home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!elroy.jpl.nasa.gov!sdd.hp.com!uakari.primate.wisc.edu!usenet.coe.montana.edu!news.u.washington.edu!serval!fili!hlu
- From: hlu@fili.eecs.wsu.edu (H.J. Lu)
- Newsgroups: comp.os.linux
- Subject: Re: Jumptable Performance (Was: Re: shared libs - can everyone be happy with this?)
- Message-ID: <1992Aug21.182052.11544@serval.net.wsu.edu>
- Date: 21 Aug 92 18:20:52 GMT
- References: <1992Aug17.144719.1961@crd.ge.com> <1992Aug17.151311.29507@ods.com> <NOP.92Aug17135014@theory.Mankato.MSUS.EDU> <1992Aug18.080437.3944@fys.ruu.nl> <1992Aug18.140858.3484@crd.ge.com> <1992Aug18.154149.26416@fys.ruu.nl> <1992Aug19.125541.865@crd.ge.com> <19
- Sender: hlu@fili (H.J. Lu)
- Organization: Washington State University
- Lines: 24
-
- In article <1992Aug21.081823.571@donau.et.tudelft.nl>, wolff@zen.et.tudelft.nl (Rogier Wolff) writes:
- |> davidsen@ariel.crd.GE.COM (william E Davidsen) writes:
- |>
- |> > If more instructions and bigger size really improved performance we'd
- |> >all leave -O off our compiles, right. Therefore it's an interesting
- |> >little puzzle.
- |> >--
- |> >bill davidsen, GE Corp. R&D Center; Box 8; Schenectady NY 12345
- |> > I admit that when I was in school I wrote COBOL. But I didn't compile.
- |>
- |> That's what optimizing is all about: The only difference between GCC codee
- |> for 486 and 386 is that in the 486 code some extra NOPs are added.
- |>
-
- It is a little bit more than that. Some code sequences are faster for 486.
- Get gcc-2.2.2.tar.Z and read i386.md for details.
-
- |> This will make the branch targets align on 16 byte boundaries, which will
- |> improve cache performance.
- |>
- |> Roger
-
-
- H.J.
-