[Bug optimization/13301] New: [tree-ssa] deficiency in jump threading

[Bug optimization/13301] New: [tree-ssa] deficiency in jump threading

[rth at gcc dot gnu dot org]

$ ./cc1 -quiet -Wall -O z.c
z.c: In function `test':
z.c:22: warning: 'idx' might be used uninitialized in this function

The test case is representative of (reportedly many) similar warnings
generated from the Linux source code.  The warning is suppressed at -O2
by rtl's jump bypassing pass.  IMO this case should be handled by the
tree-level jump threading routines.

-- 
           Summary: [tree-ssa] deficiency in jump threading
           Product: gcc
           Version: tree-ssa
            Status: UNCONFIRMED
          Severity: enhancement
          Priority: P2
         Component: optimization
        AssignedTo: law at gcc dot gnu dot org
        ReportedBy: rth at gcc dot gnu dot org
                CC: gcc-bugs at gcc dot gnu dot org
GCC target triplet: i686-linux


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=13301

[Bug optimization/13301] [tree-ssa] deficiency in jump threading

[rth at gcc dot gnu dot org]

------- Additional Comments From rth at gcc dot gnu dot org  2003-12-04 10:22 -------
Created an attachment (id=5274)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=5274&action=view)
Test case.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=13301

[Bug optimization/13301] [tree-ssa] deficiency in jump threading

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2003-12-04 10:45 -------
I do not know if the tree jump threading supports switch tables at all (or the GNU extensions for 
multiple cases) but it would be good to know the real case of this bug.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |ASSIGNED
     Ever Confirmed|                            |1
   Last reconfirmed|0000-00-00 00:00:00         |2003-12-04 10:45:19
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=13301

[Bug optimization/13301] [tree-ssa] deficiency in jump threading

[rth at redhat dot com]

------- Additional Comments From rth at redhat dot com  2003-12-04 11:12 -------
Subject: Re:  [tree-ssa] deficiency in jump threading

On Thu, Dec 04, 2003 at 10:45:19AM -0000, pinskia at gcc dot gnu dot org wrote:
> I do not know if the tree jump threading supports switch tables at all...

It does, but that's not the problem.  Past the switch, the 
"return NULL" gets compiled to "retval = NULL; goto retloc;"
and we must thread *that* goto with the if condition that
immediately follows.

> but it would be good to know the real case of this bug.

I have no doubt that there is a PHI node in the block that
contains the if condition, and that it is causing the 
existing jump threading pass to give up.

I also happen to know that Jeff said that he is working on
this type of problem, so I assigned the PR to him.


r~


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=13301

[Bug optimization/13301] [tree-ssa] deficiency in jump threading

[law at redhat dot com]

------- Additional Comments From law at redhat dot com  2003-12-04 22:19 -------
Subject: Re:  [tree-ssa] deficiency in jump threading 

In message <20031204111206.14420.qmail@sources.redhat.com>, "rth at redhat dot 
com" writes:
 >
 >------- Additional Comments From rth at redhat dot com  2003-12-04 11:12 ----
 >---
 >Subject: Re:  [tree-ssa] deficiency in jump threading
 >
 >On Thu, Dec 04, 2003 at 10:45:19AM -0000, pinskia at gcc dot gnu dot org wrot
 >e:
 >> I do not know if the tree jump threading supports switch tables at all...
 >
 >It does, but that's not the problem.  Past the switch, the 
 >"return NULL" gets compiled to "retval = NULL; goto retloc;"
 >and we must thread *that* goto with the if condition that
 >immediately follows.
 >
 >> but it would be good to know the real case of this bug.
 >
 >I have no doubt that there is a PHI node in the block that
 >contains the if condition, and that it is causing the 
 >existing jump threading pass to give up.
 >
 >I also happen to know that Jeff said that he is working on
 >this type of problem, so I assigned the PR to him.
Yup.  I've gone back to jump threading from the not-so-brief poking at
memory consumption issues. 

jeff




-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=13301

[Bug optimization/13301] [tree-ssa] deficiency in jump threading

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |tree-ssa


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=13301

[Bug optimization/13301] [tree-ssa] deficiency in jump threading

[law at gcc dot gnu dot org]

------- Additional Comments From law at gcc dot gnu dot org  2004-01-15 19:01 -------
Fixed by my Jan 15, 2004 improvement to tree-ssa-dom.c

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=13301