Subject: Re: emacs
To: Rod Morehead <>
From: Bharat S. Jhaveri <bjhaveri@Arco.COM>
List: macbsd-general
Date: 04/25/1994 09:06:23
>Well, last night I redownloaded and rebuilt emacs 19.22. What follows are
>my notes from the sucessfult attempt.
>        1. Extracted tar archive.
>        2. Did  a configure using m68k-hp-netbsd.
>        3. Had to change 
>                extern char *sys_errlist[]
>           to 
>                extern const char *const sys_errlist[]
>           the in the following files:
>                lib-src/timer.c
>                lib-src/emacsclient.c
>                src/sysdep.c
>                src/process.c
>        4. Removed Control 'L's from src/Makefile (^L). The
>           bsd make did not like them.
>        5. In file src/process.c had to change first argument to wait3
>           to an int*. Made int named tmp. The was aprox. at line 2720

> I have been trying still to compile emacs. I have followed your suggestion,
> but still no luck. It looks like everything gets built. But when I run
> emacs, I get "Abort trap". And emacs quits. My question is step 5 of your
> info. I may have not followed it right. Could you please clarify it. At
> line 2720, I added
>     int* tmp;

Very close.  What you need is.
      int  tmp;
The rest is correct. Changing this could fix your crash.

>        6. Edited Makefile to add -static option to link step. The avoids
>           the dynamic error. I added on the -nostdlib line toward the top 
>           of the Makefile.
>That was all. It seems to work ok. Let me know if the does not work for you.
>Rod Morehead

 Well, you were quite right. All I had to do was change "int *tmp" to "int
tmp" in process.c and everything went perfectly. Emacs compiled and
installed perfectly. I have been using it since friday (TERM d7) without
any problems. 

 I sincerely wnt to thank you and others, Edward Wolpert, Allen Briggs,
Lawrence Kesteloot,(and others I may have missed,) for helping me make
emacs work on macbsd. It is working really great.

 One other note. Rod, as you had mentioned, my gcc 2.5.8 was compiled
right. All I had to use was -traditional flag and it worked and compiles
most of the programs. So I went ahead and got libg++ from
gcc compiles fine most of the libg++ except when it gets to libio and finds
some VFOEK redefined. I think it is and error that even -tradional does not
handle. Well I am still trying. But, if it rings a bell that might help,
please let me know.

Allen. I switched macbsd from internal hard drive to external (SCID 1)
quantum drive. And like a miracle, all my kernel hanging and fsck problems
have disappered. I suspected that the internal Seagate disk had defects.
But the APS disk partitioning program found internal disk to be normal.
Although it found 30 defects, it said that they were normal and software
corrects them. My external hard drive is on loan for a couple of weeks. And
I am wondering, should I get a new external or internal disk. Is it
possible that macbsd kernel is more stable when installed on external disk?
Or that my problems with kernel hanging on the internal hard disk were dut
to those 30 defects that UNIX can not take into account and correct, I
know, from APS output what sectors and heads these defects are. But the
newfs command in UNIX have no input for specifying bad blocks on the hard

Anyway, on my external hard drive, macbsd is completely stable. I have been
working on it all weekend, compiled emacs, gcc, libg++ and lots of
extracting large (30 MB) tar archives, and absolutely no problems. It is
really wonderful that I do not have to reboot every few hours. It is as
stable as the AIX on my RS6000. So, if you have any suggestion about
internal vs. external disk for installing macbsd, or using newfs with
specifying known defects on a disk, please let me know.


Bharat S. Jhaveri