Regenerate the oracle for Python testing

Python wheels are built in GLIBc ~ 2.17 so some functions are not
demangled properly. So we rebuild them using the ir2vec binary and
test the Python output against it.

The ir2vec binary output is tested against the actual oracle in another
workflow. So, this should be fine.

Manylinux2014_Compliant_Source/pkg/build.sh

@@ -1,6 +1,8 @@
 #!/bin/bash
 
 set -eu
+export CXX=clang++
+export CC=clang
 
 rm -rf build || true
 mkdir build
@@ -18,3 +20,5 @@ cmake -DCMAKE_BUILD_TYPE=Release ../src && make -j"$(nproc)" && make install
 cd ..
 cp src/include/utils.h Manylinux2014_Compliant_Source/pkg/IR2Vec/
 cp vocabulary/seedEmbeddingVocab-*.txt Manylinux2014_Compliant_Source/pkg/IR2Vec/
+
+bash Manylinux2014_Compliant_Source/pkg/regen-oracle.sh

Manylinux2014_Compliant_Source/pkg/regen-oracle.sh

@@ -0,0 +1,39 @@
+#!/bin/bash
+
+set -eu
+
+cd src/test-suite
+
+rm -rf oracle
+
+SEED_VERSION="llvm12"
+SRC_WD="PE-benchmarks"
+DEST_FOLDER_LL="PE-benchmarks-llfiles-${SEED_VERSION}"
+DEST_FOLDER_SYM="oracle/SYM_${SEED_VERSION}_f"
+DEST_FOLDER_FA="oracle/FA_${SEED_VERSION}_f"
+DEST_FOLDER_SYM_P="oracle/SYM_${SEED_VERSION}"
+DEST_FOLDER_FA_P="oracle/FA_${SEED_VERSION}"
+
+mkdir -p ${DEST_FOLDER_LL}
+
+for d in ${SRC_WD}/*.c ${SRC_WD}/*.cpp ${SRC_WD}/*.cc; do
+	name=$(basename ${d}) && oname=${name%.*} && clang -S -emit-llvm -Xclang -disable-O0-optnone ${d} -o ${DEST_FOLDER}/${oname}.ll &
+done
+wait
+
+mkdir -p ${DEST_FOLDER_SYM}
+mkdir -p ${DEST_FOLDER_FA}
+mkdir -p ${DEST_FOLDER_SYM_P}
+mkdir -p ${DEST_FOLDER_FA_P}
+
+IR2VEC_PATH=../../build/bin/ir2vec
+
+VOCAB_PATH="../../vocabulary/seedEmbeddingVocab-300-${SEED_VERSION}.txt"
+
+while IFS= read -r d; do
+	${IR2VEC_PATH} -sym -vocab=${VOCAB_PATH} -o ${DEST_FOLDER_SYM}/ir2vec.txt -level f ${d} &>/dev/null
+	${IR2VEC_PATH} -fa -vocab=${VOCAB_PATH} -o ${DEST_FOLDER_FA}/ir2vec.txt -level f ${d} &>/dev/null
+	${IR2VEC_PATH} -sym -vocab=${VOCAB_PATH} -o ${DEST_FOLDER_SYM_P}/ir2vec.txt -level p ${d} >/dev/null
+	${IR2VEC_PATH} -fa -vocab=${VOCAB_PATH} -o ${DEST_FOLDER_FA_P}/ir2vec.txt -level p ${d} >/dev/null
+done <index-${SEED_VERSION}.files
+wait

Manylinux2014_Compliant_Source/pkg/setup.py

@@ -14,7 +14,7 @@
 VERSION = ""
 DESCRIPTION = ""
 
-with (pl.Path(__file__).parents[2] / "src" / "CMakeLists.txt").open() as f:
+with (pl.Path(__file__).resolve().parents[2] / "src" / "CMakeLists.txt").open() as f:
     for line in f:
         if not VERSION:
             vmatch = version_regex.match(line)  # Not using walrus because Python3.6
@@ -34,7 +34,9 @@
 
 
 def get_llvm_files():
-    out = sp.run(["llvm-config", "--libfiles"] + LLVM_LIBS, stdout=sp.PIPE)
+    out = sp.run(
+        ["llvm-config", "--link-static", "--libfiles"] + LLVM_LIBS, stdout=sp.PIPE
+    )
     files = out.stdout.decode("utf8").split()
     return files
 

Manylinux2014_Compliant_Source/pkg/tests/test_ir2vec.py

@@ -6,7 +6,7 @@
 
 ABS_ACCURACY = 1e-4
 
-TEST_SUITE_DIR = pl.Path(__file__).parents[3] / "src" / "test-suite"
+TEST_SUITE_DIR = pl.Path(__file__).resolve().parents[3] / "src" / "test-suite"
 SEED_VERSION = (
     list(IR2Vec.preparation.install_loc.glob("seed*.txt"))[0]
     .name.split(".")[0]