CGO 2022 – Proceedings

Frontmatter

GPU

A Compiler Framework for Optimizing Dynamic Parallelism on GPUs
Mhd Ghaith Olabi, Juan Gómez Luna, Onur Mutlu, Wen-mei Hwu, and Izzat El Hajj
(American University of Beirut, Lebanon; ETH Zurich, Switzerland; University of Illinois at Urbana-Champaign, USA; NVIDIA, USA)


@InProceedings{CGO22p1, author = {Mhd Ghaith Olabi and Juan Gómez Luna and Onur Mutlu and Wen-mei Hwu and Izzat El Hajj}, title = {A Compiler Framework for Optimizing Dynamic Parallelism on GPUs}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {1-0}, doi = {10.1109/CGO53902.2022.9741284}, year = {2022}, } Publisher's Version Published Artifact Artifacts Available Artifacts Reusable Results Reproduced

Automatic Horizontal Fusion for GPU Kernels
Ao Li, Bojian Zheng, Gennady Pekhimenko, and Fan Long
(Carnegie Mellon University, USA; University of Toronto, Canada)


@InProceedings{CGO22p16, author = {Ao Li and Bojian Zheng and Gennady Pekhimenko and Fan Long}, title = {Automatic Horizontal Fusion for GPU Kernels}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {16-15}, doi = {10.1109/CGO53902.2022.9741270}, year = {2022}, } Publisher's Version Published Artifact Artifacts Available Artifacts Reusable Results Reproduced

DARM: Control-Flow Melding for SIMT Thread Divergence Reduction
Charitha Saumya, Kirshanthan Sundararajah, and Milind Kulkarni
(Purdue University, USA)


@InProceedings{CGO22p31, author = {Charitha Saumya and Kirshanthan Sundararajah and Milind Kulkarni}, title = {DARM: Control-Flow Melding for SIMT Thread Divergence Reduction}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {31-30}, doi = {10.1109/CGO53902.2022.9741285}, year = {2022}, } Publisher's Version Published Artifact Artifacts Available Artifacts Reusable Results Reproduced

Efficient Execution of OpenMP on GPUs
Joseph Huber, Melanie Cornelius, Giorgis Georgakoudis, Shilei Tian, Jose M Monslave Diaz, Kuter Dinel, Barbara Chapman, and Johannes Doerfert
(Oak Ridge National Laboratory, USA; Illinois Institute of Technology, USA; Lawrence Livermore National Laboratory, USA; Stony Brook University, USA; Argonne National Laboratory, USA; Düzce University, Turkey)


@InProceedings{CGO22p46, author = {Joseph Huber and Melanie Cornelius and Giorgis Georgakoudis and Shilei Tian and Jose M Monslave Diaz and Kuter Dinel and Barbara Chapman and Johannes Doerfert}, title = {Efficient Execution of OpenMP on GPUs}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {46-45}, doi = {10.1109/CGO53902.2022.9741290}, year = {2022}, } Publisher's Version Published Artifact Artifacts Available Artifacts Functional Results Reproduced

Domain-Specific Compilation

GraphIt to CUDA Compiler in 2021 LOC: A Case for High-Performance DSL Implementation via Staging with BuilDSL
Ajay Brahmakshatriya and Saman Amarasinghe
(Massachusetts Institute of Technology, USA)


@InProceedings{CGO22p61, author = {Ajay Brahmakshatriya and Saman Amarasinghe}, title = {GraphIt to CUDA Compiler in 2021 LOC: A Case for High-Performance DSL Implementation via Staging with BuilDSL}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {61-60}, doi = {10.1109/CGO53902.2022.9741280}, year = {2022}, } Publisher's Version Published Artifact Artifacts Available Artifacts Reusable Results Reproduced

A Compiler for Sound Floating-Point Computations using Affine Arithmetic
Joao Rivera, Franz Franchetti, and Markus Püschel
(ETH Zurich, Switzerland; Carnegie Mellon University, USA)


@InProceedings{CGO22p76, author = {Joao Rivera and Franz Franchetti and Markus Püschel}, title = {A Compiler for Sound Floating-Point Computations using Affine Arithmetic}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {76-75}, doi = {10.1109/CGO53902.2022.9741286}, year = {2022}, } Publisher's Version Published Artifact Artifacts Available Artifacts Reusable Results Reproduced

Aggregate Update Problem for Multi-clocked Dataflow Languages
Hannes Kallwies, Martin Leucker, Torben Scheffel, Malte Schmitz, and Daniel Thoma
(University of Lübeck, Germany)


@InProceedings{CGO22p91, author = {Hannes Kallwies and Martin Leucker and Torben Scheffel and Malte Schmitz and Daniel Thoma}, title = {Aggregate Update Problem for Multi-clocked Dataflow Languages}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {91-90}, doi = {10.1109/CGO53902.2022.9741275}, year = {2022}, } Publisher's Version Published Artifact Artifacts Available Artifacts Reusable Results Reproduced

Performance

CompilerGym: Robust, Performant Compiler Optimization Environments for AI Research
Chris Cummins, Bram Wasti, Jiadong Guo, Brandon Cui, Jason Ansel, Sahir Gomez, Somya Jain, Jia Liu, Olivier Teytaud, Benoit Steiner, Yuandong Tian, and Hugh Leather
(Meta, USA)


@InProceedings{CGO22p106, author = {Chris Cummins and Bram Wasti and Jiadong Guo and Brandon Cui and Jason Ansel and Sahir Gomez and Somya Jain and Jia Liu and Olivier Teytaud and Benoit Steiner and Yuandong Tian and Hugh Leather}, title = {CompilerGym: Robust, Performant Compiler Optimization Environments for AI Research}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {106-105}, doi = {10.1109/CGO53902.2022.9741258}, year = {2022}, } Publisher's Version Published Artifact Artifacts Available Artifacts Reusable Results Reproduced

PALMED: Throughput Characterization for Superscalar Architectures
Nicolas Derumigny, Théophile Bastian, Fabian Gruber, Guillaume Iooss, Christophe Guillon, Louis-Noël Pouchet, and Fabrice Rastello
(Grenoble Alps University, France; Inria, France; CNRS, France; Grenoble INP, France; Colorado State University, USA; STMicroelectronics, France)


@InProceedings{CGO22p121, author = {Nicolas Derumigny and Théophile Bastian and Fabian Gruber and Guillaume Iooss and Christophe Guillon and Louis-Noël Pouchet and Fabrice Rastello}, title = {PALMED: Throughput Characterization for Superscalar Architectures}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {121-120}, doi = {10.1109/CGO53902.2022.9741289}, year = {2022}, } Publisher's Version Published Artifact Artifacts Available Artifacts Reusable Results Reproduced

SRTuner: Effective Compiler Optimization Customization by Exposing Synergistic Relations
Sunghyun Park, Salar Latifi, Yongjun Park, Armand Behroozi, Byungsoo Jeon, and Scott Mahlke
(University of Michigan at Ann Arbor, USA; Hanyang University, South Korea; Carnegie Mellon University, USA)


@InProceedings{CGO22p136, author = {Sunghyun Park and Salar Latifi and Yongjun Park and Armand Behroozi and Byungsoo Jeon and Scott Mahlke}, title = {SRTuner: Effective Compiler Optimization Customization by Exposing Synergistic Relations}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {136-135}, doi = {10.1109/CGO53902.2022.9741263}, year = {2022}, } Publisher's Version Published Artifact Artifacts Available Artifacts Functional Results Reproduced

Binary Techniques

Recovering Container Class Types in C++ Binaries
Xudong Wang, Xuezheng Xu, Qingan Li, Mengting Yuan, and Jingling Xue
(UNSW, Australia; Wuhan University, China)


@InProceedings{CGO22p151, author = {Xudong Wang and Xuezheng Xu and Qingan Li and Mengting Yuan and Jingling Xue}, title = {Recovering Container Class Types in C++ Binaries}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {151-150}, doi = {10.1109/CGO53902.2022.9741274}, year = {2022}, } Publisher's Version Published Artifact Artifacts Available Artifacts Reusable Results Reproduced

Automatic Generation of Debug Headers through BlackBox Equivalence Checking
Vaibhav Kiran Kurhe, Pratik Karia, Shubhani Gupta, Abhishek Rose, and Sorav Bansal
(IIT Delhi, India)


@InProceedings{CGO22p166, author = {Vaibhav Kiran Kurhe and Pratik Karia and Shubhani Gupta and Abhishek Rose and Sorav Bansal}, title = {Automatic Generation of Debug Headers through BlackBox Equivalence Checking}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {166-165}, doi = {10.1109/CGO53902.2022.9741273}, year = {2022}, } Publisher's Version Published Artifact Artifacts Available Artifacts Functional

Gadgets Splicing: Dynamic Binary Transformation for Precise Rewriting
Linan Tian, Yangyang Shi, Liwei Chen, Yanqi Yang, and Gang Shi
(Institute of Information Engineering at Chinese Academy of Sciences, China; University of Chinese Academy of Sciences, China)


@InProceedings{CGO22p181, author = {Linan Tian and Yangyang Shi and Liwei Chen and Yanqi Yang and Gang Shi}, title = {Gadgets Splicing: Dynamic Binary Transformation for Precise Rewriting}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {181-180}, doi = {10.1109/CGO53902.2022.9741259}, year = {2022}, } Publisher's Version Published Artifact Artifacts Available Artifacts Functional Results Reproduced

IR, Encryption, and Compression

Lambda the Ultimate SSA: Optimizing Functional Programs in SSA
Siddharth Bhat and Tobias Grosser
(IIIT Hyderabad, India; University of Edinburgh, UK)


@InProceedings{CGO22p196, author = {Siddharth Bhat and Tobias Grosser}, title = {Lambda the Ultimate SSA: Optimizing Functional Programs in SSA}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {196-195}, doi = {10.1109/CGO53902.2022.9741279}, year = {2022}, } Publisher's Version Published Artifact Artifacts Available Artifacts Functional Results Reproduced

NOELLE Offers Empowering LLVM Extensions
Angelo Matni, Enrico Armenio Deiana, Yian Su, Lukas Gross, Souradip Ghosh, Sotiris Apostolakis, Ziyang Xu, Zujun Tan, Ishita Chaturvedi, Brian Homerding, Tommy McMichen, David I. August, and Simone Campanoni
(Northwestern University, USA; Princeton University, USA)


@InProceedings{CGO22p211, author = {Angelo Matni and Enrico Armenio Deiana and Yian Su and Lukas Gross and Souradip Ghosh and Sotiris Apostolakis and Ziyang Xu and Zujun Tan and Ishita Chaturvedi and Brian Homerding and Tommy McMichen and David I. August and Simone Campanoni}, title = {NOELLE Offers Empowering LLVM Extensions}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {211-210}, doi = {10.1109/CGO53902.2022.9741276}, year = {2022}, } Publisher's Version Published Artifact Artifacts Available Artifacts Functional Results Reproduced

HECATE: Performance-Aware Scale Optimization for Homomorphic Encryption Compiler
Yongwoo Lee, Seonyeong Heo, Seonyoung Cheon, Shinnung Jeong, Changsu Kim, Eunkyung Kim, Dongyoon Lee, and Hanjun Kim
(Yonsei University, South Korea; ETH Zurich, Switzerland; Seoul National University, South Korea; Samsung SDS, South Korea; Stony Brook University, USA)


@InProceedings{CGO22p226, author = {Yongwoo Lee and Seonyeong Heo and Seonyoung Cheon and Shinnung Jeong and Changsu Kim and Eunkyung Kim and Dongyoon Lee and Hanjun Kim}, title = {HECATE: Performance-Aware Scale Optimization for Homomorphic Encryption Compiler}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {226-225}, doi = {10.1109/CGO53902.2022.9741265}, year = {2022}, } Publisher's Version

Unified Compilation for Lossless Compression and Sparse Computing
Daniel Donenfeld, Stephen Chou, and Saman Amarasinghe
(Massachusetts Institute of Technology, USA)


@InProceedings{CGO22p241, author = {Daniel Donenfeld and Stephen Chou and Saman Amarasinghe}, title = {Unified Compilation for Lossless Compression and Sparse Computing}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {241-240}, doi = {10.1109/CGO53902.2022.9741282}, year = {2022}, } Publisher's Version Artifacts Reusable Results Reproduced

Program Analysis and Optimization

Loop Rolling for Code Size Reduction
Rodrigo C. O. Rocha, Pavlos Petoumenos, Björn Franke, Pramod Bhatotia, and Michael O'Boyle
(University of Edinburgh, UK; University of Manchester, UK; TU Munich, Germany)


@InProceedings{CGO22p256, author = {Rodrigo C. O. Rocha and Pavlos Petoumenos and Björn Franke and Pramod Bhatotia and Michael O'Boyle}, title = {Loop Rolling for Code Size Reduction}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {256-255}, doi = {10.1109/CGO53902.2022.9741256}, year = {2022}, } Publisher's Version Published Artifact Artifacts Available Artifacts Reusable Results Reproduced

Solving PBQP-Based Register Allocation using Deep Reinforcement Learning
Minsu Kim, Jeong-Keun Park, and Soo-Mook Moon
(Seoul National University, South Korea)


@InProceedings{CGO22p271, author = {Minsu Kim and Jeong-Keun Park and Soo-Mook Moon}, title = {Solving PBQP-Based Register Allocation using Deep Reinforcement Learning}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {271-270}, doi = {10.1109/CGO53902.2022.9741272}, year = {2022}, } Publisher's Version

F3M: Fast Focused Function Merging
Sean Stirling, Rodrigo C. O. Rocha, Kim Hazelwood, Hugh Leather, Michael O'Boyle, and Pavlos Petoumenos
(Codeplay, UK; University of Edinburgh, UK; Facebook, USA; University of Manchester, UK)


@InProceedings{CGO22p286, author = {Sean Stirling and Rodrigo C. O. Rocha and Kim Hazelwood and Hugh Leather and Michael O'Boyle and Pavlos Petoumenos}, title = {F3M: Fast Focused Function Merging}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {286-285}, doi = {10.1109/CGO53902.2022.9741269}, year = {2022}, } Publisher's Version Published Artifact Artifacts Available Artifacts Reusable Results Reproduced

Sound, Precise, and Fast Abstract Interpretation with Tristate Numbers
Harishankar Vishwanathan, Matan Shachnai, Srinivas Narayana, and Santosh Nagarakatte
(Rutgers University, USA)


@InProceedings{CGO22p301, author = {Harishankar Vishwanathan and Matan Shachnai and Srinivas Narayana and Santosh Nagarakatte}, title = {Sound, Precise, and Fast Abstract Interpretation with Tristate Numbers}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {301-300}, doi = {10.1109/CGO53902.2022.9741267}, year = {2022}, } Publisher's Version Published Artifact Artifacts Available Artifacts Reusable Results Reproduced

Natural-Language Techniques

M3V: Multi-modal Multi-view Context Embedding for Repair Operator Prediction
Xuezheng Xu, Xudong Wang, and Jingling Xue
(UNSW, Australia)


@InProceedings{CGO22p316, author = {Xuezheng Xu and Xudong Wang and Jingling Xue}, title = {M3V: Multi-modal Multi-view Context Embedding for Repair Operator Prediction}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {316-315}, doi = {10.1109/CGO53902.2022.9741261}, year = {2022}, } Publisher's Version

Enabling Near Real-Time NLU-Driven Natural Language Programming through Dynamic Grammar Graph-Based Translation
Zifan Nan, Xipeng Shen, and Hui Guan
(North Carolina State University, USA; University of Massachusetts at Amherst, USA)


@InProceedings{CGO22p331, author = {Zifan Nan and Xipeng Shen and Hui Guan}, title = {Enabling Near Real-Time NLU-Driven Natural Language Programming through Dynamic Grammar Graph-Based Translation}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {331-330}, doi = {10.1109/CGO53902.2022.9741262}, year = {2022}, } Publisher's Version

AI Systems

SPNC: An Open-Source MLIR-Based Compiler for Fast Sum-Product Network Inference on CPUs and GPUs
Lukas Sommer, Cristian Axenie, and Andreas Koch
(TU Darmstadt, Germany; Huawei Research, Germany)


@InProceedings{CGO22p346, author = {Lukas Sommer and Cristian Axenie and Andreas Koch}, title = {SPNC: An Open-Source MLIR-Based Compiler for Fast Sum-Product Network Inference on CPUs and GPUs}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {346-345}, doi = {10.1109/CGO53902.2022.9741277}, year = {2022}, } Publisher's Version

Distill: Domain-Specific Compilation for Cognitive Models
Jan Vesely, Raghavendra Pradyumna Pothukuchi, Ketaki Joshi, Samyak Gupta, Jonathan D. Cohen, and Abhishek Bhattacharjee
(Yale University, USA; Princeton University, USA)


@InProceedings{CGO22p361, author = {Jan Vesely and Raghavendra Pradyumna Pothukuchi and Ketaki Joshi and Samyak Gupta and Jonathan D. Cohen and Abhishek Bhattacharjee}, title = {Distill: Domain-Specific Compilation for Cognitive Models}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {361-360}, doi = {10.1109/CGO53902.2022.9741278}, year = {2022}, } Publisher's Version Published Artifact Artifacts Available Artifacts Reusable Results Reproduced

Optimizing GPU Deep Learning Operators with Polyhedral Scheduling Constraint Injection
Cédric Bastoul, Zhen Zhang, Harenome Razanajato, Nelson Lossing, Adilla Susungi, Javier de Juan, Etienne Filhol, Baptiste Jarry, Gianpietro Consolaro, and Renwei Zhang
(Huawei Technologies, France; Huawei Technologies, China)


@InProceedings{CGO22p376, author = {Cédric Bastoul and Zhen Zhang and Harenome Razanajato and Nelson Lossing and Adilla Susungi and Javier de Juan and Etienne Filhol and Baptiste Jarry and Gianpietro Consolaro and Renwei Zhang}, title = {Optimizing GPU Deep Learning Operators with Polyhedral Scheduling Constraint Injection}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {376-375}, doi = {10.1109/CGO53902.2022.9741260}, year = {2022}, } Publisher's Version

Comprehensive Accelerator-Dataflow Co-design Optimization for Convolutional Neural Networks
Miheer Vaidya, Aravind Sukumaran-Rajam, Atanas Rountev, and P. Sadayappan
(University of Utah, USA; Washington State University, USA; Ohio State University, USA)


@InProceedings{CGO22p391, author = {Miheer Vaidya and Aravind Sukumaran-Rajam and Atanas Rountev and P. Sadayappan}, title = {Comprehensive Accelerator-Dataflow Co-design Optimization for Convolutional Neural Networks}, booktitle = {Proc.\ CGO}, publisher = {IEEE}, pages = {391-390}, doi = {10.1109/CGO53902.2022.9741281}, year = {2022}, } Publisher's Version

proc time: 0.05