Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!ncar!asuvax!mcdphx!udc!chant!aglew From: aglew@urbana.mcd.mot.com (Andy-Krazy-Glew) Newsgroups: comp.arch Subject: Re: flexible caches Message-ID: Date: 19 Sep 89 15:18:51 GMT References: <224@qusunr.queensu.CA> <22151@cup.portal.com> <2115@munnari.oz.au> <22211@cup.portal.com> <12907@pur-ee.UUCP> Sender: aglew@urbana.mcd.mot.com Organization: Work: Motorola MCD, Urbana Design Center; School: University of Illinois at Urbana-Champaign Lines: 19 In-reply-to: hankd@pur-ee.UUCP's message of 18 Sep 89 19:51:38 GMT [Hankk Dietz writes:] >Many optimizing-compiler writers have considered improving the >compiler's analysis by using stats collected by trial executions, but >it rarely helps. Statistics based on runtime data are notoriously >unstable. Woah! The group that I am peripherally associated with at the University of Illinois has been using runtime feedback to compiler optimization for a while (see Hwu and Chang, or Hwu, Conte, and Chang's papers in the most recent Int'l Symp on Computer Architecture) and getting good results. I think it depends on what questions you expect to answer with the runtime stats, and what sort of "trials" they are collected from. -- Andy "Krazy" Glew, Motorola MCD, aglew@urbana.mcd.mot.com 1101 E. University, Urbana, IL 61801, USA. {uunet!,}uiucuxc!udc!aglew My opinions are my own; I indicate my company only so that the reader may account for any possible bias I may have towards our products.