Files
Odin/core/runtime/core.odin

776 lines
19 KiB
Odin

// This is the runtime code required by the compiler
// IMPORTANT NOTE(bill): Do not change the order of any of this data
// The compiler relies upon this _exact_ order
package runtime
import "core:os"
import "core:mem"
// Naming Conventions:
// In general, Ada_Case for types and snake_case for values
//
// Package Name: snake_case (but prefer single word)
// Import Name: snake_case (but prefer single word)
// Types: Ada_Case
// Enum Values: Ada_Case
// Procedures: snake_case
// Local Variables: snake_case
// Constant Variables: SCREAMING_SNAKE_CASE
// IMPORTANT NOTE(bill): `type_info_of` cannot be used within a
// #shared_global_scope due to the internals of the compiler.
// This could change at a later date if the all these data structures are
// implemented within the compiler rather than in this "preload" file
// NOTE(bill): This must match the compiler's
Calling_Convention :: enum {
Invalid = 0,
Odin = 1,
Contextless = 2,
C = 3,
Std = 4,
Fast = 5,
}
Type_Info_Enum_Value :: union {
rune,
i8, i16, i32, i64, int,
u8, u16, u32, u64, uint, uintptr,
f32, f64,
};
// Variant Types
Type_Info_Named :: struct {name: string, base: ^Type_Info};
Type_Info_Integer :: struct {signed: bool};
Type_Info_Rune :: struct {};
Type_Info_Float :: struct {};
Type_Info_Complex :: struct {};
Type_Info_String :: struct {is_cstring: bool};
Type_Info_Boolean :: struct {};
Type_Info_Any :: struct {};
Type_Info_Type_Id :: struct {};
Type_Info_Pointer :: struct {
elem: ^Type_Info // nil -> rawptr
};
Type_Info_Procedure :: struct {
params: ^Type_Info, // Type_Info_Tuple
results: ^Type_Info, // Type_Info_Tuple
variadic: bool,
convention: Calling_Convention,
};
Type_Info_Array :: struct {
elem: ^Type_Info,
elem_size: int,
count: int,
};
Type_Info_Dynamic_Array :: struct {elem: ^Type_Info, elem_size: int};
Type_Info_Slice :: struct {elem: ^Type_Info, elem_size: int};
Type_Info_Tuple :: struct { // Only really used for procedures
types: []^Type_Info,
names: []string,
};
Type_Info_Struct :: struct {
types: []^Type_Info,
names: []string,
offsets: []uintptr, // offsets may not be used in tuples
usings: []bool, // usings may not be used in tuples
is_packed: bool,
is_raw_union: bool,
custom_align: bool,
};
Type_Info_Union :: struct {
variants: []^Type_Info,
tag_offset: uintptr,
tag_type: ^Type_Info,
};
Type_Info_Enum :: struct {
base: ^Type_Info,
names: []string,
values: []Type_Info_Enum_Value,
is_export: bool,
};
Type_Info_Map :: struct {
key: ^Type_Info,
value: ^Type_Info,
generated_struct: ^Type_Info,
};
Type_Info_Bit_Field :: struct {
names: []string,
bits: []i32,
offsets: []i32,
};
Type_Info :: struct {
size: int,
align: int,
id: typeid,
variant: union {
Type_Info_Named,
Type_Info_Integer,
Type_Info_Rune,
Type_Info_Float,
Type_Info_Complex,
Type_Info_String,
Type_Info_Boolean,
Type_Info_Any,
Type_Info_Type_Id,
Type_Info_Pointer,
Type_Info_Procedure,
Type_Info_Array,
Type_Info_Dynamic_Array,
Type_Info_Slice,
Type_Info_Tuple,
Type_Info_Struct,
Type_Info_Union,
Type_Info_Enum,
Type_Info_Map,
Type_Info_Bit_Field,
},
}
// NOTE(bill): only the ones that are needed (not all types)
// This will be set by the compiler
type_table: []Type_Info;
args__: []cstring;
// IMPORTANT NOTE(bill): Must be in this order (as the compiler relies upon it)
Source_Code_Location :: struct {
file_path: string,
line, column: int,
procedure: string,
}
Context :: struct {
allocator: mem.Allocator,
thread_id: int,
user_data: any,
user_index: int,
parent: ^Context,
derived: any, // May be used for derived data types
}
INITIAL_MAP_CAP :: 16;
Map_Key :: struct {
hash: u64,
str: string,
}
Map_Find_Result :: struct {
hash_index: int,
entry_prev: int,
entry_index: int,
}
Map_Entry_Header :: struct {
key: Map_Key,
next: int,
/*
value: Value_Type,
*/
}
Map_Header :: struct {
m: ^mem.Raw_Map,
is_key_string: bool,
entry_size: int,
entry_align: int,
value_offset: uintptr,
value_size: int,
}
type_info_base :: proc "contextless" (info: ^Type_Info) -> ^Type_Info {
if info == nil do return nil;
base := info;
loop: for {
switch i in base.variant {
case Type_Info_Named: base = i.base;
case: break loop;
}
}
return base;
}
type_info_base_without_enum :: proc "contextless" (info: ^Type_Info) -> ^Type_Info {
if info == nil do return nil;
base := info;
loop: for {
switch i in base.variant {
case Type_Info_Named: base = i.base;
case Type_Info_Enum: base = i.base;
case: break loop;
}
}
return base;
}
__typeid_of :: proc "contextless" (ti: ^Type_Info) -> typeid {
if ti == nil do return nil;
return ti.id;
}
__type_info_of :: proc "contextless" (id: typeid) -> ^Type_Info {
n := int(transmute(uintptr)id);
if n < 0 || n >= len(type_table) {
n = 0;
}
return &type_table[n];
}
typeid_base :: proc "contextless" (id: typeid) -> typeid {
ti := type_info_of(id);
ti = type_info_base(ti);
return typeid_of(ti);
}
typeid_base_without_enum :: proc "contextless" (id: typeid) -> typeid {
ti := type_info_base_without_enum(type_info_of(id));
return typeid_of(ti);
}
@(default_calling_convention = "c")
foreign {
@(link_name="llvm.assume")
assume :: proc(cond: bool) ---;
@(link_name="llvm.debugtrap")
__debug_trap :: proc() ---;
@(link_name="llvm.trap")
__trap :: proc() ---;
@(link_name="llvm.readcyclecounter")
read_cycle_counter :: proc() -> u64 ---;
}
__init_context_from_ptr :: proc "contextless" (c: ^Context, other: ^Context) {
if c == nil do return;
c^ = other^;
__init_context(c);
}
__init_context :: proc "contextless" (c: ^Context) {
if c == nil do return;
c.allocator = os.heap_allocator();
c.thread_id = os.current_thread_id();
}
@(builtin)
copy :: proc "contextless" (dst, src: $T/[]$E) -> int {
n := max(0, min(len(dst), len(src)));
if n > 0 do mem.copy(&dst[0], &src[0], n*size_of(E));
return n;
}
@(builtin)
pop :: proc "contextless" (array: ^$T/[dynamic]$E) -> E {
if array == nil do return E{};
assert(len(array) > 0);
res := array[len(array)-1];
(^mem.Raw_Dynamic_Array)(array).len -= 1;
return res;
}
@(builtin)
clear :: proc[clear_dynamic_array, clear_map];
@(builtin)
reserve :: proc[reserve_dynamic_array, reserve_map];
@(builtin)
new :: inline proc(T: type, loc := #caller_location) -> ^T {
ptr := (^T)(mem.alloc(size_of(T), align_of(T), loc));
ptr^ = T{};
return ptr;
}
@(builtin)
new_clone :: inline proc(data: $T, loc := #caller_location) -> ^T {
ptr := (^T)(mem.alloc(size_of(T), align_of(T), loc));
ptr^ = data;
return ptr;
}
@(builtin)
free :: proc[
mem.free_ptr,
mem.free_string,
mem.free_cstring,
mem.free_dynamic_array,
mem.free_slice,
mem.free_map,
];
@(builtin)
clear_map :: inline proc "contextless" (m: ^$T/map[$K]$V) {
if m == nil do return;
raw_map := (^mem.Raw_Map)(m);
hashes := (^mem.Raw_Dynamic_Array)(&raw_map.hashes);
entries := (^mem.Raw_Dynamic_Array)(&raw_map.entries);
hashes.len = 0;
entries.len = 0;
}
@(builtin)
reserve_map :: proc(m: ^$T/map[$K]$V, capacity: int) {
if m != nil do __dynamic_map_reserve(__get_map_header(m), capacity);
}
@(builtin)
delete :: proc(m: ^$T/map[$K]$V, key: K) {
if m != nil do __dynamic_map_delete(__get_map_header(m), __get_map_key(key));
}
@(builtin)
append :: proc(array: ^$T/[dynamic]$E, args: ...E, loc := #caller_location) -> int {
if array == nil do return 0;
arg_len := len(args);
if arg_len <= 0 do return len(array);
if cap(array) <= len(array)+arg_len {
cap := 2 * cap(array) + max(8, arg_len);
_ = reserve(array, cap, loc);
}
arg_len = min(cap(array)-len(array), arg_len);
if arg_len > 0 {
a := (^mem.Raw_Dynamic_Array)(array);
data := (^E)(a.data);
assert(data != nil);
mem.copy(mem.ptr_offset(data, a.len), &args[0], size_of(E) * arg_len);
a.len += arg_len;
}
return len(array);
}
@(builtin)
append_string :: proc(array: ^$T/[dynamic]$E/u8, args: ...string, loc := #caller_location) -> int {
for arg in args {
append(array = array, args = ([]E)(arg), loc = loc);
}
return len(array);
}
@(builtin)
clear_dynamic_array :: inline proc "contextless" (array: ^$T/[dynamic]$E) {
if array != nil do (^mem.Raw_Dynamic_Array)(array).len = 0;
}
@(builtin)
reserve_dynamic_array :: proc(array: ^$T/[dynamic]$E, capacity: int, loc := #caller_location) -> bool {
if array == nil do return false;
a := (^mem.Raw_Dynamic_Array)(array);
if capacity <= a.cap do return true;
if a.allocator.procedure == nil {
a.allocator = context.allocator;
}
assert(a.allocator.procedure != nil);
old_size := a.cap * size_of(E);
new_size := capacity * size_of(E);
allocator := a.allocator;
new_data := allocator.procedure(
allocator.data, mem.Allocator_Mode.Resize, new_size, align_of(E),
a.data, old_size, 0, loc,
);
if new_data == nil do return false;
a.data = new_data;
a.cap = capacity;
return true;
}
@(builtin)
assert :: proc "contextless" (condition: bool, message := "", using loc := #caller_location) -> bool {
if !condition {
fd := os.stderr;
__print_caller_location(fd, loc);
os.write_string(fd, " Runtime assertion");
if len(message) > 0 {
os.write_string(fd, ": ");
os.write_string(fd, message);
}
os.write_byte(fd, '\n');
__debug_trap();
}
return condition;
}
@(builtin)
panic :: proc "contextless" (message := "", using loc := #caller_location) {
fd := os.stderr;
__print_caller_location(fd, loc);
os.write_string(fd, " Panic");
if len(message) > 0 {
os.write_string(fd, ": ");
os.write_string(fd, message);
}
os.write_byte(fd, '\n');
__debug_trap();
}
// Dynamic Array
__dynamic_array_make :: proc(array_: rawptr, elem_size, elem_align: int, len, cap: int, loc := #caller_location) {
array := (^mem.Raw_Dynamic_Array)(array_);
array.allocator = context.allocator;
assert(array.allocator.procedure != nil);
if cap > 0 {
__dynamic_array_reserve(array_, elem_size, elem_align, cap, loc);
array.len = len;
}
}
__dynamic_array_reserve :: proc(array_: rawptr, elem_size, elem_align: int, cap: int, loc := #caller_location) -> bool {
array := (^mem.Raw_Dynamic_Array)(array_);
if cap <= array.cap do return true;
if array.allocator.procedure == nil {
array.allocator = context.allocator;
}
assert(array.allocator.procedure != nil);
old_size := array.cap * elem_size;
new_size := cap * elem_size;
allocator := array.allocator;
new_data := allocator.procedure(allocator.data, mem.Allocator_Mode.Resize, new_size, elem_align, array.data, old_size, 0, loc);
if new_data == nil do return false;
array.data = new_data;
array.cap = cap;
return true;
}
__dynamic_array_resize :: proc(array_: rawptr, elem_size, elem_align: int, len: int, loc := #caller_location) -> bool {
array := (^mem.Raw_Dynamic_Array)(array_);
ok := __dynamic_array_reserve(array_, elem_size, elem_align, len, loc);
if ok do array.len = len;
return ok;
}
__dynamic_array_append :: proc(array_: rawptr, elem_size, elem_align: int,
items: rawptr, item_count: int, loc := #caller_location) -> int {
array := (^mem.Raw_Dynamic_Array)(array_);
if items == nil do return 0;
if item_count <= 0 do return 0;
ok := true;
if array.cap <= array.len+item_count {
cap := 2 * array.cap + max(8, item_count);
ok = __dynamic_array_reserve(array, elem_size, elem_align, cap, loc);
}
// TODO(bill): Better error handling for failed reservation
if !ok do return array.len;
assert(array.data != nil);
data := uintptr(array.data) + uintptr(elem_size*array.len);
mem.copy(rawptr(data), items, elem_size * item_count);
array.len += item_count;
return array.len;
}
__dynamic_array_append_nothing :: proc(array_: rawptr, elem_size, elem_align: int, loc := #caller_location) -> int {
array := (^mem.Raw_Dynamic_Array)(array_);
ok := true;
if array.cap <= array.len+1 {
cap := 2 * array.cap + max(8, 1);
ok = __dynamic_array_reserve(array, elem_size, elem_align, cap, loc);
}
// TODO(bill): Better error handling for failed reservation
if !ok do return array.len;
assert(array.data != nil);
data := uintptr(array.data) + uintptr(elem_size*array.len);
mem.zero(rawptr(data), elem_size);
array.len += 1;
return array.len;
}
// Map
__get_map_header :: proc "contextless" (m: ^$T/map[$K]$V) -> Map_Header {
header := Map_Header{m = (^mem.Raw_Map)(m)};
Entry :: struct {
key: Map_Key,
next: int,
value: V,
}
_, is_string := type_info_base(type_info_of(K)).variant.(Type_Info_String);
header.is_key_string = is_string;
header.entry_size = int(size_of(Entry));
header.entry_align = int(align_of(Entry));
header.value_offset = uintptr(offset_of(Entry, value));
header.value_size = int(size_of(V));
return header;
}
__get_map_key :: proc "contextless" (key: $K) -> Map_Key {
map_key: Map_Key;
ti := type_info_base_without_enum(type_info_of(K));
switch _ in ti.variant {
case Type_Info_Integer:
switch 8*size_of(key) {
case 8: map_key.hash = u64(( ^u8)(&key)^);
case 16: map_key.hash = u64(( ^u16)(&key)^);
case 32: map_key.hash = u64(( ^u32)(&key)^);
case 64: map_key.hash = u64(( ^u64)(&key)^);
case: panic("Unhandled integer size");
}
case Type_Info_Rune:
map_key.hash = u64((^rune)(&key)^);
case Type_Info_Pointer:
map_key.hash = u64(uintptr((^rawptr)(&key)^));
case Type_Info_Float:
switch 8*size_of(key) {
case 32: map_key.hash = u64((^u32)(&key)^);
case 64: map_key.hash = u64((^u64)(&key)^);
case: panic("Unhandled float size");
}
case Type_Info_String:
str := (^string)(&key)^;
map_key.hash = __default_hash_string(str);
map_key.str = str;
case:
panic("Unhandled map key type");
}
return map_key;
}
__default_hash :: proc(data: []byte) -> u64 {
fnv64a :: proc(data: []byte) -> u64 {
h: u64 = 0xcbf29ce484222325;
for b in data {
h = (h ~ u64(b)) * 0x100000001b3;
}
return h;
}
return fnv64a(data);
}
__default_hash_string :: proc(s: string) -> u64 do return __default_hash(([]byte)(s));
__dynamic_map_reserve :: proc(using header: Map_Header, cap: int, loc := #caller_location) {
__dynamic_array_reserve(&m.hashes, size_of(int), align_of(int), cap, loc);
__dynamic_array_reserve(&m.entries, entry_size, entry_align, cap, loc);
}
__dynamic_map_rehash :: proc(using header: Map_Header, new_count: int, loc := #caller_location) #no_bounds_check {
new_header: Map_Header = header;
nm: mem.Raw_Map;
new_header.m = &nm;
header_hashes := (^mem.Raw_Dynamic_Array)(&header.m.hashes);
nm_hashes := (^mem.Raw_Dynamic_Array)(&nm.hashes);
__dynamic_array_resize(nm_hashes, size_of(int), align_of(int), new_count, loc);
__dynamic_array_reserve(&nm.entries, entry_size, entry_align, m.entries.len, loc);
for i in 0..new_count do nm.hashes[i] = -1;
for i in 0..m.entries.len {
if len(nm.hashes) == 0 do __dynamic_map_grow(new_header, loc);
entry_header := __dynamic_map_get_entry(header, i);
data := uintptr(entry_header);
fr := __dynamic_map_find(new_header, entry_header.key);
j := __dynamic_map_add_entry(new_header, entry_header.key, loc);
if fr.entry_prev < 0 {
nm.hashes[fr.hash_index] = j;
} else {
e := __dynamic_map_get_entry(new_header, fr.entry_prev);
e.next = j;
}
e := __dynamic_map_get_entry(new_header, j);
e.next = fr.entry_index;
ndata := uintptr(e);
mem.copy(rawptr(ndata+value_offset), rawptr(data+value_offset), value_size);
if __dynamic_map_full(new_header) do __dynamic_map_grow(new_header, loc);
}
mem.free_ptr_with_allocator(header_hashes.allocator, header_hashes.data, loc);
mem.free_ptr_with_allocator(header.m.entries.allocator, header.m.entries.data, loc);
header.m^ = nm;
}
__dynamic_map_get :: proc(h: Map_Header, key: Map_Key) -> rawptr {
index := __dynamic_map_find(h, key).entry_index;
if index >= 0 {
data := uintptr(__dynamic_map_get_entry(h, index));
return rawptr(data + h.value_offset);
}
return nil;
}
__dynamic_map_set :: proc(h: Map_Header, key: Map_Key, value: rawptr, loc := #caller_location) #no_bounds_check {
index: int;
assert(value != nil);
if len(h.m.hashes) == 0 {
__dynamic_map_reserve(h, INITIAL_MAP_CAP, loc);
__dynamic_map_grow(h, loc);
}
fr := __dynamic_map_find(h, key);
if fr.entry_index >= 0 {
index = fr.entry_index;
} else {
index = __dynamic_map_add_entry(h, key, loc);
if fr.entry_prev >= 0 {
entry := __dynamic_map_get_entry(h, fr.entry_prev);
entry.next = index;
} else {
h.m.hashes[fr.hash_index] = index;
}
}
{
e := __dynamic_map_get_entry(h, index);
e.key = key;
val := (^byte)(uintptr(e) + h.value_offset);
mem.copy(val, value, h.value_size);
}
if __dynamic_map_full(h) {
__dynamic_map_grow(h, loc);
}
}
__dynamic_map_grow :: proc(using h: Map_Header, loc := #caller_location) {
// TODO(bill): Determine an efficient growing rate
new_count := max(4*m.entries.cap + 7, INITIAL_MAP_CAP);
__dynamic_map_rehash(h, new_count, loc);
}
__dynamic_map_full :: inline proc(using h: Map_Header) -> bool {
return int(0.75 * f64(len(m.hashes))) <= m.entries.cap;
}
__dynamic_map_hash_equal :: proc(h: Map_Header, a, b: Map_Key) -> bool {
if a.hash == b.hash {
if h.is_key_string do return a.str == b.str;
return true;
}
return false;
}
__dynamic_map_find :: proc(using h: Map_Header, key: Map_Key) -> Map_Find_Result #no_bounds_check {
fr := Map_Find_Result{-1, -1, -1};
if len(m.hashes) > 0 {
fr.hash_index = int(key.hash % u64(len(m.hashes)));
fr.entry_index = m.hashes[fr.hash_index];
for fr.entry_index >= 0 {
entry := __dynamic_map_get_entry(h, fr.entry_index);
if __dynamic_map_hash_equal(h, entry.key, key) do return fr;
fr.entry_prev = fr.entry_index;
fr.entry_index = entry.next;
}
}
return fr;
}
__dynamic_map_add_entry :: proc(using h: Map_Header, key: Map_Key, loc := #caller_location) -> int {
prev := m.entries.len;
c := __dynamic_array_append_nothing(&m.entries, entry_size, entry_align, loc);
if c != prev {
end := __dynamic_map_get_entry(h, c-1);
end.key = key;
end.next = -1;
}
return prev;
}
__dynamic_map_delete :: proc(using h: Map_Header, key: Map_Key) {
fr := __dynamic_map_find(h, key);
if fr.entry_index >= 0 {
__dynamic_map_erase(h, fr);
}
}
__dynamic_map_get_entry :: proc(using h: Map_Header, index: int) -> ^Map_Entry_Header {
assert(0 <= index && index < m.entries.len);
return (^Map_Entry_Header)(uintptr(m.entries.data) + uintptr(index*entry_size));
}
__dynamic_map_erase :: proc(using h: Map_Header, fr: Map_Find_Result) #no_bounds_check {
if fr.entry_prev < 0 {
m.hashes[fr.hash_index] = __dynamic_map_get_entry(h, fr.entry_index).next;
} else {
__dynamic_map_get_entry(h, fr.entry_prev).next = __dynamic_map_get_entry(h, fr.entry_index).next;
}
mem.copy(__dynamic_map_get_entry(h, fr.entry_index), __dynamic_map_get_entry(h, m.entries.len-1), entry_size);
last := __dynamic_map_find(h, __dynamic_map_get_entry(h, fr.entry_index).key);
if last.entry_prev >= 0 {
__dynamic_map_get_entry(h, last.entry_prev).next = fr.entry_index;
} else {
m.hashes[last.hash_index] = fr.entry_index;
}
if fr.entry_index == m.entries.len-1 {
m.entries.len -= 1;
}
}