Region Based Structure Layout Optimization by Selective Data Copying
As the gap between processor and memory continues to grow, memory performance becomes a key performance bottleneck for many applications. Compilers therefore increasingly seek to modify an applicationpsilas data layout to improve cache locality and cache reuse. Whole program structure layout [WPSL]...
Gespeichert in:
Hauptverfasser: | , , |
---|---|
Format: | Tagungsbericht |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext bestellen |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
container_end_page | 347 |
---|---|
container_issue | |
container_start_page | 338 |
container_title | |
container_volume | |
creator | Mannarswamy, S.S. Govindarajan, R. Surendran, R. |
description | As the gap between processor and memory continues to grow, memory performance becomes a key performance bottleneck for many applications. Compilers therefore increasingly seek to modify an applicationpsilas data layout to improve cache locality and cache reuse. Whole program structure layout [WPSL] transformations can significantly increase the spatial locality of data and reduce the runtime of programs that use link-based data structures, by increasing the cache line utilization. However, in production compilers WPSL transformations do not realize the entire performance potential possible due to a number of factors. Structure layout decisions made on the basis of whole program aggregated affinity/hotness of structure fields, can be sub optimal for local code regions. WPSL is also restricted in applicability in production compilers for type unsafe languages like C/C++ due to the extensive legality checks and field sensitive pointer analysis required over the entire application. In order to overcome the issues associated with WPSL, we propose region based structure layout (RBSL) optimization framework, using selective data copying. We describe our RBSL framework, implemented in the production compiler for C/C++ on HP-UX IA-64. We show that acting in complement to the existing and mature WPSL transformation framework in our compiler, RBSL improves application performance in pointer intensive SPEC benchmarks ranging from 3% to 28% over WPSL. |
doi_str_mv | 10.1109/PACT.2009.43 |
format | Conference Proceeding |
fullrecord | <record><control><sourceid>ieee_6IE</sourceid><recordid>TN_cdi_ieee_primary_5260527</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>5260527</ieee_id><sourcerecordid>5260527</sourcerecordid><originalsourceid>FETCH-LOGICAL-i175t-d8923ceb5e92e5174621a71988b05e0176cdfcb823622f28720d392d628537823</originalsourceid><addsrcrecordid>eNotT0tLw0AYXHyAofbmzcv-gdRvv82-jjX1BYGKreCtbJKvZaWPkGyE-OuN6FwGZoZhhrEbATMhwN29zvP1DAHcLJNnLEGdidS4LDtnU2csGO2UNEa4C5YIsG701McVm3bdJ4yQTgPYhC3eaBdOR37vO6r5KrZ9FfuWeOGHUx_5sonhEL59_M2UA1_RnqoYvogvfPQ8PzVDOO6u2eXW7zua_vOEvT8-rPPntFg-veTzIg3CqJjW1qGsqFTkkJQwmUbhx4XWlqAIhNFVva1Ki1IjbtEahFo6rDXa8cooT9jtX28gok3ThoNvh41CDQqN_AG-s0tL</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>conference_proceeding</recordtype></control><display><type>conference_proceeding</type><title>Region Based Structure Layout Optimization by Selective Data Copying</title><source>IEEE Electronic Library (IEL) Conference Proceedings</source><creator>Mannarswamy, S.S. ; Govindarajan, R. ; Surendran, R.</creator><creatorcontrib>Mannarswamy, S.S. ; Govindarajan, R. ; Surendran, R.</creatorcontrib><description>As the gap between processor and memory continues to grow, memory performance becomes a key performance bottleneck for many applications. Compilers therefore increasingly seek to modify an applicationpsilas data layout to improve cache locality and cache reuse. Whole program structure layout [WPSL] transformations can significantly increase the spatial locality of data and reduce the runtime of programs that use link-based data structures, by increasing the cache line utilization. However, in production compilers WPSL transformations do not realize the entire performance potential possible due to a number of factors. Structure layout decisions made on the basis of whole program aggregated affinity/hotness of structure fields, can be sub optimal for local code regions. WPSL is also restricted in applicability in production compilers for type unsafe languages like C/C++ due to the extensive legality checks and field sensitive pointer analysis required over the entire application. In order to overcome the issues associated with WPSL, we propose region based structure layout (RBSL) optimization framework, using selective data copying. We describe our RBSL framework, implemented in the production compiler for C/C++ on HP-UX IA-64. We show that acting in complement to the existing and mature WPSL transformation framework in our compiler, RBSL improves application performance in pointer intensive SPEC benchmarks ranging from 3% to 28% over WPSL.</description><identifier>ISSN: 1089-795X</identifier><identifier>ISBN: 9780769537719</identifier><identifier>ISBN: 0769537715</identifier><identifier>EISSN: 2641-7944</identifier><identifier>DOI: 10.1109/PACT.2009.43</identifier><language>eng</language><publisher>IEEE</publisher><subject>cache locality ; compiler ; Data structures ; Frequency ; Humans ; Inspection ; Optimizing compilers ; Parallel architectures ; Performance analysis ; Production ; Program processors ; Runtime ; structure layout optimization</subject><ispartof>2009 18th International Conference on Parallel Architectures and Compilation Techniques, 2009, p.338-347</ispartof><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/5260527$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>309,310,776,780,785,786,2052,27904,54898</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/5260527$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Mannarswamy, S.S.</creatorcontrib><creatorcontrib>Govindarajan, R.</creatorcontrib><creatorcontrib>Surendran, R.</creatorcontrib><title>Region Based Structure Layout Optimization by Selective Data Copying</title><title>2009 18th International Conference on Parallel Architectures and Compilation Techniques</title><addtitle>PACT</addtitle><description>As the gap between processor and memory continues to grow, memory performance becomes a key performance bottleneck for many applications. Compilers therefore increasingly seek to modify an applicationpsilas data layout to improve cache locality and cache reuse. Whole program structure layout [WPSL] transformations can significantly increase the spatial locality of data and reduce the runtime of programs that use link-based data structures, by increasing the cache line utilization. However, in production compilers WPSL transformations do not realize the entire performance potential possible due to a number of factors. Structure layout decisions made on the basis of whole program aggregated affinity/hotness of structure fields, can be sub optimal for local code regions. WPSL is also restricted in applicability in production compilers for type unsafe languages like C/C++ due to the extensive legality checks and field sensitive pointer analysis required over the entire application. In order to overcome the issues associated with WPSL, we propose region based structure layout (RBSL) optimization framework, using selective data copying. We describe our RBSL framework, implemented in the production compiler for C/C++ on HP-UX IA-64. We show that acting in complement to the existing and mature WPSL transformation framework in our compiler, RBSL improves application performance in pointer intensive SPEC benchmarks ranging from 3% to 28% over WPSL.</description><subject>cache locality</subject><subject>compiler</subject><subject>Data structures</subject><subject>Frequency</subject><subject>Humans</subject><subject>Inspection</subject><subject>Optimizing compilers</subject><subject>Parallel architectures</subject><subject>Performance analysis</subject><subject>Production</subject><subject>Program processors</subject><subject>Runtime</subject><subject>structure layout optimization</subject><issn>1089-795X</issn><issn>2641-7944</issn><isbn>9780769537719</isbn><isbn>0769537715</isbn><fulltext>true</fulltext><rsrctype>conference_proceeding</rsrctype><creationdate>2009</creationdate><recordtype>conference_proceeding</recordtype><sourceid>6IE</sourceid><sourceid>RIE</sourceid><recordid>eNotT0tLw0AYXHyAofbmzcv-gdRvv82-jjX1BYGKreCtbJKvZaWPkGyE-OuN6FwGZoZhhrEbATMhwN29zvP1DAHcLJNnLEGdidS4LDtnU2csGO2UNEa4C5YIsG701McVm3bdJ4yQTgPYhC3eaBdOR37vO6r5KrZ9FfuWeOGHUx_5sonhEL59_M2UA1_RnqoYvogvfPQ8PzVDOO6u2eXW7zua_vOEvT8-rPPntFg-veTzIg3CqJjW1qGsqFTkkJQwmUbhx4XWlqAIhNFVva1Ki1IjbtEahFo6rDXa8cooT9jtX28gok3ThoNvh41CDQqN_AG-s0tL</recordid><startdate>200909</startdate><enddate>200909</enddate><creator>Mannarswamy, S.S.</creator><creator>Govindarajan, R.</creator><creator>Surendran, R.</creator><general>IEEE</general><scope>6IE</scope><scope>6IL</scope><scope>CBEJK</scope><scope>RIE</scope><scope>RIL</scope></search><sort><creationdate>200909</creationdate><title>Region Based Structure Layout Optimization by Selective Data Copying</title><author>Mannarswamy, S.S. ; Govindarajan, R. ; Surendran, R.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-i175t-d8923ceb5e92e5174621a71988b05e0176cdfcb823622f28720d392d628537823</frbrgroupid><rsrctype>conference_proceedings</rsrctype><prefilter>conference_proceedings</prefilter><language>eng</language><creationdate>2009</creationdate><topic>cache locality</topic><topic>compiler</topic><topic>Data structures</topic><topic>Frequency</topic><topic>Humans</topic><topic>Inspection</topic><topic>Optimizing compilers</topic><topic>Parallel architectures</topic><topic>Performance analysis</topic><topic>Production</topic><topic>Program processors</topic><topic>Runtime</topic><topic>structure layout optimization</topic><toplevel>online_resources</toplevel><creatorcontrib>Mannarswamy, S.S.</creatorcontrib><creatorcontrib>Govindarajan, R.</creatorcontrib><creatorcontrib>Surendran, R.</creatorcontrib><collection>IEEE Electronic Library (IEL) Conference Proceedings</collection><collection>IEEE Proceedings Order Plan All Online (POP All Online) 1998-present by volume</collection><collection>IEEE Xplore All Conference Proceedings</collection><collection>IEEE Electronic Library (IEL)</collection><collection>IEEE Proceedings Order Plans (POP All) 1998-Present</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Mannarswamy, S.S.</au><au>Govindarajan, R.</au><au>Surendran, R.</au><format>book</format><genre>proceeding</genre><ristype>CONF</ristype><atitle>Region Based Structure Layout Optimization by Selective Data Copying</atitle><btitle>2009 18th International Conference on Parallel Architectures and Compilation Techniques</btitle><stitle>PACT</stitle><date>2009-09</date><risdate>2009</risdate><spage>338</spage><epage>347</epage><pages>338-347</pages><issn>1089-795X</issn><eissn>2641-7944</eissn><isbn>9780769537719</isbn><isbn>0769537715</isbn><abstract>As the gap between processor and memory continues to grow, memory performance becomes a key performance bottleneck for many applications. Compilers therefore increasingly seek to modify an applicationpsilas data layout to improve cache locality and cache reuse. Whole program structure layout [WPSL] transformations can significantly increase the spatial locality of data and reduce the runtime of programs that use link-based data structures, by increasing the cache line utilization. However, in production compilers WPSL transformations do not realize the entire performance potential possible due to a number of factors. Structure layout decisions made on the basis of whole program aggregated affinity/hotness of structure fields, can be sub optimal for local code regions. WPSL is also restricted in applicability in production compilers for type unsafe languages like C/C++ due to the extensive legality checks and field sensitive pointer analysis required over the entire application. In order to overcome the issues associated with WPSL, we propose region based structure layout (RBSL) optimization framework, using selective data copying. We describe our RBSL framework, implemented in the production compiler for C/C++ on HP-UX IA-64. We show that acting in complement to the existing and mature WPSL transformation framework in our compiler, RBSL improves application performance in pointer intensive SPEC benchmarks ranging from 3% to 28% over WPSL.</abstract><pub>IEEE</pub><doi>10.1109/PACT.2009.43</doi><tpages>10</tpages></addata></record> |
fulltext | fulltext_linktorsrc |
identifier | ISSN: 1089-795X |
ispartof | 2009 18th International Conference on Parallel Architectures and Compilation Techniques, 2009, p.338-347 |
issn | 1089-795X 2641-7944 |
language | eng |
recordid | cdi_ieee_primary_5260527 |
source | IEEE Electronic Library (IEL) Conference Proceedings |
subjects | cache locality compiler Data structures Frequency Humans Inspection Optimizing compilers Parallel architectures Performance analysis Production Program processors Runtime structure layout optimization |
title | Region Based Structure Layout Optimization by Selective Data Copying |
url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-28T05%3A51%3A39IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-ieee_6IE&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=proceeding&rft.atitle=Region%20Based%20Structure%20Layout%20Optimization%20by%20Selective%20Data%20Copying&rft.btitle=2009%2018th%20International%20Conference%20on%20Parallel%20Architectures%20and%20Compilation%20Techniques&rft.au=Mannarswamy,%20S.S.&rft.date=2009-09&rft.spage=338&rft.epage=347&rft.pages=338-347&rft.issn=1089-795X&rft.eissn=2641-7944&rft.isbn=9780769537719&rft.isbn_list=0769537715&rft_id=info:doi/10.1109/PACT.2009.43&rft_dat=%3Cieee_6IE%3E5260527%3C/ieee_6IE%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rft_ieee_id=5260527&rfr_iscdi=true |