[Bug tree-optimization/103226] New: [12 Regression] Recent change to copy-headers causes execution failure for gcc.dg/torture/pr80974

[law at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103226

            Bug ID: 103226
           Summary: [12 Regression] Recent change to copy-headers causes
                    execution failure for gcc.dg/torture/pr80974
           Product: gcc
           Version: 12.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: law at gcc dot gnu.org
  Target Milestone: ---

This change:
commit e82c382971664d6fd138cc36020db4b1a91885c6
Author: Aldy Hernandez <aldyh@redhat.com>
Date:   Wed Nov 10 13:21:59 2021 +0100

    Allow loop header copying when first iteration condition is known.

Causes gcc.dg/torture/pr80974.c to fail with -O2 on bfin-elf

I haven't debugged it in any significant way other than bisection to the above
change and quickly looking at the before/after dumps.  Things diverge at the
.ch2 dump (no surprise there).  You may ultimately have to set up a cross
environment to debug this one.

The one other tidbit, before this change, the test finishes relatively fast,
afterwards it takes a LOT longer.  Maybe we're wrapping a loop or something
like that.  Dunno.  It may ultimately not be relevant.

Oh, if you end up doing a binutils, gcc & newlib build and want to test things
yourself using dejagnu you might need a suitable baseboards file.  Here's mine,
which was copied and edited from some other port...


# Copyright (C) 1997-2016 Free Software Foundation, Inc.
#
# This file is part of DejaGnu.
#
# DejaGnu is free software; you can redistribute it and/or modify it
# under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 3 of the License, or
# (at your option) any later version.
#
# DejaGnu is distributed in the hope that it will be useful, but
# WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
# General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with DejaGnu; if not, write to the Free Software Foundation,
# Inc., 51 Franklin Street - Fifth Floor, Boston, MA 02110-1301, USA.

# This is a list of toolchains that are supported on this board.
set_board_info target_install {bfin-elf}

# Load the generic configuration for this board. This will define a basic set
# of routines needed by the tool to communicate with the board.
load_generic_config "sim"

# basic-sim.exp is a basic description for the standard Cygnus simulator.
load_base_board_description "basic-sim"

# "fr30" is the name of the sim subdir.
setup_sim bfin

# No multilib options needed by default.
process_multilib_options ""

# We only support newlib on this target. We assume that all multilib
# options have been specified before we get here.

set_board_info compiler  "[find_gcc]"
set_board_info cflags    "[libgloss_include_flags] [newlib_include_flags]"
set_board_info ldflags   "-msim [libgloss_link_flags] [newlib_link_flags]"
# No linker script needed.
set_board_info ldscript ""

# The simulator doesn't return exit statuses and we need to indicate this;
# the standard GCC wrapper will work with this target.
set_board_info needs_status_wrapper  1
# Doesn't pass arguments or signals, can't return results, and doesn't
# do inferiorio.
set_board_info noargs 1
set_board_info gdb,nosignals 1
set_board_info gdb,noresults 1
set_board_info gdb,noinferiorio 1

set_board_info gcc,stack_size  4096