[Bug lto/61786] New: wrong code by LTO on x86_64-linux-gnu

[Bug lto/61786] New: wrong code by LTO on x86_64-linux-gnu

[su at cs dot ucdavis.edu]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61786

            Bug ID: 61786
           Summary: wrong code by LTO on x86_64-linux-gnu
           Product: gcc
           Version: 4.10.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: lto
          Assignee: unassigned at gcc dot gnu.org
          Reporter: su at cs dot ucdavis.edu

The current gcc trunk miscompiles the following code when using LTO on
x86_64-linux-gnu in both 32-bit and 64-bit modes. 

This is a regression from 4.9.x. 

$ gcc-trunk -v
Using built-in specs.
COLLECT_GCC=gcc-trunk
COLLECT_LTO_WRAPPER=/usr/local/gcc-trunk/libexec/gcc/x86_64-unknown-linux-gnu/4.10.0/lto-wrapper
Target: x86_64-unknown-linux-gnu
Configured with: ../gcc-trunk/configure --prefix=/usr/local/gcc-trunk
--enable-languages=c,c++ --disable-werror --enable-multilib
Thread model: posix
gcc version 4.10.0 20140711 (experimental) [trunk revision 212447] (GCC) 
$ 
$ gcc-trunk -O0 -c fn1.c
$ gcc-trunk -O0 -c main.c
$ gcc-trunk -O1 fn1.o main.o
$ a.out
$ 
$ gcc-4.9 -flto -O0 -c fn1.c
$ gcc-4.9 -flto -O0 -c main.c
$ gcc-4.9 -flto -O1 fn1.o main.o
$ a.out
$ 
$ gcc-trunk -flto -O0 -c fn1.c
$ gcc-trunk -flto -O0 -c main.c
$ gcc-trunk -flto -O1 fn1.o main.o
$ a.out
^C
$ 
$ cat fn1.c
int a, *c = &a, d; 
char b = 1;

void
fn1 (void)
{
  d = 1;
 lbl:
  if (b == d)
    {
      d = *c;
      if (b)
    goto lbl;
    }
}

int
fn2 (void)
{
  fn1 ();
  return 0;
}
$ cat main.c
extern int fn2 (void); 

int
main ()
{
  fn2 ();
  return 0;
}
$

[Bug lto/61786] wrong code by LTO on x86_64-linux-gnu

[su at cs dot ucdavis.edu]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61786

--- Comment #1 from Zhendong Su <su at cs dot ucdavis.edu> ---
Not sure whether it is due to the same root cause, but the combined code is
also miscompiled at (only) -O2 by the current trunk. 

Interestingly, -flto seems to make the miscompilation disappear. 

$ gcc-trunk -v
Using built-in specs.
COLLECT_GCC=gcc-trunk
COLLECT_LTO_WRAPPER=/usr/local/gcc-trunk/libexec/gcc/x86_64-unknown-linux-gnu/4.10.0/lto-wrapper
Target: x86_64-unknown-linux-gnu
Configured with: ../gcc-trunk/configure --prefix=/usr/local/gcc-trunk
--enable-languages=c,c++ --disable-werror --enable-multilib
Thread model: posix
gcc version 4.10.0 20140711 (experimental) [trunk revision 212447] (GCC) 
$ 
$ gcc-trunk -Os small.c; a.out
$ gcc-trunk -O3 small.c; a.out
$ gcc-4.9 -O2 small.c; a.out
$ gcc-trunk -flto -O2 small.c; a.out
$ 
$ gcc-trunk -O2 small.c
$ a.out
^C
$ 
$ 


---------------------


int a, *c = &a, d; 
char b = 1;

void
fn1 ()
{
  d = 1;
 lbl:
  if (b == d)
    {
      d = *c;
      if (b)
    goto lbl;
    }
}

int
fn2 ()
{
  fn1 ();
  return 0;
}

int
main ()
{
  fn2 ();
  return 0;
}

[Bug lto/61786] wrong code by LTO on x86_64-linux-gnu

[hp at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61786

Hans-Peter Nilsson <hp at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2014-07-14
                 CC|                            |hp at gcc dot gnu.org
     Ever confirmed|0                           |1

--- Comment #2 from Hans-Peter Nilsson <hp at gcc dot gnu.org> ---
Comfirmed at r212486 (-O2 and not -O2 -flto).  Also seen for
mmix-knuth-mmixware so apparently not target-specific. (Would be nice to have a
test-case that aborts instead of loops infinitely; fits better within the
test-harness.)

[Bug lto/61786] wrong code by LTO on x86_64-linux-gnu

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61786

Richard Biener <rguenth at gcc dot gnu.org> changed:

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

--- Comment #3 from Richard Biener <rguenth at gcc dot gnu.org> ---
Both testcases confirmed as fixed with the PR61757 fix.  I'll add the
testcases.

[Bug lto/61786] wrong code by LTO on x86_64-linux-gnu

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61786

--- Comment #4 from Richard Biener <rguenth at gcc dot gnu.org> ---
Author: rguenth
Date: Mon Jul 14 11:22:34 2014
New Revision: 212515

URL: https://gcc.gnu.org/viewcvs?rev=212515&root=gcc&view=rev
Log:
2014-07-14  Richard Biener  <rguenther@suse.de>

    PR tree-optimization/61786
    * gcc.dg/torture/pr61786.c: New testcase.
    * gcc.dg/lto/pr61786_0.c: Likewise.
    * gcc.dg/lto/pr61786_1.c: Likewise.

Added:
    trunk/gcc/testsuite/gcc.dg/lto/pr61786_0.c
    trunk/gcc/testsuite/gcc.dg/lto/pr61786_1.c
    trunk/gcc/testsuite/gcc.dg/torture/pr61786.c
Modified:
    trunk/gcc/testsuite/ChangeLog