Introduction to Computer Systems.ppt

1、Introduction to Computer Systems,Topics: Theme Five great realities of computer systems How this fits within CS curriculum Logistical issues,15-213 F 08,class01.ppt,15-213 “The Class That Gives CMU Its Zip!”,Randal E. Bryant August 26, 2008,Course Theme,Abstraction is good, but dont forget reality!

2、Most CS courses emphasize abstraction Abstract data types Asymptotic analysis These abstractions have limits Especially in the presence of bugs Need to understand underlying implementations Useful outcomes Become more effective programmers Able to find and eliminate bugs efficiently Able to tune pro

3、gram performance Prepare for later “systems” classes in CS & ECE Compilers, Operating Systems, Networks, Computer Architecture, Embedded Systems,Great Reality #1,Ints are not Integers, Floats are not RealsExamples Is x2 0? Floats: Yes! Ints:40000 * 40000 160000000050000 * 50000 ? Is (x + y) + z = x

4、+ (y + z)? Unsigned & Signed Ints: Yes! Floats: (1e20 + -1e20) + 3.14 3.141e20 + (-1e20 + 3.14) ?,Code Security Example,Similar to code found in FreeBSDs implementation of getpeername. There are legions of smart people trying to find vulnerabilities in programs Think of it as a very stringent testin

5、g environment,/* Kernel memory region holding user-accessible data */ #define KSIZE 1024 char kbufKSIZE;/* Copy at most maxlen bytes from kernel region to user buffer */ int copy_from_kernel(void *user_dest, int maxlen) /* Byte count len is minimum of buffer size and maxlen */int len = KSIZE maxlen

6、? KSIZE : maxlen;memcpy(user_dest, kbuf, len);return len; ,Typical Usage,/* Kernel memory region holding user-accessible data */ #define KSIZE 1024 char kbufKSIZE;/* Copy at most maxlen bytes from kernel region to user buffer */ int copy_from_kernel(void *user_dest, int maxlen) /* Byte count len is

7、minimum of buffer size and maxlen */int len = KSIZE maxlen ? KSIZE : maxlen;memcpy(user_dest, kbuf, len);return len; ,#define MSIZE 528void getstuff() char mybufMSIZE;copy_from_kernel(mybuf, MSIZE);printf(“%sn”, mybuf); ,Malicious Usage,/* Kernel memory region holding user-accessible data */ #define

8、 KSIZE 1024 char kbufKSIZE;/* Copy at most maxlen bytes from kernel region to user buffer */ int copy_from_kernel(void *user_dest, int maxlen) /* Byte count len is minimum of buffer size and maxlen */int len = KSIZE maxlen ? KSIZE : maxlen;memcpy(user_dest, kbuf, len);return len; ,#define MSIZE 528v

9、oid getstuff() char mybufMSIZE;copy_from_kernel(mybuf, -MSIZE);. . . ,Computer Arithmetic,Does not generate random values Arithmetic operations have important mathematical properties Cannot assume “usual” properties Due to finiteness of representations Integer operations satisfy “ring” properties Co

10、mmutativity, associativity, distributivity Floating point operations satisfy “ordering” properties Monotonicity, values of signs Observation Need to understand which abstractions apply in which contexts Important issues for compiler writers and serious application programmers,Great Reality #2,Youve

11、got to know assembly Chances are, youll never write program in assembly Compilers are much better & more patient than you are Understanding assembly key to machine-level execution model Behavior of programs in presence of bugs High-level language model breaks down Tuning program performance Understa

12、nding 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!,Assembly Code Example,Time Stamp Counter Special 64-bit register in Intel-compatible

13、machines Incremented every clock cycle Read with rdtsc instruction Application Measure time required by procedure In units of clock cycles,double t; start_counter(); P(); t = get_counter(); printf(“P required %f clock cyclesn“, t);,Code to Read Counter,Write small amount of assembly code using GCCs

14、asm facility Inserts assembly code into machine code generated by compiler,static unsigned cyc_hi = 0; static unsigned cyc_lo = 0;/* Set *hi and *lo to the high and low order bitsof the cycle counter. */ void access_counter(unsigned *hi, unsigned *lo) asm(“rdtsc; movl %edx,%0; movl %eax,%1“ : “=r“ (

15、*hi), “=r“ (*lo) : “%edx“, “%eax“); ,Great Reality #3,Memory Matters: Random Access Memory is an un-physical abstractionMemory 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 s

16、pace 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,Memory Referencing Bug Example,double fun(int i) volatile double d1 = 3.14;volatile long int a2;ai

17、= 1073741824; /* Possibly out of bounds */return d0; ,fun(0) 3.14 fun(1) 3.14 fun(2) 3.1399998664856 fun(3) 2.00000061035156 fun(4) 3.14, then segmentation fault,Referencing Bug Explanation,C does not implement bounds checking Out of range write can affect other parts of program state,0,1,2,3,4,Loca

18、tion accessed by fun(i),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 lo

19、gically 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 or ML Understand what possible interactions may occur Use or develop tools to detect referencing errors,Memory System Performance Example,Hierarchical memo

20、ry organization Performance depends on access patterns Including how step through multi-dimensional array,void copyji(int src20482048,int dst20482048) int i,j;for (j = 0; j 2048; j+)for (i = 0; i 2048; i+)dstij = srcij; ,void copyij(int src20482048,int dst20482048) int i,j;for (i = 0; i 2048; i+)for

21、 (j = 0; j 2048; j+)dstij = srcij; ,The Memory Mountain,0,200,400,600,800,1000,1200,Read throughput (MB/s),Stride (words),Working set size (bytes),Pentium III Xeon,550 MHz,16 KB on-chip L1 d-cache,16 KB on-chip L1 i-cache,512 KB off-chip unified,L2 cache,L1,L2,Mem,xe,Great Reality #4,Theres more to

22、performance than asymptotic complexity Constant factors matter too! 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 execut

23、ed How to measure program performance and identify bottlenecks How to improve performance without destroying code modularity and generality,Code Performance Example,Multiply all elements of array Performance on class machines: 7.0 clock cycles per element Latency of floating-point multiplier,/* Comp

24、ute product of array elements */ double product(double d, int n) double result = 1;int i;for (i = 0; i n; i+)result = result * di;return result; ,Loop Unrollings,Do two loop elements per iteration Reduces overhead Cycles per element: u2: 7.0 u2r: 3.6,/* Unroll by 2. Assume n is even */ double produc

25、t_u2(double d, int n) double result = 1;int i;for (i = 0; i n; i+=2)result = (result * di) * di+1;return result; ,/* Unroll by 2. Assume n is even */ double product_u2r(double d, int n) double result = 1;int i;for (i = 0; i n; i+=2)result = result * (di * di+1);return result; ,u2: Serial Computation

26、,Computation (length=12)(1 * d0) * d1) * d2) * d3) * d4) * d5) * d6) * d7) * d8) * d9) * d10) * d11) Performance N elements, D cycles/operation N*D cycles,result = (result * di) * di+1;,u2r: Reassociated Computation,Performance N elements, D cycles/operation (N/2+1)*D cycles,result = result * (di *

27、di+1);,Great Reality #5,Computers do more than execute programsThey 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 processe

28、s Coping with unreliable media Cross platform compatibility Complex performance issues,Role within Curriculum,Foundation of Computer Systems Underlying principles for hardware, software, and networking,CS 213 Systems,CS 410 Operating Systems,CS 411 Compilers,Processes Mem. Mgmt,CS 441 Networks,Netwo

29、rk Protocols,ECE 447 Architecture,ECE 349 Embedded Systems,Exec. Model Memory System,CS 412 OS Practicum,CS 123 C Programming,CS 415 Databases,Data Reps. Memory Model,CS 462 Graphics,Machine Code,Arithmetic,Course Perspective,Most Systems Courses are Builder-Centric Computer Architecture Design pipe

30、lined processor in Verilog Operating Systems Implement large portions of operating system Compilers Write compiler for simple language Networking Implement and simulate network protocols,Course Perspective (Cont.),Our Course is Programmer-Centric Purpose is to show how by knowing more about the unde

31、rlying 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 Not just a course for dedicated hackers We bring out the hidden hacker in everyone Cover mater

32、ial in this course that you wont see elsewhere,Teaching staff,Instructors Prof. Randal E. Bryant Prof. Greg GangerTAs Taiyang Chen Tessa Eng Elie Krevat Bryant Lee Christopher Lu Swapnil Patil Vijay Prakash Jiri SimsaCourse Admin Cindy Chemsak (NSH 4303),Were glad to talk with you, but please send e

33、mail or phone first.,Textbooks,Randal E. Bryant and David R. OHallaron, “Computer Systems: A Programmers Perspective”, Prentice Hall 2003. http:/csapp.cs.cmu.edu This book really matters for the course! How to solve labs Practice problems typical of exam problemsBrian Kernighan and Dennis Ritchie, “

34、The C Programming Language, Second Edition”, Prentice Hall, 1988,Course Components,Lectures Higher level concepts Recitations Applied concepts, important tools and skills for labs, clarification of lectures, exam coverage Labs The heart of the course 2 or 3 weeks Provide in-depth understanding of an

35、 aspect of systems Programming and measurement Exams Test your understanding of concepts & mathematical principles Critical component of grade,Getting Help,Class Web Page http:/www.cs.cmu.edu/213 Copies of lectures, assignments, exams, solutions Clarifications to assignmentsMessage Board http:/autol

36、ab.cs.cmu.edu Clarifications to assignments, general discussion The only board your instructors will be monitoring (No blackboard or Andrew),Getting Help,Staff mailing list 15-213-staffcs.cmu.etc “The autolab server is down!” “Who should I talk to about .” “This code ., which I dont want to post to

37、the bboard, causes my computer to melt into slag.” Teaching assistants I dont get “associativity”. Office hours, e-mail, by appointment Please send mail to 15-213-staff, not a randomly-selected TA Professors Office hour or appt. “Should I drop the class?” “A TA said . but .”,Policies: Assignments,Wo

38、rk groups You must work alone on all but final lab Handins Assignments due at 11:59pm on Tues or Thurs evening Electronic handins using Autolab (no exceptions!). Conflict exams, other irreducible conflicts OK, but must make PRIOR arrangements with Prof. Ganger. Appealing grades Within 7 days of comp

39、letion of grading. Following procedure described in syllabus Labs: Talk to the lead person on the assignment Exams: Talk to Prof. Ganger.,Timeliness,Grace Days 4 for the course Covers scheduling crunch, out-of-town trips, illnesses, minor setbacks Save them until late in the term! Lateness Penalties

40、 Once grace days used up, get penalized 15%/day Typically shut off all handins 23 days after due date Catastrophic Events Major illness, death in family, Work with your academic advisor to formulate plan for getting back on track Advice Once you start running late, its really hard to catch up,Cheati

41、ng,What is cheating? Sharing code: either by copying, retyping, looking at, or supplying a copy of 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 che

42、ating? Explaining how to use systems or tools. Helping others with high-level design issues. Penalty for cheating: Removal from course with failing grade. Detection of cheating: We do check and our tools for doing this are much better than you think!,Policies: Grading,Exam Score E (out of 100): Two

43、in class exams (25% each) Final (50%) All exams are open book / open notes. Labs Score L (out of 100): 6 labs (10-25% each) Composite Score:S = (L + E + min(L,E)/3if L E: (2L + E)/3if E L: (L + 2E)/3,Achieving Composite Score Levels,Strong labs can partially offset weak exams, but not totally,Facili

44、ties,Labs will use the Intel Computer Systems Cluster (aka “the fish machines”) 15 Pentium Xeon servers donated by Intel for CS 213 Dual 3.2 Ghz 64-bit (EM64T) Nocona Xeon processors 2 GB, 400 MHz DDR2 SDRAM memory Rack mounted in the 3rd floor Wean Hall machine room. Your accounts are ready nearing

45、 readiness. Getting help with the cluster machines: See course Web page for login directions Please direct questions to your TAs first,Programs and Data (7),Topics Bits operations, arithmetic, assembly language programs, representation of C control and data structures Includes aspects of architectur

46、e and compilers Assignments L1 (datalab): Manipulating bits L2 (bomblab): Defusing a binary bomb L3 (buflab): Hacking a buffer bomb,The Memory Hierarchy (3),Topics Memory technology, memory hierarchy, caches, disks, locality Includes aspects of architecture and OS.Assignments,Exceptional Control Flo

47、w (3),Topics Hardware exceptions, processes, process control, Unix signals, nonlocal jumps Includes aspects of compilers, OS, and architectureAssignments L4 (tshlab): Writing your own shell with job control,Virtual Memory (4),Topics Virtual memory, address translation, dynamic storage allocation Inc

48、ludes aspects of architecture and OSAssignments L5 (malloclab): Writing your own malloc package Get a real feel for systems programming,Networking, and Concurrency (6),Topics High level and low-level I/O, network programming, Internet services, Web servers concurrency, concurrent server design, thre

49、ads, I/O multiplexing with select. Includes aspects of networking, OS, and architecture.Assignments L6 (proxylab): Writing your own Web proxy,Performance (2),Topics High level processor models, code optimization (control and data), measuring time on a computer Includes aspects of architecture, compilers, and OSAssignments,