maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity

[mlir][amx] add doc to AMX dialect 

Making sure the AMX dialect webpage reads better with a short introduction on the purpose of this dialect.

Reviewed By: grosul1, bondhugula

Differential Revision: https://reviews.llvm.org/D107419
This commit is contained in:
Aart Bik 2021-08-03 18:49:52 -07:00
parent f3eb5f900d
commit 478c71bf95
1 changed files with 19 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
mlir/include/mlir/Dialect/AMX/AMX.td
View File

@ -8,7 +8,7 @@
//
// This file defines the basic operations for the AMX dialect.
//
// The Intel Advanced Matrix Extensions (AMX) provides a tile matrix
// The Intel Advanced Matrix Extensions (AMX) provide a tile matrix
// multiply unit (TMUL), a tile control register (TILECFG), and eight
// tile registers TMM0 through TMM7 (TILEDATA).
//
@ -38,6 +38,23 @@ include "mlir/Interfaces/SideEffectInterfaces.td"
def AMX_Dialect : Dialect {
let name = "amx";
let cppNamespace = "::mlir::amx";
let description = [{
The Intel Advanced Matrix Extensions (AMX) provide a tile matrix
multiply unit (TMUL), a tile control register (TILECFG), and eight
tile registers TMM0 through TMM7 (TILEDATA).
This `AMX` dialect provides a bridge between MLIR concepts such as
vectors and memrefs and the lower level LLVM IR support of AMX.
The dialect is split into user-facing AMX ops (AMX_Op) and
backend-facing intrinsic ops (AMX_IntrOp).
Note that since configuration changes (implicit at dialect level) are
costly, it is highly recommended to use the AMX dialect on same-shaped
vectors, at least within a single method.
For details, see the Intel documentation:
https://software.intel.com/content/www/us/en/develop/articles/intel-sdm.html
}];
}
//===----------------------------------------------------------------------===//
@ -203,7 +220,7 @@ def TileMulIOp : AMX_Op<"tile_muli", [NoSideEffect, AllTypesMatch<["acc", "res"]
Example:
```mlir
%0 = amx.tile_muli %a zext, %b zext, %c
%0 = amx.tile_muli %a zext, %b zext, %c
: vector<16x64xi8>, vector<16x64xi8>, vector<16x16xi32>
```
}];