i965/fs: fix compares for doubles

The destination has to have the same source as the type, or else the
simulator will complain. As a result, we need to emit a CMP that
outputs a 64-bit wide result and then do a strided MOV to pick out the
low 32 bits of each channel.

v2: Use subscript() instead of stride() (Curro)

Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
This commit is contained in:
Connor Abbott 2015-08-03 18:08:58 -07:00 committed by Samuel Iglesias Gonsálvez
parent a5d7e144ea
commit e83f51d54e

View file

@ -899,23 +899,51 @@ fs_visitor::nir_emit_alu(const fs_builder &bld, nir_alu_instr *instr)
}
case nir_op_flt:
case nir_op_fge:
case nir_op_feq:
case nir_op_fne: {
fs_reg dest = result;
if (nir_src_bit_size(instr->src[0].src) > 32) {
dest = bld.vgrf(BRW_REGISTER_TYPE_DF, 1);
}
brw_conditional_mod cond;
switch (instr->op) {
case nir_op_flt:
cond = BRW_CONDITIONAL_L;
break;
case nir_op_fge:
cond = BRW_CONDITIONAL_GE;
break;
case nir_op_feq:
cond = BRW_CONDITIONAL_Z;
break;
case nir_op_fne:
cond = BRW_CONDITIONAL_NZ;
break;
default:
unreachable("bad opcode");
}
bld.CMP(dest, op[0], op[1], cond);
if (nir_src_bit_size(instr->src[0].src) > 32) {
bld.MOV(result, subscript(dest, BRW_REGISTER_TYPE_UD, 0));
}
break;
}
case nir_op_ilt:
case nir_op_ult:
bld.CMP(result, op[0], op[1], BRW_CONDITIONAL_L);
break;
case nir_op_fge:
case nir_op_ige:
case nir_op_uge:
bld.CMP(result, op[0], op[1], BRW_CONDITIONAL_GE);
break;
case nir_op_feq:
case nir_op_ieq:
bld.CMP(result, op[0], op[1], BRW_CONDITIONAL_Z);
break;
case nir_op_fne:
case nir_op_ine:
bld.CMP(result, op[0], op[1], BRW_CONDITIONAL_NZ);
break;