This is very useful when you want to parse IR even if
its invalid (e.g. bytecode). It's also useful if you don't
want to pay the cost of verification in certain situations.
Differential Revision: https://reviews.llvm.org/D134847
Currently the generated pass declarations have to be enabled per-pass
using multiple `GEN_PASS_DECL_{PASSNAME}` defines. This adds
`GEN_PASS_DECL`, which enables the declarations for all passes in the
group with a single macro. This is convenient for cases where a single
header is used for all passes in the group.
Reviewed By: mehdi_amini, mscuttari
Differential Revision: https://reviews.llvm.org/D134766
Memref subview operation has been initially designed to work on memrefs with
strided layouts only and has never supported anything else. Port it to use the
recently added StridedLayoutAttr instead of extracting the strided from
implicitly from affine maps.
Reviewed By: nicolasvasilache
Differential Revision: https://reviews.llvm.org/D133938
This patch adds better functions for parsing MultiAffineFunctions and
PWMAFunctions in Presburger unittests.
A PWMAFunction can now be parsed as:
```
PWMAFunction result = parsePWMAF({
{"(x, y) : (x >= 10, x <= 20, y >= 1)", "(x, y) -> (x + y)"},
{"(x, y) : (x >= 21)", "(x, y) -> (x + y)"},
{"(x, y) : (x <= 9)", "(x, y) -> (x - y)"},
{"(x, y) : (x >= 10, x <= 20, y <= 0)", "(x, y) -> (x - y)"},
});
```
which is much more readable than the old format since the output can be
described as an AffineMap, instead of coefficients.
This patch also adds support for parsing divisions in MultiAffineFunctions
and PWMAFunctions which was previously not possible.
Reviewed By: arjunp
Differential Revision: https://reviews.llvm.org/D133654
This patch adds better functions for parsing MultiAffineFunctions and
PWMAFunctions in Presburger unittests.
A PWMAFunction can now be parsed as:
```
PWMAFunction result = parsePWMAF({
{"(x, y) : (x >= 10, x <= 20, y >= 1)", "(x, y) -> (x + y)"},
{"(x, y) : (x >= 21)", "(x, y) -> (x + y)"},
{"(x, y) : (x <= 9)", "(x, y) -> (x - y)"},
{"(x, y) : (x >= 10, x <= 20, y <= 0)", "(x, y) -> (x - y)"},
});
```
which is much more readable than the old format since the output can be
described as an AffineMap, instead of coefficients.
This patch also adds support for parsing divisions in MultiAffineFunctions
and PWMAFunctions which was previously not possible.
Reviewed By: arjunp
Differential Revision: https://reviews.llvm.org/D133654
This patch adds hermite normal form computation to Matrix. Part of this algorithm
lived in LinearTransform, being used for compuing column echelon form. This
patch moves the implementation to Matrix::hermiteNormalForm and generalises it
to compute the hermite normal form.
Reviewed By: arjunp
Differential Revision: https://reviews.llvm.org/D133510
Only the main Presburger library under the Presburger directory has been switched to use arbitrary precision. Users have been changed to just cast returned values back to int64_t or to use newly added convenience functions that perform the same cast internally.
The performance impact of this has been tested by checking test runtimes after copy-pasting 100 copies of each function. Affine/simplify-structures.mlir goes from 0.76s to 0.80s after this patch. Its performance sees no regression compared to its original performance at commit 18a06d4f3a before a series of patches that I landed to offset the performance overhead of switching to arbitrary precision.
Affine/canonicalize.mlir and SCF/canonicalize.mlir show no noticable difference, staying at 2.02s and about 2.35s respectively.
Also, for Affine and SCF tests as a whole (no copy-pasting), the runtime remains about 0.09s on average before and after.
Reviewed By: bondhugula
Differential Revision: https://reviews.llvm.org/D129510
Boolean splats currently can't roundtrip via the "raw" DenseElementsAttr
API. This is because internally we treat true splats in some cases as "1"(one bit set)
and in other cases as "0xFF"(all bits set). This commit cleans up this handling to
consistently use 0xFF (all bits set) as the value for a splat of true.
Differential Revision: https://reviews.llvm.org/D133743
A next step towards supporting the new dimension level types and
properties. This changes properly records the properties in the
Merger, so that subsequent computations (lattice optimizations)
and code generation (during sparsification) can do the right thing.
https://github.com/llvm/llvm-project/issues/51658
Reviewed By: Peiming
Differential Revision: https://reviews.llvm.org/D133620
This patch refactors MAF to be defined over the universe in a given space
instead of being defined over a restricted domain.
The reasoning for this refactor is to store division representation for local
variables explicitly for the function outputs. This change is required for
unionLexMax/Min to support local variables which will be upstreamed after this
patch. Another reason for this refactor is to have a flattened form of
AffineMap as MultiAffineFunction.
Reviewed By: arjunp
Differential Revision: https://reviews.llvm.org/D131864
839b436c93
changes the behavior. Based on the discussion, we also want to support
"and" behavior. The revision changes it into two functions, bitEnumContainsAny
and bitEnumContainsAll.
Reviewed By: krzysz00, antiagainst
Differential Revision: https://reviews.llvm.org/D133507
Splat of bool is encoded as a byte with all-ones in it [1]. Without this
change, this piece of code:
auto xs = builder.getI32TensorAttr({42, 42, 42, 42});
auto xs2 = xs.mapValues(builder.getI1Type(), [](const llvm::APInt &x) {
return x.isZero() ? llvm::APInt::getZero(1) : llvm::APInt::getAllOnes(1);
});
xs2.dump();
Prints:
dense<[true, false, false, false]> : tensor<4xi1>
Because only the first bit is set. This applies to both
DenseIntElementsAttr::mapValues() and DenseFPElementsAttr::mapValues().
[1]: e877b42e2c/mlir/lib/IR/BuiltinAttributes.cpp (L984)
Reviewed By: rriddle
Differential Revision: https://reviews.llvm.org/D132767
The patch introduces the required changes to update the pass declarations and definitions to use the new autogenerated files and allow dropping the old infrastructure.
Reviewed By: mehdi_amini, rriddle
Differential Review: https://reviews.llvm.org/D132838
Being the generated code macro-guarded, the autogenerated `.cpp.inc` file has been merged into the `.h.inc` to reduce the build steps.
Reviewed By: mehdi_amini, rriddle
Differential Revision: https://reviews.llvm.org/D132884
The patch addresses the linkage of the new autogenerated pass constructors, which, being declared as friend functions, resulted in having an inline nature and thus their implementations not being exported.
Reviewd By: mehdi_amini, rriddle
Differential Revision: https://reviews.llvm.org/D132572
Note:
when operating on a Type hierarchy with LeafType inheriting from MiddleType which inherits from mlir::Type.
calling LeafType::classof(MiddleType) will always return false.
because classof call the static getTypeID from its parent instead of the dynamic Type::getTypeID
so classof in this context will check if the TypeID of LeafType is the same as the TypeID of MiddleType which is always false.
It is bypassed in this commit inside CastInfo<To, From>::isPossible by calling classof with an mlir::Type.
but other unsuspecting users of LeafType::classof(MiddleType) would still get an incorrect result.
The pass tablegen backend has been reworked to remove the monolithic nature of the autogenerated declarations.
The pass public header can be generated with the -gen-pass-decls option. It contains options structs and registrations: the inclusion of options structs can be controlled individually for each pass by defining the GEN_PASS_DECL_PASSNAME macro; the declaration of the registrations have been kept together and can still be included by defining the GEN_PASS_REGISTRATION macro.
The private code used for the pass implementation (i.e. the pass base class and the constructors definitions, if missing from tablegen) can be generated with the -gen-pass-defs option. Similarly to the declarations file, the definitions of each pass can be enabled by defining the GEN_PASS_DEF_PASNAME variable.
While doing so, the pass base class has been enriched to also accept a the aformentioned struct of options and copy them to the actual pass options, thus allowing each pass to also be configurable within C++ and not only through command line.
Reviewed By: rriddle, mehdi_amini, Mogball, jpienaar
Differential Revision: https://reviews.llvm.org/D131839
Default attributes were only handled by ODS accessors generated with the
intention that these behave as if set attributes. This addresses the
long standing TODO to address this inconsistency. Moving the
initialization to construction vs every access. Removing need for
duplicated default attribute population in python bindings.
Switch some of the OpenMP ones to optional attribute with default as the
currently set default values are not legal. May need to dig more there.
Switched LinAlg generated ones to optional attribute with default as its
quite widely used and unclear where it falls on two different
interpretations.
Differential Revision: https://reviews.llvm.org/D130916
The `IR/AttributeTest.cpp` test fails to compile on Solaris:
```
/vol/llvm/src/llvm-project/local/mlir/unittests/IR/AttributeTest.cpp:223:36: error: no matching function for call to 'allocate'
AttrT::get(type, "resource", UnmanagedAsmResourceBlob::allocate(data));
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/vol/llvm/src/llvm-project/local/mlir/unittests/IR/AttributeTest.cpp:237:3: note: in instantiation of function template specialization 'checkNativeAccess<mlir::detail::DenseResourceElementsAttrBase<int8_t>, char>' requested here
checkNativeAccess<AttrT, T>(builder.getContext(), llvm::makeArrayRef(data),
^
/vol/llvm/src/llvm-project/local/mlir/unittests/IR/AttributeTest.cpp:258:3: note: in instantiation of function template specialization 'checkNativeIntAccess<mlir::detail::DenseResourceElementsAttrBase<int8_t>, char>' requested here
checkNativeIntAccess<DenseI8ResourceElementsAttr, int8_t>(builder, 8);
^
/vol/llvm/src/llvm-project/local/mlir/include/mlir/IR/AsmState.h:221:3: note: candidate template ignored: requirement '!std::is_same<char, char>::value' was not satisfied [with T = char]
allocate(ArrayRef<T> data, bool dataIsMutable = false) {
^
/vol/llvm/src/llvm-project/local/mlir/include/mlir/IR/AsmState.h:214:26: note: candidate function not viable: requires at least 2 arguments, but 1 was provided
static AsmResourceBlob allocate(ArrayRef<char> data, size_t align,
^
```
Because `char` is `signed` by default on Solaris and `int8_t` is
`char`. `std::is_same<int8_t, char>` is `true` unlike elsewhere, rejecting
the one-arg `allocate` overload.
Fixed by renaming the two overloads to avoid the ambiguity.
Tested on `amd64-pc-solaris2.11` ,`sparcv9-sun-solaris2.11`, and
`x86_64-pc-linux-gnu`.
Differential Revision: https://reviews.llvm.org/D131148
The methods to perform such operations have been implemented for the DynamicMemRefType in a way that is similar to the implementation for StridedMemRefType. Up until here one could pass an unranked memref to the library, and thus obtain a “dynamic” memref descriptor, but then there would have been no possibility to operate on its content.
A number of mlir tests `FAIL` on Solaris/sparcv9 with `Target has no JIT
support`. This patch fixes that by mimicing `clang/test/lit.cfg.py` which
implements a `host-supports-jit` keyword for this. The gtest-based unit
tests don't support `REQUIRES:`, so lack of support needs to be hardcoded
there.
Tested on `amd64-pc-solaris2.11` (`check-mlir` results unchanged) and
`sparcv9-sun-solaris2.11` (only one unrelated failure left).
Differential Revision: https://reviews.llvm.org/D131151
Implement the new sparse_tensor.reduce operation which
accepts a starting identity value and a code block
describing how to perform the reduction.
Reviewed by: aartbik
Differential Revision: https://reviews.llvm.org/D130573
This reverts commit b8ecf32f81.
This commit introduces undefined behavior according to UBSan:
UndefinedBehaviorSanitizer: nullptr-with-nonzero-offset third_party/llvm/llvm-project/mlir/include/mlir/ExecutionEngine/CRunnerUtils.h:377:5
The methods to perform such operations have been implemented for the DynamicMemRefType in a way that is similar to the implementation for StridedMemRefType. Up until here one could pass an unranked memref to the library, and thus obtain a “dynamic” memref descriptor, but then there would have been no possibility to operate on its content.
Differential Revision: https://reviews.llvm.org/D131359
The end-to-end test for this new feature also exposed a bug
in LLVM IR lowering (since then, fixed), where we need to account
for the min-poison bit as extra argument.
declare i32 @llvm.abs.i32(i32 <src>, i1 <is_int_min_poison>)
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D131712
They appeared at the wrong place in the switch, treating
them as unary op rather than binary op.
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D131509
Previously we are using IntegerAttr to back all SPIR-V enum
attributes. Therefore we all such attributes are showed like
IntegerAttr in IRs, which is barely readable and breaks
roundtripability of the IR. This commit changes to use
`EnumAttr` as the base directly so that we can have separate
attribute definitions and better IR printing.
Reviewed By: kuhar
Differential Revision: https://reviews.llvm.org/D131311
IntegerPolyhedron::findIntegerLexmin currently does not return values of
the local ids, so when a test for sampling includes a set with locals, the
result of findIntegerLexmin should be checked using containsPointNoLocal,
not containsPoint.
This reverts commit 07aaa35f74.
This breaks the Windows bot, and while the fix addressed the
`char`/`int8_t` case, it does not make sense for other cases like
`float`.
The `IR/AttributeTest.cpp` test fails to compile on Solaris:
/vol/llvm/src/llvm-project/local/mlir/unittests/IR/AttributeTest.cpp:223:36: error: no matching function for call to 'allocate'
AttrT::get(type, "resource", UnmanagedAsmResourceBlob::allocate(data));
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/vol/llvm/src/llvm-project/local/mlir/unittests/IR/AttributeTest.cpp:237:3: note: in instantiation of function template specialization 'checkNativeAccess<mlir::detail::DenseResourceElementsAttrBase<int8_t>, char>' requested here
checkNativeAccess<AttrT, T>(builder.getContext(), llvm::makeArrayRef(data),
^
/vol/llvm/src/llvm-project/local/mlir/unittests/IR/AttributeTest.cpp:258:3: note: in instantiation of function template specialization 'checkNativeIntAccess<mlir::detail::DenseResourceElementsAttrBase<int8_t>, char>' requested here
checkNativeIntAccess<DenseI8ResourceElementsAttr, int8_t>(builder, 8);
^
/vol/llvm/src/llvm-project/local/mlir/include/mlir/IR/AsmState.h:221:3: note: candidate template ignored: requirement '!std::is_same<char, char>::value' was not satisfied [with T = char]
allocate(ArrayRef<T> data, bool dataIsMutable = false) {
^
/vol/llvm/src/llvm-project/local/mlir/include/mlir/IR/AsmState.h:214:26: note: candidate function not viable: requires at least 2 arguments, but 1 was provided
static AsmResourceBlob allocate(ArrayRef<char> data, size_t align,
^
I suspect this happens because `char` is `signed` by default on Solaris.
Tested on `amd64-pc-solaris2.11` and `sparcv9-sun-solaris2.11`.
Differential Revision: https://reviews.llvm.org/D131148
Using if (TARGET ${LLVM_NATIVE_ARCH}) only works if MLIR is built
together with LLVM, but not for standalone builds of MLIR. The
correct way to check this is
if (${LLVM_NATIVE_ARCH} IN_LIST LLVM_TARGETS_TO_BUILD), as the
LLVM build system exports LLVM_TARGETS_TO_BUILD.
To avoid repeating the same check many times, add a
MLIR_ENABLE_EXECUTION_ENGINE variable.
Differential Revision: https://reviews.llvm.org/D131071
Jakub Kuderski