home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!paladin.american.edu!gatech!emory!ogicse!mintaka.lcs.mit.edu!ai-lab!mobot!bleck
- From: bleck@mobot.ai.mit.edu (Olaf Bleck)
- Newsgroups: comp.sys.m68k
- Subject: SUMMARY: PC based CPU32 compilers--need suggestions
- Message-ID: <1h5fjjINN4mn@life.ai.mit.edu>
- Date: 21 Dec 92 22:14:11 GMT
- Article-I.D.: life.1h5fjjINN4mn
- Sender: bleck@mobot (Olaf Bleck)
- Reply-To: bleck@ai.mit.edu
- Distribution: world
- Organization: MIT Artificial Intelligence Laboratory
- Lines: 149
- NNTP-Posting-Host: mobot.ai.mit.edu
-
- Thanks to everyone who replied!
- -Olaf
-
- Summary of responses: PC based CPU32 compilers--need suggestions
-
- ***********
-
- For less than the cost of the new compiler, you can buy a used sun-3, and
- then you don't have to change anything. You can have 24-hour access too! :-)
-
- PC's really are horrible to work with, avoid'em if you can!
-
- ---------------------
- ovrskeek@matt.ksu.ksu.edu
-
- ***********
-
- >I'm currently developing some things for a 68332, and I'm interested in
- >porting my development stuff to the IBM PC world.
-
- I have been using the compiler from MicroTec (located in Santa Clara, CA)
- for about 2 years now. I am very satisfied with their product. It is a
- bit expensive (about $4,000 for just the compiler) but hits just about all
- of your requirements.
-
-
- >1) My libraries make use of some of the non-standard features of GCC, such as
- >the "inline" function declaration, and inlined assembler formats. Ideally I
- >would like to find a compiler that is GCC compatible, so that I don't have to
- >rethink all the inlined stuff.
-
- I know for sure that you can use in-line assembler formats, but I'm not too
- sure about in-line function declaration. I'll check the manuals to see if
- it is supported if you want me to, but I believe it is.
-
-
- >2) Cost shouldn't be more than around a few $1000's, less ideally.
-
- This might be a bit of a problem. I think they are charging about $4,000
- for just the compiler/linker/assembler, but you will have to get a price
- quote from them for that.
-
- >3) It should have downloader capabilites, either having it's own that talks
- >to the motorola debugger, or be compatible with the motorola downloader. It
- >should also be able to generate a binary file that can be burned into an
- >eprom. (This could also be used instead of a downloader, since I could use
- >NVRAMS and to them via a programming socket).
-
- The compiler can make Motorola S-records for you and can produce IEEE-695
- (or some standards number) output with "MicroTec enhancements". This
- second format is used with their XRAY debug tool. For burning EPROMs, you
- will just want to create S-records. I have done this with their product on
- at least 10 different projects of my own, and have no problem with it. If
- you spend the bucks and get one of their XRAY source level debugger
- packages, you can use the IEEE-695 format. This proves a very nice
- environment for doing source-level debug. It is similar to using
- MicroSoft's CodeView debugger, but I happen to like using XRAY better. You
- can easily display & modify variables, set breakpoints, trace interrupt
- routines, modify registers, etc. You can intermix assembler and 'C' source
- or just work at the 'C' source level. Very nice. They have XRAY available
- in several flavors, including :
-
- 1) Simulator XRAY, which just runs on the PC host with no 68K product
- installed. XRAY entirely simulates the 68000 cpu and executes all
- instructions in an artificial environment. Kinda useful at times if
- you have the bucks, but not my first choice if you don't.
-
- 2) In-circuit monitor XRAY, which requires a 68000-family CPU and a
- dedicated serial port. You burn a small (about 11 Kbyte) program into
- EPROM and use it to boot your processor. The PC then connects via a
- serial line to the processor and you are then able to issue various
- commands to download code, debug code, etc. The user interface is
- identical to the simulator XRAY, only now you have real hardware with
- real interrupts. I find this option to be most useful, and would
- recommend this one if you have to pick one of the three XRAY versions.
- You must have enough RAM to download the code you want to debug. This
- is in addition to RAM used for stack and global or static variables.
-
- 3) ICE XRAY, which uses one of several standard ICE products from various
- hardware vendors. I have used this with the HP64700 series of emulators
- on a consulting project I am working on, and find it to be not as nice
- as the monitor XRAY (option 2). It's response is slower and is _much_
- more buggy than either of the other two XRAY packages. It does have the
- nice ability to set breakpoints in ROM and breakpoints for accessing
- locations rather than just breakpoints for locations that get executed
- (as option 2 has). It makes pretty good use of the capabilities of the
- ICE hardware.
-
- --
- Dave McMahan mcmahan@netcom.com
- {apple,amdahl,claris}!netcom!mcmahan
- **************
-
- How about GCC ported for MS-DOS 386+ with 2MB+. You can obtain such a beast
- from:
-
- Hundred Acre Consulting
- ATTN: Bruce Robertson
- 1280 Terminal Way, Suite 26
- Reno NV 89502-3243
- Tel: (702) 329-9333
- Fax: (702) 329-9566
- UUCP: info@pooh.com
- CIS: 76130,3034
-
- Unsupported package with full source and postscript doc on disk is $100 US.
- As above with printed docs (highly recommended) is $150 US.
- As above with 1 year email, phone, fax support is $500 US.
-
- I have the above unsupported package for the 68K series of targets. Bruce is
- very good to deal with. Responses and help are quick. The package includes
- gcc 2.3, as, ld, bintools, gdb. It sounds like it will match your requirements
- completely.
-
- When ordering from Bruce, tell him you heard abouth this from me (Alex Leyn,
- Leitch Video Int'l Inc.)
-
- --
- University of Waterloo, Electrical & Computer Engineering, Class of 1992
- Internet: 4BClass@EandCE.UWaterloo.CA
- c/o Alex Leyn
-
- ******************
-
- Return-Path: <benjii!chris@uunet.uu.net>
- Date: Sat, 19 Dec 92 13:54:41 PST
- From: benjii!chris@uunet.uu.net (Chris Rosebrugh)
- To: bleck@ai.mit.edu
- Subject: Re: PC based CPU32 compilers--need suggestions
-
- we've been using Intermetrics' Itools for the last 2 years.
- we develop a 68331 pen-based computer and use their Itools
- for development on both the sun and the PC. The code they
- generate is very good. We switched to CrossCode from
- SDS (Software Development Systems) for a while early
- this year because Intermetrics was dragging their feet
- in putting the PC version out which uses protected mode
- instead of real mode. But, they've got the right stuff
- up and running now (and so do we). FYI, the SDS compiler
- is slow and the generated code is generally about 15%
- slower than Itools.
-
-
-
- --
- Tear Here:
- -----------------------------------------------------------------
- _________________________________________________________________
- Olaf Bleck bleck@ai.mit.edu
-