This continues to prevent writing output buffers (out from a function),
but fixes the problem where the copy-in/out was not getting done.

Making everything work will require knowing both in/out-ness and bufferness,
but these are currently mutually exclusive, because both are storage
qualifiers.

Update spirv-tools known-good

HLSL: minor: add warning for mat() matrix size truncation

This fixes spirv-tools issue 989 where if-break can be incorrectly
deleted from a loop.

Minor change to add a compilation warning on implicit matrix size
truncations.

(This is motivated by wanting to retickle the tests.)

#1165 invalid store

HLSL: allow keyword-identifiers as cbuffer/struct names.

HLSL: add implicit mat*mat truncations

Update spirv-tools known-good

Issue #791 was partially fixed by PR #1161 (the mat mul implicit
truncations were its main point), but it still wouldn't compile due to
the use of ConstantBuffer as an identifier.  Apparently those fall into
the same class as "float float", where float is both a type and an
identifier.

This allows struct definitions with such keyword-identifiers,
and adds ConstantBuffer to the set.  'cbuffer int' is legal in HLSL,
and 'struct int' appears to only be rejected due to the redefinition
of the 'int' type.

Fixes #791

This includes an enhancement to propagate through nested structs.

Goes with PR #1161, and completes the space for mul() implicit truncations.

Note that the v*v and scalar cases are already handled by existing code.

WIP: HLSL: matrix and vector truncations for m*v, v*m, m*m

HLSL truncates the vector, or one of the two matrix dimensions if there is a
dimensional mismatch in m*v, v*m, or m*m.

This PR adds that ability.  Conversion constructors are added as required.

Completes a TODO from previous commit.

Fixes #916 (the last change covered by the PR).
3rd list item in #976.

Fixes #1158.
Fixes #1159.

Addresses 2nd item in #976.

HLSL: Accept unorm and snorm on types

HLSL: ignore geometry attributes on non-GS stages.

If a shader includes a mixture of several stages, such as HS and GS,
the non-stage output geometry should be ignored, lest it conflict
with the stage output.

Memory fixes, round 1

Appveyor: avoid double testing on master commits

This is currently parsed and ignored, save for some minor validation.

Update README about the automatic master-tot relase

WIP: HLSL: implement TextureBuffer<type>

Make key objects using the memory pool own their own pool and delete it,
such that there is not generic per-thread pool to manage.