BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260522T150119Z
LOCATION:C141/143/149
DTSTART;TZID=America/Chicago:20181115T113000
DTEND;TZID=America/Chicago:20181115T120000
UID:submissions.supercomputing.org_SC18_sess207_pap382@linklings.com
SUMMARY:Detecting MPI Usage Anomalies via Partial Program Symbolic Executi
 on
DESCRIPTION:Fangke Ye, Jisheng Zhao, and Vivek Sarkar (Georgia Institute o
 f Technology)\n\nMPI is a message passing based programming model for dist
 ributed-memory parallelism that has been had been widely used for programm
 ing supercomputers for over 25 years. However, debugging and verification 
 of MPI programs is widely recognized to be a deep technical challenge. Thi
 s challenge is further exacerbated by a recent increase in the use of nonb
 locking MPI operations that bring new classes of bugs related to data race
 s.\n\nIn this paper, we introduce a new MPI program debugging approach bas
 ed on partial symbolic execution so as to avoid the false alarms inherent 
 in the static analysis based methodology. Compared with the dynamic approa
 ch, our approach can be applied to incomplete programs and explore multipl
 e execution paths, thereby bringing more flexibility and precision. By com
 paring with well known static/dynamic tools on real-world MPI applications
 , our approach shows same precision as the dynamic tool and avoids false p
 ositive produced by the static tool.\n\nTag: Linear Algebra, Memory, MPI, 
 OpenMP, Programming Systems, Tools\n\nRegistration Category: Tech Program 
 Reg Pass\n\nSession Chair: Naoya Maruyama (NVIDIA Corporation)\n\n
END:VEVENT
END:VCALENDAR