summaryrefslogtreecommitdiff
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* ipcpd: Fix use after free and uninitalized valueSander Vrijders2019-06-212-3/+3
| | | | | | | | | | This fixes a use after free in an error condition, and makes sure that pid is set in the flow_set early on, so flow_set_destroy won't create a prefix with an uninitialized pid in case of an error in shm_flow_set_create. Signed-off-by: Sander Vrijders <sander@ouroboros.rocks> Signed-off-by: Dimitri Staessens <dimitri@ouroboros.rocks>
* lib: Add tests for the shm_rbuffDimitri Staessens2019-06-212-0/+106
| | | | | | | | | This adds some tests for the shm_rbuff after some reports that the queue length would be erroneously reported as 0 when the rbuff was full. The test passes for the reported case. Signed-off-by: Dimitri Staessens <dimitri@ouroboros.rocks> Signed-off-by: Sander Vrijders <sander@ouroboros.rocks>
* lib: Fix blocking writes to smaller rbuffDimitri Staessens2019-04-171-5/+4
| | | | | | | | | This fixes writing at high speeds when the rbuff is smaller than the rdrbuff. The pthread_cond_wait calls were blocking on the wrong condition variable. Signed-off-by: Dimitri Staessens <dimitri@ouroboros.rocks> Signed-off-by: Sander Vrijders <sander@ouroboros.rocks>
* lib: Set independent size for rbuffDimitri Staessens2019-04-066-15/+137
| | | | | | | | | | This allows setting the size of the rbuffs in a system independently of the main packet buffer using SHM_RBUFF_SIZE. The benefit of setting a smaller rbuff size is that a single process can't fully occupy the main packet buffer. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander@ouroboros.rocks>
* ipcpd: Give max priority to flow allocatorDimitri Staessens2019-03-281-2/+31
| | | | | | | | | | | The flow allocator thread was running on a low priority, causing some delay when handling packets. Usually this isn't a problem, but for congestion control updates, the delay may become problematic. The priority is now set to the maximum allowed by the scheduler policy to improve responsiveness. Signed-off-by: Dimitri Staessens <dimitri@ouroboros.rocks> Signed-off-by: Sander Vrijders <sander@ouroboros.rocks>
* ipcpd: Handle local IPCP case when parsing configDimitri Staessens2019-03-281-1/+3
| | | | | | | | | The default case in the configuration parsing prints an erroneous "Unknown IPCP type" message when a local IPCP is bootstrapped. The IPCP_LOCAL type is now handled separately (no action needed). Signed-off-by: Dimitri Staessens <dimitri@ouroboros.rocks> Signed-off-by: Sander Vrijders <sander@ouroboros.rocks>
* build: Set specific compiler flags for SWIG target0.15.2Dimitri Staessens2019-03-213-4/+4
| | | | | | | | | | The compiler flags for the SWIG target were added to the global CMAKE_C_FLAGS used for the entire project. This sets the flags uniquely for the SWIG target. The eth has a similar case for the c99 flag. There was a lingering include in dev.c that was removed. Signed-off-by: Dimitri Staessens <dimitri@ouroboros.rocks> Signed-off-by: Sander Vrijders <sander@ouroboros.rocks>
* ipcpd: Don't reuse fqueue between threadsDimitri Staessens2019-03-211-10/+17
| | | | | | | | The UDP IPCP was reusing fqueues between threads. This is not supported and lead to bad reads and buffer overflows. Signed-off-by: Dimitri Staessens <dimitri@ouroboros.rocks> Signed-off-by: Sander Vrijders <sander@ouroboros.rocks>
* ipcpd: Restrict MTU for Ethernet over loopbackDimitri Staessens2019-03-213-0/+8
| | | | | | | | | | | This restricts the MTU for the Ethernet IPCP over loopback adapters (devices named "lo*") to avoid it allocating 65K buffers per packet and quickly filling the default RDRBUFF space. The restriction is set using the build option IPCP_ETH_LO_MTU, with a default value of 1500 bytes. Signed-off-by: Dimitri Staessens <dimitri@ouroboros.rocks> Signed-off-by: Sander Vrijders <sander@ouroboros.rocks>
* ipcpd: Fix strict aliasing warning0.15.1Dimitri Staessens2019-03-181-6/+8
| | | | | | | | | | | | | Some versions of gcc seem to give a positive on the strict aliasing rule. It's absent from newer gcc versions or clang. However, rather than disabling the check for older version, this small rewrite seems to temporarily fix the false positive. Apparently, it's undefined behaviour to simply cast a char/uint8_t buffer to a variable type pointer and then dereference the type. A more elaborate patch to fix the undefined behaviour is needed. Signed-off-by: Dimitri Staessens <dimitri@ouroboros.rocks> Signed-off-by: Sander Vrijders <sander@ouroboros.rocks>
* ipcpd: Revise UDP IPCPDimitri Staessens2019-03-187-446/+421
| | | | | | | | | | | The UDP IPCP now uses a fixed server UDP port (default 3435) for all communications. This allows passing firewalls more easily since only a single port needs to be opened. The client port can be fixed as well if needed (default random). It uses an internal eid, so the MTU of the UDP layer is reduced by 4 bytes, similar to the Ethernet IPCPs. Signed-off-by: Dimitri Staessens <dimitri@ouroboros.rocks> Signed-off-by: Sander Vrijders <sander@ouroboros.rocks>
* irmd: Don't destroy shm_flow_set from IRMdDimitri Staessens2019-03-121-5/+0
| | | | | | | | | The shm_flow_set for the processes are now created and destroyed together with the irm_flow objects. This old code causes stack overflows in some (rare) conditions and must be removed. Signed-off-by: Dimitri Staessens <dimitri@ouroboros.rocks> Signed-off-by: Sander Vrijders <sander@ouroboros.rocks>
* build: Set default number of threads in eth to 1Dimitri Staessens2019-03-051-2/+2
| | | | | | | | | This makes the eth packet handler single-threaded by default, at least until stability issues that seem to be related to multi-threading are fixed. Signed-off-by: Dimitri Staessens <dimitri@ouroboros.rocks> Signed-off-by: Sander Vrijders <sander@ouroboros.rocks>
* ipcpd: Refactor create_r and flow_req_arrDimitri Staessens2019-03-047-51/+46
| | | | | | | | | | | The API calls for the IPCP to inform the IRMd of IPCP creation and incoming flow request had the pid_t in the call. This pid_t is removed and the getpid() call is now placed inside the function. Also refactors the cleanup for the main() functions of some of the lower IPCPs. Signed-off-by: Dimitri Staessens <dimitri@ouroboros.rocks> Signed-off-by: Sander Vrijders <sander@ouroboros.rocks>
* lib: Add initial rtt estimator to FRCTDimitri Staessens2019-02-082-18/+89
| | | | | | | | | | | | This adds a simple round-trip time estimator to FRCT. The estimate is a weighted average with deviation. The retransmission is scheduled after rtt + 2 times the deviation. A retransmit doubles the rtt estimate to avoid the no-update case when rtt suddenly increases. The rtt is estimated in microseconds and the granularity for retransmits is 256 microseconds. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* build: Update copyright to 2019Dimitri Staessens2019-02-05146-146/+146
| | | | | | | Updates the copyright notice in all sources to 2019. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* tools: Allow enrollment with destination IPCPDimitri Staessens2019-01-161-11/+52
| | | | | | | | | | | | | | | | | | The irm enroll tool had a semantic to enroll with a layer name, but this is not checked. Now the enroll command will retrieve the correct layer name that the IPCP got from the actual enrollment procedure. The irm enroll now has two string parameters, a dst and a layer, which cannot be both NULL. If only dst is specified, the IPCP will enroll with that name; autobind will bind with the layer name. If only layer is specified, the IPCP will enroll with the layer name, and perform a check that the layer name retrieved from enrollment is indeed the layer name before possibly autobinding. If both dst and layer are specified, the IPCP will enroll with dst and perform a check that the enrollment was in the expected layer. Basically only specifying the layer name is a shorthand for dst == layer. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* tools: Rename "normal" to "unicast"Dimitri Staessens2019-01-023-9/+9
| | | | | | | This renames the normal IPCP to unicast in the irm toolkit. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* lib: Fix deallocating non-initialized np1 flowsDimitri Staessens2018-12-293-5/+15
| | | | | | | | This fixes the deallocation of non-initialized IPCP flows. These can occur when some operations are not implemented. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* include: Add a flow_join operation for broadcastDimitri Staessens2018-12-2715-54/+173
| | | | | | | | | | | | This adds a new flow_join operaiton for broadcast, which is a much safer solution than overloading destination name semantics. The internal API now also has a different IPCP_FLOW_JOIN operation. The IRMd doesn't need to query broadcasts IPCPs for the name, it can just check if an IPCP with the layer name exists. The broadcast IPCP doesn't need to implement the query proxy call anymore. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* irmd: Manage shm_flow_set from IRMdDimitri Staessens2018-12-224-126/+111
| | | | | | | | | This moves the creation and destruction of shm_flow_set shared memory structures from the init to the IRMd. Now the management of all shared data objects is performed by the IRMd. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* ipcpd: Wait for buffer when writing to Eth deviceDimitri Staessens2018-12-081-1/+13
| | | | | | | | This will cause the Ethernet IPCP to wait for a free buffer when using raw sockets to avoid packet drops when the network is congested. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* lib: Fix reading tx queue length0.14.0Dimitri Staessens2018-12-071-1/+1
| | | | | | | The fccntl call was reading from the RX queue instead of the TX queue. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* tools: Fix server name argument in opingDimitri Staessens2018-11-301-1/+1
| | | | | | | | The --server-name option was mistyped with an underscore in the argument parser. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* lib: Add First Fragment (FFGM) bit to FRCTDimitri Staessens2018-11-211-1/+2
| | | | | | | | | This adds a First Fragment bit to FRCT. This small optimisation avoids losing two packets when there is packet loss without fragmentation, without the need to disable fragmentation at the end points. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* ipcpd: Fix uninitialized value for ECNSander Vrijders2018-11-201-9/+1
| | | | | | | | The ECN value in the packet was not set during write of a new packet. Also removes the outdated abstract syntax enum. Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be> Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be>
* ipcpd: Add Explicit Congestion Notification fieldFriedl Rubrecht2018-11-201-1/+7
| | | | | | | | This adds an extra field to the data transfer PCI which can be used to flag congestion by intermediate routers. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* lib: Revise du buff API towards a memory allocatorDimitri Staessens2018-10-263-44/+32
| | | | | | | | | This changes the API to the rdrbuff to treat it as a pool memory allocator. The head and tailspace to allocate in a buffer is now set system-wide instead of being passed as a parameter. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* tools: Add IFF_NO_PI flag to ovpnSander Vrijders2018-10-241-1/+1
| | | | | | | | The flag IFF_NO_PI is needed to make sure that no extra protocol header is added to the payload that is received on the TUN interface. Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be> Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be>
* ipcpd: Add broadcast IPCPDimitri Staessens2018-10-2421-25/+2013
| | | | | | | | | | This adds a broadcast IPCP that allows us to easily create multicast applications. The broadcast IPCP accepts flows for "<layer_name>.mc". A tool, obc (Ouroboros broadcast), is added that sends and reads a message to a broadcast layer. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* tools: Add ovpn applicationSander Vrijders2018-10-223-0/+351
| | | | | | | | | | This adds the ovpn application which allows to send TCP/IP traffic over Ouroboros. This is done by opening a TUN interface and allocating a flow to another ovpn application so that applications using TCP/IP can be used over Ouroboros as well. Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be> Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be>
* tools: Specify QoS cube for data transfer flows0.13.0Dimitri Staessens2018-10-199-32/+72
| | | | | | | | | The ipcp connect command can now set a specific qos cube for data transfer flows. For management flows, the tool ignores this and defaults to raw until data flows are stable enough. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* ipcpd: Check flow id instead of assertDimitri Staessens2018-10-191-1/+4
| | | | | | | | | This will check if the flow id is valid instead of asserting. It avoids assertion failures in the IPCP if an application crashes and the IRMd deallocates the flow while the IPCP still has pending writes. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* ipcpd: Check if there are entries to add to pffDimitri Staessens2018-10-191-3/+3
| | | | | | | | The pff didn't check if there were actually entries in the database, causing an assertion failure in some rare cases. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* ipcpd, irmd: Handle signal in main threadDimitri Staessens2018-10-186-140/+95
| | | | | | | | | The signals are now handled in the main thread instead of an asynchronous signal handler. The acceptloop is now correctly cancelled and the associated timeouts are removed. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* lib: Remove CRC flag from FRCTDimitri Staessens2018-10-152-4/+3
| | | | | | | | The integrity check mechanism was split from FRCT, this flag is not needed anymore. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* Merge branch 'testing' into beSander Vrijders2018-10-128-10/+24
|\
| * ipcpd: Send link state message before lsdb update0.12.3Dimitri Staessens2018-10-121-5/+4
| | | | | | | | | | | | | | | | | | | | When a new link is added, the link state update will be sent before the database update, so the network is a little bit more quickly aware. This improves odds of the DHT successfully enrolling at its first attempt, reducing bootstrap time of a network. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
| * ipcpd: Configure DHT slack timer at build timeDimitri Staessens2018-10-123-1/+9
| | | | | | | | | | | | | | This allows configuration of the DHT slack timer at build time. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
| * lib: Set hash algorithm correctlyDimitri Staessens2018-10-121-1/+3
| | | | | | | | | | | | | | | | | | There was a bug where the hash selection for the local, raptor and ethernet IPCPs was not passed correctly, so they were using a wrong hash. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
| * ipcpd: Make Qdisc bypass configurableSander Vrijders2018-10-123-0/+7
| | | | | | | | | | | | | | | | This will make bypassing the qdisc configurable, as it might be handy for getting fast data rates but is generally needed. Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be> Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be>
| * irmd: Don't warn if server leavesDimitri Staessens2018-10-121-3/+1
| | | | | | | | | | | | | | | | | | The irmd would log a warning when a server application quits during an accept call, but this is just a normal event. This removes the warning log. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* | Merge branch 'testing' into beDimitri Staessens2018-10-113-3/+10
|\|
| * lib: Add cleanup function in notifierSander Vrijders2018-10-111-1/+4
| | | | | | | | | | | | | | | | This adds a cleanup function in the notifier in case it gets cancelled, which is a possibility in some callbacks. Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be> Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be>
| * ipcpd: Call send_lsm under read lockSander Vrijders2018-10-111-0/+2
| | | | | | | | | | | | | | | | send_lsm was being called from the event handler as well as from lsupdate. The first one was not being locked properly. Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be> Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be>
| * ipcpd: Take correct lock when updating statsSander Vrijders2018-10-111-2/+4
| | | | | | | | | | | | | | | | The lock of the wrong fd was taken when updating the stats in the DT component. Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be> Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be>
* | Merge branch 'testing' into beSander Vrijders2018-10-119-168/+256
|\|
| * ipcpd: Speed up enrolment of DHTDimitri Staessens2018-10-112-62/+68
| | | | | | | | | | | | | | | | | | | | | | The link-state algorithm will now quickly recalculate for link additions (but not for removals, for stability). Upon notification of a new link, the DHT will wait for a brief moment to enroll. This reduces enrolment for large networks by some orders of magnitude. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
| * ipcpd: Decouple flow allocator from dt threadDimitri Staessens2018-10-111-86/+176
| | | | | | | | | | | | | | | | The flow allocator passed a blocking callback to the forwarding component, which blocks packet processing. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
| * lib: Disable CRC by defaultDimitri Staessens2018-10-111-1/+1
| | | | | | | | | | | | | | | | This sets the CRC check to be disabled by default, since it greatly hampers performance and is almost never needed. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>