For the latest version of the glfw library, you need to find a packaged
named glfw3 (not GLFW).

The GNU compiler is being strict about POSIX compliance. Explicitly
port those things outside of POSIX (like snprintf).

Most of the warnings are about using "unsafe" functions. In addition to
replacing string functions like strcpy with counting strncpy versions,
which makes sense, older MSVC versions for some reason don't support
POSIX versions of these functions. Instead, they have their weird
offshoots, which is a pain.

Compilation error on Visual Studio 10 Win 64.
Fix undefined symbol: TEST_NOT_PASSED

When you render a dense image, the load balancing should be pretty
even. At any rate, it probably won't help to shift around pixels when
there are the same amount of pixels everywhere, so ultimately this
is a waste of time.

IceT insists on using sparse images most of the time because they are
almost always better (at least at the beginning part of the
compositing). Add an option to SimpleTiming to draw images that have no
background pixels so that the "sparse" images are equivalent to dense
images. This is done by making the depth of background pixels slightly
above and the opacity never 0.

The camera angles in SimpleTiming are more-or-less perfect for IceT's
image compositing. It is a fairly common case to want to zoom into the
data, especially if there is some intersting detail. This could add
extra work and load imbalance and could make the tests more interesting.

Hopefully this can be used to test odd process group sizes. Randomly
picked group sizes should commonly have large factors.

This is added for comparison purposes only. The telescoping binary swap
(just named binary swap) is generally faster (but maybe not always).

There is now a radix-kr algorithm that should handle odd factors better
than telescoping (plus supports non-power-of-2 factors).

I had meant to add it earlier, but I did it wrong.

The radix-kr is essentially the radix-k algorithm with some ideas from
2-3 swap added in. Rather than forcing all the k's in radix-k to be
factors, radix-kr allows the k to have a remainder when splitting the
process. The remaining processes participate in that round by splitting
their image and sending them off, but they receive nothing that round
and then go idle. Although this adds some imbalance, it does not add
much. It also makes keeping track of partitions easier than 2-3 swap.

Instead of detecting the platform and guessing the right file delete
function, use the cmake -E command to do the delete in a cross-platform
way.

Previously, if OpenGL was enabled all tests would create a render window
even if it was never used because they were all crammed into the same
executable that initalized with the OpenGL context. This is an artifact
from early versions of IceT that always required OpenGL.

Creating an OpenGL context shouldn't change the results of the test, but
it was annoying to have a bunch of windows created for every test. This
change separates out the OpenGl tests from those that don't use it so
that most of the tests now just run in the background.

Some of the files in the tests directory used a dash to separate words
(e.g. test-config.h and test-util.h) whereas others used an underscore
(e.g. test_codes.h and mpi_comm.h). Renamed the files such that they all
use underscores, which matches better the filenames used elsewhere.

At the end of each test each process reported whether it got an IceT or
OpenGL error including if it is no error. That can be helpful in part to
know that the process successfully completed, but for tests on very
large processes it generated excessively large log files. Instead
suppress the report if there is no error.

The usage of the -max-image-split-study of SimpleTiming states that it
tries doubling the max split. For some reason I implemented it by
multipling by the magic k value each time. Change it so it actually just
doubles the max image split and does a more complete search.

Previously, the root process would print the log at the end of
SimpleTimingDoRender. The problem is that in the scaling studies the
world communicator is partitioned so that there are multiple root
processes running at once. I've been having trouble with log entries
stomping over each other.

This change holds all the timings in a buffer rather than print them.
After scaling studies are complete, all log entries are collected to the
world root and printed out there.

Turns out the way I was recursively dividing partitions I was getting
multiple partitions that were the same size. This change cuts out the
repeated partition sizes letting processes go idle during the testing.

The 2-3 factor scaling routine in SimpleTiming was not properly checking
that the group size was large enough to handle the number of tiles used.
This didn't cause any real problems but caused the test to report failure
even though nothing really went wrong. Fixed that problem.

The rendering in SimpleTiming is not realistic, so it does not make much
sense to measure it. Instead, by default prerender the image and time
only how long it takes IceT to composite the image.

In the RandomTransform test when rendering pre-rendered images, we were
previously always rendering to the entire window size. However, when
trying tiled displays, which break up the display, the actual global
display can be a little smaller than that due to roundoff in the
division. This change sets the viewport to be exactly the size of the
global viewport.

Modify the RandomTransform test to try a composite using a pre-rendered
image in addition to using rendering callbacks (through the OpenGL
layer). This should be better about catching bugs like the previously
fixed one.

Add to SimpleTiming another scaling study that introduces lots of
intermediate job sizes that have factors of 3. This studies around
algorithms that might be optimized for power of 2.

SimpleTiming has an option to try all processor combinations that are
powers of 2 up to the job size. Jobs of small size are run concurrently,
which should speed the overall job time but will probably result in some
communication contention.

Also fixed some glaring problems with compositing pre-rendered images.

Previously we were using GLUT to create windows and OpenGL contexts.
However, the base GLUT is no longer maintained and APPLE as deprecated
its use. There is another project, FreeGLUT, that is poised to replace
the original GLUT library, but it does not support native Mac windows
yet.

There is another library named GLFW that is also a cross-platform
library for creating OpenGL windows, and this one supports the native
Mac windows. For now we are supporting both (with the default being
GLUT, which is still more common).

GLU seems to be deprecated in OpenGL 3 and now the Apple XCode compile
is warning about using it. The only thing we were using it for was to
draw a sphere, so replace that with our own sphere drawing code (using
vertex arrays that I think are still going to be supported).

An alternative to rolling our own would be to use a canned version that
comes with GLUT. However, GLUT itself is also going through a
transitional period, so until we can be sure it is going to survive and
be available everywhere, this is a reasonable solution.

Even if IceT ignores the image created during the render callback, it
should still send appropriate rendering parameters. When rendering a
black image, I had neglected to set the projection matrix for the tile.

Thanks to Burlen Loring for finding this problem and sending the fix.

Using add_test() call with $<TARGET_FILE:..> overcomes the issues with test not
finding the executable correctly.

Setting C_TEST_PATH to include the configuration type only for generators
supporting multiple configurations.

C_TEST_PATH is now complete and correct. So all tests can simply use it without
worry about platforms or generators.