- Nov 01, 2020
-
-
Sebastian Andrzej Siewior authored
nv_update_stats() triggers a WARN_ON() when invoked from hard interrupt context because the locks in use are not hard interrupt safe. It also has an assert_spin_locked() which was the lock check before the lockdep era. Lockdep has way broader locking correctness checks and covers both issues, so replace the warning and the lock assert with lockdep_assert_held(). Signed-off-by:
Sebastian Andrzej Siewior <bigeasy@linutronix.de> Cc: Rain River <rain.1986.08.12@gmail.com> Cc: Zhu Yanjun <zyjzyj2000@gmail.com> Signed-off-by:
Jakub Kicinski <kuba@kernel.org>
-
Sebastian Andrzej Siewior authored
wait_for_cmd_complete() uses in_interrupt() to detect whether it is safe to sleep or not. The usage of in_interrupt() in drivers is phased out and Linus clearly requested that code which changes behaviour depending on context should either be seperated or the context be conveyed in an argument passed by the caller, which usually knows the context. in_interrupt() also is only partially correct because it fails to chose the correct code path when just preemption or interrupts are disabled. Add an argument 'may_block' to both functions and adjust the callers to pass the context information. The following call chains which end up invoking wait_for_cmd_complete() were analyzed to be safe to sleep: s2io_card_up() s2io_set_multicast() init_nic() init_tti() s2io_close() do_s2io_delete_unicast_mc() do_s2io_add_mac() s2io_set_mac_addr() do_s2io_prog_unicast() do_s2io_add_mac() s2io_reset() do_s2io_restore_unicast_mc() do_s2io_add_mc() do_s2io_add_mac() s2io_open() do_s2io_prog_unicast() do_s2io_add_mac() The following call chains which end up invoking wait_for_cmd_complete() were analyzed to be safe to sleep: __dev_set_rx_mode() s2io_set_multicast() s2io_txpic_intr_handle() s2io_link() init_tti() Add a may_sleep argument to wait_for_cmd_complete(), s2io_set_multicast() and init_tti() and hand the context information in from the call sites. Signed-off-by:
-
- Oct 31, 2020
-
-
Jakub Kicinski authored
Vladimir Oltean says: ==================== L2 multicast forwarding for Ocelot switch This series enables the mscc_ocelot switch to forward raw L2 (non-IP) mdb entries as configured by the bridge driver after this patch: https://patchwork.ozlabs.org/project/netdev/patch/20201028233831.610076-1-vladimir.oltean@nxp.com/ ==================== Link: https://lore.kernel.org/r/20201029022738.722794-1-vladimir.oltean@nxp.com Signed-off-by:
-
Vladimir Oltean authored
There is one main difference in mscc_ocelot between IP multicast and L2 multicast. With IP multicast, destination ports are encoded into the upper bytes of the multicast MAC address. Example: to deliver the address 01:00:5E:11:22:33 to ports 3, 8, and 9, one would need to program the address of 00:03:08:11:22:33 into hardware. Whereas for L2 multicast, the MAC table entry points to a Port Group ID (PGID), and that PGID contains the port mask that the packet will be forwarded to. As to why it is this way, no clue. My guess is that not all port combinations can be supported simultaneously with the limited number of PGIDs, and this was somehow an issue for IP multicast but not for L2 multicast. Anyway. Prior to this change, the raw L2 multicast code was bogus, due to the fact that there wasn't really any way to test it using the bridge code. There were 2 issues: - A multicast PGID was allocated for each MDB entry, but it wasn't in fact programmed to hardware. It was dummy. - In fact we don't want to reserve a multicast PGID for every single MDB entry. That would be odd because we can only have ~60 PGIDs, but thousands of MDB entries. So instead, we want to reserve a multicast PGID for every single port combination for multicast traffic. And since we can have 2 (or more) MDB entries delivered to the same port group (and therefore PGID), we need to reference-count the PGIDs. Signed-off-by:
Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by:
Florian Fainelli <f.fainelli@gmail.com> Signed-off-by:
-
Vladimir Oltean authored
This saves a re-classification of the MDB address on deletion. Signed-off-by:
-
Vladimir Oltean authored
It is Not Needed, a comment will suffice. Signed-off-by:
-
Vladimir Oltean authored
Since a helper is available for copying Ethernet addresses, let's use it. Signed-off-by:
-
Vladimir Oltean authored
ocelot.h says: /* MAC table entry types. * ENTRYTYPE_NORMAL is subject to aging. * ENTRYTYPE_LOCKED is not subject to aging. * ENTRYTYPE_MACv4 is not subject to aging. For IPv4 multicast. * ENTRYTYPE_MACv6 is not subject to aging. For IPv6 multicast. */ We don't want the permanent entries added with 'bridge mdb' to be subject to aging. Signed-off-by:
-
Vladimir Oltean authored
When creating a new multicast port group, there is implicit conversion between the __u8 state member of struct br_mdb_entry and the unsigned char flags member of struct net_bridge_port_group. This implicit conversion relies on the fact that MDB_PERMANENT is equal to MDB_PG_FLAGS_PERMANENT. Let's be more explicit and convert the state to flags manually. Signed-off-by:
-
Nikolay Aleksandrov authored
Extend the bridge multicast control and data path to configure routes for L2 (non-IP) multicast groups. The uapi struct br_mdb_entry union u is extended with another variant, mac_addr, which does not change the structure size, and which is valid when the proto field is zero. To be compatible with the forwarding code that is already in place, which acts as an IGMP/MLD snooping bridge with querier capabilities, we need to declare that for L2 MDB entries (for which there exists no such thing as IGMP/MLD snooping/querying), that there is always a querier. Otherwise, these entries would be flooded to all bridge ports and not just to those that are members of the L2 multicast group. Needless to say, only permanent L2 multicast groups can be installed on a bridge port. Signed-off-by:
Nikolay Aleksandrov <nikolay@nvidia.com> Signed-off-by:
-
Jakub Kicinski authored
Edward Cree says: ==================== sfc: EF100 TSO enhancements Support TSO over encapsulation (with GSO_PARTIAL), and over VLANs (which the code already handled but we didn't advertise). Also correct our handling of IPID mangling. I couldn't find documentation of exactly what shaped SKBs we can get given, so patch #2 is slightly guesswork, but when I tested TSO over both underlay and (VxLAN) overlay, the checksums came out correctly, so at least in those cases the edits we're making must be the right ones. Similarly, I'm not 100% sure I've correctly understood how FIXEDID and MANGLEID are supposed to work in patch #3. ==================== Link: https://lore.kernel.org/r/6e1ea05f-faeb-18df-91ef-572445691d89@solarflare.com Signed-off-by:
-
Edward Cree authored
Signed-off-by:
Edward Cree <ecree@solarflare.com> Signed-off-by:
-
Edward Cree authored
AIUI, the NETIF_F_TSO_MANGLEID flag is a signal to the stack that a driver may _need_ to mangle IDs in order to do TSO, and conversely a signal from the stack that the driver is permitted to do so. Since we support both fixed and incrementing IPIDs, we should rely on the SKB_GSO_FIXEDID flag on a per-skb basis, rather than using the MANGLEID feature to make all TSOs fixed-id. Includes other minor cleanups of ef100_make_tso_desc() coding style. Signed-off-by:
-
Edward Cree authored
The NIC only needs to know where the headers it has to edit (TCP and inner and outer IPv4) are, which fits GSO_PARTIAL nicely. It also supports non-PARTIAL offload of UDP tunnels, again just needing to be told the outer transport offset so that it can edit the UDP length field. (It's not clear to me whether the stack will ever use the non-PARTIAL version with the netdev feature flags we're setting here.) Signed-off-by:
-
Edward Cree authored
We need EFX_POPULATE_OWORD_17 for an encap TSO descriptor on EF100. Signed-off-by:
-
Jakub Kicinski authored
Alex Elder says: ==================== net: ipa: minor bug fixes This series fixes several bugs. They are minor, in that the code currently works on supported platforms even without these patches applied, but they're bugs nevertheless and should be fixed. Version 2 improves the commit message for the fourth patch. It also fixes a bug in two spots in the last patch. Both of these changes were suggested by Willem de Bruijn. ==================== Link: https://lore.kernel.org/r/20201028194148.6659-1-elder@linaro.org Signed-off-by:
-
Alex Elder authored
The minimum and maximum limits for resources assigned to a given resource group are programmed in pairs, with the limits for two groups set in a single register. If the number of supported resource groups is odd, only half of the register that defines these limits is valid for the last group; that group has no second group in the pair. Currently we ignore this constraint, and it turns out to be harmless, but it is not guaranteed to be. This patch addresses that, and adds support for programming the 5th resource group's limits. Rework how the resource group limit registers are programmed by having a single function program all group pairs rather than having one function program each pair. Add the programming of the 4-5 resource group pair limits to this function. If a resource group is not supported, pass a null pointer to ipa_resource_config_common() for that group and have that function write zeroes in that case. Tested-by:
Sujit Kautkar <sujitka@chromium.org> Signed-off-by:
Alex Elder <elder@linaro.org> Acked-by:
Willem de Bruijn <willemb@google.com> Signed-off-by:
-
Alex Elder authored
The number of resource groups supported by the hardware can be different for source and destination resources. Determine the number supported for each using separate functions. Make the functions inline end move their definitions into "ipa_reg.h", because they determine whether certain register definitions are valid. Pass just the IPA hardware version as argument. IPA_RESOURCE_GROUP_COUNT represents the maximum number of resource groups the driver supports for any hardware version. Change that symbol to be two separate constants, one for source and the other for destination resource groups. Rename them to end with "_MAX" rather than "_COUNT", to reflect their true purpose. Tested-by:
-
Alex Elder authored
The IPA hardware manages various resources (e.g. descriptors) internally to perform its functions. The resources are grouped, allowing different endpoints to use separate resource pools. This way one group of endpoints can be configured to operate unaffected by the resource use of endpoints in a different group. Endpoints should be assigned to a resource group, but we currently don't do that. Define a new resource_group field in the endpoint configuration data, and use it to assign the proper resource group to use for each AP endpoint. Tested-by:
-
Alex Elder authored
The mask for the RSRC_GRP field in the INIT_RSRC_GRP endpoint initialization register is incorrectly defined for IPA v4.2 (where it is only one bit wide). So we need to fix this. The fix is not straightforward, however. Field masks are passed to functions like u32_encode_bits(), and for that they must be constant. To address this, we define a new inline function that returns the *encoded* value to use for a given RSRC_GRP field, which depends on the IPA version. The caller can then use something like this, to assign a given endpoint resource id 1: u32 offset = IPA_REG_ENDP_INIT_RSRC_GRP_N_OFFSET(endpoint_id); u32 val = rsrc_grp_encoded(ipa->version, 1); iowrite32(val, ipa->reg_virt + offset); The next patch requires this fix. Tested-by: -
Alex Elder authored
At the end of ipa_mem_setup() we write the local packet processing context base register to tell it where the processing context memory is. But we are writing the wrong value. The value written turns out to be the offset of the modem header memory region (assigned earlier in the function). Fix this bug. Tested-by:
-
Moritz Fischer authored
The driver does not implement a shutdown handler which leads to issues when using kexec in certain scenarios. The NIC keeps on fetching descriptors which gets flagged by the IOMMU with errors like this: DMAR: DMAR:[DMA read] Request device [5e:00.0]fault addr fffff000 DMAR: DMAR:[DMA read] Request device [5e:00.0]fault addr fffff000 DMAR: DMAR:[DMA read] Request device [5e:00.0]fault addr fffff000 DMAR: DMAR:[DMA read] Request device [5e:00.0]fault addr fffff000 DMAR: DMAR:[DMA read] Request device [5e:00.0]fault addr fffff000 Signed-off-by:
Moritz Fischer <mdf@kernel.org> Link: https://lore.kernel.org/r/20201028172125.496942-1-mdf@kernel.org Signed-off-by:
-
Robert Hancock authored
The Finisar FCLF8520P2BTL 1000BaseT SFP module uses a Marvel 88E1111 PHY with a modified PHY ID. Add support for this ID using the 88E1111 methods. By default these modules do not have 1000BaseX auto-negotiation enabled, which is not generally desirable with Linux networking drivers. Add handling to enable 1000BaseX auto-negotiation when these modules are used in 1000BaseX mode. Also, some special handling is required to ensure that 1000BaseT auto-negotiation is enabled properly when desired. Based on existing handling in the AMD xgbe driver and the information in the Finisar FAQ: https://www.finisar.com/sites/default/files/resources/an-2036_1000base-t_sfp_faqreve1.pdf Signed-off-by:
Robert Hancock <robert.hancock@calian.com> Reviewed-by:
Russell King <rmk+kernel@armlinux.org.uk> Link: https://lore.kernel.org/r/20201028171540.1700032-1-robert.hancock@calian.com Signed-off-by:
-
Jakub Kicinski authored
Xin Long says: ==================== sctp: Implement RFC6951: UDP Encapsulation of SCTP Description From the RFC: The Main Reasons: o To allow SCTP traffic to pass through legacy NATs, which do not provide native SCTP support as specified in [BEHAVE] and [NATSUPP]. o To allow SCTP to be implemented on hosts that do not provide direct access to the IP layer. In particular, applications can use their own SCTP implementation if the operating system does not provide one. Implementation Notes: UDP-encapsulated SCTP is normally communicated between SCTP stacks using the IANA-assigned UDP port number 9899 (sctp-tunneling) on both ends. There are circumstances where other ports may be used on either end, and it might be required to use ports other than the registered port. Each SCTP stack uses a single local UDP encapsulation port number as the destination port for all its incoming SCTP packets, this greatly simplifies implementation design. An SCTP implementation supporting UDP encapsulation MUST maintain a remote UDP encapsulation port number per destination address for each SCTP association. Again, because the remote stack may be using ports other than the well-known port, each port may be different from each stack. However, because of remapping of ports by NATs, the remote ports associated with different remote IP addresses may not be identical, even if they are associated with the same stack. Because the well-known port might not be used, implementations need to allow other port numbers to be specified as a local or remote UDP encapsulation port number through APIs. Patches: This patchset is using the udp4/6 tunnel APIs to implement the UDP Encapsulation of SCTP with not much change in SCTP protocol stack and with all current SCTP features keeped in Linux Kernel. 1 - 4: Fix some UDP issues that may be triggered by SCTP over UDP. 5 - 7: Process incoming UDP encapsulated packets and ICMP packets. 8 -10: Remote encap port's update by sysctl, sockopt and packets. 11-14: Process outgoing pakects with UDP encapsulated and its GSO. 15-16: Add the part from draft-tuexen-tsvwg-sctp-udp-encaps-cons-03. 17: Enable this feature. Tests: - lksctp-tools/src/func_tests with UDP Encapsulation enabled/disabled: Both make v4test and v6test passed. - sctp-tests with UDP Encapsulation enabled/disabled: repeatability/procdumps/sctpdiag/gsomtuchange/extoverflow/ sctphashtable passed. Others failed as expected due to those "iptables -p sctp" rules. - netperf on lo/netns/virtio_net, with gso enabled/disabled and with ip_checksum enabled/disabled, with UDP Encapsulation enabled/disabled: No clear performance dropped. v1->v2: - Fix some incorrect code in the patches 5,6,8,10,11,13,14,17, suggested by Marcelo. - Append two patches 15-16 to add the Additional Considerations for UDP Encapsulation of SCTP from draft-tuexen-tsvwg-sctp-udp-encaps-cons-03. v2->v3: - remove the cleanup code in patch 2, suggested by Willem. - remove the patch 3 and fix the checksum in the new patch 3 after talking with Paolo, Marcelo and Guillaume. - add 'select NET_UDP_TUNNEL' in patch 4 to solve a compiling error. - fix __be16 type cast warning in patch 8. - fix the wrong endian orders when setting values in 14,16. v3->v4: - add entries in ip-sysctl.rst in patch 7,16, as Marcelo Suggested. - not create udp socks when udp_port is set to 0 in patch 16, as Marcelo noticed. v4->v5: - improve the description for udp_port and encap_port entries in patch 7, 16. - use 0 as the default udp_port. ==================== Link: https://lore.kernel.org/r/cover.1603955040.git.lucien.xin@gmail.com Signed-off-by: -
Xin Long authored
This patch is to enable udp tunneling socks by calling sctp_udp_sock_start() in sctp_ctrlsock_init(), and sctp_udp_sock_stop() in sctp_ctrlsock_exit(). Also add sysctl udp_port to allow changing the listening sock's port by users. Wit this patch, the whole sctp over udp feature can be enabled and used. v1->v2: - Also update ctl_sock udp_port in proc_sctp_do_udp_port() where netns udp_port gets changed. v2->v3: - Call htons() when setting sk udp_port from netns udp_port. v3->v4: - Not call sctp_udp_sock_start() when new_value is 0. - Add udp_port entry in ip-sysctl.rst. v4->v5: - Not call sctp_udp_sock_start/stop() in sctp_ctrlsock_init/exit(). - Improve the description of udp_port in ip-sysctl.rst. Signed-off-by:Xin Long <lucien.xin@gmail.com> Acked-by:
Marcelo Ricardo Leitner <marcelo.leitner@gmail.com> Signed-off-by:
-
Xin Long authored
This is from Section 4 of draft-tuexen-tsvwg-sctp-udp-encaps-cons-03, and it requires responding with an abort chunk with an error cause when the udp source port of the received init chunk doesn't match the encap port of the transport. Signed-off-by:
-
Xin Long authored
This patch is to add the function to make the abort chunk with the error cause for new encapsulation port restart, defined on Section 4.4 in draft-tuexen-tsvwg-sctp-udp-encaps-cons-03. v1->v2: - no change. v2->v3: - no need to call htons() when setting nep.cur_port/new_port. Signed-off-by:
-
Xin Long authored
This one basically does the similar things in sctp_v6_xmit as does for udp4 sock in the last patch, just note that: 1. label needs to be calculated, as it's the param of udp_tunnel6_xmit_skb(). 2. The 'nocheck' param of udp_tunnel6_xmit_skb() is false, as required by RFC. v1->v2: - Use sp->udp_port instead in sctp_v6_xmit(), which is more safe. Signed-off-by: -
Xin Long authored
This patch does what the rfc6951#section-5.3 says for ipv4: "Within the UDP header, the source port MUST be the local UDP encapsulation port number of the SCTP stack, and the destination port MUST be the remote UDP encapsulation port number maintained for the association and the destination address to which the packet is sent (see Section 5.1). Because the SCTP packet is the UDP payload, the length of the UDP packet MUST be the length of the SCTP packet plus the size of the UDP header. The SCTP checksum MUST be computed for IPv4 and IPv6, and the UDP checksum SHOULD be computed for IPv4 and IPv6." Some places need to be adjusted in sctp_packet_transmit(): 1. For non-gso packets, when transport's encap_port is set, sctp checksum has to be done in sctp_packet_pack(), as the outer udp will use ip_summed = CHECKSUM_PARTIAL to do the offload setting for checksum. 2. Delay calling dst_clone() and skb_dst_set() for non-udp packets until sctp_v4_xmit(), as for udp packets, skb_dst_set() is not needed before calling udp_tunnel_xmit_skb(). then in sctp_v4_xmit(): 1. Go to udp_tunnel_xmit_skb() only when transport->encap_port and net->sctp.udp_port both are set, as these are one for dst port and another for src port. 2. For gso packet, SKB_GSO_UDP_TUNNEL_CSUM is set for gso_type, and with this udp checksum can be done in __skb_udp_tunnel_segment() for each segments after the sctp gso. 3. inner_mac_header and inner_transport_header are set, as these will be needed in __skb_udp_tunnel_segment() to find the right headers. 4. df and ttl are calculated, as these are the required params by udp_tunnel_xmit_skb(). 5. nocheck param has to be false, as "the UDP checksum SHOULD be computed for IPv4 and IPv6", says in rfc6951#section-5.3. v1->v2: - Use sp->udp_port instead in sctp_v4_xmit(), which is more safe. Signed-off-by: -
Xin Long authored
sk_setup_caps() was originally called in Commit 90017acc ("sctp: Add GSO support"), as: "We have to refresh this in case we are xmiting to more than one transport at a time" This actually happens in the loop of sctp_outq_flush_transports(), and it shouldn't be tied to gso, so move it out of gso part and before sctp_packet_pack(). Signed-off-by:
-
Xin Long authored
sctp_mtu_payload() is for calculating the frag size before making chunks from a msg. So we should only add udphdr size to overhead when udp socks are listening, as only then sctp can handle the incoming sctp over udp packets and outgoing sctp over udp packets will be possible. Note that we can't do this according to transport->encap_port, as different transports may be set to different values, while the chunks were made before choosing the transport, we could not be able to meet all rfc6951#section-5.6 recommends. v1->v2: - Add udp_port for sctp_sock to avoid a potential race issue, it will be used in xmit path in the next patch. Signed-off-by: -
Xin Long authored
As rfc6951#section-5.4 says: "After finding the SCTP association (which includes checking the verification tag), the UDP source port MUST be stored as the encapsulation port for the destination address the SCTP packet is received from (see Section 5.1). When a non-encapsulated SCTP packet is received by the SCTP stack, the encapsulation of outgoing packets belonging to the same association and the corresponding destination address MUST be disabled." transport encap_port should be updated by a validated incoming packet's udp src port. We save the udp src port in sctp_input_cb->encap_port, and then update the transport in two places: 1. right after vtag is verified, which is required by RFC, and this allows the existent transports to be updated by the chunks that can only be processed on an asoc. 2. right before processing the 'init' where the transports are added, and this allows building a sctp over udp connection by client with the server not knowing the remote encap port. 3. when processing ootb_pkt and creating the temporary transport for the reply pkt. Note that sctp_input_cb->header is removed, as it's not used any more in sctp. v1->v2: - Change encap_port as __be16 for sctp_input_cb. Signed-off-by: -
Xin Long authored
This patch is to implement: rfc6951#section-6.1: Get or Set the Remote UDP Encapsulation Port Number with the param of the struct: struct sctp_udpencaps { sctp_assoc_t sue_assoc_id; struct sockaddr_storage sue_address; uint16_t sue_port; }; the encap_port of sock, assoc or transport can be changed by users, which also means it allows the different transports of the same asoc to have different encap_port value. v1->v2: - no change. v2->v3: - fix the endian warning when setting values between encap_port and sue_port. Signed-off-by: -
Xin Long authored
encap_port is added as per netns/sock/assoc/transport, and the latter one's encap_port inherits the former one's by default. The transport's encap_port value would mostly decide if one packet should go out with udp encapsulated or not. This patch also allows users to set netns' encap_port by sysctl. v1->v2: - Change to define encap_port as __be16 for sctp_sock, asoc and transport. v2->v3: - No change. v3->v4: - Add 'encap_port' entry in ip-sysctl.rst. v4->v5: - Improve the description of encap_port in ip-sysctl.rst. Signed-off-by: -
Xin Long authored
As it says in rfc6951#section-5.5: "When receiving ICMP or ICMPv6 response packets, there might not be enough bytes in the payload to identify the SCTP association that the SCTP packet triggering the ICMP or ICMPv6 packet belongs to. If a received ICMP or ICMPv6 packet cannot be related to a specific SCTP association or the verification tag cannot be verified, it MUST be discarded silently. In particular, this means that the SCTP stack MUST NOT rely on receiving ICMP or ICMPv6 messages. Implementation constraints could prevent processing received ICMP or ICMPv6 messages." ICMP or ICMPv6 packets need to be handled, and this is implemented by udp encap sock .encap_err_lookup function. The .encap_err_lookup function is called in __udp(6)_lib_err_encap() to confirm this path does need to be updated. For sctp, what we can do here is check if the corresponding asoc and transport exist. Note that icmp packet process for sctp over udp is done by udp sock .encap_err_lookup(), and it means for now we can't do as much as sctp_v4/6_err() does. Also we can't do the two mappings mentioned in rfc6951#section-5.5. Signed-off-by:
-
Xin Long authored
This patch is to add the udp6 sock part in sctp_udp_sock_start/stop(). udp_conf.use_udp6_rx_checksums is set to true, as: "The SCTP checksum MUST be computed for IPv4 and IPv6, and the UDP checksum SHOULD be computed for IPv4 and IPv6" says in rfc6951#section-5.3. v1->v2: - Add pr_err() when fails to create udp v6 sock. - Add #if IS_ENABLED(CONFIG_IPV6) not to create v6 sock when ipv6 is disabled. Signed-off-by: -
Xin Long authored
This patch is to add the functions to create/release udp4 sock, and set the sock's encap_rcv to process the incoming udp encap sctp packets. In sctp_udp_rcv(), as we can see, all we need to do is fix the transport header for sctp_rcv(), then it would implement the part of rfc6951#section-5.4: "When an encapsulated packet is received, the UDP header is removed. Then, the generic lookup is performed, as done by an SCTP stack whenever a packet is received, to find the association for the received SCTP packet" Note that these functions will be called in the last patch of this patchset when enabling this feature. v1->v2: - Add pr_err() when fails to create udp v4 sock. v2->v3: - Add 'select NET_UDP_TUNNEL' in sctp Kconfig. v3->v4: - No change. v4->v5: - Change to set udp_port to 0 by default. Signed-off-by:
-
Xin Long authored
For the gso of sctp over udp packets, sctp_gso_segment() will be called in skb_udp_tunnel_segment(), we need to set transport_header to sctp header. As all the current HWs can't handle both crc checksum and udp checksum at the same time, the crc checksum has to be done in sctp_gso_segment() by removing the NETIF_F_SCTP_CRC flag from the features. Meanwhile, if the HW can't do udp checksum, csum and csum_start has to be set correctly, and udp checksum will be done in __skb_udp_tunnel_segment() by calling gso_make_checksum(). Thanks to Paolo, Marcelo and Guillaume for helping with this one. v1->v2: - no change. v2->v3: - remove the he NETIF_F_SCTP_CRC flag from the features. - set csum and csum_start in sctp_gso_make_checksum(). Signed-off-by:
-
Xin Long authored
For some protocol's gso, like SCTP, it's using GSO_BY_FRAGS for gso_size. When using UDP to encapsulate its packet, it will return error in udp6_ufo_fragment() as skb->len < gso_size, and it will never go to the gso tunnel processing. So we should move this check after udp gso tunnel processing, the same as udp4_ufo_fragment() does. v1->v2: - no change. v2->v3: - not do any cleanup. Signed-off-by:
-
Xin Long authored
There is a chance that __udp4/6_lib_lookup() returns a udp encap sock in __udp_lib_err(), like the udp encap listening sock may use the same port as remote encap port, in which case it should go to __udp4/6_lib_err_encap() for more validation before processing the icmp packet. This patch is to check encap_type in __udp_lib_err() for the further validation for a encap sock. Signed-off-by:
-
