[Bug middle-end/55152] New: MAX_EXPR(a,-a) is really ABS_EXPR(a)

[Bug middle-end/55152] New: MAX_EXPR(a,-a) is really ABS_EXPR(a)

[glisse at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55152

             Bug #: 55152
           Summary: MAX_EXPR(a,-a) is really ABS_EXPR(a)
    Classification: Unclassified
           Product: gcc
           Version: 4.8.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: middle-end
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: glisse@gcc.gnu.org


I tried compiling this code with -Ofast, but *.optimized still only contained a
MAX_EXPR and not an ABS_EXPR (the back-end didn't find the optimization
either).

double f(double a){
  return (a>=-a)?a:-a;
}

It seems that some simple code in fold-const.c could help (fold is called from
the front-end for COND_EXPR), not sure if anything would call it if -a went
though a temporary variable though.

[Bug middle-end/55152] MAX_EXPR(a,-a) is really ABS_EXPR(a)

[pinskia at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55152

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> 2012-10-31 20:18:58 UTC ---
Only if you ignore signed zeros and maybe even NaNs.

[Bug middle-end/55152] MAX_EXPR(a,-a) is really ABS_EXPR(a)

[glisse at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55152

--- Comment #2 from Marc Glisse <glisse at gcc dot gnu.org> 2012-10-31 21:14:12 UTC ---
Ignoring signed zeros, sure, that's also required to generate a MAX_EXPR in the
first place. NaNs, I don't know, it sounds to me like they have an effect on
generating a MAX_EXPR, but turning the MAX into an ABS should be safe.

(sometimes I wish for a __builtin_max, and -fno-nans)

[Bug middle-end/55152] MAX_EXPR(a,-a) is really ABS_EXPR(a)

[glisse at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55152

--- Comment #3 from Marc Glisse <glisse at gcc dot gnu.org> 2012-10-31 21:27:27 UTC ---
Actually, looking at the doc of MAX_EXPR in tree.def (nothing in generic.texi),
we don't even need to ignore signed zeros to turn MAX_EXPR into ABS_EXPR:

/* Minimum and maximum values.  When used with floating point, if both
   operands are zeros, or if either operand is NaN, then it is unspecified
   which of the two operands is returned as the result.  */

[Bug middle-end/55152] MAX_EXPR(a,-a) is really ABS_EXPR(a)

[pinskia at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55152

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |missed-optimization
           Severity|normal                      |enhancement

[Bug middle-end/55152] MAX_EXPR(a,-a) is really ABS_EXPR(a)

[glisse at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55152

--- Comment #4 from Marc Glisse <glisse at gcc dot gnu.org> ---
Created attachment 30502
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=30502&action=edit
fold-const.c patch

I've had this patch on my hard drive for a while and it will be better here. It
needs more work. The MAX optimization can probably never be reached from C code
(and the signed overflow warning seems unnecessary in this particular case),
part of the patch should move to gimple I guess, and I shouldn't produce the
non-canonical 0 op X when I could easily use swap_tree_comparison.