Very Computer

by **Hong Zh** » Mon, 07 Nov 1994 13:12:35

Hi there,

I bought a Pentium 90 (Intel Premier PCI motherboard) system recently.
Today I downloaded Slackware from sunsite.unc.edu and installed
it from scratch. When compiling the kernel, I encountered the
following error message which I don't understand:

gcc: Internal compiler error: program cc1 got fatal signal 11
make[2]: *** [keyboard.o] Error 1
make[2]: Leaving directory `/usr/src/linux-1.1.54/drivers/char'
make[1]: *** [driversubdirs] Error 2
make[1]: Leaving directory `/usr/src/linux-1.1.54/drivers'
make: *** [linuxsubdirs] Error 2
library:/usr/src/linux# cpp: output pipe has been closed

My concern is this error may link to my new system. I would
appreciate it very much if anyone could let me know his/her
successful/unsucessful experience of compling the 1.1.54 kernel
so that I can know if there is problem with my system.

Thanks in advance,

Hong

by **Paul Ba** » Tue, 08 Nov 1994 03:41:32

>My concern is this error may link to my new system. I would
>appreciate it very much if anyone could let me know his/her
>successful/unsucessful experience of compling the 1.1.54 kernel
>so that I can know if there is problem with my system.

>Thanks in advance,

> Hong

I installed the 1.1.54 kernel on my Intel Premier PCI/II 90Mhz motherboard
and compiled the kernel with no problems. I _did_ have kernel panic's on
booting the kernel that I traced down to the WD80x3 driver through a process
of elimination (i.e. I compiled the kernel without it and it booted and ran
fine). Upgrading the kernel to 1.1.59 fixed that problem. I'm now running
1.1.61 with no problem at all.
--
Paul Bash

by **David - Morr** » Sat, 12 Nov 1994 11:00:10

>Hi there,
>I bought a Pentium 90 (Intel Premier PCI motherboard) system recently.
>Today I downloaded Slackware from sunsite.unc.edu and installed
>it from scratch. When compiling the kernel, I encountered the
>following error message which I don't understand:
>gcc: Internal compiler error: program cc1 got fatal signal 11
>make[2]: *** [keyboard.o] Error 1

There have been reported difficulties compiling the kernel using
gcc 2.6.0. Tis good to mention compiler version when reporting
compiler aborts.... Suggest you try 2.5.8 or 2.6.1 is reported
as 'well'.

by **Alan C** » Sat, 12 Nov 1994 19:35:32

>gcc: Internal compiler error: program cc1 got fatal signal 11

That normally indicates a compiler bug - except that the compiler and source
set are 'known good'. What is more likely is your machine has bad
simms/cache or wrongly set wait states and you took a hardware bug -
especially if its not reliably repeatable or occurs at different points
through the build.

Alan

--
..-----------,,----------------------------,,----------------------------,,

``----------'`----------------------------'`----------------------------''

by **Al Longye** » Sat, 12 Nov 1994 23:37:28

>>gcc: Internal compiler error: program cc1 got fatal signal 11
>That normally indicates a compiler bug - except that the compiler and source
>set are 'known good'. What is more likely is your machine has bad
>simms/cache or wrongly set wait states and you took a hardware bug -
>especially if its not reliably repeatable or occurs at different points
>through the build.

A less likely, but very common cause a few years ago, was that you simply
did not have enough memory to compile the program. If you have only 8 MB
of memory, then you should use a swap file.

(This is less likely these days because most people have at least 16 MB
of memmory. However, it is worth mentioning just in case. :))

The GCC compiler is a memory hog. It trades memory for speed and can take
a large amount of memory to compile.

If you are using XFree86 at the time, you might try to run the compile
without XFree86 to see if this solves the problem. If it does, you have
two options: increase the swap space or buy more memory.

--

by **Hong Zh** » Sun, 13 Nov 1994 05:51:43

Hi there,

I would like to thank all the people who responded to my question.
Since most of the people suggested the error would link to either
cache or memory, I tried disabled cache in the first place. No help.
After having changed memories (2 pieces of Panasonic 16Mb SIMMs),
all the problem was gone. Now the machine is very stable and fast
(a kernel compilation takes less than 9min!).

Thanks,

hong

by **Russell Leight** » Wed, 16 Nov 1994 02:30:35

I have had the same problem ... after the internal compiler
error, then more and more bad things happen, the disk gets
corrputed ... this week I swap the mother board...actually,
the 1st time this happened I thought it was the simms, they
reported parity errors after I reboooted the machine BUT if
I popped them out and exchanged them, rebooted, then put
them back it all went away...this makes me think it is the
mother board...we shall see...what is this about having
wait states set right?

Also, I cannot use >64M , if I do the pager crashes....

Any ideas...is this a generic P90/PCI thing or is
my configuration wrong/bad?

Russ

--
Russell Leighton
Taylor Computing

http://taylor.digex.net http://www.digex.net/~rrl/Welcome.html

by **Andreas Bus** » Sat, 19 Nov 1994 19:43:33

>Subject: Uh,oh P90/PCI seems like trouble (was Re: Kernel 1.1.54 woundn't compile on a Pentium PCI system)
>Date: 14 Nov 1994 12:30:35 -0500
>I have had the same problem ... after the internal compiler
>error, then more and more bad things happen, the disk gets
>corrputed ... this week I swap the mother board...actually,
>the 1st time this happened I thought it was the simms, they
>reported parity errors after I reboooted the machine BUT if
>I popped them out and exchanged them, rebooted, then put
>them back it all went away...this makes me think it is the
>mother board...we shall see...what is this about having
>wait states set right?
>Also, I cannot use >64M , if I do the pager crashes....
>Any ideas...is this a generic P90/PCI thing or is
>my configuration wrong/bad?
>Russ

Sorry for this rather sarcastic posting: You should better wait for Linux/MIPS.
Then you'll get *real* performance :-)

Andy

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

Waldorf Electronics GmbH, R&D Dep. | Phone: +49 2636-80294
Neustrasse 9-12, D-53498 Waldorf | Fax: +49 2636-80188
----------------------------------------------------------