delorie.com/archives/browse.cgi   search  
Mail Archives: cygwin-developers/1998/06/03/11:13:40

From: ggp AT informix DOT com (Guy Gascoigne - Piggford)
Subject: Re: Update on status of submitted patches
3 Jun 1998 11:13:40 -0700 :
Message-ID: <3.0.3.32.19980603105822.009b9170.cygnus.cygwin32.developers@pop.pdx.informix.com>
References: <199806031729 DOT NAA11519 AT kramden DOT cygnus DOT com>
Mime-Version: 1.0
To: cygwin32-developers AT cygnus DOT com

At 01:29 PM 6/3/98 -0400, you wrote:
>>Out of curiosity, is there a performance improvement if you a dual CPU
>>machine?
>
>I tested this on a dual Pentium 300 running Windows NT 4.0.  I could not
>discern the slightest speed improvement with -pipe.  I wonder if there
>would be a change on a different architecture, though.  Possibly I
>didn't notice a difference because my SCSI disk layout is so efficient.

Strange, let me get this right, without -pipe, the compiler uses
intermediate (temp) files to pass data between the preprocessor, compiler
and then assembler.  Thus forcing the processes to be run sequentially.
With -pipe you ought to see at least two and possibly all three processes
running simultaneously.

On a single CPU system I can see that the difference may well be negligible
(or worse), but I can't really see how on a multi cpu system it can't be
faster unless something else isn't broken stopping this from streaming
correctly.

Oh well, I don't have my second CPU yet anyway :-(

Guy

- Raw text -


  webmaster     delorie software   privacy  
  Copyright 2019   by DJ Delorie     Updated Jul 2019