Previously, we generated this for FB writes in SIMD16 mode:
load_payload(16) vgrf5@8+0.0:F, vgrf1:F, vgrf2:F, vgrf3:F, vgrf4:F
fb_write(8) (null):UD, vgrf5@8+0.0:F 1sthalf
The LOAD_PAYLOAD's destination had its register width set to 8, and the
FB_WRITE had its execution size set to 8. This seems wrong, and while
it probably doesn't affect anything, we should fix it.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
fs_reg payload = fs_reg(GRF, -1, BRW_REGISTER_TYPE_F);
load = emit(LOAD_PAYLOAD(payload, sources, length));
payload.reg = virtual_grf_alloc(load->regs_written);
+ payload.width = dispatch_width;
load->dst = payload;
write = emit(FS_OPCODE_FB_WRITE, reg_undef, payload);
write->base_mrf = -1;
load = emit(LOAD_PAYLOAD(fs_reg(MRF, 1, BRW_REGISTER_TYPE_F),
sources, length));
write = emit(FS_OPCODE_FB_WRITE);
+ write->exec_size = dispatch_width;
write->base_mrf = 1;
}