Commit Graph

341 Commits

Author SHA1 Message Date
David Goldblatt
982c10de35 TSD: Make all state access happen through a function.
Shortly, tsd state will be atomic and have some complicated enough logic down
the state-setting path that we should be aware of it.
2018-05-18 11:43:03 -07:00
Qi Wang
b293a3eb86 Fix the max_background_thread test.
We may set number of background threads separately, e.g. through
--with-malloc-conf, so avoid assuming the default number in the test.
2018-05-15 14:00:51 -07:00
Qi Wang
312352faa8 Fix background thread index issues with max_background_threads. 2018-05-15 12:25:23 -07:00
Latchesar Ionkov
a32b7bd567 Mallctl: Add arenas.lookup
Implement a new mallctl operation that allows looking up the arena a
region of memory belongs to.
2018-05-01 13:14:36 -07:00
Dave Watson
8b14f3abc0 background_thread: add max thread count config
Looking at the thread counts in our services, jemalloc's background thread
is useful, but mostly idle.  Add a config option to tune down the number of threads.
2018-04-10 14:01:45 -07:00
Qi Wang
cf2f4aac1c Fix const qualifier warnings. 2018-04-09 16:50:30 -07:00
Qi Wang
d3e0976a2c Fix type warning on Windows.
Add cast since read / write has unsigned return type on windows.
2018-04-09 16:50:30 -07:00
Qi Wang
21eb0d15a6 Fix a background_thread shutdown issue.
1) make sure background thread 0 is always created; and 2) fix synchronization
between thread 0 and the control thread.
2018-04-02 10:03:47 -07:00
David Goldblatt
ebe0b5f828 Emitter: Add support for row-based output in table mode.
This is needed for things like mutex stats in table mode.
2018-03-09 11:47:17 -08:00
David Goldblatt
27a8fe6780 Introduce the emitter module.
The emitter can be used to produce structured json or tabular output.  For now
it has no uses; in subsequent commits, I'll begin transitioning stats printing
code over.
2018-03-09 11:47:17 -08:00
Qi Wang
e4f090e8df Add opt.thp which allows explicit hugepage usage.
"always" marks all user mappings as MADV_HUGEPAGE; while "never" marks all
mappings as MADV_NOHUGEPAGE. The default setting "default" does not change any
settings.  Note that all the madvise calls are part of the default extent hooks
by design, so that customized extent hooks have complete control over the
mappings including hugepage settings.
2018-03-08 13:08:06 -08:00
David Goldblatt
26b1c13982 Background threads: fix an indexing bug.
We have a buffer overrun that manifests in the case where arena indices higher
than the number of CPUs are accessed before arena indices lower than the number
of CPUs.  This fixes the bug and adds a test.
2018-02-27 19:43:05 -08:00
David Goldblatt
21f7c13d0b Add the div module, which allows fast division by dynamic values. 2017-12-21 14:25:43 -08:00
David T. Goldblatt
4bf4a1c4ea Pull out arena_bin_info_t and arena_bin_t into their own file.
In the process, kill arena_bin_index, which is unused.  To follow are several
diffs continuing this separation.
2017-12-18 16:29:10 -08:00
Qi Wang
f70785de91 Skip test/unit/pack when profiling is enabled.
The test assumes no sampled allocations.
2017-12-18 12:47:46 -08:00
Qi Wang
fac706836f Add opt.lg_extent_max_active_fit
When allocating from dirty extents (which we always prefer if available), large
active extents can get split even if the new allocation is much smaller, in
which case the introduced fragmentation causes high long term damage.  This new
option controls the threshold to reuse and split an existing active extent.  We
avoid using a large extent for much smaller sizes, in order to reduce
fragmentation.  In some workload, adding the threshold improves virtual memory
usage by >10x.
2017-11-16 15:32:02 -08:00
Qi Wang
e422fa8e7e Add arena.i.retain_grow_limit
This option controls the max size when grow_retained.  This is useful when we
have customized extent hooks reserving physical memory (e.g. 1G huge pages).
Without this feature, the default increasing sequence could result in fragmented
and wasted physical memory.
2017-11-03 13:53:33 -07:00
David Goldblatt
d14bbf8d81 Add a "dumpable" bit to the extent state.
Currently, this is unused (i.e. all extents are always marked dumpable).  In the
future, we'll begin using this functionality.
2017-10-16 15:35:49 -07:00
David Goldblatt
8a7ee3014c Logging: capitalize log macro.
Dodge a name-conflict with the math.h logarithm function. D'oh.
2017-10-02 20:44:43 -07:00
Qi Wang
886053b966 Fix huge page test in test/unit/pages.
Huge pages could be disabled even if the kernel header has MAD_HUGEPAGE
defined.  Guard the huge pagetest with runtime detection.
2017-09-12 14:29:49 -07:00
Qi Wang
e55c3ca267 Add stats for metadata_thp.
Report number of THPs used in arena and aggregated stats.
2017-08-30 16:47:32 -07:00
Qi Wang
47b20bb654 Change opt.metadata_thp to [disabled,auto,always].
To avoid the high RSS caused by THP + low usage arena (i.e. THP becomes a
significant percentage), added a new "auto" option which will only start using
THP after a base allocator used up the first THP region.  Starting from the
second hugepage (in a single arena), "auto" behaves the same as "always",
i.e. madvise hugepage right away.
2017-08-30 16:47:32 -07:00
Qi Wang
b0825351d9 Add missing mallctl unit test for abort_conf.
The abort_conf option was missed from test/unit/mallctl.
2017-08-11 22:58:58 -07:00
Qi Wang
3ec279ba1c Fix test/unit/pages.
As part of the metadata_thp support, We now have a separate swtich
(JEMALLOC_HAVE_MADVISE_HUGE) for MADV_HUGEPAGE availability.  Use that instead
of JEMALLOC_THP (which doesn't guard pages_huge anymore) in tests.
2017-08-11 15:57:12 -07:00
Qi Wang
8fdd9a5797 Implement opt.metadata_thp
This option enables transparent huge page for base allocators (require
MADV_HUGEPAGE support).
2017-08-11 14:51:20 -07:00
David Goldblatt
e6aeceb606 Logging: log using the log var names directly.
Currently we have to log by writing something like:

  static log_var_t log_a_b_c = LOG_VAR_INIT("a.b.c");
  log (log_a_b_c, "msg");

This is sort of annoying.  Let's just write:

  log("a.b.c", "msg");
2017-07-24 14:55:54 -07:00
David Goldblatt
a9f7732d45 Logging: allow logging with empty varargs.
Currently, the log macro requires at least one argument after the format string,
because of the way the preprocessor handles varargs macros.  We can hide some of
that irritation by pushing the extra arguments into a varargs function.
2017-07-22 09:38:19 -07:00
David T. Goldblatt
9761b449c8 Add a logging facility.
This sets up a hierarchical logging facility, so that we can add logging
statements liberally, and turn them on in a fine-grained manner.
2017-07-20 17:58:37 -07:00
David Goldblatt
fb6787a78c Add a test of behavior under multi-threaded forking.
Forking a multithreaded process is dangerous but allowed, so long as the child
only executes async-signal-safe functions (e.g. exec).  Add a test to ensure
that we don't break this behavior.
2017-07-10 18:17:12 -07:00
Qi Wang
d6eb8ac8f3 Set reentrancy when invoking customized extent hooks.
Customized extent hooks may malloc / free thus trigger reentry.  Support this
behavior by adding reentrancy on hook functions.
2017-06-23 13:27:53 -07:00
Qi Wang
813643c6a7 Prevent background threads from running in post_reset().
We lookup freed extents for testing in post_reset.  Take background_thread lock
so that the extents are not modified at the same time.
2017-06-12 08:56:14 -07:00
Qi Wang
394df9519d Combine background_thread started / paused into state. 2017-06-12 08:56:14 -07:00
Qi Wang
00869e39a3 Make tsd no-cleanup during tsd reincarnation.
Since tsd cleanup isn't guaranteed when reincarnated, we set up tsd in a way
that needs no cleanup, by making it going through slow path instead.
2017-06-07 11:03:49 -07:00
Jason Evans
fd0fa003e1 Test with background_thread:true.
Add testing for background_thread:true, and condition a xallocx() -->
rallocx() escalation assertion to allow for spurious in-place rallocx()
following xallocx() failure.
2017-06-01 08:55:27 -07:00
Jason Evans
b511232fcd Refactor/fix background_thread/percpu_arena bootstrapping.
Refactor bootstrapping such that dlsym() is called during the
bootstrapping phase that can tolerate reentrant allocation.
2017-06-01 08:55:27 -07:00
Jason Evans
596b479d83 Skip default tcache testing if !opt_tcache. 2017-06-01 08:55:27 -07:00
David Goldblatt
8261e581be Header refactoring: Pull size helpers out of jemalloc module. 2017-05-31 13:08:45 -07:00
David Goldblatt
98774e64a4 Header refactoring: unify and de-catchall extent_mmap module. 2017-05-31 13:08:45 -07:00
David Goldblatt
44f9bd147a Header refactoring: unify and de-catchall rtree module. 2017-05-31 13:08:45 -07:00
Jason Evans
c606a87d2a Add the --disable-thp option to support cross compiling.
This resolves #669.
2017-05-30 11:30:54 -07:00
Jason Evans
4f0963b883 Add test for excessive retained memory. 2017-05-29 17:27:18 -07:00
Qi Wang
49505e558b Make test/unit/background_thread not flaky. 2017-05-26 21:15:15 -07:00
Qi Wang
927239b910 Cleanup smoothstep.sh / .h.
h_step_sum was used to compute moving sum.  Not in use anymore.
2017-05-25 16:52:10 -07:00
David Goldblatt
9f822a1fd7 Header refactoring: unify and de-catchall witness code. 2017-05-24 15:27:30 -07:00
Qi Wang
2c368284d2 Add tests for background threads. 2017-05-23 12:26:20 -07:00
Qi Wang
2bee0c6251 Add background thread related stats. 2017-05-23 12:26:20 -07:00
Qi Wang
b693c7868e Implementing opt.background_thread.
Added opt.background_thread to enable background threads, which handles purging
currently.  When enabled, decay ticks will not trigger purging (which will be
left to the background threads).  We limit the max number of threads to NCPUs.
When percpu arena is enabled, set CPU affinity for the background threads as
well.

The sleep interval of background threads is dynamic and determined by computing
number of pages to purge in the future (based on backlog).
2017-05-23 12:26:20 -07:00
David Goldblatt
3f685e8824 Protect the rtree/extent interactions with a mutex pool.
Instead of embedding a lock bit in rtree leaf elements, we associate extents
with a small set of mutexes.  This gets us two things:

- We can use the system mutexes.  This (hypothetically) protects us from
  priority inversion, and lets us stop doing a backoff/sleep loop, instead
  opting for precise wakeups from the mutex.
- Cuts down on the number of mutex acquisitions we have to do (from 4 in the
  worst case to two).

We end up simplifying most of the rtree code (which no longer has to deal with
locking or concurrency at all), at the cost of additional complexity in the
extent code: since the mutex protecting the rtree leaf elements is determined by
reading the extent out of those elements, the initial read is racy, so that we
may acquire an out of date mutex.  We re-check the extent in the leaf after
acquiring the mutex to protect us from this race.
2017-05-19 14:21:27 -07:00
Jason Evans
6e62c62862 Refactor *decay_time into *decay_ms.
Support millisecond resolution for decay times.  Among other use cases
this makes it possible to specify a short initial dirty-->muzzy decay
phase, followed by a longer muzzy-->clean decay phase.

This resolves #812.
2017-05-18 11:33:45 -07:00
David Goldblatt
209f2926b8 Header refactoring: tsd - cleanup and dependency breaking.
This removes the tsd macros (which are used only for tsd_t in real builds).  We
break up the circular dependencies involving tsd.

We also move all tsd access through getters and setters.  This allows us to
assert that we only touch data when tsd is in a valid state.

We simplify the usages of the x macro trick, removing all the customizability
(get/set, init, cleanup), moving the lifetime logic to tsd_init and tsd_cleanup.
This lets us make initialization order independent of order within tsd_t.
2017-05-01 10:49:56 -07:00
Jason Evans
c86c8f4ffb Add extent_destroy_t and use it during arena destruction.
Add the extent_destroy_t extent destruction hook to extent_hooks_t, and
use it during arena destruction.  This hook explicitly communicates to
the callee that the extent must be destroyed or tracked for later reuse,
lest it be permanently leaked.  Prior to this change, retained extents
could unintentionally be leaked if extent retention was enabled.

This resolves #560.
2017-04-29 09:24:12 -07:00
Jason Evans
b9ab04a191 Refactor !opt.munmap to opt.retain. 2017-04-29 09:24:12 -07:00
David Goldblatt
dab4beb277 Header refactoring: hash - unify and remove from catchall. 2017-04-25 09:51:38 -07:00
Jason Evans
c67c3e4a63 Replace --disable-munmap with opt.munmap.
Control use of munmap(2) via a run-time option rather than a
compile-time option (with the same per platform default).  The old
behavior of --disable-munmap can be achieved with
--with-malloc-conf=munmap:false.

This partially resolves #580.
2017-04-24 20:37:16 -07:00
David Goldblatt
bf2dc7e678 Header refactoring: ticker module - remove from the catchall and unify. 2017-04-24 10:33:21 -07:00
Jason Evans
b2a8453a3f Remove --disable-tls.
This option is no longer useful, because TLS is correctly configured
automatically on all supported platforms.

This partially resolves #580.
2017-04-21 11:12:29 -07:00
Jason Evans
4403c9ab44 Remove --disable-tcache.
Simplify configuration by removing the --disable-tcache option, but
replace the testing for that configuration with
--with-malloc-conf=tcache:false.

Fix the thread.arena and thread.tcache.flush mallctls to work correctly
if tcache is disabled.

This partially resolves #580.
2017-04-21 10:06:12 -07:00
Jason Evans
da4cff0279 Support --with-lg-page values larger than system page size.
All mappings continue to be PAGE-aligned, even if the system page size
is smaller.  This change is primarily intended to provide a mechanism
for supporting multiple page sizes with the same binary; smaller page
sizes work better in conjunction with jemalloc's design.

This resolves #467.
2017-04-18 19:01:04 -07:00
Jason Evans
45f087eb03 Revert "Remove BITMAP_USE_TREE."
Some systems use a native 64 KiB page size, which means that the bitmap
for the smallest size class can be 8192 bits, not just 512 bits as when
the page size is 4 KiB.  Linear search in bitmap_{sfu,ffu}() is
unacceptably slow for such large bitmaps.

This reverts commit 7c00f04ff4.
2017-04-18 19:01:04 -07:00
David Goldblatt
f692e6c214 Header refactoring: move util.h out of the catchall 2017-04-18 18:35:03 -07:00
David Goldblatt
0b00ffe55f Header refactoring: move bit_util.h out of the catchall 2017-04-18 18:35:03 -07:00
Jason Evans
76b35f4b2f Track extent structure serial number (esn) in extent_t.
This enables stable sorting of extent_t structures.
2017-04-17 14:47:45 -07:00
Qi Wang
ccfe68a916 Pass alloc_ctx down profiling path.
With this change, when profiling is enabled, we avoid doing redundant rtree
lookups. Also changed dalloc_atx_t to alloc_atx_t, as it's now used on
allocation path as well (to speed up profiling).
2017-04-12 13:55:39 -07:00
David Goldblatt
0237870c60 Header refactoring: break out ql.h dependencies 2017-04-11 11:52:30 -07:00
David Goldblatt
610cb83419 Header refactoring: break out qr.h dependencies 2017-04-11 11:52:30 -07:00
David Goldblatt
63a5cd4cc2 Header refactoring: break out rb.h dependencies 2017-04-11 11:52:30 -07:00
David Goldblatt
2f00ce4da7 Header refactoring: break out ph.h dependencies 2017-04-11 11:52:30 -07:00
David Goldblatt
b407a65401 Add basic reentrancy-checking support, and allow arena_new to reenter.
This checks whether or not we're reentrant using thread-local data, and, if we
are, moves certain internal allocations to use arena 0 (which should be properly
initialized after bootstrapping).

The immediate thing this allows is spinning up threads in arena_new, which will
enable spinning up background threads there.
2017-04-07 14:10:27 -07:00
David Goldblatt
0a0fcd3e6a Add hooking functionality
This allows us to hook chosen functions and do interesting things there (in
particular: reentrancy checking).
2017-04-07 14:10:27 -07:00
Qi Wang
fde3e20cc0 Integrate auto tcache into TSD.
The embedded tcache is initialized upon tsd initialization.  The avail arrays
for the tbins will be allocated / deallocated accordingly during init / cleanup.

With this change, the pointer to the auto tcache will always be available, as
long as we have access to the TSD.  tcache_available() (called in tcache_get())
is provided to check if we should use tcache.
2017-04-07 09:55:14 -07:00
David Goldblatt
eeabdd2466 Remove the pre-C11-atomics API, which is now unused 2017-04-05 16:25:37 -07:00
David Goldblatt
7da04a6b09 Convert prng module to use C11-style atomics 2017-04-04 16:45:52 -07:00
Qi Wang
d3cda3423c Do proper cleanup for tsd_state_reincarnated.
Also enable arena_bind under non-nominal state, as the cleanup will be handled
correctly now.
2017-04-04 00:34:49 -07:00
Qi Wang
9ed84b0d45 Add init function support to tsd members.
This will facilitate embedding tcache into tsd, which will require proper
initialization cannot be done via the static initializer.  Make tsd->rtree_ctx
to be initialized via rtree_ctx_data_init().
2017-04-04 00:19:21 -07:00
Jason Evans
7c00f04ff4 Remove BITMAP_USE_TREE.
Remove tree-structured bitmap support, in order to reduce complexity and
ease maintenance.  No bitmaps larger than 512 bits have been necessary
since before 4.0.0, and there is no current plan that would increase
maximum bitmap size.  Although tree-structured bitmaps were used on
32-bit platforms prior to this change, the overall benefits were
questionable (higher metadata overhead, higher bitmap modification cost,
marginally lower search cost).
2017-03-27 12:18:40 -07:00
Jason Evans
6258176c87 Fix bitmap_ffu() to work with 3+ levels. 2017-03-27 12:18:40 -07:00
Jason Evans
5e12223925 Fix BITMAP_USE_TREE version of bitmap_ffu().
This fixes an extent searching regression on 32-bit systems, caused by
the initial bitmap_ffu() implementation in
c8021d01f6 (Implement bitmap_ffu(), which
finds the first unset bit.), as first used in
5d33233a5e (Use a bitmap in extents_t to
speed up search.).
2017-03-25 23:29:32 -07:00
Jason Evans
c8021d01f6 Implement bitmap_ffu(), which finds the first unset bit. 2017-03-24 17:52:46 -07:00
Qi Wang
bd2006a41b Added JSON output for lock stats.
Also added option 'x' to malloc_stats() to bypass lock section.
2017-03-23 00:03:28 -07:00
Jason Evans
5e67fbc367 Push down iealloc() calls.
Call iealloc() as deep into call chains as possible without causing
redundant calls.
2017-03-22 18:33:32 -07:00
Jason Evans
ce41ab0c57 Embed root node into rtree_t.
This avoids one atomic operation per tree access.
2017-03-22 18:33:32 -07:00
Jason Evans
99d68445ef Incorporate szind/slab into rtree leaves.
Expand and restructure the rtree API such that all common operations can
be achieved with minimal work, regardless of whether the rtree leaf
fields are independent versus packed into a single atomic pointer.
2017-03-22 18:33:32 -07:00
Jason Evans
944c8a3383 Split rtree_elm_t into rtree_{node,leaf}_elm_t.
This allows leaf elements to differ in size from internal node elements.

In principle it would be more correct to use a different type for each
level of the tree, but due to implementation details related to atomic
operations, we use casts anyway, thus counteracting the value of
additional type correctness.  Furthermore, such a scheme would require
function code generation (via cpp macros), as well as either unwieldy
type names for leaves or type aliases, e.g.

  typedef struct rtree_elm_d2_s rtree_leaf_elm_t;

This alternate strategy would be more correct, and with less code
duplication, but probably not worth the complexity.
2017-03-22 18:33:32 -07:00
Jason Evans
e8921cf2eb Convert extent_t's usize to szind.
Rather than storing usize only for large (and prof-promoted)
allocations, store the size class index for allocations that reside
within the extent, such that the size class index is valid for all
extents that contain extant allocations, and invalid otherwise (mainly
to make debugging simpler).
2017-03-22 18:33:32 -07:00
Jason Evans
64e458f5cd Implement two-phase decay-based purging.
Split decay-based purging into two phases, the first of which uses lazy
purging to convert dirty pages to "muzzy", and the second of which uses
forced purging, decommit, or unmapping to convert pages to clean or
destroy them altogether.  Not all operating systems support lazy
purging, yet the application may provide extent hooks that implement
lazy purging, so care must be taken to dynamically omit the first phase
when necessary.

The mallctl interfaces change as follows:
- opt.decay_time --> opt.{dirty,muzzy}_decay_time
- arena.<i>.decay_time --> arena.<i>.{dirty,muzzy}_decay_time
- arenas.decay_time --> arenas.{dirty,muzzy}_decay_time
- stats.arenas.<i>.pdirty --> stats.arenas.<i>.p{dirty,muzzy}
- stats.arenas.<i>.{npurge,nmadvise,purged} -->
  stats.arenas.<i>.{dirty,muzzy}_{npurge,nmadvise,purged}

This resolves #521.
2017-03-15 13:13:47 -07:00
Qi Wang
ec532e2c5c Implement per-CPU arena.
The new feature, opt.percpu_arena, determines thread-arena association
dynamically based CPU id. Three modes are supported: "percpu", "phycpu"
and disabled.

"percpu" uses the current core id (with help from sched_getcpu())
directly as the arena index, while "phycpu" will assign threads on the
same physical CPU to the same arena. In other words, "percpu" means # of
arenas == # of CPUs, while "phycpu" has # of arenas == 1/2 * (# of
CPUs). Note that no runtime check on whether hyper threading is enabled
is added yet.

When enabled, threads will be migrated between arenas when a CPU change
is detected. In the current design, to reduce overhead from reading CPU
id, each arena tracks the thread accessed most recently. When a new
thread comes in, we will read CPU id and update arena if necessary.
2017-03-08 23:19:01 -08:00
Qi Wang
8721e19c04 Fix arena_prefork lock rank order for witness.
When witness is enabled, lock rank order needs to be preserved during
prefork, not only for each arena, but also across arenas. This change
breaks arena_prefork into further stages to ensure valid rank order
across arenas. Also changed test/unit/fork to use a manual arena to
catch this case.
2017-03-08 23:07:27 -08:00
Qi Wang
01f47f11a6 Store associated arena in tcache.
This fixes tcache_flush for manual tcaches, which wasn't able to find
the correct arena it associated with. Also changed the decay test to
cover this case (by using manually created arenas).
2017-03-07 12:58:11 -08:00
Jason Evans
cc75c35db5 Add any() and remove_any() to ph.
These functions select the easiest-to-remove element in the heap, which
is either the most recently inserted aux list element or the root.  If
no calls are made to first() or remove_first(), the behavior (and time
complexity) is the same as for a LIFO queue.
2017-03-07 10:25:33 -08:00
Jason Evans
8547ee11c3 Fix flakiness in test_decay_ticker.
Fix the test_decay_ticker test to carefully control slab
creation/destruction such that the decay backlog reliably reaches zero.
Use an isolated arena so that no extraneous allocation can confuse the
situation.  Speed up time during the latter part of the test so that the
entire decay time can expire in a reasonable amount of wall time.
2017-03-07 10:25:12 -08:00
David Goldblatt
438efede78 Add atomic types for ssize_t 2017-03-06 18:49:19 -08:00
David Goldblatt
e9852b5776 Disentangle assert and util
This is the first header refactoring diff, #533.  It splits the assert and util
components into separate, hermetic, header files.  In the process, it splits out
two of the large sub-components of util (the stdio.h replacement, and bit
manipulation routines) into their own components (malloc_io.h and bit_util.h).
This is mostly to break up cyclic dependencies, but it also breaks off a good
chunk of the catch-all-ness of util, which is nice.
2017-03-06 15:08:43 -08:00
David Goldblatt
d4ac7582f3 Introduce a backport of C11 atomics
This introduces a backport of C11 atomics.  It has four implementations; ranked
in order of preference, they are:
- GCC/Clang __atomic builtins
- GCC/Clang __sync builtins
- MSVC _Interlocked builtins
- C11 atomics, from <stdatomic.h>

The primary advantages are:
- Close adherence to the standard API gives us a defined memory model.
- Type safety: atomic objects are now separate types from non-atomic ones, so
  that it's impossible to mix up atomic and non-atomic updates (which is
  undefined behavior that compilers are starting to take advantage of).
- Efficiency: we can specify ordering for operations, avoiding fences and
  atomic operations on strongly ordered architectures (example:
  `atomic_write_u32(ptr, val);` involves a CAS loop, whereas
  `atomic_store(ptr, val, ATOMIC_RELEASE);` is a plain store.

This diff leaves in the current atomics API (implementing them in terms of the
backport).  This lets us transition uses over piecemeal.

Testing:
This is by nature hard to test. I've manually tested the first three options on
Linux on gcc by futzing with the #defines manually, on freebsd with gcc and
clang, on MSVC, and on OS X with clang.  All of these were x86 machines though,
and we don't have any test infrastructure set up for non-x86 platforms.
2017-03-03 13:40:59 -08:00
Jason Evans
fd058f572b Immediately purge cached extents if decay_time is 0.
This fixes a regression caused by
54269dc0ed (Remove obsolete
arena_maybe_purge() call.), as well as providing a general fix.

This resolves #665.
2017-03-02 19:43:06 -08:00
Jason Evans
de49674fbd Use MALLOC_CONF rather than malloc_conf for tests.
malloc_conf does not reliably work with MSVC, which complains of
"inconsistent dll linkage", i.e. its inability to support the
application overriding malloc_conf when dynamically linking/loading.
Work around this limitation by adding test harness support for per test
shell script sourcing, and converting all tests to use MALLOC_CONF
instead of malloc_conf.
2017-02-23 08:57:02 -08:00
Jason Evans
de8a68e853 Enhance spin_adaptive() to yield after several iterations.
This avoids worst case behavior if e.g. another thread is preempted
while owning the resource the spinning thread is waiting for.
2017-02-08 18:50:03 -08:00
Jason Evans
650c070e10 Remove rtree support for 0 (NULL) keys.
NULL can never actually be inserted in practice, and removing support
allows a branch to be removed from the fast path.
2017-02-08 18:50:03 -08:00
Jason Evans
f5cf9b19c8 Determine rtree levels at compile time.
Rather than dynamically building a table to aid per level computations,
define a constant table at compile time.  Omit both high and low
insignificant bits.  Use one to three tree levels, depending on the
number of significant bits.
2017-02-08 18:50:03 -08:00
Jason Evans
3bd6d8e41d Conditianalize lg_tcache_max use on JEMALLOC_TCACHE. 2017-02-07 12:15:36 -08:00
Jason Evans
d27f29b468 Disentangle arena and extent locking.
Refactor arena and extent locking protocols such that arena and
extent locks are never held when calling into the extent_*_wrapper()
API.  This requires extra care during purging since the arena lock no
longer protects the inner purging logic.  It also requires extra care to
protect extents from being merged with adjacent extents.

Convert extent_t's 'active' flag to an enumerated 'state', so that
retained extents are explicitly marked as such, rather than depending on
ring linkage state.

Refactor the extent collections (and their synchronization) for cached
and retained extents into extents_t.  Incorporate LRU functionality to
support purging.  Incorporate page count accounting, which replaces
arena->ndirty and arena->stats.retained.

Assert that no core locks are held when entering any internal
[de]allocation functions.  This is in addition to existing assertions
that no locks are held when entering external [de]allocation functions.

Audit and document synchronization protocols for all arena_t fields.

This fixes a potential deadlock due to recursive allocation during
gdump, in a similar fashion to b49c649bc1
(Fix lock order reversal during gdump.), but with a necessarily much
broader code impact.
2017-02-01 16:43:46 -08:00