[mlir][spirv] Add pattern matching for arith.index_cast i1 to index for ArithToSPIRV

[ianayl]

Currently, arith.index_cast gets converted to OpSConvert:

llvm-project/mlir/lib/Conversion/ArithToSPIRV/ArithToSPIRV.cpp

Line 1331 in 9bf5bf3

TypeCastingOpPattern<arith::IndexCastOp, spirv::SConvertOp>,

OpSConvert requires its operands to be of integer type, which poses an issue for i1 since SPIRV distinguishes between booleans and integers. As a result, the following example doesn't get converted, leaving behind illegal ops:

%0 = arith.index_cast %arg0 : i1 to index

This PR adds additional logic to convert arith.index_casts to SPIRV dialect when casting from i1 to index. Converting index_casts from index to i1 is a part of #156031.

[github-actions]

Thank you for submitting a Pull Request (PR) to the LLVM Project!

This PR will be automatically labeled and the relevant teams will be notified.

If you wish to, you can add reviewers by using the "Reviewers" section on this page.

If this is not working for you, it is probably because you do not have write permissions for the repository. In which case you can instead tag reviewers by name in a comment by using @ followed by their GitHub username.

If you have received no comments on your PR for a week, you can request a review by "ping"ing the PR by adding a comment “Ping”. The common courtesy "ping" rate is once a week. Please remember that you are asking for valuable time from other developers.

If you have further questions, they may be answered by the LLVM GitHub User Guide.

You can also ask questions in a comment on this PR, on the LLVM Discord or on the forums.

[llvmbot]

@llvm/pr-subscribers-mlir

@llvm/pr-subscribers-mlir-spirv

Author: Ian Li (ianayl)

Changes

Currently, arith.index_cast gets converted to OpSConvert:

llvm-project/mlir/lib/Conversion/ArithToSPIRV/ArithToSPIRV.cpp

Line 1331 in 9bf5bf3

TypeCastingOpPattern<arith::IndexCastOp, spirv::SConvertOp>,

OpSConvert requires its operands to be of integer type, which poses an issue for i1 since SPIRV distinguishes between booleans and integers. As a result, the following example doesn't get converted, leaving behind illegal ops:

%0 = arith.index_cast %arg0 : i1 to index

This PR adds additional logic to convert arith.index_casts to SPIRV dialect when converting from i1 to index. Converting index_casts from index to i1 will be submitted as an additional PR.

---

Full diff: https://github.com/llvm/llvm-project/pull/155729.diff

2 Files Affected:

• (modified) mlir/lib/Conversion/ArithToSPIRV/ArithToSPIRV.cpp (+30-1)
• (modified) mlir/test/Conversion/ArithToSPIRV/arith-to-spirv.mlir (+9)

diff --git a/mlir/lib/Conversion/ArithToSPIRV/ArithToSPIRV.cpp b/mlir/lib/Conversion/ArithToSPIRV/ArithToSPIRV.cpp
index 265293b83f84c..b9e04e456ff72 100644
--- a/mlir/lib/Conversion/ArithToSPIRV/ArithToSPIRV.cpp
+++ b/mlir/lib/Conversion/ArithToSPIRV/ArithToSPIRV.cpp
@@ -607,6 +607,35 @@ struct UIToFPI1Pattern final : public OpConversionPattern<arith::UIToFPOp> {
   }
 };
 
+//===----------------------------------------------------------------------===//
+// IndexCastOp
+//===----------------------------------------------------------------------===//
+
+/// Converts arith.index_cast to spirv.Select if the type of source is i1 or
+/// vector of i1.
+struct IndexCastI1IndexPattern final : public OpConversionPattern<arith::IndexCastOp> {
+  using OpConversionPattern::OpConversionPattern;
+
+  LogicalResult
+  matchAndRewrite(arith::IndexCastOp op, OpAdaptor adaptor,
+                  ConversionPatternRewriter &rewriter) const override {
+    Type srcType = adaptor.getOperands().front().getType();
+    if (!srcType.isInteger(1))
+      return failure();
+
+    Type dstType = getTypeConverter()->convertType(op.getType());
+    if (!dstType)
+      return getTypeConversionFailure(rewriter, op);
+
+    Location loc = op.getLoc();
+    Value zero = spirv::ConstantOp::getZero(dstType, loc, rewriter);
+    Value one = spirv::ConstantOp::getOne(dstType, loc, rewriter);
+    rewriter.replaceOpWithNewOp<spirv::SelectOp>(
+        op, dstType, adaptor.getOperands().front(), one, zero);
+    return success();
+  }
+};
+
 //===----------------------------------------------------------------------===//
 // ExtSIOp
 //===----------------------------------------------------------------------===//
@@ -1328,7 +1357,7 @@ void mlir::arith::populateArithToSPIRVPatterns(
     TypeCastingOpPattern<arith::SIToFPOp, spirv::ConvertSToFOp>,
     TypeCastingOpPattern<arith::FPToUIOp, spirv::ConvertFToUOp>,
     TypeCastingOpPattern<arith::FPToSIOp, spirv::ConvertFToSOp>,
-    TypeCastingOpPattern<arith::IndexCastOp, spirv::SConvertOp>,
+    TypeCastingOpPattern<arith::IndexCastOp, spirv::SConvertOp>, IndexCastI1IndexPattern,
     TypeCastingOpPattern<arith::IndexCastUIOp, spirv::UConvertOp>,
     TypeCastingOpPattern<arith::BitcastOp, spirv::BitcastOp>,
     CmpIOpBooleanPattern, CmpIOpPattern,
diff --git a/mlir/test/Conversion/ArithToSPIRV/arith-to-spirv.mlir b/mlir/test/Conversion/ArithToSPIRV/arith-to-spirv.mlir
index 6e2352e706acc..e86b04527383d 100644
--- a/mlir/test/Conversion/ArithToSPIRV/arith-to-spirv.mlir
+++ b/mlir/test/Conversion/ArithToSPIRV/arith-to-spirv.mlir
@@ -734,6 +734,15 @@ func.func @index_castui4(%arg0: index) {
   return
 }
 
+// CHECK-LABEL: index_casti1index_1
+func.func @index_casti1index_1(%arg0 : i1) {
+  // CHECK: %[[ZERO:.+]] = spirv.Constant 0 : i32
+  // CHECK: %[[ONE:.+]] = spirv.Constant 1 : i32
+  // CHECK: spirv.Select %{{.+}}, %[[ONE]], %[[ZERO]] : i1, i32
+  %0 = arith.index_cast %arg0 : i1 to index
+  return
+}
+
 // CHECK-LABEL: @bit_cast
 func.func @bit_cast(%arg0: vector<2xf32>, %arg1: i64) {
   // CHECK: spirv.Bitcast %{{.+}} : vector<2xf32> to vector<2xi32>

[github-actions]

✅ With the latest revision this PR passed the C/C++ code formatter.

[ianayl]

Amended PR to have support for vector<?xi1> as well

[kuhar]

@ianayl this needs to be rebased before we can merge

[github-actions]

@ianayl Congratulations on having your first Pull Request (PR) merged into the LLVM Project!

Your changes will be combined with recent changes from other authors, then tested by our build bots. If there is a problem with a build, you may receive a report in an email or a comment on this PR.

Please check whether problems have been caused by your change specifically, as the builds can include changes from many authors. It is not uncommon for your change to be included in a build that fails due to someone else's changes, or infrastructure issues.

How to do this, and the rest of the post-merge process, is covered in detail here.

If your change does cause a problem, it may be reverted, or you can revert it yourself. This is a normal part of LLVM development. You can fix your changes and open a new PR to merge them again.

If you don't get any reports, no action is required from you. Your changes are working as expected, well done!