files involved in a given VCL.

When using -v, the individual source files are output like this:

	// VCL.SHOW %d %d %s\n%s\n

First field is the source file index [0...].
Second field is the number of bytes in the source file.
Third field is the name of the source file.
Fourth field is the source file content.

Inspired by a patch from daghf

Backend restarts are now return(retry).

Fixes:	#1503

Submitted by:	Gauthier Delacroix via github

sphinx-build v0.6.6 which is on RHEL6/CentOS6 does not know about -n,
so we'll have to do without it for now.

Reported in:	#1535

This reflects the debian build files, editline will suffice.

We only need debhelper when building the debian packages,
and the systemd-units is within a conditional that awk ignores.

Updated the commands to not include these in the future, and also
the Redhat spec file so it should work on a fresh installation.

Noticed by:	fgs

The list of documented build requirement have not been updated
for a while, and was outdated.

Update the lists with recent data.

Added (commented-out) commands that extracts and formats
this information directly from the packaging files.

Reported by:	frakt

Adhere to Fedora packaging standard.

or an string-insertion-sort feels right for me.

Using reserved workspace to tracking the params and qsort(3) to sort them
solves both problems.

until it's been allocated.

This patch delays the VXID allocation until the time we know that we
will actually be logging anything for this request (not completely all
whitespace noise). This keeps the code in one place only (plus the ESI
mock request handling), instead of separated between the linger wait,
cache acceptor and pipelining code paths.

Coincidentally fixes the logging of timestamps problem in #1588.

Test case by daghf.

Fixes: #1581 #1588

Conflicts:
	bin/varnishd/cache/cache_esi_deliver.c
	bin/varnishd/cache/cache_fetch.c
	bin/varnishd/cache/cache_http1_fsm.c
	bin/varnishd/cache/cache_session.c

This enables catching batch related log problems using logexpect.

This commit will make several test cases start failing until #1588 has been fixed.

healthy backend.

Fixes: #1575

Call HTTP1_DiscardReqBody() in cnt_synth before delivery to deal with
any lingering request body.

Fixes: #1577

Submitted by:       nublaii via github

Submitted by:	Clement Gautier via github

Fixes #1532

Reinstall the readline handler when we are called in order to clear the
input line. Tested in a variety of versions by scn and myself.

Fixes #1531

beresp.cacheable is gone.

Patch by ingvar@redpill-linpro.com

https://www.varnish-cache.org/patchwork/patch/145/

busyobj if we want to examine it in any detail.

Hold that ref over all of V1D_Deliver() (different from Nils
patch) to make sure we have a consistent view on streaming
vs. no streaming throughout.

Also use the bo to clean up a hack related to ESI includes.

Diagnosed by:	Nils Goroll

As discussed on the latest VDD.

While here unify check.