X-Git-Url: https://git.saurik.com/redis.git/blobdiff_plain/794e6ce14ff85f00b0d3e86365e9cc45a90d3bcf..aa2bf6ba8bec776c02d8055ac856d96926137895:/src/db.c diff --git a/src/db.c b/src/db.c index 3135795d..a73d1fbf 100644 --- a/src/db.c +++ b/src/db.c @@ -1,3 +1,32 @@ +/* + * Copyright (c) 2009-2012, Salvatore Sanfilippo + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are met: + * + * * Redistributions of source code must retain the above copyright notice, + * this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * * Neither the name of Redis nor the names of its contributors may be used + * to endorse or promote products derived from this software without + * specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE + * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS + * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN + * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) + * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE + * POSSIBILITY OF SUCH DAMAGE. + */ + #include "redis.h" #include @@ -10,28 +39,6 @@ void SlotToKeyDel(robj *key); * C-level DB API *----------------------------------------------------------------------------*/ -/* Important notes on lookup and disk store. - * - * When disk store is enabled on lookup we can have different cases. - * - * a) The key is in memory: - * - If the key is not in IO_SAVEINPROG state we can access it. - * As if it's just IO_SAVE this means we have the key in the IO queue - * but can't be accessed by the IO thread (it requires to be - * translated into an IO Job by the cache cron function.) - * - If the key is in IO_SAVEINPROG we can't touch the key and have - * to blocking wait completion of operations. - * b) The key is not in memory: - * - If it's marked as non existing on disk as well (negative cache) - * we don't need to perform the disk access. - * - if the key MAY EXIST, but is not in memory, and it is marked as IO_SAVE - * then the key can only be a deleted one. As IO_SAVE keys are never - * evicted (dirty state), so the only possibility is that key was deleted. - * - if the key MAY EXIST we need to blocking load it. - * We check that the key is not in IO_SAVEINPROG state before accessing - * the disk object. If it is in this state, we wait. - */ - robj *lookupKey(redisDb *db, robj *key) { dictEntry *de = dictFind(db->dict,key->ptr); if (de) { @@ -40,19 +47,24 @@ robj *lookupKey(redisDb *db, robj *key) { /* Update the access time for the aging algorithm. * Don't do it if we have a saving child, as this will trigger * a copy on write madness. */ - if (server.bgsavechildpid == -1 && server.bgrewritechildpid == -1) + if (server.rdb_child_pid == -1 && server.aof_child_pid == -1) val->lru = server.lruclock; - server.stat_keyspace_hits++; return val; } else { - server.stat_keyspace_misses++; return NULL; } } robj *lookupKeyRead(redisDb *db, robj *key) { + robj *val; + expireIfNeeded(db,key); - return lookupKey(db,key); + val = lookupKey(db,key); + if (val == NULL) + server.stat_keyspace_misses++; + else + server.stat_keyspace_hits++; + return val; } robj *lookupKeyWrite(redisDb *db, robj *key) { @@ -110,7 +122,7 @@ void setKey(redisDb *db, robj *key, robj *val) { } incrRefCount(val); removeExpire(db,key); - touchWatchedKey(db,key); + signalModifiedKey(db,key); } int dbExists(redisDb *db, robj *key) { @@ -156,8 +168,6 @@ int dbDelete(redisDb *db, robj *key) { } } -/* Empty the whole database. - * If diskstore is enabled this function will just flush the in-memory cache. */ long long emptyDb() { int j; long long removed = 0; @@ -210,15 +220,15 @@ void flushallCommand(redisClient *c) { signalFlushedDb(-1); server.dirty += emptyDb(); addReply(c,shared.ok); - if (server.bgsavechildpid != -1) { - kill(server.bgsavechildpid,SIGKILL); - rdbRemoveTempFile(server.bgsavechildpid); + if (server.rdb_child_pid != -1) { + kill(server.rdb_child_pid,SIGKILL); + rdbRemoveTempFile(server.rdb_child_pid); } if (server.saveparamslen > 0) { /* Normally rdbSave() will reset dirty, but we don't want this here * as otherwise FLUSHALL will not be replicated nor put into the AOF. */ int saved_dirty = server.dirty; - rdbSave(server.dbfilename); + rdbSave(server.rdb_filename); server.dirty = saved_dirty; } server.dirty++; @@ -247,7 +257,11 @@ void existsCommand(redisClient *c) { } void selectCommand(redisClient *c) { - int id = atoi(c->argv[1]->ptr); + long id; + + if (getLongFromObjectOrReply(c, c->argv[1], &id, + "invalid DB index") != REDIS_OK) + return; if (server.cluster_enabled && id != 0) { addReplyError(c,"SELECT is not allowed in cluster mode"); @@ -280,7 +294,7 @@ void keysCommand(redisClient *c) { unsigned long numkeys = 0; void *replylen = addDeferredMultiBulkLength(c); - di = dictGetIterator(c->db->dict); + di = dictGetSafeIterator(c->db->dict); allkeys = (pattern[0] == '*' && pattern[1] == '\0'); while((de = dictNext(di)) != NULL) { sds key = dictGetKey(de); @@ -483,11 +497,12 @@ long long getExpire(redisDb *db, robj *key) { void propagateExpire(redisDb *db, robj *key) { robj *argv[2]; - argv[0] = createStringObject("DEL",3); + argv[0] = shared.del; argv[1] = key; - incrRefCount(key); + incrRefCount(argv[0]); + incrRefCount(argv[1]); - if (server.appendonly) + if (server.aof_state != REDIS_AOF_OFF) feedAppendOnlyFile(server.delCommand,db->id,argv,2); if (listLength(server.slaves)) replicationFeedSlaves(server.slaves,db->id,argv,2); @@ -512,7 +527,7 @@ int expireIfNeeded(redisDb *db, robj *key) { * that is, 0 if we think the key should be still valid, 1 if * we think the key is expired at this time. */ if (server.masterhost != NULL) { - return time(NULL) > when; + return mstime() > when; } /* Return when this key has not expired */ @@ -528,24 +543,23 @@ int expireIfNeeded(redisDb *db, robj *key) { * Expires Commands *----------------------------------------------------------------------------*/ -/* Given an string object return true if it contains exactly the "ms" - * or "MS" string. This is used in order to check if the last argument - * of EXPIRE, EXPIREAT or TTL is "ms" to switch into millisecond input/output */ -int stringObjectEqualsMs(robj *a) { - char *arg = a->ptr; - return tolower(arg[0]) == 'm' && tolower(arg[1]) == 's' && arg[2] == '\0'; -} - -void expireGenericCommand(redisClient *c, long long offset, int unit) { +/* This is the generic command implementation for EXPIRE, PEXPIRE, EXPIREAT + * and PEXPIREAT. Because the commad second argument may be relative or absolute + * the "basetime" argument is used to signal what the base time is (either 0 + * for *AT variants of the command, or the current time for relative expires). + * + * unit is either UNIT_SECONDS or UNIT_MILLISECONDS, and is only used for + * the argv[2] parameter. The basetime is always specified in milliesconds. */ +void expireGenericCommand(redisClient *c, long long basetime, int unit) { dictEntry *de; robj *key = c->argv[1], *param = c->argv[2]; - long long milliseconds; + long long when; /* unix time in milliseconds when the key will expire. */ - if (getLongLongFromObjectOrReply(c, param, &milliseconds, NULL) != REDIS_OK) + if (getLongLongFromObjectOrReply(c, param, &when, NULL) != REDIS_OK) return; - if (unit == UNIT_SECONDS) milliseconds *= 1000; - milliseconds -= offset; + if (unit == UNIT_SECONDS) when *= 1000; + when += basetime; de = dictFind(c->db->dict,key->ptr); if (de == NULL) { @@ -558,7 +572,7 @@ void expireGenericCommand(redisClient *c, long long offset, int unit) { * * Instead we take the other branch of the IF statement setting an expire * (possibly in the past) and wait for an explicit DEL from the master. */ - if (milliseconds <= 0 && !server.loading && !server.masterhost) { + if (when <= mstime() && !server.loading && !server.masterhost) { robj *aux; redisAssertWithInfo(c,key,dbDelete(c->db,key)); @@ -572,7 +586,6 @@ void expireGenericCommand(redisClient *c, long long offset, int unit) { addReply(c, shared.cone); return; } else { - long long when = mstime()+milliseconds; setExpire(c->db,key,when); addReply(c,shared.cone); signalModifiedKey(c->db,key); @@ -582,25 +595,32 @@ void expireGenericCommand(redisClient *c, long long offset, int unit) { } void expireCommand(redisClient *c) { - expireGenericCommand(c,0,UNIT_SECONDS); + expireGenericCommand(c,mstime(),UNIT_SECONDS); } void expireatCommand(redisClient *c) { - expireGenericCommand(c,mstime(),UNIT_SECONDS); + expireGenericCommand(c,0,UNIT_SECONDS); } void pexpireCommand(redisClient *c) { - expireGenericCommand(c,0,UNIT_MILLISECONDS); + expireGenericCommand(c,mstime(),UNIT_MILLISECONDS); } void pexpireatCommand(redisClient *c) { - expireGenericCommand(c,mstime(),UNIT_MILLISECONDS); + expireGenericCommand(c,0,UNIT_MILLISECONDS); } void ttlGenericCommand(redisClient *c, int output_ms) { long long expire, ttl = -1; expire = getExpire(c->db,c->argv[1]); + /* If the key does not exist at all, return -2 */ + if (expire == -1 && lookupKeyRead(c->db,c->argv[1]) == NULL) { + addReplyLongLong(c,-2); + return; + } + /* The key exists. Return -1 if it has no expire, or the actual + * TTL value otherwise. */ if (expire != -1) { ttl = expire-mstime(); if (ttl < 0) ttl = -1;