It was due to thinko in iterating through keys set - we were using key
ordinals instead of keys themselvs. Now it find such bugs:

=== RUN   TestΔBTailAllStructs/Update/T1,3/B0:d-B1:a-B3:d,4:g→T1,4/B0:e-B1:a-B4:c
    δbtail_test.go:839: 2: trackSet:
                have: map[000000000000000c:{pffffffffffffffff c2} 00000000000005c2:{p000000000000000c c0} 0000000000000618:{p000000000000000c c0}]
                want: map[000000000000000c:{pffffffffffffffff c3} 00000000000005c2:{p000000000000000c c0} 0000000000000618:{p000000000000000c c0} 0000000000000620:{p000000000000000c c0}]
    δbtail_test.go:708: T1,3/B0:d-B1:a-B3:d,4:g -> T1,4/B0:e-B1:a-B4:c  ; tracked={0 1} :

        d12:  map[0:{d e} 3:{d ø} 4:{g c}]
        δTok: map[0:{d e} 3:{d ø} 4:{g c}]
        δT:   map[0:{d e}]

        δZ:               {000000000000000c 0000000000000618 0000000000000620}
        Tracked^δZ:       {0 1}
        kadj[Tracked^δZ]: {0 1 3 4 9223372036854775807}
        kadj: map[0:{0} 1:{1 3 4 9223372036854775807} 3:{1 3 4 9223372036854775807} 4:{1 3 4 9223372036854775807} 9223372036854775807:{1 3 4 9223372036854775807}]

        δtkeycov wrong:
        have: {}
        want: {[3,∞)}
        δT    ∌  δTok[3]
        δT    ∌  δTok[4]

Currently fails with e.g.:

(neo) (z-dev) (g.env) kirr@deca:~/src/neo/src/lab.nexedi.com/nexedi/wendelin.core/wcfs/internal/xbtree$ go test -short -v -failfast -run 'TestΔBTail$/Update/T3/B1:a-B8:c→T7/B1:a-B8:c'
=== RUN   TestΔBTail
2021/06/23 14:08:45 zodb: FIXME: open /tmp/δBTail926507346/1.fs: raw cache is not ready for invalidations -> NoCache forced
=== RUN   TestΔBTail/Update/T3/B1:a-B8:c→T7/B1:a-B8:c
    δbtail_test.go:812: BUG: tkeyCov1 ⊄ tkeyCov2:
                tkeyCov1: {[3,∞)}
                tkeyCov2: {[7,∞)}
                tkeyCov1 \ tkeyCov2: {[3,7)}
    δbtail_test.go:708: T3/B1:a-B8:c -> T7/B1:a-B8:c  ; tracked={8} :

        d12:  map[]
        δTok: map[]
        δT:   map[]

        δZ:               {000000000000000c}
        Tracked^δZ:       {}
        kadj[Tracked^δZ]: {}
        kadj: map[1:{1 8 9223372036854775807} 8:{1 8 9223372036854775807} 9223372036854775807:{1 8 9223372036854775807}]

--- FAIL: TestΔBTail (0.84s)
    --- FAIL: TestΔBTail/Update/T3/B1:a-B8:c→T7/B1:a-B8:c (0.00s)

=== RUN   TestΔBTail/Update/T7/B1:a,2:b-B8:c,9:d→T3/T-T/B1:a,2:b-B8:c,9:d
    δbtail_test.go:832: 2: trackSet:
                have: map[000000000000000c:{pffffffffffffffff c2} 0000000000000041:{p0000000000000043 c0} 0000000000000042:{p0000000000000044 c0} 0000000000000043:{p000000000000000c c1} 0000000000000044:{p000000000000000c c1}]
                want: map[000000000000000c:{pffffffffffffffff c1} 0000000000000041:{p0000000000000043 c0} 0000000000000043:{p000000000000000c c1}]
    δbtail_test.go:708: T7/B1:a,2:b-B8:c,9:d -> T3/T-T/B1:a,2:b-B8:c,9:d  ; tracked={1} :

        d12:  map[]
        δTok: map[]
        δT:   map[]

        δZ:               {000000000000000c 0000000000000043 0000000000000044}
        Tracked^δZ:       {}
        kadj[Tracked^δZ]: {}
        kadj: map[1:{1 2 8 9 9223372036854775807} 2:{1 2 8 9 9223372036854775807} 8:{1 2 8 9 9223372036854775807} 9:{1 2 8 9 9223372036854775807} 9223372036854775807:{1 2 8 9 9223372036854775807}]

        δtkeycov wrong:
        have: {[7,∞)}
        want: {}

* t:
  .
  X wcfs: Fix ZSync to close wconn on zdb.close, even if zconn stays alive
  X lib/zodb: Connection += onShutdownCallback
  .
  X wcfs: lsof +D misbehaves - don't use it
  X wcfs: _fuse_unmount: Try first `kill -TERM` before `kill -QUIT` wcfs
  X wcfs: Tune _fuse_unmount to include `fusermount -u` error message into raised exception
  X wcfs: Teach start to start successfully even after unclean wcfs shutdown
  fixup! X wcfs: Run fusermount and friends with /bin:/usr/bin always on path
  X wcfs: Run fusermount and friends with /bin:/usr/bin always on path
  X wcfs: Add start to spawn a Server that can be later stopped  (draft)

Even if ZODB DB is closed and releases zconn from its .pool, zconn can
still stay alive being referenced from any other live python object -
e.g. from a frame, a traceback etc. This situation in particular happens
under ERP5's runUnitTest.

As the result, if we don't hook into DB.close and close wconn only on
zconn GC, files opened on WCFS filesystem stay opened, and WCFS server
cannot be cleanly unmounted when test run completes.

-> Fix it.

P.S. ZODB@bbd03b3a is good, but does
not really solve this problem because, once again, zconn stays
referenced from objects besides DB.

In the next patch we'll need this functionality to subscribe ZSync into
db.close to know when ZODB Connection is shut down even if it stays
alive referenced by some other objects.

For example even with live wcfs lsof +D output differes from regular lsof and is misleading:

    LSOF /dev/shm/wcfs/b53b61099c740b452b383db6df6dce4ad6d23ba2:

    COMMAND     PID       USER   FD   TYPE DEVICE SIZE/OFF NODE NAME
    runUnitTe 29137 slapuser34   24r   REG   0,48      111    4 /dev/shm/wcfs/b53b61099c740b452b383db6df6dce4ad6d23ba2/.wcfs/zurl
    runUnitTe 29137 slapuser34   25u   REG   0,48        0    7 /dev/shm/wcfs/b53b61099c740b452b383db6df6dce4ad6d23ba2/head/watch
    runUnitTe 29137 slapuser34   26r   REG   0,48  2097152    9 /dev/shm/wcfs/b53b61099c740b452b383db6df6dce4ad6d23ba2/head/bigfile/00000000000077e2

    LSOF +D /dev/shm/wcfs/b53b61099c740b452b383db6df6dce4ad6d23ba2:

    COMMAND     PID       USER   FD   TYPE DEVICE SIZE/OFF       NODE NAME
    runUnitTe 29137 slapuser34   24r   REG   0,48      111 4294967295 /dev/shm/wcfs/b53b61099c740b452b383db6df6dce4ad6d23ba2/.wcfs/zurl
    runUnitTe 29137 slapuser34   25u   REG   0,48        0 4294967295 /dev/shm/wcfs/b53b61099c740b452b383db6df6dce4ad6d23ba2/.wcfs/zurl
    runUnitTe 29137 slapuser34   26r   REG   0,48  2097152 4294967295 /dev/shm/wcfs/b53b61099c740b452b383db6df6dce4ad6d23ba2/.wcfs/zurl

Just aborting the FUSE connection does not make WCFS to exit. This abort
is needed to avoid deadlocks, but we should also signal WCFS it should
stop. Not doing so will lead to timeout in wait after FUSE connection
abort and `kill -QUIT` triggered, which makes WCFS crash and print lots
of traceback to stderr before exiting.

Previously that message was printed to stderr and it was very confusing:
we had something in the exception message, something in the log and
something in stderr.

Now exception message is self-sufficient and log provides more details.

This patch does to start what b0ca031f did for join/serve.

Stop duplicating the code and factor logic to "attach to wcsrv, or
prepare to start" into _try_attach_wcsrv function.

For example runUnitTest in ERP5 runs test program with $PATH set without those
directores. This leads to failures like

      File "/srv/slapgrid/slappart16/srv/testnode/djk/soft/950bed0366ff018b829e9bbb7dc3326f/parts/wendelin.core/wcfs/__init__.py", line 533, in _is_mountpoint
        mounted = (0 == subprocess.call(["mountpoint", "-q", path]))
      File "/srv/slapgrid/slappart16/srv/testnode/djk/shared/python2.7/a8bdae732e550b385eb04a06ed09d383/lib/python2.7/subprocess.py", line 172, in call
        return Popen(*popenargs, **kwargs).wait()
      File "/srv/slapgrid/slappart16/srv/testnode/djk/shared/python2.7/a8bdae732e550b385eb04a06ed09d383/lib/python2.7/subprocess.py", line 394, in __init__
        errread, errwrite)
      File "/srv/slapgrid/slappart16/srv/testnode/djk/shared/python2.7/a8bdae732e550b385eb04a06ed09d383/lib/python2.7/subprocess.py", line 1047, in _execute_child
        raise child_exception
    OSError: [Errno 2] No such file or directory

where those system programs are not found.

Implement Server.stop using most of the functionality that was
previously living in tWCFS test part.