libdrgn: dwarf_index: return indexed DIE entry from drgn_dwarf_index_iterator_next()

For namespace support, we will want to access the struct
drgn_dwarf_index_die for namespaces instead of the Dwarf_Die. Split
drgn_dwarf_index_get_die() out of drgn_dwarf_index_iterator_next().

Signed-off-by: Omar Sandoval <osandov@osandov.com>

libdrgn/dwarf_index.c

@@ -252,26 +252,6 @@ static inline const char *section_end(Elf_Data *data)
 	return (const char *)data->d_buf + data->d_size;
 }
 
-/*
- * An indexed DIE.
- *
- * DIEs with the same name but different tags or files are considered distinct.
- * We only compare the hash of the file name, not the string value, because a
- * 64-bit collision is unlikely enough, especially when also considering the
- * name and tag.
- */
-struct drgn_dwarf_index_die {
-	/*
-	 * The next DIE with the same name (as an index into
-	 * drgn_dwarf_index_shard::dies), or UINT32_MAX if this is the last DIE.
-	 */
-	uint32_t next;
-	uint8_t tag;
-	uint64_t file_name_hash;
-	Dwfl_Module *module;
-	size_t offset;
-};
-
 DEFINE_HASH_TABLE_FUNCTIONS(drgn_dwarf_index_die_map, string_hash, string_eq)
 DEFINE_VECTOR_FUNCTIONS(drgn_dwarf_index_die_vector)
 DEFINE_HASH_TABLE_FUNCTIONS(drgn_dwarf_index_specification_map,
@@ -2619,23 +2599,18 @@ drgn_dwarf_index_iterator_matches_tag(struct drgn_dwarf_index_iterator *it,
 	return false;
 }
 
-struct drgn_error *
+struct drgn_dwarf_index_die *
-drgn_dwarf_index_iterator_next(struct drgn_dwarf_index_iterator *it,
+drgn_dwarf_index_iterator_next(struct drgn_dwarf_index_iterator *it)
-			       Dwarf_Die *die_ret, uint64_t *bias_ret)
 {
 	struct drgn_dwarf_index *dindex = it->dindex;
 	struct drgn_dwarf_index_die *die;
-	Dwarf *dwarf;
-	Dwarf_Addr bias;
-
 	if (it->any_name) {
 		for (;;) {
-			struct drgn_dwarf_index_shard *shard;
-
 			if (it->shard >= ARRAY_SIZE(dindex->shards))
-				return &drgn_stop;
+				return NULL;
 
-			shard = &dindex->shards[it->shard];
+			struct drgn_dwarf_index_shard *shard =
+				&dindex->shards[it->shard];
 			die = &shard->dies.data[it->index];
 
 			if (++it->index >= shard->dies.size) {
@@ -2651,12 +2626,11 @@ drgn_dwarf_index_iterator_next(struct drgn_dwarf_index_iterator *it,
 		}
 	} else {
 		for (;;) {
-			struct drgn_dwarf_index_shard *shard;
-
 			if (it->index == UINT32_MAX)
-				return &drgn_stop;
+				return NULL;
 
-			shard = &dindex->shards[it->shard];
+			struct drgn_dwarf_index_shard *shard =
+				&dindex->shards[it->shard];
 			die = &shard->dies.data[it->index];
 
 			it->index = die->next;
@@ -2665,8 +2639,15 @@ drgn_dwarf_index_iterator_next(struct drgn_dwarf_index_iterator *it,
 				break;
 		}
 	}
+	return die;
+}
 
-	dwarf = dwfl_module_getdwarf(die->module, &bias);
+struct drgn_error *drgn_dwarf_index_get_die(struct drgn_dwarf_index_die *die,
+					    Dwarf_Die *die_ret,
+					    uint64_t *bias_ret)
+{
+	Dwarf_Addr bias;
+	Dwarf *dwarf = dwfl_module_getdwarf(die->module, &bias);
 	if (!dwarf)
 		return drgn_error_libdwfl();
 	if (!dwarf_offdie(dwarf, die->offset, die_ret))

libdrgn/dwarf_index.h

@@ -62,7 +62,32 @@ extern const Dwfl_Callbacks drgn_dwfl_callbacks;
 extern const Dwfl_Callbacks drgn_linux_proc_dwfl_callbacks;
 extern const Dwfl_Callbacks drgn_userspace_core_dump_dwfl_callbacks;
 
-struct drgn_dwarf_index_die;
+/*
+ * An indexed DIE.
+ *
+ * DIEs with the same name but different tags or files are considered distinct.
+ * We only compare the hash of the file name, not the string value, because a
+ * 64-bit collision is unlikely enough, especially when also considering the
+ * name and tag.
+ */
+struct drgn_dwarf_index_die {
+	/*
+	 * The next DIE with the same name (as an index into
+	 * drgn_dwarf_index_shard::dies), or UINT32_MAX if this is the last DIE.
+	 */
+	uint32_t next;
+	uint8_t tag;
+	union {
+		/* If tag != DW_TAG_namespace. */
+		uint64_t file_name_hash;
+		/* TODO: explain hash */
+		/* If tag == DW_TAG_namespace. */
+		struct drgn_dwarf_index_namespace *namespace;
+	};
+	Dwfl_Module *module;
+	size_t offset;
+};
+
 DEFINE_HASH_MAP_TYPE(drgn_dwarf_index_die_map, struct string, uint32_t)
 DEFINE_VECTOR_TYPE(drgn_dwarf_index_die_vector, struct drgn_dwarf_index_die)
 
@@ -411,18 +436,24 @@ void drgn_dwarf_index_iterator_init(struct drgn_dwarf_index_iterator *it,
  * DW_TAG_enumerator DIEs.
  *
  * @param[in] it DWARF index iterator.
+ * @return Next DIE, or @c NULL if there are no more matching DIEs.
+ */
+struct drgn_dwarf_index_die *
+drgn_dwarf_index_iterator_next(struct drgn_dwarf_index_iterator *it);
+
+/**
+ * Get a @c Dwarf_Die from a @ref drgn_dwarf_index_die.
+ *
+ * @param[in] die Indexed DIE.
  * @param[out] die_ret Returned DIE.
  * @param[out] bias_ret Returned difference between addresses in the loaded
  * module and addresses in the debugging information. This may be @c NULL if it
  * is not needed.
- * @return @c NULL on success, non-@c NULL on error. In particular, when there
+ * @return @c NULL on success, non-@c NULL on error.
- * are no more matching DIEs, @p die_ret is not modified and an error with code
- * @ref DRGN_ERROR_STOP is returned; this @ref DRGN_ERROR_STOP error does not
- * have to be passed to @ref drgn_error_destroy().
  */
-struct drgn_error *
+struct drgn_error *drgn_dwarf_index_get_die(struct drgn_dwarf_index_die *die,
-drgn_dwarf_index_iterator_next(struct drgn_dwarf_index_iterator *it,
+					    Dwarf_Die *die_ret,
-			       Dwarf_Die *die_ret, uint64_t *bias_ret);
+					    uint64_t *bias_ret);
 
 /** @} */
 

libdrgn/dwarf_info_cache.c

@@ -346,29 +346,30 @@ drgn_dwarf_info_cache_find_complete(struct drgn_dwarf_info_cache *dicache,
 				    struct drgn_type **ret)
 {
 	struct drgn_error *err;
-	struct drgn_dwarf_index_iterator it;
-	Dwarf_Die die;
-	struct drgn_qualified_type qualified_type;
 
+	struct drgn_dwarf_index_iterator it;
 	drgn_dwarf_index_iterator_init(&it, &dicache->dindex, name,
 				       strlen(name), &tag, 1);
 	/*
 	 * Find a matching DIE. Note that drgn_dwarf_index does not contain DIEs
 	 * with DW_AT_declaration, so this will always be a complete type.
 	 */
-	err = drgn_dwarf_index_iterator_next(&it, &die, NULL);
+	struct drgn_dwarf_index_die *index_die =
-	if (err)
+		drgn_dwarf_index_iterator_next(&it);
-		return err;
+	if (!index_die)
+		return &drgn_stop;
 	/*
 	 * Look for another matching DIE. If there is one, then we can't be sure
 	 * which type this is, so leave it incomplete rather than guessing.
 	 */
-	err = drgn_dwarf_index_iterator_next(&it, &die, NULL);
+	if (drgn_dwarf_index_iterator_next(&it))
-	if (!err)
 		return &drgn_stop;
-	else if (err->code != DRGN_ERROR_STOP)
-		return err;
 
+	Dwarf_Die die;
+	err = drgn_dwarf_index_get_die(index_die, &die, NULL);
+	if (err)
+		return err;
+	struct drgn_qualified_type qualified_type;
 	err = drgn_type_from_dwarf(dicache, &die, &qualified_type);
 	if (err)
 		return err;
@@ -1237,10 +1238,8 @@ struct drgn_error *drgn_dwarf_type_find(enum drgn_type_kind kind,
 {
 	struct drgn_error *err;
 	struct drgn_dwarf_info_cache *dicache = arg;
-	struct drgn_dwarf_index_iterator it;
-	Dwarf_Die die;
-	uint64_t tag;
 
+	uint64_t tag;
 	switch (kind) {
 	case DRGN_TYPE_INT:
 	case DRGN_TYPE_BOOL:
@@ -1266,9 +1265,15 @@ struct drgn_error *drgn_dwarf_type_find(enum drgn_type_kind kind,
 		UNREACHABLE();
 	}
 
+	struct drgn_dwarf_index_iterator it;
 	drgn_dwarf_index_iterator_init(&it, &dicache->dindex, name, name_len,
 				       &tag, 1);
-	while (!(err = drgn_dwarf_index_iterator_next(&it, &die, NULL))) {
+	struct drgn_dwarf_index_die *index_die;
+	while ((index_die = drgn_dwarf_index_iterator_next(&it))) {
+		Dwarf_Die die;
+		err = drgn_dwarf_index_get_die(index_die, &die, NULL);
+		if (err)
+			return err;
 		if (die_matches_filename(&die, filename)) {
 			err = drgn_type_from_dwarf(dicache, &die, ret);
 			if (err)
@@ -1281,8 +1286,6 @@ struct drgn_error *drgn_dwarf_type_find(enum drgn_type_kind kind,
 				return NULL;
 		}
 	}
-	if (err && err->code != DRGN_ERROR_STOP)
-		return err;
 	return &drgn_not_found;
 }
 
@@ -1436,13 +1439,9 @@ drgn_dwarf_object_find(const char *name, size_t name_len, const char *filename,
 {
 	struct drgn_error *err;
 	struct drgn_dwarf_info_cache *dicache = arg;
-	uint64_t tags[3];
-	size_t num_tags;
-	struct drgn_dwarf_index_iterator it;
-	Dwarf_Die die;
-	uint64_t bias;
 
-	num_tags = 0;
+	uint64_t tags[3];
+	size_t num_tags = 0;
 	if (flags & DRGN_FIND_OBJECT_CONSTANT)
 		tags[num_tags++] = DW_TAG_enumerator;
 	if (flags & DRGN_FIND_OBJECT_FUNCTION)
@@ -1450,9 +1449,16 @@ drgn_dwarf_object_find(const char *name, size_t name_len, const char *filename,
 	if (flags & DRGN_FIND_OBJECT_VARIABLE)
 		tags[num_tags++] = DW_TAG_variable;
 
+	struct drgn_dwarf_index_iterator it;
 	drgn_dwarf_index_iterator_init(&it, &dicache->dindex, name,
 				       strlen(name), tags, num_tags);
-	while (!(err = drgn_dwarf_index_iterator_next(&it, &die, &bias))) {
+	struct drgn_dwarf_index_die *index_die;
+	while ((index_die = drgn_dwarf_index_iterator_next(&it))) {
+		Dwarf_Die die;
+		uint64_t bias;
+		err = drgn_dwarf_index_get_die(index_die, &die, &bias);
+		if (err)
+			return err;
 		if (!die_matches_filename(&die, filename))
 			continue;
 		switch (dwarf_tag(&die)) {
@@ -1470,8 +1476,6 @@ drgn_dwarf_object_find(const char *name, size_t name_len, const char *filename,
 			UNREACHABLE();
 		}
 	}
-	if (err && err->code != DRGN_ERROR_STOP)
-		return err;
 	return &drgn_not_found;
 }
 

libdrgn/program.c

@@ -631,29 +631,26 @@ static void drgn_program_set_language_from_main(struct drgn_program *prog,
 						struct drgn_dwarf_index *dindex)
 {
 	struct drgn_error *err;
+
 	struct drgn_dwarf_index_iterator it;
 	static const uint64_t tags[] = { DW_TAG_subprogram };
-
 	drgn_dwarf_index_iterator_init(&it, dindex, "main", strlen("main"),
 				       tags, ARRAY_SIZE(tags));
-	for (;;) {
+	struct drgn_dwarf_index_die *index_die;
+	while ((index_die = drgn_dwarf_index_iterator_next(&it))) {
 		Dwarf_Die die;
-		const struct drgn_language *lang;
+		err = drgn_dwarf_index_get_die(index_die, &die, NULL);
-
-		err = drgn_dwarf_index_iterator_next(&it, &die, NULL);
-		if (err == &drgn_stop) {
-			break;
-		} else if (err) {
-			drgn_error_destroy(err);
-			continue;
-		}
-
-		err = drgn_language_from_die(&die, &lang);
 		if (err) {
 			drgn_error_destroy(err);
 			continue;
 		}
 
+		const struct drgn_language *lang;
+		err = drgn_language_from_die(&die, &lang);
+		if (err) {
+			drgn_error_destroy(err);
+			continue;
+		}
 		if (lang) {
 			prog->lang = lang;
 			break;