home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.amiga.programmer
- Path: sparky!uunet!ftpbox!mothost!lmpsbbs!supra.comm.mot.com!rittle
- From: rittle@supra.comm.mot.com (Loren James Rittle)
- Subject: Re: SAS/C 6.0 Details (LONG!)
- Organization: Land Mobile Products Sector, Motorola Inc.
- Date: Mon, 17 Aug 1992 16:06:12 GMT
- Message-ID: <1992Aug17.160612.5738@lmpsbbs.comm.mot.com>
- References: <BsxCu2.5A7@unx.sas.com> <VINSCI.92Aug15020642@nic.nic.funet.fi>
- Sender: news@lmpsbbs.comm.mot.com (Net News)
- Nntp-Posting-Host: 145.1.80.40
- Lines: 28
-
- In article <VINSCI.92Aug15020642@nic.nic.funet.fi> vinsci@nic.funet.fi (Leonard Norrgard) writes:
- >Doug, how does the new code generator of SAS/C 6.0 measure up against
- >GCC 2? Why not post some data?
-
- Well, I'm not Doug, but in fact since I don't work for SAS I may be a
- better judge... :-)
-
- I use GCC 2.2.2 and c89 at work (DECstation); Microtec C at work
- (crosscompiler on DECstation for 680x0); and GCC 2.1 and SAS/C at home
- (Amiga). I am quite impressed with the code generator / [auto] inliner
- / optimizer in SAS/C 6.0. Without listing a bunch of meaningless
- benchmarks, I'll say that SAS/C 6.0 puts SAS/C on par with GCC 2.x. I
- will compare SAS/C 5.1b to SAS/C 6.0 in a general sense:
-
- ls-4.5ljr when recompiled under SAS/C 6.0 goes on average
- 10-20% faster and is 10-20% smaller than the 5.1b compiled version.
-
- I have some other programs that show better speedup factors.
-
- >I have a special interest in the code generation for switch()
- >...
- >search, rather than through a linear search. Has this been
- >implemented yet?
-
- I can't answer this.
-
- Regards,
- Loren
-