On Saturday, May 30, 2015 08:11:40 AM Jason Ekstrand wrote: > On Sat, May 30, 2015 at 12:34 AM, Kenneth Graunke <kenn...@whitecape.org> > wrote: > > Now that Jason's LOAD_PAYLOAD improvements have landed, we don't need > > this. Passing 1 for the number of header registers already takes care > > of setting force_writemask_all on the header copy. > > > > Signed-off-by: Kenneth Graunke <kenn...@whitecape.org> > > --- > > src/mesa/drivers/dri/i965/brw_fs_visitor.cpp | 16 ++++------------ > > 1 file changed, 4 insertions(+), 12 deletions(-) > > > > diff --git a/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp > > b/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp > > index e336b73..c9a2644 100644 > > --- a/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp > > +++ b/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp > > @@ -1980,20 +1980,12 @@ fs_visitor::emit_urb_writes() > > fs_reg *payload_sources = ralloc_array(mem_ctx, fs_reg, length + > > 1); > > fs_reg payload = fs_reg(GRF, alloc.allocate(length + 1), > > BRW_REGISTER_TYPE_F, dispatch_width); > > - > > - /* We need WE_all on the MOV for the message header (the URB > > handles) > > - * so do a MOV to a dummy register and set force_writemask_all on > > the > > - * MOV. LOAD_PAYLOAD will preserve that. > > - */ > > Yeah, that's terrible. Thanks for cleaning this up.
Sure. I was in the area (making things more terrible) :D > > - fs_reg dummy = fs_reg(GRF, alloc.allocate(1), > > - BRW_REGISTER_TYPE_UD); > > - fs_inst *inst = emit(MOV(dummy, fs_reg(retype(brw_vec8_grf(1, 0), > > - > > BRW_REGISTER_TYPE_UD)))); > > - inst->force_writemask_all = true; > > - payload_sources[0] = dummy; > > + payload_sources[0] = > > + fs_reg(retype(brw_vec8_grf(1, 0), BRW_REGISTER_TYPE_UD)); > > > > memcpy(&payload_sources[1], sources, length * sizeof sources[0]); > > - emit(LOAD_PAYLOAD(payload, payload_sources, length + 1, 1)); > > + fs_inst *inst = > > + emit(LOAD_PAYLOAD(payload, payload_sources, length + 1, 1)); > > There's no need to save the return value here. We just throw it away > one line later. Let's move the fs_inst * declaration to the URB_WRITE > below. With that, > > Reviewed-by: Jason Ekstrand <ja...@jlekstrand.net> Oops. Yeah, I originally deleted the declaration, then went "it doesn't compile!" and put it back on some random instruction. Putting it on the URB_WRITE is clearly best :) Thanks!
signature.asc
Description: This is a digitally signed message part.
_______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
