njh@kernighan.demon.co.uk said:
> I don't know about scilab, but when I've had this sort of problem in
> the past I've generally tackled it by manually compiling the
> particular .c file in question without optimization, then going back
> up to the top level of the port and restarting the build.

Which isn't very helpful for tracking down the underlying problem in the 
compiler.

Please ensure that all bugs are reported, then we stand a fighting chance 
of stamping them out for everybody.

R.