On Thu, Sep 11, 2008 at 07:05:03PM +0200, Roland Mainz wrote:
> "Dr. Werner Fink" wrote:
> > this is what happens:
> >
> > test arith.sh begins at 2008-09-11+15:57:55
> > arith.sh[418]: compound var arithmetic failed
> > arith.sh[419]: compound var arithmetic failed
> > arith.sh[420]: compound var arithmetic failed
> >
> > and if done by hand the
> >
> > d127:/usr/src/packages/BUILD/ksh93 # chmod +x /tmp/script$$
> > d127:/usr/src/packages/BUILD/ksh93 # /tmp/script$$ 1
> > ( bar=2 baz=3 foo=1 )
> > d127:/usr/src/packages/BUILD/ksh93 # /tmp/script$$ 2
> > ( faz=0 )
> > d127:/usr/src/packages/BUILD/ksh93 # /tmp/script$$ 3
> > Memory fault
> > d127:/usr/src/packages/BUILD/ksh93 # /tmp/script$$ 4
> > Memory fault
> > d127:/usr/src/packages/BUILD/ksh93 # /tmp/script$$ 5
> > Memory fault
>
> Do you have a stack trace from the code dump or the live ksh93 process ?
The fresh buld ksh is from within the build directory:
./arch/linux.i386-64/src/cmd/ksh93/ksh
and the script is simply created by copy&paste from arith.sh
to the command line. Note that this script does not have
a she bang line that is that the running ksh simply forks
of a child which catch the SIGSEGV in its own signal handler
and returns 0176. I can attach the gdb to the ksh and use
set follow-fork-mode child
in gdb to see this, but due to the signal handler on SIGSEGV
I've currently nothing to debug.
Werner
--
"Having a smoking section in a restaurant is like having
a peeing section in a swimming pool." -- Edward Burr
_______________________________________________
ast-users mailing list
[email protected]
https://mailman.research.att.com/mailman/listinfo/ast-users
