summaryrefslogtreecommitdiffstats
path: root/net
Commit message (Collapse)AuthorAgeFilesLines
* rxrpc: Kill off the rxrpc_transport structDavid Howells2016-06-2211-451/+65Star
| | | | | | | | | | | The rxrpc_transport struct is now redundant, given that the rxrpc_peer struct is now per peer port rather than per peer host, so get rid of it. Service connection lists are transferred to the rxrpc_peer struct, as is the conn_lock. Previous patches moved the client connection handling out of the rxrpc_transport struct and discarded the connection bundling code. Signed-off-by: David Howells <dhowells@redhat.com>
* rxrpc: Kill the client connection bundle conceptDavid Howells2016-06-228-480/+288Star
| | | | | | | | | | | | Kill off the concept of maintaining a bundle of connections to a particular target service to increase the number of call slots available for any beyond four for that service (there are four call slots per connection). This will make cleaning up the connection handling code easier and facilitate removal of the rxrpc_transport struct. Bundling can be reintroduced later if necessary. Signed-off-by: David Howells <dhowells@redhat.com>
* rxrpc: Provide more refcount helper functionsDavid Howells2016-06-228-18/+26
| | | | | | | | | | Provide refcount helper functions for connections so that the code doesn't touch local or connection usage counts directly. Also make it such that local and peer put functions can take a NULL pointer. Signed-off-by: David Howells <dhowells@redhat.com>
* rxrpc: Make rxrpc_send_packet() take a connection not a transportDavid Howells2016-06-223-24/+31
| | | | | | | | | | Make rxrpc_send_packet() take a connection not a transport as part of the phasing out of the rxrpc_transport struct. Whilst we're at it, rename the function to rxrpc_send_data_packet() to differentiate it from the other packet sending functions. Signed-off-by: David Howells <dhowells@redhat.com>
* rxrpc: Calls displayed in /proc may in future lack a connectionDavid Howells2016-06-221-8/+11
| | | | | | | | | | | | | | | | | Allocated rxrpc calls displayed in /proc/net/rxrpc_calls may in future be on the proc list before they're connected or after they've been disconnected - in which case they may not have a pointer to a connection struct that can be used to get data from there. Deal with this by using stuff from the call struct in preference where possible and printing "no_connection" rather than a peer address if no connection is assigned. This change also has the added bonus that the service ID is now taken from the call rather the connection which will allow per-call service upgrades to be shown - something required for AuriStor server compatibility. Signed-off-by: David Howells <dhowells@redhat.com>
* rxrpc: Validate the net address given to rxrpc_kernel_begin_call()David Howells2016-06-221-0/+5
| | | | | | | | | | | Validate the net address given to rxrpc_kernel_begin_call() before using it. Whilst this should be mostly unnecessary for in-kernel users, it does clear the tail of the address struct in case we want to hash or compare the whole thing. Signed-off-by: David Howells <dhowells@redhat.com>
* rxrpc: Use IDR to allocate client conn IDs on a machine-wide basisDavid Howells2016-06-226-152/+196
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Use the IDR facility to allocate client connection IDs on a machine-wide basis so that each client connection has a unique identifier. When the connection ID space wraps, we advance the epoch by 1, thereby effectively having a 62-bit ID space. The IDR facility is then used to look up client connections during incoming packet routing instead of using an rbtree rooted on the transport. This change allows for the removal of the transport in the future and also means that client connections can be looked up directly in the data-ready handler by connection ID. The ID management code is placed in a new file, conn-client.c, to which all the client connection-specific code will eventually move. Note that the IDR tree gets very expensive on memory if the connection IDs are widely scattered throughout the number space, so we shall need to retire connections that have, say, an ID more than four times the maximum number of client conns away from the current allocation point to try and keep the IDs concentrated. We will also need to retire connections from an old epoch. Also note that, for the moment, a pointer to the transport has to be passed through into the ID allocation function so that we can take a BH lock to prevent a locking issue against in-BH lookup of client connections. This will go away later when RCU is used for server connections also. Signed-off-by: David Howells <dhowells@redhat.com>
* rxrpc: rxrpc_connection_lock shouldn't be a BH lock, but conn_lock isDavid Howells2016-06-221-8/+8
| | | | | | | | | | rxrpc_connection_lock shouldn't be accessed as a BH-excluding lock. It's only accessed in a few places and none of those are in BH-context. rxrpc_transport::conn_lock, however, *is* a BH-excluding lock and should be accessed so consistently. Signed-off-by: David Howells <dhowells@redhat.com>
* rxrpc: Pass sk_buff * rather than rxrpc_host_header * to functionsDavid Howells2016-06-225-43/+44
| | | | | | | | Pass a pointer to struct sk_buff rather than struct rxrpc_host_header to functions so that they can in the future get at transport protocol parameters rather than just RxRPC parameters. Signed-off-by: David Howells <dhowells@redhat.com>
* rxrpc: Fix exclusive connection handlingDavid Howells2016-06-224-72/+58Star
| | | | | | | | | | | | | | | | | | | | | | | | "Exclusive connections" are meant to be used for a single client call and then scrapped. The idea is to limit the use of the negotiated security context. The current code, however, isn't doing this: it is instead restricting the socket to a single virtual connection and doing all the calls over that. This is changed such that the socket no longer maintains a special virtual connection over which it will do all the calls, but rather gets a new one each time a new exclusive call is made. Further, using a socket option for this is a poor choice. It should be done on sendmsg with a control message marker instead so that calls can be marked exclusive individually. To that end, add RXRPC_EXCLUSIVE_CALL which, if passed to sendmsg() as a control message element, will cause the call to be done on an single-use connection. The socket option (RXRPC_EXCLUSIVE_CONNECTION) still exists and, if set, will override any lack of RXRPC_EXCLUSIVE_CALL being specified so that programs using the setsockopt() will appear to work the same. Signed-off-by: David Howells <dhowells@redhat.com>
* rxrpc: Replace conn->trans->{local,peer} with conn->params.{local,peer}David Howells2016-06-229-46/+44Star
| | | | | | | | | | Replace accesses of conn->trans->{local,peer} with conn->params.{local,peer} thus making it easier for a future commit to remove the rxrpc_transport struct. This also reduces the number of memory accesses involved. Signed-off-by: David Howells <dhowells@redhat.com>
* rxrpc: Use structs to hold connection params and protocol infoDavid Howells2016-06-2213-163/+226
| | | | | | | | | | | | | | Define and use a structure to hold connection parameters. This makes it easier to pass multiple connection parameters around. Define and use a structure to hold protocol information used to hash a connection for lookup on incoming packet. Most of these fields will be disposed of eventually, including the duplicate local pointer. Whilst we're at it rename "proto" to "family" when referring to a protocol family. Signed-off-by: David Howells <dhowells@redhat.com>
* rxrpc: fix uninitialized variable useArnd Bergmann2016-06-221-0/+3
| | | | | | | | | | | | | | | | | Hashing the peer key was introduced for AF_INET, but gcc warns about the rxrpc_peer_hash_key function returning uninitialized data for any other value of srx->transport.family: net/rxrpc/peer_object.c: In function 'rxrpc_peer_hash_key': net/rxrpc/peer_object.c:57:15: error: 'p' may be used uninitialized in this function [-Werror=maybe-uninitialized] Assuming that nothing else can be set here, this changes the function to just return zero in case of an unknown address family. Fixes: be6e6707f6ee ("rxrpc: Rework peer object handling to use hash table and RCU") Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: David Howells <dhowells@redhat.com>
* rxrpc: checking for IS_ERR() instead of NULLDan Carpenter2016-06-223-4/+4
| | | | | | | | | | | rxrpc_lookup_peer_rcu() and rxrpc_lookup_peer() return NULL on error, never error pointers, so IS_ERR() can't be used. Fix three callers of those functions. Fixes: be6e6707f6ee ('rxrpc: Rework peer object handling to use hash table and RCU') Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: David Howells <dhowells@redhat.com>
* ipv6: RFC 4884 partial support for SIT/GRE tunnelsEric Dumazet2016-06-193-6/+31
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | When receiving an ICMPv4 message containing extensions as defined in RFC 4884, and translating it to ICMPv6 at SIT or GRE tunnel, we need some extra manipulation in order to properly forward the extensions. This patch only takes care of Time Exceeded messages as they are the ones that typically carry information from various routers in a fabric during a traceroute session. It also avoids complex skb logic if the data_len is not a multiple of 8. RFC states : The "original datagram" field MUST contain at least 128 octets. If the original datagram did not contain 128 octets, the "original datagram" field MUST be zero padded to 128 octets. In practice routers use 128 bytes of original datagram, not more. Initial translation was added in commit ca15a078bd90 ("sit: generate icmpv6 error when receiving icmpv4 error") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Oussama Ghorbel <ghorbel@pivasoftware.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* gre: better support for ICMP messages for gre+ipv6Eric Dumazet2016-06-192-0/+7
| | | | | | | | | ipgre_err() can call ip6_err_gen_icmpv6_unreach() for proper support of ipv4+gre+icmp+ipv6+... frames, used for example by traceroute/mtr. Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* ipv6: translate ICMP_TIME_EXCEEDED to ICMPV6_TIME_EXCEEDEric Dumazet2016-06-192-6/+12
| | | | | | | | | | | For better traceroute/mtr support for SIT and GRE tunnels, we translate IPV4 ICMP ICMP_TIME_EXCEEDED to ICMPV6_TIME_EXCEED We also have to translate the IPv4 source IP address of ICMP message to IPv6 v4mapped. Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* ip6: move ipip6_err_gen_icmpv6_unreach()Eric Dumazet2016-06-192-37/+40
| | | | | | | | | | | We want to use this helper from GRE as well, so this is the time to move it in net/ipv6/icmp.c Also add a @nhs parameter, since SIT and GRE have different values for the header(s) to skip. Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* ipv6: icmp: add a force_saddr param to icmp6_send()Eric Dumazet2016-06-192-3/+6
| | | | | | | | | | | SIT or GRE tunnels might want to translate an IPV4 address into a v4mapped one when translating ICMP to ICMPv6. This patch adds the parameter to icmp6_send() but does not change icmpv6_send() signature. Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* RDS: TCP: Fix non static symbol warningsWei Yongjun2016-06-181-2/+2
| | | | | | | | | | | | | | Fixes the following sparse warnings: net/rds/tcp.c:59:5: warning: symbol 'rds_tcp_min_sndbuf' was not declared. Should it be static? net/rds/tcp.c:60:5: warning: symbol 'rds_tcp_min_rcvbuf' was not declared. Should it be static? Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn> Acked-by: Sowmini Varadhan <sowmini.varadhan@oracle.com> Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* Merge tag 'linux-can-next-for-4.8-20160617' of ↵David S. Miller2016-06-184-133/+193
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can-next Marc Kleine-Budde says: ==================== pull-request: can-next 2016-06-17 this is a pull request of 14 patches for net-next/master. Geert Uytterhoeven contributes a patch that adds a file patterns for CAN device tree bindings to MAINTAINERS. A patch by Alexander Aring fixes warnings when building without proc support. A patch by me improves the sample point calculation. Marek Vasut's patch converts the slcan driver to use CAN_MTU. A patch by William Breathitt Gray converts the tscan1 driver to use module_isa_driver. Two patches by Maximilian Schneider for the gs_usb driver fix coding style and add support for set_phys_id callback. 5 patches by Oliver Hartkopp add support for CANFD to the bcm. And finally two patches by Ramesh Shanmugasundaram, which add support for the rcar_canfd driver. ==================== Signed-off-by: David S. Miller <davem@davemloft.net>
| * can: bcm: add support for CAN FD framesOliver Hartkopp2016-06-171-88/+135
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The programming API of the CAN_BCM depends on struct can_frame which is given as array directly behind the bcm_msg_head structure. To follow this schema for the CAN FD frames a new flag 'CAN_FD_FRAME' in the bcm_msg_head flags indicates that the concatenated CAN frame structures behind the bcm_msg_head are defined as struct canfd_frame. This patch adds the support to handle CAN and CAN FD frames on a per BCM-op base. Main changes: - generally use struct canfd_frames instead if struct can_frames - use canfd_frame.flags instead of can_frame.can_dlc for private BCM flags - make all CAN frame sizes depending on the new CAN_FD_FRAME flags - separate between CAN and CAN FD when sending/receiving frames Due to the dependence of the CAN_FD_FRAME flag the former binary interface for classic CAN frames remains stable. Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net> Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
| * can: bcm: unify bcm_msg_head handling and prepare function parametersOliver Hartkopp2016-06-171-25/+29
| | | | | | | | | | Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net> Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
| * can: bcm: use CAN frame instead of can_frame in commentsOliver Hartkopp2016-06-171-16/+16
| | | | | | | | | | | | | | | | can_frame is the name of the struct can_frame which is not meant in the corrected comments. Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net> Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
| * can: bcm: fix indention and other minor style issuesOliver Hartkopp2016-06-171-14/+12Star
| | | | | | | | | | Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net> Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
| * can: build proc support only if CONFIG_PROC_FS is activatedAlexander Aring2016-06-173-3/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | When building can subsystem with CONFIG_PROC_FS=n I detected some unused variables warning by using proc functions. In CAN the proc handling is nicely placed in one object file. This patch adds simple add a dependency on CONFIG_PROC_FS for CAN's proc.o file and corresponding static inline no-op functions. Signed-off-by: Alexander Aring <aar@pengutronix.de> Acked-by: Oliver Hartkopp <socketcan@hartkopp.net> [mkl: provide static inline noops instead of using #ifdefs] Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
* | tipc: potential shift wrapping bug in map_set()Dan Carpenter2016-06-181-2/+2
| | | | | | | | | | | | | | | | "up_map" is a u64 type but we're not using the high 32 bits. Fixes: 35c55c9877f8 ('tipc: add neighbor monitoring framework') Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | net: ipv6: Address selection needs to consider L3 domainsDavid Ahern2016-06-181-0/+48
| | | | | | | | | | | | | | | | | | | | | | | | | | IPv6 version of 3f2fb9a834cb ("net: l3mdev: address selection should only consider devices in L3 domain") and the follow up commit, a17b693cdd876 ("net: l3mdev: prefer VRF master for source address selection"). That is, if outbound device is given then the address preference order is an address from that device, an address from the master device if it is enslaved, and then an address from a device in the same L3 domain. Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | net: vrf: Implement get_saddr for IPv6David Ahern2016-06-182-2/+34
| | | | | | | | | | | | | | | | | | | | | | | | | | IPv6 source address selection needs to consider the real egress route. Similar to IPv4 implement a get_saddr6 method which is called if source address has not been set. The get_saddr6 method does a full lookup which means pulling a route from the VRF FIB table and properly considering linklocal/multicast destination addresses. Lookup failures (eg., unreachable) then cause the source address selection to fail which gets propagated back to the caller. Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | net: ipv6: Move ip6_route_get_saddr to inlineDavid Ahern2016-06-181-17/+0Star
| | | | | | | | | | | | | | | | | | | | | | | | VRF driver needs access to ip6_route_get_saddr code. Since it does little beyond ipv6_dev_get_saddr and ipv6_dev_get_saddr is already exported for modules move ip6_route_get_saddr to the header as an inline. Code move only; no functional change. Signed-off-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | net: Remove deprecated tunnel specific UDP offload functionsAlexander Duyck2016-06-181-65/+14Star
| | | | | | | | | | | | | | | | | | | | Now that we have all the drivers using udp_tunnel_get_rx_ports, ndo_add_udp_enc_rx_port, and ndo_del_udp_enc_rx_port we can drop the function calls that were specific to VXLAN and GENEVE. Signed-off-by: Alexander Duyck <aduyck@mirantis.com> Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: David S. Miller <davem@davemloft.net>
* | net: Merge VXLAN and GENEVE push notifiers into a single notifierAlexander Duyck2016-06-181-0/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch merges the notifiers for VXLAN and GENEVE into a single UDP tunnel notifier. The idea is that we will want to only have to make one notifier call to receive the list of ports for VXLAN and GENEVE tunnels that need to be offloaded. In addition we add a new set of ndo functions named ndo_udp_tunnel_add and ndo_udp_tunnel_del that are meant to allow us to track the tunnel meta-data such as port and address family as tunnels are added and removed. The tunnel meta-data is now transported in a structure named udp_tunnel_info which for now carries the type, address family, and port number. In the future this could be updated so that we can include a tuple of values including things such as the destination IP address and other fields. I also ended up going with a naming scheme that consisted of using the prefix udp_tunnel on function names. I applied this to the notifier and ndo ops as well so that it hopefully points to the fact that these are primarily used in the udp_tunnel functions. Signed-off-by: Alexander Duyck <aduyck@mirantis.com> Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: David S. Miller <davem@davemloft.net>
* | net: Combine GENEVE and VXLAN port notifiers into single functionsAlexander Duyck2016-06-181-0/+102
|/ | | | | | | | | | | This patch merges the GENEVE and VXLAN code so that both functions pass through a shared code path. This way we can start the effort of using a single function on the network device drivers to handle both of these tunnel types. Signed-off-by: Alexander Duyck <aduyck@mirantis.com> Acked-by: Hannes Frederic Sowa <hannes@stressinduktion.org> Signed-off-by: David S. Miller <davem@davemloft.net>
* net, cls: also reject deleting all filters when TCA_KIND presentDaniel Borkmann2016-06-171-1/+1
| | | | | | | | | | | | When we check for RTM_DELTFILTER, we should also reject the request for deleting all filters under a given parent when TCA_KIND attribute is present. If present, it's currently just ignored but there's also no point to let it pass in the first place either since this doesn't have any meaning with wild-card removal. Fixes: ea7f8277f907 ("net, cls: allow for deleting all filters for given parent") Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: David S. Miller <davem@davemloft.net>
* net: xfrm: fix old-style declarationArnd Bergmann2016-06-172-6/+6
| | | | | | | | | | | | | | | | | | Modern C standards expect the '__inline__' keyword to come before the return type in a declaration, and we get a couple of warnings for this with "make W=1" in the xfrm{4,6}_policy.c files: net/ipv6/xfrm6_policy.c:369:1: error: 'inline' is not at beginning of declaration [-Werror=old-style-declaration] static int inline xfrm6_net_sysctl_init(struct net *net) net/ipv6/xfrm6_policy.c:374:1: error: 'inline' is not at beginning of declaration [-Werror=old-style-declaration] static void inline xfrm6_net_sysctl_exit(struct net *net) net/ipv4/xfrm4_policy.c:339:1: error: 'inline' is not at beginning of declaration [-Werror=old-style-declaration] static int inline xfrm4_net_sysctl_init(struct net *net) net/ipv4/xfrm4_policy.c:344:1: error: 'inline' is not at beginning of declaration [-Werror=old-style-declaration] static void inline xfrm4_net_sysctl_exit(struct net *net) Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: David S. Miller <davem@davemloft.net>
* net: the space is required after ','Wei Tang2016-06-171-6/+6
| | | | | | | | The space is missing after ',', and this will introduce much more noise when checking patch around. Signed-off-by: Wei Tang <tangwei@cmss.chinamobile.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* net: do not initialise statics to 0Wei Tang2016-06-171-1/+1
| | | | | | | | | This patch fixes the checkpatch.pl error to dev.c: ERROR: do not initialise statics to 0 Signed-off-by: Wei Tang <tangwei@cmss.chinamobile.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* mpls: allow routes on ipgre devicesSimon Horman2016-06-171-3/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This appears to be necessary and sufficient to provide MPLS in GRE (RFC4023) support. This can be used by establishing an ipgre tunnel device and then routing MPLS over it. The following example will forward MPLS frames received with an outermost MPLS label 100 over tun1, a GRE tunnel. The forwarded packet will have the outermost MPLS LSE removed and two new LSEs added with labels 200 (outermost) and 300 (next). ip link add name tun1 type gre remote 10.0.99.193 local 10.0.99.192 ttl 225 ip link set up dev tun1 ip addr add 10.0.98.192/24 dev tun1 ip route sh echo 1 > /proc/sys/net/mpls/conf/eth0/input echo 101 > /proc/sys/net/mpls/platform_labels ip -f mpls route add 100 as 200/300 via inet 10.0.98.193 ip -f mpls route sh Also remove unnecessary braces. Reviewed-by: Dinan Gunawardena <dinan.gunawardena@netronome.com> Signed-off-by: Simon Horman <simon.horman@netronome.com> Acked-by: Robert Shearman <rshearma@brocade.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* sctp: change sk state to CLOSED instead of CLOSING in sctp_sock_migrateXin Long2016-06-161-1/+1
| | | | | | | | | | | | | Commit d46e416c11c8 ("sctp: sctp should change socket state when shutdown is received") may set sk_state CLOSING in sctp_sock_migrate, but inet_accept doesn't allow the sk_state other than ESTABLISHED/ CLOSED for sctp. So we will change sk_state to CLOSED, instead of CLOSING, as actually sk is closed already there. Fixes: d46e416c11c8 ("sctp: sctp should change socket state when shutdown is received") Reported-by: Ye Xiaolong <xiaolong.ye@intel.com> Signed-off-by: Xin Long <lucien.xin@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* Merge tag 'rxrpc-rewrite-20160615' of ↵David S. Miller2016-06-1615-647/+842
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs David Howells says: ==================== rxrpc: Rework endpoint record handling Here's the next part of the AF_RXRPC rewrite. In this set I rework endpoint record handling. There are two types of endpoint record, local and peer. The local endpoint record is used as an anchor for the transport socket that AF_RXRPC uses (at the moment a UDP socket). Local endpoints can be shared between AF_RXRPC sockets under certain restricted circumstances. The peer endpoint is a record of the remote end. It is (or will be) used to keep track MTU and RTT values and, with these changes, is used to find the call(s) to abort when a network error occurs. The following significant changes are made: (1) The local endpoint event handling code is split out into its own file. (2) The local endpoint list bottom half-excluding spinlock is removed as things are arranged such that sk_user_data will not change whilst the transport socket callbacks are in progress. (3) Local endpoints can now only be shared if they have the same transport address (as before) and have a local service ID of 0 (ie. they're not listening for incoming calls). This prevents callbacks from a server to one process being picked up by another process. (4) Local endpoint destruction is now accomplished by the same work item as processes events, meaning that the destructor doesn't need to wait for the event processor. (5) Peer endpoints are now held in a hash table rather than a flat list. (6) Peer endpoints are now destroyed by RCU rather than by work item. (7) Peer endpoints are now differentiated by local endpoint and remote transport port in addition to remote transport address and transport type and family. This means that a firewall that excludes access between a particular local port and remote port won't cause calls to be aborted that use a different port pair. (8) Error report handling now no longer assumes that the source is always an IPv4 ICMP message from a UDP port and has assumptions that an ICMP message comes from an IPv4 socket removed. At some point IPv6 support will be added. (9) Peer endpoints rather than local endpoints are now the anchor point for distributing network error reports. (10) Both types of endpoint records are now disposed of as soon as all references to them are gone. There is less hanging around and once their usage counts hit zero, records can no longer be resurrected. ==================== Signed-off-by: David S. Miller <davem@davemloft.net>
| * rxrpc: Rework local endpoint managementDavid Howells2016-06-157-230/+276
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Rework the local RxRPC endpoint management. Local endpoint objects are maintained in a flat list as before. This should be okay as there shouldn't be more than one per open AF_RXRPC socket (there can be fewer as local endpoints can be shared if their local service ID is 0 and they share the same local transport parameters). Changes: (1) Local endpoints may now only be shared if they have local service ID 0 (ie. they're not being used for listening). This prevents a scenario where process A is listening of the Cache Manager port and process B contacts a fileserver - which may then attempt to send CM requests back to B. But if A and B are sharing a local endpoint, A will get the CM requests meant for B. (2) We use a mutex to handle lookups and don't provide RCU-only lookups since we only expect to access the list when opening a socket or destroying an endpoint. The local endpoint object is pointed to by the transport socket's sk_user_data for the life of the transport socket - allowing us to refer to it directly from the sk_data_ready and sk_error_report callbacks. (3) atomic_inc_not_zero() now exists and can be used to only share a local endpoint if the last reference hasn't yet gone. (4) We can remove rxrpc_local_lock - a spinlock that had to be taken with BH processing disabled given that we assume sk_user_data won't change under us. (5) The transport socket is shut down before we clear the sk_user_data pointer so that we can be sure that the transport socket's callbacks won't be invoked once the RCU destruction is scheduled. (6) Local endpoints have a work item that handles both destruction and event processing. The means that destruction doesn't then need to wait for event processing. The event queues can then be cleared after the transport socket is shut down. (7) Local endpoints are no longer available for resurrection beyond the life of the sockets that had them open. As soon as their last ref goes, they are scheduled for destruction and may not have their usage count moved from 0. Signed-off-by: David Howells <dhowells@redhat.com>
| * rxrpc: Separate local endpoint event handling out into its own fileDavid Howells2016-06-154-102/+129
| | | | | | | | | | | | | | | | Separate local endpoint event handling out into its own file preparatory to overhauling the object management aspect (which remains in the original file). Signed-off-by: David Howells <dhowells@redhat.com>
| * rxrpc: Use the peer record to distribute network errorsDavid Howells2016-06-157-94/+79Star
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Use the peer record to distribute network errors rather than the transport object (which I want to get rid of). An error from a particular peer terminates all calls on that peer. For future consideration: (1) For ICMP-induced errors it might be worth trying to extract the RxRPC header from the offending packet, if one is returned attached to the ICMP packet, to better direct the error. This may be overkill, though, since an ICMP packet would be expected to be relating to the destination port, machine or network. RxRPC ABORT and BUSY packets give notice at RxRPC level. (2) To also abort connection-level communications (such as CHALLENGE packets) where indicted by an error - but that requires some revamping of the connection event handling first. Signed-off-by: David Howells <dhowells@redhat.com>
| * rxrpc: Do a little bit of tidying in the ICMP processingDavid Howells2016-06-151-4/+2Star
| | | | | | | | | | | | | | Do a little bit of tidying in the ICMP processing code. Signed-off-by: David Howells <dhowells@redhat.com>
| * rxrpc: Don't assume anything about the address in an ICMP packetDavid Howells2016-06-151-8/+0Star
| | | | | | | | | | | | | | | | Don't assume anything about the address in an ICMP packet in rxrpc_error_report() as the address may not be IPv4 in future, especially since we're just printing these details. Signed-off-by: David Howells <dhowells@redhat.com>
| * rxrpc: Break MTU determination from ICMP into its own functionDavid Howells2016-06-151-39/+54
| | | | | | | | | | | | | | Break MTU determination from ICMP out into its own function to reduce the complexity of the error report handler. Signed-off-by: David Howells <dhowells@redhat.com>
| * rxrpc: Rename rxrpc_UDP_error_report() to rxrpc_error_report()David Howells2016-06-153-4/+4
| | | | | | | | | | | | | | Rename rxrpc_UDP_error_report() to rxrpc_error_report() as it might get called for something other than UDP. Signed-off-by: David Howells <dhowells@redhat.com>
| * rxrpc: Rework peer object handling to use hash table and RCUDavid Howells2016-06-159-203/+335
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Rework peer object handling to use a hash table instead of a flat list and to use RCU. Peer objects are no longer destroyed by passing them to a workqueue to process, but rather are just passed to the RCU garbage collector as kfree'able objects. The hash function uses the local endpoint plus all the components of the remote address, except for the RxRPC service ID. Peers thus represent a UDP port on the remote machine as contacted by a UDP port on this machine. The RCU read lock is used to handle non-creating lookups so that they can be called from bottom half context in the sk_error_report handler without having to lock the hash table against modification. rxrpc_lookup_peer_rcu() *does* take a reference on the peer object as in the future, this will be passed to a work item for error distribution in the error_report path and this function will cease being used in the data_ready path. Creating lookups are done under spinlock rather than mutex as they might be set up due to an external stimulus if the local endpoint is a server. Captured network error messages (ICMP) are handled with respect to this struct and MTU size and RTT are cached here. Signed-off-by: David Howells <dhowells@redhat.com>
* | ila: Fix checksum neutral mappingTom Herbert2016-06-161-5/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The algorithm for checksum neutral mapping is incorrect. This problem was being hidden since we were previously always performing checksum offload on the translated addresses and only with IPv6 HW csum. Enabling an ILA router shows the issue. Corrected algorithm: old_loc is the original locator in the packet, new_loc is the value to overwrite with and is found in the lookup table. old_flag is the old flag value (zero of CSUM_NEUTRAL_FLAG) and new_flag is then (old_flag ^ CSUM_NEUTRAL_FLAG) & CSUM_NEUTRAL_FLAG. Need SUM(new_id + new_flag + diff) == SUM(old_id + old_flag) for checksum neutral translation. Solving for diff gives: diff = (old_id - new_id) + (old_flag - new_flag) compute_csum_diff8(new_id, old_id) gives old_id - new_id If old_flag is set old_flag - new_flag = old_flag = CSUM_NEUTRAL_FLAG Else old_flag - new_flag = -new_flag = ~CSUM_NEUTRAL_FLAG Tested: - Implemented a user space program that creates random addresses and random locators to overwrite. Compares the checksum over the address before and after translation (must always be equal) - Enabled ILA router and showed proper operation. Signed-off-by: Tom Herbert <tom@herbertland.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | net: ipv4: Add ability to have GRE ignore DF bit in IPv4 payloadsPhilip Prindeville2016-06-162-10/+34
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In the presence of firewalls which improperly block ICMP Unreachable (including Fragmentation Required) messages, Path MTU Discovery is prevented from working. A workaround is to handle IPv4 payloads opaquely, ignoring the DF bit--as is done for other payloads like AppleTalk--and doing transparent fragmentation and reassembly. Redux includes the enforcement of mutual exclusion between this feature and Path MTU Discovery as suggested by Alexander Duyck. Cc: Alexander Duyck <alexander.duyck@gmail.com> Reviewed-by: Stephen Hemminger <stephen@networkplumber.org> Signed-off-by: Philip Prindeville <philipp@redfish-solutions.com> Signed-off-by: David S. Miller <davem@davemloft.net>
generated by cgit v1.2.3-55-g7522 (git 2.39.0) at 2024-07-13 03:27:45 +0200