'acc' Dialect

The acc dialect is an MLIR dialect for representing the OpenACC programming model. OpenACC is a standardized directive-based model which is used with C, C++, and Fortran to enable programmers to expose parallelism in their code. The descriptive approach used by OpenACC allows targeting of parallel multicore and accelerator targets like GPUs by giving the compiler the freedom of how to parallelize for specific architectures. OpenACC also provides the ability to optimize the parallelism through increasingly more prescriptive clauses.

This dialect models the constructs from the OpenACC 3.3 specification

This document describes the design of the OpenACC dialect in MLIR. It lists and explains design goals and design choices along with their rationale. It also describes specifics with regards to acc dialect operations, types, and attributes.

Dialect Design Goals ¶

• Needs to have complete representation of the OpenACC language.
  • A frontend requires this in order to properly generate a representation of possible acc pragmas in MLIR. Additionally, this dialect is expected to be further lowered when materializing its semantics. Without a complete representation, a frontend might choose a lower abstraction (such as direct runtime call) - but this would impact the ability to do analysis and optimizations on the dialect.
• Allow representation at the same semantic level as the OpenACC language while having capability to represent nuances of the source language semantics (such as Fortran descriptors) in an agnostic manner.
  • Using abstractions that closely model the OpenACC language simplifies frontend implementation. It also allows for easier debugging of the IR. However, sometimes source language specific behavior is needed when materializing OpenACC. In these cases, such as privatization of C++ objects with default constructor, the frontend fills in the recipe along with the private operation which can be packaged neatly with the acc dialect operations.
• Be able to regenerate the semantic equivalent of the user pragmas from the dialect (including bounds, names, clauses, modifiers, etc).
  • This is a strong measure of making sure that the dialect is not lossy in semantics. It also allows capability to generate appropriate and useful debug information outside of the frontend.
• Be dialect agnostic so that it can be used and coexist with other dialects including but not limited to hlfir, fir, llvm, cir.
  • Directive-based models such as OpenACC are always used with a source language, so the acc dialect coexisting with other dialect(s) is necessary by construction. Through proper abstractions, neither the acc dialect nor the source language dialect should have dependencies on each other; where needed, interfaces should be used to ensure acc dialect can verify expected properties.
• The dialect must allow dataflow to be modeled accurately and performantly using MLIR’s existing facilities.
  • Appropriate dataflow modeling is important for analyses and IR reasoning - even something as simple as walking the uses. Therefore operations, like data operations, are expected to generate results which can be used in modeling behavior. For example, consider an acc copyin clause. After the acc.copyin operation, a pointer which lives on devices should be distinguishable from one that lives in host memory.
• Be friendly to MLIR optimization passes by implementing common interfaces.
  • Interfaces, such as MemoryEffects, are the key way MLIR transformations and analyses are designed to interact with the IR. In order for the operations in the acc dialect to be optimizable (either directly or even indirectly by not blocking optimizations of nested IR), implementing relevant common interfaces is needed.

The design philosophy of the acc dialect is one where the design goals are adhered to. Current and planned operations, attributes, types must adhere to the design goals.

Operation Categories ¶

The OpenACC dialect includes both high-level operations (which retain the same semantic meaning as their OpenACC language equivalent), intermediate-level operations (which are used to decompose clauses from constructs), and low-level operations (to encode specifics associated with source language in a generic way).

The high-level operations list contains the following OpenACC language constructs and their corresponding operations:

• acc parallel → acc.parallel
• acc kernels → acc.kernels
• acc serial → acc.serial
• acc data → acc.data
• acc loop → acc.loop
• acc enter data → acc.enter_data
• acc exit data → acc.exit_data
• acc host_data → acc.host_data
• acc init → acc.init
• acc shutdown → acc.shutdown
• acc update → acc.update
• acc set → acc.set
• acc wait → acc.wait
• acc atomic read → acc.atomic.read
• acc atomic write → acc.atomic.write
• acc atomic update → acc.atomic.update
• acc atomic capture → acc.atomic.capture

This second group contains operations which are used to represent either decomposed constructs or clauses for more accurate modeling:

• acc routine → acc.routine + acc.routine_info attribute
• acc declare → acc.declare_enter + acc.declare_exit or acc.declare
• acc {construct} copyin → acc.copyin (before region) + acc.delete (after region)
• acc {construct} copy → acc.copyin (before region) + acc.copyout (after region)
• acc {construct} copyout → acc.create (before region) + acc.copyout (after region)
• acc {construct} attach → acc.attach (before region) + acc.detach (after region)
• acc {construct} create → acc.create (before region) + acc.delete (after region)
• acc {construct} present → acc.present (before region) + acc.delete (after region)
• acc {construct} no_create → acc.nocreate (before region) + acc.delete (after region)
• acc {construct} deviceptr → acc.deviceptr
• acc {construct} private → acc.private
• acc {construct} firstprivate → acc.firstprivate
• acc {construct} reduction → acc.reduction
• acc cache → acc.cache
• acc update device → acc.update_device
• acc update host → acc.update_host
• acc host_data use_device → acc.use_device
• acc declare device_resident → acc.declare_device_resident
• acc declare link → acc.declare_link
• acc exit data delete → acc.delete (with structured flag as false)
• acc exit data detach → acc.detach (with structured flag as false)
• acc {construct} {data_clause}(var[lb:ub]) → acc.bounds

The low-level operations are:

• acc.private.recipe
• acc.reduction.recipe
• acc.firstprivate.recipe
• acc.global_ctor
• acc.global_dtor
• acc.yield
• acc.terminator The low-level operations semantics and reasoning are further explained in sections below.

Data Operations ¶

Data Clause Decomposition ¶

The data clauses are decomposed from their constructs for better dataflow modeling in MLIR. There are multiple reasons for this which are consistent with the dialect goals:

• Correctly represents dataflow. Data clauses have different effects at entry to region and at exit from region.
• Friendlier to add attributes such as MemoryEffects to a single operation. This can better reflect semantics (like the fact that an acc.copyin operation only reads host memory)
• Operations can be moved or optimized individually (eg CSE).
• Easier to keep track of debug information. Line location can point to the text representing the data clause instead of the construct. Additionally, attributes can be used to keep track of variable names in clauses without having to walk the IR tree in attempt to recover the information (this makes acc dialect more agnostic with regards to what other dialect it is used with).
• Clear operation ordering since all data operations are on same list.

Each of the acc dialect data operations represents either the entry or the exit portion of the data action specification. Thus, acc.copyin represents the semantics defined in section 2.7.7 copyin clause whose wording starts with At entry to a region. The decomposed exit operation acc.delete represents the second part of that section, whose wording starts with At exit from the region. The delete action may be performed after checking and updating of the relevant reference counters noted.

The acc data operations, even when decomposed, retain their original data clause in an operation operand dataClause for possibility to recover this information during debugging. For example, acc copy, does not translate to acc.copy operation, but instead to acc.copyin for entry and acc.copyout for exit. Both the decomposed operations hold a dataClause field that specifies this was an acc copy.

The link between the decomposed entry and exit operations is the ssa value produced by the entry operation. Namely, it is the accPtr result which is used both in the dataOperands of the operation used for the construct and in the accPtr operand of the exit operation.

Bounds ¶

OpenACC data clauses allow the use of bounds specifiers as per 2.7.1 Data Specification in Data Clauses. However, array dimensions for the data are not always required in the clause if the source language’s type system captures this information - the user can just specify the variable name in the data clause. So the acc.bounds operation is an important piece to ensure uniform representation of both explicit user set dimensions and implicit type-based dimensions. It contains several key features to allow properly encoding sizes in a manner flexible and agnostic to the source language’s dialect:

• Multi-dimensional arrays can be represented by using multiple ordered acc.bounds operations.
• Bounds are required to be zero-normalized. This works well with the PointerLikeType requirement in data clauses - since a lowerbound of 0 means looking at data at the zero offset from pointer. This requirement also works well in ensuring the acc dialect is agnostic to source language dialect since it prevents ambiguity such as the case of Fortran arrays where the lower bound is not a fixed value.
• If the source dialect does not encode the dimensions in the type (eg !fir.array<?x?xi32>) but instead encodes it in some other way (such as through descriptors), then the frontend must fill in the acc.bounds operands with appropriate information (such as loads from descriptor). The acc.bounds operation also permits lossy source dialect, such as if the frontend uses aggressive pointer decay and cannot represent the dimensions in the type system (eg using !llvm.ptr for arrays). Both of these aspects show acc.bounds’ operation’s flexibility to allow the representation to be agnostic since the acc dialect is not expected to be able to understand how to extract dimension information from the types of the source dialect.
• The OpenACC specification allows either extent or upperbound in the data clause depending on whether it is Fortran or C and C++. The acc.bounds operation is rich enough to accept either or both - for convenience in lowering to the dialect and for ability to precisely capture the meaning from the clause.
• The stride, either in units or bytes, can be also captured in the acc.bounds operation. This is also an important part to be able to accept a source language’s arrays without forcing the frontend to normalize them in some way. For example, consider a case where in a parent function, a whole array is mapped to device. Then only a view of a non-1 stride is passed to child function (eg Fortran array slice with non-1 stride). A copy operation of this data in child should be able to avoid remapping this array. If instead the operation required normalizing the array (such as making it contiguous), then unexpected disjoint mapping of the same host data would be error-prone since it would result in multiple mappings to device.

Counters ¶

The data operations also maintain semantics described in the OpenACC specification related to runtime counters. More specifically, consider the specification of the entry portion of acc copyin in section 2.7.7:

At entry to a region, the structured reference counter is used. On an
enter data directive, the dynamic reference counter is used.
- If var is present and is not a null pointer, a present increment
action with the appropriate reference counter is performed.
- If var is not present, a copyin action with the appropriate reference
counter is performed.
- If var is a pointer reference, an attach action is performed.

The acc.copyin operation includes these semantics, including those related to attach, which is specified through the varPtrPtr operand. The structured flag on the operation is important since the structured reference counter should be used when the flag is true; and the dynamic reference counter should be used when it is false.

At exit from structured regions (acc data, acc kernels), the acc copyin operation is decomposed to acc.delete (with the structured flag as true). The semantics of the acc.delete are also consistent with the OpenACC specification noted for the exit portion of the acc copyin clause:

At exit from the region:
- If the structured reference counter for var is zero, no action is
taken.
- Otherwise, a detach action is performed if var is a pointer reference,
and a present decrement action with the structured reference counter is
performed if var is not a null pointer. If both structured and dynamic
reference counters are zero, a delete action is performed.

Types ¶

There are a few acc dialect type categories to describe:

• type of acc data clause operation input varPtr
  • The type of varPtr must be pointer-like. This is done by attaching the PointerLikeType interface to the appropriate MLIR type. Although memory/storage concept is a lower level abstraction, it is useful because the OpenACC model distinguishes between host and device memory explicitly - and the mapping between the two is done through pointers. Thus, by explicitly requiring it in the dialect, the appropriate language frontend must create storage or use type that satisfies the mapping constraint.
• type of result of acc data clause operations
  • The type of the acc data clause operation is exactly the same as varPtr. This was done intentionally instead of introducing an acc.ref/ptr type so that IR compatibility and the dialect’s existing strong type checking can be maintained. This is needed since the acc dialect must live within another dialect whose type system is unknown to it. The only constraint is that the appropriate dialect type must use the PointerLikeType interface.
• type of decomposed clauses
  • Decomposed clauses, such as acc.bounds and acc.declare_enter produce types to allow their results to be used only in specific operations.

Recipes ¶

Recipes are a generic way to express source language specific semantics.

There are currently two categories of recipes, but the recipe concept can be extended for any additional low-level information that needs to be captured for successful lowering of OpenACC. The two categories are:

• recipes used in the context of privatization associated with a construct
• recipes used in the context of additional specification of data semantics

The intention of the recipes is to specify how materialization of action, such as privatization, should be done when the semantics of the action needs interpreted and lowered, such as before generating LLVM dialect.

The recipes used for privatization provide a source-language independent way of specifying the creation of a local variable of that type. This means using the appropriate alloca instruction and being able to specify default initialization or default constructor.

Routine ¶

The routine directive is used to note that a procedure should be made available for the accelerator in a way that is consistent with its modifiers, such as those that describe the parallelism. In the acc dialect, an acc routine is represented through two joint pieces - an attribute and an operation:

• The acc.routine operation is simply a specifier which notes which symbol (or string) the acc routine is needed for, along with parallelism associated. This defines a symbol that can be referenced in attribute.
• The acc.routine_info attribute is an attribute used on the source dialect specific operation which specifies one or multiple acc.routine symbols. Typically, this is attached to func.func which either provides the declaration (in case of externals) or provides the actual body of the acc routine in the dialect that the source language was translated to.

Declare ¶

OpenACC declare is a mechanism which declares a definition of a global or a local to be accessible to accelerator with an implicit lifetime as that of the scope where it was declared in. Thus, declare semantics are represented through multiple operations and attributes:

• acc.declare - This is a structured operation which contains an MLIR region and can be used in similar manner as acc.data to specify an implicit data region with specific procedure lifetime. This is typically used inside func.func after variable declarations.
• acc.declare_enter - This is an unstructured operation which is used as a decomposed form of acc declare. It effectively allows the entry operation to exist in a scope different than the exit operation. It can also be used along acc.declare_exit which consumes its token to define a scoped region without using MLIR region. This operation is also used in acc.global_ctor.
• acc.declare_exit - The matching equivalent of acc.declare_enter except that it specifies exit semantics. This operation is typically used inside a func.func at the exit points or with acc.global_dtor.
• acc.global_ctor - Lives at the same level as source dialect globals and is used to specify data actions to be done at program entry. This is used in conjunction with source dialect globals whose lifetime is not just a single procedure.
• acc.global_dtor - Defines the exit data actions that should be done at program exit. Typically used to revert the actions of acc.global_ctor.

The attributes:

• acc.declare - This is a facility for easier determination of variables which are acc declare’d. This attribute is used on operations producing globals and on operations producing locals such as dialect specific alloca’s. Having this attribute is required in order to appear in a data mapping operation associated with any of the acc.declare* operations.
• acc.declare_action - Since the OpenACC specification allows declaration of variables that have yet to be allocated, this attribute is used at the allocation and deallocation points. More specifically, this attribute captures symbols of functions to be called to perform an action either pre-allocate, post-allocate, pre-deallocate, or post-deallocate. Calls to these functions should be materialized when lowering OpenACC semantics to ensure proper data actions are done after the allocation/deallocation.

OpenACC Transforms and Analyses ¶

The design goal for the acc dialect is to be friendly to MLIR optimization passes including CSE and LICM. Additionally, since it is designed to recover original clauses, it makes late verification and analysis possible in the MLIR framework outside of the frontend.

This section describes a few MLIR-level passes for which the acc dialect design should be friendly for. This section is currently solely outlining the possibilities intended by the design and not necessarily existing passes.

Verification ¶

Since the OpenACC dialect is not lossy with regards to its representation, it is possible to do OpenACC language semantic checking at the MLIR-level. What follows is a list of various semantic checks needed.

This first list is required to be done in the frontend because the acc dialect operations must be valid when constructed:

• Ensure that only listed clauses are allowed for each directive.
• Ensure that only listed modifiers are allowed for each clause.

However, the following are semantic checks that can be done at the MLIR-level (either in a separate pass or as part of the operation verifier):

• Specify the validity checks that each modifier needs. (eg num_gangs may need a positive integer).
• Ensure valid clause nesting.
• Validate clause restrictions which cannot appear with others.
• Validate that no conflicting clauses are used on variables.

Note that some of these checks can be even more precise when done at the MLIR level because optimizations like inlining and constant propagation expose detail that wouldn’t have been visible in the frontend.

Implicit Data Attributes ¶

The OpenACC specification includes a section on 2.6.2 Variables with Implicitly Determined Data Attributes. What this section describes are the data actions that should be applied to a variable for which user did not specify a data action for. The action depends on the construct being used and also on the default clause. However, the point to note here is that variables which are live-in into the acc region must employ some data mapping so the data can be passed to accelerator.

One possible optimizations that affects data attributes needed is Scalar Replacement of Aggregates (SROA). The acc dialect should not prevent this from happening on the source dialect.

Because it is intended to be possible to apply optimizations across an acc region, the analysis/transformation pass that applies the implicit data attributes should be run as late as possible - ideally right before any outlining process which uses the acc region body to create an accelerator procedure. It is expected that existing MLIR facilities, such as mlir::Liveness will work for the acc region and thus can be used to perform this analysis.

Redundant Clause Elimination ¶

The data operations are modeled in a way where data entry operations look like loads and data exit operations look like stores. Thus these operations are intended to be optimized in the following ways:

• Be able to eliminate redundant operations such as when an acc.copyin dominates another.
• Be able to hoist/sink such operations out of loops.

Operations TOC ¶

Operations ¶

source

acc.atomic.capture (acc::AtomicCaptureOp) ¶

Performs an atomic capture

Syntax:

operation ::= `acc.atomic.capture` $region attr-dict

This operation performs an atomic capture.

The region has the following allowed forms:

  acc.atomic.capture {
    acc.atomic.update ...
    acc.atomic.read ...
    acc.terminator
  }

  acc.atomic.capture {
    acc.atomic.read ...
    acc.atomic.update ...
    acc.terminator
  }

  acc.atomic.capture {
    acc.atomic.read ...
    acc.atomic.write ...
    acc.terminator
  }

Traits: RecursiveMemoryEffects, SingleBlockImplicitTerminator<TerminatorOp>, SingleBlock

Interfaces: AtomicCaptureOpInterface

acc.atomic.read (acc::AtomicReadOp) ¶

Performs an atomic read

Syntax:

operation ::= `acc.atomic.read` $v `=` $x
              `:` type($v) `,` type($x) `,` $element_type attr-dict

This operation performs an atomic read.

The operand x is the address from where the value is atomically read. The operand v is the address where the value is stored after reading.

Interfaces: AtomicReadOpInterface

Attributes: ¶

Operands: ¶

acc.atomic.update (acc::AtomicUpdateOp) ¶

Performs an atomic update

Syntax:

operation ::= `acc.atomic.update` $x `:` type($x) $region attr-dict

This operation performs an atomic update.

The operand x is exactly the same as the operand x in the OpenACC Standard (OpenACC 3.3, section 2.12). It is the address of the variable that is being updated. x is atomically read/written.

The region describes how to update the value of x. It takes the value at x as an input and must yield the updated value. Only the update to x is atomic. Generally the region must have only one instruction, but can potentially have more than one instructions too. The update is sematically similar to a compare-exchange loop based atomic update.

The syntax of atomic update operation is different from atomic read and atomic write operations. This is because only the host dialect knows how to appropriately update a value. For example, while generating LLVM IR, if there are no special atomicrmw instructions for the operation-type combination in atomic update, a compare-exchange loop is generated, where the core update operation is directly translated like regular operations by the host dialect. The front-end must handle semantic checks for allowed operations.

Traits: RecursiveMemoryEffects, SingleBlockImplicitTerminator<YieldOp>, SingleBlock

Interfaces: AtomicUpdateOpInterface

Operands: ¶

acc.atomic.write (acc::AtomicWriteOp) ¶

Performs an atomic write

Syntax:

operation ::= `acc.atomic.write` $x `=` $expr
              `:` type($x) `,` type($expr)
              attr-dict

This operation performs an atomic write.

The operand x is the address to where the expr is atomically written w.r.t. multiple threads. The evaluation of expr need not be atomic w.r.t. the write to address. In general, the type(x) must dereference to type(expr).

Interfaces: AtomicWriteOpInterface

Operands: ¶

acc.attach (acc::AttachOp) ¶

Represents acc attach semantics which updates a pointer in device memory with the corresponding device address of the pointee.

Syntax:

operation ::= `acc.attach` `varPtr` `(` $varPtr `:` custom<VarPtrType>(type($varPtr), $varType)
              oilist(
              `varPtrPtr` `(` $varPtrPtr `:` type($varPtrPtr) `)`
              | `bounds` `(` $bounds `)`
              | `async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType) `)`
              ) `->` type($accPtr) attr-dict

Description of arguments:

• varPtr: The address of variable to copy.
• varPtrPtr: Specifies the address of varPtr - only used when the variable copied is a field in a struct. This is important for OpenACC due to implicit attach semantics on data clauses (2.6.4).
• bounds: Used when copying just slice of array or array’s bounds are not encoded in type. They are in rank order where rank 0 is inner-most dimension.
• asyncOperands and asyncOperandsDeviceType: pair-wise lists of the async clause values associated with device_type’s.
• asyncOnly: a list of device_type’s for which async clause does not specify a value (default is acc_async_noval - OpenACC 3.3 2.16.1).
• dataClause: Keeps track of the data clause the user used. This is because the acc operations are decomposed. So a ‘copy’ clause is decomposed to both acc.copyin and acc.copyout operations, but both have dataClause that specifies acc_copy in this field.
• structured: Flag to note whether this is associated with structured region (parallel, kernels, data) or unstructured (enter data, exit data). This is important due to spec specifically calling out structured and dynamic reference counters (2.6.7).
• implicit: Whether this is an implicitly generated operation, such as copies done to satisfy “Variables with Implicitly Determined Data Attributes” in 2.6.2.
• name: Holds the name of variable as specified in user clause (including bounds).

The async values attached to the data entry operation imply that the data action applies to all device types specified by the device_type clauses using the activity queues on these devices as defined by the async values.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}, MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::RuntimeCounters, MemoryEffects::Write on ::mlir::acc::RuntimeCounters}

Attributes: ¶

Operands: ¶

Results: ¶

acc.bounds (acc::DataBoundsOp) ¶

Represents normalized bounds information for acc data clause.

Syntax:

operation ::= `acc.bounds` oilist(
              `lowerbound` `(` $lowerbound `:` type($lowerbound) `)`
              | `upperbound` `(` $upperbound `:` type($upperbound) `)`
              | `extent` `(` $extent `:` type($extent) `)`
              | `stride` `(` $stride `:` type($stride) `)`
              | `startIdx` `(` $startIdx `:` type($startIdx) `)`
              ) attr-dict

This operation is used to record bounds used in acc data clause in a normalized fashion (zero-based). This works well with the PointerLikeType requirement in data clauses - since a lowerbound of 0 means looking at data at the zero offset from pointer.

The operation must have an upperbound or extent (or both are allowed - but not checked for consistency). When the source language’s arrays are not zero-based, the startIdx must specify the zero-position index.

Examples below show copying a slice of 10-element array except first element. Note that the examples use extent in data clause for C++ and upperbound for Fortran (as per 2.7.1). To simplify examples, the constants are used directly in the acc.bounds operands - this is not the syntax of operation.

C++:

int array[10];
#pragma acc copy(array[1:9])

=>

acc.bounds lb(1) ub(9) extent(9) startIdx(0)

Fortran:

integer :: array(1:10)
!$acc copy(array(2:10))

=>

acc.bounds lb(1) ub(9) extent(9) startIdx(1)

Traits: AttrSizedOperandSegments

Interfaces: NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

Attributes: ¶

Operands: ¶

Results: ¶

acc.cache (acc::CacheOp) ¶

Represents the cache directive that is associated with a loop.

Syntax:

operation ::= `acc.cache` `varPtr` `(` $varPtr `:` custom<VarPtrType>(type($varPtr), $varType)
              oilist(
              `varPtrPtr` `(` $varPtrPtr `:` type($varPtrPtr) `)`
              | `bounds` `(` $bounds `)`
              | `async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType) `)`
              ) `->` type($accPtr) attr-dict

Description of arguments:

• varPtr: The address of variable to copy.
• varPtrPtr: Specifies the address of varPtr - only used when the variable copied is a field in a struct. This is important for OpenACC due to implicit attach semantics on data clauses (2.6.4).
• bounds: Used when copying just slice of array or array’s bounds are not encoded in type. They are in rank order where rank 0 is inner-most dimension.
• asyncOperands and asyncOperandsDeviceType: pair-wise lists of the async clause values associated with device_type’s.
• asyncOnly: a list of device_type’s for which async clause does not specify a value (default is acc_async_noval - OpenACC 3.3 2.16.1).
• dataClause: Keeps track of the data clause the user used. This is because the acc operations are decomposed. So a ‘copy’ clause is decomposed to both acc.copyin and acc.copyout operations, but both have dataClause that specifies acc_copy in this field.
• structured: Flag to note whether this is associated with structured region (parallel, kernels, data) or unstructured (enter data, exit data). This is important due to spec specifically calling out structured and dynamic reference counters (2.6.7).
• implicit: Whether this is an implicitly generated operation, such as copies done to satisfy “Variables with Implicitly Determined Data Attributes” in 2.6.2.
• name: Holds the name of variable as specified in user clause (including bounds).

The async values attached to the data entry operation imply that the data action applies to all device types specified by the device_type clauses using the activity queues on these devices as defined by the async values.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface), NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}, MemoryEffects::Effect{}

Attributes: ¶

Operands: ¶

Results: ¶

acc.copyin (acc::CopyinOp) ¶

Represents copyin semantics for acc data clauses like acc copyin and acc copy.

Syntax:

operation ::= `acc.copyin` `varPtr` `(` $varPtr `:` custom<VarPtrType>(type($varPtr), $varType)
              oilist(
              `varPtrPtr` `(` $varPtrPtr `:` type($varPtrPtr) `)`
              | `bounds` `(` $bounds `)`
              | `async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType) `)`
              ) `->` type($accPtr) attr-dict

Description of arguments:

• varPtr: The address of variable to copy.
• varPtrPtr: Specifies the address of varPtr - only used when the variable copied is a field in a struct. This is important for OpenACC due to implicit attach semantics on data clauses (2.6.4).
• bounds: Used when copying just slice of array or array’s bounds are not encoded in type. They are in rank order where rank 0 is inner-most dimension.
• asyncOperands and asyncOperandsDeviceType: pair-wise lists of the async clause values associated with device_type’s.
• asyncOnly: a list of device_type’s for which async clause does not specify a value (default is acc_async_noval - OpenACC 3.3 2.16.1).
• dataClause: Keeps track of the data clause the user used. This is because the acc operations are decomposed. So a ‘copy’ clause is decomposed to both acc.copyin and acc.copyout operations, but both have dataClause that specifies acc_copy in this field.
• structured: Flag to note whether this is associated with structured region (parallel, kernels, data) or unstructured (enter data, exit data). This is important due to spec specifically calling out structured and dynamic reference counters (2.6.7).
• implicit: Whether this is an implicitly generated operation, such as copies done to satisfy “Variables with Implicitly Determined Data Attributes” in 2.6.2.
• name: Holds the name of variable as specified in user clause (including bounds).

The async values attached to the data entry operation imply that the data action applies to all device types specified by the device_type clauses using the activity queues on these devices as defined by the async values.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}, MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::RuntimeCounters, MemoryEffects::Write on ::mlir::acc::RuntimeCounters}

Attributes: ¶

Operands: ¶

Results: ¶

acc.copyout (acc::CopyoutOp) ¶

Represents acc copyout semantics - reverse of copyin.

Syntax:

operation ::= `acc.copyout` `accPtr` `(` $accPtr `:` type($accPtr) `)`
              (`bounds` `(` $bounds^ `)` )?
              (`async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType)^ `)`)?
              `to` `varPtr` `(` $varPtr `:` custom<VarPtrType>(type($varPtr), $varType)
              attr-dict

• varPtr: The address of variable to copy back to.

  • accPtr: The acc address of variable. This is the link from the data-entry operation used.
  • bounds: Used when copying just slice of array or array’s bounds are not encoded in type. They are in rank order where rank 0 is inner-most dimension.
  • asyncOperands and asyncOperandsDeviceType: pair-wise lists of the async clause values associated with device_type’s.
  • asyncOnly: a list of device_type’s for which async clause does not specify a value (default is acc_async_noval - OpenACC 3.3 2.16.1).
  • dataClause: Keeps track of the data clause the user used. This is because the acc operations are decomposed. So a ‘copy’ clause is decomposed to both acc.copyin and acc.copyout operations, but both have dataClause that specifies acc_copy in this field.
  • structured: Flag to note whether this is associated with structured region (parallel, kernels, data) or unstructured (enter data, exit data). This is important due to spec specifically calling out structured and dynamic reference counters (2.6.7).
  • implicit: Whether this is an implicitly generated operation, such as copies done to satisfy “Variables with Implicitly Determined Data Attributes” in 2.6.2.
  • name: Holds the name of variable as specified in user clause (including bounds).

  The async values attached to the data exit operation imply that the data action applies to all device types specified by the device_type clauses using the activity queues on these devices as defined by the async values.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}, MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::RuntimeCounters, MemoryEffects::Write on ::mlir::acc::RuntimeCounters}

Attributes: ¶

Operands: ¶

acc.create (acc::CreateOp) ¶

Represents create semantics for acc data clauses like acc create and acc copyout.

Syntax:

operation ::= `acc.create` `varPtr` `(` $varPtr `:` custom<VarPtrType>(type($varPtr), $varType)
              oilist(
              `varPtrPtr` `(` $varPtrPtr `:` type($varPtrPtr) `)`
              | `bounds` `(` $bounds `)`
              | `async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType) `)`
              ) `->` type($accPtr) attr-dict

Description of arguments:

• varPtr: The address of variable to copy.
• varPtrPtr: Specifies the address of varPtr - only used when the variable copied is a field in a struct. This is important for OpenACC due to implicit attach semantics on data clauses (2.6.4).
• bounds: Used when copying just slice of array or array’s bounds are not encoded in type. They are in rank order where rank 0 is inner-most dimension.
• asyncOperands and asyncOperandsDeviceType: pair-wise lists of the async clause values associated with device_type’s.
• asyncOnly: a list of device_type’s for which async clause does not specify a value (default is acc_async_noval - OpenACC 3.3 2.16.1).
• dataClause: Keeps track of the data clause the user used. This is because the acc operations are decomposed. So a ‘copy’ clause is decomposed to both acc.copyin and acc.copyout operations, but both have dataClause that specifies acc_copy in this field.
• structured: Flag to note whether this is associated with structured region (parallel, kernels, data) or unstructured (enter data, exit data). This is important due to spec specifically calling out structured and dynamic reference counters (2.6.7).
• implicit: Whether this is an implicitly generated operation, such as copies done to satisfy “Variables with Implicitly Determined Data Attributes” in 2.6.2.
• name: Holds the name of variable as specified in user clause (including bounds).

The async values attached to the data entry operation imply that the data action applies to all device types specified by the device_type clauses using the activity queues on these devices as defined by the async values.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}, MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::RuntimeCounters, MemoryEffects::Write on ::mlir::acc::RuntimeCounters}

Attributes: ¶

Operands: ¶

Results: ¶

acc.data (acc::DataOp) ¶

Data construct

Syntax:

operation ::= `acc.data` oilist(
              `if` `(` $ifCond `)`
              | `async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType) `)`
              | `dataOperands` `(` $dataClauseOperands `:` type($dataClauseOperands) `)`
              | `wait` `` custom<WaitClause>($waitOperands, type($waitOperands),
              $waitOperandsDeviceType, $waitOperandsSegments, $hasWaitDevnum,
              $waitOnly)
              )
              $region attr-dict-with-keyword

The “acc.data” operation represents a data construct. It defines vars to be allocated in the current device memory for the duration of the region, whether data should be copied from local memory to the current device memory upon region entry , and copied from device memory to local memory upon region exit.

Example:

acc.data present(%a: memref<10x10xf32>, %b: memref<10x10xf32>,
    %c: memref<10xf32>, %d: memref<10xf32>) {
  // data region
}

async and wait operands are supported with device_type information. They should only be accessed by the extra provided getters. If modified, the corresponding device_type attributes must be modified as well.

Traits: AttrSizedOperandSegments, RecursiveMemoryEffects

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Write on ::mlir::acc::ConstructResource, MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}

Attributes: ¶

Operands: ¶

acc.declare (acc::DeclareOp) ¶

Declare implicit region

Syntax:

operation ::= `acc.declare` `dataOperands` `(` $dataClauseOperands `:` type($dataClauseOperands) `)`
              $region attr-dict-with-keyword

The “acc.declare” operation represents an implicit declare region in function (and subroutine in Fortran).

Example:

%pa = acc.present varPtr(%a : memref<10x10xf32>) -> memref<10x10xf32>
acc.declare dataOperands(%pa: memref<10x10xf32>) {
  // implicit region
}

Traits: RecursiveMemoryEffects

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Write on ::mlir::acc::ConstructResource}

Operands: ¶

acc.declare_device_resident (acc::DeclareDeviceResidentOp) ¶

Represents acc declare device_resident semantics.

Syntax:

operation ::= `acc.declare_device_resident` `varPtr` `(` $varPtr `:` custom<VarPtrType>(type($varPtr), $varType)
              oilist(
              `varPtrPtr` `(` $varPtrPtr `:` type($varPtrPtr) `)`
              | `bounds` `(` $bounds `)`
              | `async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType) `)`
              ) `->` type($accPtr) attr-dict

Description of arguments:

• varPtr: The address of variable to copy.
• varPtrPtr: Specifies the address of varPtr - only used when the variable copied is a field in a struct. This is important for OpenACC due to implicit attach semantics on data clauses (2.6.4).
• bounds: Used when copying just slice of array or array’s bounds are not encoded in type. They are in rank order where rank 0 is inner-most dimension.
• asyncOperands and asyncOperandsDeviceType: pair-wise lists of the async clause values associated with device_type’s.
• asyncOnly: a list of device_type’s for which async clause does not specify a value (default is acc_async_noval - OpenACC 3.3 2.16.1).
• dataClause: Keeps track of the data clause the user used. This is because the acc operations are decomposed. So a ‘copy’ clause is decomposed to both acc.copyin and acc.copyout operations, but both have dataClause that specifies acc_copy in this field.
• structured: Flag to note whether this is associated with structured region (parallel, kernels, data) or unstructured (enter data, exit data). This is important due to spec specifically calling out structured and dynamic reference counters (2.6.7).
• implicit: Whether this is an implicitly generated operation, such as copies done to satisfy “Variables with Implicitly Determined Data Attributes” in 2.6.2.
• name: Holds the name of variable as specified in user clause (including bounds).

The async values attached to the data entry operation imply that the data action applies to all device types specified by the device_type clauses using the activity queues on these devices as defined by the async values.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}, MemoryEffects::Effect{MemoryEffects::Write on ::mlir::acc::RuntimeCounters}

Attributes: ¶

Operands: ¶

Results: ¶

acc.declare_enter (acc::DeclareEnterOp) ¶

Declare directive - entry to implicit data region

Syntax:

operation ::= `acc.declare_enter` oilist(
              `dataOperands` `(` $dataClauseOperands `:` type($dataClauseOperands) `)`
              )
              attr-dict-with-keyword

The “acc.declare_enter” operation represents the OpenACC declare directive and captures the entry semantics to the implicit data region. This operation is modeled similarly to “acc.enter_data”.

Example showing acc declare create(a):

%0 = acc.create varPtr(%a : !llvm.ptr) -> !llvm.ptr
acc.declare_enter dataOperands(%0 : !llvm.ptr)

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Write on ::mlir::acc::ConstructResource, MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}

Operands: ¶

Results: ¶

acc.declare_exit (acc::DeclareExitOp) ¶

Declare directive - exit from implicit data region

Syntax:

operation ::= `acc.declare_exit` oilist(
              `token` `(` $token `)` |
              `dataOperands` `(` $dataClauseOperands `:` type($dataClauseOperands) `)`
              )
              attr-dict-with-keyword

The “acc.declare_exit” operation represents the OpenACC declare directive and captures the exit semantics from the implicit data region. This operation is modeled similarly to “acc.exit_data”.

Example showing acc declare device_resident(a):

%0 = acc.getdeviceptr varPtr(%a : !llvm.ptr) -> !llvm.ptr {dataClause = #acc<data_clause declare_device_resident>}
acc.declare_exit dataOperands(%0 : !llvm.ptr)
acc.delete accPtr(%0 : !llvm.ptr) {dataClause = #acc<data_clause declare_device_resident>}

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Write on ::mlir::acc::ConstructResource, MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}

Operands: ¶

acc.declare_link (acc::DeclareLinkOp) ¶

Represents acc declare link semantics.

Syntax:

operation ::= `acc.declare_link` `varPtr` `(` $varPtr `:` custom<VarPtrType>(type($varPtr), $varType)
              oilist(
              `varPtrPtr` `(` $varPtrPtr `:` type($varPtrPtr) `)`
              | `bounds` `(` $bounds `)`
              | `async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType) `)`
              ) `->` type($accPtr) attr-dict

Description of arguments:

• varPtr: The address of variable to copy.
• varPtrPtr: Specifies the address of varPtr - only used when the variable copied is a field in a struct. This is important for OpenACC due to implicit attach semantics on data clauses (2.6.4).
• bounds: Used when copying just slice of array or array’s bounds are not encoded in type. They are in rank order where rank 0 is inner-most dimension.
• asyncOperands and asyncOperandsDeviceType: pair-wise lists of the async clause values associated with device_type’s.
• asyncOnly: a list of device_type’s for which async clause does not specify a value (default is acc_async_noval - OpenACC 3.3 2.16.1).
• dataClause: Keeps track of the data clause the user used. This is because the acc operations are decomposed. So a ‘copy’ clause is decomposed to both acc.copyin and acc.copyout operations, but both have dataClause that specifies acc_copy in this field.
• structured: Flag to note whether this is associated with structured region (parallel, kernels, data) or unstructured (enter data, exit data). This is important due to spec specifically calling out structured and dynamic reference counters (2.6.7).
• implicit: Whether this is an implicitly generated operation, such as copies done to satisfy “Variables with Implicitly Determined Data Attributes” in 2.6.2.
• name: Holds the name of variable as specified in user clause (including bounds).

The async values attached to the data entry operation imply that the data action applies to all device types specified by the device_type clauses using the activity queues on these devices as defined by the async values.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}, MemoryEffects::Effect{MemoryEffects::Write on ::mlir::acc::RuntimeCounters}

Attributes: ¶

Operands: ¶

Results: ¶

acc.delete (acc::DeleteOp) ¶

Represents acc delete semantics - reverse of create.

Syntax:

operation ::= `acc.delete` `accPtr` `(` $accPtr `:` type($accPtr) `)`
              (`bounds` `(` $bounds^ `)` )?
              (`async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType)^ `)`)?
              attr-dict

• accPtr: The acc address of variable. This is the link from the data-entry operation used.
• bounds: Used when copying just slice of array or array’s bounds are not encoded in type. They are in rank order where rank 0 is inner-most dimension.
• asyncOperands and asyncOperandsDeviceType: pair-wise lists of the async clause values associated with device_type’s.
• asyncOnly: a list of device_type’s for which async clause does not specify a value (default is acc_async_noval - OpenACC 3.3 2.16.1).
• dataClause: Keeps track of the data clause the user used. This is because the acc operations are decomposed. So a ‘copy’ clause is decomposed to both acc.copyin and acc.copyout operations, but both have dataClause that specifies acc_copy in this field.
• structured: Flag to note whether this is associated with structured region (parallel, kernels, data) or unstructured (enter data, exit data). This is important due to spec specifically calling out structured and dynamic reference counters (2.6.7).
• implicit: Whether this is an implicitly generated operation, such as copies done to satisfy “Variables with Implicitly Determined Data Attributes” in 2.6.2.
• name: Holds the name of variable as specified in user clause (including bounds).

The async values attached to the data exit operation imply that the data action applies to all device types specified by the device_type clauses using the activity queues on these devices as defined by the async values.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}, MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::RuntimeCounters, MemoryEffects::Write on ::mlir::acc::RuntimeCounters}

Attributes: ¶

Operands: ¶

acc.detach (acc::DetachOp) ¶

Represents acc detach semantics - reverse of attach.

Syntax:

operation ::= `acc.detach` `accPtr` `(` $accPtr `:` type($accPtr) `)`
              (`bounds` `(` $bounds^ `)` )?
              (`async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType)^ `)`)?
              attr-dict

• accPtr: The acc address of variable. This is the link from the data-entry operation used.
• bounds: Used when copying just slice of array or array’s bounds are not encoded in type. They are in rank order where rank 0 is inner-most dimension.
• asyncOperands and asyncOperandsDeviceType: pair-wise lists of the async clause values associated with device_type’s.
• asyncOnly: a list of device_type’s for which async clause does not specify a value (default is acc_async_noval - OpenACC 3.3 2.16.1).
• dataClause: Keeps track of the data clause the user used. This is because the acc operations are decomposed. So a ‘copy’ clause is decomposed to both acc.copyin and acc.copyout operations, but both have dataClause that specifies acc_copy in this field.
• structured: Flag to note whether this is associated with structured region (parallel, kernels, data) or unstructured (enter data, exit data). This is important due to spec specifically calling out structured and dynamic reference counters (2.6.7).
• implicit: Whether this is an implicitly generated operation, such as copies done to satisfy “Variables with Implicitly Determined Data Attributes” in 2.6.2.
• name: Holds the name of variable as specified in user clause (including bounds).

The async values attached to the data exit operation imply that the data action applies to all device types specified by the device_type clauses using the activity queues on these devices as defined by the async values.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}, MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::RuntimeCounters, MemoryEffects::Write on ::mlir::acc::RuntimeCounters}

Attributes: ¶

Operands: ¶

acc.deviceptr (acc::DevicePtrOp) ¶

Specifies that the variable pointer is a device pointer.

Syntax:

operation ::= `acc.deviceptr` `varPtr` `(` $varPtr `:` custom<VarPtrType>(type($varPtr), $varType)
              oilist(
              `varPtrPtr` `(` $varPtrPtr `:` type($varPtrPtr) `)`
              | `bounds` `(` $bounds `)`
              | `async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType) `)`
              ) `->` type($accPtr) attr-dict

Description of arguments:

• varPtr: The address of variable to copy.
• varPtrPtr: Specifies the address of varPtr - only used when the variable copied is a field in a struct. This is important for OpenACC due to implicit attach semantics on data clauses (2.6.4).
• bounds: Used when copying just slice of array or array’s bounds are not encoded in type. They are in rank order where rank 0 is inner-most dimension.
• asyncOperands and asyncOperandsDeviceType: pair-wise lists of the async clause values associated with device_type’s.
• asyncOnly: a list of device_type’s for which async clause does not specify a value (default is acc_async_noval - OpenACC 3.3 2.16.1).
• dataClause: Keeps track of the data clause the user used. This is because the acc operations are decomposed. So a ‘copy’ clause is decomposed to both acc.copyin and acc.copyout operations, but both have dataClause that specifies acc_copy in this field.
• structured: Flag to note whether this is associated with structured region (parallel, kernels, data) or unstructured (enter data, exit data). This is important due to spec specifically calling out structured and dynamic reference counters (2.6.7).
• implicit: Whether this is an implicitly generated operation, such as copies done to satisfy “Variables with Implicitly Determined Data Attributes” in 2.6.2.
• name: Holds the name of variable as specified in user clause (including bounds).

The async values attached to the data entry operation imply that the data action applies to all device types specified by the device_type clauses using the activity queues on these devices as defined by the async values.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}, MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::RuntimeCounters}

Attributes: ¶

Operands: ¶

Results: ¶

acc.enter_data (acc::EnterDataOp) ¶

Enter data operation

Syntax:

operation ::= `acc.enter_data` oilist(
              `if` `(` $ifCond `)`
              | `async` `(` $asyncOperand `:` type($asyncOperand) `)`
              | `wait_devnum` `(` $waitDevnum `:` type($waitDevnum) `)`
              | `wait` `(` $waitOperands `:` type($waitOperands) `)`
              | `dataOperands` `(` $dataClauseOperands `:` type($dataClauseOperands) `)`
              )
              attr-dict-with-keyword

The “acc.enter_data” operation represents the OpenACC enter data directive.

Example:

acc.enter_data create(%d1 : memref<10xf32>) attributes {async}

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Write on ::mlir::acc::ConstructResource, MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}

Attributes: ¶

Operands: ¶

acc.exit_data (acc::ExitDataOp) ¶

Exit data operation

Syntax:

operation ::= `acc.exit_data` oilist(
              `if` `(` $ifCond `)`
              | `async` `(` $asyncOperand `:` type($asyncOperand) `)`
              | `wait_devnum` `(` $waitDevnum `:` type($waitDevnum) `)`
              | `wait` `(` $waitOperands `:` type($waitOperands) `)`
              | `dataOperands` `(` $dataClauseOperands `:` type($dataClauseOperands) `)`
              )
              attr-dict-with-keyword

The “acc.exit_data” operation represents the OpenACC exit data directive.

Example:

acc.exit_data delete(%d1 : memref<10xf32>) attributes {async}

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Write on ::mlir::acc::ConstructResource, MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}

Attributes: ¶

Operands: ¶

acc.firstprivate (acc::FirstprivateOp) ¶

Represents firstprivate semantic for the acc firstprivate clause.

Syntax:

operation ::= `acc.firstprivate` `varPtr` `(` $varPtr `:` custom<VarPtrType>(type($varPtr), $varType)
              oilist(
              `varPtrPtr` `(` $varPtrPtr `:` type($varPtrPtr) `)`
              | `bounds` `(` $bounds `)`
              | `async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType) `)`
              ) `->` type($accPtr) attr-dict

Description of arguments:

• varPtr: The address of variable to copy.
• varPtrPtr: Specifies the address of varPtr - only used when the variable copied is a field in a struct. This is important for OpenACC due to implicit attach semantics on data clauses (2.6.4).
• bounds: Used when copying just slice of array or array’s bounds are not encoded in type. They are in rank order where rank 0 is inner-most dimension.
• asyncOperands and asyncOperandsDeviceType: pair-wise lists of the async clause values associated with device_type’s.
• asyncOnly: a list of device_type’s for which async clause does not specify a value (default is acc_async_noval - OpenACC 3.3 2.16.1).
• dataClause: Keeps track of the data clause the user used. This is because the acc operations are decomposed. So a ‘copy’ clause is decomposed to both acc.copyin and acc.copyout operations, but both have dataClause that specifies acc_copy in this field.
• structured: Flag to note whether this is associated with structured region (parallel, kernels, data) or unstructured (enter data, exit data). This is important due to spec specifically calling out structured and dynamic reference counters (2.6.7).
• implicit: Whether this is an implicitly generated operation, such as copies done to satisfy “Variables with Implicitly Determined Data Attributes” in 2.6.2.
• name: Holds the name of variable as specified in user clause (including bounds).

The async values attached to the data entry operation imply that the data action applies to all device types specified by the device_type clauses using the activity queues on these devices as defined by the async values.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}

Attributes: ¶

Operands: ¶

Results: ¶

acc.firstprivate.recipe (acc::FirstprivateRecipeOp) ¶

Privatization recipe

Syntax:

operation ::= `acc.firstprivate.recipe` $sym_name `:` $type attr-dict-with-keyword `init` $initRegion
              `copy` $copyRegion
              (`destroy` $destroyRegion^)?

Declares an OpenACC privatization recipe with copy of the initial value. The operation requires two mandatory regions and one optional.

1. The initializer region specifies how to allocate and initialize a new private value. For example in Fortran, a derived-type might have a default initialization. The region has an argument that contains the value that need to be privatized. This is useful if the type is not known at compile time and the private value is needed to create its copy.
2. The copy region specifies how to copy the initial value to the newly created private value. It takes the initial value and the privatized value as arguments.
3. The destroy region specifies how to destruct the value when it reaches its end of life. It takes the privatized value as argument. It is optional.

A single privatization recipe can be used for multiple operand if they have the same type and do not require a specific default initialization.

Example:

acc.firstprivate.recipe @privatization_f32 : f32 init {
^bb0(%0: f32):
  // init region contains a sequence of operations to create and
  // initialize the copy if needed. It yields the create copy.
} copy {
^bb0(%0: f32, %1: !llvm.ptr):
  // copy region contains a sequence of operations to copy the initial value
  // of the firstprivate value to the newly created value.
} destroy {
^bb0(%0: f32)
  // destroy region contains a sequences of operations to destruct the
  // created copy.
}

// The privatization symbol is then used in the corresponding operation.
acc.parallel firstprivate(@privatization_f32 -> %a : f32) {
}

Traits: IsolatedFromAbove

Interfaces: RecipeInterface, Symbol

Attributes: ¶

acc.getdeviceptr (acc::GetDevicePtrOp) ¶

Gets device address if variable exists on device.

Syntax:

operation ::= `acc.getdeviceptr` `varPtr` `(` $varPtr `:` custom<VarPtrType>(type($varPtr), $varType)
              oilist(
              `varPtrPtr` `(` $varPtrPtr `:` type($varPtrPtr) `)`
              | `bounds` `(` $bounds `)`
              | `async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType) `)`
              ) `->` type($accPtr) attr-dict

This operation is used to get the accPtr for a variable. This is often used in conjunction with data exit operations when the data entry operation is not visible. This operation can have a dataClause argument that is any of the valid mlir::acc::DataClause entries. \

Description of arguments:

• varPtr: The address of variable to copy.
• varPtrPtr: Specifies the address of varPtr - only used when the variable copied is a field in a struct. This is important for OpenACC due to implicit attach semantics on data clauses (2.6.4).
• bounds: Used when copying just slice of array or array’s bounds are not encoded in type. They are in rank order where rank 0 is inner-most dimension.
• asyncOperands and asyncOperandsDeviceType: pair-wise lists of the async clause values associated with device_type’s.
• asyncOnly: a list of device_type’s for which async clause does not specify a value (default is acc_async_noval - OpenACC 3.3 2.16.1).
• dataClause: Keeps track of the data clause the user used. This is because the acc operations are decomposed. So a ‘copy’ clause is decomposed to both acc.copyin and acc.copyout operations, but both have dataClause that specifies acc_copy in this field.
• structured: Flag to note whether this is associated with structured region (parallel, kernels, data) or unstructured (enter data, exit data). This is important due to spec specifically calling out structured and dynamic reference counters (2.6.7).
• implicit: Whether this is an implicitly generated operation, such as copies done to satisfy “Variables with Implicitly Determined Data Attributes” in 2.6.2.
• name: Holds the name of variable as specified in user clause (including bounds).

The async values attached to the data entry operation imply that the data action applies to all device types specified by the device_type clauses using the activity queues on these devices as defined by the async values.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}, MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::RuntimeCounters}

Attributes: ¶

Operands: ¶

Results: ¶

acc.global_ctor (acc::GlobalConstructorOp) ¶

Used to hold construction operations associated with globals such as declare

Syntax:

operation ::= `acc.global_ctor` $sym_name $region attr-dict-with-keyword

The “acc.global_ctor” operation is used to capture OpenACC actions to apply on globals (such as acc declare) at the entry to the implicit data region. This operation is isolated and intended to be used in a module.

Example showing declare create of global:

llvm.mlir.global external @globalvar() : i32 {
  %0 = llvm.mlir.constant(0 : i32) : i32
  llvm.return %0 : i32
}
acc.global_ctor @acc_constructor {
  %0 = llvm.mlir.addressof @globalvar : !llvm.ptr
  %1 = acc.create varPtr(%0 : !llvm.ptr) -> !llvm.ptr
  acc.declare_enter dataOperands(%1 : !llvm.ptr)
}

Traits: IsolatedFromAbove

Interfaces: Symbol

Attributes: ¶

acc.global_dtor (acc::GlobalDestructorOp) ¶

Used to hold destruction operations associated with globals such as declare

Syntax:

operation ::= `acc.global_dtor` $sym_name $region attr-dict-with-keyword

The “acc.global_dtor” operation is used to capture OpenACC actions to apply on globals (such as acc declare) at the exit from the implicit data region. This operation is isolated and intended to be used in a module.

Example showing delete associated with declare create of global:

llvm.mlir.global external @globalvar() : i32 {
  %0 = llvm.mlir.constant(0 : i32) : i32
  llvm.return %0 : i32
}
acc.global_dtor @acc_destructor {
  %0 = llvm.mlir.addressof @globalvar : !llvm.ptr
  %1 = acc.getdeviceptr varPtr(%0 : !llvm.ptr) -> !llvm.ptr {dataClause = #acc<data_clause create>}
  acc.declare_exit dataOperands(%1 : !llvm.ptr)
  acc.delete accPtr(%1 : !llvm.ptr) {dataClause = #acc<data_clause create>}
}

Traits: IsolatedFromAbove

Interfaces: Symbol

Attributes: ¶

acc.host_data (acc::HostDataOp) ¶

Host_data construct

Syntax:

operation ::= `acc.host_data` oilist(
              `if` `(` $ifCond `)`
              | `dataOperands` `(` $dataClauseOperands `:` type($dataClauseOperands) `)`
              )
              $region attr-dict-with-keyword

The “acc.host_data” operation represents the OpenACC host_data construct.

Example:

%0 = acc.use_device varPtr(%a : !llvm.ptr) -> !llvm.ptr
acc.host_data dataOperands(%0 : !llvm.ptr) {

}

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Write on ::mlir::acc::ConstructResource, MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}

Attributes: ¶

Operands: ¶

acc.init (acc::InitOp) ¶

Init operation

Syntax:

operation ::= `acc.init` oilist(`device_num` `(` $deviceNumOperand `:` type($deviceNumOperand) `)`
              | `if` `(` $ifCond `)`
              ) attr-dict-with-keyword

The “acc.init” operation represents the OpenACC init executable directive.

Example:

acc.init
acc.init device_num(%dev1 : i32)

Traits: AttrSizedOperandSegments

Attributes: ¶

Operands: ¶

acc.kernels (acc::KernelsOp) ¶

Kernels construct

Syntax:

operation ::= `acc.kernels` ( `combined` `(` `loop` `)` $combined^)?
              oilist(
              `dataOperands` `(` $dataClauseOperands `:` type($dataClauseOperands) `)`
              | `async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType) `)`
              | `num_gangs` `(` custom<NumGangs>($numGangs,
              type($numGangs), $numGangsDeviceType, $numGangsSegments) `)`
              | `num_workers` `(` custom<DeviceTypeOperands>($numWorkers,
              type($numWorkers), $numWorkersDeviceType) `)`
              | `vector_length` `(` custom<DeviceTypeOperands>($vectorLength,
              type($vectorLength), $vectorLengthDeviceType) `)`
              | `wait` `` custom<WaitClause>($waitOperands, type($waitOperands),
              $waitOperandsDeviceType, $waitOperandsSegments, $hasWaitDevnum,
              $waitOnly)
              | `self` `(` $selfCond `)`
              | `if` `(` $ifCond `)`
              )
              $region attr-dict-with-keyword

The “acc.kernels” operation represents a kernels construct block. It has one region to be compiled into a sequence of kernels for execution on the current device.

Example:

acc.kernels num_gangs(%c10) num_workers(%c10)
    private(%c : memref<10xf32>) {
  // kernels region
}

collapse, gang, worker, vector, seq, independent, auto and tile operands are supported with device_type information. They should only be accessed by the extra provided getters. If modified, the corresponding device_type attributes must be modified as well.

Traits: AttrSizedOperandSegments, AutomaticAllocationScope, RecursiveMemoryEffects

Interfaces: ComputeRegionOpInterface, MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Write on ::mlir::acc::ConstructResource, MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}

Attributes: ¶

Operands: ¶

acc.loop (acc::LoopOp) ¶

Loop construct

Syntax:

operation ::= `acc.loop` custom<CombinedConstructsLoop>($combined)
              oilist(
              `gang` `` custom<GangClause>($gangOperands, type($gangOperands),
              $gangOperandsArgType, $gangOperandsDeviceType,
              $gangOperandsSegments, $gang)
              | `worker` `` custom<DeviceTypeOperandsWithKeywordOnly>(
              $workerNumOperands, type($workerNumOperands),
              $workerNumOperandsDeviceType, $worker)
              | `vector` `` custom<DeviceTypeOperandsWithKeywordOnly>($vectorOperands,
              type($vectorOperands), $vectorOperandsDeviceType, $vector)
              | `private` `(` custom<SymOperandList>(
              $privateOperands, type($privateOperands), $privatizations) `)`
              | `tile` `(` custom<DeviceTypeOperandsWithSegment>($tileOperands,
              type($tileOperands), $tileOperandsDeviceType, $tileOperandsSegments)
              `)`
              | `reduction` `(` custom<SymOperandList>(
              $reductionOperands, type($reductionOperands), $reductionRecipes)
              `)`
              | `cache` `(` $cacheOperands `:` type($cacheOperands) `)`
              )
              custom<LoopControl>($region, $lowerbound, type($lowerbound), $upperbound,
              type($upperbound), $step, type($step))
              ( `(` type($results)^ `)` )?
              attr-dict-with-keyword

The “acc.loop” operation represents the OpenACC loop construct. The lower and upper bounds specify a half-open range: the range includes the lower bound but does not include the upper bound. If the inclusive attribute is set then the upper bound is included.

Example:

acc.loop gang() vector() (%arg3 : index, %arg4 : index, %arg5 : index) = 
    (%c0, %c0, %c0 : index, index, index) to 
    (%c10, %c10, %c10 : index, index, index) step 
    (%c1, %c1, %c1 : index, index, index) {
  // Loop body
  acc.yield
} attributes { collapse = [3] }

collapse, gang, worker, vector, seq, independent, auto and tile operands are supported with device_type information. They should only be accessed by the extra provided getters. If modified, the corresponding device_type attributes must be modified as well.

Traits: AttrSizedOperandSegments, AutomaticAllocationScope, RecursiveMemoryEffects

Interfaces: ComputeRegionOpInterface, LoopLikeOpInterface, MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Write on ::mlir::acc::ConstructResource}

Attributes: ¶

Operands: ¶

Results: ¶

acc.nocreate (acc::NoCreateOp) ¶

Represents acc no_create semantics.

Syntax:

operation ::= `acc.nocreate` `varPtr` `(` $varPtr `:` custom<VarPtrType>(type($varPtr), $varType)
              oilist(
              `varPtrPtr` `(` $varPtrPtr `:` type($varPtrPtr) `)`
              | `bounds` `(` $bounds `)`
              | `async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType) `)`
              ) `->` type($accPtr) attr-dict

Description of arguments:

• varPtr: The address of variable to copy.
• varPtrPtr: Specifies the address of varPtr - only used when the variable copied is a field in a struct. This is important for OpenACC due to implicit attach semantics on data clauses (2.6.4).
• bounds: Used when copying just slice of array or array’s bounds are not encoded in type. They are in rank order where rank 0 is inner-most dimension.
• asyncOperands and asyncOperandsDeviceType: pair-wise lists of the async clause values associated with device_type’s.
• asyncOnly: a list of device_type’s for which async clause does not specify a value (default is acc_async_noval - OpenACC 3.3 2.16.1).
• dataClause: Keeps track of the data clause the user used. This is because the acc operations are decomposed. So a ‘copy’ clause is decomposed to both acc.copyin and acc.copyout operations, but both have dataClause that specifies acc_copy in this field.
• structured: Flag to note whether this is associated with structured region (parallel, kernels, data) or unstructured (enter data, exit data). This is important due to spec specifically calling out structured and dynamic reference counters (2.6.7).
• implicit: Whether this is an implicitly generated operation, such as copies done to satisfy “Variables with Implicitly Determined Data Attributes” in 2.6.2.
• name: Holds the name of variable as specified in user clause (including bounds).

The async values attached to the data entry operation imply that the data action applies to all device types specified by the device_type clauses using the activity queues on these devices as defined by the async values.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}, MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::RuntimeCounters, MemoryEffects::Write on ::mlir::acc::RuntimeCounters}

Attributes: ¶

Operands: ¶

Results: ¶

acc.parallel (acc::ParallelOp) ¶

Parallel construct

Syntax:

operation ::= `acc.parallel` ( `combined` `(` `loop` `)` $combined^)?
              oilist(
              `dataOperands` `(` $dataClauseOperands `:` type($dataClauseOperands) `)`
              | `async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType) `)`
              | `firstprivate` `(` custom<SymOperandList>($firstprivateOperands,
              type($firstprivateOperands), $firstprivatizations)
              `)`
              | `num_gangs` `(` custom<NumGangs>($numGangs,
              type($numGangs), $numGangsDeviceType, $numGangsSegments) `)`
              | `num_workers` `(` custom<DeviceTypeOperands>($numWorkers,
              type($numWorkers), $numWorkersDeviceType) `)`
              | `private` `(` custom<SymOperandList>(
              $privateOperands, type($privateOperands), $privatizations)
              `)`
              | `vector_length` `(` custom<DeviceTypeOperands>($vectorLength,
              type($vectorLength), $vectorLengthDeviceType) `)`
              | `wait` `` custom<WaitClause>($waitOperands, type($waitOperands),
              $waitOperandsDeviceType, $waitOperandsSegments, $hasWaitDevnum,
              $waitOnly)
              | `self` `(` $selfCond `)`
              | `if` `(` $ifCond `)`
              | `reduction` `(` custom<SymOperandList>(
              $reductionOperands, type($reductionOperands), $reductionRecipes)
              `)`
              )
              $region attr-dict-with-keyword

The “acc.parallel” operation represents a parallel construct block. It has one region to be executed in parallel on the current device.

Example:

acc.parallel num_gangs(%c10) num_workers(%c10)
    private(%c : memref<10xf32>) {
  // parallel region
}

async, wait, num_gangs, num_workers and vector_length operands are supported with device_type information. They should only be accessed by the extra provided getters. If modified, the corresponding device_type attributes must be modified as well.

Traits: AttrSizedOperandSegments, AutomaticAllocationScope, RecursiveMemoryEffects

Interfaces: ComputeRegionOpInterface, MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Write on ::mlir::acc::ConstructResource, MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}

Attributes: ¶

Operands: ¶

acc.present (acc::PresentOp) ¶

Specifies that the variable is already present on device.

Syntax:

operation ::= `acc.present` `varPtr` `(` $varPtr `:` custom<VarPtrType>(type($varPtr), $varType)
              oilist(
              `varPtrPtr` `(` $varPtrPtr `:` type($varPtrPtr) `)`
              | `bounds` `(` $bounds `)`
              | `async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType) `)`
              ) `->` type($accPtr) attr-dict

Description of arguments:

• varPtr: The address of variable to copy.
• varPtrPtr: Specifies the address of varPtr - only used when the variable copied is a field in a struct. This is important for OpenACC due to implicit attach semantics on data clauses (2.6.4).
• bounds: Used when copying just slice of array or array’s bounds are not encoded in type. They are in rank order where rank 0 is inner-most dimension.
• asyncOperands and asyncOperandsDeviceType: pair-wise lists of the async clause values associated with device_type’s.
• asyncOnly: a list of device_type’s for which async clause does not specify a value (default is acc_async_noval - OpenACC 3.3 2.16.1).
• dataClause: Keeps track of the data clause the user used. This is because the acc operations are decomposed. So a ‘copy’ clause is decomposed to both acc.copyin and acc.copyout operations, but both have dataClause that specifies acc_copy in this field.
• structured: Flag to note whether this is associated with structured region (parallel, kernels, data) or unstructured (enter data, exit data). This is important due to spec specifically calling out structured and dynamic reference counters (2.6.7).
• implicit: Whether this is an implicitly generated operation, such as copies done to satisfy “Variables with Implicitly Determined Data Attributes” in 2.6.2.
• name: Holds the name of variable as specified in user clause (including bounds).

The async values attached to the data entry operation imply that the data action applies to all device types specified by the device_type clauses using the activity queues on these devices as defined by the async values.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}, MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::RuntimeCounters, MemoryEffects::Write on ::mlir::acc::RuntimeCounters}

Attributes: ¶

Operands: ¶

Results: ¶

acc.private (acc::PrivateOp) ¶

Represents private semantics for acc private clause.

Syntax:

operation ::= `acc.private` `varPtr` `(` $varPtr `:` custom<VarPtrType>(type($varPtr), $varType)
              oilist(
              `varPtrPtr` `(` $varPtrPtr `:` type($varPtrPtr) `)`
              | `bounds` `(` $bounds `)`
              | `async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType) `)`
              ) `->` type($accPtr) attr-dict

Description of arguments:

• varPtr: The address of variable to copy.
• varPtrPtr: Specifies the address of varPtr - only used when the variable copied is a field in a struct. This is important for OpenACC due to implicit attach semantics on data clauses (2.6.4).
• bounds: Used when copying just slice of array or array’s bounds are not encoded in type. They are in rank order where rank 0 is inner-most dimension.
• asyncOperands and asyncOperandsDeviceType: pair-wise lists of the async clause values associated with device_type’s.
• asyncOnly: a list of device_type’s for which async clause does not specify a value (default is acc_async_noval - OpenACC 3.3 2.16.1).
• dataClause: Keeps track of the data clause the user used. This is because the acc operations are decomposed. So a ‘copy’ clause is decomposed to both acc.copyin and acc.copyout operations, but both have dataClause that specifies acc_copy in this field.
• structured: Flag to note whether this is associated with structured region (parallel, kernels, data) or unstructured (enter data, exit data). This is important due to spec specifically calling out structured and dynamic reference counters (2.6.7).
• implicit: Whether this is an implicitly generated operation, such as copies done to satisfy “Variables with Implicitly Determined Data Attributes” in 2.6.2.
• name: Holds the name of variable as specified in user clause (including bounds).

The async values attached to the data entry operation imply that the data action applies to all device types specified by the device_type clauses using the activity queues on these devices as defined by the async values.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}

Attributes: ¶

Operands: ¶

Results: ¶

acc.private.recipe (acc::PrivateRecipeOp) ¶

Privatization recipe

Syntax:

operation ::= `acc.private.recipe` $sym_name `:` $type attr-dict-with-keyword `init` $initRegion
              (`destroy` $destroyRegion^)?

Declares an OpenACC privatization recipe. The operation requires one mandatory and one optional region.

1. The initializer region specifies how to allocate and initialize a new private value. For example in Fortran, a derived-type might have a default initialization. The region has an argument that contains the value that need to be privatized. This is useful if the type is not known at compile time and the private value is needed to create its copy.
2. The destroy region specifies how to destruct the value when it reaches its end of life. It takes the privatized value as argument.

A single privatization recipe can be used for multiple operand if they have the same type and do not require a specific default initialization.

Example:

acc.private.recipe @privatization_f32 : f32 init {
^bb0(%0: f32):
  // init region contains a sequence of operations to create and
  // initialize the copy if needed. It yields the create copy.
} destroy {
^bb0(%0: f32)
  // destroy region contains a sequences of operations to destruct the
  // created copy.
}

// The privatization symbol is then used in the corresponding operation.
acc.parallel private(@privatization_f32 -> %a : f32) {
}

Traits: IsolatedFromAbove

Interfaces: RecipeInterface, Symbol

Attributes: ¶

acc.reduction (acc::ReductionOp) ¶

Represents reduction semantics for acc reduction clause.

Syntax:

operation ::= `acc.reduction` `varPtr` `(` $varPtr `:` custom<VarPtrType>(type($varPtr), $varType)
              oilist(
              `varPtrPtr` `(` $varPtrPtr `:` type($varPtrPtr) `)`
              | `bounds` `(` $bounds `)`
              | `async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType) `)`
              ) `->` type($accPtr) attr-dict

Description of arguments:

• varPtr: The address of variable to copy.
• varPtrPtr: Specifies the address of varPtr - only used when the variable copied is a field in a struct. This is important for OpenACC due to implicit attach semantics on data clauses (2.6.4).
• bounds: Used when copying just slice of array or array’s bounds are not encoded in type. They are in rank order where rank 0 is inner-most dimension.
• asyncOperands and asyncOperandsDeviceType: pair-wise lists of the async clause values associated with device_type’s.
• asyncOnly: a list of device_type’s for which async clause does not specify a value (default is acc_async_noval - OpenACC 3.3 2.16.1).
• dataClause: Keeps track of the data clause the user used. This is because the acc operations are decomposed. So a ‘copy’ clause is decomposed to both acc.copyin and acc.copyout operations, but both have dataClause that specifies acc_copy in this field.
• structured: Flag to note whether this is associated with structured region (parallel, kernels, data) or unstructured (enter data, exit data). This is important due to spec specifically calling out structured and dynamic reference counters (2.6.7).
• implicit: Whether this is an implicitly generated operation, such as copies done to satisfy “Variables with Implicitly Determined Data Attributes” in 2.6.2.
• name: Holds the name of variable as specified in user clause (including bounds).

The async values attached to the data entry operation imply that the data action applies to all device types specified by the device_type clauses using the activity queues on these devices as defined by the async values.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}

Attributes: ¶

Operands: ¶

Results: ¶

acc.reduction.recipe (acc::ReductionRecipeOp) ¶

Reduction recipe

Syntax:

operation ::= `acc.reduction.recipe` $sym_name `:` $type attr-dict-with-keyword
              `reduction_operator` $reductionOperator
              `init` $initRegion `combiner` $combinerRegion

Declares an OpenACC reduction recipe. The operation requires two mandatory regions.

1. The initializer region specifies how to initialize the local reduction value. The region has a first argument that contains the value of the reduction accumulator at the start of the reduction. It is expected to acc.yield the new value. Extra arguments can be added to deal with dynamic arrays.
2. The reduction region contains a sequences of operations to combine two values of the reduction type into one. It has at least two arguments and it is expected to acc.yield the combined value. Extra arguments can be added to deal with dynamic arrays.

Example:

acc.reduction.recipe @reduction_add_i64 : i64 reduction_operator<add> init {
^bb0(%0: i64):
  // init region contains a sequence of operations to initialize the local
  // reduction value as specified in 2.5.15
  %c0 = arith.constant 0 : i64
  acc.yield %c0 : i64
} combiner {
^bb0(%0: i64, %1: i64)
  // combiner region contains a sequence of operations to combine
  // two values into one.
  %2 = arith.addi %0, %1 : i64
  acc.yield %2 : i64
}

// The reduction symbol is then used in the corresponding operation.
acc.parallel reduction(@reduction_add_i64 -> %a : i64) {
}

The following table lists the valid operators and the initialization values according to OpenACC 3.3:

|————————————————|

Traits: IsolatedFromAbove

Interfaces: RecipeInterface, Symbol

Attributes: ¶

acc.routine (acc::RoutineOp) ¶

Acc routine operation

Syntax:

operation ::= `acc.routine` $sym_name `func` `(` $func_name `)`
              oilist (
              `bind` `(` custom<BindName>($bindName, $bindNameDeviceType) `)`
              | `gang` `` custom<RoutineGangClause>($gang, $gangDim, $gangDimDeviceType)
              | `worker` custom<DeviceTypeArrayAttr>($worker)
              | `vector` custom<DeviceTypeArrayAttr>($vector)
              | `seq` custom<DeviceTypeArrayAttr>($seq)
              | `nohost` $nohost
              | `implicit` $implicit
              ) attr-dict-with-keyword

The acc.routine operation is used to capture the clauses of acc routine directive, including the associated function name. The associated function keeps track of its corresponding routine declaration through the RoutineInfoAttr.

Example:

func.func @acc_func(%a : i64) -> () attributes 
    {acc.routine_info = #acc.routine_info<[@acc_func_rout1]>} {
  return
}
acc.routine @acc_func_rout1 func(@acc_func) gang

bind, gang, worker, vector and seq operands are supported with device_type information. They should only be accessed by the extra provided getters. If modified, the corresponding device_type attributes must be modified as well.

Traits: IsolatedFromAbove

Attributes: ¶

acc.serial (acc::SerialOp) ¶

Serial construct

Syntax:

operation ::= `acc.serial` ( `combined` `(` `loop` `)` $combined^)?
              oilist(
              `dataOperands` `(` $dataClauseOperands `:` type($dataClauseOperands) `)`
              | `async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType) `)`
              | `firstprivate` `(` custom<SymOperandList>($firstprivateOperands,
              type($firstprivateOperands), $firstprivatizations)
              `)`
              | `private` `(` custom<SymOperandList>(
              $privateOperands, type($privateOperands), $privatizations)
              `)`
              | `wait` `` custom<WaitClause>($waitOperands, type($waitOperands),
              $waitOperandsDeviceType, $waitOperandsSegments, $hasWaitDevnum,
              $waitOnly)
              | `self` `(` $selfCond `)`
              | `if` `(` $ifCond `)`
              | `reduction` `(` custom<SymOperandList>(
              $reductionOperands, type($reductionOperands), $reductionRecipes)
              `)`
              )
              $region attr-dict-with-keyword

The “acc.serial” operation represents a serial construct block. It has one region to be executed in serial on the current device.

Example:

acc.serial private(%c : memref<10xf32>) {
  // serial region
}

async and wait operands are supported with device_type information. They should only be accessed by the extra provided getters. If modified, the corresponding device_type attributes must be modified as well.

Traits: AttrSizedOperandSegments, AutomaticAllocationScope, RecursiveMemoryEffects

Interfaces: ComputeRegionOpInterface, MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Write on ::mlir::acc::ConstructResource, MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}

Attributes: ¶

Operands: ¶

acc.set (acc::SetOp) ¶

Set operation

Syntax:

operation ::= `acc.set` oilist(`default_async` `(` $defaultAsync `:` type($defaultAsync) `)`
              | `device_num` `(` $deviceNum `:` type($deviceNum) `)`
              | `if` `(` $ifCond `)`
              ) attr-dict-with-keyword

The “acc.set” operation represents the OpenACC set directive.

Example:

acc.set device_num(%dev1 : i32)

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Write on ::mlir::acc::CurrentDeviceIdResource}

Attributes: ¶

Operands: ¶

acc.shutdown (acc::ShutdownOp) ¶

Shutdown operation

Syntax:

operation ::= `acc.shutdown` oilist(`device_num` `(` $deviceNumOperand `:` type($deviceNumOperand) `)`
              |`if` `(` $ifCond `)`
              ) attr-dict-with-keyword

The “acc.shutdown” operation represents the OpenACC shutdown executable directive.

Example:

acc.shutdown
acc.shutdown device_num(%dev1 : i32)

Traits: AttrSizedOperandSegments

Attributes: ¶

Operands: ¶

acc.terminator (acc::TerminatorOp) ¶

Generic terminator for OpenACC regions

Syntax:

operation ::= `acc.terminator` attr-dict

A terminator operation for regions that appear in the body of OpenACC operation. Generic OpenACC construct regions are not expected to return any value so the terminator takes no operands. The terminator op returns control to the enclosing op.

Traits: AlwaysSpeculatableImplTrait, Terminator

Interfaces: ConditionallySpeculatable, NoMemoryEffect (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{}

acc.update (acc::UpdateOp) ¶

Update operation

Syntax:

operation ::= `acc.update` oilist(
              `if` `(` $ifCond `)`
              | `async` `` custom<DeviceTypeOperandsWithKeywordOnly>(
              $asyncOperands, type($asyncOperands),
              $asyncOperandsDeviceType, $async)
              | `wait` `` custom<WaitClause>($waitOperands, type($waitOperands),
              $waitOperandsDeviceType, $waitOperandsSegments, $hasWaitDevnum,
              $waitOnly)
              | `dataOperands` `(` $dataClauseOperands `:` type($dataClauseOperands) `)`
              )
              attr-dict-with-keyword

The acc.update operation represents the OpenACC update executable directive. As host and self clauses are synonyms, any operands for host and self are add to $hostOperands.

Example:

acc.update device(%d1 : memref<10xf32>) attributes {async}

async and wait operands are supported with device_type information. They should only be accessed by the extra provided getters. If modified, the corresponding device_type attributes must be modified as well.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Write on ::mlir::acc::ConstructResource, MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}

Attributes: ¶

Operands: ¶

acc.update_device (acc::UpdateDeviceOp) ¶

Represents acc update device semantics.

Syntax:

operation ::= `acc.update_device` `varPtr` `(` $varPtr `:` custom<VarPtrType>(type($varPtr), $varType)
              oilist(
              `varPtrPtr` `(` $varPtrPtr `:` type($varPtrPtr) `)`
              | `bounds` `(` $bounds `)`
              | `async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType) `)`
              ) `->` type($accPtr) attr-dict

Description of arguments:

• varPtr: The address of variable to copy.
• varPtrPtr: Specifies the address of varPtr - only used when the variable copied is a field in a struct. This is important for OpenACC due to implicit attach semantics on data clauses (2.6.4).
• bounds: Used when copying just slice of array or array’s bounds are not encoded in type. They are in rank order where rank 0 is inner-most dimension.
• asyncOperands and asyncOperandsDeviceType: pair-wise lists of the async clause values associated with device_type’s.
• asyncOnly: a list of device_type’s for which async clause does not specify a value (default is acc_async_noval - OpenACC 3.3 2.16.1).
• dataClause: Keeps track of the data clause the user used. This is because the acc operations are decomposed. So a ‘copy’ clause is decomposed to both acc.copyin and acc.copyout operations, but both have dataClause that specifies acc_copy in this field.
• structured: Flag to note whether this is associated with structured region (parallel, kernels, data) or unstructured (enter data, exit data). This is important due to spec specifically calling out structured and dynamic reference counters (2.6.7).
• implicit: Whether this is an implicitly generated operation, such as copies done to satisfy “Variables with Implicitly Determined Data Attributes” in 2.6.2.
• name: Holds the name of variable as specified in user clause (including bounds).

The async values attached to the data entry operation imply that the data action applies to all device types specified by the device_type clauses using the activity queues on these devices as defined by the async values.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}

Attributes: ¶

Operands: ¶

Results: ¶

acc.update_host (acc::UpdateHostOp) ¶

Represents acc update host semantics.

Syntax:

operation ::= `acc.update_host` `accPtr` `(` $accPtr `:` type($accPtr) `)`
              (`bounds` `(` $bounds^ `)` )?
              (`async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType)^ `)`)?
              `to` `varPtr` `(` $varPtr `:` custom<VarPtrType>(type($varPtr), $varType)
              attr-dict

• varPtr: The address of variable to copy back to.

  • accPtr: The acc address of variable. This is the link from the data-entry operation used.
  • bounds: Used when copying just slice of array or array’s bounds are not encoded in type. They are in rank order where rank 0 is inner-most dimension.
  • asyncOperands and asyncOperandsDeviceType: pair-wise lists of the async clause values associated with device_type’s.
  • asyncOnly: a list of device_type’s for which async clause does not specify a value (default is acc_async_noval - OpenACC 3.3 2.16.1).
  • dataClause: Keeps track of the data clause the user used. This is because the acc operations are decomposed. So a ‘copy’ clause is decomposed to both acc.copyin and acc.copyout operations, but both have dataClause that specifies acc_copy in this field.
  • structured: Flag to note whether this is associated with structured region (parallel, kernels, data) or unstructured (enter data, exit data). This is important due to spec specifically calling out structured and dynamic reference counters (2.6.7).
  • implicit: Whether this is an implicitly generated operation, such as copies done to satisfy “Variables with Implicitly Determined Data Attributes” in 2.6.2.
  • name: Holds the name of variable as specified in user clause (including bounds).

  The async values attached to the data exit operation imply that the data action applies to all device types specified by the device_type clauses using the activity queues on these devices as defined by the async values.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}, MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::RuntimeCounters, MemoryEffects::Write on ::mlir::acc::RuntimeCounters}

Attributes: ¶

Operands: ¶

acc.use_device (acc::UseDeviceOp) ¶

Represents acc use_device semantics.

Syntax:

operation ::= `acc.use_device` `varPtr` `(` $varPtr `:` custom<VarPtrType>(type($varPtr), $varType)
              oilist(
              `varPtrPtr` `(` $varPtrPtr `:` type($varPtrPtr) `)`
              | `bounds` `(` $bounds `)`
              | `async` `(` custom<DeviceTypeOperands>($asyncOperands,
              type($asyncOperands), $asyncOperandsDeviceType) `)`
              ) `->` type($accPtr) attr-dict

Description of arguments:

• varPtr: The address of variable to copy.
• varPtrPtr: Specifies the address of varPtr - only used when the variable copied is a field in a struct. This is important for OpenACC due to implicit attach semantics on data clauses (2.6.4).
• bounds: Used when copying just slice of array or array’s bounds are not encoded in type. They are in rank order where rank 0 is inner-most dimension.
• asyncOperands and asyncOperandsDeviceType: pair-wise lists of the async clause values associated with device_type’s.
• asyncOnly: a list of device_type’s for which async clause does not specify a value (default is acc_async_noval - OpenACC 3.3 2.16.1).
• dataClause: Keeps track of the data clause the user used. This is because the acc operations are decomposed. So a ‘copy’ clause is decomposed to both acc.copyin and acc.copyout operations, but both have dataClause that specifies acc_copy in this field.
• structured: Flag to note whether this is associated with structured region (parallel, kernels, data) or unstructured (enter data, exit data). This is important due to spec specifically calling out structured and dynamic reference counters (2.6.7).
• implicit: Whether this is an implicitly generated operation, such as copies done to satisfy “Variables with Implicitly Determined Data Attributes” in 2.6.2.
• name: Holds the name of variable as specified in user clause (including bounds).

The async values attached to the data entry operation imply that the data action applies to all device types specified by the device_type clauses using the activity queues on these devices as defined by the async values.

Traits: AttrSizedOperandSegments

Interfaces: MemoryEffectOpInterface (MemoryEffectOpInterface)

Effects: MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::CurrentDeviceIdResource}, MemoryEffects::Effect{MemoryEffects::Read on ::mlir::acc::RuntimeCounters}

Attributes: ¶

Operands: ¶

Results: ¶

acc.wait (acc::WaitOp) ¶

Wait operation

Syntax:

operation ::= `acc.wait` ( `(` $waitOperands^ `:` type($waitOperands) `)` )?
              oilist(`async` `(` $asyncOperand `:` type($asyncOperand) `)`
              |`wait_devnum` `(` $waitDevnum `:` type($waitDevnum) `)`
              |`if` `(` $ifCond `)`
              ) attr-dict-with-keyword

The “acc.wait” operation represents the OpenACC wait executable directive.

Example:

acc.wait(%value1: index)
acc.wait() async(%async1: i32)

acc.wait does not implement MemoryEffects interface, so it affects all the resources. This is conservatively correct. More precise modelling of the memory effects seems to be impossible without the whole program analysis.

Traits: AttrSizedOperandSegments

Attributes: ¶

Operands: ¶

acc.yield (acc::YieldOp) ¶

Acc yield and termination operation

Syntax:

operation ::= `acc.yield` attr-dict ($operands^ `:` type($operands))?

acc.yield is a special terminator operation for block inside regions in various acc ops (including parallel, loop, atomic.update). It returns values to the immediately enclosing acc op.

Traits: AlwaysSpeculatableImplTrait, HasParent<FirstprivateRecipeOp, LoopOp, ParallelOp, PrivateRecipeOp,ReductionRecipeOp, SerialOp, AtomicUpdateOp>, ReturnLike, Terminator

Interfaces: ConditionallySpeculatable, NoMemoryEffect (MemoryEffectOpInterface), RegionBranchTerminatorOpInterface

Effects: MemoryEffects::Effect{}

Operands: ¶

Attributes ¶

DeclareActionAttr ¶

Syntax:

#acc.declare_action<
  SymbolRefAttr,   # preAlloc
  SymbolRefAttr,   # postAlloc
  SymbolRefAttr,   # preDealloc
  SymbolRefAttr   # postDealloc
>

Parameters: ¶

DeclareAttr ¶

Syntax:

#acc.declare<
  DataClauseAttr,   # dataClause
  bool   # implicit
>

Parameters: ¶

ClauseDefaultValueAttr ¶

DefaultValue Clause

Syntax:

#acc.defaultvalue<
  ::mlir::acc::ClauseDefaultValue   # value
>

Enum cases:

• present (Present)
• none (None)

Parameters: ¶

CombinedConstructsTypeAttr ¶

Differentiate between combined constructs

Syntax:

#acc.combined_constructs<
  ::mlir::acc::CombinedConstructsType   # value
>

Enum cases:

• kernels_loop (KernelsLoop)
• parallel_loop (ParallelLoop)
• serial_loop (SerialLoop)

Parameters: ¶

ConstructAttr ¶

constructs supported by OpenACC

Syntax:

#acc.construct<
  ::mlir::acc::Construct   # value
>

Enum cases:

• acc_construct_parallel (acc_construct_parallel)
• acc_construct_kernels (acc_construct_kernels)
• acc_construct_loop (acc_construct_loop)
• acc_construct_data (acc_construct_data)
• acc_construct_enter_data (acc_construct_enter_data)
• acc_construct_exit_data (acc_construct_exit_data)
• acc_construct_host_data (acc_construct_host_data)
• acc_construct_atomic (acc_construct_atomic)
• acc_construct_declare (acc_construct_declare)
• acc_construct_init (acc_construct_init)
• acc_construct_shutdown (acc_construct_shutdown)
• acc_construct_set (acc_construct_set)
• acc_construct_update (acc_construct_update)
• acc_construct_routine (acc_construct_routine)
• acc_construct_wait (acc_construct_wait)
• acc_construct_runtime_api (acc_construct_runtime_api)
• acc_construct_serial (acc_construct_serial)

Parameters: ¶

DataClauseAttr ¶

data clauses supported by OpenACC

Syntax:

#acc.data_clause<
  ::mlir::acc::DataClause   # value
>

Enum cases:

• acc_copyin (acc_copyin)
• acc_copyin_readonly (acc_copyin_readonly)
• acc_copy (acc_copy)
• acc_copyout (acc_copyout)
• acc_copyout_zero (acc_copyout_zero)
• acc_present (acc_present)
• acc_create (acc_create)
• acc_create_zero (acc_create_zero)
• acc_delete (acc_delete)
• acc_attach (acc_attach)
• acc_detach (acc_detach)
• acc_no_create (acc_no_create)
• acc_private (acc_private)
• acc_firstprivate (acc_firstprivate)
• acc_deviceptr (acc_deviceptr)
• acc_getdeviceptr (acc_getdeviceptr)
• acc_update_host (acc_update_host)
• acc_update_self (acc_update_self)
• acc_update_device (acc_update_device)
• acc_use_device (acc_use_device)
• acc_reduction (acc_reduction)
• acc_declare_device_resident (acc_declare_device_resident)
• acc_declare_link (acc_declare_link)
• acc_cache (acc_cache)
• acc_cache_readonly (acc_cache_readonly)

Parameters: ¶

DeviceTypeAttr ¶

built-in device type supported by OpenACC

Syntax:

#acc.device_type<
  ::mlir::acc::DeviceType   # value
>

Enum cases:

• none (None)
• star (Star)
• default (Default)
• host (Host)
• multicore (Multicore)
• nvidia (Nvidia)
• radeon (Radeon)

Parameters: ¶

GangArgTypeAttr ¶

Differentiate the different gang arg values

Syntax:

#acc.gang_arg_type<
  ::mlir::acc::GangArgType   # value
>

Enum cases:

• Num (Num)
• Dim (Dim)
• Static (Static)

Parameters: ¶

ReductionOperatorAttr ¶

built-in reduction operations supported by OpenACC

Syntax:

#acc.reduction_operator<
  ::mlir::acc::ReductionOperator   # value
>

Enum cases:

• add (AccAdd)
• mul (AccMul)
• max (AccMax)
• min (AccMin)
• iand (AccIand)
• ior (AccIor)
• xor (AccXor)
• eqv (AccEqv)
• neqv (AccNeqv)
• land (AccLand)
• lor (AccLor)

Parameters: ¶

RoutineInfoAttr ¶

Keeps track of associated acc routine information

Syntax:

#acc.routine_info<
  ::llvm::ArrayRef<SymbolRefAttr>   # accRoutines
>

This attribute is used to create the association between a function and its acc.routine operation. A func.func uses this if its name was referenced in an acc routine directive.

Parameters: ¶

Types ¶

DataBoundsType ¶

Type for representing acc data clause bounds information

Syntax: !acc.data_bounds_ty

DeclareTokenType ¶

declare token type

Syntax: !acc.declare_token

acc.declare_token is a type returned by a declare_enter operation and can be passed to a declare_exit operation to represent an implicit data region.

Enums ¶

ClauseDefaultValue ¶

DefaultValue Clause

Cases: ¶

CombinedConstructsType ¶

Differentiate between combined constructs

Cases: ¶

Construct ¶

constructs supported by OpenACC

Cases: ¶

DataClause ¶

data clauses supported by OpenACC

Cases: ¶

DeviceType ¶

built-in device type supported by OpenACC

Cases: ¶

GangArgType ¶

Differentiate the different gang arg values

Cases: ¶

ReductionOperator ¶

built-in reduction operations supported by OpenACC

Cases: ¶