| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
|
|
|
| |
A missing else clause was missing in the fast path, causing the PCI to
be shrunk when it should not be. A double free has also been fixed.
|
|
|
|
|
| |
This adds a lock to prevent a race condition between flow_req_arr and
flow_alloc_resp.
|
|
|
|
|
|
|
| |
Since there are no SDUs on most QoS cubes, the fmgr was always timing
out on most QoS cubes, causing considerable delays in most SDUs. The
timeout was reset to a very small value (so the bug in glibc 2.25 will
cause lockups again).
|
|
|
|
|
| |
This removes some logs in the graph component and replaces prints of
uin64_t variables with an architecture agnostic variant.
|
| |
|
|
|
|
|
|
| |
The shm PCI was never initialized during flow manager init. This
commit will do that, and initialize the pdu length correctly as well,
since it was not being written into the RIB, nor read in shm_pci_init.
|
|
|
|
|
|
| |
This adds fixes the locking of the PFF which was externalized, but not
yet correctly updated within the PFF component itself and within the
flow manager.
|
| |
|
|
|
|
|
| |
The flow manager should clean up the buffer after the call to frct
create instance has either failed or succeeded.
|
|\
| |
| |
| | |
ipcpd: Increase update timeout in fmgr
|
| |
| |
| |
| | |
This works around a possible bug in glibc 2.25.
|
|/
|
|
|
| |
This initializes conn_info in the ribmgr and fmgr before it is passed
to cacep.
|
|
|
|
| |
These solve several bugfixes in the normal.
|
|
|
|
|
|
| |
This split the initialization of the flow manager into an init part
and a start part. This avoids the usage of data structures that have
not been properly initialized yet.
|
|
|
|
|
|
|
| |
This makes the routing component create a Flow State Database
(FSDB). An FSDB contains Flow State Objects (FSOs). An FSO is created
when a neighbor is added, it is deleted when a neighbor is removed and
its QoS is updated when a neighbor's QoS changes.
|
|
|
|
|
|
|
| |
Our mailserver was migrated from intec.ugent.be to the central
ugent.be emailserver. This PR updates the header files to reflect this
change as well. Some header files were also homogenized if the
parameters within the functions were badly aligned.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This refactors the different Application Entities of the normal
IPCP. They all listen to and use the connection manager to establish
new application connections.
This commit also adds a neighbors struct to the normal IPCP. It
contains neighbor structs that contain application
connection. Notifiers can be registered in case a neighbor changes
(added, removed, QoS changed).
The flow manager has an instance of this neighbors struct and listens
to these events to update its flow set. The routing component also
listens to these events so that it can update the FSDB if needed. The
flow manager now also creates the PFF instances and the routing
instances per QoS cube.
The RIB manager also uses this an instance of the neighbors struct and
listens to neighbor events as well.
|
|
|
|
|
|
|
| |
This adds the connection manager which allows the different AEs of the
normal IPCP to register with it. An AE can then use the connection
manager to allocate a flow to a neighbor, or to wait for a new
connection from a neighbor.
|
|
|
|
|
|
|
| |
By removing authentication as part of CACEP, all policies
disappear. CACEP becomes a policy-free connection establishment
protocol between Application Entities. Authentication can later be
added cleanly as a pure policy function when needed.
|
|
|
|
|
| |
The AE name should not be passed over the layer boundaries. If an
application has more than one AE it should exchange this in CACEP.
|
|
|
|
|
|
| |
The information passed to CACEP is split between the information about
the connection and the information to be used during the
authentication exchange.
|
|
|
|
|
| |
Moves the definitions of paths in the RIB for the normal IPCP to a
header ribconfig.h to avoid repetition.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
This removes the logfile and outputs log messages to the logging
system. The creation of the logfiles (as well as the ap_init() call)
were moved into ipcp_init() to simplify the IPCP creation and
shutdown.
Fixes #25
Fixes #27
|
|
|
|
|
|
|
|
|
|
|
| |
The pci function will store offsets to avoid having to recalculate
them every time. TTL handling is removed from the fmgr, the TTL is
automatically decreased when deserializing the PCI (or set to 1 if
there is no TTL present so the PDU will be forwarded). The
deserialisation function now takes a pointer to a struct pci as input
to avoid memory allocation and release each time a PDU is
processed. Some checks have been replaced with assert() and return
values replaced with void where it makes sense.
|
|
|
|
|
|
|
|
|
|
| |
This PR updates the normal IPCP to use the new RIB. The old ribmgr is
removed and replaced by a stub that needs to be implemented. All
components (dir, fmgr, frct) were adapted to the new RIB API. A lot
of functionality was moved outside of the ribmgr, such as the
addr_auth, which is now a component of the IPCP. The address is also
stored to the ipcpi struct. The irm tool has an option to set the gam
policy of the rib manager.
|
|
|
|
|
| |
This allows the selection of a policy for the graph adjacency
manager. Currently we only support constructing a complete graph.
|
|
|
|
|
| |
After a previous commit the API of the linked list was changed. This
updates the files that were not yet merged upstream.
|
|
|
|
|
| |
This commit adds the graph adjacency manager to the normal IPCP, which
sets up N-1 flows to other members.
|
|
|
|
| |
This removes a bad unlock in the flow manager.
|
|\
| |
| |
| | |
ipcpd: Fix uninitialized variables in ribmgr
|
| | |
|
|\|
| |
| |
| | |
ipcpd: Refactor normal ipcp, initial commit
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Reorganizes the normal IPCP a bit to make sure internal components do
not need to access the state of the IPCP. The IPCP has now a thread
calling accept and delegating it to the correct component based on the
AE name (this used to be in the fmgr).
Internal components are initialized upon enrollment or bootstrap of
the IPCP. If a step fails, the IPCP goes back to the INIT state, if
all components boot correctly, it goes to the operational state.
RIB synchronization is still done by sending a CDAP start/stop and
syncing with a ribmgr state, but needs revision later on.
|
|/
|
|
|
|
|
|
|
|
| |
Now correctly relays the qoscube end-to-end in the stack.
A simple function specifying the cube in the spec is used for initial
testing.
The translation is now done in dev.c, but it could be moved elsewhere
when qos cabability matures and the need arises.
|
|
|
|
|
|
| |
The main thread will wait for the IPCP_OPERATIONAL state before
starting the fmgr main thread by calling fmgr_init(), instead of the
fmgr itself waiting for that state.
|
| |
|
|
|
|
|
| |
This corrects the license statements on all files. Installed headers
are LGPLv2.1, the rest of the code is GPLv2.
|
|
|
|
|
| |
Refactors the normal IPCP fmgr and ribmgr, and modifies the API for
cdap so that no callbacks are needed.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Receiver timeouts can now be set on a flow using the flow_set_timeout
function. Specifying NULL disables the timeout. The flow_get_timeout
function gets the value for the timeout.
This commit also deprecates fcntl in favor of flow_get_flags and
flow_set_flags functions.
struct qos_spec is typedef'd as a qosspec_t.
The tools and cdap.c are updated to use the new API.
Fixes a bug in operf client where the client's writer thread wouldn't
cancel on SIGINT.
|
|\
| |
| |
| | |
ipcpd: Change IPCP_RUNNING to IPCP_OPERATIONAL
|
| |
| |
| |
| | |
Changes a state of the IPCP to a more correct terminology.
|
|/
|
|
|
| |
This fixes several bad cleanups in the normal IPCP when it is shutting
down.
|
|
|
|
|
| |
This adds a directory to the normal IPCP that maps names on IPCP
addresses.
|
|
|
|
|
|
|
|
| |
This will split the IPCP state PENDING_ENROL into IPCP_CONFIG and
IPCP_BOOTING. IPCP_CONFIG is concerned only with configuring the IPCP
with the bare essence. When in IPCP_BOOTING, the IPCP will complete
its configuration by starting its policies, and thus making the IPCP
completely functioning.
|
|
|
|
| |
This fixes some bugs in connection establishment over the normal IPCP.
|
| |
|
|
|
|
|
| |
This reduces the risk for some bugs, for instance due to
signed/unsigned mismatches and unused variables.
|
|\
| |
| |
| | |
lib: Demultiplex the fast path
|