[Bug c/39612] New: Incorrect warning issued Re variable *is* used uninitialized in this function

[Bug c/39612] New: Incorrect warning issued Re variable *is* used uninitialized in this function

[tony at bakeyournoodle dot com]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 833 bytes --]

GCC Version: gcc (gcc version 4.4.0 20090307 (Red Hat 4.4.0-0.23) (GCC)
Distribution: Fedora Rawhide.

The attached pre-processed code generates:
f.i: In function ‘f2’:
f.i:19: warning: ‘inter’ is used uninitialized in this function

I believe that this warning is incorrect.


-- 
           Summary: Incorrect warning issued Re variable *is* used
                    uninitialized in this function
           Product: gcc
           Version: 4.4.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: tony at bakeyournoodle dot com
 GCC build triplet: ppc64-redhat-linux
  GCC host triplet: ppc64-redhat-linux
GCC target triplet: ppc64-redhat-linux


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39612

[Bug c/39612] Incorrect warning issued Re variable *is* used uninitialized in this function

[tony at bakeyournoodle dot com]

------- Comment #1 from tony at bakeyournoodle dot com  2009-04-02 04:35 -------
Created an attachment (id=17577)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17577&action=view)
Preprocess code to so problem

compile with: gcc -Os -Wall -Wundef -o /dev/null -c f.i

As you can see the 2 functions f1 and f2, are the same expect for the size of
the array.  Only f2() generates the warning.  -Os is needed, other -O levels do
not show the problem


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39612

[Bug tree-optimization/39612] [4.4/4.5 Regression] Incorrect warning issued Re variable *is* used uninitialized in this function

[rguenth at gcc dot gnu dot org]

------- Comment #2 from rguenth at gcc dot gnu dot org  2009-04-02 09:43 -------
Confirmed.  store motion transforms a loop

 for (;;)
   a[0] = 1;

to

 lsmtmp = a[0];
 for (;;)
   lsmtmp = 1;
 a[0] = lsmtmp;

introducing a load from uninitialized memory.  We don't seem to DCE/DSE this
for the tree element array but only the two element one.

The fix is to avoid the load part of load-store-motion of course.  The
warning about unused aggregates is new in 4.4 which makes this a regression.

On the a-i branch we remove the stores later.


-- 

rguenth at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |rguenth at gcc dot gnu dot
                   |                            |org, rakdver at gcc dot gnu
                   |                            |dot org
             Status|UNCONFIRMED                 |NEW
          Component|c                           |tree-optimization
     Ever Confirmed|0                           |1
           Keywords|                            |missed-optimization
   Last reconfirmed|0000-00-00 00:00:00         |2009-04-02 09:43:52
               date|                            |
            Summary|Incorrect warning issued Re |[4.4/4.5 Regression]
                   |variable *is* used          |Incorrect warning issued Re
                   |uninitialized in this       |variable *is* used
                   |function                    |uninitialized in this
                   |                            |function
   Target Milestone|---                         |4.4.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39612

[Bug tree-optimization/39612] [4.4/4.5 Regression] Incorrect warning issued Re variable *is* used uninitialized in this function

[rakdver at gcc dot gnu dot org]

-- 

rakdver at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |rakdver at gcc dot gnu dot
                   |dot org                     |org
             Status|NEW                         |ASSIGNED
   Last reconfirmed|2009-04-02 09:43:52         |2009-04-03 17:41:20
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39612

[Bug tree-optimization/39612] [4.4/4.5 Regression] Incorrect warning issued Re variable *is* used uninitialized in this function

[jakub at gcc dot gnu dot org]

-- 

jakub at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P2


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39612

[Bug tree-optimization/39612] [4.4/4.5 Regression] Incorrect warning issued Re variable *is* used uninitialized in this function

[jakub at gcc dot gnu dot org]

-- 

jakub at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|4.4.0                       |4.4.1


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39612

[Bug tree-optimization/39612] [4.4/4.5 Regression] Incorrect warning issued Re variable *is* used uninitialized in this function

[rakdver at gcc dot gnu dot org]

------- Comment #3 from rakdver at gcc dot gnu dot org  2009-04-25 22:44 -------
I cannot reproduce this in 4.5; all the unnecessary loads are removed.

> The fix is to avoid the load part of load-store-motion of course.

I've considered this, but it seems much easier to just let the unnecessary
loads be dce'd (detecting the unnecessary loads in lsm would basically
duplicate the work done by ssa update for the new variables.  I thought about
doing the ssa update and then removing the loads that are unused, but since
that seems to happen anyway, there is not much point).


-- 

rakdver at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|rakdver at gcc dot gnu dot  |unassigned at gcc dot gnu
                   |org                         |dot org
             Status|ASSIGNED                    |NEW


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39612

[Bug tree-optimization/39612] [4.3/4.4/4.5 Regression] LIM inserts loads from uninitialized local memory

[rguenth at gcc dot gnu dot org]

------- Comment #4 from rguenth at gcc dot gnu dot org  2009-04-26 17:34 -------
void foo(int *);
void f2(int dst[3], int R)
{
 int i, inter[2];

 for (i = 1; i < R; i++) {
  inter[0] = 1;
  inter[1] = 1;
 }

 foo(inter);
}

Warns at -Os or also at -O2 if you disable loop header copying.  So maybe
we should avoid doing loads from local memory in places we do not know
will be executed.

On trunk simply the warning doesn't work anymore, the loads are still there,
same for 4.3.

The loads are not removed by any pass (without loop header copying) and the
loop remains.  Note that even with -Os -ftree-ch no loop header copying
is performed.


-- 

rguenth at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
      Known to work|                            |4.2.4
            Summary|[4.4/4.5 Regression]        |[4.3/4.4/4.5 Regression] LIM
                   |Incorrect warning issued Re |inserts loads from
                   |variable *is* used          |uninitialized local memory
                   |uninitialized in this       |
                   |function                    |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39612

[Bug tree-optimization/39612] [4.3/4.4/4.5 Regression] LIM inserts loads from uninitialized local memory

[rakdver at gcc dot gnu dot org]

------- Comment #5 from rakdver at gcc dot gnu dot org  2009-04-26 18:37 -------
(In reply to comment #4)
> void foo(int *);
> void f2(int dst[3], int R)
> {
>  int i, inter[2];
> 
>  for (i = 1; i < R; i++) {
>   inter[0] = 1;
>   inter[1] = 1;
>  }
> 
>  foo(inter);
> }
> 
> Warns at -Os or also at -O2 if you disable loop header copying.

Right. The initial value of inter may reach the call to foo (when R <= 1), so
we must issue (and keep) the loads if we want to perform lsm without loop
header copying.  While I recognize that adding the loads from (possibly)
uninitialized memory is somewhat annoying, limiting lsm only to loops where we
can prove that this does not happen does not make much sense, either.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39612

[Bug tree-optimization/39612] [4.3/4.4/4.5 Regression] LIM inserts loads from uninitialized local memory

[rguenther at suse dot de]

------- Comment #6 from rguenther at suse dot de  2009-04-26 18:44 -------
Subject: Re:  [4.3/4.4/4.5 Regression] LIM
 inserts loads from uninitialized local memory

On Sun, 26 Apr 2009, rakdver at gcc dot gnu dot org wrote:

> ------- Comment #5 from rakdver at gcc dot gnu dot org  2009-04-26 18:37 -------
> (In reply to comment #4)
> > void foo(int *);
> > void f2(int dst[3], int R)
> > {
> >  int i, inter[2];
> > 
> >  for (i = 1; i < R; i++) {
> >   inter[0] = 1;
> >   inter[1] = 1;
> >  }
> > 
> >  foo(inter);
> > }
> > 
> > Warns at -Os or also at -O2 if you disable loop header copying.
> 
> Right. The initial value of inter may reach the call to foo (when R <= 1), so
> we must issue (and keep) the loads if we want to perform lsm without loop
> header copying.  While I recognize that adding the loads from (possibly)
> uninitialized memory is somewhat annoying, limiting lsm only to loops where we
> can prove that this does not happen does not make much sense, either.

Ok.  I guess setting TREE_NO_WARNING on the inserted loads could avoid
the uninitialized warning in 4.4.

Richard.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39612

[Bug tree-optimization/39612] [4.3/4.4/4.5 Regression] LIM inserts loads from uninitialized local memory

[jakub at gcc dot gnu dot org]

------- Comment #7 from jakub at gcc dot gnu dot org  2009-04-27 07:24 -------
Won't that disable the warning in the valid cases though?

int
f2 (int R)
{
  int i, inter[3];

  for (i = 1; i < R; i++)
    {
      inter[0] = 1;
      inter[1] = 1;
      inter[2] = 1;
    }

  return inter[0];
}

I've tried apparently doesn't warn even without the change, because ccp for a
PHI which has undefined SSA name and defined SSA name just propagates the
defined one (without a warning), so I don't have a testcase.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39612

[Bug tree-optimization/39612] [4.3/4.4/4.5 Regression] LIM inserts loads from uninitialized local memory

[rguenther at suse dot de]

------- Comment #8 from rguenther at suse dot de  2009-04-27 08:16 -------
Subject: Re:  [4.3/4.4/4.5 Regression] LIM
 inserts loads from uninitialized local memory

On Mon, 27 Apr 2009, jakub at gcc dot gnu dot org wrote:

> ------- Comment #7 from jakub at gcc dot gnu dot org  2009-04-27 07:24 -------
> Won't that disable the warning in the valid cases though?

Good question.  I suppose it would.

> int
> f2 (int R)
> {
>   int i, inter[3];
> 
>   for (i = 1; i < R; i++)
>     {
>       inter[0] = 1;
>       inter[1] = 1;
>       inter[2] = 1;
>     }
> 
>   return inter[0];
> }
> 
> I've tried apparently doesn't warn even without the change, because ccp for a
> PHI which has undefined SSA name and defined SSA name just propagates the
> defined one (without a warning), so I don't have a testcase.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39612

[Bug tree-optimization/39612] [4.3/4.4/4.5 Regression] LIM inserts loads from uninitialized local memory

[jakub at gcc dot gnu dot org]

-- 

jakub at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|4.4.1                       |4.4.2


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39612

[Bug tree-optimization/39612] [4.3/4.4/4.5 Regression] LIM inserts loads from uninitialized local memory

[pinskia at gcc dot gnu dot org]

------- Comment #9 from pinskia at gcc dot gnu dot org  2009-10-05 20:09 -------
The warning is gone but the issue still exist.


-- 

pinskia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|2009-04-03 17:41:20         |2009-10-05 20:09:44
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39612

[Bug tree-optimization/39612] [4.3/4.4/4.5 Regression] LIM inserts loads from uninitialized local memory

[jakub at gcc dot gnu dot org]

-- 

jakub at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|4.4.2                       |4.4.3


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39612

[Bug tree-optimization/39612] [4.3/4.4/4.5 Regression] LIM inserts loads from uninitialized local memory

[jakub at gcc dot gnu dot org]

-- 

jakub at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|4.4.3                       |4.4.4


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39612

[Bug tree-optimization/39612] [4.3/4.4/4.5 Regression] LIM inserts loads from uninitialized local memory

[pinskia at gcc dot gnu dot org]

------- Comment #10 from pinskia at gcc dot gnu dot org  2010-03-03 20:21 -------
I have noticed this with some code with graphite enabled too, graphite uses
arrays for some reason.


-- 

pinskia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
  GCC build triplet|ppc64-redhat-linux          |
   GCC host triplet|ppc64-redhat-linux          |
 GCC target triplet|ppc64-redhat-linux          |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39612

[Bug tree-optimization/39612] [4.3/4.4/4.5/4.6 Regression] LIM inserts loads from uninitialized local memory

[jakub at gcc dot gnu dot org]

-- 

jakub at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|4.4.4                       |4.4.5


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39612