[Bug optimization/8681] Generates unneeded test

[Bug optimization/8681] Generates unneeded test

[pinskia@physics.uc.edu]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=8681


pinskia@physics.uc.edu changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|0000-00-00 00:00:00         |2003-06-07 20:03:47
               date|                            |


------- Additional Comments From pinskia@physics.uc.edu  2003-06-07 20:03 -------
Still happens on the mainline (20030607).

[Bug optimization/8681] Generates unneeded test

[pinskia at physics dot uc dot edu]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=8681


pinskia at physics dot uc dot edu changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|2003-06-07 20:03:47         |2003-07-08 00:50:26
               date|                            |


------- Additional Comments From pinskia at physics dot uc dot edu  2003-07-08 00:50 -------
It still happens on the mainline (20030707) and it also happens on the tree-ssa branch.

[Bug optimization/8681] Generates unneeded test

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |enhancement
   Last reconfirmed|2003-11-23 08:02:00         |2003-12-26 03:34:20
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=8681

[Bug optimization/8681] Generates unneeded test

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-03-03 16:50 -------
This is semi fixed on the tree-ssa:

sum (p)
{
  int result;
  int i;

<bb 0>:
  result = 0;
  i = 0;
  goto <bb 4> (<L5>);

<L0>:;
  if (i < 0) goto <L1>; else goto <L2>;

<L1>:;
  result = 0;

<L2>:;

<L5>:;
  result = *((int *)((unsigned int)i * 4) + p) + result;
  i = i + 1;
  if (i <= 9999) goto <L0>; else goto <L4>;

<L4>:;
  return result;

}

There is still a comparision but no longer in the asm (meaning that the RTL optimizers 
optimize it away):
_sum:
        li r0,10001
        mr r11,r3
        li r9,0
        li r3,0
        mtctr r0
        slwi r5,r9,2
        lwzx r2,r5,r11
        addi r9,r9,1
        nor r4,r9,r9
        srawi r0,r4,31
        add r3,r3,r2
        bdz L12
L8:
        slwi r5,r9,2
        addi r9,r9,1
        lwzx r2,r5,r11
        and r3,r3,r0
        nor r4,r9,r9
        srawi r0,r4,31
        add r3,r3,r2
        bdnz L8
L12:
        blr

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|2003-12-26 03:34:20         |2004-03-03 16:50:28
               date|                            |
   Target Milestone|---                         |tree-ssa


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=8681

[Bug tree-optimization/8681] Generates unneeded test

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|rtl-optimization            |tree-optimization
   Target Milestone|3.5.0                       |---


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=8681

[Bug tree-optimization/8681] Generates unneeded test

[kazu at cs dot umass dot edu]

------- Additional Comments From kazu at cs dot umass dot edu  2005-01-22 18:14 -------
Still not fixed at tree level.

sum (p)
{
  unsigned int D.1166;
  unsigned int D.1161;
  int result;
  int i;

<bb 0>:
  result = 0;
  i = 0;
  goto <bb 3> (<L10>);

<L0>:;
  if (i < 0) goto <L11>; else goto <L10>;

<L11>:;
  result = 0;

<L10>:;
  D.1161 = (unsigned int) i;
  result = result + *(p + (int *) D.1161 * 4B);
  D.1166 = D.1161 + 1;
  i = (int) D.1166;
  if (D.1166 != 10000) goto <L0>; else goto <L4>;

<L4>:;
  return result;

}


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=8681

[Bug tree-optimization/8681] Generates unneeded test

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2005-04-24 17:56 -------
i_15: [1, 2147483647]
i_16: [0, 2147483647]
i_20: VARYING 

  # i_20 = PHI <i_15(3), 0(0)>;


Looks like VRP does not understand PHI functions or it just gives up too often.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
OtherBugsDependingO|                            |18373
              nThis|                            |
   Last reconfirmed|2005-01-13 18:59:31         |2005-04-24 17:56:49
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=8681

Re: [Bug tree-optimization/8681] Generates unneeded test

[Diego Novillo]

On Sun, Apr 24, 2005 at 05:56:50PM -0000, pinskia at gcc dot gnu dot org wrote:
> 
> i_15: [1, 2147483647]
> i_16: [0, 2147483647]
> i_20: VARYING 
> 
>   # i_20 = PHI <i_15(3), 0(0)>;
> 
> 
> Looks like VRP does not understand PHI functions or it just gives up too often.
> 
i_15 and 0 have a non-empty intersection and so vrp_meet returns
VARYING.  VRP does not handle multiple ranges, but in this case
it should not be hard to merge the two adjacent ranges [1, +INF]
and [0, 0].  

What is the type of 'i'?  If it's unsigned, then we would be
wasting our time.


Diego.

[Bug tree-optimization/8681] Generates unneeded test

[dnovillo at redhat dot com]

------- Additional Comments From dnovillo at redhat dot com  2005-04-24 18:07 -------
Subject: Re:  Generates unneeded test

On Sun, Apr 24, 2005 at 05:56:50PM -0000, pinskia at gcc dot gnu dot org wrote:
> 
> i_15: [1, 2147483647]
> i_16: [0, 2147483647]
> i_20: VARYING 
> 
>   # i_20 = PHI <i_15(3), 0(0)>;
> 
> 
> Looks like VRP does not understand PHI functions or it just gives up too often.
> 
i_15 and 0 have a non-empty intersection and so vrp_meet returns
VARYING.  VRP does not handle multiple ranges, but in this case
it should not be hard to merge the two adjacent ranges [1, +INF]
and [0, 0].  

What is the type of 'i'?  If it's unsigned, then we would be
wasting our time.


Diego.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=8681

[Bug tree-optimization/8681] Generates unneeded test

[pinskia at physics dot uc dot edu]

------- Additional Comments From pinskia at physics dot uc dot edu  2005-04-24 18:10 -------
Subject: Re:  Generates unneeded test


On Apr 24, 2005, at 2:06 PM, Diego Novillo wrote:

> What is the type of 'i'?  If it's unsigned, then we would be
> wasting our time.

It is signed, otherwise "i < 0" will always be true and the conditional
would have gotten rid of already.

-- Pinski



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=8681

Re: [Bug tree-optimization/8681] Generates unneeded test

[Andrew Pinski]

On Apr 24, 2005, at 2:06 PM, Diego Novillo wrote:

> What is the type of 'i'?  If it's unsigned, then we would be
> wasting our time.

It is signed, otherwise "i < 0" will always be true and the conditional
would have gotten rid of already.

-- Pinski

[Bug tree-optimization/8681] Generates unneeded test

[kazu at cs dot umass dot edu]

------- Additional Comments From kazu at cs dot umass dot edu  2005-05-10 05:37 -------
The same patch for PR 21458 would work.


-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |kazu at cs dot umass dot edu
                   |dot org                     |
             Status|NEW                         |ASSIGNED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=8681

[Bug tree-optimization/8681] Generates unneeded test

[dnovillo at gcc dot gnu dot org]

------- Additional Comments From dnovillo at gcc dot gnu dot org  2005-06-02 18:37 -------

Fixed.  Likely with http://gcc.gnu.org/ml/gcc-patches/2005-06/msg00127.html

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=8681

[Bug tree-optimization/8681] Generates unneeded test

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |4.1.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=8681

[Bug tree-optimization/8681] Generates unneeded test

[pinskia at gcc dot gnu dot org]

-- 
Bug 8681 depends on bug 18373, which changed state.

Bug 18373 Summary: [meta-bug] VRP Value Range Propagation
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18373

           What    |Old Value                   |New Value
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=8681