/* * HLSL constant value operations for constant folding * * Copyright 2022 Francisco Casas 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 #include "hlsl.h" static bool fold_abs(struct hlsl_ctx *ctx, struct hlsl_constant_value *dst, const struct hlsl_type *dst_type, const struct hlsl_ir_constant *src) { enum hlsl_base_type type = dst_type->base_type; unsigned int k; assert(type == src->node.data_type->base_type); for (k = 0; k < dst_type->dimx; ++k) { switch (type) { case HLSL_TYPE_FLOAT: case HLSL_TYPE_HALF: dst->u[k].f = fabsf(src->value.u[k].f); break; case HLSL_TYPE_DOUBLE: dst->u[k].d = fabs(src->value.u[k].d); break; case HLSL_TYPE_INT: /* C's abs(INT_MIN) is undefined, but HLSL evaluates this to INT_MIN */ if (src->value.u[k].i == INT_MIN) dst->u[k].i = INT_MIN; else dst->u[k].i = abs(src->value.u[k].i); break; case HLSL_TYPE_UINT: dst->u[k].u = src->value.u[k].u; break; default: FIXME("Fold abs() for type %s.\n", debug_hlsl_type(ctx, dst_type)); return false; } } return true; } static bool fold_cast(struct hlsl_ctx *ctx, struct hlsl_constant_value *dst, const struct hlsl_type *dst_type, const struct hlsl_ir_constant *src) { unsigned int k; uint32_t u; int32_t i; double d; float f; if (dst_type->dimx != src->node.data_type->dimx || dst_type->dimy != src->node.data_type->dimy) { FIXME("Cast from %s to %s.\n", debug_hlsl_type(ctx, src->node.data_type), debug_hlsl_type(ctx, dst_type)); return false; } for (k = 0; k < 4; ++k) { switch (src->node.data_type->base_type) { case HLSL_TYPE_FLOAT: case HLSL_TYPE_HALF: u = src->value.u[k].f; i = src->value.u[k].f; f = src->value.u[k].f; d = src->value.u[k].f; break; case HLSL_TYPE_DOUBLE: u = src->value.u[k].d; i = src->value.u[k].d; f = src->value.u[k].d; d = src->value.u[k].d; break; case HLSL_TYPE_INT: u = src->value.u[k].i; i = src->value.u[k].i; f = src->value.u[k].i; d = src->value.u[k].i; break; case HLSL_TYPE_UINT: u = src->value.u[k].u; i = src->value.u[k].u; f = src->value.u[k].u; d = src->value.u[k].u; break; case HLSL_TYPE_BOOL: u = !!src->value.u[k].u; i = !!src->value.u[k].u; f = !!src->value.u[k].u; d = !!src->value.u[k].u; break; default: vkd3d_unreachable(); } switch (dst_type->base_type) { case HLSL_TYPE_FLOAT: case HLSL_TYPE_HALF: dst->u[k].f = f; break; case HLSL_TYPE_DOUBLE: dst->u[k].d = d; break; case HLSL_TYPE_INT: dst->u[k].i = i; break; case HLSL_TYPE_UINT: dst->u[k].u = u; break; case HLSL_TYPE_BOOL: /* Casts to bool should have already been lowered. */ default: vkd3d_unreachable(); } } return true; } static bool fold_neg(struct hlsl_ctx *ctx, struct hlsl_constant_value *dst, const struct hlsl_type *dst_type, const struct hlsl_ir_constant *src) { enum hlsl_base_type type = dst_type->base_type; unsigned int k; assert(type == src->node.data_type->base_type); for (k = 0; k < 4; ++k) { switch (type) { case HLSL_TYPE_FLOAT: case HLSL_TYPE_HALF: dst->u[k].f = -src->value.u[k].f; break; case HLSL_TYPE_DOUBLE: dst->u[k].d = -src->value.u[k].d; break; case HLSL_TYPE_INT: case HLSL_TYPE_UINT: dst->u[k].u = -src->value.u[k].u; break; default: FIXME("Fold negation for type %s.\n", debug_hlsl_type(ctx, dst_type)); return false; } } return true; } static bool fold_add(struct hlsl_ctx *ctx, struct hlsl_constant_value *dst, const struct hlsl_type *dst_type, const struct hlsl_ir_constant *src1, const struct hlsl_ir_constant *src2) { enum hlsl_base_type type = dst_type->base_type; unsigned int k; assert(type == src1->node.data_type->base_type); assert(type == src2->node.data_type->base_type); for (k = 0; k < 4; ++k) { switch (type) { case HLSL_TYPE_FLOAT: case HLSL_TYPE_HALF: dst->u[k].f = src1->value.u[k].f + src2->value.u[k].f; break; case HLSL_TYPE_DOUBLE: dst->u[k].d = src1->value.u[k].d + src2->value.u[k].d; break; /* Handling HLSL_TYPE_INT through the unsigned field to avoid * undefined behavior with signed integers in C. */ case HLSL_TYPE_INT: case HLSL_TYPE_UINT: dst->u[k].u = src1->value.u[k].u + src2->value.u[k].u; break; default: FIXME("Fold addition for type %s.\n", debug_hlsl_type(ctx, dst_type)); return false; } } return true; } static bool fold_bit_and(struct hlsl_ctx *ctx, struct hlsl_constant_value *dst, const struct hlsl_type *dst_type, const struct hlsl_ir_constant *src1, const struct hlsl_ir_constant *src2) { enum hlsl_base_type type = dst_type->base_type; unsigned int k; assert(type == src1->node.data_type->base_type); assert(type == src2->node.data_type->base_type); for (k = 0; k < dst_type->dimx; ++k) { switch (type) { case HLSL_TYPE_INT: case HLSL_TYPE_UINT: dst->u[k].u = src1->value.u[k].u & src2->value.u[k].u; break; default: FIXME("Fold bit and for type %s.\n", debug_hlsl_type(ctx, dst_type)); return false; } } return true; } static bool fold_bit_or(struct hlsl_ctx *ctx, struct hlsl_constant_value *dst, const struct hlsl_type *dst_type, const struct hlsl_ir_constant *src1, const struct hlsl_ir_constant *src2) { enum hlsl_base_type type = dst_type->base_type; unsigned int k; assert(type == src1->node.data_type->base_type); assert(type == src2->node.data_type->base_type); for (k = 0; k < dst_type->dimx; ++k) { switch (type) { case HLSL_TYPE_INT: case HLSL_TYPE_UINT: dst->u[k].u = src1->value.u[k].u | src2->value.u[k].u; break; default: FIXME("Fold bit or for type %s.\n", debug_hlsl_type(ctx, dst_type)); return false; } } return true; } static bool fold_bit_xor(struct hlsl_ctx *ctx, struct hlsl_constant_value *dst, const struct hlsl_type *dst_type, const struct hlsl_ir_constant *src1, const struct hlsl_ir_constant *src2) { enum hlsl_base_type type = dst_type->base_type; unsigned int k; assert(type == src1->node.data_type->base_type); assert(type == src2->node.data_type->base_type); for (k = 0; k < dst_type->dimx; ++k) { switch (type) { case HLSL_TYPE_INT: case HLSL_TYPE_UINT: dst->u[k].u = src1->value.u[k].u ^ src2->value.u[k].u; break; default: FIXME("Fold bit xor for type %s.\n", debug_hlsl_type(ctx, dst_type)); return false; } } return true; } static bool fold_div(struct hlsl_ctx *ctx, struct hlsl_constant_value *dst, const struct hlsl_type *dst_type, const struct hlsl_ir_constant *src1, const struct hlsl_ir_constant *src2, const struct vkd3d_shader_location *loc) { enum hlsl_base_type type = dst_type->base_type; unsigned int k; assert(type == src1->node.data_type->base_type); assert(type == src2->node.data_type->base_type); for (k = 0; k < dst_type->dimx; ++k) { switch (type) { case HLSL_TYPE_FLOAT: case HLSL_TYPE_HALF: if (ctx->profile->major_version >= 4 && src2->value.u[k].f == 0) { hlsl_warning(ctx, loc, VKD3D_SHADER_WARNING_HLSL_DIVISION_BY_ZERO, "Floating point division by zero."); } dst->u[k].f = src1->value.u[k].f / src2->value.u[k].f; if (ctx->profile->major_version < 4 && !isfinite(dst->u[k].f)) { hlsl_error(ctx, loc, VKD3D_SHADER_ERROR_HLSL_DIVISION_BY_ZERO, "Infinities and NaNs are not allowed by the shader model."); } break; case HLSL_TYPE_DOUBLE: if (src2->value.u[k].d == 0) { hlsl_warning(ctx, loc, VKD3D_SHADER_WARNING_HLSL_DIVISION_BY_ZERO, "Floating point division by zero."); } dst->u[k].d = src1->value.u[k].d / src2->value.u[k].d; break; case HLSL_TYPE_INT: if (src2->value.u[k].i == 0) { hlsl_error(ctx, loc, VKD3D_SHADER_ERROR_HLSL_DIVISION_BY_ZERO, "Division by zero."); return false; } if (src1->value.u[k].i == INT_MIN && src2->value.u[k].i == -1) dst->u[k].i = INT_MIN; else dst->u[k].i = src1->value.u[k].i / src2->value.u[k].i; break; case HLSL_TYPE_UINT: if (src2->value.u[k].u == 0) { hlsl_error(ctx, loc, VKD3D_SHADER_ERROR_HLSL_DIVISION_BY_ZERO, "Division by zero."); return false; } dst->u[k].u = src1->value.u[k].u / src2->value.u[k].u; break; default: FIXME("Fold division for type %s.\n", debug_hlsl_type(ctx, dst_type)); return false; } } return true; } static bool fold_max(struct hlsl_ctx *ctx, struct hlsl_constant_value *dst, const struct hlsl_type *dst_type, const struct hlsl_ir_constant *src1, const struct hlsl_ir_constant *src2) { enum hlsl_base_type type = dst_type->base_type; unsigned int k; assert(type == src1->node.data_type->base_type); assert(type == src2->node.data_type->base_type); for (k = 0; k < dst_type->dimx; ++k) { switch (type) { case HLSL_TYPE_INT: dst->u[k].i = max(src1->value.u[k].i, src2->value.u[k].i); break; case HLSL_TYPE_UINT: dst->u[k].u = max(src1->value.u[k].u, src2->value.u[k].u); break; default: FIXME("Fold max for type %s.\n", debug_hlsl_type(ctx, dst_type)); return false; } } return true; } static bool fold_min(struct hlsl_ctx *ctx, struct hlsl_constant_value *dst, const struct hlsl_type *dst_type, const struct hlsl_ir_constant *src1, const struct hlsl_ir_constant *src2) { enum hlsl_base_type type = dst_type->base_type; unsigned int k; assert(type == src1->node.data_type->base_type); assert(type == src2->node.data_type->base_type); for (k = 0; k < dst_type->dimx; ++k) { switch (type) { case HLSL_TYPE_INT: dst->u[k].i = min(src1->value.u[k].i, src2->value.u[k].i); break; case HLSL_TYPE_UINT: dst->u[k].u = min(src1->value.u[k].u, src2->value.u[k].u); break; default: FIXME("Fold min for type %s.\n", debug_hlsl_type(ctx, dst_type)); return false; } } return true; } static bool fold_mod(struct hlsl_ctx *ctx, struct hlsl_constant_value *dst, const struct hlsl_type *dst_type, const struct hlsl_ir_constant *src1, const struct hlsl_ir_constant *src2, const struct vkd3d_shader_location *loc) { enum hlsl_base_type type = dst_type->base_type; unsigned int k; assert(type == src1->node.data_type->base_type); assert(type == src2->node.data_type->base_type); for (k = 0; k < dst_type->dimx; ++k) { switch (type) { case HLSL_TYPE_INT: if (src2->value.u[k].i == 0) { hlsl_error(ctx, loc, VKD3D_SHADER_ERROR_HLSL_DIVISION_BY_ZERO, "Division by zero."); return false; } if (src1->value.u[k].i == INT_MIN && src2->value.u[k].i == -1) dst->u[k].i = 0; else dst->u[k].i = src1->value.u[k].i % src2->value.u[k].i; break; case HLSL_TYPE_UINT: if (src2->value.u[k].u == 0) { hlsl_error(ctx, loc, VKD3D_SHADER_ERROR_HLSL_DIVISION_BY_ZERO, "Division by zero."); return false; } dst->u[k].u = src1->value.u[k].u % src2->value.u[k].u; break; default: FIXME("Fold modulus for type %s.\n", debug_hlsl_type(ctx, dst_type)); return false; } } return true; } static bool fold_mul(struct hlsl_ctx *ctx, struct hlsl_constant_value *dst, const struct hlsl_type *dst_type, const struct hlsl_ir_constant *src1, const struct hlsl_ir_constant *src2) { enum hlsl_base_type type = dst_type->base_type; unsigned int k; assert(type == src1->node.data_type->base_type); assert(type == src2->node.data_type->base_type); for (k = 0; k < 4; ++k) { switch (type) { case HLSL_TYPE_FLOAT: case HLSL_TYPE_HALF: dst->u[k].f = src1->value.u[k].f * src2->value.u[k].f; break; case HLSL_TYPE_DOUBLE: dst->u[k].d = src1->value.u[k].d * src2->value.u[k].d; break; case HLSL_TYPE_INT: case HLSL_TYPE_UINT: dst->u[k].u = src1->value.u[k].u * src2->value.u[k].u; break; default: FIXME("Fold multiplication for type %s.\n", debug_hlsl_type(ctx, dst_type)); return false; } } return true; } static bool fold_nequal(struct hlsl_ctx *ctx, struct hlsl_constant_value *dst, const struct hlsl_type *dst_type, const struct hlsl_ir_constant *src1, const struct hlsl_ir_constant *src2) { unsigned int k; assert(dst_type->base_type == HLSL_TYPE_BOOL); assert(src1->node.data_type->base_type == src2->node.data_type->base_type); for (k = 0; k < 4; ++k) { switch (src1->node.data_type->base_type) { case HLSL_TYPE_FLOAT: case HLSL_TYPE_HALF: dst->u[k].u = src1->value.u[k].f != src2->value.u[k].f; break; case HLSL_TYPE_DOUBLE: dst->u[k].u = src1->value.u[k].d != src2->value.u[k].d; break; case HLSL_TYPE_INT: case HLSL_TYPE_UINT: case HLSL_TYPE_BOOL: dst->u[k].u = src1->value.u[k].u != src2->value.u[k].u; break; default: vkd3d_unreachable(); } dst->u[k].u *= ~0u; } return true; } bool hlsl_fold_constant_exprs(struct hlsl_ctx *ctx, struct hlsl_ir_node *instr, void *context) { struct hlsl_ir_constant *arg1, *arg2 = NULL; struct hlsl_constant_value res = {0}; struct hlsl_ir_node *res_node; struct hlsl_ir_expr *expr; unsigned int i; bool success; if (instr->type != HLSL_IR_EXPR) return false; expr = hlsl_ir_expr(instr); if (!expr->operands[0].node) return false; if (instr->data_type->class > HLSL_CLASS_VECTOR) return false; for (i = 0; i < ARRAY_SIZE(expr->operands); ++i) { if (expr->operands[i].node) { if (expr->operands[i].node->type != HLSL_IR_CONSTANT) return false; assert(expr->operands[i].node->data_type->class <= HLSL_CLASS_VECTOR); } } arg1 = hlsl_ir_constant(expr->operands[0].node); if (expr->operands[1].node) arg2 = hlsl_ir_constant(expr->operands[1].node); switch (expr->op) { case HLSL_OP1_ABS: success = fold_abs(ctx, &res, instr->data_type, arg1); break; case HLSL_OP1_CAST: success = fold_cast(ctx, &res, instr->data_type, arg1); break; case HLSL_OP1_NEG: success = fold_neg(ctx, &res, instr->data_type, arg1); break; case HLSL_OP2_ADD: success = fold_add(ctx, &res, instr->data_type, arg1, arg2); break; case HLSL_OP2_BIT_AND: success = fold_bit_and(ctx, &res, instr->data_type, arg1, arg2); break; case HLSL_OP2_BIT_OR: success = fold_bit_or(ctx, &res, instr->data_type, arg1, arg2); break; case HLSL_OP2_BIT_XOR: success = fold_bit_xor(ctx, &res, instr->data_type, arg1, arg2); break; case HLSL_OP2_DIV: success = fold_div(ctx, &res, instr->data_type, arg1, arg2, &instr->loc); break; case HLSL_OP2_MAX: success = fold_max(ctx, &res, instr->data_type, arg1, arg2); break; case HLSL_OP2_MIN: success = fold_min(ctx, &res, instr->data_type, arg1, arg2); break; case HLSL_OP2_MOD: success = fold_mod(ctx, &res, instr->data_type, arg1, arg2, &instr->loc); break; case HLSL_OP2_MUL: success = fold_mul(ctx, &res, instr->data_type, arg1, arg2); break; case HLSL_OP2_NEQUAL: success = fold_nequal(ctx, &res, instr->data_type, arg1, arg2); break; default: FIXME("Fold \"%s\" expression.\n", debug_hlsl_expr_op(expr->op)); success = false; break; } if (success) { if (!(res_node = hlsl_new_constant(ctx, instr->data_type, &res, &instr->loc))) return false; list_add_before(&expr->node.entry, &res_node->entry); hlsl_replace_node(&expr->node, res_node); } return success; } bool hlsl_fold_constant_swizzles(struct hlsl_ctx *ctx, struct hlsl_ir_node *instr, void *context) { struct hlsl_constant_value value; struct hlsl_ir_swizzle *swizzle; struct hlsl_ir_constant *src; struct hlsl_ir_node *dst; unsigned int i; if (instr->type != HLSL_IR_SWIZZLE) return false; swizzle = hlsl_ir_swizzle(instr); if (swizzle->val.node->type != HLSL_IR_CONSTANT) return false; src = hlsl_ir_constant(swizzle->val.node); for (i = 0; i < swizzle->node.data_type->dimx; ++i) value.u[i] = src->value.u[hlsl_swizzle_get_component(swizzle->swizzle, i)]; if (!(dst = hlsl_new_constant(ctx, instr->data_type, &value, &instr->loc))) return false; list_add_before(&swizzle->node.entry, &dst->entry); hlsl_replace_node(&swizzle->node, dst); return true; }