Make it possible for each thread to manage which arena it is associated
with.

Implement the 'tests' and 'check' build targets.

Move assert() calls up in arena_run_reg_alloc(), so that a corrupt
pointer will likely be caught by an assertion *before* it is
dereferenced.

If multiple threads race to initialize malloc, the loser(s) busy-wait
until initialization is complete.  Add a missing mutex lock so that the
loser(s) properly release the initialization mutex.  Under some
race conditions, this flaw could have caused one or more threads to
become permanently blocked.

Reported by Terrell Magee.

Fix the libunwind version of prof_backtrace() to set the backtrace depth
for all possible code paths.  This fixes the zero-length backtrace
problem when using libunwind.

When heap profiling is enabled but deactivated, there is no need to call
isalloc(ptr) in prof_{malloc,realloc}().  Avoid these calls, so that
profiling overhead under such conditions is negligible.

If there is more than one arena, initialize next_arena so that the
first and second threads to allocate memory use arenas 0 and 1, rather
than both using arena 0.

Add MAP_NORESERVE to the chunk_mmap() case being used by
chunk_swap_enable(), if the system supports it.

Use the size argument to tcache_dalloc_large() to control the number of
bytes set to 0x5a when junk filling is enabled, rather than accessing a
non-existent arena bin.  This bug was capable of corrupting an
arbitrarily large memory region, depending on what followed the arena
data structure in memory (typically zeroed memory, another arena_t, or a
red-black tree node for a huge object).

Properly maintain tcache_bin_t's avail pointer such that it is NULL if
no objects are cached.  This only caused problems during thread cache
destruction, since cache flushing otherwise never occurs on an empty
bin.

Properly set the context associated with each allocated object, even
when the object is not sampled.

Remove debug print code that slipped in.

Fix arena_chunk_dealloc() to put the new spare in a consistent state before
dropping the arena mutex to deallocate the previous spare.

Fix arena_run_dalloc() to insert a newly dirtied chunk into the
chunks_dirty list before potentially deallocating the chunk, so that dirty
page accounting is self-consistent.

Initialize bt2cnt_tsd so that cleanup at thread exit actually happens.

Associate (prof_ctx_t *) with allocated objects, rather than
(prof_thr_cnt_t *).  Each thread must always operate on its own
(prof_thr_cnt_t *), and an object may outlive the thread that allocated it.

Fix a compilation error due to stale data structure access code in
tcache_dalloc_large() for junk filling.

Generalize ExtractSymbols to handle all cases of library address overlap
with the main binary.

Linux kernels have been capable of concurrent page table access since
2.6.27, so this hack is not necessary for modern kernels.

Now that JEMALLOC_OPTIONS=P isn't the only way to cause stats_print() to
be called, opt_stats_print must actually be checked when reporting the
state of the P/p option.

Don't build with -march=native by default, because the generated code
may perform especially poorly on ABI-compatible, but internally
different, systems.

Split library build rules up so that parallel building works.

Fix autoconf-related dependencies.

Remove obsolete JEMALLOC_VERSION definition.

Iterated downward through both libraries and PCs.  This allows PCs
to resolve even when library address ranges overlap.

Remove a duplicate prof_leave() call in an error path through
prof_dump().

Modify ExtractSymbols to operate on sorted PCs and libraries, in order
to reduce computational complexity from O(N*M) to O(N+M).

Fix divide-by-zero error in pprof.  It is possible for sample contexts
to currently have no associated objects, but the cumulative statistics
are still useful, depending on how the user invokes pprof.  Since
jemalloc intentionally does not filter such contexts, take care not to
divide by 0 when re-scaling for v2 heap sampling.

Install pprof as part of 'make install'.

Update pprof documentation.

Leak reporting is useful even if sampling is enabled; some leaks may not
be reported, but those reported are still genuine leaks.

Add the E/e options to control whether the application starts with
sampling active/inactive (secondary control to F/f).  Add the
prof.active mallctl so that the application can activate/deactivate
sampling on the fly.

Make it possible to disable interval-triggered profile dumping, even if
profiling is enabled.  This is useful if the user only wants a single
dump at exit, or if the application manually triggers profile dumps.

If the mean heap sampling interval is larger than one page, simulate
sampled small objects with large objects.  This allows profiling context
pointers to be omitted for small objects.  As a result, the memory
overhead for sampling decreases as the sampling interval is increased.

Fix a compilation error in the profiling code.

Properly set/clear CHUNK_MAP_ZEROED for all purged pages, according to
whether the pages are (potentially) file-backed or anonymous.  This was
merely a performance pessimization for the anonymous mapping case, but
was a calloc()-related bug for the swap_enabled case.

Split arena->runs_avail into arena->runs_avail_{clean,dirty}, and
preferentially allocate dirty runs.

Remove medium size classes, because concurrent dirty page purging is
no longer capable of purging inactive dirty pages inside active runs
(due to recent arena/bin locking changes).

Enhance tcache to support caching large objects, so that the same range
of size classes is still cached, despite the removal of medium size
class support.

Initialize small run header before dropping arena->lock,
arena_chunk_purge() relies on valid small run headers during run
iteration.

Add some assertions.

Check for interior pointers in arena_[ds]alloc().

Check for corrupt pointers in tcache_alloc().

Update arena->nactive when pseudo-allocating runs in
arena_chunk_purge(), since arena_run_dalloc() subtracts from
arena->nactive.