[NFC] Fix syntax of ranked memrefs in the MLIR Language Reference.

- Eliminate incorrect |
- Eliminate memspace0 as the memory spaces currently are integer literals and memory
  space 0 is not explicitly printed.

Differential Revision: https://reviews.llvm.org/D88171
This commit is contained in:
Rahul Joshi 2020-09-23 10:57:09 -07:00
parent 0b7f03b98d
commit 8046ab04a7
1 changed files with 5 additions and 6 deletions

View File

@ -929,8 +929,7 @@ Syntax:
memref-type ::= ranked-memref-type | unranked-memref-type
ranked-memref-type ::= `memref` `<` dimension-list-ranked tensor-memref-element-type
(`,` layout-specification)? |
(`,` memory-space)? `>`
(`,` layout-specification)? (`,` memory-space)? `>`
unranked-memref-type ::= `memref` `<*x` tensor-memref-element-type
(`,` memory-space)? `>`
@ -1039,7 +1038,7 @@ Examples of memref static type
//
// { (i, j) : 0 <= i < 16, 0 <= j < 32 }
//
memref<16x32xf32, #identity, memspace0>
memref<16x32xf32, #identity>
// The dimension list "16x4x?" defines the following 3D index space:
//
@ -1049,7 +1048,7 @@ memref<16x32xf32, #identity, memspace0>
// the third dimension.
//
// %N here binds to the size of the third dimension.
%A = alloc(%N) : memref<16x4x?xf32, #col_major, memspace0>
%A = alloc(%N) : memref<16x4x?xf32, #col_major>
// A 2-d dynamic shaped memref that also has a dynamically sized tiled layout.
// The memref index space is of size %M x %N, while %B1 and %B2 bind to the
@ -1088,10 +1087,10 @@ Examples
```mlir
// Allocates a memref with 2D index space:
// { (i, j) : 0 <= i < 16, 0 <= j < 32 }
%A = alloc() : memref<16x32xf32, #imapA, memspace0>
%A = alloc() : memref<16x32xf32, #imapA>
// Loads data from memref '%A' using a 2D index: (%i, %j)
%v = load %A[%i, %j] : memref<16x32xf32, #imapA, memspace0>
%v = load %A[%i, %j] : memref<16x32xf32, #imapA>
```
##### Index Map