Re: [gfortran] Fix PRs 14771, 20894, 25048 : keep track of parentheses

[Paul Thomas <paulthomas2 at wanadoo dot fr>]

Tobi,

> Built and tested on i686-pc-linux GNU, I'm seeing a regression with
> unf_io_convert_4.f90 with and without the patch, but other than that the
> testsuite run is clean.  ChangeLog below, patch and new testcases attached.  Ok
> for the trunk and 4.1?
>  
Many thanks for this one - I was contemplating how to do it just a few 
days ago and not really getting inspired!

I have regtested on FC3/Athlon and do not see any regressions, with this 
morning's tree.

OK for both trunk and 4.1.

It seems to me that you might widen the net a bit on parens_3.f90, to 
include the value of a pointer, as a counterpart to parens_1.f90, and an 
array.  I have attached a version of this test that adds both.

Cheers

Paul
! PR 14771
! { dg-do run }
! Originally contributed by Walt Brainerd, modified for the testsuite
      PROGRAM fc107

! Submitted by Walt Brainerd, The Fortran Company
! GNU Fortran 95 (GCC 4.1.0 20050322 (experimental))
! Windows XP

! Return value should be 3

      INTEGER I, J, M(2), N(2)
      integer, pointer :: k
      integer, target :: l
      INTEGER TRYME

      interface
        FUNCTION TRYyou(RTNME,HITME)
          INTEGER RTNME(2),HITME(2), tryyou(2)
        END function tryyou
      end interface

      m = 7
      l = 5
      I = 3
      k => l

      j = tryme((i),i)
      if (j .ne. 3) call abort ()

      j = tryme((k),k)
      if (j .ne. 5) call abort ()

      n = tryyou((m),m)
      if (any(n .ne. 7)) call abort ()
      END

      INTEGER FUNCTION TRYME(RTNME,HITME)
      INTEGER RTNME,HITME
      HITME = 999
      TRYME = RTNME
      END

      FUNCTION TRYyou(RTNME,HITME)
      INTEGER RTNME(2),HITME(2), tryyou(2)
      HITME = 999
      TRYyou = RTNME
      END