Redis 2.9.56 (Redis 3.0.0 beta-7).

Test: find_available_port: check that cluster port is free as well.
The function will only return ports that have also port+10000 free, so that Redis Cluster instances can be executed at the returned port.
2014-06-30 12:11:22 +02:00 · 2014-06-30 12:11:22 +02:00 · 2014-06-30 12:11:22 +02:00 · 2014-06-27 18:39:33 +02:00 · 2014-06-27 18:39:33 +02:00 · 2014-06-27 18:39:33 +02:00
274 changed files with 28168 additions and 6957 deletions
@@ -1,83 +1,289 @@
-Redis 2.6 release notes
+Redis 3.0 release notes
+=======================

-Migrating from 2.4 to 2.6
+WARNING: Redis 3.0 is currently a BETA not suitable for production environments.
+
+--------------------------------------------------------------------------------
+Upgrade urgency levels:
+
+LOW:      No need to upgrade unless there are new features you want to use.
+MODERATE: Program an upgrade of the server, but it's not urgent.
+HIGH:     There is a critical bug that may affect a subset of users. Upgrade!
+CRITICAL: There is a critical bug affecting MOST USERS. Upgrade ASAP.
+--------------------------------------------------------------------------------
+
+--[ Redis 3.0.0 Beta 7 (version 2.9.56) ] Release date: 30 jun 2014
+
+This is the 7th beta of Redis 3.0.0.
+
+>> General changes
+
+* [FIX] Scripting fixes backported from unstable, see Redis 2.8.12 changelog
+        for more info. (Salvatore Sanfilippo)
+* [FIX] Cancel SHUTDOWN if initial AOF is being written. (Matt Stancliff)
+
+* [NEW] New command: COMMAND, for commands introspection (Matt Stancliff &
+        Salvatore Sanfilippo)
+* [NEW] hiredis: Update to latest version. (Matt Stancliff)
+* [NEW] Jemalloc updated to 3.6.0. (Salvatore Sanfilippo)
+
+>> Cluster changes
+
+* [FIX] Cluster: clear NOADDR flag when updating node address.
+        (Salvatore Sanfilippo)
+
+* [NEW] New CLUSTER SLOTS command to simplify Cluster clients operations.
+        (Matt Stancliff)
+* [NEW] More Cluster tests. (Salvatore Sanfilippo)
+* [NEW] Log when failover authorization are granted / denied.
+        (Salvatore Sanfilippo)
+
+>> Sentinel changes
+
+* [FIX] A few Sentinel bugs fixed and improvements, see Redis 2.8.12
+        changelog for more info.  (Salvatore Sanfilippo & Matt Stancliff)
+* [NEW] New Sentinel-Client handshake protocol, ROLE command, CLIENT KILL,
+        all backported to 3.0 branch. (Salvatore Sanfilippo)
+
+--[ Redis 3.0.0 Beta 6 (version 2.9.55) ] Release date: 9 jun 2014
+
+This is the 6th beta of Redis 3.0.0.
+
+>> General changes
+
+* [FIX] Fix software watchdog signal handler crash due to re-entering.
+* [FIX] Better Lua number -> string conversion for Lua scripts.
+* [FIX] Serious replication bug when min-slaves-* feature is used in slaves
+        configuration fixed.
+* [FIX] Blocking pop on lists now works when the list is created by commands
+        other than *PUSH* (for example SORT STORE).
+
+>> Cluster changes
+
+* [FIX] CRITICAL: For a bug in the implementation of CLUSTER SET-CONFIG-EPOCH
+        introduced with beta-3 (especially beta-4 where the command
+        is actually used by redis-trib), a configEpoch for a node could
+        jump backward, breaking the eventual consistency property of the
+        slots -> nodes mapping in the cluster.
+
+>> Sentinel changes
+
+* No changes for Sentinel in this release.
+
+--[ Redis 3.0.0 Beta 5 (version 2.9.54) ] Release date: 26 may 2014
+
+This is the 5th beta of Redis 3.0.0. It does not include any real
+worthwhile change (just three days passed since the previous beta), but
+fixes two stupid bugs preventing cluster tests to pass.
+
+--[ Redis 3.0.0 Beta 4 (version 2.9.53) ] Release date: 23 may 2014
+
+This is the fourth beta of Redis 3.0.0.
+
+>> General changes
+
+* [NEW] Scripting engine performances improvements.
+* [NEW] Log format slightly changed to report current node role.
+
+* [FIX] Correct the HyperLogLog stale cache flag to prevent unnecessary
+        computation.
+
+>> Cluster changes
+
+* [NEW] redis-trib: ability to import data from standalone Redis instances.
+* [NEW] redis-trib: "fix" subcommand much better at fixing errors.
+* [NEW] CLUSTER FAILOVER FORCE implemented.
+* [NEW] CLUSTER RESET implemented, it is now possible to completely reset
+        nodes and create a new cluster without restarting anything.
+* [NEW] Slave validity factor (max estimated data age to still failover)
+        is now under the control of the user, and can be configured via
+        redis.conf or CONFIG SET. Option name cluster-slave-validity-factor.
+* [NEW] Cluster test: failure detection and failover initial tests.
+* [NEW] CLUSTER MEET: better error messages when address is invalid.
+* [NEW] Bulk-accept new Cluster nodes in the Cluster bus instead of
+        performing just a single accept() per event fired.
+
+* [FIX] Bypass data_age check for manual failovers.
+* [FIX] Fixed data_age computation / check integer overflow.
+* [FIX] Various fixes to Tcl client.tcl Redis Cluster client used in tests.
+* [FIX] Better handling of stolen slots.
+* [FIX] Don't accept cluster bus connections during startup.
+
+>> Sentinel changes
+
+* [NEW] Generate +config-update-from event when a new config is received.
+* [NEW] Log when a failover will be re-attempted again.
+
+* [FIX] Sentinel: Add "dir /tmp" directive in example sentinel.conf.
+
+--[ Redis 3.0.0 Beta 3 (version 2.9.52) ] Release date: 5 may 2014
+
+This is the third beta of Redis 3.0.0.
+
+>> General changes
+
+* [NEW] New data structure: the HyperLogLog (see 2.8 release notes).
+* [NEW] Lexicograhical range queries in sorted sets (see 2.8 release notes).
+* [NEW] LRU algorithm precision greatly improved.
+
+* [FIX] Redis is now much more responsive to reply with LOADING / BUSY errors.
+
+>> Cluster changes
+
+* [NEW] Cluster testing framework and initial tests.
+* [NEW] Cluster epochs collision resolution (make Redis Cluster more resilient
+        to admin and programming errors).
+* [NEW] Persist / fsync some global state to ensure correct crash-recovery
+        semantics.
+* [NEW] New command introduced: CLUSTER SET-CONFIG-EPOCH, still not used
+        by redis-trib. Will be used to speedup the assignment of unique
+        epochs to different nodes at cluster creation time. For now this is
+        handled as a side effect of the cluster epochs collision resolution.
+
+* [FIX] Different fixes to redis-trib cluster creation.
+* [FIX] Fix an error in the CLUSTER NODES output for nodes slots.
+
+>> Sentinel changes
+
+* [NEW] Sentinels are now able to send update messages in a peer-to-peer
+        fashion even if no Redis instances are available. Now the Sentinel
+        liveness property that the most updated configuration in a given
+        partition is propagated to all the Sentinels is extended to partitions
+        without reachable instances.
+* [NEW] Sentinel safety properties are now ensured in a crash-recovery system
+        model since some state is persisted on disk before replying to other
+        nodes, and reloaded at startup.
+* [NEW] Sentinel now uses CLIENT SETNAME so that it is easy to identify
+        Sentinels using CLIENT LIST among other clients.
+* [NEW] Sentinel failure detection and reconnection code improved.
+
+--[ Redis 3.0.0 Beta 2 (version 2.9.51) ] Release date: 11 mar 2014
+
+This is the second beta of Redis 3.0.0.
+
+>> General changes
+
+* [FIX] Sometimes the absolute config file path was obtained in a wrong way.
+        This happened when there was a "dir" directive inside the config file
+        and at the same time the configuration file was given as a relative
+        path to redis-server or redis-sentinel executables.
+* [FIX] redis-cli: Automatically enter --slave mode when SYNC or PSYNC are
+        called during an interactive session.
+* [FIX] BITCOUNT: fixed unaligned access causing issues in sparc and other
+        archs not capable of dealing with unaligned accesses. This also makes
+        the code faster in archs where unaligned accesses are allowed.
+* [FIX] Force INFO used_memory_peak to match peak memory. This generated some
+        confusion among users even if it was not an actual bug.
+* [FIX] Fixed an critical EVALSHA script cache bug: scripts executed may not
+        propagate to AOF / Slaves correctly under certain conditions.
+        See issue #1549 at Github for more information.
+* [FIX] Fixed multiple bugs resulting into closing the link with master or slave
+        during replication without good reasons. This will result in useless
+        resynchronizations, or infinite loops where the replication link can't
+        be established.
+* [FIX] Don't count the time needed to populate the buffers of clients waiting
+        in MONITOR mode when populating the Slow Log entries.
+
+* [NEW] Redis-cli updated to use SCAN instead of random sampling via
+        RANDOMKEY in order to implement --bigkeys feature. Moreover the
+        implementation now supports pipelining and reports more information
+        at the end of the scan. Much faster, much better. A special thank
+        you to Michael Grunder for this improvement.
+* [NEW] redis-cli now supports a new --intrinsic-latency mode that is able
+        to meter the latency of a system due to kernel / hypervisor.
+        How to use it is explained at http://redis.io/topics/latency.
+* [NEW] New command BITPOS: find first bit set or clear in a bitmap.
+* [NEW] CONFIG REWRITE calls are now logged.
+* [NEW] AOF write errors (like no space on device) no longer abort Redis if the
+        fsync policy is none or every second. The database enters a read-only
+        mode where every write is refused with an error. Normal operations are
+        restored as soon as Redis is able to append again data to the AOF file.
+* [NEW] Sentinel now accepts SHUTDOWN command.
+
+
+>> Cluster changes
+
+* [FIX] Bind the first interface listed in the "bind" configuration directive
+        if any, in order to perform outgoing connections. This fixes Cluster
+        usage when an address is bound but there are multiple interfaces that
+        may be used to connect with other nodes.
+* [FIX] When an "Importing" slot is closed via CLUSTER SETSLOT NODE ...
+        increment the configEpoch in the special case it is zero.
+* [FIX] Current transaction is invalidated on redirection errors.
+* [FIX] Abort if port does not allow for a valid cluster bus port that is
+        always at fixed +10000 offset.
+* [FIX] Keys extraction algorithm fixed for ZUNIONSTORE/ZINTERSTORE and SORT.
+* [FIX] Better failover timeout and retry times: failover should now work
+        reliabily when node-timeout is very small (a few milliseconds).
+* [FIX] Don't allow SORT GET/BY options in Cluster mode.
+* [FIX] Clear importing/migrating state when turning from master to slave role.
+* [FIX] Set slot error if we receive an update for a busy slot.
+* [FIX] Update node configEpoch on UPDATE messages.
+
+* [NEW] Support multi-key operations as long as keys resolve to the same
+        hash slot, and the slot is not migrating, or it is migrating but all
+        the mentioned keys are available.
+* [NEW] New DEBUG command CMDKEYS available to debug / test keys identification
+        in Redis commands.
+* [NEW] redis-trib: create subcommand is now able to assign spare slaves.
+* [NEW] redis-trib: new subcommand 'call'. Exec command in all nodes.
+
+
+>> Sentinel changes
+
+* [FIX] Sentinel "IDONTKNOW" error removed as it does not made sense with the
+        new Sentinel design. This error was actually a fix for a design error
+        in the first implementation of Sentinel.
+* [FIX] Sentinel: added a missing exit() call to abort after config file
+        checks at startup. This error was introduced with an improvement in
+        a previous 2.8 release.
+* [FIX] Sentinel: better nodes fail over start time desynchronization to avoid
+        split-brain during the voting process needed to get authorization to
+        fail over. This means the system is less likely to need to retry
+        and will fail over faster. No changes in behavior / correctness.
+
+* [NEW] Sentinel unit tests and framework. More tests needed and units must
+        be improved in order to have less false positives, but it is a start
+        and features a debugging console that is useful to fix tests or to
+        inspect bugs causing tests failures.
+* [NEW] New Sentinel events: +/-monitor and +set used to monitor when an
+        instance to monitor is added or removed, or when a configuration
+        is modified via SENTINEL SET.
+
+--[ Redis 3.0.0 Beta 1 (version 2.9.50) ] Release date: 11 Feb 2014
+
+This is the first beta of Redis 3.0.0.
+
+The following is a list of improvements in Redis 3.0, compared to Redis 2.8.
+
+* [NEW] Redis Cluster: a distributed implementation of a subset of Redis.
+* [NEW] New "embedded string" object encoding resulting in less cache
+        misses. Big speed gain under certain work loads.
+* [NEW] WAIT command to block waiting for a write to be transmitted to
+        the specified number of slaves.
+* [NEW] MIGRATE connection caching. Much faster keys migraitons.
+* [NEW] MIGARTE new options COPY and REPLACE.
+* [NEW] CLIENT PAUSE command: stop processing client requests for a
+        specified amount of time.
+
+Migrating from 2.8 to 3.0
 =========================

-Redis 2.4 is mostly a strict subset of 2.6. However there are a few things
-that you should be aware of:
+Redis 3.0 is mostly a strict subset of 2.8, you should not have any problem
+upgrading your application from 2.8 to 3.0. However this is a list of small
+non-backward compatible changes introduced in the 3.0 release:

-* You can't use .rdb and AOF files generated with 2.6 into a 2.4 instance.
-* 2.6 slaves can be attached to 2.4 masters, but not the contrary, and only
-  for the time needed to perform the version upgrade.
-
-There are also a few API differences, that are unlikely to cause problems,
-but it is better to keep them in mind:
-
-* SORT now will refuse to sort in numerical mode elements that can't be parsed
-  as numbers.
-* EXPIREs now all have millisecond resolution (but this is very unlikely to
-  break code that was not conceived exploting the previous resolution error
-  in some way.)
-* INFO output is a bit different now, and contains empty lines and comments
-  starting with '#'. All the major clients should be already fixed to work
-  with the new INFO format.
-
-Also the following redis.conf and CONFIG GET / SET parameters changed name:
-
-    * hash-max-zipmap-entries, now replaced by hash-max-ziplist-entries
-    * hash-max-zipmap-value, now replaced by hash-max-ziplist-value
-    * glueoutputbuf was now completely removed as it does not make sense
-
---------
-CHANGELOG
---------
-
-What's new in Redis 2.6.0
-=========================
-
-UPGRADE URGENCY: We suggest new users to start with 2.6.0, and old users to
-                 upgrade after some testing of the application with the new
-                 Redis version.
-
-* Server side Lua scripting, see http://redis.io/commands/eval
-* Virtual Memory removed (was deprecated in 2.4)
-* Hardcoded limits about max number of clients removed.
-* AOF low level semantics is generally more sane, and especially when used
-  in slaves.
-* Milliseconds resolution expires, also added new commands with milliseconds
-  precision (PEXPIRE, PTTL, ...).
-* Clients max output buffer soft and hard limits. You can specifiy different
-  limits for different classes of clients (normal,pubsub,slave).
-* AOF is now able to rewrite aggregate data types using variadic commands,
-  often producing an AOF that is faster to save, load, and is smaller in size.
-* Every redis.conf directive is now accepted as a command line option for the
-  redis-server binary, with the same name and number of arguments.
-* Hash table seed randomization for protection against collisions attacks.
-* Performances improved when writing large objects to Redis.
-* Significant parts of the core refactored or rewritten. New internal APIs
-  and core changes allowed to develop Redis Cluster on top of the new code,
-  however for 2.6 all the cluster code was removed, and will be released with
-  Redis 3.0 when it is more complete and stable.
-* Redis ASCII art logo added at startup.
-* Crash report on memory violation or failed asserts improved significantly
-  to make debugging of hard to catch bugs simpler.
-* redis-benchmark improvements: ability to run selected tests,
-  CSV output, faster, better help.
-* redis-cli improvements: --eval for comfortable development of Lua scripts.
-* SHUTDOWN now supports two optional arguments: "SAVE" and "NOSAVE".
-* INFO output split into sections, the command is now able to just show 
-  pecific sections.
-* New statistics about how many time a command was called, and how much
-  execution time it used (INFO commandstats).
-* More predictable SORT behavior in edge cases.
-* INCRBYFLOAT and HINCRBYFLOAT commands.
+* The log format was modified. The prefix of each line included the pid
+  in the following format [1234]. Now instead it is 1234:? Where
+  '?' is actually the role of the instance. M for master, S for slave, C
+  if this process is a saving child (for RDB/AOF), and X for Sentinel.

 --------------------------------------------------------------------------------

-Credits: Where not specified the implementation and design are done by
-Salvatore Sanfilippo and Pieter Noordhuis. Thanks to VMware for making all
-this possible. Also many thanks to all the other contributors and the amazing
-community we have.
+Credits: Where not specified the implementation and design is done by
+Salvatore Sanfilippo. Thanks to Pivotal for making all this possible.
+Also many thanks to all the other contributors and the amazing community
+we have.

 See commit messages for more credits.

@@ -130,7 +130,7 @@ it the proper way for a production system, we have a script doing this
 for Ubuntu and Debian systems:

    % cd utils
-    % ./install_server
+    % ./install_server.sh

 The script will ask you a few questions and will setup everything you need
 to run Redis properly as a background daemon that will start again on
@@ -1,5 +1,5 @@
 /hiredis-test
-/hiredis-example*
+/examples/hiredis-example*
 /*.o
 /*.so
 /*.dylib
@@ -0,0 +1,6 @@
+language: c
+compiler:
+  - gcc
+  - clang
+
+script: make && make check
@@ -1,3 +1,11 @@
+### 0.11.0
+
+* Increase the maximum multi-bulk reply depth to 7.
+
+* Increase the read buffer size from 2k to 16k.
+
+* Use poll(2) instead of select(2) to support large fds (>= 1024).
+
 ### 0.10.1

 * Makefile overhaul. Important to check out if you override one or more
@@ -4,11 +4,24 @@
 # This file is released under the BSD license, see the COPYING file

 OBJ=net.o hiredis.o sds.o async.o
-BINS=hiredis-example hiredis-test
+EXAMPLES=hiredis-example hiredis-example-libevent hiredis-example-libev
+TESTS=hiredis-test
 LIBNAME=libhiredis

 HIREDIS_MAJOR=0
-HIREDIS_MINOR=10
+HIREDIS_MINOR=11
+
+# redis-server configuration used for testing
+REDIS_PORT=56379
+REDIS_SERVER=redis-server
+define REDIS_TEST_CONFIG
+	daemonize yes
+	pidfile /tmp/hiredis-test-redis.pid
+	port $(REDIS_PORT)
+	bind 127.0.0.1
+	unixsocket /tmp/hiredis-test-redis.sock
+endef
+export REDIS_TEST_CONFIG

 # Fallback to gcc when $CC is not in $PATH.
 CC:=$(shell sh -c 'type $(CC) >/dev/null 2>/dev/null && echo $(CC) || echo gcc')
@@ -41,12 +54,11 @@ ifeq ($(uname_S),Darwin)
  DYLIB_MAKE_CMD=$(CC) -shared -Wl,-install_name,$(DYLIB_MINOR_NAME) -o $(DYLIBNAME) $(LDFLAGS)
 endif

-all: $(DYLIBNAME) $(BINS)
+all: $(DYLIBNAME)

 # Deps (use make dep to generate this)
 net.o: net.c fmacros.h net.h hiredis.h
 async.o: async.c async.h hiredis.h sds.h dict.c dict.h
-example.o: example.c hiredis.h
 hiredis.o: hiredis.c fmacros.h hiredis.h net.h sds.h
 sds.o: sds.c sds.h
 test.o: test.c hiredis.h
@@ -61,36 +73,44 @@ dynamic: $(DYLIBNAME)
 static: $(STLIBNAME)

 # Binaries:
-hiredis-example-libevent: example-libevent.c adapters/libevent.h $(STLIBNAME)
-	$(CC) -o $@ $(REAL_CFLAGS) $(REAL_LDFLAGS) -levent example-libevent.c $(STLIBNAME)
+hiredis-example-libevent: examples/example-libevent.c adapters/libevent.h $(STLIBNAME)
+	$(CC) -o examples/$@ $(REAL_CFLAGS) $(REAL_LDFLAGS) -I. $< -levent $(STLIBNAME)

-hiredis-example-libev: example-libev.c adapters/libev.h $(STLIBNAME)
-	$(CC) -o $@ $(REAL_CFLAGS) $(REAL_LDFLAGS) -lev example-libev.c $(STLIBNAME)
+hiredis-example-libev: examples/example-libev.c adapters/libev.h $(STLIBNAME)
+	$(CC) -o examples/$@ $(REAL_CFLAGS) $(REAL_LDFLAGS) -I. $< -lev $(STLIBNAME)

 ifndef AE_DIR
 hiredis-example-ae:
 	@echo "Please specify AE_DIR (e.g. <redis repository>/src)"
 	@false
 else
-hiredis-example-ae: example-ae.c adapters/ae.h $(STLIBNAME)
-	$(CC) -o $@ $(REAL_CFLAGS) $(REAL_LDFLAGS) -I$(AE_DIR) $(AE_DIR)/ae.o $(AE_DIR)/zmalloc.o example-ae.c $(STLIBNAME)
+hiredis-example-ae: examples/example-ae.c adapters/ae.h $(STLIBNAME)
+	$(CC) -o examples/$@ $(REAL_CFLAGS) $(REAL_LDFLAGS) -I. -I$(AE_DIR) $< $(AE_DIR)/ae.o $(AE_DIR)/zmalloc.o $(AE_DIR)/../deps/jemalloc/lib/libjemalloc.a -pthread $(STLIBNAME)
 endif

-hiredis-%: %.o $(STLIBNAME)
+ifndef LIBUV_DIR
+hiredis-example-libuv:
+	@echo "Please specify LIBUV_DIR (e.g. ../libuv/)"
+	@false
+else
+hiredis-example-libuv: examples/example-libuv.c adapters/libuv.h $(STLIBNAME)
+	$(CC) -o examples/$@ $(REAL_CFLAGS) $(REAL_LDFLAGS) -I. -I$(LIBUV_DIR)/include $< $(LIBUV_DIR)/.libs/libuv.a -lpthread $(STLIBNAME)
+endif
+
+hiredis-example: examples/example.c $(STLIBNAME)
+	$(CC) -o examples/$@ $(REAL_CFLAGS) $(REAL_LDFLAGS) -I. $< $(STLIBNAME)
+
+examples: $(EXAMPLES)
+
+hiredis-test: test.o $(STLIBNAME)
 	$(CC) -o $@ $(REAL_LDFLAGS) $< $(STLIBNAME)

 test: hiredis-test
 	./hiredis-test

 check: hiredis-test
-	echo \
-		"daemonize yes\n" \
-		"pidfile /tmp/hiredis-test-redis.pid\n" \
-		"port 56379\n" \
-		"bind 127.0.0.1\n" \
-		"unixsocket /tmp/hiredis-test-redis.sock" \
-			| redis-server -
-	./hiredis-test -h 127.0.0.1 -p 56379 -s /tmp/hiredis-test-redis.sock || \
+	@echo "$$REDIS_TEST_CONFIG" | $(REDIS_SERVER) -
+	./hiredis-test -h 127.0.0.1 -p $(REDIS_PORT) -s /tmp/hiredis-test-redis.sock || \
 			( kill `cat /tmp/hiredis-test-redis.pid` && false )
 	kill `cat /tmp/hiredis-test-redis.pid`

@@ -98,17 +118,15 @@ check: hiredis-test
 	$(CC) -std=c99 -pedantic -c $(REAL_CFLAGS) $<

 clean:
-	rm -rf $(DYLIBNAME) $(STLIBNAME) $(BINS) hiredis-example* *.o *.gcda *.gcno *.gcov
+	rm -rf $(DYLIBNAME) $(STLIBNAME) $(TESTS) examples/hiredis-example* *.o *.gcda *.gcno *.gcov

 dep:
 	$(CC) -MM *.c

 # Installation related variables and target
 PREFIX?=/usr/local
-INCLUDE_PATH?=include/hiredis
-LIBRARY_PATH?=lib
-INSTALL_INCLUDE_PATH= $(PREFIX)/$(INCLUDE_PATH)
-INSTALL_LIBRARY_PATH= $(PREFIX)/$(LIBRARY_PATH)
+INSTALL_INCLUDE_PATH= $(PREFIX)/include/hiredis
+INSTALL_LIBRARY_PATH= $(PREFIX)/lib

 ifeq ($(uname_S),SunOS)
  INSTALL?= cp -r
@@ -1,3 +1,5 @@
+[![Build Status](https://travis-ci.org/redis/hiredis.png)](https://travis-ci.org/redis/hiredis)
+
 # HIREDIS

 Hiredis is a minimalistic C client library for the [Redis](http://redis.io/) database.
@@ -44,7 +46,7 @@ After trying to connect to Redis using `redisConnect` you should
 check the `err` field to see if establishing the connection was successful:

    redisContext *c = redisConnect("127.0.0.1", 6379);
-    if (c->err) {
+    if (c != NULL && c->err) {
        printf("Error: %s\n", c->errstr);
        // handle error
    }
@@ -66,7 +68,7 @@ When you need to pass binary safe strings in a command, the `%b` specifier can b
 used. Together with a pointer to the string, it requires a `size_t` length argument
 of the string:

-    reply = redisCommand(context, "SET foo %b", value, valuelen);
+    reply = redisCommand(context, "SET foo %b", value, (size_t) valuelen);

 Internally, Hiredis splits the command in different arguments and will
 convert it to the protocol used to communicate with Redis.
@@ -337,6 +339,9 @@ and a reply object (as described above) via `void **reply`. The returned status
 can be either `REDIS_OK` or `REDIS_ERR`, where the latter means something went
 wrong (either a protocol error, or an out of memory error).

+The parser limits the level of nesting for multi bulk payloads to 7. If the
+multi bulk nesting level is higher than this, the parser returns an error.
+
 ### Customizing replies

 The function `redisReaderGetReply` creates `redisReply` and makes the function
@@ -0,0 +1,121 @@
+#ifndef __HIREDIS_LIBUV_H__
+#define __HIREDIS_LIBUV_H__
+#include <uv.h>
+#include "../hiredis.h"
+#include "../async.h"
+#include <string.h>
+
+typedef struct redisLibuvEvents {
+  redisAsyncContext* context;
+  uv_poll_t          handle;
+  int                events;
+} redisLibuvEvents;
+
+int redisLibuvAttach(redisAsyncContext*, uv_loop_t*);
+
+static void redisLibuvPoll(uv_poll_t* handle, int status, int events) {
+  redisLibuvEvents* p = (redisLibuvEvents*)handle->data;
+
+  if (status != 0) {
+    return;
+  }
+
+  if (events & UV_READABLE) {
+    redisAsyncHandleRead(p->context);
+  }
+  if (events & UV_WRITABLE) {
+    redisAsyncHandleWrite(p->context);
+  }
+}
+
+
+static void redisLibuvAddRead(void *privdata) {
+  redisLibuvEvents* p = (redisLibuvEvents*)privdata;
+
+  p->events |= UV_READABLE;
+
+  uv_poll_start(&p->handle, p->events, redisLibuvPoll);
+}
+
+
+static void redisLibuvDelRead(void *privdata) {
+  redisLibuvEvents* p = (redisLibuvEvents*)privdata;
+
+  p->events &= ~UV_READABLE;
+
+  if (p->events) {
+    uv_poll_start(&p->handle, p->events, redisLibuvPoll);
+  } else {
+    uv_poll_stop(&p->handle);
+  }
+}
+
+
+static void redisLibuvAddWrite(void *privdata) {
+  redisLibuvEvents* p = (redisLibuvEvents*)privdata;
+
+  p->events |= UV_WRITABLE;
+
+  uv_poll_start(&p->handle, p->events, redisLibuvPoll);
+}
+
+
+static void redisLibuvDelWrite(void *privdata) {
+  redisLibuvEvents* p = (redisLibuvEvents*)privdata;
+
+  p->events &= ~UV_WRITABLE;
+
+  if (p->events) {
+    uv_poll_start(&p->handle, p->events, redisLibuvPoll);
+  } else {
+    uv_poll_stop(&p->handle);
+  }
+}
+
+
+static void on_close(uv_handle_t* handle) {
+  redisLibuvEvents* p = (redisLibuvEvents*)handle->data;
+
+  free(p);
+}
+
+
+static void redisLibuvCleanup(void *privdata) {
+  redisLibuvEvents* p = (redisLibuvEvents*)privdata;
+
+  uv_close((uv_handle_t*)&p->handle, on_close);
+}
+
+
+static int redisLibuvAttach(redisAsyncContext* ac, uv_loop_t* loop) {
+  redisContext *c = &(ac->c);
+
+  if (ac->ev.data != NULL) {
+    return REDIS_ERR;
+  }
+
+  ac->ev.addRead  = redisLibuvAddRead;
+  ac->ev.delRead  = redisLibuvDelRead;
+  ac->ev.addWrite = redisLibuvAddWrite;
+  ac->ev.delWrite = redisLibuvDelWrite;
+  ac->ev.cleanup  = redisLibuvCleanup;
+
+  redisLibuvEvents* p = (redisLibuvEvents*)malloc(sizeof(*p));
+
+  if (!p) {
+    return REDIS_ERR;
+  }
+
+  memset(p, 0, sizeof(*p));
+
+  if (uv_poll_init(loop, &p->handle, c->fd) != 0) {
+    return REDIS_ERR;
+  }
+
+  ac->ev.data    = p;
+  p->handle.data = p;
+  p->context     = ac;
+
+  return REDIS_OK;
+}
+#endif
@@ -62,7 +62,8 @@ void __redisAppendCommand(redisContext *c, char *cmd, size_t len);

 /* Functions managing dictionary of callbacks for pub/sub. */
 static unsigned int callbackHash(const void *key) {
-    return dictGenHashFunction((unsigned char*)key,sdslen((char*)key));
+    return dictGenHashFunction((const unsigned char *)key,
+                               sdslen((const sds)key));
 }

 static void *callbackValDup(void *privdata, const void *src) {
@@ -76,8 +77,8 @@ static int callbackKeyCompare(void *privdata, const void *key1, const void *key2
    int l1, l2;
    ((void) privdata);

-    l1 = sdslen((sds)key1);
-    l2 = sdslen((sds)key2);
+    l1 = sdslen((const sds)key1);
+    l2 = sdslen((const sds)key2);
    if (l1 != l2) return 0;
    return memcmp(key1,key2,l1) == 0;
 }
@@ -102,7 +103,12 @@ static dictType callbackDict = {
 };

 static redisAsyncContext *redisAsyncInitialize(redisContext *c) {
-    redisAsyncContext *ac = realloc(c,sizeof(redisAsyncContext));
+    redisAsyncContext *ac;
+
+    ac = realloc(c,sizeof(redisAsyncContext));
+    if (ac == NULL)
+        return NULL;
+
    c = &(ac->c);

    /* The regular connect functions will always set the flag REDIS_CONNECTED.
@@ -142,15 +148,45 @@ static void __redisAsyncCopyError(redisAsyncContext *ac) {
 }

 redisAsyncContext *redisAsyncConnect(const char *ip, int port) {
-    redisContext *c = redisConnectNonBlock(ip,port);
+    redisContext *c;
+    redisAsyncContext *ac;
+
+    c = redisConnectNonBlock(ip,port);
+    if (c == NULL)
+        return NULL;
+
+    ac = redisAsyncInitialize(c);
+    if (ac == NULL) {
+        redisFree(c);
+        return NULL;
+    }
+
+    __redisAsyncCopyError(ac);
+    return ac;
+}
+
+redisAsyncContext *redisAsyncConnectBind(const char *ip, int port,
+                                         const char *source_addr) {
+    redisContext *c = redisConnectBindNonBlock(ip,port,source_addr);
    redisAsyncContext *ac = redisAsyncInitialize(c);
    __redisAsyncCopyError(ac);
    return ac;
 }

 redisAsyncContext *redisAsyncConnectUnix(const char *path) {
-    redisContext *c = redisConnectUnixNonBlock(path);
-    redisAsyncContext *ac = redisAsyncInitialize(c);
+    redisContext *c;
+    redisAsyncContext *ac;
+
+    c = redisConnectUnixNonBlock(path);
+    if (c == NULL)
+        return NULL;
+
+    ac = redisAsyncInitialize(c);
+    if (ac == NULL) {
+        redisFree(c);
+        return NULL;
+    }
+
    __redisAsyncCopyError(ac);
    return ac;
 }
@@ -182,6 +218,9 @@ static int __redisPushCallback(redisCallbackList *list, redisCallback *source) {

    /* Copy callback from stack to heap */
    cb = malloc(sizeof(*cb));
+    if (cb == NULL)
+        return REDIS_ERR_OOM;
+
    if (source != NULL) {
        memcpy(cb,source,sizeof(*cb));
        cb->next = NULL;
@@ -360,7 +399,7 @@ static int __redisGetSubscribeCallback(redisAsyncContext *ac, redisReply *reply,

 void redisProcessCallbacks(redisAsyncContext *ac) {
    redisContext *c = &(ac->c);
-    redisCallback cb;
+    redisCallback cb = {NULL, NULL, NULL};
    void *reply = NULL;
    int status;

@@ -372,7 +411,7 @@ void redisProcessCallbacks(redisAsyncContext *ac) {
                __redisAsyncDisconnect(ac);
                return;
            }
-            
+
            /* If monitor mode, repush callback */
            if(c->flags & REDIS_MONITORING) {
                __redisPushCallback(&ac->replies,&cb);
@@ -442,7 +481,7 @@ void redisProcessCallbacks(redisAsyncContext *ac) {
 static int __redisAsyncHandleConnect(redisAsyncContext *ac) {
    redisContext *c = &(ac->c);

-    if (redisCheckSocketError(c,c->fd) == REDIS_ERR) {
+    if (redisCheckSocketError(c) == REDIS_ERR) {
        /* Try again later when connect(2) is still in progress. */
        if (errno == EINPROGRESS)
            return REDIS_OK;
@@ -102,6 +102,7 @@ typedef struct redisAsyncContext {

 /* Functions that proxy to hiredis */
 redisAsyncContext *redisAsyncConnect(const char *ip, int port);
+redisAsyncContext *redisAsyncConnectBind(const char *ip, int port, const char *source_addr);
 redisAsyncContext *redisAsyncConnectUnix(const char *path);
 int redisAsyncSetConnectCallback(redisAsyncContext *ac, redisConnectCallback *fn);
 int redisAsyncSetDisconnectCallback(redisAsyncContext *ac, redisDisconnectCallback *fn);
@@ -2,9 +2,10 @@
 #include <stdlib.h>
 #include <string.h>
 #include <signal.h>
-#include "hiredis.h"
-#include "async.h"
-#include "adapters/ae.h"
+
+#include <hiredis.h>
+#include <async.h>
+#include <adapters/ae.h>

 /* Put event loop in the global scope, so it can be explicitly stopped */
 static aeEventLoop *loop;
@@ -21,17 +22,22 @@ void getCallback(redisAsyncContext *c, void *r, void *privdata) {
 void connectCallback(const redisAsyncContext *c, int status) {
    if (status != REDIS_OK) {
        printf("Error: %s\n", c->errstr);
+        aeStop(loop);
        return;
    }
+
    printf("Connected...\n");
 }

 void disconnectCallback(const redisAsyncContext *c, int status) {
    if (status != REDIS_OK) {
        printf("Error: %s\n", c->errstr);
+        aeStop(loop);
        return;
    }
+
    printf("Disconnected...\n");
+    aeStop(loop);
 }

 int main (int argc, char **argv) {
@@ -44,7 +50,7 @@ int main (int argc, char **argv) {
        return 1;
    }

-    loop = aeCreateEventLoop();
+    loop = aeCreateEventLoop(64);
    redisAeAttach(loop, c);
    redisAsyncSetConnectCallback(c,connectCallback);
    redisAsyncSetDisconnectCallback(c,disconnectCallback);
@@ -2,9 +2,10 @@
 #include <stdlib.h>
 #include <string.h>
 #include <signal.h>
-#include "hiredis.h"
-#include "async.h"
-#include "adapters/libev.h"
+
+#include <hiredis.h>
+#include <async.h>
+#include <adapters/libev.h>

 void getCallback(redisAsyncContext *c, void *r, void *privdata) {
    redisReply *reply = r;
@@ -2,9 +2,10 @@
 #include <stdlib.h>
 #include <string.h>
 #include <signal.h>
-#include "hiredis.h"
-#include "async.h"
-#include "adapters/libevent.h"
+
+#include <hiredis.h>
+#include <async.h>
+#include <adapters/libevent.h>

 void getCallback(redisAsyncContext *c, void *r, void *privdata) {
    redisReply *reply = r;
@@ -0,0 +1,53 @@
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <signal.h>
+
+#include <hiredis.h>
+#include <async.h>
+#include <adapters/libuv.h>
+
+void getCallback(redisAsyncContext *c, void *r, void *privdata) {
+    redisReply *reply = r;
+    if (reply == NULL) return;
+    printf("argv[%s]: %s\n", (char*)privdata, reply->str);
+
+    /* Disconnect after receiving the reply to GET */
+    redisAsyncDisconnect(c);
+}
+
+void connectCallback(const redisAsyncContext *c, int status) {
+    if (status != REDIS_OK) {
+        printf("Error: %s\n", c->errstr);
+        return;
+    }
+    printf("Connected...\n");
+}
+
+void disconnectCallback(const redisAsyncContext *c, int status) {
+    if (status != REDIS_OK) {
+        printf("Error: %s\n", c->errstr);
+        return;
+    }
+    printf("Disconnected...\n");
+}
+
+int main (int argc, char **argv) {
+    signal(SIGPIPE, SIG_IGN);
+    uv_loop_t* loop = uv_default_loop();
+
+    redisAsyncContext *c = redisAsyncConnect("127.0.0.1", 6379);
+    if (c->err) {
+        /* Let *c leak for now... */
+        printf("Error: %s\n", c->errstr);
+        return 1;
+    }
+
+    redisLibuvAttach(c,loop);
+    redisAsyncSetConnectCallback(c,connectCallback);
+    redisAsyncSetDisconnectCallback(c,disconnectCallback);
+    redisAsyncCommand(c, NULL, NULL, "SET key %b", argv[argc-1], strlen(argv[argc-1]));
+    redisAsyncCommand(c, getCallback, (char*)"end-1", "GET key");
+    uv_run(loop, UV_RUN_DEFAULT);
+    return 0;
+}
@@ -2,17 +2,24 @@
 #include <stdlib.h>
 #include <string.h>

-#include "hiredis.h"
+#include <hiredis.h>

-int main(void) {
+int main(int argc, char **argv) {
    unsigned int j;
    redisContext *c;
    redisReply *reply;
+    const char *hostname = (argc > 1) ? argv[1] : "127.0.0.1";
+    int port = (argc > 2) ? atoi(argv[2]) : 6379;

    struct timeval timeout = { 1, 500000 }; // 1.5 seconds
-    c = redisConnectWithTimeout((char*)"127.0.0.2", 6379, timeout);
-    if (c->err) {
-        printf("Connection error: %s\n", c->errstr);
+    c = redisConnectWithTimeout(hostname, port, timeout);
+    if (c == NULL || c->err) {
+        if (c) {
+            printf("Connection error: %s\n", c->errstr);
+            redisFree(c);
+        } else {
+            printf("Connection error: can't allocate redis context\n");
+        }
        exit(1);
    }

@@ -27,7 +34,7 @@ int main(void) {
    freeReplyObject(reply);

    /* Set a key using binary safe API */
-    reply = redisCommand(c,"SET %b %b", "bar", 3, "hello", 5);
+    reply = redisCommand(c,"SET %b %b", "bar", (size_t) 3, "hello", (size_t) 5);
    printf("SET (binary API): %s\n", reply->str);
    freeReplyObject(reply);

@@ -64,5 +71,8 @@ int main(void) {
    }
    freeReplyObject(reply);

+    /* Disconnects and frees the context */
+    redisFree(c);
+
    return 0;
 }
@@ -13,4 +13,8 @@
 #define _XOPEN_SOURCE
 #endif

+#if __APPLE__ && __MACH__
+#define _OSX
+#endif
+
 #endif
@@ -917,7 +917,7 @@ err:
 * %b represents a binary safe string
 *
 * When using %b you need to provide both the pointer to the string
- * and the length in bytes. Examples:
+ * and the length in bytes as a size_t. Examples:
 *
 * len = redisFormatCommand(target, "GET %s", mykey);
 * len = redisFormatCommand(target, "SET %s %b", mykey, myval, myvallen);
@@ -1010,58 +1010,122 @@ void redisFree(redisContext *c) {
    free(c);
 }

+int redisFreeKeepFd(redisContext *c) {
+	int fd = c->fd;
+	c->fd = -1;
+	redisFree(c);
+	return fd;
+}
+
 /* Connect to a Redis instance. On error the field error in the returned
 * context will be set to the return value of the error function.
 * When no set of reply functions is given, the default set will be used. */
 redisContext *redisConnect(const char *ip, int port) {
-    redisContext *c = redisContextInit();
+    redisContext *c;
+
+    c = redisContextInit();
+    if (c == NULL)
+        return NULL;
+
    c->flags |= REDIS_BLOCK;
    redisContextConnectTcp(c,ip,port,NULL);
    return c;
 }

-redisContext *redisConnectWithTimeout(const char *ip, int port, struct timeval tv) {
-    redisContext *c = redisContextInit();
+redisContext *redisConnectWithTimeout(const char *ip, int port, const struct timeval tv) {
+    redisContext *c;
+
+    c = redisContextInit();
+    if (c == NULL)
+        return NULL;
+
    c->flags |= REDIS_BLOCK;
    redisContextConnectTcp(c,ip,port,&tv);
    return c;
 }

 redisContext *redisConnectNonBlock(const char *ip, int port) {
-    redisContext *c = redisContextInit();
+    redisContext *c;
+
+    c = redisContextInit();
+    if (c == NULL)
+        return NULL;
+
    c->flags &= ~REDIS_BLOCK;
    redisContextConnectTcp(c,ip,port,NULL);
    return c;
 }

-redisContext *redisConnectUnix(const char *path) {
+redisContext *redisConnectBindNonBlock(const char *ip, int port,
+                                       const char *source_addr) {
    redisContext *c = redisContextInit();
+    c->flags &= ~REDIS_BLOCK;
+    redisContextConnectBindTcp(c,ip,port,NULL,source_addr);
+    return c;
+}
+
+redisContext *redisConnectUnix(const char *path) {
+    redisContext *c;
+
+    c = redisContextInit();
+    if (c == NULL)
+        return NULL;
+
    c->flags |= REDIS_BLOCK;
    redisContextConnectUnix(c,path,NULL);
    return c;
 }

-redisContext *redisConnectUnixWithTimeout(const char *path, struct timeval tv) {
-    redisContext *c = redisContextInit();
+redisContext *redisConnectUnixWithTimeout(const char *path, const struct timeval tv) {
+    redisContext *c;
+
+    c = redisContextInit();
+    if (c == NULL)
+        return NULL;
+
    c->flags |= REDIS_BLOCK;
    redisContextConnectUnix(c,path,&tv);
    return c;
 }

 redisContext *redisConnectUnixNonBlock(const char *path) {
-    redisContext *c = redisContextInit();
+    redisContext *c;
+
+    c = redisContextInit();
+    if (c == NULL)
+        return NULL;
+
    c->flags &= ~REDIS_BLOCK;
    redisContextConnectUnix(c,path,NULL);
    return c;
 }

+redisContext *redisConnectFd(int fd) {
+    redisContext *c;
+
+    c = redisContextInit();
+    if (c == NULL)
+        return NULL;
+
+    c->fd = fd;
+    c->flags |= REDIS_BLOCK | REDIS_CONNECTED;
+    return c;
+}
+
 /* Set read/write timeout on a blocking socket. */
-int redisSetTimeout(redisContext *c, struct timeval tv) {
+int redisSetTimeout(redisContext *c, const struct timeval tv) {
    if (c->flags & REDIS_BLOCK)
        return redisContextSetTimeout(c,tv);
    return REDIS_ERR;
 }

+/* Enable connection KeepAlive. */
+int redisEnableKeepAlive(redisContext *c) {
+    if (redisKeepAlive(c, REDIS_KEEPALIVE_INTERVAL) != REDIS_OK)
+        return REDIS_ERR;
+    return REDIS_OK;
+}
+
 /* Use this function to handle a read event on the descriptor. It will try
 * and read some bytes from the socket and feed them to the reply parser.
 *
@@ -1077,7 +1141,7 @@ int redisBufferRead(redisContext *c) {

    nread = read(c->fd,buf,sizeof(buf));
    if (nread == -1) {
-        if (errno == EAGAIN && !(c->flags & REDIS_BLOCK)) {
+        if ((errno == EAGAIN && !(c->flags & REDIS_BLOCK)) || (errno == EINTR)) {
            /* Try again later */
        } else {
            __redisSetError(c,REDIS_ERR_IO,NULL);
@@ -1114,7 +1178,7 @@ int redisBufferWrite(redisContext *c, int *done) {
    if (sdslen(c->obuf) > 0) {
        nwritten = write(c->fd,c->obuf,sdslen(c->obuf));
        if (nwritten == -1) {
-            if (errno == EAGAIN && !(c->flags & REDIS_BLOCK)) {
+            if ((errno == EAGAIN && !(c->flags & REDIS_BLOCK)) || (errno == EINTR)) {
                /* Try again later */
            } else {
                __redisSetError(c,REDIS_ERR_IO,NULL);
@@ -1180,7 +1244,7 @@ int redisGetReply(redisContext *c, void **reply) {
 * is used, you need to call redisGetReply yourself to retrieve
 * the reply (or replies in pub/sub).
 */
-int __redisAppendCommand(redisContext *c, char *cmd, size_t len) {
+int __redisAppendCommand(redisContext *c, const char *cmd, size_t len) {
    sds newbuf;

    newbuf = sdscatlen(c->obuf,cmd,len);
@@ -1193,6 +1257,15 @@ int __redisAppendCommand(redisContext *c, char *cmd, size_t len) {
    return REDIS_OK;
 }

+int redisAppendFormattedCommand(redisContext *c, const char *cmd, size_t len) {
+
+    if (__redisAppendCommand(c, cmd, len) != REDIS_OK) {
+        return REDIS_ERR;
+    }
+
+    return REDIS_OK;
+}
+
 int redisvAppendCommand(redisContext *c, const char *format, va_list ap) {
    char *cmd;
    int len;
@@ -36,8 +36,8 @@
 #include <sys/time.h> /* for struct timeval */

 #define HIREDIS_MAJOR 0
-#define HIREDIS_MINOR 10
-#define HIREDIS_PATCH 1
+#define HIREDIS_MINOR 11
+#define HIREDIS_PATCH 0

 #define REDIS_ERR -1
 #define REDIS_OK 0
@@ -88,6 +88,8 @@

 #define REDIS_READER_MAX_BUF (1024*16)  /* Default max unused reader buffer. */

+#define REDIS_KEEPALIVE_INTERVAL 15 /* seconds */
+
 #ifdef __cplusplus
 extern "C" {
 #endif
@@ -171,13 +173,17 @@ typedef struct redisContext {
 } redisContext;

 redisContext *redisConnect(const char *ip, int port);
-redisContext *redisConnectWithTimeout(const char *ip, int port, struct timeval tv);
+redisContext *redisConnectWithTimeout(const char *ip, int port, const struct timeval tv);
 redisContext *redisConnectNonBlock(const char *ip, int port);
+redisContext *redisConnectBindNonBlock(const char *ip, int port, const char *source_addr);
 redisContext *redisConnectUnix(const char *path);
-redisContext *redisConnectUnixWithTimeout(const char *path, struct timeval tv);
+redisContext *redisConnectUnixWithTimeout(const char *path, const struct timeval tv);
 redisContext *redisConnectUnixNonBlock(const char *path);
-int redisSetTimeout(redisContext *c, struct timeval tv);
+redisContext *redisConnectFd(int fd);
+int redisSetTimeout(redisContext *c, const struct timeval tv);
+int redisEnableKeepAlive(redisContext *c);
 void redisFree(redisContext *c);
+int redisFreeKeepFd(redisContext *c);
 int redisBufferRead(redisContext *c);
 int redisBufferWrite(redisContext *c, int *done);

@@ -188,6 +194,10 @@ int redisBufferWrite(redisContext *c, int *done);
 int redisGetReply(redisContext *c, void **reply);
 int redisGetReplyFromReader(redisContext *c, void **reply);

+/* Write a formatted command to the output buffer. Use these functions in blocking mode
+ * to get a pipeline of commands. */
+int redisAppendFormattedCommand(redisContext *c, const char *cmd, size_t len);
+
 /* Write a command to the output buffer. Use these functions in blocking mode
 * to get a pipeline of commands. */
 int redisvAppendCommand(redisContext *c, const char *format, va_list ap);
@@ -54,8 +54,15 @@
 /* Defined in hiredis.c */
 void __redisSetError(redisContext *c, int type, const char *str);

+static void redisContextCloseFd(redisContext *c) {
+    if (c && c->fd >= 0) {
+        close(c->fd);
+        c->fd = -1;
+    }
+}
+
 static void __redisSetErrorFromErrno(redisContext *c, int type, const char *prefix) {
-    char buf[128];
+    char buf[128] = { 0 };
    size_t len = 0;

    if (prefix != NULL)
@@ -64,11 +71,11 @@ static void __redisSetErrorFromErrno(redisContext *c, int type, const char *pref
    __redisSetError(c,type,buf);
 }

-static int redisSetReuseAddr(redisContext *c, int fd) {
+static int redisSetReuseAddr(redisContext *c) {
    int on = 1;
-    if (setsockopt(fd, SOL_SOCKET, SO_REUSEADDR, &on, sizeof(on)) == -1) {
+    if (setsockopt(c->fd, SOL_SOCKET, SO_REUSEADDR, &on, sizeof(on)) == -1) {
        __redisSetErrorFromErrno(c,REDIS_ERR_IO,NULL);
-        close(fd);
+        redisContextCloseFd(c);
        return REDIS_ERR;
    }
    return REDIS_OK;
@@ -80,23 +87,24 @@ static int redisCreateSocket(redisContext *c, int type) {
        __redisSetErrorFromErrno(c,REDIS_ERR_IO,NULL);
        return REDIS_ERR;
    }
+    c->fd = s;
    if (type == AF_INET) {
-        if (redisSetReuseAddr(c,s) == REDIS_ERR) {
+        if (redisSetReuseAddr(c) == REDIS_ERR) {
            return REDIS_ERR;
        }
    }
-    return s;
+    return REDIS_OK;
 }

-static int redisSetBlocking(redisContext *c, int fd, int blocking) {
+static int redisSetBlocking(redisContext *c, int blocking) {
    int flags;

    /* Set the socket nonblocking.
     * Note that fcntl(2) for F_GETFL and F_SETFL can't be
     * interrupted by a signal. */
-    if ((flags = fcntl(fd, F_GETFL)) == -1) {
+    if ((flags = fcntl(c->fd, F_GETFL)) == -1) {
        __redisSetErrorFromErrno(c,REDIS_ERR_IO,"fcntl(F_GETFL)");
-        close(fd);
+        redisContextCloseFd(c);
        return REDIS_ERR;
    }

@@ -105,19 +113,61 @@ static int redisSetBlocking(redisContext *c, int fd, int blocking) {
    else
        flags |= O_NONBLOCK;

-    if (fcntl(fd, F_SETFL, flags) == -1) {
+    if (fcntl(c->fd, F_SETFL, flags) == -1) {
        __redisSetErrorFromErrno(c,REDIS_ERR_IO,"fcntl(F_SETFL)");
-        close(fd);
+        redisContextCloseFd(c);
        return REDIS_ERR;
    }
    return REDIS_OK;
 }

-static int redisSetTcpNoDelay(redisContext *c, int fd) {
+int redisKeepAlive(redisContext *c, int interval) {
+    int val = 1;
+    int fd = c->fd;
+
+    if (setsockopt(fd, SOL_SOCKET, SO_KEEPALIVE, &val, sizeof(val)) == -1){
+        __redisSetError(c,REDIS_ERR_OTHER,strerror(errno));
+        return REDIS_ERR;
+    }
+
+    val = interval;
+
+#ifdef _OSX
+    if (setsockopt(fd, IPPROTO_TCP, TCP_KEEPALIVE, &val, sizeof(val)) < 0) {
+        __redisSetError(c,REDIS_ERR_OTHER,strerror(errno));
+        return REDIS_ERR;
+    }
+#else
+#ifndef __sun
+    val = interval;
+    if (setsockopt(fd, IPPROTO_TCP, TCP_KEEPIDLE, &val, sizeof(val)) < 0) {
+        __redisSetError(c,REDIS_ERR_OTHER,strerror(errno));
+        return REDIS_ERR;
+    }
+
+    val = interval/3;
+    if (val == 0) val = 1;
+    if (setsockopt(fd, IPPROTO_TCP, TCP_KEEPINTVL, &val, sizeof(val)) < 0) {
+        __redisSetError(c,REDIS_ERR_OTHER,strerror(errno));
+        return REDIS_ERR;
+    }
+
+    val = 3;
+    if (setsockopt(fd, IPPROTO_TCP, TCP_KEEPCNT, &val, sizeof(val)) < 0) {
+        __redisSetError(c,REDIS_ERR_OTHER,strerror(errno));
+        return REDIS_ERR;
+    }
+#endif
+#endif
+
+    return REDIS_OK;
+}
+
+static int redisSetTcpNoDelay(redisContext *c) {
    int yes = 1;
-    if (setsockopt(fd, IPPROTO_TCP, TCP_NODELAY, &yes, sizeof(yes)) == -1) {
+    if (setsockopt(c->fd, IPPROTO_TCP, TCP_NODELAY, &yes, sizeof(yes)) == -1) {
        __redisSetErrorFromErrno(c,REDIS_ERR_IO,"setsockopt(TCP_NODELAY)");
-        close(fd);
+        redisContextCloseFd(c);
        return REDIS_ERR;
    }
    return REDIS_OK;
@@ -125,18 +175,19 @@ static int redisSetTcpNoDelay(redisContext *c, int fd) {

 #define __MAX_MSEC (((LONG_MAX) - 999) / 1000)

-static int redisContextWaitReady(redisContext *c, int fd, const struct timeval *timeout) {
+static int redisContextWaitReady(redisContext *c, const struct timeval *timeout) {
    struct pollfd   wfd[1];
    long msec;

    msec          = -1;
-    wfd[0].fd     = fd;
+    wfd[0].fd     = c->fd;
    wfd[0].events = POLLOUT;

    /* Only use timeout when not NULL. */
    if (timeout != NULL) {
        if (timeout->tv_usec > 1000000 || timeout->tv_sec > __MAX_MSEC) {
-            close(fd);
+            __redisSetErrorFromErrno(c, REDIS_ERR_IO, NULL);
+            redisContextCloseFd(c);
            return REDIS_ERR;
        }

@@ -152,47 +203,45 @@ static int redisContextWaitReady(redisContext *c, int fd, const struct timeval *

        if ((res = poll(wfd, 1, msec)) == -1) {
            __redisSetErrorFromErrno(c, REDIS_ERR_IO, "poll(2)");
-            close(fd);
+            redisContextCloseFd(c);
            return REDIS_ERR;
        } else if (res == 0) {
            errno = ETIMEDOUT;
            __redisSetErrorFromErrno(c,REDIS_ERR_IO,NULL);
-            close(fd);
+            redisContextCloseFd(c);
            return REDIS_ERR;
        }

-        if (redisCheckSocketError(c, fd) != REDIS_OK)
+        if (redisCheckSocketError(c) != REDIS_OK)
            return REDIS_ERR;

        return REDIS_OK;
    }

    __redisSetErrorFromErrno(c,REDIS_ERR_IO,NULL);
-    close(fd);
+    redisContextCloseFd(c);
    return REDIS_ERR;
 }

-int redisCheckSocketError(redisContext *c, int fd) {
+int redisCheckSocketError(redisContext *c) {
    int err = 0;
    socklen_t errlen = sizeof(err);

-    if (getsockopt(fd, SOL_SOCKET, SO_ERROR, &err, &errlen) == -1) {
+    if (getsockopt(c->fd, SOL_SOCKET, SO_ERROR, &err, &errlen) == -1) {
        __redisSetErrorFromErrno(c,REDIS_ERR_IO,"getsockopt(SO_ERROR)");
-        close(fd);
        return REDIS_ERR;
    }

    if (err) {
        errno = err;
        __redisSetErrorFromErrno(c,REDIS_ERR_IO,NULL);
-        close(fd);
        return REDIS_ERR;
    }

    return REDIS_OK;
 }

-int redisContextSetTimeout(redisContext *c, struct timeval tv) {
+int redisContextSetTimeout(redisContext *c, const struct timeval tv) {
    if (setsockopt(c->fd,SOL_SOCKET,SO_RCVTIMEO,&tv,sizeof(tv)) == -1) {
        __redisSetErrorFromErrno(c,REDIS_ERR_IO,"setsockopt(SO_RCVTIMEO)");
        return REDIS_ERR;
@@ -204,10 +253,12 @@ int redisContextSetTimeout(redisContext *c, struct timeval tv) {
    return REDIS_OK;
 }

-int redisContextConnectTcp(redisContext *c, const char *addr, int port, struct timeval *timeout) {
+static int _redisContextConnectTcp(redisContext *c, const char *addr, int port,
+                                   const struct timeval *timeout,
+                                   const char *source_addr) {
    int s, rv;
    char _port[6];  /* strlen("65535"); */
-    struct addrinfo hints, *servinfo, *p;
+    struct addrinfo hints, *servinfo, *bservinfo, *p, *b;
    int blocking = (c->flags & REDIS_BLOCK);

    snprintf(_port, 6, "%d", port);
@@ -231,25 +282,47 @@ int redisContextConnectTcp(redisContext *c, const char *addr, int port, struct t
        if ((s = socket(p->ai_family,p->ai_socktype,p->ai_protocol)) == -1)
            continue;

-        if (redisSetBlocking(c,s,0) != REDIS_OK)
+        c->fd = s;
+        if (redisSetBlocking(c,0) != REDIS_OK)
            goto error;
+        if (source_addr) {
+            int bound = 0;
+            /* Using getaddrinfo saves us from self-determining IPv4 vs IPv6 */
+            if ((rv = getaddrinfo(source_addr, NULL, &hints, &bservinfo)) != 0) {
+                char buf[128];
+                snprintf(buf,sizeof(buf),"Can't get addr: %s",gai_strerror(rv));
+                __redisSetError(c,REDIS_ERR_OTHER,buf);
+                goto error;
+            }
+            for (b = bservinfo; b != NULL; b = b->ai_next) {
+                if (bind(s,b->ai_addr,b->ai_addrlen) != -1) {
+                    bound = 1;
+                    break;
+                }
+            }
+            if (!bound) {
+                char buf[128];
+                snprintf(buf,sizeof(buf),"Can't bind socket: %s",strerror(errno));
+                __redisSetError(c,REDIS_ERR_OTHER,buf);
+                goto error;
+            }
+        }
        if (connect(s,p->ai_addr,p->ai_addrlen) == -1) {
            if (errno == EHOSTUNREACH) {
-                close(s);
+                redisContextCloseFd(c);
                continue;
            } else if (errno == EINPROGRESS && !blocking) {
                /* This is ok. */
            } else {
-                if (redisContextWaitReady(c,s,timeout) != REDIS_OK)
+                if (redisContextWaitReady(c,timeout) != REDIS_OK)
                    goto error;
            }
        }
-        if (blocking && redisSetBlocking(c,s,1) != REDIS_OK)
+        if (blocking && redisSetBlocking(c,1) != REDIS_OK)
            goto error;
-        if (redisSetTcpNoDelay(c,s) != REDIS_OK)
+        if (redisSetTcpNoDelay(c) != REDIS_OK)
            goto error;

-        c->fd = s;
        c->flags |= REDIS_CONNECTED;
        rv = REDIS_OK;
        goto end;
@@ -268,32 +341,41 @@ end:
    return rv;  // Need to return REDIS_OK if alright
 }

-int redisContextConnectUnix(redisContext *c, const char *path, struct timeval *timeout) {
-    int s;
+int redisContextConnectTcp(redisContext *c, const char *addr, int port,
+                           const struct timeval *timeout) {
+    return _redisContextConnectTcp(c, addr, port, timeout, NULL);
+}
+
+int redisContextConnectBindTcp(redisContext *c, const char *addr, int port,
+                               const struct timeval *timeout,
+                               const char *source_addr) {
+    return _redisContextConnectTcp(c, addr, port, timeout, source_addr);
+}
+
+int redisContextConnectUnix(redisContext *c, const char *path, const struct timeval *timeout) {
    int blocking = (c->flags & REDIS_BLOCK);
    struct sockaddr_un sa;

-    if ((s = redisCreateSocket(c,AF_LOCAL)) < 0)
+    if (redisCreateSocket(c,AF_LOCAL) < 0)
        return REDIS_ERR;
-    if (redisSetBlocking(c,s,0) != REDIS_OK)
+    if (redisSetBlocking(c,0) != REDIS_OK)
        return REDIS_ERR;

    sa.sun_family = AF_LOCAL;
    strncpy(sa.sun_path,path,sizeof(sa.sun_path)-1);
-    if (connect(s, (struct sockaddr*)&sa, sizeof(sa)) == -1) {
+    if (connect(c->fd, (struct sockaddr*)&sa, sizeof(sa)) == -1) {
        if (errno == EINPROGRESS && !blocking) {
            /* This is ok. */
        } else {
-            if (redisContextWaitReady(c,s,timeout) != REDIS_OK)
+            if (redisContextWaitReady(c,timeout) != REDIS_OK)
                return REDIS_ERR;
        }
    }

    /* Reset socket to be blocking after connect(2). */
-    if (blocking && redisSetBlocking(c,s,1) != REDIS_OK)
+    if (blocking && redisSetBlocking(c,1) != REDIS_OK)
        return REDIS_ERR;

-    c->fd = s;
    c->flags |= REDIS_CONNECTED;
    return REDIS_OK;
 }
@@ -39,9 +39,13 @@
 #define AF_LOCAL AF_UNIX
 #endif

-int redisCheckSocketError(redisContext *c, int fd);
-int redisContextSetTimeout(redisContext *c, struct timeval tv);
-int redisContextConnectTcp(redisContext *c, const char *addr, int port, struct timeval *timeout);
-int redisContextConnectUnix(redisContext *c, const char *path, struct timeval *timeout);
+int redisCheckSocketError(redisContext *c);
+int redisContextSetTimeout(redisContext *c, const struct timeval tv);
+int redisContextConnectTcp(redisContext *c, const char *addr, int port, const struct timeval *timeout);
+int redisContextConnectBindTcp(redisContext *c, const char *addr, int port,
+                               const struct timeval *timeout,
+                               const char *source_addr);
+int redisContextConnectUnix(redisContext *c, const char *path, const struct timeval *timeout);
+int redisKeepAlive(redisContext *c, int interval);

 #endif
@@ -289,27 +289,118 @@ sds sdscpy(sds s, const char *t) {
    return sdscpylen(s, t, strlen(t));
 }

+/* Helper for sdscatlonglong() doing the actual number -> string
+ * conversion. 's' must point to a string with room for at least
+ * SDS_LLSTR_SIZE bytes.
+ *
+ * The function returns the lenght of the null-terminated string
+ * representation stored at 's'. */
+#define SDS_LLSTR_SIZE 21
+int sdsll2str(char *s, long long value) {
+    char *p, aux;
+    unsigned long long v;
+    size_t l;
+
+    /* Generate the string representation, this method produces
+     * an reversed string. */
+    v = (value < 0) ? -value : value;
+    p = s;
+    do {
+        *p++ = '0'+(v%10);
+        v /= 10;
+    } while(v);
+    if (value < 0) *p++ = '-';
+
+    /* Compute length and add null term. */
+    l = p-s;
+    *p = '\0';
+
+    /* Reverse the string. */
+    p--;
+    while(s < p) {
+        aux = *s;
+        *s = *p;
+        *p = aux;
+        s++;
+        p--;
+    }
+    return l;
+}
+
+/* Identical sdsll2str(), but for unsigned long long type. */
+int sdsull2str(char *s, unsigned long long v) {
+    char *p, aux;
+    size_t l;
+
+    /* Generate the string representation, this method produces
+     * an reversed string. */
+    p = s;
+    do {
+        *p++ = '0'+(v%10);
+        v /= 10;
+    } while(v);
+
+    /* Compute length and add null term. */
+    l = p-s;
+    *p = '\0';
+
+    /* Reverse the string. */
+    p--;
+    while(s < p) {
+        aux = *s;
+        *s = *p;
+        *p = aux;
+        s++;
+        p--;
+    }
+    return l;
+}
+
+/* Create an sds string from a long long value. It is much faster than:
+ *
+ * sdscatprintf(sdsempty(),"%lld\n", value);
+ */
+sds sdsfromlonglong(long long value) {
+    char buf[SDS_LLSTR_SIZE];
+    int len = sdsll2str(buf,value);
+
+    return sdsnewlen(buf,len);
+}
+
 /* Like sdscatpritf() but gets va_list instead of being variadic. */
 sds sdscatvprintf(sds s, const char *fmt, va_list ap) {
    va_list cpy;
-    char *buf, *t;
-    size_t buflen = 16;
+    char staticbuf[1024], *buf = staticbuf, *t;
+    size_t buflen = strlen(fmt)*2;

-    while(1) {
+    /* We try to start using a static buffer for speed.
+     * If not possible we revert to heap allocation. */
+    if (buflen > sizeof(staticbuf)) {
        buf = zmalloc(buflen);
        if (buf == NULL) return NULL;
+    } else {
+        buflen = sizeof(staticbuf);
+    }
+
+    /* Try with buffers two times bigger every time we fail to
+     * fit the string in the current buffer size. */
+    while(1) {
        buf[buflen-2] = '\0';
        va_copy(cpy,ap);
        vsnprintf(buf, buflen, fmt, cpy);
        if (buf[buflen-2] != '\0') {
-            zfree(buf);
+            if (buf != staticbuf) zfree(buf);
            buflen *= 2;
+            buf = zmalloc(buflen);
+            if (buf == NULL) return NULL;
            continue;
        }
        break;
    }
+
+    /* Finally concat the obtained string to the SDS string and return it. */
    t = sdscat(s, buf);
-    zfree(buf);
+    if (buf != staticbuf) zfree(buf);
    return t;
 }

@@ -338,6 +429,122 @@ sds sdscatprintf(sds s, const char *fmt, ...) {
    return t;
 }

+/* This function is similar to sdscatprintf, but much faster as it does
+ * not rely on sprintf() family functions implemented by the libc that
+ * are often very slow. Moreover directly handling the sds string as
+ * new data is concatenated provides a performance improvement.
+ *
+ * However this function only handles an incompatible subset of printf-alike
+ * format specifiers:
+ *
+ * %s - C String
+ * %S - SDS string
+ * %i - signed int
+ * %I - 64 bit signed integer (long long, int64_t)
+ * %u - unsigned int
+ * %U - 64 bit unsigned integer (unsigned long long, uint64_t)
+ * %% - Verbatim "%" character.
+ */
+sds sdscatfmt(sds s, char const *fmt, ...) {
+    struct sdshdr *sh = (void*) (s-(sizeof(struct sdshdr)));
+    size_t initlen = sdslen(s);
+    const char *f = fmt;
+    int i;
+    va_list ap;
+
+    va_start(ap,fmt);
+    f = fmt;    /* Next format specifier byte to process. */
+    i = initlen; /* Position of the next byte to write to dest str. */
+    while(*f) {
+        char next, *str;
+        int l;
+        long long num;
+        unsigned long long unum;
+
+        /* Make sure there is always space for at least 1 char. */
+        if (sh->free == 0) {
+            s = sdsMakeRoomFor(s,1);
+            sh = (void*) (s-(sizeof(struct sdshdr)));
+        }
+
+        switch(*f) {
+        case '%':
+            next = *(f+1);
+            f++;
+            switch(next) {
+            case 's':
+            case 'S':
+                str = va_arg(ap,char*);
+                l = (next == 's') ? strlen(str) : sdslen(str);
+                if (sh->free < l) {
+                    s = sdsMakeRoomFor(s,l);
+                    sh = (void*) (s-(sizeof(struct sdshdr)));
+                }
+                memcpy(s+i,str,l);
+                sh->len += l;
+                sh->free -= l;
+                i += l;
+                break;
+            case 'i':
+            case 'I':
+                if (next == 'i')
+                    num = va_arg(ap,int);
+                else
+                    num = va_arg(ap,long long);
+                {
+                    char buf[SDS_LLSTR_SIZE];
+                    l = sdsll2str(buf,num);
+                    if (sh->free < l) {
+                        s = sdsMakeRoomFor(s,l);
+                        sh = (void*) (s-(sizeof(struct sdshdr)));
+                    }
+                    memcpy(s+i,buf,l);
+                    sh->len += l;
+                    sh->free -= l;
+                    i += l;
+                }
+                break;
+            case 'u':
+            case 'U':
+                if (next == 'u')
+                    unum = va_arg(ap,unsigned int);
+                else
+                    unum = va_arg(ap,unsigned long long);
+                {
+                    char buf[SDS_LLSTR_SIZE];
+                    l = sdsull2str(buf,unum);
+                    if (sh->free < l) {
+                        s = sdsMakeRoomFor(s,l);
+                        sh = (void*) (s-(sizeof(struct sdshdr)));
+                    }
+                    memcpy(s+i,buf,l);
+                    sh->len += l;
+                    sh->free -= l;
+                    i += l;
+                }
+                break;
+            default: /* Handle %% and generally %<unknown>. */
+                s[i++] = next;
+                sh->len += 1;
+                sh->free -= 1;
+                break;
+            }
+            break;
+        default:
+            s[i++] = *f;
+            sh->len += 1;
+            sh->free -= 1;
+            break;
+        }
+        f++;
+    }
+    va_end(ap);
+
+    /* Add null-term */
+    s[i] = '\0';
+    return s;
+}
+
 /* Remove the part of the string from left and from right composed just of
 * contiguous characters found in 'cset', that is a null terminted C string.
 *
@@ -383,7 +590,7 @@ sds sdstrim(sds s, const char *cset) {
 * Example:
 *
 * s = sdsnew("Hello World");
- * sdstrim(s,1,-1); => "ello Worl"
+ * sdsrange(s,1,-1); => "ello World"
 */
 void sdsrange(sds s, int start, int end) {
    struct sdshdr *sh = (void*) (s-(sizeof(struct sdshdr)));
@@ -525,25 +732,6 @@ void sdsfreesplitres(sds *tokens, int count) {
    zfree(tokens);
 }

-/* Create an sds string from a long long value. It is much faster than:
- *
- * sdscatprintf(sdsempty(),"%lld\n", value);
- */
-sds sdsfromlonglong(long long value) {
-    char buf[32], *p;
-    unsigned long long v;
-
-    v = (value < 0) ? -value : value;
-    p = buf+31; /* point to the last character */
-    do {
-        *p-- = '0'+(v%10);
-        v /= 10;
-    } while(v);
-    if (value < 0) *p-- = '-';
-    p++;
-    return sdsnewlen(p,32-(p-buf));
-}
-
 /* Append to the sds string "s" an escaped string representation where
 * all the non-printable characters (tested with isprint()) are turned into
 * escapes in the form "\n\r\a...." or "\x<hex-number>".
@@ -582,7 +770,7 @@ int is_hex_digit(char c) {
           (c >= 'A' && c <= 'F');
 }

-/* Helper function for sdssplitargs() that converts an hex digit into an
+/* Helper function for sdssplitargs() that converts a hex digit into an
 * integer from 0 to 15 */
 int hex_digit_to_int(char c) {
    switch(c) {
@@ -774,6 +962,7 @@ sds sdsjoin(char **argv, int argc, char *sep) {
 #ifdef SDS_TEST_MAIN
 #include <stdio.h>
 #include "testhelp.h"
+#include "limits.h"

 int main(void) {
    {
@@ -804,39 +993,61 @@ int main(void) {
        sdsfree(x);
        x = sdscatprintf(sdsempty(),"%d",123);
        test_cond("sdscatprintf() seems working in the base case",
-            sdslen(x) == 3 && memcmp(x,"123\0",4) ==0)
+            sdslen(x) == 3 && memcmp(x,"123\0",4) == 0)

        sdsfree(x);
-        x = sdstrim(sdsnew("xxciaoyyy"),"xy");
+        x = sdsnew("--");
+        x = sdscatfmt(x, "Hello %s World %I,%I--", "Hi!", LLONG_MIN,LLONG_MAX);
+        test_cond("sdscatfmt() seems working in the base case",
+            sdslen(x) == 60 &&
+            memcmp(x,"--Hello Hi! World -9223372036854775808,"
+                     "9223372036854775807--",60) == 0)
+
+        sdsfree(x);
+        x = sdsnew("--");
+        x = sdscatfmt(x, "%u,%U--", UINT_MAX, ULLONG_MAX);
+        test_cond("sdscatfmt() seems working with unsigned numbers",
+            sdslen(x) == 35 &&
+            memcmp(x,"--4294967295,18446744073709551615--",35) == 0)
+
+        sdsfree(x);
+        x = sdsnew("xxciaoyyy");
+        sdstrim(x,"xy");
        test_cond("sdstrim() correctly trims characters",
            sdslen(x) == 4 && memcmp(x,"ciao\0",5) == 0)

-        y = sdsrange(sdsdup(x),1,1);
+        y = sdsdup(x);
+        sdsrange(y,1,1);
        test_cond("sdsrange(...,1,1)",
            sdslen(y) == 1 && memcmp(y,"i\0",2) == 0)

        sdsfree(y);
-        y = sdsrange(sdsdup(x),1,-1);
+        y = sdsdup(x);
+        sdsrange(y,1,-1);
        test_cond("sdsrange(...,1,-1)",
            sdslen(y) == 3 && memcmp(y,"iao\0",4) == 0)

        sdsfree(y);
-        y = sdsrange(sdsdup(x),-2,-1);
+        y = sdsdup(x);
+        sdsrange(y,-2,-1);
        test_cond("sdsrange(...,-2,-1)",
            sdslen(y) == 2 && memcmp(y,"ao\0",3) == 0)

        sdsfree(y);
-        y = sdsrange(sdsdup(x),2,1);
+        y = sdsdup(x);
+        sdsrange(y,2,1);
        test_cond("sdsrange(...,2,1)",
            sdslen(y) == 0 && memcmp(y,"\0",1) == 0)

        sdsfree(y);
-        y = sdsrange(sdsdup(x),1,100);
+        y = sdsdup(x);
+        sdsrange(y,1,100);
        test_cond("sdsrange(...,1,100)",
            sdslen(y) == 3 && memcmp(y,"iao\0",4) == 0)

        sdsfree(y);
-        y = sdsrange(sdsdup(x),100,100);
+        y = sdsdup(x);
+        sdsrange(y,100,100);
        test_cond("sdsrange(...,100,100)",
            sdslen(y) == 0 && memcmp(y,"\0",1) == 0)

@@ -858,6 +1069,13 @@ int main(void) {
        y = sdsnew("bar");
        test_cond("sdscmp(bar,bar)", sdscmp(x,y) < 0)

+        sdsfree(y);
+        sdsfree(x);
+        x = sdsnewlen("\a\n\0foo\r",7);
+        y = sdscatrepr(sdsempty(),x,sdslen(x));
+        test_cond("sdscatrepr(...data...)",
+            memcmp(y,"\"\\a\\n\\x00foo\\r\"",15) == 0)
+
        {
            int oldfree;

@@ -76,6 +76,7 @@ sds sdscatprintf(sds s, const char *fmt, ...)
 sds sdscatprintf(sds s, const char *fmt, ...);
 #endif

+sds sdscatfmt(sds s, char const *fmt, ...);
 sds sdstrim(sds s, const char *cset);
 void sdsrange(sds s, int start, int end);
 void sdsupdatelen(sds s);
@@ -8,12 +8,14 @@
 #include <unistd.h>
 #include <signal.h>
 #include <errno.h>
+#include <limits.h>

 #include "hiredis.h"

 enum connection_type {
    CONN_TCP,
-    CONN_UNIX
+    CONN_UNIX,
+    CONN_FD
 };

 struct config {
@@ -22,6 +24,7 @@ struct config {
    struct {
        const char *host;
        int port;
+        struct timeval timeout;
    } tcp;

    struct {
@@ -62,7 +65,7 @@ static redisContext *select_database(redisContext *c) {
    return c;
 }

-static void disconnect(redisContext *c) {
+static int disconnect(redisContext *c, int keep_fd) {
    redisReply *reply;

    /* Make sure we're on DB 9. */
@@ -73,8 +76,11 @@ static void disconnect(redisContext *c) {
    assert(reply != NULL);
    freeReplyObject(reply);

-    /* Free the context as well. */
+    /* Free the context as well, but keep the fd if requested. */
+    if (keep_fd)
+        return redisFreeKeepFd(c);
    redisFree(c);
+    return -1;
 }

 static redisContext *connect(struct config config) {
@@ -84,11 +90,22 @@ static redisContext *connect(struct config config) {
        c = redisConnect(config.tcp.host, config.tcp.port);
    } else if (config.type == CONN_UNIX) {
        c = redisConnectUnix(config.unix.path);
+    } else if (config.type == CONN_FD) {
+        /* Create a dummy connection just to get an fd to inherit */
+        redisContext *dummy_ctx = redisConnectUnix(config.unix.path);
+        if (dummy_ctx) {
+            int fd = disconnect(dummy_ctx, 1);
+            printf("Connecting to inherited fd %d\n", fd);
+            c = redisConnectFd(fd);
+        }
    } else {
        assert(NULL);
    }

-    if (c->err) {
+    if (c == NULL) {
+        printf("Connection error: can't allocate redis context\n");
+        exit(1);
+    } else if (c->err) {
        printf("Connection error: %s\n", c->errstr);
        exit(1);
    }
@@ -125,13 +142,13 @@ static void test_format_commands(void) {
    free(cmd);

    test("Format command with %%b string interpolation: ");
-    len = redisFormatCommand(&cmd,"SET %b %b","foo",3,"b\0r",3);
+    len = redisFormatCommand(&cmd,"SET %b %b","foo",(size_t)3,"b\0r",(size_t)3);
    test_cond(strncmp(cmd,"*3\r\n$3\r\nSET\r\n$3\r\nfoo\r\n$3\r\nb\0r\r\n",len) == 0 &&
        len == 4+4+(3+2)+4+(3+2)+4+(3+2));
    free(cmd);

    test("Format command with %%b and an empty string: ");
-    len = redisFormatCommand(&cmd,"SET %b %b","foo",3,"",0);
+    len = redisFormatCommand(&cmd,"SET %b %b","foo",(size_t)3,"",(size_t)0);
    test_cond(strncmp(cmd,"*3\r\n$3\r\nSET\r\n$3\r\nfoo\r\n$0\r\n\r\n",len) == 0 &&
        len == 4+4+(3+2)+4+(3+2)+4+(0+2));
    free(cmd);
@@ -177,7 +194,7 @@ static void test_format_commands(void) {
    FLOAT_WIDTH_TEST(double);

    test("Format command with invalid printf format: ");
-    len = redisFormatCommand(&cmd,"key:%08p %b",(void*)1234,"foo",3);
+    len = redisFormatCommand(&cmd,"key:%08p %b",(void*)1234,"foo",(size_t)3);
    test_cond(len == -1);

    const char *argv[3];
@@ -200,10 +217,33 @@ static void test_format_commands(void) {
    free(cmd);
 }

+static void test_append_formatted_commands(struct config config) {
+    redisContext *c;
+    redisReply *reply;
+    char *cmd;
+    int len;
+
+    c = connect(config);
+
+    test("Append format command: ");
+
+    len = redisFormatCommand(&cmd, "SET foo bar");
+
+    test_cond(redisAppendFormattedCommand(c, cmd, len) == REDIS_OK);
+
+    assert(redisGetReply(c, (void*)&reply) == REDIS_OK);
+
+    free(cmd);
+    freeReplyObject(reply);
+
+    disconnect(c, 0);
+}
+
 static void test_reply_reader(void) {
    redisReader *reader;
    void *reply;
    int ret;
+    int i;

    test("Error handling in reply parser: ");
    reader = redisReaderCreate();
@@ -225,12 +265,13 @@ static void test_reply_reader(void) {
              strcasecmp(reader->errstr,"Protocol error, got \"@\" as reply type byte") == 0);
    redisReaderFree(reader);

-    test("Set error on nested multi bulks with depth > 2: ");
+    test("Set error on nested multi bulks with depth > 7: ");
    reader = redisReaderCreate();
-    redisReaderFeed(reader,(char*)"*1\r\n",4);
-    redisReaderFeed(reader,(char*)"*1\r\n",4);
-    redisReaderFeed(reader,(char*)"*1\r\n",4);
-    redisReaderFeed(reader,(char*)"*1\r\n",4);
+
+    for (i = 0; i < 9; i++) {
+        redisReaderFeed(reader,(char*)"*1\r\n",4);
+    }
+
    ret = redisReaderGetReply(reader,NULL);
    test_cond(ret == REDIS_ERR &&
              strncasecmp(reader->errstr,"No support for",14) == 0);
@@ -284,7 +325,10 @@ static void test_blocking_connection_errors(void) {
    c = redisConnect((char*)"idontexist.local", 6379);
    test_cond(c->err == REDIS_ERR_OTHER &&
        (strcmp(c->errstr,"Name or service not known") == 0 ||
-         strcmp(c->errstr,"Can't resolve: idontexist.local") == 0));
+         strcmp(c->errstr,"Can't resolve: idontexist.local") == 0 ||
+         strcmp(c->errstr,"nodename nor servname provided, or not known") == 0 ||
+         strcmp(c->errstr,"No address associated with hostname") == 0 ||
+         strcmp(c->errstr,"no address associated with name") == 0));
    redisFree(c);

    test("Returns error when the port is not open: ");
@@ -326,7 +370,7 @@ static void test_blocking_connection(struct config config) {
    freeReplyObject(reply);

    test("%%b String interpolation works: ");
-    reply = redisCommand(c,"SET %b %b","foo",3,"hello\x00world",11);
+    reply = redisCommand(c,"SET %b %b","foo",(size_t)3,"hello\x00world",(size_t)11);
    freeReplyObject(reply);
    reply = redisCommand(c,"GET foo");
    test_cond(reply->type == REDIS_REPLY_STRING &&
@@ -374,7 +418,7 @@ static void test_blocking_connection(struct config config) {
              strcasecmp(reply->element[1]->str,"pong") == 0);
    freeReplyObject(reply);

-    disconnect(c);
+    disconnect(c, 0);
 }

 static void test_blocking_io_errors(struct config config) {
@@ -428,6 +472,30 @@ static void test_blocking_io_errors(struct config config) {
    redisFree(c);
 }

+static void test_invalid_timeout_errors(struct config config) {
+    redisContext *c;
+
+    test("Set error when an invalid timeout usec value is given to redisConnectWithTimeout: ");
+
+    config.tcp.timeout.tv_sec = 0;
+    config.tcp.timeout.tv_usec = 10000001;
+
+    c = redisConnectWithTimeout(config.tcp.host, config.tcp.port, config.tcp.timeout);
+
+    test_cond(c->err == REDIS_ERR_IO);
+
+    test("Set error when an invalid timeout sec value is given to redisConnectWithTimeout: ");
+
+    config.tcp.timeout.tv_sec = (((LONG_MAX) - 999) / 1000) + 1;
+    config.tcp.timeout.tv_usec = 0;
+
+    c = redisConnectWithTimeout(config.tcp.host, config.tcp.port, config.tcp.timeout);
+
+    test_cond(c->err == REDIS_ERR_IO);
+
+    redisFree(c);
+}
+
 static void test_throughput(struct config config) {
    redisContext *c = connect(config);
    redisReply **replies;
@@ -490,7 +558,7 @@ static void test_throughput(struct config config) {
    free(replies);
    printf("\t(%dx LRANGE with 500 elements (pipelined): %.3fs)\n", num, (t2-t1)/1000000.0);

-    disconnect(c);
+    disconnect(c, 0);
 }

 // static long __test_callback_flags = 0;
@@ -603,6 +671,7 @@ int main(int argc, char **argv) {
        }
    };
    int throughput = 1;
+    int test_inherit_fd = 1;

    /* Ignore broken pipe signal (for I/O error tests). */
    signal(SIGPIPE, SIG_IGN);
@@ -621,6 +690,8 @@ int main(int argc, char **argv) {
            cfg.unix.path = argv[0];
        } else if (argc >= 1 && !strcmp(argv[0],"--skip-throughput")) {
            throughput = 0;
+        } else if (argc >= 1 && !strcmp(argv[0],"--skip-inherit-fd")) {
+            test_inherit_fd = 0;
        } else {
            fprintf(stderr, "Invalid argument: %s\n", argv[0]);
            exit(1);
@@ -636,6 +707,8 @@ int main(int argc, char **argv) {
    cfg.type = CONN_TCP;
    test_blocking_connection(cfg);
    test_blocking_io_errors(cfg);
+    test_invalid_timeout_errors(cfg);
+    test_append_formatted_commands(cfg);
    if (throughput) test_throughput(cfg);

    printf("\nTesting against Unix socket connection (%s):\n", cfg.unix.path);
@@ -644,6 +717,12 @@ int main(int argc, char **argv) {
    test_blocking_io_errors(cfg);
    if (throughput) test_throughput(cfg);

+    if (test_inherit_fd) {
+        printf("\nTesting against inherited fd (%s):\n", cfg.unix.path);
+        cfg.type = CONN_FD;
+        test_blocking_connection(cfg);
+    }
+
    if (fails) {
        printf("*** %d TESTS FAILED ***\n", fails);
        return 1;
@@ -1,20 +1,72 @@
+/*.gcov.*
+
 /autom4te.cache/
+
+/bin/jemalloc.sh
+
 /config.stamp
 /config.log
 /config.status
+/configure
+
 /doc/html.xsl
 /doc/manpages.xsl
 /doc/jemalloc.xml
+/doc/jemalloc.html
+/doc/jemalloc.3
+
 /lib/
+
 /Makefile
-/include/jemalloc/internal/jemalloc_internal\.h
-/include/jemalloc/internal/size_classes\.h
-/include/jemalloc/jemalloc\.h
-/include/jemalloc/jemalloc_defs\.h
+
+/include/jemalloc/internal/jemalloc_internal.h
+/include/jemalloc/internal/jemalloc_internal_defs.h
+/include/jemalloc/internal/private_namespace.h
+/include/jemalloc/internal/private_unnamespace.h
+/include/jemalloc/internal/public_namespace.h
+/include/jemalloc/internal/public_symbols.txt
+/include/jemalloc/internal/public_unnamespace.h
+/include/jemalloc/internal/size_classes.h
+/include/jemalloc/jemalloc.h
+/include/jemalloc/jemalloc_defs.h
+/include/jemalloc/jemalloc_macros.h
+/include/jemalloc/jemalloc_mangle.h
+/include/jemalloc/jemalloc_mangle_jet.h
+/include/jemalloc/jemalloc_protos.h
+/include/jemalloc/jemalloc_protos_jet.h
+/include/jemalloc/jemalloc_rename.h
+
 /src/*.[od]
-/test/*.[od]
-/test/*.out
-/test/[a-zA-Z_]*
-!test/*.c
-!test/*.exp
-/bin/jemalloc.sh
+/src/*.gcda
+/src/*.gcno
+
+/test/test.sh
+test/include/test/jemalloc_test.h
+test/include/test/jemalloc_test_defs.h
+
+/test/integration/[A-Za-z]*
+!/test/integration/[A-Za-z]*.*
+/test/integration/*.[od]
+/test/integration/*.gcda
+/test/integration/*.gcno
+/test/integration/*.out
+
+/test/src/*.[od]
+/test/src/*.gcda
+/test/src/*.gcno
+
+/test/stress/[A-Za-z]*
+!/test/stress/[A-Za-z]*.*
+/test/stress/*.[od]
+/test/stress/*.gcda
+/test/stress/*.gcno
+/test/stress/*.out
+
+/test/unit/[A-Za-z]*
+!/test/unit/[A-Za-z]*.*
+/test/unit/*.[od]
+/test/unit/*.gcda
+/test/unit/*.gcno
+/test/unit/*.out
+
+/VERSION
@@ -1,10 +1,10 @@
 Unless otherwise specified, files in the jemalloc source distribution are
 subject to the following license:
 --------------------------------------------------------------------------------
-Copyright (C) 2002-2012 Jason Evans <jasone@canonware.com>.
+Copyright (C) 2002-2014 Jason Evans <jasone@canonware.com>.
 All rights reserved.
 Copyright (C) 2007-2012 Mozilla Foundation.  All rights reserved.
-Copyright (C) 2009-2012 Facebook, Inc.  All rights reserved.
+Copyright (C) 2009-2014 Facebook, Inc.  All rights reserved.

 Redistribution and use in source and binary forms, with or without
 modification, are permitted provided that the following conditions are met:
@@ -3,8 +3,176 @@ bug fixes are all mentioned, but internal enhancements are omitted here for
 brevity (even though they are more fun to write about).  Much more detail can be
 found in the git revision history:

-    http://www.canonware.com/cgi-bin/gitweb.cgi?p=jemalloc.git
-    git://canonware.com/jemalloc.git
+    https://github.com/jemalloc/jemalloc
+
+* 3.6.0 (March 31, 2014)
+
+  This version contains a critical bug fix for a regression present in 3.5.0 and
+  3.5.1.
+
+  Bug fixes:
+  - Fix a regression in arena_chunk_alloc() that caused crashes during
+    small/large allocation if chunk allocation failed.  In the absence of this
+    bug, chunk allocation failure would result in allocation failure, e.g.  NULL
+    return from malloc().  This regression was introduced in 3.5.0.
+  - Fix backtracing for gcc intrinsics-based backtracing by specifying
+    -fno-omit-frame-pointer to gcc.  Note that the application (and all the
+    libraries it links to) must also be compiled with this option for
+    backtracing to be reliable.
+  - Use dss allocation precedence for huge allocations as well as small/large
+    allocations.
+  - Fix test assertion failure message formatting.  This bug did not manifect on
+    x86_64 systems because of implementation subtleties in va_list.
+  - Fix inconsequential test failures for hash and SFMT code.
+
+  New features:
+  - Support heap profiling on FreeBSD.  This feature depends on the proc
+    filesystem being mounted during heap profile dumping.
+
+* 3.5.1 (February 25, 2014)
+
+  This version primarily addresses minor bugs in test code.
+
+  Bug fixes:
+  - Configure Solaris/Illumos to use MADV_FREE.
+  - Fix junk filling for mremap(2)-based huge reallocation.  This is only
+    relevant if configuring with the --enable-mremap option specified.
+  - Avoid compilation failure if 'restrict' C99 keyword is not supported by the
+    compiler.
+  - Add a configure test for SSE2 rather than assuming it is usable on i686
+    systems.  This fixes test compilation errors, especially on 32-bit Linux
+    systems.
+  - Fix mallctl argument size mismatches (size_t vs. uint64_t) in the stats unit
+    test.
+  - Fix/remove flawed alignment-related overflow tests.
+  - Prevent compiler optimizations that could change backtraces in the
+    prof_accum unit test.
+
+* 3.5.0 (January 22, 2014)
+
+  This version focuses on refactoring and automated testing, though it also
+  includes some non-trivial heap profiling optimizations not mentioned below.
+
+  New features:
+  - Add the *allocx() API, which is a successor to the experimental *allocm()
+    API.  The *allocx() functions are slightly simpler to use because they have
+    fewer parameters, they directly return the results of primary interest, and
+    mallocx()/rallocx() avoid the strict aliasing pitfall that
+    allocm()/rallocm() share with posix_memalign().  Note that *allocm() is
+    slated for removal in the next non-bugfix release.
+  - Add support for LinuxThreads.
+
+  Bug fixes:
+  - Unless heap profiling is enabled, disable floating point code and don't link
+    with libm.  This, in combination with e.g. EXTRA_CFLAGS=-mno-sse on x64
+    systems, makes it possible to completely disable floating point register
+    use.  Some versions of glibc neglect to save/restore caller-saved floating
+    point registers during dynamic lazy symbol loading, and the symbol loading
+    code uses whatever malloc the application happens to have linked/loaded
+    with, the result being potential floating point register corruption.
+  - Report ENOMEM rather than EINVAL if an OOM occurs during heap profiling
+    backtrace creation in imemalign().  This bug impacted posix_memalign() and
+    aligned_alloc().
+  - Fix a file descriptor leak in a prof_dump_maps() error path.
+  - Fix prof_dump() to close the dump file descriptor for all relevant error
+    paths.
+  - Fix rallocm() to use the arena specified by the ALLOCM_ARENA(s) flag for
+    allocation, not just deallocation.
+  - Fix a data race for large allocation stats counters.
+  - Fix a potential infinite loop during thread exit.  This bug occurred on
+    Solaris, and could affect other platforms with similar pthreads TSD
+    implementations.
+  - Don't junk-fill reallocations unless usable size changes.  This fixes a
+    violation of the *allocx()/*allocm() semantics.
+  - Fix growing large reallocation to junk fill new space.
+  - Fix huge deallocation to junk fill when munmap is disabled.
+  - Change the default private namespace prefix from empty to je_, and change
+    --with-private-namespace-prefix so that it prepends an additional prefix
+    rather than replacing je_.  This reduces the likelihood of applications
+    which statically link jemalloc experiencing symbol name collisions.
+  - Add missing private namespace mangling (relevant when
+    --with-private-namespace is specified).
+  - Add and use JEMALLOC_INLINE_C so that static inline functions are marked as
+    static even for debug builds.
+  - Add a missing mutex unlock in a malloc_init_hard() error path.  In practice
+    this error path is never executed.
+  - Fix numerous bugs in malloc_strotumax() error handling/reporting.  These
+    bugs had no impact except for malformed inputs.
+  - Fix numerous bugs in malloc_snprintf().  These bugs were not exercised by
+    existing calls, so they had no impact.
+
+* 3.4.1 (October 20, 2013)
+
+  Bug fixes:
+  - Fix a race in the "arenas.extend" mallctl that could cause memory corruption
+    of internal data structures and subsequent crashes.
+  - Fix Valgrind integration flaws that caused Valgrind warnings about reads of
+    uninitialized memory in:
+    + arena chunk headers
+    + internal zero-initialized data structures (relevant to tcache and prof
+      code)
+  - Preserve errno during the first allocation.  A readlink(2) call during
+    initialization fails unless /etc/malloc.conf exists, so errno was typically
+    set during the first allocation prior to this fix.
+  - Fix compilation warnings reported by gcc 4.8.1.
+
+* 3.4.0 (June 2, 2013)
+
+  This version is essentially a small bugfix release, but the addition of
+  aarch64 support requires that the minor version be incremented.
+
+  Bug fixes:
+  - Fix race-triggered deadlocks in chunk_record().  These deadlocks were
+    typically triggered by multiple threads concurrently deallocating huge
+    objects.
+
+  New features:
+  - Add support for the aarch64 architecture.
+
+* 3.3.1 (March 6, 2013)
+
+  This version fixes bugs that are typically encountered only when utilizing
+  custom run-time options.
+
+  Bug fixes:
+  - Fix a locking order bug that could cause deadlock during fork if heap
+    profiling were enabled.
+  - Fix a chunk recycling bug that could cause the allocator to lose track of
+    whether a chunk was zeroed.  On FreeBSD, NetBSD, and OS X, it could cause
+    corruption if allocating via sbrk(2) (unlikely unless running with the
+    "dss:primary" option specified).  This was completely harmless on Linux
+    unless using mlockall(2) (and unlikely even then, unless the
+    --disable-munmap configure option or the "dss:primary" option was
+    specified).  This regression was introduced in 3.1.0 by the
+    mlockall(2)/madvise(2) interaction fix.
+  - Fix TLS-related memory corruption that could occur during thread exit if the
+    thread never allocated memory.  Only the quarantine and prof facilities were
+    susceptible.
+  - Fix two quarantine bugs:
+    + Internal reallocation of the quarantined object array leaked the old
+      array.
+    + Reallocation failure for internal reallocation of the quarantined object
+      array (very unlikely) resulted in memory corruption.
+  - Fix Valgrind integration to annotate all internally allocated memory in a
+    way that keeps Valgrind happy about internal data structure access.
+  - Fix building for s390 systems.
+
+* 3.3.0 (January 23, 2013)
+
+  This version includes a few minor performance improvements in addition to the
+  listed new features and bug fixes.
+
+  New features:
+  - Add clipping support to lg_chunk option processing.
+  - Add the --enable-ivsalloc option.
+  - Add the --without-export option.
+  - Add the --disable-zone-allocator option.
+
+  Bug fixes:
+  - Fix "arenas.extend" mallctl to output the number of arenas.
+  - Fix chunk_recycle() to unconditionally inform Valgrind that returned memory
+    is undefined.
+  - Fix build break on FreeBSD related to alloca.h.

 * 3.2.0 (November 9, 2012)

@@ -55,11 +55,16 @@ any of the following arguments (not a definitive list) to 'configure':
    jemalloc overlays the default malloc zone, but makes no attempt to actually
    replace the "malloc", "calloc", etc. symbols.

+--without-export
+    Don't export public APIs. This can be useful when building jemalloc as a
+    static library, or to avoid exporting public APIs when using the zone
+    allocator on OSX.
+
 --with-private-namespace=<prefix>
-    Prefix all library-private APIs with <prefix>.  For shared libraries,
+    Prefix all library-private APIs with <prefix>je_.  For shared libraries,
    symbol visibility mechanisms prevent these symbols from being exported, but
    for static libraries, naming collisions are a real possibility.  By
-    default, the prefix is "" (empty string).
+    default, <prefix> is empty, which results in a symbol prefix of je_ .

 --with-install-suffix=<suffix>
    Append <suffix> to the base name of all installed files, such that multiple
@@ -74,6 +79,25 @@ any of the following arguments (not a definitive list) to 'configure':
 --enable-debug
    Enable assertions and validation code.  This incurs a substantial
    performance hit, but is very useful during application development.
+    Implies --enable-ivsalloc.
+
+--enable-code-coverage
+    Enable code coverage support, for use during jemalloc test development.
+    Additional testing targets are available if this option is enabled:
+
+      coverage
+      coverage_unit
+      coverage_integration
+      coverage_stress
+
+    These targets do not clear code coverage results from previous runs, and
+    there are interactions between the various coverage targets, so it is
+    usually advisable to run 'make clean' between repeated code coverage runs.
+
+--enable-ivsalloc
+    Enable validation code, which verifies that pointers reside within
+    jemalloc-owned chunks before dereferencing them. This incurs a substantial
+    performance hit.

 --disable-stats
    Disable statistics gathering functionality.  See the "opt.stats_print"
@@ -136,6 +160,10 @@ any of the following arguments (not a definitive list) to 'configure':
 --disable-experimental
    Disable support for the experimental API (*allocm()).

+--disable-zone-allocator
+    Disable zone allocator for Darwin. This means jemalloc won't be hooked as
+    the default allocator on OSX/iOS.
+
 --enable-utrace
    Enable utrace(2)-based allocation tracing.  This feature is not broadly
    portable (FreeBSD has it, but Linux and OS X do not).
@@ -47,13 +47,17 @@ cfghdrs_out := @cfghdrs_out@
 cfgoutputs_in := @cfgoutputs_in@
 cfgoutputs_out := @cfgoutputs_out@
 enable_autogen := @enable_autogen@
+enable_code_coverage := @enable_code_coverage@
 enable_experimental := @enable_experimental@
+enable_zone_allocator := @enable_zone_allocator@
 DSO_LDFLAGS = @DSO_LDFLAGS@
 SOREV = @SOREV@
 PIC_CFLAGS = @PIC_CFLAGS@
 CTARGET = @CTARGET@
 LDTARGET = @LDTARGET@
 MKLIB = @MKLIB@
+AR = @AR@
+ARFLAGS = @ARFLAGS@
 CC_MM = @CC_MM@

 ifeq (macho, $(ABI))
@@ -70,18 +74,17 @@ LIBJEMALLOC := $(LIBPREFIX)jemalloc$(install_suffix)

 # Lists of files.
 BINS := $(srcroot)bin/pprof $(objroot)bin/jemalloc.sh
-CHDRS := $(objroot)include/jemalloc/jemalloc$(install_suffix).h \
-	$(objroot)include/jemalloc/jemalloc_defs$(install_suffix).h
-CSRCS := $(srcroot)src/jemalloc.c $(srcroot)src/arena.c $(srcroot)src/atomic.c \
-	$(srcroot)src/base.c $(srcroot)src/bitmap.c $(srcroot)src/chunk.c \
-	$(srcroot)src/chunk_dss.c $(srcroot)src/chunk_mmap.c \
-	$(srcroot)src/ckh.c $(srcroot)src/ctl.c $(srcroot)src/extent.c \
-	$(srcroot)src/hash.c $(srcroot)src/huge.c $(srcroot)src/mb.c \
-	$(srcroot)src/mutex.c $(srcroot)src/prof.c $(srcroot)src/quarantine.c \
-	$(srcroot)src/rtree.c $(srcroot)src/stats.c $(srcroot)src/tcache.c \
-	$(srcroot)src/util.c $(srcroot)src/tsd.c
-ifeq (macho, $(ABI))
-CSRCS += $(srcroot)src/zone.c
+C_HDRS := $(objroot)include/jemalloc/jemalloc$(install_suffix).h
+C_SRCS := $(srcroot)src/jemalloc.c $(srcroot)src/arena.c \
+	$(srcroot)src/atomic.c $(srcroot)src/base.c $(srcroot)src/bitmap.c \
+	$(srcroot)src/chunk.c $(srcroot)src/chunk_dss.c \
+	$(srcroot)src/chunk_mmap.c $(srcroot)src/ckh.c $(srcroot)src/ctl.c \
+	$(srcroot)src/extent.c $(srcroot)src/hash.c $(srcroot)src/huge.c \
+	$(srcroot)src/mb.c $(srcroot)src/mutex.c $(srcroot)src/prof.c \
+	$(srcroot)src/quarantine.c $(srcroot)src/rtree.c $(srcroot)src/stats.c \
+	$(srcroot)src/tcache.c $(srcroot)src/util.c $(srcroot)src/tsd.c
+ifeq ($(enable_zone_allocator), 1)
+C_SRCS += $(srcroot)src/zone.c
 endif
 ifeq ($(IMPORTLIB),$(SO))
 STATIC_LIBS := $(objroot)lib/$(LIBJEMALLOC).$(A)
@@ -100,27 +103,74 @@ DOCS_XML := $(objroot)doc/jemalloc$(install_suffix).xml
 DOCS_HTML := $(DOCS_XML:$(objroot)%.xml=$(srcroot)%.html)
 DOCS_MAN3 := $(DOCS_XML:$(objroot)%.xml=$(srcroot)%.3)
 DOCS := $(DOCS_HTML) $(DOCS_MAN3)
-CTESTS := $(srcroot)test/aligned_alloc.c $(srcroot)test/allocated.c \
-	$(srcroot)test/ALLOCM_ARENA.c $(srcroot)test/bitmap.c \
-	$(srcroot)test/mremap.c $(srcroot)test/posix_memalign.c \
-	$(srcroot)test/thread_arena.c $(srcroot)test/thread_tcache_enabled.c
+C_TESTLIB_SRCS := $(srcroot)test/src/math.c $(srcroot)test/src/mtx.c \
+	$(srcroot)test/src/SFMT.c $(srcroot)test/src/test.c \
+	$(srcroot)test/src/thd.c
+C_UTIL_INTEGRATION_SRCS := $(srcroot)src/util.c
+TESTS_UNIT := $(srcroot)test/unit/bitmap.c \
+	$(srcroot)test/unit/ckh.c \
+	$(srcroot)test/unit/hash.c \
+	$(srcroot)test/unit/junk.c \
+	$(srcroot)test/unit/mallctl.c \
+	$(srcroot)test/unit/math.c \
+	$(srcroot)test/unit/mq.c \
+	$(srcroot)test/unit/mtx.c \
+	$(srcroot)test/unit/prof_accum.c \
+	$(srcroot)test/unit/prof_gdump.c \
+	$(srcroot)test/unit/prof_idump.c \
+	$(srcroot)test/unit/ql.c \
+	$(srcroot)test/unit/qr.c \
+	$(srcroot)test/unit/quarantine.c \
+	$(srcroot)test/unit/rb.c \
+	$(srcroot)test/unit/rtree.c \
+	$(srcroot)test/unit/SFMT.c \
+	$(srcroot)test/unit/stats.c \
+	$(srcroot)test/unit/tsd.c \
+	$(srcroot)test/unit/util.c \
+	$(srcroot)test/unit/zero.c
+TESTS_UNIT_AUX := $(srcroot)test/unit/prof_accum_a.c \
+	$(srcroot)test/unit/prof_accum_b.c
+TESTS_INTEGRATION := $(srcroot)test/integration/aligned_alloc.c \
+	$(srcroot)test/integration/allocated.c \
+	$(srcroot)test/integration/mallocx.c \
+	$(srcroot)test/integration/mremap.c \
+	$(srcroot)test/integration/posix_memalign.c \
+	$(srcroot)test/integration/rallocx.c \
+	$(srcroot)test/integration/thread_arena.c \
+	$(srcroot)test/integration/thread_tcache_enabled.c \
+	$(srcroot)test/integration/xallocx.c
 ifeq ($(enable_experimental), 1)
-CTESTS += $(srcroot)test/allocm.c $(srcroot)test/rallocm.c
+TESTS_INTEGRATION += $(srcroot)test/integration/allocm.c \
+	$(srcroot)test/integration/MALLOCX_ARENA.c \
+	$(srcroot)test/integration/rallocm.c
 endif
+TESTS_STRESS :=
+TESTS := $(TESTS_UNIT) $(TESTS_INTEGRATION) $(TESTS_STRESS)

-COBJS := $(CSRCS:$(srcroot)%.c=$(objroot)%.$(O))
-CPICOBJS := $(CSRCS:$(srcroot)%.c=$(objroot)%.pic.$(O))
-CTESTOBJS := $(CTESTS:$(srcroot)%.c=$(objroot)%.$(O))
+C_OBJS := $(C_SRCS:$(srcroot)%.c=$(objroot)%.$(O))
+C_PIC_OBJS := $(C_SRCS:$(srcroot)%.c=$(objroot)%.pic.$(O))
+C_JET_OBJS := $(C_SRCS:$(srcroot)%.c=$(objroot)%.jet.$(O))
+C_TESTLIB_UNIT_OBJS := $(C_TESTLIB_SRCS:$(srcroot)%.c=$(objroot)%.unit.$(O))
+C_TESTLIB_INTEGRATION_OBJS := $(C_TESTLIB_SRCS:$(srcroot)%.c=$(objroot)%.integration.$(O))
+C_UTIL_INTEGRATION_OBJS := $(C_UTIL_INTEGRATION_SRCS:$(srcroot)%.c=$(objroot)%.integration.$(O))
+C_TESTLIB_STRESS_OBJS := $(C_TESTLIB_SRCS:$(srcroot)%.c=$(objroot)%.stress.$(O))
+C_TESTLIB_OBJS := $(C_TESTLIB_UNIT_OBJS) $(C_TESTLIB_INTEGRATION_OBJS) $(C_UTIL_INTEGRATION_OBJS) $(C_TESTLIB_STRESS_OBJS)

-.PHONY: all dist doc_html doc_man doc
+TESTS_UNIT_OBJS := $(TESTS_UNIT:$(srcroot)%.c=$(objroot)%.$(O))
+TESTS_UNIT_AUX_OBJS := $(TESTS_UNIT_AUX:$(srcroot)%.c=$(objroot)%.$(O))
+TESTS_INTEGRATION_OBJS := $(TESTS_INTEGRATION:$(srcroot)%.c=$(objroot)%.$(O))
+TESTS_STRESS_OBJS := $(TESTS_STRESS:$(srcroot)%.c=$(objroot)%.$(O))
+TESTS_OBJS := $(TESTS_UNIT_OBJS) $(TESTS_UNIT_AUX_OBJS) $(TESTS_INTEGRATION_OBJS) $(TESTS_STRESS_OBJS)
+
+.PHONY: all dist build_doc_html build_doc_man build_doc
 .PHONY: install_bin install_include install_lib
-.PHONY: install_html install_man install_doc install
+.PHONY: install_doc_html install_doc_man install_doc install
 .PHONY: tests check clean distclean relclean

-.SECONDARY : $(CTESTOBJS)
+.SECONDARY : $(TESTS_OBJS)

 # Default target.
-all: build
+all: build_lib

 dist: build_doc

@@ -138,30 +188,51 @@ build_doc: $(DOCS)
 # Include generated dependency files.
 #
 ifdef CC_MM
-include $(COBJS:%.$(O)=%.d)
-include $(CPICOBJS:%.$(O)=%.d)
-include $(CTESTOBJS:%.$(O)=%.d)
+-include $(C_OBJS:%.$(O)=%.d)
+-include $(C_PIC_OBJS:%.$(O)=%.d)
+-include $(C_JET_OBJS:%.$(O)=%.d)
+-include $(C_TESTLIB_OBJS:%.$(O)=%.d)
+-include $(TESTS_OBJS:%.$(O)=%.d)
 endif

-$(COBJS): $(objroot)src/%.$(O): $(srcroot)src/%.c
-$(CPICOBJS): $(objroot)src/%.pic.$(O): $(srcroot)src/%.c
-$(CPICOBJS): CFLAGS += $(PIC_CFLAGS)
-$(CTESTOBJS): $(objroot)test/%.$(O): $(srcroot)test/%.c
-$(CTESTOBJS): CPPFLAGS += -I$(objroot)test
+$(C_OBJS): $(objroot)src/%.$(O): $(srcroot)src/%.c
+$(C_PIC_OBJS): $(objroot)src/%.pic.$(O): $(srcroot)src/%.c
+$(C_PIC_OBJS): CFLAGS += $(PIC_CFLAGS)
+$(C_JET_OBJS): $(objroot)src/%.jet.$(O): $(srcroot)src/%.c
+$(C_JET_OBJS): CFLAGS += -DJEMALLOC_JET
+$(C_TESTLIB_UNIT_OBJS): $(objroot)test/src/%.unit.$(O): $(srcroot)test/src/%.c
+$(C_TESTLIB_UNIT_OBJS): CPPFLAGS += -DJEMALLOC_UNIT_TEST
+$(C_TESTLIB_INTEGRATION_OBJS): $(objroot)test/src/%.integration.$(O): $(srcroot)test/src/%.c
+$(C_TESTLIB_INTEGRATION_OBJS): CPPFLAGS += -DJEMALLOC_INTEGRATION_TEST
+$(C_UTIL_INTEGRATION_OBJS): $(objroot)src/%.integration.$(O): $(srcroot)src/%.c
+$(C_TESTLIB_STRESS_OBJS): $(objroot)test/src/%.stress.$(O): $(srcroot)test/src/%.c
+$(C_TESTLIB_STRESS_OBJS): CPPFLAGS += -DJEMALLOC_STRESS_TEST -DJEMALLOC_STRESS_TESTLIB
+$(C_TESTLIB_OBJS): CPPFLAGS += -I$(srcroot)test/include -I$(objroot)test/include
+$(TESTS_UNIT_OBJS): CPPFLAGS += -DJEMALLOC_UNIT_TEST
+$(TESTS_UNIT_AUX_OBJS): CPPFLAGS += -DJEMALLOC_UNIT_TEST
+define make-unit-link-dep
+$(1): TESTS_UNIT_LINK_OBJS += $(2)
+$(1): $(2)
+endef
+$(foreach test, $(TESTS_UNIT:$(srcroot)test/unit/%.c=$(objroot)test/unit/%$(EXE)), $(eval $(call make-unit-link-dep,$(test),$(filter $(test:%=%_a.$(O)) $(test:%=%_b.$(O)),$(TESTS_UNIT_AUX_OBJS)))))
+$(TESTS_INTEGRATION_OBJS): CPPFLAGS += -DJEMALLOC_INTEGRATION_TEST
+$(TESTS_STRESS_OBJS): CPPFLAGS += -DJEMALLOC_STRESS_TEST
+$(TESTS_OBJS): $(objroot)test/%.$(O): $(srcroot)test/%.c
+$(TESTS_OBJS): CPPFLAGS += -I$(srcroot)test/include -I$(objroot)test/include
 ifneq ($(IMPORTLIB),$(SO))
-$(COBJS): CPPFLAGS += -DDLLEXPORT
+$(C_OBJS): CPPFLAGS += -DDLLEXPORT
 endif

 ifndef CC_MM
-# Dependencies
+# Dependencies.
 HEADER_DIRS = $(srcroot)include/jemalloc/internal \
 	$(objroot)include/jemalloc $(objroot)include/jemalloc/internal
 HEADERS = $(wildcard $(foreach dir,$(HEADER_DIRS),$(dir)/*.h))
-$(COBJS) $(CPICOBJS) $(CTESTOBJS): $(HEADERS)
-$(CTESTOBJS): $(objroot)test/jemalloc_test.h
+$(C_OBJS) $(C_PIC_OBJS) $(C_JET_OBJS) $(C_TESTLIB_OBJS) $(TESTS_OBJS): $(HEADERS)
+$(TESTS_OBJS): $(objroot)test/unit/jemalloc_test.h
 endif

-$(COBJS) $(CPICOBJS) $(CTESTOBJS): %.$(O):
+$(C_OBJS) $(C_PIC_OBJS) $(C_JET_OBJS) $(C_TESTLIB_OBJS) $(TESTS_OBJS): %.$(O):
 	@mkdir -p $(@D)
 	$(CC) $(CFLAGS) -c $(CPPFLAGS) $(CTARGET) $<
 ifdef CC_MM
@@ -174,27 +245,33 @@ ifneq ($(SOREV),$(SO))
 	ln -sf $(<F) $@
 endif

-$(objroot)lib/$(LIBJEMALLOC).$(SOREV) : $(if $(PIC_CFLAGS),$(CPICOBJS),$(COBJS))
+$(objroot)lib/$(LIBJEMALLOC).$(SOREV) : $(if $(PIC_CFLAGS),$(C_PIC_OBJS),$(C_OBJS))
 	@mkdir -p $(@D)
 	$(CC) $(DSO_LDFLAGS) $(call RPATH,$(RPATH_EXTRA)) $(LDTARGET) $+ $(LDFLAGS) $(LIBS) $(EXTRA_LDFLAGS)

-$(objroot)lib/$(LIBJEMALLOC)_pic.$(A) : $(CPICOBJS)
-$(objroot)lib/$(LIBJEMALLOC).$(A) : $(COBJS)
-$(objroot)lib/$(LIBJEMALLOC)_s.$(A) : $(COBJS)
+$(objroot)lib/$(LIBJEMALLOC)_pic.$(A) : $(C_PIC_OBJS)
+$(objroot)lib/$(LIBJEMALLOC).$(A) : $(C_OBJS)
+$(objroot)lib/$(LIBJEMALLOC)_s.$(A) : $(C_OBJS)

 $(STATIC_LIBS):
 	@mkdir -p $(@D)
-	$(MKLIB) $+
+	$(AR) $(ARFLAGS)@AROUT@ $+

-$(objroot)test/bitmap$(EXE): $(objroot)src/bitmap.$(O)
-
-$(objroot)test/%$(EXE): $(objroot)test/%.$(O) $(objroot)src/util.$(O) $(DSOS)
+$(objroot)test/unit/%$(EXE): $(objroot)test/unit/%.$(O) $(TESTS_UNIT_LINK_OBJS) $(C_JET_OBJS) $(C_TESTLIB_UNIT_OBJS)
 	@mkdir -p $(@D)
-	$(CC) $(LDTARGET) $(filter %.$(O),$^) $(call RPATH,$(objroot)lib) $(objroot)lib/$(LIBJEMALLOC).$(IMPORTLIB) $(filter -lpthread,$(LIBS)) $(EXTRA_LDFLAGS)
+	$(CC) $(LDTARGET) $(filter %.$(O),$^) $(call RPATH,$(objroot)lib) $(LDFLAGS) $(filter-out -lm,$(LIBS)) -lm $(EXTRA_LDFLAGS)
+
+$(objroot)test/integration/%$(EXE): $(objroot)test/integration/%.$(O) $(C_TESTLIB_INTEGRATION_OBJS) $(C_UTIL_INTEGRATION_OBJS) $(objroot)lib/$(LIBJEMALLOC).$(IMPORTLIB)
+	@mkdir -p $(@D)
+	$(CC) $(LDTARGET) $(filter %.$(O),$^) $(call RPATH,$(objroot)lib) $(objroot)lib/$(LIBJEMALLOC).$(IMPORTLIB) $(LDFLAGS) $(filter-out -lm,$(filter -lpthread,$(LIBS))) -lm $(EXTRA_LDFLAGS)
+
+$(objroot)test/stress/%$(EXE): $(objroot)test/stress/%.$(O) $(C_JET_OBJS) $(C_TESTLIB_STRESS_OBJS) $(objroot)lib/$(LIBJEMALLOC).$(IMPORTLIB)
+	@mkdir -p $(@D)
+	$(CC) $(LDTARGET) $(filter %.$(O),$^) $(call RPATH,$(objroot)lib) $(objroot)lib/$(LIBJEMALLOC).$(IMPORTLIB) $(LDFLAGS) $(filter-out -lm,$(LIBS)) -lm $(EXTRA_LDFLAGS)

 build_lib_shared: $(DSOS)
 build_lib_static: $(STATIC_LIBS)
-build: build_lib_shared build_lib_static
+build_lib: build_lib_shared build_lib_static

 install_bin:
 	install -d $(BINDIR)
@@ -205,7 +282,7 @@ done

 install_include:
 	install -d $(INCLUDEDIR)/jemalloc
-	@for h in $(CHDRS); do \
+	@for h in $(C_HDRS); do \
 	echo "install -m 644 $$h $(INCLUDEDIR)/jemalloc"; \
 	install -m 644 $$h $(INCLUDEDIR)/jemalloc; \
 done
@@ -244,49 +321,87 @@ install_doc: install_doc_html install_doc_man

 install: install_bin install_include install_lib install_doc

-tests: $(CTESTS:$(srcroot)%.c=$(objroot)%$(EXE))
+tests_unit: $(TESTS_UNIT:$(srcroot)%.c=$(objroot)%$(EXE))
+tests_integration: $(TESTS_INTEGRATION:$(srcroot)%.c=$(objroot)%$(EXE))
+tests_stress: $(TESTS_STRESS:$(srcroot)%.c=$(objroot)%$(EXE))
+tests: tests_unit tests_integration tests_stress

-check: tests
-	@mkdir -p $(objroot)test
-	@$(SHELL) -c 'total=0; \
-		failures=0; \
-		echo "========================================="; \
-		for t in $(CTESTS:$(srcroot)%.c=$(objroot)%); do \
-			total=`expr $$total + 1`; \
-			/bin/echo -n "$${t} ... "; \
-			$(TEST_LIBRARY_PATH) $${t}$(EXE) $(abs_srcroot) \
-			  $(abs_objroot) > $(objroot)$${t}.out 2>&1; \
-			if test -e "$(srcroot)$${t}.exp"; then \
-				diff -w -u $(srcroot)$${t}.exp \
-				  $(objroot)$${t}.out >/dev/null 2>&1; \
-				fail=$$?; \
-				if test "$${fail}" -eq "1" ; then \
-					failures=`expr $${failures} + 1`; \
-					echo "*** FAIL ***"; \
-				else \
-					echo "pass"; \
-				fi; \
-			else \
-				echo "*** FAIL *** (.exp file is missing)"; \
-				failures=`expr $${failures} + 1`; \
-			fi; \
-		done; \
-		echo "========================================="; \
-		echo "Failures: $${failures}/$${total}"'
+check_unit_dir:
+	@mkdir -p $(objroot)test/unit
+check_integration_dir:
+	@mkdir -p $(objroot)test/integration
+check_stress_dir:
+	@mkdir -p $(objroot)test/stress
+check_dir: check_unit_dir check_integration_dir check_stress_dir
+
+check_unit: tests_unit check_unit_dir
+	$(SHELL) $(objroot)test/test.sh $(TESTS_UNIT:$(srcroot)%.c=$(objroot)%)
+check_integration: tests_integration check_integration_dir
+	$(SHELL) $(objroot)test/test.sh $(TESTS_INTEGRATION:$(srcroot)%.c=$(objroot)%)
+check_stress: tests_stress check_stress_dir
+	$(SHELL) $(objroot)test/test.sh $(TESTS_STRESS:$(srcroot)%.c=$(objroot)%)
+check: tests check_dir
+	$(SHELL) $(objroot)test/test.sh $(TESTS:$(srcroot)%.c=$(objroot)%)
+
+ifeq ($(enable_code_coverage), 1)
+coverage_unit: check_unit
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)src jet $(C_JET_OBJS)
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)test/src unit $(C_TESTLIB_UNIT_OBJS)
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)test/unit unit $(TESTS_UNIT_OBJS)
+
+coverage_integration: check_integration
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)src pic $(C_PIC_OBJS)
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)src integration $(C_UTIL_INTEGRATION_OBJS)
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)test/src integration $(C_TESTLIB_INTEGRATION_OBJS)
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)test/integration integration $(TESTS_INTEGRATION_OBJS)
+
+coverage_stress: check_stress
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)src pic $(C_PIC_OBJS)
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)src jet $(C_JET_OBJS)
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)test/src stress $(C_TESTLIB_STRESS_OBJS)
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)test/stress stress $(TESTS_STRESS_OBJS)
+
+coverage: check
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)src pic $(C_PIC_OBJS)
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)src jet $(C_JET_OBJS)
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)src integration $(C_UTIL_INTEGRATION_OBJS)
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)test/src unit $(C_TESTLIB_UNIT_OBJS)
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)test/src integration $(C_TESTLIB_INTEGRATION_OBJS)
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)test/src stress $(C_TESTLIB_STRESS_OBJS)
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)test/unit unit $(TESTS_UNIT_OBJS) $(TESTS_UNIT_AUX_OBJS)
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)test/integration integration $(TESTS_INTEGRATION_OBJS)
+	$(SHELL) $(srcroot)coverage.sh $(srcroot)test/stress integration $(TESTS_STRESS_OBJS)
+endif

 clean:
-	rm -f $(COBJS)
-	rm -f $(CPICOBJS)
-	rm -f $(COBJS:%.$(O)=%.d)
-	rm -f $(CPICOBJS:%.$(O)=%.d)
-	rm -f $(CTESTOBJS:%.$(O)=%$(EXE))
-	rm -f $(CTESTOBJS)
-	rm -f $(CTESTOBJS:%.$(O)=%.d)
-	rm -f $(CTESTOBJS:%.$(O)=%.out)
+	rm -f $(C_OBJS)
+	rm -f $(C_PIC_OBJS)
+	rm -f $(C_JET_OBJS)
+	rm -f $(C_TESTLIB_OBJS)
+	rm -f $(C_OBJS:%.$(O)=%.d)
+	rm -f $(C_OBJS:%.$(O)=%.gcda)
+	rm -f $(C_OBJS:%.$(O)=%.gcno)
+	rm -f $(C_PIC_OBJS:%.$(O)=%.d)
+	rm -f $(C_PIC_OBJS:%.$(O)=%.gcda)
+	rm -f $(C_PIC_OBJS:%.$(O)=%.gcno)
+	rm -f $(C_JET_OBJS:%.$(O)=%.d)
+	rm -f $(C_JET_OBJS:%.$(O)=%.gcda)
+	rm -f $(C_JET_OBJS:%.$(O)=%.gcno)
+	rm -f $(C_TESTLIB_OBJS:%.$(O)=%.d)
+	rm -f $(C_TESTLIB_OBJS:%.$(O)=%.gcda)
+	rm -f $(C_TESTLIB_OBJS:%.$(O)=%.gcno)
+	rm -f $(TESTS_OBJS:%.$(O)=%$(EXE))
+	rm -f $(TESTS_OBJS)
+	rm -f $(TESTS_OBJS:%.$(O)=%.d)
+	rm -f $(TESTS_OBJS:%.$(O)=%.gcda)
+	rm -f $(TESTS_OBJS:%.$(O)=%.gcno)
+	rm -f $(TESTS_OBJS:%.$(O)=%.out)
 	rm -f $(DSOS) $(STATIC_LIBS)
+	rm -f $(objroot)*.gcov.*

 distclean: clean
 	rm -rf $(objroot)autom4te.cache
+	rm -f $(objroot)bin/jemalloc.sh
 	rm -f $(objroot)config.log
 	rm -f $(objroot)config.status
 	rm -f $(objroot)config.stamp
@@ -1,10 +1,14 @@
-jemalloc is a general-purpose scalable concurrent malloc(3) implementation.
-This distribution is a "portable" implementation that currently targets
-FreeBSD, Linux, Apple OS X, and MinGW.  jemalloc is included as the default
-allocator in the FreeBSD and NetBSD operating systems, and it is used by the
-Mozilla Firefox web browser on Microsoft Windows-related platforms.  Depending
-on your needs, one of the other divergent versions may suit your needs better
-than this distribution.
+jemalloc is a general purpose malloc(3) implementation that emphasizes
+fragmentation avoidance and scalable concurrency support.  jemalloc first came
+into use as the FreeBSD libc allocator in 2005, and since then it has found its
+way into numerous applications that rely on its predictable behavior.  In 2010
+jemalloc development efforts broadened to include developer support features
+such as heap profiling, Valgrind integration, and extensive monitoring/tuning
+hooks.  Modern jemalloc releases continue to be integrated back into FreeBSD,
+and therefore versatility remains critical.  Ongoing development efforts trend
+toward making jemalloc among the best allocators for a broad range of demanding
+applications, and eliminating/mitigating weaknesses that have practical
+repercussions for real world applications.

 The COPYING file contains copyright and licensing information.

@@ -1 +1 @@
-3.2.0-0-g87499f6748ebe4817571e817e9f680ccb5bf54a9
+3.6.0-0-g46c0af68bd248b04df75e4f92d5fb804c3d75340
@@ -4197,8 +4197,12 @@ sub FindLibrary {
 # For libc libraries, the copy in /usr/lib/debug contains debugging symbols
 sub DebuggingLibrary {
  my $file = shift;
-  if ($file =~ m|^/| && -f "/usr/lib/debug$file") {
-    return "/usr/lib/debug$file";
+  if ($file =~ m|^/|) {
+      if (-f "/usr/lib/debug$file") {
+        return "/usr/lib/debug$file";
+      } elsif (-f "/usr/lib/debug$file.debug") {
+        return "/usr/lib/debug$file.debug";
+      }
  }
  return undef;
 }
@@ -4360,6 +4364,19 @@ sub ParseLibraries {
      $finish = HexExtend($2);
      $offset = $zero_offset;
      $lib = $3;
+    }
+    # FreeBSD 10.0 virtual memory map /proc/curproc/map as defined in
+    # function procfs_doprocmap (sys/fs/procfs/procfs_map.c)
+    #
+    # Example:
+    # 0x800600000 0x80061a000 26 0 0xfffff800035a0000 r-x 75 33 0x1004 COW NC vnode /libexec/ld-elf.s
+    # o.1 NCH -1
+    elsif ($l =~ /^(0x$h)\s(0x$h)\s\d+\s\d+\s0x$h\sr-x\s\d+\s\d+\s0x\d+\s(COW|NCO)\s(NC|NNC)\svnode\s(\S+\.so(\.\d+)*)/) {
+      $start = HexExtend($1);
+      $finish = HexExtend($2);
+      $offset = $zero_offset;
+      $lib = FindLibrary($5);
+
    } else {
      next;
    }
@@ -4382,6 +4399,7 @@ sub ParseLibraries {
      }
    }

+    if($main::opt_debug) { printf STDERR "$start:$finish ($offset) $lib\n"; }
    push(@{$result}, [$lib, $start, $finish, $offset]);
  }

@@ -4589,6 +4607,12 @@ sub ExtractSymbols {
    my $finish = $lib->[2];
    my $offset = $lib->[3];

+    # Use debug library if it exists
+    my $debug_libname = DebuggingLibrary($libname);
+    if ($debug_libname) {
+        $libname = $debug_libname;
+    }
+
    # Get list of pcs that belong in this library.
    my $contained = [];
    my ($start_pc_index, $finish_pc_index);
@@ -5019,7 +5043,7 @@ sub GetProcedureBoundariesViaNm {

      # Tag this routine with the starting address in case the image
      # has multiple occurrences of this routine.  We use a syntax
-      # that resembles template paramters that are automatically
+      # that resembles template parameters that are automatically
      # stripped out by ShortFunctionName()
      $this_routine .= "<$start_val>";

@@ -1,14 +1,12 @@
 #! /bin/sh
 # Attempt to guess a canonical system name.
-#   Copyright (C) 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
-#   2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010,
-#   2011, 2012 Free Software Foundation, Inc.
+#   Copyright 1992-2013 Free Software Foundation, Inc.

-timestamp='2012-02-10'
+timestamp='2013-06-10'

 # This file is free software; you can redistribute it and/or modify it
 # under the terms of the GNU General Public License as published by
-# the Free Software Foundation; either version 2 of the License, or
+# the Free Software Foundation; either version 3 of the License, or
 # (at your option) any later version.
 #
 # This program is distributed in the hope that it will be useful, but
@@ -22,19 +20,17 @@ timestamp='2012-02-10'
 # As a special exception to the GNU General Public License, if you
 # distribute this file as part of a program that contains a
 # configuration script generated by Autoconf, you may include it under
-# the same distribution terms that you use for the rest of that program.
-
-
-# Originally written by Per Bothner.  Please send patches (context
-# diff format) to <config-patches@gnu.org> and include a ChangeLog
-# entry.
+# the same distribution terms that you use for the rest of that
+# program.  This Exception is an additional permission under section 7
+# of the GNU General Public License, version 3 ("GPLv3").
 #
-# This script attempts to guess a canonical system name similar to
-# config.sub.  If it succeeds, it prints the system name on stdout, and
-# exits with 0.  Otherwise, it exits with 1.
+# Originally written by Per Bothner.
 #
 # You can get the latest version of this script from:
 # http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.guess;hb=HEAD
+#
+# Please send patches with a ChangeLog entry to config-patches@gnu.org.
+

 me=`echo "$0" | sed -e 's,.*/,,'`

@@ -54,9 +50,7 @@ version="\
 GNU config.guess ($timestamp)

 Originally written by Per Bothner.
-Copyright (C) 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000,
-2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012
-Free Software Foundation, Inc.
+Copyright 1992-2013 Free Software Foundation, Inc.

 This is free software; see the source for copying conditions.  There is NO
 warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE."
@@ -138,6 +132,27 @@ UNAME_RELEASE=`(uname -r) 2>/dev/null` || UNAME_RELEASE=unknown
 UNAME_SYSTEM=`(uname -s) 2>/dev/null`  || UNAME_SYSTEM=unknown
 UNAME_VERSION=`(uname -v) 2>/dev/null` || UNAME_VERSION=unknown

+case "${UNAME_SYSTEM}" in
+Linux|GNU|GNU/*)
+	# If the system lacks a compiler, then just pick glibc.
+	# We could probably try harder.
+	LIBC=gnu
+
+	eval $set_cc_for_build
+	cat <<-EOF > $dummy.c
+	#include <features.h>
+	#if defined(__UCLIBC__)
+	LIBC=uclibc
+	#elif defined(__dietlibc__)
+	LIBC=dietlibc
+	#else
+	LIBC=gnu
+	#endif
+	EOF
+	eval `$CC_FOR_BUILD -E $dummy.c 2>/dev/null | grep '^LIBC'`
+	;;
+esac
+
 # Note: order is significant - the case branches are not exclusive.

 case "${UNAME_MACHINE}:${UNAME_SYSTEM}:${UNAME_RELEASE}:${UNAME_VERSION}" in
@@ -200,6 +215,10 @@ case "${UNAME_MACHINE}:${UNAME_SYSTEM}:${UNAME_RELEASE}:${UNAME_VERSION}" in
 	# CPU_TYPE-MANUFACTURER-OPERATING_SYSTEM is used.
 	echo "${machine}-${os}${release}"
 	exit ;;
+    *:Bitrig:*:*)
+	UNAME_MACHINE_ARCH=`arch | sed 's/Bitrig.//'`
+	echo ${UNAME_MACHINE_ARCH}-unknown-bitrig${UNAME_RELEASE}
+	exit ;;
    *:OpenBSD:*:*)
 	UNAME_MACHINE_ARCH=`arch | sed 's/OpenBSD.//'`
 	echo ${UNAME_MACHINE_ARCH}-unknown-openbsd${UNAME_RELEASE}
@@ -302,7 +321,7 @@ case "${UNAME_MACHINE}:${UNAME_SYSTEM}:${UNAME_RELEASE}:${UNAME_VERSION}" in
    arm:RISC*:1.[012]*:*|arm:riscix:1.[012]*:*)
 	echo arm-acorn-riscix${UNAME_RELEASE}
 	exit ;;
-    arm:riscos:*:*|arm:RISCOS:*:*)
+    arm*:riscos:*:*|arm*:RISCOS:*:*)
 	echo arm-unknown-riscos
 	exit ;;
    SR2?01:HI-UX/MPP:*:* | SR8000:HI-UX/MPP:*:*)
@@ -801,6 +820,9 @@ EOF
    i*:CYGWIN*:*)
 	echo ${UNAME_MACHINE}-pc-cygwin
 	exit ;;
+    *:MINGW64*:*)
+	echo ${UNAME_MACHINE}-pc-mingw64
+	exit ;;
    *:MINGW*:*)
 	echo ${UNAME_MACHINE}-pc-mingw32
 	exit ;;
@@ -852,21 +874,21 @@ EOF
 	exit ;;
    *:GNU:*:*)
 	# the GNU system
-	echo `echo ${UNAME_MACHINE}|sed -e 's,[-/].*$,,'`-unknown-gnu`echo ${UNAME_RELEASE}|sed -e 's,/.*$,,'`
+	echo `echo ${UNAME_MACHINE}|sed -e 's,[-/].*$,,'`-unknown-${LIBC}`echo ${UNAME_RELEASE}|sed -e 's,/.*$,,'`
 	exit ;;
    *:GNU/*:*:*)
 	# other systems with GNU libc and userland
-	echo ${UNAME_MACHINE}-unknown-`echo ${UNAME_SYSTEM} | sed 's,^[^/]*/,,' | tr '[A-Z]' '[a-z]'``echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'`-gnu
+	echo ${UNAME_MACHINE}-unknown-`echo ${UNAME_SYSTEM} | sed 's,^[^/]*/,,' | tr '[A-Z]' '[a-z]'``echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'`-${LIBC}
 	exit ;;
    i*86:Minix:*:*)
 	echo ${UNAME_MACHINE}-pc-minix
 	exit ;;
    aarch64:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-gnu
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
 	exit ;;
    aarch64_be:Linux:*:*)
 	UNAME_MACHINE=aarch64_be
-	echo ${UNAME_MACHINE}-unknown-linux-gnu
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
 	exit ;;
    alpha:Linux:*:*)
 	case `sed -n '/^cpu model/s/^.*: \(.*\)/\1/p' < /proc/cpuinfo` in
@@ -879,59 +901,54 @@ EOF
 	  EV68*) UNAME_MACHINE=alphaev68 ;;
 	esac
 	objdump --private-headers /bin/sh | grep -q ld.so.1
-	if test "$?" = 0 ; then LIBC="libc1" ; else LIBC="" ; fi
-	echo ${UNAME_MACHINE}-unknown-linux-gnu${LIBC}
+	if test "$?" = 0 ; then LIBC="gnulibc1" ; fi
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
+    arc:Linux:*:* | arceb:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
 	exit ;;
    arm*:Linux:*:*)
 	eval $set_cc_for_build
 	if echo __ARM_EABI__ | $CC_FOR_BUILD -E - 2>/dev/null \
 	    | grep -q __ARM_EABI__
 	then
-	    echo ${UNAME_MACHINE}-unknown-linux-gnu
+	    echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
 	else
 	    if echo __ARM_PCS_VFP | $CC_FOR_BUILD -E - 2>/dev/null \
 		| grep -q __ARM_PCS_VFP
 	    then
-		echo ${UNAME_MACHINE}-unknown-linux-gnueabi
+		echo ${UNAME_MACHINE}-unknown-linux-${LIBC}eabi
 	    else
-		echo ${UNAME_MACHINE}-unknown-linux-gnueabihf
+		echo ${UNAME_MACHINE}-unknown-linux-${LIBC}eabihf
 	    fi
 	fi
 	exit ;;
    avr32*:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-gnu
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
 	exit ;;
    cris:Linux:*:*)
-	echo ${UNAME_MACHINE}-axis-linux-gnu
+	echo ${UNAME_MACHINE}-axis-linux-${LIBC}
 	exit ;;
    crisv32:Linux:*:*)
-	echo ${UNAME_MACHINE}-axis-linux-gnu
+	echo ${UNAME_MACHINE}-axis-linux-${LIBC}
 	exit ;;
    frv:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-gnu
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
 	exit ;;
    hexagon:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-gnu
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
 	exit ;;
    i*86:Linux:*:*)
-	LIBC=gnu
-	eval $set_cc_for_build
-	sed 's/^	//' << EOF >$dummy.c
-	#ifdef __dietlibc__
-	LIBC=dietlibc
-	#endif
-EOF
-	eval `$CC_FOR_BUILD -E $dummy.c 2>/dev/null | grep '^LIBC'`
-	echo "${UNAME_MACHINE}-pc-linux-${LIBC}"
+	echo ${UNAME_MACHINE}-pc-linux-${LIBC}
 	exit ;;
    ia64:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-gnu
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
 	exit ;;
    m32r*:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-gnu
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
 	exit ;;
    m68*:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-gnu
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
 	exit ;;
    mips:Linux:*:* | mips64:Linux:*:*)
 	eval $set_cc_for_build
@@ -950,54 +967,63 @@ EOF
 	#endif
 EOF
 	eval `$CC_FOR_BUILD -E $dummy.c 2>/dev/null | grep '^CPU'`
-	test x"${CPU}" != x && { echo "${CPU}-unknown-linux-gnu"; exit; }
+	test x"${CPU}" != x && { echo "${CPU}-unknown-linux-${LIBC}"; exit; }
 	;;
+    or1k:Linux:*:*)
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
+	exit ;;
    or32:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-gnu
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
 	exit ;;
    padre:Linux:*:*)
-	echo sparc-unknown-linux-gnu
+	echo sparc-unknown-linux-${LIBC}
 	exit ;;
    parisc64:Linux:*:* | hppa64:Linux:*:*)
-	echo hppa64-unknown-linux-gnu
+	echo hppa64-unknown-linux-${LIBC}
 	exit ;;
    parisc:Linux:*:* | hppa:Linux:*:*)
 	# Look for CPU level
 	case `grep '^cpu[^a-z]*:' /proc/cpuinfo 2>/dev/null | cut -d' ' -f2` in
-	  PA7*) echo hppa1.1-unknown-linux-gnu ;;
-	  PA8*) echo hppa2.0-unknown-linux-gnu ;;
-	  *)    echo hppa-unknown-linux-gnu ;;
+	  PA7*) echo hppa1.1-unknown-linux-${LIBC} ;;
+	  PA8*) echo hppa2.0-unknown-linux-${LIBC} ;;
+	  *)    echo hppa-unknown-linux-${LIBC} ;;
 	esac
 	exit ;;
    ppc64:Linux:*:*)
-	echo powerpc64-unknown-linux-gnu
+	echo powerpc64-unknown-linux-${LIBC}
 	exit ;;
    ppc:Linux:*:*)
-	echo powerpc-unknown-linux-gnu
+	echo powerpc-unknown-linux-${LIBC}
+	exit ;;
+    ppc64le:Linux:*:*)
+	echo powerpc64le-unknown-linux-${LIBC}
+	exit ;;
+    ppcle:Linux:*:*)
+	echo powerpcle-unknown-linux-${LIBC}
 	exit ;;
    s390:Linux:*:* | s390x:Linux:*:*)
-	echo ${UNAME_MACHINE}-ibm-linux
+	echo ${UNAME_MACHINE}-ibm-linux-${LIBC}
 	exit ;;
    sh64*:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-gnu
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
 	exit ;;
    sh*:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-gnu
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
 	exit ;;
    sparc:Linux:*:* | sparc64:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-gnu
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
 	exit ;;
    tile*:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-gnu
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
 	exit ;;
    vax:Linux:*:*)
-	echo ${UNAME_MACHINE}-dec-linux-gnu
+	echo ${UNAME_MACHINE}-dec-linux-${LIBC}
 	exit ;;
    x86_64:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-gnu
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
 	exit ;;
    xtensa*:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-gnu
+	echo ${UNAME_MACHINE}-unknown-linux-${LIBC}
 	exit ;;
    i*86:DYNIX/ptx:4*:*)
 	# ptx 4.0 does uname -s correctly, with DYNIX/ptx in there.
@@ -1201,6 +1227,9 @@ EOF
    BePC:Haiku:*:*)	# Haiku running on Intel PC compatible.
 	echo i586-pc-haiku
 	exit ;;
+    x86_64:Haiku:*:*)
+	echo x86_64-unknown-haiku
+	exit ;;
    SX-4:SUPER-UX:*:*)
 	echo sx4-nec-superux${UNAME_RELEASE}
 	exit ;;
@@ -1227,19 +1256,21 @@ EOF
 	exit ;;
    *:Darwin:*:*)
 	UNAME_PROCESSOR=`uname -p` || UNAME_PROCESSOR=unknown
-	case $UNAME_PROCESSOR in
-	    i386)
-		eval $set_cc_for_build
-		if [ "$CC_FOR_BUILD" != 'no_compiler_found' ]; then
-		  if (echo '#ifdef __LP64__'; echo IS_64BIT_ARCH; echo '#endif') | \
-		      (CCOPTS= $CC_FOR_BUILD -E - 2>/dev/null) | \
-		      grep IS_64BIT_ARCH >/dev/null
-		  then
-		      UNAME_PROCESSOR="x86_64"
-		  fi
-		fi ;;
-	    unknown) UNAME_PROCESSOR=powerpc ;;
-	esac
+	eval $set_cc_for_build
+	if test "$UNAME_PROCESSOR" = unknown ; then
+	    UNAME_PROCESSOR=powerpc
+	fi
+	if [ "$CC_FOR_BUILD" != 'no_compiler_found' ]; then
+	    if (echo '#ifdef __LP64__'; echo IS_64BIT_ARCH; echo '#endif') | \
+		(CCOPTS= $CC_FOR_BUILD -E - 2>/dev/null) | \
+		grep IS_64BIT_ARCH >/dev/null
+	    then
+		case $UNAME_PROCESSOR in
+		    i386) UNAME_PROCESSOR=x86_64 ;;
+		    powerpc) UNAME_PROCESSOR=powerpc64 ;;
+		esac
+	    fi
+	fi
 	echo ${UNAME_PROCESSOR}-apple-darwin${UNAME_RELEASE}
 	exit ;;
    *:procnto*:*:* | *:QNX:[0123456789]*:*)
@@ -1256,7 +1287,7 @@ EOF
    NEO-?:NONSTOP_KERNEL:*:*)
 	echo neo-tandem-nsk${UNAME_RELEASE}
 	exit ;;
-    NSE-?:NONSTOP_KERNEL:*:*)
+    NSE-*:NONSTOP_KERNEL:*:*)
 	echo nse-tandem-nsk${UNAME_RELEASE}
 	exit ;;
    NSR-?:NONSTOP_KERNEL:*:*)
@@ -1330,9 +1361,6 @@ EOF
 	exit ;;
 esac

-#echo '(No uname command or uname output not recognized.)' 1>&2
-#echo "${UNAME_MACHINE}:${UNAME_SYSTEM}:${UNAME_RELEASE}:${UNAME_VERSION}" 1>&2
-
 eval $set_cc_for_build
 cat >$dummy.c <<EOF
 #ifdef _SEQUENT_
@@ -1,24 +1,18 @@
 #! /bin/sh
 # Configuration validation subroutine script.
-#   Copyright (C) 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
-#   2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010,
-#   2011, 2012 Free Software Foundation, Inc.
+#   Copyright 1992-2013 Free Software Foundation, Inc.

-timestamp='2012-02-10'
+timestamp='2013-10-01'

-# This file is (in principle) common to ALL GNU software.
-# The presence of a machine in this file suggests that SOME GNU software
-# can handle that machine.  It does not imply ALL GNU software can.
-#
-# This file is free software; you can redistribute it and/or modify
-# it under the terms of the GNU General Public License as published by
-# the Free Software Foundation; either version 2 of the License, or
+# This file is free software; you can redistribute it and/or modify it
+# under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3 of the License, or
 # (at your option) any later version.
 #
-# This program is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-# GNU General Public License for more details.
+# This program is distributed in the hope that it will be useful, but
+# WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+# General Public License for more details.
 #
 # You should have received a copy of the GNU General Public License
 # along with this program; if not, see <http://www.gnu.org/licenses/>.
@@ -26,11 +20,12 @@ timestamp='2012-02-10'
 # As a special exception to the GNU General Public License, if you
 # distribute this file as part of a program that contains a
 # configuration script generated by Autoconf, you may include it under
-# the same distribution terms that you use for the rest of that program.
+# the same distribution terms that you use for the rest of that
+# program.  This Exception is an additional permission under section 7
+# of the GNU General Public License, version 3 ("GPLv3").


-# Please send patches to <config-patches@gnu.org>.  Submit a context
-# diff and a properly formatted GNU ChangeLog entry.
+# Please send patches with a ChangeLog entry to config-patches@gnu.org.
 #
 # Configuration subroutine to validate and canonicalize a configuration type.
 # Supply the specified configuration type as an argument.
@@ -73,9 +68,7 @@ Report bugs and patches to <config-patches@gnu.org>."
 version="\
 GNU config.sub ($timestamp)

-Copyright (C) 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000,
-2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012
-Free Software Foundation, Inc.
+Copyright 1992-2013 Free Software Foundation, Inc.

 This is free software; see the source for copying conditions.  There is NO
 warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE."
@@ -123,7 +116,7 @@ esac
 maybe_os=`echo $1 | sed 's/^\(.*\)-\([^-]*-[^-]*\)$/\2/'`
 case $maybe_os in
  nto-qnx* | linux-gnu* | linux-android* | linux-dietlibc | linux-newlib* | \
-  linux-uclibc* | uclinux-uclibc* | uclinux-gnu* | kfreebsd*-gnu* | \
+  linux-musl* | linux-uclibc* | uclinux-uclibc* | uclinux-gnu* | kfreebsd*-gnu* | \
  knetbsd*-gnu* | netbsd*-gnu* | \
  kopensolaris*-gnu* | \
  storm-chaos* | os2-emx* | rtmk-nova*)
@@ -156,7 +149,7 @@ case $os in
 	-convergent* | -ncr* | -news | -32* | -3600* | -3100* | -hitachi* |\
 	-c[123]* | -convex* | -sun | -crds | -omron* | -dg | -ultra | -tti* | \
 	-harris | -dolphin | -highlevel | -gould | -cbm | -ns | -masscomp | \
-	-apple | -axis | -knuth | -cray | -microblaze)
+	-apple | -axis | -knuth | -cray | -microblaze*)
 		os=
 		basic_machine=$1
 		;;
@@ -225,6 +218,12 @@ case $os in
 	-isc*)
 		basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'`
 		;;
+	-lynx*178)
+		os=-lynxos178
+		;;
+	-lynx*5)
+		os=-lynxos5
+		;;
 	-lynx*)
 		os=-lynxos
 		;;
@@ -253,10 +252,12 @@ case $basic_machine in
 	| alpha | alphaev[4-8] | alphaev56 | alphaev6[78] | alphapca5[67] \
 	| alpha64 | alpha64ev[4-8] | alpha64ev56 | alpha64ev6[78] | alpha64pca5[67] \
 	| am33_2.0 \
-	| arc | arm | arm[bl]e | arme[lb] | armv[2345] | armv[345][lb] | avr | avr32 \
-        | be32 | be64 \
+	| arc | arceb \
+	| arm | arm[bl]e | arme[lb] | armv[2-8] | armv[3-8][lb] | armv7[arm] \
+	| avr | avr32 \
+	| be32 | be64 \
 	| bfin \
-	| c4x | clipper \
+	| c4x | c8051 | clipper \
 	| d10v | d30v | dlx | dsp16xx \
 	| epiphany \
 	| fido | fr30 | frv \
@@ -264,10 +265,11 @@ case $basic_machine in
 	| hexagon \
 	| i370 | i860 | i960 | ia64 \
 	| ip2k | iq2000 \
+	| k1om \
 	| le32 | le64 \
 	| lm32 \
 	| m32c | m32r | m32rle | m68000 | m68k | m88k \
-	| maxq | mb | microblaze | mcore | mep | metag \
+	| maxq | mb | microblaze | microblazeel | mcore | mep | metag \
 	| mips | mipsbe | mipseb | mipsel | mipsle \
 	| mips16 \
 	| mips64 | mips64el \
@@ -285,16 +287,17 @@ case $basic_machine in
 	| mipsisa64r2 | mipsisa64r2el \
 	| mipsisa64sb1 | mipsisa64sb1el \
 	| mipsisa64sr71k | mipsisa64sr71kel \
+	| mipsr5900 | mipsr5900el \
 	| mipstx39 | mipstx39el \
 	| mn10200 | mn10300 \
 	| moxie \
 	| mt \
 	| msp430 \
 	| nds32 | nds32le | nds32be \
-	| nios | nios2 \
+	| nios | nios2 | nios2eb | nios2el \
 	| ns16k | ns32k \
 	| open8 \
-	| or32 \
+	| or1k | or32 \
 	| pdp10 | pdp11 | pj | pjl \
 	| powerpc | powerpc64 | powerpc64le | powerpcle \
 	| pyramid \
@@ -322,7 +325,7 @@ case $basic_machine in
 	c6x)
 		basic_machine=tic6x-unknown
 		;;
-	m6811 | m68hc11 | m6812 | m68hc12 | m68hcs12x | picochip)
+	m6811 | m68hc11 | m6812 | m68hc12 | m68hcs12x | nvptx | picochip)
 		basic_machine=$basic_machine-unknown
 		os=-none
 		;;
@@ -364,13 +367,13 @@ case $basic_machine in
 	| aarch64-* | aarch64_be-* \
 	| alpha-* | alphaev[4-8]-* | alphaev56-* | alphaev6[78]-* \
 	| alpha64-* | alpha64ev[4-8]-* | alpha64ev56-* | alpha64ev6[78]-* \
-	| alphapca5[67]-* | alpha64pca5[67]-* | arc-* \
+	| alphapca5[67]-* | alpha64pca5[67]-* | arc-* | arceb-* \
 	| arm-*  | armbe-* | armle-* | armeb-* | armv*-* \
 	| avr-* | avr32-* \
 	| be32-* | be64-* \
 	| bfin-* | bs2000-* \
 	| c[123]* | c30-* | [cjt]90-* | c4x-* \
-	| clipper-* | craynv-* | cydra-* \
+	| c8051-* | clipper-* | craynv-* | cydra-* \
 	| d10v-* | d30v-* | dlx-* \
 	| elxsi-* \
 	| f30[01]-* | f700-* | fido-* | fr30-* | frv-* | fx80-* \
@@ -379,11 +382,13 @@ case $basic_machine in
 	| hexagon-* \
 	| i*86-* | i860-* | i960-* | ia64-* \
 	| ip2k-* | iq2000-* \
+	| k1om-* \
 	| le32-* | le64-* \
 	| lm32-* \
 	| m32c-* | m32r-* | m32rle-* \
 	| m68000-* | m680[012346]0-* | m68360-* | m683?2-* | m68k-* \
-	| m88110-* | m88k-* | maxq-* | mcore-* | metag-* | microblaze-* \
+	| m88110-* | m88k-* | maxq-* | mcore-* | metag-* \
+	| microblaze-* | microblazeel-* \
 	| mips-* | mipsbe-* | mipseb-* | mipsel-* | mipsle-* \
 	| mips16-* \
 	| mips64-* | mips64el-* \
@@ -401,12 +406,13 @@ case $basic_machine in
 	| mipsisa64r2-* | mipsisa64r2el-* \
 	| mipsisa64sb1-* | mipsisa64sb1el-* \
 	| mipsisa64sr71k-* | mipsisa64sr71kel-* \
+	| mipsr5900-* | mipsr5900el-* \
 	| mipstx39-* | mipstx39el-* \
 	| mmix-* \
 	| mt-* \
 	| msp430-* \
 	| nds32-* | nds32le-* | nds32be-* \
-	| nios-* | nios2-* \
+	| nios-* | nios2-* | nios2eb-* | nios2el-* \
 	| none-* | np1-* | ns16k-* | ns32k-* \
 	| open8-* \
 	| orion-* \
@@ -782,11 +788,15 @@ case $basic_machine in
 		basic_machine=ns32k-utek
 		os=-sysv
 		;;
-	microblaze)
+	microblaze*)
 		basic_machine=microblaze-xilinx
 		;;
+	mingw64)
+		basic_machine=x86_64-pc
+		os=-mingw64
+		;;
 	mingw32)
-		basic_machine=i386-pc
+		basic_machine=i686-pc
 		os=-mingw32
 		;;
 	mingw32ce)
@@ -822,7 +832,7 @@ case $basic_machine in
 		basic_machine=`echo $basic_machine | sed -e 's/ms1-/mt-/'`
 		;;
 	msys)
-		basic_machine=i386-pc
+		basic_machine=i686-pc
 		os=-msys
 		;;
 	mvs)
@@ -1013,7 +1023,11 @@ case $basic_machine in
 		basic_machine=i586-unknown
 		os=-pw32
 		;;
-	rdos)
+	rdos | rdos64)
+		basic_machine=x86_64-pc
+		os=-rdos
+		;;
+	rdos32)
 		basic_machine=i386-pc
 		os=-rdos
 		;;
@@ -1340,21 +1354,21 @@ case $os in
 	-gnu* | -bsd* | -mach* | -minix* | -genix* | -ultrix* | -irix* \
 	      | -*vms* | -sco* | -esix* | -isc* | -aix* | -cnk* | -sunos | -sunos[34]*\
 	      | -hpux* | -unos* | -osf* | -luna* | -dgux* | -auroraux* | -solaris* \
-	      | -sym* | -kopensolaris* \
+	      | -sym* | -kopensolaris* | -plan9* \
 	      | -amigaos* | -amigados* | -msdos* | -newsos* | -unicos* | -aof* \
 	      | -aos* | -aros* \
 	      | -nindy* | -vxsim* | -vxworks* | -ebmon* | -hms* | -mvs* \
 	      | -clix* | -riscos* | -uniplus* | -iris* | -rtu* | -xenix* \
 	      | -hiux* | -386bsd* | -knetbsd* | -mirbsd* | -netbsd* \
-	      | -openbsd* | -solidbsd* \
+	      | -bitrig* | -openbsd* | -solidbsd* \
 	      | -ekkobsd* | -kfreebsd* | -freebsd* | -riscix* | -lynxos* \
 	      | -bosx* | -nextstep* | -cxux* | -aout* | -elf* | -oabi* \
 	      | -ptx* | -coff* | -ecoff* | -winnt* | -domain* | -vsta* \
 	      | -udi* | -eabi* | -lites* | -ieee* | -go32* | -aux* \
 	      | -chorusos* | -chorusrdb* | -cegcc* \
 	      | -cygwin* | -msys* | -pe* | -psos* | -moss* | -proelf* | -rtems* \
-	      | -mingw32* | -linux-gnu* | -linux-android* \
-	      | -linux-newlib* | -linux-uclibc* \
+	      | -mingw32* | -mingw64* | -linux-gnu* | -linux-android* \
+	      | -linux-newlib* | -linux-musl* | -linux-uclibc* \
 	      | -uxpv* | -beos* | -mpeix* | -udk* \
 	      | -interix* | -uwin* | -mks* | -rhapsody* | -darwin* | -opened* \
 	      | -openstep* | -oskit* | -conix* | -pw32* | -nonstopux* \
@@ -1486,9 +1500,6 @@ case $os in
 	-aros*)
 		os=-aros
 		;;
-	-kaos*)
-		os=-kaos
-		;;
 	-zvmoe)
 		os=-zvmoe
 		;;
@@ -1537,6 +1548,12 @@ case $basic_machine in
 	c4x-* | tic4x-*)
 		os=-coff
 		;;
+	c8051-*)
+		os=-elf
+		;;
+	hexagon-*)
+		os=-elf
+		;;
 	tic54x-*)
 		os=-coff
 		;;
@@ -1577,6 +1594,9 @@ case $basic_machine in
 	mips*-*)
 		os=-elf
 		;;
+	or1k-*)
+		os=-elf
+		;;
 	or32-*)
 		os=-coff
 		;;
@@ -19,7 +19,9 @@ AC_COMPILE_IFELSE([AC_LANG_PROGRAM(
 ]], [[
    return 0;
 ]])],
+              [je_cv_cflags_appended=$1]
              AC_MSG_RESULT([yes]),
+              [je_cv_cflags_appended=]
              AC_MSG_RESULT([no])
              [CFLAGS="${TCFLAGS}"]
 )
@@ -86,7 +88,7 @@ MANDIR=`eval echo $MANDIR`
 AC_SUBST([MANDIR])

 dnl Support for building documentation.
-AC_PATH_PROG([XSLTPROC], [xsltproc], , [$PATH])
+AC_PATH_PROG([XSLTPROC], [xsltproc], [false], [$PATH])
 if test -d "/usr/share/xml/docbook/stylesheet/docbook-xsl" ; then
  DEFAULT_XSLROOT="/usr/share/xml/docbook/stylesheet/docbook-xsl"
 elif test -d "/usr/share/sgml/docbook/xsl-stylesheets" ; then
@@ -128,6 +130,9 @@ if test "x$CFLAGS" = "x" ; then
  no_CFLAGS="yes"
  if test "x$GCC" = "xyes" ; then
    JE_CFLAGS_APPEND([-std=gnu99])
+    if test "x$je_cv_cflags_appended" = "x-std=gnu99" ; then
+      AC_DEFINE_UNQUOTED([JEMALLOC_HAS_RESTRICT])
+    fi
    JE_CFLAGS_APPEND([-Wall])
    JE_CFLAGS_APPEND([-pipe])
    JE_CFLAGS_APPEND([-g3])
@@ -145,6 +150,11 @@ if test "x$EXTRA_CFLAGS" != "x" ; then
 fi
 AC_PROG_CPP

+AC_C_BIGENDIAN([ac_cv_big_endian=1], [ac_cv_big_endian=0])
+if test "x${ac_cv_big_endian}" = "x1" ; then
+  AC_DEFINE_UNQUOTED([JEMALLOC_BIG_ENDIAN], [ ])
+fi
+
 AC_CHECK_SIZEOF([void *])
 if test "x${ac_cv_sizeof_void_p}" = "x8" ; then
  LG_SIZEOF_PTR=3
@@ -193,19 +203,24 @@ CPU_SPINWAIT=""
 case "${host_cpu}" in
  i[[345]]86)
 	;;
-  i686)
-	JE_COMPILABLE([__asm__], [], [[__asm__ volatile("pause"); return 0;]],
-	              [je_cv_asm])
-	if test "x${je_cv_asm}" = "xyes" ; then
+  i686|x86_64)
+	JE_COMPILABLE([pause instruction], [],
+	              [[__asm__ volatile("pause"); return 0;]],
+	              [je_cv_pause])
+	if test "x${je_cv_pause}" = "xyes" ; then
 	    CPU_SPINWAIT='__asm__ volatile("pause")'
 	fi
+	dnl emmintrin.h fails to compile unless MMX, SSE, and SSE2 are
+	dnl supported.
+	JE_COMPILABLE([SSE2 intrinsics], [
+#include <emmintrin.h>
+], [], [je_cv_sse2])
+	if test "x${je_cv_sse2}" = "xyes" ; then
+	  AC_DEFINE_UNQUOTED([HAVE_SSE2], [ ])
+	fi
 	;;
-  x86_64)
-	JE_COMPILABLE([__asm__ syntax], [],
-	              [[__asm__ volatile("pause"); return 0;]], [je_cv_asm])
-	if test "x${je_cv_asm}" = "xyes" ; then
-	    CPU_SPINWAIT='__asm__ volatile("pause")'
-	fi
+  powerpc)
+	AC_DEFINE_UNQUOTED([HAVE_ALTIVEC], [ ])
 	;;
  *)
 	;;
@@ -226,9 +241,15 @@ PIC_CFLAGS='-fPIC -DPIC'
 CTARGET='-o $@'
 LDTARGET='-o $@'
 EXTRA_LDFLAGS=
-MKLIB='ar crus $@'
+ARFLAGS='crus'
+AROUT=' $@'
 CC_MM=1

+AN_MAKEVAR([AR], [AC_PROG_AR])
+AN_PROGRAM([ar], [AC_PROG_AR])
+AC_DEFUN([AC_PROG_AR], [AC_CHECK_TOOL(AR, ar, :)])
+AC_PROG_AR
+
 dnl Platform-specific settings.  abi and RPATH can probably be determined
 dnl programmatically, but doing so is error-prone, which makes it generally
 dnl not worth the trouble.
@@ -250,6 +271,7 @@ case "${host}" in
 	force_tls="0"
 	DSO_LDFLAGS='-shared -Wl,-dylib_install_name,$(@F)'
 	SOREV="${rev}.${so}"
+	sbrk_deprecated="1"
 	;;
  *-*-freebsd*)
 	CFLAGS="$CFLAGS"
@@ -261,6 +283,7 @@ case "${host}" in
 	CFLAGS="$CFLAGS"
 	CPPFLAGS="$CPPFLAGS -D_GNU_SOURCE"
 	abi="elf"
+	AC_DEFINE([JEMALLOC_HAS_ALLOCA_H])
 	AC_DEFINE([JEMALLOC_PURGE_MADVISE_DONTNEED], [ ])
 	AC_DEFINE([JEMALLOC_THREADED_INIT], [ ])
 	JEMALLOC_USABLE_SIZE_CONST=""
@@ -283,6 +306,7 @@ case "${host}" in
  *-*-solaris2*)
 	CFLAGS="$CFLAGS"
 	abi="elf"
+	AC_DEFINE([JEMALLOC_PURGE_MADVISE_FREE], [ ])
 	RPATH='-Wl,-R,$(1)'
 	dnl Solaris needs this for sigwait().
 	CPPFLAGS="$CPPFLAGS -D_POSIX_PTHREAD_SEMANTICS"
@@ -309,7 +333,9 @@ case "${host}" in
 	  EXTRA_LDFLAGS="-link -DEBUG"
 	  CTARGET='-Fo$@'
 	  LDTARGET='-Fe$@'
-	  MKLIB='lib -nologo -out:$@'
+	  AR='lib'
+	  ARFLAGS='-nologo -out:'
+	  AROUT='$@'
 	  CC_MM=
        else
 	  importlib="${so}"
@@ -342,13 +368,10 @@ AC_SUBST([PIC_CFLAGS])
 AC_SUBST([CTARGET])
 AC_SUBST([LDTARGET])
 AC_SUBST([MKLIB])
+AC_SUBST([ARFLAGS])
+AC_SUBST([AROUT])
 AC_SUBST([CC_MM])

-if test "x$abi" != "xpecoff"; then
-  dnl Heap profiling uses the log(3) function.
-  LIBS="$LIBS -lm"
-fi
-
 JE_COMPILABLE([__attribute__ syntax],
              [static __attribute__((unused)) void foo(void){}],
              [],
@@ -402,11 +425,10 @@ AC_SUBST([enable_autogen])

 AC_PROG_INSTALL
 AC_PROG_RANLIB
-AC_PATH_PROG([AR], [ar], , [$PATH])
-AC_PATH_PROG([LD], [ld], , [$PATH])
-AC_PATH_PROG([AUTOCONF], [autoconf], , [$PATH])
+AC_PATH_PROG([LD], [ld], [false], [$PATH])
+AC_PATH_PROG([AUTOCONF], [autoconf], [false], [$PATH])

-public_syms="malloc_conf malloc_message malloc calloc posix_memalign aligned_alloc realloc free malloc_usable_size malloc_stats_print mallctl mallctlnametomib mallctlbymib"
+public_syms="malloc_conf malloc_message malloc calloc posix_memalign aligned_alloc realloc free mallocx rallocx xallocx sallocx dallocx nallocx mallctl mallctlnametomib mallctlbymib malloc_stats_print malloc_usable_size"

 dnl Check for allocator-related functions that should be wrapped.
 AC_CHECK_FUNC([memalign],
@@ -434,18 +456,35 @@ if test "x$enable_experimental" = "x1" ; then
 fi
 AC_SUBST([enable_experimental])

+dnl Do not compute test code coverage by default.
+GCOV_FLAGS=
+AC_ARG_ENABLE([code-coverage],
+  [AS_HELP_STRING([--enable-code-coverage],
+   [Enable code coverage])],
+[if test "x$enable_code_coverage" = "xno" ; then
+  enable_code_coverage="0"
+else
+  enable_code_coverage="1"
+fi
+],
+[enable_code_coverage="0"]
+)
+if test "x$enable_code_coverage" = "x1" ; then
+  deoptimize="no"
+  echo "$CFLAGS $EXTRA_CFLAGS" | grep '\-O' >/dev/null || deoptimize="yes"
+  if test "x${deoptimize}" = "xyes" ; then
+    JE_CFLAGS_APPEND([-O0])
+  fi
+  JE_CFLAGS_APPEND([-fprofile-arcs -ftest-coverage])
+  EXTRA_LDFLAGS="$EXTRA_LDFLAGS -fprofile-arcs -ftest-coverage"
+  AC_DEFINE([JEMALLOC_CODE_COVERAGE], [ ])
+fi
+AC_SUBST([enable_code_coverage])
+
 dnl Perform no name mangling by default.
 AC_ARG_WITH([mangling],
  [AS_HELP_STRING([--with-mangling=<map>], [Mangle symbols in <map>])],
  [mangling_map="$with_mangling"], [mangling_map=""])
-for nm in `echo ${mangling_map} |tr ',' ' '` ; do
-  k="`echo ${nm} |tr ':' ' ' |awk '{print $1}'`"
-  n="je_${k}"
-  m=`echo ${nm} |tr ':' ' ' |awk '{print $2}'`
-  AC_DEFINE_UNQUOTED([${n}], [${m}])
-  dnl Remove key from public_syms so that it isn't redefined later.
-  public_syms=`for sym in ${public_syms}; do echo "${sym}"; done |grep -v "^${k}\$" |tr '\n' ' '`
-done

 dnl Do not prefix public APIs by default.
 AC_ARG_WITH([jemalloc_prefix],
@@ -462,27 +501,23 @@ if test "x$JEMALLOC_PREFIX" != "x" ; then
  AC_DEFINE_UNQUOTED([JEMALLOC_PREFIX], ["$JEMALLOC_PREFIX"])
  AC_DEFINE_UNQUOTED([JEMALLOC_CPREFIX], ["$JEMALLOC_CPREFIX"])
 fi
-dnl Generate macros to rename public symbols.  All public symbols are prefixed
-dnl with je_ in the source code, so these macro definitions are needed even if
-dnl --with-jemalloc-prefix wasn't specified.
-for stem in ${public_syms}; do
-  n="je_${stem}"
-  m="${JEMALLOC_PREFIX}${stem}"
-  AC_DEFINE_UNQUOTED([${n}], [${m}])
-done

-dnl Do not mangle library-private APIs by default.
+AC_ARG_WITH([export],
+  [AS_HELP_STRING([--without-export], [disable exporting jemalloc public APIs])],
+  [if test "x$with_export" = "xno"; then
+  AC_DEFINE([JEMALLOC_EXPORT],[])
+fi]
+)
+
+dnl Mangle library-private APIs.
 AC_ARG_WITH([private_namespace],
  [AS_HELP_STRING([--with-private-namespace=<prefix>], [Prefix to prepend to all library-private APIs])],
-  [JEMALLOC_PRIVATE_NAMESPACE="$with_private_namespace"],
-  [JEMALLOC_PRIVATE_NAMESPACE=""]
+  [JEMALLOC_PRIVATE_NAMESPACE="${with_private_namespace}je_"],
+  [JEMALLOC_PRIVATE_NAMESPACE="je_"]
 )
-AC_DEFINE_UNQUOTED([JEMALLOC_PRIVATE_NAMESPACE], ["$JEMALLOC_PRIVATE_NAMESPACE"])
-if test "x$JEMALLOC_PRIVATE_NAMESPACE" != "x" ; then
-  AC_DEFINE_UNQUOTED([JEMALLOC_N(string_that_no_one_should_want_to_use_as_a_jemalloc_private_namespace_prefix)], [${JEMALLOC_PRIVATE_NAMESPACE}##string_that_no_one_should_want_to_use_as_a_jemalloc_private_namespace_prefix])
-else
-  AC_DEFINE_UNQUOTED([JEMALLOC_N(string_that_no_one_should_want_to_use_as_a_jemalloc_private_namespace_prefix)], [string_that_no_one_should_want_to_use_as_a_jemalloc_private_namespace_prefix])
-fi
+AC_DEFINE_UNQUOTED([JEMALLOC_PRIVATE_NAMESPACE], [$JEMALLOC_PRIVATE_NAMESPACE])
+private_namespace="$JEMALLOC_PRIVATE_NAMESPACE"
+AC_SUBST([private_namespace])

 dnl Do not add suffix to installed files by default.
 AC_ARG_WITH([install_suffix],
@@ -493,37 +528,72 @@ AC_ARG_WITH([install_suffix],
 install_suffix="$INSTALL_SUFFIX"
 AC_SUBST([install_suffix])

+dnl Substitute @je_@ in jemalloc_protos.h.in, primarily to make generation of
+dnl jemalloc_protos_jet.h easy.
+je_="je_"
+AC_SUBST([je_])
+
 cfgoutputs_in="${srcroot}Makefile.in"
 cfgoutputs_in="${cfgoutputs_in} ${srcroot}doc/html.xsl.in"
 cfgoutputs_in="${cfgoutputs_in} ${srcroot}doc/manpages.xsl.in"
 cfgoutputs_in="${cfgoutputs_in} ${srcroot}doc/jemalloc.xml.in"
-cfgoutputs_in="${cfgoutputs_in} ${srcroot}include/jemalloc/jemalloc.h.in"
+cfgoutputs_in="${cfgoutputs_in} ${srcroot}include/jemalloc/jemalloc_macros.h.in"
+cfgoutputs_in="${cfgoutputs_in} ${srcroot}include/jemalloc/jemalloc_protos.h.in"
 cfgoutputs_in="${cfgoutputs_in} ${srcroot}include/jemalloc/internal/jemalloc_internal.h.in"
-cfgoutputs_in="${cfgoutputs_in} ${srcroot}test/jemalloc_test.h.in"
+cfgoutputs_in="${cfgoutputs_in} ${srcroot}test/test.sh.in"
+cfgoutputs_in="${cfgoutputs_in} ${srcroot}test/include/test/jemalloc_test.h.in"

 cfgoutputs_out="Makefile"
 cfgoutputs_out="${cfgoutputs_out} doc/html.xsl"
 cfgoutputs_out="${cfgoutputs_out} doc/manpages.xsl"
-cfgoutputs_out="${cfgoutputs_out} doc/jemalloc${install_suffix}.xml"
-cfgoutputs_out="${cfgoutputs_out} include/jemalloc/jemalloc${install_suffix}.h"
+cfgoutputs_out="${cfgoutputs_out} doc/jemalloc.xml"
+cfgoutputs_out="${cfgoutputs_out} include/jemalloc/jemalloc_macros.h"
+cfgoutputs_out="${cfgoutputs_out} include/jemalloc/jemalloc_protos.h"
 cfgoutputs_out="${cfgoutputs_out} include/jemalloc/internal/jemalloc_internal.h"
-cfgoutputs_out="${cfgoutputs_out} test/jemalloc_test.h"
+cfgoutputs_out="${cfgoutputs_out} test/test.sh"
+cfgoutputs_out="${cfgoutputs_out} test/include/test/jemalloc_test.h"

 cfgoutputs_tup="Makefile"
 cfgoutputs_tup="${cfgoutputs_tup} doc/html.xsl:doc/html.xsl.in"
 cfgoutputs_tup="${cfgoutputs_tup} doc/manpages.xsl:doc/manpages.xsl.in"
-cfgoutputs_tup="${cfgoutputs_tup} doc/jemalloc${install_suffix}.xml:doc/jemalloc.xml.in"
-cfgoutputs_tup="${cfgoutputs_tup} include/jemalloc/jemalloc${install_suffix}.h:include/jemalloc/jemalloc.h.in"
+cfgoutputs_tup="${cfgoutputs_tup} doc/jemalloc.xml:doc/jemalloc.xml.in"
+cfgoutputs_tup="${cfgoutputs_tup} include/jemalloc/jemalloc_macros.h:include/jemalloc/jemalloc_macros.h.in"
+cfgoutputs_tup="${cfgoutputs_tup} include/jemalloc/jemalloc_protos.h:include/jemalloc/jemalloc_protos.h.in"
 cfgoutputs_tup="${cfgoutputs_tup} include/jemalloc/internal/jemalloc_internal.h"
-cfgoutputs_tup="${cfgoutputs_tup} test/jemalloc_test.h:test/jemalloc_test.h.in"
+cfgoutputs_tup="${cfgoutputs_tup} test/test.sh:test/test.sh.in"
+cfgoutputs_tup="${cfgoutputs_tup} test/include/test/jemalloc_test.h:test/include/test/jemalloc_test.h.in"

 cfghdrs_in="${srcroot}include/jemalloc/jemalloc_defs.h.in"
+cfghdrs_in="${cfghdrs_in} ${srcroot}include/jemalloc/internal/jemalloc_internal_defs.h.in"
+cfghdrs_in="${cfghdrs_in} ${srcroot}include/jemalloc/internal/private_namespace.sh"
+cfghdrs_in="${cfghdrs_in} ${srcroot}include/jemalloc/internal/private_unnamespace.sh"
+cfghdrs_in="${cfghdrs_in} ${srcroot}include/jemalloc/internal/private_symbols.txt"
+cfghdrs_in="${cfghdrs_in} ${srcroot}include/jemalloc/internal/public_namespace.sh"
+cfghdrs_in="${cfghdrs_in} ${srcroot}include/jemalloc/internal/public_unnamespace.sh"
 cfghdrs_in="${cfghdrs_in} ${srcroot}include/jemalloc/internal/size_classes.sh"
+cfghdrs_in="${cfghdrs_in} ${srcroot}include/jemalloc/jemalloc_rename.sh"
+cfghdrs_in="${cfghdrs_in} ${srcroot}include/jemalloc/jemalloc_mangle.sh"
+cfghdrs_in="${cfghdrs_in} ${srcroot}include/jemalloc/jemalloc.sh"
+cfghdrs_in="${cfghdrs_in} ${srcroot}test/include/test/jemalloc_test_defs.h.in"

-cfghdrs_out="include/jemalloc/jemalloc_defs${install_suffix}.h"
+cfghdrs_out="include/jemalloc/jemalloc_defs.h"
+cfghdrs_out="${cfghdrs_out} include/jemalloc/jemalloc${install_suffix}.h"
+cfghdrs_out="${cfghdrs_out} include/jemalloc/internal/private_namespace.h"
+cfghdrs_out="${cfghdrs_out} include/jemalloc/internal/private_unnamespace.h"
+cfghdrs_out="${cfghdrs_out} include/jemalloc/internal/public_symbols.txt"
+cfghdrs_out="${cfghdrs_out} include/jemalloc/internal/public_namespace.h"
+cfghdrs_out="${cfghdrs_out} include/jemalloc/internal/public_unnamespace.h"
 cfghdrs_out="${cfghdrs_out} include/jemalloc/internal/size_classes.h"
+cfghdrs_out="${cfghdrs_out} include/jemalloc/jemalloc_protos_jet.h"
+cfghdrs_out="${cfghdrs_out} include/jemalloc/jemalloc_rename.h"
+cfghdrs_out="${cfghdrs_out} include/jemalloc/jemalloc_mangle.h"
+cfghdrs_out="${cfghdrs_out} include/jemalloc/jemalloc_mangle_jet.h"
+cfghdrs_out="${cfghdrs_out} include/jemalloc/internal/jemalloc_internal_defs.h"
+cfghdrs_out="${cfghdrs_out} test/include/test/jemalloc_test_defs.h"

-cfghdrs_tup="include/jemalloc/jemalloc_defs${install_suffix}.h:include/jemalloc/jemalloc_defs.h.in"
+cfghdrs_tup="include/jemalloc/jemalloc_defs.h:include/jemalloc/jemalloc_defs.h.in"
+cfghdrs_tup="${cfghdrs_tup} include/jemalloc/internal/jemalloc_internal_defs.h:${srcroot}include/jemalloc/internal/jemalloc_internal_defs.h.in"
+cfghdrs_tup="${cfghdrs_tup} test/include/test/jemalloc_test_defs.h:${srcroot}test/include/test/jemalloc_test_defs.h.in"

 dnl Do not silence irrelevant compiler warnings by default, since enabling this
 dnl option incurs a performance penalty.
@@ -544,7 +614,7 @@ fi

 dnl Do not compile with debugging by default.
 AC_ARG_ENABLE([debug],
-  [AS_HELP_STRING([--enable-debug], [Build debugging code])],
+  [AS_HELP_STRING([--enable-debug], [Build debugging code (implies --enable-ivsalloc)])],
 [if test "x$enable_debug" = "xno" ; then
  enable_debug="0"
 else
@@ -555,15 +625,30 @@ fi
 )
 if test "x$enable_debug" = "x1" ; then
  AC_DEFINE([JEMALLOC_DEBUG], [ ])
-  AC_DEFINE([JEMALLOC_IVSALLOC], [ ])
+  enable_ivsalloc="1"
 fi
 AC_SUBST([enable_debug])

+dnl Do not validate pointers by default.
+AC_ARG_ENABLE([ivsalloc],
+  [AS_HELP_STRING([--enable-ivsalloc], [Validate pointers passed through the public API])],
+[if test "x$enable_ivsalloc" = "xno" ; then
+  enable_ivsalloc="0"
+else
+  enable_ivsalloc="1"
+fi
+],
+[enable_ivsalloc="0"]
+)
+if test "x$enable_ivsalloc" = "x1" ; then
+  AC_DEFINE([JEMALLOC_IVSALLOC], [ ])
+fi
+
 dnl Only optimize if not debugging.
 if test "x$enable_debug" = "x0" -a "x$no_CFLAGS" = "xyes" ; then
  dnl Make sure that an optimization flag was not specified in EXTRA_CFLAGS.
  optimize="no"
-  echo "$EXTRA_CFLAGS" | grep "\-O" >/dev/null || optimize="yes"
+  echo "$CFLAGS $EXTRA_CFLAGS" | grep '\-O' >/dev/null || optimize="yes"
  if test "x${optimize}" = "xyes" ; then
    if test "x$GCC" = "xyes" ; then
      JE_CFLAGS_APPEND([-O3])
@@ -662,22 +747,6 @@ if test "x$backtrace_method" = "x" -a "x$enable_prof_libgcc" = "x1" \
     -a "x$GCC" = "xyes" ; then
  AC_CHECK_HEADERS([unwind.h], , [enable_prof_libgcc="0"])
  AC_CHECK_LIB([gcc], [_Unwind_Backtrace], [LIBS="$LIBS -lgcc"], [enable_prof_libgcc="0"])
-  dnl The following is conservative, in that it only has entries for CPUs on
-  dnl which jemalloc has been tested.
-  AC_MSG_CHECKING([libgcc-based backtracing reliability on ${host_cpu}])
-  case "${host_cpu}" in
-    i[[3456]]86)
-      AC_MSG_RESULT([unreliable])
-      enable_prof_libgcc="0";
-      ;;
-    x86_64)
-      AC_MSG_RESULT([reliable])
-      ;;
-    *)
-      AC_MSG_RESULT([unreliable])
-      enable_prof_libgcc="0";
-      ;;
-  esac
  if test "x${enable_prof_libgcc}" = "x1" ; then
    backtrace_method="libgcc"
    AC_DEFINE([JEMALLOC_PROF_LIBGCC], [ ])
@@ -699,6 +768,7 @@ fi
 )
 if test "x$backtrace_method" = "x" -a "x$enable_prof_gcc" = "x1" \
     -a "x$GCC" = "xyes" ; then
+  JE_CFLAGS_APPEND([-fno-omit-frame-pointer])
  backtrace_method="gcc intrinsics"
  AC_DEFINE([JEMALLOC_PROF_GCC], [ ])
 else
@@ -716,6 +786,12 @@ if test "x$enable_prof" = "x1" ; then
    AC_MSG_ERROR([Heap profiling requires TLS]);
  fi
  force_tls="1"
+
+  if test "x$abi" != "xpecoff"; then
+    dnl Heap profiling uses the log(3) function.
+    LIBS="$LIBS -lm"
+  fi
+
  AC_DEFINE([JEMALLOC_PROF], [ ])
 fi
 AC_SUBST([enable_prof])
@@ -749,7 +825,7 @@ fi
 )
 if test "x$enable_mremap" = "x1" ; then
  JE_COMPILABLE([mremap(...MREMAP_FIXED...)], [
-#define _GNU_SOURCE
+#define	_GNU_SOURCE
 #include <sys/mman.h>
 ], [
 void *p = mremap((void *)0, 0, 0, MREMAP_MAYMOVE|MREMAP_FIXED, (void *)0);
@@ -793,7 +869,12 @@ fi
 dnl Check whether the BSD/SUSv1 sbrk() exists.  If not, disable DSS support.
 AC_CHECK_FUNC([sbrk], [have_sbrk="1"], [have_sbrk="0"])
 if test "x$have_sbrk" = "x1" ; then
-  AC_DEFINE([JEMALLOC_HAVE_SBRK], [ ])
+  if test "x$sbrk_deprecated" == "x1" ; then
+    AC_MSG_RESULT([Disabling dss allocation because sbrk is deprecated])
+    enable_dss="0"
+  else
+    AC_DEFINE([JEMALLOC_HAVE_SBRK], [ ])
+  fi
 else
  enable_dss="0"
 fi
@@ -906,7 +987,7 @@ AC_CACHE_CHECK([STATIC_PAGE_SHIFT],
 #include <stdio.h>
 ]],
 [[
-    long result;
+    int result;
    FILE *f;

 #ifdef _WIN32
@@ -925,7 +1006,7 @@ AC_CACHE_CHECK([STATIC_PAGE_SHIFT],
    if (f == NULL) {
 	return 1;
    }
-    fprintf(f, "%u\n", result);
+    fprintf(f, "%d\n", result);
    fclose(f);

    return 0;
@@ -1070,11 +1151,13 @@ dnl Check for ffsl(3), and fail if not found.  This function exists on all
 dnl platforms that jemalloc currently has a chance of functioning on without
 dnl modification.
 JE_COMPILABLE([a program using ffsl], [
+#include <stdio.h>
 #include <strings.h>
 #include <string.h>
 ], [
 	{
 		int rv = ffsl(0x08);
+		printf("%d\n", rv);
 	}
 ], [je_cv_function_ffsl])
 if test "x${je_cv_function_ffsl}" != "xyes" ; then
@@ -1178,7 +1261,26 @@ fi
 dnl ============================================================================
 dnl Darwin-related configuration.

-if test "x${abi}" = "xmacho" ; then
+AC_ARG_ENABLE([zone-allocator],
+  [AS_HELP_STRING([--disable-zone-allocator],
+                  [Disable zone allocator for Darwin])],
+[if test "x$enable_zone_allocator" = "xno" ; then
+  enable_zone_allocator="0"
+else
+  enable_zone_allocator="1"
+fi
+],
+[if test "x${abi}" = "xmacho"; then
+  enable_zone_allocator="1"
+fi
+]
+)
+AC_SUBST([enable_zone_allocator])
+
+if test "x${enable_zone_allocator}" = "x1" ; then
+  if test "x${abi}" != "xmacho"; then
+    AC_MSG_ERROR([--enable-zone-allocator is only supported on Darwin])
+  fi
  AC_DEFINE([JEMALLOC_IVSALLOC], [ ])
  AC_DEFINE([JEMALLOC_ZONE], [ ])

@@ -1218,9 +1320,102 @@ dnl ============================================================================
 dnl Check for typedefs, structures, and compiler characteristics.
 AC_HEADER_STDBOOL

+dnl ============================================================================
+dnl Define commands that generate output files.
+
+AC_CONFIG_COMMANDS([include/jemalloc/internal/private_namespace.h], [
+  mkdir -p "${objroot}include/jemalloc/internal"
+  "${srcdir}/include/jemalloc/internal/private_namespace.sh" "${srcdir}/include/jemalloc/internal/private_symbols.txt" > "${objroot}include/jemalloc/internal/private_namespace.h"
+], [
+  srcdir="${srcdir}"
+  objroot="${objroot}"
+])
+AC_CONFIG_COMMANDS([include/jemalloc/internal/private_unnamespace.h], [
+  mkdir -p "${objroot}include/jemalloc/internal"
+  "${srcdir}/include/jemalloc/internal/private_unnamespace.sh" "${srcdir}/include/jemalloc/internal/private_symbols.txt" > "${objroot}include/jemalloc/internal/private_unnamespace.h"
+], [
+  srcdir="${srcdir}"
+  objroot="${objroot}"
+])
+AC_CONFIG_COMMANDS([include/jemalloc/internal/public_symbols.txt], [
+  f="${objroot}include/jemalloc/internal/public_symbols.txt"
+  mkdir -p "${objroot}include/jemalloc/internal"
+  cp /dev/null "${f}"
+  for nm in `echo ${mangling_map} |tr ',' ' '` ; do
+    n=`echo ${nm} |tr ':' ' ' |awk '{print $[]1}'`
+    m=`echo ${nm} |tr ':' ' ' |awk '{print $[]2}'`
+    echo "${n}:${m}" >> "${f}"
+    dnl Remove name from public_syms so that it isn't redefined later.
+    public_syms=`for sym in ${public_syms}; do echo "${sym}"; done |grep -v "^${n}\$" |tr '\n' ' '`
+  done
+  for sym in ${public_syms} ; do
+    n="${sym}"
+    m="${JEMALLOC_PREFIX}${sym}"
+    echo "${n}:${m}" >> "${f}"
+  done
+], [
+  srcdir="${srcdir}"
+  objroot="${objroot}"
+  mangling_map="${mangling_map}"
+  public_syms="${public_syms}"
+  JEMALLOC_PREFIX="${JEMALLOC_PREFIX}"
+])
+AC_CONFIG_COMMANDS([include/jemalloc/internal/public_namespace.h], [
+  mkdir -p "${objroot}include/jemalloc/internal"
+  "${srcdir}/include/jemalloc/internal/public_namespace.sh" "${objroot}include/jemalloc/internal/public_symbols.txt" > "${objroot}include/jemalloc/internal/public_namespace.h"
+], [
+  srcdir="${srcdir}"
+  objroot="${objroot}"
+])
+AC_CONFIG_COMMANDS([include/jemalloc/internal/public_unnamespace.h], [
+  mkdir -p "${objroot}include/jemalloc/internal"
+  "${srcdir}/include/jemalloc/internal/public_unnamespace.sh" "${objroot}include/jemalloc/internal/public_symbols.txt" > "${objroot}include/jemalloc/internal/public_unnamespace.h"
+], [
+  srcdir="${srcdir}"
+  objroot="${objroot}"
+])
 AC_CONFIG_COMMANDS([include/jemalloc/internal/size_classes.h], [
-  mkdir -p "include/jemalloc/internal"
+  mkdir -p "${objroot}include/jemalloc/internal"
  "${srcdir}/include/jemalloc/internal/size_classes.sh" > "${objroot}include/jemalloc/internal/size_classes.h"
+], [
+  srcdir="${srcdir}"
+  objroot="${objroot}"
+])
+AC_CONFIG_COMMANDS([include/jemalloc/jemalloc_protos_jet.h], [
+  mkdir -p "${objroot}include/jemalloc"
+  cat "${srcdir}/include/jemalloc/jemalloc_protos.h.in" | sed -e 's/@je_@/jet_/g' > "${objroot}include/jemalloc/jemalloc_protos_jet.h"
+], [
+  srcdir="${srcdir}"
+  objroot="${objroot}"
+])
+AC_CONFIG_COMMANDS([include/jemalloc/jemalloc_rename.h], [
+  mkdir -p "${objroot}include/jemalloc"
+  "${srcdir}/include/jemalloc/jemalloc_rename.sh" "${objroot}include/jemalloc/internal/public_symbols.txt" > "${objroot}include/jemalloc/jemalloc_rename.h"
+], [
+  srcdir="${srcdir}"
+  objroot="${objroot}"
+])
+AC_CONFIG_COMMANDS([include/jemalloc/jemalloc_mangle.h], [
+  mkdir -p "${objroot}include/jemalloc"
+  "${srcdir}/include/jemalloc/jemalloc_mangle.sh" "${objroot}include/jemalloc/internal/public_symbols.txt" je_ > "${objroot}include/jemalloc/jemalloc_mangle.h"
+], [
+  srcdir="${srcdir}"
+  objroot="${objroot}"
+])
+AC_CONFIG_COMMANDS([include/jemalloc/jemalloc_mangle_jet.h], [
+  mkdir -p "${objroot}include/jemalloc"
+  "${srcdir}/include/jemalloc/jemalloc_mangle.sh" "${objroot}include/jemalloc/internal/public_symbols.txt" jet_ > "${objroot}include/jemalloc/jemalloc_mangle_jet.h"
+], [
+  srcdir="${srcdir}"
+  objroot="${objroot}"
+])
+AC_CONFIG_COMMANDS([include/jemalloc/jemalloc.h], [
+  mkdir -p "${objroot}include/jemalloc"
+  "${srcdir}/include/jemalloc/jemalloc.sh" "${objroot}" > "${objroot}include/jemalloc/jemalloc${install_suffix}.h"
+], [
+  srcdir="${srcdir}"
+  objroot="${objroot}"
+  install_suffix="${install_suffix}"
 ])

 dnl Process .in files.
@@ -1230,6 +1425,7 @@ AC_CONFIG_HEADERS([$cfghdrs_tup])

 dnl ============================================================================
 dnl Generate outputs.
+
 AC_CONFIG_FILES([$cfgoutputs_tup config.stamp bin/jemalloc.sh])
 AC_SUBST([cfgoutputs_in])
 AC_SUBST([cfgoutputs_out])
@@ -1245,6 +1441,7 @@ AC_MSG_RESULT([CC                 : ${CC}])
 AC_MSG_RESULT([CPPFLAGS           : ${CPPFLAGS}])
 AC_MSG_RESULT([CFLAGS             : ${CFLAGS}])
 AC_MSG_RESULT([LDFLAGS            : ${LDFLAGS}])
+AC_MSG_RESULT([EXTRA_LDFLAGS      : ${EXTRA_LDFLAGS}])
 AC_MSG_RESULT([LIBS               : ${LIBS}])
 AC_MSG_RESULT([RPATH_EXTRA        : ${RPATH_EXTRA}])
 AC_MSG_RESULT([])
@@ -1271,6 +1468,7 @@ AC_MSG_RESULT([autogen            : ${enable_autogen}])
 AC_MSG_RESULT([experimental       : ${enable_experimental}])
 AC_MSG_RESULT([cc-silence         : ${enable_cc_silence}])
 AC_MSG_RESULT([debug              : ${enable_debug}])
+AC_MSG_RESULT([code-coverage      : ${enable_code_coverage}])
 AC_MSG_RESULT([stats              : ${enable_stats}])
 AC_MSG_RESULT([prof               : ${enable_prof}])
 AC_MSG_RESULT([prof-libunwind     : ${enable_prof_libunwind}])
@@ -0,0 +1,16 @@
+#!/bin/sh
+
+set -e
+
+objdir=$1
+suffix=$2
+shift 2
+objs=$@
+
+gcov -b -p -f -o "${objdir}" ${objs}
+
+# Move gcov outputs so that subsequent gcov invocations won't clobber results
+# for the same sources with different compilation flags.
+for f in `find . -maxdepth 1 -type f -name '*.gcov'` ; do
+  mv "${f}" "${f}.${suffix}"
+done
@@ -1,13 +1,13 @@
 '\" t
 .\"     Title: JEMALLOC
 .\"    Author: Jason Evans
-.\" Generator: DocBook XSL Stylesheets v1.76.1 <http://docbook.sf.net/>
-.\"      Date: 11/09/2012
+.\" Generator: DocBook XSL Stylesheets v1.78.1 <http://docbook.sf.net/>
+.\"      Date: 03/31/2014
 .\"    Manual: User Manual
-.\"    Source: jemalloc 3.2.0-0-g87499f6748ebe4817571e817e9f680ccb5bf54a9
+.\"    Source: jemalloc 3.6.0-0-g46c0af68bd248b04df75e4f92d5fb804c3d75340
 .\"  Language: English
 .\"
-.TH "JEMALLOC" "3" "11/09/2012" "jemalloc 3.2.0-0-g87499f6748eb" "User Manual"
+.TH "JEMALLOC" "3" "03/31/2014" "jemalloc 3.6.0-0-g46c0af68bd24" "User Manual"
 .\" -----------------------------------------------------------------
 .\" * Define some portability stuff
 .\" -----------------------------------------------------------------
@@ -31,7 +31,7 @@
 jemalloc \- general purpose memory allocation functions
 .SH "LIBRARY"
 .PP
-This manual describes jemalloc 3\&.2\&.0\-0\-g87499f6748ebe4817571e817e9f680ccb5bf54a9\&. More information can be found at the
+This manual describes jemalloc 3\&.6\&.0\-0\-g46c0af68bd248b04df75e4f92d5fb804c3d75340\&. More information can be found at the
 \m[blue]\fBjemalloc website\fR\m[]\&\s-2\u[1]\d\s+2\&.
 .SH "SYNOPSIS"
 .sp
@@ -55,16 +55,28 @@ This manual describes jemalloc 3\&.2\&.0\-0\-g87499f6748ebe4817571e817e9f680ccb5
 .HP \w'void\ free('u
 .BI "void free(void\ *" "ptr" ");"
 .SS "Non\-standard API"
-.HP \w'size_t\ malloc_usable_size('u
-.BI "size_t malloc_usable_size(const\ void\ *" "ptr" ");"
-.HP \w'void\ malloc_stats_print('u
-.BI "void malloc_stats_print(void\ " "(*write_cb)" "\ (void\ *,\ const\ char\ *), void\ *" "cbopaque" ", const\ char\ *" "opts" ");"
+.HP \w'void\ *mallocx('u
+.BI "void *mallocx(size_t\ " "size" ", int\ " "flags" ");"
+.HP \w'void\ *rallocx('u
+.BI "void *rallocx(void\ *" "ptr" ", size_t\ " "size" ", int\ " "flags" ");"
+.HP \w'size_t\ xallocx('u
+.BI "size_t xallocx(void\ *" "ptr" ", size_t\ " "size" ", size_t\ " "extra" ", int\ " "flags" ");"
+.HP \w'size_t\ sallocx('u
+.BI "size_t sallocx(void\ *" "ptr" ", int\ " "flags" ");"
+.HP \w'void\ dallocx('u
+.BI "void dallocx(void\ *" "ptr" ", int\ " "flags" ");"
+.HP \w'size_t\ nallocx('u
+.BI "size_t nallocx(size_t\ " "size" ", int\ " "flags" ");"
 .HP \w'int\ mallctl('u
 .BI "int mallctl(const\ char\ *" "name" ", void\ *" "oldp" ", size_t\ *" "oldlenp" ", void\ *" "newp" ", size_t\ " "newlen" ");"
 .HP \w'int\ mallctlnametomib('u
 .BI "int mallctlnametomib(const\ char\ *" "name" ", size_t\ *" "mibp" ", size_t\ *" "miblenp" ");"
 .HP \w'int\ mallctlbymib('u
 .BI "int mallctlbymib(const\ size_t\ *" "mib" ", size_t\ " "miblen" ", void\ *" "oldp" ", size_t\ *" "oldlenp" ", void\ *" "newp" ", size_t\ " "newlen" ");"
+.HP \w'void\ malloc_stats_print('u
+.BI "void malloc_stats_print(void\ " "(*write_cb)" "\ (void\ *,\ const\ char\ *), void\ *" "cbopaque" ", const\ char\ *" "opts" ");"
+.HP \w'size_t\ malloc_usable_size('u
+.BI "size_t malloc_usable_size(const\ void\ *" "ptr" ");"
 .HP \w'void\ (*malloc_message)('u
 .BI "void (*malloc_message)(void\ *" "cbopaque" ", const\ char\ *" "s" ");"
 .PP
@@ -156,36 +168,105 @@ is
 .SS "Non\-standard API"
 .PP
 The
-\fBmalloc_usable_size\fR\fB\fR
-function returns the usable size of the allocation pointed to by
-\fIptr\fR\&. The return value may be larger than the size that was requested during allocation\&. The
-\fBmalloc_usable_size\fR\fB\fR
-function is not a mechanism for in\-place
-\fBrealloc\fR\fB\fR; rather it is provided solely as a tool for introspection purposes\&. Any discrepancy between the requested allocation size and the size reported by
-\fBmalloc_usable_size\fR\fB\fR
-should not be depended on, since such behavior is entirely implementation\-dependent\&.
+\fBmallocx\fR\fB\fR,
+\fBrallocx\fR\fB\fR,
+\fBxallocx\fR\fB\fR,
+\fBsallocx\fR\fB\fR,
+\fBdallocx\fR\fB\fR, and
+\fBnallocx\fR\fB\fR
+functions all have a
+\fIflags\fR
+argument that can be used to specify options\&. The functions only check the options that are contextually relevant\&. Use bitwise or (|) operations to specify one or more of the following:
+.PP
+\fBMALLOCX_LG_ALIGN(\fR\fB\fIla\fR\fR\fB) \fR
+.RS 4
+Align the memory allocation to start at an address that is a multiple of
+(1 << \fIla\fR)\&. This macro does not validate that
+\fIla\fR
+is within the valid range\&.
+.RE
+.PP
+\fBMALLOCX_ALIGN(\fR\fB\fIa\fR\fR\fB) \fR
+.RS 4
+Align the memory allocation to start at an address that is a multiple of
+\fIa\fR, where
+\fIa\fR
+is a power of two\&. This macro does not validate that
+\fIa\fR
+is a power of 2\&.
+.RE
+.PP
+\fBMALLOCX_ZERO\fR
+.RS 4
+Initialize newly allocated memory to contain zero bytes\&. In the growing reallocation case, the real size prior to reallocation defines the boundary between untouched bytes and those that are initialized to contain zero bytes\&. If this macro is absent, newly allocated memory is uninitialized\&.
+.RE
+.PP
+\fBMALLOCX_ARENA(\fR\fB\fIa\fR\fR\fB) \fR
+.RS 4
+Use the arena specified by the index
+\fIa\fR
+(and by necessity bypass the thread cache)\&. This macro has no effect for huge regions, nor for regions that were allocated via an arena other than the one specified\&. This macro does not validate that
+\fIa\fR
+specifies an arena index in the valid range\&.
+.RE
 .PP
 The
-\fBmalloc_stats_print\fR\fB\fR
-function writes human\-readable summary statistics via the
-\fIwrite_cb\fR
-callback function pointer and
-\fIcbopaque\fR
-data passed to
-\fIwrite_cb\fR, or
-\fBmalloc_message\fR\fB\fR
-if
-\fIwrite_cb\fR
+\fBmallocx\fR\fB\fR
+function allocates at least
+\fIsize\fR
+bytes of memory, and returns a pointer to the base address of the allocation\&. Behavior is undefined if
+\fIsize\fR
 is
-\fBNULL\fR\&. This function can be called repeatedly\&. General information that never changes during execution can be omitted by specifying "g" as a character within the
-\fIopts\fR
-string\&. Note that
-\fBmalloc_message\fR\fB\fR
-uses the
-\fBmallctl*\fR\fB\fR
-functions internally, so inconsistent statistics can be reported if multiple threads use these functions simultaneously\&. If
-\fB\-\-enable\-stats\fR
-is specified during configuration, \(lqm\(rq and \(lqa\(rq can be specified to omit merged arena and per arena statistics, respectively; \(lqb\(rq and \(lql\(rq can be specified to omit per size class statistics for bins and large objects, respectively\&. Unrecognized characters are silently ignored\&. Note that thread caching may prevent some statistics from being completely up to date, since extra locking would be required to merge counters that track thread cache operations\&.
+\fB0\fR, or if request size overflows due to size class and/or alignment constraints\&.
+.PP
+The
+\fBrallocx\fR\fB\fR
+function resizes the allocation at
+\fIptr\fR
+to be at least
+\fIsize\fR
+bytes, and returns a pointer to the base address of the resulting allocation, which may or may not have moved from its original location\&. Behavior is undefined if
+\fIsize\fR
+is
+\fB0\fR, or if request size overflows due to size class and/or alignment constraints\&.
+.PP
+The
+\fBxallocx\fR\fB\fR
+function resizes the allocation at
+\fIptr\fR
+in place to be at least
+\fIsize\fR
+bytes, and returns the real size of the allocation\&. If
+\fIextra\fR
+is non\-zero, an attempt is made to resize the allocation to be at least
+(\fIsize\fR + \fIextra\fR)
+bytes, though inability to allocate the extra byte(s) will not by itself result in failure to resize\&. Behavior is undefined if
+\fIsize\fR
+is
+\fB0\fR, or if
+(\fIsize\fR + \fIextra\fR > \fBSIZE_T_MAX\fR)\&.
+.PP
+The
+\fBsallocx\fR\fB\fR
+function returns the real size of the allocation at
+\fIptr\fR\&.
+.PP
+The
+\fBdallocx\fR\fB\fR
+function causes the memory referenced by
+\fIptr\fR
+to be made available for future allocations\&.
+.PP
+The
+\fBnallocx\fR\fB\fR
+function allocates no memory, but it performs the same size computation as the
+\fBmallocx\fR\fB\fR
+function, and returns the real size of the allocation that would result from the equivalent
+\fBmallocx\fR\fB\fR
+function call\&. Behavior is undefined if
+\fIsize\fR
+is
+\fB0\fR, or if request size overflows due to size class and/or alignment constraints\&.
 .PP
 The
 \fBmallctl\fR\fB\fR
@@ -229,15 +310,14 @@ that is smaller than the number of period\-separated name components, which resu
 .\}
 .nf
 unsigned nbins, i;
-
-int mib[4];
+size_t mib[4];
 size_t len, miblen;

 len = sizeof(nbins);
 mallctl("arenas\&.nbins", &nbins, &len, NULL, 0);

 miblen = 4;
-mallnametomib("arenas\&.bin\&.0\&.size", mib, &miblen);
+mallctlnametomib("arenas\&.bin\&.0\&.size", mib, &miblen);
 for (i = 0; i < nbins; i++) {
 	size_t bin_size;

@@ -250,6 +330,38 @@ for (i = 0; i < nbins; i++) {
 .if n \{\
 .RE
 .\}
+.PP
+The
+\fBmalloc_stats_print\fR\fB\fR
+function writes human\-readable summary statistics via the
+\fIwrite_cb\fR
+callback function pointer and
+\fIcbopaque\fR
+data passed to
+\fIwrite_cb\fR, or
+\fBmalloc_message\fR\fB\fR
+if
+\fIwrite_cb\fR
+is
+\fBNULL\fR\&. This function can be called repeatedly\&. General information that never changes during execution can be omitted by specifying "g" as a character within the
+\fIopts\fR
+string\&. Note that
+\fBmalloc_message\fR\fB\fR
+uses the
+\fBmallctl*\fR\fB\fR
+functions internally, so inconsistent statistics can be reported if multiple threads use these functions simultaneously\&. If
+\fB\-\-enable\-stats\fR
+is specified during configuration, \(lqm\(rq and \(lqa\(rq can be specified to omit merged arena and per arena statistics, respectively; \(lqb\(rq and \(lql\(rq can be specified to omit per size class statistics for bins and large objects, respectively\&. Unrecognized characters are silently ignored\&. Note that thread caching may prevent some statistics from being completely up to date, since extra locking would be required to merge counters that track thread cache operations\&.
+.PP
+The
+\fBmalloc_usable_size\fR\fB\fR
+function returns the usable size of the allocation pointed to by
+\fIptr\fR\&. The return value may be larger than the size that was requested during allocation\&. The
+\fBmalloc_usable_size\fR\fB\fR
+function is not a mechanism for in\-place
+\fBrealloc\fR\fB\fR; rather it is provided solely as a tool for introspection purposes\&. Any discrepancy between the requested allocation size and the size reported by
+\fBmalloc_usable_size\fR\fB\fR
+should not be depended on, since such behavior is entirely implementation\-dependent\&.
 .SS "Experimental API"
 .PP
 The experimental API is subject to change or removal without regard for backward compatibility\&. If
@@ -286,7 +398,7 @@ is a power of 2\&.
 .PP
 \fBALLOCM_ZERO\fR
 .RS 4
-Initialize newly allocated memory to contain zero bytes\&. In the growing reallocation case, the real size prior to reallocation defines the boundary between untouched bytes and those that are initialized to contain zero bytes\&. If this option is absent, newly allocated memory is uninitialized\&.
+Initialize newly allocated memory to contain zero bytes\&. In the growing reallocation case, the real size prior to reallocation defines the boundary between untouched bytes and those that are initialized to contain zero bytes\&. If this macro is absent, newly allocated memory is uninitialized\&.
 .RE
 .PP
 \fBALLOCM_NO_MOVE\fR
@@ -297,9 +409,10 @@ For reallocation, fail rather than moving the object\&. This constraint can appl
 \fBALLOCM_ARENA(\fR\fB\fIa\fR\fR\fB) \fR
 .RS 4
 Use the arena specified by the index
-\fIa\fR\&. This macro does not validate that
 \fIa\fR
-specifies an arena in the valid range\&.
+(and by necessity bypass the thread cache)\&. This macro has no effect for huge regions, nor for regions that were allocated via an arena other than the one specified\&. This macro does not validate that
+\fIa\fR
+specifies an arena index in the valid range\&.
 .RE
 .PP
 The
@@ -316,7 +429,7 @@ is not
 \fBNULL\fR\&. Behavior is undefined if
 \fIsize\fR
 is
-\fB0\fR\&.
+\fB0\fR, or if request size overflows due to size class and/or alignment constraints\&.
 .PP
 The
 \fBrallocm\fR\fB\fR
@@ -334,11 +447,11 @@ is not
 \fBNULL\fR\&. If
 \fIextra\fR
 is non\-zero, an attempt is made to resize the allocation to be at least
-\fIsize\fR + \fIextra\fR)
+(\fIsize\fR + \fIextra\fR)
 bytes, though inability to allocate the extra byte(s) will not by itself result in failure\&. Behavior is undefined if
 \fIsize\fR
 is
-\fB0\fR, or if
+\fB0\fR, if request size overflows due to size class and/or alignment constraints, or if
 (\fIsize\fR + \fIextra\fR > \fBSIZE_T_MAX\fR)\&.
 .PP
 The
@@ -368,7 +481,7 @@ to the real size of the allocation that would result from the equivalent
 function call\&. Behavior is undefined if
 \fIsize\fR
 is
-\fB0\fR\&.
+\fB0\fR, or if request size overflows due to size class and/or alignment constraints\&.
 .SH "TUNING"
 .PP
 Once, when the first call is made to one of the memory allocation routines, the allocator initializes its internals based in part on various options that can be specified at compile\- or run\-time\&.
@@ -376,7 +489,19 @@ Once, when the first call is made to one of the memory allocation routines, the
 The string pointed to by the global variable
 \fImalloc_conf\fR, the \(lqname\(rq of the file referenced by the symbolic link named
 /etc/malloc\&.conf, and the value of the environment variable
-\fBMALLOC_CONF\fR, will be interpreted, in that order, from left to right as options\&.
+\fBMALLOC_CONF\fR, will be interpreted, in that order, from left to right as options\&. Note that
+\fImalloc_conf\fR
+may be read before
+\fBmain\fR\fB\fR
+is entered, so the declaration of
+\fImalloc_conf\fR
+should specify an initializer that contains the final value to be read by jemalloc\&.
+\fImalloc_conf\fR
+is a compile\-time setting, whereas
+/etc/malloc\&.conf
+and
+\fBMALLOC_CONF\fR
+can be safely set any time prior to program invocation\&.
 .PP
 An options string is a comma\-separated list of option:value pairs\&. There is one key corresponding to each
 "opt\&.*"
@@ -634,16 +759,18 @@ in these cases\&. This option is disabled by default unless
 is specified during configuration, in which case it is enabled by default\&.
 .RE
 .PP
-"opt\&.lg_chunk" (\fBsize_t\fR) r\-
-.RS 4
-Virtual memory chunk size (log base 2)\&. The default chunk size is 4 MiB (2^22)\&.
-.RE
-.PP
 "opt\&.dss" (\fBconst char *\fR) r\-
 .RS 4
 dss (\fBsbrk\fR(2)) allocation precedence as related to
 \fBmmap\fR(2)
-allocation\&. The following settings are supported: \(lqdisabled\(rq, \(lqprimary\(rq, and \(lqsecondary\(rq (default)\&.
+allocation\&. The following settings are supported: \(lqdisabled\(rq, \(lqprimary\(rq, and \(lqsecondary\(rq\&. The default is \(lqsecondary\(rq if
+"config\&.dss"
+is true, \(lqdisabled\(rq otherwise\&.
+.RE
+.PP
+"opt\&.lg_chunk" (\fBsize_t\fR) r\-
+.RS 4
+Virtual memory chunk size (log base 2)\&. If a chunk size outside the supported size range is specified, the size is silently clipped to the minimum/maximum supported size\&. The default chunk size is 4 MiB (2^22)\&.
 .RE
 .PP
 "opt\&.narenas" (\fBsize_t\fR) r\-
@@ -698,7 +825,8 @@ option is enabled, the redzones are checked for corruption during deallocation\&
 "opt\&.zero" (\fBbool\fR) r\- [\fB\-\-enable\-fill\fR]
 .RS 4
 Zero filling enabled/disabled\&. If enabled, each byte of uninitialized allocated memory will be initialized to 0\&. Note that this initialization only happens once for each byte, so
-\fBrealloc\fR\fB\fR
+\fBrealloc\fR\fB\fR,
+\fBrallocx\fR\fB\fR
 and
 \fBrallocm\fR\fB\fR
 calls do not zero memory that was previously allocated\&. This is intended for debugging and will impact performance negatively\&. This option is disabled by default\&.
@@ -776,7 +904,7 @@ Filename prefix for profile dumps\&. If the prefix is set to the empty string, n
 jeprof\&.
 .RE
 .PP
-"opt\&.prof_active" (\fBbool\fR) r\- [\fB\-\-enable\-prof\fR]
+"opt\&.prof_active" (\fBbool\fR) rw [\fB\-\-enable\-prof\fR]
 .RS 4
 Profiling activated/deactivated\&. This is a secondary control mechanism that makes it possible to start the application with profiling enabled (see the
 "opt\&.prof"
@@ -891,7 +1019,7 @@ Purge unused dirty pages for arena <i>, or for all arenas if <i> equals
 "arena\&.<i>\&.dss" (\fBconst char *\fR) rw
 .RS 4
 Set the precedence of dss allocation as related to mmap allocation for arena <i>, or for all arenas if <i> equals
-"arenas\&.narenas"\&. See
+"arenas\&.narenas"\&. Note that even during huge allocation this setting is read from the arena that would be chosen for small or large allocation so that applications can depend on consistent dss versus mmap allocation regardless of allocation size\&. See
 "opt\&.dss"
 for supported settings\&.
 .RE
@@ -1091,7 +1219,7 @@ Number of
 or similar calls made to purge dirty pages\&.
 .RE
 .PP
-"stats\&.arenas\&.<i>\&.npurged" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
+"stats\&.arenas\&.<i>\&.purged" (\fBuint64_t\fR) r\- [\fB\-\-enable\-stats\fR]
 .RS 4
 Number of pages purged\&.
 .RE
@@ -1314,11 +1442,32 @@ function returns no value\&.
 .SS "Non\-standard API"
 .PP
 The
-\fBmalloc_usable_size\fR\fB\fR
-function returns the usable size of the allocation pointed to by
+\fBmallocx\fR\fB\fR
+and
+\fBrallocx\fR\fB\fR
+functions return a pointer to the allocated memory if successful; otherwise a
+\fBNULL\fR
+pointer is returned to indicate insufficient contiguous memory was available to service the allocation request\&.
+.PP
+The
+\fBxallocx\fR\fB\fR
+function returns the real size of the resulting resized allocation pointed to by
+\fIptr\fR, which is a value less than
+\fIsize\fR
+if the allocation could not be adequately grown in place\&.
+.PP
+The
+\fBsallocx\fR\fB\fR
+function returns the real size of the allocation pointed to by
 \fIptr\fR\&.
 .PP
 The
+\fBnallocx\fR\fB\fR
+returns the real size that would result from a successful equivalent
+\fBmallocx\fR\fB\fR
+function call, or zero if insufficient memory is available to perform the size computation\&.
+.PP
+The
 \fBmallctl\fR\fB\fR,
 \fBmallctlnametomib\fR\fB\fR, and
 \fBmallctlbymib\fR\fB\fR
@@ -1335,12 +1484,6 @@ is too large or too small\&. Alternatively,
 is too large or too small; in this case as much data as possible are read despite the error\&.
 .RE
 .PP
-ENOMEM
-.RS 4
-\fI*oldlenp\fR
-is too short to hold the requested value\&.
-.RE
-.PP
 ENOENT
 .RS 4
 \fIname\fR
@@ -1365,6 +1508,11 @@ An interface with side effects failed in some way not directly related to
 \fBmallctl*\fR\fB\fR
 read/write processing\&.
 .RE
+.PP
+The
+\fBmalloc_usable_size\fR\fB\fR
+function returns the usable size of the allocation pointed to by
+\fIptr\fR\&.
 .SS "Experimental API"
 .PP
 The
@@ -158,6 +158,7 @@ struct arena_chunk_map_s {
 };
 typedef rb_tree(arena_chunk_map_t) arena_avail_tree_t;
 typedef rb_tree(arena_chunk_map_t) arena_run_tree_t;
+typedef ql_head(arena_chunk_map_t) arena_chunk_mapelms_t;

 /* Arena chunk header. */
 struct arena_chunk_s {
@@ -174,11 +175,12 @@ struct arena_chunk_s {
 	size_t			nruns_avail;

 	/*
-	 * Number of available run adjacencies.  Clean and dirty available runs
-	 * are not coalesced, which causes virtual memory fragmentation.  The
-	 * ratio of (nruns_avail-nruns_adjac):nruns_adjac is used for tracking
-	 * this fragmentation.
-	 * */
+	 * Number of available run adjacencies that purging could coalesce.
+	 * Clean and dirty available runs are not coalesced, which causes
+	 * virtual memory fragmentation.  The ratio of
+	 * (nruns_avail-nruns_adjac):nruns_adjac is used for tracking this
+	 * fragmentation.
+	 */
 	size_t			nruns_adjac;

 	/*
@@ -400,12 +402,20 @@ extern arena_bin_info_t	arena_bin_info[NBINS];
 #define			nlclasses (chunk_npages - map_bias)

 void	arena_purge_all(arena_t *arena);
-void	arena_prof_accum(arena_t *arena, uint64_t accumbytes);
 void	arena_tcache_fill_small(arena_t *arena, tcache_bin_t *tbin,
    size_t binind, uint64_t prof_accumbytes);
 void	arena_alloc_junk_small(void *ptr, arena_bin_info_t *bin_info,
    bool zero);
+#ifdef JEMALLOC_JET
+typedef void (arena_redzone_corruption_t)(void *, size_t, bool, size_t,
+    uint8_t);
+extern arena_redzone_corruption_t *arena_redzone_corruption;
+typedef void (arena_dalloc_junk_small_t)(void *, arena_bin_info_t *);
+extern arena_dalloc_junk_small_t *arena_dalloc_junk_small;
+#else
 void	arena_dalloc_junk_small(void *ptr, arena_bin_info_t *bin_info);
+#endif
+void	arena_quarantine_junk_small(void *ptr, size_t usize);
 void	*arena_malloc_small(arena_t *arena, size_t size, bool zero);
 void	*arena_malloc_large(arena_t *arena, size_t size, bool zero);
 void	*arena_palloc(arena_t *arena, size_t size, size_t alignment, bool zero);
@@ -416,10 +426,18 @@ void	arena_dalloc_bin(arena_t *arena, arena_chunk_t *chunk, void *ptr,
    size_t pageind, arena_chunk_map_t *mapelm);
 void	arena_dalloc_small(arena_t *arena, arena_chunk_t *chunk, void *ptr,
    size_t pageind);
+#ifdef JEMALLOC_JET
+typedef void (arena_dalloc_junk_large_t)(void *, size_t);
+extern arena_dalloc_junk_large_t *arena_dalloc_junk_large;
+#endif
 void	arena_dalloc_large_locked(arena_t *arena, arena_chunk_t *chunk,
    void *ptr);
 void	arena_dalloc_large(arena_t *arena, arena_chunk_t *chunk, void *ptr);
-void	*arena_ralloc_no_move(void *ptr, size_t oldsize, size_t size,
+#ifdef JEMALLOC_JET
+typedef void (arena_ralloc_junk_large_t)(void *, size_t, size_t);
+extern arena_ralloc_junk_large_t *arena_ralloc_junk_large;
+#endif
+bool	arena_ralloc_no_move(void *ptr, size_t oldsize, size_t size,
    size_t extra, bool zero);
 void	*arena_ralloc(arena_t *arena, void *ptr, size_t oldsize, size_t size,
    size_t extra, size_t alignment, bool zero, bool try_tcache_alloc,
@@ -442,6 +460,7 @@ void	arena_postfork_child(arena_t *arena);
 #ifndef JEMALLOC_ENABLE_INLINE
 arena_chunk_map_t	*arena_mapp_get(arena_chunk_t *chunk, size_t pageind);
 size_t	*arena_mapbitsp_get(arena_chunk_t *chunk, size_t pageind);
+size_t	arena_mapbitsp_read(size_t *mapbitsp);
 size_t	arena_mapbits_get(arena_chunk_t *chunk, size_t pageind);
 size_t	arena_mapbits_unallocated_size_get(arena_chunk_t *chunk,
    size_t pageind);
@@ -452,6 +471,7 @@ size_t	arena_mapbits_dirty_get(arena_chunk_t *chunk, size_t pageind);
 size_t	arena_mapbits_unzeroed_get(arena_chunk_t *chunk, size_t pageind);
 size_t	arena_mapbits_large_get(arena_chunk_t *chunk, size_t pageind);
 size_t	arena_mapbits_allocated_get(arena_chunk_t *chunk, size_t pageind);
+void	arena_mapbitsp_write(size_t *mapbitsp, size_t mapbits);
 void	arena_mapbits_unallocated_set(arena_chunk_t *chunk, size_t pageind,
    size_t size, size_t flags);
 void	arena_mapbits_unallocated_size_set(arena_chunk_t *chunk, size_t pageind,
@@ -464,12 +484,15 @@ void	arena_mapbits_small_set(arena_chunk_t *chunk, size_t pageind,
    size_t runind, size_t binind, size_t flags);
 void	arena_mapbits_unzeroed_set(arena_chunk_t *chunk, size_t pageind,
    size_t unzeroed);
+bool	arena_prof_accum_impl(arena_t *arena, uint64_t accumbytes);
+bool	arena_prof_accum_locked(arena_t *arena, uint64_t accumbytes);
+bool	arena_prof_accum(arena_t *arena, uint64_t accumbytes);
 size_t	arena_ptr_small_binind_get(const void *ptr, size_t mapbits);
 size_t	arena_bin_index(arena_t *arena, arena_bin_t *bin);
 unsigned	arena_run_regind(arena_run_t *run, arena_bin_info_t *bin_info,
    const void *ptr);
 prof_ctx_t	*arena_prof_ctx_get(const void *ptr);
-void	arena_prof_ctx_set(const void *ptr, prof_ctx_t *ctx);
+void	arena_prof_ctx_set(const void *ptr, size_t usize, prof_ctx_t *ctx);
 void	*arena_malloc(arena_t *arena, size_t size, bool zero, bool try_tcache);
 size_t	arena_salloc(const void *ptr, bool demote);
 void	arena_dalloc(arena_t *arena, arena_chunk_t *chunk, void *ptr,
@@ -478,7 +501,7 @@ void	arena_dalloc(arena_t *arena, arena_chunk_t *chunk, void *ptr,

 #if (defined(JEMALLOC_ENABLE_INLINE) || defined(JEMALLOC_ARENA_C_))
 #  ifdef JEMALLOC_ARENA_INLINE_A
-JEMALLOC_INLINE arena_chunk_map_t *
+JEMALLOC_ALWAYS_INLINE arena_chunk_map_t *
 arena_mapp_get(arena_chunk_t *chunk, size_t pageind)
 {

@@ -488,21 +511,28 @@ arena_mapp_get(arena_chunk_t *chunk, size_t pageind)
 	return (&chunk->map[pageind-map_bias]);
 }

-JEMALLOC_INLINE size_t *
+JEMALLOC_ALWAYS_INLINE size_t *
 arena_mapbitsp_get(arena_chunk_t *chunk, size_t pageind)
 {

 	return (&arena_mapp_get(chunk, pageind)->bits);
 }

-JEMALLOC_INLINE size_t
+JEMALLOC_ALWAYS_INLINE size_t
+arena_mapbitsp_read(size_t *mapbitsp)
+{
+
+	return (*mapbitsp);
+}
+
+JEMALLOC_ALWAYS_INLINE size_t
 arena_mapbits_get(arena_chunk_t *chunk, size_t pageind)
 {

-	return (*arena_mapbitsp_get(chunk, pageind));
+	return (arena_mapbitsp_read(arena_mapbitsp_get(chunk, pageind)));
 }

-JEMALLOC_INLINE size_t
+JEMALLOC_ALWAYS_INLINE size_t
 arena_mapbits_unallocated_size_get(arena_chunk_t *chunk, size_t pageind)
 {
 	size_t mapbits;
@@ -512,7 +542,7 @@ arena_mapbits_unallocated_size_get(arena_chunk_t *chunk, size_t pageind)
 	return (mapbits & ~PAGE_MASK);
 }

-JEMALLOC_INLINE size_t
+JEMALLOC_ALWAYS_INLINE size_t
 arena_mapbits_large_size_get(arena_chunk_t *chunk, size_t pageind)
 {
 	size_t mapbits;
@@ -523,7 +553,7 @@ arena_mapbits_large_size_get(arena_chunk_t *chunk, size_t pageind)
 	return (mapbits & ~PAGE_MASK);
 }

-JEMALLOC_INLINE size_t
+JEMALLOC_ALWAYS_INLINE size_t
 arena_mapbits_small_runind_get(arena_chunk_t *chunk, size_t pageind)
 {
 	size_t mapbits;
@@ -534,7 +564,7 @@ arena_mapbits_small_runind_get(arena_chunk_t *chunk, size_t pageind)
 	return (mapbits >> LG_PAGE);
 }

-JEMALLOC_INLINE size_t
+JEMALLOC_ALWAYS_INLINE size_t
 arena_mapbits_binind_get(arena_chunk_t *chunk, size_t pageind)
 {
 	size_t mapbits;
@@ -546,7 +576,7 @@ arena_mapbits_binind_get(arena_chunk_t *chunk, size_t pageind)
 	return (binind);
 }

-JEMALLOC_INLINE size_t
+JEMALLOC_ALWAYS_INLINE size_t
 arena_mapbits_dirty_get(arena_chunk_t *chunk, size_t pageind)
 {
 	size_t mapbits;
@@ -555,7 +585,7 @@ arena_mapbits_dirty_get(arena_chunk_t *chunk, size_t pageind)
 	return (mapbits & CHUNK_MAP_DIRTY);
 }

-JEMALLOC_INLINE size_t
+JEMALLOC_ALWAYS_INLINE size_t
 arena_mapbits_unzeroed_get(arena_chunk_t *chunk, size_t pageind)
 {
 	size_t mapbits;
@@ -564,7 +594,7 @@ arena_mapbits_unzeroed_get(arena_chunk_t *chunk, size_t pageind)
 	return (mapbits & CHUNK_MAP_UNZEROED);
 }

-JEMALLOC_INLINE size_t
+JEMALLOC_ALWAYS_INLINE size_t
 arena_mapbits_large_get(arena_chunk_t *chunk, size_t pageind)
 {
 	size_t mapbits;
@@ -573,7 +603,7 @@ arena_mapbits_large_get(arena_chunk_t *chunk, size_t pageind)
 	return (mapbits & CHUNK_MAP_LARGE);
 }

-JEMALLOC_INLINE size_t
+JEMALLOC_ALWAYS_INLINE size_t
 arena_mapbits_allocated_get(arena_chunk_t *chunk, size_t pageind)
 {
 	size_t mapbits;
@@ -582,86 +612,138 @@ arena_mapbits_allocated_get(arena_chunk_t *chunk, size_t pageind)
 	return (mapbits & CHUNK_MAP_ALLOCATED);
 }

-JEMALLOC_INLINE void
+JEMALLOC_ALWAYS_INLINE void
+arena_mapbitsp_write(size_t *mapbitsp, size_t mapbits)
+{
+
+	*mapbitsp = mapbits;
+}
+
+JEMALLOC_ALWAYS_INLINE void
 arena_mapbits_unallocated_set(arena_chunk_t *chunk, size_t pageind, size_t size,
    size_t flags)
 {
-	size_t *mapbitsp;
+	size_t *mapbitsp = arena_mapbitsp_get(chunk, pageind);

-	mapbitsp = arena_mapbitsp_get(chunk, pageind);
 	assert((size & PAGE_MASK) == 0);
 	assert((flags & ~CHUNK_MAP_FLAGS_MASK) == 0);
 	assert((flags & (CHUNK_MAP_DIRTY|CHUNK_MAP_UNZEROED)) == flags);
-	*mapbitsp = size | CHUNK_MAP_BININD_INVALID | flags;
+	arena_mapbitsp_write(mapbitsp, size | CHUNK_MAP_BININD_INVALID | flags);
 }

-JEMALLOC_INLINE void
+JEMALLOC_ALWAYS_INLINE void
 arena_mapbits_unallocated_size_set(arena_chunk_t *chunk, size_t pageind,
    size_t size)
 {
-	size_t *mapbitsp;
+	size_t *mapbitsp = arena_mapbitsp_get(chunk, pageind);
+	size_t mapbits = arena_mapbitsp_read(mapbitsp);

-	mapbitsp = arena_mapbitsp_get(chunk, pageind);
 	assert((size & PAGE_MASK) == 0);
-	assert((*mapbitsp & (CHUNK_MAP_LARGE|CHUNK_MAP_ALLOCATED)) == 0);
-	*mapbitsp = size | (*mapbitsp & PAGE_MASK);
+	assert((mapbits & (CHUNK_MAP_LARGE|CHUNK_MAP_ALLOCATED)) == 0);
+	arena_mapbitsp_write(mapbitsp, size | (mapbits & PAGE_MASK));
 }

-JEMALLOC_INLINE void
+JEMALLOC_ALWAYS_INLINE void
 arena_mapbits_large_set(arena_chunk_t *chunk, size_t pageind, size_t size,
    size_t flags)
 {
-	size_t *mapbitsp;
+	size_t *mapbitsp = arena_mapbitsp_get(chunk, pageind);
+	size_t mapbits = arena_mapbitsp_read(mapbitsp);
 	size_t unzeroed;

-	mapbitsp = arena_mapbitsp_get(chunk, pageind);
 	assert((size & PAGE_MASK) == 0);
 	assert((flags & CHUNK_MAP_DIRTY) == flags);
-	unzeroed = *mapbitsp & CHUNK_MAP_UNZEROED; /* Preserve unzeroed. */
-	*mapbitsp = size | CHUNK_MAP_BININD_INVALID | flags | unzeroed |
-	    CHUNK_MAP_LARGE | CHUNK_MAP_ALLOCATED;
+	unzeroed = mapbits & CHUNK_MAP_UNZEROED; /* Preserve unzeroed. */
+	arena_mapbitsp_write(mapbitsp, size | CHUNK_MAP_BININD_INVALID | flags
+	    | unzeroed | CHUNK_MAP_LARGE | CHUNK_MAP_ALLOCATED);
 }

-JEMALLOC_INLINE void
+JEMALLOC_ALWAYS_INLINE void
 arena_mapbits_large_binind_set(arena_chunk_t *chunk, size_t pageind,
    size_t binind)
 {
-	size_t *mapbitsp;
+	size_t *mapbitsp = arena_mapbitsp_get(chunk, pageind);
+	size_t mapbits = arena_mapbitsp_read(mapbitsp);

 	assert(binind <= BININD_INVALID);
-	mapbitsp = arena_mapbitsp_get(chunk, pageind);
 	assert(arena_mapbits_large_size_get(chunk, pageind) == PAGE);
-	*mapbitsp = (*mapbitsp & ~CHUNK_MAP_BININD_MASK) | (binind <<
-	    CHUNK_MAP_BININD_SHIFT);
+	arena_mapbitsp_write(mapbitsp, (mapbits & ~CHUNK_MAP_BININD_MASK) |
+	    (binind << CHUNK_MAP_BININD_SHIFT));
 }

-JEMALLOC_INLINE void
+JEMALLOC_ALWAYS_INLINE void
 arena_mapbits_small_set(arena_chunk_t *chunk, size_t pageind, size_t runind,
    size_t binind, size_t flags)
 {
-	size_t *mapbitsp;
+	size_t *mapbitsp = arena_mapbitsp_get(chunk, pageind);
+	size_t mapbits = arena_mapbitsp_read(mapbitsp);
 	size_t unzeroed;

 	assert(binind < BININD_INVALID);
-	mapbitsp = arena_mapbitsp_get(chunk, pageind);
 	assert(pageind - runind >= map_bias);
 	assert((flags & CHUNK_MAP_DIRTY) == flags);
-	unzeroed = *mapbitsp & CHUNK_MAP_UNZEROED; /* Preserve unzeroed. */
-	*mapbitsp = (runind << LG_PAGE) | (binind << CHUNK_MAP_BININD_SHIFT) |
-	    flags | unzeroed | CHUNK_MAP_ALLOCATED;
+	unzeroed = mapbits & CHUNK_MAP_UNZEROED; /* Preserve unzeroed. */
+	arena_mapbitsp_write(mapbitsp, (runind << LG_PAGE) | (binind <<
+	    CHUNK_MAP_BININD_SHIFT) | flags | unzeroed | CHUNK_MAP_ALLOCATED);
 }

-JEMALLOC_INLINE void
+JEMALLOC_ALWAYS_INLINE void
 arena_mapbits_unzeroed_set(arena_chunk_t *chunk, size_t pageind,
    size_t unzeroed)
 {
-	size_t *mapbitsp;
+	size_t *mapbitsp = arena_mapbitsp_get(chunk, pageind);
+	size_t mapbits = arena_mapbitsp_read(mapbitsp);

-	mapbitsp = arena_mapbitsp_get(chunk, pageind);
-	*mapbitsp = (*mapbitsp & ~CHUNK_MAP_UNZEROED) | unzeroed;
+	arena_mapbitsp_write(mapbitsp, (mapbits & ~CHUNK_MAP_UNZEROED) |
+	    unzeroed);
 }

-JEMALLOC_INLINE size_t
+JEMALLOC_INLINE bool
+arena_prof_accum_impl(arena_t *arena, uint64_t accumbytes)
+{
+
+	cassert(config_prof);
+	assert(prof_interval != 0);
+
+	arena->prof_accumbytes += accumbytes;
+	if (arena->prof_accumbytes >= prof_interval) {
+		arena->prof_accumbytes -= prof_interval;
+		return (true);
+	}
+	return (false);
+}
+
+JEMALLOC_INLINE bool
+arena_prof_accum_locked(arena_t *arena, uint64_t accumbytes)
+{
+
+	cassert(config_prof);
+
+	if (prof_interval == 0)
+		return (false);
+	return (arena_prof_accum_impl(arena, accumbytes));
+}
+
+JEMALLOC_INLINE bool
+arena_prof_accum(arena_t *arena, uint64_t accumbytes)
+{
+
+	cassert(config_prof);
+
+	if (prof_interval == 0)
+		return (false);
+
+	{
+		bool ret;
+
+		malloc_mutex_lock(&arena->lock);
+		ret = arena_prof_accum_impl(arena, accumbytes);
+		malloc_mutex_unlock(&arena->lock);
+		return (ret);
+	}
+}
+
+JEMALLOC_ALWAYS_INLINE size_t
 arena_ptr_small_binind_get(const void *ptr, size_t mapbits)
 {
 	size_t binind;
@@ -822,10 +904,10 @@ arena_prof_ctx_get(const void *ptr)
 }

 JEMALLOC_INLINE void
-arena_prof_ctx_set(const void *ptr, prof_ctx_t *ctx)
+arena_prof_ctx_set(const void *ptr, size_t usize, prof_ctx_t *ctx)
 {
 	arena_chunk_t *chunk;
-	size_t pageind, mapbits;
+	size_t pageind;

 	cassert(config_prof);
 	assert(ptr != NULL);
@@ -833,10 +915,17 @@ arena_prof_ctx_set(const void *ptr, prof_ctx_t *ctx)

 	chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(ptr);
 	pageind = ((uintptr_t)ptr - (uintptr_t)chunk) >> LG_PAGE;
-	mapbits = arena_mapbits_get(chunk, pageind);
-	assert((mapbits & CHUNK_MAP_ALLOCATED) != 0);
-	if ((mapbits & CHUNK_MAP_LARGE) == 0) {
+	assert(arena_mapbits_allocated_get(chunk, pageind) != 0);
+
+	if (usize > SMALL_MAXCLASS || (prof_promote &&
+	    ((uintptr_t)ctx != (uintptr_t)1U || arena_mapbits_large_get(chunk,
+	    pageind) != 0))) {
+		assert(arena_mapbits_large_get(chunk, pageind) != 0);
+		arena_mapp_get(chunk, pageind)->prof_ctx = ctx;
+	} else {
+		assert(arena_mapbits_large_get(chunk, pageind) == 0);
 		if (prof_promote == false) {
+			size_t mapbits = arena_mapbits_get(chunk, pageind);
 			arena_run_t *run = (arena_run_t *)((uintptr_t)chunk +
 			    (uintptr_t)((pageind - (mapbits >> LG_PAGE)) <<
 			    LG_PAGE));
@@ -848,15 +937,14 @@ arena_prof_ctx_set(const void *ptr, prof_ctx_t *ctx)
 			bin_info = &arena_bin_info[binind];
 			regind = arena_run_regind(run, bin_info, ptr);

-			*((prof_ctx_t **)((uintptr_t)run + bin_info->ctx0_offset
-			    + (regind * sizeof(prof_ctx_t *)))) = ctx;
-		} else
-			assert((uintptr_t)ctx == (uintptr_t)1U);
-	} else
-		arena_mapp_get(chunk, pageind)->prof_ctx = ctx;
+			*((prof_ctx_t **)((uintptr_t)run +
+			    bin_info->ctx0_offset + (regind * sizeof(prof_ctx_t
+			    *)))) = ctx;
+		}
+	}
 }

-JEMALLOC_INLINE void *
+JEMALLOC_ALWAYS_INLINE void *
 arena_malloc(arena_t *arena, size_t size, bool zero, bool try_tcache)
 {
 	tcache_t *tcache;
@@ -887,7 +975,7 @@ arena_malloc(arena_t *arena, size_t size, bool zero, bool try_tcache)
 }

 /* Return the size of the allocation pointed to by ptr. */
-JEMALLOC_INLINE size_t
+JEMALLOC_ALWAYS_INLINE size_t
 arena_salloc(const void *ptr, bool demote)
 {
 	size_t ret;
@@ -933,7 +1021,7 @@ arena_salloc(const void *ptr, bool demote)
 	return (ret);
 }

-JEMALLOC_INLINE void
+JEMALLOC_ALWAYS_INLINE void
 arena_dalloc(arena_t *arena, arena_chunk_t *chunk, void *ptr, bool try_tcache)
 {
 	size_t pageind, mapbits;
@@ -7,7 +7,7 @@ typedef enum {
 	dss_prec_secondary = 2,

 	dss_prec_limit     = 3
-} dss_prec_t ;
+} dss_prec_t;
 #define	DSS_PREC_DEFAULT	dss_prec_secondary
 #define	DSS_DEFAULT		"secondary"

@@ -5,7 +5,7 @@ typedef struct ckh_s ckh_t;
 typedef struct ckhc_s ckhc_t;

 /* Typedefs to allow easy function pointer passing. */
-typedef void ckh_hash_t (const void *, unsigned, size_t *, size_t *);
+typedef void ckh_hash_t (const void *, size_t[2]);
 typedef bool ckh_keycomp_t (const void *, const void *);

 /* Maintain counters used to get an idea of performance. */
@@ -17,7 +17,7 @@ typedef bool ckh_keycomp_t (const void *, const void *);
 * There are 2^LG_CKH_BUCKET_CELLS cells in each hash table bucket.  Try to fit
 * one bucket per L1 cache line.
 */
-#define LG_CKH_BUCKET_CELLS (LG_CACHELINE - LG_SIZEOF_PTR - 1)
+#define	LG_CKH_BUCKET_CELLS (LG_CACHELINE - LG_SIZEOF_PTR - 1)

 #endif /* JEMALLOC_H_TYPES */
 /******************************************************************************/
@@ -75,11 +75,9 @@ bool	ckh_insert(ckh_t *ckh, const void *key, const void *data);
 bool	ckh_remove(ckh_t *ckh, const void *searchkey, void **key,
    void **data);
 bool	ckh_search(ckh_t *ckh, const void *seachkey, void **key, void **data);
-void	ckh_string_hash(const void *key, unsigned minbits, size_t *hash1,
-    size_t *hash2);
+void	ckh_string_hash(const void *key, size_t r_hash[2]);
 bool	ckh_string_keycomp(const void *k1, const void *k2);
-void	ckh_pointer_hash(const void *key, unsigned minbits, size_t *hash1,
-    size_t *hash2);
+void	ckh_pointer_hash(const void *key, size_t r_hash[2]);
 bool	ckh_pointer_keycomp(const void *k1, const void *k2);

 #endif /* JEMALLOC_H_EXTERNS */
@@ -1,3 +1,8 @@
+/*
+ * The following hash function is based on MurmurHash3, placed into the public
+ * domain by Austin Appleby.  See http://code.google.com/p/smhasher/ for
+ * details.
+ */
 /******************************************************************************/
 #ifdef JEMALLOC_H_TYPES

@@ -14,56 +19,316 @@
 #ifdef JEMALLOC_H_INLINES

 #ifndef JEMALLOC_ENABLE_INLINE
-uint64_t	hash(const void *key, size_t len, uint64_t seed);
+uint32_t	hash_x86_32(const void *key, int len, uint32_t seed);
+void	hash_x86_128(const void *key, const int len, uint32_t seed,
+    uint64_t r_out[2]);
+void	hash_x64_128(const void *key, const int len, const uint32_t seed,
+    uint64_t r_out[2]);
+void	hash(const void *key, size_t len, const uint32_t seed,
+    size_t r_hash[2]);
 #endif

 #if (defined(JEMALLOC_ENABLE_INLINE) || defined(JEMALLOC_HASH_C_))
-/*
- * The following hash function is based on MurmurHash64A(), placed into the
- * public domain by Austin Appleby.  See http://murmurhash.googlepages.com/ for
- * details.
- */
-JEMALLOC_INLINE uint64_t
-hash(const void *key, size_t len, uint64_t seed)
+/******************************************************************************/
+/* Internal implementation. */
+JEMALLOC_INLINE uint32_t
+hash_rotl_32(uint32_t x, int8_t r)
 {
-	const uint64_t m = UINT64_C(0xc6a4a7935bd1e995);
-	const int r = 47;
-	uint64_t h = seed ^ (len * m);
-	const uint64_t *data = (const uint64_t *)key;
-	const uint64_t *end = data + (len/8);
-	const unsigned char *data2;

-	assert(((uintptr_t)key & 0x7) == 0);
+	return (x << r) | (x >> (32 - r));
+}

-	while(data != end) {
-		uint64_t k = *data++;
+JEMALLOC_INLINE uint64_t
+hash_rotl_64(uint64_t x, int8_t r)
+{
+	return (x << r) | (x >> (64 - r));
+}

-		k *= m;
-		k ^= k >> r;
-		k *= m;
+JEMALLOC_INLINE uint32_t
+hash_get_block_32(const uint32_t *p, int i)
+{

-		h ^= k;
-		h *= m;
-	}
+	return (p[i]);
+}

-	data2 = (const unsigned char *)data;
-	switch(len & 7) {
-	case 7: h ^= ((uint64_t)(data2[6])) << 48;
-	case 6: h ^= ((uint64_t)(data2[5])) << 40;
-	case 5: h ^= ((uint64_t)(data2[4])) << 32;
-	case 4: h ^= ((uint64_t)(data2[3])) << 24;
-	case 3: h ^= ((uint64_t)(data2[2])) << 16;
-	case 2: h ^= ((uint64_t)(data2[1])) << 8;
-	case 1: h ^= ((uint64_t)(data2[0]));
-		h *= m;
-	}
+JEMALLOC_INLINE uint64_t
+hash_get_block_64(const uint64_t *p, int i)
+{

-	h ^= h >> r;
-	h *= m;
-	h ^= h >> r;
+	return (p[i]);
+}
+
+JEMALLOC_INLINE uint32_t
+hash_fmix_32(uint32_t h)
+{
+
+	h ^= h >> 16;
+	h *= 0x85ebca6b;
+	h ^= h >> 13;
+	h *= 0xc2b2ae35;
+	h ^= h >> 16;

 	return (h);
 }
+
+JEMALLOC_INLINE uint64_t
+hash_fmix_64(uint64_t k)
+{
+
+	k ^= k >> 33;
+	k *= QU(0xff51afd7ed558ccdLLU);
+	k ^= k >> 33;
+	k *= QU(0xc4ceb9fe1a85ec53LLU);
+	k ^= k >> 33;
+
+	return (k);
+}
+
+JEMALLOC_INLINE uint32_t
+hash_x86_32(const void *key, int len, uint32_t seed)
+{
+	const uint8_t *data = (const uint8_t *) key;
+	const int nblocks = len / 4;
+
+	uint32_t h1 = seed;
+
+	const uint32_t c1 = 0xcc9e2d51;
+	const uint32_t c2 = 0x1b873593;
+
+	/* body */
+	{
+		const uint32_t *blocks = (const uint32_t *) (data + nblocks*4);
+		int i;
+
+		for (i = -nblocks; i; i++) {
+			uint32_t k1 = hash_get_block_32(blocks, i);
+
+			k1 *= c1;
+			k1 = hash_rotl_32(k1, 15);
+			k1 *= c2;
+
+			h1 ^= k1;
+			h1 = hash_rotl_32(h1, 13);
+			h1 = h1*5 + 0xe6546b64;
+		}
+	}
+
+	/* tail */
+	{
+		const uint8_t *tail = (const uint8_t *) (data + nblocks*4);
+
+		uint32_t k1 = 0;
+
+		switch (len & 3) {
+		case 3: k1 ^= tail[2] << 16;
+		case 2: k1 ^= tail[1] << 8;
+		case 1: k1 ^= tail[0]; k1 *= c1; k1 = hash_rotl_32(k1, 15);
+			k1 *= c2; h1 ^= k1;
+		}
+	}
+
+	/* finalization */
+	h1 ^= len;
+
+	h1 = hash_fmix_32(h1);
+
+	return (h1);
+}
+
+UNUSED JEMALLOC_INLINE void
+hash_x86_128(const void *key, const int len, uint32_t seed,
+    uint64_t r_out[2])
+{
+	const uint8_t * data = (const uint8_t *) key;
+	const int nblocks = len / 16;
+
+	uint32_t h1 = seed;
+	uint32_t h2 = seed;
+	uint32_t h3 = seed;
+	uint32_t h4 = seed;
+
+	const uint32_t c1 = 0x239b961b;
+	const uint32_t c2 = 0xab0e9789;
+	const uint32_t c3 = 0x38b34ae5;
+	const uint32_t c4 = 0xa1e38b93;
+
+	/* body */
+	{
+		const uint32_t *blocks = (const uint32_t *) (data + nblocks*16);
+		int i;
+
+		for (i = -nblocks; i; i++) {
+			uint32_t k1 = hash_get_block_32(blocks, i*4 + 0);
+			uint32_t k2 = hash_get_block_32(blocks, i*4 + 1);
+			uint32_t k3 = hash_get_block_32(blocks, i*4 + 2);
+			uint32_t k4 = hash_get_block_32(blocks, i*4 + 3);
+
+			k1 *= c1; k1 = hash_rotl_32(k1, 15); k1 *= c2; h1 ^= k1;
+
+			h1 = hash_rotl_32(h1, 19); h1 += h2;
+			h1 = h1*5 + 0x561ccd1b;
+
+			k2 *= c2; k2 = hash_rotl_32(k2, 16); k2 *= c3; h2 ^= k2;
+
+			h2 = hash_rotl_32(h2, 17); h2 += h3;
+			h2 = h2*5 + 0x0bcaa747;
+
+			k3 *= c3; k3 = hash_rotl_32(k3, 17); k3 *= c4; h3 ^= k3;
+
+			h3 = hash_rotl_32(h3, 15); h3 += h4;
+			h3 = h3*5 + 0x96cd1c35;
+
+			k4 *= c4; k4 = hash_rotl_32(k4, 18); k4 *= c1; h4 ^= k4;
+
+			h4 = hash_rotl_32(h4, 13); h4 += h1;
+			h4 = h4*5 + 0x32ac3b17;
+		}
+	}
+
+	/* tail */
+	{
+		const uint8_t *tail = (const uint8_t *) (data + nblocks*16);
+		uint32_t k1 = 0;
+		uint32_t k2 = 0;
+		uint32_t k3 = 0;
+		uint32_t k4 = 0;
+
+		switch (len & 15) {
+		case 15: k4 ^= tail[14] << 16;
+		case 14: k4 ^= tail[13] << 8;
+		case 13: k4 ^= tail[12] << 0;
+			k4 *= c4; k4 = hash_rotl_32(k4, 18); k4 *= c1; h4 ^= k4;
+
+		case 12: k3 ^= tail[11] << 24;
+		case 11: k3 ^= tail[10] << 16;
+		case 10: k3 ^= tail[ 9] << 8;
+		case  9: k3 ^= tail[ 8] << 0;
+		     k3 *= c3; k3 = hash_rotl_32(k3, 17); k3 *= c4; h3 ^= k3;
+
+		case  8: k2 ^= tail[ 7] << 24;
+		case  7: k2 ^= tail[ 6] << 16;
+		case  6: k2 ^= tail[ 5] << 8;
+		case  5: k2 ^= tail[ 4] << 0;
+			k2 *= c2; k2 = hash_rotl_32(k2, 16); k2 *= c3; h2 ^= k2;
+
+		case  4: k1 ^= tail[ 3] << 24;
+		case  3: k1 ^= tail[ 2] << 16;
+		case  2: k1 ^= tail[ 1] << 8;
+		case  1: k1 ^= tail[ 0] << 0;
+			k1 *= c1; k1 = hash_rotl_32(k1, 15); k1 *= c2; h1 ^= k1;
+		}
+	}
+
+	/* finalization */
+	h1 ^= len; h2 ^= len; h3 ^= len; h4 ^= len;
+
+	h1 += h2; h1 += h3; h1 += h4;
+	h2 += h1; h3 += h1; h4 += h1;
+
+	h1 = hash_fmix_32(h1);
+	h2 = hash_fmix_32(h2);
+	h3 = hash_fmix_32(h3);
+	h4 = hash_fmix_32(h4);
+
+	h1 += h2; h1 += h3; h1 += h4;
+	h2 += h1; h3 += h1; h4 += h1;
+
+	r_out[0] = (((uint64_t) h2) << 32) | h1;
+	r_out[1] = (((uint64_t) h4) << 32) | h3;
+}
+
+UNUSED JEMALLOC_INLINE void
+hash_x64_128(const void *key, const int len, const uint32_t seed,
+    uint64_t r_out[2])
+{
+	const uint8_t *data = (const uint8_t *) key;
+	const int nblocks = len / 16;
+
+	uint64_t h1 = seed;
+	uint64_t h2 = seed;
+
+	const uint64_t c1 = QU(0x87c37b91114253d5LLU);
+	const uint64_t c2 = QU(0x4cf5ad432745937fLLU);
+
+	/* body */
+	{
+		const uint64_t *blocks = (const uint64_t *) (data);
+		int i;
+
+		for (i = 0; i < nblocks; i++) {
+			uint64_t k1 = hash_get_block_64(blocks, i*2 + 0);
+			uint64_t k2 = hash_get_block_64(blocks, i*2 + 1);
+
+			k1 *= c1; k1 = hash_rotl_64(k1, 31); k1 *= c2; h1 ^= k1;
+
+			h1 = hash_rotl_64(h1, 27); h1 += h2;
+			h1 = h1*5 + 0x52dce729;
+
+			k2 *= c2; k2 = hash_rotl_64(k2, 33); k2 *= c1; h2 ^= k2;
+
+			h2 = hash_rotl_64(h2, 31); h2 += h1;
+			h2 = h2*5 + 0x38495ab5;
+		}
+	}
+
+	/* tail */
+	{
+		const uint8_t *tail = (const uint8_t*)(data + nblocks*16);
+		uint64_t k1 = 0;
+		uint64_t k2 = 0;
+
+		switch (len & 15) {
+		case 15: k2 ^= ((uint64_t)(tail[14])) << 48;
+		case 14: k2 ^= ((uint64_t)(tail[13])) << 40;
+		case 13: k2 ^= ((uint64_t)(tail[12])) << 32;
+		case 12: k2 ^= ((uint64_t)(tail[11])) << 24;
+		case 11: k2 ^= ((uint64_t)(tail[10])) << 16;
+		case 10: k2 ^= ((uint64_t)(tail[ 9])) << 8;
+		case  9: k2 ^= ((uint64_t)(tail[ 8])) << 0;
+			k2 *= c2; k2 = hash_rotl_64(k2, 33); k2 *= c1; h2 ^= k2;
+
+		case  8: k1 ^= ((uint64_t)(tail[ 7])) << 56;
+		case  7: k1 ^= ((uint64_t)(tail[ 6])) << 48;
+		case  6: k1 ^= ((uint64_t)(tail[ 5])) << 40;
+		case  5: k1 ^= ((uint64_t)(tail[ 4])) << 32;
+		case  4: k1 ^= ((uint64_t)(tail[ 3])) << 24;
+		case  3: k1 ^= ((uint64_t)(tail[ 2])) << 16;
+		case  2: k1 ^= ((uint64_t)(tail[ 1])) << 8;
+		case  1: k1 ^= ((uint64_t)(tail[ 0])) << 0;
+			k1 *= c1; k1 = hash_rotl_64(k1, 31); k1 *= c2; h1 ^= k1;
+		}
+	}
+
+	/* finalization */
+	h1 ^= len; h2 ^= len;
+
+	h1 += h2;
+	h2 += h1;
+
+	h1 = hash_fmix_64(h1);
+	h2 = hash_fmix_64(h2);
+
+	h1 += h2;
+	h2 += h1;
+
+	r_out[0] = h1;
+	r_out[1] = h2;
+}
+
+/******************************************************************************/
+/* API. */
+JEMALLOC_INLINE void
+hash(const void *key, size_t len, const uint32_t seed, size_t r_hash[2])
+{
+#if (LG_SIZEOF_PTR == 3 && !defined(JEMALLOC_BIG_ENDIAN))
+	hash_x64_128(key, len, seed, (uint64_t *)r_hash);
+#else
+	uint64_t hashes[2];
+	hash_x86_128(key, len, seed, hashes);
+	r_hash[0] = (size_t)hashes[0];
+	r_hash[1] = (size_t)hashes[1];
+#endif
+}
 #endif

 #endif /* JEMALLOC_H_INLINES */
@@ -17,14 +17,20 @@ extern size_t		huge_allocated;
 /* Protects chunk-related data structures. */
 extern malloc_mutex_t	huge_mtx;

-void	*huge_malloc(size_t size, bool zero);
-void	*huge_palloc(size_t size, size_t alignment, bool zero);
-void	*huge_ralloc_no_move(void *ptr, size_t oldsize, size_t size,
+void	*huge_malloc(size_t size, bool zero, dss_prec_t dss_prec);
+void	*huge_palloc(size_t size, size_t alignment, bool zero,
+    dss_prec_t dss_prec);
+bool	huge_ralloc_no_move(void *ptr, size_t oldsize, size_t size,
    size_t extra);
 void	*huge_ralloc(void *ptr, size_t oldsize, size_t size, size_t extra,
-    size_t alignment, bool zero, bool try_tcache_dalloc);
+    size_t alignment, bool zero, bool try_tcache_dalloc, dss_prec_t dss_prec);
+#ifdef JEMALLOC_JET
+typedef void (huge_dalloc_junk_t)(void *, size_t);
+extern huge_dalloc_junk_t *huge_dalloc_junk;
+#endif
 void	huge_dalloc(void *ptr, bool unmap);
 size_t	huge_salloc(const void *ptr);
+dss_prec_t	huge_dss_prec_get(arena_t *arena);
 prof_ctx_t	*huge_prof_ctx_get(const void *ptr);
 void	huge_prof_ctx_set(const void *ptr, prof_ctx_t *ctx);
 bool	huge_boot(void);
@@ -1,5 +1,5 @@
 #ifndef JEMALLOC_INTERNAL_H
-#define JEMALLOC_INTERNAL_H
+#define	JEMALLOC_INTERNAL_H
 #include <math.h>
 #ifdef _WIN32
 #  include <windows.h>
@@ -54,8 +54,7 @@ typedef intptr_t ssize_t;
 #endif
 #include <fcntl.h>

-#define	JEMALLOC_NO_DEMANGLE
-#include "../jemalloc@install_suffix@.h"
+#include "jemalloc_internal_defs.h"

 #ifdef JEMALLOC_UTRACE
 #include <sys/ktrace.h>
@@ -66,13 +65,18 @@ typedef intptr_t ssize_t;
 #include <valgrind/memcheck.h>
 #endif

-#include "jemalloc/internal/private_namespace.h"
-
-#ifdef JEMALLOC_CC_SILENCE
-#define	UNUSED JEMALLOC_ATTR(unused)
+#define	JEMALLOC_NO_DEMANGLE
+#ifdef JEMALLOC_JET
+#  define JEMALLOC_N(n) jet_##n
+#  include "jemalloc/internal/public_namespace.h"
+#  define JEMALLOC_NO_RENAME
+#  include "../jemalloc@install_suffix@.h"
+#  undef JEMALLOC_NO_RENAME
 #else
-#define	UNUSED
+#  define JEMALLOC_N(n) @private_namespace@##n
+#  include "../jemalloc@install_suffix@.h"
 #endif
+#include "jemalloc/internal/private_namespace.h"

 static const bool config_debug =
 #ifdef JEMALLOC_DEBUG
@@ -221,28 +225,13 @@ static const bool config_ivsalloc =
 *   JEMALLOC_H_INLINES : Inline functions.
 */
 /******************************************************************************/
-#define JEMALLOC_H_TYPES
+#define	JEMALLOC_H_TYPES

+#include "jemalloc/internal/jemalloc_internal_macros.h"
+
+#define	MALLOCX_LG_ALIGN_MASK	((int)0x3f)
 #define	ALLOCM_LG_ALIGN_MASK	((int)0x3f)

-#define	ZU(z)	((size_t)z)
-
-#ifndef __DECONST
-#  define	__DECONST(type, var)	((type)(uintptr_t)(const void *)(var))
-#endif
-
-#ifdef JEMALLOC_DEBUG
-   /* Disable inlining to make debugging easier. */
-#  define JEMALLOC_INLINE
-#  define inline
-#else
-#  define JEMALLOC_ENABLE_INLINE
-#  define JEMALLOC_INLINE static inline
-#  ifdef _MSC_VER
-#    define inline _inline
-#  endif
-#endif
-
 /* Smallest size class to support. */
 #define	LG_TINY_MIN		3
 #define	TINY_MIN		(1U << LG_TINY_MIN)
@@ -270,6 +259,9 @@ static const bool config_ivsalloc =
 #  ifdef __arm__
 #    define LG_QUANTUM		3
 #  endif
+#  ifdef __aarch64__
+#    define LG_QUANTUM		4
+#  endif
 #  ifdef __hppa__
 #    define LG_QUANTUM		4
 #  endif
@@ -279,7 +271,7 @@ static const bool config_ivsalloc =
 #  ifdef __powerpc__
 #    define LG_QUANTUM		4
 #  endif
-#  ifdef __s390x__
+#  ifdef __s390__
 #    define LG_QUANTUM		4
 #  endif
 #  ifdef __SH4__
@@ -359,7 +351,11 @@ static const bool config_ivsalloc =
 #    include <malloc.h>
 #    define alloca _alloca
 #  else
-#    include <alloca.h>
+#    ifdef JEMALLOC_HAS_ALLOCA_H
+#      include <alloca.h>
+#    else
+#      include <stdlib.h>
+#    endif
 #  endif
 #  define VARIABLE_ARRAY(type, name, count) \
 	type *name = alloca(sizeof(type) * count)
@@ -428,15 +424,18 @@ static const bool config_ivsalloc =
 } while (0)
 #else
 #define	RUNNING_ON_VALGRIND	((unsigned)0)
-#define	VALGRIND_MALLOCLIKE_BLOCK(addr, sizeB, rzB, is_zeroed)
-#define	VALGRIND_RESIZEINPLACE_BLOCK(addr, oldSizeB, newSizeB, rzB)
-#define	VALGRIND_FREELIKE_BLOCK(addr, rzB)
-#define	VALGRIND_MAKE_MEM_UNDEFINED(_qzz_addr, _qzz_len)
-#define	VALGRIND_MAKE_MEM_DEFINED(_qzz_addr, _qzz_len)
-#define	JEMALLOC_VALGRIND_MALLOC(cond, ptr, usize, zero)
+#define	VALGRIND_MALLOCLIKE_BLOCK(addr, sizeB, rzB, is_zeroed) \
+    do {} while (0)
+#define	VALGRIND_RESIZEINPLACE_BLOCK(addr, oldSizeB, newSizeB, rzB) \
+    do {} while (0)
+#define	VALGRIND_FREELIKE_BLOCK(addr, rzB) do {} while (0)
+#define	VALGRIND_MAKE_MEM_NOACCESS(_qzz_addr, _qzz_len) do {} while (0)
+#define	VALGRIND_MAKE_MEM_UNDEFINED(_qzz_addr, _qzz_len) do {} while (0)
+#define	VALGRIND_MAKE_MEM_DEFINED(_qzz_addr, _qzz_len) do {} while (0)
+#define	JEMALLOC_VALGRIND_MALLOC(cond, ptr, usize, zero) do {} while (0)
 #define	JEMALLOC_VALGRIND_REALLOC(ptr, usize, old_ptr, old_usize,	\
-    old_rzsize, zero)
-#define	JEMALLOC_VALGRIND_FREE(ptr, rzsize)
+    old_rzsize, zero) do {} while (0)
+#define	JEMALLOC_VALGRIND_FREE(ptr, rzsize) do {} while (0)
 #endif

 #include "jemalloc/internal/util.h"
@@ -463,7 +462,7 @@ static const bool config_ivsalloc =

 #undef JEMALLOC_H_TYPES
 /******************************************************************************/
-#define JEMALLOC_H_STRUCTS
+#define	JEMALLOC_H_STRUCTS

 #include "jemalloc/internal/util.h"
 #include "jemalloc/internal/atomic.h"
@@ -492,14 +491,14 @@ typedef struct {
 	uint64_t	deallocated;
 } thread_allocated_t;
 /*
- * The JEMALLOC_CONCAT() wrapper is necessary to pass {0, 0} via a cpp macro
+ * The JEMALLOC_ARG_CONCAT() wrapper is necessary to pass {0, 0} via a cpp macro
 * argument.
 */
-#define	THREAD_ALLOCATED_INITIALIZER	JEMALLOC_CONCAT({0, 0})
+#define	THREAD_ALLOCATED_INITIALIZER	JEMALLOC_ARG_CONCAT({0, 0})

 #undef JEMALLOC_H_STRUCTS
 /******************************************************************************/
-#define JEMALLOC_H_EXTERNS
+#define	JEMALLOC_H_EXTERNS

 extern bool	opt_abort;
 extern bool	opt_junk;
@@ -559,7 +558,7 @@ void	jemalloc_postfork_child(void);

 #undef JEMALLOC_H_EXTERNS
 /******************************************************************************/
-#define JEMALLOC_H_INLINES
+#define	JEMALLOC_H_INLINES

 #include "jemalloc/internal/util.h"
 #include "jemalloc/internal/atomic.h"
@@ -591,13 +590,14 @@ arena_t	*choose_arena(arena_t *arena);
 * for allocations.
 */
 malloc_tsd_externs(arenas, arena_t *)
-malloc_tsd_funcs(JEMALLOC_INLINE, arenas, arena_t *, NULL, arenas_cleanup)
+malloc_tsd_funcs(JEMALLOC_ALWAYS_INLINE, arenas, arena_t *, NULL,
+    arenas_cleanup)

 /*
 * Compute usable size that would result from allocating an object with the
 * specified size.
 */
-JEMALLOC_INLINE size_t
+JEMALLOC_ALWAYS_INLINE size_t
 s2u(size_t size)
 {

@@ -612,7 +612,7 @@ s2u(size_t size)
 * Compute usable size that would result from allocating an object with the
 * specified size and alignment.
 */
-JEMALLOC_INLINE size_t
+JEMALLOC_ALWAYS_INLINE size_t
 sa2u(size_t size, size_t alignment)
 {
 	size_t usize;
@@ -733,32 +733,36 @@ choose_arena(arena_t *arena)
 #include "jemalloc/internal/quarantine.h"

 #ifndef JEMALLOC_ENABLE_INLINE
-void	*imallocx(size_t size, bool try_tcache, arena_t *arena);
+void	*imalloct(size_t size, bool try_tcache, arena_t *arena);
 void	*imalloc(size_t size);
-void	*icallocx(size_t size, bool try_tcache, arena_t *arena);
+void	*icalloct(size_t size, bool try_tcache, arena_t *arena);
 void	*icalloc(size_t size);
-void	*ipallocx(size_t usize, size_t alignment, bool zero, bool try_tcache,
+void	*ipalloct(size_t usize, size_t alignment, bool zero, bool try_tcache,
    arena_t *arena);
 void	*ipalloc(size_t usize, size_t alignment, bool zero);
 size_t	isalloc(const void *ptr, bool demote);
 size_t	ivsalloc(const void *ptr, bool demote);
 size_t	u2rz(size_t usize);
 size_t	p2rz(const void *ptr);
-void	idallocx(void *ptr, bool try_tcache);
+void	idalloct(void *ptr, bool try_tcache);
 void	idalloc(void *ptr);
-void	iqallocx(void *ptr, bool try_tcache);
+void	iqalloct(void *ptr, bool try_tcache);
 void	iqalloc(void *ptr);
-void	*irallocx(void *ptr, size_t size, size_t extra, size_t alignment,
-    bool zero, bool no_move, bool try_tcache_alloc, bool try_tcache_dalloc,
+void	*iralloct_realign(void *ptr, size_t oldsize, size_t size, size_t extra,
+    size_t alignment, bool zero, bool try_tcache_alloc, bool try_tcache_dalloc,
    arena_t *arena);
+void	*iralloct(void *ptr, size_t size, size_t extra, size_t alignment,
+    bool zero, bool try_tcache_alloc, bool try_tcache_dalloc, arena_t *arena);
 void	*iralloc(void *ptr, size_t size, size_t extra, size_t alignment,
-    bool zero, bool no_move);
+    bool zero);
+bool	ixalloc(void *ptr, size_t size, size_t extra, size_t alignment,
+    bool zero);
 malloc_tsd_protos(JEMALLOC_ATTR(unused), thread_allocated, thread_allocated_t)
 #endif

 #if (defined(JEMALLOC_ENABLE_INLINE) || defined(JEMALLOC_C_))
-JEMALLOC_INLINE void *
-imallocx(size_t size, bool try_tcache, arena_t *arena)
+JEMALLOC_ALWAYS_INLINE void *
+imalloct(size_t size, bool try_tcache, arena_t *arena)
 {

 	assert(size != 0);
@@ -766,35 +770,35 @@ imallocx(size_t size, bool try_tcache, arena_t *arena)
 	if (size <= arena_maxclass)
 		return (arena_malloc(arena, size, false, try_tcache));
 	else
-		return (huge_malloc(size, false));
+		return (huge_malloc(size, false, huge_dss_prec_get(arena)));
 }

-JEMALLOC_INLINE void *
+JEMALLOC_ALWAYS_INLINE void *
 imalloc(size_t size)
 {

-	return (imallocx(size, true, NULL));
+	return (imalloct(size, true, NULL));
 }

-JEMALLOC_INLINE void *
-icallocx(size_t size, bool try_tcache, arena_t *arena)
+JEMALLOC_ALWAYS_INLINE void *
+icalloct(size_t size, bool try_tcache, arena_t *arena)
 {

 	if (size <= arena_maxclass)
 		return (arena_malloc(arena, size, true, try_tcache));
 	else
-		return (huge_malloc(size, true));
+		return (huge_malloc(size, true, huge_dss_prec_get(arena)));
 }

-JEMALLOC_INLINE void *
+JEMALLOC_ALWAYS_INLINE void *
 icalloc(size_t size)
 {

-	return (icallocx(size, true, NULL));
+	return (icalloct(size, true, NULL));
 }

-JEMALLOC_INLINE void *
-ipallocx(size_t usize, size_t alignment, bool zero, bool try_tcache,
+JEMALLOC_ALWAYS_INLINE void *
+ipalloct(size_t usize, size_t alignment, bool zero, bool try_tcache,
    arena_t *arena)
 {
 	void *ret;
@@ -809,20 +813,20 @@ ipallocx(size_t usize, size_t alignment, bool zero, bool try_tcache,
 			ret = arena_palloc(choose_arena(arena), usize,
 			    alignment, zero);
 		} else if (alignment <= chunksize)
-			ret = huge_malloc(usize, zero);
+			ret = huge_malloc(usize, zero, huge_dss_prec_get(arena));
 		else
-			ret = huge_palloc(usize, alignment, zero);
+			ret = huge_palloc(usize, alignment, zero, huge_dss_prec_get(arena));
 	}

 	assert(ALIGNMENT_ADDR2BASE(ret, alignment) == ret);
 	return (ret);
 }

-JEMALLOC_INLINE void *
+JEMALLOC_ALWAYS_INLINE void *
 ipalloc(size_t usize, size_t alignment, bool zero)
 {

-	return (ipallocx(usize, alignment, zero, true, NULL));
+	return (ipalloct(usize, alignment, zero, true, NULL));
 }

 /*
@@ -830,7 +834,7 @@ ipalloc(size_t usize, size_t alignment, bool zero)
 *   void *ptr = [...]
 *   size_t sz = isalloc(ptr, config_prof);
 */
-JEMALLOC_INLINE size_t
+JEMALLOC_ALWAYS_INLINE size_t
 isalloc(const void *ptr, bool demote)
 {
 	size_t ret;
@@ -849,12 +853,12 @@ isalloc(const void *ptr, bool demote)
 	return (ret);
 }

-JEMALLOC_INLINE size_t
+JEMALLOC_ALWAYS_INLINE size_t
 ivsalloc(const void *ptr, bool demote)
 {

 	/* Return 0 if ptr is not within a chunk managed by jemalloc. */
-	if (rtree_get(chunks_rtree, (uintptr_t)CHUNK_ADDR2BASE(ptr)) == NULL)
+	if (rtree_get(chunks_rtree, (uintptr_t)CHUNK_ADDR2BASE(ptr)) == 0)
 		return (0);

 	return (isalloc(ptr, demote));
@@ -882,8 +886,8 @@ p2rz(const void *ptr)
 	return (u2rz(usize));
 }

-JEMALLOC_INLINE void
-idallocx(void *ptr, bool try_tcache)
+JEMALLOC_ALWAYS_INLINE void
+idalloct(void *ptr, bool try_tcache)
 {
 	arena_chunk_t *chunk;

@@ -896,35 +900,67 @@ idallocx(void *ptr, bool try_tcache)
 		huge_dalloc(ptr, true);
 }

-JEMALLOC_INLINE void
+JEMALLOC_ALWAYS_INLINE void
 idalloc(void *ptr)
 {

-	idallocx(ptr, true);
+	idalloct(ptr, true);
 }

-JEMALLOC_INLINE void
-iqallocx(void *ptr, bool try_tcache)
+JEMALLOC_ALWAYS_INLINE void
+iqalloct(void *ptr, bool try_tcache)
 {

 	if (config_fill && opt_quarantine)
 		quarantine(ptr);
 	else
-		idallocx(ptr, try_tcache);
+		idalloct(ptr, try_tcache);
 }

-JEMALLOC_INLINE void
+JEMALLOC_ALWAYS_INLINE void
 iqalloc(void *ptr)
 {

-	iqallocx(ptr, true);
+	iqalloct(ptr, true);
 }

-JEMALLOC_INLINE void *
-irallocx(void *ptr, size_t size, size_t extra, size_t alignment, bool zero,
-    bool no_move, bool try_tcache_alloc, bool try_tcache_dalloc, arena_t *arena)
+JEMALLOC_ALWAYS_INLINE void *
+iralloct_realign(void *ptr, size_t oldsize, size_t size, size_t extra,
+    size_t alignment, bool zero, bool try_tcache_alloc, bool try_tcache_dalloc,
+    arena_t *arena)
+{
+	void *p;
+	size_t usize, copysize;
+
+	usize = sa2u(size + extra, alignment);
+	if (usize == 0)
+		return (NULL);
+	p = ipalloct(usize, alignment, zero, try_tcache_alloc, arena);
+	if (p == NULL) {
+		if (extra == 0)
+			return (NULL);
+		/* Try again, without extra this time. */
+		usize = sa2u(size, alignment);
+		if (usize == 0)
+			return (NULL);
+		p = ipalloct(usize, alignment, zero, try_tcache_alloc, arena);
+		if (p == NULL)
+			return (NULL);
+	}
+	/*
+	 * Copy at most size bytes (not size+extra), since the caller has no
+	 * expectation that the extra bytes will be reliably preserved.
+	 */
+	copysize = (size < oldsize) ? size : oldsize;
+	memcpy(p, ptr, copysize);
+	iqalloct(ptr, try_tcache_dalloc);
+	return (p);
+}
+
+JEMALLOC_ALWAYS_INLINE void *
+iralloct(void *ptr, size_t size, size_t extra, size_t alignment, bool zero,
+    bool try_tcache_alloc, bool try_tcache_dalloc, arena_t *arena)
 {
-	void *ret;
 	size_t oldsize;

 	assert(ptr != NULL);
@@ -934,72 +970,54 @@ irallocx(void *ptr, size_t size, size_t extra, size_t alignment, bool zero,

 	if (alignment != 0 && ((uintptr_t)ptr & ((uintptr_t)alignment-1))
 	    != 0) {
-		size_t usize, copysize;
-
 		/*
 		 * Existing object alignment is inadequate; allocate new space
 		 * and copy.
 		 */
-		if (no_move)
-			return (NULL);
-		usize = sa2u(size + extra, alignment);
-		if (usize == 0)
-			return (NULL);
-		ret = ipallocx(usize, alignment, zero, try_tcache_alloc, arena);
-		if (ret == NULL) {
-			if (extra == 0)
-				return (NULL);
-			/* Try again, without extra this time. */
-			usize = sa2u(size, alignment);
-			if (usize == 0)
-				return (NULL);
-			ret = ipallocx(usize, alignment, zero, try_tcache_alloc,
-			    arena);
-			if (ret == NULL)
-				return (NULL);
-		}
-		/*
-		 * Copy at most size bytes (not size+extra), since the caller
-		 * has no expectation that the extra bytes will be reliably
-		 * preserved.
-		 */
-		copysize = (size < oldsize) ? size : oldsize;
-		memcpy(ret, ptr, copysize);
-		iqallocx(ptr, try_tcache_dalloc);
-		return (ret);
+		return (iralloct_realign(ptr, oldsize, size, extra, alignment,
+		    zero, try_tcache_alloc, try_tcache_dalloc, arena));
 	}

-	if (no_move) {
-		if (size <= arena_maxclass) {
-			return (arena_ralloc_no_move(ptr, oldsize, size,
-			    extra, zero));
-		} else {
-			return (huge_ralloc_no_move(ptr, oldsize, size,
-			    extra));
-		}
+	if (size + extra <= arena_maxclass) {
+		return (arena_ralloc(arena, ptr, oldsize, size, extra,
+		    alignment, zero, try_tcache_alloc,
+		    try_tcache_dalloc));
 	} else {
-		if (size + extra <= arena_maxclass) {
-			return (arena_ralloc(arena, ptr, oldsize, size, extra,
-			    alignment, zero, try_tcache_alloc,
-			    try_tcache_dalloc));
-		} else {
-			return (huge_ralloc(ptr, oldsize, size, extra,
-			    alignment, zero, try_tcache_dalloc));
-		}
+		return (huge_ralloc(ptr, oldsize, size, extra,
+		    alignment, zero, try_tcache_dalloc, huge_dss_prec_get(arena)));
 	}
 }

-JEMALLOC_INLINE void *
-iralloc(void *ptr, size_t size, size_t extra, size_t alignment, bool zero,
-    bool no_move)
+JEMALLOC_ALWAYS_INLINE void *
+iralloc(void *ptr, size_t size, size_t extra, size_t alignment, bool zero)
 {

-	return (irallocx(ptr, size, extra, alignment, zero, no_move, true, true,
-	    NULL));
+	return (iralloct(ptr, size, extra, alignment, zero, true, true, NULL));
+}
+
+JEMALLOC_ALWAYS_INLINE bool
+ixalloc(void *ptr, size_t size, size_t extra, size_t alignment, bool zero)
+{
+	size_t oldsize;
+
+	assert(ptr != NULL);
+	assert(size != 0);
+
+	oldsize = isalloc(ptr, config_prof);
+	if (alignment != 0 && ((uintptr_t)ptr & ((uintptr_t)alignment-1))
+	    != 0) {
+		/* Existing object alignment is inadequate. */
+		return (true);
+	}
+
+	if (size <= arena_maxclass)
+		return (arena_ralloc_no_move(ptr, oldsize, size, extra, zero));
+	else
+		return (huge_ralloc_no_move(ptr, oldsize, size, extra));
 }

 malloc_tsd_externs(thread_allocated, thread_allocated_t)
-malloc_tsd_funcs(JEMALLOC_INLINE, thread_allocated, thread_allocated_t,
+malloc_tsd_funcs(JEMALLOC_ALWAYS_INLINE, thread_allocated, thread_allocated_t,
    THREAD_ALLOCATED_INITIALIZER, malloc_tsd_no_cleanup)
 #endif

@@ -0,0 +1,205 @@
+#ifndef JEMALLOC_INTERNAL_DEFS_H_
+#define	JEMALLOC_INTERNAL_DEFS_H_
+/*
+ * If JEMALLOC_PREFIX is defined via --with-jemalloc-prefix, it will cause all
+ * public APIs to be prefixed.  This makes it possible, with some care, to use
+ * multiple allocators simultaneously.
+ */
+#undef JEMALLOC_PREFIX
+#undef JEMALLOC_CPREFIX
+
+/*
+ * JEMALLOC_PRIVATE_NAMESPACE is used as a prefix for all library-private APIs.
+ * For shared libraries, symbol visibility mechanisms prevent these symbols
+ * from being exported, but for static libraries, naming collisions are a real
+ * possibility.
+ */
+#undef JEMALLOC_PRIVATE_NAMESPACE
+
+/*
+ * Hyper-threaded CPUs may need a special instruction inside spin loops in
+ * order to yield to another virtual CPU.
+ */
+#undef CPU_SPINWAIT
+
+/* Defined if the equivalent of FreeBSD's atomic(9) functions are available. */
+#undef JEMALLOC_ATOMIC9
+
+/*
+ * Defined if OSAtomic*() functions are available, as provided by Darwin, and
+ * documented in the atomic(3) manual page.
+ */
+#undef JEMALLOC_OSATOMIC
+
+/*
+ * Defined if __sync_add_and_fetch(uint32_t *, uint32_t) and
+ * __sync_sub_and_fetch(uint32_t *, uint32_t) are available, despite
+ * __GCC_HAVE_SYNC_COMPARE_AND_SWAP_4 not being defined (which means the
+ * functions are defined in libgcc instead of being inlines)
+ */
+#undef JE_FORCE_SYNC_COMPARE_AND_SWAP_4
+
+/*
+ * Defined if __sync_add_and_fetch(uint64_t *, uint64_t) and
+ * __sync_sub_and_fetch(uint64_t *, uint64_t) are available, despite
+ * __GCC_HAVE_SYNC_COMPARE_AND_SWAP_8 not being defined (which means the
+ * functions are defined in libgcc instead of being inlines)
+ */
+#undef JE_FORCE_SYNC_COMPARE_AND_SWAP_8
+
+/*
+ * Defined if OSSpin*() functions are available, as provided by Darwin, and
+ * documented in the spinlock(3) manual page.
+ */
+#undef JEMALLOC_OSSPIN
+
+/*
+ * Defined if _malloc_thread_cleanup() exists.  At least in the case of
+ * FreeBSD, pthread_key_create() allocates, which if used during malloc
+ * bootstrapping will cause recursion into the pthreads library.  Therefore, if
+ * _malloc_thread_cleanup() exists, use it as the basis for thread cleanup in
+ * malloc_tsd.
+ */
+#undef JEMALLOC_MALLOC_THREAD_CLEANUP
+
+/*
+ * Defined if threaded initialization is known to be safe on this platform.
+ * Among other things, it must be possible to initialize a mutex without
+ * triggering allocation in order for threaded allocation to be safe.
+ */
+#undef JEMALLOC_THREADED_INIT
+
+/*
+ * Defined if the pthreads implementation defines
+ * _pthread_mutex_init_calloc_cb(), in which case the function is used in order
+ * to avoid recursive allocation during mutex initialization.
+ */
+#undef JEMALLOC_MUTEX_INIT_CB
+
+/* Defined if sbrk() is supported. */
+#undef JEMALLOC_HAVE_SBRK
+
+/* Non-empty if the tls_model attribute is supported. */
+#undef JEMALLOC_TLS_MODEL
+
+/* JEMALLOC_CC_SILENCE enables code that silences unuseful compiler warnings. */
+#undef JEMALLOC_CC_SILENCE
+
+/* JEMALLOC_CODE_COVERAGE enables test code coverage analysis. */
+#undef JEMALLOC_CODE_COVERAGE
+
+/*
+ * JEMALLOC_DEBUG enables assertions and other sanity checks, and disables
+ * inline functions.
+ */
+#undef JEMALLOC_DEBUG
+
+/* JEMALLOC_STATS enables statistics calculation. */
+#undef JEMALLOC_STATS
+
+/* JEMALLOC_PROF enables allocation profiling. */
+#undef JEMALLOC_PROF
+
+/* Use libunwind for profile backtracing if defined. */
+#undef JEMALLOC_PROF_LIBUNWIND
+
+/* Use libgcc for profile backtracing if defined. */
+#undef JEMALLOC_PROF_LIBGCC
+
+/* Use gcc intrinsics for profile backtracing if defined. */
+#undef JEMALLOC_PROF_GCC
+
+/*
+ * JEMALLOC_TCACHE enables a thread-specific caching layer for small objects.
+ * This makes it possible to allocate/deallocate objects without any locking
+ * when the cache is in the steady state.
+ */
+#undef JEMALLOC_TCACHE
+
+/*
+ * JEMALLOC_DSS enables use of sbrk(2) to allocate chunks from the data storage
+ * segment (DSS).
+ */
+#undef JEMALLOC_DSS
+
+/* Support memory filling (junk/zero/quarantine/redzone). */
+#undef JEMALLOC_FILL
+
+/* Support utrace(2)-based tracing. */
+#undef JEMALLOC_UTRACE
+
+/* Support Valgrind. */
+#undef JEMALLOC_VALGRIND
+
+/* Support optional abort() on OOM. */
+#undef JEMALLOC_XMALLOC
+
+/* Support lazy locking (avoid locking unless a second thread is launched). */
+#undef JEMALLOC_LAZY_LOCK
+
+/* One page is 2^STATIC_PAGE_SHIFT bytes. */
+#undef STATIC_PAGE_SHIFT
+
+/*
+ * If defined, use munmap() to unmap freed chunks, rather than storing them for
+ * later reuse.  This is disabled by default on Linux because common sequences
+ * of mmap()/munmap() calls will cause virtual memory map holes.
+ */
+#undef JEMALLOC_MUNMAP
+
+/*
+ * If defined, use mremap(...MREMAP_FIXED...) for huge realloc().  This is
+ * disabled by default because it is Linux-specific and it will cause virtual
+ * memory map holes, much like munmap(2) does.
+ */
+#undef JEMALLOC_MREMAP
+
+/* TLS is used to map arenas and magazine caches to threads. */
+#undef JEMALLOC_TLS
+
+/*
+ * JEMALLOC_IVSALLOC enables ivsalloc(), which verifies that pointers reside
+ * within jemalloc-owned chunks before dereferencing them.
+ */
+#undef JEMALLOC_IVSALLOC
+
+/*
+ * Darwin (OS X) uses zones to work around Mach-O symbol override shortcomings.
+ */
+#undef JEMALLOC_ZONE
+#undef JEMALLOC_ZONE_VERSION
+
+/*
+ * Methods for purging unused pages differ between operating systems.
+ *
+ *   madvise(..., MADV_DONTNEED) : On Linux, this immediately discards pages,
+ *                                 such that new pages will be demand-zeroed if
+ *                                 the address region is later touched.
+ *   madvise(..., MADV_FREE) : On FreeBSD and Darwin, this marks pages as being
+ *                             unused, such that they will be discarded rather
+ *                             than swapped out.
+ */
+#undef JEMALLOC_PURGE_MADVISE_DONTNEED
+#undef JEMALLOC_PURGE_MADVISE_FREE
+
+/*
+ * Define if operating system has alloca.h header.
+ */
+#undef JEMALLOC_HAS_ALLOCA_H
+
+/* C99 restrict keyword supported. */
+#undef JEMALLOC_HAS_RESTRICT
+
+/* For use by hash code. */
+#undef JEMALLOC_BIG_ENDIAN
+
+/* sizeof(int) == 2^LG_SIZEOF_INT. */
+#undef LG_SIZEOF_INT
+
+/* sizeof(long) == 2^LG_SIZEOF_LONG. */
+#undef LG_SIZEOF_LONG
+
+/* sizeof(intmax_t) == 2^LG_SIZEOF_INTMAX_T. */
+#undef LG_SIZEOF_INTMAX_T
+
+#endif /* JEMALLOC_INTERNAL_DEFS_H_ */
@@ -0,0 +1,51 @@
+/*
+ * JEMALLOC_ALWAYS_INLINE and JEMALLOC_INLINE are used within header files for
+ * functions that are static inline functions if inlining is enabled, and
+ * single-definition library-private functions if inlining is disabled.
+ *
+ * JEMALLOC_ALWAYS_INLINE_C and JEMALLOC_INLINE_C are for use in .c files, in
+ * which case the denoted functions are always static, regardless of whether
+ * inlining is enabled.
+ */
+#if defined(JEMALLOC_DEBUG) || defined(JEMALLOC_CODE_COVERAGE)
+   /* Disable inlining to make debugging/profiling easier. */
+#  define JEMALLOC_ALWAYS_INLINE
+#  define JEMALLOC_ALWAYS_INLINE_C static
+#  define JEMALLOC_INLINE
+#  define JEMALLOC_INLINE_C static
+#  define inline
+#else
+#  define JEMALLOC_ENABLE_INLINE
+#  ifdef JEMALLOC_HAVE_ATTR
+#    define JEMALLOC_ALWAYS_INLINE \
+	 static inline JEMALLOC_ATTR(unused) JEMALLOC_ATTR(always_inline)
+#    define JEMALLOC_ALWAYS_INLINE_C \
+	 static inline JEMALLOC_ATTR(always_inline)
+#  else
+#    define JEMALLOC_ALWAYS_INLINE static inline
+#    define JEMALLOC_ALWAYS_INLINE_C static inline
+#  endif
+#  define JEMALLOC_INLINE static inline
+#  define JEMALLOC_INLINE_C static inline
+#  ifdef _MSC_VER
+#    define inline _inline
+#  endif
+#endif
+
+#ifdef JEMALLOC_CC_SILENCE
+#  define UNUSED JEMALLOC_ATTR(unused)
+#else
+#  define UNUSED
+#endif
+
+#define	ZU(z)	((size_t)z)
+#define	QU(q)	((uint64_t)q)
+#define	QI(q)	((int64_t)q)
+
+#ifndef __DECONST
+#  define	__DECONST(type, var)	((type)(uintptr_t)(const void *)(var))
+#endif
+
+#ifndef JEMALLOC_HAS_RESTRICT
+#  define restrict
+#endif
@@ -1,367 +0,0 @@
-#define	a0calloc JEMALLOC_N(a0calloc)
-#define	a0free JEMALLOC_N(a0free)
-#define	a0malloc JEMALLOC_N(a0malloc)
-#define	arena_alloc_junk_small JEMALLOC_N(arena_alloc_junk_small)
-#define	arena_bin_index JEMALLOC_N(arena_bin_index)
-#define	arena_bin_info JEMALLOC_N(arena_bin_info)
-#define	arena_boot JEMALLOC_N(arena_boot)
-#define	arena_dalloc JEMALLOC_N(arena_dalloc)
-#define	arena_dalloc_bin JEMALLOC_N(arena_dalloc_bin)
-#define	arena_dalloc_bin_locked JEMALLOC_N(arena_dalloc_bin_locked)
-#define	arena_dalloc_junk_small JEMALLOC_N(arena_dalloc_junk_small)
-#define	arena_dalloc_large JEMALLOC_N(arena_dalloc_large)
-#define	arena_dalloc_large_locked JEMALLOC_N(arena_dalloc_large_locked)
-#define	arena_dalloc_small JEMALLOC_N(arena_dalloc_small)
-#define	arena_dss_prec_get JEMALLOC_N(arena_dss_prec_get)
-#define	arena_dss_prec_set JEMALLOC_N(arena_dss_prec_set)
-#define	arena_malloc JEMALLOC_N(arena_malloc)
-#define	arena_malloc_large JEMALLOC_N(arena_malloc_large)
-#define	arena_malloc_small JEMALLOC_N(arena_malloc_small)
-#define	arena_mapbits_allocated_get JEMALLOC_N(arena_mapbits_allocated_get)
-#define	arena_mapbits_binind_get JEMALLOC_N(arena_mapbits_binind_get)
-#define	arena_mapbits_dirty_get JEMALLOC_N(arena_mapbits_dirty_get)
-#define	arena_mapbits_get JEMALLOC_N(arena_mapbits_get)
-#define	arena_mapbits_large_binind_set JEMALLOC_N(arena_mapbits_large_binind_set)
-#define	arena_mapbits_large_get JEMALLOC_N(arena_mapbits_large_get)
-#define	arena_mapbits_large_set JEMALLOC_N(arena_mapbits_large_set)
-#define	arena_mapbits_large_size_get JEMALLOC_N(arena_mapbits_large_size_get)
-#define	arena_mapbits_small_runind_get JEMALLOC_N(arena_mapbits_small_runind_get)
-#define	arena_mapbits_small_set JEMALLOC_N(arena_mapbits_small_set)
-#define	arena_mapbits_unallocated_set JEMALLOC_N(arena_mapbits_unallocated_set)
-#define	arena_mapbits_unallocated_size_get JEMALLOC_N(arena_mapbits_unallocated_size_get)
-#define	arena_mapbits_unallocated_size_set JEMALLOC_N(arena_mapbits_unallocated_size_set)
-#define	arena_mapbits_unzeroed_get JEMALLOC_N(arena_mapbits_unzeroed_get)
-#define	arena_mapbits_unzeroed_set JEMALLOC_N(arena_mapbits_unzeroed_set)
-#define	arena_mapbitsp_get JEMALLOC_N(arena_mapbitsp_get)
-#define	arena_mapp_get JEMALLOC_N(arena_mapp_get)
-#define	arena_maxclass JEMALLOC_N(arena_maxclass)
-#define	arena_new JEMALLOC_N(arena_new)
-#define	arena_palloc JEMALLOC_N(arena_palloc)
-#define	arena_postfork_child JEMALLOC_N(arena_postfork_child)
-#define	arena_postfork_parent JEMALLOC_N(arena_postfork_parent)
-#define	arena_prefork JEMALLOC_N(arena_prefork)
-#define	arena_prof_accum JEMALLOC_N(arena_prof_accum)
-#define	arena_prof_ctx_get JEMALLOC_N(arena_prof_ctx_get)
-#define	arena_prof_ctx_set JEMALLOC_N(arena_prof_ctx_set)
-#define	arena_prof_promoted JEMALLOC_N(arena_prof_promoted)
-#define	arena_ptr_small_binind_get JEMALLOC_N(arena_ptr_small_binind_get)
-#define	arena_purge_all JEMALLOC_N(arena_purge_all)
-#define	arena_ralloc JEMALLOC_N(arena_ralloc)
-#define	arena_ralloc_no_move JEMALLOC_N(arena_ralloc_no_move)
-#define	arena_run_regind JEMALLOC_N(arena_run_regind)
-#define	arena_salloc JEMALLOC_N(arena_salloc)
-#define	arena_stats_merge JEMALLOC_N(arena_stats_merge)
-#define	arena_tcache_fill_small JEMALLOC_N(arena_tcache_fill_small)
-#define	arenas JEMALLOC_N(arenas)
-#define	arenas_booted JEMALLOC_N(arenas_booted)
-#define	arenas_cleanup JEMALLOC_N(arenas_cleanup)
-#define	arenas_extend JEMALLOC_N(arenas_extend)
-#define	arenas_initialized JEMALLOC_N(arenas_initialized)
-#define	arenas_lock JEMALLOC_N(arenas_lock)
-#define	arenas_tls JEMALLOC_N(arenas_tls)
-#define	arenas_tsd JEMALLOC_N(arenas_tsd)
-#define	arenas_tsd_boot JEMALLOC_N(arenas_tsd_boot)
-#define	arenas_tsd_cleanup_wrapper JEMALLOC_N(arenas_tsd_cleanup_wrapper)
-#define	arenas_tsd_get JEMALLOC_N(arenas_tsd_get)
-#define	arenas_tsd_set JEMALLOC_N(arenas_tsd_set)
-#define	atomic_add_u JEMALLOC_N(atomic_add_u)
-#define	atomic_add_uint32 JEMALLOC_N(atomic_add_uint32)
-#define	atomic_add_uint64 JEMALLOC_N(atomic_add_uint64)
-#define	atomic_add_z JEMALLOC_N(atomic_add_z)
-#define	atomic_sub_u JEMALLOC_N(atomic_sub_u)
-#define	atomic_sub_uint32 JEMALLOC_N(atomic_sub_uint32)
-#define	atomic_sub_uint64 JEMALLOC_N(atomic_sub_uint64)
-#define	atomic_sub_z JEMALLOC_N(atomic_sub_z)
-#define	base_alloc JEMALLOC_N(base_alloc)
-#define	base_boot JEMALLOC_N(base_boot)
-#define	base_calloc JEMALLOC_N(base_calloc)
-#define	base_node_alloc JEMALLOC_N(base_node_alloc)
-#define	base_node_dealloc JEMALLOC_N(base_node_dealloc)
-#define	base_postfork_child JEMALLOC_N(base_postfork_child)
-#define	base_postfork_parent JEMALLOC_N(base_postfork_parent)
-#define	base_prefork JEMALLOC_N(base_prefork)
-#define	bitmap_full JEMALLOC_N(bitmap_full)
-#define	bitmap_get JEMALLOC_N(bitmap_get)
-#define	bitmap_info_init JEMALLOC_N(bitmap_info_init)
-#define	bitmap_info_ngroups JEMALLOC_N(bitmap_info_ngroups)
-#define	bitmap_init JEMALLOC_N(bitmap_init)
-#define	bitmap_set JEMALLOC_N(bitmap_set)
-#define	bitmap_sfu JEMALLOC_N(bitmap_sfu)
-#define	bitmap_size JEMALLOC_N(bitmap_size)
-#define	bitmap_unset JEMALLOC_N(bitmap_unset)
-#define	bt_init JEMALLOC_N(bt_init)
-#define	buferror JEMALLOC_N(buferror)
-#define	choose_arena JEMALLOC_N(choose_arena)
-#define	choose_arena_hard JEMALLOC_N(choose_arena_hard)
-#define	chunk_alloc JEMALLOC_N(chunk_alloc)
-#define	chunk_alloc_dss JEMALLOC_N(chunk_alloc_dss)
-#define	chunk_alloc_mmap JEMALLOC_N(chunk_alloc_mmap)
-#define	chunk_boot JEMALLOC_N(chunk_boot)
-#define	chunk_dealloc JEMALLOC_N(chunk_dealloc)
-#define	chunk_dealloc_mmap JEMALLOC_N(chunk_dealloc_mmap)
-#define	chunk_dss_boot JEMALLOC_N(chunk_dss_boot)
-#define	chunk_dss_postfork_child JEMALLOC_N(chunk_dss_postfork_child)
-#define	chunk_dss_postfork_parent JEMALLOC_N(chunk_dss_postfork_parent)
-#define	chunk_dss_prec_get JEMALLOC_N(chunk_dss_prec_get)
-#define	chunk_dss_prec_set JEMALLOC_N(chunk_dss_prec_set)
-#define	chunk_dss_prefork JEMALLOC_N(chunk_dss_prefork)
-#define	chunk_in_dss JEMALLOC_N(chunk_in_dss)
-#define	chunk_npages JEMALLOC_N(chunk_npages)
-#define	chunk_postfork_child JEMALLOC_N(chunk_postfork_child)
-#define	chunk_postfork_parent JEMALLOC_N(chunk_postfork_parent)
-#define	chunk_prefork JEMALLOC_N(chunk_prefork)
-#define	chunk_unmap JEMALLOC_N(chunk_unmap)
-#define	chunks_mtx JEMALLOC_N(chunks_mtx)
-#define	chunks_rtree JEMALLOC_N(chunks_rtree)
-#define	chunksize JEMALLOC_N(chunksize)
-#define	chunksize_mask JEMALLOC_N(chunksize_mask)
-#define	ckh_bucket_search JEMALLOC_N(ckh_bucket_search)
-#define	ckh_count JEMALLOC_N(ckh_count)
-#define	ckh_delete JEMALLOC_N(ckh_delete)
-#define	ckh_evict_reloc_insert JEMALLOC_N(ckh_evict_reloc_insert)
-#define	ckh_insert JEMALLOC_N(ckh_insert)
-#define	ckh_isearch JEMALLOC_N(ckh_isearch)
-#define	ckh_iter JEMALLOC_N(ckh_iter)
-#define	ckh_new JEMALLOC_N(ckh_new)
-#define	ckh_pointer_hash JEMALLOC_N(ckh_pointer_hash)
-#define	ckh_pointer_keycomp JEMALLOC_N(ckh_pointer_keycomp)
-#define	ckh_rebuild JEMALLOC_N(ckh_rebuild)
-#define	ckh_remove JEMALLOC_N(ckh_remove)
-#define	ckh_search JEMALLOC_N(ckh_search)
-#define	ckh_string_hash JEMALLOC_N(ckh_string_hash)
-#define	ckh_string_keycomp JEMALLOC_N(ckh_string_keycomp)
-#define	ckh_try_bucket_insert JEMALLOC_N(ckh_try_bucket_insert)
-#define	ckh_try_insert JEMALLOC_N(ckh_try_insert)
-#define	ctl_boot JEMALLOC_N(ctl_boot)
-#define	ctl_bymib JEMALLOC_N(ctl_bymib)
-#define	ctl_byname JEMALLOC_N(ctl_byname)
-#define	ctl_nametomib JEMALLOC_N(ctl_nametomib)
-#define	ctl_postfork_child JEMALLOC_N(ctl_postfork_child)
-#define	ctl_postfork_parent JEMALLOC_N(ctl_postfork_parent)
-#define	ctl_prefork JEMALLOC_N(ctl_prefork)
-#define	dss_prec_names JEMALLOC_N(dss_prec_names)
-#define	extent_tree_ad_first JEMALLOC_N(extent_tree_ad_first)
-#define	extent_tree_ad_insert JEMALLOC_N(extent_tree_ad_insert)
-#define	extent_tree_ad_iter JEMALLOC_N(extent_tree_ad_iter)
-#define	extent_tree_ad_iter_recurse JEMALLOC_N(extent_tree_ad_iter_recurse)
-#define	extent_tree_ad_iter_start JEMALLOC_N(extent_tree_ad_iter_start)
-#define	extent_tree_ad_last JEMALLOC_N(extent_tree_ad_last)
-#define	extent_tree_ad_new JEMALLOC_N(extent_tree_ad_new)
-#define	extent_tree_ad_next JEMALLOC_N(extent_tree_ad_next)
-#define	extent_tree_ad_nsearch JEMALLOC_N(extent_tree_ad_nsearch)
-#define	extent_tree_ad_prev JEMALLOC_N(extent_tree_ad_prev)
-#define	extent_tree_ad_psearch JEMALLOC_N(extent_tree_ad_psearch)
-#define	extent_tree_ad_remove JEMALLOC_N(extent_tree_ad_remove)
-#define	extent_tree_ad_reverse_iter JEMALLOC_N(extent_tree_ad_reverse_iter)
-#define	extent_tree_ad_reverse_iter_recurse JEMALLOC_N(extent_tree_ad_reverse_iter_recurse)
-#define	extent_tree_ad_reverse_iter_start JEMALLOC_N(extent_tree_ad_reverse_iter_start)
-#define	extent_tree_ad_search JEMALLOC_N(extent_tree_ad_search)
-#define	extent_tree_szad_first JEMALLOC_N(extent_tree_szad_first)
-#define	extent_tree_szad_insert JEMALLOC_N(extent_tree_szad_insert)
-#define	extent_tree_szad_iter JEMALLOC_N(extent_tree_szad_iter)
-#define	extent_tree_szad_iter_recurse JEMALLOC_N(extent_tree_szad_iter_recurse)
-#define	extent_tree_szad_iter_start JEMALLOC_N(extent_tree_szad_iter_start)
-#define	extent_tree_szad_last JEMALLOC_N(extent_tree_szad_last)
-#define	extent_tree_szad_new JEMALLOC_N(extent_tree_szad_new)
-#define	extent_tree_szad_next JEMALLOC_N(extent_tree_szad_next)
-#define	extent_tree_szad_nsearch JEMALLOC_N(extent_tree_szad_nsearch)
-#define	extent_tree_szad_prev JEMALLOC_N(extent_tree_szad_prev)
-#define	extent_tree_szad_psearch JEMALLOC_N(extent_tree_szad_psearch)
-#define	extent_tree_szad_remove JEMALLOC_N(extent_tree_szad_remove)
-#define	extent_tree_szad_reverse_iter JEMALLOC_N(extent_tree_szad_reverse_iter)
-#define	extent_tree_szad_reverse_iter_recurse JEMALLOC_N(extent_tree_szad_reverse_iter_recurse)
-#define	extent_tree_szad_reverse_iter_start JEMALLOC_N(extent_tree_szad_reverse_iter_start)
-#define	extent_tree_szad_search JEMALLOC_N(extent_tree_szad_search)
-#define	get_errno JEMALLOC_N(get_errno)
-#define	hash JEMALLOC_N(hash)
-#define	huge_allocated JEMALLOC_N(huge_allocated)
-#define	huge_boot JEMALLOC_N(huge_boot)
-#define	huge_dalloc JEMALLOC_N(huge_dalloc)
-#define	huge_malloc JEMALLOC_N(huge_malloc)
-#define	huge_mtx JEMALLOC_N(huge_mtx)
-#define	huge_ndalloc JEMALLOC_N(huge_ndalloc)
-#define	huge_nmalloc JEMALLOC_N(huge_nmalloc)
-#define	huge_palloc JEMALLOC_N(huge_palloc)
-#define	huge_postfork_child JEMALLOC_N(huge_postfork_child)
-#define	huge_postfork_parent JEMALLOC_N(huge_postfork_parent)
-#define	huge_prefork JEMALLOC_N(huge_prefork)
-#define	huge_prof_ctx_get JEMALLOC_N(huge_prof_ctx_get)
-#define	huge_prof_ctx_set JEMALLOC_N(huge_prof_ctx_set)
-#define	huge_ralloc JEMALLOC_N(huge_ralloc)
-#define	huge_ralloc_no_move JEMALLOC_N(huge_ralloc_no_move)
-#define	huge_salloc JEMALLOC_N(huge_salloc)
-#define	iallocm JEMALLOC_N(iallocm)
-#define	icalloc JEMALLOC_N(icalloc)
-#define	icallocx JEMALLOC_N(icallocx)
-#define	idalloc JEMALLOC_N(idalloc)
-#define	idallocx JEMALLOC_N(idallocx)
-#define	imalloc JEMALLOC_N(imalloc)
-#define	imallocx JEMALLOC_N(imallocx)
-#define	ipalloc JEMALLOC_N(ipalloc)
-#define	ipallocx JEMALLOC_N(ipallocx)
-#define	iqalloc JEMALLOC_N(iqalloc)
-#define	iqallocx JEMALLOC_N(iqallocx)
-#define	iralloc JEMALLOC_N(iralloc)
-#define	irallocx JEMALLOC_N(irallocx)
-#define	isalloc JEMALLOC_N(isalloc)
-#define	isthreaded JEMALLOC_N(isthreaded)
-#define	ivsalloc JEMALLOC_N(ivsalloc)
-#define	jemalloc_postfork_child JEMALLOC_N(jemalloc_postfork_child)
-#define	jemalloc_postfork_parent JEMALLOC_N(jemalloc_postfork_parent)
-#define	jemalloc_prefork JEMALLOC_N(jemalloc_prefork)
-#define	malloc_cprintf JEMALLOC_N(malloc_cprintf)
-#define	malloc_mutex_init JEMALLOC_N(malloc_mutex_init)
-#define	malloc_mutex_lock JEMALLOC_N(malloc_mutex_lock)
-#define	malloc_mutex_postfork_child JEMALLOC_N(malloc_mutex_postfork_child)
-#define	malloc_mutex_postfork_parent JEMALLOC_N(malloc_mutex_postfork_parent)
-#define	malloc_mutex_prefork JEMALLOC_N(malloc_mutex_prefork)
-#define	malloc_mutex_unlock JEMALLOC_N(malloc_mutex_unlock)
-#define	malloc_printf JEMALLOC_N(malloc_printf)
-#define	malloc_snprintf JEMALLOC_N(malloc_snprintf)
-#define	malloc_strtoumax JEMALLOC_N(malloc_strtoumax)
-#define	malloc_tsd_boot JEMALLOC_N(malloc_tsd_boot)
-#define	malloc_tsd_cleanup_register JEMALLOC_N(malloc_tsd_cleanup_register)
-#define	malloc_tsd_dalloc JEMALLOC_N(malloc_tsd_dalloc)
-#define	malloc_tsd_malloc JEMALLOC_N(malloc_tsd_malloc)
-#define	malloc_tsd_no_cleanup JEMALLOC_N(malloc_tsd_no_cleanup)
-#define	malloc_vcprintf JEMALLOC_N(malloc_vcprintf)
-#define	malloc_vsnprintf JEMALLOC_N(malloc_vsnprintf)
-#define	malloc_write JEMALLOC_N(malloc_write)
-#define	map_bias JEMALLOC_N(map_bias)
-#define	mb_write JEMALLOC_N(mb_write)
-#define	mutex_boot JEMALLOC_N(mutex_boot)
-#define	narenas_auto JEMALLOC_N(narenas_auto)
-#define	narenas_total JEMALLOC_N(narenas_total)
-#define	narenas_total_get JEMALLOC_N(narenas_total_get)
-#define	ncpus JEMALLOC_N(ncpus)
-#define	nhbins JEMALLOC_N(nhbins)
-#define	opt_abort JEMALLOC_N(opt_abort)
-#define	opt_junk JEMALLOC_N(opt_junk)
-#define	opt_lg_chunk JEMALLOC_N(opt_lg_chunk)
-#define	opt_lg_dirty_mult JEMALLOC_N(opt_lg_dirty_mult)
-#define	opt_lg_prof_interval JEMALLOC_N(opt_lg_prof_interval)
-#define	opt_lg_prof_sample JEMALLOC_N(opt_lg_prof_sample)
-#define	opt_lg_tcache_max JEMALLOC_N(opt_lg_tcache_max)
-#define	opt_narenas JEMALLOC_N(opt_narenas)
-#define	opt_prof JEMALLOC_N(opt_prof)
-#define	opt_prof_accum JEMALLOC_N(opt_prof_accum)
-#define	opt_prof_active JEMALLOC_N(opt_prof_active)
-#define	opt_prof_final JEMALLOC_N(opt_prof_final)
-#define	opt_prof_gdump JEMALLOC_N(opt_prof_gdump)
-#define	opt_prof_leak JEMALLOC_N(opt_prof_leak)
-#define	opt_prof_prefix JEMALLOC_N(opt_prof_prefix)
-#define	opt_quarantine JEMALLOC_N(opt_quarantine)
-#define	opt_redzone JEMALLOC_N(opt_redzone)
-#define	opt_stats_print JEMALLOC_N(opt_stats_print)
-#define	opt_tcache JEMALLOC_N(opt_tcache)
-#define	opt_utrace JEMALLOC_N(opt_utrace)
-#define	opt_valgrind JEMALLOC_N(opt_valgrind)
-#define	opt_xmalloc JEMALLOC_N(opt_xmalloc)
-#define	opt_zero JEMALLOC_N(opt_zero)
-#define	p2rz JEMALLOC_N(p2rz)
-#define	pages_purge JEMALLOC_N(pages_purge)
-#define	pow2_ceil JEMALLOC_N(pow2_ceil)
-#define	prof_backtrace JEMALLOC_N(prof_backtrace)
-#define	prof_boot0 JEMALLOC_N(prof_boot0)
-#define	prof_boot1 JEMALLOC_N(prof_boot1)
-#define	prof_boot2 JEMALLOC_N(prof_boot2)
-#define	prof_ctx_get JEMALLOC_N(prof_ctx_get)
-#define	prof_ctx_set JEMALLOC_N(prof_ctx_set)
-#define	prof_free JEMALLOC_N(prof_free)
-#define	prof_gdump JEMALLOC_N(prof_gdump)
-#define	prof_idump JEMALLOC_N(prof_idump)
-#define	prof_interval JEMALLOC_N(prof_interval)
-#define	prof_lookup JEMALLOC_N(prof_lookup)
-#define	prof_malloc JEMALLOC_N(prof_malloc)
-#define	prof_mdump JEMALLOC_N(prof_mdump)
-#define	prof_postfork_child JEMALLOC_N(prof_postfork_child)
-#define	prof_postfork_parent JEMALLOC_N(prof_postfork_parent)
-#define	prof_prefork JEMALLOC_N(prof_prefork)
-#define	prof_promote JEMALLOC_N(prof_promote)
-#define	prof_realloc JEMALLOC_N(prof_realloc)
-#define	prof_sample_accum_update JEMALLOC_N(prof_sample_accum_update)
-#define	prof_sample_threshold_update JEMALLOC_N(prof_sample_threshold_update)
-#define	prof_tdata_booted JEMALLOC_N(prof_tdata_booted)
-#define	prof_tdata_cleanup JEMALLOC_N(prof_tdata_cleanup)
-#define	prof_tdata_get JEMALLOC_N(prof_tdata_get)
-#define	prof_tdata_init JEMALLOC_N(prof_tdata_init)
-#define	prof_tdata_initialized JEMALLOC_N(prof_tdata_initialized)
-#define	prof_tdata_tls JEMALLOC_N(prof_tdata_tls)
-#define	prof_tdata_tsd JEMALLOC_N(prof_tdata_tsd)
-#define	prof_tdata_tsd_boot JEMALLOC_N(prof_tdata_tsd_boot)
-#define	prof_tdata_tsd_cleanup_wrapper JEMALLOC_N(prof_tdata_tsd_cleanup_wrapper)
-#define	prof_tdata_tsd_get JEMALLOC_N(prof_tdata_tsd_get)
-#define	prof_tdata_tsd_set JEMALLOC_N(prof_tdata_tsd_set)
-#define	quarantine JEMALLOC_N(quarantine)
-#define	quarantine_boot JEMALLOC_N(quarantine_boot)
-#define	quarantine_tsd_boot JEMALLOC_N(quarantine_tsd_boot)
-#define	quarantine_tsd_cleanup_wrapper JEMALLOC_N(quarantine_tsd_cleanup_wrapper)
-#define	quarantine_tsd_get JEMALLOC_N(quarantine_tsd_get)
-#define	quarantine_tsd_set JEMALLOC_N(quarantine_tsd_set)
-#define	register_zone JEMALLOC_N(register_zone)
-#define	rtree_get JEMALLOC_N(rtree_get)
-#define	rtree_get_locked JEMALLOC_N(rtree_get_locked)
-#define	rtree_new JEMALLOC_N(rtree_new)
-#define	rtree_postfork_child JEMALLOC_N(rtree_postfork_child)
-#define	rtree_postfork_parent JEMALLOC_N(rtree_postfork_parent)
-#define	rtree_prefork JEMALLOC_N(rtree_prefork)
-#define	rtree_set JEMALLOC_N(rtree_set)
-#define	s2u JEMALLOC_N(s2u)
-#define	sa2u JEMALLOC_N(sa2u)
-#define	set_errno JEMALLOC_N(set_errno)
-#define	stats_cactive JEMALLOC_N(stats_cactive)
-#define	stats_cactive_add JEMALLOC_N(stats_cactive_add)
-#define	stats_cactive_get JEMALLOC_N(stats_cactive_get)
-#define	stats_cactive_sub JEMALLOC_N(stats_cactive_sub)
-#define	stats_chunks JEMALLOC_N(stats_chunks)
-#define	stats_print JEMALLOC_N(stats_print)
-#define	tcache_alloc_easy JEMALLOC_N(tcache_alloc_easy)
-#define	tcache_alloc_large JEMALLOC_N(tcache_alloc_large)
-#define	tcache_alloc_small JEMALLOC_N(tcache_alloc_small)
-#define	tcache_alloc_small_hard JEMALLOC_N(tcache_alloc_small_hard)
-#define	tcache_arena_associate JEMALLOC_N(tcache_arena_associate)
-#define	tcache_arena_dissociate JEMALLOC_N(tcache_arena_dissociate)
-#define	tcache_bin_flush_large JEMALLOC_N(tcache_bin_flush_large)
-#define	tcache_bin_flush_small JEMALLOC_N(tcache_bin_flush_small)
-#define	tcache_bin_info JEMALLOC_N(tcache_bin_info)
-#define	tcache_boot0 JEMALLOC_N(tcache_boot0)
-#define	tcache_boot1 JEMALLOC_N(tcache_boot1)
-#define	tcache_booted JEMALLOC_N(tcache_booted)
-#define	tcache_create JEMALLOC_N(tcache_create)
-#define	tcache_dalloc_large JEMALLOC_N(tcache_dalloc_large)
-#define	tcache_dalloc_small JEMALLOC_N(tcache_dalloc_small)
-#define	tcache_destroy JEMALLOC_N(tcache_destroy)
-#define	tcache_enabled_booted JEMALLOC_N(tcache_enabled_booted)
-#define	tcache_enabled_get JEMALLOC_N(tcache_enabled_get)
-#define	tcache_enabled_initialized JEMALLOC_N(tcache_enabled_initialized)
-#define	tcache_enabled_set JEMALLOC_N(tcache_enabled_set)
-#define	tcache_enabled_tls JEMALLOC_N(tcache_enabled_tls)
-#define	tcache_enabled_tsd JEMALLOC_N(tcache_enabled_tsd)
-#define	tcache_enabled_tsd_boot JEMALLOC_N(tcache_enabled_tsd_boot)
-#define	tcache_enabled_tsd_cleanup_wrapper JEMALLOC_N(tcache_enabled_tsd_cleanup_wrapper)
-#define	tcache_enabled_tsd_get JEMALLOC_N(tcache_enabled_tsd_get)
-#define	tcache_enabled_tsd_set JEMALLOC_N(tcache_enabled_tsd_set)
-#define	tcache_event JEMALLOC_N(tcache_event)
-#define	tcache_event_hard JEMALLOC_N(tcache_event_hard)
-#define	tcache_flush JEMALLOC_N(tcache_flush)
-#define	tcache_get JEMALLOC_N(tcache_get)
-#define	tcache_initialized JEMALLOC_N(tcache_initialized)
-#define	tcache_maxclass JEMALLOC_N(tcache_maxclass)
-#define	tcache_salloc JEMALLOC_N(tcache_salloc)
-#define	tcache_stats_merge JEMALLOC_N(tcache_stats_merge)
-#define	tcache_thread_cleanup JEMALLOC_N(tcache_thread_cleanup)
-#define	tcache_tls JEMALLOC_N(tcache_tls)
-#define	tcache_tsd JEMALLOC_N(tcache_tsd)
-#define	tcache_tsd_boot JEMALLOC_N(tcache_tsd_boot)
-#define	tcache_tsd_cleanup_wrapper JEMALLOC_N(tcache_tsd_cleanup_wrapper)
-#define	tcache_tsd_get JEMALLOC_N(tcache_tsd_get)
-#define	tcache_tsd_set JEMALLOC_N(tcache_tsd_set)
-#define	thread_allocated_booted JEMALLOC_N(thread_allocated_booted)
-#define	thread_allocated_initialized JEMALLOC_N(thread_allocated_initialized)
-#define	thread_allocated_tls JEMALLOC_N(thread_allocated_tls)
-#define	thread_allocated_tsd JEMALLOC_N(thread_allocated_tsd)
-#define	thread_allocated_tsd_boot JEMALLOC_N(thread_allocated_tsd_boot)
-#define	thread_allocated_tsd_cleanup_wrapper JEMALLOC_N(thread_allocated_tsd_cleanup_wrapper)
-#define	thread_allocated_tsd_get JEMALLOC_N(thread_allocated_tsd_get)
-#define	thread_allocated_tsd_set JEMALLOC_N(thread_allocated_tsd_set)
-#define	u2rz JEMALLOC_N(u2rz)
@@ -0,0 +1,5 @@
+#!/bin/sh
+
+for symbol in `cat $1` ; do
+  echo "#define	${symbol} JEMALLOC_N(${symbol})"
+done
@@ -0,0 +1,413 @@
+a0calloc
+a0free
+a0malloc
+arena_alloc_junk_small
+arena_bin_index
+arena_bin_info
+arena_boot
+arena_dalloc
+arena_dalloc_bin
+arena_dalloc_bin_locked
+arena_dalloc_junk_large
+arena_dalloc_junk_small
+arena_dalloc_large
+arena_dalloc_large_locked
+arena_dalloc_small
+arena_dss_prec_get
+arena_dss_prec_set
+arena_malloc
+arena_malloc_large
+arena_malloc_small
+arena_mapbits_allocated_get
+arena_mapbits_binind_get
+arena_mapbits_dirty_get
+arena_mapbits_get
+arena_mapbits_large_binind_set
+arena_mapbits_large_get
+arena_mapbits_large_set
+arena_mapbits_large_size_get
+arena_mapbits_small_runind_get
+arena_mapbits_small_set
+arena_mapbits_unallocated_set
+arena_mapbits_unallocated_size_get
+arena_mapbits_unallocated_size_set
+arena_mapbits_unzeroed_get
+arena_mapbits_unzeroed_set
+arena_mapbitsp_get
+arena_mapbitsp_read
+arena_mapbitsp_write
+arena_mapp_get
+arena_maxclass
+arena_new
+arena_palloc
+arena_postfork_child
+arena_postfork_parent
+arena_prefork
+arena_prof_accum
+arena_prof_accum_impl
+arena_prof_accum_locked
+arena_prof_ctx_get
+arena_prof_ctx_set
+arena_prof_promoted
+arena_ptr_small_binind_get
+arena_purge_all
+arena_quarantine_junk_small
+arena_ralloc
+arena_ralloc_junk_large
+arena_ralloc_no_move
+arena_redzone_corruption
+arena_run_regind
+arena_salloc
+arena_stats_merge
+arena_tcache_fill_small
+arenas
+arenas_booted
+arenas_cleanup
+arenas_extend
+arenas_initialized
+arenas_lock
+arenas_tls
+arenas_tsd
+arenas_tsd_boot
+arenas_tsd_cleanup_wrapper
+arenas_tsd_get
+arenas_tsd_get_wrapper
+arenas_tsd_init_head
+arenas_tsd_set
+atomic_add_u
+atomic_add_uint32
+atomic_add_uint64
+atomic_add_z
+atomic_sub_u
+atomic_sub_uint32
+atomic_sub_uint64
+atomic_sub_z
+base_alloc
+base_boot
+base_calloc
+base_node_alloc
+base_node_dealloc
+base_postfork_child
+base_postfork_parent
+base_prefork
+bitmap_full
+bitmap_get
+bitmap_info_init
+bitmap_info_ngroups
+bitmap_init
+bitmap_set
+bitmap_sfu
+bitmap_size
+bitmap_unset
+bt_init
+buferror
+choose_arena
+choose_arena_hard
+chunk_alloc
+chunk_alloc_dss
+chunk_alloc_mmap
+chunk_boot
+chunk_dealloc
+chunk_dealloc_mmap
+chunk_dss_boot
+chunk_dss_postfork_child
+chunk_dss_postfork_parent
+chunk_dss_prec_get
+chunk_dss_prec_set
+chunk_dss_prefork
+chunk_in_dss
+chunk_npages
+chunk_postfork_child
+chunk_postfork_parent
+chunk_prefork
+chunk_unmap
+chunks_mtx
+chunks_rtree
+chunksize
+chunksize_mask
+ckh_bucket_search
+ckh_count
+ckh_delete
+ckh_evict_reloc_insert
+ckh_insert
+ckh_isearch
+ckh_iter
+ckh_new
+ckh_pointer_hash
+ckh_pointer_keycomp
+ckh_rebuild
+ckh_remove
+ckh_search
+ckh_string_hash
+ckh_string_keycomp
+ckh_try_bucket_insert
+ckh_try_insert
+ctl_boot
+ctl_bymib
+ctl_byname
+ctl_nametomib
+ctl_postfork_child
+ctl_postfork_parent
+ctl_prefork
+dss_prec_names
+extent_tree_ad_first
+extent_tree_ad_insert
+extent_tree_ad_iter
+extent_tree_ad_iter_recurse
+extent_tree_ad_iter_start
+extent_tree_ad_last
+extent_tree_ad_new
+extent_tree_ad_next
+extent_tree_ad_nsearch
+extent_tree_ad_prev
+extent_tree_ad_psearch
+extent_tree_ad_remove
+extent_tree_ad_reverse_iter
+extent_tree_ad_reverse_iter_recurse
+extent_tree_ad_reverse_iter_start
+extent_tree_ad_search
+extent_tree_szad_first
+extent_tree_szad_insert
+extent_tree_szad_iter
+extent_tree_szad_iter_recurse
+extent_tree_szad_iter_start
+extent_tree_szad_last
+extent_tree_szad_new
+extent_tree_szad_next
+extent_tree_szad_nsearch
+extent_tree_szad_prev
+extent_tree_szad_psearch
+extent_tree_szad_remove
+extent_tree_szad_reverse_iter
+extent_tree_szad_reverse_iter_recurse
+extent_tree_szad_reverse_iter_start
+extent_tree_szad_search
+get_errno
+hash
+hash_fmix_32
+hash_fmix_64
+hash_get_block_32
+hash_get_block_64
+hash_rotl_32
+hash_rotl_64
+hash_x64_128
+hash_x86_128
+hash_x86_32
+huge_allocated
+huge_boot
+huge_dalloc
+huge_dalloc_junk
+huge_dss_prec_get
+huge_malloc
+huge_mtx
+huge_ndalloc
+huge_nmalloc
+huge_palloc
+huge_postfork_child
+huge_postfork_parent
+huge_prefork
+huge_prof_ctx_get
+huge_prof_ctx_set
+huge_ralloc
+huge_ralloc_no_move
+huge_salloc
+iallocm
+icalloc
+icalloct
+idalloc
+idalloct
+imalloc
+imalloct
+ipalloc
+ipalloct
+iqalloc
+iqalloct
+iralloc
+iralloct
+iralloct_realign
+isalloc
+isthreaded
+ivsalloc
+ixalloc
+jemalloc_postfork_child
+jemalloc_postfork_parent
+jemalloc_prefork
+malloc_cprintf
+malloc_mutex_init
+malloc_mutex_lock
+malloc_mutex_postfork_child
+malloc_mutex_postfork_parent
+malloc_mutex_prefork
+malloc_mutex_unlock
+malloc_printf
+malloc_snprintf
+malloc_strtoumax
+malloc_tsd_boot
+malloc_tsd_cleanup_register
+malloc_tsd_dalloc
+malloc_tsd_malloc
+malloc_tsd_no_cleanup
+malloc_vcprintf
+malloc_vsnprintf
+malloc_write
+map_bias
+mb_write
+mutex_boot
+narenas_auto
+narenas_total
+narenas_total_get
+ncpus
+nhbins
+opt_abort
+opt_dss
+opt_junk
+opt_lg_chunk
+opt_lg_dirty_mult
+opt_lg_prof_interval
+opt_lg_prof_sample
+opt_lg_tcache_max
+opt_narenas
+opt_prof
+opt_prof_accum
+opt_prof_active
+opt_prof_final
+opt_prof_gdump
+opt_prof_leak
+opt_prof_prefix
+opt_quarantine
+opt_redzone
+opt_stats_print
+opt_tcache
+opt_utrace
+opt_valgrind
+opt_xmalloc
+opt_zero
+p2rz
+pages_purge
+pow2_ceil
+prof_backtrace
+prof_boot0
+prof_boot1
+prof_boot2
+prof_bt_count
+prof_ctx_get
+prof_ctx_set
+prof_dump_open
+prof_free
+prof_gdump
+prof_idump
+prof_interval
+prof_lookup
+prof_malloc
+prof_mdump
+prof_postfork_child
+prof_postfork_parent
+prof_prefork
+prof_promote
+prof_realloc
+prof_sample_accum_update
+prof_sample_threshold_update
+prof_tdata_booted
+prof_tdata_cleanup
+prof_tdata_get
+prof_tdata_init
+prof_tdata_initialized
+prof_tdata_tls
+prof_tdata_tsd
+prof_tdata_tsd_boot
+prof_tdata_tsd_cleanup_wrapper
+prof_tdata_tsd_get
+prof_tdata_tsd_get_wrapper
+prof_tdata_tsd_init_head
+prof_tdata_tsd_set
+quarantine
+quarantine_alloc_hook
+quarantine_boot
+quarantine_booted
+quarantine_cleanup
+quarantine_init
+quarantine_tls
+quarantine_tsd
+quarantine_tsd_boot
+quarantine_tsd_cleanup_wrapper
+quarantine_tsd_get
+quarantine_tsd_get_wrapper
+quarantine_tsd_init_head
+quarantine_tsd_set
+register_zone
+rtree_delete
+rtree_get
+rtree_get_locked
+rtree_new
+rtree_postfork_child
+rtree_postfork_parent
+rtree_prefork
+rtree_set
+s2u
+sa2u
+set_errno
+small_size2bin
+stats_cactive
+stats_cactive_add
+stats_cactive_get
+stats_cactive_sub
+stats_chunks
+stats_print
+tcache_alloc_easy
+tcache_alloc_large
+tcache_alloc_small
+tcache_alloc_small_hard
+tcache_arena_associate
+tcache_arena_dissociate
+tcache_bin_flush_large
+tcache_bin_flush_small
+tcache_bin_info
+tcache_boot0
+tcache_boot1
+tcache_booted
+tcache_create
+tcache_dalloc_large
+tcache_dalloc_small
+tcache_destroy
+tcache_enabled_booted
+tcache_enabled_get
+tcache_enabled_initialized
+tcache_enabled_set
+tcache_enabled_tls
+tcache_enabled_tsd
+tcache_enabled_tsd_boot
+tcache_enabled_tsd_cleanup_wrapper
+tcache_enabled_tsd_get
+tcache_enabled_tsd_get_wrapper
+tcache_enabled_tsd_init_head
+tcache_enabled_tsd_set
+tcache_event
+tcache_event_hard
+tcache_flush
+tcache_get
+tcache_initialized
+tcache_maxclass
+tcache_salloc
+tcache_stats_merge
+tcache_thread_cleanup
+tcache_tls
+tcache_tsd
+tcache_tsd_boot
+tcache_tsd_cleanup_wrapper
+tcache_tsd_get
+tcache_tsd_get_wrapper
+tcache_tsd_init_head
+tcache_tsd_set
+thread_allocated_booted
+thread_allocated_initialized
+thread_allocated_tls
+thread_allocated_tsd
+thread_allocated_tsd_boot
+thread_allocated_tsd_cleanup_wrapper
+thread_allocated_tsd_get
+thread_allocated_tsd_get_wrapper
+thread_allocated_tsd_init_head
+thread_allocated_tsd_set
+tsd_init_check_recursion
+tsd_init_finish
+u2rz
@@ -0,0 +1,5 @@
+#!/bin/sh
+
+for symbol in `cat $1` ; do
+  echo "#undef ${symbol}"
+done
@@ -25,7 +25,7 @@
 *   uint32_t state      : Seed value.
 *   const uint32_t a, c : See above discussion.
 */
-#define prng32(r, lg_range, state, a, c) do {				\
+#define	prng32(r, lg_range, state, a, c) do {				\
 	assert(lg_range > 0);						\
 	assert(lg_range <= 32);						\
 									\
@@ -35,7 +35,7 @@
 } while (false)

 /* Same as prng32(), but 64 bits of pseudo-randomness, using uint64_t. */
-#define prng64(r, lg_range, state, a, c) do {				\
+#define	prng64(r, lg_range, state, a, c) do {				\
 	assert(lg_range > 0);						\
 	assert(lg_range <= 64);						\
 									\
@@ -8,7 +8,11 @@ typedef struct prof_ctx_s prof_ctx_t;
 typedef struct prof_tdata_s prof_tdata_t;

 /* Option defaults. */
-#define	PROF_PREFIX_DEFAULT		"jeprof"
+#ifdef JEMALLOC_PROF
+#  define PROF_PREFIX_DEFAULT		"jeprof"
+#else
+#  define PROF_PREFIX_DEFAULT		""
+#endif
 #define	LG_PROF_SAMPLE_DEFAULT		19
 #define	LG_PROF_INTERVAL_DEFAULT	-1

@@ -129,6 +133,7 @@ struct prof_ctx_s {
 	 * limbo due to one of:
 	 *   - Initializing per thread counters associated with this ctx.
 	 *   - Preparing to destroy this ctx.
+	 *   - Dumping a heap profile that includes this ctx.
 	 * nlimbo must be 1 (single destroyer) in order to safely destroy the
 	 * ctx.
 	 */
@@ -145,7 +150,11 @@ struct prof_ctx_s {
 	 * this context.
 	 */
 	ql_head(prof_thr_cnt_t)	cnts_ql;
+
+	/* Linkage for list of contexts to be dumped. */
+	ql_elm(prof_ctx_t)	dump_link;
 };
+typedef ql_head(prof_ctx_t) prof_ctx_list_t;

 struct prof_tdata_s {
 	/*
@@ -195,7 +204,12 @@ extern bool	opt_prof_gdump;       /* High-water memory dumping. */
 extern bool	opt_prof_final;       /* Final profile dumping. */
 extern bool	opt_prof_leak;        /* Dump leak summary at exit. */
 extern bool	opt_prof_accum;       /* Report cumulative bytes. */
-extern char	opt_prof_prefix[PATH_MAX + 1];
+extern char	opt_prof_prefix[
+    /* Minimize memory bloat for non-prof builds. */
+#ifdef JEMALLOC_PROF
+    PATH_MAX +
+#endif
+    1];

 /*
 * Profile dump interval, measured in bytes allocated.  Each arena triggers a
@@ -215,6 +229,11 @@ extern bool	prof_promote;
 void	bt_init(prof_bt_t *bt, void **vec);
 void	prof_backtrace(prof_bt_t *bt, unsigned nignore);
 prof_thr_cnt_t	*prof_lookup(prof_bt_t *bt);
+#ifdef JEMALLOC_JET
+size_t	prof_bt_count(void);
+typedef int (prof_dump_open_t)(bool, const char *);
+extern prof_dump_open_t *prof_dump_open;
+#endif
 void	prof_idump(void);
 bool	prof_mdump(const char *filename);
 void	prof_gdump(void);
@@ -237,7 +256,7 @@ void	prof_postfork_child(void);
 									\
 	assert(size == s2u(size));					\
 									\
-	prof_tdata = prof_tdata_get();					\
+	prof_tdata = prof_tdata_get(true);				\
 	if ((uintptr_t)prof_tdata <= (uintptr_t)PROF_TDATA_STATE_MAX) {	\
 		if (prof_tdata != NULL)					\
 			ret = (prof_thr_cnt_t *)(uintptr_t)1U;		\
@@ -286,14 +305,14 @@ void	prof_postfork_child(void);
 #ifndef JEMALLOC_ENABLE_INLINE
 malloc_tsd_protos(JEMALLOC_ATTR(unused), prof_tdata, prof_tdata_t *)

-prof_tdata_t	*prof_tdata_get(void);
+prof_tdata_t	*prof_tdata_get(bool create);
 void	prof_sample_threshold_update(prof_tdata_t *prof_tdata);
 prof_ctx_t	*prof_ctx_get(const void *ptr);
-void	prof_ctx_set(const void *ptr, prof_ctx_t *ctx);
+void	prof_ctx_set(const void *ptr, size_t usize, prof_ctx_t *ctx);
 bool	prof_sample_accum_update(size_t size);
-void	prof_malloc(const void *ptr, size_t size, prof_thr_cnt_t *cnt);
-void	prof_realloc(const void *ptr, size_t size, prof_thr_cnt_t *cnt,
-    size_t old_size, prof_ctx_t *old_ctx);
+void	prof_malloc(const void *ptr, size_t usize, prof_thr_cnt_t *cnt);
+void	prof_realloc(const void *ptr, size_t usize, prof_thr_cnt_t *cnt,
+    size_t old_usize, prof_ctx_t *old_ctx);
 void	prof_free(const void *ptr, size_t size);
 #endif

@@ -304,17 +323,15 @@ malloc_tsd_funcs(JEMALLOC_INLINE, prof_tdata, prof_tdata_t *, NULL,
    prof_tdata_cleanup)

 JEMALLOC_INLINE prof_tdata_t *
-prof_tdata_get(void)
+prof_tdata_get(bool create)
 {
 	prof_tdata_t *prof_tdata;

 	cassert(config_prof);

 	prof_tdata = *prof_tdata_tsd_get();
-	if ((uintptr_t)prof_tdata <= (uintptr_t)PROF_TDATA_STATE_MAX) {
-		if (prof_tdata == NULL)
-			prof_tdata = prof_tdata_init();
-	}
+	if (create && prof_tdata == NULL)
+		prof_tdata = prof_tdata_init();

 	return (prof_tdata);
 }
@@ -322,6 +339,20 @@ prof_tdata_get(void)
 JEMALLOC_INLINE void
 prof_sample_threshold_update(prof_tdata_t *prof_tdata)
 {
+	/*
+	 * The body of this function is compiled out unless heap profiling is
+	 * enabled, so that it is possible to compile jemalloc with floating
+	 * point support completely disabled.  Avoiding floating point code is
+	 * important on memory-constrained systems, but it also enables a
+	 * workaround for versions of glibc that don't properly save/restore
+	 * floating point registers during dynamic lazy symbol loading (which
+	 * internally calls into whatever malloc implementation happens to be
+	 * integrated into the application).  Note that some compilers (e.g.
+	 * gcc 4.8) may use floating point registers for fast memory moves, so
+	 * jemalloc must be compiled with such optimizations disabled (e.g.
+	 * -mno-sse) in order for the workaround to be complete.
+	 */
+#ifdef JEMALLOC_PROF
 	uint64_t r;
 	double u;

@@ -343,7 +374,7 @@ prof_sample_threshold_update(prof_tdata_t *prof_tdata)
 	 *   Luc Devroye
 	 *   Springer-Verlag, New York, 1986
 	 *   pp 500
-	 *   (http://cg.scs.carleton.ca/~luc/rnbookindex.html)
+	 *   (http://luc.devroye.org/rnbookindex.html)
 	 */
 	prng64(r, 53, prof_tdata->prng_state,
 	    UINT64_C(6364136223846793005), UINT64_C(1442695040888963407));
@@ -351,6 +382,7 @@ prof_sample_threshold_update(prof_tdata_t *prof_tdata)
 	prof_tdata->threshold = (uint64_t)(log(u) /
 	    log(1.0 - (1.0 / (double)((uint64_t)1U << opt_lg_prof_sample))))
 	    + (uint64_t)1U;
+#endif
 }

 JEMALLOC_INLINE prof_ctx_t *
@@ -373,7 +405,7 @@ prof_ctx_get(const void *ptr)
 }

 JEMALLOC_INLINE void
-prof_ctx_set(const void *ptr, prof_ctx_t *ctx)
+prof_ctx_set(const void *ptr, size_t usize, prof_ctx_t *ctx)
 {
 	arena_chunk_t *chunk;

@@ -383,7 +415,7 @@ prof_ctx_set(const void *ptr, prof_ctx_t *ctx)
 	chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(ptr);
 	if (chunk != ptr) {
 		/* Region. */
-		arena_prof_ctx_set(ptr, ctx);
+		arena_prof_ctx_set(ptr, usize, ctx);
 	} else
 		huge_prof_ctx_set(ptr, ctx);
 }
@@ -397,7 +429,7 @@ prof_sample_accum_update(size_t size)
 	/* Sampling logic is unnecessary if the interval is 1. */
 	assert(opt_lg_prof_sample != 0);

-	prof_tdata = *prof_tdata_tsd_get();
+	prof_tdata = prof_tdata_get(false);
 	if ((uintptr_t)prof_tdata <= (uintptr_t)PROF_TDATA_STATE_MAX)
 		return (true);

@@ -418,20 +450,20 @@ prof_sample_accum_update(size_t size)
 }

 JEMALLOC_INLINE void
-prof_malloc(const void *ptr, size_t size, prof_thr_cnt_t *cnt)
+prof_malloc(const void *ptr, size_t usize, prof_thr_cnt_t *cnt)
 {

 	cassert(config_prof);
 	assert(ptr != NULL);
-	assert(size == isalloc(ptr, true));
+	assert(usize == isalloc(ptr, true));

 	if (opt_lg_prof_sample != 0) {
-		if (prof_sample_accum_update(size)) {
+		if (prof_sample_accum_update(usize)) {
 			/*
 			 * Don't sample.  For malloc()-like allocation, it is
 			 * always possible to tell in advance how large an
 			 * object's usable size will be, so there should never
-			 * be a difference between the size passed to
+			 * be a difference between the usize passed to
 			 * PROF_ALLOC_PREP() and prof_malloc().
 			 */
 			assert((uintptr_t)cnt == (uintptr_t)1U);
@@ -439,17 +471,17 @@ prof_malloc(const void *ptr, size_t size, prof_thr_cnt_t *cnt)
 	}

 	if ((uintptr_t)cnt > (uintptr_t)1U) {
-		prof_ctx_set(ptr, cnt->ctx);
+		prof_ctx_set(ptr, usize, cnt->ctx);

 		cnt->epoch++;
 		/*********/
 		mb_write();
 		/*********/
 		cnt->cnts.curobjs++;
-		cnt->cnts.curbytes += size;
+		cnt->cnts.curbytes += usize;
 		if (opt_prof_accum) {
 			cnt->cnts.accumobjs++;
-			cnt->cnts.accumbytes += size;
+			cnt->cnts.accumbytes += usize;
 		}
 		/*********/
 		mb_write();
@@ -459,12 +491,12 @@ prof_malloc(const void *ptr, size_t size, prof_thr_cnt_t *cnt)
 		mb_write();
 		/*********/
 	} else
-		prof_ctx_set(ptr, (prof_ctx_t *)(uintptr_t)1U);
+		prof_ctx_set(ptr, usize, (prof_ctx_t *)(uintptr_t)1U);
 }

 JEMALLOC_INLINE void
-prof_realloc(const void *ptr, size_t size, prof_thr_cnt_t *cnt,
-    size_t old_size, prof_ctx_t *old_ctx)
+prof_realloc(const void *ptr, size_t usize, prof_thr_cnt_t *cnt,
+    size_t old_usize, prof_ctx_t *old_ctx)
 {
 	prof_thr_cnt_t *told_cnt;

@@ -472,15 +504,15 @@ prof_realloc(const void *ptr, size_t size, prof_thr_cnt_t *cnt,
 	assert(ptr != NULL || (uintptr_t)cnt <= (uintptr_t)1U);

 	if (ptr != NULL) {
-		assert(size == isalloc(ptr, true));
+		assert(usize == isalloc(ptr, true));
 		if (opt_lg_prof_sample != 0) {
-			if (prof_sample_accum_update(size)) {
+			if (prof_sample_accum_update(usize)) {
 				/*
-				 * Don't sample.  The size passed to
+				 * Don't sample.  The usize passed to
 				 * PROF_ALLOC_PREP() was larger than what
 				 * actually got allocated, so a backtrace was
 				 * captured for this allocation, even though
-				 * its actual size was insufficient to cross
+				 * its actual usize was insufficient to cross
 				 * the sample threshold.
 				 */
 				cnt = (prof_thr_cnt_t *)(uintptr_t)1U;
@@ -497,7 +529,7 @@ prof_realloc(const void *ptr, size_t size, prof_thr_cnt_t *cnt,
 			 */
 			malloc_mutex_lock(old_ctx->lock);
 			old_ctx->cnt_merged.curobjs--;
-			old_ctx->cnt_merged.curbytes -= old_size;
+			old_ctx->cnt_merged.curbytes -= old_usize;
 			malloc_mutex_unlock(old_ctx->lock);
 			told_cnt = (prof_thr_cnt_t *)(uintptr_t)1U;
 		}
@@ -507,23 +539,23 @@ prof_realloc(const void *ptr, size_t size, prof_thr_cnt_t *cnt,
 	if ((uintptr_t)told_cnt > (uintptr_t)1U)
 		told_cnt->epoch++;
 	if ((uintptr_t)cnt > (uintptr_t)1U) {
-		prof_ctx_set(ptr, cnt->ctx);
+		prof_ctx_set(ptr, usize, cnt->ctx);
 		cnt->epoch++;
 	} else if (ptr != NULL)
-		prof_ctx_set(ptr, (prof_ctx_t *)(uintptr_t)1U);
+		prof_ctx_set(ptr, usize, (prof_ctx_t *)(uintptr_t)1U);
 	/*********/
 	mb_write();
 	/*********/
 	if ((uintptr_t)told_cnt > (uintptr_t)1U) {
 		told_cnt->cnts.curobjs--;
-		told_cnt->cnts.curbytes -= old_size;
+		told_cnt->cnts.curbytes -= old_usize;
 	}
 	if ((uintptr_t)cnt > (uintptr_t)1U) {
 		cnt->cnts.curobjs++;
-		cnt->cnts.curbytes += size;
+		cnt->cnts.curbytes += usize;
 		if (opt_prof_accum) {
 			cnt->cnts.accumobjs++;
-			cnt->cnts.accumbytes += size;
+			cnt->cnts.accumbytes += usize;
 		}
 	}
 	/*********/
@@ -0,0 +1,6 @@
+#!/bin/sh
+
+for nm in `cat $1` ; do
+  n=`echo ${nm} |tr ':' ' ' |awk '{print $1}'`
+  echo "#define	je_${n} JEMALLOC_N(${n})"
+done
@@ -0,0 +1,6 @@
+#!/bin/sh
+
+for nm in `cat $1` ; do
+  n=`echo ${nm} |tr ':' ' ' |awk '{print $1}'`
+  echo "#undef je_${n}"
+done
@@ -1,61 +1,61 @@
 /*
 * List definitions.
 */
-#define ql_head(a_type)							\
+#define	ql_head(a_type)							\
 struct {								\
 	a_type *qlh_first;						\
 }

-#define ql_head_initializer(a_head) {NULL}
+#define	ql_head_initializer(a_head) {NULL}

-#define ql_elm(a_type)	qr(a_type)
+#define	ql_elm(a_type)	qr(a_type)

 /* List functions. */
-#define ql_new(a_head) do {						\
+#define	ql_new(a_head) do {						\
 	(a_head)->qlh_first = NULL;					\
 } while (0)

-#define ql_elm_new(a_elm, a_field) qr_new((a_elm), a_field)
+#define	ql_elm_new(a_elm, a_field) qr_new((a_elm), a_field)

-#define ql_first(a_head) ((a_head)->qlh_first)
+#define	ql_first(a_head) ((a_head)->qlh_first)

-#define ql_last(a_head, a_field)					\
+#define	ql_last(a_head, a_field)					\
 	((ql_first(a_head) != NULL)					\
 	    ? qr_prev(ql_first(a_head), a_field) : NULL)

-#define ql_next(a_head, a_elm, a_field)					\
+#define	ql_next(a_head, a_elm, a_field)					\
 	((ql_last(a_head, a_field) != (a_elm))				\
 	    ? qr_next((a_elm), a_field)	: NULL)

-#define ql_prev(a_head, a_elm, a_field)					\
+#define	ql_prev(a_head, a_elm, a_field)					\
 	((ql_first(a_head) != (a_elm)) ? qr_prev((a_elm), a_field)	\
 				       : NULL)

-#define ql_before_insert(a_head, a_qlelm, a_elm, a_field) do {		\
+#define	ql_before_insert(a_head, a_qlelm, a_elm, a_field) do {		\
 	qr_before_insert((a_qlelm), (a_elm), a_field);			\
 	if (ql_first(a_head) == (a_qlelm)) {				\
 		ql_first(a_head) = (a_elm);				\
 	}								\
 } while (0)

-#define ql_after_insert(a_qlelm, a_elm, a_field)			\
+#define	ql_after_insert(a_qlelm, a_elm, a_field)			\
 	qr_after_insert((a_qlelm), (a_elm), a_field)

-#define ql_head_insert(a_head, a_elm, a_field) do {			\
+#define	ql_head_insert(a_head, a_elm, a_field) do {			\
 	if (ql_first(a_head) != NULL) {					\
 		qr_before_insert(ql_first(a_head), (a_elm), a_field);	\
 	}								\
 	ql_first(a_head) = (a_elm);					\
 } while (0)

-#define ql_tail_insert(a_head, a_elm, a_field) do {			\
+#define	ql_tail_insert(a_head, a_elm, a_field) do {			\
 	if (ql_first(a_head) != NULL) {					\
 		qr_before_insert(ql_first(a_head), (a_elm), a_field);	\
 	}								\
 	ql_first(a_head) = qr_next((a_elm), a_field);			\
 } while (0)

-#define ql_remove(a_head, a_elm, a_field) do {				\
+#define	ql_remove(a_head, a_elm, a_field) do {				\
 	if (ql_first(a_head) == (a_elm)) {				\
 		ql_first(a_head) = qr_next(ql_first(a_head), a_field);	\
 	}								\
@@ -66,18 +66,18 @@ struct {								\
 	}								\
 } while (0)

-#define ql_head_remove(a_head, a_type, a_field) do {			\
+#define	ql_head_remove(a_head, a_type, a_field) do {			\
 	a_type *t = ql_first(a_head);					\
 	ql_remove((a_head), t, a_field);				\
 } while (0)

-#define ql_tail_remove(a_head, a_type, a_field) do {			\
+#define	ql_tail_remove(a_head, a_type, a_field) do {			\
 	a_type *t = ql_last(a_head, a_field);				\
 	ql_remove((a_head), t, a_field);				\
 } while (0)

-#define ql_foreach(a_var, a_head, a_field)				\
+#define	ql_foreach(a_var, a_head, a_field)				\
 	qr_foreach((a_var), ql_first(a_head), a_field)

-#define ql_reverse_foreach(a_var, a_head, a_field)			\
+#define	ql_reverse_foreach(a_var, a_head, a_field)			\
 	qr_reverse_foreach((a_var), ql_first(a_head), a_field)
@@ -1,28 +1,28 @@
 /* Ring definitions. */
-#define qr(a_type)							\
+#define	qr(a_type)							\
 struct {								\
 	a_type	*qre_next;						\
 	a_type	*qre_prev;						\
 }

 /* Ring functions. */
-#define qr_new(a_qr, a_field) do {					\
+#define	qr_new(a_qr, a_field) do {					\
 	(a_qr)->a_field.qre_next = (a_qr);				\
 	(a_qr)->a_field.qre_prev = (a_qr);				\
 } while (0)

-#define qr_next(a_qr, a_field) ((a_qr)->a_field.qre_next)
+#define	qr_next(a_qr, a_field) ((a_qr)->a_field.qre_next)

-#define qr_prev(a_qr, a_field) ((a_qr)->a_field.qre_prev)
+#define	qr_prev(a_qr, a_field) ((a_qr)->a_field.qre_prev)

-#define qr_before_insert(a_qrelm, a_qr, a_field) do {			\
+#define	qr_before_insert(a_qrelm, a_qr, a_field) do {			\
 	(a_qr)->a_field.qre_prev = (a_qrelm)->a_field.qre_prev;		\
 	(a_qr)->a_field.qre_next = (a_qrelm);				\
 	(a_qr)->a_field.qre_prev->a_field.qre_next = (a_qr);		\
 	(a_qrelm)->a_field.qre_prev = (a_qr);				\
 } while (0)

-#define qr_after_insert(a_qrelm, a_qr, a_field)				\
+#define	qr_after_insert(a_qrelm, a_qr, a_field)				\
    do									\
    {									\
 	(a_qr)->a_field.qre_next = (a_qrelm)->a_field.qre_next;		\
@@ -31,7 +31,7 @@ struct {								\
 	(a_qrelm)->a_field.qre_next = (a_qr);				\
    } while (0)

-#define qr_meld(a_qr_a, a_qr_b, a_field) do {				\
+#define	qr_meld(a_qr_a, a_qr_b, a_field) do {				\
 	void *t;							\
 	(a_qr_a)->a_field.qre_prev->a_field.qre_next = (a_qr_b);	\
 	(a_qr_b)->a_field.qre_prev->a_field.qre_next = (a_qr_a);	\
@@ -42,10 +42,10 @@ struct {								\

 /* qr_meld() and qr_split() are functionally equivalent, so there's no need to
 * have two copies of the code. */
-#define qr_split(a_qr_a, a_qr_b, a_field)				\
+#define	qr_split(a_qr_a, a_qr_b, a_field)				\
 	qr_meld((a_qr_a), (a_qr_b), a_field)

-#define qr_remove(a_qr, a_field) do {					\
+#define	qr_remove(a_qr, a_field) do {					\
 	(a_qr)->a_field.qre_prev->a_field.qre_next			\
 	    = (a_qr)->a_field.qre_next;					\
 	(a_qr)->a_field.qre_next->a_field.qre_prev			\
@@ -54,13 +54,13 @@ struct {								\
 	(a_qr)->a_field.qre_prev = (a_qr);				\
 } while (0)

-#define qr_foreach(var, a_qr, a_field)					\
+#define	qr_foreach(var, a_qr, a_field)					\
 	for ((var) = (a_qr);						\
 	    (var) != NULL;						\
 	    (var) = (((var)->a_field.qre_next != (a_qr))		\
 	    ? (var)->a_field.qre_next : NULL))

-#define qr_reverse_foreach(var, a_qr, a_field)				\
+#define	qr_reverse_foreach(var, a_qr, a_field)				\
 	for ((var) = ((a_qr) != NULL) ? qr_prev(a_qr, a_field) : NULL;	\
 	    (var) != NULL;						\
 	    (var) = (((var) != (a_qr))					\
@@ -1,6 +1,9 @@
 /******************************************************************************/
 #ifdef JEMALLOC_H_TYPES

+typedef struct quarantine_obj_s quarantine_obj_t;
+typedef struct quarantine_s quarantine_t;
+
 /* Default per thread quarantine size if valgrind is enabled. */
 #define	JEMALLOC_VALGRIND_QUARANTINE_DEFAULT	(ZU(1) << 24)

@@ -8,17 +11,57 @@
 /******************************************************************************/
 #ifdef JEMALLOC_H_STRUCTS

+struct quarantine_obj_s {
+	void	*ptr;
+	size_t	usize;
+};
+
+struct quarantine_s {
+	size_t			curbytes;
+	size_t			curobjs;
+	size_t			first;
+#define	LG_MAXOBJS_INIT 10
+	size_t			lg_maxobjs;
+	quarantine_obj_t	objs[1]; /* Dynamically sized ring buffer. */
+};
+
 #endif /* JEMALLOC_H_STRUCTS */
 /******************************************************************************/
 #ifdef JEMALLOC_H_EXTERNS

+quarantine_t	*quarantine_init(size_t lg_maxobjs);
 void	quarantine(void *ptr);
+void	quarantine_cleanup(void *arg);
 bool	quarantine_boot(void);

 #endif /* JEMALLOC_H_EXTERNS */
 /******************************************************************************/
 #ifdef JEMALLOC_H_INLINES

+#ifndef JEMALLOC_ENABLE_INLINE
+malloc_tsd_protos(JEMALLOC_ATTR(unused), quarantine, quarantine_t *)
+
+void	quarantine_alloc_hook(void);
+#endif
+
+#if (defined(JEMALLOC_ENABLE_INLINE) || defined(JEMALLOC_QUARANTINE_C_))
+malloc_tsd_externs(quarantine, quarantine_t *)
+malloc_tsd_funcs(JEMALLOC_ALWAYS_INLINE, quarantine, quarantine_t *, NULL,
+    quarantine_cleanup)
+
+JEMALLOC_ALWAYS_INLINE void
+quarantine_alloc_hook(void)
+{
+	quarantine_t *quarantine;
+
+	assert(config_fill && opt_quarantine);
+
+	quarantine = *quarantine_tsd_get();
+	if (quarantine == NULL)
+		quarantine_init(LG_MAXOBJS_INIT);
+}
+#endif
+
 #endif /* JEMALLOC_H_INLINES */
 /******************************************************************************/

@@ -22,10 +22,6 @@
 #ifndef RB_H_
 #define	RB_H_

-#if 0
-__FBSDID("$FreeBSD: head/lib/libc/stdlib/rb.h 204493 2010-02-28 22:57:13Z jasone $");
-#endif
-
 #ifdef RB_COMPACT
 /* Node structure. */
 #define	rb_node(a_type)							\
@@ -14,17 +14,18 @@ typedef struct rtree_s rtree_t;
 * Size of each radix tree node (must be a power of 2).  This impacts tree
 * depth.
 */
-#if (LG_SIZEOF_PTR == 2)
-#  define RTREE_NODESIZE (1U << 14)
-#else
-#  define RTREE_NODESIZE CACHELINE
-#endif
+#define	RTREE_NODESIZE (1U << 16)
+
+typedef void *(rtree_alloc_t)(size_t);
+typedef void (rtree_dalloc_t)(void *);

 #endif /* JEMALLOC_H_TYPES */
 /******************************************************************************/
 #ifdef JEMALLOC_H_STRUCTS

 struct rtree_s {
+	rtree_alloc_t	*alloc;
+	rtree_dalloc_t	*dalloc;
 	malloc_mutex_t	mutex;
 	void		**root;
 	unsigned	height;
@@ -35,7 +36,8 @@ struct rtree_s {
 /******************************************************************************/
 #ifdef JEMALLOC_H_EXTERNS

-rtree_t	*rtree_new(unsigned bits);
+rtree_t	*rtree_new(unsigned bits, rtree_alloc_t *alloc, rtree_dalloc_t *dalloc);
+void	rtree_delete(rtree_t *rtree);
 void	rtree_prefork(rtree_t *rtree);
 void	rtree_postfork_parent(rtree_t *rtree);
 void	rtree_postfork_child(rtree_t *rtree);
@@ -45,20 +47,20 @@ void	rtree_postfork_child(rtree_t *rtree);
 #ifdef JEMALLOC_H_INLINES

 #ifndef JEMALLOC_ENABLE_INLINE
-#ifndef JEMALLOC_DEBUG
-void	*rtree_get_locked(rtree_t *rtree, uintptr_t key);
+#ifdef JEMALLOC_DEBUG
+uint8_t rtree_get_locked(rtree_t *rtree, uintptr_t key);
 #endif
-void	*rtree_get(rtree_t *rtree, uintptr_t key);
-bool	rtree_set(rtree_t *rtree, uintptr_t key, void *val);
+uint8_t	rtree_get(rtree_t *rtree, uintptr_t key);
+bool	rtree_set(rtree_t *rtree, uintptr_t key, uint8_t val);
 #endif

 #if (defined(JEMALLOC_ENABLE_INLINE) || defined(JEMALLOC_RTREE_C_))
 #define	RTREE_GET_GENERATE(f)						\
 /* The least significant bits of the key are ignored. */		\
-JEMALLOC_INLINE void *							\
+JEMALLOC_INLINE uint8_t							\
 f(rtree_t *rtree, uintptr_t key)					\
 {									\
-	void *ret;							\
+	uint8_t ret;							\
 	uintptr_t subkey;						\
 	unsigned i, lshift, height, bits;				\
 	void **node, **child;						\
@@ -68,12 +70,12 @@ f(rtree_t *rtree, uintptr_t key)					\
 	    i < height - 1;						\
 	    i++, lshift += bits, node = child) {			\
 		bits = rtree->level2bits[i];				\
-		subkey = (key << lshift) >> ((ZU(1) << (LG_SIZEOF_PTR + \
+		subkey = (key << lshift) >> ((ZU(1) << (LG_SIZEOF_PTR +	\
 		    3)) - bits);					\
 		child = (void**)node[subkey];				\
 		if (child == NULL) {					\
 			RTREE_UNLOCK(&rtree->mutex);			\
-			return (NULL);					\
+			return (0);					\
 		}							\
 	}								\
 									\
@@ -84,7 +86,10 @@ f(rtree_t *rtree, uintptr_t key)					\
 	bits = rtree->level2bits[i];					\
 	subkey = (key << lshift) >> ((ZU(1) << (LG_SIZEOF_PTR+3)) -	\
 	    bits);							\
-	ret = node[subkey];						\
+	{								\
+		uint8_t *leaf = (uint8_t *)node;			\
+		ret = leaf[subkey];					\
+	}								\
 	RTREE_UNLOCK(&rtree->mutex);					\
 									\
 	RTREE_GET_VALIDATE						\
@@ -123,7 +128,7 @@ RTREE_GET_GENERATE(rtree_get)
 #undef RTREE_GET_VALIDATE

 JEMALLOC_INLINE bool
-rtree_set(rtree_t *rtree, uintptr_t key, void *val)
+rtree_set(rtree_t *rtree, uintptr_t key, uint8_t val)
 {
 	uintptr_t subkey;
 	unsigned i, lshift, height, bits;
@@ -138,14 +143,14 @@ rtree_set(rtree_t *rtree, uintptr_t key, void *val)
 		    bits);
 		child = (void**)node[subkey];
 		if (child == NULL) {
-			child = (void**)base_alloc(sizeof(void *) <<
-			    rtree->level2bits[i+1]);
+			size_t size = ((i + 1 < height - 1) ? sizeof(void *)
+			    : (sizeof(uint8_t))) << rtree->level2bits[i+1];
+			child = (void**)rtree->alloc(size);
 			if (child == NULL) {
 				malloc_mutex_unlock(&rtree->mutex);
 				return (true);
 			}
-			memset(child, 0, sizeof(void *) <<
-			    rtree->level2bits[i+1]);
+			memset(child, 0, size);
 			node[subkey] = child;
 		}
 	}
@@ -153,7 +158,10 @@ rtree_set(rtree_t *rtree, uintptr_t key, void *val)
 	/* node is a leaf, so it contains values rather than node pointers. */
 	bits = rtree->level2bits[i];
 	subkey = (key << lshift) >> ((ZU(1) << (LG_SIZEOF_PTR+3)) - bits);
-	node[subkey] = val;
+	{
+		uint8_t *leaf = (uint8_t *)node;
+		leaf[subkey] = val;
+	}
 	malloc_mutex_unlock(&rtree->mutex);

 	return (false);
@@ -140,11 +140,11 @@ void	tcache_dalloc_large(tcache_t *tcache, void *ptr, size_t size);
 #if (defined(JEMALLOC_ENABLE_INLINE) || defined(JEMALLOC_TCACHE_C_))
 /* Map of thread-specific caches. */
 malloc_tsd_externs(tcache, tcache_t *)
-malloc_tsd_funcs(JEMALLOC_INLINE, tcache, tcache_t *, NULL,
+malloc_tsd_funcs(JEMALLOC_ALWAYS_INLINE, tcache, tcache_t *, NULL,
    tcache_thread_cleanup)
 /* Per thread flag that allows thread caches to be disabled. */
 malloc_tsd_externs(tcache_enabled, tcache_enabled_t)
-malloc_tsd_funcs(JEMALLOC_INLINE, tcache_enabled, tcache_enabled_t,
+malloc_tsd_funcs(JEMALLOC_ALWAYS_INLINE, tcache_enabled, tcache_enabled_t,
    tcache_enabled_default, malloc_tsd_no_cleanup)

 JEMALLOC_INLINE void
@@ -206,7 +206,7 @@ tcache_enabled_set(bool enabled)
 	}
 }

-JEMALLOC_INLINE tcache_t *
+JEMALLOC_ALWAYS_INLINE tcache_t *
 tcache_get(bool create)
 {
 	tcache_t *tcache;
@@ -258,7 +258,7 @@ tcache_get(bool create)
 	return (tcache);
 }

-JEMALLOC_INLINE void
+JEMALLOC_ALWAYS_INLINE void
 tcache_event(tcache_t *tcache)
 {

@@ -271,7 +271,7 @@ tcache_event(tcache_t *tcache)
 		tcache_event_hard(tcache);
 }

-JEMALLOC_INLINE void *
+JEMALLOC_ALWAYS_INLINE void *
 tcache_alloc_easy(tcache_bin_t *tbin)
 {
 	void *ret;
@@ -287,7 +287,7 @@ tcache_alloc_easy(tcache_bin_t *tbin)
 	return (ret);
 }

-JEMALLOC_INLINE void *
+JEMALLOC_ALWAYS_INLINE void *
 tcache_alloc_small(tcache_t *tcache, size_t size, bool zero)
 {
 	void *ret;
@@ -297,6 +297,7 @@ tcache_alloc_small(tcache_t *tcache, size_t size, bool zero)
 	binind = SMALL_SIZE2BIN(size);
 	assert(binind < NBINS);
 	tbin = &tcache->tbins[binind];
+	size = arena_bin_info[binind].reg_size;
 	ret = tcache_alloc_easy(tbin);
 	if (ret == NULL) {
 		ret = tcache_alloc_small_hard(tcache, tbin, binind);
@@ -313,6 +314,7 @@ tcache_alloc_small(tcache_t *tcache, size_t size, bool zero)
 			} else if (opt_zero)
 				memset(ret, 0, size);
 		}
+		VALGRIND_MAKE_MEM_UNDEFINED(ret, size);
 	} else {
 		if (config_fill && opt_junk) {
 			arena_alloc_junk_small(ret, &arena_bin_info[binind],
@@ -330,7 +332,7 @@ tcache_alloc_small(tcache_t *tcache, size_t size, bool zero)
 	return (ret);
 }

-JEMALLOC_INLINE void *
+JEMALLOC_ALWAYS_INLINE void *
 tcache_alloc_large(tcache_t *tcache, size_t size, bool zero)
 {
 	void *ret;
@@ -367,6 +369,7 @@ tcache_alloc_large(tcache_t *tcache, size_t size, bool zero)
 				else if (opt_zero)
 					memset(ret, 0, size);
 			}
+			VALGRIND_MAKE_MEM_UNDEFINED(ret, size);
 		} else {
 			VALGRIND_MAKE_MEM_UNDEFINED(ret, size);
 			memset(ret, 0, size);
@@ -382,7 +385,7 @@ tcache_alloc_large(tcache_t *tcache, size_t size, bool zero)
 	return (ret);
 }

-JEMALLOC_INLINE void
+JEMALLOC_ALWAYS_INLINE void
 tcache_dalloc_small(tcache_t *tcache, void *ptr, size_t binind)
 {
 	tcache_bin_t *tbin;
@@ -406,7 +409,7 @@ tcache_dalloc_small(tcache_t *tcache, void *ptr, size_t binind)
 	tcache_event(tcache);
 }

-JEMALLOC_INLINE void
+JEMALLOC_ALWAYS_INLINE void
 tcache_dalloc_large(tcache_t *tcache, void *ptr, size_t size)
 {
 	size_t binind;
@@ -6,6 +6,12 @@

 typedef bool (*malloc_tsd_cleanup_t)(void);

+#if (!defined(JEMALLOC_MALLOC_THREAD_CLEANUP) && !defined(JEMALLOC_TLS) && \
+    !defined(_WIN32))
+typedef struct tsd_init_block_s tsd_init_block_t;
+typedef struct tsd_init_head_s tsd_init_head_t;
+#endif
+
 /*
 * TLS/TSD-agnostic macro-based implementation of thread-specific data.  There
 * are four macros that support (at least) three use cases: file-private,
@@ -75,12 +81,13 @@ extern __thread a_type	a_name##_tls;					\
 extern pthread_key_t	a_name##_tsd;					\
 extern bool		a_name##_booted;
 #elif (defined(_WIN32))
-#define malloc_tsd_externs(a_name, a_type)				\
+#define	malloc_tsd_externs(a_name, a_type)				\
 extern DWORD		a_name##_tsd;					\
 extern bool		a_name##_booted;
 #else
 #define	malloc_tsd_externs(a_name, a_type)				\
 extern pthread_key_t	a_name##_tsd;					\
+extern tsd_init_head_t	a_name##_tsd_init_head;				\
 extern bool		a_name##_booted;
 #endif

@@ -105,6 +112,10 @@ a_attr bool		a_name##_booted = false;
 #else
 #define	malloc_tsd_data(a_attr, a_name, a_type, a_initializer)		\
 a_attr pthread_key_t	a_name##_tsd;					\
+a_attr tsd_init_head_t	a_name##_tsd_init_head = {			\
+	ql_head_initializer(blocks),					\
+	MALLOC_MUTEX_INITIALIZER					\
+};									\
 a_attr bool		a_name##_booted = false;
 #endif

@@ -333,8 +344,14 @@ a_name##_tsd_get_wrapper(void)						\
 	    pthread_getspecific(a_name##_tsd);				\
 									\
 	if (wrapper == NULL) {						\
+		tsd_init_block_t block;					\
+		wrapper = tsd_init_check_recursion(			\
+		    &a_name##_tsd_init_head, &block);			\
+		if (wrapper)						\
+		    return (wrapper);					\
 		wrapper = (a_name##_tsd_wrapper_t *)			\
 		    malloc_tsd_malloc(sizeof(a_name##_tsd_wrapper_t));	\
+		block.data = wrapper;					\
 		if (wrapper == NULL) {					\
 			malloc_write("<jemalloc>: Error allocating"	\
 			    " TSD for "#a_name"\n");			\
@@ -350,6 +367,7 @@ a_name##_tsd_get_wrapper(void)						\
 			    " TSD for "#a_name"\n");			\
 			abort();					\
 		}							\
+		tsd_init_finish(&a_name##_tsd_init_head, &block);	\
 	}								\
 	return (wrapper);						\
 }									\
@@ -379,6 +397,19 @@ a_name##_tsd_set(a_type *val)						\
 /******************************************************************************/
 #ifdef JEMALLOC_H_STRUCTS

+#if (!defined(JEMALLOC_MALLOC_THREAD_CLEANUP) && !defined(JEMALLOC_TLS) && \
+    !defined(_WIN32))
+struct tsd_init_block_s {
+	ql_elm(tsd_init_block_t)	link;
+	pthread_t			thread;
+	void				*data;
+};
+struct tsd_init_head_s {
+	ql_head(tsd_init_block_t)	blocks;
+	malloc_mutex_t			lock;
+};
+#endif
+
 #endif /* JEMALLOC_H_STRUCTS */
 /******************************************************************************/
 #ifdef JEMALLOC_H_EXTERNS
@@ -388,6 +419,12 @@ void	malloc_tsd_dalloc(void *wrapper);
 void	malloc_tsd_no_cleanup(void *);
 void	malloc_tsd_cleanup_register(bool (*f)(void));
 void	malloc_tsd_boot(void);
+#if (!defined(JEMALLOC_MALLOC_THREAD_CLEANUP) && !defined(JEMALLOC_TLS) && \
+    !defined(_WIN32))
+void	*tsd_init_check_recursion(tsd_init_head_t *head,
+    tsd_init_block_t *block);
+void	tsd_init_finish(tsd_init_head_t *head, tsd_init_block_t *block);
+#endif

 #endif /* JEMALLOC_H_EXTERNS */
 /******************************************************************************/
@@ -14,7 +14,7 @@
 * Wrap a cpp argument that contains commas such that it isn't broken up into
 * multiple arguments.
 */
-#define JEMALLOC_CONCAT(...) __VA_ARGS__
+#define	JEMALLOC_ARG_CONCAT(...) __VA_ARGS__

 /*
 * Silence compiler warnings due to uninitialized values.  This is used
@@ -42,12 +42,6 @@
 } while (0)
 #endif

-/* Use to assert a particular configuration, e.g., cassert(config_debug). */
-#define	cassert(c) do {							\
-	if ((c) == false)						\
-		assert(false);						\
-} while (0)
-
 #ifndef not_reached
 #define	not_reached() do {						\
 	if (config_debug) {						\
@@ -69,10 +63,18 @@
 } while (0)
 #endif

+#ifndef assert_not_implemented
 #define	assert_not_implemented(e) do {					\
 	if (config_debug && !(e))					\
 		not_implemented();					\
 } while (0)
+#endif
+
+/* Use to assert a particular configuration, e.g., cassert(config_debug). */
+#define	cassert(c) do {							\
+	if ((c) == false)						\
+		not_reached();						\
+} while (0)

 #endif /* JEMALLOC_H_TYPES */
 /******************************************************************************/
@@ -82,8 +84,9 @@
 /******************************************************************************/
 #ifdef JEMALLOC_H_EXTERNS

-int	buferror(char *buf, size_t buflen);
-uintmax_t	malloc_strtoumax(const char *nptr, char **endptr, int base);
+int	buferror(int err, char *buf, size_t buflen);
+uintmax_t	malloc_strtoumax(const char *restrict nptr,
+    char **restrict endptr, int base);
 void	malloc_write(const char *s);

 /*
@@ -107,7 +110,6 @@ void	malloc_printf(const char *format, ...)

 #ifndef JEMALLOC_ENABLE_INLINE
 size_t	pow2_ceil(size_t x);
-void	malloc_write(const char *s);
 void	set_errno(int errnum);
 int	get_errno(void);
 #endif
@@ -1,157 +0,0 @@
-#ifndef JEMALLOC_H_
-#define	JEMALLOC_H_
-#ifdef __cplusplus
-extern "C" {
-#endif
-
-#include <limits.h>
-#include <strings.h>
-
-#define	JEMALLOC_VERSION "@jemalloc_version@"
-#define	JEMALLOC_VERSION_MAJOR @jemalloc_version_major@
-#define	JEMALLOC_VERSION_MINOR @jemalloc_version_minor@
-#define	JEMALLOC_VERSION_BUGFIX @jemalloc_version_bugfix@
-#define	JEMALLOC_VERSION_NREV @jemalloc_version_nrev@
-#define	JEMALLOC_VERSION_GID "@jemalloc_version_gid@"
-
-#include "jemalloc_defs@install_suffix@.h"
-
-#ifdef JEMALLOC_EXPERIMENTAL
-#define	ALLOCM_LG_ALIGN(la)	(la)
-#if LG_SIZEOF_PTR == 2
-#define	ALLOCM_ALIGN(a)	(ffs(a)-1)
-#else
-#define	ALLOCM_ALIGN(a)	((a < (size_t)INT_MAX) ? ffs(a)-1 : ffs(a>>32)+31)
-#endif
-#define	ALLOCM_ZERO	((int)0x40)
-#define	ALLOCM_NO_MOVE	((int)0x80)
-/* Bias arena index bits so that 0 encodes "ALLOCM_ARENA() unspecified". */
-#define	ALLOCM_ARENA(a)	((int)(((a)+1) << 8))
-
-#define	ALLOCM_SUCCESS		0
-#define	ALLOCM_ERR_OOM		1
-#define	ALLOCM_ERR_NOT_MOVED	2
-#endif
-
-/*
- * The je_ prefix on the following public symbol declarations is an artifact of
- * namespace management, and should be omitted in application code unless
- * JEMALLOC_NO_DEMANGLE is defined (see below).
- */
-extern JEMALLOC_EXPORT const char	*je_malloc_conf;
-extern JEMALLOC_EXPORT void		(*je_malloc_message)(void *cbopaque,
-    const char *s);
-
-JEMALLOC_EXPORT void	*je_malloc(size_t size) JEMALLOC_ATTR(malloc);
-JEMALLOC_EXPORT void	*je_calloc(size_t num, size_t size)
-    JEMALLOC_ATTR(malloc);
-JEMALLOC_EXPORT int	je_posix_memalign(void **memptr, size_t alignment,
-    size_t size) JEMALLOC_ATTR(nonnull(1));
-JEMALLOC_EXPORT void	*je_aligned_alloc(size_t alignment, size_t size)
-    JEMALLOC_ATTR(malloc);
-JEMALLOC_EXPORT void	*je_realloc(void *ptr, size_t size);
-JEMALLOC_EXPORT void	je_free(void *ptr);
-
-#ifdef JEMALLOC_OVERRIDE_MEMALIGN
-JEMALLOC_EXPORT void *	je_memalign(size_t alignment, size_t size)
-    JEMALLOC_ATTR(malloc);
-#endif
-
-#ifdef JEMALLOC_OVERRIDE_VALLOC
-JEMALLOC_EXPORT void *	je_valloc(size_t size) JEMALLOC_ATTR(malloc);
-#endif
-
-JEMALLOC_EXPORT size_t	je_malloc_usable_size(
-    JEMALLOC_USABLE_SIZE_CONST void *ptr);
-JEMALLOC_EXPORT void	je_malloc_stats_print(void (*write_cb)(void *,
-    const char *), void *je_cbopaque, const char *opts);
-JEMALLOC_EXPORT int	je_mallctl(const char *name, void *oldp,
-    size_t *oldlenp, void *newp, size_t newlen);
-JEMALLOC_EXPORT int	je_mallctlnametomib(const char *name, size_t *mibp,
-    size_t *miblenp);
-JEMALLOC_EXPORT int	je_mallctlbymib(const size_t *mib, size_t miblen,
-    void *oldp, size_t *oldlenp, void *newp, size_t newlen);
-
-#ifdef JEMALLOC_EXPERIMENTAL
-JEMALLOC_EXPORT int	je_allocm(void **ptr, size_t *rsize, size_t size,
-    int flags) JEMALLOC_ATTR(nonnull(1));
-JEMALLOC_EXPORT int	je_rallocm(void **ptr, size_t *rsize, size_t size,
-    size_t extra, int flags) JEMALLOC_ATTR(nonnull(1));
-JEMALLOC_EXPORT int	je_sallocm(const void *ptr, size_t *rsize, int flags)
-    JEMALLOC_ATTR(nonnull(1));
-JEMALLOC_EXPORT int	je_dallocm(void *ptr, int flags)
-    JEMALLOC_ATTR(nonnull(1));
-JEMALLOC_EXPORT int	je_nallocm(size_t *rsize, size_t size, int flags);
-#endif
-
-/*
- * By default application code must explicitly refer to mangled symbol names,
- * so that it is possible to use jemalloc in conjunction with another allocator
- * in the same application.  Define JEMALLOC_MANGLE in order to cause automatic
- * name mangling that matches the API prefixing that happened as a result of
- * --with-mangling and/or --with-jemalloc-prefix configuration settings.
- */
-#ifdef JEMALLOC_MANGLE
-#ifndef JEMALLOC_NO_DEMANGLE
-#define	JEMALLOC_NO_DEMANGLE
-#endif
-#define	malloc_conf je_malloc_conf
-#define	malloc_message je_malloc_message
-#define	malloc je_malloc
-#define	calloc je_calloc
-#define	posix_memalign je_posix_memalign
-#define	aligned_alloc je_aligned_alloc
-#define	realloc je_realloc
-#define	free je_free
-#define	malloc_usable_size je_malloc_usable_size
-#define	malloc_stats_print je_malloc_stats_print
-#define	mallctl je_mallctl
-#define	mallctlnametomib je_mallctlnametomib
-#define	mallctlbymib je_mallctlbymib
-#define	memalign je_memalign
-#define	valloc je_valloc
-#ifdef JEMALLOC_EXPERIMENTAL
-#define	allocm je_allocm
-#define	rallocm je_rallocm
-#define	sallocm je_sallocm
-#define	dallocm je_dallocm
-#define	nallocm je_nallocm
-#endif
-#endif
-
-/*
- * The je_* macros can be used as stable alternative names for the public
- * jemalloc API if JEMALLOC_NO_DEMANGLE is defined.  This is primarily meant
- * for use in jemalloc itself, but it can be used by application code to
- * provide isolation from the name mangling specified via --with-mangling
- * and/or --with-jemalloc-prefix.
- */
-#ifndef JEMALLOC_NO_DEMANGLE
-#undef je_malloc_conf
-#undef je_malloc_message
-#undef je_malloc
-#undef je_calloc
-#undef je_posix_memalign
-#undef je_aligned_alloc
-#undef je_realloc
-#undef je_free
-#undef je_malloc_usable_size
-#undef je_malloc_stats_print
-#undef je_mallctl
-#undef je_mallctlnametomib
-#undef je_mallctlbymib
-#undef je_memalign
-#undef je_valloc
-#ifdef JEMALLOC_EXPERIMENTAL
-#undef je_allocm
-#undef je_rallocm
-#undef je_sallocm
-#undef je_dallocm
-#undef je_nallocm
-#endif
-#endif
-
-#ifdef __cplusplus
-};
-#endif
-#endif /* JEMALLOC_H_ */
@@ -0,0 +1,28 @@
+#!/bin/sh
+
+objroot=$1
+
+cat <<EOF
+#ifndef JEMALLOC_H_
+#define	JEMALLOC_H_
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+EOF
+
+for hdr in jemalloc_defs.h jemalloc_rename.h jemalloc_macros.h \
+           jemalloc_protos.h jemalloc_mangle.h ; do
+  cat "${objroot}include/jemalloc/${hdr}" \
+      | grep -v 'Generated from .* by configure\.' \
+      | sed -e 's/^#define /#define	/g' \
+      | sed -e 's/ $//g'
+  echo
+done
+
+cat <<EOF
+#ifdef __cplusplus
+};
+#endif
+#endif /* JEMALLOC_H_ */
+EOF
@@ -1,222 +1,12 @@
-/*
- * If JEMALLOC_PREFIX is defined via --with-jemalloc-prefix, it will cause all
- * public APIs to be prefixed.  This makes it possible, with some care, to use
- * multiple allocators simultaneously.
- */
-#undef JEMALLOC_PREFIX
-#undef JEMALLOC_CPREFIX
-
-/*
- * Name mangling for public symbols is controlled by --with-mangling and
- * --with-jemalloc-prefix.  With default settings the je_ prefix is stripped by
- * these macro definitions.
- */
-#undef je_malloc_conf
-#undef je_malloc_message
-#undef je_malloc
-#undef je_calloc
-#undef je_posix_memalign
-#undef je_aligned_alloc
-#undef je_realloc
-#undef je_free
-#undef je_malloc_usable_size
-#undef je_malloc_stats_print
-#undef je_mallctl
-#undef je_mallctlnametomib
-#undef je_mallctlbymib
-#undef je_memalign
-#undef je_valloc
-#undef je_allocm
-#undef je_rallocm
-#undef je_sallocm
-#undef je_dallocm
-#undef je_nallocm
-
-/*
- * JEMALLOC_PRIVATE_NAMESPACE is used as a prefix for all library-private APIs.
- * For shared libraries, symbol visibility mechanisms prevent these symbols
- * from being exported, but for static libraries, naming collisions are a real
- * possibility.
- */
-#undef JEMALLOC_PRIVATE_NAMESPACE
-#undef JEMALLOC_N
-
-/*
- * Hyper-threaded CPUs may need a special instruction inside spin loops in
- * order to yield to another virtual CPU.
- */
-#undef CPU_SPINWAIT
-
-/* Defined if the equivalent of FreeBSD's atomic(9) functions are available. */
-#undef JEMALLOC_ATOMIC9
-
-/*
- * Defined if OSAtomic*() functions are available, as provided by Darwin, and
- * documented in the atomic(3) manual page.
- */
-#undef JEMALLOC_OSATOMIC
-
-/*
- * Defined if __sync_add_and_fetch(uint32_t *, uint32_t) and
- * __sync_sub_and_fetch(uint32_t *, uint32_t) are available, despite
- * __GCC_HAVE_SYNC_COMPARE_AND_SWAP_4 not being defined (which means the
- * functions are defined in libgcc instead of being inlines)
- */
-#undef JE_FORCE_SYNC_COMPARE_AND_SWAP_4
-
-/*
- * Defined if __sync_add_and_fetch(uint64_t *, uint64_t) and
- * __sync_sub_and_fetch(uint64_t *, uint64_t) are available, despite
- * __GCC_HAVE_SYNC_COMPARE_AND_SWAP_8 not being defined (which means the
- * functions are defined in libgcc instead of being inlines)
- */
-#undef JE_FORCE_SYNC_COMPARE_AND_SWAP_8
-
-/*
- * Defined if OSSpin*() functions are available, as provided by Darwin, and
- * documented in the spinlock(3) manual page.
- */
-#undef JEMALLOC_OSSPIN
-
-/*
- * Defined if _malloc_thread_cleanup() exists.  At least in the case of
- * FreeBSD, pthread_key_create() allocates, which if used during malloc
- * bootstrapping will cause recursion into the pthreads library.  Therefore, if
- * _malloc_thread_cleanup() exists, use it as the basis for thread cleanup in
- * malloc_tsd.
- */
-#undef JEMALLOC_MALLOC_THREAD_CLEANUP
-
-/*
- * Defined if threaded initialization is known to be safe on this platform.
- * Among other things, it must be possible to initialize a mutex without
- * triggering allocation in order for threaded allocation to be safe.
- */
-#undef JEMALLOC_THREADED_INIT
-
-/*
- * Defined if the pthreads implementation defines
- * _pthread_mutex_init_calloc_cb(), in which case the function is used in order
- * to avoid recursive allocation during mutex initialization.
- */
-#undef JEMALLOC_MUTEX_INIT_CB
-
 /* Defined if __attribute__((...)) syntax is supported. */
 #undef JEMALLOC_HAVE_ATTR
-#ifdef JEMALLOC_HAVE_ATTR
-#  define JEMALLOC_ATTR(s) __attribute__((s))
-#  define JEMALLOC_EXPORT JEMALLOC_ATTR(visibility("default"))
-#  define JEMALLOC_ALIGNED(s) JEMALLOC_ATTR(aligned(s))
-#  define JEMALLOC_SECTION(s) JEMALLOC_ATTR(section(s))
-#  define JEMALLOC_NOINLINE JEMALLOC_ATTR(noinline)
-#elif _MSC_VER
-#  define JEMALLOC_ATTR(s)
-#  ifdef DLLEXPORT
-#    define JEMALLOC_EXPORT __declspec(dllexport)
-#  else
-#    define JEMALLOC_EXPORT __declspec(dllimport)
-#  endif
-#  define JEMALLOC_ALIGNED(s) __declspec(align(s))
-#  define JEMALLOC_SECTION(s) __declspec(allocate(s))
-#  define JEMALLOC_NOINLINE __declspec(noinline)
-#else
-#  define JEMALLOC_ATTR(s)
-#  define JEMALLOC_EXPORT
-#  define JEMALLOC_ALIGNED(s)
-#  define JEMALLOC_SECTION(s)
-#  define JEMALLOC_NOINLINE
-#endif
-
-/* Defined if sbrk() is supported. */
-#undef JEMALLOC_HAVE_SBRK
-
-/* Non-empty if the tls_model attribute is supported. */
-#undef JEMALLOC_TLS_MODEL
-
-/* JEMALLOC_CC_SILENCE enables code that silences unuseful compiler warnings. */
-#undef JEMALLOC_CC_SILENCE
-
-/*
- * JEMALLOC_DEBUG enables assertions and other sanity checks, and disables
- * inline functions.
- */
-#undef JEMALLOC_DEBUG
-
-/* JEMALLOC_STATS enables statistics calculation. */
-#undef JEMALLOC_STATS
-
-/* JEMALLOC_PROF enables allocation profiling. */
-#undef JEMALLOC_PROF
-
-/* Use libunwind for profile backtracing if defined. */
-#undef JEMALLOC_PROF_LIBUNWIND
-
-/* Use libgcc for profile backtracing if defined. */
-#undef JEMALLOC_PROF_LIBGCC
-
-/* Use gcc intrinsics for profile backtracing if defined. */
-#undef JEMALLOC_PROF_GCC
-
-/*
- * JEMALLOC_TCACHE enables a thread-specific caching layer for small objects.
- * This makes it possible to allocate/deallocate objects without any locking
- * when the cache is in the steady state.
- */
-#undef JEMALLOC_TCACHE
-
-/*
- * JEMALLOC_DSS enables use of sbrk(2) to allocate chunks from the data storage
- * segment (DSS).
- */
-#undef JEMALLOC_DSS
-
-/* Support memory filling (junk/zero/quarantine/redzone). */
-#undef JEMALLOC_FILL

 /* Support the experimental API. */
 #undef JEMALLOC_EXPERIMENTAL

-/* Support utrace(2)-based tracing. */
-#undef JEMALLOC_UTRACE
-
-/* Support Valgrind. */
-#undef JEMALLOC_VALGRIND
-
-/* Support optional abort() on OOM. */
-#undef JEMALLOC_XMALLOC
-
-/* Support lazy locking (avoid locking unless a second thread is launched). */
-#undef JEMALLOC_LAZY_LOCK
-
-/* One page is 2^STATIC_PAGE_SHIFT bytes. */
-#undef STATIC_PAGE_SHIFT
-
 /*
- * If defined, use munmap() to unmap freed chunks, rather than storing them for
- * later reuse.  This is disabled by default on Linux because common sequences
- * of mmap()/munmap() calls will cause virtual memory map holes.
- */
-#undef JEMALLOC_MUNMAP
-
-/*
- * If defined, use mremap(...MREMAP_FIXED...) for huge realloc().  This is
- * disabled by default because it is Linux-specific and it will cause virtual
- * memory map holes, much like munmap(2) does.
- */
-#undef JEMALLOC_MREMAP
-
-/* TLS is used to map arenas and magazine caches to threads. */
-#undef JEMALLOC_TLS
-
-/*
- * JEMALLOC_IVSALLOC enables ivsalloc(), which verifies that pointers reside
- * within jemalloc-owned chunks before dereferencing them.
- */
-#undef JEMALLOC_IVSALLOC
-
-/*
- * Define overrides for non-standard allocator-related functions if they
- * are present on the system.
+ * Define overrides for non-standard allocator-related functions if they are
+ * present on the system.
 */
 #undef JEMALLOC_OVERRIDE_MEMALIGN
 #undef JEMALLOC_OVERRIDE_VALLOC
@@ -230,33 +20,5 @@
 */
 #undef JEMALLOC_USABLE_SIZE_CONST

-/*
- * Darwin (OS X) uses zones to work around Mach-O symbol override shortcomings.
- */
-#undef JEMALLOC_ZONE
-#undef JEMALLOC_ZONE_VERSION
-
-/*
- * Methods for purging unused pages differ between operating systems.
- *
- *   madvise(..., MADV_DONTNEED) : On Linux, this immediately discards pages,
- *                                 such that new pages will be demand-zeroed if
- *                                 the address region is later touched.
- *   madvise(..., MADV_FREE) : On FreeBSD and Darwin, this marks pages as being
- *                             unused, such that they will be discarded rather
- *                             than swapped out.
- */
-#undef JEMALLOC_PURGE_MADVISE_DONTNEED
-#undef JEMALLOC_PURGE_MADVISE_FREE
-
 /* sizeof(void *) == 2^LG_SIZEOF_PTR. */
 #undef LG_SIZEOF_PTR
-
-/* sizeof(int) == 2^LG_SIZEOF_INT. */
-#undef LG_SIZEOF_INT
-
-/* sizeof(long) == 2^LG_SIZEOF_LONG. */
-#undef LG_SIZEOF_LONG
-
-/* sizeof(intmax_t) == 2^LG_SIZEOF_INTMAX_T. */
-#undef LG_SIZEOF_INTMAX_T
@@ -0,0 +1,61 @@
+#include <limits.h>
+#include <strings.h>
+
+#define	JEMALLOC_VERSION "@jemalloc_version@"
+#define	JEMALLOC_VERSION_MAJOR @jemalloc_version_major@
+#define	JEMALLOC_VERSION_MINOR @jemalloc_version_minor@
+#define	JEMALLOC_VERSION_BUGFIX @jemalloc_version_bugfix@
+#define	JEMALLOC_VERSION_NREV @jemalloc_version_nrev@
+#define	JEMALLOC_VERSION_GID "@jemalloc_version_gid@"
+
+#  define MALLOCX_LG_ALIGN(la)	(la)
+#  if LG_SIZEOF_PTR == 2
+#    define MALLOCX_ALIGN(a)	(ffs(a)-1)
+#  else
+#    define MALLOCX_ALIGN(a)						\
+	 ((a < (size_t)INT_MAX) ? ffs(a)-1 : ffs(a>>32)+31)
+#  endif
+#  define MALLOCX_ZERO	((int)0x40)
+/* Bias arena index bits so that 0 encodes "MALLOCX_ARENA() unspecified". */
+#  define MALLOCX_ARENA(a)	((int)(((a)+1) << 8))
+
+#ifdef JEMALLOC_EXPERIMENTAL
+#  define ALLOCM_LG_ALIGN(la)	(la)
+#  if LG_SIZEOF_PTR == 2
+#    define ALLOCM_ALIGN(a)	(ffs(a)-1)
+#  else
+#    define ALLOCM_ALIGN(a)						\
+	 ((a < (size_t)INT_MAX) ? ffs(a)-1 : ffs(a>>32)+31)
+#  endif
+#  define ALLOCM_ZERO	((int)0x40)
+#  define ALLOCM_NO_MOVE	((int)0x80)
+/* Bias arena index bits so that 0 encodes "ALLOCM_ARENA() unspecified". */
+#  define ALLOCM_ARENA(a)	((int)(((a)+1) << 8))
+#  define ALLOCM_SUCCESS	0
+#  define ALLOCM_ERR_OOM	1
+#  define ALLOCM_ERR_NOT_MOVED	2
+#endif
+
+#ifdef JEMALLOC_HAVE_ATTR
+#  define JEMALLOC_ATTR(s) __attribute__((s))
+#  define JEMALLOC_EXPORT JEMALLOC_ATTR(visibility("default"))
+#  define JEMALLOC_ALIGNED(s) JEMALLOC_ATTR(aligned(s))
+#  define JEMALLOC_SECTION(s) JEMALLOC_ATTR(section(s))
+#  define JEMALLOC_NOINLINE JEMALLOC_ATTR(noinline)
+#elif _MSC_VER
+#  define JEMALLOC_ATTR(s)
+#  ifdef DLLEXPORT
+#    define JEMALLOC_EXPORT __declspec(dllexport)
+#  else
+#    define JEMALLOC_EXPORT __declspec(dllimport)
+#  endif
+#  define JEMALLOC_ALIGNED(s) __declspec(align(s))
+#  define JEMALLOC_SECTION(s) __declspec(allocate(s))
+#  define JEMALLOC_NOINLINE __declspec(noinline)
+#else
+#  define JEMALLOC_ATTR(s)
+#  define JEMALLOC_EXPORT
+#  define JEMALLOC_ALIGNED(s)
+#  define JEMALLOC_SECTION(s)
+#  define JEMALLOC_NOINLINE
+#endif
@@ -0,0 +1,45 @@
+#!/bin/sh
+
+public_symbols_txt=$1
+symbol_prefix=$2
+
+cat <<EOF
+/*
+ * By default application code must explicitly refer to mangled symbol names,
+ * so that it is possible to use jemalloc in conjunction with another allocator
+ * in the same application.  Define JEMALLOC_MANGLE in order to cause automatic
+ * name mangling that matches the API prefixing that happened as a result of
+ * --with-mangling and/or --with-jemalloc-prefix configuration settings.
+ */
+#ifdef JEMALLOC_MANGLE
+#  ifndef JEMALLOC_NO_DEMANGLE
+#    define JEMALLOC_NO_DEMANGLE
+#  endif
+EOF
+
+for nm in `cat ${public_symbols_txt}` ; do
+  n=`echo ${nm} |tr ':' ' ' |awk '{print $1}'`
+  echo "#  define ${n} ${symbol_prefix}${n}"
+done
+
+cat <<EOF
+#endif
+
+/*
+ * The ${symbol_prefix}* macros can be used as stable alternative names for the
+ * public jemalloc API if JEMALLOC_NO_DEMANGLE is defined.  This is primarily
+ * meant for use in jemalloc itself, but it can be used by application code to
+ * provide isolation from the name mangling specified via --with-mangling
+ * and/or --with-jemalloc-prefix.
+ */
+#ifndef JEMALLOC_NO_DEMANGLE
+EOF
+
+for nm in `cat ${public_symbols_txt}` ; do
+  n=`echo ${nm} |tr ':' ' ' |awk '{print $1}'`
+  echo "#  undef ${symbol_prefix}${n}"
+done
+
+cat <<EOF
+#endif
+EOF
@@ -0,0 +1,58 @@
+/*
+ * The @je_@ prefix on the following public symbol declarations is an artifact
+ * of namespace management, and should be omitted in application code unless
+ * JEMALLOC_NO_DEMANGLE is defined (see jemalloc_mangle@install_suffix@.h).
+ */
+extern JEMALLOC_EXPORT const char	*@je_@malloc_conf;
+extern JEMALLOC_EXPORT void		(*@je_@malloc_message)(void *cbopaque,
+    const char *s);
+
+JEMALLOC_EXPORT void	*@je_@malloc(size_t size) JEMALLOC_ATTR(malloc);
+JEMALLOC_EXPORT void	*@je_@calloc(size_t num, size_t size)
+    JEMALLOC_ATTR(malloc);
+JEMALLOC_EXPORT int	@je_@posix_memalign(void **memptr, size_t alignment,
+    size_t size) JEMALLOC_ATTR(nonnull(1));
+JEMALLOC_EXPORT void	*@je_@aligned_alloc(size_t alignment, size_t size)
+    JEMALLOC_ATTR(malloc);
+JEMALLOC_EXPORT void	*@je_@realloc(void *ptr, size_t size);
+JEMALLOC_EXPORT void	@je_@free(void *ptr);
+
+JEMALLOC_EXPORT void	*@je_@mallocx(size_t size, int flags);
+JEMALLOC_EXPORT void	*@je_@rallocx(void *ptr, size_t size, int flags);
+JEMALLOC_EXPORT size_t	@je_@xallocx(void *ptr, size_t size, size_t extra,
+    int flags);
+JEMALLOC_EXPORT size_t	@je_@sallocx(const void *ptr, int flags);
+JEMALLOC_EXPORT void	@je_@dallocx(void *ptr, int flags);
+JEMALLOC_EXPORT size_t	@je_@nallocx(size_t size, int flags);
+
+JEMALLOC_EXPORT int	@je_@mallctl(const char *name, void *oldp,
+    size_t *oldlenp, void *newp, size_t newlen);
+JEMALLOC_EXPORT int	@je_@mallctlnametomib(const char *name, size_t *mibp,
+    size_t *miblenp);
+JEMALLOC_EXPORT int	@je_@mallctlbymib(const size_t *mib, size_t miblen,
+    void *oldp, size_t *oldlenp, void *newp, size_t newlen);
+JEMALLOC_EXPORT void	@je_@malloc_stats_print(void (*write_cb)(void *,
+    const char *), void *@je_@cbopaque, const char *opts);
+JEMALLOC_EXPORT size_t	@je_@malloc_usable_size(
+    JEMALLOC_USABLE_SIZE_CONST void *ptr);
+
+#ifdef JEMALLOC_OVERRIDE_MEMALIGN
+JEMALLOC_EXPORT void *	@je_@memalign(size_t alignment, size_t size)
+    JEMALLOC_ATTR(malloc);
+#endif
+
+#ifdef JEMALLOC_OVERRIDE_VALLOC
+JEMALLOC_EXPORT void *	@je_@valloc(size_t size) JEMALLOC_ATTR(malloc);
+#endif
+
+#ifdef JEMALLOC_EXPERIMENTAL
+JEMALLOC_EXPORT int	@je_@allocm(void **ptr, size_t *rsize, size_t size,
+    int flags) JEMALLOC_ATTR(nonnull(1));
+JEMALLOC_EXPORT int	@je_@rallocm(void **ptr, size_t *rsize, size_t size,
+    size_t extra, int flags) JEMALLOC_ATTR(nonnull(1));
+JEMALLOC_EXPORT int	@je_@sallocm(const void *ptr, size_t *rsize, int flags)
+    JEMALLOC_ATTR(nonnull(1));
+JEMALLOC_EXPORT int	@je_@dallocm(void *ptr, int flags)
+    JEMALLOC_ATTR(nonnull(1));
+JEMALLOC_EXPORT int	@je_@nallocm(size_t *rsize, size_t size, int flags);
+#endif
@@ -0,0 +1,22 @@
+#!/bin/sh
+
+public_symbols_txt=$1
+
+cat <<EOF
+/*
+ * Name mangling for public symbols is controlled by --with-mangling and
+ * --with-jemalloc-prefix.  With default settings the je_ prefix is stripped by
+ * these macro definitions.
+ */
+#ifndef JEMALLOC_NO_RENAME
+EOF
+
+for nm in `cat ${public_symbols_txt}` ; do
+  n=`echo ${nm} |tr ':' ' ' |awk '{print $1}'`
+  m=`echo ${nm} |tr ':' ' ' |awk '{print $2}'`
+  echo "#  define je_${n} ${m}"
+done
+
+cat <<EOF
+#endif
+EOF
@@ -63,6 +63,7 @@ base_alloc(size_t size)
 	ret = base_next_addr;
 	base_next_addr = (void *)((uintptr_t)base_next_addr + csize);
 	malloc_mutex_unlock(&base_mtx);
+	VALGRIND_MAKE_MEM_UNDEFINED(ret, csize);

 	return (ret);
 }
@@ -88,6 +89,7 @@ base_node_alloc(void)
 		ret = base_nodes;
 		base_nodes = *(extent_node_t **)ret;
 		malloc_mutex_unlock(&base_mtx);
+		VALGRIND_MAKE_MEM_UNDEFINED(ret, sizeof(extent_node_t));
 	} else {
 		malloc_mutex_unlock(&base_mtx);
 		ret = (extent_node_t *)base_alloc(sizeof(extent_node_t));
@@ -100,6 +102,7 @@ void
 base_node_dealloc(extent_node_t *node)
 {

+	VALGRIND_MAKE_MEM_UNDEFINED(node, sizeof(extent_node_t));
 	malloc_mutex_lock(&base_mtx);
 	*(extent_node_t **)node = base_nodes;
 	base_nodes = node;
@@ -1,4 +1,4 @@
-#define JEMALLOC_BITMAP_C_
+#define	JEMALLOC_BITMAP_C_
 #include "jemalloc/internal/jemalloc_internal.h"

 /******************************************************************************/
@@ -78,6 +78,9 @@ chunk_recycle(extent_tree_t *chunks_szad, extent_tree_t *chunks_ad, size_t size,
 	assert(node->size >= leadsize + size);
 	trailsize = node->size - leadsize - size;
 	ret = (void *)((uintptr_t)node->addr + leadsize);
+	zeroed = node->zeroed;
+	if (zeroed)
+	    *zero = true;
 	/* Remove node from the tree. */
 	extent_tree_szad_remove(chunks_szad, node);
 	extent_tree_ad_remove(chunks_ad, node);
@@ -108,23 +111,26 @@ chunk_recycle(extent_tree_t *chunks_szad, extent_tree_t *chunks_ad, size_t size,
 		}
 		node->addr = (void *)((uintptr_t)(ret) + size);
 		node->size = trailsize;
+		node->zeroed = zeroed;
 		extent_tree_szad_insert(chunks_szad, node);
 		extent_tree_ad_insert(chunks_ad, node);
 		node = NULL;
 	}
 	malloc_mutex_unlock(&chunks_mtx);

-	zeroed = false;
-	if (node != NULL) {
-		if (node->zeroed) {
-			zeroed = true;
-			*zero = true;
-		}
+	if (node != NULL)
 		base_node_dealloc(node);
-	}
-	if (zeroed == false && *zero) {
-		VALGRIND_MAKE_MEM_UNDEFINED(ret, size);
-		memset(ret, 0, size);
+	if (*zero) {
+		if (zeroed == false)
+			memset(ret, 0, size);
+		else if (config_debug) {
+			size_t i;
+			size_t *p = (size_t *)(uintptr_t)ret;
+
+			VALGRIND_MAKE_MEM_DEFINED(ret, size);
+			for (i = 0; i < size / sizeof(size_t); i++)
+				assert(p[i] == 0);
+		}
 	}
 	return (ret);
 }
@@ -172,35 +178,32 @@ chunk_alloc(size_t size, size_t alignment, bool base, bool *zero,
 	/* All strategies for allocation failed. */
 	ret = NULL;
 label_return:
-	if (config_ivsalloc && base == false && ret != NULL) {
-		if (rtree_set(chunks_rtree, (uintptr_t)ret, ret)) {
-			chunk_dealloc(ret, size, true);
-			return (NULL);
+	if (ret != NULL) {
+		if (config_ivsalloc && base == false) {
+			if (rtree_set(chunks_rtree, (uintptr_t)ret, 1)) {
+				chunk_dealloc(ret, size, true);
+				return (NULL);
+			}
 		}
-	}
-	if ((config_stats || config_prof) && ret != NULL) {
-		bool gdump;
-		malloc_mutex_lock(&chunks_mtx);
-		if (config_stats)
-			stats_chunks.nchunks += (size / chunksize);
-		stats_chunks.curchunks += (size / chunksize);
-		if (stats_chunks.curchunks > stats_chunks.highchunks) {
-			stats_chunks.highchunks = stats_chunks.curchunks;
-			if (config_prof)
-				gdump = true;
-		} else if (config_prof)
-			gdump = false;
-		malloc_mutex_unlock(&chunks_mtx);
-		if (config_prof && opt_prof && opt_prof_gdump && gdump)
-			prof_gdump();
-	}
-	if (config_debug && *zero && ret != NULL) {
-		size_t i;
-		size_t *p = (size_t *)(uintptr_t)ret;
-
-		VALGRIND_MAKE_MEM_DEFINED(ret, size);
-		for (i = 0; i < size / sizeof(size_t); i++)
-			assert(p[i] == 0);
+		if (config_stats || config_prof) {
+			bool gdump;
+			malloc_mutex_lock(&chunks_mtx);
+			if (config_stats)
+				stats_chunks.nchunks += (size / chunksize);
+			stats_chunks.curchunks += (size / chunksize);
+			if (stats_chunks.curchunks > stats_chunks.highchunks) {
+				stats_chunks.highchunks =
+				    stats_chunks.curchunks;
+				if (config_prof)
+					gdump = true;
+			} else if (config_prof)
+				gdump = false;
+			malloc_mutex_unlock(&chunks_mtx);
+			if (config_prof && opt_prof && opt_prof_gdump && gdump)
+				prof_gdump();
+		}
+		if (config_valgrind)
+			VALGRIND_MAKE_MEM_UNDEFINED(ret, size);
 	}
 	assert(CHUNK_ADDR2BASE(ret) == ret);
 	return (ret);
@@ -211,9 +214,10 @@ chunk_record(extent_tree_t *chunks_szad, extent_tree_t *chunks_ad, void *chunk,
    size_t size)
 {
 	bool unzeroed;
-	extent_node_t *xnode, *node, *prev, key;
+	extent_node_t *xnode, *node, *prev, *xprev, key;

 	unzeroed = pages_purge(chunk, size);
+	VALGRIND_MAKE_MEM_NOACCESS(chunk, size);

 	/*
 	 * Allocate a node before acquiring chunks_mtx even though it might not
@@ -222,6 +226,8 @@ chunk_record(extent_tree_t *chunks_szad, extent_tree_t *chunks_ad, void *chunk,
 	 * held.
 	 */
 	xnode = base_node_alloc();
+	/* Use xprev to implement conditional deferred deallocation of prev. */
+	xprev = NULL;

 	malloc_mutex_lock(&chunks_mtx);
 	key.addr = (void *)((uintptr_t)chunk + size);
@@ -238,8 +244,6 @@ chunk_record(extent_tree_t *chunks_szad, extent_tree_t *chunks_ad, void *chunk,
 		node->size += size;
 		node->zeroed = (node->zeroed && (unzeroed == false));
 		extent_tree_szad_insert(chunks_szad, node);
-		if (xnode != NULL)
-			base_node_dealloc(xnode);
 	} else {
 		/* Coalescing forward failed, so insert a new node. */
 		if (xnode == NULL) {
@@ -249,10 +253,10 @@ chunk_record(extent_tree_t *chunks_szad, extent_tree_t *chunks_ad, void *chunk,
 			 * already been purged, so this is only a virtual
 			 * memory leak.
 			 */
-			malloc_mutex_unlock(&chunks_mtx);
-			return;
+			goto label_return;
 		}
 		node = xnode;
+		xnode = NULL; /* Prevent deallocation below. */
 		node->addr = chunk;
 		node->size = size;
 		node->zeroed = (unzeroed == false);
@@ -278,9 +282,19 @@ chunk_record(extent_tree_t *chunks_szad, extent_tree_t *chunks_ad, void *chunk,
 		node->zeroed = (node->zeroed && prev->zeroed);
 		extent_tree_szad_insert(chunks_szad, node);

-		base_node_dealloc(prev);
+		xprev = prev;
 	}
+
+label_return:
 	malloc_mutex_unlock(&chunks_mtx);
+	/*
+	 * Deallocate xnode and/or xprev after unlocking chunks_mtx in order to
+	 * avoid potential deadlock.
+	 */
+	if (xnode != NULL)
+		base_node_dealloc(xnode);
+	if (xprev != NULL)
+		base_node_dealloc(xprev);
 }

 void
@@ -307,7 +321,7 @@ chunk_dealloc(void *chunk, size_t size, bool unmap)
 	assert((size & chunksize_mask) == 0);

 	if (config_ivsalloc)
-		rtree_set(chunks_rtree, (uintptr_t)chunk, NULL);
+		rtree_set(chunks_rtree, (uintptr_t)chunk, 0);
 	if (config_stats || config_prof) {
 		malloc_mutex_lock(&chunks_mtx);
 		assert(stats_chunks.curchunks >= (size / chunksize));
@@ -342,7 +356,7 @@ chunk_boot(void)
 	extent_tree_ad_new(&chunks_ad_dss);
 	if (config_ivsalloc) {
 		chunks_rtree = rtree_new((ZU(1) << (LG_SIZEOF_PTR+3)) -
-		    opt_lg_chunk);
+		    opt_lg_chunk, base_alloc, NULL);
 		if (chunks_rtree == NULL)
 			return (true);
 	}
@@ -354,7 +368,7 @@ void
 chunk_prefork(void)
 {

-	malloc_mutex_lock(&chunks_mtx);
+	malloc_mutex_prefork(&chunks_mtx);
 	if (config_ivsalloc)
 		rtree_prefork(chunks_rtree);
 	chunk_dss_prefork();
@@ -28,16 +28,17 @@ static void		*dss_max;

 /******************************************************************************/

-#ifndef JEMALLOC_HAVE_SBRK
 static void *
-sbrk(intptr_t increment)
+chunk_dss_sbrk(intptr_t increment)
 {

+#ifdef JEMALLOC_HAVE_SBRK
+	return (sbrk(increment));
+#else
 	not_implemented();
-
 	return (NULL);
-}
 #endif
+}

 dss_prec_t
 chunk_dss_prec_get(void)
@@ -93,7 +94,7 @@ chunk_alloc_dss(size_t size, size_t alignment, bool *zero)
 		 */
 		do {
 			/* Get the current end of the DSS. */
-			dss_max = sbrk(0);
+			dss_max = chunk_dss_sbrk(0);
 			/*
 			 * Calculate how much padding is necessary to
 			 * chunk-align the end of the DSS.
@@ -117,7 +118,7 @@ chunk_alloc_dss(size_t size, size_t alignment, bool *zero)
 				return (NULL);
 			}
 			incr = gap_size + cpad_size + size;
-			dss_prev = sbrk(incr);
+			dss_prev = chunk_dss_sbrk(incr);
 			if (dss_prev == dss_max) {
 				/* Success. */
 				dss_max = dss_next;
@@ -163,7 +164,7 @@ chunk_dss_boot(void)

 	if (malloc_mutex_init(&dss_mtx))
 		return (true);
-	dss_base = sbrk(0);
+	dss_base = chunk_dss_sbrk(0);
 	dss_prev = dss_base;
 	dss_max = dss_base;

@@ -43,7 +43,7 @@ pages_map(void *addr, size_t size)
 		if (munmap(ret, size) == -1) {
 			char buf[BUFERROR_BUF];

-			buferror(buf, sizeof(buf));
+			buferror(get_errno(), buf, sizeof(buf));
 			malloc_printf("<jemalloc: Error in munmap(): %s\n",
 			    buf);
 			if (opt_abort)
@@ -69,7 +69,7 @@ pages_unmap(void *addr, size_t size)
 	{
 		char buf[BUFERROR_BUF];

-		buferror(buf, sizeof(buf));
+		buferror(get_errno(), buf, sizeof(buf));
 		malloc_printf("<jemalloc>: Error in "
 #ifdef _WIN32
 		              "VirtualFree"
@@ -49,7 +49,7 @@ static void	ckh_shrink(ckh_t *ckh);
 * Search bucket for key and return the cell number if found; SIZE_T_MAX
 * otherwise.
 */
-JEMALLOC_INLINE size_t
+JEMALLOC_INLINE_C size_t
 ckh_bucket_search(ckh_t *ckh, size_t bucket, const void *key)
 {
 	ckhc_t *cell;
@@ -67,28 +67,28 @@ ckh_bucket_search(ckh_t *ckh, size_t bucket, const void *key)
 /*
 * Search table for key and return cell number if found; SIZE_T_MAX otherwise.
 */
-JEMALLOC_INLINE size_t
+JEMALLOC_INLINE_C size_t
 ckh_isearch(ckh_t *ckh, const void *key)
 {
-	size_t hash1, hash2, bucket, cell;
+	size_t hashes[2], bucket, cell;

 	assert(ckh != NULL);

-	ckh->hash(key, ckh->lg_curbuckets, &hash1, &hash2);
+	ckh->hash(key, hashes);

 	/* Search primary bucket. */
-	bucket = hash1 & ((ZU(1) << ckh->lg_curbuckets) - 1);
+	bucket = hashes[0] & ((ZU(1) << ckh->lg_curbuckets) - 1);
 	cell = ckh_bucket_search(ckh, bucket, key);
 	if (cell != SIZE_T_MAX)
 		return (cell);

 	/* Search secondary bucket. */
-	bucket = hash2 & ((ZU(1) << ckh->lg_curbuckets) - 1);
+	bucket = hashes[1] & ((ZU(1) << ckh->lg_curbuckets) - 1);
 	cell = ckh_bucket_search(ckh, bucket, key);
 	return (cell);
 }

-JEMALLOC_INLINE bool
+JEMALLOC_INLINE_C bool
 ckh_try_bucket_insert(ckh_t *ckh, size_t bucket, const void *key,
    const void *data)
 {
@@ -120,13 +120,13 @@ ckh_try_bucket_insert(ckh_t *ckh, size_t bucket, const void *key,
 * eviction/relocation procedure until either success or detection of an
 * eviction/relocation bucket cycle.
 */
-JEMALLOC_INLINE bool
+JEMALLOC_INLINE_C bool
 ckh_evict_reloc_insert(ckh_t *ckh, size_t argbucket, void const **argkey,
    void const **argdata)
 {
 	const void *key, *data, *tkey, *tdata;
 	ckhc_t *cell;
-	size_t hash1, hash2, bucket, tbucket;
+	size_t hashes[2], bucket, tbucket;
 	unsigned i;

 	bucket = argbucket;
@@ -155,10 +155,11 @@ ckh_evict_reloc_insert(ckh_t *ckh, size_t argbucket, void const **argkey,
 #endif

 		/* Find the alternate bucket for the evicted item. */
-		ckh->hash(key, ckh->lg_curbuckets, &hash1, &hash2);
-		tbucket = hash2 & ((ZU(1) << ckh->lg_curbuckets) - 1);
+		ckh->hash(key, hashes);
+		tbucket = hashes[1] & ((ZU(1) << ckh->lg_curbuckets) - 1);
 		if (tbucket == bucket) {
-			tbucket = hash1 & ((ZU(1) << ckh->lg_curbuckets) - 1);
+			tbucket = hashes[0] & ((ZU(1) << ckh->lg_curbuckets)
+			    - 1);
 			/*
 			 * It may be that (tbucket == bucket) still, if the
 			 * item's hashes both indicate this bucket.  However,
@@ -189,22 +190,22 @@ ckh_evict_reloc_insert(ckh_t *ckh, size_t argbucket, void const **argkey,
 	}
 }

-JEMALLOC_INLINE bool
+JEMALLOC_INLINE_C bool
 ckh_try_insert(ckh_t *ckh, void const**argkey, void const**argdata)
 {
-	size_t hash1, hash2, bucket;
+	size_t hashes[2], bucket;
 	const void *key = *argkey;
 	const void *data = *argdata;

-	ckh->hash(key, ckh->lg_curbuckets, &hash1, &hash2);
+	ckh->hash(key, hashes);

 	/* Try to insert in primary bucket. */
-	bucket = hash1 & ((ZU(1) << ckh->lg_curbuckets) - 1);
+	bucket = hashes[0] & ((ZU(1) << ckh->lg_curbuckets) - 1);
 	if (ckh_try_bucket_insert(ckh, bucket, key, data) == false)
 		return (false);

 	/* Try to insert in secondary bucket. */
-	bucket = hash2 & ((ZU(1) << ckh->lg_curbuckets) - 1);
+	bucket = hashes[1] & ((ZU(1) << ckh->lg_curbuckets) - 1);
 	if (ckh_try_bucket_insert(ckh, bucket, key, data) == false)
 		return (false);

@@ -218,7 +219,7 @@ ckh_try_insert(ckh_t *ckh, void const**argkey, void const**argdata)
 * Try to rebuild the hash table from scratch by inserting all items from the
 * old table into the new.
 */
-JEMALLOC_INLINE bool
+JEMALLOC_INLINE_C bool
 ckh_rebuild(ckh_t *ckh, ckhc_t *aTab)
 {
 	size_t count, i, nins;
@@ -417,9 +418,8 @@ ckh_delete(ckh_t *ckh)
 #endif

 	idalloc(ckh->tab);
-#ifdef JEMALLOC_DEBUG
-	memset(ckh, 0x5a, sizeof(ckh_t));
-#endif
+	if (config_debug)
+		memset(ckh, 0x5a, sizeof(ckh_t));
 }

 size_t
@@ -526,31 +526,10 @@ ckh_search(ckh_t *ckh, const void *searchkey, void **key, void **data)
 }

 void
-ckh_string_hash(const void *key, unsigned minbits, size_t *hash1, size_t *hash2)
+ckh_string_hash(const void *key, size_t r_hash[2])
 {
-	size_t ret1, ret2;
-	uint64_t h;

-	assert(minbits <= 32 || (SIZEOF_PTR == 8 && minbits <= 64));
-	assert(hash1 != NULL);
-	assert(hash2 != NULL);
-
-	h = hash(key, strlen((const char *)key), UINT64_C(0x94122f335b332aea));
-	if (minbits <= 32) {
-		/*
-		 * Avoid doing multiple hashes, since a single hash provides
-		 * enough bits.
-		 */
-		ret1 = h & ZU(0xffffffffU);
-		ret2 = h >> 32;
-	} else {
-		ret1 = h;
-		ret2 = hash(key, strlen((const char *)key),
-		    UINT64_C(0x8432a476666bbc13));
-	}
-
-	*hash1 = ret1;
-	*hash2 = ret2;
+	hash(key, strlen((const char *)key), 0x94122f33U, r_hash);
 }

 bool
@@ -564,41 +543,16 @@ ckh_string_keycomp(const void *k1, const void *k2)
 }

 void
-ckh_pointer_hash(const void *key, unsigned minbits, size_t *hash1,
-    size_t *hash2)
+ckh_pointer_hash(const void *key, size_t r_hash[2])
 {
-	size_t ret1, ret2;
-	uint64_t h;
 	union {
 		const void	*v;
-		uint64_t	i;
+		size_t		i;
 	} u;

-	assert(minbits <= 32 || (SIZEOF_PTR == 8 && minbits <= 64));
-	assert(hash1 != NULL);
-	assert(hash2 != NULL);
-
 	assert(sizeof(u.v) == sizeof(u.i));
-#if (LG_SIZEOF_PTR != LG_SIZEOF_INT)
-	u.i = 0;
-#endif
 	u.v = key;
-	h = hash(&u.i, sizeof(u.i), UINT64_C(0xd983396e68886082));
-	if (minbits <= 32) {
-		/*
-		 * Avoid doing multiple hashes, since a single hash provides
-		 * enough bits.
-		 */
-		ret1 = h & ZU(0xffffffffU);
-		ret2 = h >> 32;
-	} else {
-		assert(SIZEOF_PTR == 8);
-		ret1 = h;
-		ret2 = hash(&u.i, sizeof(u.i), UINT64_C(0x5e2be9aff8709a5d));
-	}
-
-	*hash1 = ret1;
-	*hash2 = ret2;
+	hash(&u.i, sizeof(u.i), 0xd983396eU, r_hash);
 }

 bool
@@ -546,43 +546,30 @@ ctl_arena_refresh(arena_t *arena, unsigned i)
 static bool
 ctl_grow(void)
 {
-	size_t astats_size;
 	ctl_arena_stats_t *astats;
 	arena_t **tarenas;

-	/* Extend arena stats and arenas arrays. */
-	astats_size = (ctl_stats.narenas + 2) * sizeof(ctl_arena_stats_t);
-	if (ctl_stats.narenas == narenas_auto) {
-		/* ctl_stats.arenas and arenas came from base_alloc(). */
-		astats = (ctl_arena_stats_t *)imalloc(astats_size);
-		if (astats == NULL)
-			return (true);
-		memcpy(astats, ctl_stats.arenas, (ctl_stats.narenas + 1) *
-		    sizeof(ctl_arena_stats_t));
-
-		tarenas = (arena_t **)imalloc((ctl_stats.narenas + 1) *
-		    sizeof(arena_t *));
-		if (tarenas == NULL) {
-			idalloc(astats);
-			return (true);
-		}
-		memcpy(tarenas, arenas, ctl_stats.narenas * sizeof(arena_t *));
-	} else {
-		astats = (ctl_arena_stats_t *)iralloc(ctl_stats.arenas,
-		    astats_size, 0, 0, false, false);
-		if (astats == NULL)
-			return (true);
-
-		tarenas = (arena_t **)iralloc(arenas, (ctl_stats.narenas + 1) *
-		    sizeof(arena_t *), 0, 0, false, false);
-		if (tarenas == NULL)
-			return (true);
-	}
-	/* Initialize the new astats and arenas elements. */
-	memset(&astats[ctl_stats.narenas + 1], 0, sizeof(ctl_arena_stats_t));
-	if (ctl_arena_init(&astats[ctl_stats.narenas + 1]))
+	/* Allocate extended arena stats and arenas arrays. */
+	astats = (ctl_arena_stats_t *)imalloc((ctl_stats.narenas + 2) *
+	    sizeof(ctl_arena_stats_t));
+	if (astats == NULL)
 		return (true);
-	tarenas[ctl_stats.narenas] = NULL;
+	tarenas = (arena_t **)imalloc((ctl_stats.narenas + 1) *
+	    sizeof(arena_t *));
+	if (tarenas == NULL) {
+		idalloc(astats);
+		return (true);
+	}
+
+	/* Initialize the new astats element. */
+	memcpy(astats, ctl_stats.arenas, (ctl_stats.narenas + 1) *
+	    sizeof(ctl_arena_stats_t));
+	memset(&astats[ctl_stats.narenas + 1], 0, sizeof(ctl_arena_stats_t));
+	if (ctl_arena_init(&astats[ctl_stats.narenas + 1])) {
+		idalloc(tarenas);
+		idalloc(astats);
+		return (true);
+	}
 	/* Swap merged stats to their new location. */
 	{
 		ctl_arena_stats_t tstats;
@@ -593,13 +580,34 @@ ctl_grow(void)
 		memcpy(&astats[ctl_stats.narenas + 1], &tstats,
 		    sizeof(ctl_arena_stats_t));
 	}
+	/* Initialize the new arenas element. */
+	tarenas[ctl_stats.narenas] = NULL;
+	{
+		arena_t **arenas_old = arenas;
+		/*
+		 * Swap extended arenas array into place.  Although ctl_mtx
+		 * protects this function from other threads extending the
+		 * array, it does not protect from other threads mutating it
+		 * (i.e. initializing arenas and setting array elements to
+		 * point to them).  Therefore, array copying must happen under
+		 * the protection of arenas_lock.
+		 */
+		malloc_mutex_lock(&arenas_lock);
+		arenas = tarenas;
+		memcpy(arenas, arenas_old, ctl_stats.narenas *
+		    sizeof(arena_t *));
+		narenas_total++;
+		arenas_extend(narenas_total - 1);
+		malloc_mutex_unlock(&arenas_lock);
+		/*
+		 * Deallocate arenas_old only if it came from imalloc() (not
+		 * base_alloc()).
+		 */
+		if (ctl_stats.narenas != narenas_auto)
+			idalloc(arenas_old);
+	}
 	ctl_stats.arenas = astats;
 	ctl_stats.narenas++;
-	malloc_mutex_lock(&arenas_lock);
-	arenas = tarenas;
-	narenas_total++;
-	arenas_extend(narenas_total - 1);
-	malloc_mutex_unlock(&arenas_lock);

 	return (false);
 }
@@ -921,7 +929,7 @@ void
 ctl_prefork(void)
 {

-	malloc_mutex_lock(&ctl_mtx);
+	malloc_mutex_prefork(&ctl_mtx);
 }

 void
@@ -960,11 +968,11 @@ ctl_postfork_child(void)
 		if (*oldlenp != sizeof(t)) {				\
 			size_t	copylen = (sizeof(t) <= *oldlenp)	\
 			    ? sizeof(t) : *oldlenp;			\
-			memcpy(oldp, (void *)&v, copylen);		\
+			memcpy(oldp, (void *)&(v), copylen);		\
 			ret = EINVAL;					\
 			goto label_return;				\
 		} else							\
-			*(t *)oldp = v;					\
+			*(t *)oldp = (v);				\
 	}								\
 } while (0)

@@ -974,7 +982,7 @@ ctl_postfork_child(void)
 			ret = EINVAL;					\
 			goto label_return;				\
 		}							\
-		v = *(t *)newp;						\
+		(v) = *(t *)newp;					\
 	}								\
 } while (0)

@@ -995,7 +1003,7 @@ n##_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,	\
 	if (l)								\
 		malloc_mutex_lock(&ctl_mtx);				\
 	READONLY();							\
-	oldval = v;							\
+	oldval = (v);							\
 	READ(oldval, t);						\
 									\
 	ret = 0;							\
@@ -1017,7 +1025,7 @@ n##_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,	\
 		return (ENOENT);					\
 	malloc_mutex_lock(&ctl_mtx);					\
 	READONLY();							\
-	oldval = v;							\
+	oldval = (v);							\
 	READ(oldval, t);						\
 									\
 	ret = 0;							\
@@ -1036,7 +1044,7 @@ n##_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,	\
 									\
 	malloc_mutex_lock(&ctl_mtx);					\
 	READONLY();							\
-	oldval = v;							\
+	oldval = (v);							\
 	READ(oldval, t);						\
 									\
 	ret = 0;							\
@@ -1060,7 +1068,7 @@ n##_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,	\
 	if ((c) == false)						\
 		return (ENOENT);					\
 	READONLY();							\
-	oldval = v;							\
+	oldval = (v);							\
 	READ(oldval, t);						\
 									\
 	ret = 0;							\
@@ -1077,7 +1085,7 @@ n##_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,	\
 	t oldval;							\
 									\
 	READONLY();							\
-	oldval = v;							\
+	oldval = (v);							\
 	READ(oldval, t);						\
 									\
 	ret = 0;							\
@@ -1102,6 +1110,8 @@ label_return:								\
 	return (ret);							\
 }

+/******************************************************************************/
+
 CTL_RO_NL_GEN(version, JEMALLOC_VERSION, const char *)

 static int
@@ -1109,7 +1119,7 @@ epoch_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
    void *newp, size_t newlen)
 {
 	int ret;
-	uint64_t newval;
+	UNUSED uint64_t newval;

 	malloc_mutex_lock(&ctl_mtx);
 	WRITE(newval, uint64_t);
@@ -1123,49 +1133,52 @@ label_return:
 	return (ret);
 }

-static int
-thread_tcache_enabled_ctl(const size_t *mib, size_t miblen, void *oldp,
-    size_t *oldlenp, void *newp, size_t newlen)
-{
-	int ret;
-	bool oldval;
+/******************************************************************************/

-	if (config_tcache == false)
-		return (ENOENT);
+CTL_RO_BOOL_CONFIG_GEN(config_debug)
+CTL_RO_BOOL_CONFIG_GEN(config_dss)
+CTL_RO_BOOL_CONFIG_GEN(config_fill)
+CTL_RO_BOOL_CONFIG_GEN(config_lazy_lock)
+CTL_RO_BOOL_CONFIG_GEN(config_mremap)
+CTL_RO_BOOL_CONFIG_GEN(config_munmap)
+CTL_RO_BOOL_CONFIG_GEN(config_prof)
+CTL_RO_BOOL_CONFIG_GEN(config_prof_libgcc)
+CTL_RO_BOOL_CONFIG_GEN(config_prof_libunwind)
+CTL_RO_BOOL_CONFIG_GEN(config_stats)
+CTL_RO_BOOL_CONFIG_GEN(config_tcache)
+CTL_RO_BOOL_CONFIG_GEN(config_tls)
+CTL_RO_BOOL_CONFIG_GEN(config_utrace)
+CTL_RO_BOOL_CONFIG_GEN(config_valgrind)
+CTL_RO_BOOL_CONFIG_GEN(config_xmalloc)

-	oldval = tcache_enabled_get();
-	if (newp != NULL) {
-		if (newlen != sizeof(bool)) {
-			ret = EINVAL;
-			goto label_return;
-		}
-		tcache_enabled_set(*(bool *)newp);
-	}
-	READ(oldval, bool);
+/******************************************************************************/

-	ret = 0;
-label_return:
-	return (ret);
-}
+CTL_RO_NL_GEN(opt_abort, opt_abort, bool)
+CTL_RO_NL_GEN(opt_dss, opt_dss, const char *)
+CTL_RO_NL_GEN(opt_lg_chunk, opt_lg_chunk, size_t)
+CTL_RO_NL_GEN(opt_narenas, opt_narenas, size_t)
+CTL_RO_NL_GEN(opt_lg_dirty_mult, opt_lg_dirty_mult, ssize_t)
+CTL_RO_NL_GEN(opt_stats_print, opt_stats_print, bool)
+CTL_RO_NL_CGEN(config_fill, opt_junk, opt_junk, bool)
+CTL_RO_NL_CGEN(config_fill, opt_quarantine, opt_quarantine, size_t)
+CTL_RO_NL_CGEN(config_fill, opt_redzone, opt_redzone, bool)
+CTL_RO_NL_CGEN(config_fill, opt_zero, opt_zero, bool)
+CTL_RO_NL_CGEN(config_utrace, opt_utrace, opt_utrace, bool)
+CTL_RO_NL_CGEN(config_valgrind, opt_valgrind, opt_valgrind, bool)
+CTL_RO_NL_CGEN(config_xmalloc, opt_xmalloc, opt_xmalloc, bool)
+CTL_RO_NL_CGEN(config_tcache, opt_tcache, opt_tcache, bool)
+CTL_RO_NL_CGEN(config_tcache, opt_lg_tcache_max, opt_lg_tcache_max, ssize_t)
+CTL_RO_NL_CGEN(config_prof, opt_prof, opt_prof, bool)
+CTL_RO_NL_CGEN(config_prof, opt_prof_prefix, opt_prof_prefix, const char *)
+CTL_RO_CGEN(config_prof, opt_prof_active, opt_prof_active, bool) /* Mutable. */
+CTL_RO_NL_CGEN(config_prof, opt_lg_prof_sample, opt_lg_prof_sample, size_t)
+CTL_RO_NL_CGEN(config_prof, opt_prof_accum, opt_prof_accum, bool)
+CTL_RO_NL_CGEN(config_prof, opt_lg_prof_interval, opt_lg_prof_interval, ssize_t)
+CTL_RO_NL_CGEN(config_prof, opt_prof_gdump, opt_prof_gdump, bool)
+CTL_RO_NL_CGEN(config_prof, opt_prof_final, opt_prof_final, bool)
+CTL_RO_NL_CGEN(config_prof, opt_prof_leak, opt_prof_leak, bool)

-static int
-thread_tcache_flush_ctl(const size_t *mib, size_t miblen, void *oldp,
-    size_t *oldlenp, void *newp, size_t newlen)
-{
-	int ret;
-
-	if (config_tcache == false)
-		return (ENOENT);
-
-	READONLY();
-	WRITEONLY();
-
-	tcache_flush();
-
-	ret = 0;
-label_return:
-	return (ret);
-}
+/******************************************************************************/

 static int
 thread_arena_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
@@ -1227,50 +1240,49 @@ CTL_RO_NL_CGEN(config_stats, thread_deallocated,
 CTL_RO_NL_CGEN(config_stats, thread_deallocatedp,
    &thread_allocated_tsd_get()->deallocated, uint64_t *)

-/******************************************************************************/
+static int
+thread_tcache_enabled_ctl(const size_t *mib, size_t miblen, void *oldp,
+    size_t *oldlenp, void *newp, size_t newlen)
+{
+	int ret;
+	bool oldval;

-CTL_RO_BOOL_CONFIG_GEN(config_debug)
-CTL_RO_BOOL_CONFIG_GEN(config_dss)
-CTL_RO_BOOL_CONFIG_GEN(config_fill)
-CTL_RO_BOOL_CONFIG_GEN(config_lazy_lock)
-CTL_RO_BOOL_CONFIG_GEN(config_mremap)
-CTL_RO_BOOL_CONFIG_GEN(config_munmap)
-CTL_RO_BOOL_CONFIG_GEN(config_prof)
-CTL_RO_BOOL_CONFIG_GEN(config_prof_libgcc)
-CTL_RO_BOOL_CONFIG_GEN(config_prof_libunwind)
-CTL_RO_BOOL_CONFIG_GEN(config_stats)
-CTL_RO_BOOL_CONFIG_GEN(config_tcache)
-CTL_RO_BOOL_CONFIG_GEN(config_tls)
-CTL_RO_BOOL_CONFIG_GEN(config_utrace)
-CTL_RO_BOOL_CONFIG_GEN(config_valgrind)
-CTL_RO_BOOL_CONFIG_GEN(config_xmalloc)
+	if (config_tcache == false)
+		return (ENOENT);

-/******************************************************************************/
+	oldval = tcache_enabled_get();
+	if (newp != NULL) {
+		if (newlen != sizeof(bool)) {
+			ret = EINVAL;
+			goto label_return;
+		}
+		tcache_enabled_set(*(bool *)newp);
+	}
+	READ(oldval, bool);

-CTL_RO_NL_GEN(opt_abort, opt_abort, bool)
-CTL_RO_NL_GEN(opt_dss, opt_dss, const char *)
-CTL_RO_NL_GEN(opt_lg_chunk, opt_lg_chunk, size_t)
-CTL_RO_NL_GEN(opt_narenas, opt_narenas, size_t)
-CTL_RO_NL_GEN(opt_lg_dirty_mult, opt_lg_dirty_mult, ssize_t)
-CTL_RO_NL_GEN(opt_stats_print, opt_stats_print, bool)
-CTL_RO_NL_CGEN(config_fill, opt_junk, opt_junk, bool)
-CTL_RO_NL_CGEN(config_fill, opt_zero, opt_zero, bool)
-CTL_RO_NL_CGEN(config_fill, opt_quarantine, opt_quarantine, size_t)
-CTL_RO_NL_CGEN(config_fill, opt_redzone, opt_redzone, bool)
-CTL_RO_NL_CGEN(config_utrace, opt_utrace, opt_utrace, bool)
-CTL_RO_NL_CGEN(config_valgrind, opt_valgrind, opt_valgrind, bool)
-CTL_RO_NL_CGEN(config_xmalloc, opt_xmalloc, opt_xmalloc, bool)
-CTL_RO_NL_CGEN(config_tcache, opt_tcache, opt_tcache, bool)
-CTL_RO_NL_CGEN(config_tcache, opt_lg_tcache_max, opt_lg_tcache_max, ssize_t)
-CTL_RO_NL_CGEN(config_prof, opt_prof, opt_prof, bool)
-CTL_RO_NL_CGEN(config_prof, opt_prof_prefix, opt_prof_prefix, const char *)
-CTL_RO_CGEN(config_prof, opt_prof_active, opt_prof_active, bool) /* Mutable. */
-CTL_RO_NL_CGEN(config_prof, opt_lg_prof_sample, opt_lg_prof_sample, size_t)
-CTL_RO_NL_CGEN(config_prof, opt_lg_prof_interval, opt_lg_prof_interval, ssize_t)
-CTL_RO_NL_CGEN(config_prof, opt_prof_gdump, opt_prof_gdump, bool)
-CTL_RO_NL_CGEN(config_prof, opt_prof_final, opt_prof_final, bool)
-CTL_RO_NL_CGEN(config_prof, opt_prof_leak, opt_prof_leak, bool)
-CTL_RO_NL_CGEN(config_prof, opt_prof_accum, opt_prof_accum, bool)
+	ret = 0;
+label_return:
+	return (ret);
+}
+
+static int
+thread_tcache_flush_ctl(const size_t *mib, size_t miblen, void *oldp,
+    size_t *oldlenp, void *newp, size_t newlen)
+{
+	int ret;
+
+	if (config_tcache == false)
+		return (ENOENT);
+
+	READONLY();
+	WRITEONLY();
+
+	tcache_flush();
+
+	ret = 0;
+label_return:
+	return (ret);
+}

 /******************************************************************************/

@@ -1382,31 +1394,8 @@ label_return:
 	return (ret);
 }

-
 /******************************************************************************/

-CTL_RO_NL_GEN(arenas_bin_i_size, arena_bin_info[mib[2]].reg_size, size_t)
-CTL_RO_NL_GEN(arenas_bin_i_nregs, arena_bin_info[mib[2]].nregs, uint32_t)
-CTL_RO_NL_GEN(arenas_bin_i_run_size, arena_bin_info[mib[2]].run_size, size_t)
-static const ctl_named_node_t *
-arenas_bin_i_index(const size_t *mib, size_t miblen, size_t i)
-{
-
-	if (i > NBINS)
-		return (NULL);
-	return (super_arenas_bin_i_node);
-}
-
-CTL_RO_NL_GEN(arenas_lrun_i_size, ((mib[2]+1) << LG_PAGE), size_t)
-static const ctl_named_node_t *
-arenas_lrun_i_index(const size_t *mib, size_t miblen, size_t i)
-{
-
-	if (i > nlclasses)
-		return (NULL);
-	return (super_arenas_lrun_i_node);
-}
-
 static int
 arenas_narenas_ctl(const size_t *mib, size_t miblen, void *oldp,
    size_t *oldlenp, void *newp, size_t newlen)
@@ -1460,7 +1449,28 @@ CTL_RO_NL_GEN(arenas_page, PAGE, size_t)
 CTL_RO_NL_CGEN(config_tcache, arenas_tcache_max, tcache_maxclass, size_t)
 CTL_RO_NL_GEN(arenas_nbins, NBINS, unsigned)
 CTL_RO_NL_CGEN(config_tcache, arenas_nhbins, nhbins, unsigned)
+CTL_RO_NL_GEN(arenas_bin_i_size, arena_bin_info[mib[2]].reg_size, size_t)
+CTL_RO_NL_GEN(arenas_bin_i_nregs, arena_bin_info[mib[2]].nregs, uint32_t)
+CTL_RO_NL_GEN(arenas_bin_i_run_size, arena_bin_info[mib[2]].run_size, size_t)
+static const ctl_named_node_t *
+arenas_bin_i_index(const size_t *mib, size_t miblen, size_t i)
+{
+
+	if (i > NBINS)
+		return (NULL);
+	return (super_arenas_bin_i_node);
+}
+
 CTL_RO_NL_GEN(arenas_nlruns, nlclasses, size_t)
+CTL_RO_NL_GEN(arenas_lrun_i_size, ((mib[2]+1) << LG_PAGE), size_t)
+static const ctl_named_node_t *
+arenas_lrun_i_index(const size_t *mib, size_t miblen, size_t i)
+{
+
+	if (i > nlclasses)
+		return (NULL);
+	return (super_arenas_lrun_i_node);
+}

 static int
 arenas_purge_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
@@ -1492,6 +1502,7 @@ arenas_extend_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
    void *newp, size_t newlen)
 {
 	int ret;
+	unsigned narenas;

 	malloc_mutex_lock(&ctl_mtx);
 	READONLY();
@@ -1499,7 +1510,8 @@ arenas_extend_ctl(const size_t *mib, size_t miblen, void *oldp, size_t *oldlenp,
 		ret = EAGAIN;
 		goto label_return;
 	}
-	READ(ctl_stats.narenas - 1, unsigned);
+	narenas = ctl_stats.narenas - 1;
+	READ(narenas, unsigned);

 	ret = 0;
 label_return:
@@ -1565,6 +1577,11 @@ CTL_RO_NL_CGEN(config_prof, prof_interval, prof_interval, uint64_t)

 /******************************************************************************/

+CTL_RO_CGEN(config_stats, stats_cactive, &stats_cactive, size_t *)
+CTL_RO_CGEN(config_stats, stats_allocated, ctl_stats.allocated, size_t)
+CTL_RO_CGEN(config_stats, stats_active, ctl_stats.active, size_t)
+CTL_RO_CGEN(config_stats, stats_mapped, ctl_stats.mapped, size_t)
+
 CTL_RO_CGEN(config_stats, stats_chunks_current, ctl_stats.chunks.current,
    size_t)
 CTL_RO_CGEN(config_stats, stats_chunks_total, ctl_stats.chunks.total, uint64_t)
@@ -1572,6 +1589,20 @@ CTL_RO_CGEN(config_stats, stats_chunks_high, ctl_stats.chunks.high, size_t)
 CTL_RO_CGEN(config_stats, stats_huge_allocated, huge_allocated, size_t)
 CTL_RO_CGEN(config_stats, stats_huge_nmalloc, huge_nmalloc, uint64_t)
 CTL_RO_CGEN(config_stats, stats_huge_ndalloc, huge_ndalloc, uint64_t)
+
+CTL_RO_GEN(stats_arenas_i_dss, ctl_stats.arenas[mib[2]].dss, const char *)
+CTL_RO_GEN(stats_arenas_i_nthreads, ctl_stats.arenas[mib[2]].nthreads, unsigned)
+CTL_RO_GEN(stats_arenas_i_pactive, ctl_stats.arenas[mib[2]].pactive, size_t)
+CTL_RO_GEN(stats_arenas_i_pdirty, ctl_stats.arenas[mib[2]].pdirty, size_t)
+CTL_RO_CGEN(config_stats, stats_arenas_i_mapped,
+    ctl_stats.arenas[mib[2]].astats.mapped, size_t)
+CTL_RO_CGEN(config_stats, stats_arenas_i_npurge,
+    ctl_stats.arenas[mib[2]].astats.npurge, uint64_t)
+CTL_RO_CGEN(config_stats, stats_arenas_i_nmadvise,
+    ctl_stats.arenas[mib[2]].astats.nmadvise, uint64_t)
+CTL_RO_CGEN(config_stats, stats_arenas_i_purged,
+    ctl_stats.arenas[mib[2]].astats.purged, uint64_t)
+
 CTL_RO_CGEN(config_stats, stats_arenas_i_small_allocated,
    ctl_stats.arenas[mib[2]].allocated_small, size_t)
 CTL_RO_CGEN(config_stats, stats_arenas_i_small_nmalloc,
@@ -1635,19 +1666,6 @@ stats_arenas_i_lruns_j_index(const size_t *mib, size_t miblen, size_t j)
 	return (super_stats_arenas_i_lruns_j_node);
 }

-CTL_RO_GEN(stats_arenas_i_nthreads, ctl_stats.arenas[mib[2]].nthreads, unsigned)
-CTL_RO_GEN(stats_arenas_i_dss, ctl_stats.arenas[mib[2]].dss, const char *)
-CTL_RO_GEN(stats_arenas_i_pactive, ctl_stats.arenas[mib[2]].pactive, size_t)
-CTL_RO_GEN(stats_arenas_i_pdirty, ctl_stats.arenas[mib[2]].pdirty, size_t)
-CTL_RO_CGEN(config_stats, stats_arenas_i_mapped,
-    ctl_stats.arenas[mib[2]].astats.mapped, size_t)
-CTL_RO_CGEN(config_stats, stats_arenas_i_npurge,
-    ctl_stats.arenas[mib[2]].astats.npurge, uint64_t)
-CTL_RO_CGEN(config_stats, stats_arenas_i_nmadvise,
-    ctl_stats.arenas[mib[2]].astats.nmadvise, uint64_t)
-CTL_RO_CGEN(config_stats, stats_arenas_i_purged,
-    ctl_stats.arenas[mib[2]].astats.purged, uint64_t)
-
 static const ctl_named_node_t *
 stats_arenas_i_index(const size_t *mib, size_t miblen, size_t i)
 {
@@ -1664,8 +1682,3 @@ label_return:
 	malloc_mutex_unlock(&ctl_mtx);
 	return (ret);
 }
-
-CTL_RO_CGEN(config_stats, stats_cactive, &stats_cactive, size_t *)
-CTL_RO_CGEN(config_stats, stats_allocated, ctl_stats.allocated, size_t)
-CTL_RO_CGEN(config_stats, stats_active, ctl_stats.active, size_t)
-CTL_RO_CGEN(config_stats, stats_mapped, ctl_stats.mapped, size_t)
@@ -16,14 +16,14 @@ malloc_mutex_t	huge_mtx;
 static extent_tree_t	huge;

 void *
-huge_malloc(size_t size, bool zero)
+huge_malloc(size_t size, bool zero, dss_prec_t dss_prec)
 {

-	return (huge_palloc(size, chunksize, zero));
+	return (huge_palloc(size, chunksize, zero, dss_prec));
 }

 void *
-huge_palloc(size_t size, size_t alignment, bool zero)
+huge_palloc(size_t size, size_t alignment, bool zero, dss_prec_t dss_prec)
 {
 	void *ret;
 	size_t csize;
@@ -48,8 +48,7 @@ huge_palloc(size_t size, size_t alignment, bool zero)
 	 * it is possible to make correct junk/zero fill decisions below.
 	 */
 	is_zeroed = zero;
-	ret = chunk_alloc(csize, alignment, false, &is_zeroed,
-	    chunk_dss_prec_get());
+	ret = chunk_alloc(csize, alignment, false, &is_zeroed, dss_prec);
 	if (ret == NULL) {
 		base_node_dealloc(node);
 		return (NULL);
@@ -78,7 +77,7 @@ huge_palloc(size_t size, size_t alignment, bool zero)
 	return (ret);
 }

-void *
+bool
 huge_ralloc_no_move(void *ptr, size_t oldsize, size_t size, size_t extra)
 {

@@ -89,28 +88,23 @@ huge_ralloc_no_move(void *ptr, size_t oldsize, size_t size, size_t extra)
 	    && CHUNK_CEILING(oldsize) >= CHUNK_CEILING(size)
 	    && CHUNK_CEILING(oldsize) <= CHUNK_CEILING(size+extra)) {
 		assert(CHUNK_CEILING(oldsize) == oldsize);
-		if (config_fill && opt_junk && size < oldsize) {
-			memset((void *)((uintptr_t)ptr + size), 0x5a,
-			    oldsize - size);
-		}
-		return (ptr);
+		return (false);
 	}

 	/* Reallocation would require a move. */
-	return (NULL);
+	return (true);
 }

 void *
 huge_ralloc(void *ptr, size_t oldsize, size_t size, size_t extra,
-    size_t alignment, bool zero, bool try_tcache_dalloc)
+    size_t alignment, bool zero, bool try_tcache_dalloc, dss_prec_t dss_prec)
 {
 	void *ret;
 	size_t copysize;

 	/* Try to avoid moving the allocation. */
-	ret = huge_ralloc_no_move(ptr, oldsize, size, extra);
-	if (ret != NULL)
-		return (ret);
+	if (huge_ralloc_no_move(ptr, oldsize, size, extra) == false)
+		return (ptr);

 	/*
 	 * size and oldsize are different enough that we need to use a
@@ -118,18 +112,18 @@ huge_ralloc(void *ptr, size_t oldsize, size_t size, size_t extra,
 	 * space and copying.
 	 */
 	if (alignment > chunksize)
-		ret = huge_palloc(size + extra, alignment, zero);
+		ret = huge_palloc(size + extra, alignment, zero, dss_prec);
 	else
-		ret = huge_malloc(size + extra, zero);
+		ret = huge_malloc(size + extra, zero, dss_prec);

 	if (ret == NULL) {
 		if (extra == 0)
 			return (NULL);
 		/* Try again, this time without extra. */
 		if (alignment > chunksize)
-			ret = huge_palloc(size, alignment, zero);
+			ret = huge_palloc(size, alignment, zero, dss_prec);
 		else
-			ret = huge_malloc(size, zero);
+			ret = huge_malloc(size, zero, dss_prec);

 		if (ret == NULL)
 			return (NULL);
@@ -169,23 +163,56 @@ huge_ralloc(void *ptr, size_t oldsize, size_t size, size_t extra,
 			 */
 			char buf[BUFERROR_BUF];

-			buferror(buf, sizeof(buf));
+			buferror(get_errno(), buf, sizeof(buf));
 			malloc_printf("<jemalloc>: Error in mremap(): %s\n",
 			    buf);
 			if (opt_abort)
 				abort();
 			memcpy(ret, ptr, copysize);
 			chunk_dealloc_mmap(ptr, oldsize);
+		} else if (config_fill && zero == false && opt_junk && oldsize
+		    < newsize) {
+			/*
+			 * mremap(2) clobbers the original mapping, so
+			 * junk/zero filling is not preserved.  There is no
+			 * need to zero fill here, since any trailing
+			 * uninititialized memory is demand-zeroed by the
+			 * kernel, but junk filling must be redone.
+			 */
+			memset(ret + oldsize, 0xa5, newsize - oldsize);
 		}
 	} else
 #endif
 	{
 		memcpy(ret, ptr, copysize);
-		iqallocx(ptr, try_tcache_dalloc);
+		iqalloct(ptr, try_tcache_dalloc);
 	}
 	return (ret);
 }

+#ifdef JEMALLOC_JET
+#undef huge_dalloc_junk
+#define	huge_dalloc_junk JEMALLOC_N(huge_dalloc_junk_impl)
+#endif
+static void
+huge_dalloc_junk(void *ptr, size_t usize)
+{
+
+	if (config_fill && config_dss && opt_junk) {
+		/*
+		 * Only bother junk filling if the chunk isn't about to be
+		 * unmapped.
+		 */
+		if (config_munmap == false || (config_dss && chunk_in_dss(ptr)))
+			memset(ptr, 0x5a, usize);
+	}
+}
+#ifdef JEMALLOC_JET
+#undef huge_dalloc_junk
+#define	huge_dalloc_junk JEMALLOC_N(huge_dalloc_junk)
+huge_dalloc_junk_t *huge_dalloc_junk = JEMALLOC_N(huge_dalloc_junk_impl);
+#endif
+
 void
 huge_dalloc(void *ptr, bool unmap)
 {
@@ -208,8 +235,8 @@ huge_dalloc(void *ptr, bool unmap)

 	malloc_mutex_unlock(&huge_mtx);

-	if (unmap && config_fill && config_dss && opt_junk)
-		memset(node->addr, 0x5a, node->size);
+	if (unmap)
+		huge_dalloc_junk(node->addr, node->size);

 	chunk_dealloc(node->addr, node->size, unmap);

@@ -236,6 +263,13 @@ huge_salloc(const void *ptr)
 	return (ret);
 }

+dss_prec_t
+huge_dss_prec_get(arena_t *arena)
+{
+
+	return (arena_dss_prec_get(choose_arena(arena)));
+}
+
 prof_ctx_t *
 huge_prof_ctx_get(const void *ptr)
 {
@@ -6,7 +6,7 @@
 #endif

 #ifndef _CRT_SPINCOUNT
-#define _CRT_SPINCOUNT 4000
+#define	_CRT_SPINCOUNT 4000
 #endif

 /******************************************************************************/
@@ -1,3 +1,4 @@
+#define	JEMALLOC_QUARANTINE_C_
 #include "jemalloc/internal/jemalloc_internal.h"

 /*
@@ -11,39 +12,18 @@
 /******************************************************************************/
 /* Data. */

-typedef struct quarantine_obj_s quarantine_obj_t;
-typedef struct quarantine_s quarantine_t;
-
-struct quarantine_obj_s {
-	void	*ptr;
-	size_t	usize;
-};
-
-struct quarantine_s {
-	size_t			curbytes;
-	size_t			curobjs;
-	size_t			first;
-#define	LG_MAXOBJS_INIT 10
-	size_t			lg_maxobjs;
-	quarantine_obj_t	objs[1]; /* Dynamically sized ring buffer. */
-};
-
-static void	quarantine_cleanup(void *arg);
-
-malloc_tsd_data(static, quarantine, quarantine_t *, NULL)
-malloc_tsd_funcs(JEMALLOC_INLINE, quarantine, quarantine_t *, NULL,
-    quarantine_cleanup)
+malloc_tsd_data(, quarantine, quarantine_t *, NULL)

 /******************************************************************************/
 /* Function prototypes for non-inline static functions. */

-static quarantine_t	*quarantine_init(size_t lg_maxobjs);
 static quarantine_t	*quarantine_grow(quarantine_t *quarantine);
+static void	quarantine_drain_one(quarantine_t *quarantine);
 static void	quarantine_drain(quarantine_t *quarantine, size_t upper_bound);

 /******************************************************************************/

-static quarantine_t *
+quarantine_t *
 quarantine_init(size_t lg_maxobjs)
 {
 	quarantine_t *quarantine;
@@ -68,8 +48,10 @@ quarantine_grow(quarantine_t *quarantine)
 	quarantine_t *ret;

 	ret = quarantine_init(quarantine->lg_maxobjs + 1);
-	if (ret == NULL)
+	if (ret == NULL) {
+		quarantine_drain_one(quarantine);
 		return (quarantine);
+	}

 	ret->curbytes = quarantine->curbytes;
 	ret->curobjs = quarantine->curobjs;
@@ -89,23 +71,29 @@ quarantine_grow(quarantine_t *quarantine)
 		memcpy(&ret->objs[ncopy_a], quarantine->objs, ncopy_b *
 		    sizeof(quarantine_obj_t));
 	}
+	idalloc(quarantine);

 	return (ret);
 }

+static void
+quarantine_drain_one(quarantine_t *quarantine)
+{
+	quarantine_obj_t *obj = &quarantine->objs[quarantine->first];
+	assert(obj->usize == isalloc(obj->ptr, config_prof));
+	idalloc(obj->ptr);
+	quarantine->curbytes -= obj->usize;
+	quarantine->curobjs--;
+	quarantine->first = (quarantine->first + 1) & ((ZU(1) <<
+	    quarantine->lg_maxobjs) - 1);
+}
+
 static void
 quarantine_drain(quarantine_t *quarantine, size_t upper_bound)
 {

-	while (quarantine->curbytes > upper_bound && quarantine->curobjs > 0) {
-		quarantine_obj_t *obj = &quarantine->objs[quarantine->first];
-		assert(obj->usize == isalloc(obj->ptr, config_prof));
-		idalloc(obj->ptr);
-		quarantine->curbytes -= obj->usize;
-		quarantine->curobjs--;
-		quarantine->first = (quarantine->first + 1) & ((ZU(1) <<
-		    quarantine->lg_maxobjs) - 1);
-	}
+	while (quarantine->curbytes > upper_bound && quarantine->curobjs > 0)
+		quarantine_drain_one(quarantine);
 }

 void
@@ -119,24 +107,16 @@ quarantine(void *ptr)

 	quarantine = *quarantine_tsd_get();
 	if ((uintptr_t)quarantine <= (uintptr_t)QUARANTINE_STATE_MAX) {
-		if (quarantine == NULL) {
-			if ((quarantine = quarantine_init(LG_MAXOBJS_INIT)) ==
-			    NULL) {
-				idalloc(ptr);
-				return;
-			}
-		} else {
-			if (quarantine == QUARANTINE_STATE_PURGATORY) {
-				/*
-				 * Make a note that quarantine() was called
-				 * after quarantine_cleanup() was called.
-				 */
-				quarantine = QUARANTINE_STATE_REINCARNATED;
-				quarantine_tsd_set(&quarantine);
-			}
-			idalloc(ptr);
-			return;
+		if (quarantine == QUARANTINE_STATE_PURGATORY) {
+			/*
+			 * Make a note that quarantine() was called after
+			 * quarantine_cleanup() was called.
+			 */
+			quarantine = QUARANTINE_STATE_REINCARNATED;
+			quarantine_tsd_set(&quarantine);
 		}
+		idalloc(ptr);
+		return;
 	}
 	/*
 	 * Drain one or more objects if the quarantine size limit would be
@@ -161,15 +141,24 @@ quarantine(void *ptr)
 		obj->usize = usize;
 		quarantine->curbytes += usize;
 		quarantine->curobjs++;
-		if (opt_junk)
-			memset(ptr, 0x5a, usize);
+		if (config_fill && opt_junk) {
+			/*
+			 * Only do redzone validation if Valgrind isn't in
+			 * operation.
+			 */
+			if ((config_valgrind == false || opt_valgrind == false)
+			    && usize <= SMALL_MAXCLASS)
+				arena_quarantine_junk_small(ptr, usize);
+			else
+				memset(ptr, 0x5a, usize);
+		}
 	} else {
 		assert(quarantine->curbytes == 0);
 		idalloc(ptr);
 	}
 }

-static void
+void
 quarantine_cleanup(void *arg)
 {
 	quarantine_t *quarantine = *(quarantine_t **)arg;
@@ -2,42 +2,55 @@
 #include "jemalloc/internal/jemalloc_internal.h"

 rtree_t *
-rtree_new(unsigned bits)
+rtree_new(unsigned bits, rtree_alloc_t *alloc, rtree_dalloc_t *dalloc)
 {
 	rtree_t *ret;
-	unsigned bits_per_level, height, i;
+	unsigned bits_per_level, bits_in_leaf, height, i;
+
+	assert(bits > 0 && bits <= (sizeof(uintptr_t) << 3));

 	bits_per_level = ffs(pow2_ceil((RTREE_NODESIZE / sizeof(void *)))) - 1;
-	height = bits / bits_per_level;
-	if (height * bits_per_level != bits)
-		height++;
-	assert(height * bits_per_level >= bits);
+	bits_in_leaf = ffs(pow2_ceil((RTREE_NODESIZE / sizeof(uint8_t)))) - 1;
+	if (bits > bits_in_leaf) {
+		height = 1 + (bits - bits_in_leaf) / bits_per_level;
+		if ((height-1) * bits_per_level + bits_in_leaf != bits)
+			height++;
+	} else {
+		height = 1;
+	}
+	assert((height-1) * bits_per_level + bits_in_leaf >= bits);

-	ret = (rtree_t*)base_alloc(offsetof(rtree_t, level2bits) +
+	ret = (rtree_t*)alloc(offsetof(rtree_t, level2bits) +
 	    (sizeof(unsigned) * height));
 	if (ret == NULL)
 		return (NULL);
 	memset(ret, 0, offsetof(rtree_t, level2bits) + (sizeof(unsigned) *
 	    height));

+	ret->alloc = alloc;
+	ret->dalloc = dalloc;
 	if (malloc_mutex_init(&ret->mutex)) {
-		/* Leak the rtree. */
+		if (dalloc != NULL)
+			dalloc(ret);
 		return (NULL);
 	}
 	ret->height = height;
-	if (bits_per_level * height > bits)
-		ret->level2bits[0] = bits % bits_per_level;
-	else
-		ret->level2bits[0] = bits_per_level;
-	for (i = 1; i < height; i++)
-		ret->level2bits[i] = bits_per_level;
+	if (height > 1) {
+		if ((height-1) * bits_per_level + bits_in_leaf > bits) {
+			ret->level2bits[0] = (bits - bits_in_leaf) %
+			    bits_per_level;
+		} else
+			ret->level2bits[0] = bits_per_level;
+		for (i = 1; i < height-1; i++)
+			ret->level2bits[i] = bits_per_level;
+		ret->level2bits[height-1] = bits_in_leaf;
+	} else
+		ret->level2bits[0] = bits;

-	ret->root = (void**)base_alloc(sizeof(void *) << ret->level2bits[0]);
+	ret->root = (void**)alloc(sizeof(void *) << ret->level2bits[0]);
 	if (ret->root == NULL) {
-		/*
-		 * We leak the rtree here, since there's no generic base
-		 * deallocation.
-		 */
+		if (dalloc != NULL)
+			dalloc(ret);
 		return (NULL);
 	}
 	memset(ret->root, 0, sizeof(void *) << ret->level2bits[0]);
@@ -45,6 +58,31 @@ rtree_new(unsigned bits)
 	return (ret);
 }

+static void
+rtree_delete_subtree(rtree_t *rtree, void **node, unsigned level)
+{
+
+	if (level < rtree->height - 1) {
+		size_t nchildren, i;
+
+		nchildren = ZU(1) << rtree->level2bits[level];
+		for (i = 0; i < nchildren; i++) {
+			void **child = (void **)node[i];
+			if (child != NULL)
+				rtree_delete_subtree(rtree, child, level + 1);
+		}
+	}
+	rtree->dalloc(node);
+}
+
+void
+rtree_delete(rtree_t *rtree)
+{
+
+	rtree_delete_subtree(rtree, rtree->root, 0);
+	rtree->dalloc(rtree);
+}
+
 void
 rtree_prefork(rtree_t *rtree)
 {
@@ -345,25 +345,25 @@ stats_print(void (*write_cb)(void *, const char *), void *cbopaque,
 		malloc_cprintf(write_cb, cbopaque, "Assertions %s\n",
 		    bv ? "enabled" : "disabled");

-#define OPT_WRITE_BOOL(n)						\
+#define	OPT_WRITE_BOOL(n)						\
 		if ((err = je_mallctl("opt."#n, &bv, &bsz, NULL, 0))	\
 		    == 0) {						\
 			malloc_cprintf(write_cb, cbopaque,		\
 			    "  opt."#n": %s\n", bv ? "true" : "false");	\
 		}
-#define OPT_WRITE_SIZE_T(n)						\
+#define	OPT_WRITE_SIZE_T(n)						\
 		if ((err = je_mallctl("opt."#n, &sv, &ssz, NULL, 0))	\
 		    == 0) {						\
 			malloc_cprintf(write_cb, cbopaque,		\
 			"  opt."#n": %zu\n", sv);			\
 		}
-#define OPT_WRITE_SSIZE_T(n)						\
+#define	OPT_WRITE_SSIZE_T(n)						\
 		if ((err = je_mallctl("opt."#n, &ssv, &sssz, NULL, 0))	\
 		    == 0) {						\
 			malloc_cprintf(write_cb, cbopaque,		\
 			    "  opt."#n": %zd\n", ssv);			\
 		}
-#define OPT_WRITE_CHAR_P(n)						\
+#define	OPT_WRITE_CHAR_P(n)						\
 		if ((err = je_mallctl("opt."#n, &cpv, &cpsz, NULL, 0))	\
 		    == 0) {						\
 			malloc_cprintf(write_cb, cbopaque,		\
@@ -97,9 +97,8 @@ tcache_bin_flush_small(tcache_bin_t *tbin, size_t binind, unsigned rem,
 		arena_bin_t *bin = &arena->bins[binind];

 		if (config_prof && arena == tcache->arena) {
-			malloc_mutex_lock(&arena->lock);
-			arena_prof_accum(arena, tcache->prof_accumbytes);
-			malloc_mutex_unlock(&arena->lock);
+			if (arena_prof_accum(arena, tcache->prof_accumbytes))
+				prof_idump();
 			tcache->prof_accumbytes = 0;
 		}

@@ -176,11 +175,14 @@ tcache_bin_flush_large(tcache_bin_t *tbin, size_t binind, unsigned rem,
 		arena_chunk_t *chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(
 		    tbin->avail[0]);
 		arena_t *arena = chunk->arena;
+		UNUSED bool idump;

+		if (config_prof)
+			idump = false;
 		malloc_mutex_lock(&arena->lock);
 		if ((config_prof || config_stats) && arena == tcache->arena) {
 			if (config_prof) {
-				arena_prof_accum(arena,
+				idump = arena_prof_accum_locked(arena,
 				    tcache->prof_accumbytes);
 				tcache->prof_accumbytes = 0;
 			}
@@ -212,6 +214,8 @@ tcache_bin_flush_large(tcache_bin_t *tbin, size_t binind, unsigned rem,
 			}
 		}
 		malloc_mutex_unlock(&arena->lock);
+		if (config_prof && idump)
+			prof_idump();
 	}
 	if (config_stats && merged_stats == false) {
 		/*
@@ -256,8 +260,8 @@ tcache_arena_dissociate(tcache_t *tcache)
 		/* Unlink from list of extant tcaches. */
 		malloc_mutex_lock(&tcache->arena->lock);
 		ql_remove(&tcache->arena->tcache_ql, tcache, link);
-		malloc_mutex_unlock(&tcache->arena->lock);
 		tcache_stats_merge(tcache, tcache->arena);
+		malloc_mutex_unlock(&tcache->arena->lock);
 	}
 }

@@ -288,7 +292,7 @@ tcache_create(arena_t *arena)
 	else if (size <= tcache_maxclass)
 		tcache = (tcache_t *)arena_malloc_large(arena, size, true);
 	else
-		tcache = (tcache_t *)icallocx(size, false, arena);
+		tcache = (tcache_t *)icalloct(size, false, arena);

 	if (tcache == NULL)
 		return (NULL);
@@ -343,11 +347,9 @@ tcache_destroy(tcache_t *tcache)
 		}
 	}

-	if (config_prof && tcache->prof_accumbytes > 0) {
-		malloc_mutex_lock(&tcache->arena->lock);
-		arena_prof_accum(tcache->arena, tcache->prof_accumbytes);
-		malloc_mutex_unlock(&tcache->arena->lock);
-	}
+	if (config_prof && tcache->prof_accumbytes > 0 &&
+	    arena_prof_accum(tcache->arena, tcache->prof_accumbytes))
+		prof_idump();

 	tcache_size = arena_salloc(tcache, false);
 	if (tcache_size <= SMALL_MAXCLASS) {
@@ -364,7 +366,7 @@ tcache_destroy(tcache_t *tcache)

 		arena_dalloc_large(arena, chunk, tcache);
 	} else
-		idallocx(tcache, false);
+		idalloct(tcache, false);
 }

 void
@@ -397,11 +399,14 @@ tcache_thread_cleanup(void *arg)
 	}
 }

+/* Caller must own arena->lock. */
 void
 tcache_stats_merge(tcache_t *tcache, arena_t *arena)
 {
 	unsigned i;

+	cassert(config_stats);
+
 	/* Merge and reset tcache stats. */
 	for (i = 0; i < NBINS; i++) {
 		arena_bin_t *bin = &arena->bins[i];
@@ -21,7 +21,7 @@ void
 malloc_tsd_dalloc(void *wrapper)
 {

-	idalloc(wrapper);
+	idalloct(wrapper, false);
 }

 void
@@ -105,3 +105,37 @@ JEMALLOC_SECTION(".CRT$XLY") JEMALLOC_ATTR(used)
 static const BOOL	(WINAPI *tls_callback)(HINSTANCE hinstDLL,
    DWORD fdwReason, LPVOID lpvReserved) = _tls_callback;
 #endif
+
+#if (!defined(JEMALLOC_MALLOC_THREAD_CLEANUP) && !defined(JEMALLOC_TLS) && \
+    !defined(_WIN32))
+void *
+tsd_init_check_recursion(tsd_init_head_t *head, tsd_init_block_t *block)
+{
+	pthread_t self = pthread_self();
+	tsd_init_block_t *iter;
+
+	/* Check whether this thread has already inserted into the list. */
+	malloc_mutex_lock(&head->lock);
+	ql_foreach(iter, &head->blocks, link) {
+		if (iter->thread == self) {
+			malloc_mutex_unlock(&head->lock);
+			return (iter->data);
+		}
+	}
+	/* Insert block into list. */
+	ql_elm_new(block, link);
+	block->thread = self;
+	ql_tail_insert(&head->blocks, block, link);
+	malloc_mutex_unlock(&head->lock);
+	return (NULL);
+}
+
+void
+tsd_init_finish(tsd_init_head_t *head, tsd_init_block_t *block)
+{
+
+	malloc_mutex_lock(&head->lock);
+	ql_remove(&head->blocks, block, link);
+	malloc_mutex_unlock(&head->lock);
+}
+#endif
@@ -77,7 +77,7 @@ malloc_write(const char *s)
 * provide a wrapper.
 */
 int
-buferror(char *buf, size_t buflen)
+buferror(int err, char *buf, size_t buflen)
 {

 #ifdef _WIN32
@@ -85,34 +85,36 @@ buferror(char *buf, size_t buflen)
 	    (LPSTR)buf, buflen, NULL);
 	return (0);
 #elif defined(_GNU_SOURCE)
-	char *b = strerror_r(errno, buf, buflen);
+	char *b = strerror_r(err, buf, buflen);
 	if (b != buf) {
 		strncpy(buf, b, buflen);
 		buf[buflen-1] = '\0';
 	}
 	return (0);
 #else
-	return (strerror_r(errno, buf, buflen));
+	return (strerror_r(err, buf, buflen));
 #endif
 }

 uintmax_t
-malloc_strtoumax(const char *nptr, char **endptr, int base)
+malloc_strtoumax(const char *restrict nptr, char **restrict endptr, int base)
 {
 	uintmax_t ret, digit;
 	int b;
 	bool neg;
 	const char *p, *ns;

+	p = nptr;
 	if (base < 0 || base == 1 || base > 36) {
+		ns = p;
 		set_errno(EINVAL);
-		return (UINTMAX_MAX);
+		ret = UINTMAX_MAX;
+		goto label_return;
 	}
 	b = base;

 	/* Swallow leading whitespace and get sign, if any. */
 	neg = false;
-	p = nptr;
 	while (true) {
 		switch (*p) {
 		case '\t': case '\n': case '\v': case '\f': case '\r': case ' ':
@@ -146,7 +148,7 @@ malloc_strtoumax(const char *nptr, char **endptr, int base)
 			if (b == 8)
 				p++;
 			break;
-		case 'x':
+		case 'X': case 'x':
 			switch (p[2]) {
 			case '0': case '1': case '2': case '3': case '4':
 			case '5': case '6': case '7': case '8': case '9':
@@ -164,7 +166,9 @@ malloc_strtoumax(const char *nptr, char **endptr, int base)
 			}
 			break;
 		default:
-			break;
+			p++;
+			ret = 0;
+			goto label_return;
 		}
 	}
 	if (b == 0)
@@ -181,13 +185,22 @@ malloc_strtoumax(const char *nptr, char **endptr, int base)
 		if (ret < pret) {
 			/* Overflow. */
 			set_errno(ERANGE);
-			return (UINTMAX_MAX);
+			ret = UINTMAX_MAX;
+			goto label_return;
 		}
 		p++;
 	}
 	if (neg)
 		ret = -ret;

+	if (p == ns) {
+		/* No conversion performed. */
+		set_errno(EINVAL);
+		ret = UINTMAX_MAX;
+		goto label_return;
+	}
+
+label_return:
 	if (endptr != NULL) {
 		if (p == ns) {
 			/* No characters were converted. */
@@ -195,7 +208,6 @@ malloc_strtoumax(const char *nptr, char **endptr, int base)
 		} else
 			*endptr = (char *)p;
 	}
-
 	return (ret);
 }

@@ -331,7 +343,7 @@ malloc_vsnprintf(char *str, size_t size, const char *format, va_list ap)
 			APPEND_C(' ');					\
 	}								\
 } while (0)
-#define GET_ARG_NUMERIC(val, len) do {					\
+#define	GET_ARG_NUMERIC(val, len) do {					\
 	switch (len) {							\
 	case '?':							\
 		val = va_arg(ap, int);					\
@@ -354,6 +366,9 @@ malloc_vsnprintf(char *str, size_t size, const char *format, va_list ap)
 	case 'j':							\
 		val = va_arg(ap, intmax_t);				\
 		break;							\
+	case 'j' | 0x80:						\
+		val = va_arg(ap, uintmax_t);				\
+		break;							\
 	case 't':							\
 		val = va_arg(ap, ptrdiff_t);				\
 		break;							\
@@ -385,11 +400,6 @@ malloc_vsnprintf(char *str, size_t size, const char *format, va_list ap)
 			unsigned char len = '?';

 			f++;
-			if (*f == '%') {
-				/* %% */
-				APPEND_C(*f);
-				break;
-			}
 			/* Flags. */
 			while (true) {
 				switch (*f) {
@@ -419,6 +429,10 @@ malloc_vsnprintf(char *str, size_t size, const char *format, va_list ap)
 			case '*':
 				width = va_arg(ap, int);
 				f++;
+				if (width < 0) {
+					left_justify = true;
+					width = -width;
+				}
 				break;
 			case '0': case '1': case '2': case '3': case '4':
 			case '5': case '6': case '7': case '8': case '9': {
@@ -428,19 +442,16 @@ malloc_vsnprintf(char *str, size_t size, const char *format, va_list ap)
 				assert(uwidth != UINTMAX_MAX || get_errno() !=
 				    ERANGE);
 				width = (int)uwidth;
-				if (*f == '.') {
-					f++;
-					goto label_precision;
-				} else
-					goto label_length;
 				break;
-			} case '.':
-				f++;
-				goto label_precision;
-			default: goto label_length;
+			} default:
+				break;
 			}
+			/* Width/precision separator. */
+			if (*f == '.')
+				f++;
+			else
+				goto label_length;
 			/* Precision. */
-			label_precision:
 			switch (*f) {
 			case '*':
 				prec = va_arg(ap, int);
@@ -469,16 +480,8 @@ malloc_vsnprintf(char *str, size_t size, const char *format, va_list ap)
 				} else
 					len = 'l';
 				break;
-			case 'j':
-				len = 'j';
-				f++;
-				break;
-			case 't':
-				len = 't';
-				f++;
-				break;
-			case 'z':
-				len = 'z';
+			case 'q': case 'j': case 't': case 'z':
+				len = *f;
 				f++;
 				break;
 			default: break;
@@ -487,6 +490,11 @@ malloc_vsnprintf(char *str, size_t size, const char *format, va_list ap)
 			switch (*f) {
 				char *s;
 				size_t slen;
+			case '%':
+				/* %% */
+				APPEND_C(*f);
+				f++;
+				break;
 			case 'd': case 'i': {
 				intmax_t val JEMALLOC_CC_SILENCE_INIT(0);
 				char buf[D2S_BUFSIZE];
@@ -540,7 +548,7 @@ malloc_vsnprintf(char *str, size_t size, const char *format, va_list ap)
 				assert(len == '?' || len == 'l');
 				assert_not_implemented(len != 'l');
 				s = va_arg(ap, char *);
-				slen = (prec == -1) ? strlen(s) : prec;
+				slen = (prec < 0) ? strlen(s) : prec;
 				APPEND_PADDED_S(s, slen, width, left_justify);
 				f++;
 				break;
@@ -553,8 +561,7 @@ malloc_vsnprintf(char *str, size_t size, const char *format, va_list ap)
 				APPEND_PADDED_S(s, slen, width, left_justify);
 				f++;
 				break;
-			}
-			default: not_implemented();
+			} default: not_reached();
 			}
 			break;
 		} default: {
@@ -137,7 +137,7 @@ zone_destroy(malloc_zone_t *zone)
 {

 	/* This function should never be called. */
-	assert(false);
+	not_reached();
 	return (NULL);
 }

@@ -1,66 +0,0 @@
-#define	JEMALLOC_MANGLE
-#include "jemalloc_test.h"
-
-#define	NTHREADS 10
-
-void *
-je_thread_start(void *arg)
-{
-	unsigned thread_ind = (unsigned)(uintptr_t)arg;
-	unsigned arena_ind;
-	int r;
-	void *p;
-	size_t rsz, sz;
-
-	sz = sizeof(arena_ind);
-	if (mallctl("arenas.extend", &arena_ind, &sz, NULL, 0)
-	    != 0) {
-		malloc_printf("Error in arenas.extend\n");
-		abort();
-	}
-
-	if (thread_ind % 4 != 3) {
-		size_t mib[3];
-		size_t miblen = sizeof(mib) / sizeof(size_t);
-		const char *dss_precs[] = {"disabled", "primary", "secondary"};
-		const char *dss = dss_precs[thread_ind % 4];
-		if (mallctlnametomib("arena.0.dss", mib, &miblen) != 0) {
-			malloc_printf("Error in mallctlnametomib()\n");
-			abort();
-		}
-		mib[1] = arena_ind;
-		if (mallctlbymib(mib, miblen, NULL, NULL, (void *)&dss,
-		    sizeof(const char *))) {
-			malloc_printf("Error in mallctlbymib()\n");
-			abort();
-		}
-	}
-
-	r = allocm(&p, &rsz, 1, ALLOCM_ARENA(arena_ind));
-	if (r != ALLOCM_SUCCESS) {
-		malloc_printf("Unexpected allocm() error\n");
-		abort();
-	}
-
-	return (NULL);
-}
-
-int
-main(void)
-{
-	je_thread_t threads[NTHREADS];
-	unsigned i;
-
-	malloc_printf("Test begin\n");
-
-	for (i = 0; i < NTHREADS; i++) {
-		je_thread_create(&threads[i], je_thread_start,
-		    (void *)(uintptr_t)i);
-	}
-
-	for (i = 0; i < NTHREADS; i++)
-		je_thread_join(threads[i], NULL);
-
-	malloc_printf("Test end\n");
-	return (0);
-}
@@ -1,2 +0,0 @@
-Test begin
-Test end
@@ -1,25 +0,0 @@
-Test begin
-Alignment: 8
-Alignment: 16
-Alignment: 32
-Alignment: 64
-Alignment: 128
-Alignment: 256
-Alignment: 512
-Alignment: 1024
-Alignment: 2048
-Alignment: 4096
-Alignment: 8192
-Alignment: 16384
-Alignment: 32768
-Alignment: 65536
-Alignment: 131072
-Alignment: 262144
-Alignment: 524288
-Alignment: 1048576
-Alignment: 2097152
-Alignment: 4194304
-Alignment: 8388608
-Alignment: 16777216
-Alignment: 33554432
-Test end
@@ -1,118 +0,0 @@
-#define	JEMALLOC_MANGLE
-#include "jemalloc_test.h"
-
-void *
-je_thread_start(void *arg)
-{
-	int err;
-	void *p;
-	uint64_t a0, a1, d0, d1;
-	uint64_t *ap0, *ap1, *dp0, *dp1;
-	size_t sz, usize;
-
-	sz = sizeof(a0);
-	if ((err = mallctl("thread.allocated", &a0, &sz, NULL, 0))) {
-		if (err == ENOENT) {
-#ifdef JEMALLOC_STATS
-			assert(false);
-#endif
-			goto label_return;
-		}
-		malloc_printf("%s(): Error in mallctl(): %s\n", __func__,
-		    strerror(err));
-		exit(1);
-	}
-	sz = sizeof(ap0);
-	if ((err = mallctl("thread.allocatedp", &ap0, &sz, NULL, 0))) {
-		if (err == ENOENT) {
-#ifdef JEMALLOC_STATS
-			assert(false);
-#endif
-			goto label_return;
-		}
-		malloc_printf("%s(): Error in mallctl(): %s\n", __func__,
-		    strerror(err));
-		exit(1);
-	}
-	assert(*ap0 == a0);
-
-	sz = sizeof(d0);
-	if ((err = mallctl("thread.deallocated", &d0, &sz, NULL, 0))) {
-		if (err == ENOENT) {
-#ifdef JEMALLOC_STATS
-			assert(false);
-#endif
-			goto label_return;
-		}
-		malloc_printf("%s(): Error in mallctl(): %s\n", __func__,
-		    strerror(err));
-		exit(1);
-	}
-	sz = sizeof(dp0);
-	if ((err = mallctl("thread.deallocatedp", &dp0, &sz, NULL, 0))) {
-		if (err == ENOENT) {
-#ifdef JEMALLOC_STATS
-			assert(false);
-#endif
-			goto label_return;
-		}
-		malloc_printf("%s(): Error in mallctl(): %s\n", __func__,
-		    strerror(err));
-		exit(1);
-	}
-	assert(*dp0 == d0);
-
-	p = malloc(1);
-	if (p == NULL) {
-		malloc_printf("%s(): Error in malloc()\n", __func__);
-		exit(1);
-	}
-
-	sz = sizeof(a1);
-	mallctl("thread.allocated", &a1, &sz, NULL, 0);
-	sz = sizeof(ap1);
-	mallctl("thread.allocatedp", &ap1, &sz, NULL, 0);
-	assert(*ap1 == a1);
-	assert(ap0 == ap1);
-
-	usize = malloc_usable_size(p);
-	assert(a0 + usize <= a1);
-
-	free(p);
-
-	sz = sizeof(d1);
-	mallctl("thread.deallocated", &d1, &sz, NULL, 0);
-	sz = sizeof(dp1);
-	mallctl("thread.deallocatedp", &dp1, &sz, NULL, 0);
-	assert(*dp1 == d1);
-	assert(dp0 == dp1);
-
-	assert(d0 + usize <= d1);
-
-label_return:
-	return (NULL);
-}
-
-int
-main(void)
-{
-	int ret = 0;
-	je_thread_t thread;
-
-	malloc_printf("Test begin\n");
-
-	je_thread_start(NULL);
-
-	je_thread_create(&thread, je_thread_start, NULL);
-	je_thread_join(thread, (void *)&ret);
-
-	je_thread_start(NULL);
-
-	je_thread_create(&thread, je_thread_start, NULL);
-	je_thread_join(thread, (void *)&ret);
-
-	je_thread_start(NULL);
-
-	malloc_printf("Test end\n");
-	return (ret);
-}
@@ -1,2 +0,0 @@
-Test begin
-Test end
@@ -1,194 +0,0 @@
-#define	JEMALLOC_MANGLE
-#include "jemalloc_test.h"
-
-#define CHUNK 0x400000
-/* #define MAXALIGN ((size_t)UINT64_C(0x80000000000)) */
-#define MAXALIGN ((size_t)0x2000000LU)
-#define NITER 4
-
-int
-main(void)
-{
-	int r;
-	void *p;
-	size_t nsz, rsz, sz, alignment, total;
-	unsigned i;
-	void *ps[NITER];
-
-	malloc_printf("Test begin\n");
-
-	sz = 42;
-	nsz = 0;
-	r = nallocm(&nsz, sz, 0);
-	if (r != ALLOCM_SUCCESS) {
-		malloc_printf("Unexpected nallocm() error\n");
-		abort();
-	}
-	rsz = 0;
-	r = allocm(&p, &rsz, sz, 0);
-	if (r != ALLOCM_SUCCESS) {
-		malloc_printf("Unexpected allocm() error\n");
-		abort();
-	}
-	if (rsz < sz)
-		malloc_printf("Real size smaller than expected\n");
-	if (nsz != rsz)
-		malloc_printf("nallocm()/allocm() rsize mismatch\n");
-	if (dallocm(p, 0) != ALLOCM_SUCCESS)
-		malloc_printf("Unexpected dallocm() error\n");
-
-	r = allocm(&p, NULL, sz, 0);
-	if (r != ALLOCM_SUCCESS) {
-		malloc_printf("Unexpected allocm() error\n");
-		abort();
-	}
-	if (dallocm(p, 0) != ALLOCM_SUCCESS)
-		malloc_printf("Unexpected dallocm() error\n");
-
-	nsz = 0;
-	r = nallocm(&nsz, sz, ALLOCM_ZERO);
-	if (r != ALLOCM_SUCCESS) {
-		malloc_printf("Unexpected nallocm() error\n");
-		abort();
-	}
-	rsz = 0;
-	r = allocm(&p, &rsz, sz, ALLOCM_ZERO);
-	if (r != ALLOCM_SUCCESS) {
-		malloc_printf("Unexpected allocm() error\n");
-		abort();
-	}
-	if (nsz != rsz)
-		malloc_printf("nallocm()/allocm() rsize mismatch\n");
-	if (dallocm(p, 0) != ALLOCM_SUCCESS)
-		malloc_printf("Unexpected dallocm() error\n");
-
-#if LG_SIZEOF_PTR == 3
-	alignment = UINT64_C(0x8000000000000000);
-	sz        = UINT64_C(0x8000000000000000);
-#else
-	alignment = 0x80000000LU;
-	sz        = 0x80000000LU;
-#endif
-	nsz = 0;
-	r = nallocm(&nsz, sz, ALLOCM_ALIGN(alignment));
-	if (r == ALLOCM_SUCCESS) {
-		malloc_printf(
-		    "Expected error for nallocm(&nsz, %zu, %#x)\n",
-		    sz, ALLOCM_ALIGN(alignment));
-	}
-	rsz = 0;
-	r = allocm(&p, &rsz, sz, ALLOCM_ALIGN(alignment));
-	if (r == ALLOCM_SUCCESS) {
-		malloc_printf(
-		    "Expected error for allocm(&p, %zu, %#x)\n",
-		    sz, ALLOCM_ALIGN(alignment));
-	}
-	if (nsz != rsz)
-		malloc_printf("nallocm()/allocm() rsize mismatch\n");
-
-#if LG_SIZEOF_PTR == 3
-	alignment = UINT64_C(0x4000000000000000);
-	sz        = UINT64_C(0x8400000000000001);
-#else
-	alignment = 0x40000000LU;
-	sz        = 0x84000001LU;
-#endif
-	nsz = 0;
-	r = nallocm(&nsz, sz, ALLOCM_ALIGN(alignment));
-	if (r != ALLOCM_SUCCESS)
-		malloc_printf("Unexpected nallocm() error\n");
-	rsz = 0;
-	r = allocm(&p, &rsz, sz, ALLOCM_ALIGN(alignment));
-	if (r == ALLOCM_SUCCESS) {
-		malloc_printf(
-		    "Expected error for allocm(&p, %zu, %#x)\n",
-		    sz, ALLOCM_ALIGN(alignment));
-	}
-
-	alignment = 0x10LU;
-#if LG_SIZEOF_PTR == 3
-	sz = UINT64_C(0xfffffffffffffff0);
-#else
-	sz = 0xfffffff0LU;
-#endif
-	nsz = 0;
-	r = nallocm(&nsz, sz, ALLOCM_ALIGN(alignment));
-	if (r == ALLOCM_SUCCESS) {
-		malloc_printf(
-		    "Expected error for nallocm(&nsz, %zu, %#x)\n",
-		    sz, ALLOCM_ALIGN(alignment));
-	}
-	rsz = 0;
-	r = allocm(&p, &rsz, sz, ALLOCM_ALIGN(alignment));
-	if (r == ALLOCM_SUCCESS) {
-		malloc_printf(
-		    "Expected error for allocm(&p, %zu, %#x)\n",
-		    sz, ALLOCM_ALIGN(alignment));
-	}
-	if (nsz != rsz)
-		malloc_printf("nallocm()/allocm() rsize mismatch\n");
-
-	for (i = 0; i < NITER; i++)
-		ps[i] = NULL;
-
-	for (alignment = 8;
-	    alignment <= MAXALIGN;
-	    alignment <<= 1) {
-		total = 0;
-		malloc_printf("Alignment: %zu\n", alignment);
-		for (sz = 1;
-		    sz < 3 * alignment && sz < (1U << 31);
-		    sz += (alignment >> (LG_SIZEOF_PTR-1)) - 1) {
-			for (i = 0; i < NITER; i++) {
-				nsz = 0;
-				r = nallocm(&nsz, sz,
-				    ALLOCM_ALIGN(alignment) | ALLOCM_ZERO);
-				if (r != ALLOCM_SUCCESS) {
-					malloc_printf(
-					    "nallocm() error for size %zu"
-					    " (%#zx): %d\n",
-					    sz, sz, r);
-					exit(1);
-				}
-				rsz = 0;
-				r = allocm(&ps[i], &rsz, sz,
-				    ALLOCM_ALIGN(alignment) | ALLOCM_ZERO);
-				if (r != ALLOCM_SUCCESS) {
-					malloc_printf(
-					    "allocm() error for size %zu"
-					    " (%#zx): %d\n",
-					    sz, sz, r);
-					exit(1);
-				}
-				if (rsz < sz) {
-					malloc_printf(
-					    "Real size smaller than"
-					    " expected\n");
-				}
-				if (nsz != rsz) {
-					malloc_printf(
-					    "nallocm()/allocm() rsize"
-					    " mismatch\n");
-				}
-				if ((uintptr_t)p & (alignment-1)) {
-					malloc_printf(
-					    "%p inadequately aligned for"
-					    " alignment: %zu\n", p, alignment);
-				}
-				sallocm(ps[i], &rsz, 0);
-				total += rsz;
-				if (total >= (MAXALIGN << 1))
-					break;
-			}
-			for (i = 0; i < NITER; i++) {
-				if (ps[i] != NULL) {
-					dallocm(ps[i], 0);
-					ps[i] = NULL;
-				}
-			}
-		}
-	}
-
-	malloc_printf("Test end\n");
-	return (0);
-}
@@ -1,25 +0,0 @@
-Test begin
-Alignment: 8
-Alignment: 16
-Alignment: 32
-Alignment: 64
-Alignment: 128
-Alignment: 256
-Alignment: 512
-Alignment: 1024
-Alignment: 2048
-Alignment: 4096
-Alignment: 8192
-Alignment: 16384
-Alignment: 32768
-Alignment: 65536
-Alignment: 131072
-Alignment: 262144
-Alignment: 524288
-Alignment: 1048576
-Alignment: 2097152
-Alignment: 4194304
-Alignment: 8388608
-Alignment: 16777216
-Alignment: 33554432
-Test end
@@ -1,2 +0,0 @@
-Test begin
-Test end
@@ -0,0 +1,186 @@
+/*
+ * This file derives from SFMT 1.3.3
+ * (http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html), which was
+ * released under the terms of the following license:
+ *
+ *   Copyright (c) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima
+ *   University. 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 the Hiroshima University 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.
+ */
+/** 
+ * @file SFMT-alti.h 
+ *
+ * @brief SIMD oriented Fast Mersenne Twister(SFMT)
+ * pseudorandom number generator
+ *
+ * @author Mutsuo Saito (Hiroshima University)
+ * @author Makoto Matsumoto (Hiroshima University)
+ *
+ * Copyright (C) 2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima
+ * University. All rights reserved.
+ *
+ * The new BSD License is applied to this software.
+ * see LICENSE.txt
+ */
+
+#ifndef SFMT_ALTI_H
+#define SFMT_ALTI_H
+
+/**
+ * This function represents the recursion formula in AltiVec and BIG ENDIAN.
+ * @param a a 128-bit part of the interal state array
+ * @param b a 128-bit part of the interal state array
+ * @param c a 128-bit part of the interal state array
+ * @param d a 128-bit part of the interal state array
+ * @return output
+ */
+JEMALLOC_ALWAYS_INLINE
+vector unsigned int vec_recursion(vector unsigned int a,
+						vector unsigned int b,
+						vector unsigned int c,
+						vector unsigned int d) {
+
+    const vector unsigned int sl1 = ALTI_SL1;
+    const vector unsigned int sr1 = ALTI_SR1;
+#ifdef ONLY64
+    const vector unsigned int mask = ALTI_MSK64;
+    const vector unsigned char perm_sl = ALTI_SL2_PERM64;
+    const vector unsigned char perm_sr = ALTI_SR2_PERM64;
+#else
+    const vector unsigned int mask = ALTI_MSK;
+    const vector unsigned char perm_sl = ALTI_SL2_PERM;
+    const vector unsigned char perm_sr = ALTI_SR2_PERM;
+#endif
+    vector unsigned int v, w, x, y, z;
+    x = vec_perm(a, (vector unsigned int)perm_sl, perm_sl);
+    v = a;
+    y = vec_sr(b, sr1);
+    z = vec_perm(c, (vector unsigned int)perm_sr, perm_sr);
+    w = vec_sl(d, sl1);
+    z = vec_xor(z, w);
+    y = vec_and(y, mask);
+    v = vec_xor(v, x);
+    z = vec_xor(z, y);
+    z = vec_xor(z, v);
+    return z;
+}
+
+/**
+ * This function fills the internal state array with pseudorandom
+ * integers.
+ */
+JEMALLOC_INLINE void gen_rand_all(sfmt_t *ctx) {
+    int i;
+    vector unsigned int r, r1, r2;
+
+    r1 = ctx->sfmt[N - 2].s;
+    r2 = ctx->sfmt[N - 1].s;
+    for (i = 0; i < N - POS1; i++) {
+	r = vec_recursion(ctx->sfmt[i].s, ctx->sfmt[i + POS1].s, r1, r2);
+	ctx->sfmt[i].s = r;
+	r1 = r2;
+	r2 = r;
+    }
+    for (; i < N; i++) {
+	r = vec_recursion(ctx->sfmt[i].s, ctx->sfmt[i + POS1 - N].s, r1, r2);
+	ctx->sfmt[i].s = r;
+	r1 = r2;
+	r2 = r;
+    }
+}
+
+/**
+ * This function fills the user-specified array with pseudorandom
+ * integers.
+ *
+ * @param array an 128-bit array to be filled by pseudorandom numbers.  
+ * @param size number of 128-bit pesudorandom numbers to be generated.
+ */
+JEMALLOC_INLINE void gen_rand_array(sfmt_t *ctx, w128_t *array, int size) {
+    int i, j;
+    vector unsigned int r, r1, r2;
+
+    r1 = ctx->sfmt[N - 2].s;
+    r2 = ctx->sfmt[N - 1].s;
+    for (i = 0; i < N - POS1; i++) {
+	r = vec_recursion(ctx->sfmt[i].s, ctx->sfmt[i + POS1].s, r1, r2);
+	array[i].s = r;
+	r1 = r2;
+	r2 = r;
+    }
+    for (; i < N; i++) {
+	r = vec_recursion(ctx->sfmt[i].s, array[i + POS1 - N].s, r1, r2);
+	array[i].s = r;
+	r1 = r2;
+	r2 = r;
+    }
+    /* main loop */
+    for (; i < size - N; i++) {
+	r = vec_recursion(array[i - N].s, array[i + POS1 - N].s, r1, r2);
+	array[i].s = r;
+	r1 = r2;
+	r2 = r;
+    }
+    for (j = 0; j < 2 * N - size; j++) {
+	ctx->sfmt[j].s = array[j + size - N].s;
+    }
+    for (; i < size; i++) {
+	r = vec_recursion(array[i - N].s, array[i + POS1 - N].s, r1, r2);
+	array[i].s = r;
+	ctx->sfmt[j++].s = r;
+	r1 = r2;
+	r2 = r;
+    }
+}
+
+#ifndef ONLY64
+#if defined(__APPLE__)
+#define ALTI_SWAP (vector unsigned char) \
+	(4, 5, 6, 7, 0, 1, 2, 3, 12, 13, 14, 15, 8, 9, 10, 11)
+#else
+#define ALTI_SWAP {4, 5, 6, 7, 0, 1, 2, 3, 12, 13, 14, 15, 8, 9, 10, 11}
+#endif
+/**
+ * This function swaps high and low 32-bit of 64-bit integers in user
+ * specified array.
+ *
+ * @param array an 128-bit array to be swaped.
+ * @param size size of 128-bit array.
+ */
+JEMALLOC_INLINE void swap(w128_t *array, int size) {
+    int i;
+    const vector unsigned char perm = ALTI_SWAP;
+
+    for (i = 0; i < size; i++) {
+	array[i].s = vec_perm(array[i].s, (vector unsigned int)perm, perm);
+    }
+}
+#endif
+
+#endif
@@ -0,0 +1,132 @@
+/*
+ * This file derives from SFMT 1.3.3
+ * (http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html), which was
+ * released under the terms of the following license:
+ *
+ *   Copyright (c) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima
+ *   University. 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 the Hiroshima University 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.
+ */
+#ifndef SFMT_PARAMS_H
+#define SFMT_PARAMS_H
+
+#if !defined(MEXP)
+#ifdef __GNUC__
+  #warning "MEXP is not defined. I assume MEXP is 19937."
+#endif
+  #define MEXP 19937
+#endif
+/*-----------------
+  BASIC DEFINITIONS
+  -----------------*/
+/** Mersenne Exponent. The period of the sequence 
+ *  is a multiple of 2^MEXP-1.
+ * #define MEXP 19937 */
+/** SFMT generator has an internal state array of 128-bit integers,
+ * and N is its size. */
+#define N (MEXP / 128 + 1)
+/** N32 is the size of internal state array when regarded as an array
+ * of 32-bit integers.*/
+#define N32 (N * 4)
+/** N64 is the size of internal state array when regarded as an array
+ * of 64-bit integers.*/
+#define N64 (N * 2)
+
+/*----------------------
+  the parameters of SFMT
+  following definitions are in paramsXXXX.h file.
+  ----------------------*/
+/** the pick up position of the array.
+#define POS1 122 
+*/
+
+/** the parameter of shift left as four 32-bit registers.
+#define SL1 18
+ */
+
+/** the parameter of shift left as one 128-bit register. 
+ * The 128-bit integer is shifted by (SL2 * 8) bits. 
+#define SL2 1 
+*/
+
+/** the parameter of shift right as four 32-bit registers.
+#define SR1 11
+*/
+
+/** the parameter of shift right as one 128-bit register. 
+ * The 128-bit integer is shifted by (SL2 * 8) bits. 
+#define SR2 1 
+*/
+
+/** A bitmask, used in the recursion.  These parameters are introduced
+ * to break symmetry of SIMD.
+#define MSK1 0xdfffffefU
+#define MSK2 0xddfecb7fU
+#define MSK3 0xbffaffffU
+#define MSK4 0xbffffff6U 
+*/
+
+/** These definitions are part of a 128-bit period certification vector.
+#define PARITY1	0x00000001U
+#define PARITY2	0x00000000U
+#define PARITY3	0x00000000U
+#define PARITY4	0xc98e126aU
+*/
+
+#if MEXP == 607
+  #include "test/SFMT-params607.h"
+#elif MEXP == 1279
+  #include "test/SFMT-params1279.h"
+#elif MEXP == 2281
+  #include "test/SFMT-params2281.h"
+#elif MEXP == 4253
+  #include "test/SFMT-params4253.h"
+#elif MEXP == 11213
+  #include "test/SFMT-params11213.h"
+#elif MEXP == 19937
+  #include "test/SFMT-params19937.h"
+#elif MEXP == 44497
+  #include "test/SFMT-params44497.h"
+#elif MEXP == 86243
+  #include "test/SFMT-params86243.h"
+#elif MEXP == 132049
+  #include "test/SFMT-params132049.h"
+#elif MEXP == 216091
+  #include "test/SFMT-params216091.h"
+#else
+#ifdef __GNUC__
+  #error "MEXP is not valid."
+  #undef MEXP
+#else
+  #undef MEXP
+#endif
+
+#endif
+
+#endif /* SFMT_PARAMS_H */
@@ -0,0 +1,81 @@
+/*
+ * This file derives from SFMT 1.3.3
+ * (http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/index.html), which was
+ * released under the terms of the following license:
+ *
+ *   Copyright (c) 2006,2007 Mutsuo Saito, Makoto Matsumoto and Hiroshima
+ *   University. 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 the Hiroshima University 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.
+ */
+#ifndef SFMT_PARAMS11213_H
+#define SFMT_PARAMS11213_H
+
+#define POS1	68
+#define SL1	14
+#define SL2	3
+#define SR1	7
+#define SR2	3
+#define MSK1	0xeffff7fbU
+#define MSK2	0xffffffefU
+#define MSK3	0xdfdfbfffU
+#define MSK4	0x7fffdbfdU
+#define PARITY1	0x00000001U
+#define PARITY2	0x00000000U
+#define PARITY3	0xe8148000U
+#define PARITY4	0xd0c7afa3U
+
+
+/* PARAMETERS FOR ALTIVEC */
+#if defined(__APPLE__)	/* For OSX */
+    #define ALTI_SL1	(vector unsigned int)(SL1, SL1, SL1, SL1)
+    #define ALTI_SR1	(vector unsigned int)(SR1, SR1, SR1, SR1)
+    #define ALTI_MSK	(vector unsigned int)(MSK1, MSK2, MSK3, MSK4)
+    #define ALTI_MSK64 \
+	(vector unsigned int)(MSK2, MSK1, MSK4, MSK3)
+    #define ALTI_SL2_PERM \
+	(vector unsigned char)(3,21,21,21,7,0,1,2,11,4,5,6,15,8,9,10)
+    #define ALTI_SL2_PERM64 \
+	(vector unsigned char)(3,4,5,6,7,29,29,29,11,12,13,14,15,0,1,2)
+    #define ALTI_SR2_PERM \
+	(vector unsigned char)(5,6,7,0,9,10,11,4,13,14,15,8,19,19,19,12)
+    #define ALTI_SR2_PERM64 \
+	(vector unsigned char)(13,14,15,0,1,2,3,4,19,19,19,8,9,10,11,12)
+#else	/* For OTHER OSs(Linux?) */
+    #define ALTI_SL1	{SL1, SL1, SL1, SL1}
+    #define ALTI_SR1	{SR1, SR1, SR1, SR1}
+    #define ALTI_MSK	{MSK1, MSK2, MSK3, MSK4}
+    #define ALTI_MSK64	{MSK2, MSK1, MSK4, MSK3}
+    #define ALTI_SL2_PERM	{3,21,21,21,7,0,1,2,11,4,5,6,15,8,9,10}
+    #define ALTI_SL2_PERM64	{3,4,5,6,7,29,29,29,11,12,13,14,15,0,1,2}
+    #define ALTI_SR2_PERM	{5,6,7,0,9,10,11,4,13,14,15,8,19,19,19,12}
+    #define ALTI_SR2_PERM64	{13,14,15,0,1,2,3,4,19,19,19,8,9,10,11,12}
+#endif	/* For OSX */
+#define IDSTR	"SFMT-11213:68-14-3-7-3:effff7fb-ffffffef-dfdfbfff-7fffdbfd"
+
+#endif /* SFMT_PARAMS11213_H */
--- a/Show More
+++ b/Show More