[Bug c++/58788] OMP linear clause parsing problem

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58788

--- Comment #1 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
2.8.1 has:
linear(list[:linear-step])
and appendix b has:
  linear ( variable-list : expression )
variable-list:
  id-expression
  variable-list , id-expression
so I linear(j : 4, k) is IMNSHO equivalent to linear(j : (4, k)), not to
linear(j: 4) linear(k : 1)
Even compound expression is an expression, if the standard wanted to allow
anything else, it would need to use very different wording.
If you want to specify linear(j:4) linear(k), just write it that way (one can
group variables with the same step together of course).