[MLGO] Add tests for scripts

[boomanaiden154]

This patch adds integration tests for the script entry points. The tests don't exercise all functionality, as that case is better covered by the unit testing already checked in. This ensures that things like flag parsing work and that the scripts are syntactically valid.

[boomanaiden154]

This does a couple additional dependencies, but they're relatively small, will already be present when running check-all, and actually fully exercises the common path of extract_ir.py which we weren't able to do previously in ml-compiler-opt.

[llvmbot]

@llvm/pr-subscribers-mlgo

Author: Aiden Grossman (boomanaiden154)

Changes

This patch adds integration tests for the script entry points. The tests don't exercise all functionality, as that case is better covered by the unit testing already checked in. This ensures that things like flag parsing work and that the scripts are syntactically valid.

---

Full diff: https://github.com/llvm/llvm-project/pull/78878.diff

5 Files Affected:

• (modified) llvm/utils/mlgo-utils/CMakeLists.txt (+1-1)
• (added) llvm/utils/mlgo-utils/tests/corpus/combine_training_corpus_script.test (+29)
• (added) llvm/utils/mlgo-utils/tests/corpus/extract_ir_script.test (+44)
• (added) llvm/utils/mlgo-utils/tests/corpus/make_corpus_script.test (+22)
• (modified) llvm/utils/mlgo-utils/tests/lit.cfg (+9-1)

diff --git a/llvm/utils/mlgo-utils/CMakeLists.txt b/llvm/utils/mlgo-utils/CMakeLists.txt
index 7b303c7639401a..3129331d58c75b 100644
--- a/llvm/utils/mlgo-utils/CMakeLists.txt
+++ b/llvm/utils/mlgo-utils/CMakeLists.txt
@@ -5,7 +5,7 @@ configure_lit_site_cfg(
 
 add_lit_testsuite(check-mlgo-utils "Running mlgo-utils tests"
   ${CMAKE_CURRENT_BINARY_DIR}
-  DEPENDS "FileCheck" "not" "count"
+  DEPENDS "FileCheck" "not" "count" "split-file" "yaml2obj" "llvm-objcopy"
 )
 
 set_target_properties(check-mlgo-utils PROPERTIES FOLDER "Tests")
diff --git a/llvm/utils/mlgo-utils/tests/corpus/combine_training_corpus_script.test b/llvm/utils/mlgo-utils/tests/corpus/combine_training_corpus_script.test
new file mode 100644
index 00000000000000..1aa182146a49ee
--- /dev/null
+++ b/llvm/utils/mlgo-utils/tests/corpus/combine_training_corpus_script.test
@@ -0,0 +1,29 @@
+## Testing that the combine_trainig_corpus script works as expected when
+## invoked.
+
+# RUN: rm -rf %t.dir && mkdir %t.dir
+# RUN: split-file %s %t.dir
+# RUN: %python %scripts_dir/corpus/combine_training_corpus.py --root_dir=%t.dir
+# RUN: cat %t.dir/corpus_description.json | FileCheck %s
+
+## Check that we end up with the same properties as the original corpora
+# CHECK: "has_thinlto": false
+
+## Check that the modules end up in the combined corpus. Order does not matter.
+# CHECK-DAG: "subcorpus1/test1.o"
+# CHECK-DAG: "subcorpus2/test2.o"
+
+#--- subcorpus1/corpus_description.json
+{
+  "has_thinlto": false,
+  "modules": [
+    "test1.o"
+  ]
+}
+#--- subcorpus2/corpus_description.json
+{
+  "has_thinlto": false,
+  "modules": [
+    "test2.o"
+  ]
+}
diff --git a/llvm/utils/mlgo-utils/tests/corpus/extract_ir_script.test b/llvm/utils/mlgo-utils/tests/corpus/extract_ir_script.test
new file mode 100644
index 00000000000000..a7629eb629219d
--- /dev/null
+++ b/llvm/utils/mlgo-utils/tests/corpus/extract_ir_script.test
@@ -0,0 +1,44 @@
+## Test that invoking the extract_ir script work as expected.
+
+# RUN: rm -rf %t.dir && mkdir %t.dir
+# RUN: yaml2obj %s -o %t.dir/test1.o
+# RUN: yaml2obj %s -o %t.dir/test2.o
+# RUN: rm -rf %t.dir.out && mkdir %t.dir.out
+
+# RUN: %python %scripts_dir/corpus/extract_ir.py --input=%t.dir --input_type=directory --output_dir=%t.dir.out --llvm_objcopy_path=llvm-objcopy
+# RUN: cat %t.dir.out/corpus_description.json | FileCheck %s
+
+## Check that this is not a thinLTO build
+# CHECK: "has_thinlto": false
+## Check that the expected modules end up in the corpus description
+# CHECK-DAG: "test1.o"
+# CHECK-DAG: "test2.o"
+
+# RUN: ls %t.dir.out | FileCheck %s --check-prefix CHECK-DIR
+
+# CHECK-DIR: test1.o.bc
+# CHECK-DIR: test1.o.cmd
+# CHECK-DIR: test2.o.bc
+# CHECK-DIR: test2.o.cmd
+
+--- !ELF
+FileHeader:
+  Class:           ELFCLASS64
+  Data:            ELFDATA2LSB
+  Type:            ET_REL
+  Machine:         EM_X86_64
+  SectionHeaderStringTable: .strtab
+Sections:
+  - Name:            .text
+    Type:            SHT_PROGBITS
+    Flags:           [ SHF_ALLOC, SHF_EXECINSTR ]
+    AddressAlign:    0x10
+    Content:         55
+  - Name:            .llvmbc
+    Type:            SHT_PROGBITS
+    AddressAlign:    0x1
+    Content:         55
+  - Name:            .llvmcmd
+    Type:            SHT_PROGBITS
+    AddressAlign:    0x1
+    Content:         ff
diff --git a/llvm/utils/mlgo-utils/tests/corpus/make_corpus_script.test b/llvm/utils/mlgo-utils/tests/corpus/make_corpus_script.test
new file mode 100644
index 00000000000000..f4f97544bce47d
--- /dev/null
+++ b/llvm/utils/mlgo-utils/tests/corpus/make_corpus_script.test
@@ -0,0 +1,22 @@
+## Testing that the make_corpus script works as expected when invoked.
+
+# RUN: rm -rf %t.dir && mkdir %t.dir
+# RUN: touch %t.dir/test1.bc
+# RUN: touch %t.dir/test2.bc
+# RUN: rm -rf %t.out.dir && mkdir %t.out.dir
+
+# RUN: %python %scripts_dir/corpus/make_corpus.py --input_dir=%t.dir --output_dir=%t.out.dir --default_args="-test"
+
+# RUN: cat %t.out.dir/corpus_description.json | FileCheck %s
+
+## Check that we get the expected command in the global command override
+# CHECK: "-test"
+# CHECK: "has_thinlto": false
+## Check that the modules are in the corpus description
+# CHECK: "test1"
+# CHECK: "test2"
+
+# RUN: ls %t.out.dir | FileCheck %s --check-prefix CHECK-DIR
+
+# CHECK-DIR: test1.bc
+# CHECK-DIR: test2.bc
diff --git a/llvm/utils/mlgo-utils/tests/lit.cfg b/llvm/utils/mlgo-utils/tests/lit.cfg
index 055f0945942fc1..0f6137e5e91383 100644
--- a/llvm/utils/mlgo-utils/tests/lit.cfg
+++ b/llvm/utils/mlgo-utils/tests/lit.cfg
@@ -1,3 +1,5 @@
+import os
+
 import lit.formats
 
 from lit.llvm import llvm_config
@@ -5,7 +7,7 @@ from lit.llvm import llvm_config
 config.name = "mlgo-utils"
 config.test_format = lit.formats.ShTest(execute_external=False)
 
-config.suffixes = [".py"]
+config.suffixes = [".py", ".test"]
 
 config.test_source_root = os.path.dirname(__file__)
 config.test_exec_root = config.obj_root
@@ -13,3 +15,9 @@ config.test_exec_root = config.obj_root
 config.environment["PYTHONPATH"] = os.path.join(config.src_root, "utils", "mlgo-utils")
 
 llvm_config.use_default_substitutions()
+config.substitutions.append(("split-file", llvm_config.use_llvm_tool("split-file")))
+config.substitutions.append(("yaml2obj", llvm_config.use_llvm_tool("yaml2obj")))
+config.substitutions.append(("llvm-objcopy", llvm_config.use_llvm_tool("llvm-objcopy")))
+
+scripts_dir = os.path.join(config.src_root, "utils/mlgo-utils/mlgo")
+config.substitutions.append(("%scripts_dir", scripts_dir))