[Bug c/37485] New: [graphite] Disconnecting exit edge in process of code generation

[Bug c/37485] New: [graphite] Disconnecting exit edge in process of code generation

[harsha dot jagasia at amd dot com]

In tranlate_clast when clast stmt is a stmt_user, we can end up disconnecting
the edge that is the exit_edge of the scop that is transformed. This state
creates problems because the exit_edge no longer has a destination. Hence when
after translate_clast, the exit_edge is redirected with redirect_edge_succ, the
edge is already disconnected.


-- 
           Summary:  [graphite] Disconnecting exit edge in process of code
                    generation
           Product: gcc
           Version: 4.4.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: harsha dot jagasia at amd dot com
 GCC build triplet: x86_64-unknown-linux-gnu
  GCC host triplet: x86_64-unknown-linux-gnu
GCC target triplet: x86_64-unknown-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37485

[Bug middle-end/37485] [graphite] Disconnecting exit edge in process of code generation

[harsha dot jagasia at amd dot com]

------- Comment #1 from harsha dot jagasia at amd dot com  2008-09-11 20:05 -------
Created an attachment (id=16298)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=16298&action=view)
Reduced test case for bug


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37485

[Bug middle-end/37485] [graphite] Disconnecting exit edge in process of code generation

[grosser at gcc dot gnu dot org]

------- Comment #2 from grosser at gcc dot gnu dot org  2008-09-12 12:55 -------
Created an attachment (id=16302)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=16302&action=view)
Just a simple loop over an array

I can confirm this one using this simple test case.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37485

[Bug middle-end/37485] [graphite] Disconnecting exit edge in process of code generation

[hjagasia at gcc dot gnu dot org]

-- 

hjagasia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |hjagasia at gcc dot gnu dot
                   |dot org                     |org
             Status|UNCONFIRMED                 |ASSIGNED
     Ever Confirmed|0                           |1
   Last reconfirmed|0000-00-00 00:00:00         |2008-09-12 15:42:40
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37485

[Bug middle-end/37485] [graphite] Disconnecting exit edge in process of code generation

[hjagasia at gcc dot gnu dot org]

------- Comment #3 from hjagasia at gcc dot gnu dot org  2008-09-12 15:44 -------
Jan and I have a bug fix for this, which will be posted in some time.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37485

[Bug middle-end/37485] [graphite] Disconnecting exit edge in process of code generation

[hjagasia at gcc dot gnu dot org]

------- Comment #4 from hjagasia at gcc dot gnu dot org  2008-09-12 16:52 -------
The reduced test case block-2.c triggers 3 seperate bugs.

First:
In tranlate_clast when clast stmt is a stmt_user, we can end up disconnecting
the edge that is the exit_edge of the scop that is transformed. This state
creates problems because the exit_edge no longer has a destination. Hence when
after translate_clast, the exit_edge is redirected with redirect_edge_succ, the
edge is already disconnected.

Second:
Graphite does not handle clast user stmts with constant arguments.

Third:
Graphite can create scops which overlap. This is an issue if one of the
overlapping scops is transformed which can cause some edges to be redirected.
When the successive overlapping scops are attempted to be transformed, the
basic blocks in the scop are no longer the same.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37485

[Bug middle-end/37485] [graphite] Disconnecting exit edge in process of code generation

[hjagasia at gcc dot gnu dot org]

------- Comment #5 from hjagasia at gcc dot gnu dot org  2008-09-12 16:54 -------
Created an attachment (id=16304)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=16304&action=view)
This patch fixes all three problems in the reduced test case block-2.c


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37485

[Bug middle-end/37485] [graphite] Disconnecting exit edge in process of code generation

[hjagasia at gcc dot gnu dot org]

------- Comment #6 from hjagasia at gcc dot gnu dot org  2008-09-12 16:59 -------
2008-09-05  Jan Sjodin  <jan.sjodin@amd.com>
            Harsha Jagasia  <harsha.jagasia@amd.com>

        * graphite.c (gmp_cst_to_tree): Moved.
        (iv_stack_entry_is_constant): New.
        (iv_stack_entry_is_iv): New.
        (loop_iv_stack_push): Renamed to loop_iv_stack_push_iv.
        (loop_iv_stack_insert_constant): New.
        (loop_iv_stack_pop): Use new datatpype.
        (loop_iv_stack_get_iv): Same.
        (loop_iv_stack_get_iv_from_name): Same.
        (loop_iv_stack_debug): Renamed to debug_loop_iv_stack.
        (loop_iv_stack_patch_for_consts): New.
        (loop_iv_stack_remove_constants): New.
        (graphite_create_new_loop): Use loop_iv_stack_push_iv.
        (translate_clast): Call loop_iv_stack_patch_for_consts and
        loop_iv_stack_remove_constants.
        (gloog): Use new datatype.  Redirect construction edge to end
        block to avoid accidental deletion.
        (limit_scops): Prevent overlapping scops.
        * graphite.h (enum iv_stack_entry_kind): New.  Tag for data in
        iv stack entry.
        (union iv_stack_entry_data): New.  Data in iv stack entry.
        (struct iv_stack_entry): New.  Datatype for iv stack entries.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37485

[Bug middle-end/37485] [graphite] Disconnecting exit edge in process of code generation

[rguenth at gcc dot gnu dot org]

------- Comment #7 from rguenth at gcc dot gnu dot org  2008-09-13 12:43 -------
Fixed?


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37485

[Bug middle-end/37485] [graphite] Disconnecting exit edge in process of code generation

[grosser at gcc dot gnu dot org]

------- Comment #8 from grosser at gcc dot gnu dot org  2008-09-29 13:21 -------
Commit 140746 should have fixed Bug 3. Can you confirm this?


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37485

[Bug middle-end/37485] [graphite] Disconnecting exit edge in process of code generation

[spop at gcc dot gnu dot org]

------- Comment #9 from spop at gcc dot gnu dot org  2008-09-29 14:50 -------
Subject: Re:  [graphite] Disconnecting exit edge in process of code generation

> Commit 140746 should have fixed Bug 3. Can you confirm this?

This is a side effect of your patch.  This bug is not yet fixed, and
the patch is not yet committed, as it triggered another bug in IVOPTS,
and then another bug in VRP.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37485

[Bug middle-end/37485] [graphite] Disconnecting exit edge in process of code generation

[spop at gcc dot gnu dot org]

------- Comment #10 from spop at gcc dot gnu dot org  2008-10-02 22:50 -------
Subject: Bug 37485

Author: spop
Date: Thu Oct  2 22:48:46 2008
New Revision: 140844

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=140844
Log:
2008-10-02  Jan Sjodin  <jan.sjodin@amd.com>
            Harsha Jagasia  <harsha.jagasia@amd.com>

        PR tree-optimization/37485
        * gcc.dg/graphite/block-2.c: New
        * graphite.c (gmp_cst_to_tree): Moved.
        (iv_stack_entry_is_constant): New.
        (iv_stack_entry_is_iv): New.
        (loop_iv_stack_push): Renamed to loop_iv_stack_push_iv.
        (loop_iv_stack_insert_constant): New.
        (loop_iv_stack_pop): Use new datatpype.
        (loop_iv_stack_get_iv): Same.
        (loop_iv_stack_get_iv_from_name): Same.
        (loop_iv_stack_debug): Renamed to debug_loop_iv_stack.
        (loop_iv_stack_patch_for_consts): New.
        (loop_iv_stack_remove_constants): New.
        (graphite_create_new_loop): Use loop_iv_stack_push_iv.
        (translate_clast): Call loop_iv_stack_patch_for_consts and
        loop_iv_stack_remove_constants.
        (gloog): Use new datatype.  Redirect construction edge to end
        block to avoid accidental deletion.
        * graphite.h (enum iv_stack_entry_kind): New.  Tag for data in
        iv stack entry.
        (union iv_stack_entry_data): New.  Data in iv stack entry.
        (struct iv_stack_entry): New.  Datatype for iv stack entries.


Added:
    branches/graphite/gcc/testsuite/gcc.dg/graphite/block-2.c
Modified:
    branches/graphite/gcc/ChangeLog.graphite
    branches/graphite/gcc/graphite.c
    branches/graphite/gcc/graphite.h


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37485

[Bug middle-end/37485] [graphite] Disconnecting exit edge in process of code generation

[spop at gcc dot gnu dot org]

------- Comment #11 from spop at gcc dot gnu dot org  2008-10-03 20:28 -------
*** Bug 37587 has been marked as a duplicate of this bug. ***


-- 

spop at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
OtherBugsDependingO|                            |37587
              nThis|                            |
                 CC|                            |linuxl4 at sohu dot com


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37485

[Bug middle-end/37485] [graphite] Disconnecting exit edge in process of code generation

[grosser at gcc dot gnu dot org]

------- Comment #12 from grosser at gcc dot gnu dot org  2008-10-10 19:40 -------
(In reply to comment #9)
> Subject: Re:  [graphite] Disconnecting exit edge in process of code generation
> 
> > Commit 140746 should have fixed Bug 3. Can you confirm this?
> 
> This is a side effect of your patch.  

I am sorry. With "Bug 3" I ment:

Third:
Graphite can create scops which overlap. This is an issue if one of the
overlapping scops is transformed which can cause some edges to be redirected.
When the successive overlapping scops are attempted to be transformed, the
basic blocks in the scop are no longer the same.

My commit fixed a bug, that created overlapping scops, and inserted a check for
overlapping scops. Are there any overlapping scops with the current code?


-- 

grosser at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |grosser at gcc dot gnu dot
                   |                            |org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37485

[Bug middle-end/37485] [graphite] Disconnecting exit edge in process of code generation

[spop at gcc dot gnu dot org]

------- Comment #13 from spop at gcc dot gnu dot org  2008-10-11 19:26 -------
Subject: Re:  [graphite] Disconnecting exit edge in process of code generation

>Are there any overlapping scops with the current code?
I haven't seen overlapping scops anymore.  Thanks.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37485

[Bug middle-end/37485] [graphite] Disconnecting exit edge in process of code generation

[spop at gcc dot gnu dot org]

------- Comment #14 from spop at gcc dot gnu dot org  2008-10-15 16:25 -------
Subject: Bug 37485

Author: spop
Date: Wed Oct 15 16:24:15 2008
New Revision: 141143

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=141143
Log:
2008-10-15  Jan Sjodin  <jan.sjodin@amd.com>
            Harsha Jagasia  <harsha.jagasia@amd.com>

        PR tree-optimization/37485
        * graphite.c (gmp_cst_to_tree): Moved.
        (iv_stack_entry_is_constant): New.
        (iv_stack_entry_is_iv): New.
        (loop_iv_stack_push): Renamed to loop_iv_stack_push_iv.
        (loop_iv_stack_insert_constant): New.
        (loop_iv_stack_pop): Use new datatpype.
        (loop_iv_stack_get_iv): Same.
        (loop_iv_stack_get_iv_from_name): Same.
        (loop_iv_stack_debug): Renamed to debug_loop_iv_stack.
        (loop_iv_stack_patch_for_consts): New.
        (loop_iv_stack_remove_constants): New.
        (graphite_create_new_loop): Use loop_iv_stack_push_iv.
        (translate_clast): Call loop_iv_stack_patch_for_consts and
        loop_iv_stack_remove_constants.
        (gloog): Use new datatype.  Redirect construction edge to end
        block to avoid accidental deletion.
        * graphite.h (enum iv_stack_entry_kind): New.  Tag for data in
        iv stack entry.
        (union iv_stack_entry_data): New.  Data in iv stack entry.
        (struct iv_stack_entry): New.  Datatype for iv stack entries.

2008-10-15  Sebastian Pop  <sebastian.pop@amd.com>

        * tree-data-ref.c (stmt_simple_memref_p): Fix memleak.
        * graphite.c (free_scop): Same.
        (build_scops): Same.
        (free_scattering): New.
        (build_cloog_prog): Call free_scattering.
        (patch_phis_for_virtual_defs): Fix memleak.
        (graphite_trans_bb_strip_mine): Same.
        (limit_scops): Same.


Added:
    trunk/gcc/testsuite/gcc.dg/graphite/pr37485.c
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/graphite.c
    trunk/gcc/graphite.h
    trunk/gcc/testsuite/ChangeLog
    trunk/gcc/tree-data-ref.c


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37485

[Bug middle-end/37485] [graphite] Disconnecting exit edge in process of code generation

[spop at gcc dot gnu dot org]

------- Comment #15 from spop at gcc dot gnu dot org  2008-10-15 16:27 -------
Fixed.


-- 

spop at gcc dot gnu dot org changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37485