Fix OOM cleanup in huge_palloc() to call idalloct() rather than
base_node_dalloc(). This bug is a result of incomplete refactoring, and
has no impact other than leaking memory during OOM.
This provides in-place expansion of huge allocations when the end of the
allocation is at the end of the sbrk heap. There's already the ability
to extend in-place via recycled chunks but this handles the initial
growth of the heap via repeated vector / string reallocations.
A possible future extension could allow realloc to go from the following:
| huge allocation | recycled chunks |
^ dss_end
To a larger allocation built from recycled *and* new chunks:
| huge allocation |
^ dss_end
Doing that would involve teaching the chunk recycling code to request
new chunks to satisfy the request. The chunk_dss code wouldn't require
any further changes.
#include <stdlib.h>
int main(void) {
size_t chunk = 4 * 1024 * 1024;
void *ptr = NULL;
for (size_t size = chunk; size < chunk * 128; size *= 2) {
ptr = realloc(ptr, size);
if (!ptr) return 1;
}
}
dss:secondary: 0.083s
dss:primary: 0.083s
After:
dss:secondary: 0.083s
dss:primary: 0.003s
The dss heap grows in the upwards direction, so the oldest chunks are at
the low addresses and they are used first. Linux prefers to grow the
mmap heap downwards, so the trick will not work in the *current* mmap
chunk allocator as a huge allocation will only be at the top of the heap
in a contrived case.
Fix quarantine to actually update tsd when expanding, and to avoid
double initialization (leaking the first quarantine) due to recursive
initialization.
This resolves#161.
* use sized deallocation in iralloct_realign
* iralloc and ixalloc always need the old size, so pass it in from the
caller where it's often already calculated
The size of the source allocation is known at this point, so reading the
chunk header can be avoided for the small size class fast path. This is
not very useful right now, but it provides a significant performance
boost with an alternate ralloc entry point taking the old size.
Purge trailing pages during shrinking huge reallocation when resulting
size is not a multiple of the chunk size. Similarly, zero pages if
necessary during growing huge reallocation when the resulting size is
not a multiple of the chunk size.
Add the 'util' column, which reports the proportion of available regions
that are currently in use for each small size class. Small run
utilization is the complement of external fragmentation. For example,
utilization of 0.75 indicates that 25% of small run memory is consumed
by external fragmentation, in other (more obtuse) words, 33% external
fragmentation overhead.
This resolves#27.
Add per size class huge allocation statistics, and normalize various
stats:
- Change the arenas.nlruns type from size_t to unsigned.
- Add the arenas.nhchunks and arenas.hchunks.<i>.size mallctl's.
- Replace the stats.arenas.<i>.bins.<j>.allocated mallctl with
stats.arenas.<i>.bins.<j>.curregs .
- Add the stats.arenas.<i>.hchunks.<j>.nmalloc,
stats.arenas.<i>.hchunks.<j>.ndalloc,
stats.arenas.<i>.hchunks.<j>.nrequests, and
stats.arenas.<i>.hchunks.<j>.curhchunks mallctl's.
Fix a prof_tctx_t/prof_tdata_t cleanup race by storing a copy of thr_uid
in prof_tctx_t, so that the associated tdata need not be present during
tctx teardown.
Remove code in arena_dalloc_bin_run() that preserved the "clean" state
of trailing clean pages by splitting them into a separate run during
deallocation. This was a useful mechanism for reducing dirty page
churn when bin runs comprised many pages, but bin runs are now quite
small.
Remove the nextind field from arena_run_t now that it is no longer
needed, and change arena_run_t's bin field (arena_bin_t *) to binind
(index_t). These two changes remove 8 bytes of chunk header overhead
per page, which saves 1/512 of all arena chunk memory.
Add:
--with-lg-page
--with-lg-page-sizes
--with-lg-size-class-group
--with-lg-quantum
Get rid of STATIC_PAGE_SHIFT, in favor of directly setting LG_PAGE.
Fix various edge conditions exposed by the configure options.
atexit(3) can deadlock internally during its own initialization if
jemalloc calls atexit() during jemalloc initialization. Mitigate the
impact by restructuring prof initialization to avoid calling atexit()
unless the registered function will actually dump a final heap profile.
Additionally, disable prof_final by default so that this land mine is
opt-in rather than opt-out.
This resolves#144.
This avoids grabbing the base mutex, as a step towards fine-grained
locking for huge allocations. The thread cache also provides a tiny
(~3%) improvement for serial huge allocations.
Abstract arenas access to use arena_get() (or a0get() where appropriate)
rather than directly reading e.g. arenas[ind]. Prior to the addition of
the arenas.extend mallctl, the worst possible outcome of directly
accessing arenas was a stale read, but arenas.extend may allocate and
assign a new array to arenas.
Add a tsd-based arenas_cache, which amortizes arenas reads. This
introduces some subtle bootstrapping issues, with tsd_boot() now being
split into tsd_boot[01]() to support tsd wrapper allocation
bootstrapping, as well as an arenas_cache_bypass tsd variable which
dynamically terminates allocation of arenas_cache itself.
Promote a0malloc(), a0calloc(), and a0free() to be generally useful for
internal allocation, and use them in several places (more may be
appropriate).
Abstract arena->nthreads management and fix a missing decrement during
thread destruction (recent tsd refactoring left arenas_cleanup()
unused).
Change arena_choose() to propagate OOM, and handle OOM in all callers.
This is important for providing consistent allocation behavior when the
MALLOCX_ARENA() flag is being used. Prior to this fix, it was possible
for an OOM to result in allocation silently allocating from a different
arena than the one specified.
Normalize size classes to use the same number of size classes per size
doubling (currently hard coded to 4), across the intire range of size
classes. Small size classes already used this spacing, but in order to
support this change, additional small size classes now fill [4 KiB .. 16
KiB). Large size classes range from [16 KiB .. 4 MiB). Huge size
classes now support non-multiples of the chunk size in order to fill (4
MiB .. 16 MiB).
This adds support for expanding huge allocations in-place by requesting
memory at a specific address from the chunk allocator.
It's currently only implemented for the chunk recycling path, although
in theory it could also be done by optimistically allocating new chunks.
On Linux, it could attempt an in-place mremap. However, that won't work
in practice since the heap is grown downwards and memory is not unmapped
(in a normal build, at least).
Repeated vector reallocation micro-benchmark:
#include <string.h>
#include <stdlib.h>
int main(void) {
for (size_t i = 0; i < 100; i++) {
void *ptr = NULL;
size_t old_size = 0;
for (size_t size = 4; size < (1 << 30); size *= 2) {
ptr = realloc(ptr, size);
if (!ptr) return 1;
memset(ptr + old_size, 0xff, size - old_size);
old_size = size;
}
free(ptr);
}
}
The glibc allocator fails to do any in-place reallocations on this
benchmark once it passes the M_MMAP_THRESHOLD (default 128k) but it
elides the cost of copies via mremap, which is currently not something
that jemalloc can use.
With this improvement, jemalloc still fails to do any in-place huge
reallocations for the first outer loop, but then succeeds 100% of the
time for the remaining 99 iterations. The time spent doing allocations
and copies drops down to under 5%, with nearly all of it spent doing
purging + faulting (when huge pages are disabled) and the array memset.
An improved mremap API (MREMAP_RETAIN - #138) would be far more general
but this is a portable optimization and would still be useful on Linux
for xallocx.
Numbers with transparent huge pages enabled:
glibc (copies elided via MREMAP_MAYMOVE): 8.471s
jemalloc: 17.816s
jemalloc + no-op madvise: 13.236s
jemalloc + this commit: 6.787s
jemalloc + this commit + no-op madvise: 6.144s
Numbers with transparent huge pages disabled:
glibc (copies elided via MREMAP_MAYMOVE): 15.403s
jemalloc: 39.456s
jemalloc + no-op madvise: 12.768s
jemalloc + this commit: 15.534s
jemalloc + this commit + no-op madvise: 6.354s
Closes#137
Fix an OOM-related regression in arena_tcache_fill_small() that caused
cache corruption that would almost certainly expose the application to
undefined behavior, usually in the form of an allocation request
returning an already-allocated region, or somewhat less likely, a freed
region that had already been returned to the arena, thus making it
available to the arena for any purpose.
This regression was introduced by
9c43c13a35 (Reverse tcache fill order.),
and was present in all releases from 2.2.0 through 3.6.0.
This resolves#98.
Fix prof regressions related to tdata (main per thread profiling data
structure) destruction:
- Deadlock. The fix for this was intended to be part of
20c31deaae (Test prof.reset mallctl and
fix numerous discovered bugs.) but the fix was left incomplete.
- Destruction race. Detaching tdata just prior to destruction without
holding the tdatas lock made it possible for another thread to destroy
the tdata out from under the thread that was on its way to doing so.
Fix tsd cleanup regressions that were introduced in
5460aa6f66 (Convert all tsd variables to
reside in a single tsd structure.). These regressions were twofold:
1) tsd_tryget() should never (and need never) return NULL. Rename it to
tsd_fetch() and simplify all callers.
2) tsd_*_set() must only be called when tsd is in the nominal state,
because cleanup happens during the nominal-->purgatory transition,
and re-initialization must not happen while in the purgatory state.
Add tsd_nominal() and use it as needed. Note that tsd_*{p,}_get()
can still be used as long as no re-initialization that would require
cleanup occurs. This means that e.g. the thread_allocated counter
can be updated unconditionally.
Implement/test/fix the opt.prof_thread_active_init,
prof.thread_active_init, and thread.prof.active mallctl's.
Test/fix the thread.prof.name mallctl.
Refactor opt_prof_active to be read-only and move mutable state into the
prof_active variable. Stop leaning on ctl-related locking for
protection.
Move small run metadata into the arena chunk header, with multiple
expected benefits:
- Lower run fragmentation due to reduced run sizes; runs are more likely
to completely drain when there are fewer total regions.
- Improved cache behavior. Prior to this change, run headers were
always page-aligned, which put extra pressure on some CPU cache sets.
The degree to which this was a problem was hardware dependent, but it
likely hurt some even for the most advanced modern hardware.
- Buffer overruns/underruns are less likely to corrupt allocator
metadata.
- Size classes between 4 KiB and 16 KiB become reasonable to support
without any special handling, and the runs are small enough that dirty
unused pages aren't a significant concern.
Fix a race that caused a non-critical assertion failure. To trigger the
race, a thread had to be part way through initializing a new sample,
such that it was discoverable by the dumping thread, but not yet linked
into its gctx by the time a later dump phase would normally have reset
its state to 'nominal'.
Additionally, lock access to the state field during modification to
transition to the dumping state. It's not apparent that this oversight
could have caused an actual problem due to outer locking that protects
the dumping machinery, but the added locking pedantically follows the
stated locking protocol for the state field.
Don't use atomic_add_uint64(), because it isn't available on 32-bit
platforms.
Fix forking support functions to manage all prof-related mutexes.
These regressions were introduced by
602c8e0971 (Implement per thread heap
profiling.), which did not make it into any releases prior to these
fixes.
Fix irallocx_prof() sample logic to only update the threshold counter
after it knows what size the allocation ended up being. This regression
was caused by 6e73dc194e (Fix a profile
sampling race.), which did not make it into any releases prior to this
fix.
* assertion failure
* malloc_init failure
* malloc not already initialized (in malloc_init)
* running in valgrind
* thread cache disabled at runtime
Clang and GCC already consider a comparison with NULL or -1 to be cold,
so many branches (out-of-memory) are already correctly considered as
cold and marking them is not important.
Fix a profile sampling race that was due to preparing to sample, yet
doing nothing to assure that the context remains valid until the stats
are updated.
These regressions were caused by
602c8e0971 (Implement per thread heap
profiling.), which did not make it into any releases prior to these
fixes.
Fix prof_tdata_get() to avoid dereferencing an invalid tdata pointer
(when it's PROF_TDATA_STATE_{REINCARNATED,PURGATORY}).
Fix prof_tdata_get() callers to check for invalid results besides NULL
(PROF_TDATA_STATE_{REINCARNATED,PURGATORY}).
These regressions were caused by
602c8e0971 (Implement per thread heap
profiling.), which did not make it into any releases prior to these
fixes.
This adds a new `sdallocx` function to the external API, allowing the
size to be passed by the caller. It avoids some extra reads in the
thread cache fast path. In the case where stats are enabled, this
avoids the work of calculating the size from the pointer.
An assertion validates the size that's passed in, so enabling debugging
will allow users of the API to debug cases where an incorrect size is
passed in.
The performance win for a contrived microbenchmark doing an allocation
and immediately freeing it is ~10%. It may have a different impact on a
real workload.
Closes#28
Optimize [nmd]alloc() fast paths such that the (flags == 0) case is
streamlined, flags decoding only happens to the minimum degree
necessary, and no conditionals are repeated.
Junk filling is done in arena_dalloc_bin_locked(), so arena_alloc_junk_small()
is redundant. Also, we should use arena_dalloc_junk_small() instead of
arena_alloc_junk_small().
__*_hook() is glibc, but on at least one glibc platform (homebrew),
the __GLIBC__ define isn't set correctly and we miss being able to
use these hooks.
Do a feature test for it during configuration so that we enable it
anywhere the hooks are actually available.
Rename data structures (prof_thr_cnt_t-->prof_tctx_t,
prof_ctx_t-->prof_gctx_t), and convert to storing a prof_tctx_t for
sampled objects.
Convert PROF_ALLOC_PREP() to prof_alloc_prep(), since precise backtrace
depth within jemalloc functions is no longer an issue (pprof prunes
irrelevant frames).
Implement mallctl's:
- prof.reset implements full sample data reset, and optional change of
sample interval.
- prof.lg_sample reads the current sample interval (opt.lg_prof_sample
was the permanent source of truth prior to prof.reset).
- thread.prof.name provides naming capability for threads within heap
profile dumps.
- thread.prof.active makes it possible to activate/deactivate heap
profiling for individual threads.
Modify the heap dump files to contain per thread heap profile data.
This change is incompatible with the existing pprof, which will require
enhancements to read and process the enriched data.
Treat prof_tdata_t's bt2cnt as a comprehensive map of the thread's
extant allocation samples (do not limit the total number of entries).
This helps prepare the way for per thread heap profiling.
Fix runs_dirty-based purging to also purge dirty pages in the spare
chunk.
Refactor runs_dirty manipulation into arena_dirty_{insert,remove}(), and
move the arena->ndirty accounting into those functions.
Remove the u.ql_link field from arena_chunk_map_t, and get rid of the
enclosing union for u.rb_link, since only rb_link remains.
Remove the ndirty field from arena_chunk_t.
Fix the cactive statistic to decrease (rather than increase) when active
memory decreases. This regression was introduced by
aa5113b1fd (Refactor overly large/complex
functions) and first released in 3.5.0.
Some platforms, such as Google's Portable Native Client, use Newlib and
thus lack access to madvise(2). In those instances, pages_purge() is
transformed into a no-op.
Native Client doesn't allow forking, thus there is no need to catch
fork()ing events for Native Client.
Additionally, without this commit, jemalloc will introduce an unresolved
pthread_atfork() in PNaCl Rust bins.
Some platforms (like those using Newlib) don't have ffs/ffsl. This
commit adds a check to configure.ac for __builtin_ffsl if ffsl isn't
found. __builtin_ffsl performs the same function as ffsl, and has the
added benefit of being available on any platform utilizing
Gcc-compatible compiler.
This change does not address the used of ffs in the MALLOCX_ARENA()
macro.
Add size class computation capability, currently used only as validation
of the size class lookup tables. Generalize the size class spacing used
for bins, for eventual use throughout the full range of allocation
sizes.
Refactor huge allocation to be managed by arenas (though the global
red-black tree of huge allocations remains for lookup during
deallocation). This is the logical conclusion of recent changes that 1)
made per arena dss precedence apply to huge allocation, and 2) made it
possible to replace the per arena chunk allocation/deallocation
functions.
Remove the top level huge stats, and replace them with per arena huge
stats.
Normalize function names and types to *dalloc* (some were *dealloc*).
Remove the --enable-mremap option. As jemalloc currently operates, this
is a performace regression for some applications, but planned work to
logarithmically space huge size classes should provide similar amortized
performance. The motivation for this change was that mremap-based huge
reallocation forced leaky abstractions that prevented refactoring.
Add new mallctl endpoints "arena<i>.chunk.alloc" and
"arena<i>.chunk.dealloc" to allow userspace to configure
jemalloc's chunk allocator and deallocator on a per-arena
basis.
Simplify backtracing to not ignore any frames, and compensate for this
in pprof in order to increase flexibility with respect to function-based
refactoring even in the presence of non-deterministic inlining. Modify
pprof to blacklist all jemalloc allocation entry points including
non-standard ones like mallocx(), and ignore all allocator-internal
frames. Prior to this change, pprof excluded the specifically
blacklisted functions from backtraces, but it left allocator-internal
frames intact.
Forcefully disable tcache if running inside Valgrind, and remove
Valgrind calls in tcache-specific code.
Restructure Valgrind-related code to move most Valgrind calls out of the
fast path functions.
Take advantage of static knowledge to elide some branches in
JEMALLOC_VALGRIND_REALLOC().
Make dss non-optional on all platforms which support sbrk(2).
Fix the "arena.<i>.dss" mallctl to return an error if "primary" or
"secondary" precedence is specified, but sbrk(2) is not supported.
Make promotion of sampled small objects to large objects mandatory, so
that profiling metadata can always be stored in the chunk map, rather
than requiring one pointer per small region in each small-region page
run. In practice the non-prof-promote code was only useful when using
jemalloc to track all objects and report them as leaks at program exit.
However, Valgrind is at least as good a tool for this particular use
case.
Furthermore, the non-prof-promote code is getting in the way of
some optimizations that will make heap profiling much cheaper for the
predominant use case (sampling a small representative proportion of all
allocations).
When you call free() we load chunk->arena even though that
data isn't used on the tcache hot path.
In profiling some FB applications, I found that ~30% of the
dTLB misses in the free() function come from this line. With
4 MB chunks, the arena_chunk_t->map is ~ 32 KB (1024 pages
in the chunk, 4 8 byte pointers in arena_chunk_map_t). This
means there's only a 1/8 chance of the page containing
chunk->arena also comtaining the map bits.
This happens when it fails to allocate a new chunk. Which
arena_chunk_alloc then passes into arena_avail_insert without any
checks. This then causes a crash when arena_avail_insert tries
to check chunk->ndirty.
This was introduced by the refactoring of arena_chunk_alloc
which previously would have returned NULL immediately after
calling chunk_alloc. This is now the return from
arena_chunk_init_hard so we need to check that return, and
not continue if it was NULL.
If mremap(2) is used for huge reallocation, physical pages are mapped to
new virtual addresses rather than data being copied to new pages. This
bypasses the normal junk filling that would happen during allocation, so
add junk filling that is specific to this case.
Refactor prof_dump() to use a two pass algorithm, and prof_leave() prior
to the second pass. This avoids write(2) system calls while holding
critical prof resources.
Fix prof_dump() to close the dump file descriptor for all relevant error
paths.
Minimize the size of prof-related static buffers when prof is disabled.
This saves roughly 65 KiB of application memory for non-prof builds.
Refactor prof_ctx_init() out of prof_lookup_global().
Refactor overly large functions by breaking out helper functions.
Refactor overly complex multi-purpose functions into separate more
specific functions.
Extract profiling code from malloc(), imemalign(), calloc(), realloc(),
mallocx(), rallocx(), and xallocx(). This slightly reduces the amount
of code compiled into the fast paths, but the primary benefit is the
combinatorial complexity reduction.
Simplify iralloc[t]() by creating a separate ixalloc() that handles the
no-move cases.
Further simplify [mrxn]allocx() (and by implication [mrn]allocm()) to
make request size overflows due to size class and/or alignment
constraints trigger undefined behavior (detected by debug-only
assertions).
Report ENOMEM rather than EINVAL if an OOM occurs during heap profiling
backtrace creation in imemalign(). This bug impacted posix_memalign()
and aligned_alloc().
Add unit tests for pow2_ceil(), malloc_strtoumax(), and
malloc_snprintf().
Fix numerous bugs in malloc_strotumax() error handling/reporting. These
bugs could have caused application-visible issues for some seldom used
(0X... and 0... prefixes) or malformed MALLOC_CONF or mallctl() argument
strings, but otherwise they had no impact.
Fix numerous bugs in malloc_snprintf(). These bugs were not exercised
by existing malloc_*printf() calls, so they had no impact.
Reduce rtree memory usage by storing booleans (1 byte each) rather than
pointers. The rtree code is only used to record whether jemalloc manages
a chunk of memory, so there's no need to store pointers in the rtree.
Increase rtree node size to 64 KiB in order to reduce tree depth from 13
to 3 on 64-bit systems. The conversion to more compact leaf nodes was
enough by itself to make the rtree depth 1 on 32-bit systems; due to the
fact that root nodes are smaller than the specified node size if
possible, the node size change has no impact on 32-bit systems (assuming
default chunk size).
Verify that freed regions are quarantined, and that redzone corruption
is detected.
Introduce a testing idiom for intercepting/replacing internal functions.
In this case the replaced function is ordinarily a static function, but
the idiom should work similarly for library-private functions.
Don't junk fill reallocations for which the request size is less than
the current usable size, but not enough smaller to cause a size class
change. Unlike malloc()/calloc()/realloc(), *allocx() contractually
treats the full usize as the allocation, so a caller can ask for zeroed
memory via mallocx() and a series of rallocx() calls that all specify
MALLOCX_ZERO, and be assured that all newly allocated bytes will be
zeroed and made available to the application without danger of allocator
mutation until the size class decreases enough to cause usize reduction.
Refactor such that arena_prof_ctx_set() receives usize as an argument,
and use it to determine whether to handle ptr as a small region, rather
than reading the chunk page map.
Implement the *allocx() API, which is a successor to the *allocm() API.
The *allocx() functions are slightly simpler to use because they have
fewer parameters, they directly return the results of primary interest,
and mallocx()/rallocx() avoid the strict aliasing pitfall that
allocm()/rallocx() share with posix_memalign(). The following code
violates strict aliasing rules:
foo_t *foo;
allocm((void **)&foo, NULL, 42, 0);
whereas the following is safe:
foo_t *foo;
void *p;
allocm(&p, NULL, 42, 0);
foo = (foo_t *)p;
mallocx() does not have this problem:
foo_t *foo = (foo_t *)mallocx(42, 0);
Add JEMALLOC_INLINE_C and use it instead of JEMALLOC_INLINE in .c files,
so that the annotated functions are always static.
Remove SFMT's inline-related macros and use jemalloc's instead, so that
there's no danger of interactions with jemalloc's definitions that
disable inlining for debug builds.
Refactor tests to use explicit testing assertions, rather than diff'ing
test output. This makes the test code a bit shorter, more explicitly
encodes testing intent, and makes test failure diagnosis more
straightforward.
Fix malloc_tsd_dalloc() to bypass tcache when dallocating, so that there
is no danger of causing tcache reincarnation during thread exit.
Whether this infinite loop occurs depends on the pthreads TSD
implementation; it is known to occur on Solaris.
Submitted by Markus Eberspächer.
When using LinuxThreads pthread_setspecific triggers recursive
allocation on all threads. Work around this by creating a global linked
list of in-progress tsd initializations.
This modifies the _tsd_get_wrapper macro-generated function. When it has
to initialize an TSD object it will push the item to the linked list
first. If this causes a recursive allocation then the _get_wrapper
request is satisfied from the list. When pthread_setspecific returns the
item is removed from the list.
This effectively adds a very poor substitute for real TLS used only
during pthread_setspecific allocation recursion.
Signed-off-by: Crestez Dan Leonard <lcrestez@ixiacom.com>
Add a missing mutex unlock in a malloc_init_hard() error path (failed
mutex initialization). In practice this bug was very unlikely to ever
trigger, but if it did, application deadlock would likely result.
Reported by Pat Lynch.
Fix a compiler warning in chunk_record() that was due to reading node
rather than xnode. In practice this did not cause any correctness
issue, but dataflow analysis in some compilers cannot tell that node and
xnode are always equal in cases that the read is reached.
Fix a race condition in the "arenas.extend" mallctl that could lead to
internal data structure corruption. The race could be hit if one
thread called the "arenas.extend" mallctl while another thread
concurrently triggered initialization of one of the lazily created
arenas.
Fix a Valgrind integration flaw that caused Valgrind warnings about
reads of uninitialized memory in internal zero-initialized data
structures (relevant to tcache and prof code).
