sdjournal: fix concurrent map access

libsystemdFunctions is shared between Journal instances than may cause
concurrency issues accessing the same functions at the same time for the
very first time

sdjournal: fix race condition in test

since the go builtin command  creates a closure, it uses the same err variable than causes race condition in tests

sdjournal: free correct pointer

Since defer creates a closure, it ends up freeing the wrong
pointer in these situations (i.e. not the one that ends up
getting allocated by asprintf during the actual sdjournal
call).

sdjournal: fix memory leak in cursor handling

free memory that's allocated by the sd_journal_get_cursor function

bugfix

activation: don't specify NextProtos so that we can take advantage of HTTP/2

util: fix import path

dbus: add PropPids to set initial pids of transient scope units

util: conditionally build CGO functions

Some projects don't need the CGO functions in the util package;
split out the functions into a separate file so CGO_ENABLED=0
works if those functions are never used.

Fixes #171

dbus: add mask/unmask for unit files

daemon: improve SdNotify errors signaling

After this patch, SdNotify returns values compliant to sd_notify(3).

Return values:
* (false, nil) - notification not supported
* (false, err) - notification supported, but failure happened
* (true, nil) - notification supported, data has been sent

sdjournal: add GetDataBytes() and GetDataValueBytes() to Journal

This allows to avoid extra string → []byte conversion overhead for cases
when journal field value is e.g. parsed (which usually requires []byte).

sdjournal: fix a race in GetEntry test

travis: run whole testsuite in containers

dbus: fix testsuite

sdjournal: fix an implicit seek to head

When a JournalReader is initialized with a NumFromTail bigger than
the number of available entries, cursor will be positioned on the
first entry. Upon initial (mandatory) Next invocation, it will
incorrectly start reading from the second entry.

This fixes a pathological case where the first entry of a journal
is always ignored when using NumFromTail.

Some assumptions do not hold anymore on recent debian-derivatives.
Tests updated to avoid failures in modern dbus/systemd environments.

sdjournal: add test for incorrect Read() length

sdjournal: fix incomplete reads in JournalReader.Read()

Rework configuration to run testsuite into containerized
environments. This allows to run all tests on travis.

Hard-sleep instead of waiting for a journal event.
This fixes a race due to waiting for any events but enumerating only
matching ones.

sdjournal: fix Writer call

sdjournal: Break the loop if error happens during Write().

This patch fixes two issues in JournalReader.Read():

1. The return value of copy() was ignored and JournalReader.Read() was
returning the wrong amount of bytes copied.

2. JournalReader implements the Reader interface so it must be able to read a
specific amount of bytes. Journald on the other hand only allows us to read
full entries of arbitrary size (without byte granularity). JournalReader
therefore needs to internally buffer entries that don't fit in the read buffer.
Callers should keep calling until 0 and/or an error is returned.

Fixes https://github.com/coreos/go-systemd/issues/172