Details within these bear even more fleshing out, but would like to have
that driven by actual need.

SPV: Fix pointer address spaces for sampler function parameters

For opaque types such as samplers, images, and atomic counters, we want to
reference the actual object in the child function.  For a long time, we
used a shadow variable and made a copy of the image/sampler.  In 76d0ac1a,
this was changed to not shadow samplers.  However, this didn't cover all
opaque types and it also didn't get the pointer storage classes right.
This commit fixes both of these issues.

Fixes #324

SPV: Defer capability decl for ClipDistance, CullDistance, PointSize until…

The compiler will mark struct members with those builtins, but won't
declare the capability until that member is accessed by some executable
instruction.

Test changes:
- spv.430.vert: was missing ClipDistance capability.
- spv.precise.tese: remove TessellationPointSize capability.

Build: Fix Ubuntu 16.10 build

Infrastructure: Add .clang-format.

The grammar now accepts type casts, like "(int)x", but that
has to be disambiguated from "(a + b)", needed deeper lookahead
and backing up than what existed so far.

HLSL: Add negative intrinsics tests

HLSL: Fix for empty statement segfault

remapper: do not eliminate interface variables in dce

HLSL: Initial implementation of direct-mapped subset of HLSL intrinsics

this is done by counting op_entrypoint as a use/def

This checkin implements about half of the HLSL intrinsics for a subset of their
entire type support (but a useful subset).  The uncommented lines in
TBuiltInParseablesHlsl::identifyBuiltIns shows which are connected.

Build: For MinGW, statically link exes against basic runtimes

GLSL takes a traditional sampler, but SPIR-V wants just the
sampled image, not the combined sampler and image.

Avoids the need to ship basic runtimes like libgcc_s_sjlj-1.dll
with the executables.

Build: Produce different names for files in debug builds on MSVC

SPV: Fix double matrix creation

This amends the previous commit, which ommitted barriers in version 420 for compute shader.

Core compute shaders were working, but the extension wasn't implemented.

Also beefed up support for running compute shaders is #version 420, but this
work is only partially done.