[RTG] Add ISA assembly emission pass

include/circt/Dialect/RTG/Transforms/RTGPasses.td

@@ -32,4 +32,36 @@ def ElaborationPass : Pass<"rtg-elaborate", "mlir::ModuleOp"> {
   let dependentDialects = ["mlir::arith::ArithDialect"];
 }
 
+def EmitRTGISAAssemblyPass : Pass<"rtg-emit-isa-assembly", "mlir::ModuleOp"> {
+  let summary = "Elaborate the contexts of RTG";
+  let description = [{
+    Emits all 'rtg.test's in the IR in a format understood by assemblers.
+
+    There are two options to specify lists of instructions that are not
+    supported by the assembler. For instructions in any of those lists, this
+    pass will emit the equivalent binary representation.
+
+    This pass operates on the `InstructionOpInterface` and folds constant-like
+    operations to support downstream dialects.
+  }];
+
+  let options = [
+    Option<"splitOutput", "split-output", "bool", /*default=*/"false",
+      "If 'true' emits one file per 'rtg.test' in the IR. The name of the "
+      "file matches the test name and is placed in 'path'. Otherwise, path "
+      "is interpreted as the full file path including filename.">,
+    Option<"path", "path", "std::string", /*default=*/"",
+      "The directory or file path in which the output files should be "
+      "created. If empty is is emitted to stderr (not allowed if "
+      "'split-output' is set to 'true')">,
+    Option<"unsupportedInstructionsFile", "unsupported-instructions-file",
+      "std::string", /*default=*/"",
+      "An absolute path to a file with a list of instruction names not "
+      "supported by the assembler.">,
+    ListOption<"unsupportedInstructions", "unsupported-instructions",
+      "std::string",
+      "A list of ISA instruction names not supported by the assembler.">,
+  ];
+}
+
 #endif // CIRCT_DIALECT_RTG_TRANSFORMS_RTGPASSES_TD

lib/Dialect/RTG/Transforms/CMakeLists.txt

@@ -1,5 +1,6 @@
 add_circt_dialect_library(CIRCTRTGTransforms
   ElaborationPass.cpp
+  EmitRTGISAAssemblyPass.cpp
 
   DEPENDS
   CIRCTRTGTransformsIncGen
@@ -9,6 +10,7 @@ add_circt_dialect_library(CIRCTRTGTransforms
 
   LINK_LIBS PRIVATE
   CIRCTRTGDialect
+  CIRCTSupport
   MLIRArithDialect
   MLIRIR
   MLIRPass

lib/Dialect/RTG/Transforms/EmitRTGISAAssemblyPass.cpp

@@ -0,0 +1,249 @@
+//===- EmitRTGISAAssemblyPass.cpp - RTG Assembly Emitter ------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+//
+// This is the main ISA Assembly emitter implementation for the RTG dialect.
+//
+//===----------------------------------------------------------------------===//
+
+#include "circt/Dialect/RTG/IR/RTGISAAssemblyOpInterfaces.h"
+#include "circt/Dialect/RTG/IR/RTGOps.h"
+#include "circt/Dialect/RTG/Transforms/RTGPasses.h"
+#include "circt/Support/Path.h"
+#include "mlir/IR/Threading.h"
+#include "mlir/Support/FileUtilities.h"
+#include "llvm/ADT/SmallString.h"
+#include "llvm/ADT/StringRef.h"
+#include "llvm/ADT/StringSet.h"
+#include "llvm/Support/FileSystem.h"
+#include "llvm/Support/Path.h"
+#include "llvm/Support/ToolOutputFile.h"
+#include "llvm/Support/raw_ostream.h"
+#include <fstream>
+
+namespace circt {
+namespace rtg {
+#define GEN_PASS_DEF_EMITRTGISAASSEMBLYPASS
+#include "circt/Dialect/RTG/Transforms/RTGPasses.h.inc"
+} // namespace rtg
+} // namespace circt
+
+using namespace circt;
+using namespace rtg;
+
+#define DEBUG_TYPE "emit-rtg-isa-assembly"
+
+namespace {
+
+class Emitter {
+public:
+  Emitter(llvm::raw_ostream &os,
+          const llvm::StringSet<llvm::MallocAllocator> &unsupportedInstr)
+      : os(os), unsupportedInstr(unsupportedInstr) {}
+
+  LogicalResult emitInstruction(InstructionOpInterface instr) {
+    os << llvm::indent(4);
+    bool useBinary = unsupportedInstr.contains(instr->getName().getStringRef());
+
+    // TODO: we cannot just assume that double-slash is the way to do a line
+    // comment
+    if (useBinary)
+      os << "// ";
+
+    SmallVector<Attribute> operands;
+    for (auto operand : instr->getOperands()) {
+      auto attr = state.lookup(operand);
+      if (!attr)
+        return failure();
+
+      operands.push_back(attr);
+    }
+
+    instr.printInstructionAssembly(os, operands);
+    os << "\n";
+
+    if (!useBinary)
+      return success();
+
+    os << llvm::indent(4);
+    // TODO: don't hardcode '.word'
+    os << ".word 0x";
+    instr.printInstructionBinary(os, operands);
+    os << "\n";
+
+    return success();
+  }
+
+  LogicalResult emitTest(rtg::TestOp test, bool emitHeaderFooter = false) {
+    if (emitHeaderFooter)
+      os << "// Begin of " << test.getSymName() << "\n\n";
+
+    for (auto &op : *test.getBody()) {
+      if (op.hasTrait<OpTrait::ConstantLike>()) {
+        SmallVector<OpFoldResult> results;
+        if (failed(op.fold(results)))
+          return failure();
+
+        for (auto [val, res] : llvm::zip(op.getResults(), results)) {
+          auto attr = res.dyn_cast<Attribute>();
+          if (!attr)
+            return failure();
+
+          state[val] = attr;
+        }
+
+        continue;
+      }
+
+      if (auto instr = dyn_cast<InstructionOpInterface>(&op)) {
+        if (failed(emitInstruction(instr)))
+          return failure();
+
+        continue;
+      }
+
+      return op.emitError("emitter unknown RTG operation");
+    }
+
+    state.clear();
+
+    if (emitHeaderFooter)
+      os << "\n// End of " << test.getSymName() << "\n\n";
+
+    return success();
+  }
+
+private:
+  /// Output Stream.
+  llvm::raw_ostream &os;
+
+  /// Instructions to emit in binary.
+  const llvm::StringSet<llvm::MallocAllocator> &unsupportedInstr;
+
+  /// Evaluated values.
+  DenseMap<Value, Attribute> state;
+};
+
+} // namespace
+
+static void parseUnsupportedInstructionsFile(
+    const std::string &unsupportedInstructionsFile,
+    llvm::StringSet<llvm::MallocAllocator> &unsupportedInstrs) {
+  if (!unsupportedInstructionsFile.empty()) {
+    std::ifstream input(unsupportedInstructionsFile);
+    std::string token;
+    while (std::getline(input, token, ',')) {
+      auto trimmed = StringRef(token).trim();
+      if (!trimmed.empty())
+        unsupportedInstrs.insert(trimmed);
+    }
+  }
+}
+
+static std::unique_ptr<llvm::ToolOutputFile>
+createOutputFile(StringRef filename, StringRef dirname,
+                 function_ref<InFlightDiagnostic()> emitError) {
+  // Determine the output path from the output directory and filename.
+  SmallString<128> outputFilename(dirname);
+  appendPossiblyAbsolutePath(outputFilename, filename);
+  auto outputDir = llvm::sys::path::parent_path(outputFilename);
+
+  // Create the output directory if needed.
+  std::error_code error = llvm::sys::fs::create_directories(outputDir);
+  if (error) {
+    emitError() << "cannot create output directory \"" << outputDir
+                << "\": " << error.message();
+    return {};
+  }
+
+  // Open the output file.
+  std::string errorMessage;
+  auto output = mlir::openOutputFile(outputFilename, &errorMessage);
+  if (!output)
+    emitError() << errorMessage;
+  return output;
+}
+
+//===----------------------------------------------------------------------===//
+// EmitRTGISAAssemblyPass
+//===----------------------------------------------------------------------===//
+
+namespace {
+struct EmitRTGISAAssemblyPass
+    : public rtg::impl::EmitRTGISAAssemblyPassBase<EmitRTGISAAssemblyPass> {
+  using Base::Base;
+
+  void runOnOperation() override;
+  /// Emit each 'rtg.test' into a separate file using the test's name as the
+  /// filename.
+  LogicalResult
+  emitSplit(const llvm::StringSet<llvm::MallocAllocator> &unsupportedInstr);
+  /// Emit all tests into a single file (or print them to stderr if no file path
+  /// is given).
+  LogicalResult
+  emit(const llvm::StringSet<llvm::MallocAllocator> &unsupportedInstr);
+};
+} // namespace
+
+void EmitRTGISAAssemblyPass::runOnOperation() {
+  if ((!path.hasValue() || path.empty()) && splitOutput) {
+    getOperation().emitError("'split-output' option only valid in combination "
+                             "with a valid 'path' argument");
+    return signalPassFailure();
+  }
+
+  llvm::StringSet<llvm::MallocAllocator> unsupportedInstr(
+      unsupportedInstructions);
+  parseUnsupportedInstructionsFile(unsupportedInstructionsFile.getValue(),
+                                   unsupportedInstr);
+
+  if (splitOutput) {
+    if (failed(emitSplit(unsupportedInstr)))
+      return signalPassFailure();
+
+    return;
+  }
+
+  if (failed(emit(unsupportedInstr)))
+    return signalPassFailure();
+}
+
+LogicalResult EmitRTGISAAssemblyPass::emit(
+    const llvm::StringSet<llvm::MallocAllocator> &unsupportedInstr) {
+  std::unique_ptr<llvm::ToolOutputFile> file;
+  bool emitToFile = path.hasValue() && !path.empty();
+  if (emitToFile) {
+    file = createOutputFile(path, std::string(),
+                            [&]() { return getOperation().emitError(); });
+    if (!file)
+      return failure();
+
+    file->keep();
+  }
+
+  Emitter emitter(emitToFile ? file->os() : llvm::errs(), unsupportedInstr);
+  for (auto test : getOperation().getOps<TestOp>())
+    if (failed(emitter.emitTest(test, true)))
+      return failure();
+
+  return success();
+}
+
+LogicalResult EmitRTGISAAssemblyPass::emitSplit(
+    const llvm::StringSet<llvm::MallocAllocator> &unsupportedInstr) {
+  auto tests = getOperation().getOps<TestOp>();
+  return failableParallelForEach(
+      &getContext(), tests.begin(), tests.end(), [&](rtg::TestOp test) {
+        auto res = createOutputFile(test.getSymName().str() + ".s", path,
+                                    [&]() { return test.emitError(); });
+        if (!res)
+          return failure();
+
+        res->keep();
+        return Emitter(res->os(), unsupportedInstr).emitTest(test);
+      });
+}

test/Dialect/RTG/Transform/emit-rtg-isa-assembly-split-invalid.mlir

@@ -0,0 +1,6 @@
+// RUN: circt-opt --rtg-emit-isa-assembly=split-output=true --verify-diagnostics %s
+
+// expected-error @below {{'split-output' option only valid in combination with a valid 'path' argument}}
+module {
+  rtg.test @test0 : !rtg.dict<> {}
+}

test/Dialect/RTG/Transform/emit-rtg-isa-assembly-split.mlir

@@ -0,0 +1,24 @@
+// RUN: circt-opt --rtg-emit-isa-assembly="path=%T split-output=true" %s && FileCheck %s --input-file=%T/test0.s --check-prefix=CHECK-TEST0 && FileCheck %s --input-file=%T/test1.s --check-prefix=CHECK-TEST1
+// RUN: circt-opt --rtg-emit-isa-assembly="path=%t split-output=false" %s && FileCheck %s --input-file=%t --check-prefixes=CHECK,CHECK-TEST0,CHECK-TEST1
+
+// CHECK: Begin of test0
+// CHECK-EMPTY:
+
+rtg.test @test0 : !rtg.dict<> {
+  // CHECK-TEST0: ebreak
+  rtgtest.ebreak
+}
+
+// CHECK-EMPTY:
+// CHECK: End of test0
+// CHECK-EMPTY:
+// CHECK-NEXT: Begin of test1
+// CHECK-EMPTY:
+
+rtg.test @test1 : !rtg.dict<> {
+  // CHECK-TEST1: ecall
+  rtgtest.ecall
+}
+
+// CHECK-EMPTY:
+// CHECK-NEXT: End of test1

test/Dialect/RTG/Transform/emit-rtg-isa-assembly.mlir

@@ -0,0 +1,58 @@
+// RUN: circt-opt --rtg-emit-isa-assembly %s 2>&1 >/dev/null | FileCheck %s --check-prefix=CHECK-ALLOWED --match-full-lines --strict-whitespace
+// RUN: circt-opt --rtg-emit-isa-assembly="unsupported-instructions=rtgtest.ebreak,rtgtest.ecall unsupported-instructions-file=%S/unsupported-instr.txt" %s 2>&1 >/dev/null | FileCheck %s --match-full-lines --strict-whitespace
+
+// CHECK:// Begin of test0
+// CHECK-EMPTY:
+// CHECK-ALLOWED:// Begin of test0
+// CHECK-ALLOWED-EMPTY:
+
+rtg.test @test0 : !rtg.dict<> {
+  %rd = rtg.fixed_reg #rtgtest.ra
+  %rs = rtg.fixed_reg #rtgtest.s0
+  %imm = rtgtest.immediate #rtgtest.imm12<0>
+
+  // CHECK-ALLOWED-NEXT:    jalr ra, 0(s0)
+  // CHECK-NEXT:    // jalr ra, 0(s0)
+  // CHECK-NEXT:    .word 0x400E7
+  rtgtest.jalr %rd, %rs, %imm
+
+  // CHECK-ALLOWED-NEXT:    lb ra, 0(s0)
+  // CHECK-NEXT:    // lb ra, 0(s0)
+  // CHECK-NEXT:    .word 0x40083
+  rtgtest.lb %rd, %rs, %imm
+
+  // CHECK-ALLOWED-NEXT:    lh ra, 0(s0)
+  // CHECK-NEXT:    // lh ra, 0(s0)
+  // CHECK-NEXT:    .word 0x41083
+  rtgtest.lh %rd, %rs, %imm
+
+  // CHECK-ALLOWED-NEXT:    lw ra, 0(s0)
+  // CHECK-NEXT:    // lw ra, 0(s0)
+  // CHECK-NEXT:    .word 0x42083
+  rtgtest.lw %rd, %rs, %imm
+
+  // CHECK-ALLOWED-NEXT:    lbu ra, 0(s0)
+  // CHECK-NEXT:    // lbu ra, 0(s0)
+  // CHECK-NEXT:    .word 0x44083
+  rtgtest.lbu %rd, %rs, %imm
+
+  // CHECK-ALLOWED-NEXT:    lhu ra, 0(s0)
+  // CHECK-NEXT:    // lhu ra, 0(s0)
+  // CHECK-NEXT:    .word 0x45083
+  rtgtest.lhu %rd, %rs, %imm
+
+  // CHECK-ALLOWED-NEXT:    ebreak
+  // CHECK-NEXT:    // ebreak
+  // CHECK-NEXT:    .word 0x100073
+  rtgtest.ebreak
+
+  // CHECK-ALLOWED-NEXT:    ecall
+  // CHECK-NEXT:    // ecall
+  // CHECK-NEXT:    .word 0x73
+  rtgtest.ecall
+}
+
+// CHECK-EMPTY:
+// CHECK-NEXT:// End of test0
+// CHECK-ALLOWED-EMPTY:
+// CHECK-ALLOWED-NEXT:// End of test0

test/Dialect/RTG/Transform/unsupported-instr.txt

@@ -0,0 +1 @@
+rtgtest.jalr,rtgtest.lb,rtgtest.lh,rtgtest.lw,rtgtest.lbu,rtgtest.lhu