summaryrefslogtreecommitdiff
path: root/tests/compat_reqttl.c
Commit message (Collapse)AuthorAge
* Move compatibility header so that <nanomsg/nn.h> works.Garrett D'Amore2018-02-23
| | | | | | Basically, we have moved the compat stuff into a separate directory. Compatibility layer users will have to update their compile flags, but should be able to avoid changing any *source* files with this change.
* REP drops peers a little too aggressively.Garrett D'Amore2017-08-14
| | | | | | | | We noticed that certain failure modes were exposed in tests that were caused by us closing the underlying pipe when certain messaging errors occurred. Discarding the pipe is the wrong answer; instead we should discard the message and keep the pipe open (unless the message is so malformed that the remote party cannot be trusted.)
* fixes #47 compat_reqttls fails sometimesGarrett D'Amore2017-08-09
| | | | fixes #23 Restore the old idhash logic for sockets
* Use inproc to minimize timeouts on TCP (Windows).Garrett D'Amore2017-08-08
|
* Increase the timeouts for compat_reqttl (context switches?)Garrett D'Amore2017-08-08
| | | | | Also, we need to nn_close(), because the close can reasonably race against nn_term(), causing false test failures.
* Added nn_compat code for option handling, fixed other bugs.Garrett D'Amore2017-08-08
Hop counts for REQ were busted (bad TTL), and imported the compat_reqtll test. At the same time, added code to nn_term to shut down completely, discarding sockets. (Note that some things, such as globals, may still be left around; that's ok.)