Very Computer

by **Greg Patt** » Mon, 03 Jan 1994 17:37:58

I'm having an enormous amount of trouble with the afore mentioned error.
(recap: Internal compiler error: program cc1 got fatal signal 11)
I assume (probably incorrectly) that this is some kind of memory
problem and I seem to have lowered the frequency of the problem by
upping the waitstates in my ami bios.

Does anyone (PLEASE) have any solutions to this problem?

info:
My machine is a i486DX-50 and I've been running Linux since pl9.
The problem was rife then and seemed to almost disappear from
pl13 -> pl14f and now it's back bigger and badder than ever.

Thanks for reading (and hopefully responding),
--
Greg Patten

by **Dennis L** » Mon, 03 Jan 1994 23:55:53

I saw the same problems on my 386. The frequency of the symptom increased
until eventually I wasn't able to compile anything. Also noticed intermittent
file corruption problems.

I resolved the problem by replacing my IDE controller ($15).

by **Daeshik K** » Tue, 04 Jan 1994 05:51:20

Please do check your file system time to time.
(fsck or e2fsck which ever you have) Especially the root partition.

I used to have same problems as well as pipe error on shell,
which is gone now..

--
* Daeshik Kim (angma at IRC) \ O: (703) 689-5878 *

* uunet.UU.NET!loony!Hell!Devil crash at HiTel *

by **Rob Janss** » Tue, 04 Jan 1994 06:00:47

Quote:>I'm having an enormous amount of trouble with the afore mentioned error.
>(recap: Internal compiler error: program cc1 got fatal signal 11)
>I assume (probably incorrectly) that this is some kind of memory
>problem and I seem to have lowered the frequency of the problem by
>upping the waitstates in my ami bios.
>Does anyone (PLEASE) have any solutions to this problem?
>info:
>My machine is a i486DX-50 and I've been running Linux since pl9.
>The problem was rife then and seemed to almost disappear from
>pl13 -> pl14f and now it's back bigger and badder than ever.

I have seen this on a system at the club, and I was unable to fix it.
It happened when installing a distribution that worked flawlessly on
all other systems there, and seems to be hardware-related. But indeed
it could not be fixed with the setup. Strange...

Rob
--
-------------------------------------------------------------------------

-------------------------------------------------------------------------

by **Hannes Reinec** » Tue, 04 Jan 1994 20:17:25

: I'm having an enormous amount of trouble with the afore mentioned error.
: (recap: Internal compiler error: program cc1 got fatal signal 11)
: I assume (probably incorrectly) that this is some kind of memory
: problem and I seem to have lowered the frequency of the problem by
: upping the waitstates in my ami bios.

: Does anyone (PLEASE) have any solutions to this problem?

: info:
: My machine is a i486DX-50 and I've been running Linux since pl9.
: The problem was rife then and seemed to almost disappear from
: pl13 -> pl14f and now it's back bigger and badder than ever.

: Thanks for reading (and hopefully responding),
: --
: Greg Patten

Hi all,
As far as I know, it IS a memory problem. (It has happened to myself as I
tried to compile libtiff ... ).
Hint: start 'top' in another VC and watch the memory usage. I think it'll
be a good idea to increase the swap space ( 32 MB or something ).

Hope this helps

Greetings from Heidelberg

Hannes Reinecke

by **Oliver Raupa** » Tue, 04 Jan 1994 21:25:36

>I'm having an enormous amount of trouble with the afore mentioned error.
>(recap: Internal compiler error: program cc1 got fatal signal 11)
>I assume (probably incorrectly) that this is some kind of memory
>problem and I seem to have lowered the frequency of the problem by
>upping the waitstates in my ami bios.
>Does anyone (PLEASE) have any solutions to this problem?

Yes. It's a memory problem. Use the single-user-mode (init s) and
compile again or(and) increase swapsize.

-Oliver

by **Greg Patt** » Wed, 05 Jan 1994 00:17:10

>>I'm having an enormous amount of trouble with the afore mentioned error.
>>(recap: Internal compiler error: program cc1 got fatal signal 11)
>>I assume (probably incorrectly) that this is some kind of memory
>>problem and I seem to have lowered the frequency of the problem by
>>upping the waitstates in my ami bios.
>>Does anyone (PLEASE) have any solutions to this problem?
>Yes. It's a memory problem. Use the single-user-mode (init s) and
>compile again or(and) increase swapsize.

I have 8 MB of RAM and 16 MB of swap which is rarely ever half full.
Is this not enough swap?
--
Greg Patten Melbourne, Australia

by **Steve Wampl** » Wed, 05 Jan 1994 01:04:00

I just started (a few minutes ago) getting a similar message on all
my compiles (in the middle of building Tk...). In my case it was
the preprocessor (cpp) instead of cc1, but it is now happening on
every single compile I try, even with trivial programs. I have
20MB RAM and a 32MB swap, and have plenty of room on both (after killing
a lot of processes). I may have been low at the time the error first
occured (3 xservers running and one remote login), so it's possible
that cpp became corrupted then, and now it's still trying to run
the one in the RAM-resident disk buffers. Is there a way to force
a reread of a program from the actual disk, to see if that's the
case? (I.e. a way to invalidate the disk buffers?)

Other programs remain unaffected, for what that's worth.

by **Thomas Aaron Ins** » Wed, 05 Jan 1994 04:39:10

> : I'm having an enormous amount of trouble with the afore mentioned error.
> : (recap: Internal compiler error: program cc1 got fatal signal 11)
> : I assume (probably incorrectly) that this is some kind of memory
> : problem and I seem to have lowered the frequency of the problem by
> : upping the waitstates in my ami bios.
> As far as I know, it IS a memory problem. (It has happened to myself as I
> tried to compile libtiff ... ).

I'm not convinced. I had a similar problem (with cpp) and upgrading
my memory to 8MB real, 16MB swap didn't solve anything. Then,
someone suggested that my binary might be corrupt. Sure enough, I
replaced it with a "known working" copy & all my problems went away.

In defense of your position, a friend of mine has had the same problem
through several installations of different Linux distributions.
--

"I don't want to achieve immortality through my work. I want to achieve it
through not dying." -- Woody Allen

by **Gordon Soukore** » Mon, 03 Jan 1994 22:56:07

: I saw the same problems on my 386. The frequency of the symptom increased
: until eventually I wasn't able to compile anything. Also noticed intermittent
: file corruption problems.

: I resolved the problem by replacing my IDE controller ($15).

THis condition also happens on my SCSI drive and/or controller and my i486.

by **Rob Janss** » Wed, 05 Jan 1994 18:53:09

>>>I'm having an enormous amount of trouble with the afore mentioned error.
>>>(recap: Internal compiler error: program cc1 got fatal signal 11)
>>>I assume (probably incorrectly) that this is some kind of memory
>>>problem and I seem to have lowered the frequency of the problem by
>>>upping the waitstates in my ami bios.
>>>Does anyone (PLEASE) have any solutions to this problem?
>>Yes. It's a memory problem. Use the single-user-mode (init s) and
>>compile again or(and) increase swapsize.
>I have 8 MB of RAM and 16 MB of swap which is rarely ever half full.
>Is this not enough swap?

It cannot be fixed by adding swap, I've tried that as well.
It is also not related to bad disk sectors, I think...
But it is certainly related to the machine. In the case I researched
there were some machines, all loaded from the same set of disks, of which
only one had the behaviour. After redoing installation from scratch
the problem was still there.

Rob
--
-------------------------------------------------------------------------

-------------------------------------------------------------------------

by **Bob Mart** » Wed, 05 Jan 1994 10:20:49

: I'm not convinced. I had a similar problem (with cpp) and upgrading
: my memory to 8MB real, 16MB swap didn't solve anything. Then,
: someone suggested that my binary might be corrupt. Sure enough, I
: replaced it with a "known working" copy & all my problems went away.

: In defense of your position, a friend of mine has had the same problem
: through several installations of different Linux distributions.
: --

: "I don't want to achieve immortality through my work. I want to achieve it
: through not dying." -- Woody Allen

I had strange problem last year with ( I think ) gcc-2.2.2 . It was
getting random parse errors on perfectly fine statements. Turned out
the tar file from where I downloaded was corrupt. Got a copy from another
site and problems went away. So a corrupted binary could be a possibility.
--

bob martin

by **Bill Mitche** » Wed, 05 Jan 1994 23:34:22

Quote:>: >>Does anyone (PLEASE) have any solutions to this problem?

>: >Yes. It's a memory problem. Use the single-user-mode (init s) and
>: >compile again or(and) increase swapsize.

>: I have 8 MB of RAM and 16 MB of swap which is rarely ever half full.
>: Is this not enough swap?

>I think you're missing perhaps an obvious thing. The only time I ever
>got a fatal signal 11 from gcc was when I had an older IDE drive that
>had bad sectors, and when I downloaded the gcc stuff, it landed on
>some bad sectors. I've since replaced that drive, and everything is
>in good shape (knock wood).

>If it is a memory problem then it is a problem with either your
>SIMMs being bad or perhaps the bios settings, not with
>[...]

I also see this problem. Rarely am I able to recompile the kernel
without the compiler dying two or three times during the build. If
I just type "make zImage" again, it picks up and compiles several
more files before dying again. Eventually it gets through it.

I have linux on a SCSI disk which checks good with mkfs -c. I have
17 Mb of RAM. The RAM checks good with several memory test programs
I've tried which start up from MSDOS and then purport to test all of
base + extended + expanded RAM. Is a trusted RAM test program available
somewhere?

Responses by email as well as followup, please. I'm leaving on a long
trip in a few hours and any net.responses will expire before I'm able to
check them.

--

by **Tim Morl** » Thu, 06 Jan 1994 00:05:52

> . Is a trusted RAM test program available
>somewhere?

Hmm How about the real system thrasher, run linux and do a kernel
rebuild, if you've got enough memory that this doesn't start to thrash
then run 3 or 4 of them at once.... This will the test the machine in
a very good simulation of real use 8-))))

Tim M

by **Roland Ki** » Thu, 06 Jan 1994 21:13:55

>>>I'm having an enormous amount of trouble with the afore mentioned error.
>>>(recap: Internal compiler error: program cc1 got fatal signal 11)
>>>I assume (probably incorrectly) that this is some kind of memory
>>>problem and I seem to have lowered the frequency of the problem by
>>>upping the waitstates in my ami bios.

>>>Does anyone (PLEASE) have any solutions to this problem?

>>Yes. It's a memory problem. Use the single-user-mode (init s) and
>>compile again or(and) increase swapsize.

>I have 8 MB of RAM and 16 MB of swap which is rarely ever half full.
>Is this not enough swap?

It cannot be fixed by adding swap, I've tried that as well.
It is also not related to bad disk sectors, I think...
But it is certainly related to the machine. In the case I researched
there were some machines, all loaded from the same set of disks, of which
only one had the behaviour. After redoing installation from scratch
the problem was still there.

There's been a lot of discussion on this kind of problem before and the most
usual cause seems to have been a bad SIMM. Linux really does give the memory
a good thrashing (better than the boot-up test and better than MS-DOG or WINDOG)
so a bad SIMM will show up under conditions of extreme stress. Compiling a kernel
is one of those extreme stress conditions. So you could try swapping around the
SIMMs or even pulling one out and running without it.