summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Add documentation for background_thread related options.Qi Wang2017-05-231-0/+78
|
* Add profiling for the background thread mutex.Qi Wang2017-05-234-6/+21
|
* Add background thread related stats.Qi Wang2017-05-239-21/+214
|
* Implementing opt.background_thread.Qi Wang2017-05-2323-348/+1245
| | | | | | | | | | | 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).
* Protect the rtree/extent interactions with a mutex pool.David Goldblatt2017-05-1918-536/+340
| | | | | | | | | | | | | | | | | | 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elements is determined by reading the extent out of those elements, the initial read is racy, so that we may acquire an out of date mutex.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.
* Allow mutexes to take a lock ordering enum at construction.David Goldblatt2017-05-1912-30/+75
| | | | | | | This lets us specify whether and how mutexes of the same rank are allowed to be acquired. Currently, we only allow two polices (only a single mutex at a given rank at a time, and mutexes acquired in ascending order), but we can plausibly allow more (e.g. the "release uncontended mutexes before blocking").
* Refactor *decay_time into *decay_ms.Jason Evans2017-05-1816-328/+320
| | | | | | | | 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.
* Add stats: arena uptime.Qi Wang2017-05-186-0/+42
|
* Avoid over-rebuilding due to namespace mangling.Jason Evans2017-05-172-3/+10
| | | | | | | Take care not to touch generated namespace mangling headers unless their contents would change. This resolves #838.
* Use srcroot path for private_namespace.sh.Qi Wang2017-05-161-2/+2
|
* Refactor (MALLOCX_ARENA_MAX + 1) to be MALLOCX_ARENA_LIMIT.Jason Evans2017-05-145-11/+11
| | | | This resolves #673.
* Automatically generate private symbol name mangling macros.Jason Evans2017-05-129-671/+239
| | | | | | | | Rather than using a manually maintained list of internal symbols to drive name mangling, add a compilation phase to automatically extract the list of internal symbols. This resolves #677.
* Remove unused private_unnamespace infrastructure.Jason Evans2017-05-123-15/+0
|
* Stop depending on JEMALLOC_N() for function interception during testing.Jason Evans2017-05-1213-281/+157
| | | | | | Instead, always define function pointers for interceptable functions, but mark them const unless testing, so that the compiler can optimize out the pointer dereferences.
* Do not build in parallel on AppVeyor.Jason Evans2017-05-121-2/+2
| | | | | | The compiler database used by MSVC is increasingly becoming corrupt, presumably due to concurrency-related corruption, despite the -FS compiler flag being specified as recommended.
* Update and rename INSTALL to INSTALL.mdArkady Shapkin2017-05-121-57/+107
|
* Specify -Werror for run_tests builds.Jason Evans2017-05-121-1/+1
|
* Avoid compiler warnings on Windows.Jason Evans2017-05-121-5/+5
|
* Remove mutex_prof_data_t redeclaration.Jason Evans2017-05-112-2/+1
| | | | | | | Redeclaration causes compilations failures with e.g. gcc 4.2.1 on FreeBSD. This regression was introduced by 89e2d3c12b573310e60b97beaf178007a71d83a3 (Header refactoring: ctl - unify and remove from catchall.).
* Add --with-version=VERSION .Jason Evans2017-05-032-6/+21
| | | | | | | This simplifies configuration when embedding a jemalloc release into another project's git repository. This resolves #811.
* Fix rtree_leaf_elm_szind_slab_update().Jason Evans2017-05-031-8/+7
| | | | | | | | | | | Re-read the leaf element when atomic CAS fails due to a race with another thread that has locked the leaf element, since atomic_compare_exchange_strong_p() overwrites the expected value with the actual value on failure. This regression was introduced by 0ee0e0c155a05d0d028a9972ad86b9eaac4ccabd (Implement compact rtree leaf element representation.). This resolves #798.
* rtree_leaf_elm_extent_write() --> rtree_leaf_elm_extent_lock_write()Jason Evans2017-05-031-8/+9
| | | | | | | Refactor rtree_leaf_elm_extent_write() as rtree_leaf_elm_extent_lock_write(), so that whether the leaf element is currently acquired is separate from what lock state to write. This allows for a relaxed atomic read when releasing the lock.
* Make VS2015 project work againrustyx2017-05-022-238/+9
|
* Revert "Use trylock in tcache_bin_flush when possible."Qi Wang2017-05-013-142/+63
| | | | | This reverts commit 8584adc451f31adfc4ab8693d9189cf3a7e5d858. Production results not favorable. Will investigate separately.
* Header refactoring: tsd - cleanup and dependency breaking.David Goldblatt2017-05-0129-913/+870
| | | | | | | | | | | | 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.
* Add extent_destroy_t and use it during arena destruction.Jason Evans2017-04-2910-17/+120
| | | | | | | | | | 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.
* Refactor !opt.munmap to opt.retain.Jason Evans2017-04-2916-43/+42
|
* Revert "Use try_flush first in tcache_dalloc."Qi Wang2017-04-281-10/+4
| | | | | | This reverts commit b0c2a28280d363fc85aa8b4fdbe7814ef46cb17b. Production benchmark shows this caused significant regression in both CPU and memory consumption. Will investigate separately later on.
* Inline tcache_bin_flush_small_impl / _large_impl.Qi Wang2017-04-281-2/+2
|
* Use try_flush first in tcache_dalloc.Qi Wang2017-04-261-4/+10
| | | | Only do must_flush if try_flush didn't manage to free anything.
* Use trylock in tcache_bin_flush when possible.Qi Wang2017-04-263-63/+142
| | | | | During tcache gc, use tcache_bin_try_flush_small / _large so that we can skip items with their bins locked already.
* Remove redundant extent lookup in tcache_bin_flush_large.Qi Wang2017-04-251-1/+1
|
* Avoid prof_dump during reentrancy.Qi Wang2017-04-252-12/+23
|
* Header refactoring: pages.h - unify and remove from catchall.David Goldblatt2017-04-256-34/+33
|
* Header refactoring: hash - unify and remove from catchall.David Goldblatt2017-04-255-4/+6
|
* Header refactoring: ctl - unify and remove from catchall.David Goldblatt2017-04-2511-264/+257
| | | | | In order to do this, we introduce the mutex_prof module, which breaks a circular dependency between ctl and prof.
* Replace --disable-munmap with opt.munmap.Jason Evans2017-04-2516-85/+77
| | | | | | | | | 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.
* Remove --enable-code-coverage.Jason Evans2017-04-246-109/+1
| | | | | | | This option hasn't been particularly useful since the original pre-3.0.0 push to broaden test coverage. This partially resolves #580.
* Remove --disable-cc-silence.Jason Evans2017-04-245-41/+6
| | | | | | | The explicit compiler warning suppression controlled by this option is universally desirable, so remove the ability to disable suppression. This partially resolves #580.
* Use trylock in arena_decay_impl().Qi Wang2017-04-241-8/+16
| | | | If another thread is working on decay, we don't have to wait for the mutex.
* Implement malloc_mutex_trylock() w/ proper stats update.Qi Wang2017-04-243-14/+36
|
* Remove --with-lg-tiny-min.Jason Evans2017-04-244-41/+11
| | | | | | This option isn't useful in practice. This partially resolves #580.
* Remove --with-lg-size-class-group.Jason Evans2017-04-242-39/+2
| | | | | | | | Four size classes per size doubling has proven to be a universally good choice for the entire 4.x release series, so there's little point to preserving this configurability. This partially resolves #580.
* Header refactoring: bitmap - unify and remove from catchall.David Goldblatt2017-04-249-387/+373
|
* Header refactoring: stats - unify and remove from catchallDavid Goldblatt2017-04-248-35/+29
|
* Header refactoring: move smoothstep.h out of the catchall.David Goldblatt2017-04-242-1/+1
|
* Header refactoring: size_classes module - remove from the catchallDavid Goldblatt2017-04-2420-1/+28
|
* Header refactoring: ckh module - remove from the catchall and unify.David Goldblatt2017-04-248-84/+106
|
* Header refactoring: ticker module - remove from the catchall and unify.David Goldblatt2017-04-2412-21/+26
|
* Header refactoring: prng module - remove from the catchall and unify.David Goldblatt2017-04-246-35/+49
|
generated by cgit v0.12 at 2025-04-02 14:25:29 (GMT)