[gnu.bash.bug] ISI68K

ray@GIBBS.PHYSICS.PURDUE.EDU (Ray Moody) (06/14/89)

>Also, Has anyone gotten bash to compile on an IS68K (68020/<math co-pr>) 
>running under Bsd 4.2..

    I have been running bash on an ISI68K with an 68881 math
co-processor under under BSD4.3  I would hazzard a guess that my
4.3BSD binary would run on your 4.2BSD machine.  Yes, strchr() is in
libc on my machine.  You can make your own out of index().  See line
331 in file "general.c".

    Bash gives off "billiyons and billiyons" of compiler warnings as
it compiles, but it works anyway.  I plan to fix this someday.

    Also, if someone is working on a port to a UMAX machine, please
use "#ifdef UMAX" only inside "#ifndef ISI68K" because <sys/param.h>
defines a macro called UMAX.  Yes, I know, this is gross, but I didn't
write this OS, I just make it go.

								Ray

composer@BU-CS.BU.EDU (Jeff Kellem) (06/14/89)

|   Date: Tue, 13 Jun 89 22:41:47 EST
|   From: ray@gibbs.physics.purdue.edu (Ray Moody)
|
|       Bash gives off "billiyons and billiyons" of compiler warnings as
|   it compiles, but it works anyway.  I plan to fix this someday.

Actually, I believe that James Van Artsdalen is doing a good amount of
work towards cleaning up that aspect.  So, you may not have to worry
about it.  (Actually, with gcc 1.34 and 1.35 straight [no -Wall], bash
0.99 compiles fine with no errors whatsoever.  It's just when you get
to "standard" vendor compilers that you have trouble.   :)

|       Also, if someone is working on a port to a UMAX machine, please
|   use "#ifdef UMAX" only inside "#ifndef ISI68K" because <sys/param.h>
|   defines a macro called UMAX.  Yes, I know, this is gross, but I didn't
|   write this OS, I just make it go.

Well, I've had bash running on an Encore Multimax running UMAX 3.3 since
version 0.95.  So far, no machine specific changes need to be made.  :)
Out of curiosity, what is the UMAX macro on the ISI68K for?  Thanks for
the info, though.

|								   Ray

				-jeff

Jeff Kellem
INTERNET: composer@bu-cs.bu.edu  (or composer%bu-cs.bu.edu@bu-it.bu.edu)
UUCP: ...!harvard!bu-cs!composer

james@bigtex.cactus.org (James Van Artsdalen) (06/15/89)

> Date: Wed, 14 Jun 89 00:01:53 EDT
> From: composer@bu-cs.bu.edu (Jeff Kellem)

|   Date: Tue, 13 Jun 89 22:41:47 EST
|   From: ray@gibbs.physics.purdue.edu (Ray Moody)

|       Bash gives off "billiyons and billiyons" of compiler warnings as
|   it compiles, but it works anyway.  I plan to fix this someday.

> Actually, I believe that James Van Artsdalen is doing a good amount of
> work towards cleaning up that aspect.

Trying anyway.  Turns out to be quite difficult.  bash will not work
on anything where sizeof int != sizeof char *, or where initialized
char variables take up less than sizeof int bytes, and probably not
with a really smart globally optimizing compiler (I wouldn't recommend
using gcc's inlining just yet).

I have a set of ANSI function prototypes and a complete set of extern
declarations for variables.  My eventual goal is to be able to compile
with PCC (one that supports void) without warnings, and to compile
with gcc -Wall without warnings.

The function types have had to undergo a lot of changes, and some of
the variables have also needed new types.  I don't know when or how
the function prototypes will get merged in to the real source - it may
be a while.
---
James R. Van Artsdalen          james@bigtex.cactus.org   "Live Free or Die"
Dell Computer Co    9505 Arboretum Blvd Austin TX 78759         512-338-8789