Like we did before commit 067e6deee4aba447c6627ce02b58c6bdcd90470e.
These skips aren't all that interesting; it's entirely intentional that
e.g. a 2.0-3.0 run wouldn't run 4.0 shaders.
Adjust the algorithm for deciding for which profiles to test compilation.
We first ensure that if the compilation result changes (most often as the result
of a feature introduced in a specific version), we test the versions immediately
on either side of the change, to validate that vkd3d-shader is emulating the
same version behaviour.
We then ensure that we are testing at least one version from each set of sm1,
sm4, and sm6.
Mainly in order to not waste time compile-testing the same version
more than once [as we do with e.g. the d3d11 and d3d12 runners, or
d3d12, GL, and Vulkan].
When a shader fails to compile for a range of versions, we want to validate that
we are correctly implementing that behaviour. E.g. if a feature requires shader
model 5.0, we should validate that it compiles correctly with 5.0 (which we do),
but also that it *fails* to compile with 4.1 (which we do not).
The obvious and simple solution is to simply run compile tests for each version.
There are, however, at least 12 versions of HLSL up to and including 6.0, at
least 10 of which are known to introduce new features. Shader compilation takes
about 10-15% of the time that draw and dispatch does, both for native and
(currently) vkd3d. Testing every version for every shader would add a
noticeable amount of time to the tests.
In practice, the interesting versions to test for most shaders are:
* At least one from each range 1-3, 4-5, and 6. It's common enough for the
semantics of the HLSL to differ between bytecode formats, or for features to
be added or removed across those boundaries.
* If the shader requires a given feature, we want to test both sides of the cusp
to ensure we're requiring the same version for the feature.
In practice this is 3 or 4 versions, which is measurably less than the 12 we'd
otherwise be running.
In order to achieve this goal of testing only the 3 or 4 interesting versions
for a shader, we need to know what version is actually required for a feature.
This is encoded in the shader itself using e.g. [pixel shader fail(sm<5)].
This patch therefore implements the first step towards this goal, namely,
determining which versions succeed and fail, so we can figure out which ones are
interesting.
We could require the test writer to specify which versions are interesting ahead
of time (e.g. "for version in 2.0 4.1 5.0 6.0") but this is both redundant (and
there are a *lot* of tests that need some feature gate) and easy for a test
writer to get wrong by missing interesting versions.
Most Vulkan drivers are happy to compile a pipeline with interstage
interface mismatches, even if the pipeline will be broken. Metal,
and therefore MoltenVK, instead doesn't and errors out immediately.
This allows the Vulkan runner to execute in environments which do
not support geometry shaders, most notably MoltenVK. In particular,
that enables SM<4 shaders to run in such environments
Since recent versions seem to have more features and possibly
fewer bugs it is useful to know what you're dealing with.
Currently the latest WARP drivers are distributed here:
https://www.nuget.org/packages/Microsoft.Direct3D.WARP
The used UAV formats are explicitly added in the [require] section of
every test that uses them.
Some of these tests were failing on Intel UHD graphics 770 because of
missing support for additional UAV load types, explicitly requiring
these formats allows these tests to be skipped.
Currently we are skipping compilation tests ([* shader] blocks) on
missing capabilities. This is not really necessary and it is useful to
test shader compilation even on machines that can't run these tests.
So, this commit moves the checks to the [test] blocks, skipping them
entirely when caps are not met.
To make the effects of enum test_action values more explicit, these
are renamed to TEST_ACTION_SKIP_COMPILATION and
TEST_ACTION_SKIP_EXECUTION.
The current default is r32g32b32a32 but it requires special support
which is not available on all GPUs, so it is not a very convenient
default.
Instead of changing the default making it different from RTV resoures,
the format is required to always be explicit for UAVs.
The exceptions are counter_buffer and buffers with "stride", which don't
require a format because it is already implied.
