[Bug rtl-optimization/50078] [4.6/4.7 Regression] combine wrong code: volatile accesses optimized out

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50078

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |wrong-code
             Status|ASSIGNED                    |NEW
          Component|tree-optimization           |rtl-optimization
      Known to work|                            |4.5.3
         AssignedTo|rguenth at gcc dot gnu.org  |unassigned at gcc dot
                   |                            |gnu.org
   Target Milestone|---                         |4.6.2
            Summary|[4.6.1+ SVN 2011-08-13] ARM |[4.6/4.7 Regression]
                   |wrong code: volatile        |combine wrong code:
                   |accesses optimized out      |volatile accesses optimized
                   |                            |out

--- Comment #3 from Richard Guenther <rguenth at gcc dot gnu.org> 2011-08-15 
09:11:50 UTC ---
Actually I was wrong guessing - the tree level is fine, it is combine that
removes the "noop" move completely:

Trying 6 -> 11:
Failed to match this instruction:
(set (mem/s:SI (plus:DI (ashiftrt:DI (mult:DI (subreg:DI (reg/v:SI 60 [ arg ])
0)
                    (const_int 4294967296 [0x100000000]))
                (const_int 30 [0x1e]))
            (symbol_ref:DI ("var")  <var_decl 0x7ffff7ee3140 var>)) [2 var S4
A32])
    (mem/s/v:SI (plus:DI (ashiftrt:DI (mult:DI (subreg:DI (reg/v:SI 60 [ arg ])
0)
                    (const_int 4294967296 [0x100000000]))
                (const_int 30 [0x1e]))
            (symbol_ref:DI ("var")  <var_decl 0x7ffff7ee3140 var>)) [2 var S4
A32]))
rejecting combination of insns 6 and 11
original costs 4 + 9 = 13
replacement cost 31
deleting noop move 11

Confirmed on x86_64 as well.