fester@CPAC.WASHINGTON.EDU (01/22/91)
What is the best way to get really accurate program timings ? Is Microsoft C's CLOCK() library routine good enough, does TurboC have something better, should I be writing an assembly routine that uses a totally undocumented system call (:-)), what ? If anyone out there has had to work with really accurate timings, how did you do it ? LeaF fester@yin.cpac.washington.edu (mail or posting responses is fine. thanks.)
raymond@math.berkeley.edu (Raymond Chen) (01/22/91)
In article <9101220039.AA06964@yin.cpac.washington.edu>, fester@CPAC writes: >What is the best way to get really accurate program timings ? Most people will refer you to the book `The Zen of Assembler'. Here are the reviews it has received from the net. @book{ma:zen author = "Michael Abrash", title = "The {Z}en of Assembler", publisher = {Scott, Foresman and Company}, address = {Glenview, Illinois}, year = 1990, ISBN = {0-673-38602-3} } Volume 2 is scheduled for sometime in 1990. "A good discussion of reprogramming timer 0 in Chapter 2." -- David J. Birnbaum (djb@wjh12.harvard.edu) "A super accurate timer is presented (actually a short- and a long-term timer) along with an explanation of their operation." -- John Mitchell (johnm@sim.Berkeley.EDU) "Abrash actually develops two timers in his book. The higher accuracy timer requires you to turn off all interupts. Its primary purpose is to time small sections of assembly code. For timing full programs Abrash has a long term timer of more questionable accuracy. However, it is still a book that anyone with more than a passing interest in assembly should read. ... If you are considering learning assembly you might look at this series." --- Tom Webb (webb@uhccux.uhcc.Hawaii.Edu) -------------------- Other books and net.recommendations can be obtained from the comp.sys.ibm.pc.misc ftp site (math.princeton.edu) or via mail server (see the Frequently Asked Questions file) in the file `books'.