For the sake of consistency, function definitions and their
corresponding declarations should use the same names for parameters.
I've enabled this check in static analysis to prevent this issue from
occurring again in the future.
Previously if a thread does only allocations, it stays on the slow path /
minimal initialized state forever. However, dealloc-only is a valid pattern for
dedicated reclamation threads -- this means thread cache is disabled (no batched
flush) for them, which causes high overhead and contention.
Added the condition to fully initialize TSD when a fair amount of dealloc
activities are observed.
Before this commit, in case FreeBSD libc jemalloc was overridden by another
jemalloc, proper thread shutdown callback was involved only for the overriding
jemalloc. A call to _malloc_thread_cleanup from libthr would be redirected to
user jemalloc, leaving data about dead threads hanging in system jemalloc. This
change tackles the issue in two ways. First, for current and old system
jemallocs, which we can not modify, the overriding jemalloc would locate and
invoke system cleanup routine. For upcoming jemalloc integrations, the cleanup
registering function will also be redirected to user jemalloc, which means that
system jemalloc's cleanup routine will be registered in user's jemalloc and a
single call to _malloc_thread_cleanup will be sufficient to invoke both
callbacks.
Adding guarded extents, which are regular extents surrounded by guard pages
(mprotected). To reduce syscalls, small guarded extents are cached as a
separate eset in ecache, and decay through the dirty / muzzy / retained pipeline
as usual.
This (experimental, undocumented) functionality can be used by users to track
various statistics of interest at a finer level of granularity than the thread.
Performance-sensitive users will use sized deallocation facilities, so that
actually touching the rtree_ctx is unnecessary. We make it the last element of
the slow data, so that it is for practical purposes almost-fast.
Make the event module to accept two event types, and pass around the event
context. Use bytes-based events to trigger tcache GC on deallocation, and get
rid of the tcache ticker.
Add options stats_interval and stats_interval_opts to allow interval based stats
printing. This provides an easy way to collect stats without code changes,
because opt.stats_print may not work (some binaries never exit).
We have to work to circumvent the safety checks in pre_reentrancy when going
down extent hook pathways. Instead, let's explicitly have checked and unchecked
guards.
Fold the tsd_state check onto the event threshold check. The fast threshold is
set to 0 when tsd switch to non-nominal.
The fast_threshold can be reset by remote threads, to refect the non nominal tsd
state change.
Augmented the tsd layout graph so that the two recently added fields,
`offset_state` and `bytes_until_sample`, are properly reflected.
As is shown, the cache footprint is 16 bytes larger than before.
This makes it possible to have multiple set of bins in an arena, which improves
arena scalability because the bins (especially the small ones) are always the
limiting factor in production workload.
A bin shard is picked on allocation; each extent tracks the bin shard id for
deallocation. The shard size will be determined using runtime options.
generation of sub bytes_until_sample, usize; je; for x86 arch.
Subtraction is unconditional, and only flags are checked for the jump,
no extra compare is necessary. This also reduces register pressure.
In case of multithreaded fork, we want to leave the child in a reasonable state,
in which tsd_nominal_tsds is either empty or contains only the forking thread.
Previously, we made the user deal with this themselves, but that's not good
enough; if hooks may allocate, we should test the allocation pathways down
hooks. If we're doing that, we might as well actually implement the protection
for the user.
While working on #852, I noticed the prng state is atomic. This is the only
atomic use of prng in all of jemalloc. Instead, use a threadlocal prng
state if possible to avoid unnecessary cache line contention.
We use the minimal_initilized tsd (which requires no cleanup) for free()
specifically, if tsd hasn't been initialized yet.
Any other activity will transit the state from minimal to normal. This is to
workaround the case where a thread has no malloc calls in its lifetime until
during thread termination, free() happens after tls destructors.
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.
