[Bug c/30375] New: tree-ssa-dse incorrectly removes struct initialization

[Bug c/30375] New: tree-ssa-dse incorrectly removes struct initialization

[marcus at jet dot franken dot de]

A wine testcase started failing with the trunk compiler some days ago.

The difference is pretty obivous in the tree dump:
diff -u encode.c.056t.dce3 encode.c.057t.dse1 
--- encode.c.056t.dce3  2007-01-04 23:03:14.000000000 +0100
+++ encode.c.057t.dse1  2007-01-04 23:03:14.000000000 +0100
@@ -52,7 +52,6 @@
   int iftmp.507;

 <bb 2>:
-  info = {};
   point = {};
   entry = {};
   buf = 0B;


However, if you look at it, info is clearly used right below.

I will attach .i, and the two dump files.


-- 
           Summary: tree-ssa-dse incorrectly removes struct initialization
           Product: gcc
           Version: 4.3.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: marcus at jet dot franken dot de
 GCC build triplet: i686-pc-linux-gnu
  GCC host triplet: i686-pc-linux-gnu
GCC target triplet: i686-pc-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug c/30375] tree-ssa-dse incorrectly removes struct initialization

[marcus at jet dot franken dot de]

------- Comment #2 from marcus at jet dot franken dot de  2007-01-04 22:08 -------
Created an attachment (id=12857)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=12857&action=view)
encode.c.056t.dce3


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug c/30375] tree-ssa-dse incorrectly removes struct initialization

[marcus at jet dot franken dot de]

------- Comment #1 from marcus at jet dot franken dot de  2007-01-04 22:08 -------
Created an attachment (id=12856)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=12856&action=view)
encode.i.bz2


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug c/30375] tree-ssa-dse incorrectly removes struct initialization

[marcus at jet dot franken dot de]

------- Comment #3 from marcus at jet dot franken dot de  2007-01-04 22:09 -------
Created an attachment (id=12858)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=12858&action=view)
encode.c.057t.dse1


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug c/30375] tree-ssa-dse incorrectly removes struct initialization

[marcus at jet dot franken dot de]

------- Comment #4 from marcus at jet dot franken dot de  2007-01-04 22:10 -------
generated by this cmd:

/home/marcus/projects/gcc/BIN/bin/gcc -m32 -c -I. -I. -I../../../include
-I../../../include   -D_REENTRANT -fPIC -Wall -pipe -fno-strict-aliasing
-Wdeclaration-after-statement -Wwrite-strings -Wpointer-arith  -g  -O2 -Wall -W
-D_FORTIFY_SOURCE=2  -o encode.o encode.c --save-temps -fverbose-asm
-fdump-rtl-all -fdump-tree-all


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[pinskia at gcc dot gnu dot org]

------- Comment #5 from pinskia at gcc dot gnu dot org  2007-01-04 22:21 -------
I think this was exposed/caused by the mem-ssa and the specific DSE
optimization for aggregate DSE:
2006-12-11  Aldy Hernandez  <aldyh@redhat.com>

        * tree-ssa-dse.c (aggregate_vardecl_d): New.
        (dse_global_data): Add aggregate_vardecl field.
        (dse_possible_dead_store_p): New.
        Add prev_defvar variable.
        Allow immediate uses and previous immediate uses to differ
        if they are setting different parts of the whole.
        (get_aggregate_vardecl): New.
        (dse_record_partial_aggregate_store): New.
        (dse_whole_aggregate_clobbered_p): New.
        (dse_partial_kill_p): New.
        (dse_optimize_stmt): Abstract code checking a possible dead store
        into new function dse_possible_dead_store_p().
        Call dse_maybe_record_aggregate_store().
        When checking whether a STMT and its USE_STMT refer to the
        same memory address, check also for partial kills that clobber
        the whole.
        Move some variable definitions to the block where they are used.
        (aggregate_vardecl_hash): New.
        (aggregate_vardecl_eq): New.
        (aggregate_vardecl_free): New.
        (aggregate_whole_store_p): New.
        (tree_ssa_dse): Initialize and free aggregate_vardecl.
        Mark which aggregate stores we care about.


I think we have:
  info_1 = V_DEF<info_0(D)>
  info = {};
...
  info_2 = V_DEF<info_1(D)>
  winetest_set_location (&"encode.c"[0], 2886);

Which thinks info is fully clobbered by winetest_set_location which is not
true.


-- 

pinskia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |aldyh at gcc dot gnu dot
                   |                            |org, dnovillo at gcc dot gnu
                   |                            |dot org
          Component|c                           |tree-optimization
  GCC build triplet|i686-pc-linux-gnu           |
   GCC host triplet|i686-pc-linux-gnu           |
 GCC target triplet|i686-pc-linux-gnu           |
           Keywords|                            |wrong-code
            Summary|tree-ssa-dse incorrectly    |[4.3 Regression] tree-ssa-
                   |removes struct              |dse incorrectly removes
                   |initialization              |struct initialization
   Target Milestone|---                         |4.3.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[pinskia at gcc dot gnu dot org]

------- Comment #6 from pinskia at gcc dot gnu dot org  2007-01-05 05:29 -------
I am in correct in saying the problem is related to the call but instead it is
related to:
  # MPT.775_180 = VDEF <MPT.775_179(D)>
  info = {};
  # MPT.775_181 = VDEF <MPT.775_180>
  point = {};


They have the same memory tags which causes the new part of DSE to think the
first store of info to be fully redundant which is wrong.


-- 

pinskia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |blocker


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[steven at gcc dot gnu dot org]

------- Comment #7 from steven at gcc dot gnu dot org  2007-02-03 15:28 -------
Is this now being looked into by Diego or Aldy?


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[steven at gcc dot gnu dot org]

-- 

steven 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         |2007-02-03 15:28:16
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[dnovillo at gcc dot gnu dot org]

------- Comment #8 from dnovillo at gcc dot gnu dot org  2007-02-03 15:49 -------
(In reply to comment #7)
> Is this now being looked into by Diego or Aldy?
> 
It wasn't.  It is now.


-- 

dnovillo at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |dnovillo at gcc dot gnu dot
                   |dot org                     |org
             Status|NEW                         |ASSIGNED
   Last reconfirmed|2007-02-03 15:28:16         |2007-02-03 15:49:29
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[aldyh at gcc dot gnu dot org]

------- Comment #9 from aldyh at gcc dot gnu dot org  2007-03-09 20:05 -------
Subject: Bug 30375

Author: aldyh
Date: Fri Mar  9 20:05:08 2007
New Revision: 122758

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=122758
Log:
        PR tree-optimization/30375
        * tree-ssa-dse.c (dse_possible_dead_store_p): Do not eliminate if
        LHS of statements is not the same.
        * testsuite/gcc.dg/tree-ssa/ssa-dse-10.c: New.

Added:
    trunk/gcc/testsuite/gcc.dg/tree-ssa/ssa-dse-10.c
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/tree-ssa-dse.c


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[pinskia at gcc dot gnu dot org]

------- Comment #10 from pinskia at gcc dot gnu dot org  2007-03-09 23:30 -------
Fixed.


-- 

pinskia at gcc dot gnu dot org changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[marcus at jet dot franken dot de]

------- Comment #11 from marcus at jet dot franken dot de  2007-07-24 20:29 -------
has resurfaced in gcc-Version 4.3.0 20070724 (experimental)
but in a different file.


-- 

marcus at jet dot franken dot de changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[marcus at jet dot franken dot de]

------- Comment #13 from marcus at jet dot franken dot de  2007-07-24 20:31 -------
Created an attachment (id=13968)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=13968&action=view)
msg.c.074t.dce4.bz2


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[marcus at jet dot franken dot de]

------- Comment #12 from marcus at jet dot franken dot de  2007-07-24 20:31 -------
Created an attachment (id=13967)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=13967&action=view)
msg.i.bz2

the .i.bz2 file


-- 

marcus at jet dot franken dot de changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
  Attachment #12856|0                           |1
        is obsolete|                            |
  Attachment #12857|0                           |1
        is obsolete|                            |
  Attachment #12858|0                           |1
        is obsolete|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[marcus at jet dot franken dot de]

------- Comment #15 from marcus at jet dot franken dot de  2007-07-24 20:32 -------
diff -u msg.c.074t.dce4 msg.c.075t.dse1 |less
--- msg.c.074t.dce4     2007-07-24 22:23:47.000000000 +0200
+++ msg.c.075t.dse1     2007-07-24 22:23:47.000000000 +0200
@@ -2760,7 +2760,6 @@
   unsigned int D.12275;

 <bb 2>:
-  signInfo = {};
   signInfo.cbSize = 28;
   signer = {};
   signer.cbSize = 48;


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[marcus at jet dot franken dot de]

------- Comment #14 from marcus at jet dot franken dot de  2007-07-24 20:31 -------
Created an attachment (id=13969)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=13969&action=view)
msg.c.075t.dse1.bz2


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[marcus at jet dot franken dot de]

------- Comment #16 from marcus at jet dot franken dot de  2007-07-25 05:10 -------
Created an attachment (id=13970)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=13970&action=view)
pr30375.c

I created a testcase for later addition to gcc.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[pinskia at gcc dot gnu dot org]

------- Comment #17 from pinskia at gcc dot gnu dot org  2007-08-06 05:52 -------
Reconfirmed.


-- 

pinskia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|REOPENED                    |ASSIGNED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[mmitchel at gcc dot gnu dot org]

-- 

mmitchel at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P1


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[rguenth at gcc dot gnu dot org]

------- Comment #18 from rguenth at gcc dot gnu dot org  2007-08-22 20:23 -------
Note alias info is correct:

  # signInfo_2 = VDEF <signInfo_1(D)>
  signInfo = {};
  # signInfo_3 = VDEF <signInfo_2>
  signInfo.a = 16;
  # signInfo_4 = VDEF <signInfo_3>
  signInfo.b = 1;
  # signInfo_5 = VDEF <signInfo_4>
  signInfo.c = 0;

but somehow DSE figures that the a/b/c inits do a def killing the initial one.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[rguenth at gcc dot gnu dot org]

------- Comment #19 from rguenth at gcc dot gnu dot org  2007-08-22 20:25 -------
Or more like, the use on the call is missed.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[rguenth at gcc dot gnu dot org]

------- Comment #20 from rguenth at gcc dot gnu dot org  2007-08-31 10:50 -------
Works now but still broken with --param max-aliased-vops=0.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[rguenth at gcc dot gnu dot org]

------- Comment #21 from rguenth at gcc dot gnu dot org  2007-08-31 10:59 -------
Runtime testcase that fails with --param max-aliased-vops=0:

typedef struct _s {
    int a;
    int b;
    int c;
    int d;
} s;

extern void abort(void);

void __attribute__((noinline)) g(s *p)
{
  if (p->d != 0)
    abort ();
}

char *c = (void*)0;
void __attribute__((noinline)) f(void) { if (c) *c = 1; }

void test_signed_msg_encoding(void)
{
    s signInfo = { sizeof(signInfo), 0 };

    signInfo.b = 1;
    signInfo.c = 0;
    g(&signInfo);
    signInfo.d = 1;
    f();
}

int main()
{
  test_signed_msg_encoding ();
  test_signed_msg_encoding ();
  return 0;
}


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[rguenth at gcc dot gnu dot org]

------- Comment #22 from rguenth at gcc dot gnu dot org  2007-08-31 13:56 -------
We are clearly missing the use of signInfo.d by the call, thus we should
from the bitmap of clobbered parts of signInfo remove the used parts once we
encounter the call.  But somehow we never remove bits from this bitmap and
I have no clue how this should work with other intermediate uses, but it
somehow does.  And indeed, it also happily removes the whole-aggregate store
for

    s signInfo = { sizeof(signInfo), 0 };

    signInfo.b = 1;
    signInfo.c = 0;
    h(signInfo.d);
    signInfo.d = 1;

if you disable other optimizations before DSE.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[rguenth at gcc dot gnu dot org]

------- Comment #23 from rguenth at gcc dot gnu dot org  2007-08-31 14:13 -------
DSE would need to keep track of the "last" partial store to the aggregate and
use that for memory_address_same () check to make

"   The single use of the store's virtual definition ensures that
   there are no intervening aliased loads and the requirement that
   the second load post dominate the first ensures that if the earlier
   store executes, then the later stores will execute before the function
   exits."

true for aggregate stores.  "last" as in not otherwise dead last store, in the
testcase that would be signInfo.d = 1; stmt.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[rguenth at gcc dot gnu dot org]

------- Comment #24 from rguenth at gcc dot gnu dot org  2007-08-31 14:26 -------
Ugh.  But of course memory_address_same only checks what the name suggests.
Nowhere do we check for intermediate uses.  That whole-aggregate DSE just
cannot work at all.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[rguenth at gcc dot gnu dot org]

------- Comment #25 from rguenth at gcc dot gnu dot org  2007-08-31 14:34 -------
Aldy, can you look into this again?  I tried to simply revert the
whole-aggregate
DSE change, but it went in with the mem-ssa merge :/


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[marcus at jet dot franken dot de]

------- Comment #26 from marcus at jet dot franken dot de  2007-09-02 18:35 -------
btw, this went latent again with commit r127834:

+2007-08-27  Daniel Berlin  <dberlin@dberlin.org>
+       
+       Fix PR tree-optimization/33173
+       * tree-ssa-alias.c (find_used_portions): Fix reversed test.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[rguenth at gcc dot gnu dot org]

------- Comment #27 from rguenth at gcc dot gnu dot org  2007-09-25 13:58 -------
tree-dse also miscompiles perls pp_pack.c.
It removes the store to symptr->patend here:

  # MPT.498_212 = VDEF <MPT.498_210>
  symptr_19(D)->patend = D.14885_146;
  lenptr.50_147 = (char *) patptr_128;
  # MPT.498_213 = VDEF <MPT.498_212>
  symptr_19(D)->patptr = lenptr.50_147;
  # MPT.498_214 = VDEF <MPT.498_213>
  D.14887_148 = S_measure_struct (my_perl_44(D), symptr_19(D));

ignoring the use by S_measure_struct.  Which causes t/op/pack.t to fail.


-- 

rguenth at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
OtherBugsDependingO|                            |33383
              nThis|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[rguenth at gcc dot gnu dot org]

------- Comment #28 from rguenth at gcc dot gnu dot org  2007-09-25 14:48 -------
I'll take it for now.


-- 

rguenth at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|dnovillo at gcc dot gnu dot |rguenth at gcc dot gnu dot
                   |org                         |org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[belyshev at depni dot sinp dot msu dot ru]

------- Comment #29 from belyshev at depni dot sinp dot msu dot ru  2007-09-26 09:31 -------
Another testcase:

/* { dg-do run } */
/* { dg-options "-O1 --param max-aliased-vops=0" } */

struct T
{
  int a, b;
} t;

__attribute__((noinline)) struct T *f (struct T *p)
{
  struct T *q = __builtin_malloc (sizeof (struct T));
  *q = *p;
  return q;
}

int main (void)
{
  struct T *p;

  t.a = 1;
  t.b = 2;
  p = f (&t);
  t.a = 3;

  if (p->a != 1)
    __builtin_abort ();

  return 0;
}


-- 

belyshev at depni dot sinp dot msu dot ru changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |belyshev at depni dot sinp
                   |                            |dot msu dot ru


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[belyshev at depni dot sinp dot msu dot ru]

------- Comment #30 from belyshev at depni dot sinp dot msu dot ru  2007-09-26 09:46 -------
(In reply to comment #29)
> Another testcase:
Ignore this one, I filed it as a separate report, see bug 33560


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[rguenth at gcc dot gnu dot org]

------- Comment #31 from rguenth at gcc dot gnu dot org  2007-09-26 11:55 -------
Subject: Bug 30375

Author: rguenth
Date: Wed Sep 26 11:55:17 2007
New Revision: 128810

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=128810
Log:
2007-09-26  Richard Guenther  <rguenther@suse.de>

        PR tree-optimization/30375
        PR tree-optimization/33560
        * tree-ssa-dse.c (get_use_of_stmt_lhs): Give up on uses
        with calls.

        Revert
        2006-05-22  Aldy Hernandez  <aldyh@redhat.com>

        * tree-ssa-dse.c (aggregate_vardecl_d): New.
        (dse_global_data): Add aggregate_vardecl field.
        (dse_possible_dead_store_p): New.
        Add prev_defvar variable.
        Allow immediate uses and previous immediate uses to differ
        if they are setting different parts of the whole.
        (get_aggregate_vardecl): New.
        (dse_record_partial_aggregate_store): New.
        (dse_whole_aggregate_clobbered_p): New.
        (dse_partial_kill_p): New.
        Call dse_maybe_record_aggregate_store().
        When checking whether a STMT and its USE_STMT refer to the
        same memory address, check also for partial kills that clobber
        the whole.
        Move some variable definitions to the block where they are used.
        (aggregate_vardecl_hash): New.
        (aggregate_vardecl_eq): New.
        (aggregate_vardecl_free): New.
        (aggregate_whole_store_p): New.
        (tree_ssa_dse): Initialize and free aggregate_vardecl.
        Mark which aggregate stores we care about.

        * gcc.dg/tree-ssa/complex-4.c: XFAIL.
        * gcc.dg/tree-ssa/complex-5.c: Likewise.
        * gcc.dg/tree-ssa/ssa-dse-9.c: Likewise.
        * gcc.dg/torture/pr30375.c: New testcase.
        * gcc.dg/torture/pr33560.c: New testcase.
        * gcc.dg/tree-ssa/pr30375.c: Likewise.

Added:
    trunk/gcc/testsuite/gcc.dg/torture/pr30375.c
    trunk/gcc/testsuite/gcc.dg/torture/pr33560.c
    trunk/gcc/testsuite/gcc.dg/tree-ssa/pr30375.c
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/testsuite/ChangeLog
    trunk/gcc/testsuite/gcc.dg/tree-ssa/complex-4.c
    trunk/gcc/testsuite/gcc.dg/tree-ssa/complex-5.c
    trunk/gcc/testsuite/gcc.dg/tree-ssa/ssa-dse-9.c
    trunk/gcc/tree-ssa-dse.c


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375

[Bug tree-optimization/30375] [4.3 Regression] tree-ssa-dse incorrectly removes struct initialization

[rguenth at gcc dot gnu dot org]

------- Comment #32 from rguenth at gcc dot gnu dot org  2007-09-26 11:56 -------
Fixed.


-- 

rguenth at gcc dot gnu dot org changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30375