... to save stack space

Fixes #2643

Ref #2643

Yes, I know, this is far from perfect, but it does not require
any changes outside the signal handler and will hopefully simplify
error report handling significantly.

Ref #2643

Spotted by:	fgs

purports to say on the tin, and by removing the quantum-test
it trips over.

better sense of it now.

Ref #2622

running the transaction.

almost certainly an error, so make it one.

Fixes #2613

fixes #2639

Not real ids; for consistency with the other platforms.

note on jail_solaris:

ideally, vmods could tell the jails about privileges they require, but
for now (and because vmod_unix lives in varnish-cache), just add the
required privilege to the permitted set.

I have also considered the option to add additional privileges via the
-j argument (and actually would still want to add that somewhen), but
for this purpose, varnish should really DTRT by default.

note on priv_allocset:

Most of the sun folk had proven to be good interface designers, but
an API which requires dynamic allocation/deallocation really does not
play well with my efficiency fetish. So good we got library
constructors/destructors.

note on getpeerucred:

basically the same thing, but this time they at least added
ucred_size(). Not intended for use like this, but anyway....

Negative character constant

Last -need-bytes kept to test that it still works.

Also reappropriate test a00001 to exercise the Teken code with
the vttest (http://invisible-island.net/vttest/) program, if installed.

This allows us to use TERM=xterm.