Free Essay

Cmu 213 Ppt

In: Computers and Technology

Submitted By yg1365
Words 2285
Pages 10
Carnegie Mellon

Course Overview
15-213 /18-213: Introduction to Computer Systems 1st Lecture, Jan. 14, 2014 Instructors: Seth Copen Goldstein, Anthony Rowe, Greg Kesden

The course that gives CMU its “Zip”!

Carnegie Mellon

   

Course theme Five realities How the course fits into the CS/ECE curriculum Logistics


Carnegie Mellon

Course Theme: Abstraction Is Good But Don’t Forget Reality

Most CS and CE courses emphasize abstraction
 Abstract data types  Asymptotic analysis

These abstractions have limits
 Especially in the presence of bugs  Need to understand details of underlying implementations

Useful outcomes from taking 213
 Become more effective programmers
Able to find and eliminate bugs efficiently  Able to understand and tune for program performance  Prepare for later “systems” classes in CS & ECE  Compilers, Operating Systems, Networks, Computer Architecture, Embedded Systems, Storage Systems, etc.


Carnegie Mellon

Great Reality #1: Ints are not Integers, Floats are not Reals

Example 1: Is x2 ≥ 0?
 Float’s: Yes!

 Int’s:

40000 * 40000  1600000000 50000 * 50000  ??

Source: 4

Carnegie Mellon

Great Reality #1: Ints are not Integers, Floats are not Reals

Example 1: Is x2 ≥ 0?
 Float’s: Yes!

 Int’s:

40000 * 40000  1600000000 50000 * 50000  ??

Example 2: Is (x + y) + z = x + (y + z)?
 Unsigned & Signed Int’s: Yes!  Float’s:

(1e20 + -1e20) + 3.14  3.14 1e20 + (-1e20 + 3.14)  ??
Source: 5

Carnegie Mellon

Computer Arithmetic

Does not generate random values
 Arithmetic operations have important mathematical properties

Cannot assume all “usual” mathematical properties
 Due to finiteness of representations  Integer operations satisfy “ring” properties
Commutativity, associativity, distributivity  Floating point operations satisfy “ordering” properties  Monotonicity, values of signs

 Need to understand which abstractions apply in which contexts  Important issues for compiler writers and serious application programmers


Carnegie Mellon

Great Reality #2: You’ve Got to Know Assembly

Chances are, you’ll never write programs in assembly
 Compilers are much better & more patient than you are

But: Understanding assembly is key to machine-level execution model
 Behavior of programs in presence of bugs
High-level language models break down  Tuning program performance  Understand optimizations done / not done by the compiler  Understanding sources of program inefficiency  Implementing system software  Compiler has machine code as target  Operating systems must manage process state  Creating / fighting malware  x86 assembly is the language of choice!


Carnegie Mellon

Great Reality #3: Memory Matters
Random Access Memory Is an Unphysical Abstraction

Memory is not unbounded
 It must be allocated and managed  Many applications are memory dominated

Memory referencing bugs especially pernicious
 Effects are distant in both time and space

Memory performance is not uniform
 Cache and virtual memory effects can greatly affect program performance  Adapting program to characteristics of memory system can lead to major speed improvements


Carnegie Mellon

Memory Referencing Bug Example double fun(int i) { volatile double d[1] = {3.14}; volatile long int a[2]; a[i] = 1073741824; /* Possibly out of bounds */ return d[0]; }

fun(0) fun(1) fun(2) fun(3) fun(4)

    

3.14 3.14 5.30499e-315 3.14 segmentation fault

Result is architecture, compiler, and OS specific


Carnegie Mellon

Memory Referencing Bug Example double fun(int i) { volatile double d[1] = {3.14}; volatile long int a[2]; a[i] = 1073741824; /* Possibly out of bounds */ return d[0]; }

fun(0) fun(1) fun(2) fun(3) fun(4)

    

3.14 3.14 5.30499e-315 3.14 segmentation fault Saved State d[0] a[1] a[0] 3 2 1 0


Location accessed by fun(i)

Carnegie Mellon

Memory Referencing Errors

C and C++ do not provide any memory protection
 Out of bounds array references  Invalid pointer values  Abuses of malloc/free

Can lead to nasty bugs
 Whether or not bug has any effect depends on system and compiler  Action at a distance
 

Corrupted object logically unrelated to one being accessed Effect of bug may be first observed long after it is generated

How can I deal with this?
 Program in Java, Ruby or ML  Understand what possible interactions may occur  Use or develop tools to detect referencing errors (e.g. Valgrind)

Carnegie Mellon

Great Reality #4: There’s more to performance than asymptotic complexity
 

Constant factors matter too! And even exact op count does not predict performance
 Easily see 10:1 performance range depending on how code written  Must optimize at multiple levels: algorithm, data representations, procedures, and loops

Must understand system to optimize performance
 How programs compiled and executed  How to measure program performance and identify bottlenecks  How to improve performance without destroying code modularity and generality 12

Carnegie Mellon

Memory System Performance Example void copyij(int src[2048][2048], int dst[2048][2048]) { int i,j; for (i = 0; i < 2048; i++) for (j = 0; j < 2048; j++) dst[i][j] = src[i][j]; } void copyji(int src[2048][2048], int dst[2048][2048]) { int i,j; for (j = 0; j < 2048; j++) for (i = 0; i < 2048; i++) dst[i][j] = src[i][j]; }

Same instructions, but different order → 21x slower! (Pentium 4)
 

Hierarchical memory organization Performance depends on access patterns
 Including how step through multi-dimensional array

Carnegie Mellon

Great Reality #5: Computers do more than execute programs

They need to get data in and out
 I/O system critical to program reliability and performance

They communicate with each other over networks
 Many system-level issues arise in presence of network
   

Concurrent operations by autonomous processes Coping with unreliable media Cross platform compatibility Complex performance issues


Carnegie Mellon

Role within CS/ECE Curriculum
CS 412 OS Practicum

ECE 545/549 Capstone CS 411 Compilers
ECE 340 Digital Computation

CS 415 Databases

CS 441 Networks

CS 410 Operating Systems

ECE 447 Architecture

ECE 349 Embedded Systems

ECE 348 Embedded System Eng.

Data Reps. Memory Model

Network Protocols

Processes Machine Mem. Mgmt Code Arithmetic

Execution Model Memory System


Foundation of Computer Systems Underlying principles for hardware, software, and networking

CS 122 Imperative Programming

Carnegie Mellon

Course Perspective

Most Systems Courses are Builder-Centric
 Computer Architecture
Design pipelined processor in Verilog  Operating Systems  Implement large portions of operating system  Compilers  Write compiler for simple language  Networking  Implement and simulate network protocols


Carnegie Mellon

Course Perspective (Cont.)

Our Course is Programmer-Centric
 Purpose is to show that by knowing more about the underlying system, one can be more effective as a programmer  Enable you to  Write programs that are more reliable and efficient  Incorporate features that require hooks into OS – E.g., concurrency, signal handlers  Cover material in this course that you won’t see elsewhere  Not just a course for dedicated hackers  We bring out the hidden hacker in everyone!


Power Programmers

Manage the flow of data
 Inside the computer (memory hierarchy)  Between computers and devices (I/O)

Manage concurrency
 Inside the computer (multiple cores, threads, vectors, events, …)  Between computers (web servers, distributed apps, …)


Teaching staff

Carnegie Mellon

Seth Copen Goldstein

Anthony Rowe

Greg Kesden

Carnegie Mellon


Randal E. Bryant and David R. O’Hallaron,
 “Computer Systems: A Programmer’s Perspective, Second Edition”
(CS:APP2e), Prentice Hall, 2011   This book really matters for the course!  How to solve labs  Practice problems typical of exam problems

Brian Kernighan and Dennis Ritchie,
 “The C Programming Language, Second Edition”, Prentice Hall, 1988


Carnegie Mellon

Course Components

 Higher level concepts

 Applied concepts, important tools and skills for labs, clarification of lectures, exam coverage

Labs (7)
   
The heart of the course 1-2 weeks each Provide in-depth understanding of an aspect of systems Programming and measurement

Exams (midterm + final)
 Test your understanding of concepts & mathematical principles  Online this semester

Carnegie Mellon

Getting Help

Class Web page:
 Complete schedule of lectures, exams, and assignments  Copies of lectures, assignments, exams, solutions  Clarifications to assignments

 We won’t be using Blackboard for the course

 We won’t be using Piazza for this course


Carnegie Mellon

Getting Help

Staff mailing list:
 Use this for all communication with the teaching staff  Always CC staff mailing list during email exchanges  Send email to individual instructors only to schedule appointments

Office hours (starting Sunday Jan 19th):
 SMTWR, 6:00-8:00pm, WeH 5207

1:1 Appointments
 You can schedule 1:1 appointments with any of the teaching staff


Carnegie Mellon

Policies: Assignments (Labs) And Exams

Work groups
 You must work alone on all assignments

 Assignments due at 11:59pm on Tues or Thurs evening (except L7, which is due on a Sunday)  Electronic handins using Autolab (no exceptions!)

Conflict exams, other irreducible conflicts
 OK, but must make PRIOR arrangements with Professors  Notifying us well ahead of time shows maturity and makes us like you more (and thus to work harder to help you out of your problem)

Appealing grades
 In writing within 7 days of completion of grading  Follow formal procedure described in syllabus

Carnegie Mellon


Labs will use the Intel Computer Systems Cluster (aka “the shark machines”)
 linux> ssh  21 servers donated by Intel for 213
10 student machines (for student logins)  1 head node (for Autolab server and instructor logins)  10 grading machines (for autograding)  Each server: iCore 7: 8 Nehalem cores, 32 GB DRAM, RHEL 6.1  Rack mounted in Gates machine room  Login using your Andrew ID and password

Getting help with the cluster machines:
 Please direct questions to staff mailing list

Carnegie Mellon


Grace days
   
5 grace days for the course (none for L7) Limit of 2 grace days per lab used automatically Covers scheduling crunch, out-of-town trips, illnesses, minor setbacks Save them until late in the term!

Lateness penalties
 Once grace day(s) used up, get penalized 15% per day  No handins later than 3 days after due date

Catastrophic events
 Major illness, death in family, …  Formulate a plan (with your academic advisor) to get back on track

 Once you start running late, it’s really hard to catch up

Carnegie Mellon


What is cheating?
 Sharing code: by copying, retyping, looking at, or supplying a file  Coaching: helping your friend to write a lab, line-by-line  Copying code from previous course or from elsewhere on WWW

Only allowed to use code we supply, or from CS:APP website

What is NOT cheating?
 Explaining how to use systems or tools  Helping others with high-level design issues

Penalty for cheating:
 Removal from course with failing grade  Permanent mark on your record

Detection of cheating:
 Our tools for doing this are much better than most cheaters think!  Last Fall, 12 students were caught cheating and failed the course.

Carnegie Mellon

Other Rules of the Lecture Hall

Laptops: permitted Electronic communications: forbidden
 No email, instant messaging, cell phone calls, etc

Presence in lectures, recitations: voluntary, recommended No recordings of ANY KIND


Carnegie Mellon

Policies: Grading

Exams (50%): midterm (20%), final (30%) Labs (50%): weighted according to effort Final grades based on a combination of straight scale and possibly a tiny amount of curving.


Carnegie Mellon

Programs and Data

 Bits operations, arithmetic, assembly language programs  Representation of C control and data structures  Includes aspects of architecture and compilers

 L1 (datalab): Manipulating bits  L2 (bomblab): Defusing a binary bomb  L3 (buflab): Hacking a buffer bomb


Carnegie Mellon

The Memory Hierarchy

 Memory technology, memory hierarchy, caches, disks, locality  Includes aspects of architecture and OS

 L4 (cachelab): Building a cache simulator and optimizing for locality.

Learn how to exploit locality in your programs.


Carnegie Mellon

Exceptional Control Flow

 Hardware exceptions, processes, process control, Unix signals, nonlocal jumps  Includes aspects of compilers, OS, and architecture

 L5 (tshlab): Writing your own Unix shell.

A first introduction to concurrency


Carnegie Mellon

Virtual Memory

 Virtual memory, address translation, dynamic storage allocation  Includes aspects of architecture and OS

 L6 (malloclab): Writing your own malloc package

Get a real feel for systems-level programming


Carnegie Mellon

Networking, and Concurrency

    
High level and low-level I/O, network programming Internet services, Web servers concurrency, concurrent server design, threads I/O multiplexing with select Includes aspects of networking, OS, and architecture

 L7 (proxylab): Writing your own Web proxy

Learn network programming and more about concurrency and synchronization.


Carnegie Mellon

Lab Rationale

Each lab has a well-defined goal such as solving a puzzle or winning a contest

Doing the lab should result in new skills and concepts
We try to use competition in a fun and healthy way
 Set a reasonable threshold for full credit  Post intermediate results (anonymized) on Web page for glory!


Carnegie Mellon

Labs are provided by the CMU Autolab system
 Developed by CMU faculty and students  Key ideas: Autograding and Scoreboards
Autograding: Using VMs on-demand to evaluate untrusted code.  Scoreboards: Real-time, rank-ordered, and anonymous summary.  Used by 2,500+ students each semester, since Fall, 2010

With Autolab you can use your Web browser to:
   
Download the lab materials Handin your code for autograding by the Autolab server View the class scoreboard View the complete history of your code handins, autograded result, instructor’s evaluations, and gradebook.


Carnegie Mellon

Autolab accounts

Students enrolled as of 10am on Mon, Jan 14th have accounts You must be enrolled to get an account
 Autolab is not tied in to the Hub’s rosters  If you add in, contact for an account  Put “waitlist add” in the email subject


Carnegie Mellon

Waitlist questions
 

15-213: Catherine Fichtner ( 18-213: Jennifer Loughran ( Please don’t contact the instructors with waitlist questions.


Carnegie Mellon

Welcome and Enjoy!


Similar Documents

Free Essay


...Hundreds(of(real(personal(accounts(of Group'Discussions'&'Personal'Interviews during(MBA(admissions(to(India’s(best(B9schools Written'by Compiled'by Loads'of'MBA'Aspirants The'PaGaLGuY'MadCapz'Group Antholo gy Hundreds of real personal accounts of Group Discussions and Personal Interviews during MBA admissions to India’s best business schools. In this edition: The IIMs at Ahmedabad, Bangalore, Calcutta, Lucknow, Indore & Kozhikode. Written by Loads of MBA aspirants Compiled by The PaGaLGuY MadCapz Team PaGaLGuY GD-PI Anthology Copyright © 2011, All text and content in this document is solely owned by Reproduction without permission in any form or means is illegal. Special copy prepared exclusively for mustafa rokerya Get your own Free personalized copy (with your name on it) of this book from What this book is about What is a real IIM interview like? What kind of questions do they ask and what judgments do applicants have to make while answering them? Since 2003, those with real Group Discussion and Personal Interview calls from India’s top bschools have been posting entire and detailed transcripts of their admission interviews immediately after they happen, so that others slotted for later interviews can learn what GDPI is going to be like this year. This book is a collection of dozens of handpicked GDPI experiences from the country’s top bschools during the......

Words: 178933 - Pages: 716