/* * Copyright © 2021 Google * * SPDX-License-Identifier: MIT */ #include "nir/radv_nir_rt_common.h" #include "bvh/bvh.h" #include "nir_builder.h" #include "radv_debug.h" static nir_def *build_node_to_addr(struct radv_device *device, nir_builder *b, nir_def *node, bool skip_type_and); bool radv_use_bvh_stack_rtn(const struct radv_physical_device *pdevice) { /* gfx12 requires using the bvh4 ds_bvh_stack_rtn differently - enable hw stack instrs on gfx12 only with bvh8 */ return ((pdevice->info.gfx_level >= GFX11 && pdevice->info.gfx_level < GFX12) || pdevice->cache_key.bvh8) && !pdevice->cache_key.emulate_rt; } nir_def * radv_build_bvh_stack_rtn_addr(nir_builder *b, nir_def *stack_idx, const struct radv_physical_device *pdev, uint32_t workgroup_size, uint32_t stack_base, uint32_t max_stack_entries) { assert(stack_base % 4 == 0); /* RDNA3's ds_bvh_stack_rtn instruction uses a special encoding for the stack address. * Bits 0-17 encode the current stack index (set to 0 initially) * Bits 18-31 encodes the stack base in multiples of 4 * * The hardware uses a stride of 128 bytes (32 entries) for the stack index so the upper 32 threads need a different * base offset with wave64. */ if (workgroup_size > 32) { nir_def *wave32_thread_id = nir_iand_imm(b, stack_idx, 0x1f); nir_def *wave32_group_id = nir_ushr_imm(b, stack_idx, 5); uint32_t stack_entries_per_group = max_stack_entries * 32; nir_def *group_stack_base = nir_imul_imm(b, wave32_group_id, stack_entries_per_group); stack_idx = nir_iadd(b, wave32_thread_id, group_stack_base); } stack_idx = nir_iadd_imm(b, stack_idx, stack_base / 4); /* There are 4 bytes in each stack entry so no further arithmetic is needed. */ if (pdev->info.gfx_level >= GFX12) stack_idx = nir_ishl_imm(b, stack_idx, 15); else stack_idx = nir_ishl_imm(b, stack_idx, 18); return stack_idx; } static void nir_sort_hit_pair(nir_builder *b, nir_variable *var_distances, nir_variable *var_indices, uint32_t chan_1, uint32_t chan_2) { nir_def *ssa_distances = nir_load_var(b, var_distances); nir_def *ssa_indices = nir_load_var(b, var_indices); /* if (distances[chan_2] < distances[chan_1]) { */ nir_push_if(b, nir_flt(b, nir_channel(b, ssa_distances, chan_2), nir_channel(b, ssa_distances, chan_1))); { /* swap(distances[chan_2], distances[chan_1]); */ nir_def *new_distances[4] = {nir_undef(b, 1, 32), nir_undef(b, 1, 32), nir_undef(b, 1, 32), nir_undef(b, 1, 32)}; nir_def *new_indices[4] = {nir_undef(b, 1, 32), nir_undef(b, 1, 32), nir_undef(b, 1, 32), nir_undef(b, 1, 32)}; new_distances[chan_2] = nir_channel(b, ssa_distances, chan_1); new_distances[chan_1] = nir_channel(b, ssa_distances, chan_2); new_indices[chan_2] = nir_channel(b, ssa_indices, chan_1); new_indices[chan_1] = nir_channel(b, ssa_indices, chan_2); nir_store_var(b, var_distances, nir_vec(b, new_distances, 4), (1u << chan_1) | (1u << chan_2)); nir_store_var(b, var_indices, nir_vec(b, new_indices, 4), (1u << chan_1) | (1u << chan_2)); } /* } */ nir_pop_if(b, NULL); } static nir_def * intersect_ray_amd_software_box(struct radv_device *device, nir_builder *b, nir_def *bvh_node, nir_def *ray_tmax, nir_def *origin, nir_def *dir, nir_def *inv_dir) { const struct glsl_type *vec4_type = glsl_vector_type(GLSL_TYPE_FLOAT, 4); const struct glsl_type *uvec4_type = glsl_vector_type(GLSL_TYPE_UINT, 4); unsigned old_math_ctrl = b->fp_math_ctrl; b->fp_math_ctrl |= nir_fp_exact | nir_fp_preserve_nan | nir_fp_preserve_inf; nir_def *node_addr = build_node_to_addr(device, b, bvh_node, false); /* vec4 distances = vec4(INF, INF, INF, INF); */ nir_variable *distances = nir_variable_create(b->shader, nir_var_shader_temp, vec4_type, "distances"); nir_store_var(b, distances, nir_imm_vec4(b, INFINITY, INFINITY, INFINITY, INFINITY), 0xf); /* uvec4 child_indices = uvec4(0xffffffff, 0xffffffff, 0xffffffff, 0xffffffff); */ nir_variable *child_indices = nir_variable_create(b->shader, nir_var_shader_temp, uvec4_type, "child_indices"); nir_store_var(b, child_indices, nir_imm_ivec4(b, 0xffffffffu, 0xffffffffu, 0xffffffffu, 0xffffffffu), 0xf); /* Need to remove infinities here because otherwise we get nasty NaN propagation * if the direction has 0s in it. */ /* inv_dir = clamp(inv_dir, -FLT_MAX, FLT_MAX); */ inv_dir = nir_fclamp(b, inv_dir, nir_imm_float(b, -FLT_MAX), nir_imm_float(b, FLT_MAX)); for (int i = 0; i < 4; i++) { const uint32_t child_offset = offsetof(struct radv_bvh_box32_node, children[i]); const uint32_t coord_offsets[2] = { offsetof(struct radv_bvh_box32_node, coords[i].min.x), offsetof(struct radv_bvh_box32_node, coords[i].max.x), }; /* node->children[i] -> uint */ nir_def *child_index = nir_load_global(b, 1, 32, nir_iadd_imm(b, node_addr, child_offset), .align_mul = 64, .align_offset = child_offset % 64); /* node->coords[i][0], node->coords[i][1] -> vec3 */ nir_def *node_coords[2] = { nir_load_global(b, 3, 32, nir_iadd_imm(b, node_addr, coord_offsets[0]), .align_mul = 64, .align_offset = coord_offsets[0] % 64), nir_load_global(b, 3, 32, nir_iadd_imm(b, node_addr, coord_offsets[1]), .align_mul = 64, .align_offset = coord_offsets[1] % 64), }; /* If x of the aabb min is NaN, then this is an inactive aabb. * We don't need to care about any other components being NaN as that is UB. * https://docs.vulkan.org/spec/latest/chapters/accelstructures.html#acceleration-structure-inactive-prims */ nir_def *min_x = nir_channel(b, node_coords[0], 0); nir_def *min_x_is_not_nan = nir_inot(b, nir_fneu(b, min_x, min_x)); /* NaN != NaN -> true */ /* vec3 bound0 = (node->coords[i][0] - origin) * inv_dir; */ nir_def *bound0 = nir_fmul(b, nir_fsub(b, node_coords[0], origin), inv_dir); /* vec3 bound1 = (node->coords[i][1] - origin) * inv_dir; */ nir_def *bound1 = nir_fmul(b, nir_fsub(b, node_coords[1], origin), inv_dir); /* float tmin = max(max(min(bound0.x, bound1.x), min(bound0.y, bound1.y)), min(bound0.z, * bound1.z)); */ nir_def *tmin = nir_fmax(b, nir_fmax(b, nir_fmin(b, nir_channel(b, bound0, 0), nir_channel(b, bound1, 0)), nir_fmin(b, nir_channel(b, bound0, 1), nir_channel(b, bound1, 1))), nir_fmin(b, nir_channel(b, bound0, 2), nir_channel(b, bound1, 2))); /* float tmax = min(min(max(bound0.x, bound1.x), max(bound0.y, bound1.y)), max(bound0.z, * bound1.z)); */ nir_def *tmax = nir_fmin(b, nir_fmin(b, nir_fmax(b, nir_channel(b, bound0, 0), nir_channel(b, bound1, 0)), nir_fmax(b, nir_channel(b, bound0, 1), nir_channel(b, bound1, 1))), nir_fmax(b, nir_channel(b, bound0, 2), nir_channel(b, bound1, 2))); /* if (!isnan(node->coords[i][0].x) && tmax >= max(0.0f, tmin) && tmin < ray_tmax) { */ nir_push_if(b, nir_iand(b, min_x_is_not_nan, nir_iand(b, nir_fge(b, tmax, nir_fmax(b, nir_imm_float(b, 0.0f), tmin)), nir_flt(b, tmin, ray_tmax)))); { /* child_indices[i] = node->children[i]; */ nir_def *new_child_indices[4] = {child_index, child_index, child_index, child_index}; nir_store_var(b, child_indices, nir_vec(b, new_child_indices, 4), 1u << i); /* distances[i] = tmin; */ nir_def *new_distances[4] = {tmin, tmin, tmin, tmin}; nir_store_var(b, distances, nir_vec(b, new_distances, 4), 1u << i); } /* } */ nir_pop_if(b, NULL); } /* Sort our distances with a sorting network. */ nir_sort_hit_pair(b, distances, child_indices, 0, 1); nir_sort_hit_pair(b, distances, child_indices, 2, 3); nir_sort_hit_pair(b, distances, child_indices, 0, 2); nir_sort_hit_pair(b, distances, child_indices, 1, 3); nir_sort_hit_pair(b, distances, child_indices, 1, 2); b->fp_math_ctrl = old_math_ctrl; return nir_load_var(b, child_indices); } static nir_def * radv_build_intersect_edge(nir_builder *b, nir_def *v0_x, nir_def *v0_y, nir_def *v1_x, nir_def *v1_y) { /* Test (1 0 0) direction: t = */ nir_def *t_x = nir_fsub(b, v1_x, v0_x); nir_def *test_y = nir_feq_imm(b, t_x, 0.0); /* Test (0 1 0) direction: t = */ nir_def *t_y = nir_fsub(b, v1_y, v0_y); return nir_bcsel(b, test_y, nir_flt_imm(b, t_y, 0.0), nir_flt_imm(b, t_x, 0.0)); } static nir_def * radv_build_intersect_vertex(nir_builder *b, nir_def *v0_x, nir_def *v1_x, nir_def *v2_x) { /* Choose n=(1 0 0) to simplify the dot product. */ nir_def *edge0 = nir_fsub(b, v1_x, v0_x); nir_def *edge1 = nir_fsub(b, v2_x, v0_x); return nir_iand(b, nir_fle_imm(b, edge0, 0.0), nir_fgt_imm(b, edge1, 0.0)); } static nir_def * intersect_ray_amd_software_tri(struct radv_device *device, nir_builder *b, nir_def *bvh_node, nir_def *ray_tmax, nir_def *origin, nir_def *dir, nir_def *inv_dir) { const struct glsl_type *vec4_type = glsl_vector_type(GLSL_TYPE_FLOAT, 4); unsigned old_math_ctrl = b->fp_math_ctrl; b->fp_math_ctrl |= nir_fp_exact | nir_fp_preserve_nan | nir_fp_preserve_inf; nir_def *node_addr = build_node_to_addr(device, b, bvh_node, false); const uint32_t coord_offsets[3] = { offsetof(struct radv_bvh_triangle_node, coords[0]), offsetof(struct radv_bvh_triangle_node, coords[1]), offsetof(struct radv_bvh_triangle_node, coords[2]), }; /* node->coords[0], node->coords[1], node->coords[2] -> vec3 */ nir_def *node_coords[3] = { nir_load_global(b, 3, 32, nir_iadd_imm(b, node_addr, coord_offsets[0]), .align_mul = 64, .align_offset = coord_offsets[0] % 64), nir_load_global(b, 3, 32, nir_iadd_imm(b, node_addr, coord_offsets[1]), .align_mul = 64, .align_offset = coord_offsets[1] % 64), nir_load_global(b, 3, 32, nir_iadd_imm(b, node_addr, coord_offsets[2]), .align_mul = 64, .align_offset = coord_offsets[2] % 64), }; nir_variable *result = nir_variable_create(b->shader, nir_var_shader_temp, vec4_type, "result"); nir_store_var(b, result, nir_imm_vec4(b, INFINITY, 1.0f, 0.0f, 0.0f), 0xf); /* Based on watertight Ray/Triangle intersection from * http://jcgt.org/published/0002/01/05/paper.pdf */ /* Calculate the dimension where the ray direction is largest */ nir_def *abs_dir = nir_fabs(b, dir); nir_def *abs_dirs[3] = { nir_channel(b, abs_dir, 0), nir_channel(b, abs_dir, 1), nir_channel(b, abs_dir, 2), }; /* Find index of greatest value of abs_dir and put that as kz. */ nir_def *packed_k = nir_bcsel(b, nir_fge(b, abs_dirs[0], abs_dirs[1]), nir_bcsel(b, nir_fge(b, abs_dirs[0], abs_dirs[2]), nir_imm_int(b, (0 << 4) | (2 << 2) | (1 << 0)), nir_imm_int(b, (2 << 4) | (1 << 2) | (0 << 0))), nir_bcsel(b, nir_fge(b, abs_dirs[1], abs_dirs[2]), nir_imm_int(b, (1 << 4) | (0 << 2) | (2 << 0)), nir_imm_int(b, (2 << 4) | (1 << 2) | (0 << 0)))); nir_def *kx = nir_iand_imm(b, packed_k, 0x3); nir_def *ky = nir_ubfe_imm(b, packed_k, 2, 2); nir_def *kz = nir_ishr_imm(b, packed_k, 4); nir_def *k_indices[3] = {kx, ky, kz}; nir_def *k = nir_vec(b, k_indices, 3); /* Swap kx and ky dimensions to preserve winding order */ unsigned swap_xy_swizzle[4] = {1, 0, 2, 3}; k = nir_bcsel(b, nir_flt_imm(b, nir_vector_extract(b, dir, kz), 0.0f), nir_swizzle(b, k, swap_xy_swizzle, 3), k); kx = nir_channel(b, k, 0); ky = nir_channel(b, k, 1); kz = nir_channel(b, k, 2); /* Calculate shear constants */ nir_def *sz = nir_vector_extract(b, inv_dir, kz); nir_def *sx = nir_fmul(b, nir_vector_extract(b, dir, kx), sz); nir_def *sy = nir_fmul(b, nir_vector_extract(b, dir, ky), sz); /* Calculate vertices relative to ray origin */ nir_def *v_a = nir_fsub(b, node_coords[0], origin); nir_def *v_b = nir_fsub(b, node_coords[1], origin); nir_def *v_c = nir_fsub(b, node_coords[2], origin); /* Perform shear and scale */ nir_def *ax = nir_fsub(b, nir_vector_extract(b, v_a, kx), nir_fmul(b, sx, nir_vector_extract(b, v_a, kz))); nir_def *ay = nir_fsub(b, nir_vector_extract(b, v_a, ky), nir_fmul(b, sy, nir_vector_extract(b, v_a, kz))); nir_def *bx = nir_fsub(b, nir_vector_extract(b, v_b, kx), nir_fmul(b, sx, nir_vector_extract(b, v_b, kz))); nir_def *by = nir_fsub(b, nir_vector_extract(b, v_b, ky), nir_fmul(b, sy, nir_vector_extract(b, v_b, kz))); nir_def *cx = nir_fsub(b, nir_vector_extract(b, v_c, kx), nir_fmul(b, sx, nir_vector_extract(b, v_c, kz))); nir_def *cy = nir_fsub(b, nir_vector_extract(b, v_c, ky), nir_fmul(b, sy, nir_vector_extract(b, v_c, kz))); nir_def *u = nir_fsub(b, nir_fmul(b, cx, by), nir_fmul(b, cy, bx)); nir_def *v = nir_fsub(b, nir_fmul(b, ax, cy), nir_fmul(b, ay, cx)); nir_def *w = nir_fsub(b, nir_fmul(b, bx, ay), nir_fmul(b, by, ax)); /* Perform edge tests. */ nir_def *cond_back = nir_ior(b, nir_ior(b, nir_flt_imm(b, u, 0.0f), nir_flt_imm(b, v, 0.0f)), nir_flt_imm(b, w, 0.0f)); nir_def *cond_front = nir_ior(b, nir_ior(b, nir_fgt_imm(b, u, 0.0f), nir_fgt_imm(b, v, 0.0f)), nir_fgt_imm(b, w, 0.0f)); nir_def *cond = nir_inot(b, nir_iand(b, cond_back, cond_front)); /* When an edge is hit, we have to ensure that it is not hit twice in case it is shared. * * Vulkan 1.4.322, Section 40.1.1 Watertightness: * * Any set of two triangles with two shared vertices that were specified in the same * winding order in each triangle have a shared edge defined by those vertices. * * This means we can decide which triangle should intersect by comparing the shared edge * to two arbitrary directions because the shared edges are antiparallel. The triangle * vertices are transformed so the ray direction is (0 0 1). Therefore it makes sense to * choose (1 0 0) and (0 1 0) as reference directions. * * Hitting edges is extremely rare so an if should be worth. */ nir_def *is_edge_a = nir_feq_imm(b, u, 0.0f); nir_def *is_edge_b = nir_feq_imm(b, v, 0.0f); nir_def *is_edge_c = nir_feq_imm(b, w, 0.0f); nir_def *cond_edge = nir_ior(b, is_edge_a, nir_ior(b, is_edge_b, is_edge_c)); nir_def *intersect_edge = cond; nir_push_if(b, cond_edge); { nir_def *intersect_edge_a = nir_iand(b, is_edge_a, radv_build_intersect_edge(b, bx, by, cx, cy)); nir_def *intersect_edge_b = nir_iand(b, is_edge_b, radv_build_intersect_edge(b, cx, cy, ax, ay)); nir_def *intersect_edge_c = nir_iand(b, is_edge_c, radv_build_intersect_edge(b, ax, ay, bx, by)); intersect_edge = nir_iand(b, intersect_edge, nir_ior(b, nir_ior(b, intersect_edge_a, intersect_edge_b), intersect_edge_c)); /* For vertices, special handling is needed to avoid double hits. The spec defines * shared vertices as follows (Vulkan 1.4.322, Section 40.1.1 Watertightness): * * Any set of two or more triangles where all triangles have one vertex with an * identical position value, that vertex is a shared vertex. * * Since the no double hit/miss requirement of a shared vertex is only formulated for * closed fans * * Implementations should not double-hit or miss when a ray intersects a shared edge, * or a shared vertex of a closed fan. * * it is possible to choose an arbitrary direction n that defines which triangle in the * closed fan should intersect the shared vertex with the ray. * * All edges that include the above vertex are shared edges. * * Implies that all triangles have the same winding order. It is therefore sufficiant * to choose the triangle where the other vertices are on both sides of a plane * perpendicular to n (relying on winding order to get one instead of two triangles * that meet said condition). */ nir_def *is_vertex_a = nir_iand(b, is_edge_b, is_edge_c); nir_def *is_vertex_b = nir_iand(b, is_edge_a, is_edge_c); nir_def *is_vertex_c = nir_iand(b, is_edge_a, is_edge_b); nir_def *intersect_vertex_a = nir_iand(b, is_vertex_a, radv_build_intersect_vertex(b, ax, bx, cx)); nir_def *intersect_vertex_b = nir_iand(b, is_vertex_b, radv_build_intersect_vertex(b, bx, cx, ax)); nir_def *intersect_vertex_c = nir_iand(b, is_vertex_c, radv_build_intersect_vertex(b, cx, ax, bx)); nir_def *is_vertex = nir_ior(b, nir_ior(b, is_vertex_a, is_vertex_b), is_vertex_c); nir_def *intersect_vertex = nir_ior(b, nir_ior(b, intersect_vertex_a, intersect_vertex_b), intersect_vertex_c); intersect_vertex = nir_ior(b, nir_inot(b, is_vertex), intersect_vertex); intersect_edge = nir_iand(b, intersect_edge, intersect_vertex); } nir_pop_if(b, NULL); cond = nir_if_phi(b, intersect_edge, cond); nir_push_if(b, cond); { nir_def *det = nir_fadd(b, u, nir_fadd(b, v, w)); nir_def *az = nir_fmul(b, sz, nir_vector_extract(b, v_a, kz)); nir_def *bz = nir_fmul(b, sz, nir_vector_extract(b, v_b, kz)); nir_def *cz = nir_fmul(b, sz, nir_vector_extract(b, v_c, kz)); nir_def *t = nir_fadd(b, nir_fadd(b, nir_fmul(b, u, az), nir_fmul(b, v, bz)), nir_fmul(b, w, cz)); nir_def *t_signed = nir_fmul(b, nir_fsign(b, det), t); nir_def *det_cond_front = nir_inot(b, nir_flt_imm(b, t_signed, 0.0f)); nir_push_if(b, det_cond_front); { nir_def *det_abs = nir_fabs(b, det); t = nir_fdiv(b, t, det_abs); v = nir_fdiv(b, v, det_abs); w = nir_fdiv(b, w, det_abs); nir_def *indices[4] = {t, nir_fsign(b, det), v, w}; nir_store_var(b, result, nir_vec(b, indices, 4), 0xf); } nir_pop_if(b, NULL); } nir_pop_if(b, NULL); b->fp_math_ctrl = old_math_ctrl; return nir_load_var(b, result); } nir_def * build_addr_to_node(struct radv_device *device, nir_builder *b, nir_def *addr, nir_def *flags) { const struct radv_physical_device *pdev = radv_device_physical(device); const uint64_t bvh_size = 1ull << 42; nir_def *node = nir_ushr_imm(b, addr, 3); node = nir_iand_imm(b, node, (bvh_size - 1) << 3); if (pdev->cache_key.bvh8) { /* The HW ray flags are the same bits as the API flags. * - SpvRayFlagsTerminateOnFirstHitKHRMask, SpvRayFlagsSkipClosestHitShaderKHRMask are handled in shader code. * - SpvRayFlagsSkipTrianglesKHRMask, SpvRayFlagsSkipAABBsKHRMask do not work. */ flags = nir_iand_imm(b, flags, SpvRayFlagsOpaqueKHRMask | SpvRayFlagsNoOpaqueKHRMask | SpvRayFlagsCullBackFacingTrianglesKHRMask | SpvRayFlagsCullFrontFacingTrianglesKHRMask | SpvRayFlagsCullOpaqueKHRMask | SpvRayFlagsCullNoOpaqueKHRMask); node = nir_ior(b, node, nir_ishl_imm(b, nir_u2u64(b, flags), 54)); } return node; } static nir_def * build_node_to_addr(struct radv_device *device, nir_builder *b, nir_def *node, bool skip_type_and) { const struct radv_physical_device *pdev = radv_device_physical(device); nir_def *addr = skip_type_and ? node : nir_iand_imm(b, node, ~7ull); addr = nir_ishl_imm(b, addr, 3); /* Assumes everything is in the top half of address space, which is true in * GFX9+ for now. */ return pdev->info.gfx_level >= GFX9 ? nir_ior_imm(b, addr, 0xffffull << 48) : addr; } nir_def * nir_build_vec3_mat_mult(nir_builder *b, nir_def *vec, nir_def *matrix[], bool translation) { nir_def *result_components[3] = { nir_channel(b, matrix[0], 3), nir_channel(b, matrix[1], 3), nir_channel(b, matrix[2], 3), }; for (unsigned i = 0; i < 3; ++i) { for (unsigned j = 0; j < 3; ++j) { nir_def *v = nir_fmul(b, nir_channels(b, vec, 1 << j), nir_channels(b, matrix[i], 1 << j)); result_components[i] = (translation || j) ? nir_fadd(b, result_components[i], v) : v; } } return nir_vec(b, result_components, 3); } nir_def * radv_load_vertex_position(struct radv_device *device, nir_builder *b, nir_def *primitive_addr, uint32_t index) { const struct radv_physical_device *pdev = radv_device_physical(device); if (pdev->cache_key.bvh8) { /* Assume that vertices are uncompressed. */ uint32_t offset = ROUND_DOWN_TO(RADV_GFX12_PRIMITIVE_NODE_HEADER_SIZE / 8, 4) + index * 3 * sizeof(float); nir_def *data[4]; for (uint32_t i = 0; i < ARRAY_SIZE(data); i++) { data[i] = nir_load_global(b, 1, 32, nir_iadd_imm(b, primitive_addr, offset)); offset += 4; } uint32_t subdword_offset = RADV_GFX12_PRIMITIVE_NODE_HEADER_SIZE % 32; nir_def *vertices[3]; for (uint32_t i = 0; i < ARRAY_SIZE(vertices); i++) { nir_def *lo = nir_ubitfield_extract_imm(b, data[i], subdword_offset, 32 - subdword_offset); nir_def *hi = nir_ubitfield_extract_imm(b, data[i + 1], 0, subdword_offset); vertices[i] = nir_ior(b, lo, nir_ishl_imm(b, hi, 32 - subdword_offset)); } return nir_vec3(b, vertices[0], vertices[1], vertices[2]); } uint32_t offset = index * 3 * sizeof(float); return nir_load_global(b, 3, 32, nir_iadd_imm(b, primitive_addr, offset)); } void radv_load_wto_matrix(struct radv_device *device, nir_builder *b, nir_def *instance_addr, nir_def **out) { const struct radv_physical_device *pdev = radv_device_physical(device); unsigned offset = offsetof(struct radv_bvh_instance_node, wto_matrix); if (pdev->cache_key.bvh8) offset = offsetof(struct radv_gfx12_instance_node, wto_matrix); for (unsigned i = 0; i < 3; ++i) { out[i] = nir_load_global(b, 4, 32, nir_iadd_imm(b, instance_addr, offset + i * 16), .align_mul = 64, .align_offset = (offset + i * 16) % 64); } } void radv_load_otw_matrix(struct radv_device *device, nir_builder *b, nir_def *instance_addr, nir_def **out) { const struct radv_physical_device *pdev = radv_device_physical(device); unsigned offset = offsetof(struct radv_bvh_instance_node, otw_matrix); if (pdev->cache_key.bvh8) offset = sizeof(struct radv_gfx12_instance_node) + offsetof(struct radv_gfx12_instance_node_user_data, otw_matrix); for (unsigned i = 0; i < 3; ++i) { out[i] = nir_load_global(b, 4, 32, nir_iadd_imm(b, instance_addr, offset + i * 16), .align_mul = 64, .align_offset = (offset + i * 16) % 64); } } nir_def * radv_load_custom_instance(struct radv_device *device, nir_builder *b, nir_def *instance_addr) { const struct radv_physical_device *pdev = radv_device_physical(device); if (pdev->cache_key.bvh8) { return nir_load_global(b, 1, 32, nir_iadd_imm(b, instance_addr, sizeof(struct radv_gfx12_instance_node) + offsetof(struct radv_gfx12_instance_node_user_data, custom_instance))); } return nir_iand_imm( b, nir_load_global( b, 1, 32, nir_iadd_imm(b, instance_addr, offsetof(struct radv_bvh_instance_node, custom_instance_and_mask))), 0xFFFFFF); } nir_def * radv_load_instance_id(struct radv_device *device, nir_builder *b, nir_def *instance_addr) { const struct radv_physical_device *pdev = radv_device_physical(device); if (pdev->cache_key.bvh8) { return nir_load_global(b, 1, 32, nir_iadd_imm(b, instance_addr, sizeof(struct radv_gfx12_instance_node) + offsetof(struct radv_gfx12_instance_node_user_data, instance_index))); } return nir_load_global(b, 1, 32, nir_iadd_imm(b, instance_addr, offsetof(struct radv_bvh_instance_node, instance_id))); } /* When a hit is opaque the any_hit shader is skipped for this hit and the hit * is assumed to be an actual hit. */ static nir_def * hit_is_opaque(nir_builder *b, nir_def *sbt_offset_and_flags, const struct radv_ray_flags *ray_flags, nir_def *geometry_id_and_flags) { nir_def *opaque = nir_uge_imm(b, nir_ior(b, geometry_id_and_flags, sbt_offset_and_flags), RADV_INSTANCE_FORCE_OPAQUE | RADV_INSTANCE_NO_FORCE_NOT_OPAQUE); opaque = nir_bcsel(b, ray_flags->force_opaque, nir_imm_true(b), opaque); opaque = nir_bcsel(b, ray_flags->force_not_opaque, nir_imm_false(b), opaque); return opaque; } static nir_def * create_bvh_descriptor(nir_builder *b, const struct radv_physical_device *pdev, struct radv_ray_flags *ray_flags) { /* We create a BVH descriptor that covers the entire memory range. That way we can always * use the same descriptor, which avoids divergence when different rays hit different * instances at the cost of having to use 64-bit node ids. */ const uint64_t bvh_size = 1ull << 42; const uint32_t sort_triangles_first = pdev->cache_key.bvh8 ? BITFIELD_BIT(52 - 32) : 0; const uint32_t box_sort_enable = BITFIELD_BIT(63 - 32); const uint32_t triangle_return_mode = BITFIELD_BIT(120 - 96); /* Return IJ for triangles */ uint32_t dword0 = 0; nir_def *dword1 = nir_imm_intN_t(b, sort_triangles_first | box_sort_enable, 32); uint32_t dword2 = (bvh_size - 1) & 0xFFFFFFFFu; uint32_t dword3 = ((bvh_size - 1) >> 32) | triangle_return_mode | (1u << 31); if (pdev->info.gfx_level >= GFX11) { /* Enable pointer flags on GFX11+ */ dword3 |= BITFIELD_BIT(119 - 96); /* Instead of the default box sorting (closest point), use largest for terminate_on_first_hit rays; * this makes it more likely that the ray traversal will visit fewer nodes. */ const uint32_t box_sort_largest = 1; /* Only use largest sorting when all invocations have the same ray flags, otherwise * fall back to the default closest point. */ dword1 = nir_bcsel(b, nir_vote_all(b, 1, ray_flags->terminate_on_first_hit), nir_imm_int(b, (box_sort_largest << 21) | sort_triangles_first | box_sort_enable), dword1); } if (pdev->cache_key.bvh8) { /* compressed_format_en */ dword3 |= BITFIELD_BIT(115 - 96); /* wide_sort_en */ dword3 |= BITFIELD_BIT(117 - 96); /* instance_en */ dword3 |= BITFIELD_BIT(118 - 96); } return nir_vec4(b, nir_imm_intN_t(b, dword0, 32), dword1, nir_imm_intN_t(b, dword2, 32), nir_imm_intN_t(b, dword3, 32)); } static void insert_traversal_triangle_case(struct radv_device *device, nir_builder *b, const struct radv_ray_traversal_args *args, const struct radv_ray_flags *ray_flags, nir_def *result, nir_def *bvh_node) { struct radv_physical_device *pdev = radv_device_physical(device); if (!args->triangle_cb) return; struct radv_triangle_intersection intersection; intersection.t = nir_channel(b, result, 0); nir_def *div = nir_channel(b, result, 1); intersection.t = nir_fdiv(b, intersection.t, div); nir_def *tmax = nir_load_deref(b, args->vars.tmax); nir_push_if(b, nir_flt(b, intersection.t, tmax)); { intersection.frontface = nir_fgt_imm(b, div, 0); nir_def *not_cull; if (pdev->info.gfx_level < GFX11 || pdev->cache_key.emulate_rt) { nir_def *switch_ccw = nir_test_mask(b, nir_load_deref(b, args->vars.sbt_offset_and_flags), RADV_INSTANCE_TRIANGLE_FLIP_FACING); intersection.frontface = nir_ixor(b, intersection.frontface, switch_ccw); not_cull = ray_flags->no_skip_triangles; nir_def *not_facing_cull = nir_bcsel(b, intersection.frontface, ray_flags->no_cull_front, ray_flags->no_cull_back); not_cull = nir_iand(b, not_cull, nir_ior(b, not_facing_cull, nir_test_mask(b, nir_load_deref(b, args->vars.sbt_offset_and_flags), RADV_INSTANCE_TRIANGLE_FACING_CULL_DISABLE))); } else { not_cull = nir_imm_true(b); } nir_push_if(b, nir_iand(b, nir_flt(b, args->tmin, intersection.t), not_cull)); { intersection.base.node_addr = build_node_to_addr(device, b, bvh_node, false); nir_def *triangle_info = nir_load_global( b, 2, 32, nir_iadd_imm(b, intersection.base.node_addr, offsetof(struct radv_bvh_triangle_node, triangle_id))); intersection.base.primitive_id = nir_channel(b, triangle_info, 0); intersection.base.geometry_id_and_flags = nir_channel(b, triangle_info, 1); intersection.base.opaque = hit_is_opaque(b, nir_load_deref(b, args->vars.sbt_offset_and_flags), ray_flags, intersection.base.geometry_id_and_flags); not_cull = nir_bcsel(b, intersection.base.opaque, ray_flags->no_cull_opaque, ray_flags->no_cull_no_opaque); nir_push_if(b, not_cull); { nir_def *divs[2] = {div, div}; intersection.barycentrics = nir_fdiv(b, nir_channels(b, result, 0xc), nir_vec(b, divs, 2)); args->triangle_cb(b, &intersection, args, ray_flags); } nir_pop_if(b, NULL); } nir_pop_if(b, NULL); } nir_pop_if(b, NULL); } static void insert_traversal_triangle_case_gfx12(struct radv_device *device, nir_builder *b, const struct radv_ray_traversal_args *args, const struct radv_ray_flags *ray_flags, nir_variable *intrinsic_result, nir_def *result, nir_def *global_bvh_node, nir_def *bvh_node) { if (!args->triangle_cb) return; nir_def *t[2] = { nir_channel(b, result, 0), nir_channel(b, result, 4), }; nir_def *triangle0_first = nir_flt(b, t[0], t[1]); nir_def *second_t = nir_bcsel(b, triangle0_first, t[1], t[0]); nir_def *second_iteration = nir_load_deref(b, args->vars.second_iteration); nir_def *revisit = nir_iand(b, nir_inot(b, second_iteration), nir_flt(b, second_t, nir_load_deref(b, args->vars.tmax))); nir_store_deref(b, args->vars.second_iteration, revisit, 0x1); if (args->use_bvh_stack_rtn) { nir_def *next_node = nir_bcsel(b, revisit, bvh_node, nir_imm_int(b, RADV_BVH_STACK_SKIP_0_TO_7)); nir_def *comps[8]; for (unsigned i = 0; i < 6; ++i) comps[i] = nir_channel(b, result, i); comps[6] = nir_imm_int(b, RADV_BVH_STACK_SKIP_0_TO_7); comps[7] = next_node; nir_store_var(b, intrinsic_result, nir_vec(b, comps, 8), 0xff); } else { nir_def *next_node = nir_bcsel(b, revisit, bvh_node, nir_imm_int(b, RADV_BVH_INVALID_NODE)); nir_store_deref(b, args->vars.current_node, next_node, 0x1); } nir_def *triangle0 = nir_ixor(b, triangle0_first, second_iteration); struct radv_triangle_intersection intersection; intersection.t = nir_bcsel(b, triangle0, t[0], t[1]); nir_push_if(b, nir_iand(b, nir_flt(b, intersection.t, nir_load_deref(b, args->vars.tmax)), nir_flt(b, args->tmin, intersection.t))); { nir_def *dword1 = nir_bcsel(b, triangle0, nir_channel(b, result, 1), nir_channel(b, result, 5)); nir_def *dword2 = nir_bcsel(b, triangle0, nir_channel(b, result, 2), nir_channel(b, result, 6)); nir_def *dword3 = nir_bcsel(b, triangle0, nir_channel(b, result, 3), nir_channel(b, result, 7)); intersection.frontface = nir_inot(b, nir_test_mask(b, dword3, 1)); intersection.base.node_addr = build_node_to_addr(device, b, global_bvh_node, false); intersection.base.primitive_id = nir_ishr_imm(b, dword3, 1); intersection.base.geometry_id_and_flags = nir_ishr_imm(b, nir_bcsel(b, triangle0, nir_channel(b, result, 8), nir_channel(b, result, 9)), 2); intersection.base.opaque = nir_inot(b, nir_test_mask(b, dword2, 1u << 31)); intersection.barycentrics = nir_fabs(b, nir_vec2(b, dword1, dword2)); nir_push_if(b, nir_bcsel(b, intersection.base.opaque, ray_flags->no_cull_opaque, ray_flags->no_cull_no_opaque)); { args->triangle_cb(b, &intersection, args, ray_flags); } nir_pop_if(b, NULL); } nir_pop_if(b, NULL); } static void insert_traversal_aabb_case(struct radv_device *device, nir_builder *b, const struct radv_ray_traversal_args *args, const struct radv_ray_flags *ray_flags, nir_def *bvh_node) { if (!args->aabb_cb) return; nir_push_if(b, ray_flags->no_skip_aabbs); { struct radv_leaf_intersection intersection; intersection.node_addr = build_node_to_addr(device, b, bvh_node, false); nir_def *triangle_info = nir_load_global( b, 2, 32, nir_iadd_imm(b, intersection.node_addr, offsetof(struct radv_bvh_aabb_node, primitive_id))); intersection.primitive_id = nir_channel(b, triangle_info, 0); intersection.geometry_id_and_flags = nir_channel(b, triangle_info, 1); intersection.opaque = hit_is_opaque(b, nir_load_deref(b, args->vars.sbt_offset_and_flags), ray_flags, intersection.geometry_id_and_flags); nir_push_if(b, nir_bcsel(b, intersection.opaque, ray_flags->no_cull_opaque, ray_flags->no_cull_no_opaque)); { args->aabb_cb(b, &intersection, args); } nir_pop_if(b, NULL); } nir_pop_if(b, NULL); } static void insert_traversal_aabb_case_gfx12(struct radv_device *device, nir_builder *b, const struct radv_ray_traversal_args *args, const struct radv_ray_flags *ray_flags, nir_def *result, nir_def *bvh_node) { if (!args->aabb_cb) return; struct radv_leaf_intersection intersection; intersection.node_addr = build_node_to_addr(device, b, bvh_node, false); intersection.primitive_id = nir_ishr_imm(b, nir_channel(b, result, 3), 1); intersection.geometry_id_and_flags = nir_ishr_imm(b, nir_channel(b, result, 8), 2); intersection.opaque = nir_inot(b, nir_test_mask(b, nir_channel(b, result, 2), 1u << 31)); nir_push_if(b, nir_bcsel(b, intersection.opaque, ray_flags->no_cull_opaque, ray_flags->no_cull_no_opaque)); { args->aabb_cb(b, &intersection, args); } nir_pop_if(b, NULL); } static nir_def * fetch_parent_node(struct radv_device *device, nir_builder *b, nir_def *bvh, nir_def *node) { const struct radv_physical_device *pdev = radv_device_physical(device); nir_def *offset = nir_iadd_imm(b, nir_imul_imm(b, nir_udiv_imm(b, node, pdev->cache_key.bvh8 ? 16 : 8), 4), 4); return nir_load_global(b, 1, 32, nir_isub(b, bvh, nir_u2u64(b, offset)), .align_mul = 4); } static nir_def * radv_test_flag(nir_builder *b, const struct radv_ray_traversal_args *args, uint32_t flag, bool set) { nir_def *result; if (args->set_flags & flag) result = nir_imm_true(b); else if (args->unset_flags & flag) result = nir_imm_false(b); else result = nir_test_mask(b, args->flags, flag); return set ? result : nir_inot(b, result); } static nir_def * build_bvh_base(nir_builder *b, const struct radv_physical_device *pdev, nir_def *base_addr, nir_def *ptr_flags, bool overwrite) { if (pdev->info.gfx_level < GFX11 || pdev->cache_key.emulate_rt) return base_addr; nir_def *base_addr_vec = nir_unpack_64_2x32(b, base_addr); nir_def *addr_hi = nir_channel(b, base_addr_vec, 1); if (overwrite) addr_hi = nir_bitfield_insert(b, addr_hi, ptr_flags, nir_imm_int(b, 22), nir_imm_int(b, 10)); else addr_hi = nir_ior(b, addr_hi, nir_ishl_imm(b, ptr_flags, 22)); return nir_pack_64_2x32(b, nir_vector_insert_imm(b, base_addr_vec, addr_hi, 1)); } static void build_instance_exit(nir_builder *b, const struct radv_physical_device *pdev, const struct radv_ray_traversal_args *args, nir_def *stack_instance_exit, nir_def *ptr_flags) { nir_def *root_instance_exit = nir_iand( b, nir_ieq_imm(b, nir_load_deref(b, args->vars.current_node), RADV_BVH_INVALID_NODE), nir_ieq(b, nir_load_deref(b, args->vars.previous_node), nir_load_deref(b, args->vars.instance_bottom_node))); nir_if *instance_exit = nir_push_if(b, nir_ior(b, stack_instance_exit, root_instance_exit)); instance_exit->control = nir_selection_control_dont_flatten; { if (pdev->cache_key.bvh8 && args->use_bvh_stack_rtn) nir_store_deref(b, args->vars.stack, nir_ior_imm(b, nir_load_deref(b, args->vars.stack), RADV_BVH_STACK_FLAG_TLAS_POP), 0x1); else nir_store_deref(b, args->vars.top_stack, nir_imm_int(b, -1), 1); nir_store_deref(b, args->vars.previous_node, nir_load_deref(b, args->vars.instance_top_node), 1); nir_store_deref(b, args->vars.instance_bottom_node, nir_imm_int(b, RADV_BVH_NO_INSTANCE_ROOT), 1); nir_def *root_bvh_base = pdev->cache_key.bvh8 ? args->root_bvh_base : build_bvh_base(b, pdev, args->root_bvh_base, ptr_flags, true); nir_store_deref(b, args->vars.bvh_base, root_bvh_base, 0x1); nir_store_deref(b, args->vars.origin, args->origin, 7); nir_store_deref(b, args->vars.dir, args->dir, 7); nir_store_deref(b, args->vars.inv_dir, nir_frcp(b, args->dir), 7); } nir_pop_if(b, NULL); } nir_def * radv_build_ray_traversal(struct radv_device *device, nir_builder *b, const struct radv_ray_traversal_args *args) { const struct radv_physical_device *pdev = radv_device_physical(device); nir_variable *incomplete = nir_local_variable_create(b->impl, glsl_bool_type(), "incomplete"); nir_store_var(b, incomplete, nir_imm_true(b), 0x1); nir_variable *intrinsic_result = nir_local_variable_create(b->impl, glsl_uvec4_type(), "intrinsic_result"); nir_variable *last_visited_node = nir_local_variable_create(b->impl, glsl_uint_type(), "last_visited_node"); struct radv_ray_flags ray_flags = { .force_opaque = radv_test_flag(b, args, SpvRayFlagsOpaqueKHRMask, true), .force_not_opaque = radv_test_flag(b, args, SpvRayFlagsNoOpaqueKHRMask, true), .terminate_on_first_hit = radv_test_flag(b, args, SpvRayFlagsTerminateOnFirstHitKHRMask, true), .no_cull_front = radv_test_flag(b, args, SpvRayFlagsCullFrontFacingTrianglesKHRMask, false), .no_cull_back = radv_test_flag(b, args, SpvRayFlagsCullBackFacingTrianglesKHRMask, false), .no_cull_opaque = radv_test_flag(b, args, SpvRayFlagsCullOpaqueKHRMask, false), .no_cull_no_opaque = radv_test_flag(b, args, SpvRayFlagsCullNoOpaqueKHRMask, false), .no_skip_triangles = radv_test_flag(b, args, SpvRayFlagsSkipTrianglesKHRMask, false), .no_skip_aabbs = radv_test_flag(b, args, SpvRayFlagsSkipAABBsKHRMask, false), }; nir_def *ptr_flags = nir_iand_imm(b, args->flags, ~(SpvRayFlagsTerminateOnFirstHitKHRMask | SpvRayFlagsSkipClosestHitShaderKHRMask)); nir_store_deref(b, args->vars.bvh_base, build_bvh_base(b, pdev, nir_load_deref(b, args->vars.bvh_base), ptr_flags, true), 0x1); nir_def *desc = create_bvh_descriptor(b, pdev, &ray_flags); nir_def *vec3ones = nir_imm_vec3(b, 1.0, 1.0, 1.0); nir_loop *loop = nir_push_loop(b); { if (!args->use_bvh_stack_rtn) nir_loop_add_continue_construct(loop); /* When exiting instances via stack, current_node won't ever be invalid with ds_bvh_stack_rtn */ if (args->use_bvh_stack_rtn) { /* Early-exit when the stack is empty and there are no more nodes to process. */ nir_push_if(b, nir_ieq_imm(b, nir_load_deref(b, args->vars.current_node), RADV_BVH_STACK_TERMINAL_NODE)); { nir_store_var(b, incomplete, nir_imm_false(b), 0x1); nir_jump(b, nir_jump_break); } nir_pop_if(b, NULL); build_instance_exit(b, pdev, args, nir_ilt(b, nir_load_deref(b, args->vars.stack), nir_load_deref(b, args->vars.top_stack)), ptr_flags); } nir_push_if(b, nir_ieq_imm(b, nir_load_deref(b, args->vars.current_node), RADV_BVH_INVALID_NODE)); { /* Early exit if we never overflowed the stack, to avoid having to backtrack to * the root for no reason. */ if (!args->use_bvh_stack_rtn) { nir_push_if(b, nir_ilt_imm(b, nir_load_deref(b, args->vars.stack), args->stack_stride)); { nir_store_var(b, incomplete, nir_imm_false(b), 0x1); nir_jump(b, nir_jump_break); } nir_pop_if(b, NULL); build_instance_exit( b, pdev, args, nir_ige(b, nir_load_deref(b, args->vars.top_stack), nir_load_deref(b, args->vars.stack)), ptr_flags); } nir_def *overflow_cond = nir_ige(b, nir_load_deref(b, args->vars.stack_low_watermark), nir_load_deref(b, args->vars.stack)); /* ds_bvh_stack_rtn returns 0xFFFFFFFF if and only if there was a stack overflow. */ if (args->use_bvh_stack_rtn) overflow_cond = nir_imm_true(b); nir_push_if(b, overflow_cond); { /* Fix up the stack pointer if we overflowed. The HW will decrement the stack pointer by one in that case. */ if (args->use_bvh_stack_rtn) nir_store_deref(b, args->vars.stack, nir_iadd_imm(b, nir_load_deref(b, args->vars.stack), 1), 0x1); nir_def *prev = nir_load_deref(b, args->vars.previous_node); nir_def *bvh_addr = build_node_to_addr(device, b, nir_load_deref(b, args->vars.bvh_base), true); nir_def *parent = fetch_parent_node(device, b, bvh_addr, prev); nir_push_if(b, nir_ieq_imm(b, parent, RADV_BVH_INVALID_NODE)); { nir_store_var(b, incomplete, nir_imm_false(b), 0x1); nir_jump(b, nir_jump_break); } nir_pop_if(b, NULL); nir_store_deref(b, args->vars.current_node, parent, 0x1); } nir_push_else(b, NULL); { if (!args->use_bvh_stack_rtn) { nir_store_deref(b, args->vars.stack, nir_iadd_imm(b, nir_load_deref(b, args->vars.stack), -args->stack_stride), 1); nir_def *stack_ptr = nir_umod_imm(b, nir_load_deref(b, args->vars.stack), args->stack_stride * args->stack_entries); nir_def *bvh_node = args->stack_load_cb(b, stack_ptr, args); nir_store_deref(b, args->vars.current_node, bvh_node, 0x1); } nir_store_deref(b, args->vars.previous_node, nir_imm_int(b, RADV_BVH_INVALID_NODE), 0x1); } nir_pop_if(b, NULL); } nir_push_else(b, NULL); { nir_store_deref(b, args->vars.previous_node, nir_imm_int(b, RADV_BVH_INVALID_NODE), 0x1); } nir_pop_if(b, NULL); nir_def *bvh_node = nir_load_deref(b, args->vars.current_node); if (args->use_bvh_stack_rtn) nir_store_var(b, last_visited_node, nir_imm_int(b, RADV_BVH_STACK_TERMINAL_NODE), 0x1); else nir_store_deref(b, args->vars.current_node, nir_imm_int(b, RADV_BVH_INVALID_NODE), 0x1); nir_def *prev_node = nir_load_deref(b, args->vars.previous_node); nir_store_deref(b, args->vars.previous_node, bvh_node, 0x1); nir_def *global_bvh_node = nir_iadd(b, nir_load_deref(b, args->vars.bvh_base), nir_u2u64(b, bvh_node)); bool has_result = false; if (pdev->info.compiler_info.has_image_bvh_intersect_ray && !pdev->cache_key.emulate_rt) { nir_store_var( b, intrinsic_result, nir_bvh64_intersect_ray_amd(b, 32, desc, nir_unpack_64_2x32(b, global_bvh_node), nir_load_deref(b, args->vars.tmax), nir_load_deref(b, args->vars.origin), nir_load_deref(b, args->vars.dir), nir_load_deref(b, args->vars.inv_dir)), 0xf); has_result = true; } nir_push_if(b, nir_test_mask(b, bvh_node, BITFIELD64_BIT(ffs(radv_bvh_node_box16) - 1))); { nir_push_if(b, nir_test_mask(b, bvh_node, BITFIELD64_BIT(ffs(radv_bvh_node_instance) - 1))); { nir_push_if(b, nir_test_mask(b, bvh_node, BITFIELD64_BIT(ffs(radv_bvh_node_aabb) - 1))); { insert_traversal_aabb_case(device, b, args, &ray_flags, global_bvh_node); } nir_push_else(b, NULL); { if (args->vars.iteration_instance_count) { nir_def *iteration_instance_count = nir_load_deref(b, args->vars.iteration_instance_count); iteration_instance_count = nir_iadd_imm(b, iteration_instance_count, 1 << 16); nir_store_deref(b, args->vars.iteration_instance_count, iteration_instance_count, 0x1); } /* instance */ nir_def *instance_node_addr = build_node_to_addr(device, b, global_bvh_node, false); nir_store_deref(b, args->vars.instance_addr, instance_node_addr, 1); nir_def *instance_data = nir_load_global(b, 4, 32, instance_node_addr, .align_mul = 64, .align_offset = 0); nir_def *wto_matrix[3]; radv_load_wto_matrix(device, b, instance_node_addr, wto_matrix); nir_store_deref(b, args->vars.sbt_offset_and_flags, nir_channel(b, instance_data, 3), 1); if (!args->ignore_cull_mask) { nir_def *instance_and_mask = nir_channel(b, instance_data, 2); nir_push_if(b, nir_ult(b, nir_iand(b, instance_and_mask, args->cull_mask), nir_imm_int(b, 1 << 24))); { if (!args->use_bvh_stack_rtn) nir_jump(b, nir_jump_continue); } nir_push_else(b, NULL); } nir_store_deref(b, args->vars.top_stack, nir_load_deref(b, args->vars.stack), 1); /* If ray flags dictate a forced opaqueness/nonopaqueness, instance flags dictating the same are * meaningless. */ uint32_t forced_opaqueness_mask = SpvRayFlagsOpaqueKHRMask | SpvRayFlagsNoOpaqueKHRMask; nir_def *instance_flag_mask = nir_bcsel(b, nir_test_mask(b, ptr_flags, forced_opaqueness_mask), nir_imm_int64(b, ~((uint64_t)forced_opaqueness_mask << 54ull)), nir_imm_int64(b, ~0ull)); nir_def *instance_pointer = nir_pack_64_2x32(b, nir_trim_vector(b, instance_data, 2)); instance_pointer = nir_iand(b, instance_pointer, instance_flag_mask); nir_store_deref(b, args->vars.bvh_base, build_bvh_base(b, pdev, instance_pointer, ptr_flags, false), 0x1); /* Push the instance root node onto the stack */ if (args->use_bvh_stack_rtn) { nir_store_var(b, last_visited_node, nir_imm_int(b, RADV_BVH_INVALID_NODE), 0x1); nir_store_var(b, intrinsic_result, nir_imm_ivec4(b, RADV_BVH_ROOT_NODE, RADV_BVH_INVALID_NODE, RADV_BVH_INVALID_NODE, RADV_BVH_INVALID_NODE), 0xf); } else { nir_store_deref(b, args->vars.current_node, nir_imm_int(b, RADV_BVH_ROOT_NODE), 0x1); } nir_store_deref(b, args->vars.instance_bottom_node, nir_imm_int(b, RADV_BVH_ROOT_NODE), 1); nir_store_deref(b, args->vars.instance_top_node, bvh_node, 1); /* Transform the ray into object space */ nir_store_deref(b, args->vars.origin, nir_build_vec3_mat_mult(b, args->origin, wto_matrix, true), 7); nir_store_deref(b, args->vars.dir, nir_build_vec3_mat_mult(b, args->dir, wto_matrix, false), 7); nir_store_deref(b, args->vars.inv_dir, nir_fdiv(b, vec3ones, nir_load_deref(b, args->vars.dir)), 7); if (!args->ignore_cull_mask) nir_pop_if(b, NULL); } nir_pop_if(b, NULL); } nir_push_else(b, NULL); { nir_def *result; if (has_result) { result = nir_load_var(b, intrinsic_result); } else { /* If we didn't run the intrinsic cause the hardware didn't support it, * emulate ray/box intersection here */ result = intersect_ray_amd_software_box( device, b, global_bvh_node, nir_load_deref(b, args->vars.tmax), nir_load_deref(b, args->vars.origin), nir_load_deref(b, args->vars.dir), nir_load_deref(b, args->vars.inv_dir)); } /* box */ if (args->use_bvh_stack_rtn) { nir_store_var(b, last_visited_node, prev_node, 0x1); } else { nir_push_if(b, nir_ieq_imm(b, prev_node, RADV_BVH_INVALID_NODE)); { nir_def *new_nodes[4]; for (unsigned i = 0; i < 4; ++i) new_nodes[i] = nir_channel(b, result, i); for (unsigned i = 1; i < 4; ++i) nir_push_if(b, nir_ine_imm(b, new_nodes[i], RADV_BVH_INVALID_NODE)); for (unsigned i = 4; i-- > 1;) { nir_def *stack = nir_load_deref(b, args->vars.stack); nir_def *stack_ptr = nir_umod_imm(b, stack, args->stack_entries * args->stack_stride); args->stack_store_cb(b, stack_ptr, new_nodes[i], args); nir_store_deref(b, args->vars.stack, nir_iadd_imm(b, stack, args->stack_stride), 1); if (i == 1) { nir_def *new_watermark = nir_iadd_imm(b, nir_load_deref(b, args->vars.stack), -args->stack_entries * args->stack_stride); new_watermark = nir_imax(b, nir_load_deref(b, args->vars.stack_low_watermark), new_watermark); nir_store_deref(b, args->vars.stack_low_watermark, new_watermark, 0x1); } nir_pop_if(b, NULL); } nir_store_deref(b, args->vars.current_node, new_nodes[0], 0x1); } nir_push_else(b, NULL); { nir_def *next = nir_imm_int(b, RADV_BVH_INVALID_NODE); for (unsigned i = 0; i < 3; ++i) { next = nir_bcsel(b, nir_ieq(b, prev_node, nir_channel(b, result, i)), nir_channel(b, result, i + 1), next); } nir_store_deref(b, args->vars.current_node, next, 0x1); } nir_pop_if(b, NULL); } } nir_pop_if(b, NULL); } nir_push_else(b, NULL); { nir_def *result; if (has_result) { result = nir_load_var(b, intrinsic_result); } else { /* If we didn't run the intrinsic cause the hardware didn't support it, * emulate ray/tri intersection here */ result = intersect_ray_amd_software_tri( device, b, global_bvh_node, nir_load_deref(b, args->vars.tmax), nir_load_deref(b, args->vars.origin), nir_load_deref(b, args->vars.dir), nir_load_deref(b, args->vars.inv_dir)); } insert_traversal_triangle_case(device, b, args, &ray_flags, result, global_bvh_node); } nir_pop_if(b, NULL); if (args->vars.iteration_instance_count) { nir_def *iteration_instance_count = nir_load_deref(b, args->vars.iteration_instance_count); iteration_instance_count = nir_iadd_imm(b, iteration_instance_count, 1); nir_store_deref(b, args->vars.iteration_instance_count, iteration_instance_count, 0x1); } if (args->use_bvh_stack_rtn) { nir_def *stack_result = nir_bvh_stack_rtn_amd(b, 32, nir_load_deref(b, args->vars.stack), nir_load_var(b, last_visited_node), nir_load_var(b, intrinsic_result), .stack_size = args->stack_entries); nir_store_deref(b, args->vars.stack, nir_channel(b, stack_result, 0), 0x1); nir_store_deref(b, args->vars.current_node, nir_channel(b, stack_result, 1), 0x1); } if (args->vars.break_flag) { nir_push_if(b, nir_load_deref(b, args->vars.break_flag)); { nir_jump(b, nir_jump_break); } nir_pop_if(b, NULL); } } nir_pop_loop(b, NULL); return nir_load_var(b, incomplete); } nir_def * radv_build_ray_traversal_gfx12(struct radv_device *device, nir_builder *b, const struct radv_ray_traversal_args *args) { const struct radv_physical_device *pdev = radv_device_physical(device); nir_variable *incomplete = nir_local_variable_create(b->impl, glsl_bool_type(), "incomplete"); nir_store_var(b, incomplete, nir_imm_true(b), 0x1); nir_variable *intrinsic_result = nir_local_variable_create(b->impl, glsl_uvec_type(8), "intrinsic_result"); nir_variable *last_visited_node = nir_local_variable_create(b->impl, glsl_uint_type(), "last_visited_node"); struct radv_ray_flags ray_flags = { .force_opaque = radv_test_flag(b, args, SpvRayFlagsOpaqueKHRMask, true), .force_not_opaque = radv_test_flag(b, args, SpvRayFlagsNoOpaqueKHRMask, true), .terminate_on_first_hit = radv_test_flag(b, args, SpvRayFlagsTerminateOnFirstHitKHRMask, true), .no_cull_front = radv_test_flag(b, args, SpvRayFlagsCullFrontFacingTrianglesKHRMask, false), .no_cull_back = radv_test_flag(b, args, SpvRayFlagsCullBackFacingTrianglesKHRMask, false), .no_cull_opaque = radv_test_flag(b, args, SpvRayFlagsCullOpaqueKHRMask, false), .no_cull_no_opaque = radv_test_flag(b, args, SpvRayFlagsCullNoOpaqueKHRMask, false), .no_skip_triangles = radv_test_flag(b, args, SpvRayFlagsSkipTrianglesKHRMask, false), .no_skip_aabbs = radv_test_flag(b, args, SpvRayFlagsSkipAABBsKHRMask, false), }; nir_def *desc = create_bvh_descriptor(b, pdev, &ray_flags); nir_loop *loop = nir_push_loop(b); { if (!args->use_bvh_stack_rtn) nir_loop_add_continue_construct(loop); /* When exiting instances via stack, current_node won't ever be invalid with ds_bvh_stack_rtn */ if (args->use_bvh_stack_rtn) { /* Early-exit when the stack is empty and there are no more nodes to process. */ nir_push_if(b, nir_ieq_imm(b, nir_load_deref(b, args->vars.current_node), RADV_BVH_STACK_TERMINAL_NODE)); { nir_store_var(b, incomplete, nir_imm_false(b), 0x1); nir_jump(b, nir_jump_break); } nir_pop_if(b, NULL); build_instance_exit(b, pdev, args, nir_test_mask(b, nir_load_deref(b, args->vars.stack), RADV_BVH_STACK_FLAG_TLAS_POP), NULL); } nir_push_if(b, nir_ieq_imm(b, nir_load_deref(b, args->vars.current_node), RADV_BVH_INVALID_NODE)); { /* Early exit if we never overflowed the stack, to avoid having to backtrack to * the root for no reason. */ if (!args->use_bvh_stack_rtn) { nir_push_if(b, nir_ilt_imm(b, nir_load_deref(b, args->vars.stack), args->stack_stride)); { nir_store_var(b, incomplete, nir_imm_false(b), 0x1); nir_jump(b, nir_jump_break); } nir_pop_if(b, NULL); build_instance_exit( b, pdev, args, nir_ige(b, nir_load_deref(b, args->vars.top_stack), nir_load_deref(b, args->vars.stack)), NULL); } nir_def *overflow_cond = nir_ige(b, nir_load_deref(b, args->vars.stack_low_watermark), nir_load_deref(b, args->vars.stack)); /* ds_bvh_stack_rtn returns 0xFFFFFFFF if and only if there was a stack overflow. */ if (args->use_bvh_stack_rtn) overflow_cond = nir_imm_true(b); nir_push_if(b, overflow_cond); { nir_def *prev = nir_load_deref(b, args->vars.previous_node); nir_def *loaded_parent_id; nir_def *primitive_parent_id; nir_push_if(b, nir_test_mask(b, prev, BITFIELD64_BIT(ffs(radv_bvh_node_box16) - 1))); { nir_def *is_instance = nir_test_mask(b, prev, BITFIELD64_BIT(ffs(radv_bvh_node_instance) - 1)); nir_def *field_offset = nir_bcsel( b, is_instance, nir_imm_int( b, (int32_t)offsetof(struct radv_gfx12_instance_node, parent_id) - (radv_bvh_node_instance << 3)), nir_imm_int(b, (int32_t)offsetof(struct radv_gfx12_box_node, parent_id) - (radv_bvh_node_box32 << 3))); nir_def *offset = nir_iadd(b, nir_ishl_imm(b, prev, 3), field_offset); nir_def *bvh_addr = build_node_to_addr(device, b, nir_load_deref(b, args->vars.bvh_base), true); loaded_parent_id = nir_load_global(b, 1, 32, nir_iadd(b, bvh_addr, nir_u2u64(b, offset))); } nir_push_else(b, NULL); { primitive_parent_id = nir_load_deref(b, args->vars.parent_node); } nir_pop_if(b, NULL); nir_def *parent = nir_if_phi(b, loaded_parent_id, primitive_parent_id); nir_push_if(b, nir_ieq_imm(b, parent, RADV_BVH_INVALID_NODE)); { nir_store_var(b, incomplete, nir_imm_false(b), 0x1); nir_jump(b, nir_jump_break); } nir_pop_if(b, NULL); nir_store_deref(b, args->vars.current_node, parent, 0x1); } nir_push_else(b, NULL); { if (!args->use_bvh_stack_rtn) { nir_store_deref(b, args->vars.stack, nir_iadd_imm(b, nir_load_deref(b, args->vars.stack), -args->stack_stride), 1); nir_def *stack_ptr = nir_umod_imm(b, nir_load_deref(b, args->vars.stack), args->stack_stride * args->stack_entries); nir_def *bvh_node = args->stack_load_cb(b, stack_ptr, args); nir_store_deref(b, args->vars.current_node, bvh_node, 0x1); } nir_store_deref(b, args->vars.previous_node, nir_imm_int(b, RADV_BVH_INVALID_NODE), 0x1); } nir_pop_if(b, NULL); } nir_push_else(b, NULL); { nir_store_deref(b, args->vars.previous_node, nir_imm_int(b, RADV_BVH_INVALID_NODE), 0x1); } nir_pop_if(b, NULL); nir_def *bvh_node = nir_load_deref(b, args->vars.current_node); nir_def *prev_node = nir_load_deref(b, args->vars.previous_node); nir_store_deref(b, args->vars.previous_node, bvh_node, 0x1); if (args->use_bvh_stack_rtn) nir_store_var(b, last_visited_node, nir_imm_int(b, RADV_BVH_INVALID_NODE), 0x1); else nir_store_deref(b, args->vars.current_node, nir_imm_int(b, RADV_BVH_INVALID_NODE), 0x1); nir_def *global_bvh_node = nir_iadd(b, nir_load_deref(b, args->vars.bvh_base), nir_u2u64(b, bvh_node)); nir_def *result = nir_bvh8_intersect_ray_amd(b, 32, desc, nir_unpack_64_2x32(b, nir_load_deref(b, args->vars.bvh_base)), nir_ishr_imm(b, args->cull_mask, 24), nir_load_deref(b, args->vars.tmax), nir_load_deref(b, args->vars.origin), nir_load_deref(b, args->vars.dir), bvh_node); nir_store_var(b, intrinsic_result, nir_channels(b, result, 0xff), 0xff); nir_store_deref(b, args->vars.origin, nir_channels(b, result, 0x7 << 10), 0x7); nir_store_deref(b, args->vars.dir, nir_channels(b, result, 0x7 << 13), 0x7); nir_push_if(b, nir_test_mask(b, bvh_node, BITFIELD64_BIT(ffs(radv_bvh_node_box16) - 1))); { nir_push_if(b, nir_test_mask(b, bvh_node, BITFIELD64_BIT(ffs(radv_bvh_node_instance) - 1))); { if (args->vars.iteration_instance_count) { nir_def *iteration_instance_count = nir_load_deref(b, args->vars.iteration_instance_count); iteration_instance_count = nir_iadd_imm(b, iteration_instance_count, 1 << 16); nir_store_deref(b, args->vars.iteration_instance_count, iteration_instance_count, 0x1); } nir_def *next_node = nir_iand_imm(b, nir_channel(b, result, 7), 0xff); nir_push_if(b, nir_ieq_imm(b, next_node, 0xff)); { nir_store_deref(b, args->vars.origin, args->origin, 7); nir_store_deref(b, args->vars.dir, args->dir, 7); if (args->use_bvh_stack_rtn) { nir_def *skip_0_7 = nir_imm_int(b, RADV_BVH_STACK_SKIP_0_TO_7); nir_store_var(b, intrinsic_result, nir_vector_insert_imm(b, nir_load_var(b, intrinsic_result), skip_0_7, 7), 0xff); } else { nir_jump(b, nir_jump_continue); } } nir_push_else(b, NULL); { /* instance */ nir_def *instance_node_addr = build_node_to_addr(device, b, global_bvh_node, false); nir_store_deref(b, args->vars.instance_addr, instance_node_addr, 1); nir_store_deref(b, args->vars.sbt_offset_and_flags, nir_channel(b, result, 6), 1); nir_store_deref(b, args->vars.top_stack, nir_load_deref(b, args->vars.stack), 1); nir_store_deref(b, args->vars.bvh_base, nir_pack_64_2x32(b, nir_channels(b, result, 0x3 << 2)), 1); /* Push the instance root node onto the stack */ if (args->use_bvh_stack_rtn) { nir_def *comps[8]; for (unsigned i = 0; i < 6; ++i) comps[i] = nir_channel(b, result, i); comps[6] = nir_imm_int(b, RADV_BVH_STACK_SKIP_0_TO_7); comps[7] = next_node; nir_store_var(b, intrinsic_result, nir_vec(b, comps, 8), 0xff); } else { nir_store_deref(b, args->vars.current_node, next_node, 0x1); } nir_store_deref(b, args->vars.instance_bottom_node, next_node, 1); nir_store_deref(b, args->vars.instance_top_node, bvh_node, 1); } nir_pop_if(b, NULL); } nir_push_else(b, NULL); { nir_store_deref(b, args->vars.parent_node, bvh_node, 0x1); /* box */ if (args->use_bvh_stack_rtn) { nir_store_var(b, last_visited_node, prev_node, 0x1); } else { nir_push_if(b, nir_ieq_imm(b, prev_node, RADV_BVH_INVALID_NODE)); { nir_def *new_nodes[8]; for (unsigned i = 0; i < 8; ++i) new_nodes[i] = nir_channel(b, result, i); for (unsigned i = 1; i < 8; ++i) nir_push_if(b, nir_ine_imm(b, new_nodes[i], RADV_BVH_INVALID_NODE)); for (unsigned i = 8; i-- > 1;) { nir_def *stack = nir_load_deref(b, args->vars.stack); nir_def *stack_ptr = nir_umod_imm(b, stack, args->stack_entries * args->stack_stride); args->stack_store_cb(b, stack_ptr, new_nodes[i], args); nir_store_deref(b, args->vars.stack, nir_iadd_imm(b, stack, args->stack_stride), 1); if (i == 1) { nir_def *new_watermark = nir_iadd_imm(b, nir_load_deref(b, args->vars.stack), -args->stack_entries * args->stack_stride); new_watermark = nir_imax(b, nir_load_deref(b, args->vars.stack_low_watermark), new_watermark); nir_store_deref(b, args->vars.stack_low_watermark, new_watermark, 0x1); } nir_pop_if(b, NULL); } nir_store_deref(b, args->vars.current_node, new_nodes[0], 0x1); } nir_push_else(b, NULL); { nir_def *next = nir_imm_int(b, RADV_BVH_INVALID_NODE); for (unsigned i = 0; i < 7; ++i) { next = nir_bcsel(b, nir_ieq(b, prev_node, nir_channel(b, result, i)), nir_channel(b, result, i + 1), next); } nir_store_deref(b, args->vars.current_node, next, 0x1); } nir_pop_if(b, NULL); } } nir_pop_if(b, NULL); } nir_push_else(b, NULL); { if (args->use_bvh_stack_rtn) { nir_def *skip_0_7 = nir_imm_int(b, RADV_BVH_STACK_SKIP_0_TO_7); nir_store_var(b, intrinsic_result, nir_vector_insert_imm(b, nir_load_var(b, intrinsic_result), skip_0_7, 7), 0xff); } nir_push_if(b, nir_test_mask(b, nir_channel(b, result, 1), 1u << 31)); { nir_push_if(b, ray_flags.no_skip_aabbs); insert_traversal_aabb_case_gfx12(device, b, args, &ray_flags, result, global_bvh_node); nir_pop_if(b, NULL); } nir_push_else(b, NULL); { nir_push_if(b, ray_flags.no_skip_triangles); insert_traversal_triangle_case_gfx12(device, b, args, &ray_flags, intrinsic_result, result, global_bvh_node, bvh_node); nir_pop_if(b, NULL); } nir_pop_if(b, NULL); } nir_pop_if(b, NULL); if (args->vars.iteration_instance_count) { nir_def *iteration_instance_count = nir_load_deref(b, args->vars.iteration_instance_count); iteration_instance_count = nir_iadd_imm(b, iteration_instance_count, 1); nir_store_deref(b, args->vars.iteration_instance_count, iteration_instance_count, 0x1); } if (args->use_bvh_stack_rtn) { nir_def *stack_result; stack_result = nir_bvh_stack_rtn_amd(b, 32, nir_load_deref(b, args->vars.stack), nir_load_var(b, last_visited_node), nir_load_var(b, intrinsic_result), .stack_size = args->stack_entries); nir_store_deref(b, args->vars.stack, nir_channel(b, stack_result, 0), 0x1); nir_store_deref(b, args->vars.current_node, nir_channel(b, stack_result, 1), 0x1); } if (args->vars.break_flag) { nir_push_if(b, nir_load_deref(b, args->vars.break_flag)); { nir_jump(b, nir_jump_break); } nir_pop_if(b, NULL); } } nir_pop_loop(b, NULL); return nir_load_var(b, incomplete); }