Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with
or
.
Download ZIP
Fetching contributors…

Cannot retrieve contributors at this time

1628 lines (1539 sloc) 55.669 kB
/*
Generic Functions
. method table and lookup
. GF constructor, add_method
. dispatch
. static parameter inference
. method specialization, invoking type inference
*/
#include <stdlib.h>
#include <string.h>
#include <assert.h>
#ifdef __WIN32__
#include <malloc.h>
#endif
#include "julia.h"
#include "builtin_proto.h"
#define ENABLE_INFERENCE
// debugging options
//#define TRACE_INFERENCE
//#define JL_TRACE
//#define JL_GF_PROFILE
static jl_methtable_t *new_method_table(jl_sym_t *name)
{
jl_methtable_t *mt = (jl_methtable_t*)allocobj(sizeof(jl_methtable_t));
mt->type = (jl_type_t*)jl_methtable_type;
mt->name = name;
mt->defs = JL_NULL;
mt->cache = JL_NULL;
mt->cache_arg1 = JL_NULL;
mt->cache_targ = JL_NULL;
mt->max_args = 0;
#ifdef JL_GF_PROFILE
mt->ncalls = 0;
#endif
return mt;
}
static int cache_match_by_type(jl_value_t **types, size_t n, jl_tuple_t *sig,
int va)
{
if (!va && n > jl_tuple_len(sig))
return 0;
if (jl_tuple_len(sig) > n) {
if (!(n == jl_tuple_len(sig)-1 && va))
return 0;
}
size_t i;
for(i=0; i < n; i++) {
jl_value_t *decl = jl_tupleref(sig, i);
if (i == jl_tuple_len(sig)-1) {
if (va) {
jl_value_t *t = jl_tparam0(decl);
for(; i < n; i++) {
if (!jl_subtype(types[i], t, 0))
return 0;
}
return 1;
}
}
jl_value_t *a = types[i];
if (jl_is_tuple(decl)) {
// tuples don't have to match exactly, to avoid caching
// signatures for tuples of every length
if (!jl_subtype(a, decl, 0))
return 0;
}
else if (jl_is_tag_type(a) && jl_is_tag_type(decl) &&
((jl_tag_type_t*)decl)->name == jl_type_type->name &&
((jl_tag_type_t*)a )->name == jl_type_type->name) {
jl_value_t *tp0 = jl_tparam0(decl);
if (tp0 == (jl_value_t*)jl_typetype_tvar) {
// in the case of Type{T}, the types don't have
// to match exactly either. this is cached as Type{T}.
// analogous to the situation with tuples.
}
else {
if (!jl_types_equal(jl_tparam0(a), tp0))
return 0;
}
}
else if (decl == (jl_value_t*)jl_any_type) {
}
else {
if (!jl_types_equal(a, decl))
return 0;
}
}
return 1;
}
static inline int cache_match(jl_value_t **args, size_t n, jl_tuple_t *sig,
int va)
{
// NOTE: This function is a huge performance hot spot!!
if (jl_tuple_len(sig) > n) {
if (n != jl_tuple_len(sig)-1)
return 0;
}
size_t i;
for(i=0; i < n; i++) {
jl_value_t *decl = jl_tupleref(sig, i);
if (i == jl_tuple_len(sig)-1) {
if (va) {
jl_value_t *t = jl_tparam0(decl);
for(; i < n; i++) {
if (!jl_subtype(args[i], t, 1))
return 0;
}
return 1;
}
}
jl_value_t *a = args[i];
if (jl_is_tuple(decl)) {
// tuples don't have to match exactly, to avoid caching
// signatures for tuples of every length
if (!jl_is_tuple(a) || !jl_subtype(a, decl, 1))
return 0;
}
else if (jl_is_type_type(decl) &&
jl_is_nontuple_type(a)) { //***
jl_value_t *tp0 = jl_tparam0(decl);
if (tp0 == (jl_value_t*)jl_typetype_tvar) {
// in the case of Type{T}, the types don't have
// to match exactly either. this is cached as Type{T}.
// analogous to the situation with tuples.
}
else {
if (a!=tp0 && !jl_types_equal(a,tp0))
return 0;
}
}
else if (decl == (jl_value_t*)jl_any_type) {
}
else {
/*
we know there are only concrete types here, and types are
hash-consed, so pointer comparison should work.
*/
if ((jl_value_t*)jl_typeof(a) != decl)
return 0;
}
}
return 1;
}
static inline
jl_methlist_t *mtcache_hash_lookup(jl_array_t *a, jl_value_t *ty, int tparam)
{
uptrint_t uid;
if ((jl_is_struct_type(ty) && (uid = ((jl_struct_type_t*)ty)->uid)) ||
(jl_is_bits_type(ty) && (uid = ((jl_bits_type_t*)ty)->uid))) {
jl_methlist_t *ml = (jl_methlist_t*)jl_cellref(a, uid & (a->length-1));
if (ml && ml!=JL_NULL) {
jl_value_t *t = jl_tupleref(ml->sig, 0);
if (tparam) t = jl_tparam0(t);
if (t == ty)
return ml;
}
}
return JL_NULL;
}
static void mtcache_rehash(jl_array_t **pa)
{
size_t len = (*pa)->length;
jl_value_t **d = (jl_value_t**)(*pa)->data;
jl_array_t *n = jl_alloc_cell_1d(len*2);
jl_value_t **nd = (jl_value_t**)n->data;
size_t i;
for(i=0; i < len; i++) {
jl_methlist_t *ml = (jl_methlist_t*)d[i];
if (ml && ml!=JL_NULL) {
jl_value_t *t = jl_tupleref(ml->sig,0);
if (jl_is_type_type(t))
t = jl_tparam0(t);
uptrint_t uid;
if (jl_is_struct_type(t))
uid = ((jl_struct_type_t*)t)->uid;
else
uid = ((jl_bits_type_t*)t)->uid;
nd[uid & (len*2-1)] = (jl_value_t*)ml;
}
}
*pa = n;
}
static jl_methlist_t **mtcache_hash_bp(jl_array_t **pa, jl_value_t *ty,
int tparam)
{
uptrint_t uid;
if ((jl_is_struct_type(ty) && (uid = ((jl_struct_type_t*)ty)->uid)) ||
(jl_is_bits_type(ty) && (uid = ((jl_bits_type_t*)ty)->uid))) {
while (1) {
jl_methlist_t **pml = (jl_methlist_t**)&jl_cellref(*pa, uid & ((*pa)->length-1));
if (*pml == NULL || *pml == JL_NULL) {
*pml = JL_NULL;
return pml;
}
jl_value_t *t = jl_tupleref((*pml)->sig,0);
if (tparam) t = jl_tparam0(t);
if (t == ty)
return pml;
mtcache_rehash(pa);
}
}
return NULL;
}
/*
Method caches are divided into three parts: one for signatures where
the first argument is a singleton kind (Type{Foo}), one indexed by the
UID of the first argument's type in normal cases, and a fallback
table of everything else.
*/
static jl_function_t *jl_method_table_assoc_exact_by_type(jl_methtable_t *mt,
jl_tuple_t *types)
{
jl_methlist_t *ml = JL_NULL;
if (jl_tuple_len(types) > 0) {
jl_value_t *ty = jl_t0(types);
if (jl_is_type_type(ty)) {
jl_value_t *a0 = jl_tparam0(ty);
if (mt->cache_targ != JL_NULL) {
ml = mtcache_hash_lookup(mt->cache_targ, a0, 1);
if (ml!=JL_NULL)
goto mt_assoc_bt_lkup;
}
}
if (mt->cache_arg1 != JL_NULL) {
ml = mtcache_hash_lookup(mt->cache_arg1, ty, 0);
}
}
if (ml == JL_NULL)
ml = mt->cache;
mt_assoc_bt_lkup:
while (ml != JL_NULL) {
if (cache_match_by_type(&jl_tupleref(types,0), jl_tuple_len(types),
(jl_tuple_t*)ml->sig, ml->va)) {
return ml->func;
}
ml = ml->next;
}
return jl_bottom_func;
}
static jl_function_t *jl_method_table_assoc_exact(jl_methtable_t *mt,
jl_value_t **args, size_t n)
{
// NOTE: This function is a huge performance hot spot!!
jl_methlist_t *ml = JL_NULL;
if (n > 0) {
jl_value_t *a0 = args[0];
jl_value_t *ty = (jl_value_t*)jl_typeof(a0);
if (ty == (jl_value_t*)jl_struct_kind ||
ty == (jl_value_t*)jl_bits_kind) {
if (mt->cache_targ != JL_NULL) {
ml = mtcache_hash_lookup(mt->cache_targ, a0, 1);
if (ml != JL_NULL)
goto mt_assoc_lkup;
}
}
if (mt->cache_arg1 != JL_NULL) {
ml = mtcache_hash_lookup(mt->cache_arg1, ty, 0);
if (ml != JL_NULL) {
if (ml->next==JL_NULL && n==1 && jl_tuple_len(ml->sig)==1)
return ml->func;
if (n==2) {
// some manually-unrolled common special cases
jl_value_t *a1 = args[1];
jl_methlist_t *mn = ml;
if (jl_tuple_len(mn->sig)==2 &&
jl_tupleref(mn->sig,1)==(jl_value_t*)jl_typeof(a1))
return mn->func;
mn = mn->next;
if (mn!=JL_NULL && jl_tuple_len(mn->sig)==2 &&
jl_tupleref(mn->sig,1)==(jl_value_t*)jl_typeof(a1))
return mn->func;
}
}
}
}
if (ml == JL_NULL)
ml = mt->cache;
mt_assoc_lkup:
while (ml != JL_NULL) {
if (jl_tuple_len(ml->sig) == n || ml->va) {
if (cache_match(args, n, (jl_tuple_t*)ml->sig, ml->va)) {
return ml->func;
}
}
ml = ml->next;
}
return jl_bottom_func;
}
// return a new lambda-info that has some extra static parameters
// merged in.
jl_lambda_info_t *jl_add_static_parameters(jl_lambda_info_t *l, jl_tuple_t *sp)
{
JL_GC_PUSH(&sp);
if (jl_tuple_len(l->sparams) > 0)
sp = jl_tuple_append(sp, l->sparams);
jl_lambda_info_t *nli = jl_new_lambda_info(l->ast, sp);
nli->name = l->name;
nli->fptr = l->fptr;
nli->module = l->module;
nli->file = l->file;
nli->line = l->line;
JL_GC_POP();
return nli;
}
JL_CALLABLE(jl_trampoline);
jl_function_t *jl_instantiate_method(jl_function_t *f, jl_tuple_t *sp)
{
if (f->linfo == NULL)
return f;
jl_function_t *nf = jl_new_closure(f->fptr, f->env, NULL);
JL_GC_PUSH(&nf);
nf->linfo = jl_add_static_parameters(f->linfo, sp);
JL_GC_POP();
return nf;
}
// make a new method that calls the generated code from the given linfo
jl_function_t *jl_reinstantiate_method(jl_function_t *f, jl_lambda_info_t *li)
{
return jl_new_closure(NULL, f->env, li);
}
static
jl_methlist_t *jl_method_list_insert(jl_methlist_t **pml, jl_tuple_t *type,
jl_function_t *method, jl_tuple_t *tvars,
int check_amb);
static
jl_function_t *jl_method_cache_insert(jl_methtable_t *mt, jl_tuple_t *type,
jl_function_t *method)
{
jl_methlist_t **pml = &mt->cache;
if (jl_tuple_len(type) > 0) {
jl_value_t *t0 = jl_t0(type);
uptrint_t uid=0;
// if t0 != jl_typetype_type and the argument is Type{...}, this
// method has specializations for singleton kinds and we use
// the table indexed for that purpose.
if (t0 != (jl_value_t*)jl_typetype_type && jl_is_type_type(t0)) {
jl_value_t *a0 = jl_tparam0(t0);
if (jl_is_struct_type(a0))
uid = ((jl_struct_type_t*)a0)->uid;
else if (jl_is_bits_type(a0))
uid = ((jl_bits_type_t*)a0)->uid;
if (uid > 0) {
if (mt->cache_targ == JL_NULL)
mt->cache_targ = jl_alloc_cell_1d(16);
pml = mtcache_hash_bp(&mt->cache_targ, a0, 1);
goto ml_do_insert;
}
}
if (jl_is_struct_type(t0))
uid = ((jl_struct_type_t*)t0)->uid;
else if (jl_is_bits_type(t0))
uid = ((jl_bits_type_t*)t0)->uid;
if (uid > 0) {
if (mt->cache_arg1 == JL_NULL)
mt->cache_arg1 = jl_alloc_cell_1d(16);
pml = mtcache_hash_bp(&mt->cache_arg1, t0, 0);
}
}
ml_do_insert:
return jl_method_list_insert(pml, type, method, jl_null, 0)->func;
}
#if defined(JL_TRACE) || defined(TRACE_INFERENCE)
static char *type_summary(jl_value_t *t)
{
if (jl_is_tuple(t)) return "Tuple";
if (jl_is_some_tag_type(t))
return ((jl_tag_type_t*)t)->name->name->name;
JL_PRINTF(JL_STDERR, "unexpected argument type: ");
jl_show(jl_stderr_obj(), t);
JL_PRINTF(JL_STDERR, "\n");
assert(0);
return NULL;
}
#endif
#ifdef TRACE_INFERENCE
static void print_sig(jl_tuple_t *type)
{
size_t i;
for(i=0; i < jl_tuple_len(type); i++) {
if (i > 0) JL_PRINTF(JL_STDERR, ", ");
jl_value_t *v = jl_tupleref(type,i);
if (jl_is_tuple(v)) {
JL_PUTC('(', JL_STDERR);
print_sig((jl_tuple_t*)v);
JL_PUTC(')', JL_STDERR);
}
else {
JL_PRINTF(JL_STDERR, "%s", type_summary(v));
}
}
}
#endif
extern jl_function_t *jl_typeinf_func;
/*
run type inference on lambda "li" in-place, for given argument types.
"def" is the original method definition of which this is an instance;
can be equal to "li" if not applicable.
*/
int jl_in_inference = 0;
void jl_type_infer(jl_lambda_info_t *li, jl_tuple_t *argtypes,
jl_lambda_info_t *def)
{
int last_ii = jl_in_inference;
jl_in_inference = 1;
if (jl_typeinf_func != NULL) {
// TODO: this should be done right before code gen, so if it is
// interrupted we can try again the next time the function is
// called
assert(li->inInference == 0);
li->inInference = 1;
jl_value_t *fargs[4];
fargs[0] = (jl_value_t*)li;
fargs[1] = (jl_value_t*)argtypes;
fargs[2] = (jl_value_t*)jl_null;
fargs[3] = (jl_value_t*)def;
#ifdef TRACE_INFERENCE
JL_PRINTF(JL_STDERR,"inference on %s(", li->name->name);
print_sig(argtypes);
JL_PRINTF(JL_STDERR, ")\n");
#endif
#ifdef ENABLE_INFERENCE
jl_value_t *newast = jl_apply(jl_typeinf_func, fargs, 4);
li->ast = jl_tupleref(newast, 0);
li->inferred = 1;
#endif
li->inInference = 0;
}
jl_in_inference = last_ii;
}
static jl_value_t *nth_slot_type(jl_tuple_t *sig, size_t i)
{
size_t len = jl_tuple_len(sig);
if (len == 0)
return NULL;
if (i < len-1)
return jl_tupleref(sig, i);
if (jl_is_seq_type(jl_tupleref(sig,len-1))) {
return jl_tparam0(jl_tupleref(sig,len-1));
}
if (i == len-1)
return jl_tupleref(sig, i);
return NULL;
}
static int very_general_type(jl_value_t *t)
{
return (t && (t==(jl_value_t*)jl_any_type ||
(jl_is_typevar(t) &&
((jl_tvar_t*)t)->ub==(jl_value_t*)jl_any_type)));
}
static int tuple_all_Any(jl_tuple_t *t)
{
for(int i=0; i < jl_tuple_len(t); i++) {
if (jl_tupleref(t,i) != (jl_value_t*)jl_any_type)
return 0;
}
return 1;
}
static int is_kind(jl_value_t *v)
{
return (v==(jl_value_t*)jl_union_kind || v==(jl_value_t*)jl_struct_kind ||
v==(jl_value_t*)jl_bits_kind || v==(jl_value_t*)jl_typector_type ||
v==(jl_value_t*)jl_tag_kind);
}
static jl_value_t *ml_matches(jl_methlist_t *ml, jl_value_t *type,
jl_sym_t *name, int lim);
static jl_function_t *cache_method(jl_methtable_t *mt, jl_tuple_t *type,
jl_function_t *method, jl_tuple_t *decl,
jl_tuple_t *sparams)
{
size_t i;
int need_guard_entries = 0;
jl_value_t *temp=NULL;
jl_function_t *newmeth=NULL;
JL_GC_PUSH(&type, &temp, &newmeth);
for (i=0; i < jl_tuple_len(type); i++) {
jl_value_t *elt = jl_tupleref(type,i);
jl_value_t *decl_i = nth_slot_type(decl,i);
int set_to_any = 0;
if (decl_i == jl_ANY_flag) {
// don't specialize on slots marked ANY
temp = jl_tupleref(type, i);
jl_tupleset(type, i, (jl_value_t*)jl_any_type);
int nintr=0;
jl_methlist_t *curr = mt->defs;
// if this method is the only match even with the current slot
// set to Any, then it is safe to cache it that way.
while (curr != JL_NULL && curr->func!=method) {
if (jl_type_intersection((jl_value_t*)curr->sig,
(jl_value_t*)type) !=
(jl_value_t*)jl_bottom_type) {
nintr++;
break;
}
curr = curr->next;
}
if (nintr) {
// TODO: even if different specializations of this slot need
// separate cache entries, have them share code.
jl_tupleset(type, i, temp);
}
else {
set_to_any = 1;
}
}
if (set_to_any) {
}
else if (jl_is_tuple(elt)) {
/*
don't cache tuple type exactly; just remember that it was
a tuple, unless the declaration asks for something more
specific. determined with a type intersection.
*/
int might_need_guard=0;
temp = jl_tupleref(type, i);
if (i < jl_tuple_len(decl)) {
jl_value_t *declt = jl_tupleref(decl,i);
// for T..., intersect with T
if (jl_is_seq_type(declt))
declt = jl_tparam0(declt);
if (declt == (jl_value_t*)jl_tuple_type ||
jl_subtype((jl_value_t*)jl_tuple_type, declt, 0)) {
// don't specialize args that matched (Any...) or Any
jl_tupleset(type, i, (jl_value_t*)jl_tuple_type);
might_need_guard = 1;
}
else {
declt = jl_type_intersection(declt,
(jl_value_t*)jl_tuple_type);
if (jl_tuple_len(elt) > 3 ||
tuple_all_Any((jl_tuple_t*)declt)) {
jl_tupleset(type, i, declt);
might_need_guard = 1;
}
}
}
else {
jl_tupleset(type, i, (jl_value_t*)jl_tuple_type);
might_need_guard = 1;
}
assert(jl_tupleref(type,i) != (jl_value_t*)jl_bottom_type);
if (might_need_guard) {
jl_methlist_t *curr = mt->defs;
// can't generalize type if there's an overlapping definition
// with typevars.
// TODO: it seems premature to take these intersections
// before the whole signature has been generalized.
// example ((T...,),S,S,S,S,S,S,S,S,S,S,S,S,S,S,S,S,...)
while (curr != JL_NULL && curr->func!=method) {
if (curr->tvars!=jl_null &&
jl_type_intersection((jl_value_t*)curr->sig,
(jl_value_t*)type) !=
(jl_value_t*)jl_bottom_type) {
jl_tupleset(type, i, temp);
might_need_guard = 0;
break;
}
curr = curr->next;
}
}
if (might_need_guard) {
jl_methlist_t *curr = mt->defs;
while (curr != JL_NULL && curr->func!=method) {
jl_tuple_t *sig = curr->sig;
if (jl_tuple_len(sig) > i &&
jl_is_tuple(jl_tupleref(sig,i))) {
need_guard_entries = 1;
break;
}
curr = curr->next;
}
}
}
else if (jl_is_type_type(elt) && jl_is_type_type(jl_tparam0(elt))) {
/*
actual argument was Type{...}, we computed its type as
Type{Type{...}}. we must avoid unbounded nesting here, so
cache the signature as Type{T}, unless something more
specific like Type{Type{Int32}} was actually declared.
this can be determined using a type intersection.
*/
if (i < jl_tuple_len(decl)) {
jl_value_t *declt = jl_tupleref(decl,i);
// for T..., intersect with T
if (jl_is_seq_type(declt))
declt = jl_tparam0(declt);
jl_tupleset(type, i,
jl_type_intersection(declt, (jl_value_t*)jl_typetype_type));
}
else {
jl_tupleset(type, i, (jl_value_t*)jl_typetype_type);
}
assert(jl_tupleref(type,i) != (jl_value_t*)jl_bottom_type);
}
else if (jl_is_type_type(elt) && very_general_type(decl_i)) {
/*
here's a fairly complex heuristic: if this argument slot's
declared type is Any, and no definition overlaps with Type
for this slot, then don't specialize for every Type that
might be passed.
Since every type x has its own type Type{x}, this would be
excessive specialization for an Any slot.
TypeConstructors are problematic because they can be alternate
representations of any type. Extensionally, TC == TC.body, but
typeof(TC) != typeof(TC.body). This creates an ambiguity:
Type{TC} is type-equal to Type{TC.body}, yet a slot
x::TypeConstructor matches the first but not the second, while
also matching all other TypeConstructors. This means neither
Type{TC} nor TypeConstructor is more specific.
To solve this, we identify "kind slots", which are slots
for which some definition specifies a kind (e.g. AbstractKind).
Those tend to be in reflective functions that look at types
themselves. For these slots we specialize on jl_typeof(T) instead
of Type{T}, i.e. the kind of the type rather than the specific
type.
*/
int ok=1, kindslot=0;
jl_methlist_t *curr = mt->defs;
jl_value_t *kind = (jl_value_t*)jl_typeof(jl_tparam0(elt));
while (curr != JL_NULL) {
jl_value_t *slottype = nth_slot_type(curr->sig, i);
if (slottype && curr->func!=method) {
if (slottype == kind) {
ok=0;
break;
}
if (is_kind(slottype))
kindslot=1;
}
curr = curr->next;
}
if (ok) {
if (kindslot) {
jl_tupleset(type, i, kind);
}
else {
curr = mt->defs;
while (curr != JL_NULL) {
jl_value_t *slottype = nth_slot_type(curr->sig, i);
if (slottype && curr->func!=method) {
if (!very_general_type(slottype) &&
jl_type_intersection(slottype, (jl_value_t*)jl_type_type) !=
(jl_value_t*)jl_bottom_type) {
ok=0;
break;
}
}
curr = curr->next;
}
if (ok) {
jl_tupleset(type, i, jl_typetype_type);
}
}
}
}
else if (is_kind(decl_i)) {
// if a slot is specialized for a particular kind, it can be
// considered a reflective method and so only needs to be
// specialized for type representation, not type extent.
jl_methlist_t *curr = mt->defs;
int ok=1;
while (curr != JL_NULL) {
jl_value_t *slottype = nth_slot_type(curr->sig, i);
if (slottype && curr->func!=method) {
if (jl_subtype(slottype, decl_i, 0)) {
ok=0;
break;
}
}
curr = curr->next;
}
if (ok)
jl_tupleset(type, i, decl_i);
}
}
// for varargs methods, only specialize up to max_args.
// in general, here we want to find the biggest type that's not a
// supertype of any other method signatures. so far we are conservative
// and the types we find should be bigger.
if (jl_tuple_len(type) > mt->max_args &&
jl_is_seq_type(jl_tupleref(decl,jl_tuple_len(decl)-1))) {
size_t nspec = mt->max_args + 2;
jl_tuple_t *limited = jl_alloc_tuple(nspec);
for(i=0; i < nspec-1; i++) {
jl_tupleset(limited, i, jl_tupleref(type, i));
}
jl_value_t *lasttype = jl_tupleref(type,i-1);
// if all subsequent arguments are subtypes of lasttype, specialize
// on that instead of decl. for example, if decl is
// (Any...)
// and type is
// (Symbol, Symbol, Symbol)
// then specialize as (Symbol...), but if type is
// (Symbol, Int32, Expr)
// then specialize as (Any...)
size_t j = i;
int all_are_subtypes=1;
for(; j < jl_tuple_len(type); j++) {
if (!jl_subtype(jl_tupleref(type,j), lasttype, 0)) {
all_are_subtypes = 0;
break;
}
}
type = limited;
if (all_are_subtypes) {
// avoid Type{Type{...}...}...
if (jl_is_type_type(lasttype))
lasttype = (jl_value_t*)jl_type_type;
temp = (jl_value_t*)jl_tuple1(lasttype);
jl_tupleset(type, i, jl_apply_type((jl_value_t*)jl_seq_type,
(jl_tuple_t*)temp));
}
else {
jl_value_t *lastdeclt = jl_tupleref(decl,jl_tuple_len(decl)-1);
if (jl_tuple_len(sparams) > 0) {
lastdeclt = (jl_value_t*)
jl_instantiate_type_with((jl_type_t*)lastdeclt,
sparams->data,
jl_tuple_len(sparams)/2);
}
jl_tupleset(type, i, lastdeclt);
}
// now there is a problem: the computed signature is more
// general than just the given arguments, so it might conflict
// with another definition that doesn't have cache instances yet.
// to fix this, we insert guard cache entries for all intersections
// of this signature and definitions. those guard entries will
// supersede this one in conflicted cases, alerting us that there
// should actually be a cache miss.
need_guard_entries = 1;
}
if (need_guard_entries) {
temp = ml_matches(mt->defs, (jl_value_t*)type, lambda_sym, -1);
for(i=0; i < jl_array_len(temp); i++) {
jl_value_t *m = jl_cellref(temp, i);
if (jl_tupleref(m,2) != (jl_value_t*)method->linfo) {
jl_method_cache_insert(mt, (jl_tuple_t*)jl_tupleref(m, 0),
jl_bottom_func);
}
}
}
// here we infer types and specialize the method
/*
if (sparams==jl_null)
newmeth = method;
else
*/
jl_array_t *lilist=NULL;
jl_lambda_info_t *li=NULL;
if (method->linfo && method->linfo->specializations!=NULL) {
// reuse code already generated for this combination of lambda and
// arguments types. this happens for inner generic functions where
// a new closure is generated on each call to the enclosing function.
lilist = method->linfo->specializations;
int k;
for(k=0; k < lilist->length; k++) {
li = (jl_lambda_info_t*)jl_cellref(lilist, k);
if (jl_types_equal(li->specTypes, (jl_value_t*)type))
break;
}
if (k == lilist->length) lilist=NULL;
}
if (lilist != NULL && !li->inInference) {
assert(li);
newmeth = jl_reinstantiate_method(method, li);
(void)jl_method_cache_insert(mt, type, newmeth);
JL_GC_POP();
return newmeth;
}
else {
newmeth = jl_instantiate_method(method, sparams);
}
/*
if "method" itself can ever be compiled, for example for use as
an unspecialized method (see below), then newmeth->fptr might point
to some slow compiled code instead of jl_trampoline, meaning our
type-inferred code would never get compiled. this can be fixed with
the commented-out snippet below.
*/
assert(!(newmeth->linfo && newmeth->linfo->ast) ||
newmeth->fptr == &jl_trampoline);
/*
if (newmeth->linfo&&newmeth->linfo->ast&&newmeth->fptr!=&jl_trampoline) {
newmeth->fptr = &jl_trampoline;
}
*/
(void)jl_method_cache_insert(mt, type, newmeth);
if (newmeth->linfo != NULL && newmeth->linfo->sparams == jl_null) {
// when there are no static parameters, one unspecialized version
// of a function can be shared among all cached specializations.
if (method->linfo->unspecialized == NULL) {
method->linfo->unspecialized =
jl_instantiate_method(method, jl_null);
}
newmeth->linfo->unspecialized = method->linfo->unspecialized;
}
if (newmeth->linfo != NULL && newmeth->linfo->ast != NULL) {
newmeth->linfo->specTypes = (jl_value_t*)type;
jl_array_t *spe = method->linfo->specializations;
if (spe == NULL) {
spe = jl_alloc_cell_1d(1);
jl_cellset(spe, 0, newmeth->linfo);
}
else {
jl_cell_1d_push(spe, (jl_value_t*)newmeth->linfo);
}
method->linfo->specializations = spe;
jl_type_infer(newmeth->linfo, type, method->linfo);
}
JL_GC_POP();
return newmeth;
}
static jl_value_t *lookup_match(jl_value_t *a, jl_value_t *b, jl_tuple_t **penv,
jl_tuple_t *tvars)
{
jl_value_t *ti = jl_type_intersection_matching(a, b, penv, tvars);
if (ti == (jl_value_t*)jl_bottom_type)
return ti;
jl_value_t **ee = alloca(sizeof(void*) * jl_tuple_len(*penv));
int n=0;
// only keep vars in tvars list
jl_value_t **tvs;
int tvarslen;
if (jl_is_typevar(tvars)) {
tvs = (jl_value_t**)&tvars;
tvarslen = 1;
}
else {
tvs = &jl_t0(tvars);
tvarslen = jl_tuple_len(tvars);
}
for(int i=0; i < jl_tuple_len(*penv); i+=2) {
jl_value_t *v = jl_tupleref(*penv,i);
jl_value_t *val = jl_tupleref(*penv,i+1);
for(int j=0; j < tvarslen; j++) {
if (v == tvs[j]) {
ee[n++] = v;
ee[n++] = val;
}
}
}
if (n != jl_tuple_len(*penv)) {
jl_tuple_t *en = jl_alloc_tuple_uninit(n);
memcpy(en->data, ee, n*sizeof(void*));
*penv = en;
}
return ti;
}
static jl_function_t *jl_mt_assoc_by_type(jl_methtable_t *mt, jl_tuple_t *tt, int cache)
{
jl_methlist_t *m = mt->defs;
size_t nargs = jl_tuple_len(tt);
size_t i;
jl_value_t *ti=(jl_value_t*)jl_bottom_type;
jl_tuple_t *newsig=NULL, *env = jl_null;
JL_GC_PUSH(&env, &newsig);
while (m != JL_NULL) {
if (m->tvars!=jl_null) {
ti = lookup_match((jl_value_t*)tt, (jl_value_t*)m->sig,
&env, m->tvars);
if (ti != (jl_value_t*)jl_bottom_type) {
// parametric methods only match if all typevars are matched by
// non-typevars.
for(i=1; i < jl_tuple_len(env); i+=2) {
if (jl_is_typevar(jl_tupleref(env,i)))
break;
}
if (i >= jl_tuple_len(env))
break;
ti = (jl_value_t*)jl_bottom_type;
}
}
else if (jl_tuple_subtype(&jl_tupleref(tt,0), nargs,
&jl_tupleref(m->sig,0),
jl_tuple_len(m->sig), 0, 0)) {
break;
}
m = m->next;
}
if (ti == (jl_value_t*)jl_bottom_type) {
JL_GC_POP();
if (m != JL_NULL) {
if (!cache)
return m->func;
return cache_method(mt, tt, m->func, (jl_tuple_t*)m->sig, jl_null);
}
return jl_bottom_func;
}
assert(jl_is_tuple(env));
// don't bother computing this if no arguments are tuples
for(i=0; i < jl_tuple_len(tt); i++) {
if (jl_is_tuple(jl_tupleref(tt,i)))
break;
}
if (i < jl_tuple_len(tt)) {
newsig = (jl_tuple_t*)jl_instantiate_type_with((jl_type_t*)m->sig,
&jl_tupleref(env,0),
jl_tuple_len(env)/2);
}
else {
newsig = (jl_tuple_t*)m->sig;
}
assert(jl_is_tuple(newsig));
jl_function_t *nf;
if (!cache)
nf = m->func;
else
nf = cache_method(mt, tt, m->func, newsig, env);
JL_GC_POP();
return nf;
}
jl_tag_type_t *jl_wrap_Type(jl_value_t *t);
static int sigs_eq(jl_value_t *a, jl_value_t *b)
{
if (jl_has_typevars(a) || jl_has_typevars(b)) {
return jl_types_equal_generic(a,b);
}
return jl_subtype(a, b, 0) && jl_subtype(b, a, 0);
}
int jl_args_morespecific(jl_value_t *a, jl_value_t *b)
{
int msp = jl_type_morespecific(a,b,0);
if (jl_has_typevars(b)) {
if (jl_type_match_morespecific(a,b) == (jl_value_t*)jl_false) {
if (jl_has_typevars(a))
return 0;
return msp;
}
if (jl_has_typevars(a)) {
//if (jl_type_match_morespecific(b,a) == (jl_value_t*)jl_false)
// return 1;
// this rule seems to work better:
if (jl_type_match(b,a) == (jl_value_t*)jl_false)
return 1;
}
int nmsp = jl_type_morespecific(b,a,0);
if (nmsp == msp)
return 0;
}
if (jl_has_typevars((jl_value_t*)a)) {
int nmsp = jl_type_morespecific(b,a,0);
if (nmsp && msp)
return 1;
if (jl_type_match_morespecific(b,a) != (jl_value_t*)jl_false) {
return 0;
}
}
return msp;
}
static int is_va_tuple(jl_tuple_t *t)
{
return (jl_tuple_len(t)>0 && jl_is_seq_type(jl_tupleref(t,jl_tuple_len(t)-1)));
}
/*
warn about ambiguous method priorities
the relative priority of A and B is ambiguous if
!subtype(A,B) && !subtype(B,A) && no corresponding tuple
elements are disjoint.
for example, (AbstractArray, AbstractMatrix) and (AbstractMatrix, AbstractArray) are ambiguous.
however, (AbstractArray, AbstractMatrix, Foo) and (AbstractMatrix, AbstractArray, Bar) are fine
since Foo and Bar are disjoint, so there would be no confusion over
which one to call.
There is also this kind of ambiguity: foo{T,S}(T, S) vs. foo(Any,Any)
In this case jl_types_equal() is true, but one is jl_type_morespecific
or jl_type_match_morespecific than the other.
To check this, jl_types_equal_generic needs to be more sophisticated
so (T,T) is not equivalent to (Any,Any). (TODO)
*/
static void check_ambiguous(jl_methlist_t *ml, jl_tuple_t *type,
jl_tuple_t *sig, jl_sym_t *fname)
{
size_t tl = jl_tuple_len(type);
size_t sl = jl_tuple_len(sig);
// we know !jl_args_morespecific(type, sig)
if ((tl==sl ||
(tl==sl+1 && is_va_tuple(type)) ||
(tl+1==sl && is_va_tuple(sig))) &&
!jl_args_morespecific((jl_value_t*)sig, (jl_value_t*)type)) {
jl_value_t *isect = jl_type_intersection((jl_value_t*)type,
(jl_value_t*)sig);
if (isect == (jl_value_t*)jl_bottom_type)
return;
JL_GC_PUSH(&isect);
jl_methlist_t *l = ml;
while (l != JL_NULL) {
if (sigs_eq(isect, (jl_value_t*)l->sig))
goto done_chk_amb; // ok, intersection is covered
l = l->next;
}
char *n = fname->name;
jl_value_t *errstream = jl_stderr_obj();
JL_STREAM *s = JL_STDERR;
JL_PRINTF(s, "Warning: New definition %s", n);
jl_show(errstream, (jl_value_t*)type);
JL_PRINTF(s, " is ambiguous with %s", n);
jl_show(errstream, (jl_value_t*)sig);
JL_PRINTF(s, ".\n Make sure %s", n);
jl_show(errstream, isect);
JL_PRINTF(s, " is defined first.\n");
done_chk_amb:
JL_GC_POP();
}
}
static int has_unions(jl_tuple_t *type)
{
int i;
for(i=0; i < jl_tuple_len(type); i++) {
jl_value_t *t = jl_tupleref(type,i);
if (jl_is_union_type(t) ||
(jl_is_seq_type(t) && jl_is_union_type(jl_tparam0(t))))
return 1;
}
return 0;
}
static
jl_methlist_t *jl_method_list_insert(jl_methlist_t **pml, jl_tuple_t *type,
jl_function_t *method, jl_tuple_t *tvars,
int check_amb)
{
jl_methlist_t *l, **pl;
assert(jl_is_tuple(type));
l = *pml;
while (l != JL_NULL) {
if (((l->tvars==jl_null) == (tvars==jl_null)) &&
sigs_eq((jl_value_t*)type, (jl_value_t*)l->sig)) {
// method overwritten
JL_SIGATOMIC_BEGIN();
l->sig = type;
l->tvars = tvars;
l->va = (jl_tuple_len(type) > 0 &&
jl_is_seq_type(jl_tupleref(type,jl_tuple_len(type)-1))) ?
1 : 0;
l->invokes = JL_NULL;
l->func = method;
JL_SIGATOMIC_END();
return l;
}
l = l->next;
}
pl = pml;
l = *pml;
while (l != JL_NULL) {
if (jl_args_morespecific((jl_value_t*)type, (jl_value_t*)l->sig))
break;
if (check_amb) {
check_ambiguous(*pml, (jl_tuple_t*)type, (jl_tuple_t*)l->sig,
method->linfo ? method->linfo->name :
anonymous_sym);
}
pl = &l->next;
l = l->next;
}
jl_methlist_t *newrec = (jl_methlist_t*)allocobj(sizeof(jl_methlist_t));
newrec->type = (jl_type_t*)jl_method_type;
newrec->sig = type;
newrec->tvars = tvars;
newrec->va = (jl_tuple_len(type) > 0 &&
jl_is_seq_type(jl_tupleref(type,jl_tuple_len(type)-1))) ?
1 : 0;
newrec->func = method;
newrec->invokes = JL_NULL;
newrec->next = l;
JL_SIGATOMIC_BEGIN();
*pl = newrec;
// if this contains Union types, methods after it might actually be
// more specific than it. we need to re-sort them.
if (has_unions(type)) {
jl_methlist_t *item = newrec->next, *next;
jl_methlist_t **pitem = &newrec->next, **pnext;
while (item != JL_NULL) {
pl = pml;
l = *pml;
next = item->next;
pnext = &item->next;
while (l != newrec->next) {
if (jl_args_morespecific((jl_value_t*)item->sig,
(jl_value_t*)l->sig)) {
// reinsert item earlier in the list
*pitem = next;
item->next = l;
*pl = item;
pnext = pitem;
break;
}
pl = &l->next;
l = l->next;
}
item = next;
pitem = pnext;
}
}
JL_SIGATOMIC_END();
return newrec;
}
static void remove_conflicting(jl_methlist_t **pl, jl_value_t *type)
{
jl_methlist_t *l = *pl;
while (l != JL_NULL) {
if (jl_type_intersection(type, (jl_value_t*)l->sig) !=
(jl_value_t*)jl_bottom_type) {
*pl = l->next;
}
else {
pl = &l->next;
}
l = l->next;
}
}
jl_methlist_t *jl_method_table_insert(jl_methtable_t *mt, jl_tuple_t *type,
jl_function_t *method, jl_tuple_t *tvars)
{
if (jl_tuple_len(tvars) == 1)
tvars = (jl_tuple_t*)jl_t0(tvars);
JL_SIGATOMIC_BEGIN();
jl_methlist_t *ml = jl_method_list_insert(&mt->defs,type,method,tvars,1);
// invalidate cached methods that overlap this definition
remove_conflicting(&mt->cache, (jl_value_t*)type);
if (mt->cache_arg1 != JL_NULL) {
for(int i=0; i < jl_array_len(mt->cache_arg1); i++) {
jl_methlist_t **pl = (jl_methlist_t**)&jl_cellref(mt->cache_arg1,i);
if (*pl && *pl != JL_NULL)
remove_conflicting(pl, (jl_value_t*)type);
}
}
if (mt->cache_targ != JL_NULL) {
for(int i=0; i < jl_array_len(mt->cache_targ); i++) {
jl_methlist_t **pl = (jl_methlist_t**)&jl_cellref(mt->cache_targ,i);
if (*pl && *pl != JL_NULL)
remove_conflicting(pl, (jl_value_t*)type);
}
}
// update max_args
jl_tuple_t *t = (jl_tuple_t*)type;
size_t na = jl_tuple_len(t);
if (is_va_tuple(t))
na--;
if (na > mt->max_args) {
mt->max_args = na;
}
JL_SIGATOMIC_END();
return ml;
}
jl_value_t *jl_no_method_error(jl_function_t *f, jl_value_t **args, size_t na)
{
jl_value_t **a = alloca(sizeof(jl_value_t*)*(na+1));
a[0] = (jl_value_t*)f;
int i;
for(i=0; i < na; i++)
a[i+1] = args[i];
return jl_apply(jl_method_missing_func, a, na+1);
}
static jl_tuple_t *arg_type_tuple(jl_value_t **args, size_t nargs)
{
jl_tuple_t *tt = jl_alloc_tuple(nargs);
JL_GC_PUSH(&tt);
size_t i;
for(i=0; i < jl_tuple_len(tt); i++) {
jl_value_t *a;
if (jl_is_nontuple_type(args[i])) { //***
a = (jl_value_t*)jl_wrap_Type(args[i]);
}
else {
a = (jl_value_t*)jl_full_type(args[i]);
}
jl_tupleset(tt, i, a);
}
JL_GC_POP();
return tt;
}
jl_function_t *jl_method_lookup_by_type(jl_methtable_t *mt, jl_tuple_t *types,
int cache)
{
jl_function_t *sf = jl_method_table_assoc_exact_by_type(mt, types);
if (sf == jl_bottom_func) {
sf = jl_mt_assoc_by_type(mt, types, cache);
}
return sf;
}
jl_function_t *jl_method_lookup(jl_methtable_t *mt, jl_value_t **args, size_t nargs, int cache)
{
jl_function_t *sf = jl_method_table_assoc_exact(mt, args, nargs);
if (sf == jl_bottom_func) {
jl_tuple_t *tt = arg_type_tuple(args, nargs);
JL_GC_PUSH(&tt);
sf = jl_mt_assoc_by_type(mt, tt, cache);
JL_GC_POP();
}
return sf;
}
// compile-time method lookup
jl_function_t *jl_get_specialization(jl_function_t *f, jl_tuple_t *types)
{
assert(jl_is_gf(f));
if (!jl_is_leaf_type((jl_value_t*)types))
return NULL;
jl_methtable_t *mt = jl_gf_mtable(f);
jl_function_t *sf = jl_method_lookup_by_type(mt, types, 1);
if (sf == jl_bottom_func) {
return NULL;
}
if (sf->linfo == NULL || sf->linfo->ast == NULL) {
return NULL;
}
if (sf->linfo->inInference) return NULL;
if (sf->linfo->functionObject == NULL) {
if (sf->fptr != &jl_trampoline)
return NULL;
jl_compile(sf);
}
return sf;
}
DLLEXPORT void jl_compile_hint(jl_function_t *f, jl_tuple_t *types)
{
(void)jl_get_specialization(f, types);
}
#ifdef JL_TRACE
static int trace_en = 0;
static int error_en = 1;
static void __attribute__ ((unused)) enable_trace(int x) { trace_en=x; }
static void show_call(jl_value_t *F, jl_value_t **args, uint32_t nargs)
{
JL_PRINTF(JL_STDOUT, "%s(", jl_gf_name(F)->name);
for(size_t i=0; i < nargs; i++) {
if (i > 0) JL_PRINTF(JL_STDOUT, ", ");
JL_PRINTF(JL_STDOUT, "%s", type_summary((jl_value_t*)jl_typeof(args[i])));
}
JL_PRINTF(JL_STDOUT, ")\n");
}
#endif
JL_CALLABLE(jl_apply_generic)
{
jl_methtable_t *mt = jl_gf_mtable(F);
#ifdef JL_GF_PROFILE
mt->ncalls++;
#endif
#ifdef JL_TRACE
if (trace_en) {
show_call(F, args, nargs);
}
#endif
/*
search order:
look at concrete signatures
if there is an exact match, return it
otherwise look for a matching generic signature
if no concrete or generic match, raise error
if no generic match, use the concrete one even if inexact
otherwise instantiate the generic method and use it
*/
jl_function_t *mfunc = jl_method_table_assoc_exact(mt, args, nargs);
if (mfunc != jl_bottom_func) {
if (mfunc->linfo != NULL &&
(mfunc->linfo->inInference || mfunc->linfo->inCompile)) {
// if inference is running on this function, return a copy
// of the function to be compiled without inference and run.
jl_lambda_info_t *li = mfunc->linfo;
if (li->unspecialized == NULL) {
li->unspecialized = jl_instantiate_method(mfunc, li->sparams);
}
mfunc = li->unspecialized;
}
}
else {
jl_tuple_t *tt = arg_type_tuple(args, nargs);
JL_GC_PUSH(&tt);
mfunc = jl_mt_assoc_by_type(mt, tt, 1);
JL_GC_POP();
}
if (mfunc == jl_bottom_func) {
#ifdef JL_TRACE
if (error_en) {
show_call(F, args, nargs);
}
#endif
return jl_no_method_error((jl_function_t*)F, args, nargs);
}
assert(!mfunc->linfo || !mfunc->linfo->inInference);
return jl_apply(mfunc, args, nargs);
}
// invoke()
// this does method dispatch with a set of types to match other than the
// types of the actual arguments. this means it sometimes does NOT call the
// most specific method for the argument types, so we need different logic.
// first we use the given types to look up a definition, then we perform
// caching and specialization within just that definition.
// every definition has its own private method table for this purpose.
//
// NOTE: assumes argument type is a subtype of the lookup type.
jl_value_t *jl_gf_invoke(jl_function_t *gf, jl_tuple_t *types,
jl_value_t **args, size_t nargs)
{
assert(jl_is_gf(gf));
jl_methtable_t *mt = jl_gf_mtable(gf);
jl_methlist_t *m = mt->defs;
size_t typelen = jl_tuple_len(types);
size_t i;
jl_value_t *env = (jl_value_t*)jl_false;
while (m != JL_NULL) {
if (m->tvars!=jl_null) {
env = jl_type_match((jl_value_t*)types, (jl_value_t*)m->sig);
if (env != (jl_value_t*)jl_false) break;
}
else if (jl_tuple_subtype(&jl_tupleref(types,0), typelen,
&jl_tupleref(m->sig,0),
jl_tuple_len(m->sig), 0, 0)) {
break;
}
m = m->next;
}
if (m == JL_NULL) {
return jl_no_method_error(gf, args, nargs);
}
// now we have found the matching definition.
// next look for or create a specialization of this definition.
jl_function_t *mfunc;
if (m->invokes == JL_NULL)
mfunc = jl_bottom_func;
else
mfunc = jl_method_table_assoc_exact(m->invokes, args, nargs);
if (mfunc != jl_bottom_func) {
if (mfunc->linfo != NULL &&
(mfunc->linfo->inInference || mfunc->linfo->inCompile)) {
// if inference is running on this function, return a copy
// of the function to be compiled without inference and run.
jl_lambda_info_t *li = mfunc->linfo;
if (li->unspecialized == NULL) {
li->unspecialized = jl_instantiate_method(mfunc, li->sparams);
}
mfunc = li->unspecialized;
}
}
else {
jl_tuple_t *tpenv=jl_null;
jl_tuple_t *newsig=NULL;
jl_tuple_t *tt=NULL;
JL_GC_PUSH(&env, &newsig, &tt);
if (m->invokes == JL_NULL) {
m->invokes = new_method_table(mt->name);
// this private method table has just this one definition
jl_method_list_insert(&m->invokes->defs,m->sig,m->func,m->tvars,0);
}
tt = arg_type_tuple(args, nargs);
newsig = (jl_tuple_t*)m->sig;
if (env != (jl_value_t*)jl_false) {
tpenv = (jl_tuple_t*)env;
// don't bother computing this if no arguments are tuples
for(i=0; i < jl_tuple_len(tt); i++) {
if (jl_is_tuple(jl_tupleref(tt,i)))
break;
}
if (i < jl_tuple_len(tt)) {
newsig =
(jl_tuple_t*)jl_instantiate_type_with((jl_type_t*)m->sig,
&jl_tupleref(tpenv,0),
jl_tuple_len(tpenv)/2);
}
}
mfunc = cache_method(m->invokes, tt, m->func, newsig, tpenv);
JL_GC_POP();
}
return jl_apply(mfunc, args, nargs);
}
static void print_methlist(jl_value_t *outstr, char *name, jl_methlist_t *ml)
{
JL_STREAM *s = (JL_STREAM*)jl_iostr_data(outstr);
while (ml != JL_NULL) {
JL_PRINTF(s, "%s", name);
if (ml->tvars != jl_null) {
if (jl_is_typevar(ml->tvars)) {
JL_PUTC('{', s); jl_show(outstr, (jl_value_t*)ml->tvars);
JL_PUTC('}', s);
}
else {
jl_show_tuple(outstr, ml->tvars, '{', '}', 0);
}
}
jl_show(outstr, (jl_value_t*)ml->sig);
if (ml->func == jl_bottom_func) {
// mark guard cache entries
JL_PRINTF(s, " *");
}
else {
jl_lambda_info_t *li = ml->func->linfo;
assert(li);
long lno = li->line;
if (lno > 0) {
char *fname = ((jl_sym_t*)li->file)->name;
JL_PRINTF(s, " at %s:%d", fname, lno);
}
}
if (ml->next != JL_NULL)
JL_PRINTF(s, "\n");
ml = ml->next;
}
}
void jl_show_method_table(jl_value_t *outstr, jl_function_t *gf)
{
char *name = jl_gf_name(gf)->name;
jl_methtable_t *mt = jl_gf_mtable(gf);
print_methlist(outstr, name, mt->defs);
//JL_PRINTF(JL_STDOUT, "\ncache:\n");
//print_methlist(outstr, name, mt->cache);
}
void jl_initialize_generic_function(jl_function_t *f, jl_sym_t *name)
{
f->fptr = jl_apply_generic;
f->env = (jl_value_t*)new_method_table(name);
}
jl_function_t *jl_new_generic_function(jl_sym_t *name)
{
jl_function_t *f = jl_new_closure(jl_apply_generic, NULL, NULL);
JL_GC_PUSH(&f);
jl_initialize_generic_function(f, name);
JL_GC_POP();
return f;
}
void jl_add_method(jl_function_t *gf, jl_tuple_t *types, jl_function_t *meth,
jl_tuple_t *tvars)
{
assert(jl_is_function(gf));
assert(jl_is_tuple(types));
assert(jl_is_func(meth));
assert(jl_is_mtable(jl_gf_mtable(gf)));
if (meth->linfo != NULL)
meth->linfo->name = jl_gf_name(gf);
(void)jl_method_table_insert(jl_gf_mtable(gf), types, meth, tvars);
}
DLLEXPORT jl_tuple_t *jl_match_method(jl_value_t *type, jl_value_t *sig,
jl_tuple_t *tvars)
{
jl_tuple_t *env = jl_null;
jl_value_t *ti=NULL;
JL_GC_PUSH(&env, &ti);
ti = lookup_match(type, (jl_value_t*)sig, &env, tvars);
jl_tuple_t *result = jl_tuple2(ti, env);
JL_GC_POP();
return result;
}
static jl_tuple_t *match_method(jl_value_t *type, jl_function_t *func,
jl_tuple_t *sig, jl_tuple_t *tvars)
{
jl_tuple_t *env = jl_null;
jl_value_t *ti=NULL;
JL_GC_PUSH(&env, &ti);
ti = lookup_match(type, (jl_value_t*)sig, &env, tvars);
jl_tuple_t *result = NULL;
if (ti != (jl_value_t*)jl_bottom_type) {
assert(func->linfo); // no builtin methods
jl_value_t *cenv;
if (func->env != NULL) {
cenv = func->env;
}
else {
cenv = (jl_value_t*)jl_null;
}
result = jl_tuple(4, ti, env, func->linfo, cenv);
}
JL_GC_POP();
return result;
}
// returns linked tuples (argtypes, static_params, lambdainfo, cloenv, next)
static jl_value_t *ml_matches(jl_methlist_t *ml, jl_value_t *type,
jl_sym_t *name, int lim)
{
jl_array_t *t = (jl_array_t*)jl_an_empty_cell;
jl_tuple_t *matc=NULL;
JL_GC_PUSH(&t, &matc);
int len=0;
while (ml != JL_NULL) {
// a method is shadowed if type <: S <: m->sig where S is the
// signature of another applicable method
/*
more generally, we can stop when the type is a subtype of the
union of all the signatures examined so far.
*/
matc = match_method(type, ml->func, ml->sig, ml->tvars);
if (matc != NULL) {
len++;
if (lim >= 0 && len > lim) {
JL_GC_POP();
return jl_false;
}
if (len == 1) {
t = jl_alloc_cell_1d(1);
jl_cellref(t,0) = (jl_value_t*)matc;
}
else {
jl_cell_1d_push(t, (jl_value_t*)matc);
}
// (type ∩ ml->sig == type) ⇒ (type ⊆ ml->sig)
// NOTE: jl_subtype check added in case the intersection is
// over-approximated.
if (jl_types_equal(jl_t0(matc), type) &&
jl_subtype(type, (jl_value_t*)ml->sig, 0)) {
JL_GC_POP();
return (jl_value_t*)t;
}
}
ml = ml->next;
}
JL_GC_POP();
return (jl_value_t*)t;
}
void jl_add_constructors(jl_struct_type_t *t);
JL_CALLABLE(jl_f_ctor_trampoline);
// return linked tuples (t1, M1, (t2, M2, (... ()))) of types and methods.
// t is the intersection of the type argument and the method signature,
// and M is the corresponding LambdaStaticData (jl_lambda_info_t)
// lim is the max # of methods to return. if there are more return jl_false.
// -1 for no limit.
DLLEXPORT
jl_value_t *jl_matching_methods(jl_function_t *gf, jl_value_t *type, int lim)
{
assert(jl_is_func(gf));
if (gf->fptr == jl_f_no_function)
return (jl_value_t*)jl_an_empty_cell;
if (gf->fptr == jl_f_ctor_trampoline)
jl_add_constructors((jl_struct_type_t*)gf);
if (!jl_is_gf(gf)) {
return (jl_value_t*)jl_an_empty_cell;
}
jl_methtable_t *mt = jl_gf_mtable(gf);
return ml_matches(mt->defs, type, jl_gf_name(gf), lim);
}
DLLEXPORT
int jl_is_builtin(jl_value_t *v)
{
return ((jl_is_func(v) && (((jl_function_t*)v)->linfo==NULL) &&
!jl_is_gf(v)) ||
jl_typeis(v,jl_intrinsic_type));
}
DLLEXPORT
int jl_is_genericfunc(jl_value_t *v)
{
return (jl_is_func(v) && jl_is_gf(v));
}
DLLEXPORT
jl_sym_t *jl_genericfunc_name(jl_value_t *v)
{
return jl_gf_name(v);
}
Jump to Line
Something went wrong with that request. Please try again.