[Bug target/13958] Conversion from unsigned to double is painfully slow on P4

[ubizjak at gmail dot com]

------- Comment #3 from ubizjak at gmail dot com  2008-03-21 13:45 -------
This is due to partial memory access penalty. For TARGET_INTER_UNIT_MOVES, we
can create:

        movd    a(,%eax,4), %xmm0
        movq    %xmm0, (%esp)
        fildll  (%esp)

instead of:

        movl    a(,%eax,4), %eax
        xorl    %edx, %edx
        pushl   %edx
        pushl   %eax
        fildll  (%esp)

Otherwise just use -mfpmat=sse.


-- 

ubizjak at gmail dot com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |ubizjak at gmail dot com
                   |dot org                     |
             Status|NEW                         |ASSIGNED
   Last reconfirmed|2006-01-16 06:17:55         |2008-03-21 13:45:57
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=13958