Note XXX relative to #3202

Usage pattern:

     struct vsb vsb[1];
     char *p;

     WS_VSB_new(vsb, ctx->ws);
     VSB_printf(vsb, "blablabla");
     p = WS_VSB_finish(vsb);
     if (p == NULL)
             return (FAILURE);

Does this stabilize a vtest false negative which I do not understand?

There are two one-second delays on the server side, but varnishd might
read the response a bit longer after the first delay such that the
second interval is shortened from the varnish perspective.

Ref #3210

Other man pages seem to have bogus dependencies.

And harmonize the reStructuredText format of default values entries.

The assert on WS_ReserveSize() in ses_handle() can not trip because
sizeof (struct pool_task) is less than sizeof (struct waited). But to safe
guard against future problems if that were to change, this patch makes
sure that the session workspace can hold the largest of them before
entering the waiter, erroring out if not.

Proxy TLVs claiming to have PP2_TYPE_SSL sub-TLVs without complete payload
would cause a Varnish assert. This patch fixes the parsing of the TLVs.

h2_init_sess can only be reached through H1 with either previous knowledge
or opportunistic upgrade. Because of this the proto_priv session attribute
will always be set before entry. This patch simplifies and removes dead
code containing a call to SES_Reserve_proto_priv.

Note: Better diff with the --ignore-all-space option

In h2_init_sess, an extra call was always made to SES_Reseve_proto_priv(),
even though it was already reserved. This wasted a pointer worth of
session workspace. This patch removes the extra call.

If proxy protocol is in use, it is possible to fill the session workspace
exactly before entering http1_new_session(), which will cause it to assert
when calling SES_Reserve_proto_priv().

with this patch we will close the session gracefully.

as decided by bugwash.

Closes #3203

Sorry, I had missed a fixup before push

Basically taken from #3204 by @gquintard

- sphinx install for release 8
- included libunwind though it is optional
- left out libconfig libcurl and openssh-clients because
  varnish-cache does not need them

Looking at 3ec9b582 I realized that this single sentence was packed
with mistakes: allocation should also have been plural and it wasn't the
third but fourth allocation that would fail (and of course the frist
typo).

I would have wanted to use a protocol agnostic URL, but it seems sphinx
does not support them (//varnish-cache.org/... is interpreted as a URL).
If anyone knows better, please improve.

(I'm assuming these two were typos, if they were intentional, I would
have expected explanatory comments.)

(actually not on 32bit, but where the default workspace is low)

Ref #3203

related to vtest failures

Ref #3194

It now bails out due to a workspace overflow which it acually did
trigger all the time, but that remained unnoticed due to #3194

The test to multiply the cookie header 128 times (8x in the first
regex, 16x in the second) was not realistic even for "modern times",
so I do not think we should have workspaces sized by such an example.

Since 8baf4a69 we lost information about
an overflowed workspace by calling std.log(), which resulted from use of
the workspace for construction of a contiguous string from the argument
constituents.

Since then, we have changed the interface to STRANDS, but this issue
remained.

We now solve the case for real by pushing the string concatenation down
to VSL: New versions of VSL and VSLb (coded by example of VSLv() and
VSLbt()) take a strands argument and create a log record without
additional copy overhead.

These solve #3194 for std.log(), make logging more efficient and, in
particular, allow use of std.log() in low workspace conditions (because
they do not require any).

Also improve test coverage for std.log()

Ref #3194

note: no change to vrt.h, this is a private interface between vcl and
vrt_vcl/vpi.

... potentially saving a lock/unlock on the critical vcl_mtx

Both callers of vcl_get(), VPI_vcl_select() and VCL_Refresh(),
potentially have to unref another vcl first.

For VCL_Refresh(), this happens when the active vcl changes and the
previously active vcl is cached in (struct worker).vcl, which can be
assumed to always be the case on a busy system.

For VPI_vcl_select(), this happens when switching vcls at
esi_level > 0.

To summarize, this patch will primarily reduce contention on the
vcl_mtx after a 'vcl.use' CLI command on a busy system and for
'return(vcl(...))' with ESI.

This is already implied in the code (because otherwise we would
potentially leak a vcl reference (busy counter)).

(yes, I have run make check with this)

We should not consider an early close of the backend connection a
failure of the test, this vtc is not concerned with bodies.