home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.os.linux
- Path: sparky!uunet!ods!david
- From: david@ods.com (David Engel)
- Subject: Re: find 3.7 and fileutils 3.3 uploaded
- Message-ID: <1992Jul29.210622.5397@ods.com>
- Organization: Optical Data Systems, Inc.
- X-Newsreader: Tin 1.1 PL4
- References: <1992Jul29.193543.1119@crd.ge.com>
- Date: Wed, 29 Jul 1992 21:06:22 GMT
- Lines: 27
-
- davidsen@ariel.crd.GE.COM (william E Davidsen) writes:
- : BTW: how much os a hit does the library take on using the jump table?
- : Between the extra level of indirect, dumping the cache and pipeline
- : twice as often, and paging on small systems, this looks like a potential
- : 5% increase in clock time. Do you have numbers? I see people taking the
-
- The very rough measurements I did a couple of months ago showed a hit of
- about 2-4%. The programs I tested were mostly I/O bound and the method
- I used to time them was rather crude though, so I don't know how realistic
- those numbers are. If someone has any compute bound programs that exercise
- libc, I like to know about it so I can try them.
-
- : Note, I'm looking for clarification rather than saying "this is bad."
- : My guess of what it adds to small systems calls could be way off, etc.
- : I'm well aware of the benefits of jump tables, just concerned about
- : possible performance costs. Maybe the only hit is the extra pieline
- : dump, I suspect the table would be memory, if not in cache, all the
- : time.
-
- Yes, if all or most all progarms use the jump-table libs (like they do on
- my system :), the jump-table would almost always be in memory.
-
- David
- --
- David Engel Optical Data Systems, Inc.
- david@ods.com 1101 E. Arapaho Road
- (214) 234-6400 Richardson, TX 75081
-