Path: utzoo!attcan!uunet!husc6!spdcc!ima!compilers-sender From: daryl@ihlpe.att.com Newsgroups: comp.compilers Subject: Re: Compiler complexity (was: VAX Always Uses Fewer Instructions) Message-ID: <1129@ima.ISC.COM> Date: 21 Jun 88 05:26:26 GMT Sender: compilers-sender@ima.ISC.COM Reply-To: daryl@ihlpe.att.com Lines: 32 Approved: compilers@ima.UUCP In-Reply-To: your article <1117@ima.ISC.COM> > In article <20338@beta.lanl.gov>, jlg@beta.lanl.gov (Jim Giles) writes: > > Do modern compilers for CISC make good > > use of the variety in the instruction set? > > J. Giles > The current state of the art in retargetable code generators is such > that the addressing modes of CISC machines can be easily handled > (with the possible exeption of addressing mode with side-effects). > Complex instructions like the VAX polynomial instruction and string > move instructions are still problematic. One thing I noticed on a VMS machine once that most UNIX hacks wouldn't see is that languages such as PL/I and PASCAL with the internal strings and nested subroutines made far better use of the VAX instruction set than the C compiler. Have any RISC studies been done with languages with more abstractions than C? Daryl Monge UUCP: ...!ihnp4!ihcae!daryl AT&T CIS: 72717,65 Bell Labs, Naperville, Ill AT&T 312-979-3603 [For the IBM 801 work, their working language was PL.8, the 80% of PL/I that they found useful. As far as I know, the conclusions they came to were much the same as the other RISC efforts, even though unlike many of the other groups they from the first included some of the best compiler people in the world. They have since grafted C and Pascal front ends on to the PL.8 compiler and they apparently produce good code, too, though some of their early misunderstandings about the semantics of C were pretty amazing. -John] -- Send compilers articles to ima!compilers or, in a pinch, to Levine@YALE.EDU Plausible paths are { ihnp4 | decvax | cbosgd | harvard | yale | bbn}!ima Please send responses to the originator of the message -- I cannot forward mail accidentally sent back to compilers. Meta-mail to ima!compilers-request