This commit adds two resource limits, max_compute_workgroup_size
and max_compute_workgroup_invocations as resource limits to
the target environment. They are not used at the current moment,
but they will affect the SPIR-V CodeGen. Adding for now to have
a proper target environment modelling.
Differential Revision: https://reviews.llvm.org/D73905
Summary:
Currently BuildableType is assumed to be preceded by a builder. This prevents constructing types that don't have a callable 'get' method with the builder. This revision reworks the format to be like attribute builders, i.e. by accepting $_builder within the format itself.
Differential Revision: https://reviews.llvm.org/D73736
Summary: This revision add support for accepting a few type constraints, e.g. AllTypesMatch, when inferring types for operands and results. This is used to remove the c++ parsers for several additional operations.
Differential Revision: https://reviews.llvm.org/D73735
Summary:
Replace the generic zero- and one-result builders in LLVM::CallOp with a custom
builder that takes an LLVMFuncOp, which can be used to extract the result type
and create the symbol reference attribute. This is merely a convenience for
upcoming changes. The ODS-generated builders remain present.
Introduce LLVM::LLVMType::isVoidTy by analogy with the underlying LLVM type.
Differential Revision: https://reviews.llvm.org/D73895
The patterns for converting `std.alloc` and `std.dealoc` can be configured to
use `llvm.alloca` instead of calling `malloc` and `free`. This configuration
has been only possible through a command-line flag, despite the presence of a
(misleading) parameter in the pass constructor. Use the parameter instead and
only initalize it from the command line flags if the pass is constructed from
the mlir-opt registration.
Summary:
These hooks were originally introduced to support passes deriving the
StandardToLLVM conversion, in particular converting types from different
dialects to LLVM types in a single-step conversion. They are no longer in use
since the pass and conversion infrastructure has evolved sufficiently to make
defining new passes with exactly the same functionality simple through the use
of populate* functions, conversion targets and type converters. Remove the
hooks. Any users of this hooks can call the dialect conversion infrastructure
directly instead, which is likely to require less LoC than these hooks.
Differential Revision: https://reviews.llvm.org/D73795
Summary:
In the original design, gpu.launch required explicit capture of uses
and passing them as operands to the gpu.launch operation. This was
motivated by infrastructure restrictions rather than design. This
change lifts the requirement and removes the concept of kernel
arguments from gpu.launch. Instead, the kernel outlining
transformation now does the explicit capturing.
This is a breaking change for users of gpu.launch.
Differential Revision: https://reviews.llvm.org/D73769
Summary:
This patch introduces an alternative calling convention for
MemRef function arguments in LLVM dialect. It converts MemRef
function arguments to LLVM bare pointers to the MemRef element
type instead of creating a MemRef descriptor. Bare pointers are
then promoted to a MemRef descriptors at the beginning of the
function. This calling convention is only enabled with a flag.
Reviewers: ftynse, bondhugula, nicolasvasilache, rriddle, mehdi_amini
Reviewed By: ftynse, rriddle, mehdi_amini
Subscribers: Joonsoo, flaub, merge_guards_bot, jholewinski, mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, csigg, arpith-jacob, mgester, lucyrfox, herhut, aartbik, liufengdb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72802
Summary:
After the `subview` operation was migrated from Linalg to Standard, it changed
semantics and does not guarantee the absence of out-of-bounds accesses through
the created view anymore. Compute the size of the subview to make sure it
always fits within the view (subviews in last iterations of the loops may be
smaller than those in other iterations).
Differential Revision: https://reviews.llvm.org/D73614
This commit adds a pattern to lower linalg.generic for reduction
to spv.GroupNonUniform* ops. Right now this only supports integer
reduction on 1-D input memref. Shader entry point ABI is queried
to make sure that the input memref's shape matches the local
workgroup's invocation configuration. This makes sure that the
workload fits in one local workgroup so that we can leverage
SPIR-V group non-uniform operations.
linglg.generic is a structured op that preserves the right level
of information. It is easier to recognize reduction at this level
than performing analysis on loops.
This commit also exposes `getElementPtr` in SPIRVLowering.h given
that it's a generally useful utility function.
Differential Revision: https://reviews.llvm.org/D73437
The refactored MemRefType::get() calls all intend to clone from another
memref type, with some modifications. In fact, some calls dropped memory space
during the cloning. Migrate them to the cloning API so that nothing gets
dropped if they are not explicitly listed.
It's close to NFC but not quite, as it helps with propagating memory spaces in
some places.
Differential Revision: https://reviews.llvm.org/D73296
Summary:
MLIR materializes various enumeration-based LLVM IR operands as enumeration
attributes using ODS. This requires bidirectional conversion between different
but very similar enums, currently hardcoded. Extend the ODS modeling of
LLVM-specific enumeration attributes to include the name of the corresponding
enum in the LLVM C++ API as well as the names of specific enumerants. Use this
new information to automatically generate the conversion functions between enum
attributes and LLVM API enums in the two-way conversion between the LLVM
dialect and LLVM IR proper.
Differential Revision: https://reviews.llvm.org/D73468
Summary:
This revision switches over many operations to use the declarative methods for defining the assembly specification. This updates operations in the NVVM, ROCDL, Standard, and VectorOps dialects.
Differential Revision: https://reviews.llvm.org/D73407
Summary:
This revision add support, and testing, for generating the parser and printer from the declarative operation format.
Differential Revision: https://reviews.llvm.org/D73406
Summary:
This is the first revision in a series that adds support for declaratively specifying the asm format of an operation. This revision
focuses solely on parsing the format. Future revisions will add support for generating the proper parser/printer, as well as
transitioning the syntax definition of many existing operations.
This was originally proposed here:
https://llvm.discourse.group/t/rfc-declarative-op-assembly-format/340
Differential Revision: https://reviews.llvm.org/D73405
Summary:
In some cases, one may want to use different names for C++ symbol of an
enumerand from its string representation. In particular, in the LLVM dialect
for, e.g., Linkage, we would like to preserve the same enumerand names as LLVM
API and the same textual IR form as LLVM IR, yet the two are different
(CamelCase vs snake_case with additional limitations on not being a C++
keyword).
Modify EnumAttrCaseInfo in OpBase.td to include both the integer value and its
string representation. By default, this representation is the same as C++
symbol name. Introduce new IntStrAttrCaseBase that allows one to use different
names. Exercise it for LLVM Dialect Linkage attribute. Other attributes will
follow as separate changes.
Differential Revision: https://reviews.llvm.org/D73362
Summary:
The 'gpu.terminator' operation is used as the terminator for the
regions of gpu.launch. This is to disambugaute them from the
return operation on 'gpu.func' functions.
This is a breaking change and users of the gpu dialect will need
to adapt their code when producting 'gpu.launch' operations.
Reviewers: nicolasvasilache
Subscribers: mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, csigg, arpith-jacob, mgester, lucyrfox, liufengdb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73620
Summary:
This will help catch improper use of the MLIR API's. In particular, this
catches an error that was manifesting as nondeterministic assertion
failures (the nondeterminism was due to the failure happening only when the
StorageUniquer's DenseMap's probing happened to compare two specific
keys).
No test. The fact that all the existing tests pass with this additional
invariant gives confidence that it is correct/useful.
Differential Revision: https://reviews.llvm.org/D73645
Summary:
Canonicalization and folding patterns in StandardOps may interfere with the needs
of Linalg. This revision introduces specific foldings for dynamic memrefs that can
be proven to be static.
Very concretely:
Determines whether it is possible to fold it away in the parent Linalg op:
```mlir
%1 = memref_cast %0 : memref<8x16xf32> to memref<?x?xf32>
%2 = linalg.slice %1 ... : memref<?x?xf32> ...
// or
%1 = memref_cast %0 : memref<8x16xf32, affine_map<(i, j)->(16 * i + j)>>
to memref<?x?xf32>
linalg.generic(%1 ...) : memref<?x?xf32> ...
```
into
```mlir
%2 = linalg.slice %0 ... : memref<8x16xf32> ...
// or
linalg.generic(%0 ... : memref<8x16xf32, affine_map<(i, j)->(16 * i + j)>>
```
Reviewers: ftynse, aartbik, jsetoain, tetuante, asaadaldien
Subscribers: mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, liufengdb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73565
Summary:
Barrier is a simple operation that takes no arguments and returns
nothing, but implies a side effect (synchronization of all threads)
Reviewers: jdoerfert
Subscribers: mgorny, guansong, mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, nicolasvasilache, arpith-jacob, mgester, lucyrfox, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72400
This is how it should've been and brings it more in line with
std::string_view. There should be no functional change here.
This is mostly mechanical from a custom clang-tidy check, with a lot of
manual fixups. It uncovers a lot of minor inefficiencies.
This doesn't actually modify StringRef yet, I'll do that in a follow-up.
Summary:
Operation represents all of the uses of each result with one use list, so manipulating the use list of a specific result requires filtering the main use list. This revision adds an optimization for the case of single result operations to avoid this filtering.
Differential Revision: https://reviews.llvm.org/D73430
Summary:
This also removes the explicit pattern for loop.terminator to ensure
that the terminator is only erased if the parent op is rewritten.
Reductions are not yet supported.
Reviewers: nicolasvasilache
Subscribers: mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, aartbik, liufengdb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73348
Summary:
The intrinsic operation added multiple type annotations to the llvm intrinsic operations, but only one is needed.
The related tests in llvmir-intrinsics.mlir checked the wrong number and are adjusted as well.
Reviewers: nicolasvasilache, ftynse
Reviewed By: ftynse
Subscribers: merge_guards_bot, ftynse, mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, liufengdb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73470
Summary: This pass deletes all symbols that are found to be unreachable. This is done by computing the set of operations that are known to be live, propagating that liveness to other symbols, and then deleting all symbols that are not within this live set.
Differential Revision: https://reviews.llvm.org/D72482
Summary: This revision refactors the implementation of the symbol use-list functionality to be a bit cleaner, as well as easier to reason about. Aside from code cleanup, this revision updates the user contract to never recurse into operations if they define a symbol table. The current functionality, which does recurse, makes it difficult to examine the uses held by a symbol table itself. Moving forward users may provide a specific region to examine for uses instead.
Differential Revision: https://reviews.llvm.org/D73427
Summary: The new internal representation of operation results now allows for accessing the result types to be more efficient. Changing the API to ArrayRef is more efficient and removes the need to explicitly materialize vectors in several places.
Differential Revision: https://reviews.llvm.org/D73429
Summary: This allows for providing a default "catchall" legality check that is not dependent on specific operations or dialects. For example, this can be useful to check legality based on the specific types of operation operands or results.
Differential Revision: https://reviews.llvm.org/D73379
Summary:
Remove 'valuesToRemoveIfDead' from PatternRewriter API. The removal
functionality wasn't implemented and we decided [1] not to implement it in
favor of having more powerful DCE approaches.
[1] https://github.com/tensorflow/mlir/pull/212
Reviewers: rriddle, bondhugula
Reviewed By: rriddle
Subscribers: liufengdb, mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, nicolasvasilache, arpith-jacob, mgester, lucyrfox, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72545
This commit changes the logic of `getBuiltinVariableValue` to get
or create the builtin variable in the nearest symbol table. This
will allow us to use this function in other partial conversion
cases where we haven't created the spv.module yet.
Differential Revision: https://reviews.llvm.org/D73416
This commit exposes the func op conversion pattern via a new
`populateBuiltinFuncToSPIRVPatterns` function from the standard
to SPIR-V conversion passs. This is structurally better given
that func op belongs to the builtin dialect. More importantly,
this makes the pattern reusable to other dialect to SPIR-V
dialect conversion as other dialect can well adopt builtin
func op instead of having its own. Besides, it's very common
to use func ops as test wrappers in lit tests, so test passes
will need to handle func ops too.
Differential Revision: https://reviews.llvm.org/D73421
Thus far certain SPIR-V ops have been required to be in spv.module.
While this provides strong verification to catch unexpected errors,
it's quite rigid and makes progressive lowering difficult. Sometimes
we would like to partially lower ops from other dialects, which may
involve creating ops like global variables that should be placed in
other module-like ops. So this commit relaxes the requirement of
such SPIR-V ops' scope to module-like ops. Similarly for function-
like ops.
Differential Revision: https://reviews.llvm.org/D73415
Summary:
Rewrites the extract/insert_slices operation in terms of
strided_slice/insert_strided_slice ops with intermediate
tuple uses (that should get optimimized away with typical
usage). This is done in a separate "pass" to enable testing
this particular rewriting in isolation.
Reviewers: nicolasvasilache, andydavis1, ftynse
Reviewed By: nicolasvasilache
Subscribers: merge_guards_bot, mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, liufengdb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73295
Summary:
Rationale:
Some examples were using "offsets : [0, 2]" syntax which
should use a "=" instead. The same examples were referring
to the integer attribute array as k-dimensional, which is
a bit confusing (it is 1-dimensional, with k elements).
Changed to "k-sized".
Reviewers: nicolasvasilache, andydavis1, ftynse
Reviewed By: nicolasvasilache
Subscribers: merge_guards_bot, mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, liufengdb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73293
Summary:
LLVMIRIntrinsicGen is using LLVM_Op as the base class for intrinsics.
This works for LLVM intrinsics in the LLVM Dialect, but when we are
trying to convert custom intrinsics that originate from a custom
LLVM dialect (like NVVM or ROCDL) these usually have a different
"cppNamespace" that needs to be applied to these dialect.
These dialect specific characteristics (like "cppNamespace")
are typically organized by creating a custom op (like NVVM_Op or
ROCDL_Op) that passes the correct dialect to the LLVM_OpBase class.
It seems natural to allow LLVMIRIntrinsicGen to take that into
consideration when generating the conversion code from one of these
dialect to a set of target specific intrinsics.
Reviewers: rriddle, andydavis1, antiagainst, nicolasvasilache, ftynse
Subscribers: jdoerfert, mehdi_amini, jpienaar, burmako, shauheen, arpith-jacob, mgester, lucyrfox, aartbik, liufengdb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73233
This reverts commit eec36909c1.
This modeling is incorrect. baseAttr is intended for attribute
decorators that are not backed by C++ attribute classes. It essentially
says DerivedAttr isa BaseAttr, which is wrong for ArrayAttr classes.
If one needs to store the element type, it should be stored as a
separate filed in the tablegen class.
Summary:
This diff extends the Linalg EDSC builders so we can easily create mixed
tensor/buffer linalg.generic ops. This is expected to be useful for
HLO -> Linalg lowering.
The StructuredIndexed struct is made to derive from ValueHandle and can
now capture a type + indexing expressions. This is used to represent return
tensors.
Pointwise unary and binary builders are extended to allow both output buffers
and return tensors. This has implications on the number of region arguments.
Reviewers: ftynse, hanchung, asaadaldien
Subscribers: mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, aartbik, liufengdb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73149
Summary:
Add method in ODS to specify verification for operations implementing a
OpInterface. Use this with infer type op interface to verify that the
inferred type matches the return type and remove special case in
TestPatterns.
This could also have been achieved by using OpInterfaceMethod but verify
seems pretty common and it is not an arbitrary method that just happened
to be named verifyTrait, so having it be defined in special way seems
appropriate/better documenting.
Differential Revision: https://reviews.llvm.org/D73122
Summary:
This diff extends the Linalg EDSC builders so we can easily create mixed
tensor/buffer linalg.generic ops. This is expected to be useful for
HLO -> Linalg lowering.
The `StructuredIndexed` struct is made to derive from `ValueHandle` and can
now capture a type + indexing expressions. This is used to represent return
tensors.
Pointwise unary and binary builders are extended to allow both output buffers
and return tensors. This has implications on the number of region arguments.
Reviewers: ftynse, herhut, hanchung, asaadaldien, stellaraccident
Reviewed By: asaadaldien
Subscribers: merge_guards_bot, mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, aartbik, liufengdb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72863
For the generated builder taking in unwrapped attribute values,
if the argument is a string, we should avoid wrapping it in quotes;
otherwise we are always setting the string attribute to contain
the string argument's name. The quotes come from StrinAttr's
`constBuilderCall`, which is reasonable for string literals, but
not function arguments containing strings.
Differential Revision: https://reviews.llvm.org/D72977
Summary:
This is based on the use of code constantly checking for an attribute on
a model and instead represents the distinct operaion with a different
op. Instead, this op can be used to provide better filtering.
Reverts "Revert "[mlir] Create a gpu.module operation for the GPU Dialect.""
This reverts commit ac446302ca4145cdc89f377c0c364c29ee303be5 after
fixing internal Google issues.
This additionally updates ROCDL lowering to use the new gpu.module.
Reviewers: herhut, mravishankar, antiagainst, nicolasvasilache
Subscribers: jholewinski, mgorny, mehdi_amini, jpienaar, burmako, shauheen, csigg, arpith-jacob, mgester, lucyrfox, aartbik, liufengdb, llvm-commits, mravishankar, rriddle, antiagainst, bkramer
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72921
Summary:
Add a `llvm.cmpxchg` op as a counterpart to LLVM IR's `cmpxchg` instruction.
Note that the `weak`, `volatile`, and `syncscope` attributes are not yet supported.
This will be useful for upcoming parallel versions of affine.for and generally
for reduction-like semantics (especially for reductions that can't make use
of `atomicrmw`, e.g. `fmax`).
Reviewers: ftynse, nicolasvasilache
Reviewed By: ftynse
Subscribers: merge_guards_bot, jfb, mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, liufengdb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72995
Summary:
Generalize broadcastable trait to variadic operands. Update the
documentation that still talked about element type as part of
broadcastable trait (that bug was already fixed). Also rename
Broadcastable to ResultBroadcastableShape to be more explicit that the
trait affects the result shape (it is possible for op to allow
broadcastable operands but not have result shape that is broadcast
compatible with operands).
Doing some intermediate work to have getBroadcastedType take an optional
elementType as input and use that if specified, instead of the common
element type of type1 and type2 in this function.
Differential Revision: https://reviews.llvm.org/D72559
mlir currently fails to build on Solaris:
/vol/llvm/src/llvm-project/dist/mlir/lib/Conversion/VectorToLoops/ConvertVectorToLoops.cpp:78:20: error: reference to 'index_t' is ambiguous
IndexHandle zero(index_t(0)), one(index_t(1));
^
/usr/include/sys/types.h:103:16: note: candidate found by name lookup is 'index_t'
typedef short index_t;
^
/vol/llvm/src/llvm-project/dist/mlir/include/mlir/EDSC/Builders.h:27:8: note: candidate found by name lookup is 'mlir::edsc::index_t'
struct index_t {
^
and many more.
Given that POSIX reserves all identifiers ending in `_t` 2.2.2 The Name Space <https://pubs.opengroup.org/onlinepubs/9699919799/functions/V2_chap02.html>, it seems
quite unwise to use such identifiers in user code, even more so without a distinguished
prefix.
The following patch fixes this by renaming `index_t` to `index_type`.
cases.
Tested on `amd64-pc-solaris2.11` and `sparcv9-sun-solaris2.11`.
Differential Revision: https://reviews.llvm.org/D72619
Summary:
This is a simple extension to allow vectorization to work not only on GenericLinalgOp
but more generally across named ops too.
For now, this still only vectorizes matmul-like ops but is a step towards more
generic vectorization of Linalg ops.
Reviewers: ftynse
Subscribers: mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, aartbik, liufengdb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72942
Summary:
First step towards the consolidation
of a lot of vector related utilities
that are now all over the place
(or even duplicated).
Reviewers: nicolasvasilache, andydavis1
Reviewed By: nicolasvasilache, andydavis1
Subscribers: merge_guards_bot, mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, liufengdb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72955
Summary:
This op is the counterpart to LLVM's atomicrmw instruction. Note that
volatile and syncscope attributes are not yet supported.
This will be useful for upcoming parallel versions of `affine.for` and generally
for reduction-like semantics.
Differential Revision: https://reviews.llvm.org/D72741
Summary:
This is a temporary implementation to support Flang. The LLVM-IR parser
will need to be extended in some way to support recursive types. The
exact approach here is still a work-in-progress.
Unfortunately, this won't pass roundtrip testing yet. Adding a comment
to the test file as a reminder.
Differential Revision: https://reviews.llvm.org/D72542
Summary: This field is currently not used by anything, and using a ClassID instance provides better support for more efficient classof.
Reviewers: mehdi_amini, nicolasvasilache
Reviewed By: mehdi_amini
Subscribers: merge_guards_bot, jpienaar, burmako, shauheen, antiagainst, nicolasvasilache, arpith-jacob, mgester, lucyrfox, liufengdb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72822
Introduce a new generator for MLIR tablegen driver that consumes LLVM IR
intrinsic definitions and produces MLIR ODS definitions. This is useful to
bulk-generate MLIR operations equivalent to existing LLVM IR intrinsics, such
as additional arithmetic instructions or NVVM.
A test exercising the generation is also added. It reads the main LLVM
intrinsics file and produces ODS to make sure the TableGen model remains in
sync with what is used in LLVM.
Differential Revision: https://reviews.llvm.org/D72926
In SPIR-V, when a new version is introduced, it is possible some
existing extensions will be incorporated into it so that it becomes
implicitly declared if targeting the new version. This affects
conversion target specification because we need to take this into
account when allowing what extensions to use.
For a capability, it may also implies some other capabilities,
for example, the `Shader` capability implies `Matrix` the capability.
This should also be taken into consideration when preparing the
conversion target: when we specify an capability is allowed, all
its recursively implied capabilities are also allowed.
This commit adds utility functions to query implied extensions for
a given version and implied capabilities for a given capability
and updated SPIRVConversionTarget to use them.
This commit also fixes a bug in availability spec. When a symbol
(op or enum case) can be enabled by an extension, we should drop
it's minimal version requirement. Being enabled by an extension
naturally means the symbol can be used by *any* SPIR-V version
as long as the extension is supported. The grammar still encodes
the 'version' field for such cases, but it should be interpreted
as a different way: rather than meaning a minimal version
requirement, it says the symbol becomes core at that specific
version.
Differential Revision: https://reviews.llvm.org/D72765
By default, for an enum attribute, we will generate a list of equality
comparisons for all supported cases inside it's predicate. This list
can be fairly large for certain SPIR-V enum attributes. Instead, we
already have such a list generated by EnumsGen in the symbolize
functions. Leverage that to simplify the generated C++ code.
Differential Revision: https://reviews.llvm.org/D72763
Certain SPIR-V capabilities are only available in certain SPIR-V versions
or extensions. Also a SPIR-V capability may implicitly declares other
capabilities.
This commit updates gen_spirv_dialect.py to support generating such
information into SPIRVBase.td. It requires us to topologically sort
all capabilities because now a capability can refer to another one.
This commits also registers a few extensions because their symbols are
used by capability availability.
Note that this commit hasn't updated SPIRVConversionTarget to take
into consideration such relationship yet. That will be done in a
following-up commit.
Differential Revision: https://reviews.llvm.org/D72760
This is (more?) usable by GDB pretty printers and seems nicer to write.
There's one tricky caveat that in C++14 (LLVM's codebase today) the
static constexpr member declaration is not a definition - so odr use of
this constant requires an out of line definition, which won't be
provided (that'd make all these trait classes more annoyidng/expensive
to maintain). But the use of this constant in the library implementation
is/should always be in a non-odr context - only two unit tests needed to
be touched to cope with this/avoid odr using these constants.
Based on/expanded from D72590 by Christian Sigg.
Summary:
* Add shaped container type interface which allows infering the shape, element
type and attribute of shaped container type separately. Show usage by way of
tensor type inference trait which combines the shape & element type in
infering a tensor type;
- All components need not be specified;
- Attribute is added to allow for layout attribute that was previously
discussed;
* Expand the test driver to make it easier to test new creation instances
(adding new operands or ops with attributes or regions would trigger build
functions/type inference methods);
- The verification part will be moved out of the test and to verify method
instead of ops implementing the type inference interface in a follow up;
* Add MLIRContext as arg to possible to create type for ops without arguments,
region or location;
* Also move out the section in OpDefinitions doc to separate ShapeInference doc
where the shape function requirements can be captured;
- Part of this would move to the shape dialect and/or shape dialect ops be
included as subsection of this doc;
* Update ODS's variable usage to match camelBack format for builder,
state and arg variables;
- I could have split this out, but I had to make some changes around
these and the inconsistency bugged me :)
Differential Revision: https://reviews.llvm.org/D72432
Summary:
This diff moves the conversion pass declaration closer to its definition
and makes the namespacing of passes consistent with the rest of the
infrastructure (i.e. `mlir::linalg::createXXXPass` -> `mlir::createXXXPass`).
Reviewers: ftynse, jpienaar, mehdi_amini
Subscribers: rriddle, burmako, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, aartbik, liufengdb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72766
This commit defines a new SPIR-V dialect attribute for specifying
a SPIR-V target environment. It is a dictionary attribute containing
the SPIR-V version, supported extension list, and allowed capability
list. A SPIRVConversionTarget subclass is created to take in the
target environment and sets proper dynmaically legal ops by querying
the op availability interface of SPIR-V ops to make sure they are
available in the specified target environment. All existing conversions
targeting SPIR-V is changed to use this SPIRVConversionTarget. It
probes whether the input IR has a `spv.target_env` attribute,
otherwise, it uses the default target environment: SPIR-V 1.0 with
Shader capability and no extra extensions.
Differential Revision: https://reviews.llvm.org/D72256
Summary:
This allows for users to cache printer state, which can be costly to recompute. Each of the IR print methods gain a new overload taking this new state class.
Depends On D72293
Reviewed By: jpienaar
Differential Revision: https://reviews.llvm.org/D72294
Summary:
This diff fixes issues with the semantics of linalg.generic on tensors that appeared when converting directly from HLO to linalg.generic.
The changes are self-contained within MLIR and can be captured and tested independently of XLA.
The linalg.generic and indexed_generic are updated to:
To allow progressive lowering from the value world (a.k.a tensor values) to
the buffer world (a.k.a memref values), a linalg.generic op accepts
mixing input and output ranked tensor values with input and output memrefs.
```
%1 = linalg.generic #trait_attribute %A, %B {other-attributes} :
tensor<?x?xf32>,
memref<?x?xf32, stride_specification>
-> (tensor<?x?xf32>)
```
In this case, the number of outputs (args_out) must match the sum of (1) the
number of output buffer operands and (2) the number of tensor return values.
The semantics is that the linalg.indexed_generic op produces (i.e.
allocates and fills) its return values.
Tensor values must be legalized by a buffer allocation pass before most
transformations can be applied. Such legalization moves tensor return values
into output buffer operands and updates the region argument accordingly.
Transformations that create control-flow around linalg.indexed_generic
operations are not expected to mix with tensors because SSA values do not
escape naturally. Still, transformations and rewrites that take advantage of
tensor SSA values are expected to be useful and will be added in the near
future.
Subscribers: bmahjour, mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72555
Summary:
This is based on the use of code constantly checking for an attribute on
a model and instead represents the distinct operaion with a different
op. Instead, this op can be used to provide better filtering.
Reviewers: herhut, mravishankar, antiagainst, rriddle
Reviewed By: herhut, antiagainst, rriddle
Subscribers: liufengdb, aartbik, jholewinski, mgorny, mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, nicolasvasilache, csigg, arpith-jacob, mgester, lucyrfox, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72336
Summary:
The visibility defines the structural reachability of the symbol within the IR. Symbols can define one of three visibilities:
* Public
The symbol \may be accessed from outside of the visible IR. We cannot assume that we can observe all of the uses of this symbol.
* Private
The symbol may only be referenced from within the operations in the current symbol table, via SymbolRefAttr.
* Nested
The symbol may be referenced by operations in symbol tables above the current symbol table, as long as each symbol table parent also defines a non-private symbol. This allows or referencing the symbol from outside of the defining symbol table, while retaining the ability for the compiler to see all uses.
These properties help to reason about the properties of a symbol, and will be used in a follow up to implement a dce pass on dead symbols.
A few examples of what this would look like in the IR are shown below:
module @public_module {
// This function can be accessed by 'live.user'
func @nested_function() attributes { sym_visibility = "nested" }
// This function cannot be accessed outside of 'public_module'
func @private_function() attributes { sym_visibility = "private" }
}
// This function can only be accessed from within this module.
func @private_function() attributes { sym_visibility = "private" }
// This function may be referenced externally.
func @public_function()
"live.user"() {uses = [@public_module::@nested_function,
@private_function,
@public_function]} : () -> ()
Depends On D72043
Reviewed By: mehdi_amini
Differential Revision: https://reviews.llvm.org/D72044
Summary:
This enables tracking calls that cross symbol table boundaries. It also simplifies some of the implementation details of CallableOpInterface, i.e. there can only be one region within the callable operation.
Depends On D72042
Reviewed By: jpienaar
Differential Revision: https://reviews.llvm.org/D72043
Summary: This updates the use list algorithms to support querying from a specific symbol, allowing for the collection and detection of nested references. This works by walking the parent "symbol scopes" and applying the existing algorithm at each level.
Reviewed By: jpienaar
Differential Revision: https://reviews.llvm.org/D72042
Summary: The current syntax for AffineMapAttr and IntegerSetAttr conflict with function types, making it currently impossible to round-trip function types(and e.g. FuncOp) in the IR. This revision changes the syntax for the attributes by wrapping them in a keyword. AffineMapAttr is wrapped with `affine_map<>` and IntegerSetAttr is wrapped with `affine_set<>`.
Reviewed By: nicolasvasilache, ftynse
Differential Revision: https://reviews.llvm.org/D72429
Summary: Introduce m_Constant() which allows matching a constant operation without forcing the user also to capture the attribute value.
Differential Revision: https://reviews.llvm.org/D72397
Summary:
This diff makes it easier to create a `linalg.reshape` op
and adds an EDSC builder api test to exercise the new builders.
Reviewers: ftynse, jpienaar
Subscribers: mehdi_amini, rriddle, burmako, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, aartbik, liufengdb, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72580
Summary:
- update zero_extendi and sign_extendi in edsc/intrinsic namespace
- Builder API test for zero_extendi and sign_extendi
Differential Revision: https://reviews.llvm.org/D72298
Thus far we can only generate the same set of methods even for
operations in different dialects. This is problematic for dialects that
want to generate additional operation class methods programmatically,
e.g., a special builder method or attribute getter method. Apparently
we cannot update the OpDefinitionsGen backend every time when such
a need arises. So this CL introduces a hook into the OpDefinitionsGen
backend to allow dialects to emit additional methods and traits to
operation classes.
Differential Revision: https://reviews.llvm.org/D72514
Introduce a set of function that promote a memref argument of a `gpu.func` to
workgroup memory using memory attribution. The promotion boils down to
additional loops performing the copy from the original argument to the
attributed memory in the beginning of the function, and back at the end of the
function using all available threads. The loop bounds are specified so as to
adapt to any size of the workgroup. These utilities are intended to compose
with other existing utilities (loop coalescing and tiling) in cases where the
distribution of work across threads is uneven, e.g. copying a 2D memref with
only the threads along the "x" dimension. Similarly, specialization of the
kernel to specific launch sizes should be implemented as a separate pass
combining constant propagation and canonicalization.
Introduce a simple attribute-driven pass to test the promotion transformation
since we don't have a heuristic at the moment.
Differential revision: https://reviews.llvm.org/D71904
Summary:
This diff implements the progressive lowering of strided_slice to either:
1. extractelement + insertelement for the 1-D case
2. extract + optional strided_slice + insert for the n-D case.
This combines properly with the other conversion patterns to lower all the way to LLVM.
Appropriate tests are added.
Reviewers: ftynse, rriddle, AlexEichenberger, andydavis1, tetuante
Reviewed By: andydavis1
Subscribers: merge_guards_bot, mehdi_amini, jpienaar, burmako, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72310
Summary:
This diff adds a new operation to linalg to allow reshaping of an
existing view into a new view in the same buffer at the same offset.
More specifically:
The `linalg.reshape` op produces a new view whose sizes are a reassociation
of the original `view`. Depending on whether or not the reassociated
MemRefType is contiguous, the resulting memref may require explicit alloc
and copies.
A reassociation is defined as a continous grouping of dimensions and is
represented with a affine map array attribute. In the future, non-continous
groupings may be allowed (i.e. permutations, reindexings etc).
For now, it is assumed that either:
1. a reassociation produces and consumes contiguous MemRefType or,
2. the reshape op will be folded into its consumers (by changing the shape
of the computations).
All other cases are undefined behavior and a reshape op may not lower to
LLVM if it cannot be proven statically that it does not require alloc+copy.
A reshape may either collapse or expand dimensions, depending on the
relationship between source and target memref ranks. The verification rule
is that the reassociation maps are applied to the memref with the larger
rank to obtain the memref with the smaller rank. In the case of a dimension
expansion, the reassociation maps can be interpreted as inverse maps.
Examples:
```mlir
// Dimension collapse (i, j) -> i' and k -> k'
%1 = linalg.reshape %0 [(i, j, k) -> (i, j),
(i, j, k) -> (k)] :
memref<?x?x?xf32, stride_spec> into memref<?x?xf32, stride_spec_2>
```
```mlir
// Dimension expansion i -> (i', j') and (k) -> (k')
%1 = linalg.reshape %0 [(i, j, k) -> (i, j),
(i, j, k) -> (k)] :
memref<?x?xf32, stride_spec> into memref<?x?x?xf32, stride_spec_2>
```
The relevant invalid and roundtripping tests are added.
Reviewers: AlexEichenberger, ftynse, rriddle, asaadaldien, yangjunpro
Subscribers: kiszk, merge_guards_bot, mehdi_amini, jpienaar, burmako, shauheen, antiagainst, arpith-jacob, mgester, lucyrfox, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72168
Summary: This fixes the return value of helper methods on the base range class.
Reviewed By: jpienaar
Differential Revision: https://reviews.llvm.org/D72127
This commit fixes shader ABI attributes to use `spv.` as the prefix
so that they match the dialect's namespace. This enables us to add
verification hooks in the SPIR-V dialect to verify them.
Reviewed By: mravishankar
Differential Revision: https://reviews.llvm.org/D72062
Summary:
This changes the implementation of OpResult to have some of the results be represented inline in Value, via a pointer int pair of Operation*+result number, and the rest being trailing objects on the main operation. The full details of the new representation is detailed in the proposal here:
https://groups.google.com/a/tensorflow.org/g/mlir/c/XXzzKhqqF_0/m/v6bKb08WCgAJ
The only difference between here and the above proposal is that we only steal 2-bits for the Value kind instead of 3. This means that we can only fit 2-results inline instead of 6. This allows for other users to steal the final bit for PointerUnion/etc. If necessary, we can always steal this bit back in the future to save more space if 3-6 results are common enough.
Reviewed By: jpienaar
Differential Revision: https://reviews.llvm.org/D72020
This commit updates gen_spirv_dialect.py to query the grammar and
generate availability spec for various enum attribute definitions
and all defined ops.
Reviewed By: mravishankar
Differential Revision: https://reviews.llvm.org/D72095
Summary:
This diff adds support to allow `linalg.generic` and
`linalg.indexed_generic` to take tensor input and output
arguments.
The subset of output tensor operand types must appear
verbatim in the result types after an arrow. The parser,
printer and verifier are extended to accomodate this
behavior.
The Linalg operations now support variadic ranked tensor
return values. This extension exhibited issues with the
current handling of NativeCall in RewriterGen.cpp. As a
consequence, an explicit cast to `SmallVector<Value, 4>`
is added in the proper place to support the new behavior
(better suggestions are welcome).
Relevant cleanups and name uniformization are applied.
Relevant invalid and roundtrip test are added.
Reviewers: mehdi_amini, rriddle, jpienaar, antiagainst, ftynse
Subscribers: burmako, shauheen, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72022
Lots of SPIR-V ops take enum attributes and certain enum cases
need extra capabilities or extensions to be available. This commit
extends to allow specifying availability spec on enum cases.
Extra utility functions are generated for the corresponding enum
classes to return the availability requirement. The availability
interface implemention for a SPIR-V op now goes over all enum
attributes to collect the availability requirements.
Reviewed By: mravishankar
Differential Revision: https://reviews.llvm.org/D71947
* replaceAllUsesWith may be supplied with a null value.
* some compilers fail to implicitly convert single result operations to
OpaqueValue, so add an explicit OpOperand::set(Value) method.
Summary:
This is part of an ongoing cleanup and uniformization work.
This diff performs 3 types of cleanups:
1. Uniformize transformation names.
2. Replace all pattern operands that need not be captured by `$_`
3. Replace all usage of pattern captured op by the normalized `op` name (instead of positional parameters such as `$0`)
Reviewers: ftynse
Subscribers: mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72081
Summary: This is part of an ongoing cleanup and uniformization work.
Reviewers: ftynse
Subscribers: mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72079
Summary: This is part of an ongoing cleanup and uniformization work.
Subscribers: mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D72078
This allows us to include the definitions of these attributes in
other files without pulling in all dependencies for lowering.
Reviewed By: mravishankar
Differential Revision: https://reviews.llvm.org/D72054
Summary:
This commit fixes links to code directories and uses doc links on
mlir.llvm.org where possible. The docs in TableGen dialect definition
is also updated to reflect recent developments.
Reviewed By: mravishankar
Differential Revision: https://reviews.llvm.org/D72051
Summary: A new class is added, IRMultiObjectWithUseList, that allows for representing an IR use list that holds multiple sub values(used in this case for OpResults). This class provides all of the same functionality as the base IRObjectWithUseList, but for specific sub-values. This saves a word per operation result and is a necessary step in optimizing the layout of operation results. For now the use list is placed on the operation itself, so zero-result operations grow by a word. When the work for optimizing layout is finished, this can be moved back to being a trailing object based on memory/runtime benchmarking.
Reviewed By: jpienaar
Differential Revision: https://reviews.llvm.org/D71955
Summary: The successor operand counts are directly tied to block operands anyways, and this simplifies the trailing objects of Operation(i.e. one less computation to perform).
Reviewed By: mehdi_amini
Differential Revision: https://reviews.llvm.org/D71949
SPIR-V has a few mechanisms to control op availability: version,
extension, and capabilities. These mechanisms are considered as
different availability classes.
This commit introduces basic definitions for modelling SPIR-V
availability classes. Specifically, an `Availability` class is
added to SPIRVBase.td, along with two subclasses: MinVersion
and MaxVersion for versioning. SPV_Op is extended to take a
list of `Availability`. Each `Availability` instance carries
information for generating op interfaces for the corresponding
availability class and also the concrete availability
requirements.
With the availability spec on ops, we can now auto-generate the
op interfaces of all SPIR-V availability classes and also
synthesize the op's implementations of these interfaces. The
interface generation is done via new TableGen backends
-gen-avail-interface-{decls|defs}. The op's implementation is
done via -gen-spirv-avail-impls.
Differential Revision: https://reviews.llvm.org/D71930
MSVC has trouble resolving the static 'printOptionValue' from the method on llvm:🆑:opt/list. This change renames the static method to avoid this conflict.
This change refactors pass options to be more similar to how statistics are modeled. More specifically, the options are specified directly on the pass instead of in a separate options class. (Note that the behavior and specification for pass pipelines remains the same.) This brings about several benefits:
* The specification of options is much simpler
* The round-trip format of a pass can be generated automatically
* This gives a somewhat deeper integration with "configuring" a pass, which we could potentially expose to users in the future.
PiperOrigin-RevId: 286953824
This means that in-place, or root, updates need to use explicit calls to `startRootUpdate`, `finalizeRootUpdate`, and `cancelRootUpdate`. The major benefit of this change is that it enables in-place updates in DialectConversion, which simplifies the FuncOp pattern for example. The major downside to this is that the cases that *may* modify an operation in-place will need an explicit cancel on the failure branches(assuming that they started an update before attempting the transformation).
PiperOrigin-RevId: 286933674
This will enable future commits to reimplement the internal implementation of OpResult without needing to change all of the existing users. This is part of a chain of commits optimizing the size of operation results.
PiperOrigin-RevId: 286930047
This will enable future commits to reimplement the internal implementation of OpResult without needing to change all of the existing users. This is part of a chain of commits optimizing the size of operation results.
PiperOrigin-RevId: 286919966
This is an initial step to refactoring the representation of OpResult as proposed in: https://groups.google.com/a/tensorflow.org/g/mlir/c/XXzzKhqqF_0/m/v6bKb08WCgAJ
This change will make it much simpler to incrementally transition all of the existing code to use value-typed semantics.
PiperOrigin-RevId: 286844725
Rename the 'shlis' operation in the standard dialect to 'shift_left'. Add tests
for this operation (these have been missing so far) and add a lowering to the
'shl' operation in the LLVM dialect.
Add also 'shift_right_signed' (lowered to LLVM's 'ashr') and 'shift_right_unsigned'
(lowered to 'lshr').
The original plan was to name these operations 'shift.left', 'shift.right.signed'
and 'shift.right.unsigned'. This works if the operations are prefixed with 'std.'
in MLIR assembly. Unfortunately during import the short form is ambigous with
operations from a hypothetical 'shift' dialect. The best solution seems to omit
dots in standard operations for now.
Closestensorflow/mlir#226
PiperOrigin-RevId: 286803388
This requires using explicitly default copy constructor and copy assignment
operator instead of hand-rolled ones. These classes are indeed cheap to copy
since they are wrappers around a pointer to the implementation. This change
makes sure templated code can use standard type traits to understand that
copying such objects is cheap and appeases analysis tools such as clang-tidy.
PiperOrigin-RevId: 286725565
This will allow us to lower most of gpu.all_reduce (when all_reduce
doesn't exist in the target dialect) within the GPU dialect, and only do
target-specific lowering for the shuffle op.
PiperOrigin-RevId: 286548256
This is the block argument equivalent of the existing `getAsmResultNames` hook.
Closestensorflow/mlir#329
COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/mlir/pull/329 from plaidml:flaub-region-arg-names fc7876f2d1335024e441083cd25263fd6247eb7d
PiperOrigin-RevId: 286523299
Concatting lists in TableGen is easy, creating unique lists less so. There is no reason for duplicated op traits so we could throw an error instead but duplicates could occur due to concatting different list of traits in ODS (e.g., for convenience reasons), so just dedup them during Operator trait construction instead.
PiperOrigin-RevId: 286488423
Update vector transfer_read/write ops to operatate on memrefs with vector element type.
This handle cases where the memref vector element type represents the minimal memory transfer unit (or multiple of the minimal memory transfer unit).
PiperOrigin-RevId: 286482115
This CL allows specifying an additional name for specifying the .td file that is used to generate the doc for a dialect. This is necessary for a dialect like Linalg which has different "types" of ops that are used in different contexts.
This CL also restructures the Linalg documentation and renames LinalgLibraryOps -> LinalgStructuredOps but is otherwise NFC.
PiperOrigin-RevId: 286450414
Adds vector ReshapeOp to the VectorOps dialect. An aggregate vector reshape operation, which aggregates multiple hardware vectors, can enable optimizations during decomposition (e.g. loading one input hardware vector and performing multiple rotate and scatter store operations to the vector output).
PiperOrigin-RevId: 286440658
Introduces some centralized methods to move towards
consistent use of i32 as vector subscripts.
Note: sizes/strides/offsets attributes are still i64
PiperOrigin-RevId: 286434133
This function template has been introduced in the early days of MLIR to work
around the absence of common type for ranges of values (operands, block
argumeents, vectors, etc). Core IR now provides ValueRange for exactly this
purpose. Use it instead of the template parameter.
PiperOrigin-RevId: 286431338
Added test cases for the newly added LLVM operations and lowering features.
Closestensorflow/mlir#300
COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/mlir/pull/300 from dfki-jugr:std_to_llvm da6168bbc1a369ae2e99ad3881fdddd82f075dd4
PiperOrigin-RevId: 286231169
This enables providing a default implementation of an interface method. This method is defined on the Trait that is attached to the operation, and thus has all of the same constraints and properties as any other interface method. This allows for interface authors to provide a conservative default implementation for certain methods, without requiring that all users explicitly define it. The default implementation can be specified via the argument directly after the interface method body:
StaticInterfaceMethod<
/*desc=*/"Returns whether two array of types are compatible result types for an op.",
/*retTy=*/"bool",
/*methodName=*/"isCompatibleReturnTypes",
/*args=*/(ins "ArrayRef<Type>":$lhs, "ArrayRef<Type>":$rhs),
/*methodBody=*/[{
return ConcreteOp::isCompatibleReturnTypes(lhs, rhs);
}],
/*defaultImplementation=*/[{
/// Returns whether two arrays are equal as strongest check for
/// compatibility by default.
return lhs == rhs;
}]
PiperOrigin-RevId: 286226054
'```mlir' is used to indicate the code block is MLIR code/should use MLIR syntax
highlighting, while '{.mlir}' was a markdown extension that used a style file
to color the background differently of the code block. The background color
extension was a custom one that we can retire given we have syntax
highlighting.
Also change '```td' to '```tablegen' to match chroma syntax highlighting
designation.
PiperOrigin-RevId: 286222976
Introduce affine.prefetch: op to prefetch using a multi-dimensional
subscript on a memref; similar to affine.load but has no effect on
semantics, but only on performance.
Provide lowering through std.prefetch, llvm.prefetch and map to llvm's
prefetch instrinsic. All attributes reflected through the lowering -
locality hint, rw, and instr/data cache.
affine.prefetch %0[%i, %j + 5], false, 3, true : memref<400x400xi32>
Signed-off-by: Uday Bondhugula <uday@polymagelabs.com>
Closestensorflow/mlir#225
COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/mlir/pull/225 from bondhugula:prefetch 4c3b4e93bc64d9a5719504e6d6e1657818a2ead0
PiperOrigin-RevId: 286212997
The definition of the function template LLVM::ModuleTranslation::lookupValues
has been located in a source file. As long as it has been the only file that
actually called into the function, this did not cause any problem. However, it
creates linking issues if the function is used from other translation units.
PiperOrigin-RevId: 286203078
When memory attributions are present in `gpu.func`, require that they are of
memref type and live in memoryspaces 3 and 5 for workgroup and private memory
attributions, respectively. Adapt the conversion from the GPU dialect to the
NVVM dialect to drop the private memory space from attributions as NVVM is able
to model them as local `llvm.alloca`s in the default memory space.
PiperOrigin-RevId: 286161763
The lowering of MemRef types to the LLVM dialect is connected to the underlying
runtime representation of structured memory buffers. It has changed several
times in the past and reached the current state of a LLVM structured-typed
descriptor containing two pointers and all sizes. In several reported use
cases, a different, often simpler, lowering scheme is required. For example,
lowering statically-shaped memrefs to bare LLVM pointers to simplify aliasing
annotation. Split the pattern population functions into those include
memref-related operations and the remaining ones. Users are expected to extend
TypeConverter::convertType to handle the memref types differently.
PiperOrigin-RevId: 286030610
This class provides a simplified mechanism for defining a switch over a set of types using llvm casting functionality. More specifically, this allows for defining a switch over a value of type T where each case corresponds to a type(CaseT) that can be used with dyn_cast<CaseT>(...). An example is shown below:
// Traditional piece of code:
Operation *op = ...;
if (auto constant = dyn_cast<ConstantOp>(op))
...;
else if (auto return = dyn_cast<ReturnOp>(op))
...;
else
...;
// New piece of code:
Operation *op = ...;
TypeSwitch<Operation *>(op)
.Case<ConstantOp>([](ConstantOp constant) { ... })
.Case<ReturnOp>([](ReturnOp return) { ... })
.Default([](Operation *op) { ... });
Aside from the above, TypeSwitch supports return values, void return, multiple types per case, etc. The usability is intended to be very similar to StringSwitch.
(Using c++14 template lambdas makes everything even nicer)
More complex example of how this makes certain things easier:
LogicalResult process(Constant op);
LogicalResult process(ReturnOp op);
LogicalResult process(FuncOp op);
TypeSwitch<Operation *, LogicalResult>(op)
.Case<ConstantOp, ReturnOp, FuncOp>([](auto op) { return process(op); })
.Default([](Operation *op) { return op->emitError() << "could not be processed"; });
PiperOrigin-RevId: 286003613
This CL adds more Linalg EDSC ops and tests to support building pointwise operations along with conv and dilated_conv.
This also fixes a bug in the existing linalg_matmul EDSC and beefs up the test.
The current set of ops is already enough to build an interesting, albeit simple, model used internally.
PiperOrigin-RevId: 285838012
This updates the lowering pipelines from the GPU dialect to lower-level
dialects (NVVM, SPIRV) to use the recently introduced gpu.func operation
instead of a standard function annotated with an attribute. In particular, the
kernel outlining is updated to produce gpu.func instead of std.func and the
individual conversions are updated to consume gpu.funcs and disallow standard
funcs after legalization, if necessary. The attribute "gpu.kernel" is preserved
in the generic syntax, but can also be used with the custom syntax on
gpu.funcs. The special kind of function for GPU allows one to use additional
features such as memory attribution.
PiperOrigin-RevId: 285822272
This keeps the IR valid and consistent as it is expected that each block should have a valid parent region/operation. Previously, converted blocks were kept floating without a valid parent region.
PiperOrigin-RevId: 285821687
The conversion from the Loops dialect to the Standard dialect, also known as
loop-to-cfg lowering, has been historically a function pass. It can be required
on non-Standard function Ops, in particular the recently introduced GPU
functions. Make the conversion an operation pass instead of a function pass.
PiperOrigin-RevId: 285814560
This PR targest issue tensorflow/mlir#295. It exposes the already existing
subiew promotion pass as a declarative pattern
Change-Id: If901ebef9fb53fcd0b12ecc536f6b174ce320b92
Closestensorflow/mlir#315
COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/mlir/pull/315 from tetuante:issue295 8e5f268b6d85f31015c33505329dbd7a4db97ac5
PiperOrigin-RevId: 285801463
Similar to insert/extract vector instructions but
(1) work on 1-D vectors only
(2) allow for a dynamic index
%c3 = constant 3 : index
%0 = vector.insertelement %arg0, %arg1[%c : index] : vector<4xf32>
%1 = vector.extractelement %arg0[%c3 : index] : vector<4xf32>
PiperOrigin-RevId: 285792205
ExtractSlicesOp extracts slices of its vector operand and with a specified tiling scheme.
This operation centralizes the tiling scheme around a single op, which simplifies vector op unrolling and subsequent pattern rewrite transformations.
PiperOrigin-RevId: 285761129
- bring op description comment in sync with the doc
- fix misformat in doc
Signed-off-by: Uday Bondhugula <uday@polymagelabs.com>
Closestensorflow/mlir#317
COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/mlir/pull/317 from bondhugula:quickfix 7fcd945b318c973b2488b702874c87526855c8ef
PiperOrigin-RevId: 285574527
This will be evolved into a simple programming model for custom ops and custom layers in followup CLs.
This CL also deletes the obsolete tablegen's reference-impl.td that was using EDSCs.
PiperOrigin-RevId: 285459545
This change allows for DialectConversion to attempt folding as a mechanism to legalize illegal operations. This also expands folding support in OpBuilder::createOrFold to generate new constants when folding, and also enables it to work in the context of a PatternRewriter.
PiperOrigin-RevId: 285448440
This cleans up the implementation of the various operation print methods. This is done via a combination of code cleanup, adding new streaming methods to the printer(e.g. operand ranges), etc.
PiperOrigin-RevId: 285285181
This type is not used anymore now that Linalg view and subview have graduated to std and that alignment is supported on alloc.
PiperOrigin-RevId: 285213424
Both work for the current use case, but the latter allows implementing
prefix sums and is a little easier to understand for partial warps.
PiperOrigin-RevId: 285145287
It is sometimes useful to create operations separately from the builder before insertion as it may be easier to erase them in isolation if necessary. One example use case for this is folding, as we will only want to insert newly generated constant operations on success. This has the added benefit of fixing some silent PatternRewriter failures related to cloning, as the OpBuilder 'clone' methods don't call createOperation.
PiperOrigin-RevId: 285086242
This CL adds more common information to StructuredOpsUtils.h
The n_view attribute is retired in favor of args_in + args_out but the CL is otherwise NFC.
PiperOrigin-RevId: 285000621
Add one more simplification for floordiv and mod affine expressions.
Examples:
(2*d0 + 1) floordiv 2 is simplified to d0
(8*d0 + 4*d1 + d2) floordiv 4 simplified to 4*d0 + d1 + d2 floordiv 4.
etc.
Similarly, (4*d1 + 1) mod 2 is simplified to 1,
(2*d0 + 8*d1) mod 8 simplified to 2*d0 mod 8.
Change getLargestKnownDivisor to return int64_t to be consistent and
to avoid casting at call sites (since the return value is used in expressions
of int64_t/index type).
Signed-off-by: Uday Bondhugula <uday@polymagelabs.com>
Closestensorflow/mlir#202
COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/mlir/pull/202 from bondhugula:affine b13fcb2f1c00a39ca5434613a02408e085a80e77
PiperOrigin-RevId: 284866710
Move the definition of gpu.launch_func operation from hand-rolled C++
implementation to the ODS framework. Also move the documentation. This only
performs the move and remains a non-functional change, a follow-up will clean
up the custom functions that can be auto-generated using ODS.
PiperOrigin-RevId: 284842252
This has several benefits:
* The implementation is much cleaner and more efficient.
* The ranges now have support for many useful operations: operator[], slice, drop_front, size, etc.
* Value ranges can now directly query a range for their types via 'getTypes()': e.g:
void foo(Operation::operand_range operands) {
auto operandTypes = operands.getTypes();
}
PiperOrigin-RevId: 284834912
This patch closes issue tensorflow/mlir#272
We add a standalone iterator permutation transformation to Linalg.
This transformation composes a permutation map with the maps in the
"indexing_maps" attribute. It also permutes "iterator_types"
accordingly.
Change-Id: I7c1e693b8203aeecc595a7c012e738ca1100c857
Closestensorflow/mlir#307
COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/mlir/pull/307 from tetuante:issue272 f7908d58792f4111119721885e247045104f1131
PiperOrigin-RevId: 284824102
This reorganizes the vector transformations to be more easily testable as patterns and more easily composable into fused passes in the future.
PiperOrigin-RevId: 284817474
Add some convenience build methods to SPIR-V ops and update the
lowering to use these methods where possible.
For SPIRV::CompositeExtractOp move the method to deduce type of
element based on base and indices into a convenience function. Some
additional functionality needed to handle differences between parsing
and verification methods.
PiperOrigin-RevId: 284794404
These come from a non-standard extenion that is not available on Github, so it
only clutters the documentation source with {.mlir} or {.ebnf} tags.
PiperOrigin-RevId: 284733003
Avoid `error: could not convert ?(const char*)"reduction"? from ?const char*? to ?llvm::StringLiteral?`. Tested with gcc-5.5.
PiperOrigin-RevId: 284677810
For example
%0 = vector.shuffle %x, %y [3 : i32, 2 : i32, 1 : i32, 0 : i32] : vector<2xf32>, vector<2xf32>
yields a vector<4xf32> result with a permutation of the elements of %x and %y
PiperOrigin-RevId: 284657191
Each of the support classes for Block are now moved into a new header BlockSupport.h. The successor iterator class is also reimplemented as an indexed_accessor_range. This makes the class more efficient, and expands on its available functionality.
PiperOrigin-RevId: 284646792
Many ranges want similar functionality from a range type(e.g. slice/drop_front/operator[]/etc.), so these classes provide a generic implementation that may be used by many different types of ranges. This removes some code duplication, and also empowers many of the existing range types in MLIR(e.g. result type ranges, operand ranges, ElementsAttr ranges, etc.). This change only updates RegionRange and ValueRange, more ranges will be updated in followup commits.
PiperOrigin-RevId: 284615679
This CL starts extracting commonalities between dialects that use the structured ops abstractions. Also fixes an OSS build issue where StringRef were incorrectly used with constexpr.
PiperOrigin-RevId: 284591114
The existing GPU to SPIR-V lowering created a spv.module for every
function with gpu.kernel attribute. A better approach is to lower the
module that the function lives in (which has the attribute
gpu.kernel_module) to a spv.module operation. This better captures the
host-device separation modeled by GPU dialect and simplifies the
lowering as well.
PiperOrigin-RevId: 284574688
Unifies vector op unrolling transformation, by using the same unrolling implementation for contraction and elementwise operations.
Removes fakefork/join operations which are non longer needed now that we have the InsertStridedSlice operation.
PiperOrigin-RevId: 284570784
This CL uses the newly expanded matcher support to easily detect when a linalg.generic has a multiply-accumulate body. A linalg.generic with such a body is rewritten as a vector contraction.
This CL additionally limits the rewrite to the case of matrix multiplication on contiguous and statically shaped memrefs for now.
Before expanding further, we should harden the infrastructure for expressing custom ops with the structured ops abstraction.
PiperOrigin-RevId: 284566659
Follows ValueRange in representing a generic abstraction over the different
ways to represent a range of Regions. This wrapper is not as ValueRange and only
considers the current cases of interest: MutableArrayRef<Region> and
ArrayRef<std::unique_ptr<Region>> as occurs during op construction vs op region
querying.
Note: ArrayRef<std::unique_ptr<Region>> allows for unset regions, so this range
returns a pointer to a Region instead of a Region.
PiperOrigin-RevId: 284563229
This CL adds support for building matchers recursively.
The following matchers are provided:
1. `m_any()` can match any value
2. `m_val(Value *)` binds to a value and must match it
3. `RecursivePatternMatcher<OpType, Matchers...>` n-arity pattern that matches `OpType` and whose operands must be matched exactly by `Matchers...`.
This allows building expression templates for patterns, declaratively, in a very natural fashion.
For example pattern `p9` defined as follows:
```
auto mul_of_muladd = m_Op<MulFOp>(m_Op<MulFOp>(), m_Op<AddFOp>());
auto mul_of_anyadd = m_Op<MulFOp>(m_any(), m_Op<AddFOp>());
auto p9 = m_Op<MulFOp>(m_Op<MulFOp>(
mul_of_muladd, m_Op<MulFOp>()),
m_Op<MulFOp>(mul_of_anyadd, mul_of_anyadd));
```
Successfully matches `%6` in:
```
%0 = addf %a, %b: f32
%1 = addf %a, %c: f32 // matched
%2 = addf %c, %b: f32
%3 = mulf %a, %2: f32 // matched
%4 = mulf %3, %1: f32 // matched
%5 = mulf %4, %4: f32 // matched
%6 = mulf %5, %5: f32 // matched
```
Note that 0-ary matchers can be used as leaves in place of n-ary matchers. This alleviates from passing explicit `m_any()` leaves.
In the future, we may add extra patterns to specify that operands may be matched in any order.
PiperOrigin-RevId: 284469446
This allows for users to provide operand_range and result_range in builder.create<> calls, instead of requiring an explicit copy into a separate data structure like SmallVector/std::vector.
PiperOrigin-RevId: 284360710
This class represents a generic abstraction over the different ways to represent a range of Values: ArrayRef<Value *>, operand_range, result_range. This class will allow for removing the many instances of explicit SmallVector<Value *, N> construction. It has the same memory cost as ArrayRef, and only suffers cost from indexing(if+elsing the different underlying representations).
This change only updates a few of the existing usages, with more to be changed in followups; e.g. 'build' API.
PiperOrigin-RevId: 284307996
This adds an additional filtering mode for printing after a pass that checks to see if the pass actually changed the IR before printing it. This "change" detection is implemented using a SHA1 hash of the current operation and its children.
PiperOrigin-RevId: 284291089
- for the symbol rules, the code was updated but the doc wasn't.
Signed-off-by: Uday Bondhugula <uday@polymagelabs.com>
Closestensorflow/mlir#284
COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/mlir/pull/284 from bondhugula:doc 9aad8b8a715559f7ce61265f3da3f8a3c11b45ea
PiperOrigin-RevId: 284283712
Previously the error case was using a sentinel in the error case which was bad. Also make the one `build` invoke the other `build` to reuse verification there.
And follow up on suggestion to use formatv which I missed during previous review.
PiperOrigin-RevId: 284265762
Move the definition of the GPU launch opreation from hand-rolled C++ code to
ODS framework. This only does the moves, a follow-up is necessary to clean up
users of custom functions that could be auto-generated by ODS.
PiperOrigin-RevId: 284261856
The "FunctionLike" and "IsIsolatedFromAbove" op traits are now defined as named
records in base ODS file. Use those instead of NativeOpTrait referring to the
C++ class name in the ODS definition of LLVMFuncOp. NFC.
PiperOrigin-RevId: 284260891
Since these operations lower to [insert|extract][element|value] at LLVM
dialect level, neither element nor value would correctly reflect the meaning.
PiperOrigin-RevId: 284240727
Accept the address space of the global as a builder argument when constructing
an LLVM::GlobalOp instance. This decreases the reliance of LLVM::GlobalOp users
on the internal name of the attribute used for this purpose. Update several
uses of the address space in GPU to NVVM conversion.
PiperOrigin-RevId: 284233254
Move the definition of the GPU function opreation from hand-rolled C++ code to
ODS framework. This only does the moves, a follow-up is necessary to clean up
users of custom functions that could be auto-generated by ODS.
PiperOrigin-RevId: 284233245
For ops with infer type op interface defined, generate version that calls the inferal method on build. This is intermediate step to removing special casing of SameOperandsAndResultType & FirstAttrDereivedResultType. After that would be generating the inference code, with the initial focus on shaped container types. In between I plan to refactor these a bit to reuse generated paths. The intention would not be to add the type inference trait in multiple places, but rather to take advantage of the current modelling in ODS where possible to emit it instead.
Switch the `inferReturnTypes` method to be static.
Skipping ops with regions here as I don't like the Region vs unique_ptr<Region> difference at the moment, and I want the infer return type trait to be useful for verification too. So instead, just skip it for now to avoid churn.
PiperOrigin-RevId: 284217913
GPU functions use memory attributions, a combination of Op attributes and
region arguments, to specify function-wide buffers placed in workgroup or
private memory spaces. Introduce a lowering pattern for GPU functions to be
converted to LLVM functions taking into account memory attributions. Workgroup
attributions get transformed into module-level globals with unique names
derived from function names. Private attributions get converted into
llvm.allocas inside the function body. In both cases, we inject at the
beginning of the function the IR that obtains the raw pointer to the data and
populates a MemRef descriptor based on the MemRef type of buffer, making
attributions compose with the rest of the MemRef lowering and transparent for
use with std.load and std.store. While using raw pointers instead of
descriptors might have been more efficient, it is better implemented as a
canonicalization or a separate transformation so that non-attribution memrefs
could also benefit from it.
PiperOrigin-RevId: 284208396
Updates vector ContractionOp to use proper vector masks (produced by CreateMaskOp/ConstantMaskOp).
Leverages the following canonicalizations in unrolling unit test: CreateMaskOp -> ConstantMaskOp, StridedSliceOp(ConstantMaskOp) -> ConstantMaskOp
Removes IndexTupleOp (no longer needed now that we have vector mask ops).
Updates all unit tests.
PiperOrigin-RevId: 284182168
The AddressOf operation in the LLVM dialect return a pointer to a global
variable. The latter may be in a non-default address space as indicated by the
"addr_space" attribute. Check that the address space of the pointer returned by
AddressOfOp matches that of the referenced GlobalOp. Update the AddressOfOp
builder to respect this constraint.
PiperOrigin-RevId: 284138860
This patch closes issue tensorflow/mlir#271.
It adds an optional permutation map to declarative tiling transformations.
The map is expressed as a list of integers.
Closestensorflow/mlir#288
COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/mlir/pull/288 from tetuante:issue271 2df2938d6a1f01b3bc404ded08dea2dd1e10b588
PiperOrigin-RevId: 284064151
This allows for more interesting behavior from users, e.g. enabling the ability to dump the IR to a separate file for each pass invocation.
PiperOrigin-RevId: 284059447
A CompositeInsertOp operation make a copy of a composite object,
while modifying one part of it.
Closestensorflow/mlir#292
COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/mlir/pull/292 from denis0x0D:sandbox/composite_insert 2200962b9057bda53cd2f2866b461e2797196380
PiperOrigin-RevId: 284036551
Statistics are a way to keep track of what the compiler is doing and how effective various optimizations are. It is useful to see what optimizations are contributing to making a particular program run faster. Pass-instance specific statistics take this even further as you can see the effect of placing a particular pass at specific places within the pass pipeline, e.g. they could help answer questions like "what happens if I run CSE again here".
Statistics can be added to a pass by simply adding members of type 'Pass::Statistics'. This class takes as a constructor arguments: the parent pass pointer, a name, and a description. Statistics can be dumped by the pass manager in a similar manner to how pass timing information is dumped, i.e. via PassManager::enableStatistics programmatically; or -pass-statistics and -pass-statistics-display via the command line pass manager options.
Below is an example:
struct MyPass : public OperationPass<MyPass> {
Statistic testStat{this, "testStat", "A test statistic"};
void runOnOperation() {
...
++testStat;
...
}
};
$ mlir-opt -pass-pipeline='func(my-pass,my-pass)' foo.mlir -pass-statistics
Pipeline Display:
===-------------------------------------------------------------------------===
... Pass statistics report ...
===-------------------------------------------------------------------------===
'func' Pipeline
MyPass
(S) 15 testStat - A test statistic
MyPass
(S) 6 testStat - A test statistic
List Display:
===-------------------------------------------------------------------------===
... Pass statistics report ...
===-------------------------------------------------------------------------===
MyPass
(S) 21 testStat - A test statistic
PiperOrigin-RevId: 284022014
SPIR-V/Vulkan spec requires the workgroups size to be specified with
the spv.ExecutionMode operation. This was hard-wired to be set to a
particular value. It is now changed to be configurable by clients of
the pass or of the patterns that implement the lowering from GPU to
SPIRV.
PiperOrigin-RevId: 284017482
This change adds support for non-congruent indices in the operation ordering within a basic block. This effect of this is that insertions are less likely to cause an invalidation of the ordering within a block. This has a big effect on modules that have very large basic blocks.
PiperOrigin-RevId: 283858136
In some situations a diagnostic may optionally be emitted by the presence of a location, e.g. attribute and type verification. These situations currently require extra 'if(loc) emitError(...); return failure()' wrappers that make verification clunky. These new overloads take an optional location and a list of arguments to the diagnostic, and return a LogicalResult. We take the arguments directly and return LogicalResult instead of returning InFlightDiagnostic because we cannot create a valid diagnostic with a null location. This creates an awkward situation where a user may try to treat the, potentially null, diagnostic as a valid one and encounter crashes when attaching notes/etc. Below is an example of how these methods simplify some existing usages:
Before:
if (loc)
emitError(*loc, "this is my diagnostic with argument: ") << 5;
return failure();
After:
return emitOptionalError(loc, "this is my diagnostic with argument: ", 5);
PiperOrigin-RevId: 283853599
In the future, a more configurable malloc and free interface should be used and exposed via
extra parameters to the `createLowerToLLVMPass`. Until requirements are gathered, a simple CL flag allows generating code that runs successfully on hardware that cannot use the stdlib.
PiperOrigin-RevId: 283833424
Adds a ConstantMaskOp to the vector ops dialect.
Adds the following canonicalization patterns:
CreateMaskOp -> ConstantMaskOp
StridedSliceOp(ConstantMaskOp) -> ConstantMaskOp
PiperOrigin-RevId: 283816752
- the name was misleading; this is really checking if a Value being used
to index was loop IV invariant. Update comment.
- the method is only used locally; what can be exposed in the future is
isAccessInvariant(LoadOrStoreOp op, Value *iv)
Signed-off-by: Uday Bondhugula <uday@polymagelabs.com>
Closestensorflow/mlir#285
COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/mlir/pull/285 from bondhugula:quickfix fe5837abe987980c4ab469a9aa7de8e4f0007d9f
PiperOrigin-RevId: 283771923
This CL refactors some of the MLIR vector dependencies to allow decoupling VectorOps, vector analysis, vector transformations and vector conversions from each other.
This makes the system more modular and allows extracting VectorToVector into VectorTransforms that do not depend on vector conversions.
This refactoring exhibited a bunch of cyclic library dependencies that have been cleaned up.
PiperOrigin-RevId: 283660308
This CL also did the following cleanup:
- Moved the test for spv.SubgroupBallotKHR to its own file
- Wrapped generated canonicalization patterns in anonymous namespace
- Updated header comments in SPVOps.td
PiperOrigin-RevId: 283650091
Not all StandardOps can be lowered to SPIR-V. For example, subview op
implementation requires use of pointer bitcasts which is not valid
according to SPIR-V spec (or at least is ambiguous about it). Such ops
need to be removed/transformed before lowering to SPIR-V. The
SPIRVLegalizationPass is added a place where such legalizations can be
added. Current implementation folds the subview ops with load/stores
so that the lowering itself does not have to convert a subview op.
PiperOrigin-RevId: 283642981
The hook has the following form:
* `bool isInvalidated(const AnalysisManager::PreservedAnalyses &)`
Given a preserved analysis set, the analysis returns true if it should truly be
invalidated. This allows for more fine-tuned invalidation in cases where an
analysis wasn't explicitly marked preserved, but may be preserved(or
invalidated) based upon other properties; such as analyses sets.
PiperOrigin-RevId: 283582889
A recent commit introduced the Linkage attribute to the LLVM dialect and used
it in the Global Op. Also use it in LLVMFuncOp. As per LLVM Language Reference,
if the linkage attribute is omitted, the function is assumed to have external
linkage.
PiperOrigin-RevId: 283493299
Existing builders generated by ODS require attributes to be passed
in as mlir::Attribute or its subclasses. This is okay foraggregate-
parameter builders, which is primarily to be used by programmatic
C++ code generation; it is inconvenient for separate-parameter
builders meant to be called in manually written C++ code because
it requires developers to wrap raw values into mlir::Attribute by
themselves.
This CL extends to generate additional builder methods that
take raw values for attributes and handles the wrapping in the
builder implementation. Additionally, if an attribute appears
late in the arguments list and has a default value, the default
value is supplied in the declaration if possible.
PiperOrigin-RevId: 283355919
LLVM IR supports linkage on global objects such as global variables and
functions. Introduce the Linkage attribute into the LLVM dialect, backed by an
integer storage. Use this attribute on LLVM::GlobalOp and make it mandatory.
Implement parsing/printing of the attribute and conversion to LLVM IR.
See tensorflow/mlir#277.
PiperOrigin-RevId: 283309328
* Had leftover call that would result in converting to dictionary attr before
being implicitedly converted back to NamedAttributeList;
* NamedAttributeList is value typed, so don't use const reference;
PiperOrigin-RevId: 283072576
Helper utilies for parsing and printing FunctionLike Ops are only relevant to
the implementation of the Op, not its definition. They depend on
OpImplementation.h and increase the inclusion footprint of FunctionSupport.h,
and do so only to provide some utilities in the "impl" namespace. Move them to
a separate files, similarly to OpDefinition/OpImplementation distinction, and
make only Op implementations use them while keeping headers cleaner. NFC.
PiperOrigin-RevId: 282964556
Updated comments and used static instead of anonymous namspace
to hide functions to be consistent with the existing codebase.
PiperOrigin-RevId: 282847784
Adding zero and multiplying one can be common when generating code
for index calculation.
This CL also sorted canonicalize.mlir to alphabetical order.
PiperOrigin-RevId: 282828055
This CL rewrites the linalg ops to loops transformations as patterns that can be targeted directly from Tablegen. Reliance on OpFolder is removed and to cope with it we introduce local folding patterns that are applied greedily.
PiperOrigin-RevId: 282765550
Since second argument is always fully overwritten and
shape is define in "to" clause, it is not needed.
Also renamed "into" to "to" now that arg is dropped.
PiperOrigin-RevId: 282686475
This method is close to creating an OperationState first and then unpacking it
but avoids creating the OperationState and takes a NamedAttributeList for
attributes rather than array of NamedAttribute (to enable reusing an already
created NamedAttributeList).
Reuse this new method via create that takes OperationState. I'll update inferReturnTypes in follow up to also take NamedAttributeList and so a build method that uses both inferReturnTypes and create can reuse the same list.
PiperOrigin-RevId: 282651642
These changes to SPIR-V lowering while adding support for lowering
SUbViewOp, but are not directly related.
- Change the lowering of MemRefType to
!spv.ptr<!spv.struct<!spv.array<...>[offset]>, ..>
This is consistent with the Vulkan spec.
- To enable testing a simple pattern of lowering functions is added to
ConvertStandardToSPIRVPass. This is just used to convert the type of
the arguments of the function. The added function lowering itself is
not meant to be the way functions are eventually lowered into SPIR-V
dialect.
PiperOrigin-RevId: 282589644
Certain operations can have multiple variadic operands and their size
relationship is not always known statically. For such cases, we need
a per-op-instance specification to divide the operands into logical
groups or segments. This can be modeled by attributes.
This CL introduces C++ trait AttrSizedOperandSegments for operands and
AttrSizedResultSegments for results. The C++ trait just guarantees
such size attribute has the correct type (1D vector) and values
(non-negative), etc. It serves as the basis for ODS sugaring that
with ODS argument declarations we can further verify the number of
elements match the number of ODS-declared operands and we can generate
handy getter methods.
PiperOrigin-RevId: 282467075
This new op is the counterpart of vector.StridedSliceOp and will be used for in the pattern rewrites for vector unrolling.
PiperOrigin-RevId: 282447414
To simplify the lowering into SPIR-V, while still respecting the ABI
requirements of SPIR-V/Vulkan, split the process into two
1) While lowering a function to SPIR-V (when the function is an entry
point function), allow specifying attributes on arguments and
function itself that describe the ABI of the function.
2) Add a pass that materializes the ABI described in the function.
Two attributes are needed.
1) Attribute on arguments of the entry point function that describe
the descriptor_set, binding, storage class, etc, of the
spv.globalVariable this argument will be replaced by
2) Attribute on function that specifies workgroup size, etc. (for now
only workgroup size).
Add the pass -spirv-lower-abi-attrs to materialize the ABI described
by the attributes.
This change makes the SPIRVBasicTypeConverter class unnecessary and is
removed, further simplifying the SPIR-V lowering path.
PiperOrigin-RevId: 282387587
Memref_cast supports cast from static shape to dynamic shape
memrefs. The same should be true for strides as well, i.e a memref
with static strides can be casted to a memref with dynamic strides.
PiperOrigin-RevId: 282381862
This is the counterpart of vector.extractelement op and has the same
limitations at the moment (static I64IntegerArrayAttr to express position).
This restriction will be filterd in the future.
LLVM lowering will be added in a subsequent commit.
PiperOrigin-RevId: 282365760
Introduce a new function-like operation to the GPU dialect to provide a
placeholder for the execution semantic description and to add support for GPU
memory hierarchy. This aligns with the overall goal of the dialect to expose
the common abstraction layer for GPU devices, in particular by providing an
MLIR unit of semantics (i.e. an operation) for memory modeling.
This proposal has been discussed in the mailing list:
https://groups.google.com/a/tensorflow.org/d/msg/mlir/RfXNP7Hklsc/MBNN7KhjAgAJ
As decided, the "convergence" aspect of the execution model will be factored
out into a new discussion and therefore is not included in this commit. This
commit only introduces the operation but does not hook it up with the remaining
flow. The intention is to develop the new flow while keeping the old flow
operational and do the switch in a simple, separately reversible commit.
PiperOrigin-RevId: 282357599
The check in isValidSymbol, as far as a DimOp result went, checked if
the dim op was on a top-level memref. However, any alloc'ed, view, or
subview memref would be fine as long as the corresponding dimension of
that memref is either a static one or was in turn created using a valid
symbol in the case of dynamic dimensions.
Reported-by: Jose Gomez
Signed-off-by: Uday Bondhugula <uday@polymagelabs.com>
Closestensorflow/mlir#252
COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/mlir/pull/252 from bondhugula:symbol 7b57dc394df9375e651f497231c6e4525a32a662
PiperOrigin-RevId: 282097114
Support for including a file multiple times was added in tablegen, removing the need for these extra guards. This is because we already insert c/c++ style header guards within each of the specific .td files.
PiperOrigin-RevId: 282076728
Add a canonicalizer for `spirv::LogicalNotOp`.
Converts:
* spv.LogicalNot(spv.IEqual(...)) -> spv.INotEqual(...)
* spv.LogicalNot(spv.INotEqual(...)) -> spv.IEqual(...)
* spv.LogicalNot(spv.LogicalEqual(...)) -> spv.LogicalNotEqual(...)
* spv.LogicalNot(spv.LogicalNotEqual(...)) -> spv.LogicalEqual(...)
Also moved the test for spv.IMul to arithemtic tests.
Closestensorflow/mlir#256
COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/mlir/pull/256 from denis0x0D:sandbox/canon_logical_not 76ab5787b2c777f948c8978db061d99e76453d44
PiperOrigin-RevId: 282012356
Change vector op names from VectorFooOp to Vector_FooOp and from
vector::VectorFooOp to vector::FooOp.
Closestensorflow/mlir#257
COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/mlir/pull/257 from Kayjukh:master dfc3a0e04114885aaec8740d5951d6984d6e1577
PiperOrigin-RevId: 281967461
This avoids the need to cast back to the derived type when calling get, i.e. removes the need to do DenseIntElementsAttr::get(...).cast<DenseIntElementsAttr>().
PiperOrigin-RevId: 281772163
This will make it easier to scale out test patterns and build specific passes that do not interfere with independent testing.
PiperOrigin-RevId: 281736335
This moves the different canonicalizations of regions into one place and invokes them in the fixed-point iteration of the canonicalizer.
PiperOrigin-RevId: 281617072
This is a simple multi-level DCE pass that operates pretty generically on
the IR. Its key feature compared to the existing peephole dead op folding
that happens during canonicalization is being able to delete recursively
dead cycles of the use-def graph, including block arguments.
PiperOrigin-RevId: 281568202
The current SubViewOp specification allows for either all offsets,
shape and stride to be dynamic or all of them to be static. There are
opportunities for more fine-grained canonicalization based on which of
these are static. For example, if the sizes are static, the result
memref is of static shape. The specification of SubViewOp is modified
to allow on or more of offsets, shapes and strides to be statically
specified. The verification is updated to ensure that the result type
of the subview op is consistent with which of these are static and
which are dynamic.
PiperOrigin-RevId: 281560457
This CL uses the pattern rewrite infrastructure to implement a simple VectorOps -> VectorOps legalization strategy to unroll coarse-grained vector operations into finer grained ones.
The transformation is written using local pattern rewrites to allow composition with other rewrites. It proceeds by iteratively introducing fake cast ops and cleaning canonicalizing or lowering them away where appropriate.
This is an example of writing transformations as compositions of local pattern rewrites that should enable us to make them significantly more declarative.
PiperOrigin-RevId: 281555100
This interface provides more fine-grained hooks into the AsmPrinter than the dialect interface, allowing for operations to define the asm name to use for results directly on the operations themselves. The hook is also expanded to enable defining named result "groups". Get a special name to use when printing the results of this operation.
The given callback is invoked with a specific result value that starts a
result "pack", and the name to give this result pack. To signal that a
result pack should use the default naming scheme, a None can be passed
in instead of the name.
For example, if you have an operation that has four results and you want
to split these into three distinct groups you could do the following:
setNameFn(getResult(0), "first_result");
setNameFn(getResult(1), "middle_results");
setNameFn(getResult(3), ""); // use the default numbering.
This would print the operation as follows:
%first_result, %middle_results:2, %0 = "my.op" ...
PiperOrigin-RevId: 281546873
The `vector.strided_slice` takes an n-D vector, k-D `offsets` integer array attribute, a
k-D `sizes` integer array attribute, a k-D `strides` integer array attribute and extracts
the n-D subvector at the proper offset.
Returns an n-D vector where the first k-D dimensions match the `sizes` attribute.
The returned subvector contains the elements starting at offset `offsets` and ending at
`offsets + sizes`.
Example:
```
%1 = vector.strided_slice %0
{offsets : [0, 2], sizes : [2, 4], strides : [1, 1]}:
vector<4x8x16xf32> // returns a vector<2x4x16xf32>
```
This op will be useful for progressive lowering within the VectorOp dialect.
PiperOrigin-RevId: 281352749
This method is needed for N->1 conversion patterns to retrieve remapped
Values used in the original N operations.
Closestensorflow/mlir#237
COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/mlir/pull/237 from dcaballe:dcaballe/getRemappedValue 1f64fadcf2b203f7b336ff0c5838b116ae3625db
PiperOrigin-RevId: 281321881
The variant that accepts a type will check that the parsed attribute is a valid instance of AttrType. The non-type variant would silently fail in this case, leading to garbage attribute values.
PiperOrigin-RevId: 281136528
Convert chained `spirv::BitcastOp` operations into
one `spirv::BitcastOp` operation.
Closestensorflow/mlir#238
COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/mlir/pull/238 from denis0x0D:sandbox/canon_bitcast 4352ed4f81b959ec92f849c599e733b62a99c010
PiperOrigin-RevId: 281129234
This improves consistency and will concretely avoid collisions between VectorExtractElementOp and ExtractElementOp when they are included in the same transforms / rewrites.
PiperOrigin-RevId: 281101588
This CL added op definitions for a few bit operations:
* OpBitFieldInsert
* OpBitFieldSExtract
* OpBitFieldUExtract
Closestensorflow/mlir#233
COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/mlir/pull/233 from denis0x0D:sandbox/bit_field_ops e7fd85b00d72d483d7992dc42b9cc4d673903455
PiperOrigin-RevId: 280691816
In essence, std.subview is just an abstract indexing transformation (somewhat
akin to a gep in llvm) and by itself has no effect. From a practical perspective
this helps, as it allows to remove dead subview operations.
PiperOrigin-RevId: 280630046
This is step 1/n in refactoring infrastructure along the Vector dialect to make it ready for retargetability and composable progressive lowering.
PiperOrigin-RevId: 280529784
Refactoring the conversion from StandardOps/GPU dialect to SPIR-V
dialect:
1) Move the SPIRVTypeConversion and SPIRVOpLowering class into SPIR-V
dialect.
2) Add header files that expose functions to add patterns for the
dialects to SPIR-V lowering, as well as a pass that does the
dialect to SPIR-V lowering.
3) Make SPIRVOpLowering derive from OpLowering class.
PiperOrigin-RevId: 280486871
The `Operator` class keeps an `arguments` field, which contains pointers
to `operands` and `attributes` elements. Thus it must be populated after
`operands` and `attributes` are finalized so to have stable pointers.
SmallVector may re-allocate when still having new elements added, which
will invalidate pointers.
PiperOrigin-RevId: 280466896
Previous commits removed all uses of LLVMTypeConverter::k*PosInMemRefDescriptor
outside of the MemRefDescriptor class. These numbers are an implementation
detail and can be hidden under a layer of more semantic APIs.
PiperOrigin-RevId: 280442444
Following up on the consolidation of MemRef descriptor conversion, update
Vector-to-LLVM conversion to use the helper class that abstracts away the
implementation details of the MemRef descriptor. This also makes the types of
the attributes in emitted llvm.insert/extractelement operations consistently
i64 instead of a mix of index and i64.
PiperOrigin-RevId: 280441451
This CL moves VectorOps to Tablegen and cleans up the implementation.
This is almost NFC but 2 changes occur:
1. an interface change occurs in the padding value specification in vector_transfer_read:
the value becomes non-optional. As a shortcut we currently use %f0 for all paddings.
This should become an OpInterface for vectorization in the future.
2. the return type of vector.type_cast is trivial and simplified to `memref<vector<...>>`
Relevant roundtrip and invalid tests that used to sit in core are moved to the vector dialect.
The op documentation is moved to the .td file.
PiperOrigin-RevId: 280430869
Following up on the consolidation of MemRef descriptor conversion, update
Linalg-to-LLVM conversion to use the helper class that abstracts away the
implementation details of the MemRef descriptor. This required MemRefDescriptor
to become publicly visible. Since this conversion is heavily EDSC-based,
introduce locally an additional wrapper that uses builder and location pointed
to by the EDSC context while emitting descriptor manipulation operations.
PiperOrigin-RevId: 280429228
This CL uses the now standard std.subview in linalg.
Two shortcuts are currently taken to allow this port:
1. the type resulting from a view is currently degraded to fully dynamic to pass the SubViewOp verifier.
2. indexing into SubViewOp may access out of bounds since lowering to LLVM does not currently enforce it by construction.
These will be fixed in subsequent commits after discussions.
PiperOrigin-RevId: 280250129
This is a quite complex operation that users are likely to attempt to write
themselves and get wrong (citation: users=me).
Ideally, we could pull this into FunctionLike, but for now, the
FunctionType rewriting makes it FuncOp specific. We would need some hook
for rewriting the function type (which for LLVM's func op, would need to
rewrite the underlying LLVM type).
PiperOrigin-RevId: 280234164
This refactors the implementation of block signature(type) conversion to not insert fake cast operations to perform the type conversion, but to instead create a new block containing the proper signature. This has the benefit of enabling the use of pre-computed analyses that rely on mapping values. It also leads to a much cleaner implementation overall. The major user facing change is that applySignatureConversion will now replace the entry block of the region, meaning that blocks generally shouldn't be cached over calls to applySignatureConversion.
PiperOrigin-RevId: 280226936
The current implementation silently fails if the '@' identifier isn't present, making it similar to the 'optional' parse methods. This change renames the current implementation to 'Optional' and adds a new 'parseSymbolName' that emits an error.
PiperOrigin-RevId: 280214610
During deserialization, the loop header block will be moved into the
spv.loop's region. If the loop header block has block arguments,
we need to make sure it is correctly carried over to the block where
the new spv.loop resides.
During serialization, we need to make sure block arguments from the
spv.loop's entry block are not silently dropped.
PiperOrigin-RevId: 280021777
This causes the AsmPrinter to use a local value numbering when printing the IR, allowing for the printer to be used safely in a local context, e.g. to ensure thread-safety when printing the IR. This means that the IR printing instrumentation can also be used during multi-threading when module-scope is disabled. Operation::dump and DiagnosticArgument(Operation*) are also updated to always print local scope, as this is the most common use case when debugging.
PiperOrigin-RevId: 279988203
This CL adds an extra pointer to the memref descriptor to allow specifying alignment.
In a previous implementation, we used 2 types: `linalg.buffer` and `view` where the buffer type was the unit of allocation/deallocation/alignment and `view` was the unit of indexing.
After multiple discussions it was decided to use a single type, which conflates both, so the memref descriptor now needs to carry both pointers.
This is consistent with the [RFC-Proposed Changes to MemRef and Tensor MLIR Types](https://groups.google.com/a/tensorflow.org/forum/#!searchin/mlir/std.view%7Csort:date/mlir/-wKHANzDNTg/4K6nUAp8AAAJ).
PiperOrigin-RevId: 279959463
This change allows for adding additional nested references to a SymbolRefAttr to allow for further resolving a symbol if that symbol also defines a SymbolTable. If a referenced symbol also defines a symbol table, a nested reference can be used to refer to a symbol within that table. Nested references are printed after the main reference in the following form:
symbol-ref-attribute ::= symbol-ref-id (`::` symbol-ref-id)*
Example:
module @reference {
func @nested_reference()
}
my_reference_op @reference::@nested_reference
Given that SymbolRefAttr is now more general, the existing functionality centered around a single reference is moved to a derived class FlatSymbolRefAttr. Followup commits will add support to lookups, rauw, etc. for scoped references.
PiperOrigin-RevId: 279860501
This operation is a companion operation to the std.view operation added as proposed in "Updates to the MLIR MemRefType" RFC.
PiperOrigin-RevId: 279766410
This code should be exercised using the existing kernel outlining unit test, but
let me know if I should add a dedicated unit test using a fake call instruction
as well.
PiperOrigin-RevId: 279436321
This CL added op definitions for a few bit operations:
* OpShiftLeftLogical
* OpShiftRightArithmetic
* OpShiftRightLogical
* OpBitCount
* OpBitReverse
* OpNot
Also moved the definition of spv.BitwiseAnd to follow the
lexicographical order.
Closestensorflow/mlir#215
COPYBARA_INTEGRATE_REVIEW=https://github.com/tensorflow/mlir/pull/215 from denis0x0D:sandbox/bit_ops d9b0852b689ac6c4879a9740b1740a2357f44d24
PiperOrigin-RevId: 279350470
MLIR translation tools can emit diagnostics and we want to be able to check if
it is indeed the case in tests. Reuse the source manager error handlers
provided for mlir-opt to support the verification in mlir-translate. This
requires us to change the signature of the functions that are registered to
translate sources to MLIR: it now takes a source manager instead of a memory
buffer.
PiperOrigin-RevId: 279132972
A return type that differs from the inferred return type need not indicate that an operation is invalid (e.g., tensor<*xf32> vs tensor<10xf32>) but they should be compatible for the operation to be considered valid. Add method to query if inferred type is compatible with return type.
Also add InferTypeOpIntefaceDefault trait that considers equality and compatibility as the same. Currently an op has to opt in to using it explicitly.
PiperOrigin-RevId: 279085639
Now that a view op has graduated to the std dialect, we can update Linalg to use it and remove ops that have become obsolete. As a byproduct, the linalg buffer and associated ops can also disappear.
PiperOrigin-RevId: 279073591
Alexander Belyaev