summaryrefslogtreecommitdiffstats
path: root/include
Commit message (Collapse)AuthorAgeFilesLines
...
* Do not use syscall(2) on OS X 10.12 (deprecated).Jason Evans2016-11-031-0/+3
|
* Add os_unfair_lock support.Jason Evans2016-11-033-0/+17
| | | | | OS X 10.12 deprecated OSSpinLock; os_unfair_lock is the recommended replacement.
* Fix/refactor zone allocator integration code.Jason Evans2016-11-031-1/+1
| | | | | | | | | Fix zone_force_unlock() to reinitialize, rather than unlocking mutexes, since OS X 10.12 cannot tolerate a child unlocking mutexes that were locked by its parent. Refactor; this was a side effect of experimenting with zone {de,re}registration during fork(2).
* Add an assertion in witness_owner().Jason Evans2016-10-311-0/+3
|
* Refactor witness_unlock() to fix undefined test behavior.Jason Evans2016-10-312-11/+29
| | | | This resolves #396.
* Use CLOCK_MONOTONIC_COARSE rather than COARSE_MONOTONIC_RAW.Jason Evans2016-10-301-2/+2
| | | | | | | | The raw clock variant is slow (even relative to plain CLOCK_MONOTONIC), whereas the coarse clock variant is faster than CLOCK_MONOTONIC, but still has resolution (~1ms) that is adequate for our purposes. This resolves #479.
* Support static linking of jemalloc with glibcDave Watson2016-10-281-0/+3
| | | | | | | | | | | | | | | | | | | | | | | glibc defines its malloc implementation with several weak and strong symbols: strong_alias (__libc_calloc, __calloc) weak_alias (__libc_calloc, calloc) strong_alias (__libc_free, __cfree) weak_alias (__libc_free, cfree) strong_alias (__libc_free, __free) strong_alias (__libc_free, free) strong_alias (__libc_malloc, __malloc) strong_alias (__libc_malloc, malloc) The issue is not with the weak symbols, but that other parts of glibc depend on __libc_malloc explicitly. Defining them in terms of jemalloc API's allows the linker to drop glibc's malloc.o completely from the link, and static linking no longer results in symbol collisions. Another wrinkle: jemalloc during initialization calls sysconf to get the number of CPU's. GLIBC allocates for the first time before setting up isspace (and other related) tables, which are used by sysconf. Instead, use the pthread API to get the number of CPUs with GLIBC, which seems to work. This resolves #442.
* Avoid negation of unsigned numbers.Jason Evans2016-10-281-2/+2
| | | | | | Rather than relying on two's complement negation for alignment mask generation, use bitwise not and addition. This dodges warnings from MSVC, and should be strength-reduced by compiler optimization anyway.
* Do not (recursively) allocate within tsd_fetch().Jason Evans2016-10-216-33/+71
| | | | | | | Refactor tsd so that tsdn_fetch() does not trigger allocation, since allocation could cause infinite recursion. This resolves #458.
* Make dss operations lockless.Jason Evans2016-10-134-18/+8
| | | | | | | | | | | | | | Rather than protecting dss operations with a mutex, use atomic operations. This has negligible impact on synchronization overhead during typical dss allocation, but is a substantial improvement for extent_in_dss() and the newly added extent_dss_mergeable(), which can be called multiple times during extent deallocations. This change also has the advantage of avoiding tsd in deallocation paths associated with purging, which resolves potential deadlocks during thread exit due to attempted tsd resurrection. This resolves #425.
* Add/use adaptive spinning.Jason Evans2016-10-132-0/+55
| | | | | | | | Add spin_t and spin_{init,adaptive}(), which provide a simple abstraction for adaptive spinning. Adaptively spin during busy waits in bootstrapping and rtree node initialization.
* Remove all vestiges of chunks.Jason Evans2016-10-125-90/+2
| | | | | | | | Remove mallctls: - opt.lg_chunk - stats.cactive This resolves #464.
* Remove ratio-based purging.Jason Evans2016-10-123-41/+5
| | | | | | | | | | | | | Make decay-based purging the default (and only) mode. Remove associated mallctls: - opt.purge - opt.lg_dirty_mult - arena.<i>.lg_dirty_mult - arenas.lg_dirty_mult - stats.arenas.<i>.lg_dirty_mult This resolves #385.
* Fix and simplify decay-based purging.Jason Evans2016-10-111-18/+11
| | | | | | | | | | | | | | | | | | | | | Simplify decay-based purging attempts to only be triggered when the epoch is advanced, rather than every time purgeable memory increases. In a correctly functioning system (not previously the case; see below), this only causes a behavior difference if during subsequent purge attempts the least recently used (LRU) purgeable memory extent is initially too large to be purged, but that memory is reused between attempts and one or more of the next LRU purgeable memory extents are small enough to be purged. In practice this is an arbitrary behavior change that is within the set of acceptable behaviors. As for the purging fix, assure that arena->decay.ndirty is recorded *after* the epoch advance and associated purging occurs. Prior to this fix, it was possible for purging during epoch advance to cause a substantially underrepresentative (arena->ndirty - arena->decay.ndirty), i.e. the number of dirty pages attributed to the current epoch was too low, and a series of unintended purges could result. This fix is also relevant in the context of the simplification described above, but the bug's impact would be limited to over-purging at epoch advances.
* Do not advance decay epoch when time goes backwards.Jason Evans2016-10-112-0/+4
| | | | | | Instead, move the epoch backward in time. Additionally, add nstime_monotonic() and use it in debug builds to assert that time only goes backward if nstime_update() is using a non-monotonic time source.
* Refactor arena->decay_* into arena->decay.* (arena_decay_t).Jason Evans2016-10-111-46/+53
|
* Refine nstime_update().Jason Evans2016-10-103-3/+19
| | | | | | | | | | | | | | | | | | | | | Add missing #include <time.h>. The critical time facilities appear to have been transitively included via unistd.h and sys/time.h, but in principle this omission was capable of having caused clock_gettime(CLOCK_MONOTONIC, ...) to have been overlooked in favor of gettimeofday(), which in turn could cause spurious non-monotonic time updates. Refactor nstime_get() out of nstime_update() and add configure tests for all variants. Add CLOCK_MONOTONIC_RAW support (Linux-specific) and mach_absolute_time() support (OS X-specific). Do not fall back to clock_gettime(CLOCK_REALTIME, ...). This was a fragile Linux-specific workaround, which we're unlikely to use at all now that clock_gettime(CLOCK_MONOTONIC_RAW, ...) is supported, and if we have no choice besides non-monotonic clocks, gettimeofday() is only incrementally worse.
* Fix size class overflow bugs.Jason Evans2016-10-032-3/+21
| | | | | | | Avoid calling s2u() on raw extent sizes in extent_recycle(). Clamp psz2ind() (implemented as psz2ind_clamp()) when inserting/removing into/from size-segregated extent heaps.
* Update extent hook function prototype comments.Jason Evans2016-09-291-13/+14
|
* Fix LG_QUANTUM definition for sparc64Eric Le Bihan2016-09-261-1/+1
| | | | | | GCC 4.9.3 cross-compiled for sparc64 defines __sparc_v9__, not __sparc64__ nor __sparcv9. This prevents LG_QUANTUM from being defined properly. Adding this new value to the check solves the issue.
* Avoid self assignment in tsd_set().Jason Evans2016-09-231-4/+8
|
* Add various mutex ownership assertions.Jason Evans2016-09-231-0/+2
|
* Fix extent_{before,last,past}() to return page-aligned results.Jason Evans2016-09-231-4/+5
|
* Protect extents_dirty access with extents_mtx.Jason Evans2016-09-223-11/+14
| | | | This fixes race conditions during purging.
* Fix -Wundef in _MSC_VER check.Josh Gao2016-09-151-1/+1
|
* Check for __builtin_unreachable at configure timeElliot Ronaghan2016-07-072-16/+12
| | | | | | | | | | | | | | | | | | | Add a configure check for __builtin_unreachable instead of basing its availability on the __GNUC__ version. On OS X using gcc (a real gcc, not the bundled version that's just a gcc front-end) leads to a linker assertion: https://github.com/jemalloc/jemalloc/issues/266 It turns out that this is caused by a gcc bug resulting from the use of __builtin_unreachable(): https://gcc.gnu.org/bugzilla/show_bug.cgi?id=57438 To work around this bug, check that __builtin_unreachable() actually works at configure time, and if it doesn't use abort() instead. The check is based on https://gcc.gnu.org/bugzilla/show_bug.cgi?id=57438#c21. With this `make check` passes with a homebrew installed gcc-5 and gcc-6.
* Define 64-bits atomics unconditionallyMike Hommey2016-06-091-10/+8
| | | | They are used on all platforms in prng.h.
* Fix extent_*_get to build with MSVCMike Hommey2016-06-091-4/+4
|
* Don't use compact red-black trees with the pgi compilerElliot Ronaghan2016-06-081-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Some bug (either in the red-black tree code, or in the pgi compiler) seems to cause red-black trees to become unbalanced. This issue seems to go away if we don't use compact red-black trees. Since red-black trees don't seem to be used much anymore, I opted for what seems to be an easy fix here instead of digging in and trying to find the root cause of the bug. Some context in case it's helpful: I experienced a ton of segfaults while using pgi as Chapel's target compiler with jemalloc 4.0.4. The little bit of debugging I did pointed me somewhere deep in red-black tree manipulation, but I didn't get a chance to investigate further. It looks like 4.2.0 replaced most uses of red-black trees with pairing-heaps, which seems to avoid whatever bug I was hitting. However, `make check_unit` was still failing on the rb test, so I figured the core issue was just being masked. Here's the `make check_unit` failure: ```sh === test/unit/rb === test_rb_empty: pass tree_recurse:test/unit/rb.c:90: Failed assertion: (((_Bool) (((uintptr_t) (left_node)->link.rbn_right_red) & ((size_t)1)))) == (false) --> true != false: Node should be black test_rb_random:test/unit/rb.c:274: Failed assertion: (imbalances) == (0) --> 1 != 0: Tree is unbalanced tree_recurse:test/unit/rb.c:90: Failed assertion: (((_Bool) (((uintptr_t) (left_node)->link.rbn_right_red) & ((size_t)1)))) == (false) --> true != false: Node should be black test_rb_random:test/unit/rb.c:274: Failed assertion: (imbalances) == (0) --> 1 != 0: Tree is unbalanced node_remove:test/unit/rb.c:190: Failed assertion: (imbalances) == (0) --> 2 != 0: Tree is unbalanced <jemalloc>: test/unit/rb.c:43: Failed assertion: "pathp[-1].cmp < 0" test/test.sh: line 22: 12926 Aborted Test harness error ``` While starting to debug I saw the RB_COMPACT option and decided to check if turning that off resolved the bug. It seems to have fixed it (`make check_unit` passes and the segfaults under Chapel are gone) so it seems like on okay work-around. I'd imagine this has performance implications for red-black trees under pgi, but if they're not going to be used much anymore it's probably not a big deal.
* Fix potential VM map fragmentation regression.Jason Evans2016-06-071-1/+1
| | | | | | | | Revert 245ae6036c09cc11a72fab4335495d95cddd5beb (Support --with-lg-page values larger than actual page size.), because it could cause VM map fragmentation if the kernel grows mmap()ed memory downward. This resolves #391.
* Fix extent_alloc_dss() regressions.Jason Evans2016-06-062-0/+2
| | | | | Page-align the gap, if any, and add/use extent_dalloc_gap(), which registers the gap extent before deallocation.
* Remove a stray memset(), and fix a junk filling test regression.Jason Evans2016-06-062-0/+4
|
* Modify extent hook functions to take an (extent_t *) argument.Jason Evans2016-06-063-37/+40
| | | | | | | This facilitates the application accessing its own extent allocator metadata during hook invocations. This resolves #259.
* Add rtree lookup path caching.Jason Evans2016-06-064-36/+176
| | | | | | | | | rtree-based extent lookups remain more expensive than chunk-based run lookups, but with this optimization the fast path slowdown is ~3 CPU cycles per metadata lookup (on Intel Core i7-4980HQ), versus ~11 cycles prior. The path caching speedup tends to degrade gracefully unless allocated memory is spread far apart (as is the case when using a mixture of sbrk() and mmap()).
* Make tsd cleanup functions optional, remove noop cleanup functions.Jason Evans2016-06-066-31/+17
|
* Add a missing prof_alloc_rollback() call.Jason Evans2016-06-061-0/+1
| | | | | | In the case where prof_alloc_prep() is called with an over-estimate of allocation size, and sampling doesn't end up being triggered, the tctx must be discarded.
* Miscellaneous s/chunk/extent/ updates.Jason Evans2016-06-063-5/+4
|
* Relax NBINS constraint (max 255 --> max 256).Jason Evans2016-06-061-4/+2
|
* Remove obsolete stats.arenas.<i>.metadata.mapped mallctl.Jason Evans2016-06-064-22/+18
| | | | | Rename stats.arenas.<i>.metadata.allocated mallctl to stats.arenas.<i>.metadata .
* Better document --enable-ivsalloc.Jason Evans2016-06-062-3/+10
|
* Rename most remaining *chunk* APIs to *extent*.Jason Evans2016-06-066-88/+87
|
* s/chunk_lookup/extent_lookup/g, s/chunks_rtree/extents_rtree/gJason Evans2016-06-064-20/+18
|
* s/CHUNK_HOOKS_INITIALIZER/EXTENT_HOOKS_INITIALIZER/gJason Evans2016-06-062-10/+10
|
* Rename chunks_{cached,retained,mtx} to extents_{cached,retained,mtx}.Jason Evans2016-06-062-9/+9
|
* Rename chunk_*_t hooks to extent_*_t.Jason Evans2016-06-061-21/+22
|
* s/chunk_hook/extent_hook/gJason Evans2016-06-064-22/+25
|
* Rename huge to large.Jason Evans2016-06-0611-115/+115
|
* Update private symbols.Jason Evans2016-06-061-11/+19
|
* Move slabs out of chunks.Jason Evans2016-06-068-863/+147
|
* Improve interval-based profile dump triggering.Jason Evans2016-06-061-1/+1
| | | | | | | | | | | When an allocation is large enough to trigger multiple dumps, use modular math rather than subtraction to reset the interval counter. Prior to this change, it was possible for a single allocation to cause many subsequent allocations to all trigger profile dumps. When updating usable size for a sampled object, try to cancel out the difference between LARGE_MINCLASS and usable size from the interval counter.
generated by cgit v0.12 at 2025-04-17 22:17:12 (GMT)