| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
| |
The RIB will now show some stats for the flow allocator, including
congestion avoidance statistics. This is needed before decoupling the
data transfer component and the flow allocator as some current stats
show in DT will move to FA.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
| |
The mb-ecn policy has a couple of divisions in the math, which I
wanted to avoid. Now it measures the number of bytes sent in a window,
and updates the next window with AIMD logic. If the number of bytes in
the window is reached, the call blocks. To avoid long packet bursts,
the window size continually scales to contain between CA_MINPS (8) and
CA_MAXPS (64) packets.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
| |
The dt component bypasses the flow allocator on the receiver side, and
may try to update congestion context when the flow has already been
deallocated by the receiver. I will fix this bypass and always pass
through the flow allocator sometime soon; for now, I added a check in
the flow allocator call to avoid the SEGV.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
| |
The enrollment procedure was not passing the policy for congestion
avoidance.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This adds congestion avoidance policies to the unicast IPCP. The
default policy is a multi-bit explicit congestion avoidance algorithm
based on data-center TCP congestion avoidance (DCTCP) to relay
information about the maximum queue depth that packets experienced to
the receiver. There's also a "nop" policy to disable congestion
avoidance for testing and benchmarking purposes.
The (initial) API for congestion avoidance policies is:
void * (* ctx_create)(void);
void (* ctx_destroy)(void * ctx);
These calls create / and or destroy a context for congestion control
for a specific flow. Thread-safety of the context is the
responsability of the flow allocator (operations on the ctx should be
performed under a lock).
ca_wnd_t (* ctx_update_snd)(void * ctx,
size_t len);
This is the sender call to update the context, and should be called
for every packet that is sent on the flow. The len parameter in this
API is the packet length, which allows calculating the bandwidth. It
returns an opaque union type that is used for the call to check/wait
if the congestion window is open or closed (and allowing to release
locks before waiting).
bool (* ctx_update_rcv)(void * ctx,
size_t len,
uint8_t ecn,
uint16_t * ece);
This is the call to update the flow congestion context on the receiver
side. It should be called for every received packet. It gets the ecn
value from the packet and its length, and returns the ECE (explicit
congestion experienced) value to be sent to the sender in case of
congestion. The boolean returned signals whether or not a congestion
update needs to be sent.
void (* ctx_update_ece)(void * ctx,
uint16_t ece);
This is the call for the sending side top update the context when it
receives an ECE update from the receiver.
void (* wnd_wait)(ca_wnd_t wnd);
This is a (blocking) call that waits for the congestion window to
clear. It should be stateless (to avoid waiting under locks). This may
change later on if passing the context is needed for different algorithms.
uint8_t (* calc_ecn)(int fd,
size_t len);
This is the call that intermediate IPCPs(routers) should use to update
the ECN field on passing packets.
The multi-bit ECN policy bases the value for the ECN field on the
depth of the rbuff queue packets will be sent on. I created another
call to grab the queue depth as fccntl is write-locking the
application. We can further optimize this to avoid most locking on the
rbuff.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
| |
The flow stats had quite a lot of duplication.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
| |
The condition variable was not initialized correctly and using the
wrong clock for pthread_cond_timedwait.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
| |
A flow_set is thread-safe and doesn't need to be protected by a lock.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
| |
Fix assignment instead of comparison operator.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
| |
There were some issues identified by the Clang static analyzer that
are now fixed.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
| |
GCC 10 static analyzer found that the wrong index was used in the fail
path of psched_create, causing double (multiple) frees.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
| |
GCC 10 defaults to -fno-common, so some variables that were defined in
the headers needed to be declared "extern". The GCC 10 static analyzer
can now be invoked using the DebugAnalyzer build option.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
| |
The compiler spotted some variables that weren't really used.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
| |
This is more in line with the write() system call and prepares for
partial writes. Partial writes are disabled by default (and not yet
implemented).
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
| |
There were updates under rdlock instead of wrlock, causing data races
and trouble. Also speeds up shutdown a bit.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
| |
The initial implementation for the ECDHE key exchange was doing the
key exchange after a flow was established. The public keys are now
sent allowg on the flow allocation messages, so that an encrypted
tunnel can be created within 1 RTT. The flow allocation steps had to
be extended to pass the opaque data ('piggybacking').
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
| |
This adds tests for LFA and ECMP to the graph_test routine.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
| |
There were bugs introduced in the LFA algorithm refactor causing
infinite recursion and SEGV. The infinite recursion check was added as
an explicit compiler flag to the build.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
| |
Otherwise the compile will complain that the comparison of an unsigned
enum expression < 0 is always false.
Signed-off-by: Sander Vrijders <[email protected]>
Signed-off-by: Dimitri Staessens <[email protected]>
|
|
|
|
|
|
|
|
| |
Some fixes in the multipath implementation related to memory
management that showed up with the static analyzer.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
|
| |
This adds an equal-cost multipath routing policy to Ouroboros, based
on Nick Aerts' code. When selected, flows will send packets over all
paths with equal cost (hop count). Path selection is round-robin. It
does not yet take into account flows that are down.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
| |
During refactor a variable initialization was removed.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
| |
This makes the hashtable more tailored to a packet forwarding table
(PFT). In the end not much of a change was needed, but now it's clear
the pft maps a destination address to a list of (outgoing) fds.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
| |
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
|
| |
The vertex was used before definition in the graph
implementation. Fixed potential data race in link_state
algorithm. Added missing asserts. Removed initialization of variables
where not needed to let compiler warn about uninitialized uses.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
| |
The Packet Forwarding Function (PFF) was user-configurable using the
irm tool. However, this isn't really wanted since the PFF is dictated
by the routing algorithm. This moves the responsability for selecting
the correct PFF from the network admin to the unicast IPCP
implementation. Each routing policy now has to specify which PFF it
will use.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
|
| |
The hashtable is only used for forwarding tables in the unicast
IPCP. This moves the generic hashtable out of the library into the
unicast IPCP to prepare a more tailored implementation specific to
routing tables containing address lists.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
|
| |
The LFA algorithm modifies the output of the simple routing algorithm,
but the output was mixed in the general call. This moves the LFA
subroutine to be self-contained. This makes for a cleaner entry point
when adding more routing algorithms.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
|
| |
The eth, udp and local IPCPs were not filtering out the event types
from the flow, causing some reads when there are no packets in the
queue. The types are now also organized as flags so they can be
filtered more quickly if needed.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
| |
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
| |
The ip address was stored twice, which is not needed. Only keeping the
ip address in the sockaddr struct will make IPv6 support more
homogeneous.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
| |
The cypher_s field in QoS was sometimes 32 and sometimes 16 bits. This
is now corrected to be 16 bits.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This adds a per-message symmetric encryption using the OpenSSL
library. At flow allocation, an Elliptic Curve Diffie-Hellman exchange
is performed to derive a shared secret, which is then hashed using
SHA3-256 to be used as a key for symmetric AES-256 encryption. Each
message on an encrypted flow adds a small crypto header that includes
a random 128-bit Initialization Vector (IV). If the server does not
have OpenSSL enabled, the flow allocation will fail with an -ECRYPT
error.
Future optimizations are to piggyback the public keys on the flow
allocation message, and to enable per-flow encryption that maintains
the context of the encryption over multiple packets and doesn't
require sending IVs.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
| |
This completes the renaming of the normal IPCP to the unicast IPCP in
the sources, to get everything consistent with the documentation.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
|
| |
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 <[email protected]>
Signed-off-by: Dimitri Staessens <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
| |
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 <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
| |
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 <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
|
| |
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 <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
| |
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 <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
| |
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 <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
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 <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
| |
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 <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
| |
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 <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
| |
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 <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
| |
Updates the copyright notice in all sources to 2019.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
| |
This fixes the deallocation of non-initialized IPCP flows. These can
occur when some operations are not implemented.
Signed-off-by: Dimitri Staessens <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
| |
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 <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
| |
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 <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|
|
|
|
|
|
|
|
| |
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 <[email protected]>
Signed-off-by: Dimitri Staessens <[email protected]>
|
|
|
|
|
|
|
|
| |
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 <[email protected]>
Signed-off-by: Sander Vrijders <[email protected]>
|