fatal signal 11

[lawson_whitney at juno.com]

On Sat, 10 Mar 2001 cfk at pacbell.net wrote:

> I downloaded the latest wine from the cvs site and ran its
> ./tools/wineinstall. It ran for quite a while and croaked on
> /wine/server/semaphore.c. Basically, gcc says "Internal compiler error:
> program cc1 bog fatal signal 11". The compiler is 2.91.66 (aka egcs-1.1.2
> release) and compiles other programs including 2.4.1 & 2.4.2 Linux kernels
> just fine. I am puzzled what 'fatal signal 11' may mean. Can anyone offer a
> suggestion for a course of  investigation.
>
I don't like egcs; as soon as slackware 7.0 installed it I replaced it
witn gcc-2.95.1 from contrrib, which I have since patched to 2.95.2, and
haven't had any trouble.  Just applied the last of the patches to get to
the current CVS and ran make, so if you get this, I still don't have any
trouble (I use wine to send mail).

A signal 11 is _either_ a program bug or flaky hardware.  I have a p-120
that throws sig 11's when the humidity gets extremely high; running the
air conditioner for a few minutes fixes it.  My junk-pentium 66 is
unaffected by humidity, but a bit slower.

If it happens consistently in the same spot, it is probably a program
bug.  gcc-2.95.2 has the strength-reduce bug, but Wine knows how to cope
with that.  Wine build complete.  Nothing in the current CVS provokes
gcc-2.95.2 to segment fault.

If you want to look into hardware-generated segment faults,

http://www.bitwizard.nl/sig11/

has a lot of good advice.

As youcan see by "kill -l" signal 11 is SIGSEGV - a program tried to
operate on memory that isn't there.  The X86 is a segmented machine, so
it detects that as a segment fault.

Lawson

panic: detected rogue system administrator!!
---cut here


________________________________________________________________
GET INTERNET ACCESS FROM JUNO!
Juno offers FREE or PREMIUM Internet access for less!
Join Juno today!  For your FREE software, visit:
http://dl.www.juno.com/get/tagj.