Summary:
This revision adds a tensor_reshape operation that operates on tensors.
In the tensor world the constraints are less stringent and we can allow more
arbitrary dynamic reshapes, as long as they are contractions.
The expansion of a dynamic dimension into multiple dynamic dimensions is under-specified and is punted on for now.
Differential Revision: https://reviews.llvm.org/D77360
This will fix the case:
$ toyc -emit=jit test.toy
$ cat test.toy
def main() {
var a = 1;
print(a);
}
Without this patch it would trigger an assertion.
Differential Revision: https://reviews.llvm.org/D77464
The current return type sometimes leads to code like
to_vector<2>(ValueRange(loop.getInductionIvs())). It would be nice to
shorten it. Users who need access to Block::BlockArgListType (if there
are any), can always call getBody()->getArguments(); if needed.
Also remove getNumInductionVars(), since there is getNumLoops().
Differential Revision: https://reviews.llvm.org/D77526
Summary:
This revision performs several cleanups on the translation infra:
* Removes the TranslateCLParser library and consolidates into Translation
- This was a weird library that existed in Support, and didn't really justify being a standalone library.
* Cleans up the internal registration and consolidates all of the translation functions within one registry.
Differential Revision: https://reviews.llvm.org/D77514
Summary: Blocks are numbered locally within a region, so numbering above the parent region is unnecessary.
Differential Revision: https://reviews.llvm.org/D77510
Summary: This updates the canonicalization documentation, and properly documents the different ways of canonicalizing operations.
Differential Revision: https://reviews.llvm.org/D77490
Summary:
This revision adds a section to WritingAPass to document the declarative specification, and how to use it.
Differential Revision: https://reviews.llvm.org/D77102
Even if this indicates in general a problem at call sites, the printer
is used for debugging and avoiding crashing is friendlier for example
when used in diagnostics or other printer.
Differential Revision: https://reviews.llvm.org/D77481
Add a pattern rewriter utility to erase blocks (while notifying the
pattern rewriting driver of the erased ops). Use this to remove trivial
else blocks in affine.if ops.
Differential Revision: https://reviews.llvm.org/D77083
Removing dead ops should make the outer loop of the pattern rewriting
driver run again. Although its operands are added to the worklist, if no
changes happenned to them or remaining ops in the worklist, the driver
wouldn't run once again - but it should be.
Differential Revision: https://reviews.llvm.org/D77483
The ForOp::build ensures that there is a block terminator which is great for
the default use case when there are no iter_args and loop.for returns no
results. In non-zero results case we always need to call replaceOpWithNewOp
which is not the nicest thing in the world. We can stop inserting YieldOp when
iter_args is non-empty. IfOp::build already behaves similarly.
Summary: This revision adds support for marking the last region as variadic in the ODS region list with the VariadicRegion directive.
Differential Revision: https://reviews.llvm.org/D77455
This adds a minimal out-of-tree dialect template which can be used to start work on a standalone dialect implementation without having to integrate it in the main LLVM tree.
It mostly sets up the directory structure and provides CMakeLists.txt files to build a dialect library, an opt-like tool to operate on that dialect as well as tests. It could be expanded in the future to add examples of more user-defined operations, types, attributes, generated enums, transforms, etc. and linked to a tutorial.
Differential Revision: https://reviews.llvm.org/D77133
The implementation of shape inference in the toy tutorial did not conform to the correct algorithmic description.
The result was only correct because all operations appear to be processed in sequence.
Differential Revision: https://reviews.llvm.org/D77382
Summary: The attribute grammar includes an optional trailing colon type, so for attributes without a constant buildable type this will generally lead to unexpected and undesired behavior. Given that, it's better to just error out on these cases.
Differential Revision: https://reviews.llvm.org/D77293
Summary: It is a very common user trap to think that the location printed along with the diagnostic is the same as the current operation that caused the error. This revision changes the behavior to always print the current operation, except for when diagnostics are being verified. This is achieved by moving the command line flags in IR/ to be options on the MLIRContext.
Differential Revision: https://reviews.llvm.org/D77095
Summary:
A recent extension allowed the `loop.if` operation to return results yielded by
its regions. However, such operations could not be lowered to a CFG of standard
operations because it would have required to modify the argument list of a
block, which is not allowed in a conversion pattern. Now that the conversion
infrastructure supports block creation, use it to create a block with an
argument list that dominates the operations following the `loop.if` and forward
the results as arguments of this block.
Depends On D77416
Differential Revision: https://reviews.llvm.org/D77418
Summary:
Linalg makes it possible to interface codegen with externally precompiled HPC libraries. The mechanism to allow such interop uses a normalized ABI and the emission of C interface wrappers.
The mechanism controlling these C interface emission is too aggressive and makes it very easy to obtained undefined symbols for external function (e.g. the ones coming from libm).
This revision uses the newly introduced llvm.emit_c_interface function attribute which allows controlling this behavior at a function granularity. As a consequence LinalgToLLVM does not need to activate the C wrapper emission when adding the StdToLLVM patterns.
Differential Revision: https://reviews.llvm.org/D77364
PatternRewriter and derived classes provide a set of virtual methods to
manipulate blocks, which ConversionPatternRewriter overrides to keep track of
the manipulations and undo them in case the conversion fails. However, one can
currently create a block only by splitting another block into two. This not
only makes the API inconsistent (`splitBlock` is allowed in conversion
patterns, but `createBlock` is not), but it also make it impossible for one to
create blocks with argument lists different from those of already existing
blocks since in-place block updates are not supported either. Such
functionality precludes dialect conversion infrastructure from being used more
extensively on region-containing ops, for example, for value-returning "if"
operations. At the same time, ConversionPatternRewriter already allows one to
undo block creation as block creation is one of the primitive operations in
already supported region inlining.
Support block creation in conversion patterns by hooking `createBlock` on the
block action undo mechanism. This requires to make `Builder::createBlock`
virtual, similarly to Op insertion. This is a minimal change to the Builder
infrastructure that will later help support additional use cases such as block
signature changes. `createBlock` now additionally takes the types of the block
arguments that are added immediately so as to avoid in-place argument list
manipulation that would be illegal in conversion patterns.
Previously, the tablegen() cmake command, which defines custom
commands for running tablegen, included several hardcoded paths. This
becomes unwieldy as there are more users for which these paths are
insufficient. For most targets, cmake uses include_directories() and
the INCLUDE_DIRECTORIES directory property to specify include paths.
This change picks up the INCLUDE_DIRECTORIES property and adds it
to the include path used when running tablegen. As a side effect, this
allows us to remove several hard coded paths to tablegen that are redundant
with specified include_directories().
I haven't removed the hardcoded path to CMAKE_CURRENT_SOURCE_DIR, which
seems generically useful. There are several users in clang which apparently
don't have the current directory as an include_directories(). This could
be considered separately.
The new version of this path uses list APPEND rather than list TRANSFORM,
in order to be compatible with cmake 3.4.3. If we update to cmake 3.12 then
we can use list TRANSFORM instead.
Differential Revision: https://reviews.llvm.org/D77156
Previously, the tablegen() cmake command, which defines custom
commands for running tablegen, included several hardcoded paths. This
becomes unwieldy as there are more users for which these paths are
insufficient. For most targets, cmake uses include_directories() and
the INCLUDE_DIRECTORIES directory property to specify include paths.
This change picks up the INCLUDE_DIRECTORIES property and adds it
to the include path used when running tablegen. As a side effect, this
allows us to remove several hard coded paths to tablegen that are redundant
with specified include_directories().
I haven't removed the hardcoded path to CMAKE_CURRENT_SOURCE_DIR, which
seems generically useful. There are several users in clang which apparently
don't have the current directory as an include_directories(). This could
be considered separately.
Differential Revision: https://reviews.llvm.org/D77156
Two back-to-back transpose operations are combined into a single transpose, which uses a combination of their permutation vectors.
Differential Revision: https://reviews.llvm.org/D77331
A certain number of EDSCs have a named form (e.g. `linalg.matmul`) and a generic form (e.g. `linalg.generic` with matmul traits).
Despite living in different namespaces, using the same name is confusiong in clients.
Rename them as `linalg_matmul` and `linalg_generic_matmul` respectively.
This slightly tweaks the generated code from:
#ifdef GEN_PASS_REGISTRATION
::mlir::registerPass("flag1", ...
::mlir::registerPass("flag2", ...
#endif // GEN_PASS_REGISTRATION
to:
#ifdef GEN_PASS_REGISTRATION
#define GEN_PASS_REGISTRATION_Pass1
#define GEN_PASS_REGISTRATION_Pass2
#endif // GEN_PASS_REGISTRATION
#ifdef GEN_PASS_REGISTRATION_Pass1
::mlir::registerPass("flag1", ...
#endif
#ifdef GEN_PASS_REGISTRATION_Pass1
::mlir::registerPass("flag2", ...
#endif
That way the generated code can be included by defining the
`GEN_PASS_REGISTRATION` macro as currenty and register all the passes,
but one can also define only `GEN_PASS_REGISTRATION_Pass1` to register a
subset of the passes.
Differential Revision: https://reviews.llvm.org/D77322
C interface emission is controlled by a flag and has coarse granularity.
With this coarse control, interfaces are emitted for all external functions.
This makes is easy to get undefined symbols.
This revision adds support for controlling per-function emission with an "emit_c_interface" attribute.
Summary:
LLVM IR functions can have arbitrary attributes attached to them, some of which
affect may affect code transformations. Until we can model all attributes
consistently, provide a pass-through mechanism that forwards attributes from
the LLVMFuncOp in MLIR to LLVM IR functions during translation. This mechanism
relies on LLVM IR being able to recognize string representations of the
attributes and performs some additional checking to avoid hitting assertions
within LLVM code.
Differential Revision: https://reviews.llvm.org/D77072
Add a method that given an affine map returns another with just its unique
results. Use this to drop redundant bounds in max/min for affine.for. Update
affine.for's canonicalization pattern and createCanonicalizedForOp to use
this.
Differential Revision: https://reviews.llvm.org/D77237
There is no need to directly depends on this from mlir-opt, some library
may transitively depend on a subset of the targets when enabled (like
NVPTX for Cuda codegen tests) but this is handled by CMake already.
Modernize/cleanup code in loop transforms utils - a lot of this code was
written prior to the currently available IR support / code style. This
patch also does some variable renames including inst -> op, comment
updates, turns getCleanupLoopLowerBound into a local function.
Differential Revision: https://reviews.llvm.org/D77175
Summary:
This is to allow optimizations like loop invariant code motion to work
on the ParallelOp.
Additional small cleanup on the ForOp implementation of
LoopLikeInterface and the test file of loop-invariant-code-motion.
Differential Revision: https://reviews.llvm.org/D77128
Summary:
This revision adds support for auto-generating pass documentation, replacing the need to manually keep Passes.md up-to-date. This matches the behavior already in place for dialect and interface documentation.
Differential Revision: https://reviews.llvm.org/D76660
This revision adds support for generating utilities for passes such as options/statistics/etc. that can be inferred from the tablegen definition. This removes additional boilerplate from the pass, and also makes it easier to remove the reliance on the pass registry to provide certain things(e.g. the pass argument).
Differential Revision: https://reviews.llvm.org/D76659
This removes the need to statically register conversion passes, and also puts all of the conversions within one centralized file.
Differential Revision: https://reviews.llvm.org/D76658
This generates a Passes.td for all of the dialects that have transformation passes. This removes the need for global registration for all of the dialect passes.
Differential Revision: https://reviews.llvm.org/D76657
This will greatly simplify a number of things related to passes:
* Enables generation of pass registration
* Enables generation of boiler plate pass utilities
* Enables generation of pass documentation
This revision focuses on adding the basic structure and adds support for generating the registration for passes in the Transforms/ directory. Future revisions will add more support and move more passes over.
Differential Revision: https://reviews.llvm.org/D76656
Summary:
The commit provides a single method to build affine maps with zero or more
results. Users of mlir::AffineMap previously had to dispatch between two methods
depending on the number of results.
At the same time, this commit fixes the method for building affine map with zero
results that was previously ignoring its `symbolCount` argument.
Differential Revision: https://reviews.llvm.org/D77126
Summary:
OpBuilder(Block) is specifically replaced with
OpBuilder::atBlockEnd(Block);
This is to make insertion behavior clear due to there being no one
correct answer for which location in a block the default insertion
point should be.
Differential Revision: https://reviews.llvm.org/D77060
Summary:
The RAW fusion happens only if the produecer block dominates the consumer block.
The WAW pattern also works with the precondition. I.e., if a producer can
dominate the consumer, they can fairly fuse together.
Since they are all tilable, we can think the pattern like this way:
Input:
```
linalg_op1 view
tile_loop
subview_2
linalg_op2 subview_2
```
Tile the first Linalg op as same as the second Linalg.
```
tile_loop
subview_1
linalg_op1 subview_1
tile_loop
subview_2
liangl_op2 subview_2
```
Since the first Linalg op is tilable in the same way and the computation are
independently, it's fair to fuse it with the second Linalg op.
```
tile_loop
subview_1
linalg_op1 subview_1
linalg_op2 subview_2
```
In short, this patch includes:
- Handling both RAW and WAW pattern.
- Adding a interface method to get input and output buffers.
- Exposing a method to get a StringRef of a dependency type.
- Fixing existing WAW tests and add one more use case: initialize the buffer
before conv op.
Differential Revision: https://reviews.llvm.org/D76897
Summary:
Performs an N-D pooling operation similarly to the description in the TF
documentation:
https://www.tensorflow.org/api_docs/python/tf/nn/pool
Different from the description, this operation doesn't perform on batch and
channel. It only takes tensors of rank `N`.
```
output[x[0], ..., x[N-1]] =
REDUCE_{z[0], ..., z[N-1]}
input[
x[0] * strides[0] - pad_before[0] + dilation_rate[0]*z[0],
...
x[N-1]*strides[N-1] - pad_before[N-1] + dilation_rate[N-1]*z[N-1]
],
```
The required optional arguments are:
- strides: an i64 array specifying the stride (i.e. step) for window
loops.
- dilations: an i64 array specifying the filter upsampling/input
downsampling rate
- padding: an i64 array of pairs (low, high) specifying the number of
elements to pad along a dimension.
If strides or dilations attributes are missing then the default value is
one for each of the input dimensions. Similarly, padding values are zero
for both low and high in each of the dimensions, if not specified.
Differential Revision: https://reviews.llvm.org/D76414
This commit changes the separator line for dividing auto-generated
docs from spec and manually added appendix from "### Custom assembly
form" to "<!-- End of AutoGen section -->". This is in preparation
to use the declarative assembly form in MLIR core. We will replace
more and more manually written assembly forms to be autogenerated.
Differential Revision: https://reviews.llvm.org/D77158
Move lower-affine.mlir from test/Transforms to
test/Conversion/AffineToStandard/. Other related NFC.
Signed-off-by: Uday Bondhugula <uday@polymagelabs.com>
Differential Revision: https://reviews.llvm.org/D77008
Existing tiling implementation of Linalg would still work for tiling
the batch dimensions of the convolution op.
Differential Revision: https://reviews.llvm.org/D76637
Summary:
Add support for TupleGetOp folding through InsertSlicesOp and ExtractSlicesOp.
Vector-to-vector transformations for unrolling and lowering to hardware vectors
can generate chains of structured vector operations (InsertSlicesOp,
ExtractSlicesOp and ShapeCastOp) between the producer of a hardware vector
value and its consumer. Because InsertSlicesOp, ExtractSlicesOp and ShapeCastOp
are structured, we can track the location (tuple index and vector offsets) of
the consumer vector value through the chain of structured operations to the
producer, enabling a much more powerful producer-consumer fowarding of values
through structured ops and tuple, which in turn enables a more powerful
TupleGetOp folding transformation.
Reviewers: nicolasvasilache, aartbik
Reviewed By: aartbik
Subscribers: grosul1, mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, liufengdb, Joonsoo, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D76889
Summary: This object file has grown beyond the default limit, and elsewhere in LLVM, we seem to be setting this flag as a one-off, so continuing that here.
Reviewers: mravishankar, antiagainst
Subscribers: mgorny, mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, nicolasvasilache, arpith-jacob, mgester, lucyrfox, liufengdb, Joonsoo, bader, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D77002
Rewrite mlir::permuteLoops (affine loop permutation utility) to fix
incorrect approach. Avoiding using sinkLoops entirely - use single move
approach. Add test pass.
This fixes https://bugs.llvm.org/show_bug.cgi?id=45328
Depends on D77003.
Signed-off-by: Uday Bondhugula <uday@polymagelabs.com>
Differential Revision: https://reviews.llvm.org/D77004
Summary:
This revision performs a lot of different cleanups on operation documentation to ensure that they are consistent, e.g. using mlir code blocks, formatting, etc.
This revision also includes the auto-generated documentation into the hand-written documentation for the dialects that have a specific top-level dialect file. This updates the documentation for all dialects aside from SPIRV and STD. These dialects will be updated in a followup.
Differential Revision: https://reviews.llvm.org/D76734
Summary: This revision updates the dialect documentation to use the auto-generated markdown for operations. This allows for updating some out-of-date bits of documentation, and allows for displaying a large of number of newly added operations that did not have a counter part in Standard.md.
Differential Revision: https://reviews.llvm.org/D76743
Summary: This revision updates the SourceMgrDiagnosticHandler to not print the source location of a note if it is the same location as the previously printed diagnostic. This helps avoid redundancy, and potential confusion, when looking at the diagnostic output.
Differential Revision: https://reviews.llvm.org/D76787
Add missing assert checks for input to mlir::interchangeLoops utility.
Rename interchangeLoops -> permuteLoops; update doc comments to clarify
inputs / return val. Other than the assert checks, this is NFC.
Signed-off-by: Uday Bondhugula <uday@polymagelabs.com>
Differential Revision: https://reviews.llvm.org/D77003
Move test/lib/TestDialect to test/lib/Dialect/Test - makes the dir
structure more uniform.
Signed-off-by: Uday Bondhugula <uday@polymagelabs.com>
Differential Revision: https://reviews.llvm.org/D76677
This patch introduces a utility to separate full tiles from partial
tiles when tiling affine loop nests where trip counts are unknown or
where tile sizes don't divide trip counts. A conditional guard is
generated to separate out the full tile (with constant trip count loops)
into the then block of an 'affine.if' and the partial tile to the else
block. The separation allows the 'then' block (which has constant trip
count loops) to be optimized better subsequently: for eg. for
unroll-and-jam, register tiling, vectorization without leading to
cleanup code, or to offload to accelerators. Among techniques from the
literature, the if/else based separation leads to the most compact
cleanup code for multi-dimensional cases (because a single version is
used to model all partial tiles).
INPUT
affine.for %i0 = 0 to %M {
affine.for %i1 = 0 to %N {
"foo"() : () -> ()
}
}
OUTPUT AFTER TILING W/O SEPARATION
map0 = affine_map<(d0) -> (d0)>
map1 = affine_map<(d0)[s0] -> (d0 + 32, s0)>
affine.for %arg2 = 0 to %M step 32 {
affine.for %arg3 = 0 to %N step 32 {
affine.for %arg4 = #map0(%arg2) to min #map1(%arg2)[%M] {
affine.for %arg5 = #map0(%arg3) to min #map1(%arg3)[%N] {
"foo"() : () -> ()
}
}
}
}
OUTPUT AFTER TILING WITH SEPARATION
map0 = affine_map<(d0) -> (d0)>
map1 = affine_map<(d0) -> (d0 + 32)>
map2 = affine_map<(d0)[s0] -> (d0 + 32, s0)>
#set0 = affine_set<(d0, d1)[s0, s1] : (-d0 + s0 - 32 >= 0, -d1 + s1 - 32 >= 0)>
affine.for %arg2 = 0 to %M step 32 {
affine.for %arg3 = 0 to %N step 32 {
affine.if #set0(%arg2, %arg3)[%M, %N] {
// Full tile.
affine.for %arg4 = #map0(%arg2) to #map1(%arg2) {
affine.for %arg5 = #map0(%arg3) to #map1(%arg3) {
"foo"() : () -> ()
}
}
} else {
// Partial tile.
affine.for %arg4 = #map0(%arg2) to min #map2(%arg2)[%M] {
affine.for %arg5 = #map0(%arg3) to min #map2(%arg3)[%N] {
"foo"() : () -> ()
}
}
}
}
}
The separation is tested via a cmd line flag on the loop tiling pass.
The utility itself allows one to pass in any band of contiguously nested
loops, and can be used by other transforms/utilities. The current
implementation works for hyperrectangular loop nests.
Signed-off-by: Uday Bondhugula <uday@polymagelabs.com>
Differential Revision: https://reviews.llvm.org/D76700
- add `to_extent_tensor`
- rename `create_shape` to `from_extent_tensor` for symmetry
- add `split_at` and `concat` ops for basic shape manipulations
This set of ops is inspired by the requirements of lowering a dynamic-shape-aware batch matmul op. For such an op, the "matrix" dimensions aren't subject to broadcasting but the others are, and so we need to slice, broadcast, and reconstruct the final output shape. Furthermore, the actual broadcasting op used downstream uses a tensor of extents as its preferred shape interface for the actual op that does the broadcasting.
However, this functionality is quite general. It's obvious that `to_extent_tensor` is needed long-term to support many common patterns that involve computations on shapes. We can evolve the shape manipulation ops introduced here. The specific choices made here took into consideration the potentially unranked nature of the !shape.shape type, which means that a simple listing of dimensions to extract isn't possible in general.
Differential Revision: https://reviews.llvm.org/D76817
This fixes a number of warnings, where a function is re-defined after it is tagged as "being imported":
D:\llvm-project\mlir\lib\ExecutionEngine\CRunnerUtils.cpp(24,17): warning: 'print_i32' redeclared without 'dllimport' attribute: 'dllexport' attribute added [-Winconsistent-dllimport]
extern "C" void print_i32(int32_t i) { fprintf(stdout, "%" PRId32, i); }
^
D:\llvm-project\mlir\include\mlir/ExecutionEngine/CRunnerUtils.h(168,42): note: previous declaration is here
extern "C" MLIR_CRUNNERUTILS_EXPORT void print_i32(int32_t i);
^
Differential Revision: https://reviews.llvm.org/D76654
The Dominance analysis currently misses a utility function to find the nearest common dominator of two given blocks. This is required for a huge variety of different control-flow analyses and transformations. This commit adds this function and moves the getNode function from DominanceInfo to DominanceInfoBase, as it also works for post dominators.
Differential Revision: https://reviews.llvm.org/D75507
This change adds a new option to the StandardToLLVM lowering to configure
the bitwidth of the index type independently of the target architecture's
pointer size.
Differential revision: https://reviews.llvm.org/D76353
Multiple operation conversions from the Standard dialect to the LLVM dialect
are trivial one-to-one conversions that use only the pattern defined in base
utility classes such as OneToOneConvertToLLVMPattern and
VectorConvertToLLVMPattern. Use template aliases ("using" declarations) instead
of creating derived classes without new functionality.
With commit 4d60f47 VectorOps was renamed to Vector and the naming of
the CMake target was adjusted. With commit 363dd3f QuantOps was
renamed to Quant, but the naming of the CMake target is left
untouched. This renames the CMake target.
It's common in many dialects to use tensors to themselves hold tensor shapes (for example, the shape is itself the result of some non-trivial calculation). Currently, such dialects have to use `tensor<?xi64>` or worse (like allowing either i32 or i64 tensors to represent shapes). `tensor<?xindex>` is the natural type to represent this, but is currently disallowed. This patch allows it.
Differential Revision: https://reviews.llvm.org/D76726
Summary:
Provide a public VectorConvertToLLVMPattern utility class to implement
conversions with automatic unrolling of operation on multidimensional vectors
to lists of operations on single-dimensional vectors when lowering to the LLVM
dialect. Drop the template-based check on the number of operands since the
actual implementation does not depend on the operand number anymore. This check
only creates spurious concepts (UnaryOpLowering, BinaryOpLowering, etc).
Differential Revision: https://reviews.llvm.org/D76865
Summary:
The Standard-to-LLVM dialect convresion has a set of utility classes that
simplify conversions, including patterns that provide one-to-one conversion
operation conversion with optional result packing. Expose these classes in a
public header so that conversions other than Standard-to-LLVM (e.g. vectors, or
LLVM-based intrinsics) could also use them. Since the patterns are implemented
as class templates and in order to keep the code size limited, keep the
implementation private by resorting to op identifiers instead of template-based
builders.
Differential Revision: https://reviews.llvm.org/D76864
This allows conversion of a ParallelLoop from N induction variables to
some nuber of induction variables less than N.
The first intended use of this is for the GPUDialect to convert
ParallelLoops to iterate over 3 dimensions so they can be launched as
GPU Kernels.
To implement this:
- Normalize each iteration space of the ParallelLoop
- Use the same induction variable in a new ParallelLoop for multiple
original iterations.
- Split the new induction variable back into the original set of values
inside the body of the ParallelLoop.
Subscribers: mgorny, mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, nicolasvasilache, arpith-jacob, mgester, lucyrfox, aartbik, liufengdb, Joonsoo, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D76363
- add method to get back an integer set from flat affine constraints;
this allows a round trip
- use this to complete the simplification of integer sets in
-simplify-affine-structures
- update FlatAffineConstraints::removeTrivialRedundancy to also do GCD
tightening and normalize by GCD (while still keeping it linear time).
Signed-off-by: Uday Bondhugula <uday@polymagelabs.com>
Summary:
The test performs add on vector<16384xf32> with
number of workgroups = (128, 1, 1)
local workgroup size = (128, 1, 1)
On a NVIDIA Quadro P1000, I see the following results:
Command buffer submit time: 13us
Compute shader execution time: 19.616us
Differential Revision: https://reviews.llvm.org/D76799
Summary:
Some operations have custom syntax where an operand is always followed by a
specific token of streams if the operand is present. Parsing such operations
requires the ability to optionally parse an operand. Provide a relevant
function in the custom Op parser.
Differential Revision: https://reviews.llvm.org/D76779
Summary:
The attribute parser fails to correctly parse unsigned 64 bit
attributes as the check `isNegative ? (int64_t)-val.getValue() >= 0
: (int64_t)val.getValue() < 0` will falsely detect an overflow for
unsigned values larger than 2^63-1.
This patch reworks the overflow logic to instead of doing arithmetic
on int64_t use APInt::isSignBitSet() and knowledge of the attribute
type.
Test-cases which verify the de-facto behavior of the parser and
triggered the previous faulty handing of unsigned 64 bit attrbutes are
also added.
Differential Revision: https://reviews.llvm.org/D76493
Summary: The current ConvertStandardToLLVM phase lowers the standard TanHOp to function calls to external tanh symbols. However, this leads to misunderstandings since these external symbols are not defined anywhere. This commit removes the TanHOp lowering functionality from ConvertStandardToLLVM, adapts the LowerGpuOpsToNVVMOps and LowerGpuOpsToROCDLOps passes and adjusts the affected test cases.
Reviewers: mravishankar, herhut
Subscribers: jholewinski, mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, nicolasvasilache, csigg, arpith-jacob, mgester, lucyrfox, aartbik, liufengdb, Joonsoo, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75509
gpu.launch
Current implementation of lowering from loop.parallel to gpu.launch
uses a DictionaryAttr to specify the mapping. Moving this attribute to
be auto-generated from specification as a StructAttr. This simplifies
a lot the logic of looking up and creating this attribute.
Differential Revision: https://reviews.llvm.org/D76165
Summary:
The restriction that a derived attribute should represent an
attribute/be materializable as an attribute was not made clear.
Differential Revision: https://reviews.llvm.org/D76715
Summary:
This revisions performs several cleanups to the generated dialect documentation:
* Standardizes format of attributes/operands/results sections
* Splits out operation/type/dialect documentation generation to allow for composing generated and hand-written documentation
* Add section for declarative assembly syntax and successors
* General cleanup
Differential Revision: https://reviews.llvm.org/D76573
The declarations for these were already part of transforms utils, but
the definitions were left in affine transforms. Move definitions to loop
transforms utils.
Signed-off-by: Uday Bondhugula <uday@polymagelabs.com>
Differential Revision: https://reviews.llvm.org/D76633
Summary:
While here, simplify the lexer a bit by eliminating the unneeded 'operator'
classification of certain sigils, they can just be treated as 'punctuation'.
Reviewers: rriddle!
Subscribers: mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, nicolasvasilache, arpith-jacob, mgester, lucyrfox, liufengdb, Joonsoo, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D76647
When trying to fold an operation during operation creation check that
the operation folding succeeds before inserting the op.
Differential Revision: https://reviews.llvm.org/D76415
Summary:
This allows the custom parser/printer hooks to do interesting things with
the SSA names. This patch:
- Adds a new 'getResultName' method to OpAsmParser that allows a parser
implementation to get information about its result names, along with
a getNumResults() method that allows op parser impls to know how many
results are expected.
- Adds a OpAsmPrinter::printOperand overload that takes an explicit stream.
- Adds a test.string_attr_pretty_name operation that uses these hooks to
do fancy things with the result name.
Reviewers: rriddle!
Subscribers: mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, nicolasvasilache, arpith-jacob, mgester, lucyrfox, liufengdb, Joonsoo, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D76205
Move some of the affine transforms and their test cases to their
respective dialect directory. This patch does not complete the move, but
takes care of a good part.
Renames: prefix 'affine' to affine loop tiling cl options,
vectorize -> super-vectorize
Signed-off-by: Uday Bondhugula <uday@polymagelabs.com>
Differential Revision: https://reviews.llvm.org/D76565
Summary:
This removes the static pass registration, and also cleans up some lingering technical debt.
Differential Revision: https://reviews.llvm.org/D76554
Summary:
This file only contains references to test passes, and was never removed when the test passes were moved to the test/ directory.
Differential Revision: https://reviews.llvm.org/D76553
Fix memref type doc comment to state that -1 indicates a dynamically
shaped dimension and not any negative number.
Signed-off-by: Uday Bondhugula <uday@polymagelabs.com>
Differential Revision: https://reviews.llvm.org/D76557
Two MLIR examples do not link because the library path is different when using Xcode vs Makefiles.
This change adds the configuration name to the library path when building with Xcode.
i.e. LLVM_BUILD_DIR/debug/lib
Summary:
Change AffineOps Dialect structure to better group both IR and Tranforms. This included extracting transforms directly related to AffineOps. Also move AffineOps to Affine.
Differential Revision: https://reviews.llvm.org/D76161
Summary:
After D75831 has been landed, both the generic op and indexed_generic op can
handle 0-D edge case. In the previous patch, only generic op has been updated.
This patch updates the lowering to loops for indexed_generic op. Since they are
almost the sanme, the patch also refactors the common part.
Differential Revision: https://reviews.llvm.org/D76413
The Vector Dialect [document](https://mlir.llvm.org/docs/Dialects/Vector/) discusses the vector abstractions that MLIR supports and the various tradeoffs involved.
One of the layer that is missing in OSS atm is the Hardware Vector Ops (HWV) level.
This revision proposes an AVX512-specific to add a new Dialect/Targets/AVX512 Dialect that would directly target AVX512-specific intrinsics.
Atm, we rely too much on LLVM’s peephole optimizer to do a good job from small insertelement/extractelement/shufflevector. In the future, when possible, generic abstractions such as VP intrinsics should be preferred.
The revision will allow trading off HW-specific vs generic abstractions in MLIR.
Differential Revision: https://reviews.llvm.org/D75987
OperationFolder::tryToFold was running the pre-replacement
action even when there was no constant folding, i.e., when the operation
was just being updated in place but was not going to be replaced. This
led to nested ops being unnecessarily removed from the worklist and only
being processed in the next outer iteration of the greedy pattern
rewriter, which is also why this didn't affect the final output IR but
only the convergence rate. It also led to an op's results' users to be
unnecessarily added to the worklist.
Signed-off-by: Uday Bondhugula <uday@polymagelabs.com>
Differential Revision: https://reviews.llvm.org/D76268
Because MLIR_HAS_EXPORTS is not set, MLIRTarget.cmake is not delivered
to the install area. When this happens, the delivered MLIRConfig.cmake
should not reference it. Independently, we need to determine under what
conditions MLIR_HAS_EXPORTS should be set. Probably we are not exporting
all the libraries correctly.
The Interface libraries were moved from Analysis, but declared in
cmake using add_llvm_library(). This breaks LLVM_BUILD_LLVM_DYLIB
builds.
Differential Revision: https://reviews.llvm.org/D76463
Builder::get{I32,I64}VectorAttr are actually of limited applicability since
vector types can't have zero elements, whereas many uses of this kind of
attribute (such as dimension lists for "transpose"-like and other tensor
ops) often can result in empty lists.
Differential Revision: https://reviews.llvm.org/D76403
Summary: This patch add tests when lowering multiple `gpu.all_reduce` operations in the same kernel. This was previously failing.
Differential Revision: https://reviews.llvm.org/D75930
Summary:
These are not supported by any of the code using `type_cast`. In the general
case, such casting would require memrefs to handle a non-contiguous vector
representation or misaligned vectors (e.g., if the offset of the source memref
is not divisible by vector size, since offset in the target memref is expressed
in the number of elements).
Differential Revision: https://reviews.llvm.org/D76349
Summary:
Utility to perform CallOp Dialect conversion, specifically handling cases where
an argument type has changed and the corresponding CallOp needs to be updated.
Differential Revision: https://reviews.llvm.org/D76326
Summary:
With the move towards dialect registration that does not depend only use
static initialization, we are running into more cases where the dialects
are registered by different methods. For example, TensorFlow still uses
static initialization to register all MLIR core dialects, which prevents
explicit registration of any of them when linking it in. We ran into this
issue in https://github.com/google/iree/pull/982.
To address potential issues with conflicts from non-standard
allocators passed to registerDialectAllocator, made this method
private. Now all dialects can only be registered with their
constructor.
Similarly deduplicates DialectHooks for consistency and makes their
registration follow the same pattern.
Differential Revision: https://reviews.llvm.org/D76329
This commit merges the DRR pattern for std.constant to spv.constant
conversion into the C++ OpConversionPattern. This allows us to have
remove the DRR pattern file. Along the way, this commit enhanced
std.constant to spv.constant conversion to consider type conversions,
which means converting the underlying attributes if necessary.
Differential Revision: https://reviews.llvm.org/D76246
Previously we have a few patterns that were written with DRR. DRR
at the moment does not work nicely with dialect conversion framework.
It generates normal RewritePatterns, while the dialect conversion
framework requires ConversionPatterns to take into consideration
the type conversion. So this commit starts to change existing DRR
patterns for standard ops to OpConversionPattern to incorporate the
SPIR-V type conversion. All patterns are converted except the one
for constant ops, which will happen in a subsequent commit.
Differential Revision: https://reviews.llvm.org/D76245
Non-32-bit scalar types requires special hardware support that may
not exist on all Vulkan-capable GPUs. This is reflected as non-32-bit
scalar types require special capabilities or extensions to be used.
This commit makes SPIRVTypeConverter target environment aware so
that it can properly convert standard types to what is accepted on
the target environment.
Right now if a scalar type bitwidth is not supported in the target
environment, we use 32-bit unconditionally. This requires Vulkan
runtime to also feed in data with a matched bitwidth and layout,
especially for interface types. The Vulkan runtime can do that by
inspecting the SPIR-V module. Longer term, we might want to introduce
a way to control how such case are handled and explicitly fail
if wanted.
Differential Revision: https://reviews.llvm.org/D76244
Types should be checked with the type hierarchy. This should result in
better responsibility division and API surface.
Differential Revision: https://reviews.llvm.org/D76243
This commit unifies target environment queries into a new wrapper
class spirv::TargetEnv and shares across various places needing
the functionality. We still create multiple instances of TargetEnv
though given the parent components (type converters, passes,
conversion targets) have different lifetimes.
In the meantime, LowerABIAttributesPass is updated to take into
consideration the target environment, which requires updates to
tests to provide that.
Differential Revision: https://reviews.llvm.org/D76242
Previously we only consider the version/extension/capability requirement
on the op itself. This commit updates SPIRVConversionTarget to also
take into consideration the values' types when deciding op legality.
Differential Revision: https://reviews.llvm.org/D75876
Previously in SPIRVTypeConverter, we always convert memref types
to StorageBuffer regardless of their memory spaces. This commit
fixes that to let the conversion to look into memory space
properly. For this purpose, a mapping between SPIR-V storage class
and memref memory space is introduced. The mapping is arbitary
decided at the moment and the hope is that we can leverage
string memory space later to be more clear.
Now spv.interface_var_abi cannot contain storage class unless it's
attached to a scalar value, where we need the storage class as side
channel information. Verifications and tests are properly adjusted.
Differential Revision: https://reviews.llvm.org/D76241
Summary:
Although bool and int1 are the same sometimes, using bool constant matches the
semantic better. In any cases, we don't have to care the type of conditions if
we remove the intial value. The type is determined automatically by the returned
type of logical operations.
Differential Revision: https://reviews.llvm.org/D76338
Summary: The usage story in for NDEBUG isn't fleshed out yet, so this revision ensures that none of the diagnostic code exists in the binary.
Differential Revision: https://reviews.llvm.org/D76372
Summary: This is somewhat complex(annoying) as it involves directly tracking the uses within each of the callgraph nodes, and updating them as needed during inlining. The benefit of this is that we can have a more exact cost model, enable inlining some otherwise non-inlinable cases, and also ensure that newly dead callables are properly disposed of.
Differential Revision: https://reviews.llvm.org/D75476
Summary:
This adds support in RewriterGen for calling into the new `PatternRewriter::notifyMatchFailure` hook. This lets derived pattern rewriters display this information to users, an example from DialectConversion is shown below:
```
Legalizing operation : 'std.and'(0x60e0000066a0) {
* Fold {
} -> FAILURE : unable to fold
* Pattern : 'std.and -> (spv.BitwiseAnd)' {
** Failure : operand 0 of op 'std.and' failed to satisfy constraint: '8/16/32/64-bit integer or vector of 8/16/32/64-bit integer values of length 2/3/4'
} -> FAILURE : pattern failed to match
* Pattern : 'std.and -> (spv.LogicalAnd)' {
** Failure : operand 0 of op 'std.and' failed to satisfy constraint: 'bool or vector of bool values of length 2/3/4'
} -> FAILURE : pattern failed to match
} -> FAILURE : no matched legalization pattern
```
Differential Revision: https://reviews.llvm.org/D76335
Summary:
This revision restructures the calling of vector transforms to make it more flexible to ask for lowering through LLVM matrix intrinsics.
This also makes sure we bail out in degenerate cases (i.e. 1) in which LLVM complains about not being able to scalarize.
Differential Revision: https://reviews.llvm.org/D76266
LLVM has a documented mechanism for passing configuration information
to an out of tree project using cmake. See
https://llvm.org/docs/CMake.html#embedding-llvm-in-your-project. This
patch adds similar support for MLIR.
Using this requires something like:
cmake_minimum_required(VERSION 3.4.3)
project(SimpleProject)
find_package(MLIR REQUIRED CONFIG)
include_directories(${LLVM_INCLUDE_DIRS})
include_directories(${MLIR_INCLUDE_DIRS})
link_directories(${LLVM_BUILD_LIBRARY_DIR})
add_definitions(${LLVM_DEFINITIONS})
set(CMAKE_MODULE_PATH
${LLVM_CMAKE_DIR}
${MLIR_CMAKE_DIR}
)
include(AddLLVM)
include(TableGen)
include(AddMLIR)
add_executable(test-opt test-opt.cpp)
llvm_update_compile_flags(test-opt)
get_property(dialect_libs GLOBAL PROPERTY MLIR_DIALECT_LIBS)
get_property(conversion_libs GLOBAL PROPERTY MLIR_CONVERSION_LIBS)
message(dialects=${dialect_libs})
set(LIBS
${dialect_libs}
${conversion_libs}
MLIRLoopOpsTransforms
MLIRLoopAnalysis
MLIRAnalysis
MLIRDialect
MLIREDSC
MLIROptLib
MLIRParser
MLIRPass
MLIRQuantizerFxpMathConfig
MLIRQuantizerSupport
MLIRQuantizerTransforms
MLIRSPIRV
MLIRSPIRVTestPasses
MLIRSPIRVTransforms
MLIRTransforms
MLIRTransformUtils
MLIRTestDialect
MLIRTestIR
MLIRTestPass
MLIRTestTransforms
MLIRSupport
MLIRIR
MLIROptLib
LLVMSupport
LLVMCore
LLVMAsmParser
)
target_link_libraries(test-opt ${LIBS})
Differential Revision: https://reviews.llvm.org/D76047
Summary:
Renamed QuantOps to Quant to avoid the Ops suffix. All dialects will contain
ops, so the Ops suffix is redundant.
Differential Revision: https://reviews.llvm.org/D76318
Summary:
This revision adds a new hook, `notifyMatchFailure`, that allows for notifying the rewriter that a match failure is coming with the provided reason. This hook takes as a parameter a callback that fills a `Diagnostic` instance with the reason why the match failed. This allows for the rewriter to decide how this information can be displayed to the end-user, and may completely ignore it if desired(opt mode). For now, DialectConversion is updated to include this information in the debug output.
Differential Revision: https://reviews.llvm.org/D76203
MLIR supports terminators that have the same successor block with different
block operands, which cannot be expressed in the LLVM's phi-notation as the
block identifier is used to tell apart the predecessors. This limitation can be
worked around by branching to a new block instead, with this new block
unconditionally branching to the original successor and forwarding the
argument. Until now, this transformation was performed during the conversion
from the Standard to the LLVM dialect. This does not scale well to multiple
dialects targeting the LLVM dialect as all of them would have to be aware of
this limitation and perform the preparatory transformation. Instead, do it as a
separate pass and run it immediately before the translation.
Differential Revision: https://reviews.llvm.org/D75619
A memref argument is converted into a pointer-to-struct argument
of type `{T*, T*, i64, i64[N], i64[N]}*` in the wrapper function,
where T is the converted element type and N is the memref rank.
Differential Revision: https://reviews.llvm.org/D76059
Summary: This adds bitfields that map to the dialect attribute verifier hooks. This also moves over the Test dialect to have its declaration generated.
Differential Revision: https://reviews.llvm.org/D76254
Summary: PatternState was a mechanism to pass state between the match and rewrite calls of a RewritePattern. With the rise of matchAndRewrite, this class is unused and unnecessary. This revision removes PatternState and simplifies PatternMatchResult to just be a LogicalResult. A future revision will replace all usages of PatternMatchResult/matchSuccess/matchFailure with LogicalResult equivalents.
Differential Revision: https://reviews.llvm.org/D76202
- rename vars that had inst suffixes (due to ops earlier being
known as insts); other renames for better readability
- drop unnecessary matches in test cases
- iterate without block terminator
- comment/doc updates
- instBodySkew -> affineForOpBodySkew
Signed-off-by: Uday Bondhugula <uday@polymagelabs.com>
Differential Revision: https://reviews.llvm.org/D76214
Summary:
This regional op in the QuantOps dialect will be used to wrap
high-precision ops into atomic units for quantization. All the values
used by the internal ops are captured explicitly by the op inputs. The
quantization parameters of the inputs and outputs are stored in the
attributes.
Subscribers: jfb, mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, nicolasvasilache, arpith-jacob, mgester, lucyrfox, aartbik, Joonsoo, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75972
Summary: This generates the class declarations for dialects using the existing 'Dialect' tablegen classes.
Differential Revision: https://reviews.llvm.org/D76185
Setting MLIR_TABLEGEN_EXE would prevent building the native tool which is used in cross-compiling
Differential Revision: https://reviews.llvm.org/D75299
Summary:
- remove stale declarations on flat affine constraints
- avoid allocating small vectors where possible
- clean up code comments, rename some variables
Differential Revision: https://reviews.llvm.org/D76117
Summary:
To enable this, two changes are needed:
1) Add an optional attribute `padding` to linalg.conv.
2) Compute if the indices accessing is out of bound in the loops. If so, use the
padding value `0`. Otherwise, use the value derived from load.
In the patch, the padding only works for lowering without other transformations,
e.g., tiling, fusion, etc.
Differential Revision: https://reviews.llvm.org/D75722
Summary:
This revision adds lowering of vector.contract to llvm.intr.matrix_multiply.
Note that there is currently a mismatch between the MLIR vector dialect which
expects row-major layout and the LLVM matrix intrinsics which expect column
major layout.
As a consequence, we currently only match a vector.contract with indexing maps
that express column-major matrix multiplication.
Other cases would require additional transposes and it is better to wait for
LLVM intrinsics to provide a per-operation attribute that would specify which
layout is expected.
A separate integration test, not submitted to MLIR core, has independently
verified that correct execution occurs on a 2x2x2 matrix multiplication.
Differential Revision: https://reviews.llvm.org/D76014
Summary:
The direct lowering of vector.broadcast into LLVM has been replaced by
progressive lowering into elementary vector ops. This also required a
small refactoring of a llvm.mlir test that used a direct vector.broadcast
operator (just to define a matmul).
Reviewers: nicolasvasilache, andydavis1, rriddle
Reviewed By: nicolasvasilache
Subscribers: mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, nicolasvasilache, arpith-jacob, mgester, lucyrfox, liufengdb, Joonsoo, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D76143
Summary: A number of transform import StandardOps despite not being dependent on it. Cleaned it up to better understand what dialects each of these transforms depend on.
Differential Revision: https://reviews.llvm.org/D76112
Previously we only consider the version/capability/extension requirements
on ops themselves. Some types in SPIR-V also require special extensions
or capabilities to be used. For example, non-32-bit integers/floats
will require different capabilities and/or extensions depending on
where they are used because it may mean special hardware abilities.
This commit adds query methods to SPIR-V type class hierarchy to support
querying extensions and capabilities. We don't go through ODS for
auto-generating such information given that we don't have them in
SPIR-V machine readable grammar and there are just a few types.
Differential Revision: https://reviews.llvm.org/D75875
Nicolas Vasilache