Update jemalloc to 4.0.2.

20 files changed, 619 insertions, 427 deletions

diff --git a/contrib/jemalloc/ChangeLog b/contrib/jemalloc/ChangeLog
index 0cf887c..58e4462 100644
--- a/contrib/jemalloc/ChangeLog
+++ b/contrib/jemalloc/ChangeLog
@@ -4,6 +4,76 @@ brevity.  Much more detail can be found in the git revision history:
 
     https://github.com/jemalloc/jemalloc
 
+* 4.0.2 (September 21, 2015)
+
+  This bugfix release addresses a few bugs specific to heap profiling.
+
+  Bug fixes:
+  - Fix ixallocx_prof_sample() to never modify nor create sampled small
+    allocations.  xallocx() is in general incapable of moving small allocations,
+    so this fix removes buggy code without loss of generality.
+  - Fix irallocx_prof_sample() to always allocate large regions, even when
+    alignment is non-zero.
+  - Fix prof_alloc_rollback() to read tdata from thread-specific data rather
+    than dereferencing a potentially invalid tctx.
+
+* 4.0.1 (September 15, 2015)
+
+  This is a bugfix release that is somewhat high risk due to the amount of
+  refactoring required to address deep xallocx() problems.  As a side effect of
+  these fixes, xallocx() now tries harder to partially fulfill requests for
+  optional extra space.  Note that a couple of minor heap profiling
+  optimizations are included, but these are better thought of as performance
+  fixes that were integral to disovering most of the other bugs.
+
+  Optimizations:
+  - Avoid a chunk metadata read in arena_prof_tctx_set(), since it is in the
+    fast path when heap profiling is enabled.  Additionally, split a special
+    case out into arena_prof_tctx_reset(), which also avoids chunk metadata
+    reads.
+  - Optimize irallocx_prof() to optimistically update the sampler state.  The
+    prior implementation appears to have been a holdover from when
+    rallocx()/xallocx() functionality was combined as rallocm().
+
+  Bug fixes:
+  - Fix TLS configuration such that it is enabled by default for platforms on
+    which it works correctly.
+  - Fix arenas_cache_cleanup() and arena_get_hard() to handle
+    allocation/deallocation within the application's thread-specific data
+    cleanup functions even after arenas_cache is torn down.
+  - Fix xallocx() bugs related to size+extra exceeding HUGE_MAXCLASS.
+  - Fix chunk purge hook calls for in-place huge shrinking reallocation to
+    specify the old chunk size rather than the new chunk size.  This bug caused
+    no correctness issues for the default chunk purge function, but was
+    visible to custom functions set via the "arena.<i>.chunk_hooks" mallctl.
+  - Fix heap profiling bugs:
+    + Fix heap profiling to distinguish among otherwise identical sample sites
+      with interposed resets (triggered via the "prof.reset" mallctl).  This bug
+      could cause data structure corruption that would most likely result in a
+      segfault.
+    + Fix irealloc_prof() to prof_alloc_rollback() on OOM.
+    + Make one call to prof_active_get_unlocked() per allocation event, and use
+      the result throughout the relevant functions that handle an allocation
+      event.  Also add a missing check in prof_realloc().  These fixes protect
+      allocation events against concurrent prof_active changes.
+    + Fix ixallocx_prof() to pass usize_max and zero to ixallocx_prof_sample()
+      in the correct order.
+    + Fix prof_realloc() to call prof_free_sampled_object() after calling
+      prof_malloc_sample_object().  Prior to this fix, if tctx and old_tctx were
+      the same, the tctx could have been prematurely destroyed.
+  - Fix portability bugs:
+    + Don't bitshift by negative amounts when encoding/decoding run sizes in
+      chunk header maps.  This affected systems with page sizes greater than 8
+      KiB.
+    + Rename index_t to szind_t to avoid an existing type on Solaris.
+    + Add JEMALLOC_CXX_THROW to the memalign() function prototype, in order to
+      match glibc and avoid compilation errors when including both
+      jemalloc/jemalloc.h and malloc.h in C++ code.
+    + Don't assume that /bin/sh is appropriate when running size_classes.sh
+      during configuration.
+    + Consider __sparcv9 a synonym for __sparc64__ when defining LG_QUANTUM.
+    + Link tests to librt if it contains clock_gettime(2).
+
 * 4.0.0 (August 17, 2015)
 
   This version contains many speed and space optimizations, both minor and
diff --git a/contrib/jemalloc/FREEBSD-diffs b/contrib/jemalloc/FREEBSD-diffs
index 7d6955d..25e823c 100644
--- a/contrib/jemalloc/FREEBSD-diffs
+++ b/contrib/jemalloc/FREEBSD-diffs
@@ -47,7 +47,7 @@ index 8fc774b..fdbef95 100644
 +  </refsect1>
  </refentry>
 diff --git a/include/jemalloc/internal/jemalloc_internal.h.in b/include/jemalloc/internal/jemalloc_internal.h.in
-index 7a137b6..b0001e9 100644
+index 8536a3e..0c2a81f 100644
 --- a/include/jemalloc/internal/jemalloc_internal.h.in
 +++ b/include/jemalloc/internal/jemalloc_internal.h.in
 @@ -8,6 +8,9 @@
@@ -111,10 +111,10 @@ index f051f29..561378f 100644
  
  #endif /* JEMALLOC_H_EXTERNS */
 diff --git a/include/jemalloc/internal/private_symbols.txt b/include/jemalloc/internal/private_symbols.txt
-index dbf6aa7..f87dba8 100644
+index a90021a..34904bf 100644
 --- a/include/jemalloc/internal/private_symbols.txt
 +++ b/include/jemalloc/internal/private_symbols.txt
-@@ -277,7 +277,6 @@ iralloct_realign
+@@ -280,7 +280,6 @@ iralloct_realign
  isalloc
  isdalloct
  isqalloc
@@ -282,7 +282,7 @@ index f943891..47d032c 100755
 +#include "jemalloc_FreeBSD.h"
  EOF
 diff --git a/src/jemalloc.c b/src/jemalloc.c
-index ed7863b..d078a1f 100644
+index 5a2d324..b6cbb79 100644
 --- a/src/jemalloc.c
 +++ b/src/jemalloc.c
 @@ -4,6 +4,10 @@
@@ -296,7 +296,7 @@ index ed7863b..d078a1f 100644
  /* Runtime configuration options. */
  const char	*je_malloc_conf JEMALLOC_ATTR(weak);
  bool	opt_abort =
-@@ -2475,6 +2479,107 @@ je_malloc_usable_size(JEMALLOC_USABLE_SIZE_CONST void *ptr)
+@@ -2490,6 +2494,107 @@ je_malloc_usable_size(JEMALLOC_USABLE_SIZE_CONST void *ptr)
   */
  /******************************************************************************/
  /*
@@ -404,7 +404,7 @@ index ed7863b..d078a1f 100644
   * The following functions are used by threading libraries for protection of
   * malloc during fork().
   */
-@@ -2575,4 +2680,11 @@ jemalloc_postfork_child(void)
+@@ -2590,4 +2695,11 @@ jemalloc_postfork_child(void)
  	ctl_postfork_child();
  }
  
diff --git a/contrib/jemalloc/VERSION b/contrib/jemalloc/VERSION
index 6802c56..210a4fe 100644
--- a/contrib/jemalloc/VERSION
+++ b/contrib/jemalloc/VERSION
@@ -1 +1 @@
-4.0.0-0-g6e98caf8f064482b9ab292ef3638dea67420bbc2
+4.0.2-0-g486d249fb4715fd3de679b6c2a04f7e657883111
diff --git a/contrib/jemalloc/doc/jemalloc.3 b/contrib/jemalloc/doc/jemalloc.3
index 2afb862..2582062 100644
--- a/contrib/jemalloc/doc/jemalloc.3
+++ b/contrib/jemalloc/doc/jemalloc.3
@@ -2,12 +2,12 @@
 .\"     Title: JEMALLOC
 .\"    Author: Jason Evans
 .\" Generator: DocBook XSL Stylesheets v1.76.1 <http://docbook.sf.net/>
-.\"      Date: 08/18/2015
+.\"      Date: 09/21/2015
 .\"    Manual: User Manual
-.\"    Source: jemalloc 4.0.0-0-g6e98caf8f064482b9ab292ef3638dea67420bbc2
+.\"    Source: jemalloc 4.0.2-0-g486d249fb4715fd3de679b6c2a04f7e657883111
 .\"  Language: English
 .\"
-.TH "JEMALLOC" "3" "08/18/2015" "jemalloc 4.0.0-0-g6e98caf8f064" "User Manual"
+.TH "JEMALLOC" "3" "09/21/2015" "jemalloc 4.0.2-0-g486d249fb471" "User Manual"
 .\" -----------------------------------------------------------------
 .\" * Define some portability stuff
 .\" -----------------------------------------------------------------
@@ -31,7 +31,7 @@
 jemalloc \- general purpose memory allocation functions
 .SH "LIBRARY"
 .PP
-This manual describes jemalloc 4\&.0\&.0\-0\-g6e98caf8f064482b9ab292ef3638dea67420bbc2\&. More information can be found at the
+This manual describes jemalloc 4\&.0\&.2\-0\-g486d249fb4715fd3de679b6c2a04f7e657883111\&. More information can be found at the
 \m[blue]\fBjemalloc website\fR\m[]\&\s-2\u[1]\d\s+2\&.
 .PP
 The following configuration options are enabled in libc\*(Aqs built\-in jemalloc:
diff --git a/contrib/jemalloc/include/jemalloc/internal/arena.h b/contrib/jemalloc/include/jemalloc/internal/arena.h
index cb015ee..12c6179 100644
--- a/contrib/jemalloc/include/jemalloc/internal/arena.h
+++ b/contrib/jemalloc/include/jemalloc/internal/arena.h
@@ -39,7 +39,7 @@ typedef struct arena_s arena_t;
 #ifdef JEMALLOC_ARENA_STRUCTS_A
 struct arena_run_s {
 	/* Index of bin this run is associated with. */
-	index_t		binind;
+	szind_t		binind;
 
 	/* Number of free regions in run. */
 	unsigned	nfree;
@@ -424,7 +424,7 @@ extern arena_bin_info_t	arena_bin_info[NBINS];
 extern size_t		map_bias; /* Number of arena chunk header pages. */
 extern size_t		map_misc_offset;
 extern size_t		arena_maxrun; /* Max run size for arenas. */
-extern size_t		arena_maxclass; /* Max size class for arenas. */
+extern size_t		large_maxclass; /* Max large size class. */
 extern unsigned		nlclasses; /* Number of large size classes. */
 extern unsigned		nhclasses; /* Number of huge size classes. */
 
@@ -448,7 +448,7 @@ bool	arena_lg_dirty_mult_set(arena_t *arena, ssize_t lg_dirty_mult);
 void	arena_maybe_purge(arena_t *arena);
 void	arena_purge_all(arena_t *arena);
 void	arena_tcache_fill_small(arena_t *arena, tcache_bin_t *tbin,
-    index_t binind, uint64_t prof_accumbytes);
+    szind_t binind, uint64_t prof_accumbytes);
 void	arena_alloc_junk_small(void *ptr, arena_bin_info_t *bin_info,
     bool zero);
 #ifdef JEMALLOC_JET
@@ -488,7 +488,7 @@ extern arena_ralloc_junk_large_t *arena_ralloc_junk_large;
 bool	arena_ralloc_no_move(void *ptr, size_t oldsize, size_t size,
     size_t extra, bool zero);
 void	*arena_ralloc(tsd_t *tsd, arena_t *arena, void *ptr, size_t oldsize,
-    size_t size, size_t extra, size_t alignment, bool zero, tcache_t *tcache);
+    size_t size, size_t alignment, bool zero, tcache_t *tcache);
 dss_prec_t	arena_dss_prec_get(arena_t *arena);
 bool	arena_dss_prec_set(arena_t *arena, dss_prec_t dss_prec);
 ssize_t	arena_lg_dirty_mult_default_get(void);
@@ -519,17 +519,19 @@ arena_chunk_map_misc_t	*arena_run_to_miscelm(arena_run_t *run);
 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_size_decode(size_t mapbits);
 size_t	arena_mapbits_unallocated_size_get(arena_chunk_t *chunk,
     size_t pageind);
 size_t	arena_mapbits_large_size_get(arena_chunk_t *chunk, size_t pageind);
 size_t	arena_mapbits_small_runind_get(arena_chunk_t *chunk, size_t pageind);
-index_t	arena_mapbits_binind_get(arena_chunk_t *chunk, size_t pageind);
+szind_t	arena_mapbits_binind_get(arena_chunk_t *chunk, size_t pageind);
 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_decommitted_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);
+size_t	arena_mapbits_size_encode(size_t size);
 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,
@@ -539,21 +541,23 @@ void	arena_mapbits_internal_set(arena_chunk_t *chunk, size_t pageind,
 void	arena_mapbits_large_set(arena_chunk_t *chunk, size_t pageind,
     size_t size, size_t flags);
 void	arena_mapbits_large_binind_set(arena_chunk_t *chunk, size_t pageind,
-    index_t binind);
+    szind_t binind);
 void	arena_mapbits_small_set(arena_chunk_t *chunk, size_t pageind,
-    size_t runind, index_t binind, size_t flags);
+    size_t runind, szind_t binind, size_t flags);
 void	arena_metadata_allocated_add(arena_t *arena, size_t size);
 void	arena_metadata_allocated_sub(arena_t *arena, size_t size);
 size_t	arena_metadata_allocated_get(arena_t *arena);
 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);
-index_t	arena_ptr_small_binind_get(const void *ptr, size_t mapbits);
-index_t	arena_bin_index(arena_t *arena, arena_bin_t *bin);
+szind_t	arena_ptr_small_binind_get(const void *ptr, size_t mapbits);
+szind_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_tctx_t	*arena_prof_tctx_get(const void *ptr);
-void	arena_prof_tctx_set(const void *ptr, prof_tctx_t *tctx);
+void	arena_prof_tctx_set(const void *ptr, size_t usize, prof_tctx_t *tctx);
+void	arena_prof_tctx_reset(const void *ptr, size_t usize,
+    const void *old_ptr, prof_tctx_t *old_tctx);
 void	*arena_malloc(tsd_t *tsd, arena_t *arena, size_t size, bool zero,
     tcache_t *tcache);
 arena_t	*arena_aalloc(const void *ptr);
@@ -653,13 +657,29 @@ arena_mapbits_get(arena_chunk_t *chunk, size_t pageind)
 }
 
 JEMALLOC_ALWAYS_INLINE size_t
+arena_mapbits_size_decode(size_t mapbits)
+{
+	size_t size;
+
+#if CHUNK_MAP_SIZE_SHIFT > 0
+	size = (mapbits & CHUNK_MAP_SIZE_MASK) >> CHUNK_MAP_SIZE_SHIFT;
+#elif CHUNK_MAP_SIZE_SHIFT == 0
+	size = mapbits & CHUNK_MAP_SIZE_MASK;
+#else
+	size = (mapbits & CHUNK_MAP_SIZE_MASK) << -CHUNK_MAP_SIZE_SHIFT;
+#endif
+
+	return (size);
+}
+
+JEMALLOC_ALWAYS_INLINE size_t
 arena_mapbits_unallocated_size_get(arena_chunk_t *chunk, size_t pageind)
 {
 	size_t mapbits;
 
 	mapbits = arena_mapbits_get(chunk, pageind);
 	assert((mapbits & (CHUNK_MAP_LARGE|CHUNK_MAP_ALLOCATED)) == 0);
-	return ((mapbits & CHUNK_MAP_SIZE_MASK) >> CHUNK_MAP_SIZE_SHIFT);
+	return (arena_mapbits_size_decode(mapbits));
 }
 
 JEMALLOC_ALWAYS_INLINE size_t
@@ -670,7 +690,7 @@ arena_mapbits_large_size_get(arena_chunk_t *chunk, size_t pageind)
 	mapbits = arena_mapbits_get(chunk, pageind);
 	assert((mapbits & (CHUNK_MAP_LARGE|CHUNK_MAP_ALLOCATED)) ==
 	    (CHUNK_MAP_LARGE|CHUNK_MAP_ALLOCATED));
-	return ((mapbits & CHUNK_MAP_SIZE_MASK) >> CHUNK_MAP_SIZE_SHIFT);
+	return (arena_mapbits_size_decode(mapbits));
 }
 
 JEMALLOC_ALWAYS_INLINE size_t
@@ -684,11 +704,11 @@ arena_mapbits_small_runind_get(arena_chunk_t *chunk, size_t pageind)
 	return (mapbits >> CHUNK_MAP_RUNIND_SHIFT);
 }
 
-JEMALLOC_ALWAYS_INLINE index_t
+JEMALLOC_ALWAYS_INLINE szind_t
 arena_mapbits_binind_get(arena_chunk_t *chunk, size_t pageind)
 {
 	size_t mapbits;
-	index_t binind;
+	szind_t binind;
 
 	mapbits = arena_mapbits_get(chunk, pageind);
 	binind = (mapbits & CHUNK_MAP_BININD_MASK) >> CHUNK_MAP_BININD_SHIFT;
@@ -754,6 +774,23 @@ arena_mapbitsp_write(size_t *mapbitsp, size_t mapbits)
 	*mapbitsp = mapbits;
 }
 
+JEMALLOC_ALWAYS_INLINE size_t
+arena_mapbits_size_encode(size_t size)
+{
+	size_t mapbits;
+
+#if CHUNK_MAP_SIZE_SHIFT > 0
+	mapbits = size << CHUNK_MAP_SIZE_SHIFT;
+#elif CHUNK_MAP_SIZE_SHIFT == 0
+	mapbits = size;
+#else
+	mapbits = size >> -CHUNK_MAP_SIZE_SHIFT;
+#endif
+
+	assert((mapbits & ~CHUNK_MAP_SIZE_MASK) == 0);
+	return (mapbits);
+}
+
 JEMALLOC_ALWAYS_INLINE void
 arena_mapbits_unallocated_set(arena_chunk_t *chunk, size_t pageind, size_t size,
     size_t flags)
@@ -761,11 +798,10 @@ arena_mapbits_unallocated_set(arena_chunk_t *chunk, size_t pageind, size_t size,
 	size_t *mapbitsp = arena_mapbitsp_get(chunk, pageind);
 
 	assert((size & PAGE_MASK) == 0);
-	assert(((size << CHUNK_MAP_SIZE_SHIFT) & ~CHUNK_MAP_SIZE_MASK) == 0);
 	assert((flags & CHUNK_MAP_FLAGS_MASK) == flags);
 	assert((flags & CHUNK_MAP_DECOMMITTED) == 0 || (flags &
 	    (CHUNK_MAP_DIRTY|CHUNK_MAP_UNZEROED)) == 0);
-	arena_mapbitsp_write(mapbitsp, (size << CHUNK_MAP_SIZE_SHIFT) |
+	arena_mapbitsp_write(mapbitsp, arena_mapbits_size_encode(size) |
 	    CHUNK_MAP_BININD_INVALID | flags);
 }
 
@@ -777,10 +813,9 @@ arena_mapbits_unallocated_size_set(arena_chunk_t *chunk, size_t pageind,
 	size_t mapbits = arena_mapbitsp_read(mapbitsp);
 
 	assert((size & PAGE_MASK) == 0);
-	assert(((size << CHUNK_MAP_SIZE_SHIFT) & ~CHUNK_MAP_SIZE_MASK) == 0);
 	assert((mapbits & (CHUNK_MAP_LARGE|CHUNK_MAP_ALLOCATED)) == 0);
-	arena_mapbitsp_write(mapbitsp, (size << CHUNK_MAP_SIZE_SHIFT) | (mapbits
-	    & ~CHUNK_MAP_SIZE_MASK));
+	arena_mapbitsp_write(mapbitsp, arena_mapbits_size_encode(size) |
+	    (mapbits & ~CHUNK_MAP_SIZE_MASK));
 }
 
 JEMALLOC_ALWAYS_INLINE void
@@ -799,18 +834,17 @@ arena_mapbits_large_set(arena_chunk_t *chunk, size_t pageind, size_t size,
 	size_t *mapbitsp = arena_mapbitsp_get(chunk, pageind);
 
 	assert((size & PAGE_MASK) == 0);
-	assert(((size << CHUNK_MAP_SIZE_SHIFT) & ~CHUNK_MAP_SIZE_MASK) == 0);
 	assert((flags & CHUNK_MAP_FLAGS_MASK) == flags);
 	assert((flags & CHUNK_MAP_DECOMMITTED) == 0 || (flags &
 	    (CHUNK_MAP_DIRTY|CHUNK_MAP_UNZEROED)) == 0);
-	arena_mapbitsp_write(mapbitsp, (size << CHUNK_MAP_SIZE_SHIFT) |
+	arena_mapbitsp_write(mapbitsp, arena_mapbits_size_encode(size) |
 	    CHUNK_MAP_BININD_INVALID | flags | CHUNK_MAP_LARGE |
 	    CHUNK_MAP_ALLOCATED);
 }
 
 JEMALLOC_ALWAYS_INLINE void
 arena_mapbits_large_binind_set(arena_chunk_t *chunk, size_t pageind,
-    index_t binind)
+    szind_t binind)
 {
 	size_t *mapbitsp = arena_mapbitsp_get(chunk, pageind);
 	size_t mapbits = arena_mapbitsp_read(mapbitsp);
@@ -824,7 +858,7 @@ arena_mapbits_large_binind_set(arena_chunk_t *chunk, size_t pageind,
 
 JEMALLOC_ALWAYS_INLINE void
 arena_mapbits_small_set(arena_chunk_t *chunk, size_t pageind, size_t runind,
-    index_t binind, size_t flags)
+    szind_t binind, size_t flags)
 {
 	size_t *mapbitsp = arena_mapbitsp_get(chunk, pageind);
 
@@ -901,10 +935,10 @@ arena_prof_accum(arena_t *arena, uint64_t accumbytes)
 	}
 }
 
-JEMALLOC_ALWAYS_INLINE index_t
+JEMALLOC_ALWAYS_INLINE szind_t
 arena_ptr_small_binind_get(const void *ptr, size_t mapbits)
 {
-	index_t binind;
+	szind_t binind;
 
 	binind = (mapbits & CHUNK_MAP_BININD_MASK) >> CHUNK_MAP_BININD_SHIFT;
 
@@ -916,7 +950,7 @@ arena_ptr_small_binind_get(const void *ptr, size_t mapbits)
 		size_t rpages_ind;
 		arena_run_t *run;
 		arena_bin_t *bin;
-		index_t run_binind, actual_binind;
+		szind_t run_binind, actual_binind;
 		arena_bin_info_t *bin_info;
 		arena_chunk_map_misc_t *miscelm;
 		void *rpages;
@@ -950,10 +984,10 @@ arena_ptr_small_binind_get(const void *ptr, size_t mapbits)
 #  endif /* JEMALLOC_ARENA_INLINE_A */
 
 #  ifdef JEMALLOC_ARENA_INLINE_B
-JEMALLOC_INLINE index_t
+JEMALLOC_INLINE szind_t
 arena_bin_index(arena_t *arena, arena_bin_t *bin)
 {
-	index_t binind = bin - arena->bins;
+	szind_t binind = bin - arena->bins;
 	assert(binind < NBINS);
 	return (binind);
 }
@@ -1060,7 +1094,7 @@ arena_prof_tctx_get(const void *ptr)
 }
 
 JEMALLOC_INLINE void
-arena_prof_tctx_set(const void *ptr, prof_tctx_t *tctx)
+arena_prof_tctx_set(const void *ptr, size_t usize, prof_tctx_t *tctx)
 {
 	arena_chunk_t *chunk;
 
@@ -1070,17 +1104,59 @@ arena_prof_tctx_set(const void *ptr, prof_tctx_t *tctx)
 	chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(ptr);
 	if (likely(chunk != ptr)) {
 		size_t pageind = ((uintptr_t)ptr - (uintptr_t)chunk) >> LG_PAGE;
+
 		assert(arena_mapbits_allocated_get(chunk, pageind) != 0);
 
-		if (unlikely(arena_mapbits_large_get(chunk, pageind) != 0)) {
-			arena_chunk_map_misc_t *elm = arena_miscelm_get(chunk,
-			    pageind);
+		if (unlikely(usize > SMALL_MAXCLASS || (uintptr_t)tctx >
+		    (uintptr_t)1U)) {
+			arena_chunk_map_misc_t *elm;
+
+			assert(arena_mapbits_large_get(chunk, pageind) != 0);
+
+			elm = arena_miscelm_get(chunk, pageind);
 			atomic_write_p(&elm->prof_tctx_pun, tctx);
+		} else {
+			/*
+			 * tctx must always be initialized for large runs.
+			 * Assert that the surrounding conditional logic is
+			 * equivalent to checking whether ptr refers to a large
+			 * run.
+			 */
+			assert(arena_mapbits_large_get(chunk, pageind) == 0);
 		}
 	} else
 		huge_prof_tctx_set(ptr, tctx);
 }
 
+JEMALLOC_INLINE void
+arena_prof_tctx_reset(const void *ptr, size_t usize, const void *old_ptr,
+    prof_tctx_t *old_tctx)
+{
+
+	cassert(config_prof);
+	assert(ptr != NULL);
+
+	if (unlikely(usize > SMALL_MAXCLASS || (ptr == old_ptr &&
+	    (uintptr_t)old_tctx > (uintptr_t)1U))) {
+		arena_chunk_t *chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(ptr);
+		if (likely(chunk != ptr)) {
+			size_t pageind;
+			arena_chunk_map_misc_t *elm;
+
+			pageind = ((uintptr_t)ptr - (uintptr_t)chunk) >>
+			    LG_PAGE;
+			assert(arena_mapbits_allocated_get(chunk, pageind) !=
+			    0);
+			assert(arena_mapbits_large_get(chunk, pageind) != 0);
+
+			elm = arena_miscelm_get(chunk, pageind);
+			atomic_write_p(&elm->prof_tctx_pun,
+			    (prof_tctx_t *)(uintptr_t)1U);
+		} else
+			huge_prof_tctx_reset(ptr);
+	}
+}
+
 JEMALLOC_ALWAYS_INLINE void *
 arena_malloc(tsd_t *tsd, arena_t *arena, size_t size, bool zero,
     tcache_t *tcache)
@@ -1098,7 +1174,7 @@ arena_malloc(tsd_t *tsd, arena_t *arena, size_t size, bool zero,
 			    zero));
 		} else
 			return (arena_malloc_small(arena, size, zero));
-	} else if (likely(size <= arena_maxclass)) {
+	} else if (likely(size <= large_maxclass)) {
 		/*
 		 * Initialize tcache after checking size in order to avoid
 		 * infinite recursion during tcache initialization.
@@ -1131,7 +1207,7 @@ arena_salloc(const void *ptr, bool demote)
 	size_t ret;
 	arena_chunk_t *chunk;
 	size_t pageind;
-	index_t binind;
+	szind_t binind;
 
 	assert(ptr != NULL);
 
@@ -1190,7 +1266,7 @@ arena_dalloc(tsd_t *tsd, void *ptr, tcache_t *tcache)
 		if (likely((mapbits & CHUNK_MAP_LARGE) == 0)) {
 			/* Small allocation. */
 			if (likely(tcache != NULL)) {
-				index_t binind = arena_ptr_small_binind_get(ptr,
+				szind_t binind = arena_ptr_small_binind_get(ptr,
 				    mapbits);
 				tcache_dalloc_small(tsd, tcache, ptr, binind);
 			} else {
@@ -1242,7 +1318,7 @@ arena_sdalloc(tsd_t *tsd, void *ptr, size_t size, tcache_t *tcache)
 		if (likely(size <= SMALL_MAXCLASS)) {
 			/* Small allocation. */
 			if (likely(tcache != NULL)) {
-				index_t binind = size2index(size);
+				szind_t binind = size2index(size);
 				tcache_dalloc_small(tsd, tcache, ptr, binind);
 			} else {
 				size_t pageind = ((uintptr_t)ptr -
diff --git a/contrib/jemalloc/include/jemalloc/internal/huge.h b/contrib/jemalloc/include/jemalloc/internal/huge.h
index 8b6c6ce..ece7af9 100644
--- a/contrib/jemalloc/include/jemalloc/internal/huge.h
+++ b/contrib/jemalloc/include/jemalloc/internal/huge.h
@@ -13,11 +13,10 @@ void	*huge_malloc(tsd_t *tsd, arena_t *arena, size_t size, bool zero,
     tcache_t *tcache);
 void	*huge_palloc(tsd_t *tsd, arena_t *arena, size_t size, size_t alignment,
     bool zero, tcache_t *tcache);
-bool	huge_ralloc_no_move(void *ptr, size_t oldsize, size_t size,
-    size_t extra, bool zero);
+bool	huge_ralloc_no_move(void *ptr, size_t oldsize, size_t usize_min,
+    size_t usize_max, bool zero);
 void	*huge_ralloc(tsd_t *tsd, arena_t *arena, void *ptr, size_t oldsize,
-    size_t size, size_t extra, size_t alignment, bool zero,
-    tcache_t *tcache);
+    size_t usize, size_t alignment, bool zero, tcache_t *tcache);
 #ifdef JEMALLOC_JET
 typedef void (huge_dalloc_junk_t)(void *, size_t);
 extern huge_dalloc_junk_t *huge_dalloc_junk;
@@ -27,6 +26,7 @@ arena_t	*huge_aalloc(const void *ptr);
 size_t	huge_salloc(const void *ptr);
 prof_tctx_t	*huge_prof_tctx_get(const void *ptr);
 void	huge_prof_tctx_set(const void *ptr, prof_tctx_t *tctx);
+void	huge_prof_tctx_reset(const void *ptr);
 
 #endif /* JEMALLOC_H_EXTERNS */
 /******************************************************************************/
diff --git a/contrib/jemalloc/include/jemalloc/internal/jemalloc_internal.h b/contrib/jemalloc/include/jemalloc/internal/jemalloc_internal.h
index d9d8b5b..6c3cc18 100644
--- a/contrib/jemalloc/include/jemalloc/internal/jemalloc_internal.h
+++ b/contrib/jemalloc/include/jemalloc/internal/jemalloc_internal.h
@@ -181,7 +181,7 @@ static const bool config_cache_oblivious =
 #include "jemalloc/internal/jemalloc_internal_macros.h"
 
 /* Size class index type. */
-typedef unsigned index_t;
+typedef unsigned szind_t;
 
 /*
  * Flags bits:
@@ -229,7 +229,7 @@ typedef unsigned index_t;
 #  ifdef __alpha__
 #    define LG_QUANTUM		4
 #  endif
-#  ifdef __sparc64__
+#  if (defined(__sparc64__) || defined(__sparcv9))
 #    define LG_QUANTUM		4
 #  endif
 #  if (defined(__amd64__) || defined(__x86_64__) || defined(_M_X64))
@@ -508,12 +508,12 @@ void	jemalloc_postfork_child(void);
 #include "jemalloc/internal/huge.h"
 
 #ifndef JEMALLOC_ENABLE_INLINE
-index_t	size2index_compute(size_t size);
-index_t	size2index_lookup(size_t size);
-index_t	size2index(size_t size);
-size_t	index2size_compute(index_t index);
-size_t	index2size_lookup(index_t index);
-size_t	index2size(index_t index);
+szind_t	size2index_compute(size_t size);
+szind_t	size2index_lookup(size_t size);
+szind_t	size2index(size_t size);
+size_t	index2size_compute(szind_t index);
+size_t	index2size_lookup(szind_t index);
+size_t	index2size(szind_t index);
 size_t	s2u_compute(size_t size);
 size_t	s2u_lookup(size_t size);
 size_t	s2u(size_t size);
@@ -524,7 +524,7 @@ arena_t	*arena_get(tsd_t *tsd, unsigned ind, bool init_if_missing,
 #endif
 
 #if (defined(JEMALLOC_ENABLE_INLINE) || defined(JEMALLOC_C_))
-JEMALLOC_INLINE index_t
+JEMALLOC_INLINE szind_t
 size2index_compute(size_t size)
 {
 
@@ -555,7 +555,7 @@ size2index_compute(size_t size)
 	}
 }
 
-JEMALLOC_ALWAYS_INLINE index_t
+JEMALLOC_ALWAYS_INLINE szind_t
 size2index_lookup(size_t size)
 {
 
@@ -568,7 +568,7 @@ size2index_lookup(size_t size)
 	}
 }
 
-JEMALLOC_ALWAYS_INLINE index_t
+JEMALLOC_ALWAYS_INLINE szind_t
 size2index(size_t size)
 {
 
@@ -579,7 +579,7 @@ size2index(size_t size)
 }
 
 JEMALLOC_INLINE size_t
-index2size_compute(index_t index)
+index2size_compute(szind_t index)
 {
 
 #if (NTBINS > 0)
@@ -606,7 +606,7 @@ index2size_compute(index_t index)
 }
 
 JEMALLOC_ALWAYS_INLINE size_t
-index2size_lookup(index_t index)
+index2size_lookup(szind_t index)
 {
 	size_t ret = (size_t)index2size_tab[index];
 	assert(ret == index2size_compute(index));
@@ -614,7 +614,7 @@ index2size_lookup(index_t index)
 }
 
 JEMALLOC_ALWAYS_INLINE size_t
-index2size(index_t index)
+index2size(szind_t index)
 {
 
 	assert(index < NSIZES);
@@ -702,7 +702,7 @@ sa2u(size_t size, size_t alignment)
 	}
 
 	/* Try for a large size class. */
-	if (likely(size <= arena_maxclass) && likely(alignment < chunksize)) {
+	if (likely(size <= large_maxclass) && likely(alignment < chunksize)) {
 		/*
 		 * We can't achieve subpage alignment, so round up alignment
 		 * to the minimum that can actually be supported.
@@ -973,7 +973,7 @@ u2rz(size_t usize)
 	size_t ret;
 
 	if (usize <= SMALL_MAXCLASS) {
-		index_t binind = size2index(usize);
+		szind_t binind = size2index(usize);
 		ret = arena_bin_info[binind].redzone_size;
 	} else
 		ret = 0;
@@ -1093,7 +1093,7 @@ iralloct(tsd_t *tsd, void *ptr, size_t oldsize, size_t size, size_t alignment,
 		    zero, tcache, arena));
 	}
 
-	return (arena_ralloc(tsd, arena, ptr, oldsize, size, 0, alignment, zero,
+	return (arena_ralloc(tsd, arena, ptr, oldsize, size, alignment, zero,
 	    tcache));
 }
 
diff --git a/contrib/jemalloc/include/jemalloc/internal/private_namespace.h b/contrib/jemalloc/include/jemalloc/internal/private_namespace.h
index 10fab65..ffc4ea2 100644
--- a/contrib/jemalloc/include/jemalloc/internal/private_namespace.h
+++ b/contrib/jemalloc/include/jemalloc/internal/private_namespace.h
@@ -50,13 +50,14 @@
 #define	arena_mapbitsp_get JEMALLOC_N(arena_mapbitsp_get)
 #define	arena_mapbitsp_read JEMALLOC_N(arena_mapbitsp_read)
 #define	arena_mapbitsp_write JEMALLOC_N(arena_mapbitsp_write)
+#define	arena_mapbits_size_decode JEMALLOC_N(arena_mapbits_size_decode)
+#define	arena_mapbits_size_encode JEMALLOC_N(arena_mapbits_size_encode)
 #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_maxclass JEMALLOC_N(arena_maxclass)
 #define	arena_maxrun JEMALLOC_N(arena_maxrun)
 #define	arena_maybe_purge JEMALLOC_N(arena_maybe_purge)
 #define	arena_metadata_allocated_add JEMALLOC_N(arena_metadata_allocated_add)
@@ -79,6 +80,7 @@
 #define	arena_prof_accum_locked JEMALLOC_N(arena_prof_accum_locked)
 #define	arena_prof_promoted JEMALLOC_N(arena_prof_promoted)
 #define	arena_prof_tctx_get JEMALLOC_N(arena_prof_tctx_get)
+#define	arena_prof_tctx_reset JEMALLOC_N(arena_prof_tctx_reset)
 #define	arena_prof_tctx_set JEMALLOC_N(arena_prof_tctx_set)
 #define	arena_ptr_small_binind_get JEMALLOC_N(arena_ptr_small_binind_get)
 #define	arena_purge_all JEMALLOC_N(arena_purge_all)
@@ -249,6 +251,7 @@
 #define	huge_malloc JEMALLOC_N(huge_malloc)
 #define	huge_palloc JEMALLOC_N(huge_palloc)
 #define	huge_prof_tctx_get JEMALLOC_N(huge_prof_tctx_get)
+#define	huge_prof_tctx_reset JEMALLOC_N(huge_prof_tctx_reset)
 #define	huge_prof_tctx_set JEMALLOC_N(huge_prof_tctx_set)
 #define	huge_ralloc JEMALLOC_N(huge_ralloc)
 #define	huge_ralloc_no_move JEMALLOC_N(huge_ralloc_no_move)
@@ -282,6 +285,7 @@
 #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	large_maxclass JEMALLOC_N(large_maxclass)
 #define	lg_floor JEMALLOC_N(lg_floor)
 #define	malloc_cprintf JEMALLOC_N(malloc_cprintf)
 #define	malloc_mutex_init JEMALLOC_N(malloc_mutex_init)
@@ -376,6 +380,7 @@
 #define	prof_sample_accum_update JEMALLOC_N(prof_sample_accum_update)
 #define	prof_sample_threshold_update JEMALLOC_N(prof_sample_threshold_update)
 #define	prof_tctx_get JEMALLOC_N(prof_tctx_get)
+#define	prof_tctx_reset JEMALLOC_N(prof_tctx_reset)
 #define	prof_tctx_set JEMALLOC_N(prof_tctx_set)
 #define	prof_tdata_cleanup JEMALLOC_N(prof_tdata_cleanup)
 #define	prof_tdata_get JEMALLOC_N(prof_tdata_get)
diff --git a/contrib/jemalloc/include/jemalloc/internal/prof.h b/contrib/jemalloc/include/jemalloc/internal/prof.h
index 2e22711..e5198c3 100644
--- a/contrib/jemalloc/include/jemalloc/internal/prof.h
+++ b/contrib/jemalloc/include/jemalloc/internal/prof.h
@@ -90,10 +90,11 @@ struct prof_tctx_s {
 	prof_tdata_t		*tdata;
 
 	/*
-	 * Copy of tdata->thr_uid, necessary because tdata may be defunct during
-	 * teardown.
+	 * Copy of tdata->thr_{uid,discrim}, necessary because tdata may be
+	 * defunct during teardown.
 	 */
 	uint64_t		thr_uid;
+	uint64_t		thr_discrim;
 
 	/* Profiling counters, protected by tdata->lock. */
 	prof_cnt_t		cnts;
@@ -330,14 +331,18 @@ bool	prof_gdump_get_unlocked(void);
 prof_tdata_t	*prof_tdata_get(tsd_t *tsd, bool create);
 bool	prof_sample_accum_update(tsd_t *tsd, size_t usize, bool commit,
     prof_tdata_t **tdata_out);
-prof_tctx_t	*prof_alloc_prep(tsd_t *tsd, size_t usize, bool update);
+prof_tctx_t	*prof_alloc_prep(tsd_t *tsd, size_t usize, bool prof_active,
+    bool update);
 prof_tctx_t	*prof_tctx_get(const void *ptr);
-void	prof_tctx_set(const void *ptr, prof_tctx_t *tctx);
+void	prof_tctx_set(const void *ptr, size_t usize, prof_tctx_t *tctx);
+void	prof_tctx_reset(const void *ptr, size_t usize, const void *old_ptr,
+    prof_tctx_t *tctx);
 void	prof_malloc_sample_object(const void *ptr, size_t usize,
     prof_tctx_t *tctx);
 void	prof_malloc(const void *ptr, size_t usize, prof_tctx_t *tctx);
 void	prof_realloc(tsd_t *tsd, const void *ptr, size_t usize,
-    prof_tctx_t *tctx, bool updated, size_t old_usize, prof_tctx_t *old_tctx);
+    prof_tctx_t *tctx, bool prof_active, bool updated, const void *old_ptr,
+    size_t old_usize, prof_tctx_t *old_tctx);
 void	prof_free(tsd_t *tsd, const void *ptr, size_t usize);
 #endif
 
@@ -402,13 +407,24 @@ prof_tctx_get(const void *ptr)
 }
 
 JEMALLOC_ALWAYS_INLINE void
-prof_tctx_set(const void *ptr, prof_tctx_t *tctx)
+prof_tctx_set(const void *ptr, size_t usize, prof_tctx_t *tctx)
 {
 
 	cassert(config_prof);
 	assert(ptr != NULL);
 
-	arena_prof_tctx_set(ptr, tctx);
+	arena_prof_tctx_set(ptr, usize, tctx);
+}
+
+JEMALLOC_ALWAYS_INLINE void
+prof_tctx_reset(const void *ptr, size_t usize, const void *old_ptr,
+    prof_tctx_t *old_tctx)
+{
+
+	cassert(config_prof);
+	assert(ptr != NULL);
+
+	arena_prof_tctx_reset(ptr, usize, old_ptr, old_tctx);
 }
 
 JEMALLOC_ALWAYS_INLINE bool
@@ -442,7 +458,7 @@ prof_sample_accum_update(tsd_t *tsd, size_t usize, bool update,
 }
 
 JEMALLOC_ALWAYS_INLINE prof_tctx_t *
-prof_alloc_prep(tsd_t *tsd, size_t usize, bool update)
+prof_alloc_prep(tsd_t *tsd, size_t usize, bool prof_active, bool update)
 {
 	prof_tctx_t *ret;
 	prof_tdata_t *tdata;
@@ -450,8 +466,8 @@ prof_alloc_prep(tsd_t *tsd, size_t usize, bool update)
 
 	assert(usize == s2u(usize));
 
-	if (!prof_active_get_unlocked() || likely(prof_sample_accum_update(tsd,
-	    usize, update, &tdata)))
+	if (!prof_active || likely(prof_sample_accum_update(tsd, usize, update,
+	    &tdata)))
 		ret = (prof_tctx_t *)(uintptr_t)1U;
 	else {
 		bt_init(&bt, tdata->vec);
@@ -473,22 +489,24 @@ prof_malloc(const void *ptr, size_t usize, prof_tctx_t *tctx)
 	if (unlikely((uintptr_t)tctx > (uintptr_t)1U))
 		prof_malloc_sample_object(ptr, usize, tctx);
 	else
-		prof_tctx_set(ptr, (prof_tctx_t *)(uintptr_t)1U);
+		prof_tctx_set(ptr, usize, (prof_tctx_t *)(uintptr_t)1U);
 }
 
 JEMALLOC_ALWAYS_INLINE void
 prof_realloc(tsd_t *tsd, const void *ptr, size_t usize, prof_tctx_t *tctx,
-    bool updated, size_t old_usize, prof_tctx_t *old_tctx)
+    bool prof_active, bool updated, const void *old_ptr, size_t old_usize,
+    prof_tctx_t *old_tctx)
 {
+	bool sampled, old_sampled;
 
 	cassert(config_prof);
 	assert(ptr != NULL || (uintptr_t)tctx <= (uintptr_t)1U);
 
-	if (!updated && ptr != NULL) {
+	if (prof_active && !updated && ptr != NULL) {
 		assert(usize == isalloc(ptr, true));
 		if (prof_sample_accum_update(tsd, usize, true, NULL)) {
 			/*
-			 * Don't sample.  The usize passed to PROF_ALLOC_PREP()
+			 * 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 usize was insufficient to cross the
@@ -498,12 +516,16 @@ prof_realloc(tsd_t *tsd, const void *ptr, size_t usize, prof_tctx_t *tctx,
 		}
 	}
 
-	if (unlikely((uintptr_t)old_tctx > (uintptr_t)1U))
-		prof_free_sampled_object(tsd, old_usize, old_tctx);
-	if (unlikely((uintptr_t)tctx > (uintptr_t)1U))
+	sampled = ((uintptr_t)tctx > (uintptr_t)1U);
+	old_sampled = ((uintptr_t)old_tctx > (uintptr_t)1U);
+
+	if (unlikely(sampled))
 		prof_malloc_sample_object(ptr, usize, tctx);
 	else
-		prof_tctx_set(ptr, (prof_tctx_t *)(uintptr_t)1U);
+		prof_tctx_reset(ptr, usize, old_ptr, old_tctx);
+
+	if (unlikely(old_sampled))
+		prof_free_sampled_object(tsd, old_usize, old_tctx);
 }
 
 JEMALLOC_ALWAYS_INLINE void
diff --git a/contrib/jemalloc/include/jemalloc/internal/size_classes.h b/contrib/jemalloc/include/jemalloc/internal/size_classes.h
index eed6fbd..4ca58a1 100644
--- a/contrib/jemalloc/include/jemalloc/internal/size_classes.h
+++ b/contrib/jemalloc/include/jemalloc/internal/size_classes.h
@@ -25,6 +25,7 @@
  *   LOOKUP_MAXCLASS: Maximum size class included in lookup table.
  *   SMALL_MAXCLASS: Maximum small size class.
  *   LG_LARGE_MINCLASS: Lg of minimum large size class.
+ *   HUGE_MAXCLASS: Maximum (huge) size class.
  */
 
 #define	LG_SIZE_CLASS_GROUP	2
@@ -180,6 +181,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 13) + (((size_t)3) << 11))
 #define	LG_LARGE_MINCLASS	14
+#define	HUGE_MAXCLASS		((((size_t)1) << 31) + (((size_t)3) << 29))
 #endif
 
 #if (LG_SIZEOF_PTR == 2 && LG_TINY_MIN == 3 && LG_QUANTUM == 3 && LG_PAGE == 13)
@@ -333,6 +335,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 14) + (((size_t)3) << 12))
 #define	LG_LARGE_MINCLASS	15
+#define	HUGE_MAXCLASS		((((size_t)1) << 31) + (((size_t)3) << 29))
 #endif
 
 #if (LG_SIZEOF_PTR == 2 && LG_TINY_MIN == 3 && LG_QUANTUM == 3 && LG_PAGE == 14)
@@ -486,6 +489,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 15) + (((size_t)3) << 13))
 #define	LG_LARGE_MINCLASS	16
+#define	HUGE_MAXCLASS		((((size_t)1) << 31) + (((size_t)3) << 29))
 #endif
 
 #if (LG_SIZEOF_PTR == 2 && LG_TINY_MIN == 3 && LG_QUANTUM == 3 && LG_PAGE == 16)
@@ -639,6 +643,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 17) + (((size_t)3) << 15))
 #define	LG_LARGE_MINCLASS	18
+#define	HUGE_MAXCLASS		((((size_t)1) << 31) + (((size_t)3) << 29))
 #endif
 
 #if (LG_SIZEOF_PTR == 2 && LG_TINY_MIN == 3 && LG_QUANTUM == 4 && LG_PAGE == 12)
@@ -789,6 +794,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 13) + (((size_t)3) << 11))
 #define	LG_LARGE_MINCLASS	14
+#define	HUGE_MAXCLASS		((((size_t)1) << 31) + (((size_t)3) << 29))
 #endif
 
 #if (LG_SIZEOF_PTR == 2 && LG_TINY_MIN == 3 && LG_QUANTUM == 4 && LG_PAGE == 13)
@@ -939,6 +945,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 14) + (((size_t)3) << 12))
 #define	LG_LARGE_MINCLASS	15
+#define	HUGE_MAXCLASS		((((size_t)1) << 31) + (((size_t)3) << 29))
 #endif
 
 #if (LG_SIZEOF_PTR == 2 && LG_TINY_MIN == 3 && LG_QUANTUM == 4 && LG_PAGE == 14)
@@ -1089,6 +1096,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 15) + (((size_t)3) << 13))
 #define	LG_LARGE_MINCLASS	16
+#define	HUGE_MAXCLASS		((((size_t)1) << 31) + (((size_t)3) << 29))
 #endif
 
 #if (LG_SIZEOF_PTR == 2 && LG_TINY_MIN == 3 && LG_QUANTUM == 4 && LG_PAGE == 16)
@@ -1239,6 +1247,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 17) + (((size_t)3) << 15))
 #define	LG_LARGE_MINCLASS	18
+#define	HUGE_MAXCLASS		((((size_t)1) << 31) + (((size_t)3) << 29))
 #endif
 
 #if (LG_SIZEOF_PTR == 2 && LG_TINY_MIN == 4 && LG_QUANTUM == 4 && LG_PAGE == 12)
@@ -1387,6 +1396,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 13) + (((size_t)3) << 11))
 #define	LG_LARGE_MINCLASS	14
+#define	HUGE_MAXCLASS		((((size_t)1) << 31) + (((size_t)3) << 29))
 #endif
 
 #if (LG_SIZEOF_PTR == 2 && LG_TINY_MIN == 4 && LG_QUANTUM == 4 && LG_PAGE == 13)
@@ -1535,6 +1545,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 14) + (((size_t)3) << 12))
 #define	LG_LARGE_MINCLASS	15
+#define	HUGE_MAXCLASS		((((size_t)1) << 31) + (((size_t)3) << 29))
 #endif
 
 #if (LG_SIZEOF_PTR == 2 && LG_TINY_MIN == 4 && LG_QUANTUM == 4 && LG_PAGE == 14)
@@ -1683,6 +1694,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 15) + (((size_t)3) << 13))
 #define	LG_LARGE_MINCLASS	16
+#define	HUGE_MAXCLASS		((((size_t)1) << 31) + (((size_t)3) << 29))
 #endif
 
 #if (LG_SIZEOF_PTR == 2 && LG_TINY_MIN == 4 && LG_QUANTUM == 4 && LG_PAGE == 16)
@@ -1831,6 +1843,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 17) + (((size_t)3) << 15))
 #define	LG_LARGE_MINCLASS	18
+#define	HUGE_MAXCLASS		((((size_t)1) << 31) + (((size_t)3) << 29))
 #endif
 
 #if (LG_SIZEOF_PTR == 3 && LG_TINY_MIN == 3 && LG_QUANTUM == 3 && LG_PAGE == 12)
@@ -2144,6 +2157,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 13) + (((size_t)3) << 11))
 #define	LG_LARGE_MINCLASS	14
+#define	HUGE_MAXCLASS		((((size_t)1) << 63) + (((size_t)3) << 61))
 #endif
 
 #if (LG_SIZEOF_PTR == 3 && LG_TINY_MIN == 3 && LG_QUANTUM == 3 && LG_PAGE == 13)
@@ -2457,6 +2471,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 14) + (((size_t)3) << 12))
 #define	LG_LARGE_MINCLASS	15
+#define	HUGE_MAXCLASS		((((size_t)1) << 63) + (((size_t)3) << 61))
 #endif
 
 #if (LG_SIZEOF_PTR == 3 && LG_TINY_MIN == 3 && LG_QUANTUM == 3 && LG_PAGE == 14)
@@ -2770,6 +2785,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 15) + (((size_t)3) << 13))
 #define	LG_LARGE_MINCLASS	16
+#define	HUGE_MAXCLASS		((((size_t)1) << 63) + (((size_t)3) << 61))
 #endif
 
 #if (LG_SIZEOF_PTR == 3 && LG_TINY_MIN == 3 && LG_QUANTUM == 3 && LG_PAGE == 16)
@@ -3083,6 +3099,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 17) + (((size_t)3) << 15))
 #define	LG_LARGE_MINCLASS	18
+#define	HUGE_MAXCLASS		((((size_t)1) << 63) + (((size_t)3) << 61))
 #endif
 
 #if (LG_SIZEOF_PTR == 3 && LG_TINY_MIN == 3 && LG_QUANTUM == 4 && LG_PAGE == 12)
@@ -3393,6 +3410,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 13) + (((size_t)3) << 11))
 #define	LG_LARGE_MINCLASS	14
+#define	HUGE_MAXCLASS		((((size_t)1) << 63) + (((size_t)3) << 61))
 #endif
 
 #if (LG_SIZEOF_PTR == 3 && LG_TINY_MIN == 3 && LG_QUANTUM == 4 && LG_PAGE == 13)
@@ -3703,6 +3721,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 14) + (((size_t)3) << 12))
 #define	LG_LARGE_MINCLASS	15
+#define	HUGE_MAXCLASS		((((size_t)1) << 63) + (((size_t)3) << 61))
 #endif
 
 #if (LG_SIZEOF_PTR == 3 && LG_TINY_MIN == 3 && LG_QUANTUM == 4 && LG_PAGE == 14)
@@ -4013,6 +4032,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 15) + (((size_t)3) << 13))
 #define	LG_LARGE_MINCLASS	16
+#define	HUGE_MAXCLASS		((((size_t)1) << 63) + (((size_t)3) << 61))
 #endif
 
 #if (LG_SIZEOF_PTR == 3 && LG_TINY_MIN == 3 && LG_QUANTUM == 4 && LG_PAGE == 16)
@@ -4323,6 +4343,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 17) + (((size_t)3) << 15))
 #define	LG_LARGE_MINCLASS	18
+#define	HUGE_MAXCLASS		((((size_t)1) << 63) + (((size_t)3) << 61))
 #endif
 
 #if (LG_SIZEOF_PTR == 3 && LG_TINY_MIN == 4 && LG_QUANTUM == 4 && LG_PAGE == 12)
@@ -4631,6 +4652,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 13) + (((size_t)3) << 11))
 #define	LG_LARGE_MINCLASS	14
+#define	HUGE_MAXCLASS		((((size_t)1) << 63) + (((size_t)3) << 61))
 #endif
 
 #if (LG_SIZEOF_PTR == 3 && LG_TINY_MIN == 4 && LG_QUANTUM == 4 && LG_PAGE == 13)
@@ -4939,6 +4961,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 14) + (((size_t)3) << 12))
 #define	LG_LARGE_MINCLASS	15
+#define	HUGE_MAXCLASS		((((size_t)1) << 63) + (((size_t)3) << 61))
 #endif
 
 #if (LG_SIZEOF_PTR == 3 && LG_TINY_MIN == 4 && LG_QUANTUM == 4 && LG_PAGE == 14)
@@ -5247,6 +5270,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 15) + (((size_t)3) << 13))
 #define	LG_LARGE_MINCLASS	16
+#define	HUGE_MAXCLASS		((((size_t)1) << 63) + (((size_t)3) << 61))
 #endif
 
 #if (LG_SIZEOF_PTR == 3 && LG_TINY_MIN == 4 && LG_QUANTUM == 4 && LG_PAGE == 16)
@@ -5555,6 +5579,7 @@
 #define	LOOKUP_MAXCLASS		((((size_t)1) << 11) + (((size_t)4) << 9))
 #define	SMALL_MAXCLASS		((((size_t)1) << 17) + (((size_t)3) << 15))
 #define	LG_LARGE_MINCLASS	18
+#define	HUGE_MAXCLASS		((((size_t)1) << 63) + (((size_t)3) << 61))
 #endif
 
 #ifndef SIZE_CLASSES_DEFINED
diff --git a/contrib/jemalloc/include/jemalloc/internal/tcache.h b/contrib/jemalloc/include/jemalloc/internal/tcache.h
index 493f457..5079cd2 100644
--- a/contrib/jemalloc/include/jemalloc/internal/tcache.h
+++ b/contrib/jemalloc/include/jemalloc/internal/tcache.h
@@ -77,7 +77,7 @@ struct tcache_s {
 	ql_elm(tcache_t) link;		/* Used for aggregating stats. */
 	uint64_t	prof_accumbytes;/* Cleared after arena_prof_accum(). */
 	unsigned	ev_cnt;		/* Event count since incremental GC. */
-	index_t		next_gc_bin;	/* Next bin to GC. */
+	szind_t		next_gc_bin;	/* Next bin to GC. */
 	tcache_bin_t	tbins[1];	/* Dynamically sized. */
 	/*
 	 * The pointer stacks associated with tbins follow as a contiguous
@@ -126,10 +126,10 @@ extern tcaches_t	*tcaches;
 size_t	tcache_salloc(const void *ptr);
 void	tcache_event_hard(tsd_t *tsd, tcache_t *tcache);
 void	*tcache_alloc_small_hard(tsd_t *tsd, arena_t *arena, tcache_t *tcache,
-    tcache_bin_t *tbin, index_t binind);
+    tcache_bin_t *tbin, szind_t binind);
 void	tcache_bin_flush_small(tsd_t *tsd, tcache_t *tcache, tcache_bin_t *tbin,
-    index_t binind, unsigned rem);
-void	tcache_bin_flush_large(tsd_t *tsd, tcache_bin_t *tbin, index_t binind,
+    szind_t binind, unsigned rem);
+void	tcache_bin_flush_large(tsd_t *tsd, tcache_bin_t *tbin, szind_t binind,
     unsigned rem, tcache_t *tcache);
 void	tcache_arena_associate(tcache_t *tcache, arena_t *arena);
 void	tcache_arena_reassociate(tcache_t *tcache, arena_t *oldarena,
@@ -161,7 +161,7 @@ void	*tcache_alloc_small(tsd_t *tsd, arena_t *arena, tcache_t *tcache,
 void	*tcache_alloc_large(tsd_t *tsd, arena_t *arena, tcache_t *tcache,
     size_t size, bool zero);
 void	tcache_dalloc_small(tsd_t *tsd, tcache_t *tcache, void *ptr,
-    index_t binind);
+    szind_t binind);
 void	tcache_dalloc_large(tsd_t *tsd, tcache_t *tcache, void *ptr,
     size_t size);
 tcache_t	*tcaches_get(tsd_t *tsd, unsigned ind);
@@ -267,7 +267,7 @@ tcache_alloc_small(tsd_t *tsd, arena_t *arena, tcache_t *tcache, size_t size,
     bool zero)
 {
 	void *ret;
-	index_t binind;
+	szind_t binind;
 	size_t usize;
 	tcache_bin_t *tbin;
 
@@ -312,7 +312,7 @@ tcache_alloc_large(tsd_t *tsd, arena_t *arena, tcache_t *tcache, size_t size,
     bool zero)
 {
 	void *ret;
-	index_t binind;
+	szind_t binind;
 	size_t usize;
 	tcache_bin_t *tbin;
 
@@ -360,7 +360,7 @@ tcache_alloc_large(tsd_t *tsd, arena_t *arena, tcache_t *tcache, size_t size,
 }
 
 JEMALLOC_ALWAYS_INLINE void
-tcache_dalloc_small(tsd_t *tsd, tcache_t *tcache, void *ptr, index_t binind)
+tcache_dalloc_small(tsd_t *tsd, tcache_t *tcache, void *ptr, szind_t binind)
 {
 	tcache_bin_t *tbin;
 	tcache_bin_info_t *tbin_info;
@@ -386,7 +386,7 @@ tcache_dalloc_small(tsd_t *tsd, tcache_t *tcache, void *ptr, index_t binind)
 JEMALLOC_ALWAYS_INLINE void
 tcache_dalloc_large(tsd_t *tsd, tcache_t *tcache, void *ptr, size_t size)
 {
-	index_t binind;
+	szind_t binind;
 	tcache_bin_t *tbin;
 	tcache_bin_info_t *tbin_info;
 
diff --git a/contrib/jemalloc/include/jemalloc/internal/tsd.h b/contrib/jemalloc/include/jemalloc/internal/tsd.h
index 62a887e..eed7aa0 100644
--- a/contrib/jemalloc/include/jemalloc/internal/tsd.h
+++ b/contrib/jemalloc/include/jemalloc/internal/tsd.h
@@ -190,7 +190,7 @@ a_name##tsd_boot0(void)							\
 	return (false);							\
 }									\
 a_attr void								\
-a_name##tsd_boot1()							\
+a_name##tsd_boot1(void)							\
 {									\
 									\
 	/* Do nothing. */						\
@@ -235,7 +235,7 @@ a_name##tsd_boot0(void)							\
 	return (false);							\
 }									\
 a_attr void								\
-a_name##tsd_boot1()							\
+a_name##tsd_boot1(void)							\
 {									\
 									\
 	/* Do nothing. */						\
@@ -345,7 +345,7 @@ a_name##tsd_boot0(void)							\
 	return (false);							\
 }									\
 a_attr void								\
-a_name##tsd_boot1()							\
+a_name##tsd_boot1(void)							\
 {									\
 	a_name##tsd_wrapper_t *wrapper;					\
 	wrapper = (a_name##tsd_wrapper_t *)				\
@@ -467,7 +467,7 @@ a_name##tsd_boot0(void)							\
 	return (false);							\
 }									\
 a_attr void								\
-a_name##tsd_boot1()							\
+a_name##tsd_boot1(void)							\
 {									\
 	a_name##tsd_wrapper_t *wrapper;					\
 	wrapper = (a_name##tsd_wrapper_t *)				\
diff --git a/contrib/jemalloc/include/jemalloc/jemalloc.h b/contrib/jemalloc/include/jemalloc/jemalloc.h
index cd8f7da..95a3a51 100644
--- a/contrib/jemalloc/include/jemalloc/jemalloc.h
+++ b/contrib/jemalloc/include/jemalloc/jemalloc.h
@@ -79,12 +79,12 @@ extern "C" {
 #include <limits.h>
 #include <strings.h>
 
-#define	JEMALLOC_VERSION "4.0.0-0-g6e98caf8f064482b9ab292ef3638dea67420bbc2"
+#define	JEMALLOC_VERSION "4.0.2-0-g486d249fb4715fd3de679b6c2a04f7e657883111"
 #define	JEMALLOC_VERSION_MAJOR 4
 #define	JEMALLOC_VERSION_MINOR 0
-#define	JEMALLOC_VERSION_BUGFIX 0
+#define	JEMALLOC_VERSION_BUGFIX 2
 #define	JEMALLOC_VERSION_NREV 0
-#define	JEMALLOC_VERSION_GID "6e98caf8f064482b9ab292ef3638dea67420bbc2"
+#define	JEMALLOC_VERSION_GID "486d249fb4715fd3de679b6c2a04f7e657883111"
 
 #  define MALLOCX_LG_ALIGN(la)	(la)
 #  if LG_SIZEOF_PTR == 2
@@ -234,7 +234,7 @@ JEMALLOC_EXPORT size_t JEMALLOC_NOTHROW	je_malloc_usable_size(
 #ifdef JEMALLOC_OVERRIDE_MEMALIGN
 JEMALLOC_EXPORT JEMALLOC_ALLOCATOR JEMALLOC_RESTRICT_RETURN
     void JEMALLOC_NOTHROW	*je_memalign(size_t alignment, size_t size)
-    JEMALLOC_ATTR(malloc);
+    JEMALLOC_CXX_THROW JEMALLOC_ATTR(malloc);
 #endif
 
 #ifdef JEMALLOC_OVERRIDE_VALLOC
diff --git a/contrib/jemalloc/src/arena.c b/contrib/jemalloc/src/arena.c
index af48b39..7f4a6ca 100644
--- a/contrib/jemalloc/src/arena.c
+++ b/contrib/jemalloc/src/arena.c
@@ -11,7 +11,7 @@ arena_bin_info_t	arena_bin_info[NBINS];
 size_t		map_bias;
 size_t		map_misc_offset;
 size_t		arena_maxrun; /* Max run size for arenas. */
-size_t		arena_maxclass; /* Max size class for arenas. */
+size_t		large_maxclass; /* Max large size class. */
 static size_t	small_maxrun; /* Max run size used for small size classes. */
 static bool	*small_run_tab; /* Valid small run page multiples. */
 unsigned	nlclasses; /* Number of large size classes. */
@@ -39,7 +39,7 @@ JEMALLOC_INLINE_C arena_chunk_map_misc_t *
 arena_miscelm_key_create(size_t size)
 {
 
-	return ((arena_chunk_map_misc_t *)((size << CHUNK_MAP_SIZE_SHIFT) |
+	return ((arena_chunk_map_misc_t *)(arena_mapbits_size_encode(size) |
 	    CHUNK_MAP_KEY));
 }
 
@@ -58,8 +58,7 @@ arena_miscelm_key_size_get(const arena_chunk_map_misc_t *miscelm)
 
 	assert(arena_miscelm_is_key(miscelm));
 
-	return (((uintptr_t)miscelm & CHUNK_MAP_SIZE_MASK) >>
-	    CHUNK_MAP_SIZE_SHIFT);
+	return (arena_mapbits_size_decode((uintptr_t)miscelm));
 }
 
 JEMALLOC_INLINE_C size_t
@@ -73,7 +72,7 @@ arena_miscelm_size_get(arena_chunk_map_misc_t *miscelm)
 	chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(miscelm);
 	pageind = arena_miscelm_to_pageind(miscelm);
 	mapbits = arena_mapbits_get(chunk, pageind);
-	return ((mapbits & CHUNK_MAP_SIZE_MASK) >> CHUNK_MAP_SIZE_SHIFT);
+	return (arena_mapbits_size_decode(mapbits));
 }
 
 JEMALLOC_INLINE_C int
@@ -315,7 +314,7 @@ arena_run_reg_dalloc(arena_run_t *run, void *ptr)
 	arena_chunk_t *chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(run);
 	size_t pageind = ((uintptr_t)ptr - (uintptr_t)chunk) >> LG_PAGE;
 	size_t mapbits = arena_mapbits_get(chunk, pageind);
-	index_t binind = arena_ptr_small_binind_get(ptr, mapbits);
+	szind_t binind = arena_ptr_small_binind_get(ptr, mapbits);
 	arena_bin_info_t *bin_info = &arena_bin_info[binind];
 	unsigned regind = arena_run_regind(run, bin_info, ptr);
 
@@ -426,7 +425,7 @@ arena_run_split_large_helper(arena_t *arena, arena_run_t *run, size_t size,
 {
 	arena_chunk_t *chunk;
 	arena_chunk_map_misc_t *miscelm;
-	size_t flag_dirty, flag_decommitted, run_ind, need_pages, i;
+	size_t flag_dirty, flag_decommitted, run_ind, need_pages;
 	size_t flag_unzeroed_mask;
 
 	chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(run);
@@ -460,6 +459,7 @@ arena_run_split_large_helper(arena_t *arena, arena_run_t *run, size_t size,
 			 * The run is clean, so some pages may be zeroed (i.e.
 			 * never before touched).
 			 */
+			size_t i;
 			for (i = 0; i < need_pages; i++) {
 				if (arena_mapbits_unzeroed_get(chunk, run_ind+i)
 				    != 0)
@@ -508,7 +508,7 @@ arena_run_init_large(arena_t *arena, arena_run_t *run, size_t size, bool zero)
 
 static bool
 arena_run_split_small(arena_t *arena, arena_run_t *run, size_t size,
-    index_t binind)
+    szind_t binind)
 {
 	arena_chunk_t *chunk;
 	arena_chunk_map_misc_t *miscelm;
@@ -780,7 +780,7 @@ arena_chunk_dalloc(arena_t *arena, arena_chunk_t *chunk)
 static void
 arena_huge_malloc_stats_update(arena_t *arena, size_t usize)
 {
-	index_t index = size2index(usize) - nlclasses - NBINS;
+	szind_t index = size2index(usize) - nlclasses - NBINS;
 
 	cassert(config_stats);
 
@@ -793,7 +793,7 @@ arena_huge_malloc_stats_update(arena_t *arena, size_t usize)
 static void
 arena_huge_malloc_stats_update_undo(arena_t *arena, size_t usize)
 {
-	index_t index = size2index(usize) - nlclasses - NBINS;
+	szind_t index = size2index(usize) - nlclasses - NBINS;
 
 	cassert(config_stats);
 
@@ -806,7 +806,7 @@ arena_huge_malloc_stats_update_undo(arena_t *arena, size_t usize)
 static void
 arena_huge_dalloc_stats_update(arena_t *arena, size_t usize)
 {
-	index_t index = size2index(usize) - nlclasses - NBINS;
+	szind_t index = size2index(usize) - nlclasses - NBINS;
 
 	cassert(config_stats);
 
@@ -819,7 +819,7 @@ arena_huge_dalloc_stats_update(arena_t *arena, size_t usize)
 static void
 arena_huge_dalloc_stats_update_undo(arena_t *arena, size_t usize)
 {
-	index_t index = size2index(usize) - nlclasses - NBINS;
+	szind_t index = size2index(usize) - nlclasses - NBINS;
 
 	cassert(config_stats);
 
@@ -1125,7 +1125,7 @@ arena_run_alloc_large(arena_t *arena, size_t size, bool zero)
 }
 
 static arena_run_t *
-arena_run_alloc_small_helper(arena_t *arena, size_t size, index_t binind)
+arena_run_alloc_small_helper(arena_t *arena, size_t size, szind_t binind)
 {
 	arena_run_t *run = arena_run_first_best_fit(arena, size);
 	if (run != NULL) {
@@ -1136,7 +1136,7 @@ arena_run_alloc_small_helper(arena_t *arena, size_t size, index_t binind)
 }
 
 static arena_run_t *
-arena_run_alloc_small(arena_t *arena, size_t size, index_t binind)
+arena_run_alloc_small(arena_t *arena, size_t size, szind_t binind)
 {
 	arena_chunk_t *chunk;
 	arena_run_t *run;
@@ -1889,7 +1889,7 @@ static arena_run_t *
 arena_bin_nonfull_run_get(arena_t *arena, arena_bin_t *bin)
 {
 	arena_run_t *run;
-	index_t binind;
+	szind_t binind;
 	arena_bin_info_t *bin_info;
 
 	/* Look for a usable run. */
@@ -1939,8 +1939,7 @@ arena_bin_nonfull_run_get(arena_t *arena, arena_bin_t *bin)
 static void *
 arena_bin_malloc_hard(arena_t *arena, arena_bin_t *bin)
 {
-	void *ret;
-	index_t binind;
+	szind_t binind;
 	arena_bin_info_t *bin_info;
 	arena_run_t *run;
 
@@ -1953,6 +1952,7 @@ arena_bin_malloc_hard(arena_t *arena, arena_bin_t *bin)
 		 * Another thread updated runcur while this one ran without the
 		 * bin lock in arena_bin_nonfull_run_get().
 		 */
+		void *ret;
 		assert(bin->runcur->nfree > 0);
 		ret = arena_run_reg_alloc(bin->runcur, bin_info);
 		if (run != NULL) {
@@ -1986,13 +1986,11 @@ arena_bin_malloc_hard(arena_t *arena, arena_bin_t *bin)
 }
 
 void
-arena_tcache_fill_small(arena_t *arena, tcache_bin_t *tbin, index_t binind,
+arena_tcache_fill_small(arena_t *arena, tcache_bin_t *tbin, szind_t binind,
     uint64_t prof_accumbytes)
 {
 	unsigned i, nfill;
 	arena_bin_t *bin;
-	arena_run_t *run;
-	void *ptr;
 
 	assert(tbin->ncached == 0);
 
@@ -2002,6 +2000,8 @@ arena_tcache_fill_small(arena_t *arena, tcache_bin_t *tbin, index_t binind,
 	malloc_mutex_lock(&bin->lock);
 	for (i = 0, nfill = (tcache_bin_info[binind].ncached_max >>
 	    tbin->lg_fill_div); i < nfill; i++) {
+		arena_run_t *run;
+		void *ptr;
 		if ((run = bin->runcur) != NULL && run->nfree > 0)
 			ptr = arena_run_reg_alloc(run, &arena_bin_info[binind]);
 		else
@@ -2076,12 +2076,13 @@ arena_redzone_corruption_t *arena_redzone_corruption =
 static void
 arena_redzones_validate(void *ptr, arena_bin_info_t *bin_info, bool reset)
 {
-	size_t size = bin_info->reg_size;
-	size_t redzone_size = bin_info->redzone_size;
-	size_t i;
 	bool error = false;
 
 	if (opt_junk_alloc) {
+		size_t size = bin_info->reg_size;
+		size_t redzone_size = bin_info->redzone_size;
+		size_t i;
+
 		for (i = 1; i <= redzone_size; i++) {
 			uint8_t *byte = (uint8_t *)((uintptr_t)ptr - i);
 			if (*byte != 0xa5) {
@@ -2131,7 +2132,7 @@ arena_dalloc_junk_small_t *arena_dalloc_junk_small =
 void
 arena_quarantine_junk_small(void *ptr, size_t usize)
 {
-	index_t binind;
+	szind_t binind;
 	arena_bin_info_t *bin_info;
 	cassert(config_fill);
 	assert(opt_junk_free);
@@ -2149,7 +2150,7 @@ arena_malloc_small(arena_t *arena, size_t size, bool zero)
 	void *ret;
 	arena_bin_t *bin;
 	arena_run_t *run;
-	index_t binind;
+	szind_t binind;
 
 	binind = size2index(size);
 	assert(binind < NBINS);
@@ -2233,7 +2234,7 @@ arena_malloc_large(arena_t *arena, size_t size, bool zero)
 	ret = (void *)((uintptr_t)arena_miscelm_to_rpages(miscelm) +
 	    random_offset);
 	if (config_stats) {
-		index_t index = size2index(usize) - NBINS;
+		szind_t index = size2index(usize) - NBINS;
 
 		arena->stats.nmalloc_large++;
 		arena->stats.nrequests_large++;
@@ -2326,7 +2327,7 @@ arena_palloc_large(tsd_t *tsd, arena_t *arena, size_t usize, size_t alignment,
 	ret = arena_miscelm_to_rpages(miscelm);
 
 	if (config_stats) {
-		index_t index = size2index(usize) - NBINS;
+		szind_t index = size2index(usize) - NBINS;
 
 		arena->stats.nmalloc_large++;
 		arena->stats.nrequests_large++;
@@ -2356,7 +2357,7 @@ arena_palloc(tsd_t *tsd, arena_t *arena, size_t usize, size_t alignment,
 	    && (usize & PAGE_MASK) == 0))) {
 		/* Small; alignment doesn't require special run placement. */
 		ret = arena_malloc(tsd, arena, usize, zero, tcache);
-	} else if (usize <= arena_maxclass && alignment <= PAGE) {
+	} else if (usize <= large_maxclass && alignment <= PAGE) {
 		/*
 		 * Large; alignment doesn't require special run placement.
 		 * However, the cached pointer may be at a random offset from
@@ -2367,7 +2368,7 @@ arena_palloc(tsd_t *tsd, arena_t *arena, size_t usize, size_t alignment,
 		if (config_cache_oblivious)
 			ret = (void *)((uintptr_t)ret & ~PAGE_MASK);
 	} else {
-		if (likely(usize <= arena_maxclass)) {
+		if (likely(usize <= large_maxclass)) {
 			ret = arena_palloc_large(tsd, arena, usize, alignment,
 			    zero);
 		} else if (likely(alignment <= chunksize))
@@ -2385,7 +2386,7 @@ arena_prof_promoted(const void *ptr, size_t size)
 {
 	arena_chunk_t *chunk;
 	size_t pageind;
-	index_t binind;
+	szind_t binind;
 
 	cassert(config_prof);
 	assert(ptr != NULL);
@@ -2413,7 +2414,7 @@ arena_dissociate_bin_run(arena_chunk_t *chunk, arena_run_t *run,
 	if (run == bin->runcur)
 		bin->runcur = NULL;
 	else {
-		index_t binind = arena_bin_index(extent_node_arena_get(
+		szind_t binind = arena_bin_index(extent_node_arena_get(
 		    &chunk->node), bin);
 		arena_bin_info_t *bin_info = &arena_bin_info[binind];
 
@@ -2477,7 +2478,7 @@ arena_dalloc_bin_locked_impl(arena_t *arena, arena_chunk_t *chunk, void *ptr,
 	arena_run_t *run;
 	arena_bin_t *bin;
 	arena_bin_info_t *bin_info;
-	index_t binind;
+	szind_t binind;
 
 	pageind = ((uintptr_t)ptr - (uintptr_t)chunk) >> LG_PAGE;
 	rpages_ind = pageind - arena_mapbits_small_runind_get(chunk, pageind);
@@ -2559,7 +2560,7 @@ arena_dalloc_junk_large_t *arena_dalloc_junk_large =
     JEMALLOC_N(arena_dalloc_junk_large_impl);
 #endif
 
-void
+static void
 arena_dalloc_large_locked_impl(arena_t *arena, arena_chunk_t *chunk,
     void *ptr, bool junked)
 {
@@ -2574,7 +2575,7 @@ arena_dalloc_large_locked_impl(arena_t *arena, arena_chunk_t *chunk,
 		if (!junked)
 			arena_dalloc_junk_large(ptr, usize);
 		if (config_stats) {
-			index_t index = size2index(usize) - NBINS;
+			szind_t index = size2index(usize) - NBINS;
 
 			arena->stats.ndalloc_large++;
 			arena->stats.allocated_large -= usize;
@@ -2621,8 +2622,8 @@ arena_ralloc_large_shrink(arena_t *arena, arena_chunk_t *chunk, void *ptr,
 	arena_run_trim_tail(arena, chunk, run, oldsize + large_pad, size +
 	    large_pad, true);
 	if (config_stats) {
-		index_t oldindex = size2index(oldsize) - NBINS;
-		index_t index = size2index(size) - NBINS;
+		szind_t oldindex = size2index(oldsize) - NBINS;
+		szind_t index = size2index(size) - NBINS;
 
 		arena->stats.ndalloc_large++;
 		arena->stats.allocated_large -= oldsize;
@@ -2641,42 +2642,42 @@ arena_ralloc_large_shrink(arena_t *arena, arena_chunk_t *chunk, void *ptr,
 
 static bool
 arena_ralloc_large_grow(arena_t *arena, arena_chunk_t *chunk, void *ptr,
-    size_t oldsize, size_t size, size_t extra, bool zero)
+    size_t oldsize, size_t usize_min, size_t usize_max, bool zero)
 {
 	size_t pageind = ((uintptr_t)ptr - (uintptr_t)chunk) >> LG_PAGE;
 	size_t npages = (oldsize + large_pad) >> LG_PAGE;
 	size_t followsize;
-	size_t usize_min = s2u(size);
 
 	assert(oldsize == arena_mapbits_large_size_get(chunk, pageind) -
 	    large_pad);
 
 	/* Try to extend the run. */
-	assert(usize_min > oldsize);
 	malloc_mutex_lock(&arena->lock);
-	if (pageind+npages < chunk_npages &&
-	    arena_mapbits_allocated_get(chunk, pageind+npages) == 0 &&
-	    (followsize = arena_mapbits_unallocated_size_get(chunk,
-	    pageind+npages)) >= usize_min - oldsize) {
+	if (pageind+npages >= chunk_npages || arena_mapbits_allocated_get(chunk,
+	    pageind+npages) != 0)
+		goto label_fail;
+	followsize = arena_mapbits_unallocated_size_get(chunk, pageind+npages);
+	if (oldsize + followsize >= usize_min) {
 		/*
 		 * The next run is available and sufficiently large.  Split the
 		 * following run, then merge the first part with the existing
 		 * allocation.
 		 */
 		arena_run_t *run;
-		size_t flag_dirty, flag_unzeroed_mask, splitsize, usize;
+		size_t usize, splitsize, size, flag_dirty, flag_unzeroed_mask;
 
-		usize = s2u(size + extra);
+		usize = usize_max;
 		while (oldsize + followsize < usize)
 			usize = index2size(size2index(usize)-1);
 		assert(usize >= usize_min);
+		assert(usize >= oldsize);
 		splitsize = usize - oldsize;
+		if (splitsize == 0)
+			goto label_fail;
 
 		run = &arena_miscelm_get(chunk, pageind+npages)->run;
-		if (arena_run_split_large(arena, run, splitsize, zero)) {
-			malloc_mutex_unlock(&arena->lock);
-			return (true);
-		}
+		if (arena_run_split_large(arena, run, splitsize, zero))
+			goto label_fail;
 
 		size = oldsize + splitsize;
 		npages = (size + large_pad) >> LG_PAGE;
@@ -2700,8 +2701,8 @@ arena_ralloc_large_grow(arena_t *arena, arena_chunk_t *chunk, void *ptr,
 		    pageind+npages-1)));
 
 		if (config_stats) {
-			index_t oldindex = size2index(oldsize) - NBINS;
-			index_t index = size2index(size) - NBINS;
+			szind_t oldindex = size2index(oldsize) - NBINS;
+			szind_t index = size2index(size) - NBINS;
 
 			arena->stats.ndalloc_large++;
 			arena->stats.allocated_large -= oldsize;
@@ -2718,8 +2719,8 @@ arena_ralloc_large_grow(arena_t *arena, arena_chunk_t *chunk, void *ptr,
 		malloc_mutex_unlock(&arena->lock);
 		return (false);
 	}
+label_fail:
 	malloc_mutex_unlock(&arena->lock);
-
 	return (true);
 }
 
@@ -2748,98 +2749,107 @@ arena_ralloc_junk_large_t *arena_ralloc_junk_large =
  * always fail if growing an object, and the following run is already in use.
  */
 static bool
-arena_ralloc_large(void *ptr, size_t oldsize, size_t size, size_t extra,
-    bool zero)
+arena_ralloc_large(void *ptr, size_t oldsize, size_t usize_min,
+    size_t usize_max, bool zero)
 {
-	size_t usize;
-
-	/* Make sure extra can't cause size_t overflow. */
-	if (unlikely(extra >= arena_maxclass))
-		return (true);
+	arena_chunk_t *chunk;
+	arena_t *arena;
 
-	usize = s2u(size + extra);
-	if (usize == oldsize) {
-		/* Same size class. */
+	if (oldsize == usize_max) {
+		/* Current size class is compatible and maximal. */
 		return (false);
-	} else {
-		arena_chunk_t *chunk;
-		arena_t *arena;
+	}
 
-		chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(ptr);
-		arena = extent_node_arena_get(&chunk->node);
+	chunk = (arena_chunk_t *)CHUNK_ADDR2BASE(ptr);
+	arena = extent_node_arena_get(&chunk->node);
 
-		if (usize < oldsize) {
-			/* Fill before shrinking in order avoid a race. */
-			arena_ralloc_junk_large(ptr, oldsize, usize);
-			arena_ralloc_large_shrink(arena, chunk, ptr, oldsize,
-			    usize);
-			return (false);
-		} else {
-			bool ret = arena_ralloc_large_grow(arena, chunk, ptr,
-			    oldsize, size, extra, zero);
-			if (config_fill && !ret && !zero) {
-				if (unlikely(opt_junk_alloc)) {
-					memset((void *)((uintptr_t)ptr +
-					    oldsize), 0xa5, isalloc(ptr,
-					    config_prof) - oldsize);
-				} else if (unlikely(opt_zero)) {
-					memset((void *)((uintptr_t)ptr +
-					    oldsize), 0, isalloc(ptr,
-					    config_prof) - oldsize);
-				}
+	if (oldsize < usize_max) {
+		bool ret = arena_ralloc_large_grow(arena, chunk, ptr, oldsize,
+		    usize_min, usize_max, zero);
+		if (config_fill && !ret && !zero) {
+			if (unlikely(opt_junk_alloc)) {
+				memset((void *)((uintptr_t)ptr + oldsize), 0xa5,
+				    isalloc(ptr, config_prof) - oldsize);
+			} else if (unlikely(opt_zero)) {
+				memset((void *)((uintptr_t)ptr + oldsize), 0,
+				    isalloc(ptr, config_prof) - oldsize);
 			}
-			return (ret);
 		}
+		return (ret);
 	}
+
+	assert(oldsize > usize_max);
+	/* Fill before shrinking in order avoid a race. */
+	arena_ralloc_junk_large(ptr, oldsize, usize_max);
+	arena_ralloc_large_shrink(arena, chunk, ptr, oldsize, usize_max);
+	return (false);
 }
 
 bool
 arena_ralloc_no_move(void *ptr, size_t oldsize, size_t size, size_t extra,
     bool zero)
 {
+	size_t usize_min, usize_max;
 
-	if (likely(size <= arena_maxclass)) {
+	usize_min = s2u(size);
+	usize_max = s2u(size + extra);
+	if (likely(oldsize <= large_maxclass && usize_min <= large_maxclass)) {
 		/*
 		 * Avoid moving the allocation if the size class can be left the
 		 * same.
 		 */
-		if (likely(oldsize <= arena_maxclass)) {
-			if (oldsize <= SMALL_MAXCLASS) {
-				assert(
-				    arena_bin_info[size2index(oldsize)].reg_size
-				    == oldsize);
-				if ((size + extra <= SMALL_MAXCLASS &&
-				    size2index(size + extra) ==
-				    size2index(oldsize)) || (size <= oldsize &&
-				    size + extra >= oldsize))
+		if (oldsize <= SMALL_MAXCLASS) {
+			assert(arena_bin_info[size2index(oldsize)].reg_size ==
+			    oldsize);
+			if ((usize_max <= SMALL_MAXCLASS &&
+			    size2index(usize_max) == size2index(oldsize)) ||
+			    (size <= oldsize && usize_max >= oldsize))
+				return (false);
+		} else {
+			if (usize_max > SMALL_MAXCLASS) {
+				if (!arena_ralloc_large(ptr, oldsize, usize_min,
+				    usize_max, zero))
 					return (false);
-			} else {
-				assert(size <= arena_maxclass);
-				if (size + extra > SMALL_MAXCLASS) {
-					if (!arena_ralloc_large(ptr, oldsize,
-					    size, extra, zero))
-						return (false);
-				}
 			}
 		}
 
 		/* Reallocation would require a move. */
 		return (true);
-	} else
-		return (huge_ralloc_no_move(ptr, oldsize, size, extra, zero));
+	} else {
+		return (huge_ralloc_no_move(ptr, oldsize, usize_min, usize_max,
+		    zero));
+	}
+}
+
+static void *
+arena_ralloc_move_helper(tsd_t *tsd, arena_t *arena, size_t usize,
+    size_t alignment, bool zero, tcache_t *tcache)
+{
+
+	if (alignment == 0)
+		return (arena_malloc(tsd, arena, usize, zero, tcache));
+	usize = sa2u(usize, alignment);
+	if (usize == 0)
+		return (NULL);
+	return (ipalloct(tsd, usize, alignment, zero, tcache, arena));
 }
 
 void *
 arena_ralloc(tsd_t *tsd, arena_t *arena, void *ptr, size_t oldsize, size_t size,
-    size_t extra, size_t alignment, bool zero, tcache_t *tcache)
+    size_t alignment, bool zero, tcache_t *tcache)
 {
 	void *ret;
+	size_t usize;
 
-	if (likely(size <= arena_maxclass)) {
+	usize = s2u(size);
+	if (usize == 0)
+		return (NULL);
+
+	if (likely(usize <= large_maxclass)) {
 		size_t copysize;
 
 		/* Try to avoid moving the allocation. */
-		if (!arena_ralloc_no_move(ptr, oldsize, size, extra, zero))
+		if (!arena_ralloc_no_move(ptr, oldsize, usize, 0, zero))
 			return (ptr);
 
 		/*
@@ -2847,53 +2857,23 @@ arena_ralloc(tsd_t *tsd, arena_t *arena, void *ptr, size_t oldsize, size_t size,
 		 * the object.  In that case, fall back to allocating new space
 		 * and copying.
 		 */
-		if (alignment != 0) {
-			size_t usize = sa2u(size + extra, alignment);
-			if (usize == 0)
-				return (NULL);
-			ret = ipalloct(tsd, usize, alignment, zero, tcache,
-			    arena);
-		} else {
-			ret = arena_malloc(tsd, arena, size + extra, zero,
-			    tcache);
-		}
-
-		if (ret == NULL) {
-			if (extra == 0)
-				return (NULL);
-			/* Try again, this time without extra. */
-			if (alignment != 0) {
-				size_t usize = sa2u(size, alignment);
-				if (usize == 0)
-					return (NULL);
-				ret = ipalloct(tsd, usize, alignment, zero,
-				    tcache, arena);
-			} else {
-				ret = arena_malloc(tsd, arena, size, zero,
-				    tcache);
-			}
-
-			if (ret == NULL)
-				return (NULL);
-		}
+		ret = arena_ralloc_move_helper(tsd, arena, usize, alignment,
+		    zero, tcache);
+		if (ret == NULL)
+			return (NULL);
 
 		/*
 		 * Junk/zero-filling were already done by
 		 * ipalloc()/arena_malloc().
 		 */
 
-		/*
-		 * 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;
+		copysize = (usize < oldsize) ? usize : oldsize;
 		JEMALLOC_VALGRIND_MAKE_MEM_UNDEFINED(ret, copysize);
 		memcpy(ret, ptr, copysize);
 		isqalloc(tsd, ptr, oldsize, tcache);
 	} else {
-		ret = huge_ralloc(tsd, arena, ptr, oldsize, size, extra,
-		    alignment, zero, tcache);
+		ret = huge_ralloc(tsd, arena, ptr, oldsize, usize, alignment,
+		    zero, tcache);
 	}
 	return (ret);
 }
@@ -3241,7 +3221,6 @@ small_run_size_init(void)
 bool
 arena_boot(void)
 {
-	size_t header_size;
 	unsigned i;
 
 	arena_lg_dirty_mult_default_set(opt_lg_dirty_mult);
@@ -3260,7 +3239,7 @@ arena_boot(void)
 	 */
 	map_bias = 0;
 	for (i = 0; i < 3; i++) {
-		header_size = offsetof(arena_chunk_t, map_bits) +
+		size_t header_size = offsetof(arena_chunk_t, map_bits) +
 		    ((sizeof(arena_chunk_map_bits_t) +
 		    sizeof(arena_chunk_map_misc_t)) * (chunk_npages-map_bias));
 		map_bias = (header_size + PAGE_MASK) >> LG_PAGE;
@@ -3272,17 +3251,17 @@ arena_boot(void)
 
 	arena_maxrun = chunksize - (map_bias << LG_PAGE);
 	assert(arena_maxrun > 0);
-	arena_maxclass = index2size(size2index(chunksize)-1);
-	if (arena_maxclass > arena_maxrun) {
+	large_maxclass = index2size(size2index(chunksize)-1);
+	if (large_maxclass > arena_maxrun) {
 		/*
 		 * For small chunk sizes it's possible for there to be fewer
 		 * non-header pages available than are necessary to serve the
 		 * size classes just below chunksize.
 		 */
-		arena_maxclass = arena_maxrun;
+		large_maxclass = arena_maxrun;
 	}
-	assert(arena_maxclass > 0);
-	nlclasses = size2index(arena_maxclass) - size2index(SMALL_MAXCLASS);
+	assert(large_maxclass > 0);
+	nlclasses = size2index(large_maxclass) - size2index(SMALL_MAXCLASS);
 	nhclasses = NSIZES - nlclasses - NBINS;
 
 	bin_info_init();
diff --git a/contrib/jemalloc/src/chunk_dss.c b/contrib/jemalloc/src/chunk_dss.c
index de0546d..61fc916 100644
--- a/contrib/jemalloc/src/chunk_dss.c
+++ b/contrib/jemalloc/src/chunk_dss.c
@@ -69,8 +69,6 @@ void *
 chunk_alloc_dss(arena_t *arena, void *new_addr, size_t size, size_t alignment,
     bool *zero, bool *commit)
 {
-	void *ret;
-
 	cassert(have_dss);
 	assert(size > 0 && (size & chunksize_mask) == 0);
 	assert(alignment > 0 && (alignment & chunksize_mask) == 0);
@@ -84,9 +82,6 @@ chunk_alloc_dss(arena_t *arena, void *new_addr, size_t size, size_t alignment,
 
 	malloc_mutex_lock(&dss_mtx);
 	if (dss_prev != (void *)-1) {
-		size_t gap_size, cpad_size;
-		void *cpad, *dss_next;
-		intptr_t incr;
 
 		/*
 		 * The loop is necessary to recover from races with other
@@ -94,6 +89,9 @@ chunk_alloc_dss(arena_t *arena, void *new_addr, size_t size, size_t alignment,
 		 * malloc.
 		 */
 		do {
+			void *ret, *cpad, *dss_next;
+			size_t gap_size, cpad_size;
+			intptr_t incr;
 			/* Avoid an unnecessary system call. */
 			if (new_addr != NULL && dss_max != new_addr)
 				break;
diff --git a/contrib/jemalloc/src/chunk_mmap.c b/contrib/jemalloc/src/chunk_mmap.c
index 36eb075..b9ba741 100644
--- a/contrib/jemalloc/src/chunk_mmap.c
+++ b/contrib/jemalloc/src/chunk_mmap.c
@@ -6,14 +6,16 @@
 static void *
 chunk_alloc_mmap_slow(size_t size, size_t alignment, bool *zero, bool *commit)
 {
-	void *ret, *pages;
-	size_t alloc_size, leadsize;
+	void *ret;
+	size_t alloc_size;
 
 	alloc_size = size + alignment - PAGE;
 	/* Beware size_t wrap-around. */
 	if (alloc_size < size)
 		return (NULL);
 	do {
+		void *pages;
+		size_t leadsize;
 		pages = pages_map(NULL, alloc_size);
 		if (pages == NULL)
 			return (NULL);
diff --git a/contrib/jemalloc/src/huge.c b/contrib/jemalloc/src/huge.c
index 54c2114..f8778db 100644
--- a/contrib/jemalloc/src/huge.c
+++ b/contrib/jemalloc/src/huge.c
@@ -126,18 +126,19 @@ huge_dalloc_junk_t *huge_dalloc_junk = JEMALLOC_N(huge_dalloc_junk_impl);
 #endif
 
 static void
-huge_ralloc_no_move_similar(void *ptr, size_t oldsize, size_t usize,
-    size_t size, size_t extra, bool zero)
+huge_ralloc_no_move_similar(void *ptr, size_t oldsize, size_t usize_min,
+    size_t usize_max, bool zero)
 {
-	size_t usize_next;
+	size_t usize, usize_next;
 	extent_node_t *node;
 	arena_t *arena;
 	chunk_hooks_t chunk_hooks = CHUNK_HOOKS_INITIALIZER;
 	bool zeroed;
 
 	/* Increase usize to incorporate extra. */
-	while (usize < s2u(size+extra) && (usize_next = s2u(usize+1)) < oldsize)
-		usize = usize_next;
+	for (usize = usize_min; usize < usize_max && (usize_next = s2u(usize+1))
+	    <= oldsize; usize = usize_next)
+		; /* Do nothing. */
 
 	if (oldsize == usize)
 		return;
@@ -148,11 +149,12 @@ huge_ralloc_no_move_similar(void *ptr, size_t oldsize, size_t usize,
 	/* Fill if necessary (shrinking). */
 	if (oldsize > usize) {
 		size_t sdiff = oldsize - usize;
-		zeroed = !chunk_purge_wrapper(arena, &chunk_hooks, ptr,
-		    CHUNK_CEILING(usize), usize, sdiff);
 		if (config_fill && unlikely(opt_junk_free)) {
 			memset((void *)((uintptr_t)ptr + usize), 0x5a, sdiff);
 			zeroed = false;
+		} else {
+			zeroed = !chunk_purge_wrapper(arena, &chunk_hooks, ptr,
+			    CHUNK_CEILING(oldsize), usize, sdiff);
 		}
 	} else
 		zeroed = true;
@@ -194,6 +196,8 @@ huge_ralloc_no_move_shrink(void *ptr, size_t oldsize, size_t usize)
 	arena = extent_node_arena_get(node);
 	chunk_hooks = chunk_hooks_get(arena);
 
+	assert(oldsize > usize);
+
 	/* Split excess chunks. */
 	cdiff = CHUNK_CEILING(oldsize) - CHUNK_CEILING(usize);
 	if (cdiff != 0 && chunk_hooks.split(ptr, CHUNK_CEILING(oldsize),
@@ -202,14 +206,15 @@ huge_ralloc_no_move_shrink(void *ptr, size_t oldsize, size_t usize)
 
 	if (oldsize > usize) {
 		size_t sdiff = oldsize - usize;
-		zeroed = !chunk_purge_wrapper(arena, &chunk_hooks,
-		    CHUNK_ADDR2BASE((uintptr_t)ptr + usize),
-		    CHUNK_CEILING(usize), CHUNK_ADDR2OFFSET((uintptr_t)ptr +
-		    usize), sdiff);
 		if (config_fill && unlikely(opt_junk_free)) {
 			huge_dalloc_junk((void *)((uintptr_t)ptr + usize),
 			    sdiff);
 			zeroed = false;
+		} else {
+			zeroed = !chunk_purge_wrapper(arena, &chunk_hooks,
+			    CHUNK_ADDR2BASE((uintptr_t)ptr + usize),
+			    CHUNK_CEILING(oldsize),
+			    CHUNK_ADDR2OFFSET((uintptr_t)ptr + usize), sdiff);
 		}
 	} else
 		zeroed = true;
@@ -228,18 +233,11 @@ huge_ralloc_no_move_shrink(void *ptr, size_t oldsize, size_t usize)
 }
 
 static bool
-huge_ralloc_no_move_expand(void *ptr, size_t oldsize, size_t size, bool zero) {
-	size_t usize;
+huge_ralloc_no_move_expand(void *ptr, size_t oldsize, size_t usize, bool zero) {
 	extent_node_t *node;
 	arena_t *arena;
 	bool is_zeroed_subchunk, is_zeroed_chunk;
 
-	usize = s2u(size);
-	if (usize == 0) {
-		/* size_t overflow. */
-		return (true);
-	}
-
 	node = huge_node_get(ptr);
 	arena = extent_node_arena_get(node);
 	malloc_mutex_lock(&arena->huge_mtx);
@@ -280,89 +278,76 @@ huge_ralloc_no_move_expand(void *ptr, size_t oldsize, size_t size, bool zero) {
 }
 
 bool
-huge_ralloc_no_move(void *ptr, size_t oldsize, size_t size, size_t extra,
-    bool zero)
+huge_ralloc_no_move(void *ptr, size_t oldsize, size_t usize_min,
+    size_t usize_max, bool zero)
 {
-	size_t usize;
+
+	assert(s2u(oldsize) == oldsize);
 
 	/* Both allocations must be huge to avoid a move. */
-	if (oldsize < chunksize)
+	if (oldsize < chunksize || usize_max < chunksize)
 		return (true);
 
-	assert(s2u(oldsize) == oldsize);
-	usize = s2u(size);
-	if (usize == 0) {
-		/* size_t overflow. */
-		return (true);
+	if (CHUNK_CEILING(usize_max) > CHUNK_CEILING(oldsize)) {
+		/* Attempt to expand the allocation in-place. */
+		if (!huge_ralloc_no_move_expand(ptr, oldsize, usize_max, zero))
+			return (false);
+		/* Try again, this time with usize_min. */
+		if (usize_min < usize_max && CHUNK_CEILING(usize_min) >
+		    CHUNK_CEILING(oldsize) && huge_ralloc_no_move_expand(ptr,
+		    oldsize, usize_min, zero))
+			return (false);
 	}
 
 	/*
 	 * Avoid moving the allocation if the existing chunk size accommodates
 	 * the new size.
 	 */
-	if (CHUNK_CEILING(oldsize) >= CHUNK_CEILING(usize)
-	    && CHUNK_CEILING(oldsize) <= CHUNK_CEILING(s2u(size+extra))) {
-		huge_ralloc_no_move_similar(ptr, oldsize, usize, size, extra,
+	if (CHUNK_CEILING(oldsize) >= CHUNK_CEILING(usize_min)
+	    && CHUNK_CEILING(oldsize) <= CHUNK_CEILING(usize_max)) {
+		huge_ralloc_no_move_similar(ptr, oldsize, usize_min, usize_max,
 		    zero);
 		return (false);
 	}
 
 	/* Attempt to shrink the allocation in-place. */
-	if (CHUNK_CEILING(oldsize) >= CHUNK_CEILING(usize))
-		return (huge_ralloc_no_move_shrink(ptr, oldsize, usize));
+	if (CHUNK_CEILING(oldsize) > CHUNK_CEILING(usize_max))
+		return (huge_ralloc_no_move_shrink(ptr, oldsize, usize_max));
+	return (true);
+}
 
-	/* Attempt to expand the allocation in-place. */
-	if (huge_ralloc_no_move_expand(ptr, oldsize, size + extra, zero)) {
-		if (extra == 0)
-			return (true);
+static void *
+huge_ralloc_move_helper(tsd_t *tsd, arena_t *arena, size_t usize,
+    size_t alignment, bool zero, tcache_t *tcache)
+{
 
-		/* Try again, this time without extra. */
-		return (huge_ralloc_no_move_expand(ptr, oldsize, size, zero));
-	}
-	return (false);
+	if (alignment <= chunksize)
+		return (huge_malloc(tsd, arena, usize, zero, tcache));
+	return (huge_palloc(tsd, arena, usize, alignment, zero, tcache));
 }
 
 void *
-huge_ralloc(tsd_t *tsd, arena_t *arena, void *ptr, size_t oldsize, size_t size,
-    size_t extra, size_t alignment, bool zero, tcache_t *tcache)
+huge_ralloc(tsd_t *tsd, arena_t *arena, void *ptr, size_t oldsize, size_t usize,
+    size_t alignment, bool zero, tcache_t *tcache)
 {
 	void *ret;
 	size_t copysize;
 
 	/* Try to avoid moving the allocation. */
-	if (!huge_ralloc_no_move(ptr, oldsize, size, extra, zero))
+	if (!huge_ralloc_no_move(ptr, oldsize, usize, usize, zero))
 		return (ptr);
 
 	/*
-	 * size and oldsize are different enough that we need to use a
+	 * usize and oldsize are different enough that we need to use a
 	 * different size class.  In that case, fall back to allocating new
 	 * space and copying.
 	 */
-	if (alignment > chunksize) {
-		ret = huge_palloc(tsd, arena, size + extra, alignment, zero,
-		    tcache);
-	} else
-		ret = huge_malloc(tsd, arena, size + extra, zero, tcache);
-
-	if (ret == NULL) {
-		if (extra == 0)
-			return (NULL);
-		/* Try again, this time without extra. */
-		if (alignment > chunksize) {
-			ret = huge_palloc(tsd, arena, size, alignment, zero,
-			    tcache);
-		} else
-			ret = huge_malloc(tsd, arena, size, zero, tcache);
-
-		if (ret == NULL)
-			return (NULL);
-	}
+	ret = huge_ralloc_move_helper(tsd, arena, usize, alignment, zero,
+	    tcache);
+	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;
+	copysize = (usize < oldsize) ? usize : oldsize;
 	memcpy(ret, ptr, copysize);
 	isqalloc(tsd, ptr, oldsize, tcache);
 	return (ret);
@@ -439,3 +424,10 @@ huge_prof_tctx_set(const void *ptr, prof_tctx_t *tctx)
 	extent_node_prof_tctx_set(node, tctx);
 	malloc_mutex_unlock(&arena->huge_mtx);
 }
+
+void
+huge_prof_tctx_reset(const void *ptr)
+{
+
+	huge_prof_tctx_set(ptr, (prof_tctx_t *)(uintptr_t)1U);
+}
diff --git a/contrib/jemalloc/src/jemalloc.c b/contrib/jemalloc/src/jemalloc.c
index d078a1f..b6cbb79 100644
--- a/contrib/jemalloc/src/jemalloc.c
+++ b/contrib/jemalloc/src/jemalloc.c
@@ -183,13 +183,24 @@ static bool			malloc_initializer = NO_INITIALIZER;
 static malloc_mutex_t	init_lock = SRWLOCK_INIT;
 #else
 static malloc_mutex_t	init_lock;
+static bool init_lock_initialized = false;
 
 JEMALLOC_ATTR(constructor)
 static void WINAPI
 _init_init_lock(void)
 {
 
-	malloc_mutex_init(&init_lock);
+	/* If another constructor in the same binary is using mallctl to
+	 * e.g. setup chunk hooks, it may end up running before this one,
+	 * and malloc_init_hard will crash trying to lock the uninitialized
+	 * lock. So we force an initialization of the lock in
+	 * malloc_init_hard as well. We don't try to care about atomicity
+	 * of the accessed to the init_lock_initialized boolean, since it
+	 * really only matters early in the process creation, before any
+	 * separate thread normally starts doing anything. */
+	if (!init_lock_initialized)
+		malloc_mutex_init(&init_lock);
+	init_lock_initialized = true;
 }
 
 #ifdef _MSC_VER
@@ -514,17 +525,17 @@ arena_get_hard(tsd_t *tsd, unsigned ind, bool init_if_missing)
 		assert(ind < narenas_actual || !init_if_missing);
 		narenas_cache = (ind < narenas_actual) ? narenas_actual : ind+1;
 
-		if (!*arenas_cache_bypassp) {
+		if (tsd_nominal(tsd) && !*arenas_cache_bypassp) {
 			*arenas_cache_bypassp = true;
 			arenas_cache = (arena_t **)a0malloc(sizeof(arena_t *) *
 			    narenas_cache);
 			*arenas_cache_bypassp = false;
-		} else
-			arenas_cache = NULL;
+		}
 		if (arenas_cache == NULL) {
 			/*
 			 * This function must always tell the truth, even if
-			 * it's slow, so don't let OOM or recursive allocation
+			 * it's slow, so don't let OOM, thread cleanup (note
+			 * tsd_nominal check), nor recursive allocation
 			 * avoidance (note arenas_cache_bypass check) get in the
 			 * way.
 			 */
@@ -535,6 +546,7 @@ arena_get_hard(tsd_t *tsd, unsigned ind, bool init_if_missing)
 			malloc_mutex_unlock(&arenas_lock);
 			return (arena);
 		}
+		assert(tsd_nominal(tsd) && !*arenas_cache_bypassp);
 		tsd_arenas_cache_set(tsd, arenas_cache);
 		tsd_narenas_cache_set(tsd, narenas_cache);
 	}
@@ -653,8 +665,10 @@ arenas_cache_cleanup(tsd_t *tsd)
 	arena_t **arenas_cache;
 
 	arenas_cache = tsd_arenas_cache_get(tsd);
-	if (arenas_cache != NULL)
+	if (arenas_cache != NULL) {
+		tsd_arenas_cache_set(tsd, NULL);
 		a0dalloc(arenas_cache);
+	}
 }
 
 void
@@ -1301,6 +1315,9 @@ static bool
 malloc_init_hard(void)
 {
 
+#if defined(_WIN32) && _WIN32_WINNT < 0x0600
+	_init_init_lock();
+#endif
 	malloc_mutex_lock(&init_lock);
 	if (!malloc_init_hard_needed()) {
 		malloc_mutex_unlock(&init_lock);
@@ -1365,7 +1382,7 @@ imalloc_prof(tsd_t *tsd, size_t usize)
 	void *p;
 	prof_tctx_t *tctx;
 
-	tctx = prof_alloc_prep(tsd, usize, true);
+	tctx = prof_alloc_prep(tsd, usize, prof_active_get_unlocked(), true);
 	if (unlikely((uintptr_t)tctx != (uintptr_t)1U))
 		p = imalloc_prof_sample(tsd, usize, tctx);
 	else
@@ -1455,7 +1472,7 @@ imemalign_prof(tsd_t *tsd, size_t alignment, size_t usize)
 	void *p;
 	prof_tctx_t *tctx;
 
-	tctx = prof_alloc_prep(tsd, usize, true);
+	tctx = prof_alloc_prep(tsd, usize, prof_active_get_unlocked(), true);
 	if (unlikely((uintptr_t)tctx != (uintptr_t)1U))
 		p = imemalign_prof_sample(tsd, alignment, usize, tctx);
 	else
@@ -1586,7 +1603,7 @@ icalloc_prof(tsd_t *tsd, size_t usize)
 	void *p;
 	prof_tctx_t *tctx;
 
-	tctx = prof_alloc_prep(tsd, usize, true);
+	tctx = prof_alloc_prep(tsd, usize, prof_active_get_unlocked(), true);
 	if (unlikely((uintptr_t)tctx != (uintptr_t)1U))
 		p = icalloc_prof_sample(tsd, usize, tctx);
 	else
@@ -1669,7 +1686,7 @@ label_return:
 }
 
 static void *
-irealloc_prof_sample(tsd_t *tsd, void *oldptr, size_t old_usize, size_t usize,
+irealloc_prof_sample(tsd_t *tsd, void *old_ptr, size_t old_usize, size_t usize,
     prof_tctx_t *tctx)
 {
 	void *p;
@@ -1677,31 +1694,36 @@ irealloc_prof_sample(tsd_t *tsd, void *oldptr, size_t old_usize, size_t usize,
 	if (tctx == NULL)
 		return (NULL);
 	if (usize <= SMALL_MAXCLASS) {
-		p = iralloc(tsd, oldptr, old_usize, LARGE_MINCLASS, 0, false);
+		p = iralloc(tsd, old_ptr, old_usize, LARGE_MINCLASS, 0, false);
 		if (p == NULL)
 			return (NULL);
 		arena_prof_promoted(p, usize);
 	} else
-		p = iralloc(tsd, oldptr, old_usize, usize, 0, false);
+		p = iralloc(tsd, old_ptr, old_usize, usize, 0, false);
 
 	return (p);
 }
 
 JEMALLOC_ALWAYS_INLINE_C void *
-irealloc_prof(tsd_t *tsd, void *oldptr, size_t old_usize, size_t usize)
+irealloc_prof(tsd_t *tsd, void *old_ptr, size_t old_usize, size_t usize)
 {
 	void *p;
+	bool prof_active;
 	prof_tctx_t *old_tctx, *tctx;
 
-	old_tctx = prof_tctx_get(oldptr);
-	tctx = prof_alloc_prep(tsd, usize, true);
+	prof_active = prof_active_get_unlocked();
+	old_tctx = prof_tctx_get(old_ptr);
+	tctx = prof_alloc_prep(tsd, usize, prof_active, true);
 	if (unlikely((uintptr_t)tctx != (uintptr_t)1U))
-		p = irealloc_prof_sample(tsd, oldptr, old_usize, usize, tctx);
+		p = irealloc_prof_sample(tsd, old_ptr, old_usize, usize, tctx);
 	else
-		p = iralloc(tsd, oldptr, old_usize, usize, 0, false);
-	if (p == NULL)
+		p = iralloc(tsd, old_ptr, old_usize, usize, 0, false);
+	if (unlikely(p == NULL)) {
+		prof_alloc_rollback(tsd, tctx, true);
 		return (NULL);
-	prof_realloc(tsd, p, usize, tctx, true, old_usize, old_tctx);
+	}
+	prof_realloc(tsd, p, usize, tctx, prof_active, true, old_ptr, old_usize,
+	    old_tctx);
 
 	return (p);
 }
@@ -1905,6 +1927,7 @@ imallocx_flags_decode_hard(tsd_t *tsd, size_t size, int flags, size_t *usize,
 		*alignment = MALLOCX_ALIGN_GET_SPECIFIED(flags);
 		*usize = sa2u(size, *alignment);
 	}
+	assert(*usize != 0);
 	*zero = MALLOCX_ZERO_GET(flags);
 	if ((flags & MALLOCX_TCACHE_MASK) != 0) {
 		if ((flags & MALLOCX_TCACHE_MASK) == MALLOCX_TCACHE_NONE)
@@ -1947,41 +1970,29 @@ imallocx_flags(tsd_t *tsd, size_t usize, size_t alignment, bool zero,
     tcache_t *tcache, arena_t *arena)
 {
 
-	if (alignment != 0)
+	if (unlikely(alignment != 0))
 		return (ipalloct(tsd, usize, alignment, zero, tcache, arena));
-	if (zero)
+	if (unlikely(zero))
 		return (icalloct(tsd, usize, tcache, arena));
 	return (imalloct(tsd, usize, tcache, arena));
 }
 
-JEMALLOC_ALWAYS_INLINE_C void *
-imallocx_maybe_flags(tsd_t *tsd, size_t size, int flags, size_t usize,
-    size_t alignment, bool zero, tcache_t *tcache, arena_t *arena)
-{
-
-	if (likely(flags == 0))
-		return (imalloc(tsd, size));
-	return (imallocx_flags(tsd, usize, alignment, zero, tcache, arena));
-}
-
 static void *
-imallocx_prof_sample(tsd_t *tsd, size_t size, int flags, size_t usize,
-    size_t alignment, bool zero, tcache_t *tcache, arena_t *arena)
+imallocx_prof_sample(tsd_t *tsd, size_t usize, size_t alignment, bool zero,
+    tcache_t *tcache, arena_t *arena)
 {
 	void *p;
 
 	if (usize <= SMALL_MAXCLASS) {
 		assert(((alignment == 0) ? s2u(LARGE_MINCLASS) :
 		    sa2u(LARGE_MINCLASS, alignment)) == LARGE_MINCLASS);
-		p = imallocx_maybe_flags(tsd, LARGE_MINCLASS, flags,
-		    LARGE_MINCLASS, alignment, zero, tcache, arena);
+		p = imallocx_flags(tsd, LARGE_MINCLASS, alignment, zero, tcache,
+		    arena);
 		if (p == NULL)
 			return (NULL);
 		arena_prof_promoted(p, usize);
-	} else {
-		p = imallocx_maybe_flags(tsd, size, flags, usize, alignment,
-		    zero, tcache, arena);
-	}
+	} else
+		p = imallocx_flags(tsd, usize, alignment, zero, tcache, arena);
 
 	return (p);
 }
@@ -1999,13 +2010,12 @@ imallocx_prof(tsd_t *tsd, size_t size, int flags, size_t *usize)
 	if (unlikely(imallocx_flags_decode(tsd, size, flags, usize, &alignment,
 	    &zero, &tcache, &arena)))
 		return (NULL);
-	tctx = prof_alloc_prep(tsd, *usize, true);
-	if (likely((uintptr_t)tctx == (uintptr_t)1U)) {
-		p = imallocx_maybe_flags(tsd, size, flags, *usize, alignment,
-		    zero, tcache, arena);
-	} else if ((uintptr_t)tctx > (uintptr_t)1U) {
-		p = imallocx_prof_sample(tsd, size, flags, *usize, alignment,
-		    zero, tcache, arena);
+	tctx = prof_alloc_prep(tsd, *usize, prof_active_get_unlocked(), true);
+	if (likely((uintptr_t)tctx == (uintptr_t)1U))
+		p = imallocx_flags(tsd, *usize, alignment, zero, tcache, arena);
+	else if ((uintptr_t)tctx > (uintptr_t)1U) {
+		p = imallocx_prof_sample(tsd, *usize, alignment, zero, tcache,
+		    arena);
 	} else
 		p = NULL;
 	if (unlikely(p == NULL)) {
@@ -2080,8 +2090,8 @@ label_oom:
 }
 
 static void *
-irallocx_prof_sample(tsd_t *tsd, void *oldptr, size_t old_usize, size_t size,
-    size_t alignment, size_t usize, bool zero, tcache_t *tcache, arena_t *arena,
+irallocx_prof_sample(tsd_t *tsd, void *old_ptr, size_t old_usize,
+    size_t usize, size_t alignment, bool zero, tcache_t *tcache, arena_t *arena,
     prof_tctx_t *tctx)
 {
 	void *p;
@@ -2089,13 +2099,13 @@ irallocx_prof_sample(tsd_t *tsd, void *oldptr, size_t old_usize, size_t size,
 	if (tctx == NULL)
 		return (NULL);
 	if (usize <= SMALL_MAXCLASS) {
-		p = iralloct(tsd, oldptr, old_usize, LARGE_MINCLASS, alignment,
+		p = iralloct(tsd, old_ptr, old_usize, LARGE_MINCLASS, alignment,
 		    zero, tcache, arena);
 		if (p == NULL)
 			return (NULL);
 		arena_prof_promoted(p, usize);
 	} else {
-		p = iralloct(tsd, oldptr, old_usize, size, alignment, zero,
+		p = iralloct(tsd, old_ptr, old_usize, usize, alignment, zero,
 		    tcache, arena);
 	}
 
@@ -2103,28 +2113,30 @@ irallocx_prof_sample(tsd_t *tsd, void *oldptr, size_t old_usize, size_t size,
 }
 
 JEMALLOC_ALWAYS_INLINE_C void *
-irallocx_prof(tsd_t *tsd, void *oldptr, size_t old_usize, size_t size,
+irallocx_prof(tsd_t *tsd, void *old_ptr, size_t old_usize, size_t size,
     size_t alignment, size_t *usize, bool zero, tcache_t *tcache,
     arena_t *arena)
 {
 	void *p;
+	bool prof_active;
 	prof_tctx_t *old_tctx, *tctx;
 
-	old_tctx = prof_tctx_get(oldptr);
-	tctx = prof_alloc_prep(tsd, *usize, false);
+	prof_active = prof_active_get_unlocked();
+	old_tctx = prof_tctx_get(old_ptr);
+	tctx = prof_alloc_prep(tsd, *usize, prof_active, true);
 	if (unlikely((uintptr_t)tctx != (uintptr_t)1U)) {
-		p = irallocx_prof_sample(tsd, oldptr, old_usize, size,
-		    alignment, *usize, zero, tcache, arena, tctx);
+		p = irallocx_prof_sample(tsd, old_ptr, old_usize, *usize,
+		    alignment, zero, tcache, arena, tctx);
 	} else {
-		p = iralloct(tsd, oldptr, old_usize, size, alignment, zero,
+		p = iralloct(tsd, old_ptr, old_usize, size, alignment, zero,
 		    tcache, arena);
 	}
 	if (unlikely(p == NULL)) {
-		prof_alloc_rollback(tsd, tctx, false);
+		prof_alloc_rollback(tsd, tctx, true);
 		return (NULL);
 	}
 
-	if (p == oldptr && alignment != 0) {
+	if (p == old_ptr && alignment != 0) {
 		/*
 		 * The allocation did not move, so it is possible that the size
 		 * class is smaller than would guarantee the requested
@@ -2135,7 +2147,8 @@ irallocx_prof(tsd_t *tsd, void *oldptr, size_t old_usize, size_t size,
 		 */
 		*usize = isalloc(p, config_prof);
 	}
-	prof_realloc(tsd, p, *usize, tctx, false, old_usize, old_tctx);
+	prof_realloc(tsd, p, *usize, tctx, prof_active, true, old_ptr,
+	    old_usize, old_tctx);
 
 	return (p);
 }
@@ -2230,26 +2243,13 @@ ixallocx_helper(void *ptr, size_t old_usize, size_t size, size_t extra,
 
 static size_t
 ixallocx_prof_sample(void *ptr, size_t old_usize, size_t size, size_t extra,
-    size_t alignment, size_t max_usize, bool zero, prof_tctx_t *tctx)
+    size_t alignment, bool zero, prof_tctx_t *tctx)
 {
 	size_t usize;
 
 	if (tctx == NULL)
 		return (old_usize);
-	/* Use minimum usize to determine whether promotion may happen. */
-	if (((alignment == 0) ? s2u(size) : sa2u(size, alignment)) <=
-	    SMALL_MAXCLASS) {
-		if (ixalloc(ptr, old_usize, SMALL_MAXCLASS+1,
-		    (SMALL_MAXCLASS+1 >= size+extra) ? 0 : size+extra -
-		    (SMALL_MAXCLASS+1), alignment, zero))
-			return (old_usize);
-		usize = isalloc(ptr, config_prof);
-		if (max_usize < LARGE_MINCLASS)
-			arena_prof_promoted(ptr, usize);
-	} else {
-		usize = ixallocx_helper(ptr, old_usize, size, extra, alignment,
-		    zero);
-	}
+	usize = ixallocx_helper(ptr, old_usize, size, extra, alignment, zero);
 
 	return (usize);
 }
@@ -2258,9 +2258,11 @@ JEMALLOC_ALWAYS_INLINE_C size_t
 ixallocx_prof(tsd_t *tsd, void *ptr, size_t old_usize, size_t size,
     size_t extra, size_t alignment, bool zero)
 {
-	size_t max_usize, usize;
+	size_t usize_max, usize;
+	bool prof_active;
 	prof_tctx_t *old_tctx, *tctx;
 
+	prof_active = prof_active_get_unlocked();
 	old_tctx = prof_tctx_get(ptr);
 	/*
 	 * usize isn't knowable before ixalloc() returns when extra is non-zero.
@@ -2268,21 +2270,23 @@ ixallocx_prof(tsd_t *tsd, void *ptr, size_t old_usize, size_t size,
 	 * prof_alloc_prep() to decide whether to capture a backtrace.
 	 * prof_realloc() will use the actual usize to decide whether to sample.
 	 */
-	max_usize = (alignment == 0) ? s2u(size+extra) : sa2u(size+extra,
+	usize_max = (alignment == 0) ? s2u(size+extra) : sa2u(size+extra,
 	    alignment);
-	tctx = prof_alloc_prep(tsd, max_usize, false);
+	assert(usize_max != 0);
+	tctx = prof_alloc_prep(tsd, usize_max, prof_active, false);
 	if (unlikely((uintptr_t)tctx != (uintptr_t)1U)) {
 		usize = ixallocx_prof_sample(ptr, old_usize, size, extra,
-		    alignment, zero, max_usize, tctx);
+		    alignment, zero, tctx);
 	} else {
 		usize = ixallocx_helper(ptr, old_usize, size, extra, alignment,
 		    zero);
 	}
-	if (unlikely(usize == old_usize)) {
+	if (usize == old_usize) {
 		prof_alloc_rollback(tsd, tctx, false);
 		return (usize);
 	}
-	prof_realloc(tsd, ptr, usize, tctx, false, old_usize, old_tctx);
+	prof_realloc(tsd, ptr, usize, tctx, prof_active, false, ptr, old_usize,
+	    old_tctx);
 
 	return (usize);
 }
@@ -2304,6 +2308,17 @@ je_xallocx(void *ptr, size_t size, size_t extra, int flags)
 	tsd = tsd_fetch();
 
 	old_usize = isalloc(ptr, config_prof);
+
+	/* Clamp extra if necessary to avoid (size + extra) overflow. */
+	if (unlikely(size + extra > HUGE_MAXCLASS)) {
+		/* Check for size overflow. */
+		if (unlikely(size > HUGE_MAXCLASS)) {
+			usize = old_usize;
+			goto label_not_resized;
+		}
+		extra = HUGE_MAXCLASS - size;
+	}
+
 	if (config_valgrind && unlikely(in_valgrind))
 		old_rzsize = u2rz(old_usize);
 
diff --git a/contrib/jemalloc/src/prof.c b/contrib/jemalloc/src/prof.c
index a05792f..0a08062 100644
--- a/contrib/jemalloc/src/prof.c
+++ b/contrib/jemalloc/src/prof.c
@@ -139,9 +139,16 @@ prof_tctx_comp(const prof_tctx_t *a, const prof_tctx_t *b)
 	uint64_t b_thr_uid = b->thr_uid;
 	int ret = (a_thr_uid > b_thr_uid) - (a_thr_uid < b_thr_uid);
 	if (ret == 0) {
-		uint64_t a_tctx_uid = a->tctx_uid;
-		uint64_t b_tctx_uid = b->tctx_uid;
-		ret = (a_tctx_uid > b_tctx_uid) - (a_tctx_uid < b_tctx_uid);
+		uint64_t a_thr_discrim = a->thr_discrim;
+		uint64_t b_thr_discrim = b->thr_discrim;
+		ret = (a_thr_discrim > b_thr_discrim) - (a_thr_discrim <
+		    b_thr_discrim);
+		if (ret == 0) {
+			uint64_t a_tctx_uid = a->tctx_uid;
+			uint64_t b_tctx_uid = b->tctx_uid;
+			ret = (a_tctx_uid > b_tctx_uid) - (a_tctx_uid <
+			    b_tctx_uid);
+		}
 	}
 	return (ret);
 }
@@ -202,7 +209,7 @@ prof_alloc_rollback(tsd_t *tsd, prof_tctx_t *tctx, bool updated)
 		 */
 		tdata = prof_tdata_get(tsd, true);
 		if (tdata != NULL)
-			prof_sample_threshold_update(tctx->tdata);
+			prof_sample_threshold_update(tdata);
 	}
 
 	if ((uintptr_t)tctx > (uintptr_t)1U) {
@@ -219,7 +226,7 @@ void
 prof_malloc_sample_object(const void *ptr, size_t usize, prof_tctx_t *tctx)
 {
 
-	prof_tctx_set(ptr, tctx);
+	prof_tctx_set(ptr, usize, tctx);
 
 	malloc_mutex_lock(tctx->tdata->lock);
 	tctx->cnts.curobjs++;
@@ -791,6 +798,7 @@ prof_lookup(tsd_t *tsd, prof_bt_t *bt)
 		}
 		ret.p->tdata = tdata;
 		ret.p->thr_uid = tdata->thr_uid;
+		ret.p->thr_discrim = tdata->thr_discrim;
 		memset(&ret.p->cnts, 0, sizeof(prof_cnt_t));
 		ret.p->gctx = gctx;
 		ret.p->tctx_uid = tdata->tctx_uid_next++;
@@ -1569,7 +1577,6 @@ prof_idump(void)
 {
 	tsd_t *tsd;
 	prof_tdata_t *tdata;
-	char filename[PATH_MAX + 1];
 
 	cassert(config_prof);
 
@@ -1585,6 +1592,7 @@ prof_idump(void)
 	}
 
 	if (opt_prof_prefix[0] != '\0') {
+		char filename[PATH_MAX + 1];
 		malloc_mutex_lock(&prof_dump_seq_mtx);
 		prof_dump_filename(filename, 'i', prof_dump_iseq);
 		prof_dump_iseq++;
@@ -1623,7 +1631,6 @@ prof_gdump(void)
 {
 	tsd_t *tsd;
 	prof_tdata_t *tdata;
-	char filename[DUMP_FILENAME_BUFSIZE];
 
 	cassert(config_prof);
 
@@ -1639,6 +1646,7 @@ prof_gdump(void)
 	}
 
 	if (opt_prof_prefix[0] != '\0') {
+		char filename[DUMP_FILENAME_BUFSIZE];
 		malloc_mutex_lock(&prof_dump_seq_mtx);
 		prof_dump_filename(filename, 'u', prof_dump_useq);
 		prof_dump_useq++;
diff --git a/contrib/jemalloc/src/tcache.c b/contrib/jemalloc/src/tcache.c
index 3814365..fdafd0c 100644
--- a/contrib/jemalloc/src/tcache.c
+++ b/contrib/jemalloc/src/tcache.c
@@ -32,7 +32,7 @@ size_t	tcache_salloc(const void *ptr)
 void
 tcache_event_hard(tsd_t *tsd, tcache_t *tcache)
 {
-	index_t binind = tcache->next_gc_bin;
+	szind_t binind = tcache->next_gc_bin;
 	tcache_bin_t *tbin = &tcache->tbins[binind];
 	tcache_bin_info_t *tbin_info = &tcache_bin_info[binind];
 
@@ -72,7 +72,7 @@ tcache_event_hard(tsd_t *tsd, tcache_t *tcache)
 
 void *
 tcache_alloc_small_hard(tsd_t *tsd, arena_t *arena, tcache_t *tcache,
-    tcache_bin_t *tbin, index_t binind)
+    tcache_bin_t *tbin, szind_t binind)
 {
 	void *ret;
 
@@ -87,7 +87,7 @@ tcache_alloc_small_hard(tsd_t *tsd, arena_t *arena, tcache_t *tcache,
 
 void
 tcache_bin_flush_small(tsd_t *tsd, tcache_t *tcache, tcache_bin_t *tbin,
-    index_t binind, unsigned rem)
+    szind_t binind, unsigned rem)
 {
 	arena_t *arena;
 	void *ptr;
@@ -166,7 +166,7 @@ tcache_bin_flush_small(tsd_t *tsd, tcache_t *tcache, tcache_bin_t *tbin,
 }
 
 void
-tcache_bin_flush_large(tsd_t *tsd, tcache_bin_t *tbin, index_t binind,
+tcache_bin_flush_large(tsd_t *tsd, tcache_bin_t *tbin, szind_t binind,
     unsigned rem, tcache_t *tcache)
 {
 	arena_t *arena;
@@ -496,13 +496,13 @@ tcache_boot(void)
 	unsigned i;
 
 	/*
-	 * If necessary, clamp opt_lg_tcache_max, now that arena_maxclass is
+	 * If necessary, clamp opt_lg_tcache_max, now that large_maxclass is
 	 * known.
 	 */
 	if (opt_lg_tcache_max < 0 || (1U << opt_lg_tcache_max) < SMALL_MAXCLASS)
 		tcache_maxclass = SMALL_MAXCLASS;
-	else if ((1U << opt_lg_tcache_max) > arena_maxclass)
-		tcache_maxclass = arena_maxclass;
+	else if ((1U << opt_lg_tcache_max) > large_maxclass)
+		tcache_maxclass = large_maxclass;
 	else
 		tcache_maxclass = (1U << opt_lg_tcache_max);