The check is added to the builtin logic for now.

Fixes #2631.

or should we use "%" PRIu64 in the format ?

reported by Travis via @fgsch

The blind caching is OBE for six years now, see
344a709c

the schema intended for JSON output.

to avoid micro-managing memory

Follow up cf20e04e

required after the constification in
8f83087f

I have no idea why the umem authors did not make this argument
a const, they call strncpy on it

A show stopper used to be that the top request's workspace
was reserved for delivery during ESI processing, but since
beeaa19c the path is clear.

spotted by @lkarsten

Ref #2672

After recent failures discovered by travis-ci it should be safe to
have this back on.

rely on h2_errcheck() calls anyway.

it will ever be useful for.

It complains that &pfh_vut is never NULL...

For out-of-tree code that may set up multiple VUTs in a single process,
we don't want the VUT that initialized a PID file to remove it if it
finishes before other VUTs.

A global pseudo-VUT is introduced to keep track of the PID file and
the error callback in order to use them in the atexit(3) callback.

Original test case by Geoff, slightly modified.

Closes #2651

This can happen if a signal is caught after `VUT_Fini` was called, but
at this point we should no longer manipulate the VUT being dismantled.

Original test case by Geoff, slightly modified.

Closes #2650