While at it, get rid of the hard-coded package name.

Refs #2251

for the gcov vtester to finish in 60 seconds.

So far this what I find in the commit log from varnish-5.0.0 until
the beginning of 2017. Some "biggies" still to be mentioned are:

* varnishd command-line options
* VCl type conversions, stringification etc.
* ranges?

It closes a race for barriers because the list could be updated by the
top thread while walked by another thread. It also saves useless lookups
by rejecting names earlier.

This is not the case for servers because a) they have proper locking of
the servers list and b) there is the dispatch special case. It would
otherwise be harmless to do it for the server too because dispatched
servers don't depend on user input.

We aren't building a vrt_test binary, it's a script. We also don't need
to distribute vrt.c, the cost of a copy doesn't matter and it ensures
that this file is always in the $(buildir). I also cleaned other wrong
assumptions regarding $(srcdir) vs $(builddir).

This reverts commit e706b22a. The real
problem was solved in d906981a. The
extension adds noise considering the make rules below don't comply with
the target name.

With a more assertive test this time.

Primarily to document the varieties of pass (pass, hit-for-miss and
hit-for-pass).
Reviewed by Dridi.

It had subsections of the Cookies chapter that had nothing to do with them.

Fixes #2250  (again)

Instead of passing it by value, make sure to have one variable holding
the file descriptor. This should help Coverity stop complaining about a
resource leak.

The steps:

    9.  Condition vbp_write_proxy_v1(vt, s) != 0, taking true branch.
    10. leaked_handle: Handle variable s going out of scope leaks
        the handle.

If the return value is non-zero the socket is VTCP_close'd.

Refs #2244

Fixes #2250

Some lowhanging Flexelint nitpicking.

The qualifiers were accidentally lost in 17a456bd, because barriers
became the template and were the only non-const occurrence.

A command like `barrier r1` would fail while holding the lock because of
the invalid name, but the reset of the test case would then deadlock. It
leads to a timeout like we see sometimes on VTEST so getting rid of them
may just require to review when `vtc_fatal` is called while locking.

In this very case we can only create barriers in the top thread so the
lock is useless.

It takes the barrier as a template and provides a consistent error
message logged with the caller's vtclog. The name is the first thing
checked when a new named entity is created, saving useless setup.

Fixes #2249