nginx.git/auto/unix, branch release-1.9.2

The "reuseport" option of the "listen" directive.

2015-05-20T12:51:56+00:00

When configured, an individual listen socket on a given address is
created for each worker process.  This allows to reduce in-kernel lock
contention on configurations with high accept rates, resulting in better
performance.  As of now it works on Linux and DragonFly BSD.

Note that on Linux incoming connection requests are currently tied up
to a specific listen socket, and if some sockets are closed, connection
requests will be reset, see https://lwn.net/Articles/542629/.  With
nginx, this may happen if the number of worker processes is reduced.
There is no such problem on DragonFly BSD.

Based on previous work by Sepherosa Ziehau and Yingqi Lu.

Configure: style.

2015-05-20T12:51:13+00:00

Events: implemented epoll notification mechanism.

2015-03-14T14:37:13+00:00

Core: expose maximum values of time_t and ngx_int_t.

2015-03-16T21:24:34+00:00

These are needed to detect overflows.

Events: use eventfd() instead of syscall(SYS_eventfd) if possible.

2014-05-23T12:37:05+00:00

This fixes --with-file-aio support on systems that lack eventfd()
syscall, notably aarch64 Linux.

The syscall(SYS_eventfd) may still be necessary on systems that
have eventfd() syscall in the kernel but lack it in glibc, e.g.
as seen in the current CentOS 5 release.

Added support for TCP_FASTOPEN supported in Linux >= 3.7.1.

2013-12-03T18:07:03+00:00

---
 auto/unix                       | 12 ++++++++++++
 src/core/ngx_connection.c       | 32 ++++++++++++++++++++++++++++++++
 src/core/ngx_connection.h       |  4 ++++
 src/http/ngx_http.c             |  4 ++++
 src/http/ngx_http_core_module.c | 21 +++++++++++++++++++++
 src/http/ngx_http_core_module.h |  3 +++
 6 files changed, 76 insertions(+)

Configure: TCP_KEEPIDLE test name simplified.

2013-09-04T17:17:05+00:00

Core: guard use of AI_ADDRCONFIG.

2013-08-05T09:44:56+00:00

Some systems (notably NetBSD and OpenBSD) lack AI_ADDRCONFIG support.

Reported by Piotr Sikora.

Core: only resolve address families configured on the local system.

2013-08-05T06:55:59+00:00

This is done by passing AI_ADDRCONFIG to getaddrinfo().

On Linux, setting net.ipv6.conf.all.disable_ipv6 to 1 will now be
respected.

On FreeBSD, AI_ADDRCONFIG filtering is currently implemented by
attempting to create a datagram socket for the corresponding family,
which succeeds even if the system doesn't in fact have any addresses
of that family configured.  That is, if the system with IPv6 support
in the kernel doesn't have IPv6 addresses configured, AI_ADDRCONFIG
will filter out IPv6 only inside a jail without IPv6 addresses or
with IPv6 disabled.

Implemented IPv6 support for URLs specified using domain names.

2012-12-17T12:08:53+00:00

This includes "debug_connection", upstreams, "proxy_pass", etc.
(ticket #92)

To preserve compatibility, "listen" specified with a domain name
selects the first IPv4 address, if available.  If not available,
the first IPv6 address will be used (ticket #186).