On 30 December 2014 at 08:14, Roland Scheidegger <srol...@vmware.com> wrote: > Just minor nits, looks good to me otherwise. > I agree with others that probably the round family of functions should > be added too (but could be done in another patch). Maybe could have one > cap bit then (so some implementations only doing what's required by sm5, > hence missing things like round and rsq and everybody else being able to > do everything). > > Roland > >> - const struct tgsi_full_src_register *reg, >> - const uint chan_index, >> - enum tgsi_exec_datatype src_datatype) >> +fetch_source_d(const struct tgsi_exec_machine *mach, > I think the _d in the name is a bit misleading here since this fetches > any type, not just floats. Unless this stands for something else...
Its one those extending an API functions, for non-doubles you should still be using fetch_source which wraps this. I'm not sure I can think of a meaningful name, fetch_source2 or fetch_source_ex, kinda sound as pointless. I'll leave it as-is for now as I'm not sure it really makes things worse. Dave. _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
