shadowing).

client has closed the connection and gone away. If so, release the
session early.

Fixes: #1252

This will help in determining remote hang up of the connection for
situations where we still are not able to send any reply, but freeing
the session will reduce resource overhead (e.g. when staying on
waitinglists for extended periods).

On platforms that support it also add runtime parameters to control
the keep-alive packet settings through socket options. On platforms
that don't support these socket options, the values must be set system
wide.

The Varnish runtime parameters will only be applied when they are less
than the system default.

duplicate counters consistent on reloads.

Add a test case for this.

Add a section on bans and offline silos in the docs.

be recognized as such on reload

the header, not 0).

Add a test case for reload of truncated bans.

Fixes: #1266

This isn't really functional right now, due to the log record changes,
but it will allow you to specify -b again and you can have a different
format for backend requests than for client requests.

accept-sockets can actually be bound before we muck about with
daemon() or SHM files.

Belived to be implicated in #1262

Fixes #1259

namespace for vmod.

once I get the other compiler to cope too.

This snapshot is committed to solicit feedback, so please do :-)

features.

As per discussion at the Varnish dev meeting, seeing Cache-control:
no-cache (and friends) now disable caching by default.  As a twist, we
respect Surrogate-control if that is present, in the default VCL.

Fixes: 1124
References: 477

If we don't add -c and -b in there, varnishlog will print all
transactions that aren't associated with a session.  That is likely to
not be what the user wants.

Fixes: #1071

poll can return EINTR, so we need to handle that.

Fixes: #1166

Fixes #1198

good, but at least it belongs here.

Fixes	#1255

Eliminate SIGCHLD usage, it's icky, at best, when we have other
child processes (See #1256)

Instead of reaping the child on SIGCHLD, we do it explicitly,
and there is now a 10 second wait to give the child a chance to
shut down gracefully, before we take a bat to the kneecaps.

More work may be warranted, but I want to get some feedback on
this bit first.

Fixes	#1256

Add a read-method to http_conn, and use it from all VFPs, which
should be protocol agnostic.

from HTTP1 specific functionality.

resp.*

Submitted by: jonathan.huot@thomsonreuters.com