|
|
|
Modern Operating Systems by Herbert Bos and Andrew S. Tanenb...
Modern_Operating_Systems_by_Herbert_Bos_and_Andrew_S._Tanenbaum_4th_Ed.pdf
Showing 1043 out of 1137
Modern Operating Systems by Herbert Bos and Andrew...
Modern_Operating_Systems_by_Herbert_Bos_and_Andrew_S._Tanenbaum_4th_Ed.pdf-M ODERN O PERATING S YSTEMS
Modern Operating Systems by Herbert...
Modern_Operating_Systems_by_Herbert_Bos_and_Andrew_S._Tanenbaum_4th_Ed.pdf-M ODERN O PERATING S YSTEMS
Page 1043
1012
OPERATING SYSTEM DESIGN
CHAP. 12
Here is a true story of where an optimization did more harm than good. One of
the authors (AST) had a former student (who shall here remain nameless) who
wrote the original MINIX
mkfs
program. This program lays down a fresh file sys-
tem on a newly formatted disk.
The student spent about 6 months optimizing it,
including putting in disk caching. When he turned it in, it did not work and it re-
quired several additional months of debugging. This program typically runs on the
hard disk once during the life of the computer, when the system is installed.
It also
runs once for each disk that is formatted.
Each run takes about 2 sec. Even if the
unoptimized version had taken 1 minute, it was a poor use of resources to spend so
much time optimizing a program that is used so infrequently.
A slogan that has considerable applicability to performance optimization is
Good enough is good enough.
By this we mean that once the performance has achieved a reasonable level, it is
probably not worth the effort and complexity to squeeze out the last few percent.
If the scheduling algorithm is reasonably fair and keeps the CPU busy 90% of the
time, it is doing its job. Devising a far more complex one that is 5% better is proba-
bly a bad idea.
Similarly, if the page rate is low enough that it is not a bottleneck,
jumping through hoops to get optimal performance is usually not worth it. Avoid-
ing disaster is far more important than getting optimal performance, especially
since what is optimal with one load may not be optimal with another.
Another concern is what to optimize when. Some programmers have a tenden-
cy to optimize to death whatever they develop, as soon as it is appears to work. The
problem is that after optimization, the system may be less clean, making it harder
to maintain and debug. Also, it makes it harder to adapt it, and perhaps do more
fruitful optimization later. The problem is known as premature optimization. Don-
ald Knuth, sometimes referred to as the father of the analysis of algorithms, once
said that ‘‘premature optimization is the root of all evil.’’
12.4.3 Space-Time Trade-offs
One general approach to improving performance is to trade off time vs.
space.
It frequently occurs in computer science that there is a choice between an algo-
rithm that uses little memory but is slow and an algorithm that uses much more
memory but is faster. When making an important optimization, it is worth looking
for algorithms that gain speed by using more memory or conversely save precious
memory by doing more computation.
One technique that is sometimes helpful is to replace small procedures by
macros. Using a macro eliminates the overhead that is associated with a procedure
call. The gain is especially significant if the call occurs inside a loop.
As an ex-
ample, suppose we use bitmaps to keep track of resources and frequently need to
know how many units are free in some portion of the bitmap. For this purpose we
will need a procedure,
bit
count
, that counts the number of 1 bits in a byte. The
Ace your assessments! Get Better Grades
Browse thousands of Study Materials & Solutions from your Favorite Schools
Concordia University
Concordia_University
School:
Operating_Systems
Course:
Great resource for chem class. Had all the past labs and assignments
Leland P.
Santa Clara University
Introducing Study Plan
Using AI Tools to Help you understand and remember your course concepts better and faster than any other resource.
Find the best videos to learn every concept in that course from Youtube and Tiktok without searching.
Save All Relavent Videos & Materials and access anytime and anywhere
Prepare Smart and Guarantee better grades
Know more
Students also viewed documents
lab 18.docx
lab_18.docx
Course
Course
3
Module5QuizSTA2023.d...
Module5QuizSTA2023.docx.docx
Course
Course
10
Week 7 Test Math302....
Week_7_Test_Math302.docx.docx
Course
Course
30
Chapter 1 Assigment ...
Chapter_1_Assigment_Questions.docx.docx
Course
Course
5
Week 4 tests.docx.do...
Week_4_tests.docx.docx
Course
Course
23
Week 6 tests.docx.do...
Week_6_tests.docx.docx
Course
Course
106
