[Bug middle-end/11700] New: [3.3/3.4 regression] [M68K] ICE in subreg_hard_regno

[Bug middle-end/11700] New: [3.3/3.4 regression] [M68K] ICE in subreg_hard_regno

[schwab at suse dot de]

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

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11700

           Summary: [3.3/3.4 regression] [M68K] ICE in subreg_hard_regno
           Product: gcc
           Version: 3.4
            Status: UNCONFIRMED
          Keywords: ice-on-valid-code
          Severity: normal
          Priority: P2
         Component: middle-end
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: schwab at suse dot de
                CC: gcc-bugs at gcc dot gnu dot org
 GCC build triplet: powerpc-linux
  GCC host triplet: powerpc-linux
GCC target triplet: m68k-linux

This is similar to bug 10557.  Either subreg_offset_representable_p is still to strict, or it should 
not be called at all when !check_mode. 
 
$ m68k-linux-gcc -O2 double.c -v 
Reading specs from /usr/local/m68k-linux/lib/gcc-lib/m68k-linux/3.4/specs 
Configured with: ../gcc/configure --host=powerpc-linux --prefix=/usr/local/m68k-linux 
--enable-languages=c,c++ --target=m68k-linux --enable-shared --enable-clocale=gnu 
--enable-__cxa_atexit --with-as=/usr/local/m68k-linux/m68k-linux/bin/as 
--with-ld=/usr/local/m68k-linux/m68k-linux/bin/ld 
--with-gcc-version-trigger=/home/andreas/src/gcc/gcc/gcc/version.c --no-create --no-recursion : 
(reconfigured)  
Thread model: posix 
gcc version 3.4 20030727 (experimental) 
 /usr/local/m68k-linux/lib/gcc-lib/m68k-linux/3.4/cc1 -quiet -v -D__HAVE_68881__ double.c 
-quiet -dumpbase double.c -auxbase double -O2 -version -o /tmp/ccsdwhgS.s 
ignoring nonexistent directory "/usr/local/m68k-linux/m68k-linux/sys-include" 
#include "..." search starts here: 
#include <...> search starts here: 
 /usr/local/m68k-linux/lib/gcc-lib/m68k-linux/3.4/include 
 /usr/local/m68k-linux/m68k-linux/include 
End of search list. 
GNU C version 3.4 20030727 (experimental) (m68k-linux) 
	compiled by GNU C version 3.4 20030712 (experimental). 
GGC heuristics: --param ggc-min-expand=47 --param ggc-min-heapsize=31832 
double.c: In function `conj_test': 
 
double.c:12: internal compiler error: in subreg_hard_regno, at emit-rtl.c:1023

[Bug middle-end/11700] [3.3/3.4 regression] [M68K] ICE in subreg_hard_regno

[schwab at suse dot de]

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

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11700



------- Additional Comments From schwab at suse dot de  2003-07-28 23:18 -------
Created an attachment (id=4500)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=4500&action=view)
Testcase

[Bug optimization/11700] [3.3/3.4 regression] [M68K] ICE in subreg_hard_regno

[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=11700


pinskia at physics dot uc dot edu changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
          Component|middle-end                  |optimization
     Ever Confirmed|                            |1
   Last reconfirmed|0000-00-00 00:00:00         |2003-07-29 02:53:13
               date|                            |


------- Additional Comments From pinskia at physics dot uc dot edu  2003-07-29 02:53 -------
I can confirm this on the mainline (20030728).

[Bug optimization/11700] [3.3/3.4 regression] [M68K] ICE in subreg_hard_regno

[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=11700


pinskia at physics dot uc dot edu changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |critical
           Priority|P2                          |P1

[Bug optimization/11700] [3.3/3.4 regression] [M68K] ICE in subreg_hard_regno

[janis187 at us dot ibm dot com]

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

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11700


janis187 at us dot ibm dot com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |roger at eyesopen dot com


------- Additional Comments From janis187 at us dot ibm dot com  2003-08-29 22:43 -------
The regression in PR 11700 was introduced or exposed with this patch:

--- gcc/gcc/ChangeLog ---

2003-05-30  Roger Sayle  <roger@eyesopen.com>

        * emit-rtl.c (gen_complex_constant_part): New function for getting
        the constant real or imaginary part of a complex constant.
        (gen_realpart): Use it.
        (gen_imagpart): Likewise.

The regression hunt took place on i686-pc-linux-gnu with a cross cc1
for m68k-linux using the test case from comment #1 compiled with -O2.

[Bug optimization/11700] [3.3/3.4 regression] [M68K] ICE in subreg_hard_regno

[cvs-commit at gcc dot gnu dot org]

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

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11700



------- Additional Comments From cvs-commit at gcc dot gnu dot org  2003-09-04 01:53 -------
Subject: Bug 11700

CVSROOT:	/cvs/gcc
Module name:	gcc
Changes by:	sayle@gcc.gnu.org	2003-09-04 01:53:01

Modified files:
	gcc            : ChangeLog simplify-rtx.c 
	gcc/testsuite  : ChangeLog 
Added files:
	gcc/testsuite/gcc.c-torture/compile: 20030903-1.c 

Log message:
	PR optimization/11700.
	* simplify-rtx.c (simplify_subreg): Check that the subreg offset
	of a hard register is representable before trying to simplify it
	using subreg_hard_regno.
	
	* gcc.c-torture/compile/20030903-1.c: New test case.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/ChangeLog.diff?cvsroot=gcc&r1=2.957&r2=2.958
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/simplify-rtx.c.diff?cvsroot=gcc&r1=1.156&r2=1.157
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&r1=1.3020&r2=1.3021
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gcc.c-torture/compile/20030903-1.c.diff?cvsroot=gcc&r1=NONE&r2=1.1

[Bug optimization/11700] [3.3/3.4 regression] [M68K] ICE in subreg_hard_regno

[roger at eyesopen dot com]

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

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11700


roger at eyesopen dot com changed:

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


------- Additional Comments From roger at eyesopen dot com  2003-09-05 13:12 -------
This should now be fixed on mainline.