Update Fork

README.md

@@ -435,7 +435,7 @@ top comment inside `server.c`.
 After the command operates in some way, it returns a reply to the client,
 usually using `addReply()` or a similar function defined inside `networking.c`.
 
-There are tons of commands implementations inside th Redis source code
+There are tons of commands implementations inside the Redis source code
 that can serve as examples of actual commands implementations. To write
 a few toy commands can be a good exercise to familiarize with the code base.
 

deps/jemalloc/include/jemalloc/internal/jemalloc_internal_inlines_c.h

@@ -215,4 +215,32 @@ ixalloc(tsdn_t *tsdn, void *ptr, size_t oldsize, size_t size, size_t extra,
 	return arena_ralloc_no_move(tsdn, ptr, oldsize, size, extra, zero);
 }
 
+JEMALLOC_ALWAYS_INLINE int
+iget_defrag_hint(tsdn_t *tsdn, void* ptr, int *bin_util, int *run_util) {
+	int defrag = 0;
+	rtree_ctx_t rtree_ctx_fallback;
+	rtree_ctx_t *rtree_ctx = tsdn_rtree_ctx(tsdn, &rtree_ctx_fallback);
+	szind_t szind;
+	bool is_slab;
+	rtree_szind_slab_read(tsdn, &extents_rtree, rtree_ctx, (uintptr_t)ptr, true, &szind, &is_slab);
+	if (likely(is_slab)) {
+		/* Small allocation. */
+		extent_t *slab = iealloc(tsdn, ptr);
+		arena_t *arena = extent_arena_get(slab);
+		szind_t binind = extent_szind_get(slab);
+		bin_t *bin = &arena->bins[binind];
+		malloc_mutex_lock(tsdn, &bin->lock);
+		/* don't bother moving allocations from the slab currently used for new allocations */
+		if (slab != bin->slabcur) {
+			const bin_info_t *bin_info = &bin_infos[binind];
+			size_t availregs = bin_info->nregs * bin->stats.curslabs;
+			*bin_util = (bin->stats.curregs<<16) / availregs;
+			*run_util = ((bin_info->nregs - extent_nfree_get(slab))<<16) / bin_info->nregs;
+			defrag = 1;
+		}
+		malloc_mutex_unlock(tsdn, &bin->lock);
+	}
+	return defrag;
+}
+
 #endif /* JEMALLOC_INTERNAL_INLINES_C_H */

deps/jemalloc/include/jemalloc/jemalloc_macros.h.in

@@ -120,3 +120,7 @@
 #  define JEMALLOC_RESTRICT_RETURN
 #  define JEMALLOC_ALLOCATOR
 #endif
+
+/* This version of Jemalloc, modified for Redis, has the je_get_defrag_hint()
+ * function. */
+#define JEMALLOC_FRAG_HINT

deps/jemalloc/src/jemalloc.c

@@ -3324,3 +3324,14 @@ jemalloc_postfork_child(void) {
 }
 
 /******************************************************************************/
+
+/* Helps the application decide if a pointer is worth re-allocating in order to reduce fragmentation.
+ * returns 0 if the allocation is in the currently active run,
+ * or when it is not causing any frag issue (large or huge bin)
+ * returns the bin utilization and run utilization both in fixed point 16:16.
+ * If the application decides to re-allocate it should use MALLOCX_TCACHE_NONE when doing so. */
+JEMALLOC_EXPORT int JEMALLOC_NOTHROW
+get_defrag_hint(void* ptr, int *bin_util, int *run_util) {
+	assert(ptr != NULL);
+	return iget_defrag_hint(TSDN_NULL, ptr, bin_util, run_util);
+}

redis.conf

@@ -1211,6 +1211,12 @@ hz 10
 # big latency spikes.
 aof-rewrite-incremental-fsync yes
 
+# When redis saves RDB file, if the following option is enabled
+# the file will be fsync-ed every 32 MB of data generated. This is useful
+# in order to commit the file to the disk more incrementally and avoid
+# big latency spikes.
+rdb-save-incremental-fsync yes
+
 # Redis LFU eviction (see maxmemory setting) can be tuned. However it is a good
 # idea to start with the default settings and only change them after investigating
 # how to improve the performances and how the keys LFU change over time, which

sentinel.conf

@@ -194,3 +194,31 @@ sentinel failover-timeout mymaster 180000
 #
 # sentinel client-reconfig-script mymaster /var/redis/reconfig.sh
 
+# SECURITY
+#
+# By default SENTINEL SET will not be able to change the notification-script
+# and client-reconfig-script at runtime. This avoids a trivial security issue
+# where clients can set the script to anything and trigger a failover in order
+# to get the program executed.
+
+sentinel deny-scripts-reconfig yes
+
+# REDIS COMMANDS RENAMING
+#
+# Sometimes the Redis server has certain commands, that are needed for Sentinel
+# to work correctly, renamed to unguessable strings. This is often the case
+# of CONFIG and SLAVEOF in the context of providers that provide Redis as
+# a service, and don't want the customers to reconfigure the instances outside
+# of the administration console.
+#
+# In such case it is possible to tell Sentinel to use different command names
+# instead of the normal ones. For example if the master "mymaster", and the
+# associated slaves, have "CONFIG" all renamed to "GUESSME", I could use:
+#
+# sentinel rename-command mymaster CONFIG GUESSME
+#
+# After such configuration is set, every time Sentinel would use CONFIG it will
+# use GUESSME instead. Note that there is no actual need to respect the command
+# case, so writing "config guessme" is the same in the example above.
+#
+# SENTINEL SET can also be used in order to perform this configuration at runtime.

src/aof.c

@@ -228,7 +228,7 @@ static void killAppendOnlyChild(void) {
 void stopAppendOnly(void) {
     serverAssert(server.aof_state != AOF_OFF);
     flushAppendOnlyFile(1);
-    aof_fsync(server.aof_fd);
+    redis_fsync(server.aof_fd);
     close(server.aof_fd);
 
     server.aof_fd = -1;
@@ -476,10 +476,10 @@ void flushAppendOnlyFile(int force) {
 
     /* Perform the fsync if needed. */
     if (server.aof_fsync == AOF_FSYNC_ALWAYS) {
-        /* aof_fsync is defined as fdatasync() for Linux in order to avoid
+        /* redis_fsync is defined as fdatasync() for Linux in order to avoid
          * flushing metadata. */
         latencyStartMonitor(latency);
-        aof_fsync(server.aof_fd); /* Let's try to get this data on the disk */
+        redis_fsync(server.aof_fd); /* Let's try to get this data on the disk */
         latencyEndMonitor(latency);
         latencyAddSampleIfNeeded("aof-fsync-always",latency);
         server.aof_last_fsync = server.unixtime;
@@ -1221,7 +1221,6 @@ int rewriteAppendOnlyFileRio(rio *aof) {
     dictIterator *di = NULL;
     dictEntry *de;
     size_t processed = 0;
-    long long now = mstime();
     int j;
 
     for (j = 0; j < server.dbnum; j++) {
@@ -1247,9 +1246,6 @@ int rewriteAppendOnlyFileRio(rio *aof) {
 
             expiretime = getExpire(db,&key);
 
-            /* If this key is already expired skip it */
-            if (expiretime != -1 && expiretime < now) continue;
-
             /* Save the key and associated value */
             if (o->type == OBJ_STRING) {
                 /* Emit a SET command */
@@ -1322,7 +1318,7 @@ int rewriteAppendOnlyFile(char *filename) {
     rioInitWithFile(&aof,fp);
 
     if (server.aof_rewrite_incremental_fsync)
-        rioSetAutoSync(&aof,AOF_AUTOSYNC_BYTES);
+        rioSetAutoSync(&aof,REDIS_AUTOSYNC_BYTES);
 
     if (server.aof_use_rdb_preamble) {
         int error;
@@ -1690,7 +1686,7 @@ void backgroundRewriteDoneHandler(int exitcode, int bysignal) {
             oldfd = server.aof_fd;
             server.aof_fd = newfd;
             if (server.aof_fsync == AOF_FSYNC_ALWAYS)
-                aof_fsync(newfd);
+                redis_fsync(newfd);
             else if (server.aof_fsync == AOF_FSYNC_EVERYSEC)
                 aof_background_fsync(newfd);
             server.aof_selected_db = -1; /* Make sure SELECT is re-issued */

src/bio.c

@@ -187,7 +187,7 @@ void *bioProcessBackgroundJobs(void *arg) {
         if (type == BIO_CLOSE_FILE) {
             close((long)job->arg1);
         } else if (type == BIO_AOF_FSYNC) {
-            aof_fsync((long)job->arg1);
+            redis_fsync((long)job->arg1);
         } else if (type == BIO_LAZY_FREE) {
             /* What we free changes depending on what arguments are set:
              * arg1 -> free the object at pointer.

src/cluster.c

@@ -4835,15 +4835,39 @@ void dumpCommand(client *c) {
 
 /* RESTORE key ttl serialized-value [REPLACE] */
 void restoreCommand(client *c) {
-    long long ttl;
+    long long ttl, lfu_freq = -1, lru_idle = -1, lru_clock = -1;
     rio payload;
-    int j, type, replace = 0;
+    int j, type, replace = 0, absttl = 0;
     robj *obj;
 
     /* Parse additional options */
     for (j = 4; j < c->argc; j++) {
+        int additional = c->argc-j-1;
         if (!strcasecmp(c->argv[j]->ptr,"replace")) {
             replace = 1;
+        } else if (!strcasecmp(c->argv[j]->ptr,"absttl")) {
+            absttl = 1;
+        } else if (!strcasecmp(c->argv[j]->ptr,"idletime") && additional >= 1 &&
+                   lfu_freq == -1)
+        {
+            if (getLongLongFromObjectOrReply(c,c->argv[j+1],&lru_idle,NULL)
+                    != C_OK) return;
+            if (lru_idle < 0) {
+                addReplyError(c,"Invalid IDLETIME value, must be >= 0");
+                return;
+            }
+            lru_clock = LRU_CLOCK();
+            j++; /* Consume additional arg. */
+        } else if (!strcasecmp(c->argv[j]->ptr,"freq") && additional >= 1 &&
+                   lru_idle == -1)
+        {
+            if (getLongLongFromObjectOrReply(c,c->argv[j+1],&lfu_freq,NULL)
+                    != C_OK) return;
+            if (lfu_freq < 0 || lfu_freq > 255) {
+                addReplyError(c,"Invalid FREQ value, must be >= 0 and <= 255");
+                return;
+            }
+            j++; /* Consume additional arg. */
         } else {
             addReply(c,shared.syntaxerr);
             return;
@@ -4884,7 +4908,11 @@ void restoreCommand(client *c) {
 
     /* Create the key and set the TTL if any */
     dbAdd(c->db,c->argv[1],obj);
-    if (ttl) setExpire(c,c->db,c->argv[1],mstime()+ttl);
+    if (ttl) {
+        if (!absttl) ttl+=mstime();
+        setExpire(c,c->db,c->argv[1],ttl);
+    }
+    objectSetLRUOrLFU(obj,lfu_freq,lru_idle,lru_clock);
     signalModifiedKey(c->db,c->argv[1]);
     addReply(c,shared.ok);
     server.dirty++;
@@ -5589,7 +5617,11 @@ void clusterRedirectClient(client *c, clusterNode *n, int hashslot, int error_co
  * longer handles, the client is sent a redirection error, and the function
  * returns 1. Otherwise 0 is returned and no operation is performed. */
 int clusterRedirectBlockedClientIfNeeded(client *c) {
-    if (c->flags & CLIENT_BLOCKED && c->btype == BLOCKED_LIST) {
+    if (c->flags & CLIENT_BLOCKED &&
+        (c->btype == BLOCKED_LIST ||
+         c->btype == BLOCKED_ZSET ||
+         c->btype == BLOCKED_STREAM))
+    {
         dictEntry *de;
         dictIterator *di;