On Fri, 11 Apr 1997, Erich Rast wrote:

> Hi,
>=20
> I don=B4t want you to get bored, but i still haven=B4t been able to solve
> my problem with gcc - it almost always exits with the error "cc1 got fata=
l
> signal 4" or signal 11. This is what i=B4ve done so far to get things wor=
k:
> 1.) checked swap area - it=B4s 10meg, not enough for programs like dt? my=
 real
> memory is 5meg (IIci)

Since your memory is really low, you may want a bit more swap but I doubt=
=20
it would fix this particular problem.

I got that error once, a few months ago while compiling some large=20
package. I tryied again (from where it left off) and it worked fine. That=
=20
was the only time this ever happenned to me. I wish I could tell you the=20
cause.

Is anybody else getting this?

> 2.) completely reinstalled the comp12 binaries on my zip (mounted on /zip=
)
> 3.) once again linked all new libexec (cc1, cc1plus etc.) to the correspo=
nding
> /usr/libexec/... on my root partition (hd 40meg)
> 4.) checked many, many gcc options e.g. i tried
>      cc -I/dt-1.1.5 -I- -c grf.c
> for this source for dt - because it still crashes i guess it=B4s not a
> problem with symbolic links (grf.c only includes from his own directory n=
o
> system includes)
> 5.) read all manual pages corresponding to gcc, signals, devices (zip...)=
, etc.
> 6.) crashed my zip installation because i thought playing with disklabel =
would
> help improve the data security, repeated step 1-5 (i don=B4t longer think
> this is a zip problem - something must be wrong with the gcc binaries on
> _my_ configuration)
>=20
> I don=B4t have to resources to do crosscompiling for NetBSD on another
> system, so please don`t tell me i should recompile gcc.
>=20
> Does anyone have any clue what the **** is wrong with my system?
> Maybe it has something to do with the lack of an FPU?

I agree with whoever said before that IIci normally has an FPU...

>=20
> Please, can anyone who knows this problem at least give me an hint how
> to find it`s source or give me advice what i could check further?
> - NetBSD is great, but without a C compiler it=B4s hmmm....blind? deaf?
> If i only have to uppgrade to -current, that=B4s ok, but i don=B4t want t=
o do
> this as long as i=B4m unsure if this solves the problem...
>=20
> Yours,
>        ---erich
>=20

-----------------------------------------------------------------
"Because user errors often produce unpredictable results, the user
should try to avoid them."

IBM MVS/XA System Programming Library.