i n t e r p r o c e d u r a l s p e c i a l i z a t i o n

I n t e r p r o c e d u r a l S p e c i a l i - PowerPoint PPT Presentation

May 31, 2023 •161 likes •935 views

I n t e r p r o c e d u r a l S p e c i a l i z a t i o n o f H i g h e r - O r d e r D y n a m i c L a n g u a g e s W i t h o u t S t a t i c A n a l y s i s ECOOP

  1. I n t e r p r o c e d u r a l S p e c i a l i z a t i o n o f H i g h e r - O r d e r D y n a m i c L a n g u a g e s W i t h o u t S t a t i c A n a l y s i s ECOOP June 22, 2017 Baptiste Saleil baptiste.saleil@umontreal.ca Marc Feeley feeley@iro.umontreal.ca

  2. I n t r o d u c t i o n  Research on JIT compilation  Dynamic languages  Dynamic techniques  LC : Research oriented Scheme compiler  Scheme and Functional Programming Workshop 2014 & 2015  https://github.com/bsaleil/lc 1

  3. I n t r o d u c t i o n  Dynamic languages  Work done at compilation  Work done at execution  Dynamic type checking  Ensures safety of the primitives :)  Impact on performance :( 2

  4. D y n a m i c t y p e c h e c k i n g  Example ( d e f i n e ( s u m - t o - 1 0 x ) ( i f ( > x 1 0 ) 0 ( + x ( s u m - t o - 1 0 ( + x 1 ) ) ) ) ) 3

  5. D y n a m i c t y p e c h e c k i n g  Example ( d e f i n e ( s u m - t o - 1 0 x ) ( i f ( > x 1 0 ) 0 ( + x ( s u m - t o - 1 0 ( + x 1 ) ) ) ) ) 4 type checks in this code 3

  6. D y n a m i c t y p e c h e c k i n g  Example ( d e f i n e ( s u m - t o - 1 0 x ) ( i f ( > x 1 0 ) 0 ( + x ( s u m - t o - 1 0 ( + x 1 ) ) ) ) ) 4 type checks in this code  How can we remove them ? 3

  7. I n t e r p r o c e d u r a l S p e c i a l i z a t i o n ( d e f i n e ( s u m - t o - 1 0 x ) ( i f ( > x 1 0 ) 0 ( + x ( s u m - t o - 1 0 ( + x 1 ) ) ) ) ) ( p r i n t ( s u m - t o - 1 0 6 ) ) ( p r i n t ( s u m - t o - 1 0 7 . 5 ) ) 4

  8. I n t e r p r o c e d u r a l S p e c i a l i z a t i o n ( d e f i n e ( s u m - t o - 1 0 x ) ( i f ( > x 1 0 ) s u m - t o - 1 0 : 0 ( x : i n t ) ( + x ( s u m - t o - 1 0 ( + x 1 ) ) ) ) ) + i n t ( p r i n t ( s u m - t o - 1 0 6 ) ) ( p r i n t ( s u m - t o - 1 0 7 . 5 ) ) 4

  9. I n t e r p r o c e d u r a l S p e c i a l i z a t i o n ( d e f i n e ( s u m - t o - 1 0 x ) ( i f ( > x 1 0 ) s u m - t o - 1 0 : 0 ( x : i n t ) ( + x ( s u m - t o - 1 0 ( + x 1 ) ) ) ) ) + i n t ( p r i n t ( s u m - t o - 1 0 6 ) ) s u m - t o - 1 0 : ( p r i n t ( s u m - t o - 1 0 7 . 5 ) ) ( x : f l o a t ) + f l o a t 4

  10. R u n n i n g E x a m p l e : H i g h e r o r d e r f u n c t i o n ( d e f i n e ( m a k e - s u m e r n ) Closure that captures n ( l e t r e c ( ( f ( l a m b d a ( x ) ( i f ( > x n ) 0 ( + x ( f ( + x 1 ) ) ) ) ) ) ) f ) ) ( d e f i n e s u m - t o - 1 0 ( m a k e - s u m e r 1 0 ) ) ( d e f i n e s u m - t o - p i ( m a k e - s u m e r 3 . 1 4 ) ) ; 6 + 7 + 8 + 9 + 1 0 ( p r i n t ( s u m - t o - 1 0 6 ) ) ; 7 . 5 + 8 . 5 + 9 . 5 ( p r i n t ( s u m - t o - 1 0 7 . 5 ) ) ; 1 . 1 0 + 2 . 1 0 + 3 . 1 0 5 ( p r i n t ( s u m - t o - p i 1 . 1 0 ) )

  11. R u n n i n g E x a m p l e : H i g h e r o r d e r f u n c t i o n ( d e f i n e ( m a k e - s u m e r n ) Closure that captures n ( l e t r e c ( ( f ( l a m b d a ( x ) ( i f ( > x n ) 0 ( + x ( f ( + x 1 ) ) ) ) ) ) ) f ) ) ( d e f i n e s u m - t o - 1 0 ( m a k e - s u m e r 1 0 ) ) ( d e f i n e s u m - t o - p i ( m a k e - s u m e r 3 . 1 4 ) ) ; 6 + 7 + 8 + 9 + 1 0 ( p r i n t ( s u m - t o - 1 0 6 ) ) ; 7 . 5 + 8 . 5 + 9 . 5 ( p r i n t ( s u m - t o - 1 0 7 . 5 ) ) ; 1 . 1 0 + 2 . 1 0 + 3 . 1 0 5 ( p r i n t ( s u m - t o - p i 1 . 1 0 ) )

  12. R u n n i n g E x a m p l e : H i g h e r o r d e r f u n c t i o n ( d e f i n e ( m a k e - s u m e r n ) → Static analysis (e.g. 0 -CFA) ( l e t r e c ( ( f ( l a m b d a ( x ) ( i f ( > x n ) 0 ( + x ( f ( + x 1 ) ) ) ) ) ) ) f ) ) ( d e f i n e s u m - t o - 1 0 ( m a k e - s u m e r 1 0 ) ) ( d e f i n e s u m - t o - p i ( m a k e - s u m e r 3 . 1 4 ) ) ; 6 + 7 + 8 + 9 + 1 0 ( p r i n t ( s u m - t o - 1 0 6 ) ) ; 7 . 5 + 8 . 5 + 9 . 5 ( p r i n t ( s u m - t o - 1 0 7 . 5 ) ) ; 1 . 1 0 + 2 . 1 0 + 3 . 1 0 6 ( p r i n t ( s u m - t o - p i 1 . 1 0 ) )

  13. R u n n i n g E x a m p l e : H i g h e r o r d e r f u n c t i o n ( d e f i n e ( m a k e - s u m e r n ) → Static analysis (e.g. 0 -CFA)  JIT incompatible ( l e t r e c ( ( f ( l a m b d a ( x ) ( i f ( > x n )  Lacks precision 0 ( + x ( f ( + x 1 ) ) ) ) ) ) ) f ) ) ( d e f i n e s u m - t o - 1 0 ( m a k e - s u m e r 1 0 ) ) ( d e f i n e s u m - t o - p i ( m a k e - s u m e r 3 . 1 4 ) ) ; 6 + 7 + 8 + 9 + 1 0 ( p r i n t ( s u m - t o - 1 0 6 ) ) ; 7 . 5 + 8 . 5 + 9 . 5 ( p r i n t ( s u m - t o - 1 0 7 . 5 ) ) ; 1 . 1 0 + 2 . 1 0 + 3 . 1 0 6 ( p r i n t ( s u m - t o - p i 1 . 1 0 ) )

  14. R u n n i n g E x a m p l e : H i g h e r o r d e r f u n c t i o n ( d e f i n e ( m a k e - s u m e r n ) → Static analysis (e.g. 0 -CFA)  JIT incompatible ( l e t r e c ( ( f ( l a m b d a ( x ) ( i f ( > x n )  Lacks precision 0 ( + x ( f ( + x 1 ) ) ) ) ) ) ) f ) ) ( d e f i n e s u m - t o - 1 0 ( m a k e - s u m e r 1 0 ) ) ( d e f i n e s u m - t o - p i ( m a k e - s u m e r 3 . 1 4 ) ) ; 6 + 7 + 8 + 9 + 1 0 ( p r i n t ( s u m - t o - 1 0 6 ) ) ; 7 . 5 + 8 . 5 + 9 . 5 ( p r i n t ( s u m - t o - 1 0 7 . 5 ) ) ; 1 . 1 0 + 2 . 1 0 + 3 . 1 0 6 ( p r i n t ( s u m - t o - p i 1 . 1 0 ) )

  15. I n t e r p r o c e d u r a l S p e c i a l i z a t i o n o f H i g h e r - O r d e r D y n a m i c L a n g u a g e s W i t h o u t S t a t i c A n a l y s i s 7

  16. I n t e r p r o c e d u r a l S p e c i a l i z a t i o n o f H i g h e r - O r d e r D y n a m i c L a n g u a g e s W i t h o u t S t a t i c A n a l y s i s 7

  17. I n t e r p r o c e d u r a l S p e c i a l i z a t i o n o f H i g h e r - O r d e r D y n a m i c L a n g u a g e s W i t h o u t S t a t i c A n a l y s i s 7

  18. I n t e r p r o c e d u r a l S p e c i a l i z a t i o n o f H i g h e r - O r d e r D y n a m i c L a n g u a g e s W i t h o u t S t a t i c A n a l y s i s 7

  19. I n t e r p r o c e d u r a l S p e c i a l i z a t i o n o f H i g h e r - O r d e r D y n a m i c L a n g u a g e s W i t h o u t S t a t i c A n a l y s i s → Basic Block Versioning (BBV)  Simple and Efgective Type Check Removal through Lazy Basic Block Versioning Maxime Chevalier-Boisvert and Marc Feeley, ECOOP 2015  Lazy intraprocedural code specialization (JIT)  No static analysis  Dynamic languages (JavaScript, Scheme, ...) 7

  20. N a i v e c o m p i l a t i o n ( d e f i n e ( m a k e - s u m e r n ) ( l e t r e c ( ( f ( l a m b d a ( x ) ( i f ( > x n ) 0 ( + x ( f ( + x 1 ) ) ) ) ) ) ) f ) ) ( d e f i n e s u m - t o - 1 0 ( m a k e - s u m e r 1 0 ) ) ( d e f i n e s u m - t o - p i ( m a k e - s u m e r 3 . 1 4 ) ) ; 6 + 7 + 8 + 9 + 1 0 ( p r i n t ( s u m - t o - 1 0 6 ) ) ; 7 . 5 + 8 . 5 + 9 . 5 ( p r i n t ( s u m - t o - 1 0 7 . 5 ) ) ; 1 . 1 0 + 2 . 1 0 + 3 . 1 0 ( p r i n t ( s u m - t o - p i 1 . 1 0 ) ) 8

Recommend

Relation between Agreement Measures on Human Labeling and Machine Learning Performance: Results

Relation between Agreement Measures on Human Labeling and Machine Learning Performance: Results

Relation between Agreement Measures on Human Labeling and Machine Learning Performance: Results from an Art History Domain Becky Passonneau, Columbia University Tom Lippincott, Columbia University Tae Yano, Carnegie Mellon University Judith Klavans,

374 views • 21 slides
Different Contributions to Cost-Effective Transcription and Translation of Video Lectures Thesis

Different Contributions to Cost-Effective Transcription and Translation of Video Lectures Thesis

Different Contributions to Cost-Effective Transcription and Translation of Video Lectures Thesis Presented by: Joan Albert Silvestre Cerd Supervisors: Dr. Alfons Juan Cscar Dr. Jorge Civera Saiz Machine Learning and Language Processing

407 views • 37 slides
CS 395T: Visual Recognition Exploiting Context for Object Detection 5 th October 2012 Aashish

CS 395T: Visual Recognition Exploiting Context for Object Detection 5 th October 2012 Aashish

CS 395T: Visual Recognition Exploiting Context for Object Detection 5 th October 2012 Aashish Sheshadri Components Analyzed 1. Scene Classification using GIST Descriptors. 2. Contextual Priming. Scene Classification Dataset : 15 Scene

532 views • 28 slides
WORKING GROUP MEETING August 27, 2020 1 TRAVEL DEMAND MODELING 2 Impacts on Regional Roadway

WORKING GROUP MEETING August 27, 2020 1 TRAVEL DEMAND MODELING 2 Impacts on Regional Roadway

WORKING GROUP MEETING August 27, 2020 1 TRAVEL DEMAND MODELING 2 Impacts on Regional Roadway Network (Daily) 2017 2045 % 2045 % Description Base Baseline Chang Baseline Chang Year w/o Tech e* w/Tech e* Vehicle-Miles 43,150,459

462 views • 13 slides
THE AGENDA THE PAST Challenges faced from past campaigns THE IDEA Issue Insight

THE AGENDA THE PAST Challenges faced from past campaigns THE IDEA Issue Insight

THE AGENDA THE PAST Challenges faced from past campaigns THE IDEA Issue Insight Objectives THE CAMPAIGN Key Message & Objectives Target Audience Moodboard Visuals Deliverables Timeline

524 views • 29 slides
Introduction to the summer school on Clouds and Big Data 2015 Seif Haridi KTH/SICS

Introduction to the summer school on Clouds and Big Data 2015 Seif Haridi KTH/SICS

Introduction to the summer school on Clouds and Big Data 2015 Seif Haridi KTH/SICS Big Data a collection of massive amount of information whether structured or unstructured Big Compute large scale processing required to extract

459 views • 10 slides
Change point detection in Python with ruptures Digital French-German Summer School with Industry

Change point detection in Python with ruptures Digital French-German Summer School with Industry

Change point detection in Python with ruptures Digital French-German Summer School with Industry 2020 Charles Truong 1 1Centre Borelli Universit Paris-Saclay ENS Paris-Saclay, CNRS Wednesday 24 th June Introduction Change point detection

1.21k views • 32 slides
Google Summer of Code 2005 Google Summer of Code 2005 Bernard Li Bernard Li SoC 2005 SoC 2005

Google Summer of Code 2005 Google Summer of Code 2005 Bernard Li Bernard Li SoC 2005 SoC 2005

Google Summer of Code 2005 Google Summer of Code 2005 Bernard Li Bernard Li SoC 2005 SoC 2005 Google set aside $1 million USD to pay Google set aside $1 million USD to pay students to work on Open Source projects students to work on

67 views • 5 slides
Midland Section ACS Board Meeting March 2, 2020 Agenda Time Topic Presenter 7:00 Call to

Midland Section ACS Board Meeting March 2, 2020 Agenda Time Topic Presenter 7:00 Call to

American Chemical Society Midland Section ACS Board Meeting March 2, 2020 Agenda Time Topic Presenter 7:00 Call to Order, Approve Minutes (February) Mark Jones 7:05 Chair Updates Mark Jones Meet Anne Catherine Bedard, new Diversity and

337 views • 17 slides
Department of Computer Science Lehman College, City University of New York Summer 2020 CMP 334:

Department of Computer Science Lehman College, City University of New York Summer 2020 CMP 334:

Department of Computer Science Lehman College, City University of New York Summer 2020 CMP 334: Computer Organization Lecture Schedule: Mondays, Tuesdays, Wednesdays, and Thursdays, 10:00 am - 1:25 pm Lecture Location:

55 views • 4 slides
Multifa mily MUL T IF AMIL Y Ma rke t Re vie w MARKE T RE VIE W Dwig ht Dunto n Re nte

Multifa mily MUL T IF AMIL Y Ma rke t Re vie w MARKE T RE VIE W Dwig ht Dunto n Re nte

Multifa mily MUL T IF AMIL Y Ma rke t Re vie w MARKE T RE VIE W Dwig ht Dunto n Re nte rs b y Ne c e ssity a nd b y Cho ic e T oug h E c onomic Conditions Mille nnia ls une mplo yme nt ra te : 10.9% Na tio na l Une mplo yme

140 views • 12 slides
Radial-velocity variability of the Sun as a star with HARPS and HARPS-N Raphalle D. Haywood

Radial-velocity variability of the Sun as a star with HARPS and HARPS-N Raphalle D. Haywood

Radial-velocity variability of the Sun as a star with HARPS and HARPS-N Raphalle D. Haywood (Harvard Observatory) D. Charbonneau, A. Collier Cameron, X. Dumusque, A. Glenday, D. W. Latham, C. Lovis, J. Maldonado, G. Micela, E. Molinari, A.

189 views • 14 slides
Phased Transactional Memory Dan Nussbaum Scalable Synchronization Research Group Joint work

Phased Transactional Memory Dan Nussbaum Scalable Synchronization Research Group Joint work

Phased Transactional Memory Dan Nussbaum Scalable Synchronization Research Group Joint work with Yossi Lev and Mark Moir Sun Microsystems Labs August 16, 2007 1 1 Transactional Memory (TM) Replace locks with atomic sections.

255 views • 21 slides
Background Simulations for the IXO Wide Field Imager Steffen Hauf, Markus Kuster, Dieter H.H.

Background Simulations for the IXO Wide Field Imager Steffen Hauf, Markus Kuster, Dieter H.H.

Background Simulations for the IXO Wide Field Imager Steffen Hauf, Markus Kuster, Dieter H.H. Hoffmann, Maria Grazia Pia, Eckhard Kendziorra, Philipp Lang, Alexander Stephanescu, Lothar Strder, Chris Tenzer and Georg Weidenspointner Credit:

582 views • 21 slides
ECE590-03 Enterprise Storage Architecture Fall 2016 Storage devices Tyler Bletsch Duke

ECE590-03 Enterprise Storage Architecture Fall 2016 Storage devices Tyler Bletsch Duke

ECE590-03 Enterprise Storage Architecture Fall 2016 Storage devices Tyler Bletsch Duke University Slides include material from Vince Freeh (NCSU) Basic storage device history From

708 views • 56 slides
The bluetides simulation Tiziana DiMatteo (CMU ) Yu Feng (Berkeley), Rupert Croft (CMU ), Aklant

The bluetides simulation Tiziana DiMatteo (CMU ) Yu Feng (Berkeley), Rupert Croft (CMU ), Aklant

The next quasars and galaxies frontier: The bluetides simulation Tiziana DiMatteo (CMU ) Yu Feng (Berkeley), Rupert Croft (CMU ), Aklant Bhowmick, Kuan-Wei Hunag, (CMU) Simeon Bird (JHU), Steve Wilkins (Sussex),Ananth Tenneti (CMU ) Nick

948 views • 52 slides
GALILEO AND THE LAWS OF MOTION Daniel Garber (Princeton

GALILEO AND THE LAWS OF MOTION Daniel Garber (Princeton

GALILEO AND THE LAWS OF MOTION Daniel Garber (Princeton University) Boston Colloquium for the Philosophy of Science GALILEO 450TH ANNIVERSARY October

357 views • 19 slides
Information Visualization Software Visualization Different Techniques for structured data

Information Visualization Software Visualization Different Techniques for structured data

Information Visualization Software Visualization Different Techniques for structured data Textual data and tables Information Visualization Hierarchical data Techniques Networks Graphs (later!) A closer look at tree

307 views • 4 slides
Deconstructing Data Science David Bamman, UC Berkeley Info 290 Lecture 2: Survey of

Deconstructing Data Science David Bamman, UC Berkeley Info 290 Lecture 2: Survey of

Deconstructing Data Science David Bamman, UC Berkeley Info 290 Lecture 2: Survey of Methods Jan 25, 2016 Announcements Python and Jupyter workshop (sign up through bCourses) Linear regression Deep learning Decision trees

709 views • 38 slides
Automatically Extracting, Analyzing, and Visualizing Information on Music Artists from the Web

Automatically Extracting, Analyzing, and Visualizing Information on Music Artists from the Web

Automatically Extracting, Analyzing, and Visualizing Information on Music Artists from the Web Workshop on Learning the Semantics of Audio Signals (LSAS) 21 st June 2008, Paris, France Markus Schedl, Peter Knees Department of Computational

691 views • 26 slides
Updates in Neuromodulation 2020 Kenneth B. Chapman, MD Director, Pain Medicine Staten Island

Updates in Neuromodulation 2020 Kenneth B. Chapman, MD Director, Pain Medicine Staten Island

Updates in Neuromodulation 2020 Kenneth B. Chapman, MD Director, Pain Medicine Staten Island University Hospital Assistant Clinical Professor, NYU Langone Medical Center Adjunct Assistant Professor, Zucker School of Medicine at

590 views • 44 slides
Parents Day Secondary 2 3 March 2018 Objectives of Todays Session To provide an

Parents Day Secondary 2 3 March 2018 Objectives of Todays Session To provide an

Parents Day Secondary 2 3 March 2018 Objectives of Todays Session To provide an update on the school's policies, programmes and post-secondary options. To highlight key learning programmes that enhance students learning.

925 views • 76 slides
The British Sundial Society Chris Williams David Brown Jen Brown Bill Visick

The British Sundial Society Chris Williams David Brown Jen Brown Bill Visick

The British Sundial Society Chris Williams David Brown Jen Brown Bill Visick www.sundialsoc.org.uk Greek and Roman Sundials Chris Williams Sun path Stick shadow Typical Dials Forum Stadium Theatre Making a Hemicyclium dial David Brown

496 views • 22 slides
Contents 1. The structure equations 1 1.1. CayleyDickson construction 1 1.2. Spin(7) and

Contents 1. The structure equations 1 1.1. CayleyDickson construction 1 1.2. Spin(7) and

Contents 1. The structure equations 1 1.1. CayleyDickson construction 1 1.2. Spin(7) and the octonions 1 1.3. The standard basis 2 1.4. Lie algebra of Spin(7) 3 1.5. MaurerCartan form 4 1.6. The first structure equations 5

259 views • 8 slides

More recommend