fdo-mirrors/mesa
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2026-05-06 02:58:05 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

i965/vec4: implement HW workaround for align16 double to float conversion

Browse source 
From the BDW PRM, Workarounds chapter:

   "DF->f format conversion for Align16 has wrong emask calculation when
    source is immediate."

Notice that Broadwell and later are strictly scalar at the moment though, so
this is not really necessary.

v2: Instead of moving the immediate to a vgrf and converting from there, just
    convert the double immediate to float in the compiler and move the result
    to the destination (Matt)

Reviewed-by: Matt Turner <mattst88@gmail.com>
This commit is contained in:
Iago Toral Quiroga 2016-06-29 13:08:25 +02:00 • committed by Samuel Iglesias Gonsálvez
parent bfc1f0f017
commit 4b22576234
1 changed files with 11 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

11
src/mesa/drivers/dri/i965/brw_vec4_nir.cpp
View file

@ -1071,6 +1071,17 @@ vec4_visitor::emit_conversion_from_double(dst_reg dst, src_reg src,
bool saturate,
brw_reg_type single_type)
{
/* BDW PRM vol 15 - workarounds:
* DF->f format conversion for Align16 has wrong emask calculation when
* source is immediate.
*/
if (devinfo->gen == 8 && single_type == BRW_REGISTER_TYPE_F &&
src.file == BRW_IMMEDIATE_VALUE) {
vec4_instruction *inst = emit(MOV(dst, brw_imm_f(src.df)));
inst->saturate = saturate;
return;
}
dst_reg temp = dst_reg(this, glsl_type::dvec4_type);
emit(MOV(temp, src));