| Commit message (Collapse) | Author | Age |
|---|
| ... | |
| |
|
|
|
|
| |
This both makes new functions available to the core,
and addresses a bug which would have prevented building the
ZeroTier transport on Windows.
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
| |
* fixes #634 for tests/tls by explicitly using IPv4 for the "Valid verify works" test
|
| | |
|
| | |
|
| |
|
|
|
| |
This also fixes a leaked TCP connection on a failure path, which we
noticed while working this change.
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
| |
Well, actually, #506 fixed that, but this fixes the commentary
that indicates otherwise in the code.
|
| | |
|
| |
|
|
|
|
|
|
|
| |
This changes the signature of the aio cancellation routines
to take the argument for cancellation directly, so we do not
need to lookup the argument using the nni_aio_get_prov_data.
We should probably consider eliminating nni_aio_get_prov_data,
and co, and changing the prov_extra to reflect prov_data. Later.
|
| | |
|
| | |
|
| |
|
|
| |
fixes #651 Use after free in websocket
|
| | |
|
| |
|
|
|
|
|
| |
The POLLHUP (or rather EPOLLHUP) flag does not quite mean the same
thing in Linux, and we've seen random failures where we will sometimes
get this event on a socket that is freshly connected. This might be
a bug in Linux, but it is easy enough to workaround -- we just don't
watch for it at all.
|
| |
|
| |
fixed compilation error on OpenBSD, missing sockpeercred
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
| |
Actually the problem was in socket core, in particular in the
shutdown code. The socket shutdown is supposed to ensure that
no pipes were present on the socket, so that protocols need not
concern themselves with this. The code unfortunately was busted,
due to an ordering problem compounded by a race condition. This
fixes that, and changes the REQ protocol to avoid the blocking
condition altogether, and sprinkles a few assertions to validate
these rules are being adhered to.
|
| |
|
|
|
|
|
|
|
|
| |
fixes #599 nng_dial sync should not return until added to socket
This reintroduces the changes for the above fixes, building upon the
transport modifications that we have made to eliminate the separate
transport pipe start entry point. It also includes slightly reworked
code during start to put a hold on the pipe when it is created, which
we we drop at the end, hopefully fixing a use-after-free.
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
| |
While here, perform a more aggressive close of the pipe on
reaping (IPC).
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
| |
fixes #615 IPC close on Windows leaves handle open
This reintroduces the code to eradicate the separate transport
start function for IPC, as a incremental step towards the full
fix for 599 and 208. It also addresses 615, by including revised
logic for the handling of close.
|
| |
|
|
|
|
| |
This adds a proper boolean condition for the pipe being closed
(removing the unused sending flag), and adds checks for both the
pipe closed and the socket closed flags at key points.
|
| |
|
|
|
|
|
|
|
| |
This converts the tests to use async I/O callbacks instead of
threads for running tests. This should greatly reduce the
amount of pressure we apply to the system. On macOS the
time to start up with a pressure of 500 is significantly less
than under the old system. Plus, as we are no longer at the mercy
of the scheduler, we're far more likely to get a successful test.
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
fixes #622 incorrect assumptions about malloc(0)
Windows actually allocates an object of size zero when calling
malloc on size zero. This is unusual behavior, and we just
add logic to work more like malloc on POSIX systems.
Other systems can return non-NULL objects to fixed pages here.
We think the best option here is to uniformly return NULL from
our APIs in these circumstances, and to include testing to validate
that.
|
| | |
|
| | |
|
| |
|
|
|
| |
This changeset needs work. We are seeing errors described by
This reverts commit d7f7c896c0ede24249ef63b1e45b1878bf4bd473.
|
| |
|
|
|
|
|
|
|
|
| |
fixes #208 pipe start should occur before connect / accept
fixes #616 Race condition closing between header & body
This refactors the transports to handle their own connection
handshaking before passing the pipe to the socket. This
changes and simplifies the setup. This also fixes a rather
challenging race condition described by #616.
|
| | |
|
| |
|
|
| |
On QNX, specifying a numeric servname while leaving ai_socktype unspecified would result in EAI_SERVICE.
|
| | |
|
