AlexIndustrial/mesa
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

r600: Use nir_builder helpers for load/store_shared_r600

Browse Source 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24762>
This commit is contained in:
Faith Ekstrand
2023-08-17 14:01:42 -05:00
committed by Marge Bot
parent 28137a187b
commit c04f313cc5
1 changed files with 14 additions and 38 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/gallium/drivers/r600/sfn/sfn_nir_lower_tess_io.cpp
+14 -38
View File
@@ -277,17 +277,10 @@ emit_store_lds(nir_builder *b, nir_intrinsic_instr *op, nir_def *addr)
uint32_t writemask = wmask >> nir_intrinsic_component(op);
auto store_tcs_out =
nir_intrinsic_instr_create(b->shader, nir_intrinsic_store_local_shared_r600);
nir_intrinsic_set_write_mask(store_tcs_out, writemask);
store_tcs_out->src[0] = nir_src_for_ssa(op->src[0].ssa);
store_tcs_out->num_components = store_tcs_out->src[0].ssa->num_components;
bool start_even = (orig_writemask & (1u << (2 * i)));
auto addr2 = nir_iadd_imm(b, addr, 8 * i + (start_even ? 0 : 4));
store_tcs_out->src[1] = nir_src_for_ssa(addr2);
nir_builder_instr_insert(b, &store_tcs_out->instr);
nir_def *addr2 = nir_iadd_imm(b, addr, 8 * i + (start_even ? 0 : 4));
nir_store_local_shared_r600(b, op->src[0].ssa, addr2,
.write_mask = writemask);
}
}
@@ -419,21 +412,16 @@ r600_lower_tess_io_impl(nir_builder *b, nir_instr *instr, enum mesa_prim prim_ty
nir_def *addr_outer =
nir_iadd(b, addr0, load_offset_group(b, tf_inner_address_offset + ncomps));
auto tf =
nir_intrinsic_instr_create(b->shader, nir_intrinsic_load_local_shared_r600);
tf->num_components = ncomps;
tf->src[0] = nir_src_for_ssa(addr_outer);
nir_def_init(&tf->instr, &tf->def, tf->num_components, 32);
nir_builder_instr_insert(b, &tf->instr);
nir_def *tf = nir_load_local_shared_r600(b, 32, addr_outer);
if (ncomps < 4 && b->shader->info.stage != MESA_SHADER_TESS_EVAL) {
auto undef = nir_undef(b, 1, 32);
nir_def *srcs[4] = {undef, undef, undef, undef};
for (unsigned i = 0; i < ncomps; ++i)
srcs[i] = nir_channel(b, &tf->def, i);
srcs[i] = nir_channel(b, tf, i);
auto help = nir_vec(b, srcs, 4);
nir_def_rewrite_uses(&op->def, help);
} else {
nir_def_rewrite_uses(&op->def, &tf->def);
nir_def_rewrite_uses(&op->def, tf);
}
nir_instr_remove(instr);
return true;
@@ -523,13 +511,7 @@ r600_append_tcs_TF_emission(nir_shader *shader, enum mesa_prim prim_type)
nir_def *addr0 = r600_tcs_base_address(b, base, rel_patch_id);
nir_def *addr_outer = nir_iadd(b, addr0, load_offset_group(b, outer_comps));
auto tf_outer =
nir_intrinsic_instr_create(b->shader, nir_intrinsic_load_local_shared_r600);
tf_outer->num_components = outer_comps;
tf_outer->src[0] = nir_src_for_ssa(addr_outer);
nir_def_init(
&tf_outer->instr, &tf_outer->def, tf_outer->num_components, 32);
nir_builder_instr_insert(b, &tf_outer->instr);
nir_def *tf_outer = nir_load_local_shared_r600(b, 32, addr_outer);
std::vector<nir_def *> tf_out;
@@ -548,45 +530,39 @@ r600_append_tcs_TF_emission(nir_shader *shader, enum mesa_prim prim_type)
tf_out.push_back(nir_vec2(b,
out_addr0,
nir_channel(b, &tf_outer->def, chanx)));
nir_channel(b, tf_outer, chanx)));
tf_out.push_back(nir_vec2(b, nir_iadd_imm(b, out_addr0, 4),
nir_channel(b, &tf_outer->def, chany)));
nir_channel(b, tf_outer, chany)));
if (outer_comps > 2) {
tf_out.push_back(nir_vec2(b,
nir_iadd_imm(b, out_addr0, 8),
nir_channel(b, &tf_outer->def, 2)));
nir_channel(b, tf_outer, 2)));
}
if (outer_comps > 3) {
tf_out.push_back(nir_vec2(b,
nir_iadd_imm(b, out_addr0, 12),
nir_channel(b, &tf_outer->def, 3)));
nir_channel(b, tf_outer, 3)));
inner_base = 16;
}
if (inner_comps) {
nir_def *addr1 = nir_iadd(b, addr0, load_offset_group(b, 4 + inner_comps));
auto tf_inner =
nir_intrinsic_instr_create(b->shader, nir_intrinsic_load_local_shared_r600);
tf_inner->num_components = inner_comps;
tf_inner->src[0] = nir_src_for_ssa(addr1);
nir_def_init(
&tf_inner->instr, &tf_inner->def, tf_inner->num_components, 32);
nir_builder_instr_insert(b, &tf_inner->instr);
nir_def *tf_inner = nir_load_local_shared_r600(b, 32, addr1);
tf_out.push_back(nir_vec2(b,
nir_iadd_imm(b, out_addr0, inner_base),
nir_channel(b, &tf_inner->def, 0)));
nir_channel(b, tf_inner, 0)));
if (inner_comps > 1) {
tf_out.push_back(nir_vec2(b,
nir_iadd_imm(b, out_addr0, inner_base + 4),
nir_channel(b, &tf_inner->def, 1)));
nir_channel(b, tf_inner, 1)));
}
}