A mirror of https://github.com/nanomsg/nng https://cogarr.net/source/cgit.cgi/nng/atom?h=v1.0.0-beta.1 2018-05-18T15:16:01Z 2018-05-18T15:16:01Z Garrett D'Amore garrett@damore.org 2018-05-18T15:16:01Z urn:sha1:c13d79054e219a8739db4e2dda668a785211c758 We've added a new macro, NNG_RELEASE_SUFFIX. If this is not empty, then the release should be considered a pre-release, with no specific stability guarantees. 2018-05-18T04:52:07Z Garrett D'Amore garrett@damore.org 2018-05-18T03:41:06Z urn:sha1:846c30081a67e961b4a060bdca192ddafb87cce9 This changes nni_task_fini to always run synchronously, waiting for the task to finish before cleaning up. Much simpler code. Additionally, we've refactored the resolver code to avoid the use of taskqs, which added complexity and inefficiency. The approach of just allocating its own threads and a work queue to process them turns out to be vastly simpler, and actually reduces extra allocations and context switches. wip POSIX resolv threads. (Taskqs are just overhead and complexity here.) Windows resolver changes. Task cleanup. fix up windows mutex. 2018-05-18T03:20:17Z Garrett D'Amore garrett@damore.org 2018-05-18T03:20:17Z urn:sha1:109a559590abfe3017dd317c3068e2457188541c 2018-05-18T02:29:37Z Garrett D'Amore garrett@damore.org 2018-05-17T19:54:01Z urn:sha1:70d478f5d185e147ca8d3dcba4cbd8bb6da3719a This changes the signature of nng_pipe_notify(), and the associated events. The documentation is updated to reflect this. We have also broken the lock up so that we don't hold the master socket lock for some of these things, which may have beneficial impact on performance. 2018-05-17T16:44:39Z Garrett D'Amore garrett@damore.org 2018-05-17T04:46:56Z urn:sha1:e490aa3353f05e158a0f1f485f371cd49e70b4f5 2018-05-17T04:08:55Z Garrett D'Amore garrett@damore.org 2018-05-16T21:12:29Z urn:sha1:eab6a4d2d96d11d3926e927c135362fc166895f0 This changes the array of flags, which was confusing, brittle, and racy, into a much simpler reference (busy) count on the task structures. This allows us to support certain kinds of "reentrant" dispatching, where either a synchronous or asynchronous task can reschedule / dispatch itself. The new code also helps reduce certain lock pressure, as a bonus. 2018-05-16T18:42:06Z Garrett D'Amore garrett@damore.org 2018-05-16T17:52:55Z urn:sha1:ec84c5a8406bb203d5f8830e280f93cf1f63cd6a We also fixed some warnings found by gcc8. 2018-05-16T17:03:38Z Garrett D'Amore garrett@damore.org 2018-05-16T17:03:38Z urn:sha1:84c32ea006f01f10cea335a9b152ec68714f46c9 2018-05-16T02:23:05Z Garrett D'Amore garrett@damore.org 2018-05-16T02:23:05Z urn:sha1:59ee3628d5db511ce5527ce4ddbf126e929d7da8 fixes #438 Consider dropping AI_V4MAPPED 2018-05-16T01:06:15Z Garrett D'Amore garrett@damore.org 2018-05-15T22:43:10Z urn:sha1:f69b4c1da69d1e9c513ad63da488a49b6b46a706 While here, there is a little more endpoint termination hardening. Running this code base seems to not incur any faults beyond the very rare TCP port conflict that can occur from our random port selection in the test suite.