[Bug fortran/40413] New: Internal error in connection with optimization and allocatable objects

[Bug fortran/40413] New: Internal error in connection with optimization and allocatable objects

[juergen dot reuter at desy dot de]

The code below does compile with gfortran 4.5.0 without any specification or
with -O0, but generates an internal error when compiling with -O1, -O2, or -O3:
state_matrices.f90: In function 'polarization_copy':
state_matrices.f90:39:0: internal compiler error: in
load_assign_lhs_subreplacements, at tree-sra.c:1954
Please submit a full bug report,
with preprocessed source if appropriate.

The code compiles when using the version in comments, i.e. the usage of the
dummy variable state_dummy. The internal error is avoided also when commenting
out the line 38 with the state_matrix_copy function call. Removing the
multiplicity or the spin_type components also makes the code compile with
optimization. 


module state_matrices

  implicit none
  private

  public :: state_matrix_copy
  public :: state_matrix_t
  public :: matrix_element_t

  type :: matrix_element_t
     private
     integer, dimension(:), allocatable :: f
  end type matrix_element_t

  type :: state_matrix_t
     private
     type(matrix_element_t), dimension(:), allocatable :: me
  end type state_matrix_t

  type :: polarization_t
     logical :: polarized = .false.
     integer :: spin_type = 0
     integer :: multiplicity = 0
     type(state_matrix_t) :: state
  end type polarization_t

contains

  function polarization_copy (pol_in) result (pol)
    type(polarization_t) :: pol
    type(polarization_t), intent(in) :: pol_in
    !!! type(state_matrix_t) :: state_dummy
    pol%polarized = pol_in%polarized
    pol%spin_type = pol_in%spin_type
    pol%multiplicity = pol_in%multiplicity
    !!! state_dummy = state_matrix_copy (pol_in%state)
    !!! pol%state = state_dummy
    pol%state = state_matrix_copy (pol_in%state)
  end function polarization_copy

  function state_matrix_copy (state_in) result (state)
    type(state_matrix_t) :: state
    type(state_matrix_t), intent(in), target :: state_in
  end function state_matrix_copy

end module state_matrices


-- 
           Summary: Internal error in connection with optimization and
                    allocatable objects
           Product: gcc
           Version: 4.5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: fortran
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: juergen dot reuter at desy dot de
 GCC build triplet: Build version svn v148346
  GCC host triplet: MAC OS X Darwin and Linux x86


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40413

[Bug fortran/40413] Internal error in connection with optimization and allocatable objects

[dominiq at lps dot ens dot fr]

------- Comment #1 from dominiq at lps dot ens dot fr  2009-06-11 10:34 -------
Confirmed on i686-apple-darwin9. The ICE disappears with -m64 and it is a
regression: the code compiles with 4.4.0 and 4.4.1 without warning and with
4.2.3 and 4.3.3 with:

pr40413.f90: In function 'state_matrix_copy':
pr40413.f90:29: warning: Function return value not set


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40413

[Bug fortran/40413] Internal error in connection with optimization and allocatable objects

[reuter at mail dot desy dot de]

------- Comment #2 from reuter at mail dot desy dot de  2009-06-11 14:54 -------
Subject: Re:  Internal error in connection with optimization
 and allocatable objects

On Thu, 11 Jun 2009, dominiq at lps dot ens dot fr wrote:

Salut, Dominiq!
I'm not so much of an export in the terminology. What exactly does 
regression mean? The code compiled even with 4.5.0 and optimization up to
svn around v148080. The option -m64 does work. What exactly does it do?
Thanks in advance for you help!
Je suis desole pour non parler Francais! 
Cheers,
        Juergen

>
>
> ------- Comment #1 from dominiq at lps dot ens dot fr  2009-06-11 10:34 -------
> Confirmed on i686-apple-darwin9. The ICE disappears with -m64 and it is a
> regression: the code compiles with 4.4.0 and 4.4.1 without warning and with
> 4.2.3 and 4.3.3 with:
>
> pr40413.f90: In function 'state_matrix_copy':
> pr40413.f90:29: warning: Function return value not set
>
>
> -- 
>
>
> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40413
>
> ------- You are receiving this mail because: -------
> You reported the bug, or are watching the reporter.
>

------------------------------------
Juergen Reuter
Albrecht-Ludwigs-Universitaet Freiburg
Physikalisches Institut
Hermann-Herder-Str. 3
79104 Freiburg, GERMANY
Tel.: +49-761-203-8407
Skype: jr_reuter
------------------------------------


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40413

[Bug fortran/40413] Internal error in connection with optimization and allocatable objects

[dominiq at lps dot ens dot fr]

------- Comment #3 from dominiq at lps dot ens dot fr  2009-06-11 15:23 -------
> I'm not so much of an export in the terminology. What exactly does 
> regression mean?

It means that a code worked at revision n (here 4.4.0) and no longer at
revision n+1 (here 4.5.0). This means that either something has been broken by
some change or a bug has been exposed by it. Since changes are supposed to
improve any piece of code, regressions have in general a higher priority than
other bugs, so it is important to mark them in the subject line (look at
http://gcc.gnu.org/ml/gcc-bugs/2009-06/ for examples). Since I don't have the
rights to change the subject line, you have to do it yourself or wait for some
maintainer to do it.

> The option -m64 does work. What exactly does it do?

It forces the compilation in "64 bit mode" while the default in Darwin is 32
bit. I don't know the detail of the differences besides the fact that the
addresses are coded on 64 bits instead of 32 and in general the code runs
faster (~10 to 20%) with -m64.

Note that I provided the information in the hope to help those who will fix the
bug, meanwhile you can use -m64 as a workaround.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40413

[Bug fortran/40413] Internal error in connection with optimization and allocatable objects

[reuter at mail dot desy dot de]

------- Comment #4 from reuter at mail dot desy dot de  2009-06-11 19:31 -------
Subject: Re:  Internal error in connection with optimization
 and allocatable objects

On Thu, 11 Jun 2009, dominiq at lps dot ens dot fr wrote:


Thanks for the explanations. I actually cannot change the priority,
only the assignee or a specifically empowered user can change the status.

>
>
> ------- Comment #3 from dominiq at lps dot ens dot fr  2009-06-11 15:23 -------
>> I'm not so much of an export in the terminology. What exactly does
>> regression mean?
>
> It means that a code worked at revision n (here 4.4.0) and no longer at
> revision n+1 (here 4.5.0). This means that either something has been broken by
> some change or a bug has been exposed by it. Since changes are supposed to
> improve any piece of code, regressions have in general a higher priority than
> other bugs, so it is important to mark them in the subject line (look at
> http://gcc.gnu.org/ml/gcc-bugs/2009-06/ for examples). Since I don't have the
> rights to change the subject line, you have to do it yourself or wait for some
> maintainer to do it.
>
>> The option -m64 does work. What exactly does it do?
>
> It forces the compilation in "64 bit mode" while the default in Darwin is 32
> bit. I don't know the detail of the differences besides the fact that the
> addresses are coded on 64 bits instead of 32 and in general the code runs
> faster (~10 to 20%) with -m64.
>
> Note that I provided the information in the hope to help those who will fix the
> bug, meanwhile you can use -m64 as a workaround.
>
>
> -- 
>
>
> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40413
>
> ------- You are receiving this mail because: -------
> You reported the bug, or are watching the reporter.
>

------------------------------------
Juergen Reuter
Albrecht-Ludwigs-Universitaet Freiburg
Physikalisches Institut
Hermann-Herder-Str. 3
79104 Freiburg, GERMANY
Tel.: +49-761-203-8407
Skype: jr_reuter
------------------------------------


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40413

[Bug fortran/40413] Internal error in connection with optimization and allocatable objects

[dominiq at lps dot ens dot fr]

------- Comment #5 from dominiq at lps dot ens dot fr  2009-06-11 21:47 -------
The test compiles with 4.5.0 revision 147438.

> I actually cannot change the priority

You don't have to change the priority, just add "[4.5 Regression]" at the
beginning of the subject line.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40413

[Bug fortran/40413] [4.5 Regression] Internal error in connection with optimization and allocatable objects

[pault at gcc dot gnu dot org]

------- Comment #6 from pault at gcc dot gnu dot org  2009-06-12 05:20 -------
Confirmed on FC8/x86 running in a VMware on an XP professional machine.... if
you get that:-)

Anyway, it's not limited to Darwin nor to 64 bits.  It's not present, as far as
I can tell with gcc-4.4 20080624; annoyingly, the code produced looks no
different.

Cheers

Paul


-- 

pault at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|0                           |1
   Last reconfirmed|0000-00-00 00:00:00         |2009-06-12 05:20:52
               date|                            |
            Summary|Internal error in connection|[4.5 Regression] Internal
                   |with optimization and       |error in connection with
                   |allocatable objects         |optimization and allocatable
                   |                            |objects


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40413

[Bug tree-optimization/40413] [4.5 Regression] Internal error in connection with optimization and allocatable objects

[rguenth at gcc dot gnu dot org]

-- 

rguenth at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mjambor at suse dot cz
          Component|fortran                     |tree-optimization
   Target Milestone|---                         |4.5.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40413

[Bug tree-optimization/40413] [4.5 Regression] Internal error in connection with optimization and allocatable objects

[pault at gcc dot gnu dot org]

------- Comment #7 from pault at gcc dot gnu dot org  2009-06-13 15:52 -------
(In reply to comment #6)

Changing
  type :: polarization_t
     logical :: polarized = .false.
     integer :: spin_type = 0
     integer :: multiplicity = 0
     type(state_matrix_t) :: state
  end type polarization_t

to
  type :: polarization_t
     type(state_matrix_t) :: state ! put structure here
     logical :: polarized = .false.
     integer :: spin_type = 0
     integer :: multiplicity = 0
  end type polarization_t

compiles at -O3 for me.

Paul


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40413

[Bug tree-optimization/40413] [4.5 Regression] Internal error in connection with optimization and allocatable objects

[jamborm at gcc dot gnu dot org]

------- Comment #8 from jamborm at gcc dot gnu dot org  2009-06-14 01:30 -------
Mine, will look into it shortly.


-- 

jamborm at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |jamborm at gcc dot gnu dot
                   |dot org                     |org
             Status|NEW                         |ASSIGNED
   Last reconfirmed|2009-06-12 05:20:52         |2009-06-14 01:30:09
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40413

[Bug tree-optimization/40413] [4.5 Regression] Internal error in connection with optimization and allocatable objects

[jamborm at gcc dot gnu dot org]

------- Comment #9 from jamborm at gcc dot gnu dot org  2009-06-15 09:07 -------
Created an attachment (id=18001)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18001&action=view)
Fix

This was quite a serious oversight on my part, I wonder how it went
for so long unnoticed.  I am about to bootstrap and regression test
this patch and will submit it if both succeed.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40413

[Bug tree-optimization/40413] [4.5 Regression] Internal error in connection with optimization and allocatable objects

[jamborm at gcc dot gnu dot org]

------- Comment #10 from jamborm at gcc dot gnu dot org  2009-06-16 09:54 -------
Bootstrap and testing passed, submitted in
http://gcc.gnu.org/ml/gcc-patches/2009-06/msg01179.html


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40413

[Bug tree-optimization/40413] [4.5 Regression] Internal error in connection with optimization and allocatable objects

[jamborm at gcc dot gnu dot org]

------- Comment #11 from jamborm at gcc dot gnu dot org  2009-06-16 10:12 -------
Subject: Bug 40413

Author: jamborm
Date: Tue Jun 16 10:11:55 2009
New Revision: 148520

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=148520
Log:
2009-06-16  Martin Jambor  <mjambor@suse.cz>

        PR tree-optimization/40413
        * tree-sra.c (load_assign_lhs_subreplacements): Pass offset to
        build_ref_for_offset.
        (propagate_subacesses_accross_link): Fix a typo in a comment.

        * testsuite/gfortran.fortran-torture/compile/pr40413.f90: New file.


Added:
    trunk/gcc/testsuite/gfortran.fortran-torture/compile/pr40413.f90
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/testsuite/ChangeLog
    trunk/gcc/tree-sra.c


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40413

[Bug tree-optimization/40413] [4.5 Regression] Internal error in connection with optimization and allocatable objects

[jamborm at gcc dot gnu dot org]

------- Comment #12 from jamborm at gcc dot gnu dot org  2009-06-16 10:24 -------
Fixed


-- 

jamborm at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40413