Re: fortran/9972: [3.4 regression] Failure of g77.f-torture/execute/980520-1.f

Re: fortran/9972: [3.4 regression] Failure of g77.f-torture/execute/980520-1.f

[dje]

Synopsis: [3.4 regression] Failure of g77.f-torture/execute/980520-1.f

Responsible-Changed-From-To: toon->steven
Responsible-Changed-By: dje
Responsible-Changed-When: Tue May 13 18:07:33 2003
Responsible-Changed-Why:
    related to PR 7257
State-Changed-From-To: open->analyzed
State-Changed-By: dje
State-Changed-When: Tue May 13 18:07:33 2003
State-Changed-Why:
    flag_inline_functions test for RTL deferral conflicts with Fortran assumption for unreferenced function elimination matching data symbol elimination.

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=9972

Re: fortran/9972: [3.4 regression] Failure of g77.f-torture/execute/980520-1.f

[toon]

Old Synopsis: [3.3/3.4 regression] Failure of g77.f-torture/execute/980520-1.f
New Synopsis: [3.4 regression] Failure of g77.f-torture/execute/980520-1.f

Responsible-Changed-From-To: unassigned->toon
Responsible-Changed-By: toon
Responsible-Changed-When: Fri Mar 14 10:39:12 2003
Responsible-Changed-Why:
    Fortran Maintainer.
    I changed the synopsis to match the actual situation after
    Jan Hubicka revoked his "Defer RTL" patch on the 3.3 branch.
    It will be high priority again as soon as 3.3 is out.

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=9972

Re: fortran/9972: [3.4 regression] Failure of g77.f-torture/execute/980520-1.f

[Billinghurst, David (CRTS)]

The following reply was made to PR fortran/9972; it has been noted by GNATS.

From: "Billinghurst, David (CRTS)" <David.Billinghurst@riotinto.com>
To: <gcc-gnats@gcc.gnu.org>,
	<gcc-bugs@gcc.gnu.org>,
	<toon@gcc.gnu.org>
Cc:  
Subject: Re: fortran/9972: [3.4 regression] Failure of g77.f-torture/execute/980520-1.f
Date: Tue, 11 Mar 2003 11:08:52 +1100

 http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=3Dview%20audit-trail&database=3D=
 gcc&pr=3D9972
 
 This regression has now been "ported" to the 3.3 branch
  - Sun Mar  9 11:07:10 GMT 2003 is OK
  - Mon Mar 10 11:07:45 GMT 2003 fails
 
 

fortran/9972: [3.4 regression] Failure of g77.f-torture/execute/980520-1.f

[David.Billinghurst]

>Number:         9972
>Category:       fortran
>Synopsis:       [3.4 regression] Failure of g77.f-torture/execute/980520-1.f
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    unassigned
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Thu Mar 06 01:46:00 UTC 2003
>Closed-Date:
>Last-Modified:
>Originator:     David Billinghurst
>Release:        gcc-3.4 20030305
>Organization:
>Environment:
mips-sgi-irix6.5
>Description:
Test g77.f-torture/execute/980520-1.f compile -O0 
 - passed Tue Mar 4 07:07:13 UTC 2003
 - fails  Wed Mar 5 07:07:04 UTC 2003

The failure is

ld32: ERROR 33: Unresolved text symbol "val_"

The (reformatted) test is:

Produced a link error through not eliminating the unused
statement function after 1998-05-15 change to gcc/toplev.c.  
Fixed by 1998-05-23 change to f/com.c.
values(i,j) = val((i-1)*n+j)
end


This new behaviour is not unreasonable of -O0, so we may prefer to mark it as XFAIL in the testsuite.
>How-To-Repeat:

>Fix:

>Release-Note:
>Audit-Trail:
>Unformatted: