ComputerSurge/bind9
3
0
Fork 0
Code Pull Requests 1 Activity

reacquire the node lock the caller held before returning from no_references()

Browse Source
This commit is contained in:
David Lawrence
2001-04-18 08:06:11 +00:00
parent ebea68ec9c
commit f6f8d61a77
1 changed files with 29 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files

48
lib/dns/rbtdb.c
View File

@@ -15,7 +15,7 @@
* WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
*/
/* $Id: rbtdb.c,v 1.156 2001/04/12 22:58:34 tale Exp $ */
/* $Id: rbtdb.c,v 1.157 2001/04/18 08:06:11 tale Exp $ */
/*
* Principal Author: Bob Halley
@@ -129,6 +129,7 @@ typedef struct rdatasetheader {
typedef struct {
isc_mutex_t lock;
/* Locked by lock. */
unsigned int references;
isc_boolean_t exiting;
} rbtdb_nodelock_t;
@@ -742,6 +743,7 @@ no_references(dns_rbtdb_t *rbtdb, dns_rbtnode_t *node,
{
isc_result_t result;
isc_boolean_t write_locked;
unsigned int locknum;
/*
* Caller must be holding the node lock.
@@ -766,9 +768,10 @@ no_references(dns_rbtdb_t *rbtdb, dns_rbtnode_t *node,
}
}
INSIST(rbtdb->node_locks[node->locknum].references > 0);
rbtdb->node_locks[node->locknum].references--;
UNLOCK(&rbtdb->node_locks[node->locknum].lock);
locknum = node->locknum;
INSIST(rbtdb->node_locks[locknum].references > 0);
rbtdb->node_locks[locknum].references--;
/*
* XXXDCL should this only be done for cache zones?
@@ -780,6 +783,12 @@ no_references(dns_rbtdb_t *rbtdb, dns_rbtnode_t *node,
* XXXDCL need to add a deferred delete method for ISC_R_LOCKBUSY.
*/
if (lock != isc_rwlocktype_write) {
/*
* Free the node lock before acquiring a tree write lock, per
* the lock hierarchy rules stated at the start of this file.
*/
UNLOCK(&rbtdb->node_locks[locknum].lock);
if (lock == isc_rwlocktype_read)
RWUNLOCK(&rbtdb->tree_lock, isc_rwlocktype_read);
@@ -822,6 +831,12 @@ no_references(dns_rbtdb_t *rbtdb, dns_rbtnode_t *node,
if (lock == isc_rwlocktype_read)
RWLOCK(&rbtdb->tree_lock, isc_rwlocktype_read);
/*
* Reacquire the node lock that the caller held.
*/
LOCK(&rbtdb->node_locks[locknum].lock);
}
static inline void
@@ -1006,9 +1021,8 @@ closeversion(dns_db_t *db, dns_dbversion_t **versionp, isc_boolean_t commit) {
if (rbtnode->references == 0)
no_references(rbtdb, rbtnode, least_serial,
isc_rwlocktype_none);
else
UNLOCK(&rbtdb->
node_locks[rbtnode->locknum].lock);
UNLOCK(&rbtdb->node_locks[rbtnode->locknum].lock);
isc_mem_put(rbtdb->common.mctx, changed,
sizeof *changed);
@@ -2095,9 +2109,8 @@ zone_find(dns_db_t *db, dns_name_t *name, dns_dbversion_t *version,
if (node->references == 0)
no_references(search.rbtdb, node, 0,
isc_rwlocktype_none);
else
UNLOCK(&(search.rbtdb->
node_locks[node->locknum].lock));
UNLOCK(&(search.rbtdb->node_locks[node->locknum].lock));
}
if (close_version)
@@ -2631,9 +2644,7 @@ cache_find(dns_db_t *db, dns_name_t *name, dns_dbversion_t *version,
if (node->references == 0)
no_references(search.rbtdb, node, 0,
isc_rwlocktype_none);
else
UNLOCK(&(search.rbtdb->
node_locks[node->locknum].lock));
UNLOCK(&(search.rbtdb->node_locks[node->locknum].lock));
}
dns_rbtnodechain_reset(&search.chain);
@@ -2816,8 +2827,9 @@ detachnode(dns_db_t *db, dns_dbnode_t **targetp) {
if (rbtdb->node_locks[node->locknum].references == 0 &&
rbtdb->node_locks[node->locknum].exiting)
maybe_free = ISC_TRUE;
} else
UNLOCK(&rbtdb->node_locks[node->locknum].lock);
}
UNLOCK(&rbtdb->node_locks[node->locknum].lock);
*targetp = NULL;
@@ -4604,8 +4616,7 @@ dereference_iter_node(rbtdb_dbiterator_t *rbtdbiter) {
INSIST(rbtdbiter->node->references > 0);
if (--node->references == 0)
no_references(rbtdb, node, 0, rbtdbiter->tree_locked);
else
UNLOCK(&rbtdb->node_locks[node->locknum].lock);
UNLOCK(&rbtdb->node_locks[node->locknum].lock);
rbtdbiter->node = NULL;
}
@@ -4646,8 +4657,7 @@ flush_deletions(rbtdb_dbiterator_t *rbtdbiter) {
if (node->references == 0)
no_references(rbtdb, node, 0,
rbtdbiter->tree_locked);
else
UNLOCK(&rbtdb->node_locks[node->locknum].lock);
UNLOCK(&rbtdb->node_locks[node->locknum].lock);
}
rbtdbiter->delete = 0;