New c-torture execute testcase

[Jeffrey A Law]

  In message <199710241319.OAA22049@phal.cygnus.co.uk>you write:
  > This is modified from 931004-12.c, but by using a struct of size 5,
  > a lot of different code is executed for typical 32 bit targets -
  > I was looking for the reason why the MUST_PASS_IN_STACK problem didn't
  > bite me on the SH as on the d10v, and found out that it was only a matter
  > of the structure being larger than PARM_BOUNDARY.
I've added this to the egcs testsuite as strct-stdarg-1.c.  We can rename
as necessary if it's included under a different name in the next c-torture
release.

jeff