Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!wuarchive!brutus.cs.uiuc.edu!psuvax1!rutgers!mit-eddie!bu-cs!encore!maxzilla.encore.com
From: goldman@maxzilla.encore.com (Steve Goldman)
Newsgroups: comp.sys.encore
Subject: Re: Getting the most for a process.
Message-ID: <10131@encore.Encore.COM>
Date: 15 Oct 89 01:37:34 GMT
References: <3314@gouda.quad.com>
Sender: news@Encore.COM
Lines: 26

From article <3314@gouda.quad.com>, by david@quad1.quad.com (David A. Fox):
> In article <20140@mimsy.UUCP> chris@mimsy.UUCP (Chris Torek) writes:
>>
>>Unfortunately, the Encore version of cc, which is apparently a Greenhills
>>C compiler, has all of its `phases' built in.  Thus, if you are compiling
>>a single file, you cannot preprocess on cpu 0, compile on cpu 1, and
>>assemble on cpu 2 all at the same time.

  Several years ago we built versions of the C and f77 compilers that ran
as group of processes, the results were not impressive and those versions
were never released. Parallel make was much more effective.

> Tsk, tsk. You are forgetting about the -X182 option to cc :-) : 
> 
> ..c.o:
> 	$(CC) $(CFLAGS) -S -c -Wc,-X182 $< | as -o $*.o 

Which is entirely equivalent to:
  	$(CC) $(CFLAGS) -c -q nodirect $< 
and less likely to get bad results when switch numbers change :-)
Though, I'd be surprised if this beats direct code generation in cc...

Steve Goldman, Encore Computer Corp          (919) 481-3730
901 Kildaire Farm rd., bldg D  Cary, NC  27511       USA
arpa: goldman@encore.com 
uucp: {bu-cs,decvax,gould}!encore!goldman