The number of operations dedicated to CUF grew and where all still in
FIR. In order to have a better organization, the CUF operations,
attributes and code is moved into their specific dialect and files. CUF
dialect is tightly coupled with HLFIR/FIR and their types.
The CUF attributes are bundled into their own library since some
HLFIR/FIR operations depend on them and the CUF dialect depends on the
FIR types. Without having the attributes into a separate library there
would be a dependency cycle.
This was using a StringRef, which is very unsafe because the method name
might just get disposed due to the async nature of the response.
This was causing weird characters being printed in the output logs.
Previously, these rewrites would drop scalable dimensions and treated
`[1]` (scalable one dim) as a unit dimension. This patch propagates
scalable dimensions and ensures `[1]` is not treated as a unit
dimension.
This commit renames the name of the block sorting utility function to
`getBlocksSortedByDominance`. A topological order is not defined on a
general directed graph, so the previous name did not make sense.
This adds support for `dense_resource` in arith to spirv.
Note that this inlines the blob into the IR. Another possibility would
be to add proper dense_resource support to spirv, but there is a lot of
special handling going on to convert a `DenseElementsAttr` to the
correct SPIRV type. Some of that even iterates over all the values in
the Attribute. For proper support of a `DenseResourceElementsAttr` this
probably needs a redesign. I would like to hear some opinions on that!
The test is disabled on non little Endian machines. See
https://github.com/llvm/llvm-project/issues/63469 for more information.
Summary:
No other project has these in the CMake itself, and they're wildly
inconsistent even within the project. These don't really add anything so
I think they should be removed.
[Recommit of e88ba6d975]
According to the specification in
https://github.com/ARM-software/acle/pull/309 this adds the intrinsics
void_svadd_za16_vg1x2_f16(uint32_t slice, svfloat16x2_t zn)
__arm_streaming __arm_inout("za");
void_svadd_za16_vg1x4_f16(uint32_t slice, svfloat16x4_t zn)
__arm_streaming __arm_inout("za");
void_svsub_za16_vg1x2_f16(uint32_t slice, svfloat16x2_t zn)
__arm_streaming __arm_inout("za");
void_svsub_za16_vg1x4_f16(uint32_t slice, svfloat16x4_t zn)
__arm_streaming __arm_inout("za");
as well as the corresponding `bf16` variants.
Composite OpenMP constructs where DISTRIBUTE is the first leaf
construct, as well as standalone DISTRIBUTE constructs, are allowed
inside of TEAMS regions.
Before this patch, nesting a DISTRIBUTE construct inside of a combined
TARGET TEAMS construct was disallowed, which it shouldn't be. Now both
TEAMS and TARGET TEAMS constructs can be immediate parents of DISTRIBUTE
constructs.
Scalable types are only available when:
* The function is compiled with +sve
* The function is compiled with +sme and the function is executed in
Streaming-SVE mode.
Instead add a proper attribute in clang, and add convert it to function
metadata to keep the information in the IR. The goal is to remove the
dependency on __attribute__((assume)) that should have not be there in
the first place.
Ref https://github.com/llvm/llvm-project/pull/84934
This functionality was added about three years ago, but has been in a
significantly broken state since it was added. It has begun to cause a
maintenance burden for work in Clang (largely due to the complexity of
having two levels of code generation involved), and the original author
is unable to help maintain it. Because it only worked under limited
circumstances and because of the maintenance burden, it is being
removed. If someone wishes to resurrect the functionality, they should
hopefully be able to do so from this one commit.
Fixes#82591
For all of the following reductions:
vector.reduce.or
vector.reduce.and
vector.reduce.xor
vector.reduce.add
vector.reduce.mul
vector.reduce.umin
vector.reduce.umax
vector.reduce.smin
vector.reduce.smax
vector.reduce.fmin
vector.reduce.fmax
if the input operand is the result of a vector.reverse then we can
perform a reduction on the vector.reverse input instead since the answer
is the same. If the reassociation is permitted we can also do the same
folds for these:
vector.reduce.fadd
vector.reduce.fmul
This commit turns the 1:N dialect conversion pattern for function
signatures into a pattern for `FunctionOpInterface`. This is similar to
the interface-based pattern that is provided with the 1:1 dialect
conversion (`populateFunctionOpInterfaceTypeConversionPattern`). No
change in functionality apart from supporting all `FunctionOpInterface`
ops and not just `func::FuncOp`.
The class `Lattice` should automatically delegate invocations of the
meet operator to the meet operation of the associated lattice value
class if that class provides a static function called `meet`. This
process fails for two reasons:
1. `Lattice::has_meet` checks for a member function `meet` without
arguments of the lattice value class, although it should check for a
static member function.
2. The function template `Lattice::meet<VT>()` implementing the default
meet operation directly in the lattice is always present and takes
precedence over the delegating function template `Lattice::meet<VT,
std::integral_constant<bool, true>>()`.
This change fixes the automatic delegation of the meet operation of a
lattice to the lattice value class in the presence of a static `meet`
function by conditionally enabling either the delegating function
template or the non-delegating function template and by changing
`Lattice::has_meet` so that it checks for a static `meet` member
function in the lattice value type.
The test from `TestSparseBackwardDataFlowAnalysis.cpp` is changed, such
that the `meet` function is not provided directly in the `WrittenTo`
lattice, but by the `Lattice` base class in order to trigger delegation
to a lattice value class.
Instead of widening e.g. i8 cttz(x) to i16 cttz(x | 0x100), use the more
optimizable form cttz_zero_undef(x | 0x100) since the widened operand is
definitely not zero.
This commit deletes the "simple" constructor of `CallDescription` which
did not require a `CallDescription::Mode` argument and always used the
"wildcard" mode `CDM::Unspecified`.
A few months ago, this vague matching mode was used by many checkers,
which caused bugs like https://github.com/llvm/llvm-project/issues/81597
and https://github.com/llvm/llvm-project/issues/88181. Since then, my
commits improved the available matching modes and ensured that all
checkers explicitly specify the right matching mode.
After those commits, the only remaining references to the "simple"
constructor were some unit tests; this commit updates them to use an
explicitly specified matching mode (often `CDM::SimpleFunc`).
The mode `CDM::Unspecified` was not deleted in this commit because it's
still a reasonable choice in `GenericTaintChecker` and a few unit tests.
It did not make sense that this said "all tile operations will go
through memory". Only the operations where the warning is emitted will
go through memory. The message has been updated to reflect that.
See the test file. At head, this crashes with
```
assertion failed at llvm/lib/Support/APInt.cpp:492 in
uint64_t llvm::APInt::extractBitsAsZExtValue(unsigned int, unsigned int) const:
bitPosition < BitWidth && (numBits + bitPosition) <= BitWidth &&
"Illegal bit extraction"
```
After internal intrinsic 'spv_switch' is processed we need to delete
G_BLOCK_ADDR instructions that were generated to keep track of the
corresponding basic blocks. If we just delete G_BLOCK_ADDR instructions
with BlockAddress operands, this leaves their BasicBlock counterparts in
a "address taken" status. This would make AsmPrinter to generate a
series of unneeded labels of a `"Address of block that was removed by
CodeGen"` kind. This PR is to ensure that we don't have a dangling
BlockAddress constants by zapping the BlockAddress nodes, and only after
that proceed with erasing G_BLOCK_ADDR instructions.
See also https://github.com/llvm/llvm-project/pull/87823 for more
details.
This PR generation of argument types of internal intrinsic functions
`spv_const_composite` and `spv_track_constant`, so that composite
constants of ConstantVector type preserve their correct type in
transformation passes and can be successfully used further by LLVM
intrinsic functions.
The added test case serves two purposes: it is to check the above
mentioned fix and to demonstrate that a call to __builtin_alloca() maps
to instructions from SPV_INTEL_variable_length_array when this extension
is available.
Previously we only generated declaration fragments for template type
parameters/arguments, this adds supports for most other possible
template parameters/arguments.
rdar://127732598
ResolveCommand will not succeed for an alias command with arguments, and
the code wasn't providing any. Replace that with explicit query(ies) for
the existence of a command with the given name.
