IITH-Compilers · nishant-sachdeva · Feb 11, 2025 · Feb 13, 2025 · svkeerthy · Feb 15, 2025
diff --git a/.github/workflows/publish.yml b/.github/workflows/publish.yml
@@ -20,10 +20,10 @@ jobs:
       run: |
         sudo python3 -m pip install lit
         sudo python3 -m pip install filecheck
-    - name: Install LLVM-18
+    - name: Install LLVM-20
       run: |
         wget https://apt.llvm.org/llvm.sh
-        sudo bash llvm.sh 18
+        sudo bash llvm.sh 20
     - uses: actions/checkout@v2
     - name: Eigen-setup
       run: |

diff --git a/.github/workflows/test.yml b/.github/workflows/test.yml
@@ -23,10 +23,10 @@ jobs:
       run: |
         sudo python3 -m pip install lit
         sudo python3 -m pip install filecheck
-    - name: Install LLVM-19
+    - name: Install LLVM-20
       run: |
         wget https://apt.llvm.org/llvm.sh
-        sudo bash llvm.sh 19
+        sudo bash llvm.sh 20
     - uses: actions/checkout@v2
     - name: Eigen-setup
       run: |

diff --git a/Manylinux2014_Compliant_Source/manylinux-llvm/Dockerfile b/Manylinux2014_Compliant_Source/manylinux-llvm/Dockerfile
@@ -36,7 +36,7 @@ RUN cmake -DCMAKE_BUILD_TYPE=Release -G Ninja -B build \
 	&& cmake --install build --prefix /root/destdir
 
 WORKDIR /root/llvm
-ARG LLVM_VERSION="18.1.8"
+ARG LLVM_VERSION="20.1.0"
 RUN wget -q "https://github.com/llvm/llvm-project/releases/download/llvmorg-${LLVM_VERSION}/llvm-project-${LLVM_VERSION}.src.tar.xz" \
 	&& tar -xf "llvm-project-${LLVM_VERSION}.src.tar.xz"
 WORKDIR /root/llvm/llvm-project-${LLVM_VERSION}.src/build

diff --git a/Manylinux2014_Compliant_Source/pkg/regen-oracle.sh b/Manylinux2014_Compliant_Source/pkg/regen-oracle.sh
@@ -6,7 +6,7 @@ cd src/test-suite
 
 rm -rf oracle
 
-SEED_VERSION="llvm19"
+SEED_VERSION="llvm20"
 SRC_WD="PE-benchmarks"
 DEST_FOLDER_LL="PE-benchmarks-llfiles-${SEED_VERSION}"
 DEST_FOLDER_SYM="oracle/SYM_${SEED_VERSION}_f"
@@ -18,7 +18,7 @@ mkdir -p ${DEST_FOLDER_LL}
 
 for d in ${SRC_WD}/*.c ${SRC_WD}/*.cpp ${SRC_WD}/*.cc; do
 	echo "Compiling ${d} to IR"
-	name=$(basename ${d}) && oname=${name%.*} && clang-18 -S -emit-llvm -Xclang -disable-O0-optnone ${d} -o ${DEST_FOLDER_LL}/${oname}.ll &
+	name=$(basename ${d}) && oname=${name%.*} && clang-20 -S -emit-llvm -Xclang -disable-O0-optnone ${d} -o ${DEST_FOLDER_LL}/${oname}.ll &
 done
 wait
 

diff --git a/Manylinux2014_Compliant_Source/pkg/tests/test_ir2vec.py b/Manylinux2014_Compliant_Source/pkg/tests/test_ir2vec.py
@@ -12,7 +12,7 @@
 ABS_ACCURACY = 1e-4
 
 TEST_SUITE_DIR = pl.Path(__file__).resolve().parents[3] / "src" / "test-suite"
-SEED_VERSION = "llvm19"
+SEED_VERSION = "llvm20"
 
 ll_files = []
 path = TEST_SUITE_DIR / f"index-{SEED_VERSION}.files"

diff --git a/README.md b/README.md
@@ -6,7 +6,7 @@ Please see [here](https://compilers.cse.iith.ac.in/projects/ir2vec/) for more de
 
 > IR2Vec: LLVM IR Based Scalable Program Embeddings, S. VenkataKeerthy, Rohit Aggarwal, Shalini Jain, Maunendra Sankar Desarkar, Ramakrishna Upadrasta, and Y. N. Srikant
 
-[![LLVM](https://img.shields.io/badge/LLVM-v19.1.7-blue)](https://github.com/llvm/llvm-project/releases/tag/llvmorg-19.1.7)
+[![LLVM](https://img.shields.io/badge/LLVM-v20.1.0-blue)](https://github.com/llvm/llvm-project/releases/tag/llvmorg-20.1.0)
 [![PyPI Version](https://img.shields.io/pypi/v/IR2Vec)](https://pypi.org/project/IR2Vec/)
 ![Tests](https://github.com/IITH-Compilers/IR2Vec/workflows/Tests/badge.svg)
 ![Publish](https://github.com/IITH-Compilers/IR2Vec/workflows/Publish/badge.svg)
@@ -18,7 +18,8 @@ Please see [here](https://compilers.cse.iith.ac.in/projects/ir2vec/) for more de
 
 | LLVM Version | Branch |
 | ------------ | ------ |
-| LLVM 19.1.7 | [main](https://github.com/IITH-Compilers/IR2Vec) |
+| LLVM 20.1.0 | [main](https://github.com/IITH-Compilers/IR2Vec) |
+| LLVM 19.1.7 | [main](https://github.com/IITH-Compilers/IR2Vec/tree/llvm19) |
 | LLVM 18.1.8 | [llvm18](https://github.com/IITH-Compilers/IR2Vec/tree/llvm18) |
 | LLVM 17.0.6 | [llvm17](https://github.com/IITH-Compilers/IR2Vec/tree/llvm17) |
 | LLVM 16.0.1 | [llvm16](https://github.com/IITH-Compilers/IR2Vec/tree/llvm16) |
@@ -29,31 +30,31 @@ Please see [here](https://compilers.cse.iith.ac.in/projects/ir2vec/) for more de
 
 ## Table Of Contents
 - [IR2Vec](#ir2vec)
-  - [LLVM Version Archive](#llvm-version-archive)
-  - [Table Of Contents](#table-of-contents)
-  - [Installation](#installation)
-  - [Python](#python)
-  - [Cpp](#cpp)
-  - [Requirements](#requirements)
-  - [Building from source](#building-from-source)
-  - [Generating program representations](#generating-program-representations)
-    - [Using Binary](#using-binary)
-      - [Command-Line options](#command-line-options)
-      - [Flow-Aware Embeddings](#flow-aware-embeddings)
-      - [Symbolic Embeddings](#symbolic-embeddings)
-  - [Using Libraries](#using-libraries)
-  - [Using Python package (IR2Vec-Wheels)](#using-python-package-ir2vec-wheels)
-    - [Initialization -ir2vec.initEmbedding](#initialization--ir2vecinitembedding)
-    - [getProgramVector](#getprogramvector)
-    - [getFunctionVectors](#getfunctionvectors)
-    - [getInstructionVectors](#getinstructionvectors)
-  - [Example](#example)
-  - [Binaries, Libraries and Wheels - Artifacts](#binaries-libraries-and-wheels---artifacts)
-  - [Experiments](#experiments)
-    - [Note](#note)
-  - [Citation](#citation)
-  - [Contributions](#contributions)
-  - [License](#license)
+    - [LLVM Version Archive](#llvm-version-archive)
+    - [Table Of Contents](#table-of-contents)
+    - [Installation](#installation)
+    - [Python](#python)
+    - [Cpp](#cpp)
+    - [Requirements](#requirements)
+    - [Building from source](#building-from-source)
+    - [Generating program representations](#generating-program-representations)
+        - [Using Binary](#using-binary)
+            - [Command-Line options](#command-line-options)
+            - [Flow-Aware Embeddings](#flow-aware-embeddings)
+            - [Symbolic Embeddings](#symbolic-embeddings)
+    - [Using Libraries](#using-libraries)
+    - [Using Python package (IR2Vec-Wheels)](#using-python-package-ir2vec-wheels)
+        - [Initialization -ir2vec.initEmbedding](#initialization--ir2vecinitembedding)
+        - [getProgramVector](#getprogramvector)
+        - [getFunctionVectors](#getfunctionvectors)
+        - [getInstructionVectors](#getinstructionvectors)
+    - [Example](#example)
+    - [Binaries, Libraries and Wheels - Artifacts](#binaries-libraries-and-wheels---artifacts)
+    - [Experiments](#experiments)
+        - [Note](#note)
+    - [Citation](#citation)
+    - [Contributions](#contributions)
+    - [License](#license)
 
 ## Installation
 
@@ -77,7 +78,11 @@ If you're a C++ developer and require low-level control, optimization, or integr
 ## Requirements
 * cmake (>= 3.13.4)
 * GNU Make (4.2.1)
+<<<<<<< HEAD
 * LLVM (19.1.7) - [src](https://github.com/llvm/llvm-project/tree/release/17.x), [release](https://releases.llvm.org/download.html#19.1.7)
+=======
+* LLVM (20.1.0) - [src](https://github.com/llvm/llvm-project/tree/release/20.x), [release](https://releases.llvm.org/download.html#20.1.0)
+>>>>>>> initial commit - change from llvm18 to llvm20
     * Support for latest LLVM versions would be added soon
 * Eigen library (3.3.7) (Optional)
 * Python (3.6.7)

diff --git a/seed_embeddings/OpenKE/config/Trainer.py b/seed_embeddings/OpenKE/config/Trainer.py
@@ -57,6 +57,7 @@ def __init__(
         self.analogies = analogy.AnalogyScorer(analogy_file=analogy_file)
 
     def load_entity_names(self, index_dir):
+        print("Loading entity names...", index_dir)
         with open(os.path.join(index_dir, "entity2id.txt")) as fEntity:
             content = fEntity.read()
 

diff --git a/seed_embeddings/OpenKE/generate_embedding.py b/seed_embeddings/OpenKE/generate_embedding.py
@@ -95,6 +95,7 @@ def train(arg_conf):
         alpha=0.00729,
         opt_method="Adam",
         checkpoint_dir=checkpoint_dir,
+        index_dir=arg_conf.index_dir,
     )
     trainer.run(
         link_prediction=False,
@@ -137,10 +138,10 @@ def findRep(src, dest, index_dir):
         metavar="DIRECTORY",
         help="Location of the directory entity2id.txt, train2id.txt and relation2id.txt",
         required=False,
-        default="../seed_embeddings/preprocessed/",
+        default="../preprocessed/",
     )
     parser.add_argument(
-        "--epoch", dest="epoch", help="Epochs", required=False, type=int, default=1000
+        "--epoch", dest="epoch", help="Epochs", required=False, type=int, default=20
     )
 
     parser.add_argument(
@@ -188,7 +189,7 @@ def findRep(src, dest, index_dir):
         ),
     )
 
-    findRep(outfilejson, seedfile, arg_conf.index_dir)
+    findRep(outfile, seedfile, arg_conf.index_dir)
 
     print("Training finished...")
     print("seed file : ", seedfile)
diff --git a/seed_embeddings/OpenKE/generate_embedding_ray.py b/seed_embeddings/OpenKE/generate_embedding_ray.py
@@ -261,6 +261,8 @@ def reformat_embeddings(input_str):
         metric = "loss"
         mode = "min"
 
+    print(metric, mode)
+
     scheduler = ASHAScheduler(
         time_attr="training_iteration",
         max_t=arg_conf.epoch,
@@ -270,11 +272,12 @@ def reformat_embeddings(input_str):
         mode=mode,
     )
     optuna = OptunaSearch(metric="loss", mode="min")
+    print("GPU usage ", arg_conf.use_gpu)
 
     if arg_conf.use_gpu:
         train_with_resources = tune.with_resources(
             tune.with_parameters(train, args=arg_conf),
-            resources={"cpu": 8, "gpu": 0.15},
+            resources={"cpu": 2, "gpu": 0.11},
         )
     else:
         train_with_resources = tune.with_resources(
@@ -286,9 +289,9 @@ def reformat_embeddings(input_str):
         param_space=search_space,
         tune_config=TuneConfig(
             search_alg=optuna,
-            max_concurrent_trials=12,
+            max_concurrent_trials=4,
             scheduler=scheduler,
-            num_samples=128,
+            num_samples=16,
         ),
         run_config=RunConfig(
             storage_path=arg_conf.storage_path,
@@ -336,6 +339,7 @@ def reformat_embeddings(input_str):
     # Construct the output file name using the best hyperparameters
     outfile = os.path.join(
         index_dir,
+        "embeddings/",
         "seedEmbedding_{}_{}Dim_{}Alpha_{}batchsize_{}margin.ckpt".format(
             metric,
             dim,
@@ -357,6 +361,7 @@ def reformat_embeddings(input_str):
 
         embeddings_path = os.path.join(
             index_dir,
+            "embeddings/",
             "seedEmbedding_{}_{}Dim_{}Alpha_{}batchsize_{}margin.ckpt".format(
                 metric,
                 dim,

diff --git a/seed_embeddings/OpenKE/launch_training.sh b/seed_embeddings/OpenKE/launch_training.sh
@@ -0,0 +1,4 @@
+#! /bin/bash
+
+python3 generate_embedding_ray.py --is_analogy True --use_gpu True --dim 75 --index_dir "/home/intern23002/iitH/ir2vec/IR2Vec/seed_embeddings/preprocessed" \
+	--analogy_file "/home/intern23002/iitH/ir2vec/IR2Vec/seed_embeddings/OpenKE/analogies.txt"
diff --git a/seed_embeddings/triplets.sh b/seed_embeddings/triplets.sh
@@ -34,7 +34,7 @@ if [ -z $4 ]; then
 	exit
 fi
 
-LLVM_BUILD=$5
+LLVM_BUILD=/usr
 
 if [ -z $LLVM_BUILD ]; then
 	echo "5st arg should have a valid Build path"
@@ -70,13 +70,13 @@ while read p; do
 		fi
 		USED_OPT[$a]=$opt
 		DEBUG echo "opt from $opt"
-		${LLVM_BUILD}/bin/opt-19 -S -$opt $p -o $tmpfile
+		${LLVM_BUILD}/bin/opt-20 -S -$opt $p -o $tmpfile
 		$COLLECT_BUILD/bin/ir2vec -collectIR -o $4 $tmpfile &>/dev/null
 		let "a++"
 		rm "$tmpfile"
 	done &
 	if [ $counter == 100 ]; then
-		sleep 20
+		sleep 3
 		counter=0
 	fi
 

diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt
@@ -69,7 +69,7 @@ if(NOT LLVM_IR2VEC)
   set(LT_LLVM_INSTALL_DIR "" CACHE PATH "LLVM installation directory")
   list(APPEND CMAKE_PREFIX_PATH "${LT_LLVM_INSTALL_DIR}/lib/cmake/llvm/")
 
-  find_package(LLVM 19.1.7 REQUIRED CONFIG)
+  find_package(LLVM 20.1.0 REQUIRED CONFIG)
   message(STATUS "Found LLVM ${LLVM_PACKAGE_VERSION}")
   message(STATUS "Using LLVMConfig.cmake in: ${LLVM_DIR}")
 

diff --git a/src/test-suite/CMakeLists.txt b/src/test-suite/CMakeLists.txt
@@ -70,11 +70,11 @@ endif()
 
 # sanity checks and lit configs
 configure_file(sanity_check.sh.cmake sanity_check.sh @ONLY)
-file(COPY PE-benchmarks-llfiles-llvm19 DESTINATION ./)
+file(COPY PE-benchmarks-llfiles-llvm20 DESTINATION ./)
 file(COPY sqlite3.ll DESTINATION ./)
 file(COPY oracle DESTINATION ./)
 file(COPY ../../vocabulary DESTINATION ./)
-file(COPY index-llvm19.files DESTINATION ./)
+file(COPY index-llvm20.files DESTINATION ./)
 
 
 configure_file(lit.site.cfg.py.in lit.site.cfg.py @ONLY)

diff --git a/...he_closest_pair_from_two_sorted_arrays.ll → ...he_closest_pair_from_two_sorted_arrays.ll b/...he_closest_pair_from_two_sorted_arrays.ll → ...he_closest_pair_from_two_sorted_arrays.ll
@@ -1,7 +1,7 @@
 ; ModuleID = 'PE-benchmarks/Find_the_closest_pair_from_two_sorted_arrays.cpp'
 source_filename = "PE-benchmarks/Find_the_closest_pair_from_two_sorted_arrays.cpp"
 target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-i128:128-f80:128-n8:16:32:64-S128"
-target triple = "x86_64-unknown-linux-gnu"
+target triple = "x86_64-pc-linux-gnu"
 
 module asm ".globl _ZSt21ios_base_library_initv"
 
@@ -195,6 +195,6 @@ attributes #4 = { nocallback nofree nounwind willreturn memory(argmem: readwrite
 !2 = !{i32 7, !"PIE Level", i32 2}
 !3 = !{i32 7, !"uwtable", i32 2}
 !4 = !{i32 7, !"frame-pointer", i32 2}
-!5 = !{!"clang version 19.1.7 (https://github.com/llvm/llvm-project.git cd708029e0b2869e80abe31ddb175f7c35361f90)"}
+!5 = !{!"Ubuntu clang version 20.1.0 (++20250204023448+d185bd94ff77-1~exp1~20250204023606.14)"}
 !6 = distinct !{!6, !7}
 !7 = !{!"llvm.loop.mustprogress"}
diff --git a/...rks-llfiles-llvm19/Iterative_QuickSort.ll → ...rks-llfiles-llvm20/Iterative_QuickSort.ll b/...rks-llfiles-llvm19/Iterative_QuickSort.ll → ...rks-llfiles-llvm20/Iterative_QuickSort.ll
@@ -1,7 +1,7 @@
 ; ModuleID = 'PE-benchmarks/Iterative_QuickSort.cpp'
 source_filename = "PE-benchmarks/Iterative_QuickSort.cpp"
 target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-i128:128-f80:128-n8:16:32:64-S128"
-target triple = "x86_64-unknown-linux-gnu"
+target triple = "x86_64-pc-linux-gnu"
 
 module asm ".globl _ZSt21ios_base_library_initv"
 
@@ -319,7 +319,7 @@ attributes #5 = { nocallback nofree nounwind willreturn memory(argmem: readwrite
 !2 = !{i32 7, !"PIE Level", i32 2}
 !3 = !{i32 7, !"uwtable", i32 2}
 !4 = !{i32 7, !"frame-pointer", i32 2}
-!5 = !{!"clang version 19.1.7 (https://github.com/llvm/llvm-project.git cd708029e0b2869e80abe31ddb175f7c35361f90)"}
+!5 = !{!"Ubuntu clang version 20.1.0 (++20250204023448+d185bd94ff77-1~exp1~20250204023606.14)"}
 !6 = distinct !{!6, !7}
 !7 = !{!"llvm.loop.mustprogress"}
 !8 = distinct !{!8, !7}