home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.lang.c++
- Path: sparky!uunet!usc!sdd.hp.com!ux1.cso.uiuc.edu!m.cs.uiuc.edu!vela!linx!sygnus!tpurves
- From: tpurves@sygnus.uucp (Timothy Purves)
- Subject: Re: BC++ 3.1 and 386 code generation
- Message-ID: <1992Aug29.004046.21126@sygnus.uucp>
- Organization: None
- References: <36863@uflorida.cis.ufl.edu>
- Date: Sat, 29 Aug 1992 00:40:46 GMT
- Lines: 30
-
- In article <36863@uflorida.cis.ufl.edu> asher@qtp.ufl.edu (Robert L. Asher) writes:
- >
- >Recently a friend of mine purchased the BC++ 3.1 upgrade and I thought I might
- >be interested as well. But I had some doubts and wanted to know about the
- >speed improvements and wheather 386/486 code generation had been added. He
- >indicated that it had and so did the blurb from Borland I got via e-mail. So I
- >had some programs which are fairly cpu intensive. We compiled them and made
- >some tests. 3.1 does not seem to make 386 code for me. I compiled using
- >bcc -Ot -3 -f287 code.cpp
- >and the execution file then ran slower than the 3.0 version of the code
- >compiled with the same options except a -2 used. This was suspicous in itself.
- >Then I tried running the code on a 286 machine and it did. What's going on
- >here? Did we miss something?
- >
- >Thanks for any help / info.
- >Robert
- >
- >--
- >Robert L. Asher |
- >Department of Chemistry |
- >University of Florida |
- >asher@qtp.ufl.edu |
-
- As I understand borland's 386 code generation. is just for Long int's, nothing
- else.
-
- --
-
- tpurves@sygnus (Timothy Purves)
-
-