| /external/llvm-project/mlir/docs/ |
| D | doxygen-mainpage.dox | 1 /// \mainpage MLIR
4 /// Welcome to the MLIR project.
7 /// up MLIR, not the **external** use of MLIR. There are no instructions
8 /// here on how to use MLIR, only the APIs that make up the software. For
14 /// Since MLIR is constantly under active development, what you're about to
|
| /external/llvm-project/mlir/docs/Rationale/ |
| D | MLIRForGraphAlgorithms.md | 1 # MLIR: Incremental Application to Graph Algorithms in ML Frameworks
3 The existing documentation about MLIR focuses on long term vision, how its
10 This document explains that adoption of MLIR to solve graph based problems
15 One note: even though a major advantage of MLIR is that it can span the full
17 focuses on the use of MLIR for **graph-level algorithms**. MLIR will also unlock
20 MLIR's relationship to XLA, Eigen, etc, are out of scope for this particular
26 MLIR in the future.
28 ### How is MLIR relevant?
30 MLIR is an overloaded acronym which unpacks as "Multi-Level Intermediate
44 MLIR infrastructure, but that isn't a focus of this doc.))
[all …]
|
| /external/llvm-project/mlir/docs/Tutorials/Toy/ |
| D | _index.md | 4 MLIR. The goal of this tutorial is to introduce the concepts of MLIR; in
11 This tutorial assumes you have cloned and built MLIR; if you have not yet done
13 [Getting started with MLIR](https://mlir.llvm.org/getting_started/).
19 - [Chapter #2](Ch-2.md): Traversing the AST to emit a dialect in MLIR,
20 introducing base MLIR concepts. Here we show how to start attaching
21 semantics to our custom operations in MLIR.
34 We'll demonstrate how to add a custom type to MLIR, and how it fits in the
|
| /external/tensorflow/tensorflow/compiler/mlir/g3doc/ |
| D | overview.md | 1 # MLIR chapter
5 MLIR, or Multi-Level Intermediate Representation, is a representation format
9 MLIR is, at its heart, a flexible infrastructure for modern optimizing
16 MLIR is highly influenced by [LLVM](https://llvm.org/) and unabashedly reuses
23 We expect MLIR to be of interest to many groups, including:
34 gradual adoption of MLIR to simplify every aspect of this stack.
36 <img alt="MLIR overview diagram" src="./images/mlir-infra.svg"/>
|
| D | xla_gpu_codegen.md | 1 # MLIR CodeGen for XLA
18 right after XLA completes the buffer-assignment phase and emit instead an MLIR
19 module in the `lhlo` dialect. From there we perform the codegen using MLIR
36 * **Task 3** migrates device output from LLVM IR to some form of MLIR. It's
47 * Switch a legacy emitter to an MLIR-based emitter (e.g. Linalg), or
48 * Mechanically translate the existing emitter to take MLIR representation
111 1. Refactor the old emitter to be like LHLO -> MLIR LLVM Dialect:
116 1. Refactor old emitters to be like LHLO -> MLIR GPU + Standard + Loops:
139 still achieve the goal of enabling XLA to pick up MLIR emitters. This is by
155 3 (MLIR emitters).
[all …]
|
| D | dialects.md | 1 # MLIR dialects
6 To separate different hardware and software targets, MLIR has “dialects”,
16 allowing MLIR to emit GPU and CPU code through LLVM.
23 ## Adding to MLIR
25 MLIR has no fixed/built-in list of globally known operations (no “intrinsics”).
26 Dialects can define entirely custom types, which is how MLIR can model things
|
| /external/llvm-project/mlir/tools/mlir-shlib/ |
| D | CMakeLists.txt | 24 # libClang needs this, but it causes problems for MLIR (probably
25 # because we use public library dependencies within MLIR.)
35 MLIR
46 target_link_libraries(MLIR PRIVATE ${LLVM_PTHREAD_LIB})
52 mlir_check_all_link_libraries(MLIR)
|
| /external/tensorflow/tensorflow/compiler/mlir/lite/tests/end2end/ |
| D | quant_stats.pbtxt | 2 #| FileCheck --check-prefix=MLIR %s
42 # MLIR-LABEL: func @main(%arg0: tensor<4x!quant.uniform<u8:f32, 0.015686274509803921:128>>, %arg1: …
43 # MLIR-NEXT: attributes {tf.entry_function = {inputs = "input0,input1", outputs = "Add"}} {
44 # MLIR-NEXT: %[[add:.*]] = "tfl.add"(%arg0, %arg1) {fused_activation_function = "NONE"} : (tenso…
45 # MLIR-NEXT: return %[[add]] : tensor<4x!quant.uniform<u8:f32, 0.0078431372549019607:128>>
46 # MLIR-NEXT: }
94 # CHECK-NEXT: description: "MLIR Converted.",
|
| D | fake_quant_per_channel.pbtxt | 1 …s='21.54917' -tf-output-arrays=output %s -o - --output-mlir 2>&1 | FileCheck --check-prefix=MLIR %s
446 # MLIR-LABEL: func @main
447 # MLIR-SAME: (%[[ARG_0:[a-z0-9]+]]: tensor<1x1x1x256x!quant.uniform<i8:f32, 0.21632751372549019:27…
448 # MLIR-SAME: control_outputs = ""
449 # MLIR-SAME: inputs = "input"
450 # MLIR-SAME: outputs = "output"
451 # MLIR: %[[shape:.*]] = constant dense<[1, -1, 31]> : tensor<3xi32>
452 # MLIR: %[[bias:.*]] = "tfl.pseudo_qconst"() {qtype = tensor<186x!quant.uniform<i32:f32:0
453 # MLIR: %[[weight:.*]] = "tfl.pseudo_qconst"() {qtype = tensor<186x1x1x256x!quant.uniform<i…
454 # MLIR: %[[conv:.*]] = "tfl.conv_2d"(%[[ARG_0]], %[[weight]], %[[bias]]) {dilation_h_factor…
[all …]
|
| /external/llvm-project/mlir/docs/Dialects/ |
| D | Vector.md | 5 MLIR supports multi-dimensional `vector` types and custom operations on those
9 in MLIR today and operate on ssa-values of type `vector` along with pattern
13 new proposal but rather a textual documentation of existing MLIR components
19 captures the current codegen paths implemented in MLIR in the various existing
27 abstractions in MLIR.
29 ![`vector` Abstractions in MLIR](https://user-images.githubusercontent.com/10148468/71176949-e85ad0…
40 The existing MLIR `vector`-level dialects are related to the following
60 1. MLIR Rewrite Patterns applied by the MLIR `PatternRewrite` infrastructure
64 2. `Virtual Vector -> Hardware Vector` lowering is specified as a set of MLIR
89 abstraction for all targets and perform cost-based lowering decisions in MLIR
[all …]
|
| D | LLVM.md | 3 This dialect wraps the LLVM IR types and instructions into MLIR types and
5 cover for the differences in the IR structure (e.g., MLIR does not have `phi`
10 "LLVM IR _dialect_" to refer to the MLIR dialect reflecting LLVM instructions
25 The LLVM IR dialect defines a single MLIR type, `LLVM::LLVMType`, that can wrap
50 All operations in the LLVM IR dialect have a custom form in MLIR. The mnemonic
55 MLIR functions are defined by an operation that is not built into the IR itself.
58 use MLIR syntax to express it. They are required to have exactly one result
61 differently by the built-in MLIR function.
233 In order to comply with MLIR design, branch operations in the LLVM IR dialect
234 pass arguments to basic blocks. Successors must be valid block MLIR identifiers
[all …]
|
| /external/tensorflow/tensorflow/compiler/mlir/lite/ |
| D | README.md | 1 # The new [MLIR](https://github.com/llvm/llvm-project/tree/master/mlir) based
6 1. MLIR dialects, transformation passes and utilities for TensorFlow Lite.
17 - Import from GraphDef, in .pb or .pbtxt format, into MLIR.
30 operates on MLIR TensorFlow Lite dialect and is simple/direct translation.
33 for the full list of MLIR passes for conversion from TensorFlow to
|
| /external/llvm-project/mlir/include/mlir/Dialect/PDL/IR/ |
| D | PDLBase.td | 9 // Defines base support for MLIR PDL operations.
26 available in MLIR. This abstraction allows for representing patterns
27 transforming MLIR, as MLIR. This allows for applying all of the benefits
28 that the general MLIR infrastructure provides, to the infrastructure itself.
32 PDL abstracts over various different aspects of patterns and core MLIR data
37 are handle types to MLIR C++ types, with `!pdl.attribute`, `!pdl.operation`,
|
| /external/tensorflow/tensorflow/compiler/mlir/hlo/ |
| D | CMakeLists.txt | 47 # MLIR/LLVM Configuration
50 # Find MLIR to install if we are building standalone. If building as part of
51 # another project, let it handle the MLIR dependency. The dependent project
52 # might use a bundled version of MLIR instead of installing, for instance.
54 find_package(MLIR REQUIRED CONFIG)
|
| D | README.md | 1 # MLIR-HLO: A Standalone "HLO" MLIR-based Compiler
15 MLIR components. It is designed to provide an end-to-end flow independent of
19 [MLIR Developer Guide](https://mlir.llvm.org/getting_started/DeveloperGuide/) in
27 To build the code in this repository, you need a clone of the LLVM/MLIR git
38 We provide a script to configure and build LLVM/MLIR:
58 MLIR-HLO aims to provide an end-to-end compiler for CPU and GPU, as well as
79 [MLIR](https://mlir.llvm.org/) is a compiler infrastructure which intends to
82 roadmap for MLIR is to provide a
83 [Tensor Compute Primitive](https://llvm.discourse.group/c/mlir/MLIR-TCP-WG/36)
90 The work on MLIR-HLO can be seen as a stepping stone towards building TCP, while
[all …]
|
| /external/tensorflow/tensorflow/compiler/mlir/ |
| D | README.md | 1 # MLIR dialects and utilities for TensorFlow, TensorFlow Lite and XLA.
3 This module contains the MLIR
11 See [MLIR's website](https://mlir.llvm.org) for complete documentation.
20 To develop across MLIR core and TensorFlow, it is useful to override the repo
|
| /external/llvm-project/mlir/include/mlir/Reducer/ |
| D | Passes.td | 1 //===-- Passes.td - MLIR Reduce pass definition file -------*- tablegen -*-===//
9 // This file contains definitions of the passes for the MLIR Reduce Tool.
19 let summary = "A general reduction tree pass for the MLIR Reduce Tool";
|
| /external/llvm-project/mlir/examples/toy/Ch2/ |
| D | toyc.cpp | 40 enum InputType { Toy, MLIR }; enumerator
45 cl::values(clEnumValN(MLIR, "mlir",
76 if (inputType != InputType::MLIR && in dumpMLIR()
111 if (inputType == InputType::MLIR) { in dumpAST()
|
| /external/llvm-project/flang/include/flang/Optimizer/Dialect/ |
| D | CMakeLists.txt | 1 # This replicates part of the add_mlir_dialect cmake function from MLIR that
2 # cannot be used her because it expects to be run inside MLIR directory which
13 tablegen(MLIR ${dialect_doc_filename}.md -gen-op-doc "-I${MLIR_MAIN_SRC_DIR}" "-I${MLIR_INCLUDE_DIR…
|
| /external/llvm-project/mlir/examples/standalone/include/Standalone/ |
| D | StandaloneDialect.td | 20 let summary = "A standalone out-of-tree MLIR dialect.";
22 This dialect is an example of an out-of-tree MLIR dialect designed to
23 illustrate the basic setup required to develop MLIR-based tools without
|
| /external/llvm-project/mlir/examples/standalone/ |
| D | README.md | 1 # An out-of-tree MLIR dialect
3 This is an example of an out-of-tree [MLIR](https://mlir.llvm.org/) dialect along with a standalone…
7 This setup assumes that you have built LLVM and MLIR in `$BUILD_DIR` and installed them to `$PREFIX…
|
| /external/llvm-project/mlir/examples/toy/Ch3/ |
| D | toyc.cpp | 42 enum InputType { Toy, MLIR }; enumerator
47 cl::values(clEnumValN(MLIR, "mlir",
77 if (inputType != InputType::MLIR && in loadMLIR()
131 if (inputType == InputType::MLIR) { in dumpAST()
|
| /external/llvm-project/mlir/cmake/modules/ |
| D | MLIRConfig.cmake.in | 1 # This file allows users to call find_package(MLIR) and pick up our targets.
25 # By creating these targets here, subprojects that depend on MLIR's
27 # in-tree with MLIR or not.
|
| /external/llvm-project/mlir/docs/Tutorials/ |
| D | _index.md | 3 This section contains multiple MLIR tutorials.
4 See [Toy tutorial](toy) for an introduction to using MLIR infrastructure.
|
| /external/llvm-project/mlir/examples/toy/Ch4/ |
| D | toyc.cpp | 43 enum InputType { Toy, MLIR }; enumerator
48 cl::values(clEnumValN(MLIR, "mlir",
78 if (inputType != InputType::MLIR && in loadMLIR()
140 if (inputType == InputType::MLIR) { in dumpAST()
|