[mlir][ODS] Rename several uses of 'description' to 'summary'

[rriddle]

Summary

Right now constraint/predicate traits/etc. use their "description" field as a one line human readable string. This breaks the current convention, by which a "description" may be multi-line. This revision renames the "description" field in these cases to "summary" which matches what the string is actually used as. This also unbreaks the use of TypeDefs(and eventually AttrDefs) in conjunction with existing type constraint facilities like Optional.

[rriddle]

Rebase

mlir/docs/OpDefinitions.md

 # Operation Definition Specification (ODS)
 
 In addition to specializing the `mlir::Op` C++ template, MLIR also supports
 defining operations and data types in a table-driven manner. This is achieved
 via [TableGen][TableGen], which is both a generic language and its tooling to
 maintain records of domain-specific information. Facts regarding an operation
-are specified concisely into a TableGen record, which will be expanded into
-an equivalent `mlir::Op` C++ template specialization at compiler build time.
+are specified concisely into a TableGen record, which will be expanded into an
+equivalent `mlir::Op` C++ template specialization at compiler build time.
 
 This manual explains in detail all the available mechanisms for defining
 operations in such a table-driven manner. It aims to be a specification instead
-of a tutorial. Please refer to [Quickstart tutorial to adding MLIR graph
-rewrite](Tutorials/QuickstartRewrites.md) for the latter.
+of a tutorial. Please refer to
+[Quickstart tutorial to adding MLIR graph rewrite](Tutorials/QuickstartRewrites.md)
+for the latter.
 
-In addition to detailing each mechanism, this manual also tries to capture
-best practices. They are rendered as quoted bullet points.
+In addition to detailing each mechanism, this manual also tries to capture best
+practices. They are rendered as quoted bullet points.
 
 ## Motivation
 
 MLIR allows pluggable dialects, and dialects contain, among others, a list of
 operations. This open and extensible ecosystem leads to the "stringly" type IR
 problem, e.g., repetitive string comparisons during optimization and analysis
 passes, unintuitive accessor methods (e.g., generic/error prone `getOperand(3)`
 vs self-documenting `getStride()`) with more generic return types, verbose and
-generic constructors without default arguments, verbose textual IR dump, and
-so on. Furthermore, operation verification is:
+generic constructors without default arguments, verbose textual IR dump, and so
+on. Furthermore, operation verification is:
 
-1. best case: a central string-to-verification-function map,
-1. middle case: duplication of verification across the code base, or
-1. worst case: no verification functions.
+1.  best case: a central string-to-verification-function map,
+1.  middle case: duplication of verification across the code base, or
+1.  worst case: no verification functions.
 
 The fix is to support defining ops in a table-driven manner. Then for each
 dialect, we can have a central place that contains everything you need to know
 about each op, including its constraints, custom assembly form, etc. This
 description is also used to generate helper functions and classes to allow
 building, verification, parsing, printing, analysis, and many more.
 
 ## Benefits
 
 Compared to the C++ template, this table-driven approach has several benefits
 including but not limited to:
 
-* **Single source of truth**: We strive to encode all facts regarding an
+*   **Single source of truth**: We strive to encode all facts regarding an
     operation into the record, so that readers don't need to jump among code
     snippets to fully understand an operation.
-* **Removing boilerplate**: We can automatically generate
+*   **Removing boilerplate**: We can automatically generate
     operand/attribute/result getter methods, operation build methods, operation
-  verify methods, and many more utilities from the record. This greatly reduces
-  the boilerplate needed for defining a new op.
-* **Facilitating auto-generation**: The usage of these operation information
+    verify methods, and many more utilities from the record. This greatly
+    reduces the boilerplate needed for defining a new op.
+*   **Facilitating auto-generation**: The usage of these operation information
     records are by no means limited to op definition itself. We can use them to
     drive the auto-generation of many other components, like computation graph
     serialization.
 
 ## TableGen Syntax
 
 We use TableGen as the language for specifying operation information. TableGen
 itself just provides syntax for writing records; the syntax and constructs
 allowed in a TableGen file (typically with filename suffix `.td`) can be found
 [here][TableGenProgRef].
 
window in `value`.
   let results = (outs
     TF_FpTensor:$output
   );
 
   TF_DerivedOperandTypeAttr T = TF_DerivedOperandTypeAttr<0>;
 }
 ```
 
-In the following we describe all the fields needed. Please see the definition
-of the `Op` class for the complete list of fields supported.
+In the following we describe all the fields needed. Please see the definition of
+the `Op` class for the complete list of fields supported.
 
 ### Operation name
 
 The operation name is a unique identifier of the operation within MLIR, e.g.,
 `tf.Add` for addition operation in the TensorFlow dialect. This is the
 equivalent of the mnemonic in assembly language. It is used for parsing and
 printing in the textual format. It is also used for pattern matching in graph
 rewrites.

 
 let description = [{
 ...
 }];
 ```
 
 `description` should be written in Markdown syntax.
 
-Placing the documentation at the beginning is recommended since
-it helps in understanding the operation.
+Placing the documentation at the beginning is recommended since it helps in
+understanding the operation.
 
-> * Place documentation at the beginning of the operation definition
-> * The summary should be short and concise. It should be a one-liner without
->   trailing punctuation. Put expanded explanation in description.
+> *   Place documentation at the beginning of the operation definition
+> *   The summary should be short and concise. It should be a one-liner without
+>     trailing punctuation. Put expanded explanation in description.
 
 ### Operation arguments
 
 There are two kinds of arguments: operands and attributes. Operands are runtime
 values produced by other ops; while attributes are compile-time known constant
 values, including two categories:
 
-1. Natural attributes: these attributes affect the behavior of the operations
+1.  Natural attributes: these attributes affect the behavior of the operations
     (e.g., padding for convolution);
-1. Derived attributes: these attributes are not needed to define the operation
+1.  Derived attributes: these attributes are not needed to define the operation
     but are instead derived from information of the operation. E.g., the output
     shape of type. This is mostly used for convenience interface generation or
     interaction with other frameworks/translation.
 
     All derived attributes should be materializable as an Attribute. That is,
-   even though they are not materialized, it should be possible to store as
-   an attribute.
+    even though they are not materialized, it should be possible to store as an
+    attribute.
 
 Both operands and attributes are specified inside the `dag`-typed `arguments`,
 led by `ins`:
 
 ```tablegen
 let arguments = (ins
   <type-constraint>:$<operand-name>,
   ...
   <attr-constraint>:$<attr-name>,
   ...
 );
 ```
 
 Here `<type-constraint>` is a TableGen `def` from the `TypeConstraint` class
 hierarchy. Similarly, `<attr-constraint>` is a TableGen `def` from the
 `AttrConstraint` class hierarchy. See [Constraints](#constraints) for more
 information.
 
 There is no requirements on the relative order of operands and attributes; they
 can mix freely. The relative order of operands themselves matters. From each
 named argument a named getter will be generated that returns the argument with
-the return type (in the case of attributes the return type will be
-constructed from the storage type, while for operands it will be `Value`). Each
-attribute's raw value (e.g., as stored) can also be accessed via generated
-`<name>Attr` getters for use in transformation passes where the more user
-friendly return type is less suitable.
+the return type (in the case of attributes the return type will be constructed
+from the storage type, while for operands it will be `Value`). Each attribute's
+raw value (e.g., as stored) can also be accessed via generated `<name>Attr`
+getters for use in transformation passes where the more user friendly return
+type is less suitable.
 
 All the arguments should be named to 1) provide documentation, 2) drive
 auto-generation of getter methods, 3) provide a handle to reference for other
 places like constraints.
 
 #### Variadic operands
 
 To declare a variadic operand, wrap the `TypeConstraint` for the operand with

 let results = (outs
   <type-constraint>:$<result-name>,
   ...
 );
 ```
 
 #### Variadic results
 
-Similar to variadic operands, `Variadic<...>` can also be used for results.
-And similarly, `SameVariadicResultSize` for multiple variadic results in the
-same operation.
+Similar to variadic operands, `Variadic<...>` can also be used for results. And
+similarly, `SameVariadicResultSize` for multiple variadic results in the same
+operation.
 
 ### Operation successors
 
 For terminator operations, the successors are specified inside of the
 `dag`-typed `successors`, led by `successor`:
 
 ```tablegen
 let successors = (successor
   <successor-constraint>:$<successor-name>,
   ...
 );
 ```
 
 #### Variadic successors
 
 Similar to the `Variadic` class used for variadic operands and results,
 `VariadicSuccessor<...>` can be used for successors. Variadic successors can
 currently only be specified as the last successor in the successor list.
 
 ### Operation traits and constraints
 
-Traits are operation properties that affect syntax or semantics. MLIR C++
-models various traits in the `mlir::OpTrait` namespace.
+Traits are operation properties that affect syntax or semantics. MLIR C++ models
+various traits in the `mlir::OpTrait` namespace.
 
 Both operation traits, [interfaces](Interfaces.md#utilizing-the-ods-framework),
 and constraints involving multiple operands/attributes/results are provided as
 the second template parameter to the `Op` class. They should be deriving from
 the `OpTrait` class. See [Constraints](#constraints) for more information.
 
 ### Builder methods
 

 
 ### Custom parser and printer methods
 
 Functions to parse and print the operation's custom assembly form.
 
 ### Custom verifier code
 
 Verification code will be automatically generated for
-[constraints](#constraints) specified on various entities of the op. To
-perform _additional_ verification, you can use
+[constraints](#constraints) specified on various entities of the op. To perform
+_additional_ verification, you can use
 
 ```tablegen
 let verifier = [{
   ...
 }];
 ```
 
 Code placed in `verifier` will be called after the auto-generated verification

 
 A custom directive has two main parts: The `UserDirective` and the `Params`. A
 custom directive is transformed into a call to a `print*` and a `parse*` method
 when generating the C++ code for the format. The `UserDirective` is an
 identifier used as a suffix to these two calls, i.e., `custom<MyDirective>(...)`
 would result in calls to `parseMyDirective` and `printMyDirective` within the
 parser and printer respectively. `Params` may be any combination of variables
 (i.e. Attribute, Operand, Successor, etc.), type directives, and `attr-dict`.
-The type directives must refer to a variable, but that variable need not also
-be a parameter to the custom directive.
+The type directives must refer to a variable, but that variable need not also be
+a parameter to the custom directive.
 
 The arguments to the `parse<UserDirective>` method are firstly a reference to
 the `OpAsmParser`(`OpAsmParser &`), and secondly a set of output parameters
 corresponding to the parameters specified in the format. The mapping of
 declarative parameter to `parse` method argument is detailed below:
 
 *   Attribute Variables
     -   Single: `<Attribute-Storage-Type>(e.g. Attribute) &`
*   TypeRef Directives
     -   Single: `Type`
     -   Optional: `Type`
     -   Variadic: `const SmallVectorImpl<Type> &`
 *   `attr-dict` Directive: `NamedAttrList &`
 
 When a variable is optional, the value should only be specified if the variable
 is present. Otherwise, the value should remain `None` or null.
 
-The arguments to the `print<UserDirective>` method is firstly a reference to
-the `OpAsmPrinter`(`OpAsmPrinter &`), second the op (e.g. `FooOp op` which
-can be `Operation *op` alternatively), and finally a set of output parameters
+The arguments to the `print<UserDirective>` method is firstly a reference to the
+`OpAsmPrinter`(`OpAsmPrinter &`), second the op (e.g. `FooOp op` which can be
+`Operation *op` alternatively), and finally a set of output parameters
 corresponding to the parameters specified in the format. The mapping of
 declarative parameter to `print` method argument is detailed below:
 
 *   Attribute Variables
     -   Single: `<Attribute-Storage-Type>(e.g. Attribute)`
     -   Optional: `<Attribute-Storage-Type>(e.g. Attribute)`
 *   Operand Variables
     -   Single: `Value`

 
 ##### Type Inference
 
 One requirement of the format is that the types of operands and results must
 always be present. In certain instances, the type of a variable may be deduced
 via type constraints or other information available. In these cases, the type of
 that variable may be elided from the format.
 
-* Buildable Types
+*   Buildable Types
 
-Some type constraints may only have one representation, allowing for them to
-be directly buildable; for example the `I32` or `Index` types. Types in `ODS`
-may mark themselves as buildable by setting the `builderCall` field or
-inheriting from the `BuildableType` class.
+Some type constraints may only have one representation, allowing for them to be
+directly buildable; for example the `I32` or `Index` types. Types in `ODS` may
+mark themselves as buildable by setting the `builderCall` field or inheriting
+from the `BuildableType` class.
 
-* Trait Equality Constraints
+*   Trait Equality Constraints
 
 There are many operations that have known type equality constraints registered
 as traits on the operation; for example the true, false, and result values of a
 `select` operation often have the same type. The assembly format may inspect
 these equal constraints to discern the types of missing variables. The currently
-supported traits are: `AllTypesMatch`, `TypesMatchWith`, `SameTypeOperands`,
-and `SameOperandsAndResultType`.
+supported traits are: `AllTypesMatch`, `TypesMatchWith`, `SameTypeOperands`, and
+`SameOperandsAndResultType`.
 
 ### `hasCanonicalizer`
 
 This boolean field indicate whether canonicalization patterns have been defined
 for this operation. If it is `1`, then `::getCanonicalizationPatterns()` should
 be defined.
 
 ### `hasFolder`
 
-This boolean field indicate whether general folding rules have been defined
-for this operation. If it is `1`, then `::fold()` should be defined.
+This boolean field indicate whether general folding rules have been defined for
+this operation. If it is `1`, then `::fold()` should be defined.
 
 ### Extra declarations
 
 One of the goals of table-driven op definition is to auto-generate as much logic
 and methods needed for each op as possible. With that said, there will always be
 long-tail cases that won't be covered. For such cases, you can use
 `extraClassDeclaration`. Code in `extraClassDeclaration` will be copied
 literally to the generated C++ op class.
 
-Note that `extraClassDeclaration` is a mechanism intended for long-tail cases
-by power users; for not-yet-implemented widely-applicable cases, improving the
+Note that `extraClassDeclaration` is a mechanism intended for long-tail cases by
+power users; for not-yet-implemented widely-applicable cases, improving the
 infrastructure is preferable.
 
 ### Generated C++ code
 
 [OpDefinitionsGen][OpDefinitionsGen] processes the op definition spec file and
 generates two files containing the corresponding C++ code: one for declarations,
 the other for definitions. The former is generated via the `-gen-op-decls`
 command-line option, while the latter is via the `-gen-op-defs` option.
 
 The definition file contains all the op method definitions, which can be
 included and enabled by defining `GET_OP_CLASSES`. For each operation,
 OpDefinitionsGen generates an operation class and an
 [operand adaptor](#operand-adaptors) class. Besides, it also contains a
 comma-separated list of all defined ops, which can be included and enabled by
 defining `GET_OP_LIST`.
 
 #### Class name and namespaces
 
 For each operation, its generated C++ class name is the symbol `def`ed with
-TableGen with dialect prefix removed. The first `_` serves as the delimiter.
-For example, for `def TF_AddOp`, the C++ class name would be `AddOp`.
-We remove the `TF` prefix because it is for scoping ops; other dialects
-may as well define their own `AddOp`s.
+TableGen with dialect prefix removed. The first `_` serves as the delimiter. For
+example, for `def TF_AddOp`, the C++ class name would be `AddOp`. We remove the
+`TF` prefix because it is for scoping ops; other dialects may as well define
+their own `AddOp`s.
 
 The namespaces of the generated C++ class will come from the dialect's
-`cppNamespace` field. For example, if a dialect's `cppNamespace` is `A::B`,
-then an op of that dialect will be placed in
-`namespace A { namespace B { ... } }`. If a dialect does not specify a
-`cppNamespace`, we then use the dialect's name as the namespace.
+`cppNamespace` field. For example, if a dialect's `cppNamespace` is `A::B`, then
+an op of that dialect will be placed in `namespace A { namespace B { ... } }`.
+If a dialect does not specify a `cppNamespace`, we then use the dialect's name
+as the namespace.
 
 This means the qualified name of the generated C++ class does not necessarily
 match exactly with the operation name as explained in
 [Operation name](#operation-name). This is to allow flexible naming to satisfy
 coding style requirements.
 
 #### Operand adaptors
 

 Constraint is a core concept in table-driven operation definition: operation
 verification and graph operation matching are all based on satisfying
 constraints. So both the operation definition and rewrite rules specification
 significantly involve writing constraints. We have the `Constraint` class in
 [`OpBase.td`][OpBase] has the common base class for all constraints.
 
 An operation's constraint can cover different range; it may
 
-* Only concern a single attribute (e.g. being a 32-bit integer greater than 5),
-* Multiple operands and results (e.g., the 1st result's shape must be the same
+*   Only concern a single attribute (e.g. being a 32-bit integer greater than
+    5),
+*   Multiple operands and results (e.g., the 1st result's shape must be the same
     as the 1st operand), or
-* Intrinsic to the operation itself (e.g., having no side effect).
+*   Intrinsic to the operation itself (e.g., having no side effect).
 
 We call them as single-entity constraint, multi-entity constraint, and traits,
 respectively.
 
 ### Single-entity constraint
 
 Constraints scoped to a single operand, attribute, or result are specified at
 the entity's declaration place as described in
 [Operation arguments](#operation-arguments) and
 [Operation results](#operation-results).
 
 To help modelling constraints of common types, a set of `TypeConstraint`s are
 created; they are the `Type` subclass hierarchy. It includes `F32` for the
-constraints of being a float, `TensorOf<[F32]>` for the constraints of being
-a float tensor, and so on.
+constraints of being a float, `TensorOf<[F32]>` for the constraints of being a
+float tensor, and so on.
 
 Similarly, a set of `AttrConstraint`s are created for helping modelling
 constraints of common attribute kinds. They are the `Attr` subclass hierarchy.
 It includes `F32Attr` for the constraints of being a float attribute,
 `F32ArrayAttr` for the constraints of being a float array attribute, and so on.
 
 ### Multi-entity constraint
 
-Constraints involving more than one operand/attribute/result are quite common
-on operations, like the element type and shape relation between operands and
+Constraints involving more than one operand/attribute/result are quite common on
+operations, like the element type and shape relation between operands and
 results. These constraints should be specified as the `Op` class template
 parameter as described in
 [Operation traits and constraints](#operation-traits-and-constraints).
 
 Multi-entity constraints are modeled as `PredOpTrait` (a subclass of `OpTrait`)
 in [`OpBase.td`][OpBase].A bunch of constraint primitives are provided to help
 specification. See [`OpBase.td`][OpBase] for the complete list.
 

 in a nested manner, using the two categories of predicates:
 
 1.  `CPred`: the primitive leaf predicate.
 2.  Compound predicate: a predicate composed from child predicates using
     predicate combiners (conjunction: `And`, disjunction: `Or`, negation: `Neg`,
     substitution: `SubstLeaves`, concatenation: `Concat`).
 
 `CPred` is the basis for composing more complex predicates. It is the "atom"
-predicate from the perspective of TableGen and the "interface" between
-TableGen and C++. What is inside is already C++ code, which will be treated
-as opaque strings with special placeholders to be substituted.
+predicate from the perspective of TableGen and the "interface" between TableGen
+and C++. What is inside is already C++ code, which will be treated as opaque
+strings with special placeholders to be substituted.
 
 You can put any C++ code that returns a boolean value inside a `CPred`,
-including evaluating expressions, calling functions, calling class methods,
-and so on.
+including evaluating expressions, calling functions, calling class methods, and
+so on.
 
 To help interaction with the C++ environment, there are a few special
 placeholders provided to refer to entities in the context where this predicate
-is used. They serve as "hooks" to the enclosing environment.  This includes
+is used. They serve as "hooks" to the enclosing environment. This includes
 `$_builder`, `$_op`, and `$_self`:
 
-* `$_builder` will be replaced by a `mlir::Builder` instance so that you can
+*   `$_builder` will be replaced by a `mlir::Builder` instance so that you can
     access common build methods.
-* `$_op` will be replaced by the current operation so that you can access
+*   `$_op` will be replaced by the current operation so that you can access
     information of the current operation.
-* `$_self` will be replaced with the entity this predicate is attached to.
+*   `$_self` will be replaced with the entity this predicate is attached to.
     E.g., `BoolAttr` is an attribute constraint that wraps a
     `CPred<"$_self.isa<BoolAttr>()">`. Then for `F32:$attr`,`$_self` will be
     replaced by `$attr`. For type constraints, it's a little bit special since
     we want the constraints on each type definition reads naturally and we want
     to attach type constraints directly to an operand/result, `$_self` will be
-  replaced by the operand/result's type. E.g., for `F32` in `F32:$operand`, its
-  `$_self` will be expanded as `getOperand(...).getType()`.
+    replaced by the operand/result's type. E.g., for `F32` in `F32:$operand`,
+    its `$_self` will be expanded as `getOperand(...).getType()`.
 
 TODO: Reconsider the leading symbol for special placeholders. Eventually we want
 to allow referencing operand/result $-names; such $-names can start with
 underscore.
 
 For example, to write an attribute `attr` is an `IntegerAttr`, in C++ you can
 just call `attr.isa<IntegerAttr>()`. The code can be wrapped in a `CPred` as
 `$_self.isa<IntegerAttr>()`, with `$_self` as the special placeholder to be
 replaced by the current attribute `attr` at expansion time.
 
-For more complicated predicates, you can wrap it in a single `CPred`, or you
-can use predicate combiners to combine them. For example, to write the
-constraint that an attribute `attr` is a 32-bit or 64-bit integer, you can
-write it as
+For more complicated predicates, you can wrap it in a single `CPred`, or you can
+use predicate combiners to combine them. For example, to write the constraint
+that an attribute `attr` is a 32-bit or 64-bit integer, you can write it as
 
 ```tablegen
 And<[
   CPred<"$_self.isa<IntegerAttr>()">,
   Or<[
     CPred<"$_self.cast<IntegerAttr>().getType().isInteger(32)">,
     CPred<"$_self.cast<IntegerAttr>().getType().isInteger(64)">
   ]>
 ]>
 ```
 
 (Note that the above is just to show with a familiar example how you can use
 `CPred` and predicate combiners to write complicated predicates. For integer
 attributes specifically, [`OpBase.td`][OpBase] already defines `I32Attr` and
 `I64Attr`. So you can actually reuse them to write it as `Or<[I32Attr.predicate,
 I64Attr.predicate]>`.)
 
 TODO: Build up a library of reusable primitive constraints
 
 If the predicate is very complex to write with `CPred` together with predicate
-combiners, you can also write it as a normal C++ function and use the `CPred`
-as a way to "invoke" the function. For example, to verify an attribute `attr`
-has some property, you can write a C++ function like
+combiners, you can also write it as a normal C++ function and use the `CPred` as
+a way to "invoke" the function. For example, to verify an attribute `attr` has
+some property, you can write a C++ function like
 
 ```cpp
 bool HasSomeProperty(Attribute attr) { ... }
 ```
 
 and then define the op as:
 
 ```tablegen
 def HasSomeProperty : AttrConstraint<CPred<"HasSomeProperty($_self)">,
                                      "has some property">;
 
 def MyOp : Op<...> {
   let arguments = (ins
     ...
     HasSomeProperty:$attr
   );
 }
 ```
 
-As to whether we should define the predicate using a single `CPred` wrapping
-the whole expression, multiple `CPred`s with predicate combiners, or a single
+As to whether we should define the predicate using a single `CPred` wrapping the
+whole expression, multiple `CPred`s with predicate combiners, or a single
 `CPred` "invoking" a function, there are no clear-cut criteria. Defining using
 `CPred` and predicate combiners is preferable since it exposes more information
 (instead hiding all the logic behind a C++ function) into the op definition spec
-so that it can potentially drive more auto-generation cases. But it will
-require a nice library of common predicates as the building blocks to avoid the
+so that it can potentially drive more auto-generation cases. But it will require
+a nice library of common predicates as the building blocks to avoid the
 duplication, which is being worked on right now.
 
 ## Attribute Definition
 
 An attribute is a compile-time known constant of an operation.
 
 ODS provides attribute wrappers over C++ attribute classes. There are a few
 common C++ [attribute classes][AttrClasses] defined in MLIR's core IR library
-and one is free to define dialect-specific attribute classes. ODS allows one
-to use these attributes in TableGen to define operations, potentially with
-more fine-grained constraints. For example, `StrAttr` directly maps to
-`StringAttr`; `F32Attr`/`F64Attr` requires the `FloatAttr` to additionally
-be of a certain bitwidth.
+and one is free to define dialect-specific attribute classes. ODS allows one to
+use these attributes in TableGen to define operations, potentially with more
+fine-grained constraints. For example, `StrAttr` directly maps to `StringAttr`;
+`F32Attr`/`F64Attr` requires the `FloatAttr` to additionally be of a certain
+bitwidth.
 
 ODS attributes are defined as having a storage type (corresponding to a backing
 `mlir::Attribute` that _stores_ the attribute), a return type (corresponding to
 the C++ _return_ type of the generated of the helper getters) as well as method
 to convert between the internal storage and the helper method.
 
 ### Attribute decorators
 

 -   Set `hasCustomStorageConstructor` to generate the TypeStorage class with a
     constructor which is just declared -- no definition -- so you can write it
     yourself.
 -   Use the `TypeParameter` tablegen class instead of the "c++Type" string.
 
 ### TypeParameter tablegen class
 
 This is used to further specify attributes about each of the types parameters.
-It includes documentation (`description` and `syntax`), the C++ type to use, and
-a custom allocator to use in the storage constructor method.
+It includes documentation (`summary` and `syntax`), the C++ type to use, and a
+custom allocator to use in the storage constructor method.
 
 ```tablegen
 // DO NOT DO THIS!
 let parameters = (ins "ArrayRef<int>":$dims);
 ```
 
 The default storage constructor blindly copies fields by value. It does not know
 anything about the types. In this case, the ArrayRef<int> requires allocation
-   The `storageNamespace` field is used to set the namespace where the storage
     class should sit. Defaults to "detail".
 -   The `extraClassDeclaration` field is used to include extra code in the class
     declaration.
 
 ## Debugging Tips
 
 ### Run `mlir-tblgen` to see the generated content
 
-TableGen syntax sometimes can be obscure; reading the generated content can be
-a very helpful way to understand and debug issues. To build `mlir-tblgen`, run
+TableGen syntax sometimes can be obscure; reading the generated content can be a
+very helpful way to understand and debug issues. To build `mlir-tblgen`, run
 `cmake --build . --target mlir-tblgen` in your build directory and find the
 `mlir-tblgen` binary in the `bin/` subdirectory. All the supported generators
 can be found via `mlir-tblgen --help`. For example, `--gen-op-decls` and
 `--gen-op-defs` as explained in [Generated C++ code](#generated-c++-code).
 
 To see the generated code, invoke `mlir-tblgen` with a specific generator by
 providing include paths via `-I`. For example,
 

 ```
 
 ## Appendix
 
 ### Requirements and existing mechanisms analysis
 
 The op description should as declarative as possible to allow a wide range of
 tools to work with them and query methods generated from them. In particular
-this means specifying traits, constraints and shape inference information in
-a way that is easily analyzable (e.g., avoid opaque calls to C++ functions where
+this means specifying traits, constraints and shape inference information in a
+way that is easily analyzable (e.g., avoid opaque calls to C++ functions where
 possible).
 
 We considered the approaches of several contemporary systems and focused on
 requirements that were desirable:
 
 *   Ops registered using a registry separate from C++ code.
     *   Unknown ops are allowed in MLIR, so ops need not be registered. The
         ability of the compiler to optimize those ops or graphs containing those

mlir/include/mlir/Dialect/ArmSVE/ArmSVE.td

 //===----------------------------------------------------------------------===//
 // ArmSVE type definitions
 //===----------------------------------------------------------------------===//
 
 def ArmSVE_ScalableVectorType : DialectType<ArmSVE_Dialect,
     CPred<"$_self.isa<ScalableVectorType>()">,
               "scalable vector type">,
     BuildableType<"$_builder.getType<ScalableVectorType>()"> {
-  let typeDescription = [{
+  let description = [{
     `arm_sve.vector` represents vectors that will be processed by a scalable
     vector architecture.
   }];
 }
 
 class ArmSVE_Type<string name> : TypeDef<ArmSVE_Dialect, name> { }
 
 def ScalableVectorType : ArmSVE_Type<"ScalableVector"> {

 class ScalableVectorOfLength<list<int> allowedLengths> : Type<
   IsScalableVectorOfLengthPred<allowedLengths>,
   " of length " # !interleave(allowedLengths, "/")>;
 
 class ScalableVectorOfLengthAndType<list<int> allowedLengths,
                                     list<Type> allowedTypes> : Type<
   And<[ScalableVectorOf<allowedTypes>.predicate,
        ScalableVectorOfLength<allowedLengths>.predicate]>,
-  ScalableVectorOf<allowedTypes>.description #
-  ScalableVectorOfLength<allowedLengths>.description>;
+  ScalableVectorOf<allowedTypes>.summary #
+  ScalableVectorOfLength<allowedLengths>.summary>;
 
 //===----------------------------------------------------------------------===//
 // ArmSVE op definitions
 //===----------------------------------------------------------------------===//
 
 class ArmSVE_Op<string mnemonic, list<OpTrait> traits = []> :
   Op<ArmSVE_Dialect, mnemonic, traits> {}
 

mlir/include/mlir/Dialect/Async/IR/AsyncBase.td

def AsyncDialect : Dialect {
   }];
 
   let cppNamespace = "::mlir::async";
 }
 
 def Async_TokenType : DialectType<AsyncDialect,
     CPred<"$_self.isa<::mlir::async::TokenType>()">, "token type">,
     BuildableType<"$_builder.getType<::mlir::async::TokenType>()"> {
-  let typeDescription = [{
+  let description = [{
     `async.token` is a type returned by asynchronous operations, and it becomes
     `ready` when the asynchronous operations that created it is completed.
   }];
 }
 
 class Async_ValueType<Type type>
     : DialectType<AsyncDialect,
         And<[
           CPred<"$_self.isa<::mlir::async::ValueType>()">,
           SubstLeaves<"$_self",
                       "$_self.cast<::mlir::async::ValueType>().getValueType()",
                       type.predicate>
-       ]>, "async value type with " # type.description # " underlying type"> {
-  let typeDescription = [{
+       ]>, "async value type with " # type.summary # " underlying type"> {
+  let description = [{
     `async.value` represents a value returned by asynchronous operations,
     which may or may not be available currently, but will be available at some
     point in the future.
   }];
 
   Type valueType = type;
 }
 
 def Async_GroupType : DialectType<AsyncDialect,
     CPred<"$_self.isa<::mlir::async::GroupType>()">, "group type">,
     BuildableType<"$_builder.getType<::mlir::async::GroupType>()"> {
-  let typeDescription = [{
+  let description = [{
     `async.group` represent a set of async tokens or values and allows to
     execute async operations on all of them together (e.g. wait for the
     completion of all/any of them).
   }];
 }
 
 def Async_AnyValueType : DialectType<AsyncDialect,
                            CPred<"$_self.isa<::mlir::async::ValueType>()">,

mlir/include/mlir/Dialect/LLVMIR/LLVMOpBase.td

 // Type constraint accepting LLVM pointer type with an additional constraint
 // on the element type.
 class LLVM_PointerTo<Type pointee> : Type<
   And<[LLVM_AnyPointer.predicate,
        SubstLeaves<
          "$_self",
          "$_self.cast<::mlir::LLVM::LLVMPointerType>().getElementType()",
          pointee.predicate>]>,
-  "LLVM pointer to " # pointee.description>;
+  "LLVM pointer to " # pointee.summary>;
 
 // Type constraint accepting any LLVM structure type.
 def LLVM_AnyStruct : Type<CPred<"$_self.isa<::mlir::LLVM::LLVMStructType>()">,
                          "LLVM structure type">;
 
 // Type constraint accepting opaque LLVM structure type.
 def LLVM_OpaqueStruct : Type<
   And<[LLVM_AnyStruct.predicate,

 // Type constraint accepting an LLVM vector type with an additional constraint
 // on the vector element type.
 class LLVM_VectorOf<Type element> : Type<
   And<[LLVM_AnyVector.predicate,
        SubstLeaves<
          "$_self",
          "$_self.cast<::mlir::LLVM::LLVMVectorType>().getElementType()",
          element.predicate>]>,
-  "LLVM vector of " # element.description>;
+  "LLVM vector of " # element.summary>;
 
 // Type constraint accepting a constrained type, or a vector of such types.
 class LLVM_ScalarOrVectorOf<Type element> :
     AnyTypeOf<[element, LLVM_VectorOf<element>]>;
 
 // Base class for LLVM operations. Defines the interface to the llvm::IRBuilder
 // used to translate to LLVM IR proper.
 class LLVM_OpBase<Dialect dialect, string mnemonic, list<OpTrait> traits = []> :

mlir/include/mlir/Dialect/Quant/QuantOpsBase.td

 //===----------------------------------------------------------------------===//
 // Quantization type definitions
 //===----------------------------------------------------------------------===//
 
 class quant_TypedPrimitiveOrContainer<Type etype> :
     Type<Or<[etype.predicate,
                 TensorOf<[etype]>.predicate,
                 VectorOf<[etype]>.predicate]>,
-         "primitive/tensor/vector of " # etype.description>;
+         "primitive/tensor/vector of " # etype.summary>;
 
 // An implementation of QuantizedType.
 def quant_QuantizedType :
     Type<CPred<"$_self.isa<mlir::quant::QuantizedType>()">, "QuantizedType">;
 
 // A primitive type that can represent a real value. This is either a
 // floating point value or a quantized type.
 def quant_RealPrimitiveType :

mlir/include/mlir/Dialect/Shape/IR/ShapeBase.td

def ShapeDialect : Dialect {
 
   let hasConstantMaterializer = 1;
   let hasOperationAttrVerify = 1;
 }
 
 def Shape_ShapeType : DialectType<ShapeDialect,
     CPred<"$_self.isa<::mlir::shape::ShapeType>()">, "shape">,
     BuildableType<"$_builder.getType<::mlir::shape::ShapeType>()"> {
-  let typeDescription = [{
+  let description = [{
     `shape.shape` represents either an unranked shape, a ranked shape with
     possibly unknown dimensions or an invalid shape. The rank is of type
     `shape.size` and, if rank is known, the extent is a 1D tensor of type
     `shape.size`.
 
     Shape is printed:
     * `[*]` if it is an unranked shape
     * `[?, 2]` if a rank 2 tensor with one unknown dimension
     * `[3, 4]` is a rank 2 static tensor
     * `[]` is a scalar
     * `[1]` is a rank 1 tensor with 1 element
     * `[invalid]` for an invalid shape
   }];
 }
 
 def Shape_SizeType : DialectType<ShapeDialect,
     CPred<"$_self.isa<::mlir::shape::SizeType>()">, "size">,
     BuildableType<"$_builder.getType<::mlir::shape::SizeType>()"> {
-  let typeDescription = [{
+  let description = [{
     `shape.size` represents a non-negative integer with support for being
     unknown and invalid.
 
     Operations on `shape.size` types are specialized to handle unknown/dynamic
     value. So, for example, `<unknown> + x == <unknown>` for all non-error `x :
     !shape.size` (e.g., an unknown value does not become known due to addition).
   }];
 }
 
 def Shape_ValueShapeType : DialectType<ShapeDialect,
     CPred<"$_self.isa<::mlir::shape::ValueShapeType>()">, "value shape">,
     BuildableType<"::mlir::shape::ValueShapeType::get($_builder.getContext())">
 {
-  let typeDescription = [{
+  let description = [{
     `shape.value_shape` represents the value produced by an operation (this
     corresponds to `Value` in the compiler) and a shape. Conceptually this is a
     tuple of a value (potentially unknown) and `shape.shape`. The value and
     shape can either or both be unknown. If both the `value` and `shape` are
     known, then the shape of `value` is conformant with `shape`. That is, the
     shape of the value conforms to the shape of the ValueShape, so that if we
     have ValueShape `(value, shape)` then `join(shape_of(value), shape)` would
     be error free and in particular it means that if both are statically known,
     then they are equal.
   }];
 }
 
 def Shape_ExtentTensorType :
     1DTensorOf<[Index]>,
     BuildableType<"::mlir::RankedTensorType::get({ShapedType::kDynamicSize}, "
                   "$_builder.getType<::mlir::IndexType>())"> {
-  let typeDescription = [{
+  let description = [{
     The extent tensor is a tensor of rank one with arbitrarily many index
     elements (tensor<?xindex>). Like `!shape.shape`, it is used to represent
     shapes with the difference that it is guaranteed to be error-free.
   }];
 }
 
 def Shape_ShapeOrSizeType : AnyTypeOf<[Shape_SizeType, Shape_ShapeType],
   "shape or size">;
 
 def Shape_ShapeOrExtentTensorType : AnyTypeOf<[Shape_ShapeType,
                                                Shape_ExtentTensorType],
                                               "shape or extent tensor">;
 
 def Shape_SizeOrIndexType : AnyTypeOf<[Shape_SizeType, Index], "size or index">;
 
 def Shape_WitnessType : DialectType<ShapeDialect,
     CPred<"$_self.isa<::mlir::shape::WitnessType>()">, "witness">,
     BuildableType<"$_builder.getType<::mlir::shape::WitnessType>()"> {
-  let typeDescription = [{
+  let description = [{
     A witness is a structural device in the compiler to maintain ordering of
     code relying on information obtained from passing assertions. Witnesses do
     not represent any physical data.
 
     "cstr_" operations will return witnesses and be lowered into assertion logic
     when not resolvable at compile time.
 
     "assuming_" operations will take witnesses as input and represent only

mlir/include/mlir/Dialect/Tosa/IR/TosaOpBase.td

 // This quantization attribute expresses numerical behavior of operators where
 // the operator has a numerical relationship between a single input and output.
 // For example: tosa.negate.
 def Tosa_UnaryOpQuantizationAttr : StructAttr<"UnaryOpQuantizationAttr",
   Tosa_Dialect, [
     StructFieldAttr<"input_zp",         I32Attr>,
     StructFieldAttr<"output_zp",        I32Attr>
     ]> {
-  let description = "Attribute for UnaryOp quantization information.";
+  let summary = "Attribute for UnaryOp quantization information.";
 }
 
 // There is no explicit BinaryOpQuantizationAttr for 2-input/1-output ops. In
 // this case, a tosa.rescale is used to express the inputs to the same scale.
 // TODO: Upload WIP legalization document describing this construction by
 // example.
 
 // This quantization attribute holds input and weight zero point. Both the
 // ConvOp and MatMulOp QuantizationAttrs follow a common design semantic where
 // their ownquantization attribute only expresses the numerical behavior at
 // the inputs.
 // The scaling of their accumulator output is done using an explicit
 // tosa.rescale operator that scales the accumulator result to output scale.
 def Tosa_ConvOpQuantizationAttr : StructAttr<"ConvOpQuantizationAttr",
   Tosa_Dialect, [
     StructFieldAttr<"input_zp",         I32Attr>,
     StructFieldAttr<"weight_zp",        I32Attr>
     ]> {
-  let description = "Attribute for Conv type op quantization information.";
+  let summary = "Attribute for Conv type op quantization information.";
 }
 
 def Tosa_MatMulOpQuantizationAttr : StructAttr<"MatMulOpQuantizationAttr",
   Tosa_Dialect, [
     StructFieldAttr<"a_zp",             I32Attr>,
     StructFieldAttr<"b_zp",             I32Attr>
     ]> {
-  let description = "Attribute for MatMulOp quantization information.";
+  let summary = "Attribute for MatMulOp quantization information.";
 }
 
 // This attribute holds input zero point correction applied to the padding
 // zeros to ensure numerical accuracy in the subsequent TOSA operations.
 // Its functional application is described in the tosa.pad() operator
 // description in the specification.
 def Tosa_PadOpQuantizationAttr : StructAttr<"PadOpQuantizationAttr",
   Tosa_Dialect, [
     StructFieldAttr<"input_zp",         I32Attr>
     ]> {
-  let description = "Attribute for PadOp quantization information.";
+  let summary = "Attribute for PadOp quantization information.";
 }
 
 //===----------------------------------------------------------------------===//
 // TOSA Operator Quantization Builders.
 //===----------------------------------------------------------------------===//
 
 // This builder is called on all convolution operators except for TransposeConv,
 // which has specialized output shape semantics. The builder also defines the

mlir/include/mlir/IR/OpBase.td

 // this class are used to generate verification code in op verifier, and
 // matching code in pattern matcher.
 //
 // Constraints are predicates with descriptive names, to facilitate inspection,
 // provide nice error messages, etc.
 class Constraint<Pred pred, string desc = ""> {
   // The predicates that this constraint requires.
   Pred predicate = pred;
-  // User-readable description used in error reporting messages. If empty, a
-  // generic message will be used.
-  string description = desc;
+  // User-readable one line summary used in error reporting messages. If empty,
+  // a generic message will be used.
+  string summary = desc;
 }
 
 // Subclasses used to differentiate different constraint kinds. These are used
 // as markers for the TableGen backend to handle different constraint kinds
 // differently if needed. Constraints not deriving from the following subclasses
 // are considered as uncategorized constraints.
 
 // Subclass for constraints on a type.
-class TypeConstraint<Pred predicate, string description = "",
+class TypeConstraint<Pred predicate, string summary = "",
                      string cppClassNameParam = "::mlir::Type"> :
-    Constraint<predicate, description> {
+    Constraint<predicate, summary> {
   // The name of the C++ Type class if known, or Type if not.
   string cppClassName = cppClassNameParam;
 }
 
 // Subclass for constraints on an attribute.
-class AttrConstraint<Pred predicate, string description = ""> :
-    Constraint<predicate, description>;
+class AttrConstraint<Pred predicate, string summary = ""> :
+    Constraint<predicate, summary>;
 
 // Subclass for constraints on a region.
-class RegionConstraint<Pred predicate, string description = ""> :
-    Constraint<predicate, description>;
+class RegionConstraint<Pred predicate, string summary = ""> :
+    Constraint<predicate, summary>;
 
 // Subclass for constraints on a successor.
-class SuccessorConstraint<Pred predicate, string description = ""> :
-    Constraint<predicate, description>;
+class SuccessorConstraint<Pred predicate, string summary = ""> :
+    Constraint<predicate, summary>;
 
 // How to use these constraint categories:
 //
 // * Use TypeConstraint to specify
 //   * Constraints on an op's operand/result definition
 //   * Further constraints to match an op's operand/result in source pattern
 //
 // * Use Attr (a subclass for AttrConstraint) for

 //===----------------------------------------------------------------------===//
 // Type definitions
 //===----------------------------------------------------------------------===//
 
 // A type, carries type constraints.
 class Type<Pred condition, string descr = "",
            string cppClassName = "::mlir::Type"> :
     TypeConstraint<condition, descr, cppClassName> {
-  string typeDescription = "";
+  string description = "";
   string builderCall = "";
 }
 
-// Allows providing an alternative name and description to an existing type def.
-class TypeAlias<Type t, string description = t.description> :
-    Type<t.predicate, description> {
-  let typeDescription = t.typeDescription;
+// Allows providing an alternative name and summary to an existing type def.
+class TypeAlias<Type t, string summary = t.summary> :
+    Type<t.predicate, summary> {
+  let description = t.description;
   let builderCall = t.builderCall;
 }
 
 // A type of a specific dialect.
 class DialectType<Dialect d, Pred condition, string descr = "",
                   string cppClassName = "::mlir::Type"> :
     Type<condition, descr, cppClassName> {
   Dialect dialect = d;
 }
 
 // A variadic type constraint. It expands to zero or more of the base type. This
 // class is used for supporting variadic operands/results.
-class Variadic<Type type> : TypeConstraint<type.predicate, type.description> {
+class Variadic<Type type> : TypeConstraint<type.predicate, type.summary> {
   Type baseType = type;
 }
 
 // An optional type constraint. It expands to either zero or one of the base
 // type. This class is used for supporting optional operands/results.
-class Optional<Type type> : TypeConstraint<type.predicate, type.description> {
+class Optional<Type type> : TypeConstraint<type.predicate, type.summary> {
   Type baseType = type;
 }
 
 // A type that can be constructed using MLIR::Builder.
 // Note that this does not "inherit" from Type because it would require
 // duplicating Type subclasses for buildable and non-buildable cases to avoid
 // diamond "inheritance".
 // TODO: we may extend this to a more general 'Buildable' trait, making some

 def AnyType : Type<CPred<"true">, "any type">;
 
 // None type
 def NoneType : Type<CPred<"$_self.isa<::mlir::NoneType>()">, "none type",
                     "::mlir::NoneType">,
       BuildableType<"$_builder.getType<::mlir::NoneType>()">;
 
 // Any type from the given list
-class AnyTypeOf<list<Type> allowedTypes, string description = "",
+class AnyTypeOf<list<Type> allowedTypes, string summary = "",
                 string cppClassName = "::mlir::Type"> : Type<
     // Satisfy any of the allowed type's condition
     Or<!foreach(allowedtype, allowedTypes, allowedtype.predicate)>,
-    !if(!eq(description, ""),
-        !interleave(!foreach(t, allowedTypes, t.description), " or "),
-        description)>;
+    !if(!eq(summary, ""),
+        !interleave(!foreach(t, allowedTypes, t.summary), " or "),
+        summary)>;
 
 // Integer types.
 
 // Any integer type irrespective of its width and signedness semantics.
 def AnyInteger : Type<CPred<"$_self.isa<::mlir::IntegerType>()">, "integer",
                       "::mlir::IntegerType">;
 
 // Any integer type (regardless of signedness semantics) of a specific width.
def BF16 : Type<CPred<"$_self.isBF16()">, "bfloat16 type">,
            BuildableType<"$_builder.getBF16Type()">;
 
 class Complex<Type type>
     : Type<And<[
           CPred<"$_self.isa<::mlir::ComplexType>()">,
           SubstLeaves<"$_self",
                       "$_self.cast<::mlir::ComplexType>().getElementType()",
            type.predicate>]>,
-           "complex type with " # type.description # " elements",
+           "complex type with " # type.summary # " elements",
            "::mlir::ComplexType">,
       SameBuildabilityAs<type, "::mlir::ComplexType::get($_builder.get" # type #
                                "Type())"> {
   Type elementType = type;
 }
 
 def AnyComplex : Type<CPred<"$_self.isa<::mlir::ComplexType>()">,
                       "complex-type", "::mlir::ComplexType">;
 
-class OpaqueType<string dialect, string name, string description>
+class OpaqueType<string dialect, string name, string summary>
   : Type<CPred<"isOpaqueTypeWithName($_self, \""#dialect#"\", \""#name#"\")">,
-         description, "::mlir::OpaqueType">,
+         summary, "::mlir::OpaqueType">,
     BuildableType<"::mlir::OpaqueType::get($_builder.getContext(), "
                   "$_builder.getIdentifier(\"" # dialect # "\"), \""
                   # name # "\")">;
 
 // Function Type
 
 // Any function type.
 def FunctionType : Type<CPred<"$_self.isa<::mlir::FunctionType>()">,
                               "function type", "::mlir::FunctionType">;
 
 // A container type is a type that has another type embedded within it.
 class ContainerType<Type etype, Pred containerPred, code elementTypeCall,
                     string descr, string cppClassName = "::mlir::Type"> :
     // First, check the container predicate.  Then, substitute the extracted
     // element into the element type checker.
     Type<And<[containerPred,
                 SubstLeaves<"$_self", !cast<string>(elementTypeCall),
                 etype.predicate>]>,
-         descr # " of " # etype.description # " values", cppClassName> {
+         descr # " of " # etype.summary # " values", cppClassName> {
   // The type of elements in the container.
   Type elementType = etype;
 
   // Call to retrieve.
   code getElementTypeCall = elementTypeCall;
 }
 
 class ShapedContainerType<list<Type> allowedTypes,
class VectorOfRank<list<int> allowedRanks> : Type<
   " of ranks " # !interleave(allowedRanks, "/"), "::mlir::VectorType">;
 
 // Any vector where the rank is from the given `allowedRanks` list and the type
 // is from the given `allowedTypes` list
 class VectorOfRankAndType<list<int> allowedRanks,
                           list<Type> allowedTypes> : Type<
   And<[VectorOf<allowedTypes>.predicate,
        VectorOfRank<allowedRanks>.predicate]>,
-  VectorOf<allowedTypes>.description #
-  VectorOfRank<allowedRanks>.description,
+  VectorOf<allowedTypes>.summary # VectorOfRank<allowedRanks>.summary,
   "::mlir::VectorType">;
 
 // Whether the number of elements of a vector is from the given
 // `allowedLengths` list
 class IsVectorOfLengthPred<list<int> allowedLengths> :
   And<[IsVectorTypePred,
        Or<!foreach(allowedlength, allowedLengths,
                    CPred<[{$_self.cast<::mlir::VectorType>().getNumElements()

 
 // Any vector where the number of elements is from the given
 // `allowedLengths` list and the type is from the given `allowedTypes`
 // list
 class VectorOfLengthAndType<list<int> allowedLengths,
                             list<Type> allowedTypes> : Type<
   And<[VectorOf<allowedTypes>.predicate,
        VectorOfLength<allowedLengths>.predicate]>,
-  VectorOf<allowedTypes>.description #
-  VectorOfLength<allowedLengths>.description,
+  VectorOf<allowedTypes>.summary # VectorOfLength<allowedLengths>.summary,
   "::mlir::VectorType">;
 
 def AnyVector : VectorOf<[AnyType]>;
 
 // Shaped types.
 
 def AnyShaped: ShapedContainerType<[AnyType], IsShapedTypePred, "shaped",
                                    "::mlir::ShapedType">;

 
 def AnyRankedTensor :
   ShapedContainerType<[AnyType], And<[IsTensorTypePred, HasRankPred]>,
   "ranked tensor", "::mlir::TensorType">;
 
 // TODO: Have an easy way to add another constraint to a type.
 class StaticShapeTensorOf<list<Type> allowedTypes>
     : Type<And<[TensorOf<allowedTypes>.predicate, HasStaticShapePred]>,
-           "statically shaped " # TensorOf<allowedTypes>.description,
+           "statically shaped " # TensorOf<allowedTypes>.summary,
            "::mlir::TensorType">;
 
 def AnyStaticShapeTensor : StaticShapeTensorOf<[AnyType]>;
 
 def I1Tensor   : TensorOf<[I1]>;
 def I8Tensor   : TensorOf<[I8]>;
 def I16Tensor  : TensorOf<[I16]>;
 def I32Tensor  : TensorOf<[I32]>;
 def I64Tensor  : TensorOf<[I64]>;
 def IndexTensor: TensorOf<[Index]>;
 
 def BF16Tensor : TensorOf<[BF16]>;
 def F16Tensor  : TensorOf<[F16]>;
 def F32Tensor  : TensorOf<[F32]>;
 def F64Tensor  : TensorOf<[F64]>;
 
 // Ranked tensor type with one of the specified types and ranks.
 class TensorRankOf<list<Type> allowedTypes, list<int> ranks> :
     Type<And<[TensorOf<allowedTypes>.predicate, HasAnyRankOfPred<ranks>]>,
          !interleave(!foreach(rank, ranks, rank # "D"), "/") # " " #
-         TensorOf<allowedTypes>.description, "::mlir::TensorType">;
+         TensorOf<allowedTypes>.summary, "::mlir::TensorType">;
 
 class 0DTensorOf<list<Type> allowedTypes> : TensorRankOf<allowedTypes, [0]>;
 class 1DTensorOf<list<Type> allowedTypes> : TensorRankOf<allowedTypes, [1]>;
 class 2DTensorOf<list<Type> allowedTypes> : TensorRankOf<allowedTypes, [2]>;
 class 3DTensorOf<list<Type> allowedTypes> : TensorRankOf<allowedTypes, [3]>;
 class 4DTensorOf<list<Type> allowedTypes> : TensorRankOf<allowedTypes, [4]>;
 
 // Unranked Memref type

 def F16MemRef  : MemRefOf<[F16]>;
 def F32MemRef  : MemRefOf<[F32]>;
 def F64MemRef  : MemRefOf<[F64]>;
 
 // TODO: Have an easy way to add another constraint to a type.
 class MemRefRankOf<list<Type> allowedTypes, list<int> ranks> :
     Type<And<[MemRefOf<allowedTypes>.predicate, HasAnyRankOfPred<ranks>]>,
          !interleave(!foreach(rank, ranks, rank # "D"), "/") # " " #
-         MemRefOf<allowedTypes>.description>;
+         MemRefOf<allowedTypes>.summary>;
 
 class StaticShapeMemRefOf<list<Type> allowedTypes>
     : Type<And<[MemRefOf<allowedTypes>.predicate, HasStaticShapePred]>,
-           "statically shaped " # MemRefOf<allowedTypes>.description>;
+           "statically shaped " # MemRefOf<allowedTypes>.summary>;
 
 def AnyStaticShapeMemRef : StaticShapeMemRefOf<[AnyType]>;
 
 // For a MemRefType, verify that it has strides.
 def HasStridesPred : CPred<[{ isStrided($_self.cast<::mlir::MemRefType>()) }]>;
 
 class StridedMemRefOf<list<Type> allowedTypes>
     : Type<And<[MemRefOf<allowedTypes>.predicate, HasStridesPred]>,
-           "strided " # MemRefOf<allowedTypes>.description>;
+           "strided " # MemRefOf<allowedTypes>.summary>;
 
 def AnyStridedMemRef : StridedMemRefOf<[AnyType]>;
 
 class AnyStridedMemRefOfRank<int rank> :
   Type<And<[AnyStridedMemRef.predicate,
             MemRefRankOf<[AnyType], [rank]>.predicate]>,
-       AnyStridedMemRef.description # " of rank " # rank>;
+       AnyStridedMemRef.summary # " of rank " # rank>;
 
 class StridedMemRefRankOf<list<Type> allowedTypes, list<int> ranks> :
     Type<And<[MemRefOf<allowedTypes>.predicate, HasAnyRankOfPred<ranks>]>,
          !interleave(!foreach(rank, ranks, rank # "D"), "/") # " " #
-         MemRefOf<allowedTypes>.description>;
+         MemRefOf<allowedTypes>.summary>;
 
 // This represents a generic tuple without any constraints on element type.
 def AnyTuple : Type<IsTupleTypePred, "tuple", "::mlir::TupleType">;
 
 // A container type that has other types embedded in it, but (unlike
 // ContainerType) can hold elements with a mix of types. Requires a call that
 // produces a list of all elements' types.
 class MixedContainerType<Type etype, Pred containerPred, code elementTypesCall,
                          string descr> :
     Type<
         And<[
             containerPred,
             Concat<
                 "::llvm::all_of(" # elementTypesCall # ", [](Type t) { return ",
                 SubstLeaves<"$_self", "t", etype.predicate>,
                 "; })"
             >
         ]>,
-        descr # " with any combination of " # etype.description # " values"> {
+        descr # " with any combination of " # etype.summary # " values"> {
   // The type of elements in the container.
   Type elementType = etype;
 
   // Call to retrieve.
   code getElementTypesCall = elementTypesCall;
 }
 
 // A Tuple that holds a mix of elements of the allowed types.
class DialectAttr<Dialect d, Pred condition, string descr = ""> :
   let cppNamespace = d.cppNamespace;
 }
 
 //===----------------------------------------------------------------------===//
 // Attribute modifier definition
 
 // Decorates an attribute to have an (unvalidated) default value if not present.
 class DefaultValuedAttr<Attr attr, string val> :
-    Attr<attr.predicate, attr.description> {
+    Attr<attr.predicate, attr.summary> {
   // Construct this attribute with the input attribute and change only
   // the default value.
   // Note: this has to be kept up to date with Attr above.
   let storageType = attr.storageType;
   let returnType = attr.returnType;
   let convertFromStorage = attr.convertFromStorage;
   let constBuilderCall = attr.constBuilderCall;
   let defaultValue = val;
   let valueType = attr.valueType;
 
   let baseAttr = attr;
 }
 
 // Decorates an attribute as optional. The return type of the generated
 // attribute accessor method will be Optional<>.
-class OptionalAttr<Attr attr> : Attr<attr.predicate, attr.description> {
+class OptionalAttr<Attr attr> : Attr<attr.predicate, attr.summary> {
   // Rewrite the attribute to be optional.
   // Note: this has to be kept up to date with Attr above.
   let storageType = attr.storageType;
   let returnType = "::llvm::Optional<" # attr.returnType #">";
   let convertFromStorage = "$_self ? " # returnType # "(" #
                            attr.convertFromStorage # ") : (::llvm::None)";
   let valueType = attr.valueType;
   let isOptional = 1;
class TypedStrAttr<Type ty>
     : StringBasedAttr<CPred<"$_self.isa<::mlir::StringAttr>()">,
                             "string attribute"> {
   let valueType = ty;
 }
 
 // Base class for attributes containing types. Example:
 //   def IntTypeAttr : TypeAttrBase<"IntegerType", "integer type attribute">
 // defines a type attribute containing an integer type.
-class TypeAttrBase<string retType, string description> :
+class TypeAttrBase<string retType, string summary> :
     Attr<And<[
       CPred<"$_self.isa<::mlir::TypeAttr>()">,
       CPred<"$_self.cast<::mlir::TypeAttr>().getValue().isa<"
             # retType # ">()">]>,
-    description> {
+    summary> {
   let storageType = [{ ::mlir::TypeAttr }];
   let returnType = retType;
   let valueType = NoneType;
   let convertFromStorage = "$_self.getValue().cast<" # retType # ">()";
 }
 
 def TypeAttr : TypeAttrBase<"::mlir::Type", "any type attribute">;
 
class EnumAttrInfo<string name, list<EnumAttrCaseInfo> cases> {
   string maxEnumValFnName = "getMaxEnumValFor" # name;
 }
 
 // An enum attribute backed by StringAttr.
 //
 // Op attributes of this kind are stored as StringAttr. Extra verification will
 // be generated on the string though: only the symbols of the allowed cases are
 // permitted as the string value.
-class StrEnumAttr<string name, string description,
-                  list<StrEnumAttrCase> cases> :
-    EnumAttrInfo<name, cases>,
+class StrEnumAttr<string name, string summary, list<StrEnumAttrCase> cases>
+  : EnumAttrInfo<name, cases>,
     StringBasedAttr<
       And<[StrAttr.predicate, Or<!foreach(case, cases, case.predicate)>]>,
-      !if(!empty(description), "allowed string cases: " #
+      !if(!empty(summary), "allowed string cases: " #
           !interleave(!foreach(case, cases, "'" # case.symbol # "'"), ", "),
-          description)>;
+          summary)>;
 
 // An enum attribute backed by IntegerAttr.
 //
 // Op attributes of this kind are stored as IntegerAttr. Extra verification will
 // be generated on the integer though: only the values of the allowed cases are
 // permitted as the integer value.
-class IntEnumAttr<I intType, string name, string description,
+class IntEnumAttr<I intType, string name, string summary,
                   list<IntEnumAttrCaseBase> cases> :
     EnumAttrInfo<name, cases>,
     SignlessIntegerAttrBase<intType,
-      !if(!empty(description), "allowed " # intType.description # " cases: " #
-          !interleave(!foreach(case, cases, case.value), ", "), description)> {
+      !if(!empty(summary), "allowed " # intType.summary # " cases: " #
+          !interleave(!foreach(case, cases, case.value), ", "), summary)> {
   let predicate = And<[
     SignlessIntegerAttrBase<intType, "">.predicate,
     Or<!foreach(case, cases, case.predicate)>]>;
 }
 
-class I32EnumAttr<string name, string description,
+class I32EnumAttr<string name, string summary,
                   list<I32EnumAttrCase> cases> :
-    IntEnumAttr<I32, name, description, cases> {
+    IntEnumAttr<I32, name, summary, cases> {
   let returnType = cppNamespace # "::" # name;
   let underlyingType = "uint32_t";
   let convertFromStorage = "static_cast<" # returnType # ">($_self.getInt())";
   let constBuilderCall =
           "$_builder.getI32IntegerAttr(static_cast<int32_t>($0))";
 }
-class I64EnumAttr<string name, string description,
+class I64EnumAttr<string name, string summary,
                   list<I64EnumAttrCase> cases> :
-    IntEnumAttr<I64, name, description, cases> {
+    IntEnumAttr<I64, name, summary, cases> {
   let returnType = cppNamespace # "::" # name;
   let underlyingType = "uint64_t";
   let convertFromStorage = "static_cast<" # returnType # ">($_self.getInt())";
   let constBuilderCall =
           "$_builder.getI64IntegerAttr(static_cast<int64_t>($0))";
 }
 
 // A bit enum stored with 32-bit IntegerAttr.
 //
 // Op attributes of this kind are stored as IntegerAttr. Extra verification will
 // be generated on the integer to make sure only allowed bit are set. Besides,
 // helper methods are generated to parse a string separated with a specified
 // delimiter to a symbol and vice versa.
-class BitEnumAttr<string name, string description,
+class BitEnumAttr<string name, string summary,
                   list<BitEnumAttrCase> cases> :
-    EnumAttrInfo<name, cases>, SignlessIntegerAttrBase<I32, description> {
+    EnumAttrInfo<name, cases>, SignlessIntegerAttrBase<I32, summary> {
   let predicate = And<[
     I32Attr.predicate,
     // Make sure we don't have unknown bit set.
     CPred<"!($_self.cast<::mlir::IntegerAttr>().getValue().getZExtValue() & (~("
           # !interleave(!foreach(case, cases, case.value # "u"), "|") #
           ")))">
   ]>;
 
class BitEnumAttr<string name, string summary,
 
   // The delimiter used to separate bit enum cases in strings.
   string separator = "|";
 }
 
 //===----------------------------------------------------------------------===//
 // Composite attribute kinds
 
-class DictionaryAttrBase<Pred condition, string description> :
-    Attr<condition, description> {
+class DictionaryAttrBase<Pred condition, string summary> :
+    Attr<condition, summary> {
   let storageType = [{ ::mlir::DictionaryAttr }];
   let returnType = [{ ::mlir::DictionaryAttr }];
   let valueType = NoneType;
   let convertFromStorage = "$_self";
 }
 
 def DictionaryAttr
     : DictionaryAttrBase<CPred<"$_self.isa<::mlir::DictionaryAttr>()">,
                                "dictionary of named attribute values">;
 
-class ElementsAttrBase<Pred condition, string description> :
-    Attr<condition, description> {
+class ElementsAttrBase<Pred condition, string summary> :
+    Attr<condition, summary> {
   let storageType = [{ ::mlir::ElementsAttr }];
   let returnType = [{ ::mlir::ElementsAttr }];
   let convertFromStorage = "$_self";
 }
 
 def ElementsAttr : ElementsAttrBase<CPred<"$_self.isa<::mlir::ElementsAttr>()">,
                                     "constant vector/tensor attribute">;
 
-class IntElementsAttrBase<Pred condition, string description> :
+class IntElementsAttrBase<Pred condition, string summary> :
     ElementsAttrBase<And<[CPred<"$_self.isa<::mlir::DenseIntElementsAttr>()">,
                           condition]>,
-                     description> {
+                     summary> {
   let storageType = [{ ::mlir::DenseIntElementsAttr }];
   let returnType = [{ ::mlir::DenseIntElementsAttr }];
 
   let convertFromStorage = "$_self";
 }
 
 def IndexElementsAttr
     : IntElementsAttrBase<CPred<[{$_self.cast<::mlir::DenseIntElementsAttr>()

 class RankedSignlessIntElementsAttr<int width, list<int> dims> :
     SignlessIntElementsAttr<width> {
   // Check that this has the specified shape.
   let predicate = And<[
     SignlessIntElementsAttr<width>.predicate,
     CPred<"$_self.cast<::mlir::DenseIntElementsAttr>().getType().getShape() == "
         "::mlir::ArrayRef<int64_t>({" # !interleave(dims, ", ") # "})">]>;
 
-  let description = width # "-bit signless int elements attribute of shape [" #
-                    !interleave(dims, ", ") # "]";
+  let summary = width # "-bit signless int elements attribute of shape [" #
 
   let constBuilderCall = "::mlir::DenseIntElementsAttr::get("
     "::mlir::RankedTensorType::get({" # !interleave(dims, ", ") #
     "}, $_builder.getIntegerType(" # width # ")), ::llvm::makeArrayRef($0))";
 }
 
 class RankedI32ElementsAttr<list<int> dims> :
     RankedSignlessIntElementsAttr<32, dims>;

 CPred<"$_self.isa<::mlir::AffineMapAttr>()">, "AffineMap attribute"> {
   let storageType = [{::mlir::AffineMapAttr }];
   let returnType = [{ ::mlir::AffineMap }];
   let valueType = Index;
   let constBuilderCall = "::mlir::AffineMapAttr::get($0)";
 }
 
 // Base class for array attributes.
-class ArrayAttrBase<Pred condition, string description> :
-    Attr<condition, description> {
+class ArrayAttrBase<Pred condition, string summary> : Attr<condition, summary> {
   let storageType = [{ ::mlir::ArrayAttr }];
   let returnType = [{ ::mlir::ArrayAttr }];
   let valueType = NoneType;
   let convertFromStorage = "$_self";
 }
 
 def ArrayAttr : ArrayAttrBase<CPred<"$_self.isa<::mlir::ArrayAttr>()">,
                               "array attribute">;
 
 // Base class for array attributes whose elements are of the same kind.
 // `element` specifies the element attribute kind stored in this array.
-class TypedArrayAttrBase<Attr element, string description>: ArrayAttrBase<
+class TypedArrayAttrBase<Attr element, string summary>: ArrayAttrBase<
     And<[
       // Guarantee this is an ArrayAttr first
       CPred<"$_self.isa<::mlir::ArrayAttr>()">,
       // Guarantee all elements satisfy the constraints from `element`
       Concat<"::llvm::all_of($_self.cast<::mlir::ArrayAttr>(), "
                             "[](::mlir::Attribute attr) { return ",
                                SubstLeaves<"$_self", "attr", element.predicate>,
                             "; })">]>,
-    description> {
+    summary> {
   let constBuilderCall = "$_builder.getArrayAttr($0)";
 
   Attr elementAttr = element;
 }
 
 def AffineMapArrayAttr : TypedArrayAttrBase<AffineMapAttr,
                                       "AffineMap array attribute"> {
   let constBuilderCall = "$_builder.getAffineMapArrayAttr($0)";

 //===----------------------------------------------------------------------===//
 
 // A general mechanism to further confine the given `attr` with all the
 // `constraints`. This allows to compose complex constraints out of a series
 // of more primitive ones.
 class Confined<Attr attr, list<AttrConstraint> constraints> : Attr<
     And<!listconcat([attr.predicate],
                       !foreach(pred, constraints, pred.predicate))>,
-    !foldl(/*init*/attr.description, /*list*/constraints,
-           prev, cur, prev # " " # cur.description)> {
+    !foldl(/*init*/attr.summary, /*list*/constraints,
+           prev, cur, prev # " " # cur.summary)> {
   let storageType = attr.storageType;
   let returnType = attr.returnType;
   let convertFromStorage = attr.convertFromStorage;
   let constBuilderCall = attr.constBuilderCall;
   let defaultValue = attr.defaultValue;
   let valueType = attr.valueType;
   let isOptional = attr.isOptional;
 
   let baseAttr = attr;
 }
 
 // An AttrConstraint that holds if all attr constraints specified in
 // 'constraints' hold.
 class AllAttrConstraintsOf<list<AttrConstraint> constraints> : AttrConstraint<
     And<!listconcat([!head(constraints).predicate],
                       !foreach(pred, !tail(constraints), pred.predicate))>,
-    !interleave(!foreach(con, constraints, con.description), " and ")> {
+    !interleave(!foreach(con, constraints, con.summary), " and ")> {
 }
 
 class IntMinValue<int n> : AttrConstraint<
     CPred<"$_self.cast<::mlir::IntegerAttr>().getInt() >= " # n>,
     "whose minimum value is " # n>;
 
 class IntMaxValue<int n> : AttrConstraint<
     CPred<"$_self.cast<::mlir::IntegerAttr>().getInt() <= " # n>,

 
 // A region with the given number of blocks.
 class SizedRegion<int numBlocks> : Region<
   CPred<"::llvm::hasNItems($_self, " # numBlocks # ")">,
   "region with " # numBlocks # " blocks">;
 
 // A variadic region constraint. It expands to zero or more of the base region.
 class VariadicRegion<Region region>
-  : Region<region.predicate, region.description>;
+  : Region<region.predicate, region.summary>;
 
 //===----------------------------------------------------------------------===//
 // Successor definitions
 //===----------------------------------------------------------------------===//
 
 class Successor<Pred condition, string descr = ""> :
     SuccessorConstraint<condition, descr>;
 
 // Any successor.
 def AnySuccessor : Successor<?, "any successor">;
 
 // A variadic successor constraint. It expands to zero or more of the base
 // successor.
 class VariadicSuccessor<Successor successor>
-  : Successor<successor.predicate, successor.description>;
+  : Successor<successor.predicate, successor.summary>;
 
 //===----------------------------------------------------------------------===//
 // OpTrait definitions
 //===----------------------------------------------------------------------===//
 
 // OpTrait represents a trait regarding an op.
 class OpTrait;
 

 // affects op definition generator internals, like how op builders and
 // operand/attribute/result getters are generated.
 class GenInternalOpTrait<string prop> : OpTrait {
   string trait = "::mlir::OpTrait::" # prop;
 }
 
 // PredOpTrait is an op trait implemented by way of a predicate on the op.
 class PredOpTrait<string descr, Pred pred> : OpTrait {
-  string description = descr;
+  string summary = descr;
   Pred predicate = pred;
 }
 
 // Op defines an affine scope.
 def AffineScope : NativeOpTrait<"AffineScope">;
 // Op defines an automatic allocation scope.
 def AutomaticAllocationScope : NativeOpTrait<"AutomaticAllocationScope">;
 // Op supports operand broadcast behavior.

 
 // Class for providing additional information on the variables, i.e. arguments
 // and results, of an operation.
 class OpVariable<Constraint varConstraint, string desc = "",
                  list<OpVariableDecorator> varDecorators = []> {
   // The constraint, either attribute or type, of the argument.
   Constraint constraint = varConstraint;
 
-  // A description for the argument.
-  string description = desc;
+  // One-line human-readable description of the argument.
+  string summary = desc;
 
   // The list of decorators for this variable, e.g. side effects.
   list<OpVariableDecorator> decorators = varDecorators;
 }
 class Arg<Constraint constraint, string desc = "",
           list<OpVariableDecorator> decorators = []>
   : OpVariable<constraint, desc, decorators>;
 class Res<Constraint constraint, string desc = "",
StrFunc<"$" # name # ".getType().cast<::mlir::ShapedType>()"
                                  ".getNumElements()">;
 
 class ElementType<string name> : StrFunc<"getElementTypeOrSelf($" # name # ")">;
 
 class AllMatchPred<list<string> values> :
     CPred<"::llvm::is_splat(::llvm::makeArrayRef({"
           # !interleave(values, ", ") #"}))">;
 
-class AllMatch<list<string> values, string description> :
-    PredOpTrait<description, AllMatchPred<values>>;
+class AllMatch<list<string> values, string summary> :
+    PredOpTrait<summary, AllMatchPred<values>>;
 
 // TODO: Only works for non-variadic.
 class AllMatchSameOperatorPred<list<string> names, string operator> :
     AllMatchPred<!foreach(n, names, !subst("$_self", "$" # n, operator))>;
 
 class AllMatchSameOperatorTrait<list<string> names, string operator,
-                                string description> :
+                                string summary> :
     PredOpTrait<
-        "all of {" # !interleave(names, ", ") # "} have same " # description,
+        "all of {" # !interleave(names, ", ") # "} have same " # summary,
         AllMatchSameOperatorPred<names, operator>> {
   list<string> values = names;
 }
 
 class AllElementCountsMatch<list<string> names> :
     AllMatchSameOperatorTrait<names, ElementCount<"_self">.result,
                               "element count">;
 
 class AllElementTypesMatch<list<string> names> :
     AllMatchSameOperatorTrait<names, ElementType<"_self">.result,
                               "element type">;
 
 class AllRanksMatch<list<string> names> :
     AllMatchSameOperatorTrait<names, Rank<"_self">.result, "rank">;
 
 class AllShapesMatch<list<string> names> :
     AllMatchSameOperatorTrait<names, Shape<"_self">.result, "shape">;
 
 class AllTypesMatch<list<string> names> :
     AllMatchSameOperatorTrait<names, "$_self.getType()", "type">;
 
 // A type constraint that denotes `transform(lhs.getType()) == rhs.getType()`.
-class TypesMatchWith<string description, string lhsArg, string rhsArg,
+class TypesMatchWith<string summary, string lhsArg, string rhsArg,
                      string transform> :
-    PredOpTrait<description, CPred<
+    PredOpTrait<summary, CPred<
       !subst("$_self", "$" # lhsArg # ".getType()", transform)
       # " == $" # rhsArg # ".getType()">> {
   string lhs = lhsArg;
   string rhs = rhsArg;
   string transformer = transform;
 }
 
 // Type Constraint operand `idx`'s Element type is `type`.
 class TCopVTEtIs<int idx, Type type> : And<[
    CPred<"$_op.getNumOperands() > " # idx>,
    SubstLeaves<"$_self", "$_op.getOperand(" # idx # ").getType()",
      IsShapedTypePred>,
    SubstLeaves<"$_self", "getElementTypeOrSelf($_op.getOperand(" # idx # "))",
      type.predicate>]>;
 
 // Predicate to verify that a named argument or result's element type matches a
 // given type.
 class TypeIsPred<string name, Type type> :
    SubstLeaves<"$_self", "$" # name # ".getType()", type.predicate>;
 class TypeIs<string name, Type type> : PredOpTrait<
-  "'" # name # "' is " # type.description, TypeIsPred<name, type>>;
+  "'" # name # "' is " # type.summary, TypeIsPred<name, type>>;
 
 // Predicate to verify that a named argument or result's element type matches a
 // given type.
 class ElementTypeIsPred<string name, Type type> : And<[
    SubstLeaves<"$_self", "$" # name # ".getType()", IsShapedTypePred>,
    SubstLeaves<"$_self", "getElementTypeOrSelf($" # name # ")",
      type.predicate>]>;
 class ElementTypeIs<string name, Type type> : PredOpTrait<
-  "'" # name # "' is " # type.description, ElementTypeIsPred<name, type>>;
+  "'" # name # "' is " # type.summary, ElementTypeIsPred<name, type>>;
 
 // Predicate to verify that the i'th operand and the j'th operand have the same
 // elemental type.
 // Type Constraint operand `i`'s Element type is Same As operand `j`'s Element
 // type.
 class TCopVTEtIsSameAs<int i, int j> : And<[
     CPred<"$_op.getNumOperands() > " # !if(!gt(i,j),i,j)>,
     SubstLeaves<"$_self", "$_op.getOperand(" # i # ").getType()",

 // Define a new type, named `name`, belonging to `dialect` that inherits from
 // the given C++ base class.
 class TypeDef<Dialect dialect, string name,
               string baseCppClass = "::mlir::Type">
     : DialectType<dialect, CPred<"">, /*descr*/"", name # "Type"> {
   // The name of the C++ base class to use for this Type.
   string cppBaseClassName = baseCppClass;
 
-  // Short summary of the type.
-  string summary = ?;
+  // Additional, longer human-readable description of what the op does.
+  string description = "";
 
   // Name of storage class to generate or use.
   string storageClass = name # "TypeStorage";
   // Namespace (withing dialect c++ namespace) in which the storage class
   // resides.
   string storageNamespace = "detail";
   // Specify if the storage class is to be generated.
   bit genStorageClass = 1;

 
 // 'Parameters' should be subclasses of this or simple strings (which is a
 // shorthand for TypeParameter<"C++Type">).
 class TypeParameter<string type, string desc> {
   // Custom memory allocation code for storage constructor.
   code allocator = ?;
   // The C++ type of this parameter.
   string cppType = type;
-  // A description of this parameter.
-  string description = desc;
+  // One-line human-readable description of the argument.
+  string summary = desc;
   // The format string for the asm syntax (documentation only).
   string syntax = ?;
 }
 
 // For StringRefs, which require allocation.
 class StringRefParameter<string desc> :
     TypeParameter<"::llvm::StringRef", desc> {
   let allocator = [{$_dst = $_allocator.copyInto($_self);}];

mlir/include/mlir/TableGen/Constraint.h

public:
   // Returns the condition template that can be used to check if a type or
   // attribute satisfies this constraint.  The template may contain "{0}" that
   // must be substituted with an expression returning an mlir::Type or
   // mlir::Attribute.
   std::string getConditionTemplate() const;
 
   // Returns the user-readable description of this constraint. If the
   // description is not provided, returns the TableGen def name.
-  StringRef getDescription() const;
+  StringRef getSummary() const;
 
   // Constraint kind
   enum Kind { CK_Attr, CK_Region, CK_Successor, CK_Type, CK_Uncategorized };
 
   Kind getKind() const { return kind; }
 
   /// Get an opaque pointer to the constraint.
   const void *getAsOpaquePointer() const { return def; }

mlir/include/mlir/TableGen/OpTrait.h

 
 // OpTrait corresponding to a predicate on the operation.
 class PredOpTrait : public OpTrait {
 public:
   // Returns the template for constructing the predicate.
   std::string getPredTemplate() const;
 
   // Returns the description of what the predicate is verifying.
-  StringRef getDescription() const;
+  StringRef getSummary() const;
 
   static bool classof(const OpTrait *t) { return t->getKind() == Kind::Pred; }
 };
 
 // OpTrait controlling op definition generator internals.
 class InternalOpTrait : public OpTrait {
 public:
   // Returns the trait controlling op definition generator internals.

mlir/include/mlir/TableGen/Type.h

 };
 
 // Wrapper class with helper methods for accessing Types defined in TableGen.
 class Type : public TypeConstraint {
 public:
   explicit Type(const llvm::Record *record);
 
   // Returns the description of the type.
-  StringRef getTypeDescription() const;
+  StringRef getDescription() const;
 
   // Returns the dialect for the type if defined.
   Dialect getDialect() const;
 };
 
 } // end namespace tblgen
 } // end namespace mlir
 
 #endif // MLIR_TABLEGEN_TYPE_H_

mlir/include/mlir/TableGen/TypeDef.h

 class TypeParameter {
 public:
   explicit TypeParameter(const llvm::DagInit *def, unsigned num)
       : def(def), num(num) {}
 
   // Get the parameter name.
   StringRef getName() const;
   // If specified, get the custom allocator code for this parameter.
-  llvm::Optional<StringRef> getAllocator() const;
+  Optional<StringRef> getAllocator() const;
   // Get the C++ type of this parameter.
   StringRef getCppType() const;
   // Get a description of this parameter for documentation purposes.
-  llvm::Optional<StringRef> getDescription() const;
+  Optional<StringRef> getSummary() const;
   // Get the assembly syntax documentation.
   StringRef getSyntax() const;
 
 private:
   const llvm::DagInit *def;
   const unsigned num;
 };
 
 } // end namespace tblgen
 } // end namespace mlir
 
 #endif // MLIR_TABLEGEN_TYPEDEF_H

mlir/lib/TableGen/Constraint.cpp

Pred Constraint::getPredicate() const {
   const auto *pred = dyn_cast<llvm::DefInit>(val->getValue());
   return Pred(pred);
 }
 
 std::string Constraint::getConditionTemplate() const {
   return getPredicate().getCondition();
 }
 
-StringRef Constraint::getDescription() const {
-  // If a summary is found, we use that given that it is a focused single line
-  // comment.
+StringRef Constraint::getSummary() const {
   if (Optional<StringRef> summary = def->getValueAsOptionalString("summary"))
     return *summary;
-  // If a summary can't be found, look for a specific description field to use
-  // for the constraint.
-  StringRef desc = def->getValueAsString("description");
-  if (!desc.empty())
-    return desc;
-  // Otherwise, fallback to the name of the constraint definition.
   return def->getName();
 }
 
 AppliedConstraint::AppliedConstraint(Constraint &&constraint,
                                      llvm::StringRef self,
                                      std::vector<std::string> &&entities)
     : constraint(constraint), self(std::string(self)),
       entities(std::move(entities)) {}

mlir/lib/TableGen/OpTrait.cpp

llvm::StringRef InternalOpTrait::getTrait() const {
   return def->getValueAsString("trait");
 }
 
 std::string PredOpTrait::getPredTemplate() const {
   auto pred = Pred(def->getValueInit("predicate"));
   return pred.getCondition();
 }
 
-llvm::StringRef PredOpTrait::getDescription() const {
-  return def->getValueAsString("description");
+llvm::StringRef PredOpTrait::getSummary() const {
+  return def->getValueAsString("summary");
 }
 
 OpInterface InterfaceOpTrait::getOpInterface() const {
   return OpInterface(def);
 }
 
 std::string InterfaceOpTrait::getTrait() const {
   llvm::StringRef trait = def->getValueAsString("trait");

mlir/lib/TableGen/Type.cpp

 
 // Return the C++ class name for this type (which may just be ::mlir::Type).
 StringRef TypeConstraint::getCPPClassName() const {
   return def->getValueAsString("cppClassName");
 }
 
 Type::Type(const llvm::Record *record) : TypeConstraint(record) {}
 
-StringRef Type::getTypeDescription() const {
-  return def->getValueAsString("typeDescription");
+StringRef Type::getDescription() const {
+  return def->getValueAsString("description");
 }
 
 Dialect Type::getDialect() const {
   return Dialect(def->getValueAsDef("dialect"));
 }

mlir/lib/TableGen/TypeDef.cpp

 
 bool TypeDef::operator<(const TypeDef &other) const {
   return getName() < other.getName();
 }
 
 StringRef TypeParameter::getName() const {
   return def->getArgName(num)->getValue();
 }
-llvm::Optional<StringRef> TypeParameter::getAllocator() const {
+Optional<StringRef> TypeParameter::getAllocator() const {
   llvm::Init *parameterType = def->getArg(num);
   if (isa<llvm::StringInit>(parameterType))
     return llvm::Optional<StringRef>();
 
   if (auto *typeParameter = dyn_cast<llvm::DefInit>(parameterType)) {
     llvm::RecordVal *code = typeParameter->getDef()->getValue("allocator");
     if (!code)
       return llvm::Optional<StringRef>();
StringRef TypeParameter::getCppType() const {
   if (auto *stringType = dyn_cast<llvm::StringInit>(parameterType))
     return stringType->getValue();
   if (auto *typeParameter = dyn_cast<llvm::DefInit>(parameterType))
     return typeParameter->getDef()->getValueAsString("cppType");
   llvm::PrintFatalError(
       "Parameters DAG arguments must be either strings or defs "
       "which inherit from TypeParameter\n");
 }
-llvm::Optional<StringRef> TypeParameter::getDescription() const {
+Optional<StringRef> TypeParameter::getSummary() const {
   auto *parameterType = def->getArg(num);
   if (auto *typeParameter = dyn_cast<llvm::DefInit>(parameterType)) {
     const auto *desc = typeParameter->getDef()->getValue("description");
     if (llvm::StringInit *ci = dyn_cast<llvm::StringInit>(desc->getValue()))
       return ci->getValue();
   }
-  return llvm::Optional<StringRef>();
+  return Optional<StringRef>();
 }
 StringRef TypeParameter::getSyntax() const {
   auto *parameterType = def->getArg(num);
   if (auto *stringType = dyn_cast<llvm::StringInit>(parameterType))
     return stringType->getValue();
   if (auto *typeParameter = dyn_cast<llvm::DefInit>(parameterType)) {
     const auto *syntax = typeParameter->getDef()->getValue("syntax");
     if (syntax && isa<llvm::StringInit>(syntax->getValue()))
       return dyn_cast<llvm::StringInit>(syntax->getValue())->getValue();
     return getCppType();
   }
   llvm::PrintFatalError("Parameters DAG arguments must be either strings or "
                         "defs which inherit from TypeParameter");
 }

mlir/test/mlir-tblgen/op-attribute.td

 // RECORD-LABEL: def SomeTypedArrayAttr
 // RECORD: Attr elementAttr = SomeAttr;
 
 def Test_Dialect_2 : Dialect {
   let name = "dialect_2";
 }
 def MyStruct : StructAttr<"MyStruct", Test_Dialect_2,
 [StructFieldAttr<"potatoes", I64ElementsAttr>]> {
-  let description = "A structure describing a number of potatoes.";
+  let summary = "A structure describing a number of potatoes.";
 }
 
 def StructAttrOp : NS_Op<"struct_attr_op", []> {
   let arguments = (ins
     MyStruct:$potatoes
   );
 }
 
 // DECL: dialect_2::MyStruct potatoes();

mlir/tools/mlir-tblgen/EnumsGen.cpp

static void emitUnderlyingToSymFnForBitEnum(const Record &enumDef,
   os << "}\n";
 }
 
 static void emitEnumDecl(const Record &enumDef, raw_ostream &os) {
   EnumAttr enumAttr(enumDef);
   StringRef enumName = enumAttr.getEnumClassName();
   StringRef cppNamespace = enumAttr.getCppNamespace();
   std::string underlyingType = std::string(enumAttr.getUnderlyingType());
-  StringRef description = enumAttr.getDescription();
+  StringRef description = enumAttr.getSummary();
   StringRef strToSymFnName = enumAttr.getStringToSymbolFnName();
   StringRef symToStrFnName = enumAttr.getSymbolToStringFnName();
   StringRef symToStrFnRetType = enumAttr.getSymbolToStringFnRetType();
   StringRef underlyingToSymFnName = enumAttr.getUnderlyingToSymbolFnName();
   auto enumerants = enumAttr.getAllCases();
 
   llvm::SmallVector<StringRef, 2> namespaces;
   llvm::SplitString(cppNamespace, namespaces, "::");

mlir/tools/mlir-tblgen/OpDefinitionsGen.cpp

os << "static ::mlir::LogicalResult " << name
           "valueKind, unsigned valueGroupStartIndex) {\n";
 
     os << "  if (!("
        << tgfmt(constraint.getConditionTemplate(), &fctx.withSelf("type"))
        << ")) {\n"
        << formatv(
               "    return op->emitOpError(valueKind) << \" #\" << "
               "valueGroupStartIndex << \" must be {0}, but got \" << type;\n",
-              constraint.getDescription())
+              constraint.getSummary())
        << "  }\n"
        << "  return ::mlir::success();\n"
        << "}\n\n";
   }
 }
 
 //===----------------------------------------------------------------------===//
 // Utility structs and functions
if (allowMissingAttr) {
       // set. This does effectively assume that the default value is valid.
       // TODO: verify the debug value is valid (perhaps in debug mode only).
       body << "  if (" << varName << ") {\n";
     }
 
     body << tgfmt("    if (!($0)) return $1\"attribute '$2' "
                   "failed to satisfy constraint: $3\");\n",
                   /*ctx=*/nullptr, tgfmt(condition, &ctx.withSelf(varName)),
-                  emitErrorPrefix, attrName, attr.getDescription());
+                  emitErrorPrefix, attrName, attr.getSummary());
     if (allowMissingAttr)
       body << "  }\n";
     body << "  }\n";
   }
 }
 
 OpEmitter::OpEmitter(const Operator &op,
                      const StaticVerifierFunctionEmitter &staticVerifierEmitter)
void OpEmitter::genVerifier() {
   genOperandResultVerifier(body, op.getOperands(), "operand");
   genOperandResultVerifier(body, op.getResults(), "result");
 
   for (auto &trait : op.getTraits()) {
     if (auto *t = dyn_cast<tblgen::PredOpTrait>(&trait)) {
       body << tgfmt("  if (!($0))\n    "
                     "return emitOpError(\"failed to verify that $1\");\n",
                     &verifyCtx, tgfmt(t->getPredTemplate(), &verifyCtx),
-                    t->getDescription());
+                    t->getSummary());
     }
   }
 
   genRegionVerifier(body);
   genSuccessorVerifier(body);
 
   if (hasCustomVerify) {
     FmtContext fctx;
for (unsigned i = 0; i < numRegions; ++i) {
 
     body << formatv("      (void)region;\n"
                     "      if (!({0})) {\n        "
                     "return emitOpError(\"region #\") << index << \" {1}"
                     "failed to "
                     "verify constraint: {2}\";\n      }\n",
                     constraint,
                     region.name.empty() ? "" : "('" + region.name + "') ",
-                    region.constraint.getDescription())
+                    region.constraint.getSummary())
          << "      ++index;\n"
          << "    }\n";
   }
   body << "  }\n";
 }
 
 void OpEmitter::genSuccessorVerifier(OpMethodBody &body) {
   // If we have no successors, there is nothing more to do.
auto constraint = tgfmt(successor.constraint.getConditionTemplate(),
                           .str();
 
     body << formatv("      (void)successor;\n"
                     "      if (!({0})) {\n        "
                     "return emitOpError(\"successor #\") << index << \"('{1}') "
                     "failed to "
                     "verify constraint: {2}\";\n      }\n",
                     constraint, successor.name,
-                    successor.constraint.getDescription())
+                    successor.constraint.getSummary())
          << "      ++index;\n"
          << "    }\n";
   }
   body << "  }\n";
 }
 
 /// Add a size count trait to the given operation class.
 static void addSizeCountTrait(OpClass &opClass, StringRef traitKind,

mlir/tools/mlir-tblgen/OpDocGen.cpp

 
 /// Emit the given named constraint.
 template <typename T>
 static void emitNamedConstraint(const T &it, raw_ostream &os) {
   if (!it.name.empty())
     os << "`" << it.name << "`";
   else
     os << "&laquo;unnamed&raquo;";
-  os << " | " << it.constraint.getDescription() << "\n";
+  os << " | " << it.constraint.getSummary() << "\n";
 }
 
 //===----------------------------------------------------------------------===//
 // Operation Documentation
 //===----------------------------------------------------------------------===//
 
 /// Emit the assembly format of an operation.
 static void emitAssemblyFormat(StringRef opName, StringRef format,
if (op.getNumAttributes() != 0) {
     // TODO: Attributes are only documented by TableGen name, with no further
     // info. This should be improved.
     os << "\n#### Attributes:\n\n";
     os << "| Attribute | MLIR Type | Description |\n"
        << "| :-------: | :-------: | ----------- |\n";
     for (const auto &it : op.getAttributes()) {
       StringRef storageType = it.attr.getStorageType();
       os << "`" << it.name << "` | " << storageType << " | "
-         << it.attr.getDescription() << "\n";
+         << it.attr.getSummary() << "\n";
     }
   }
 
   // Emit each of the operands.
   if (op.getNumOperands() != 0) {
     os << "\n#### Operands:\n\n";
     os << "| Operand | Description |\n"
        << "| :-----: | ----------- |\n";
for (const llvm::Record *opDef : opDefs)
     emitOpDoc(Operator(opDef), os);
 }
 
 //===----------------------------------------------------------------------===//
 // Type Documentation
 //===----------------------------------------------------------------------===//
 
 static void emitTypeDoc(const Type &type, raw_ostream &os) {
-  os << "### " << type.getDescription() << "\n";
-  emitDescription(type.getTypeDescription(), os);
+  os << "### " << type.getSummary() << "\n";
+  emitDescription(type.getDescription(), os);
   os << "\n";
 }
 
 //===----------------------------------------------------------------------===//
 // TypeDef Documentation
 //===----------------------------------------------------------------------===//
 
 /// Emit the assembly format of a type.
if (td.getMnemonic() && td.getPrinterCode() && *td.getPrinterCode() == "" &&
       td.getParserCode() && *td.getParserCode() == "")
     emitTypeAssemblyFormat(td, os);
   if (td.hasDescription())
     mlir::tblgen::emitDescription(td.getDescription(), os);
 
   // Emit attribute documentation.
   SmallVector<TypeParameter, 4> parameters;
   td.getParameters(parameters);
-  if (parameters.size() != 0) {
+  if (!parameters.empty()) {
     os << "\n#### Type parameters:\n\n";
     os << "| Parameter | C++ type | Description |\n"
        << "| :-------: | :-------: | ----------- |\n";
     for (const auto &it : parameters) {
-      auto desc = it.getDescription();
+      auto desc = it.getSummary();
       os << "| " << it.getName() << " | `" << td.getCppClassName() << "` | "
          << (desc ? *desc : "") << " |\n";
     }
   }
 
   os << "\n";
 }
 

mlir/tools/mlir-tblgen/OpFormatGen.cpp

for (TypeResolution &resolver :
     body << "  for (::mlir::Type type : " << variable->name << "Types) {\n"
          << "    (void)type;\n"
          << "    if (!("
          << tgfmt(constraint.getConditionTemplate(),
                   &verifierFCtx.withSelf("type"))
          << ")) {\n"
          << formatv("      return parser.emitError(parser.getNameLoc()) << "
                     "\"'{0}' must be {1}, but got \" << type;\n",
-                    variable->name, constraint.getDescription())
+                    variable->name, constraint.getSummary())
          << "    }\n"
          << "  }\n";
   }
 
   // Initialize the set of buildable types.
   if (!buildableTypes.empty()) {
     FmtContext typeBuilderCtx;
     typeBuilderCtx.withBuilder("parser.getBuilder()");

mlir/tools/mlir-tblgen/RewriterGen.cpp

for (int i = 0, e = tree.getNumArgs(); i != e; ++i) {
 
     auto self = formatv("{0}", argName);
     emitMatchCheck(
         opName,
         tgfmt(constraint.getConditionTemplate(), &fmtCtx.withSelf(self)),
         formatv("\"operand {0} of native code call '{1}' failed to satisfy "
                 "constraint: "
                 "'{2}'\"",
-                i, tree.getNativeCodeTemplate(), constraint.getDescription()));
+                i, tree.getNativeCodeTemplate(), constraint.getSummary()));
   }
 
   LLVM_DEBUG(llvm::dbgs() << "done emitting match for native code call\n");
 }
 
 // Helper function to match patterns.
 void PatternEmitter::emitOpMatch(DagNode tree, StringRef opName, int depth) {
   Operator &op = tree.getDialectOp(opMap);
if (operand->constraint != constraint) {
       auto self = formatv("(*{0}.getODSOperands({1}).begin()).getType()",
                           opName, argIndex);
       emitMatchCheck(
           opName,
           tgfmt(constraint.getConditionTemplate(), &fmtCtx.withSelf(self)),
           formatv("\"operand {0} of op '{1}' failed to satisfy constraint: "
                   "'{2}'\"",
                   operand - op.operand_begin(), op.getOperationName(),
-                  constraint.getDescription()));
+                  constraint.getSummary()));
     }
   }
 
   // Capture the value
   auto name = tree.getArgName(argIndex);
   // `$_` is a special symbol to ignore op argument matching.
   if (!name.empty() && name != "_") {
     // We need to subtract the number of attributes before this operand to get
if (!matcher.isUnspecified()) {
     // If a constraint is specified, we need to generate C++ statements to
     // check the constraint.
     emitMatchCheck(
         opName,
         tgfmt(matcher.getConditionTemplate(), &fmtCtx.withSelf("tblgen_attr")),
         formatv("\"op '{0}' attribute '{1}' failed to satisfy constraint: "
                 "{2}\"",
                 op.getOperationName(), namedAttr->name,
-                matcher.getAsConstraint().getDescription()));
+                matcher.getAsConstraint().getSummary()));
   }
 
   // Capture the value
   auto name = tree.getArgName(argIndex);
   // `$_` is a special symbol to ignore op argument matching.
   if (!name.empty() && name != "_") {
     os << formatv("{0} = tblgen_attr;\n", name);
   }
for (auto &appliedConstraint : pattern.getConstraints()) {
 
     auto condition = constraint.getConditionTemplate();
     if (isa<TypeConstraint>(constraint)) {
       auto self = formatv("({0}.getType())",
                           symbolInfoMap.getValueAndRangeUse(entities.front()));
       emitMatchCheck(
           opName, tgfmt(condition, &fmtCtx.withSelf(self.str())),
           formatv("\"value entity '{0}' failed to satisfy constraint: {1}\"",
-                  entities.front(), constraint.getDescription()));
+                  entities.front(), constraint.getSummary()));
 
     } else if (isa<AttrConstraint>(constraint)) {
       PrintFatalError(
           loc, "cannot use AttrConstraint in Pattern multi-entity constraints");
     } else {
       // TODO: replace formatv arguments with the exact specified
       // args.
       if (entities.size() > 4) {
if (isa<TypeConstraint>(constraint)) {
       for (; i < 4; ++i)
         names.push_back("<unused>");
       emitMatchCheck(opName,
                      tgfmt(condition, &fmtCtx.withSelf(self), names[0],
                            names[1], names[2], names[3]),
                      formatv("\"entities '{0}' failed to satisfy constraint: "
                              "{1}\"",
                              llvm::join(entities, ", "),
-                             constraint.getDescription()));
+                             constraint.getSummary()));
     }
   }
 
   // Some of the operands could be bound to the same symbol name, we need
   // to enforce equality constraint on those.
   // TODO: we should be able to emit equality checks early
   // and short circuit unnecessary work if vars are not equal.
   for (auto symbolInfoIt = symbolInfoMap.begin();

mlir/tools/mlir-tblgen/StructsGen.cpp

)";
 
   os << "};\n\n";
 }
 
 static void emitStructDecl(const Record &structDef, raw_ostream &os) {
   StructAttr structAttr(&structDef);
   StringRef structName = structAttr.getStructClassName();
   StringRef cppNamespace = structAttr.getCppNamespace();
-  StringRef description = structAttr.getDescription();
+  StringRef description = structAttr.getSummary();
   auto fields = structAttr.getAllFields();
 
   // Wrap in the appropriate namespace.
   llvm::SmallVector<StringRef, 2> namespaces;
   llvm::SplitString(cppNamespace, namespaces, "::");
 
   for (auto ns : namespaces)
     os << "namespace " << ns << " {\n";

mlir/unittests/TableGen/structs.td

 def Test_Struct : StructAttr<"TestStruct", Test_Dialect, [
                 StructFieldAttr<"sample_integer", I32Attr>,
                 StructFieldAttr<"sample_float", F32Attr>,
                 StructFieldAttr<"sample_elements", I32ElementsAttr>,
                 StructFieldAttr<"sample_optional_integer",
                                 OptionalAttr<I32Attr>>,
                 StructFieldAttr<"sample_default_valued_integer",
                                 DefaultValuedAttr<I32Attr, "42">>] > {
-  let description = "Structure for test data";
+  let summary = "Structure for test data";
 }