Write a Blog >>
Mon 19 Jun 2017 14:00 - 14:25 at Aula Master - Concurrency Analysis Chair(s): Jeff Huang

GPU programming models enable and encourage massively parallel programming with over a million threads, requiring extreme parallelism to achieve good performance. Massive parallelism brings significant correctness challenges by increasing the possibility for bugs as the number of thread interleavings balloons. Conventional dynamic safety analyses struggle to run at this scale.

We present Barracuda, a data race detector for GPU programs written in Nvidia’s CUDA language. Barracuda handles a wider range of parallelism constructs than previous work, including branch operations, low-level atomics and memory fences, which allows Barracuda to detect new classes of races. Barracuda operates at the binary level for increased compatibility with existing code, leveraging a new binary instrumentation framework that is extensible to other dynamic analyses. Barracuda incorporates a number of novel optimizations that are crucial for scaling data race detection to over a million threads.

Mon 19 Jun

Displayed time zone: Amsterdam, Berlin, Bern, Rome, Stockholm, Vienna change

14:00 - 15:40
Concurrency AnalysisPLDI Research Papers at Aula Master
Chair(s): Jeff Huang Texas A&M University
14:00
25m
Talk
BARRACUDA: Binary-level Analysis of Runtime RAces in CUDA programs
PLDI Research Papers
Ariel Eizenberg University of Pennsylvania, Yuanfeng Peng University of Pennsylvania, Toma Pigli University of Pennsylvania, William Mansky Princeton University, Joseph Devietti University of Pennsylvania
14:25
25m
Talk
BigFoot: Static Check Placement for Dynamic Race Detection
PLDI Research Papers
Dustin Rhodes , Cormac Flanagan University of California, Santa Cruz, Stephen N. Freund Williams College
14:50
25m
Talk
Dynamic Race Prediction in Linear Time
PLDI Research Papers
Dileep Kini University of Illinois at Urbana-Champaign, Umang Mathur University of Illinois at Urbana-Champaign, Mahesh Viswanathan University of Illinois at Urbana-Champaign
Media Attached
15:15
25m
Talk
Systematic Black-Box Analysis of Collaborative Web Applications
PLDI Research Papers
Marina Billes TU Darmstadt, Anders Møller Aarhus University, Michael Pradel TU Darmstadt
Media Attached