Re: c/5250: ICE in print_rtl_and_abort, at flow.c:6458 with -O3

Re: c/5250: ICE in print_rtl_and_abort, at flow.c:6458 with -O3

[Gwenole Beauchesne]

The following reply was made to PR c/5250; it has been noted by GNATS.

From: Gwenole Beauchesne <gbeauchesne@mandrakesoft.com>
To: dmuell@gmx.net, gerold.jury@utanet.at, joel@gcc.gnu.org, maciej@imsa.edu,
        utx@penguin.cz, relf@os2.ru, snowwolf@one2one-networks.com,
        gcc-gnats@gcc.gnu.org, gcc-prs@gcc.gnu.org, gcc-bugs@gcc.gnu.org,
        nobody@gcc.gnu.org, nottelm@uni-muenster.de
Cc:  
Subject: Re: c/5250: ICE in print_rtl_and_abort, at flow.c:6458 with -O3
Date: Fri, 08 Feb 2002 23:09:39 +0100

 Hi,
 
 PR c/4257, PR c/5409, PR c/5250 appear to have been fixed in gcc-3.1 
 when Jan Hubicka killed jump_optimize: 
 <http://gcc.gnu.org/ml/gcc-patches/2001-07/msg01430.html> .
 
 I don't know if I get it right but the following patch attempts to 
 disable the NOOP_MOVES jump_optimization prior to register renaming.
 
 2002-02-08  Gwenole Beauchesne  <gbeauchesne@mandrakesoft.com>
 
     PR c/4257, PR c/5049, PR c/5250
     * toplev.c (rest_of_compilation): Don't perform NOOP_MOVES
     jump_optimization prior to register renaming. However, tidy the
     CFG by deleting unreachable code.
     * gcc.c-torture/compile/20020208-1.c: New test.
 
 --- 
 gcc-3.0.4/gcc/testsuite/gcc.c-torture/compile/20020208-1.c.no-noop_moves    
 Fri Feb  8 20:28:37 2002
 +++ gcc-3.0.4/gcc/testsuite/gcc.c-torture/compile/20020208-1.c    Fri 
 Feb  8 20:28:37 2002
 @@ -0,0 +1,11 @@
 +/* Origin: PR c/4257 from Volker Reichelt <reichelt@igpm.rwth-aachen.de>. 
 */
 +/* This testcase caused gcc-3.0.1 to ICE if compiled with -O1
 +   -frename-registers.  */
 +
 +void f ()
 +{
 +  unsigned int *p;
 +  p++;
 +  *p += *p / (*p < 2);
 +}
 +
 --- gcc-3.0.4/gcc/toplev.c.no-noop_moves    Fri Feb  8 20:01:51 2002
 +++ gcc-3.0.4/gcc/toplev.c    Fri Feb  8 20:01:51 2002
 @@ -3546,9 +3546,8 @@
    timevar_push (TV_FLOW2);
    open_dump_file (DFI_flow2, decl);
  
 -  jump_optimize (insns, !JUMP_CROSS_JUMP,
 -         JUMP_NOOP_MOVES, !JUMP_AFTER_REGSCAN);
    find_basic_blocks (insns, max_reg_num (), rtl_dump_file);
 +  cleanup_cfg ();
  
    /* On some machines, the prologue and epilogue code, or parts thereof,
       can be represented as RTL.  Doing so lets us schedule insns between
 
 
 http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=5250
 

Re: c/5250: ICE in print_rtl_and_abort, at flow.c:6458 with -O3

[rodrigc]

Synopsis: ICE in print_rtl_and_abort, at flow.c:6458 with -O3

State-Changed-From-To: analyzed->closed
State-Changed-By: rodrigc
State-Changed-When: Sun Mar  3 18:17:48 2002
State-Changed-Why:
    Problem reproduced with gcc 3.0.4.
    
    The problem does not occur with gcc 3.1 20020217 (experimental).

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=5250

Re: c/5250: ICE in print_rtl_and_abort, at flow.c:6458 with -O3

[Craig Rodrigues]

The following reply was made to PR c/5250; it has been noted by GNATS.

From: Craig Rodrigues <rodrigc@attbi.com>
To: dmuell@gmx.net, gerold.jury@utanet.at, joel@gcc.gnu.org, maciej@imsa.edu,
   utx@penguin.cz, relf@os2.ru, snowwolf@one2one-networks.com, usul@gmx.li,
   gcc-gnats@gcc.gnu.org, gcc-prs@gcc.gnu.org, gcc-bugs@gcc.gnu.org,
   nobody@gcc.gnu.org, nottelm@uni-muenster.de
Cc:  
Subject: Re: c/5250: ICE in print_rtl_and_abort, at flow.c:6458 with -O3
Date: Sun, 03 Mar 2002 20:58:06 -0500

 The following two PR's are duplicates of this one:
 http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=2293
 http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=5822
 
 
 http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=5250 
 
 

Re: c/5250: ICE in print_rtl_and_abort, at flow.c:6458 with -O3

[rodrigc]

Synopsis: ICE in print_rtl_and_abort, at flow.c:6458 with -O3

State-Changed-From-To: open->analyzed
State-Changed-By: rodrigc
State-Changed-When: Wed Jan  2 12:38:12 2002
State-Changed-Why:
    The following testcase contributed by Volker Reichelt
    in PR 4257 duplicates your ICE at -O3:
    
    /* From Volker Reichelt <reichelt@igpm.rwth-aachen.de>,
     * PR 4257 */
     
    void
    f ()
    {
      unsigned int *p;
      p++;
      *p += *p / (*p < 2);
    }
    
    The workaround is to add the flag -fno-rename-registers if
    compiling with -O3.
    
    This bug is fixed in gcc 3.1.
    

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=5250

c/5250: ICE in print_rtl_and_abort, at flow.c:6458 with -O3

[nottelm]

>Number:         5250
>Category:       c
>Synopsis:       ICE in print_rtl_and_abort, at flow.c:6458 with -O3
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    unassigned
>State:          open
>Class:          ice-on-legal-code
>Submitter-Id:   net
>Arrival-Date:   Wed Jan 02 10:06:00 PST 2002
>Closed-Date:
>Last-Modified:
>Originator:     nottelm@uni-muenster.de
>Release:        gcc-3.0.3
>Organization:
>Environment:
Linux pt2037 2.4.17 #1 SMP Fri Dec 21 20:50:49 MET 2001 i686 unknown
>Description:
See also PR5049 and PR4257.
The code is taken from kdemultimedia-3.0beta1 and reduced
to a testcase. Like in the prob reports above the ICE
vanishes if I choose -O2 as optimization level.
I hope the testcase is possibly helpful.
>How-To-Repeat:
gcc -O3 -c testcase.c
>Fix:

>Release-Note:
>Audit-Trail:
>Unformatted:
----gnatsweb-attachment----
Content-Type: application/octet-stream; name="testcase.c"
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="testcase.c"
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