Guus Sliepen [Mon, 15 Feb 2021 22:48:09 +0000 (23:48 +0100)]
Allow a different location for the lock file.
In order to reduce the write load on the configuration directory as much as
possible, allow the location of the lock file to be set by the application
to somewhere outside the configuration directory. On Linux, it can be put
in /dev/shm.
Guus Sliepen [Mon, 8 Feb 2021 17:14:38 +0000 (18:14 +0100)]
Add missing config file update on receiving a proactive REQ_KEY request.
A REQ_KEY request from a node that wants our key can contain their key as
well. The node status was not set to dirty in this case, which could
prevent their key from being written to disk with the default settings, but
this actually always forgot to write the key to disk with storage policy
set to KEYS_ONLY.
Guus Sliepen [Fri, 5 Feb 2021 22:27:27 +0000 (23:27 +0100)]
Immediately write host config file when a new key is received.
When we receive a new public key, we should immediately call
node_write_config() to ensure it gets saved if the storage policy is
KEYS_ONLY.
Guus Sliepen [Thu, 4 Feb 2021 21:47:25 +0000 (22:47 +0100)]
Expand the storage policy test to check for correctness during invitations.
Guus Sliepen [Thu, 4 Feb 2021 21:46:58 +0000 (22:46 +0100)]
Don't allow meshlink_join() when the storage policy is DISABLED.
It does not make sense to allow this, since we need to write host config
files during a join, otherwise MeshLink's directory would be left in an
invalid state.
Guus Sliepen [Wed, 3 Feb 2021 22:31:18 +0000 (23:31 +0100)]
Add a test for the various storage policies.
Guus Sliepen [Wed, 3 Feb 2021 22:30:36 +0000 (23:30 +0100)]
Don't clear node dirty flag in meshlink_stop().
It's node_write_config() itself that should unset the dirty flag, depending
on the storage policy and if storage succeeded. Otherwise, we risk that
setting the storage policy back to ENABLED after calling meshlink_stop()
will not cause pending updates to be written out by meshlink_close().
Guus Sliepen [Fri, 29 Jan 2021 21:59:57 +0000 (22:59 +0100)]
Add meshlink_set_storage_policy().
This allows control over when MeshLink stores configuration files:
- MESHLINK_STORAGE_ENABLED: on all updates
- MESHLINK_STORAGE_KEYS_ONLY: only on new keys and black/whitelist updates
- MESHLINK_STORAGE_DISABLED: never
Guus Sliepen [Thu, 18 Feb 2021 19:57:23 +0000 (20:57 +0100)]
Allow the application to provide channel send/receive buffers.
This adds two new functions to allow the application to allocate send
and/or receive buffers for channels:
meshlink_set_channel_rcvbuf_storage(mesh, channel, buf, size)
meshlink_set_channel_sndbuf_storage(mesh, channel, buf, size)
Guus Sliepen [Tue, 23 Feb 2021 18:56:31 +0000 (19:56 +0100)]
Add meshlink_set_channel_flags().
This allows changing some of the flags of a channel. Currently it can be
used to change MESHLINK_CHANNEL_NO_PARTIAL and MESHLINK_CHANNEL_DROP_LATE.
Guus Sliepen [Tue, 23 Feb 2021 18:55:01 +0000 (19:55 +0100)]
Add missing checks for NULL mesh pointers in the public API.
Guus Sliepen [Sun, 21 Feb 2021 21:11:52 +0000 (22:11 +0100)]
Add warning messages when checking for bad fds after select() returns an error.
Guus Sliepen [Mon, 15 Feb 2021 21:16:04 +0000 (22:16 +0100)]
Never call timeout_set() outside callbacks if no callback is set.
We should never try to update a timer if it was never added to the event
loop. There were a few places that didn't have an explicit check to
prevent this from happening.
Guus Sliepen [Tue, 9 Feb 2021 20:00:35 +0000 (21:00 +0100)]
Try to recover from select() returning an error.
If we get an error back from select(), it could be because one of the
filedescriptors is in a bad state. This shouldn't occur, but if it does,
just call all the registered I/O callbacks to have them check for errors.
Only quit the event loop if the error persists.
Also fix some I/O callbacks that didn't expect to be called when the
filedescriptor isn't signalled.
Guus Sliepen [Mon, 1 Feb 2021 13:07:54 +0000 (14:07 +0100)]
Fix a socket fd leak.
Due to a logic error, whenever an interface went down and up, one or
more socket fds could be leaked. A long running process on a roaming
device or an unstable network could therefore run out of fds.
Guus Sliepen [Mon, 25 Jan 2021 20:00:47 +0000 (21:00 +0100)]
Check the return value of catta_simple_poll_get().
This ensures we gracefully handle the case of Catta not being able to bind
to the NETLINK_ROUTE socket.
Guus Sliepen [Thu, 14 Jan 2021 21:10:45 +0000 (22:10 +0100)]
Use AI_* instead of NI_* constants for getaddrinfo().
Guus Sliepen [Wed, 30 Dec 2020 13:55:35 +0000 (14:55 +0100)]
Send our canonical address to nodes we want to communicate with.
This ensures canonical addresses are propagated automatically, and can be
used when making meta and UDP connections.
Guus Sliepen [Wed, 30 Dec 2020 13:53:46 +0000 (14:53 +0100)]
Clean up the request jump table.
Use C99 designated array initializers to ensure the jump table and list of
request names is kept in sync with the enums.
Guus Sliepen [Wed, 30 Dec 2020 13:50:56 +0000 (14:50 +0100)]
Use the canonical address during UDP probes.
It is possible that a node has a canonical address, but due to NAT or other
reasons, the meta-connections with that node use other addresses. If UDP is
only possible to the canonical address, then we need to include that during
the initial UDP probing phase.
Guus Sliepen [Sun, 6 Dec 2020 15:18:18 +0000 (16:18 +0100)]
Add missing calls to init_sync_flag() in test cases.
Guus Sliepen [Sun, 6 Dec 2020 15:17:58 +0000 (16:17 +0100)]
Fix deleting configuration directories with hidden files.
We incorrectly skipped files starting with a dot when deleting directories.
This is a problem on macOS, which might add .DS_Store files.
Guus Sliepen [Sun, 22 Nov 2020 10:59:54 +0000 (11:59 +0100)]
Fix potential NULL pointer dereference.
It is possible that an attempt is made to forward a request to a node
that is not reachable via meta-connections. This would trigger an assert
in debug builds, or cause a segmentation fault in release builds. Add
checks before attempts to send a request to node->nexthop->connection.
Guus Sliepen [Mon, 16 Nov 2020 19:57:46 +0000 (20:57 +0100)]
Don't try to renew SPTPS keys for unreachable nodes.
This caused other functions that incorrectly used status.validkey to check
if they could send data to a node to pass a NULL pointer to send_request(),
causing a crash about an hour after a node went offline.
Guus Sliepen [Tue, 10 Nov 2020 20:10:00 +0000 (21:10 +0100)]
Only reset UDP SPTPS sessions if the session ID changed.
Previously we reset the SPTPS session if we detected if a node was
unreachable. However, that node might not think it was unreachable,
leading to only one side to reset the SPTPS connection. This would then
take some time to resolve itself.
We already had code to detect whether a node was restarted, so we use
that to detect if, once a node becomes reachable again, it remembers the
old SPTPS session or whether we have to start a new one. This should be
deterministic and not depend on the exact timing of events.
Guus Sliepen [Thu, 29 Oct 2020 22:38:22 +0000 (23:38 +0100)]
Also send the blacklist notification when we already have a connection.
Instead of just closing the connection, and having to wait for the
reconnection to happen to send the blacklist notification, we do it
immediately when meshlink_blacklist() is called.
Guus Sliepen [Thu, 29 Oct 2020 22:30:41 +0000 (23:30 +0100)]
Add a callback that notifies the application when it is blacklisted.
If an outgoing connection is made to a node that blacklists the local node,
a callback can now be set that informs the application of it.
Guus Sliepen [Sun, 25 Oct 2020 21:20:43 +0000 (22:20 +0100)]
Fix Doxygen errors.
Guus Sliepen [Sun, 25 Oct 2020 21:18:53 +0000 (22:18 +0100)]
Add API functions to query the blacklist status of nodes.
Guus Sliepen [Sun, 25 Oct 2020 21:17:29 +0000 (22:17 +0100)]
Check blacklist status before committing an invitation.
Although we delete invitation files when blacklisting a node, there is a
race condition where an invitation connection is created right before the
invitee is blacklisted. So check that the node is blacklisted right before
committing the node config file to disk.
Guus Sliepen [Sun, 25 Oct 2020 21:15:04 +0000 (22:15 +0100)]
Remove invitation files when blacklisting or forgetting a node.
Guus Sliepen [Sun, 11 Oct 2020 14:16:31 +0000 (16:16 +0200)]
When a new connection is activated, terminate any pending connections to the same peer.
This prevents issues mainly in the test suite where peers try to connect to
each other simultaneously, and have to terminate one of the connections.
Before both connections would succeed, and both would be terminated, leading
to a loop of reconnections until enough randomness got in to break the tie.
Guus Sliepen [Sun, 11 Oct 2020 13:40:34 +0000 (15:40 +0200)]
Don't reset the UDP SPTPS session when a node becomes reachable.
Only do this when it becomes unreachable. This fixes an issue where right
after a meta-connection is established, the initiator sends a proactive
REQ_KEY, before the peer really becomes reachable according to the graph.
When the latter happened, it would reset the session so far, causing a new
REQ_KEY to be sent, which could cross the ANS_KEY from the peer. This would
resolve itself after a few seconds, but causes an unnecessary delay that is
easy to trigger.
Guus Sliepen [Sun, 11 Oct 2020 13:19:01 +0000 (15:19 +0200)]
Fix waiting for peer node to become reachable in test suite.
Guus Sliepen [Wed, 30 Sep 2020 20:16:22 +0000 (22:16 +0200)]
Fix cornercases closing channels.
Closing a channel while there was data in the receive buffer would cause a
RST to be sent instead of a FIN. We now always send a FIN, and let data
in the receive buffer be handled for a later data handling (which would
then send a RST if necessary).
The RST could be dropped if the ACK seqno was not in the correct range.
We now always accept RSTs for established connections.
Finally, when receiving more data after closing the channel, we would just
accept the data but discard it, instead of sending a RST back. Now we do
send a RST back.
Guus Sliepen [Tue, 29 Sep 2020 20:38:55 +0000 (22:38 +0200)]
Send RST packets when receiving data after we closed a UDP channel.
If the application closed a channel, we keep the UTCP connection alive for
a bit longer to handle resends of FIN packets. However, if this is missed
for some reason, either because the FIN got lost or the peer ignored the
receive callback, and the peer is sending new data, we need to inform it
that we are no longer listening. To do this, send a RST back.
Guus Sliepen [Fri, 25 Sep 2020 20:00:55 +0000 (22:00 +0200)]
Extend the timeout period of the authentication phase on progress.
If there is progress during the authentication phase of connections, we reset
the ping timer to give extra time to complete the authentication.
We do the same for invitation connections.
Guus Sliepen [Fri, 25 Sep 2020 19:53:12 +0000 (21:53 +0200)]
Don't use fast timeouts for fully established connections.
During the fast retry period, we want to have a fast ping timeout until we have
a fully working connection. However, the code still used fast timeouts during
the fast retry window even if the connection was fully established.
Guus Sliepen [Fri, 25 Sep 2020 19:48:59 +0000 (21:48 +0200)]
Fix timeouts of 1 second expiring in less than one second.
Guus Sliepen [Thu, 10 Sep 2020 21:18:26 +0000 (23:18 +0200)]
Allow sptps_force_kex() in all situations.
Also allow sptps_force_kex() during the initial key exchange.
Guus Sliepen [Thu, 10 Sep 2020 21:13:39 +0000 (23:13 +0200)]
Allow sptps_force_kex() while a key exchange is in progress
We should not do anything if we are already exchanging a new key, and
just return true. This change prevents higher layers in MeshLink from
terminating a connection between two nodes if both peers call
sptps_force_kex() at nearly the same time.
Guus Sliepen [Mon, 7 Sep 2020 19:12:28 +0000 (21:12 +0200)]
Use the canonical address exclusively for making outgoing meta-connections.
If we have a node's canonical address, we now always use that as a source
for addresses for outgoing meta-connection attempts. This commit also adds
the function meshlink_clear_canonical_address() to ensure the canonical
address can be removed if it is no longer valid.
Guus Sliepen [Mon, 7 Sep 2020 19:10:22 +0000 (21:10 +0200)]
Update the invite-join test to update the canonical address after a port change.
Guus Sliepen [Mon, 7 Sep 2020 15:58:23 +0000 (17:58 +0200)]
Always ensure we store a port number when setting the canonical address.
Guus Sliepen [Sat, 5 Sep 2020 09:52:12 +0000 (11:52 +0200)]
Use the fast retry period of the destination node's device class.
Guus Sliepen [Tue, 4 Aug 2020 13:24:07 +0000 (15:24 +0200)]
Remove temporary files at startup.
When something happens while a host config files is written, a temporary
file might be left over. Clean these up when we find them when starting
MeshLink.
Guus Sliepen [Wed, 29 Jul 2020 12:44:42 +0000 (14:44 +0200)]
Add meshlink_set_channel_listen_cb().
The accept callback is called when the peer has already fully established a
connection. The listen callback is called earlier, when there is no
fully established channel yet. However, the listen callback itself does not
get a channel handle, it can only make a decision based on the peer node
and port number whether to accept the channel, and if so the accept callback
will be called later.
Guus Sliepen [Wed, 29 Jul 2020 12:40:27 +0000 (14:40 +0200)]
Ensure the poll callback is called when a UDP channel has finished connecting.
Guus Sliepen [Thu, 23 Jul 2020 22:06:53 +0000 (00:06 +0200)]
Speed up the import-export test by using the PMTU callback.
Guus Sliepen [Thu, 23 Jul 2020 22:06:26 +0000 (00:06 +0200)]
Fix the import-export test failing sporadically.
Since two nodes connecting simultaneously can cause a node to be seen as
temporarily being down, thus setting the last_unreachable status.
Guus Sliepen [Thu, 23 Jul 2020 22:04:38 +0000 (00:04 +0200)]
Always let the initiator send a REQ_KEY once a connection is activated.
Before, the logic was to do this when the graph reported a bidirectional
edge. However, there was a possibility that if two nodes connect to each
other simultaneously, causing a second connection to be activated while the
first was also still active, which caused the REQ_KEY to not be sent.
Guus Sliepen [Thu, 23 Jul 2020 21:13:57 +0000 (23:13 +0200)]
Fix the channels-udp test case.
Close channels gracefully after the server has finished sending the streams.
Guus Sliepen [Thu, 23 Jul 2020 14:19:29 +0000 (16:19 +0200)]
Combine blackbox join test cases into the invite-join test.
Guus Sliepen [Wed, 22 Jul 2020 21:34:06 +0000 (23:34 +0200)]
Don't attempt to sync confbase for ephemeral nodes during a join.
Guus Sliepen [Wed, 22 Jul 2020 19:07:02 +0000 (21:07 +0200)]
Port the blackbox status_cb test.
Use ephemeral MeshLink instances to speed up the test.
Guus Sliepen [Wed, 22 Jul 2020 18:36:18 +0000 (20:36 +0200)]
Port the blackbox meta-connections test using network namespaces.
Guus Sliepen [Wed, 22 Jul 2020 18:35:18 +0000 (20:35 +0200)]
When activating a meta-connection, remember the address of the peer.
There was a possible situation where we did not remember the address of a
peer when there is only a TCP connection.
Guus Sliepen [Wed, 22 Jul 2020 18:31:59 +0000 (20:31 +0200)]
Reset last_connect_try for all nodes when starting the mesh.
This helps the autoconnect algorithm reconnect faster if the mesh was
stopped and restarted in quick succession.
Guus Sliepen [Wed, 22 Jul 2020 18:30:14 +0000 (20:30 +0200)]
Add reset_sync_flag().
This is for the test suite to reset a flag without broadcasting that it
changed state.
Guus Sliepen [Wed, 22 Jul 2020 18:29:22 +0000 (20:29 +0200)]
Add devtool_set_meta_status_cb().
This function is similar to meshlink_set_node_status_cb(), except that
this callback will only be called when a meta-connection to a node is
activated or terminated. This is mainly useful for the test suite.
Guus Sliepen [Mon, 13 Jul 2020 20:40:06 +0000 (22:40 +0200)]
Don't store empty canonical addresses.
Use a NULL pointer instead of an empty string to signal the lack of a
canonical address.
Guus Sliepen [Wed, 22 Jul 2020 13:00:59 +0000 (15:00 +0200)]
Fix invitation URL generation when running in a network namespace.
MeshLink could call getifaddrs() in the namespace of the caller instead of
the MeshLink thread, causing the wrong addresses to be put in the inviation
URL.
Guus Sliepen [Wed, 22 Jul 2020 12:59:44 +0000 (14:59 +0200)]
Export missing meshlink_open_params_* symbols.
Guus Sliepen [Fri, 10 Jul 2020 20:15:39 +0000 (22:15 +0200)]
Fix a crash with some network configurations.
It is apparently possible for getifaddrs() to return a struct ifaddrs that
contains a NULL ifa_addr pointer.
Guus Sliepen [Fri, 10 Jul 2020 20:13:23 +0000 (22:13 +0200)]
Add meshlink_reset_timers().
This resets the outgoing connections timers and ping timers, exactly what
is done when the network intterface status changes.
Guus Sliepen [Tue, 7 Jul 2020 08:26:21 +0000 (10:26 +0200)]
Check for the presence of stdatomic.h before using it.
Guus Sliepen [Sat, 4 Jul 2020 11:59:52 +0000 (13:59 +0200)]
Don't use assert() to check the results of pthread_*() calls.
This was done to debug the code, but it fails when MeshLink is compiled with
-DNDEBUG. Remove all assert()s from calls to pthread functions, and instead
add explicit checks to only those functions that can fail.
Guus Sliepen [Sat, 4 Jul 2020 10:26:15 +0000 (12:26 +0200)]
Export meshlink_set_dev_class_maxtimeout().
Guus Sliepen [Sat, 4 Jul 2020 10:26:00 +0000 (12:26 +0200)]
Ensure the maxtimeout value is taken from the destination node's device class.
Guus Sliepen [Wed, 24 Jun 2020 20:22:01 +0000 (22:22 +0200)]
Make the maximum outgoing connection timeout runtime configurable.
This moves maxtimeout to the dev_class_traits, and adds a function to
change it, similar to how we handle other configurable timeouts.
Guus Sliepen [Tue, 16 Jun 2020 19:34:52 +0000 (21:34 +0200)]
Fix compiling the C++ examples.
Guus Sliepen [Tue, 16 Jun 2020 18:07:12 +0000 (20:07 +0200)]
Revert "Fix warnings caused by C-only flags passed to the C++ compiler."
This reverts commit
9382f55b3f1c14c74e3bda229e277400743b11cc.
Guus Sliepen [Tue, 16 Jun 2020 18:06:59 +0000 (20:06 +0200)]
Revert "Don't overwrite CFLAGS when adding -std=c11."
This reverts commit
6d0820fda557de28c014b8ad4c40c385d23029a7.
Guus Sliepen [Mon, 15 Jun 2020 08:03:36 +0000 (10:03 +0200)]
Don't overwrite CFLAGS when adding -std=c11.
Guus Sliepen [Sun, 14 Jun 2020 12:45:18 +0000 (14:45 +0200)]
React faster to network changes, including point-to-point links.
Tell Catta to also include point-to-point links, and when we get an
update from the Catta thread, wake up the main MeshLink thread so we
react to it immediately.
Guus Sliepen [Sat, 13 Jun 2020 19:45:08 +0000 (21:45 +0200)]
Only call setitimer if ITIMER_REAL is defined.
Musl doesn't implement setitimer().
Guus Sliepen [Sat, 13 Jun 2020 19:39:39 +0000 (21:39 +0200)]
Set NOSIGPIPE on all sockets.
MacOS can raise a SIGPIPE when a local socket gets disconnected.
Guus Sliepen [Fri, 12 Jun 2020 22:41:47 +0000 (00:41 +0200)]
Fix warnings and missing mutex/cond initialization in the test suite.
Guus Sliepen [Fri, 12 Jun 2020 22:40:43 +0000 (00:40 +0200)]
Initialize the adns_done_queue.
Guus Sliepen [Fri, 12 Jun 2020 22:40:09 +0000 (00:40 +0200)]
Fix warnings caused by C-only flags passed to the C++ compiler.
Guus Sliepen [Fri, 12 Jun 2020 07:09:52 +0000 (09:09 +0200)]
Add missing initialization of a condition variable.
Guus Sliepen [Thu, 11 Jun 2020 20:31:25 +0000 (22:31 +0200)]
Fix some test cases using the same configuration directory.
Guus Sliepen [Thu, 11 Jun 2020 19:52:00 +0000 (21:52 +0200)]
Use atomic operations to check whether to write to the signal pipe.
We need to do an atomic test-and-set operation to check whether we can
avoid writing to the signal pipe. Use C11 atomics to do this in a portable
way (hopefully).
Guus Sliepen [Thu, 11 Jun 2020 20:22:37 +0000 (22:22 +0200)]
Remove gettimeofday() usage from test cases.
Guus Sliepen [Thu, 11 Jun 2020 20:17:23 +0000 (22:17 +0200)]
Add asserts() to all pthread related function calls.
We normally expect all pthread-related functions to succeed, so in all
places where we didn't already explicitly check the return value, assert()
that the functions return 0.
Guus Sliepen [Wed, 10 Jun 2020 20:25:12 +0000 (22:25 +0200)]
Properly initialize mutexes and condition variables.
On Linux, zeroing a pthread_mutex_t or pthread_cond_t variable ensures
the mutex/cond is properly initialized, however this is not the case om
some other platforms. Ensure we always call pthread_mutex/cond_init().
Guus Sliepen [Tue, 9 Jun 2020 21:18:04 +0000 (23:18 +0200)]
Fix compiler warnings on macOS.
Guus Sliepen [Tue, 9 Jun 2020 21:12:44 +0000 (23:12 +0200)]
Check whether IP_MTU is defined.
Guus Sliepen [Fri, 5 Jun 2020 16:09:38 +0000 (18:09 +0200)]
Fix meshlink_join() failing on Android.
The adns_blocking_request() function did not pass a hint to
getaddrinfo(). With glibc, the resulting struct addrinfo sets socktype
and protocol to SOCK_STREAM and IPPROTO_TCP, and the call to connect()
copied these values. However, bionic doesn't set the socktype and
protocol to those values if no hint was specified.
Guus Sliepen [Fri, 5 Jun 2020 07:40:26 +0000 (09:40 +0200)]
Ensure utcp-test is compiled with the same flags as libmeshlink.
Guus Sliepen [Wed, 3 Jun 2020 19:23:33 +0000 (21:23 +0200)]
Include "system.h" in the UTCP sources.
The UTCP sources were not using the system.h header file like the rest
of the code, resulting in issues with some compilers.
Guus Sliepen [Wed, 3 Jun 2020 19:19:09 +0000 (21:19 +0200)]
Don't crash when trying to connect a channel to port 0.
Guus Sliepen [Thu, 21 May 2020 12:48:02 +0000 (14:48 +0200)]
Explicitly set the stack size for the MeshLink thread.
Different libcs have different default sizes for newly created threads. In
particular, Musl defaults to 80 kB, which is too small for MeshLink. We now
request 1 MB, which should be more than enough to handle the deepest call
stacks.
Guus Sliepen [Wed, 20 May 2020 22:13:50 +0000 (00:13 +0200)]
Limit the size of the fd read buffer in channel_poll().
Guus Sliepen [Mon, 18 May 2020 21:25:50 +0000 (23:25 +0200)]
Add a test for AIO callback cornercases.
Guus Sliepen [Mon, 18 May 2020 21:24:56 +0000 (23:24 +0200)]
Fix assert that could incorrectly be triggered when a peer closed the channel.
If we are sending AIO data on a channel and the peer closed the connection,
we could trigger an assert incorrectly.
Guus Sliepen [Mon, 18 May 2020 21:23:32 +0000 (23:23 +0200)]
Report the amount of actual data sent/received in AIO callbacks.
We did this for fds but not for memory buffers.
Guus Sliepen [Mon, 18 May 2020 21:22:06 +0000 (23:22 +0200)]
Reset closed connections if there is incoming data.
Guus Sliepen [Fri, 15 May 2020 21:12:34 +0000 (23:12 +0200)]
Include our own key in REQ_PUBKEY requests.
If we don't know a peer's public key, it most likely means the peer
doesn't know our public key, so proactively send it along with the
REQ_PUBKEY request.
Guus Sliepen [Fri, 15 May 2020 07:39:59 +0000 (09:39 +0200)]
Handle UTCP receive buffer wraparound corner cases.
Before we allowed buf->offset to be equal to buf->size. This caused an
issue where buffer_call() would call the callback twice, once for 0
bytes at the end of the buffer, and once for len bytes at the start of
the buffer. This would cause the callback function to think the channel
had encountered an error.
If the data in the ringbuffer wraps around, and we call the receive
callback for the first part of the data, the callback function might
close the channel, so we must not call the callback for the second part
of the data.