.e40908ff:08 Before: 0.e40908ff:08 Signed-off-by:
Heiner Kallweit <hkallweit1@gmail.com> Reviewed-by:
-
Zheyu Ma authored
During the process of driver probing, probe function should return < 0 for failure, otherwise kernel will treat value >= 0 as success. Therefore, the driver should set 'err' to -ENODEV when 'adapter->registered_device_map' is NULL. Otherwise kernel will assume that the driver has been successfully probed and will cause unexpected errors. Signed-off-by:
Zheyu Ma <zheyuma97@gmail.com> Reviewed-by:
-
Maciej Fijalkowski authored
Pragma unroll was introduced around GCC 8, whereas current xsk code in ice that prepares loop_unrolled_for macro that is based on mentioned pragma, compares GCC version against 4, which is wrong and Stephen found this out by compiling kernel with GCC 5.4 [0]. Fix this mistake and check if GCC version is >= 8. [0]: https://lore.kernel.org/netdev/20220307213659.47658125@canb.auug.org.au/ Fixes: 126cdfe1 ("ice: xsk: Improve AF_XDP ZC Tx and use batching API") Reported-by:
Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by:
Maciej Fijalkowski <maciej.fijalkowski@intel.com> Link: https://lore.kernel.org/r/20220307231353.56638-1-maciej.fijalkowski@intel.comSigned-off-by:
-
Dan Carpenter authored
The previous check handled the "if (!nh)" condition so we know "nh" is non-NULL here. Delete the check and pull the code in one tab. Signed-off-by:
Dan Carpenter <dan.carpenter@oracle.com> Reviewed-by:
Nikolay Aleksandrov <razor@blackwall.org> Reviewed-by:
Roopa Prabhu <roopa@nvidia.com> Link: https://lore.kernel.org/r/20220307125735.GC16710@kiliSigned-off-by:
-
- 07 Mar, 2022 5 commits
-
-
Dan Carpenter authored
The > ARRAY_SIZE() needs to be >= ARRAY_SIZE() to prevent an out of bounds access. Fixes: 9f492c4c ("ptp: ocp: add TOD debug information") Signed-off-by:
Richard Cochran <richardcochran@gmail.com> Acked-by:
-
Tom Rix authored
The clang static analyzer reports this issue rtnetlink.c:5481:2: warning: Undefined or garbage value returned to caller return err; ^~~~~~~~~~ There is a function level err variable, in the list_for_each_entry_rcu block there is a shadow err. Remove the shadow. In the same block, the call to nla_nest_start_noflag() can fail without setting an err. Set the err to -EMSGSIZE. Fixes: 216e6906 ("net: rtnetlink: rtnl_fill_statsinfo(): Permit non-EMSGSIZE error returns") Signed-off-by:
Tom Rix <trix@redhat.com> Reviewed-by:
Ido Schimmel <idosch@nvidia.com> Reviewed-by:
Petr Machata <petrm@nvidia.com> Signed-off-by:
-
Tom Rix authored
Clang static analysis reports this representative issue dsa.c:486:2: warning: Undefined or garbage value returned to caller return err; ^~~~~~~~~~ err is only set in the loop. If the loop is empty, garbage will be returned. So initialize err to 0 to handle this noop case. Signed-off-by:
-
Guo Zhengkui authored
Fit the following coccicheck warning: tools/testing/selftests/net/reuseport_bpf_numa.c:89:28-29: WARNING: Use ARRAY_SIZE. It has been tested with gcc (Debian 8.3.0-6) 8.3.0 on x86_64. Signed-off-by:
Guo Zhengkui <guozhengkui@vivo.com> Signed-off-by:
-
Guo Zhengkui authored
Fix the following coccicheck warning: drivers/net/ethernet/netronome/nfp/nfp_net_common.c:3434:8-48: WARNING avoid newline at end of message in NL_SET_ERR_MSG_MOD Signed-off-by:
Niklas Söderlund <niklas.soderlund@corigine.com> Signed-off-by:
-
