But leave LLTempRedirect available in the code base.

LLSD::emptyMap() is a factory for an empty map instance, NOT a predicate on
any particular instance. In fact checking configuration.isUndefined() and
testing whether the map is empty are both subsumed by (! configuration).

That way, if there's a problem, a developer can rerun the same command.

Specify all of msvcp$VER.dll, msvcr$VER.dll and vcruntime$VER.dll -- but check
each of them individually, because any given VS release has only a subset of
those. Add messaging to clarify what we're doing.

Introduce to_staging_dirs CMake macro to cut down on redundant boilerplate:
the idiom in which we use copy_if_different twice, once to the Release staging
directory and once to the RelWithDebInfo staging directory, each time
appending the target pathnames to third_party_targets. Replace that idiom with
calls to to_staging_dirs.

However, this is not the right moment to perform that refactoring.

Evidently, with VS 2017, what would have been msvcr140.dll has become
vcruntime140.dll instead. msvcr140.dll is no longer a good sample DLL for
which to search.

Thanks NickyD.

Also forget obsolete references to VS 2010 runtime DLLs.

LLAppViewer's heap LLWatchdogTimeout might be destroyed very late -- as late
as in LLAppViewer's destructor. By that time, LLAppViewer::cleanup() has
already called LLSingletonBase::deleteAll(), destroying the LLWatchdog
LLSingleton instance.

But LLWatchdogTimeout isa LLWatchdogEntry, and ~LLWatchdogEntry() calls
stop(), and stop() tries to remove that instance from LLWatchdog, thus
inadvertently resurrecting the deleted LLWatchdog. Which is pointless because
the resurrected LLWatchdog has never heard of the LLWatchdogTimeout instance
trying to remove itself.

Defend LLWatchdogEntry::stop() against the case in which LLWatchdog has
already been deleted.

Make the LLError::Settings LLSingleton duplicate the file handle for stderr
(usually 2) on construction. Make its destructor restore the original target
for that file handle. Provide a getDupStderr() method to obtain the duplicate
file handle.

Move Settings declaration up to the top of the file so other code can
reference it.

Make RecordToFile (the Recorder subclass engaged by LLError::logToFile()),
instead of duplicating stderr's file handle itself, capture the duplicate
stderr file handle from Settings to revert stderr redirection on destruction.

Make RecordToStderr (the Recorder subclass engaged by LLError::logToStderr())
use fdopen() to create an LLFILE* targeting the duplicate file handle from
Settings. Write output to that instead of to stderr so logToStderr() continues
to provide output for the user instead of duplicating each line into the log
file.

Add ndof_dump_list() call, to enumerate available devices, when
ndof_init_first() returns failure.

Add "joystick" tags to existing LL_INFOS() (etc.) calls in
llviewerjoystick.cpp to make it easier to enable and disable such log
messages.

Add a specialized operator<<() function to log the contents of an NDOF_Device
struct. Add a couple LL_DEBUGS() calls for more visibility into library
operations.

Some of the libraries we use produce log output to stderr. Such output can be
informative, but is invisible unless you launch the viewer from a console. In
particular, it's invisible to anyone trying to diagnose a problem by reading
someone else's SecondLife.log file.

Make RecordToFile -- the Recorder subclass engaged by LLError::logToFile() --
redirect STDERR_FILENO to the newly-opened log file so that any subsequent
writes to stderr (or cerr, for that matter) will be captured in the log file.
But first duplicate the original stderr file handle, and restore it when
RecordToFile is destroyed. That way, output written to stderr during the final
moments of application shutdown should still appear on (console) stderr.

LLUniqueFile wraps an LLFILE* in a move-only class that closes the wrapped
LLFILE* on destruction. It provides conversion operators to permit idiomatic
usage as an LLFILE* value.

Since the consuming coroutine LLCoprocedurePool::coprocedureInvokerCoro() has
been observed to outlive the LLCoprocedurePool instance that owns the
CoprocQueue_t, closing that queue isn't enough to keep the coroutine from
crashing at shutdown: accessing a deleted CoprocQueue_t is fatal whether or
not it's been closed.

Make LLCoprocedurePool store a shared_ptr to a heap CoprocQueue_t instance,
and pass that shared_ptr by value to consuming coroutines. That way the
CoprocQueue_t instance is guaranteed to live as long as the last interested
party.

Instead of heap-allocating a CoroData instance per coroutine, storing the
pointer in a ptr_map and deleting it from the ptr_map once the
fiber_specific_ptr for that coroutine is cleaned up -- just declare a stack
instance on the top-level stack frame, the simplest C++ lifespan management.
Derive CoroData from LLInstanceTracker to detect potential name collisions and
to enumerate instances.

Continue registering each coroutine's CoroData instance in our
fiber_specific_ptr, but use a no-op deleter function.

Make ~LLCoros() directly pump the fiber scheduler a few times, instead of
having a special "LLApp" listener.

By the time "LLApp" listeners are notified that the app is quitting, the
mainloop is no longer running. Even though those listeners do things like
close work queues and inject exceptions into pending promises, any coroutines
waiting on those resources must regain control before they can notice and shut
down properly. Add a final "LLApp" listener that resumes ready coroutines a
few more times.

Make sure every other "LLApp" listener is positioned before that new one.

unless we're going to reference it.

Add LLCoros::TempStatus instances around known suspension points so
printActiveCoroutines() can report what each suspended coroutine is waiting
for.

Similarly, sprinkle checkStop() calls at known suspension points.

Make LLApp::setStatus() post an event to a new LLEventPump "LLApp" with a
string corresponding to the status value being set, but only until
~LLEventPumps() -- since setStatus() also gets called very late in the
application's lifetime.

Make postAndSuspendSetup() (used by postAndSuspend(), suspendUntilEventOn(),
postAndSuspendWithTimeout(), suspendUntilEventOnWithTimeout()) add a listener
on the new "LLApp" LLEventPump that pushes the new LLCoros::Stopping exception
to the coroutine waiting on the LLCoros::Promise. Make it return the new
LLBoundListener along with the previous one.

Accordingly, make postAndSuspend() and postAndSuspendWithTimeout() store the
new LLBoundListener returned by postAndSuspendSetup() in a LLTempBoundListener
(as with the previous one) so it will automatically disconnect once the wait
is over.

Make each LLCoprocedurePool instance listen on "LLApp" with a listener that
closes the queue on which new work items are dispatched. Closing the queue
causes the waiting dispatch coroutine to terminate. Store the connection in an
LLTempBoundListener on the LLCoprocedurePool so it will disconnect
automatically on destruction.

Refactor the loop in coprocedureInvokerCoro() to instantiate TempStatus around
the suspending call.

Change a couple spammy LL_INFOS() calls to LL_DEBUGS(). Give all logging calls
in that module a "CoProcMgr" tag to make it straightforward to re-enable the
LL_DEBUGS() calls as desired.

Introduce LLCoros::Stop exception, with subclasses Stopping, Stopped and
Shutdown. Add LLCoros::checkStop(), intended to be called periodically by any
coroutine with nontrivial lifespan. It checks the LLApp status and, unless
isRunning(), throws one of these new exceptions.

Make LLCoros::toplevel() catch Stop specially and log forcible coroutine
termination.

Now that LLApp status matters even in a test program, introduce a trivial
LLTestApp subclass whose sole function is to make isRunning() true.
(LLApp::setStatus() is protected: only a subclass can call it.) Add LLTestApp
instances to lleventcoro_test.cpp and lllogin_test.cpp.

Make LLCoros::toplevel() accept parameters by value rather than by const
reference so we can continue using them even after context switches.

Make private LLCoros::get_CoroData() static. Given that we've observed some
coroutines living past LLCoros destruction, making the caller call
LLCoros::instance() is more dangerous than encapsulating it within a static
method -- since the encapsulated call can check LLCoros::wasDeleted() first
and do something reasonable instead. This also eliminates the need for both a
const and non-const overload.

Defend LLCoros::delete_CoroData() (cleanup function for fiber_specific_ptr for
CoroData, implicitly called after coroutine termination) against calls after
~LLCoros().

Add a status string to coroutine-local data, with LLCoro::setStatus(),
getStatus() and RAII class TempStatus.

Add an optional 'when' string argument to LLCoros::printActiveCoroutines().
Make ~LLCoros() print the coroutines still active at destruction.

which performs "by hand" the same sequence of calls found in stepFrame().

Why not simply call stepFrame()? Hysterical reasons?

~LLEventPumps() deletes every LLEventPump instance it created itself. However,
many classes themselves contain LLEventPump subclass instances. These are
registered with LLEventPumps without it managing their lifespan.

But LLEventPump::reset() frees the LLStandardSignal aka
boost::signals2::signal instance owned by the LLEventPump, perforce
disconnecting all current listeners and disabling the LLEventPump. Even though
the instance still exists, if someone subsequently calls post(), nothing will
happen -- which is better than control trying to reach a method of a deleted
object.

Also add corresponding LLEventTimeout::post_every(), post_at(), post_after()
methods.

Having a map from std::string to a factory function returning LLEventPump* is
a cool idea, especially since you could statically populate such a map with
string literals and little lambdas.

Unfortunately, static initialization of any data is a bad idea when control
can reach consuming code before that module's static data are constructed.

Since LLEventPumps is already an LLSingleton, it's simple enough to make its
map non-static and initialize it in the constructor.

But another recent static factory-function map was introduced in
llleaplistener.cpp to support the LLLeapListener::newpump() operation. That
involves no LLSingletons.

Introduce LLEventPumps::make(name, tweak, type) to instantiate an LLEventPump
subclass of the specified type with specified (name, tweak) parameters.
Instances returned by make() are owned by LLEventPumps, as with obtain().
Introduce LLEventPumps::BadType exception for when the type string isn't
recognized.

LLEventPumps::obtain() can then simply call make() when the specified instance
name doesn't already exist. The configuration data used internally by obtain()
becomes { string instance name, string subclass name }. Although this too is
currently initialized in the LLEventPumps constructor, migrating it to a
configuration file would now be far more straightforward than before.

LLLeapListener::newpump(), too, can call LLEventPumps::make() with the
caller-specified type string. This eliminates that static factory map.
newpump() must catch BadType and report the error back to its invoker.

Given that the LLEventPump subclass instances returned by make() are owned by
LLEventPumps rather than LLLeapListener, there is no further need for the
LLLeapListener::mEventPumps ptr_map, which was used only to manage lifetime.
Also remove LLLeapListener's "killpump" operation since LLEventPumps provides
no corresponding functionality.

Although this is legal, apparently there was a bug in the C++11 standard (to
which gcc 4.8 conforms) that was subsequently fixed in the standard (and thus
in gcc 4.9). Thanks Henri Beauchamp.