[Bug optimization/15077] New: ICE in make_decl_rtl when inlining tail recursive nested function

[Bug optimization/15077] New: ICE in make_decl_rtl when inlining tail recursive nested function

[falk at debian dot org]

gcc version 3.5-tree-ssa 20040416 (merged 20040414)

% cat test.c
int popcount(int x) {
    int loop(int x, int accu) {
        if (x == 0)
            return accu;
        else
            return loop(x & (x - 1), accu + 1);
    }
    return loop(x, 0);
}

% gcc -c -O2 -finline-functions test.c  
test.c: In function `popcount':
test.c:6: internal compiler error: in make_decl_rtl, at varasm.c:758
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.

-- 
           Summary: ICE in make_decl_rtl when inlining tail recursive nested
                    function
           Product: gcc
           Version: tree-ssa
            Status: UNCONFIRMED
          Keywords: ice-on-valid-code
          Severity: normal
          Priority: P2
         Component: optimization
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: falk at debian dot org
                CC: gcc-bugs at gcc dot gnu dot org
 GCC build triplet: alphaev68-unknown-linux-gnu
  GCC host triplet: alphaev68-unknown-linux-gnu
GCC target triplet: alphaev68-unknown-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15077

[Bug optimization/15077] ICE in make_decl_rtl when inlining tail recursive nested function

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-04-22 18:41 -------
Confirmed, inlining loop into itself is causing this failure.
Jan could you look into this because you know the inliner more than most people.  Also this is related 
to another bug which Pessimizes code at -O3 compared to -O2 becuase of inlining into it self.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |hubicka at gcc dot gnu dot
                   |                            |org
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|                            |1
   Last reconfirmed|0000-00-00 00:00:00         |2004-04-22 18:41:59
               date|                            |
   Target Milestone|---                         |tree-ssa


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15077

[Bug tree-optimization/15077] [3.5 Regression] ICE in make_decl_rtl when inlining tail recursive nested function

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2004-07-28 02:27 -------
Subject: Bug 15077

CVSROOT:	/cvs/gcc
Module name:	gcc
Changes by:	kenner@gcc.gnu.org	2004-07-28 02:27:20

Modified files:
	gcc            : ChangeLog function.h tree-inline.c 
	                 tree-inline.h tree-optimize.c 

Log message:
	PR optimization/15077
	* function.h (struct function): Add field saved_static_chain_decl.
	Fix comment for static_chain_decl.
	* tree-inline.c (save_body): Add new arg and handle static_chain_decl.
	* tree-inline.h (save_body): Add new arg.
	* tree-optimize.c (tree_rest_of_compilation): Handle saving
	static_chain_decl.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/ChangeLog.diff?cvsroot=gcc&r1=2.4713&r2=2.4714
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/function.h.diff?cvsroot=gcc&r1=1.130&r2=1.131
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/tree-inline.c.diff?cvsroot=gcc&r1=1.130&r2=1.131
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/tree-inline.h.diff?cvsroot=gcc&r1=1.11&r2=1.12
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/tree-optimize.c.diff?cvsroot=gcc&r1=2.33&r2=2.34



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15077

[Bug tree-optimization/15077] [3.5 Regression] ICE in make_decl_rtl when inlining tail recursive nested function

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-07-28 05:51 -------
Fixed.

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15077

[Bug tree-optimization/15077] [3.5 Regression] ICE in make_decl_rtl when inlining tail recursive nested function

[bangerth at dealii dot org]

------- Additional Comments From bangerth at dealii dot org  2004-07-28 13:04 -------
Kenner, this PR has a nice small testcase, but it seems as if you haven't 
added it to the testsuite. Any reason why? 
 
W. 

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15077

[Bug tree-optimization/15077] [3.5 Regression] ICE in make_decl_rtl when inlining tail recursive nested function

[kenner at vlsi1 dot ultra dot nyu dot edu]

------- Additional Comments From kenner at vlsi1 dot ultra dot nyu dot edu  2004-07-28 13:07 -------
Subject: Re:   [3.5 Regression] ICE in make_decl_rtl when inlining tail recursive nested function

    Kenner, this PR has a nice small testcase, but it seems as if you haven't 
    added it to the testsuite. Any reason why? 

No good one.  I originally saw this as an Ada bug and was informed it was
also an open PR right around the time I commited the fix.  The Ada test
case was already in the test suite and I'd forgotten about this one.

Indeed this one should be added as well.  I'll do that.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15077

[Bug tree-optimization/15077] [3.5 Regression] ICE in make_decl_rtl when inlining tail recursive nested function

[bangerth at dealii dot org]

------- Additional Comments From bangerth at dealii dot org  2004-07-28 13:13 -------
> Indeed this one should be added as well.  I'll do that. 
 
Thanks. Reducing testcases can be very painful and time consuming. 
It is disappointing to see if they are not added to the testsuite then, 
as it may mean that we have to do it again if the same bug reappears for 
some reason. 
 
W. 

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15077