Push the calculation of the tuple's kernel part down to kernel's build scripts:
- update the kernel script's API with the function CT_DiKernelTupleValues
- update doc accordingly (also with the architecture change, missing in the previous commit)
- small clean-up in the main script, remove useless test

/trunk/scripts/build/kernel/bare-metal.sh | 5 5 0 0 +++++
/trunk/scripts/build/kernel/linux.sh | 5 5 0 0 +++++
/trunk/scripts/crosstool.sh | 31 13 18 0 +++++++++++++------------------
/trunk/scripts/functions | 15 8 7 0 ++++++++-------
/trunk/docs/overview.txt | 13 9 4 0 +++++++++----
5 files changed, 40 insertions(+), 29 deletions(-)

Original patch from H.J. Lu, as reported there:

http://www.nabble.com/PATCH:-Use-short-for-fnstsw-td14775171.html

fnstsw stores 16bit into %ax. The upper 16bit of %eax is unchanged.

The new assembler (binutils-2.18.50.0.4 and up) will disallow "fnstsw %eax".

diff -dur glibc-2.7.orig/sysdeps/i386/fpu/ftestexcept.c glibc-2.7/sysdeps/i386/fpu/ftestexcept.c

--- glibc-2.7.orig/sysdeps/i386/fpu/ftestexcept.c	2004-03-05 11:14:48.000000000 +0100

+++ glibc-2.7/sysdeps/i386/fpu/ftestexcept.c	2008-06-24 10:54:43.000000000 +0200

@@ -26,7 +26,7 @@

 int

 fetestexcept (int excepts)

 {

-  int temp;

+  short temp;

   int xtemp = 0;

   /* Get current exceptions.  */