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:20181221T160731Z
LOCATION:C141/143/149
DTSTART;TZID=America/Chicago:20181115T103000
DTEND;TZID=America/Chicago:20181115T110000
UID:submissions.supercomputing.org_SC18_sess207_pap179@linklings.com
SUMMARY:Dynamic Data Race Detection for OpenMP Programs
DESCRIPTION:Paper\nLinear Algebra, Memory, MPI, OpenMP, Programming System
 s, Tools, Tech Program Reg Pass\n\nDynamic Data Race Detection for OpenMP 
 Programs\n\nGu, Mellor-Crummey\n\nTwo concurrent accesses to a shared vari
 able that are unordered by synchronization are said to be a data race if a
 t least one access is a write. Data races cause shared memory parallel pro
 grams to behave unpredictably. This paper describes ROMP -- a tool for det
 ecting data races in executions of scalable parallel applications that emp
 loy OpenMP for node-level parallelism. The complexity of OpenMP, which inc
 ludes primitives for managing data environments, SPMD and SIMD parallelism
 , work sharing, tasking, mutual exclusion, and ordering, presents a formid
 able challenge for data race detection. ROMP is a hybrid data race detecto
 r that tracks accesses, access orderings, and mutual exclusion. Unlike oth
 er OpenMP race detectors, ROMP detects races with respect to logical paral
 lelism rather than implementation threads. Experiments show that ROMP yiel
 ds precise race reports for a broader set of OpenMP constructs than prior 
 state-of-the-art race detectors.
URL:https://sc18.supercomputing.org/presentation/?id=pap179&sess=sess207
END:VEVENT
END:VCALENDAR