We made the following two presentations at the JSSST 2019 Annual Conference.
Li received the student research award (学生奨励賞) and the best presentation award (優秀発表賞)!


Youyou Cong will present the following paper at ICFP 2019.
Title: Compiling with Continuations, or without? Whatever.
Authors: Youyou Cong, Leo Osvald, Gregory Essertel, and Tiark Rompf
Abstract:
What makes a good compiler IR? In the context of functional languages, there has been an extensive debate on the advantages and disadvantages of continuation-passing-style (CPS). The consensus seems to be that some form of explicit continuations is necessary to model jumps in a functional style, but that they should have a 2nd-class status, separate from regular functions, to ensure efficient code generation. Building on this observation, a recent study from PLDI 2017 proposed a direct-style IR with explicit join points, which essentially represent local continuations, i.e., functions that do not return or escape. While this IR can work well in practice, as evidenced by the implementation of join points in the Glasgow Haskell Compiler (GHC), there still seems to be room for improvement, especially with regard to the way continuations are handled in the course of optimization.
In this paper, we contribute to the CPS debate by developing a novel IR with the following features. First, we integrate a control operator that resembles Felleisen’s C, eliminating certain redundant rewrites observed in the previous study. Second, we treat the non-returning and non-escaping aspects of continuations separately, allowing efficient compilation of well-behaved functions defined by the user. Third, we define a selective CPS translation of our IR, which erases control operators while preserving the meaning and typing of programs. These features enable optimizations in both direct style and full CPS, as well as in any intermediate style with selectively exposed continuations. Thus, we change the spectrum of available options from “CPS yes or no” to “as much or as little CPS as you want, when you want it”.

We are sending off five members in these months.
- Jeanie Adkisson, who is to finish her Master’s degree, will move to Canada and work in a public sector.
- Tomoyuki Aotani, who has been served as an assistant professor from the foundation of the group, will pursuit professional experiences in industry.
- Li Dongfang, who is to finish his Master’s degree, starts working at Institute of Advanced Artificial Intelligence in Nanjing.
- Matthias Springer, who is to finish his Doctoral degree, will work for Google.
- Chengkai Yang, who is to finish the Tokyo Tech Summer Program, will go back Georgia Tech, his home university.
We wish for their bright future, and hope we can see each other soon!
Chengkai Yang, an exchange student from Georgia Tech, gave the final presentation on his working during the Tokyo Tech Summer Program. He worked on KaniCUDA, a program synthesis-based optimization tool for CUDA programs, especially on accelerating the synthesizer speed.
Two members presented their Master’s theses.
- Jeanine Adkisson, Magritte: A Language for Pipe-Based Programming (PDF)
- Li Dongfang, Deep Learning based Code Completion with ASTToken2Vec (PDF)
(official announcement)


The Computer Software journal of JSSST published Masuhara’s foreword titled “お前まだアレやってるの? (Are you still working on that?)” in the issue of August 2019.
Our paper “DynaSOAr: A Parallel Memory Allocator for Object-oriented Programming on GPUs with Efficient Memory Access”, authored by Matthias Springer and Hidehiko Masuhara, was accepted at the 33rd European Conference on Object-Oriented Programming (ECOOP 2019).

Matthias Springer had a public defense of his doctoral dissertation on “Memory-Efficient Object-Oriented Programming on GPUs.” His doctoral degree is almost there!
(more…)
Our paper “Massively Parallel GPU Memory Compaction” (PDF), authored by Matthias Springer and Hidehiko Masuhara, was accepted at the ACM SIGPLAN International Symposium on Memory Management (ISMM) 2019, co-located with PLDI 2019.
The poster “CompactGpu: Massively Parallel Memory Defragmentation on GPUs“, authored by Matthias Springer, was accepted at the ACM Student Research Competition (SRC) of the ACM SIGPLAN Conference on Programming Language Design and Implementation (PLDI) 2019.
Matthias Springer received the Best Student Presentation Award at ISMM and the first place in the graduate category of the SRC.
The following research will be presented at the IPSJ-PRO (124th, Nagoya).
“A Reformalization of Quantified Types Based on Coeffect Calculus”
Yudai Tanabe, Luthfan Anshar Lubis, Tomoyuki Aotani, Hidehiko Masuhara