On Sun, 2014-05-04 at 17:39 -0400, David Edelsohn wrote: > On Fri, May 2, 2014 at 7:34 PM, Peter Bergner <berg...@vnet.ibm.com> wrote: > > @@ -2517,6 +2518,7 @@ extern int frame_pointer_needed; > > #define RS6000_BTM_POPCNTD MASK_POPCNTD /* Target supports ISA > > 2.06. */ > > #define RS6000_BTM_CELL MASK_FPRND /* Target is cell > > powerpc. */ > > #define RS6000_BTM_DFP MASK_DFP /* Decimal floating point. > > */ > > +#define RS6000_BTM_HARD_FLOAT MASK_SOFT_FLOAT /* Hardware floating point. > > */ > > It's a little confusing that the use reverses the polarity of the mask > meaning.
Agreed, but since there is no MASK_HARD_FLOAT, we're stuck with that. However, the use of the MASK here is only used to define which bit within the builtin bitmask corresponds to hard float. The actual setting of the builtin bitmask occurs below in rs6000_builtin_mask_calculate and there we use the BTM mask we defined above. > > Index: gcc/config/rs6000/rs6000.c > > =================================================================== > > --- gcc/config/rs6000/rs6000.c (revision 209990) > > +++ gcc/config/rs6000/rs6000.c (working copy) > > @@ -3039,7 +3039,8 @@ rs6000_builtin_mask_calculate (void) > > | ((TARGET_P8_VECTOR) ? RS6000_BTM_P8_VECTOR : 0) > > | ((TARGET_CRYPTO) ? RS6000_BTM_CRYPTO : 0) > > | ((TARGET_HTM) ? RS6000_BTM_HTM : 0) > > - | ((TARGET_DFP) ? RS6000_BTM_DFP : 0)); > > + | ((TARGET_DFP) ? RS6000_BTM_DFP : 0) > > + | ((TARGET_HARD_FLOAT) ? RS6000_BTM_HARD_FLOAT: 0)); > > This is missing a space between RS6000_BTM_HARD_FLOAT and ":". Fixed. > This is okay on trunk and all branches. Thanks, committed to trunk as revision 210054, 4.9 as revision 210055 and to 4.8 as revision 210056. Peter
