summaryrefslogtreecommitdiff
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
...
* ipcpd: Create new SDUs for DHT join retriesDimitri Staessens2017-11-071-15/+15
| | | | | | | | | | The DHT join operation left an SDU in the rdrbuff for a couple of seconds during retries, which may cause the system to block if there is heavy traffic. This patch releases the sdb when the write fails and creates a new packet for every retry. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* lib: Refactor FRCT implementationDimitri Staessens2017-11-077-543/+683
| | | | | | | | | | | | | | | | | | | | The frct_pci and rq headers are moved from include/ouroboros to src/lib since they are only needed in the library. FRCT is moved to its own source file. FRCT takes the application PDUs, encapsulates and processes them and hands them back. This makes it easier to disable FRCT should the application want to write to a "raw" flow. An FRCT instance is now allocated upon alloc and released upon dealloc. The FRCT data structure is split into a sender and receiver connection record. Setting a new configuration will now be done upon sending the next data PDU, which will flag the DRF for a new run and use that configuration. This avoids some issues should packets arrive out-of-order, and simplifies setting a configuration. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* dev: Split nonblocking read and writeDimitri Staessens2017-10-301-2/+2
| | | | | | | | | | The FLOWFNONBLOCK flag now has two subflags FLOWFRNOBLOCK and FLOWFWNOBLOCK which allows setting the behavior of read and write independently. The default behavior is unchanged (blocking read and write). Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* dev: Bind python scripts to script nameDimitri Staessens2017-10-251-0/+13
| | | | | | | | | | This revises the library init script so that it checks whether the program is run by a python executable (python, python2 or python3) and announces the script name instead of the python interpreter. This enables binding python scripts. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* tools: Remove comment lines in CBRSander Vrijders2017-10-241-2/+0
| | | | | | | There were unnecessary comment lines in cbr.c, this removes them. Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be> Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be>
* tools: Change license to 3 clause BSDDimitri Staessens2017-10-2431-310/+808
| | | | | | | | | A more permissive license is needed on the tools since they serve as example code for programs built on top of the ouroboros-dev and ouroboros-irm libraries. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* tools: Refactor echo to single source fileDimitri Staessens2017-10-243-127/+73
| | | | | | | | This simplifies the echo-app tool implementation to a single source file since it's meant to be a small example application. Signed-off-by: Dimitri Staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* lib: Fix exit bug in dev.cdimitri staessens2017-10-231-0/+4
| | | | | | | | | | If the library init() fails, the fini() call tries to access unreleased resources. The fix resets the ai struct to 0 if init() fails and checks a heap pointer for NULL before the fini() accesses internals. Signed-off-by: dimitri staessens <dimitri.staessens@ugent.be> Signed-off-by: Sander Vrijders <sander.vrijders@ugent.be>
* Merged in dstaesse/ouroboros/be-deprecate-ouroboros-init-fini (pull request ↵dimitri staessens2017-10-1428-140/+86
|\ | | | | | | | | | | #626) Be deprecate ouroboros init fini
| * lib: Deprecate ouroboros_init and ourboros_finidimitri staessens2017-10-1428-140/+86
| | | | | | | | | | | | | | | | | | | | | | | | | | | | This commit deprecates ouroboros_init and ouroboros_fini and adds them as a constructor or destructor, causing these function to be run automatically when a program that links to the library calls and exits main(). For this to fully work, the library had to be split so that we can avoid the irmd calling these functions (the IRMd has to create the shm structures on which these calls depend). The library is split in 3 parts: libouroboros-dev, libouroboros-irm and libouroboros-common. The latter is linked to the other two so that including libouroboros-dev or libouroboros-irm will also link libouroboros-common.
* | Merged in dstaesse/ouroboros/be-fix-multiblock (pull request #627)dimitri staessens2017-10-144-2/+7
|\ \ | |/ |/| | | lib: Re-enable multi-block support
| * lib: Re-enable multi-block supportdimitri staessens2017-10-134-2/+7
| | | | | | | | | | | | Multi-block support was forgotten during the update of the build system. This enables it again and fixes some bugs when it is disabled and larger SDUs are sent.
* | ipcpd: Fix unreachable code in shim-eth-llcdimitri staessens2017-10-111-1/+9
|/ | | | | Closing the file descriptor after a while(true) loop led to clang compiler errors. This is fixed by adding a cleanup handler.
* Merged in dstaesse/ouroboros/be-fix-ls (pull request #624)dimitri staessens2017-10-071-1/+4
|\ | | | | | | ipcpd: Handle cancellation while writing LSAs
| * ipcpd: Handle cancellation while writing LSAsdimitri staessens2017-10-071-1/+4
| | | | | | | | | | | | | | | | | | | | This adds a cleanup handler to ensure the lock is released in the event that the thread is cancelled during a write. An alternative solution could be to make the flow write in a non-blocking way (only the blocking write contains an execution path that has a cancellation point). Fixes #55
* | irmd: Add option to print versiondimitri staessens2017-10-062-27/+37
|/ | | | | | | | | This adds the --version option to the IRMd. With this option it will print the version of Ouroboros to stdout and exit. Checking the version does not require root privileges. Also fixes calling log functions before log_init() and after log_fini().
* irmd: Don't destroy flows before IPCPsdimitri staessens2017-10-041-22/+1
| | | | | | This allows the IPCPs to call ouroboros_fini() and clean up their buffers and drastically reduce chances of getting an assertion error that some rbuffs are not empty when shutting down the irmd.
* lib: Cancel tpm threads instead of marking exitdimitri staessens2017-09-308-251/+261
| | | | | | | | | This makes the threadpool use pthread_cancel instead of setting an exit flag that threadpool managed threads check periodically. This drastically reduces CPU consumption in the irmd when running a lot of applications. It requires cancellation handlers in the ipcp and irmd to be implemented to ensure safe cancellation during operation and shutdown.
* ipcpd: normal: React to flow events in link state policySander Vrijders2017-09-293-4/+56
| | | | | | This will let the link state policy react to flow up and down events by notifying the PFFs of the routing instances of this event so they can take an appropriate action.
* ipcpd: normal: Add Loop-Free Alternates routingSander Vrijders2017-09-297-55/+227
| | | | | | This adds the Loop-Free Alternates (LFA) policy. In case a link goes down a LFA may be selected to route the SDUs on without causing loops instead of the main hop that just went down.
* Merged in sandervrijders/ouroboros/be-routing (pull request #617)Sander Vrijders2017-09-264-189/+288
|\ | | | | | | Be routing
| * ipcpd: normal: Add refcount to graph edgesSander Vrijders2017-09-263-22/+53
| | | | | | | | | | This adds a refcount to the graph edges so that it is only included in the calculation if both sides announced it.
| * ipcpd: normal: Keep index in vertex structSander Vrijders2017-09-252-27/+24
| | | | | | | | | | This keeps the index in the vertex struct so that is more easily available during Dijkstra.
| * ipcpd: normal: Simplify Dijkstra implementationSander Vrijders2017-09-251-25/+17
| | | | | | | | | | | | This simplifies the Dijkstra implementation by immediately setting the correct next hop during Dijkstra instead of looping through the list of predecessors afterwards.
| * ipcpd: normal: Simplify internal graph functionsSander Vrijders2017-09-251-38/+26
| | | | | | | | | | This simplifies several internal graph functions by passing an array of bools instead of an array of vertices.
| * ipcpd: normal: Return list as routing tableSander Vrijders2017-09-254-106/+197
| | | | | | | | | | | | This returns a list as routing table instead of a pointer to a pointer to a pointer, which simplifies the looping through the routing table and makes it more extensible for future additions.
* | Merged in sandervrijders/ouroboros/be-dht-fix (pull request #618)Sander Vrijders2017-09-261-3/+3
|\ \ | | | | | | | | | ipcpd: normal: Set clock to realtime clock
| * | ipcpd: normal: Set clock to realtime clockSander Vrijders2017-09-261-3/+3
| |/ | | | | | | | | | | A pthread_cond_wait was using the pthread_cond_clock and thereby timing out constantly, consuming 100% CPU. This changes it to use the realtime clock.
* / ipcpd: Add multiplier to scale scheduler threadsdimitri staessens2017-09-253-10/+14
|/
* ipcpd: Fix deadlock in DHTdimitri staessens2017-09-241-27/+41
| | | | | | | | | | | | | The lookup_update was waiting for the LU_INIT state to resolve under dht->lock which prevented that update. This PR fixes this bug, but leaves a (very rare) bug when the lookup_destroy is called while the lookup_update is waiting for the LU_INIT state to resolve. The solution also is a (harmless) lock inversion, but this is also not the best. Fixes #51 Fixes #52
* ipcpd: Add threadpool manager to DHTdimitri staessens2017-09-245-309/+404
| | | | | | | | | | | This adds a threadpool manager to the DHT. This was needed because the detached thread could cause a data race on shutdown. The threadpool manager is revised to allow multiple instances in a single program. The irmd and ipcp now store commands in a buffer (list) instead of a single buffer before passing it to handler threads.
* Merged in dstaesse/ouroboros/be-bugfixes (pull request #614)dimitri staessens2017-09-232-76/+124
|\ | | | | | | Some bugfixes
| * ipcpd: Don't publish DHT entries under lockdimitri staessens2017-09-231-30/+85
| |
| * lib: Clean up bitmap implementationdimitri staessens2017-09-231-46/+39
| | | | | | | | | | | | There was a return -1 in the allocate call which is unsafe since -1 may be a valid id in the bitmap. Since it's a data structure for internal use, I replaced the NULL checks with assertions.
* | Merged in dstaesse/ouroboros/be-bugfixes (pull request #613)dimitri staessens2017-09-232-9/+17
|\| | | | | | | Some bugfixes
| * ipcpd: Lock dht before reading parametersdimitri staessens2017-09-231-7/+14
| |
| * ipcpd: Fix data race in shim-udpdimitri staessens2017-09-231-2/+3
| | | | | | | | | | This fixes a data race in the shim-udp where the alloc_lock was released before the ipcpi.alloc_id was set.
* | lib: Disable lockless rbuff by defaultSander Vrijders2017-09-221-2/+2
| | | | | | | | | | This disables the lockless rbuff by default since it is less stable and tested.
* | ipcpd: Fix DHT recursive lookupdimitri staessens2017-09-221-1/+4
|/
* ipcpd: Mark len parameter as unused in simple_pffdimitri staessens2017-09-211-0/+4
|
* lib: Include string.h in hashtable.c for memcpydimitri staessens2017-09-211-0/+1
|
* Merged in dstaesse/ouroboros/be-cancel (pull request #609)dimitri staessens2017-09-213-72/+85
|\ | | | | | | ipcpd: Cancel threads in IPCPs
| * ipcpd: Cancel threads in IPCPsdimitri staessens2017-09-213-72/+85
| | | | | | | | | | | | This reduces CPU consumption and shutdown times. Invalidates #43
* | Merged in sandervrijders/ouroboros/be-lfa-pff (pull request #607)Sander Vrijders2017-09-2110-71/+569
|\ \ | |/ |/| | | ipcpd: normal: Add alternate hop PFF
| * ipcpd: normal: Add alternate hop PFFSander Vrijders2017-09-2110-71/+569
| | | | | | | | | | This adds a PFF that returns an alternate hop as next hop in case the hop that would have been returned is down.
* | ipcpd: Use the kernel scheduler for QoSdimitri staessens2017-09-213-61/+106
|/ | | | | | | | | This revises the sdu_scheduler of the normal IPCP to create a scheduler thread per QoS cube and let the kernel scheduler schedule them based on a priority. Priorities can be set at build time in a range of 0-100, which will be mapped onto a suitable priority for the kernel scheduler. The current scheduler policy is fixed to SCHED_RR since it is the most suitable scheduler defined by POSIX.
* lib: Pass the fuse struct to the handling threaddimitri staessens2017-09-212-5/+3
| | | | This avoids potential data races on rib.fuse.
* Merged in dstaesse/ouroboros/be-dht-revision (pull request #605)dimitri staessens2017-09-211-15/+58
|\ | | | | | | ipcpd: Handle DHT SDUs in different thread
| * ipcpd: Handle DHT SDUs in different threaddimitri staessens2017-09-201-15/+58
| | | | | | | | | | | | | | The DHT will now spawn a thread when receiving SDUs to avoid starvation of sdu scheduler threads. Also fixes some locking issues.
* | Merged in sandervrijders/ouroboros/be-pff (pull request #604)Sander Vrijders2017-09-2018-140/+417
|\ \ | |/ |/| | | ipcpd: normal: Make PFF policy-based