[Bug middle-end/21032] [4.0 Regression] With -frounding-math, incorrectly reorders unary minus

[bagnara at cs dot unipr dot it]

------- Comment #23 from bagnara at cs dot unipr dot it  2007-09-10 12:32 -------
My fault: I forgot to use the -frounding-math option.  So, for the "wrong-code"
aspect there is no problem.  But the "missed-optimization" bit is still there:
why do we have

        fldl    12(%ebp)
        fchs
        movl    8(%ebp), %eax
        fstps   -20(%ebp)
        flds    -20(%ebp)
        fstps   -4(%ebp)
        flds    -4(%ebp)
        fchs
        fstps   (%eax)

insted of simply

        fldl    12(%ebp)
        fchs
        fstps   -4(%ebp)
        flds    -4(%ebp)
        fchs
        fstps   (%eax)

?


-- 

bagnara at cs dot unipr dot it changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |abramobagnara at tin dot it


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21032