Integer division or modulo is undefined behaviour in SPIR-V when the
divisor is zero, or when the dividend is the most negative number
representable by the result type and the divisor is negative one.
This patch makes us avoid this undefined behaviour and instead ensures
we adhere to the WGSL spec in these cases: for divisions the
expression evaluates to the value of the dividend, and for modulos the
expression evaluates to zero.
Similarily to how we handle these cases for the MSL and HLSL backends,
prior to emitting each function we emit code for any helper functions
required by that function's expressions. In this case that is helper
functions for integer division and modulo. Then, when emitting the
actual function's body, if we encounter an expression which needs
wrapped we instead emit a function call to the helper.
Emit helper functions for MathFunction::Abs and UnaryOperator::Negate
with a signed integer scalar or vector operand. And for
BinaryOperator::Divide and BinaryOperator::Modulo with signed or
unsigned integer scalar or vector operands.
Abs and Negate are written to avoid signed integer overflow when the
operand equals INT_MIN. This is achieved by bitcasting the value to
unsigned, using the negation operator, then bitcasting the result back
to signed. As HLSL's bitcast functions asint() and asuint() only work
for 32-bit types, we only use this workaround in such cases.
Division and Modulo avoid undefined behaviour for INT_MIN / -1 and
divide-by-zero by using 1 for the divisor instead. Additionally we
avoid undefined behaviour when using the modulo operator on operands
of mixed signedness by using the equation from the WGSL spec, using
division, subtraction and multiplication, rather than HLSL's modulus
operator.
Though not explicitly specified one way or the other, we have been
informed by DirectX engineers that signed integer overflow may be
undefined behaviour in some cases. To avoid this, we therefore bitcast
signed operands to unsigned prior to performing addition, subtraction,
or multiplication, then bitcast the result back to signed. As signed
types are represented as two's complement, this gives the correct
result whilst avoid any potential undefined behaviour.
Unfortunately HLSL's bitcast functions asint() and asuint() only work
for the 32-bit int and uint types. We therefore only apply this
workaround for 32-bit signed arithmetic. Support for other bit widths
could be added in the future, but extra care must be taken when
converting from unsigned to signed to avoid undefined or
implemented-defined behaviour.
There is no literal suffix in HLSL for the integer type. We can,
however, wrap integer literals in a constructor style cast. This
avoids ambiguity when passing literals to overloaded functions, which
we'll make use of in the subsequent patch in this PR.
Adds infrastructure to the MSL backend for emitting helper functions,
based upon the existing HLSL backend equivalent. Emit helper functions
for MathFunction::Abs and UnaryOperator::Negate with a signed integer
scalar or vector operand. And for BinaryOperator::Divide and
BinaryOperator::Modulo with signed or unsigned integer scalar or
vector operands.
Abs and Negate are written to avoid signed integer overflow when the
operand equals INT_MIN. This is achieved by bitcasting the value to
unsigned and using the negation operator, then bitcasting the result
back to signed.
Division and Modulo avoid undefined bevaviour for INT_MIN / -1 and
divide-by-zero by using 1 for the divisor instead. Additionally we
avoid undefined behaviour when using the modulo operator when either
or both operands are negative by using the equation from the WGSL
spec, using division, subtraction and multiplication, rather than
MSL's modulus operator.
Lastly, as the usage of the wrapper function for unary integer
negation makes the negation_avoids_prefix_decrement() testcase less
interesting, we extend it to additionally test negating floats.
Attempting to convert a non-abstract type will always fail, which can
result in unrelated errors being misreported as type conversion
errors. For example, in #7035 we made it so that abstract types can be
used as return values. This changed the final testcase in the
invalid_functions() test to fail due to failing to convert a u32 to an
atomic<u32>, rather than with a NonConstructibleReturnType error.
This patch makes it so that we do not attempt to convert non-abstract
types in the first place. This avoids the AutoConversion error for
that testcase, meaning the NonConstructibleReturnType is once again
reported.
Various callsites of try_automatic_conversions() have been updated to
ensure they still return an InitializationTypeMismatch error if the
types are mismatched, even if try_automatic_conversions() succeeds (eg
when conversion was not attempted due to the type being concrete). To
reduce code duplication these callsites were all adapted to use the
existing type_and_init() helper function.
Lastly, a couple of tests that expected to result in a AutoConversion
error have been adjusted to ensure this still occurs, by ensuring the
type used is abstract.
When converting the underlying scalar type of an array during const
evaluation, we currently use the resulting base type's size as the array
stride for the resulting type. For certain types, this may not match the
required alignment and will therefore result in a validation error.
For example, `array<vec3<f32>, N>` should have a stride of 16. But if
declared with an abstract initializer, eg `array(vec3(0.0))` we will
incorrectly determine the stride to be 12.
To solve this, we use the proc::Layouter struct to determine the
required array stride during const evaluation. To avoid repeating
layouting work, we reuse the Lowerer's layouter, passing it through the
various *Contexts through to the ConstantEvaluator.
When lowering a return statement, call expression_for_abstract()
rather than expression() to avoid concretizing the return value. Then,
if the function has a return type, call try_automatic_conversions() to
attempt to convert our return value to the correct type.
This has the unfortunate side effect that some errors that would have
been caught by the validator are instead encountered as conversion
errors by the parser. This may result in a slightly less descriptive
error message in some cases. (See the change to the invalid_functions()
test, for example.)
Signed integer overflow is undefined behaviour in MSL. However, signed
integers are defined to be two's complement. This allows us to cast
signed values to their corresponding unsigned type, perform the
arithmetic on these unsigned values (which has defined overflow
behaviour) then cast the result back to signed.
Care must be taken when emitting the isign polyfill, which uses
metal::select(). We must ensure the -1, 0, and 1 literals used as
inputs to select() have the correct width, else bitcasting the output
of select() will fail due to mismatched widths.
* feat: Add 32-bit floating-point atomics
* Current supported platforms: Metal
* Platforms to support in the future: Vulkan
Related issues or PRs:
* gfx-rs/wgpu#1020
* Add changelog
* Edit changelog
* feat: Add 32-bit float atomics support for Vulkan (SPIR-V shaders)
* atomicSub for f32 in the previous commits is removed.
* Update test
* chore: doc type link
* refactor: Revise float atomics on msl and spv
* Make branches tidy
* Also revise old codes
* Ensure the implementations are supported by Metal and Vulkan backends
* refactor: Renaming flt32 atomics to float32 atomics
* chore: Add link to Vulkan feature
* fix: cargo fmt
* chore: hack comment
* Revert changelog
* Fix: Cargo advisory
* Update wgpu-hal/src/metal/adapter.rs
Co-authored-by: Teodor Tanasoaia <28601907+teoxoy@users.noreply.github.com>
* Update naga/src/lib.rs
Co-authored-by: Teodor Tanasoaia <28601907+teoxoy@users.noreply.github.com>
* Adjust feature flag position
---------
Co-authored-by: Teodor Tanasoaia <28601907+teoxoy@users.noreply.github.com>
The WGSL spec only allows named component expressions when the base
type is a vector or a structure, so this patch removes support for it
for matrices. Additionally tests which used this for matrices have
been updated to use indexing expressions instead, and a test has been
added to ensure a named component expression on a matrix results in an
error.
Apply automatic conversions to the right-hand sides of assignment and
compound assignment statements.
Fix `try_automatic_conversion_for_leaf_scalar` to handle arrays,
not just scalars, vectors, and matrices. Previously this was only used
for converting the right-hand sides of bit shift operators to `u32`,
and bit shift operators don't support arrays. But now we're using
`try_automatic_conversion_for_leaf_scalar` for ordinary assignments,
and you can assign arrays.
Update a test of propagating source code spans through compaction.
This test needs validation to report an error, but this commit's fix
causes the error in the test to be detected in the front end, which is
too early to exercise compaction. Use a new error that the front end
still won't notice.