Re: Ping: [ARM] Loosen MODES_TIEABLE_P

2011-09-29 Thread Ramana Radhakrishnan
On 20 September 2011 11:12, Richard Sandiford wrote: > Ping for: > >    http://gcc.gnu.org/ml/gcc-patches/2011-09/msg00626.html > > which allows taking the vector subreg of a structure mode to be treated > as "cheap" (because direct register references are allowed).  At the > moment we treat them

Ping: [ARM] Loosen MODES_TIEABLE_P

2011-09-20 Thread Richard Sandiford
Ping for: http://gcc.gnu.org/ml/gcc-patches/2011-09/msg00626.html which allows taking the vector subreg of a structure mode to be treated as "cheap" (because direct register references are allowed). At the moment we treat them as expensive. Richard

[ARM] Loosen MODES_TIEABLE_P

2011-09-09 Thread Richard Sandiford
ARM's MODES_TIEABLE_P only allows classes of the same mode to be tied: #define MODES_TIEABLE_P(MODE1, MODE2) \ (GET_MODE_CLASS (MODE1) == GET_MODE_CLASS (MODE2)) But for NEON, we'd like structure modes to be tied to their vector elements. In particular, a vector subreg of a structure