TU Darmstadt / ULB / TUbiblio

Runtime-adaptable selective performance instrumentation

Kreutzer, Sebastian ; Iwainsky, Christian ; Garcia-Gasulla, Marta ; Lopez, Victor ; Bischof, Christian (2023)
Runtime-adaptable selective performance instrumentation.
37th IEEE International Parallel and Distributed Processing Symposium. St.Petersburg, USA (15.-19.05.2023)
doi: 10.1109/IPDPSW59300.2023.00073
Konferenzveröffentlichung, Bibliographie

Kurzbeschreibung (Abstract)

Automated code instrumentation, i.e. the insertion of measurement hooks into a target application by the compiler, is an established technique for collecting reliable, fine-grained performance data. The set of functions to instrument has to be selected with care, as instrumenting every available function typically yields too large a runtime overhead, thus skewing the measurement. No “one-suits-all” selection mechanism exists, since the instrumentation decision is dependent on the measurement objective, the limit for tolerable runtime overhead and peculiarities of the target application. The Compiler-assisted Performance Instrumentation (CaPI) tool assists in creating such instrumentation configurations, by enabling the user to combine different selection mechanisms as part of a configurable selection pipeline, operating on a statically constructed whole-program call-graph. Previously, CaPI relied on a static instrumentation worktlow which made the process of refining the initial selection quite cumbersome for large-scale codes, as the application had to be recompiled after each adjustment. In this work, we present new runtime-adaptable instrumentation capabilities for CaPI which do not require recompilation when instrumentation changes are made. To this end, the XRay instrumentation feature of the LLVM compiler was extended to support the instrumentation of shared dynamic objects. An XRay-compatible runtime system was added to CaPI that instruments selected functions at program start, thereby significantly reducing the required time for selection refinements. Furthermore, an interface to the TALP tool for recording parallel efficiency metrics was implemented, alongside a specialized selection module for creating suitable coarse-grained region instrumentations.

Typ des Eintrags: Konferenzveröffentlichung
Erschienen: 2023
Autor(en): Kreutzer, Sebastian ; Iwainsky, Christian ; Garcia-Gasulla, Marta ; Lopez, Victor ; Bischof, Christian
Art des Eintrags: Bibliographie
Titel: Runtime-adaptable selective performance instrumentation
Sprache: Englisch
Publikationsjahr: 4 August 2023
Verlag: IEEE
Buchtitel: 2023 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW)
Veranstaltungstitel: 37th IEEE International Parallel and Distributed Processing Symposium
Veranstaltungsort: St.Petersburg, USA
Veranstaltungsdatum: 15.-19.05.2023
DOI: 10.1109/IPDPSW59300.2023.00073
Kurzbeschreibung (Abstract):

Automated code instrumentation, i.e. the insertion of measurement hooks into a target application by the compiler, is an established technique for collecting reliable, fine-grained performance data. The set of functions to instrument has to be selected with care, as instrumenting every available function typically yields too large a runtime overhead, thus skewing the measurement. No “one-suits-all” selection mechanism exists, since the instrumentation decision is dependent on the measurement objective, the limit for tolerable runtime overhead and peculiarities of the target application. The Compiler-assisted Performance Instrumentation (CaPI) tool assists in creating such instrumentation configurations, by enabling the user to combine different selection mechanisms as part of a configurable selection pipeline, operating on a statically constructed whole-program call-graph. Previously, CaPI relied on a static instrumentation worktlow which made the process of refining the initial selection quite cumbersome for large-scale codes, as the application had to be recompiled after each adjustment. In this work, we present new runtime-adaptable instrumentation capabilities for CaPI which do not require recompilation when instrumentation changes are made. To this end, the XRay instrumentation feature of the LLVM compiler was extended to support the instrumentation of shared dynamic objects. An XRay-compatible runtime system was added to CaPI that instruments selected functions at program start, thereby significantly reducing the required time for selection refinements. Furthermore, an interface to the TALP tool for recording parallel efficiency metrics was implemented, alongside a specialized selection module for creating suitable coarse-grained region instrumentations.

Fachbereich(e)/-gebiet(e): 20 Fachbereich Informatik
20 Fachbereich Informatik > Scientific Computing
Zentrale Einrichtungen
Zentrale Einrichtungen > Hochschulrechenzentrum (HRZ)
Zentrale Einrichtungen > Hochschulrechenzentrum (HRZ) > Hochleistungsrechner
Hinterlegungsdatum: 18 Jan 2024 13:39
Letzte Änderung: 07 Mär 2024 13:15
PPN: 516021346
Export:
Suche nach Titel in: TUfind oder in Google
Frage zum Eintrag Frage zum Eintrag

Optionen (nur für Redakteure)
Redaktionelle Details anzeigen Redaktionelle Details anzeigen