Merge branch 'dev'

Author: jasone

ChangeLog

@@ -5,6 +5,30 @@ found in the git revision history:
 
     https://github.com/jemalloc/jemalloc
 
+* 3.6.0 (March 31, 2014)
+
+  This version contains a critical bug fix for a regression present in 3.5.0 and
+  3.5.1.
+
+  Bug fixes:
+  - Fix a regression in arena_chunk_alloc() that caused crashes during
+    small/large allocation if chunk allocation failed.  In the absence of this
+    bug, chunk allocation failure would result in allocation failure, e.g.  NULL
+    return from malloc().  This regression was introduced in 3.5.0.
+  - Fix backtracing for gcc intrinsics-based backtracing by specifying
+    -fno-omit-frame-pointer to gcc.  Note that the application (and all the
+    libraries it links to) must also be compiled with this option for
+    backtracing to be reliable.
+  - Use dss allocation precedence for huge allocations as well as small/large
+    allocations.
+  - Fix test assertion failure message formatting.  This bug did not manifect on
+    x86_64 systems because of implementation subtleties in va_list.
+  - Fix inconsequential test failures for hash and SFMT code.
+
+  New features:
+  - Support heap profiling on FreeBSD.  This feature depends on the proc
+    filesystem being mounted during heap profile dumping.
+
 * 3.5.1 (February 25, 2014)
 
   This version primarily addresses minor bugs in test code.

Makefile.in

@@ -141,7 +141,7 @@ TESTS_INTEGRATION := $(srcroot)test/integration/aligned_alloc.c \
 	$(srcroot)test/integration/xallocx.c
 ifeq ($(enable_experimental), 1)
 TESTS_INTEGRATION += $(srcroot)test/integration/allocm.c \
-	$(srcroot)test/integration/ALLOCM_ARENA.c \
+	$(srcroot)test/integration/MALLOCX_ARENA.c \
 	$(srcroot)test/integration/rallocm.c
 endif
 TESTS_STRESS :=

bin/pprof

@@ -4197,8 +4197,12 @@ sub FindLibrary {
 # For libc libraries, the copy in /usr/lib/debug contains debugging symbols
 sub DebuggingLibrary {
   my $file = shift;
-  if ($file =~ m|^/| && -f "/usr/lib/debug$file") {
-    return "/usr/lib/debug$file";
+  if ($file =~ m|^/|) {
+      if (-f "/usr/lib/debug$file") {
+        return "/usr/lib/debug$file";
+      } elsif (-f "/usr/lib/debug$file.debug") {
+        return "/usr/lib/debug$file.debug";
+      }
   }
   return undef;
 }
@@ -4360,6 +4364,19 @@ sub ParseLibraries {
       $finish = HexExtend($2);
       $offset = $zero_offset;
       $lib = $3;
+    }
+    # FreeBSD 10.0 virtual memory map /proc/curproc/map as defined in
+    # function procfs_doprocmap (sys/fs/procfs/procfs_map.c)
+    #
+    # Example:
+    # 0x800600000 0x80061a000 26 0 0xfffff800035a0000 r-x 75 33 0x1004 COW NC vnode /libexec/ld-elf.s
+    # o.1 NCH -1
+    elsif ($l =~ /^(0x$h)\s(0x$h)\s\d+\s\d+\s0x$h\sr-x\s\d+\s\d+\s0x\d+\s(COW|NCO)\s(NC|NNC)\svnode\s(\S+\.so(\.\d+)*)/) {
+      $start = HexExtend($1);
+      $finish = HexExtend($2);
+      $offset = $zero_offset;
+      $lib = FindLibrary($5);
+
     } else {
       next;
     }
@@ -4382,6 +4399,7 @@ sub ParseLibraries {
       }
     }
 
+    if($main::opt_debug) { printf STDERR "$start:$finish ($offset) $lib\n"; }
     push(@{$result}, [$lib, $start, $finish, $offset]);
   }
 
@@ -4589,6 +4607,12 @@ sub ExtractSymbols {
     my $finish = $lib->[2];
     my $offset = $lib->[3];
 
+    # Use debug library if it exists
+    my $debug_libname = DebuggingLibrary($libname);
+    if ($debug_libname) {
+        $libname = $debug_libname;
+    }
+
     # Get list of pcs that belong in this library.
     my $contained = [];
     my ($start_pc_index, $finish_pc_index);
@@ -5019,7 +5043,7 @@ sub GetProcedureBoundariesViaNm {
 
       # Tag this routine with the starting address in case the image
       # has multiple occurrences of this routine.  We use a syntax
-      # that resembles template paramters that are automatically
+      # that resembles template parameters that are automatically
       # stripped out by ShortFunctionName()
       $this_routine .= "<$start_val>";
 

configure.ac

@@ -150,6 +150,11 @@ if test "x$EXTRA_CFLAGS" != "x" ; then
 fi
 AC_PROG_CPP
 
+AC_C_BIGENDIAN([ac_cv_big_endian=1], [ac_cv_big_endian=0])
+if test "x${ac_cv_big_endian}" = "x1" ; then
+  AC_DEFINE_UNQUOTED([JEMALLOC_BIG_ENDIAN], [ ])
+fi
+
 AC_CHECK_SIZEOF([void *])
 if test "x${ac_cv_sizeof_void_p}" = "x8" ; then
   LG_SIZEOF_PTR=3
@@ -742,22 +747,6 @@ if test "x$backtrace_method" = "x" -a "x$enable_prof_libgcc" = "x1" \
      -a "x$GCC" = "xyes" ; then
   AC_CHECK_HEADERS([unwind.h], , [enable_prof_libgcc="0"])
   AC_CHECK_LIB([gcc], [_Unwind_Backtrace], [LIBS="$LIBS -lgcc"], [enable_prof_libgcc="0"])
-  dnl The following is conservative, in that it only has entries for CPUs on
-  dnl which jemalloc has been tested.
-  AC_MSG_CHECKING([libgcc-based backtracing reliability on ${host_cpu}])
-  case "${host_cpu}" in
-    i[[3456]]86)
-      AC_MSG_RESULT([unreliable])
-      enable_prof_libgcc="0";
-      ;;
-    x86_64)
-      AC_MSG_RESULT([reliable])
-      ;;
-    *)
-      AC_MSG_RESULT([unreliable])
-      enable_prof_libgcc="0";
-      ;;
-  esac
   if test "x${enable_prof_libgcc}" = "x1" ; then
     backtrace_method="libgcc"
     AC_DEFINE([JEMALLOC_PROF_LIBGCC], [ ])
@@ -779,6 +768,7 @@ fi
 )
 if test "x$backtrace_method" = "x" -a "x$enable_prof_gcc" = "x1" \
      -a "x$GCC" = "xyes" ; then
+  JE_CFLAGS_APPEND([-fno-omit-frame-pointer])
   backtrace_method="gcc intrinsics"
   AC_DEFINE([JEMALLOC_PROF_GCC], [ ])
 else

doc/jemalloc.xml.in

@@ -1439,8 +1439,12 @@ malloc_conf = "xmalloc:true";]]></programlisting>
         <listitem><para>Set the precedence of dss allocation as related to mmap
         allocation for arena &lt;i&gt;, or for all arenas if &lt;i&gt; equals
         <link
-        linkend="arenas.narenas"><mallctl>arenas.narenas</mallctl></link>.  See
-        <link linkend="opt.dss"><mallctl>opt.dss</mallctl></link> for supported
+        linkend="arenas.narenas"><mallctl>arenas.narenas</mallctl></link>.  Note
+        that even during huge allocation this setting is read from the arena
+        that would be chosen for small or large allocation so that applications
+        can depend on consistent dss versus mmap allocation regardless of
+        allocation size.  See <link
+        linkend="opt.dss"><mallctl>opt.dss</mallctl></link> for supported
         settings.
         </para></listitem>
       </varlistentry>

include/jemalloc/internal/hash.h

@@ -320,7 +320,7 @@ hash_x64_128(const void *key, const int len, const uint32_t seed,
 JEMALLOC_INLINE void
 hash(const void *key, size_t len, const uint32_t seed, size_t r_hash[2])
 {
-#if (LG_SIZEOF_PTR == 3)
+#if (LG_SIZEOF_PTR == 3 && !defined(JEMALLOC_BIG_ENDIAN))
 	hash_x64_128(key, len, seed, (uint64_t *)r_hash);
 #else
 	uint64_t hashes[2];

include/jemalloc/internal/huge.h

@@ -17,18 +17,20 @@ extern size_t		huge_allocated;
 /* Protects chunk-related data structures. */
 extern malloc_mutex_t	huge_mtx;
 
-void	*huge_malloc(size_t size, bool zero);
-void	*huge_palloc(size_t size, size_t alignment, bool zero);
+void	*huge_malloc(size_t size, bool zero, dss_prec_t dss_prec);
+void	*huge_palloc(size_t size, size_t alignment, bool zero,
+    dss_prec_t dss_prec);
 bool	huge_ralloc_no_move(void *ptr, size_t oldsize, size_t size,
     size_t extra);
 void	*huge_ralloc(void *ptr, size_t oldsize, size_t size, size_t extra,
-    size_t alignment, bool zero, bool try_tcache_dalloc);
+    size_t alignment, bool zero, bool try_tcache_dalloc, dss_prec_t dss_prec);
 #ifdef JEMALLOC_JET
 typedef void (huge_dalloc_junk_t)(void *, size_t);
 extern huge_dalloc_junk_t *huge_dalloc_junk;
 #endif
 void	huge_dalloc(void *ptr, bool unmap);
 size_t	huge_salloc(const void *ptr);
+dss_prec_t	huge_dss_prec_get(arena_t *arena);
 prof_ctx_t	*huge_prof_ctx_get(const void *ptr);
 void	huge_prof_ctx_set(const void *ptr, prof_ctx_t *ctx);
 bool	huge_boot(void);

include/jemalloc/internal/jemalloc_internal.h.in

@@ -770,7 +770,7 @@ imalloct(size_t size, bool try_tcache, arena_t *arena)
 	if (size <= arena_maxclass)
 		return (arena_malloc(arena, size, false, try_tcache));
 	else
-		return (huge_malloc(size, false));
+		return (huge_malloc(size, false, huge_dss_prec_get(arena)));
 }
 
 JEMALLOC_ALWAYS_INLINE void *
@@ -787,7 +787,7 @@ icalloct(size_t size, bool try_tcache, arena_t *arena)
 	if (size <= arena_maxclass)
 		return (arena_malloc(arena, size, true, try_tcache));
 	else
-		return (huge_malloc(size, true));
+		return (huge_malloc(size, true, huge_dss_prec_get(arena)));
 }
 
 JEMALLOC_ALWAYS_INLINE void *
@@ -813,9 +813,9 @@ ipalloct(size_t usize, size_t alignment, bool zero, bool try_tcache,
 			ret = arena_palloc(choose_arena(arena), usize,
 			    alignment, zero);
 		} else if (alignment <= chunksize)
-			ret = huge_malloc(usize, zero);
+			ret = huge_malloc(usize, zero, huge_dss_prec_get(arena));
 		else
-			ret = huge_palloc(usize, alignment, zero);
+			ret = huge_palloc(usize, alignment, zero, huge_dss_prec_get(arena));
 	}
 
 	assert(ALIGNMENT_ADDR2BASE(ret, alignment) == ret);
@@ -984,7 +984,7 @@ iralloct(void *ptr, size_t size, size_t extra, size_t alignment, bool zero,
 		    try_tcache_dalloc));
 	} else {
 		return (huge_ralloc(ptr, oldsize, size, extra,
-		    alignment, zero, try_tcache_dalloc));
+		    alignment, zero, try_tcache_dalloc, huge_dss_prec_get(arena)));
 	}
 }
 

include/jemalloc/internal/jemalloc_internal_defs.h.in

@@ -190,6 +190,9 @@
 /* C99 restrict keyword supported. */
 #undef JEMALLOC_HAS_RESTRICT
 
+/* For use by hash code. */
+#undef JEMALLOC_BIG_ENDIAN
+
 /* sizeof(int) == 2^LG_SIZEOF_INT. */
 #undef LG_SIZEOF_INT
 

include/jemalloc/internal/private_symbols.txt

@@ -197,6 +197,7 @@ huge_allocated
 huge_boot
 huge_dalloc
 huge_dalloc_junk
+huge_dss_prec_get
 huge_malloc
 huge_mtx
 huge_ndalloc