/* * Copyright 2023 Conor McCarthy for CodeWeavers * * This library is free software; you can redistribute it and/or * modify it under the terms of the GNU Lesser General Public * License as published by the Free Software Foundation; either * version 2.1 of the License, or (at your option) any later version. * * This library is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU * Lesser General Public License for more details. * * You should have received a copy of the GNU Lesser General Public * License along with this library; if not, write to the Free Software * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA */ #include "vkd3d_shader_private.h" #define VKD3D_SM6_VERSION_MAJOR(version) (((version) >> 4) & 0xf) #define VKD3D_SM6_VERSION_MINOR(version) (((version) >> 0) & 0xf) #define BITCODE_MAGIC VKD3D_MAKE_TAG('B', 'C', 0xc0, 0xde) #define DXIL_OP_MAX_OPERANDS 17 enum bitcode_block_id { BLOCKINFO_BLOCK = 0, MODULE_BLOCK = 8, PARAMATTR_BLOCK = 9, PARAMATTR_GROUP_BLOCK = 10, CONSTANTS_BLOCK = 11, FUNCTION_BLOCK = 12, VALUE_SYMTAB_BLOCK = 14, METADATA_BLOCK = 15, METADATA_ATTACHMENT_BLOCK = 16, TYPE_BLOCK = 17, USELIST_BLOCK = 18, }; enum bitcode_blockinfo_code { SETBID = 1, BLOCKNAME = 2, SETRECORDNAME = 3, }; enum bitcode_block_abbreviation { END_BLOCK = 0, ENTER_SUBBLOCK = 1, DEFINE_ABBREV = 2, UNABBREV_RECORD = 3, }; enum bitcode_abbrev_type { ABBREV_FIXED = 1, ABBREV_VBR = 2, ABBREV_ARRAY = 3, ABBREV_CHAR = 4, ABBREV_BLOB = 5, }; enum bitcode_address_space { ADDRESS_SPACE_DEFAULT, ADDRESS_SPACE_DEVICEMEM, ADDRESS_SPACE_CBUFFER, ADDRESS_SPACE_GROUPSHARED, }; enum bitcode_module_code { MODULE_CODE_VERSION = 1, MODULE_CODE_GLOBALVAR = 7, MODULE_CODE_FUNCTION = 8, }; enum bitcode_constant_code { CST_CODE_SETTYPE = 1, CST_CODE_NULL = 2, CST_CODE_UNDEF = 3, CST_CODE_INTEGER = 4, CST_CODE_FLOAT = 6, CST_CODE_STRING = 8, CST_CODE_CE_GEP = 12, CST_CODE_CE_INBOUNDS_GEP = 20, CST_CODE_DATA = 22, }; enum bitcode_function_code { FUNC_CODE_DECLAREBLOCKS = 1, FUNC_CODE_INST_BINOP = 2, FUNC_CODE_INST_CAST = 3, FUNC_CODE_INST_RET = 10, FUNC_CODE_INST_BR = 11, FUNC_CODE_INST_SWITCH = 12, FUNC_CODE_INST_PHI = 16, FUNC_CODE_INST_ALLOCA = 19, FUNC_CODE_INST_LOAD = 20, FUNC_CODE_INST_EXTRACTVAL = 26, FUNC_CODE_INST_CMP2 = 28, FUNC_CODE_INST_VSELECT = 29, FUNC_CODE_INST_CALL = 34, FUNC_CODE_INST_ATOMICRMW = 38, FUNC_CODE_INST_LOADATOMIC = 41, FUNC_CODE_INST_GEP = 43, FUNC_CODE_INST_STORE = 44, FUNC_CODE_INST_STOREATOMIC = 45, FUNC_CODE_INST_CMPXCHG = 46, }; enum bitcode_type_code { TYPE_CODE_NUMENTRY = 1, TYPE_CODE_VOID = 2, TYPE_CODE_FLOAT = 3, TYPE_CODE_DOUBLE = 4, TYPE_CODE_LABEL = 5, TYPE_CODE_INTEGER = 7, TYPE_CODE_POINTER = 8, TYPE_CODE_HALF = 10, TYPE_CODE_ARRAY = 11, TYPE_CODE_VECTOR = 12, TYPE_CODE_METADATA = 16, TYPE_CODE_STRUCT_ANON = 18, TYPE_CODE_STRUCT_NAME = 19, TYPE_CODE_STRUCT_NAMED = 20, TYPE_CODE_FUNCTION = 21, }; enum bitcode_value_symtab_code { VST_CODE_ENTRY = 1, VST_CODE_BBENTRY = 2, }; enum dx_intrinsic_opcode { DX_LOAD_INPUT = 4, DX_STORE_OUTPUT = 5, }; struct sm6_pointer_info { const struct sm6_type *type; enum bitcode_address_space addr_space; }; struct sm6_struct_info { const char *name; unsigned int elem_count; const struct sm6_type *elem_types[]; }; struct sm6_function_info { const struct sm6_type *ret_type; unsigned int param_count; const struct sm6_type *param_types[]; }; struct sm6_array_info { unsigned int count; const struct sm6_type *elem_type; }; enum sm6_type_class { TYPE_CLASS_VOID, TYPE_CLASS_INTEGER, TYPE_CLASS_FLOAT, TYPE_CLASS_POINTER, TYPE_CLASS_STRUCT, TYPE_CLASS_FUNCTION, TYPE_CLASS_VECTOR, TYPE_CLASS_ARRAY, TYPE_CLASS_LABEL, TYPE_CLASS_METADATA, }; struct sm6_type { enum sm6_type_class class; union { unsigned int width; struct sm6_pointer_info pointer; struct sm6_struct_info *struc; struct sm6_function_info *function; struct sm6_array_info array; } u; }; enum sm6_value_type { VALUE_TYPE_FUNCTION, VALUE_TYPE_REG, }; struct sm6_function_data { const char *name; bool is_prototype; unsigned int attribs_id; }; struct sm6_value { const struct sm6_type *type; enum sm6_value_type value_type; bool is_undefined; union { struct sm6_function_data function; struct vkd3d_shader_register reg; } u; }; struct dxil_record { unsigned int code; unsigned int operand_count; uint64_t operands[]; }; struct sm6_symbol { unsigned int id; const char *name; }; struct sm6_block { struct vkd3d_shader_instruction *instructions; size_t instruction_capacity; size_t instruction_count; }; struct sm6_function { const struct sm6_value *declaration; struct sm6_block *blocks[1]; size_t block_count; size_t value_count; }; struct dxil_block { const struct dxil_block *parent; enum bitcode_block_id id; unsigned int abbrev_len; unsigned int start; unsigned int length; unsigned int level; /* The abbrev, block and record structs are not relocatable. */ struct dxil_abbrev **abbrevs; size_t abbrev_capacity; size_t abbrev_count; unsigned int blockinfo_bid; bool has_bid; struct dxil_block **child_blocks; size_t child_block_capacity; size_t child_block_count; struct dxil_record **records; size_t record_capacity; size_t record_count; }; struct sm6_parser { const uint32_t *ptr, *start, *end; unsigned int bitpos; struct dxil_block root_block; struct dxil_block *current_block; struct dxil_global_abbrev **abbrevs; size_t abbrev_capacity; size_t abbrev_count; struct sm6_type *types; size_t type_count; struct sm6_symbol *global_symbols; size_t global_symbol_count; struct vkd3d_shader_dst_param *output_params; struct vkd3d_shader_dst_param *input_params; struct sm6_function *functions; size_t function_count; struct sm6_value *values; size_t value_count; size_t value_capacity; size_t cur_max_value; unsigned int ssa_next_id; struct vkd3d_shader_parser p; }; struct dxil_abbrev_operand { uint64_t context; bool (*read_operand)(struct sm6_parser *sm6, uint64_t context, uint64_t *operand); }; struct dxil_abbrev { unsigned int count; bool is_array; struct dxil_abbrev_operand operands[]; }; struct dxil_global_abbrev { unsigned int block_id; struct dxil_abbrev abbrev; }; static const uint64_t CALL_CONV_FLAG_EXPLICIT_TYPE = 1ull << 15; static size_t size_add_with_overflow_check(size_t a, size_t b) { size_t i = a + b; return (i < a) ? SIZE_MAX : i; } static struct sm6_parser *sm6_parser(struct vkd3d_shader_parser *parser) { return CONTAINING_RECORD(parser, struct sm6_parser, p); } static bool sm6_parser_is_end(struct sm6_parser *sm6) { return sm6->ptr == sm6->end; } static uint32_t sm6_parser_read_uint32(struct sm6_parser *sm6) { if (sm6_parser_is_end(sm6)) { sm6->p.failed = true; return 0; } return *sm6->ptr++; } static uint32_t sm6_parser_read_bits(struct sm6_parser *sm6, unsigned int length) { unsigned int l, prev_len = 0; uint32_t bits; if (!length) return 0; assert(length < 32); if (sm6_parser_is_end(sm6)) { sm6->p.failed = true; return 0; } assert(sm6->bitpos < 32); bits = *sm6->ptr >> sm6->bitpos; l = 32 - sm6->bitpos; if (l <= length) { ++sm6->ptr; if (sm6_parser_is_end(sm6) && l < length) { sm6->p.failed = true; return bits; } sm6->bitpos = 0; bits |= *sm6->ptr << l; prev_len = l; } sm6->bitpos += length - prev_len; return bits & ((1 << length) - 1); } static uint64_t sm6_parser_read_vbr(struct sm6_parser *sm6, unsigned int length) { unsigned int bits, flag, mask, shift = 0; uint64_t result = 0; if (!length) return 0; if (sm6_parser_is_end(sm6)) { sm6->p.failed = true; return 0; } flag = 1 << (length - 1); mask = flag - 1; do { bits = sm6_parser_read_bits(sm6, length); result |= (uint64_t)(bits & mask) << shift; shift += length - 1; } while ((bits & flag) && !sm6->p.failed && shift < 64); sm6->p.failed |= !!(bits & flag); return result; } static void sm6_parser_align_32(struct sm6_parser *sm6) { if (!sm6->bitpos) return; if (sm6_parser_is_end(sm6)) { sm6->p.failed = true; return; } ++sm6->ptr; sm6->bitpos = 0; } static bool dxil_block_handle_blockinfo_record(struct dxil_block *block, struct dxil_record *record) { /* BLOCKINFO blocks must only occur immediately below the module root block. */ if (block->level > 1) { WARN("Invalid blockinfo block level %u.\n", block->level); return false; } switch (record->code) { case SETBID: if (!record->operand_count) { WARN("Missing id operand.\n"); return false; } if (record->operands[0] > UINT_MAX) WARN("Truncating block id %"PRIu64".\n", record->operands[0]); block->blockinfo_bid = record->operands[0]; block->has_bid = true; break; case BLOCKNAME: case SETRECORDNAME: break; default: FIXME("Unhandled BLOCKINFO record type %u.\n", record->code); break; } return true; } static enum vkd3d_result dxil_block_add_record(struct dxil_block *block, struct dxil_record *record) { unsigned int reserve; switch (block->id) { /* Rough initial reserve sizes for small shaders. */ case CONSTANTS_BLOCK: reserve = 32; break; case FUNCTION_BLOCK: reserve = 128; break; case METADATA_BLOCK: reserve = 32; break; case TYPE_BLOCK: reserve = 32; break; default: reserve = 8; break; } reserve = max(reserve, block->record_count + 1); if (!vkd3d_array_reserve((void **)&block->records, &block->record_capacity, reserve, sizeof(*block->records))) { ERR("Failed to allocate %u records.\n", reserve); return VKD3D_ERROR_OUT_OF_MEMORY; } if (block->id == BLOCKINFO_BLOCK && !dxil_block_handle_blockinfo_record(block, record)) return VKD3D_ERROR_INVALID_SHADER; block->records[block->record_count++] = record; return VKD3D_OK; } static enum vkd3d_result sm6_parser_read_unabbrev_record(struct sm6_parser *sm6) { struct dxil_block *block = sm6->current_block; enum vkd3d_result ret = VKD3D_OK; unsigned int code, count, i; struct dxil_record *record; code = sm6_parser_read_vbr(sm6, 6); count = sm6_parser_read_vbr(sm6, 6); if (!(record = vkd3d_malloc(sizeof(*record) + count * sizeof(record->operands[0])))) { ERR("Failed to allocate record with %u operands.\n", count); return VKD3D_ERROR_OUT_OF_MEMORY; } record->code = code; record->operand_count = count; for (i = 0; i < count; ++i) record->operands[i] = sm6_parser_read_vbr(sm6, 6); if (sm6->p.failed) ret = VKD3D_ERROR_INVALID_SHADER; if (ret < 0 || (ret = dxil_block_add_record(block, record)) < 0) vkd3d_free(record); return ret; } static bool sm6_parser_read_literal_operand(struct sm6_parser *sm6, uint64_t context, uint64_t *op) { *op = context; return !sm6->p.failed; } static bool sm6_parser_read_fixed_operand(struct sm6_parser *sm6, uint64_t context, uint64_t *op) { *op = sm6_parser_read_bits(sm6, context); return !sm6->p.failed; } static bool sm6_parser_read_vbr_operand(struct sm6_parser *sm6, uint64_t context, uint64_t *op) { *op = sm6_parser_read_vbr(sm6, context); return !sm6->p.failed; } static bool sm6_parser_read_char6_operand(struct sm6_parser *sm6, uint64_t context, uint64_t *op) { *op = "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789._"[sm6_parser_read_bits(sm6, 6)]; return !sm6->p.failed; } static bool sm6_parser_read_blob_operand(struct sm6_parser *sm6, uint64_t context, uint64_t *op) { int count = sm6_parser_read_vbr(sm6, 6); sm6_parser_align_32(sm6); for (; count > 0; count -= 4) sm6_parser_read_uint32(sm6); FIXME("Unhandled blob operand.\n"); return false; } static enum vkd3d_result dxil_abbrev_init(struct dxil_abbrev *abbrev, unsigned int count, struct sm6_parser *sm6) { enum bitcode_abbrev_type prev_type, type; unsigned int i; abbrev->is_array = false; for (i = 0, prev_type = 0; i < count && !sm6->p.failed; ++i) { if (sm6_parser_read_bits(sm6, 1)) { if (prev_type == ABBREV_ARRAY) { WARN("Unexpected literal abbreviation after array.\n"); return VKD3D_ERROR_INVALID_SHADER; } abbrev->operands[i].context = sm6_parser_read_vbr(sm6, 8); abbrev->operands[i].read_operand = sm6_parser_read_literal_operand; continue; } switch (type = sm6_parser_read_bits(sm6, 3)) { case ABBREV_FIXED: case ABBREV_VBR: abbrev->operands[i].context = sm6_parser_read_vbr(sm6, 5); abbrev->operands[i].read_operand = (type == ABBREV_FIXED) ? sm6_parser_read_fixed_operand : sm6_parser_read_vbr_operand; break; case ABBREV_ARRAY: if (prev_type == ABBREV_ARRAY || i != count - 2) { WARN("Unexpected array abbreviation.\n"); return VKD3D_ERROR_INVALID_SHADER; } abbrev->is_array = true; --i; --count; break; case ABBREV_CHAR: abbrev->operands[i].read_operand = sm6_parser_read_char6_operand; break; case ABBREV_BLOB: if (prev_type == ABBREV_ARRAY || i != count - 1) { WARN("Unexpected blob abbreviation.\n"); return VKD3D_ERROR_INVALID_SHADER; } abbrev->operands[i].read_operand = sm6_parser_read_blob_operand; break; } prev_type = type; } abbrev->count = count; return sm6->p.failed ? VKD3D_ERROR_INVALID_SHADER : VKD3D_OK; } static enum vkd3d_result sm6_parser_add_global_abbrev(struct sm6_parser *sm6) { struct dxil_block *block = sm6->current_block; unsigned int count = sm6_parser_read_vbr(sm6, 5); struct dxil_global_abbrev *global_abbrev; enum vkd3d_result ret; assert(block->id == BLOCKINFO_BLOCK); if (!vkd3d_array_reserve((void **)&sm6->abbrevs, &sm6->abbrev_capacity, sm6->abbrev_count + 1, sizeof(*sm6->abbrevs)) || !(global_abbrev = vkd3d_malloc(sizeof(*global_abbrev) + count * sizeof(global_abbrev->abbrev.operands[0])))) { ERR("Failed to allocate global abbreviation.\n"); return VKD3D_ERROR_OUT_OF_MEMORY; } if ((ret = dxil_abbrev_init(&global_abbrev->abbrev, count, sm6)) < 0) { vkd3d_free(global_abbrev); return ret; } if (!block->has_bid) { WARN("Missing blockinfo block id.\n"); return VKD3D_ERROR_INVALID_SHADER; } if (block->blockinfo_bid == MODULE_BLOCK) { FIXME("Unhandled global abbreviation for module block.\n"); return VKD3D_ERROR_INVALID_SHADER; } global_abbrev->block_id = block->blockinfo_bid; sm6->abbrevs[sm6->abbrev_count++] = global_abbrev; return VKD3D_OK; } static enum vkd3d_result sm6_parser_add_block_abbrev(struct sm6_parser *sm6) { struct dxil_block *block = sm6->current_block; struct dxil_abbrev *abbrev; enum vkd3d_result ret; unsigned int count; if (block->id == BLOCKINFO_BLOCK) return sm6_parser_add_global_abbrev(sm6); count = sm6_parser_read_vbr(sm6, 5); if (!vkd3d_array_reserve((void **)&block->abbrevs, &block->abbrev_capacity, block->abbrev_count + 1, sizeof(*block->abbrevs)) || !(abbrev = vkd3d_malloc(sizeof(*abbrev) + count * sizeof(abbrev->operands[0])))) { ERR("Failed to allocate block abbreviation.\n"); return VKD3D_ERROR_OUT_OF_MEMORY; } if ((ret = dxil_abbrev_init(abbrev, count, sm6)) < 0) { vkd3d_free(abbrev); return ret; } block->abbrevs[block->abbrev_count++] = abbrev; return VKD3D_OK; } static enum vkd3d_result sm6_parser_read_abbrev_record(struct sm6_parser *sm6, unsigned int abbrev_id) { enum vkd3d_result ret = VKD3D_ERROR_INVALID_SHADER; struct dxil_block *block = sm6->current_block; struct dxil_record *temp, *record; unsigned int i, count, array_len; struct dxil_abbrev *abbrev; uint64_t code; if (abbrev_id >= block->abbrev_count) { WARN("Invalid abbreviation id %u.\n", abbrev_id); return VKD3D_ERROR_INVALID_SHADER; } abbrev = block->abbrevs[abbrev_id]; if (!(count = abbrev->count)) return VKD3D_OK; if (count == 1 && abbrev->is_array) return VKD3D_ERROR_INVALID_SHADER; /* First operand is the record code. The array is included in the count, but will be done separately. */ count -= abbrev->is_array + 1; if (!(record = vkd3d_malloc(sizeof(*record) + count * sizeof(record->operands[0])))) { ERR("Failed to allocate record with %u operands.\n", count); return VKD3D_ERROR_OUT_OF_MEMORY; } if (!abbrev->operands[0].read_operand(sm6, abbrev->operands[0].context, &code)) goto fail; if (code > UINT_MAX) FIXME("Truncating 64-bit record code %#"PRIx64".\n", code); record->code = code; for (i = 0; i < count; ++i) if (!abbrev->operands[i + 1].read_operand(sm6, abbrev->operands[i + 1].context, &record->operands[i])) goto fail; record->operand_count = count; /* An array can occur only as the last operand. */ if (abbrev->is_array) { array_len = sm6_parser_read_vbr(sm6, 6); if (!(temp = vkd3d_realloc(record, sizeof(*record) + (count + array_len) * sizeof(record->operands[0])))) { ERR("Failed to allocate record with %u operands.\n", count + array_len); ret = VKD3D_ERROR_OUT_OF_MEMORY; goto fail; } record = temp; for (i = 0; i < array_len; ++i) { if (!abbrev->operands[count + 1].read_operand(sm6, abbrev->operands[count + 1].context, &record->operands[count + i])) { goto fail; } } record->operand_count += array_len; } if ((ret = dxil_block_add_record(block, record)) < 0) goto fail; return VKD3D_OK; fail: vkd3d_free(record); return ret; } static enum vkd3d_result dxil_block_init(struct dxil_block *block, const struct dxil_block *parent, struct sm6_parser *sm6); static enum vkd3d_result dxil_block_read(struct dxil_block *parent, struct sm6_parser *sm6) { unsigned int reserve = (parent->id == MODULE_BLOCK) ? 12 : 2; struct dxil_block *block; enum vkd3d_result ret; sm6->current_block = parent; do { unsigned int abbrev_id = sm6_parser_read_bits(sm6, parent->abbrev_len); switch (abbrev_id) { case END_BLOCK: sm6_parser_align_32(sm6); return VKD3D_OK; case ENTER_SUBBLOCK: if (parent->id != MODULE_BLOCK && parent->id != FUNCTION_BLOCK) { WARN("Invalid subblock parent id %u.\n", parent->id); return VKD3D_ERROR_INVALID_SHADER; } if (!vkd3d_array_reserve((void **)&parent->child_blocks, &parent->child_block_capacity, max(reserve, parent->child_block_count + 1), sizeof(*parent->child_blocks)) || !(block = vkd3d_calloc(1, sizeof(*block)))) { ERR("Failed to allocate block.\n"); return VKD3D_ERROR_OUT_OF_MEMORY; } if ((ret = dxil_block_init(block, parent, sm6)) < 0) { vkd3d_free(block); return ret; } parent->child_blocks[parent->child_block_count++] = block; sm6->current_block = parent; break; case DEFINE_ABBREV: if ((ret = sm6_parser_add_block_abbrev(sm6)) < 0) return ret; break; case UNABBREV_RECORD: if ((ret = sm6_parser_read_unabbrev_record(sm6)) < 0) { WARN("Failed to read unabbreviated record.\n"); return ret; } break; default: if ((ret = sm6_parser_read_abbrev_record(sm6, abbrev_id - 4)) < 0) { WARN("Failed to read abbreviated record.\n"); return ret; } break; } } while (!sm6->p.failed); return VKD3D_ERROR_INVALID_SHADER; } static size_t sm6_parser_compute_global_abbrev_count_for_block_id(struct sm6_parser *sm6, unsigned int block_id) { size_t i, count; for (i = 0, count = 0; i < sm6->abbrev_count; ++i) count += sm6->abbrevs[i]->block_id == block_id; return count; } static void dxil_block_destroy(struct dxil_block *block) { size_t i; for (i = 0; i < block->record_count; ++i) vkd3d_free(block->records[i]); vkd3d_free(block->records); for (i = 0; i < block->child_block_count; ++i) { dxil_block_destroy(block->child_blocks[i]); vkd3d_free(block->child_blocks[i]); } vkd3d_free(block->child_blocks); block->records = NULL; block->record_count = 0; block->child_blocks = NULL; block->child_block_count = 0; } static enum vkd3d_result dxil_block_init(struct dxil_block *block, const struct dxil_block *parent, struct sm6_parser *sm6) { size_t i, abbrev_count = 0; enum vkd3d_result ret; block->parent = parent; block->level = parent ? parent->level + 1 : 0; block->id = sm6_parser_read_vbr(sm6, 8); block->abbrev_len = sm6_parser_read_vbr(sm6, 4); sm6_parser_align_32(sm6); block->length = sm6_parser_read_uint32(sm6); block->start = sm6->ptr - sm6->start; if (sm6->p.failed) return VKD3D_ERROR_INVALID_SHADER; if ((block->abbrev_count = sm6_parser_compute_global_abbrev_count_for_block_id(sm6, block->id))) { if (!vkd3d_array_reserve((void **)&block->abbrevs, &block->abbrev_capacity, block->abbrev_count, sizeof(*block->abbrevs))) { ERR("Failed to allocate block abbreviations.\n"); return VKD3D_ERROR_OUT_OF_MEMORY; } for (i = 0; i < sm6->abbrev_count; ++i) if (sm6->abbrevs[i]->block_id == block->id) block->abbrevs[abbrev_count++] = &sm6->abbrevs[i]->abbrev; assert(abbrev_count == block->abbrev_count); } if ((ret = dxil_block_read(block, sm6)) < 0) dxil_block_destroy(block); for (i = abbrev_count; i < block->abbrev_count; ++i) vkd3d_free(block->abbrevs[i]); vkd3d_free(block->abbrevs); block->abbrevs = NULL; block->abbrev_count = 0; return ret; } static size_t dxil_block_compute_function_count(const struct dxil_block *root) { size_t i, count; for (i = 0, count = 0; i < root->child_block_count; ++i) count += root->child_blocks[i]->id == FUNCTION_BLOCK; return count; } static size_t dxil_block_compute_module_decl_count(const struct dxil_block *block) { size_t i, count; for (i = 0, count = 0; i < block->record_count; ++i) count += block->records[i]->code == MODULE_CODE_FUNCTION; return count; } static size_t dxil_block_compute_constants_count(const struct dxil_block *block) { size_t i, count; for (i = 0, count = 0; i < block->record_count; ++i) count += block->records[i]->code != CST_CODE_SETTYPE; return count; } static void dxil_global_abbrevs_cleanup(struct dxil_global_abbrev **abbrevs, size_t count) { size_t i; for (i = 0; i < count; ++i) vkd3d_free(abbrevs[i]); vkd3d_free(abbrevs); } static const struct dxil_block *sm6_parser_get_level_one_block(const struct sm6_parser *sm6, enum bitcode_block_id id, bool *is_unique) { const struct dxil_block *block, *found = NULL; size_t i; for (i = 0, *is_unique = true; i < sm6->root_block.child_block_count; ++i) { block = sm6->root_block.child_blocks[i]; if (block->id != id) continue; if (!found) found = block; else *is_unique = false; } return found; } static char *dxil_record_to_string(const struct dxil_record *record, unsigned int offset) { unsigned int i; char *str; assert(offset <= record->operand_count); if (!(str = vkd3d_calloc(record->operand_count - offset + 1, 1))) return NULL; for (i = offset; i < record->operand_count; ++i) str[i - offset] = record->operands[i]; return str; } static bool dxil_record_validate_operand_min_count(const struct dxil_record *record, unsigned int min_count, struct sm6_parser *sm6) { if (record->operand_count >= min_count) return true; WARN("Invalid operand count %u for code %u.\n", record->operand_count, record->code); vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_INVALID_OPERAND_COUNT, "Invalid operand count %u for record code %u.", record->operand_count, record->code); return false; } static void dxil_record_validate_operand_max_count(const struct dxil_record *record, unsigned int max_count, struct sm6_parser *sm6) { if (record->operand_count <= max_count) return; WARN("Ignoring %u extra operands for code %u.\n", record->operand_count - max_count, record->code); vkd3d_shader_parser_warning(&sm6->p, VKD3D_SHADER_WARNING_DXIL_IGNORING_OPERANDS, "Ignoring %u extra operands for record code %u.", record->operand_count - max_count, record->code); } static bool dxil_record_validate_operand_count(const struct dxil_record *record, unsigned int min_count, unsigned int max_count, struct sm6_parser *sm6) { dxil_record_validate_operand_max_count(record, max_count, sm6); return dxil_record_validate_operand_min_count(record, min_count, sm6); } static enum vkd3d_result sm6_parser_type_table_init(struct sm6_parser *sm6) { const struct dxil_record *record; size_t i, type_count, type_index; const struct dxil_block *block; char *struct_name = NULL; unsigned int j, count; struct sm6_type *type; uint64_t type_id; bool is_unique; sm6->p.location.line = 0; sm6->p.location.column = 0; if (!(block = sm6_parser_get_level_one_block(sm6, TYPE_BLOCK, &is_unique))) { WARN("No type definitions found.\n"); return VKD3D_OK; } if (!is_unique) WARN("Ignoring invalid extra type table(s).\n"); sm6->p.location.line = block->id; type_count = 0; for (i = 0; i < block->record_count; ++i) type_count += block->records[i]->code != TYPE_CODE_NUMENTRY && block->records[i]->code != TYPE_CODE_STRUCT_NAME; /* The type array must not be relocated. */ if (!(sm6->types = vkd3d_calloc(type_count, sizeof(*sm6->types)))) { ERR("Failed to allocate type array.\n"); return VKD3D_ERROR_OUT_OF_MEMORY; } for (i = 0; i < block->record_count; ++i) { sm6->p.location.column = i; record = block->records[i]; type = &sm6->types[sm6->type_count]; type_index = sm6->type_count; switch (record->code) { case TYPE_CODE_ARRAY: case TYPE_CODE_VECTOR: if (!dxil_record_validate_operand_count(record, 2, 2, sm6)) return VKD3D_ERROR_INVALID_SHADER; type->class = record->code == TYPE_CODE_ARRAY ? TYPE_CLASS_ARRAY : TYPE_CLASS_VECTOR; if (!(type->u.array.count = record->operands[0])) { TRACE("Setting unbounded for type %zu.\n", type_index); type->u.array.count = UINT_MAX; } if ((type_id = record->operands[1]) >= type_count) { WARN("Invalid contained type id %"PRIu64" for type %zu.\n", type_id, type_index); return VKD3D_ERROR_INVALID_SHADER; } type->u.array.elem_type = &sm6->types[type_id]; break; case TYPE_CODE_DOUBLE: dxil_record_validate_operand_max_count(record, 0, sm6); type->class = TYPE_CLASS_FLOAT; type->u.width = 64; break; case TYPE_CODE_FLOAT: dxil_record_validate_operand_max_count(record, 0, sm6); type->class = TYPE_CLASS_FLOAT; type->u.width = 32; break; case TYPE_CODE_FUNCTION: if (!dxil_record_validate_operand_min_count(record, 2, sm6)) return VKD3D_ERROR_INVALID_SHADER; if (record->operands[0]) FIXME("Unhandled vararg function type %zu.\n", type_index); type->class = TYPE_CLASS_FUNCTION; if ((type_id = record->operands[1]) >= type_count) { WARN("Invalid return type id %"PRIu64" for type %zu.\n", type_id, type_index); return VKD3D_ERROR_INVALID_SHADER; } count = record->operand_count - 2; if (vkd3d_object_range_overflow(sizeof(type->u.function), count, sizeof(type->u.function->param_types[0])) || !(type->u.function = vkd3d_malloc(offsetof(struct sm6_function_info, param_types[count])))) { ERR("Failed to allocate function parameter types.\n"); return VKD3D_ERROR_OUT_OF_MEMORY; } type->u.function->ret_type = &sm6->types[type_id]; type->u.function->param_count = count; for (j = 0; j < count; ++j) { if ((type_id = record->operands[j + 2]) >= type_count) { WARN("Invalid parameter type id %"PRIu64" for type %zu.\n", type_id, type_index); vkd3d_free(type->u.function); return VKD3D_ERROR_INVALID_SHADER; } type->u.function->param_types[j] = &sm6->types[type_id]; } break; case TYPE_CODE_HALF: dxil_record_validate_operand_max_count(record, 0, sm6); type->class = TYPE_CLASS_FLOAT; type->u.width = 16; break; case TYPE_CODE_INTEGER: { uint64_t width; if (!dxil_record_validate_operand_count(record, 1, 1, sm6)) return VKD3D_ERROR_INVALID_SHADER; type->class = TYPE_CLASS_INTEGER; switch ((width = record->operands[0])) { case 1: case 8: case 16: case 32: case 64: break; default: WARN("Invalid integer width %"PRIu64" for type %zu.\n", width, type_index); return VKD3D_ERROR_INVALID_SHADER; } type->u.width = width; break; } case TYPE_CODE_LABEL: type->class = TYPE_CLASS_LABEL; break; case TYPE_CODE_METADATA: type->class = TYPE_CLASS_METADATA; break; case TYPE_CODE_NUMENTRY: continue; case TYPE_CODE_POINTER: if (!dxil_record_validate_operand_count(record, 1, 2, sm6)) return VKD3D_ERROR_INVALID_SHADER; type->class = TYPE_CLASS_POINTER; if ((type_id = record->operands[0]) >= type_count) { WARN("Invalid pointee type id %"PRIu64" for type %zu.\n", type_id, type_index); return VKD3D_ERROR_INVALID_SHADER; } type->u.pointer.type = &sm6->types[type_id]; type->u.pointer.addr_space = (record->operand_count > 1) ? record->operands[1] : ADDRESS_SPACE_DEFAULT; break; case TYPE_CODE_STRUCT_ANON: case TYPE_CODE_STRUCT_NAMED: if (!dxil_record_validate_operand_min_count(record, 2, sm6)) return VKD3D_ERROR_INVALID_SHADER; if (record->code == TYPE_CODE_STRUCT_NAMED && !struct_name) { WARN("Missing struct name before struct type %zu.\n", type_index); return VKD3D_ERROR_INVALID_SHADER; } type->class = TYPE_CLASS_STRUCT; count = record->operand_count - 1; if (vkd3d_object_range_overflow(sizeof(type->u.struc), count, sizeof(type->u.struc->elem_types[0])) || !(type->u.struc = vkd3d_malloc(offsetof(struct sm6_struct_info, elem_types[count])))) { ERR("Failed to allocate struct element types.\n"); return VKD3D_ERROR_OUT_OF_MEMORY; } if (record->operands[0]) FIXME("Ignoring struct packed attribute.\n"); type->u.struc->elem_count = count; for (j = 0; j < count; ++j) { if ((type_id = record->operands[j + 1]) >= type_count) { WARN("Invalid contained type id %"PRIu64" for type %zu.\n", type_id, type_index); vkd3d_free(type->u.struc); return VKD3D_ERROR_INVALID_SHADER; } type->u.struc->elem_types[j] = &sm6->types[type_id]; } if (record->code == TYPE_CODE_STRUCT_ANON) { type->u.struc->name = NULL; break; } type->u.struc->name = struct_name; struct_name = NULL; break; case TYPE_CODE_STRUCT_NAME: if (!(struct_name = dxil_record_to_string(record, 0))) { ERR("Failed to allocate struct name.\n"); return VKD3D_ERROR_OUT_OF_MEMORY; } if (!struct_name[0]) WARN("Struct name is empty for type %zu.\n", type_index); continue; case TYPE_CODE_VOID: dxil_record_validate_operand_max_count(record, 0, sm6); type->class = TYPE_CLASS_VOID; break; default: FIXME("Unhandled type %u at index %zu.\n", record->code, type_index); return VKD3D_ERROR_INVALID_SHADER; } ++sm6->type_count; } assert(sm6->type_count == type_count); if (struct_name) { WARN("Unused struct name %s.\n", struct_name); vkd3d_free(struct_name); } return VKD3D_OK; } static inline bool sm6_type_is_void(const struct sm6_type *type) { return type->class == TYPE_CLASS_VOID; } static inline bool sm6_type_is_integer(const struct sm6_type *type) { return type->class == TYPE_CLASS_INTEGER; } static inline bool sm6_type_is_i8(const struct sm6_type *type) { return type->class == TYPE_CLASS_INTEGER && type->u.width == 8; } static inline bool sm6_type_is_i32(const struct sm6_type *type) { return type->class == TYPE_CLASS_INTEGER && type->u.width == 32; } static inline bool sm6_type_is_floating_point(const struct sm6_type *type) { return type->class == TYPE_CLASS_FLOAT; } static inline bool sm6_type_is_numeric(const struct sm6_type *type) { return type->class == TYPE_CLASS_INTEGER || type->class == TYPE_CLASS_FLOAT; } static inline bool sm6_type_is_pointer(const struct sm6_type *type) { return type->class == TYPE_CLASS_POINTER; } static bool sm6_type_is_numeric_aggregate(const struct sm6_type *type) { unsigned int i; switch (type->class) { case TYPE_CLASS_ARRAY: case TYPE_CLASS_VECTOR: return sm6_type_is_numeric(type->u.array.elem_type); case TYPE_CLASS_STRUCT: /* Do not handle nested structs. Support can be added if they show up. */ for (i = 0; i < type->u.struc->elem_count; ++i) if (!sm6_type_is_numeric(type->u.struc->elem_types[i])) return false; return true; default: return false; } } static inline bool sm6_type_is_struct(const struct sm6_type *type) { return type->class == TYPE_CLASS_STRUCT; } static inline bool sm6_type_is_function(const struct sm6_type *type) { return type->class == TYPE_CLASS_FUNCTION; } static inline bool sm6_type_is_function_pointer(const struct sm6_type *type) { return sm6_type_is_pointer(type) && sm6_type_is_function(type->u.pointer.type); } static inline bool sm6_type_is_handle(const struct sm6_type *type) { return sm6_type_is_struct(type) && !strcmp(type->u.struc->name, "dx.types.Handle"); } static inline const struct sm6_type *sm6_type_get_element_type(const struct sm6_type *type) { return (type->class == TYPE_CLASS_ARRAY || type->class == TYPE_CLASS_VECTOR) ? type->u.array.elem_type : type; } static const struct sm6_type *sm6_type_get_pointer_to_type(const struct sm6_type *type, enum bitcode_address_space addr_space, struct sm6_parser *sm6) { size_t i, start = type - sm6->types; const struct sm6_type *pointer_type; /* DXC seems usually to place the pointer type immediately after its pointee. */ for (i = (start + 1) % sm6->type_count; i != start; i = (i + 1) % sm6->type_count) { pointer_type = &sm6->types[i]; if (sm6_type_is_pointer(pointer_type) && pointer_type->u.pointer.type == type && pointer_type->u.pointer.addr_space == addr_space) return pointer_type; } return NULL; } /* Never returns null for elem_idx 0. */ static const struct sm6_type *sm6_type_get_scalar_type(const struct sm6_type *type, unsigned int elem_idx) { switch (type->class) { case TYPE_CLASS_ARRAY: case TYPE_CLASS_VECTOR: if (elem_idx >= type->u.array.count) return NULL; return sm6_type_get_scalar_type(type->u.array.elem_type, 0); case TYPE_CLASS_POINTER: return sm6_type_get_scalar_type(type->u.pointer.type, 0); case TYPE_CLASS_STRUCT: if (elem_idx >= type->u.struc->elem_count) return NULL; return sm6_type_get_scalar_type(type->u.struc->elem_types[elem_idx], 0); default: return type; } } static const struct sm6_type *sm6_parser_get_type(struct sm6_parser *sm6, uint64_t type_id) { if (type_id >= sm6->type_count) { WARN("Invalid type index %"PRIu64" at %zu.\n", type_id, sm6->value_count); vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_INVALID_TYPE_ID, "DXIL type id %"PRIu64" is invalid.", type_id); return NULL; } return &sm6->types[type_id]; } static int global_symbol_compare(const void *a, const void *b) { return vkd3d_u32_compare(((const struct sm6_symbol *)a)->id, ((const struct sm6_symbol *)b)->id); } static enum vkd3d_result sm6_parser_symtab_init(struct sm6_parser *sm6) { const struct dxil_record *record; const struct dxil_block *block; struct sm6_symbol *symbol; size_t i, count; bool is_unique; sm6->p.location.line = 0; sm6->p.location.column = 0; if (!(block = sm6_parser_get_level_one_block(sm6, VALUE_SYMTAB_BLOCK, &is_unique))) { /* There should always be at least one symbol: the name of the entry point function. */ WARN("No value symtab block found.\n"); return VKD3D_ERROR_INVALID_SHADER; } if (!is_unique) FIXME("Ignoring extra value symtab block(s).\n"); sm6->p.location.line = block->id; for (i = 0, count = 0; i < block->record_count; ++i) count += block->records[i]->code == VST_CODE_ENTRY; if (!(sm6->global_symbols = vkd3d_calloc(count, sizeof(*sm6->global_symbols)))) { ERR("Failed to allocate global symbols.\n"); return VKD3D_ERROR_OUT_OF_MEMORY; } for (i = 0; i < block->record_count; ++i) { sm6->p.location.column = i; record = block->records[i]; if (record->code != VST_CODE_ENTRY) { FIXME("Unhandled symtab code %u.\n", record->code); continue; } if (!dxil_record_validate_operand_min_count(record, 1, sm6)) continue; symbol = &sm6->global_symbols[sm6->global_symbol_count]; symbol->id = record->operands[0]; if (!(symbol->name = dxil_record_to_string(record, 1))) { ERR("Failed to allocate symbol name.\n"); return VKD3D_ERROR_OUT_OF_MEMORY; } ++sm6->global_symbol_count; } sm6->p.location.column = block->record_count; qsort(sm6->global_symbols, sm6->global_symbol_count, sizeof(*sm6->global_symbols), global_symbol_compare); for (i = 1; i < sm6->global_symbol_count; ++i) { if (sm6->global_symbols[i].id == sm6->global_symbols[i - 1].id) { WARN("Invalid duplicate symbol id %u.\n", sm6->global_symbols[i].id); return VKD3D_ERROR_INVALID_SHADER; } } return VKD3D_OK; } static const char *sm6_parser_get_global_symbol_name(const struct sm6_parser *sm6, size_t id) { size_t i, start; /* id == array index is normally true */ i = start = id % sm6->global_symbol_count; do { if (sm6->global_symbols[i].id == id) return sm6->global_symbols[i].name; i = (i + 1) % sm6->global_symbol_count; } while (i != start); return NULL; } static unsigned int register_get_uint_value(const struct vkd3d_shader_register *reg) { if (!register_is_constant(reg) || !data_type_is_integer(reg->data_type)) return UINT_MAX; if (reg->dimension == VSIR_DIMENSION_VEC4) WARN("Returning vec4.x.\n"); if (reg->type == VKD3DSPR_IMMCONST64) { if (reg->u.immconst_uint64[0] > UINT_MAX) FIXME("Truncating 64-bit value.\n"); return reg->u.immconst_uint64[0]; } return reg->u.immconst_uint[0]; } static inline bool sm6_value_is_function_dcl(const struct sm6_value *value) { return value->value_type == VALUE_TYPE_FUNCTION; } static inline bool sm6_value_is_dx_intrinsic_dcl(const struct sm6_value *fn) { assert(sm6_value_is_function_dcl(fn)); return fn->u.function.is_prototype && !strncmp(fn->u.function.name, "dx.op.", 6); } static inline struct sm6_value *sm6_parser_get_current_value(const struct sm6_parser *sm6) { assert(sm6->value_count < sm6->value_capacity); return &sm6->values[sm6->value_count]; } static inline bool sm6_value_is_register(const struct sm6_value *value) { return value->value_type == VALUE_TYPE_REG; } static inline bool sm6_value_is_constant(const struct sm6_value *value) { return sm6_value_is_register(value) && register_is_constant(&value->u.reg); } static inline bool sm6_value_is_undef(const struct sm6_value *value) { return sm6_value_is_register(value) && value->u.reg.type == VKD3DSPR_UNDEF; } static inline unsigned int sm6_value_get_constant_uint(const struct sm6_value *value) { if (!sm6_value_is_constant(value)) return UINT_MAX; return register_get_uint_value(&value->u.reg); } static unsigned int sm6_parser_alloc_ssa_id(struct sm6_parser *sm6) { return sm6->ssa_next_id++; } static struct vkd3d_shader_src_param *instruction_src_params_alloc(struct vkd3d_shader_instruction *ins, unsigned int count, struct sm6_parser *sm6) { struct vkd3d_shader_src_param *params = shader_parser_get_src_params(&sm6->p, count); if (!params) { ERR("Failed to allocate src params.\n"); vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_OUT_OF_MEMORY, "Out of memory allocating instruction src paramaters."); return NULL; } ins->src = params; ins->src_count = count; return params; } static struct vkd3d_shader_dst_param *instruction_dst_params_alloc(struct vkd3d_shader_instruction *ins, unsigned int count, struct sm6_parser *sm6) { struct vkd3d_shader_dst_param *params = shader_parser_get_dst_params(&sm6->p, count); if (!params) { ERR("Failed to allocate dst params.\n"); vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_OUT_OF_MEMORY, "Out of memory allocating instruction dst paramaters."); return NULL; } ins->dst = params; ins->dst_count = count; return params; } static void register_init_with_id(struct vkd3d_shader_register *reg, enum vkd3d_shader_register_type reg_type, enum vkd3d_data_type data_type, unsigned int id) { vsir_register_init(reg, reg_type, data_type, 1); reg->idx[0].offset = id; } static enum vkd3d_data_type vkd3d_data_type_from_sm6_type(const struct sm6_type *type) { if (type->class == TYPE_CLASS_INTEGER) { switch (type->u.width) { case 8: return VKD3D_DATA_UINT8; case 32: return VKD3D_DATA_UINT; default: FIXME("Unhandled width %u.\n", type->u.width); return VKD3D_DATA_UINT; } } else if (type->class == TYPE_CLASS_FLOAT) { switch (type->u.width) { case 32: return VKD3D_DATA_FLOAT; case 64: return VKD3D_DATA_DOUBLE; default: FIXME("Unhandled width %u.\n", type->u.width); return VKD3D_DATA_FLOAT; } } FIXME("Unhandled type %u.\n", type->class); return VKD3D_DATA_UINT; } static void register_init_ssa_scalar(struct vkd3d_shader_register *reg, const struct sm6_type *type, struct sm6_parser *sm6) { enum vkd3d_data_type data_type; unsigned int id; id = sm6_parser_alloc_ssa_id(sm6); data_type = vkd3d_data_type_from_sm6_type(sm6_type_get_scalar_type(type, 0)); register_init_with_id(reg, VKD3DSPR_SSA, data_type, id); } static void dst_param_init(struct vkd3d_shader_dst_param *param) { param->write_mask = VKD3DSP_WRITEMASK_0; param->modifiers = 0; param->shift = 0; } static inline void dst_param_init_scalar(struct vkd3d_shader_dst_param *param, unsigned int component_idx) { param->write_mask = 1u << component_idx; param->modifiers = 0; param->shift = 0; } static void dst_param_init_ssa_scalar(struct vkd3d_shader_dst_param *param, const struct sm6_type *type, struct sm6_parser *sm6) { dst_param_init(param); register_init_ssa_scalar(¶m->reg, type, sm6); } static inline void src_param_init(struct vkd3d_shader_src_param *param) { param->swizzle = VKD3D_SHADER_SWIZZLE(X, X, X, X); param->modifiers = VKD3DSPSM_NONE; } static void src_param_init_scalar(struct vkd3d_shader_src_param *param, unsigned int component_idx) { param->swizzle = vkd3d_shader_create_swizzle(component_idx, component_idx, component_idx, component_idx); param->modifiers = VKD3DSPSM_NONE; } static void src_param_init_from_value(struct vkd3d_shader_src_param *param, const struct sm6_value *src) { src_param_init(param); param->reg = src->u.reg; } static void register_address_init(struct vkd3d_shader_register *reg, const struct sm6_value *address, unsigned int idx, struct sm6_parser *sm6) { assert(idx < ARRAY_SIZE(reg->idx)); if (sm6_value_is_constant(address)) { reg->idx[idx].offset = sm6_value_get_constant_uint(address); } else if (sm6_value_is_undef(address)) { reg->idx[idx].offset = 0; } else { struct vkd3d_shader_src_param *rel_addr = shader_parser_get_src_params(&sm6->p, 1); if (rel_addr) src_param_init_from_value(rel_addr, address); reg->idx[idx].offset = 0; reg->idx[idx].rel_addr = rel_addr; } } static void instruction_dst_param_init_ssa_scalar(struct vkd3d_shader_instruction *ins, struct sm6_parser *sm6) { struct vkd3d_shader_dst_param *param = instruction_dst_params_alloc(ins, 1, sm6); struct sm6_value *dst = sm6_parser_get_current_value(sm6); dst_param_init_ssa_scalar(param, dst->type, sm6); param->write_mask = VKD3DSP_WRITEMASK_0; dst->u.reg = param->reg; } /* Recurse through the block tree while maintaining a current value count. The current * count is the sum of the global count plus all declarations within the current function. * Store into value_capacity the highest count seen. */ static size_t sm6_parser_compute_max_value_count(struct sm6_parser *sm6, const struct dxil_block *block, size_t value_count) { size_t i, old_value_count = value_count; if (block->id == MODULE_BLOCK) value_count = size_add_with_overflow_check(value_count, dxil_block_compute_module_decl_count(block)); for (i = 0; i < block->child_block_count; ++i) value_count = sm6_parser_compute_max_value_count(sm6, block->child_blocks[i], value_count); switch (block->id) { case CONSTANTS_BLOCK: /* Function local constants are contained in a child block of the function block. */ value_count = size_add_with_overflow_check(value_count, dxil_block_compute_constants_count(block)); break; case FUNCTION_BLOCK: /* A function must start with a block count, which emits no value. This formula is likely to * overestimate the value count somewhat, but this should be no problem. */ value_count = size_add_with_overflow_check(value_count, max(block->record_count, 1u) - 1); sm6->value_capacity = max(sm6->value_capacity, value_count); sm6->functions[sm6->function_count].value_count = value_count; /* The value count returns to its previous value after handling a function. */ if (value_count < SIZE_MAX) value_count = old_value_count; break; default: break; } return value_count; } static size_t sm6_parser_get_value_index(struct sm6_parser *sm6, uint64_t idx) { size_t i; /* The value relative index is 32 bits. */ if (idx > UINT32_MAX) WARN("Ignoring upper 32 bits of relative index.\n"); i = (uint32_t)sm6->value_count - (uint32_t)idx; /* This may underflow to produce a forward reference, but it must not exceeed the final value count. */ if (i >= sm6->cur_max_value) { WARN("Invalid value index %"PRIx64" at %zu.\n", idx, sm6->value_count); vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_INVALID_OPERAND, "Invalid value relative index %u.", (unsigned int)idx); return SIZE_MAX; } if (i == sm6->value_count) { WARN("Invalid value self-reference at %zu.\n", sm6->value_count); vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_INVALID_OPERAND, "Invalid value self-reference."); return SIZE_MAX; } return i; } static size_t sm6_parser_get_value_idx_by_ref(struct sm6_parser *sm6, const struct dxil_record *record, const struct sm6_type *fwd_type, unsigned int *rec_idx) { unsigned int idx; uint64_t val_ref; size_t operand; idx = *rec_idx; if (!dxil_record_validate_operand_min_count(record, idx + 1, sm6)) return SIZE_MAX; val_ref = record->operands[idx++]; operand = sm6_parser_get_value_index(sm6, val_ref); if (operand == SIZE_MAX) return SIZE_MAX; if (operand >= sm6->value_count) { if (!fwd_type) { /* Forward references are followed by a type id unless an earlier operand set the type, * or it is contained in a function declaration. */ if (!dxil_record_validate_operand_min_count(record, idx + 1, sm6)) return SIZE_MAX; if (!(fwd_type = sm6_parser_get_type(sm6, record->operands[idx++]))) return SIZE_MAX; } FIXME("Forward value references are not supported yet.\n"); vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_INVALID_OPERAND, "Unsupported value forward reference."); return SIZE_MAX; } *rec_idx = idx; return operand; } static const struct sm6_value *sm6_parser_get_value_by_ref(struct sm6_parser *sm6, const struct dxil_record *record, const struct sm6_type *type, unsigned int *rec_idx) { size_t operand = sm6_parser_get_value_idx_by_ref(sm6, record, type, rec_idx); return operand == SIZE_MAX ? NULL : &sm6->values[operand]; } static bool sm6_parser_declare_function(struct sm6_parser *sm6, const struct dxil_record *record) { const unsigned int max_count = 15; const struct sm6_type *ret_type; struct sm6_value *fn; unsigned int i, j; if (!dxil_record_validate_operand_count(record, 8, max_count, sm6)) return false; fn = sm6_parser_get_current_value(sm6); fn->value_type = VALUE_TYPE_FUNCTION; if (!(fn->u.function.name = sm6_parser_get_global_symbol_name(sm6, sm6->value_count))) { WARN("Missing symbol name for function %zu.\n", sm6->value_count); fn->u.function.name = ""; } if (!(fn->type = sm6_parser_get_type(sm6, record->operands[0]))) return false; if (!sm6_type_is_function(fn->type)) { WARN("Type is not a function.\n"); return false; } ret_type = fn->type->u.function->ret_type; if (!(fn->type = sm6_type_get_pointer_to_type(fn->type, ADDRESS_SPACE_DEFAULT, sm6))) { WARN("Failed to get pointer type for type %u.\n", fn->type->class); return false; } if (record->operands[1]) WARN("Ignoring calling convention %#"PRIx64".\n", record->operands[1]); fn->u.function.is_prototype = !!record->operands[2]; if (record->operands[3]) WARN("Ignoring linkage %#"PRIx64".\n", record->operands[3]); if (record->operands[4] > UINT_MAX) WARN("Invalid attributes id %#"PRIx64".\n", record->operands[4]); /* 1-based index. */ if ((fn->u.function.attribs_id = record->operands[4])) TRACE("Ignoring function attributes.\n"); /* These always seem to be zero. */ for (i = 5, j = 0; i < min(record->operand_count, max_count); ++i) j += !!record->operands[i]; if (j) WARN("Ignoring %u operands.\n", j); if (sm6_value_is_dx_intrinsic_dcl(fn) && !sm6_type_is_void(ret_type) && !sm6_type_is_numeric(ret_type) && !sm6_type_is_numeric_aggregate(ret_type) && !sm6_type_is_handle(ret_type)) { WARN("Unexpected return type for dx intrinsic function '%s'.\n", fn->u.function.name); } ++sm6->value_count; return true; } static inline uint64_t decode_rotated_signed_value(uint64_t value) { if (value != 1) { bool neg = value & 1; value >>= 1; return neg ? -value : value; } return value << 63; } static inline float bitcast_uint64_to_float(uint64_t value) { union { uint32_t uint32_value; float float_value; } u; u.uint32_value = value; return u.float_value; } static inline double bitcast_uint64_to_double(uint64_t value) { union { uint64_t uint64_value; double double_value; } u; u.uint64_value = value; return u.double_value; } static enum vkd3d_result sm6_parser_constants_init(struct sm6_parser *sm6, const struct dxil_block *block) { enum vkd3d_shader_register_type reg_type = VKD3DSPR_INVALID; const struct sm6_type *type, *elem_type; enum vkd3d_data_type reg_data_type; const struct dxil_record *record; struct sm6_value *dst; size_t i, value_idx; uint64_t value; for (i = 0, type = NULL; i < block->record_count; ++i) { sm6->p.location.column = i; record = block->records[i]; value_idx = sm6->value_count; if (record->code == CST_CODE_SETTYPE) { if (!dxil_record_validate_operand_count(record, 1, 1, sm6)) return VKD3D_ERROR_INVALID_SHADER; if (!(type = sm6_parser_get_type(sm6, record->operands[0]))) return VKD3D_ERROR_INVALID_SHADER; elem_type = sm6_type_get_element_type(type); if (sm6_type_is_numeric(elem_type)) { reg_data_type = vkd3d_data_type_from_sm6_type(elem_type); reg_type = elem_type->u.width > 32 ? VKD3DSPR_IMMCONST64 : VKD3DSPR_IMMCONST; } else { reg_data_type = VKD3D_DATA_UNUSED; reg_type = VKD3DSPR_INVALID; } if (i == block->record_count - 1) WARN("Unused SETTYPE record.\n"); continue; } if (!type) { WARN("Constant record %zu has no type.\n", value_idx); return VKD3D_ERROR_INVALID_SHADER; } dst = sm6_parser_get_current_value(sm6); dst->type = type; dst->value_type = VALUE_TYPE_REG; dst->u.reg.type = reg_type; dst->u.reg.dimension = VSIR_DIMENSION_SCALAR; dst->u.reg.data_type = reg_data_type; switch (record->code) { case CST_CODE_NULL: /* Register constant data is already zero-filled. */ break; case CST_CODE_INTEGER: if (!dxil_record_validate_operand_count(record, 1, 1, sm6)) return VKD3D_ERROR_INVALID_SHADER; if (!sm6_type_is_integer(type)) { WARN("Invalid integer of non-integer type %u at constant idx %zu.\n", type->class, value_idx); return VKD3D_ERROR_INVALID_SHADER; } value = decode_rotated_signed_value(record->operands[0]); if (type->u.width <= 32) dst->u.reg.u.immconst_uint[0] = value & ((1ull << type->u.width) - 1); else dst->u.reg.u.immconst_uint64[0] = value; break; case CST_CODE_FLOAT: if (!dxil_record_validate_operand_count(record, 1, 1, sm6)) return VKD3D_ERROR_INVALID_SHADER; if (!sm6_type_is_floating_point(type)) { WARN("Invalid float of non-fp type %u at constant idx %zu.\n", type->class, value_idx); return VKD3D_ERROR_INVALID_SHADER; } if (type->u.width == 16) FIXME("Half float type is not supported yet.\n"); else if (type->u.width == 32) dst->u.reg.u.immconst_float[0] = bitcast_uint64_to_float(record->operands[0]); else if (type->u.width == 64) dst->u.reg.u.immconst_double[0] = bitcast_uint64_to_double(record->operands[0]); else vkd3d_unreachable(); break; case CST_CODE_DATA: WARN("Unhandled constant array.\n"); break; case CST_CODE_UNDEF: dxil_record_validate_operand_max_count(record, 0, sm6); dst->u.reg.type = VKD3DSPR_UNDEF; /* Mark as explicitly undefined, not the result of a missing constant code or instruction. */ dst->is_undefined = true; break; default: FIXME("Unhandled constant code %u.\n", record->code); dst->u.reg.type = VKD3DSPR_UNDEF; break; } ++sm6->value_count; } return VKD3D_OK; } static struct vkd3d_shader_instruction *sm6_parser_require_space(struct sm6_parser *sm6, size_t extra) { if (!shader_instruction_array_reserve(&sm6->p.instructions, sm6->p.instructions.count + extra)) { ERR("Failed to allocate instruction.\n"); return NULL; } return &sm6->p.instructions.elements[sm6->p.instructions.count]; } /* Space should be reserved before calling this. It is intended to require no checking of the returned pointer. */ static struct vkd3d_shader_instruction *sm6_parser_add_instruction(struct sm6_parser *sm6, enum vkd3d_shader_opcode handler_idx) { struct vkd3d_shader_instruction *ins = sm6_parser_require_space(sm6, 1); assert(ins); vsir_instruction_init(ins, &sm6->p.location, handler_idx); ++sm6->p.instructions.count; return ins; } static enum vkd3d_result sm6_parser_globals_init(struct sm6_parser *sm6) { const struct dxil_block *block = &sm6->root_block; const struct dxil_record *record; uint64_t version; size_t i; sm6->p.location.line = block->id; sm6->p.location.column = 0; for (i = 0; i < block->record_count; ++i) { sm6->p.location.column = i; record = block->records[i]; switch (record->code) { case MODULE_CODE_FUNCTION: if (!sm6_parser_declare_function(sm6, record)) { vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_INVALID_FUNCTION_DCL, "A DXIL function declaration is invalid."); return VKD3D_ERROR_INVALID_SHADER; } break; case MODULE_CODE_GLOBALVAR: FIXME("Global variables are not implemented yet.\n"); break; case MODULE_CODE_VERSION: if (!dxil_record_validate_operand_count(record, 1, 1, sm6)) return VKD3D_ERROR_INVALID_SHADER; if ((version = record->operands[0]) != 1) { FIXME("Unsupported format version %#"PRIx64".\n", version); vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_UNSUPPORTED_BITCODE_FORMAT, "Bitcode format version %#"PRIx64" is unsupported.", version); return VKD3D_ERROR_INVALID_SHADER; } break; default: break; } } return VKD3D_OK; } static void dst_param_io_init(struct vkd3d_shader_dst_param *param, const struct signature_element *e, enum vkd3d_shader_register_type reg_type) { enum vkd3d_shader_component_type component_type; param->write_mask = e->mask; param->modifiers = 0; param->shift = 0; /* DXIL types do not have signedness. Load signed elements as unsigned. */ component_type = e->component_type == VKD3D_SHADER_COMPONENT_INT ? VKD3D_SHADER_COMPONENT_UINT : e->component_type; vsir_register_init(¶m->reg, reg_type, vkd3d_data_type_from_component_type(component_type), 0); } static void sm6_parser_init_signature(struct sm6_parser *sm6, const struct shader_signature *s, enum vkd3d_shader_register_type reg_type, struct vkd3d_shader_dst_param *params) { struct vkd3d_shader_dst_param *param; const struct signature_element *e; unsigned int i; for (i = 0; i < s->element_count; ++i) { e = &s->elements[i]; param = ¶ms[i]; dst_param_io_init(param, e, reg_type); param->reg.idx[0].offset = i; param->reg.idx_count = 1; } } static void sm6_parser_emit_signature(struct sm6_parser *sm6, const struct shader_signature *s, enum vkd3d_shader_opcode handler_idx, enum vkd3d_shader_opcode siv_handler_idx, struct vkd3d_shader_dst_param *params) { struct vkd3d_shader_instruction *ins; struct vkd3d_shader_dst_param *param; const struct signature_element *e; unsigned int i; for (i = 0; i < s->element_count; ++i) { e = &s->elements[i]; /* Do not check e->used_mask because in some cases it is zero for used elements. * TODO: scan ahead for used I/O elements. */ if (e->sysval_semantic != VKD3D_SHADER_SV_NONE && e->sysval_semantic != VKD3D_SHADER_SV_TARGET) { ins = sm6_parser_add_instruction(sm6, siv_handler_idx); param = &ins->declaration.register_semantic.reg; ins->declaration.register_semantic.sysval_semantic = vkd3d_siv_from_sysval(e->sysval_semantic); } else { ins = sm6_parser_add_instruction(sm6, handler_idx); param = &ins->declaration.dst; } /* TODO: set the interpolation mode when signatures are loaded from DXIL metadata. */ ins->flags = (handler_idx == VKD3DSIH_DCL_INPUT_PS) ? VKD3DSIM_LINEAR_NOPERSPECTIVE : 0; *param = params[i]; } } static void sm6_parser_init_output_signature(struct sm6_parser *sm6, const struct shader_signature *output_signature) { sm6_parser_init_signature(sm6, output_signature, (sm6->p.shader_version.type == VKD3D_SHADER_TYPE_PIXEL) ? VKD3DSPR_COLOROUT : VKD3DSPR_OUTPUT, sm6->output_params); } static void sm6_parser_init_input_signature(struct sm6_parser *sm6, const struct shader_signature *input_signature) { sm6_parser_init_signature(sm6, input_signature, VKD3DSPR_INPUT, sm6->input_params); } static void sm6_parser_emit_output_signature(struct sm6_parser *sm6, const struct shader_signature *output_signature) { sm6_parser_emit_signature(sm6, output_signature, VKD3DSIH_DCL_OUTPUT, VKD3DSIH_DCL_OUTPUT_SIV, sm6->output_params); } static void sm6_parser_emit_input_signature(struct sm6_parser *sm6, const struct shader_signature *input_signature) { sm6_parser_emit_signature(sm6, input_signature, (sm6->p.shader_version.type == VKD3D_SHADER_TYPE_PIXEL) ? VKD3DSIH_DCL_INPUT_PS : VKD3DSIH_DCL_INPUT, (sm6->p.shader_version.type == VKD3D_SHADER_TYPE_PIXEL) ? VKD3DSIH_DCL_INPUT_PS_SIV : VKD3DSIH_DCL_INPUT_SIV, sm6->input_params); } static const struct sm6_value *sm6_parser_next_function_definition(struct sm6_parser *sm6) { size_t i, count = sm6->function_count; for (i = 0; i < sm6->value_count; ++i) { if (sm6_type_is_function_pointer(sm6->values[i].type) && !sm6->values[i].u.function.is_prototype && !count--) break; } if (i == sm6->value_count) return NULL; ++sm6->function_count; return &sm6->values[i]; } static struct sm6_block *sm6_block_create() { struct sm6_block *block = vkd3d_calloc(1, sizeof(*block)); return block; } static void sm6_parser_emit_dx_load_input(struct sm6_parser *sm6, struct sm6_block *code_block, enum dx_intrinsic_opcode op, const struct sm6_value **operands, struct vkd3d_shader_instruction *ins) { struct vkd3d_shader_src_param *src_param; const struct shader_signature *signature; unsigned int row_index, column_index; const struct signature_element *e; row_index = sm6_value_get_constant_uint(operands[0]); column_index = sm6_value_get_constant_uint(operands[2]); vsir_instruction_init(ins, &sm6->p.location, VKD3DSIH_MOV); signature = &sm6->p.shader_desc.input_signature; if (row_index >= signature->element_count) { WARN("Invalid row index %u.\n", row_index); vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_INVALID_OPERAND, "Invalid input row index %u.", row_index); return; } e = &signature->elements[row_index]; src_param = instruction_src_params_alloc(ins, 1, sm6); src_param->reg = sm6->input_params[row_index].reg; src_param_init_scalar(src_param, column_index); if (e->register_count > 1) register_address_init(&src_param->reg, operands[1], 0, sm6); instruction_dst_param_init_ssa_scalar(ins, sm6); } static void sm6_parser_emit_dx_store_output(struct sm6_parser *sm6, struct sm6_block *code_block, enum dx_intrinsic_opcode op, const struct sm6_value **operands, struct vkd3d_shader_instruction *ins) { struct vkd3d_shader_src_param *src_param; struct vkd3d_shader_dst_param *dst_param; const struct shader_signature *signature; unsigned int row_index, column_index; const struct signature_element *e; const struct sm6_value *value; row_index = sm6_value_get_constant_uint(operands[0]); column_index = sm6_value_get_constant_uint(operands[2]); signature = &sm6->p.shader_desc.output_signature; if (row_index >= signature->element_count) { WARN("Invalid row index %u.\n", row_index); vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_INVALID_OPERAND, "Invalid output row index %u.", row_index); return; } e = &signature->elements[row_index]; if (column_index >= VKD3D_VEC4_SIZE) { WARN("Invalid column index %u.\n", column_index); vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_INVALID_OPERAND, "Invalid output column index %u.", column_index); return; } value = operands[3]; if (!sm6_value_is_register(value)) { WARN("Source value is not a register.\n"); vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_INVALID_OPERAND, "Expected store operation source to be a register."); return; } vsir_instruction_init(ins, &sm6->p.location, VKD3DSIH_MOV); if (!(dst_param = instruction_dst_params_alloc(ins, 1, sm6))) return; dst_param_init_scalar(dst_param, column_index); dst_param->reg = sm6->output_params[row_index].reg; if (e->register_count > 1) register_address_init(&dst_param->reg, operands[1], 0, sm6); if ((src_param = instruction_src_params_alloc(ins, 1, sm6))) src_param_init_from_value(src_param, value); } struct sm6_dx_opcode_info { const char ret_type; const char *operand_info; void (*handler)(struct sm6_parser *, struct sm6_block *, enum dx_intrinsic_opcode, const struct sm6_value **, struct vkd3d_shader_instruction *); }; /* 8 -> int8 i -> int32 v -> void o -> overloaded */ static const struct sm6_dx_opcode_info sm6_dx_op_table[] = { [DX_LOAD_INPUT ] = {'o', "ii8i", sm6_parser_emit_dx_load_input}, [DX_STORE_OUTPUT ] = {'v', "ii8o", sm6_parser_emit_dx_store_output}, }; static bool sm6_parser_validate_operand_type(struct sm6_parser *sm6, const struct sm6_type *type, char info_type) { switch (info_type) { case 0: FIXME("Invalid operand count.\n"); return false; case '8': return sm6_type_is_i8(type); case 'i': return sm6_type_is_i32(type); case 'v': return !type; case 'o': /* TODO: some type checking may be possible */ return true; default: FIXME("Unhandled operand code '%c'.\n", info_type); return false; } } static bool sm6_parser_validate_dx_op(struct sm6_parser *sm6, enum dx_intrinsic_opcode op, const char *name, const struct sm6_value **operands, unsigned int operand_count, struct sm6_value *dst) { const struct sm6_dx_opcode_info *info; unsigned int i; info = &sm6_dx_op_table[op]; if (!sm6_parser_validate_operand_type(sm6, dst->type, info->ret_type)) { WARN("Failed to validate return type for dx intrinsic id %u, '%s'.\n", op, name); /* Return type validation failure is not so critical. We only need to set * a data type for the SSA result. */ } for (i = 0; i < operand_count; ++i) { const struct sm6_value *value = operands[i]; if (!sm6_value_is_register(value) || !sm6_parser_validate_operand_type(sm6, value->type, info->operand_info[i])) { WARN("Failed to validate operand %u for dx intrinsic id %u, '%s'.\n", i + 1, op, name); vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_INVALID_OPERAND, "Operand %u for call to dx intrinsic function '%s' is invalid.", i + 1, name); return false; } } if (info->operand_info[operand_count]) { WARN("Missing operands for dx intrinsic id %u, '%s'.\n", op, name); vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_INVALID_OPERAND_COUNT, "Call to dx intrinsic function '%s' has missing operands.", name); return false; } return true; } static void sm6_parser_emit_unhandled(struct sm6_parser *sm6, struct vkd3d_shader_instruction *ins, struct sm6_value *dst) { const struct sm6_type *type; ins->handler_idx = VKD3DSIH_NOP; if (!dst->type) return; type = sm6_type_get_scalar_type(dst->type, 0); vsir_register_init(&dst->u.reg, VKD3DSPR_UNDEF, vkd3d_data_type_from_sm6_type(type), 0); /* dst->is_undefined is not set here because it flags only explicitly undefined values. */ } static void sm6_parser_decode_dx_op(struct sm6_parser *sm6, struct sm6_block *code_block, enum dx_intrinsic_opcode op, const char *name, const struct sm6_value **operands, unsigned int operand_count, struct vkd3d_shader_instruction *ins, struct sm6_value *dst) { if (op >= ARRAY_SIZE(sm6_dx_op_table) || !sm6_dx_op_table[op].operand_info) { FIXME("Unhandled dx intrinsic function id %u, '%s'.\n", op, name); vkd3d_shader_parser_warning(&sm6->p, VKD3D_SHADER_WARNING_DXIL_UNHANDLED_INTRINSIC, "Call to intrinsic function %s is unhandled.", name); sm6_parser_emit_unhandled(sm6, ins, dst); return; } if (sm6_parser_validate_dx_op(sm6, op, name, operands, operand_count, dst)) sm6_dx_op_table[op].handler(sm6, code_block, op, operands, ins); else sm6_parser_emit_unhandled(sm6, ins, dst); } static void sm6_parser_emit_call(struct sm6_parser *sm6, const struct dxil_record *record, struct sm6_block *code_block, struct vkd3d_shader_instruction *ins, struct sm6_value *dst) { const struct sm6_value *operands[DXIL_OP_MAX_OPERANDS]; const struct sm6_value *fn_value, *op_value; unsigned int i = 1, j, operand_count; const struct sm6_type *type = NULL; uint64_t call_conv; if (!dxil_record_validate_operand_min_count(record, 2, sm6)) return; /* TODO: load the 1-based attributes index from record->operands[0] and validate against attribute count. */ if ((call_conv = record->operands[i++]) & CALL_CONV_FLAG_EXPLICIT_TYPE) type = sm6_parser_get_type(sm6, record->operands[i++]); if (call_conv &= ~CALL_CONV_FLAG_EXPLICIT_TYPE) WARN("Ignoring calling convention %#"PRIx64".\n", call_conv); if (!(fn_value = sm6_parser_get_value_by_ref(sm6, record, NULL, &i))) return; if (!sm6_value_is_function_dcl(fn_value)) { WARN("Function target value is not a function declaration.\n"); vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_INVALID_OPERAND, "Function call target value is not a function declaration."); return; } if (type && type != fn_value->type->u.pointer.type) WARN("Explicit call type does not match function type.\n"); type = fn_value->type->u.pointer.type; if (!sm6_type_is_void(type->u.function->ret_type)) dst->type = type->u.function->ret_type; operand_count = type->u.function->param_count; if (operand_count > ARRAY_SIZE(operands)) { WARN("Ignoring %zu operands.\n", operand_count - ARRAY_SIZE(operands)); vkd3d_shader_parser_warning(&sm6->p, VKD3D_SHADER_WARNING_DXIL_IGNORING_OPERANDS, "Ignoring %zu operands for function call.", operand_count - ARRAY_SIZE(operands)); operand_count = ARRAY_SIZE(operands); } for (j = 0; j < operand_count; ++j) { if (!(operands[j] = sm6_parser_get_value_by_ref(sm6, record, type->u.function->param_types[j], &i))) return; } if ((j = record->operand_count - i)) { WARN("Ignoring %u operands beyond the function parameter list.\n", j); vkd3d_shader_parser_warning(&sm6->p, VKD3D_SHADER_WARNING_DXIL_IGNORING_OPERANDS, "Ignoring %u function call operands beyond the parameter list.", j); } if (!fn_value->u.function.is_prototype) { FIXME("Unhandled call to local function.\n"); vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_INVALID_OPERAND, "Call to a local function is unsupported."); return; } if (!sm6_value_is_dx_intrinsic_dcl(fn_value)) WARN("External function is not a dx intrinsic.\n"); if (!operand_count) { WARN("Missing dx intrinsic function id.\n"); vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_INVALID_OPERAND_COUNT, "The id for a dx intrinsic function is missing."); return; } op_value = operands[0]; if (!sm6_value_is_constant(op_value) || !sm6_type_is_integer(op_value->type)) { WARN("dx intrinsic function id is not a constant int.\n"); vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_INVALID_OPERAND, "Expected a constant integer dx intrinsic function id."); return; } sm6_parser_decode_dx_op(sm6, code_block, register_get_uint_value(&op_value->u.reg), fn_value->u.function.name, &operands[1], operand_count - 1, ins, dst); } static void sm6_parser_emit_ret(struct sm6_parser *sm6, const struct dxil_record *record, struct sm6_block *code_block, struct vkd3d_shader_instruction *ins) { if (!dxil_record_validate_operand_count(record, 0, 1, sm6)) return; if (record->operand_count) FIXME("Non-void return is not implemented.\n"); ins->handler_idx = VKD3DSIH_NOP; } static enum vkd3d_result sm6_parser_function_init(struct sm6_parser *sm6, const struct dxil_block *block, struct sm6_function *function) { struct vkd3d_shader_instruction *ins; const struct dxil_record *record; bool ret_found, is_terminator; struct sm6_block *code_block; struct sm6_value *dst; size_t i, block_idx; if (sm6->function_count) { FIXME("Multiple functions are not supported yet.\n"); return VKD3D_ERROR_INVALID_SHADER; } if (!(function->declaration = sm6_parser_next_function_definition(sm6))) { WARN("Failed to find definition to match function body.\n"); return VKD3D_ERROR_INVALID_SHADER; } if (block->record_count < 2) { /* It should contain at least a block count and a RET instruction. */ WARN("Invalid function block record count %zu.\n", block->record_count); return VKD3D_ERROR_INVALID_SHADER; } if (block->records[0]->code != FUNC_CODE_DECLAREBLOCKS || !block->records[0]->operand_count || block->records[0]->operands[0] > UINT_MAX) { WARN("Block count declaration not found or invalid.\n"); return VKD3D_ERROR_INVALID_SHADER; } if (!(function->block_count = block->records[0]->operands[0])) { WARN("Function contains no blocks.\n"); return VKD3D_ERROR_INVALID_SHADER; } if (function->block_count > 1) { FIXME("Branched shaders are not supported yet.\n"); return VKD3D_ERROR_INVALID_SHADER; } if (!(function->blocks[0] = sm6_block_create())) { ERR("Failed to allocate code block.\n"); return VKD3D_ERROR_OUT_OF_MEMORY; } code_block = function->blocks[0]; sm6->cur_max_value = function->value_count; for (i = 1, block_idx = 0, ret_found = false; i < block->record_count; ++i) { sm6->p.location.column = i; if (!code_block) { WARN("Invalid block count %zu.\n", function->block_count); vkd3d_shader_parser_error(&sm6->p, VKD3D_SHADER_ERROR_DXIL_INVALID_OPERAND, "Invalid block count %zu.", function->block_count); return VKD3D_ERROR_INVALID_SHADER; } /* block->record_count - 1 is the instruction count, but some instructions * can emit >1 IR instruction, so extra may be used. */ if (!vkd3d_array_reserve((void **)&code_block->instructions, &code_block->instruction_capacity, max(code_block->instruction_count + 1, block->record_count), sizeof(*code_block->instructions))) { ERR("Failed to allocate instructions.\n"); return VKD3D_ERROR_OUT_OF_MEMORY; } ins = &code_block->instructions[code_block->instruction_count]; ins->handler_idx = VKD3DSIH_INVALID; dst = sm6_parser_get_current_value(sm6); dst->type = NULL; dst->value_type = VALUE_TYPE_REG; is_terminator = false; record = block->records[i]; switch (record->code) { case FUNC_CODE_INST_CALL: sm6_parser_emit_call(sm6, record, code_block, ins, dst); break; case FUNC_CODE_INST_RET: sm6_parser_emit_ret(sm6, record, code_block, ins); is_terminator = true; ret_found = true; break; default: FIXME("Unhandled dxil instruction %u.\n", record->code); return VKD3D_ERROR_INVALID_SHADER; } if (sm6->p.failed) return VKD3D_ERROR; assert(ins->handler_idx != VKD3DSIH_INVALID); if (is_terminator) { ++block_idx; code_block = (block_idx < function->block_count) ? function->blocks[block_idx] : NULL; } if (code_block) code_block->instruction_count += ins->handler_idx != VKD3DSIH_NOP; else assert(ins->handler_idx == VKD3DSIH_NOP); sm6->value_count += !!dst->type; } if (!ret_found) { WARN("Function contains no RET instruction.\n"); return VKD3D_ERROR_INVALID_SHADER; } return VKD3D_OK; } static bool sm6_block_emit_instructions(struct sm6_block *block, struct sm6_parser *sm6) { struct vkd3d_shader_instruction *ins = sm6_parser_require_space(sm6, block->instruction_count + 1); if (!ins) return false; memcpy(ins, block->instructions, block->instruction_count * sizeof(*block->instructions)); sm6->p.instructions.count += block->instruction_count; sm6_parser_add_instruction(sm6, VKD3DSIH_RET); return true; } static enum vkd3d_result sm6_parser_module_init(struct sm6_parser *sm6, const struct dxil_block *block, unsigned int level) { size_t i, old_value_count = sm6->value_count; struct sm6_function *function; enum vkd3d_result ret; for (i = 0; i < block->child_block_count; ++i) { if ((ret = sm6_parser_module_init(sm6, block->child_blocks[i], level + 1)) < 0) return ret; } sm6->p.location.line = block->id; sm6->p.location.column = 0; switch (block->id) { case CONSTANTS_BLOCK: function = &sm6->functions[sm6->function_count]; sm6->cur_max_value = function->value_count; return sm6_parser_constants_init(sm6, block); case FUNCTION_BLOCK: function = &sm6->functions[sm6->function_count]; if ((ret = sm6_parser_function_init(sm6, block, function)) < 0) return ret; /* The value index returns to its previous value after handling a function. It's usually nonzero * at the start because of global constants/variables/function declarations. Function constants * occur in a child block, so value_count is already saved before they are emitted. */ memset(&sm6->values[old_value_count], 0, (sm6->value_count - old_value_count) * sizeof(*sm6->values)); sm6->value_count = old_value_count; break; case BLOCKINFO_BLOCK: case MODULE_BLOCK: case PARAMATTR_BLOCK: case PARAMATTR_GROUP_BLOCK: case VALUE_SYMTAB_BLOCK: case METADATA_BLOCK: case METADATA_ATTACHMENT_BLOCK: case TYPE_BLOCK: break; default: FIXME("Unhandled block id %u.\n", block->id); break; } return VKD3D_OK; } static void sm6_type_table_cleanup(struct sm6_type *types, size_t count) { size_t i; if (!types) return; for (i = 0; i < count; ++i) { switch (types[i].class) { case TYPE_CLASS_STRUCT: vkd3d_free((void *)types[i].u.struc->name); vkd3d_free(types[i].u.struc); break; case TYPE_CLASS_FUNCTION: vkd3d_free(types[i].u.function); break; default: break; } } vkd3d_free(types); } static void sm6_symtab_cleanup(struct sm6_symbol *symbols, size_t count) { size_t i; for (i = 0; i < count; ++i) vkd3d_free((void *)symbols[i].name); vkd3d_free(symbols); } static void sm6_block_destroy(struct sm6_block *block) { vkd3d_free(block->instructions); vkd3d_free(block); } static void sm6_functions_cleanup(struct sm6_function *functions, size_t count) { size_t i, j; for (i = 0; i < count; ++i) { for (j = 0; j < functions[i].block_count; ++j) sm6_block_destroy(functions[i].blocks[j]); } vkd3d_free(functions); } static void sm6_parser_destroy(struct vkd3d_shader_parser *parser) { struct sm6_parser *sm6 = sm6_parser(parser); dxil_block_destroy(&sm6->root_block); dxil_global_abbrevs_cleanup(sm6->abbrevs, sm6->abbrev_count); shader_instruction_array_destroy(&parser->instructions); sm6_type_table_cleanup(sm6->types, sm6->type_count); sm6_symtab_cleanup(sm6->global_symbols, sm6->global_symbol_count); sm6_functions_cleanup(sm6->functions, sm6->function_count); vkd3d_free(sm6->values); free_shader_desc(&parser->shader_desc); vkd3d_free(sm6); } static const struct vkd3d_shader_parser_ops sm6_parser_ops = { .parser_destroy = sm6_parser_destroy, }; static enum vkd3d_result sm6_parser_init(struct sm6_parser *sm6, const uint32_t *byte_code, size_t byte_code_size, const char *source_name, struct vkd3d_shader_message_context *message_context) { const struct shader_signature *output_signature = &sm6->p.shader_desc.output_signature; const struct shader_signature *input_signature = &sm6->p.shader_desc.input_signature; const struct vkd3d_shader_location location = {.source_name = source_name}; uint32_t version_token, dxil_version, token_count, magic; unsigned int chunk_offset, chunk_size; size_t count, length, function_count; enum bitcode_block_abbreviation abbr; struct vkd3d_shader_version version; struct dxil_block *block; enum vkd3d_result ret; unsigned int i; count = byte_code_size / sizeof(*byte_code); if (count < 6) { WARN("Invalid data size %zu.\n", byte_code_size); vkd3d_shader_error(message_context, &location, VKD3D_SHADER_ERROR_DXIL_INVALID_SIZE, "DXIL chunk size %zu is smaller than the DXIL header size.", byte_code_size); return VKD3D_ERROR_INVALID_SHADER; } version_token = byte_code[0]; TRACE("Compiler version: 0x%08x.\n", version_token); token_count = byte_code[1]; TRACE("Token count: %u.\n", token_count); if (token_count < 6 || count < token_count) { WARN("Invalid token count %u (word count %zu).\n", token_count, count); vkd3d_shader_error(message_context, &location, VKD3D_SHADER_ERROR_DXIL_INVALID_CHUNK_SIZE, "DXIL chunk token count %#x is invalid (word count %zu).", token_count, count); return VKD3D_ERROR_INVALID_SHADER; } if (byte_code[2] != TAG_DXIL) WARN("Unknown magic number 0x%08x.\n", byte_code[2]); dxil_version = byte_code[3]; if (dxil_version > 0x102) WARN("Unknown DXIL version: 0x%08x.\n", dxil_version); else TRACE("DXIL version: 0x%08x.\n", dxil_version); chunk_offset = byte_code[4]; if (chunk_offset < 16 || chunk_offset >= byte_code_size) { WARN("Invalid bitcode chunk offset %#x (data size %zu).\n", chunk_offset, byte_code_size); vkd3d_shader_error(message_context, &location, VKD3D_SHADER_ERROR_DXIL_INVALID_CHUNK_OFFSET, "DXIL bitcode chunk has invalid offset %#x (data size %#zx).", chunk_offset, byte_code_size); return VKD3D_ERROR_INVALID_SHADER; } chunk_size = byte_code[5]; if (chunk_size > byte_code_size - chunk_offset) { WARN("Invalid bitcode chunk size %#x (data size %zu, chunk offset %#x).\n", chunk_size, byte_code_size, chunk_offset); vkd3d_shader_error(message_context, &location, VKD3D_SHADER_ERROR_DXIL_INVALID_CHUNK_SIZE, "DXIL bitcode chunk has invalid size %#x (data size %#zx, chunk offset %#x).", chunk_size, byte_code_size, chunk_offset); return VKD3D_ERROR_INVALID_SHADER; } sm6->start = (const uint32_t *)((const char*)&byte_code[2] + chunk_offset); if ((magic = sm6->start[0]) != BITCODE_MAGIC) { WARN("Unknown magic number 0x%08x.\n", magic); vkd3d_shader_parser_warning(&sm6->p, VKD3D_SHADER_WARNING_DXIL_UNKNOWN_MAGIC_NUMBER, "DXIL bitcode chunk magic number 0x%08x is not the expected 0x%08x.", magic, BITCODE_MAGIC); } sm6->end = &sm6->start[(chunk_size + sizeof(*sm6->start) - 1) / sizeof(*sm6->start)]; if ((version.type = version_token >> 16) >= VKD3D_SHADER_TYPE_COUNT) { FIXME("Unknown shader type %#x.\n", version.type); vkd3d_shader_parser_warning(&sm6->p, VKD3D_SHADER_WARNING_DXIL_UNKNOWN_SHADER_TYPE, "Unknown shader type %#x.", version.type); } version.major = VKD3D_SM6_VERSION_MAJOR(version_token); version.minor = VKD3D_SM6_VERSION_MINOR(version_token); if ((abbr = sm6->start[1] & 3) != ENTER_SUBBLOCK) { WARN("Initial block abbreviation %u is not ENTER_SUBBLOCK.\n", abbr); vkd3d_shader_error(message_context, &location, VKD3D_SHADER_ERROR_DXIL_INVALID_BITCODE, "DXIL bitcode chunk has invalid initial block abbreviation %u.", abbr); return VKD3D_ERROR_INVALID_SHADER; } /* Estimate instruction count to avoid reallocation in most shaders. */ count = max(token_count, 400) - 400; vkd3d_shader_parser_init(&sm6->p, message_context, source_name, &version, &sm6_parser_ops, (count + (count >> 2)) / 2u + 10); sm6->ptr = &sm6->start[1]; sm6->bitpos = 2; block = &sm6->root_block; if ((ret = dxil_block_init(block, NULL, sm6)) < 0) { if (ret == VKD3D_ERROR_OUT_OF_MEMORY) vkd3d_shader_error(message_context, &location, VKD3D_SHADER_ERROR_DXIL_OUT_OF_MEMORY, "Out of memory parsing DXIL bitcode chunk."); else if (ret == VKD3D_ERROR_INVALID_SHADER) vkd3d_shader_error(message_context, &location, VKD3D_SHADER_ERROR_DXIL_INVALID_BITCODE, "DXIL bitcode chunk has invalid bitcode."); else vkd3d_unreachable(); return ret; } dxil_global_abbrevs_cleanup(sm6->abbrevs, sm6->abbrev_count); sm6->abbrevs = NULL; sm6->abbrev_count = 0; length = sm6->ptr - sm6->start - block->start; if (length != block->length) { WARN("Invalid block length %zu; expected %u.\n", length, block->length); vkd3d_shader_parser_warning(&sm6->p, VKD3D_SHADER_WARNING_DXIL_INVALID_BLOCK_LENGTH, "Root block ends with length %zu but indicated length is %u.", length, block->length); } if (sm6->ptr != sm6->end) { size_t expected_length = sm6->end - sm6->start; length = sm6->ptr - sm6->start; WARN("Invalid module length %zu; expected %zu.\n", length, expected_length); vkd3d_shader_parser_warning(&sm6->p, VKD3D_SHADER_WARNING_DXIL_INVALID_MODULE_LENGTH, "Module ends with length %zu but indicated length is %zu.", length, expected_length); } if ((ret = sm6_parser_type_table_init(sm6)) < 0) { if (ret == VKD3D_ERROR_OUT_OF_MEMORY) vkd3d_shader_error(message_context, &location, VKD3D_SHADER_ERROR_DXIL_OUT_OF_MEMORY, "Out of memory parsing DXIL type table."); else if (ret == VKD3D_ERROR_INVALID_SHADER) vkd3d_shader_error(message_context, &location, VKD3D_SHADER_ERROR_DXIL_INVALID_TYPE_TABLE, "DXIL type table is invalid."); else vkd3d_unreachable(); return ret; } if ((ret = sm6_parser_symtab_init(sm6)) < 0) { if (ret == VKD3D_ERROR_OUT_OF_MEMORY) vkd3d_shader_error(message_context, &location, VKD3D_SHADER_ERROR_DXIL_OUT_OF_MEMORY, "Out of memory parsing DXIL value symbol table."); else if (ret == VKD3D_ERROR_INVALID_SHADER) vkd3d_shader_error(message_context, &location, VKD3D_SHADER_ERROR_DXIL_INVALID_VALUE_SYMTAB, "DXIL value symbol table is invalid."); else vkd3d_unreachable(); return ret; } if (!(sm6->output_params = shader_parser_get_dst_params(&sm6->p, output_signature->element_count)) || !(sm6->input_params = shader_parser_get_dst_params(&sm6->p, input_signature->element_count))) { ERR("Failed to allocate input/output parameters.\n"); vkd3d_shader_error(message_context, &location, VKD3D_SHADER_ERROR_DXIL_OUT_OF_MEMORY, "Out of memory allocating input/output parameters."); return VKD3D_ERROR_OUT_OF_MEMORY; } function_count = dxil_block_compute_function_count(&sm6->root_block); if (!(sm6->functions = vkd3d_calloc(function_count, sizeof(*sm6->functions)))) { ERR("Failed to allocate function array.\n"); vkd3d_shader_error(message_context, &location, VKD3D_SHADER_ERROR_DXIL_OUT_OF_MEMORY, "Out of memory allocating DXIL function array."); return VKD3D_ERROR_OUT_OF_MEMORY; } if (sm6_parser_compute_max_value_count(sm6, &sm6->root_block, 0) == SIZE_MAX) { WARN("Value array count overflowed.\n"); vkd3d_shader_error(message_context, &location, VKD3D_SHADER_ERROR_DXIL_INVALID_MODULE, "Overflow occurred in the DXIL module value count."); return VKD3D_ERROR_INVALID_SHADER; } if (!(sm6->values = vkd3d_calloc(sm6->value_capacity, sizeof(*sm6->values)))) { ERR("Failed to allocate value array.\n"); vkd3d_shader_error(message_context, &location, VKD3D_SHADER_ERROR_DXIL_OUT_OF_MEMORY, "Out of memory allocating DXIL value array."); return VKD3D_ERROR_OUT_OF_MEMORY; } sm6->ssa_next_id = 1; if ((ret = sm6_parser_globals_init(sm6)) < 0) { WARN("Failed to load global declarations.\n"); return ret; } sm6_parser_init_output_signature(sm6, output_signature); sm6_parser_init_input_signature(sm6, input_signature); if ((ret = sm6_parser_module_init(sm6, &sm6->root_block, 0)) < 0) { if (ret == VKD3D_ERROR_OUT_OF_MEMORY) vkd3d_shader_error(message_context, &location, VKD3D_SHADER_ERROR_DXIL_OUT_OF_MEMORY, "Out of memory parsing DXIL module."); else if (ret == VKD3D_ERROR_INVALID_SHADER) vkd3d_shader_error(message_context, &location, VKD3D_SHADER_ERROR_DXIL_INVALID_MODULE, "DXIL module is invalid."); return ret; } if (!sm6_parser_require_space(sm6, output_signature->element_count + input_signature->element_count)) { vkd3d_shader_error(message_context, &location, VKD3D_SHADER_ERROR_DXIL_OUT_OF_MEMORY, "Out of memory emitting shader signature declarations."); return VKD3D_ERROR_OUT_OF_MEMORY; } sm6_parser_emit_output_signature(sm6, output_signature); sm6_parser_emit_input_signature(sm6, input_signature); sm6->p.shader_desc.ssa_count = sm6->ssa_next_id; for (i = 0; i < sm6->function_count; ++i) { if (!sm6_block_emit_instructions(sm6->functions[i].blocks[0], sm6)) { vkd3d_shader_error(message_context, &location, VKD3D_SHADER_ERROR_DXIL_OUT_OF_MEMORY, "Out of memory emitting shader instructions."); return VKD3D_ERROR_OUT_OF_MEMORY; } } dxil_block_destroy(&sm6->root_block); return VKD3D_OK; } int vkd3d_shader_sm6_parser_create(const struct vkd3d_shader_compile_info *compile_info, struct vkd3d_shader_message_context *message_context, struct vkd3d_shader_parser **parser) { struct vkd3d_shader_desc *shader_desc; uint32_t *byte_code = NULL; struct sm6_parser *sm6; int ret; ERR("Creating a DXIL parser. This is unsupported; you get to keep all the pieces if it breaks.\n"); if (!(sm6 = vkd3d_calloc(1, sizeof(*sm6)))) { ERR("Failed to allocate parser.\n"); return VKD3D_ERROR_OUT_OF_MEMORY; } shader_desc = &sm6->p.shader_desc; shader_desc->is_dxil = true; if ((ret = shader_extract_from_dxbc(&compile_info->source, message_context, compile_info->source_name, shader_desc)) < 0) { WARN("Failed to extract shader, vkd3d result %d.\n", ret); vkd3d_free(sm6); return ret; } sm6->p.shader_desc = *shader_desc; shader_desc = &sm6->p.shader_desc; if (((uintptr_t)shader_desc->byte_code & (VKD3D_DXBC_CHUNK_ALIGNMENT - 1))) { /* LLVM bitcode should be 32-bit aligned, but before dxc v1.7.2207 this was not always the case in the DXBC * container due to missing padding after signature names. Get an aligned copy to prevent unaligned access. */ if (!(byte_code = vkd3d_malloc(align(shader_desc->byte_code_size, VKD3D_DXBC_CHUNK_ALIGNMENT)))) ERR("Failed to allocate aligned chunk. Unaligned access will occur.\n"); else memcpy(byte_code, shader_desc->byte_code, shader_desc->byte_code_size); } ret = sm6_parser_init(sm6, byte_code ? byte_code : shader_desc->byte_code, shader_desc->byte_code_size, compile_info->source_name, message_context); vkd3d_free(byte_code); if (!sm6->p.failed && ret >= 0) vsir_validate(&sm6->p); if (sm6->p.failed && ret >= 0) ret = VKD3D_ERROR_INVALID_SHADER; if (ret < 0) { WARN("Failed to initialise shader parser.\n"); sm6_parser_destroy(&sm6->p); return ret; } *parser = &sm6->p; return ret; }