Commit Graph

1827 Commits

Author SHA1 Message Date
Amaury Séchet
f743690739 Remove unused mutex from hpa_central 2023-03-10 11:25:47 -08:00
Qi Wang
c7805f1eb5 Add a header in HPA stats for the nonfull slabs. 2023-02-17 13:31:27 -08:00
Qi Wang
b6125120ac Add an explicit name to the dedicated oversize arena. 2023-02-17 13:31:09 -08:00
Qi Wang
5fd55837bb Fix thread_name updating for heap profiling.
The current thread name reading path updates the name every time, which requires
both alloc and dalloc -- and the temporary NULL value in the middle causes races
where the prof dump read path gets NULLed in the middle.

Minimize the changes in this commit to isolate the bugfix testing; will also
refactor the whole thread name paths later.
2023-02-15 17:49:40 -08:00
Qi Wang
8580c65f81 Implement prof sample hooks "experimental.hooks.prof_sample(_free)".
The added hooks hooks.prof_sample and hooks.prof_sample_free are intended to
allow advanced users to track additional information, to enable new ways of
profiling on top of the jemalloc heap profile and sample features.

The sample hook is invoked after the allocation and backtracing, and forwards
the both the allocation and backtrace to the user hook; the sample_free hook
happens before the actual deallocation, and forwards only the ptr and usz to the
hook.
2022-12-07 16:06:49 -08:00
Guangli Dai
e8f9f13811 Inline free and sdallocx into operator delete 2022-11-21 11:14:05 -08:00
Qi Wang
481bbfc990 Add a configure option --enable-force-getenv.
Allows the use of getenv() rather than secure_getenv() to read MALLOC_CONF.
This helps in situations where hosts are under full control, and setting
MALLOC_CONF is needed while also setuid.  Disabled by default.
2022-11-04 13:37:14 -07:00
Qi Wang
143e9c4a2f Enable fast thread locals for dealloc-only threads.
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.
2022-10-25 09:54:38 -07:00
David Carlier
4c95c953e2 fix build for non linux/BSD platforms. 2022-10-03 10:42:09 -07:00
Guangli Dai
ba19d2cb78 Add arena-level name.
An arena-level name can help identify manual arenas.
2022-09-16 15:04:59 -07:00
Guangli Dai
a0734fd6ee Making jemalloc max stack depth a runtime option 2022-09-12 13:56:22 -07:00
Abael He
56ddbea270 error: implicit declaration of function 'pthread_create_fptr_init' is invalid in C99
./autogen.sh \
&& ./configure --prefix=/usr/local  --enable-static   --enable-autogen --enable-xmalloc --with-static-libunwind=/usr/local/lib/libunwind.a --enable-lazy-lock --with-jemalloc-prefix='' \
&& make -j16

...
gcc -std=gnu11 -Werror=unknown-warning-option -Wall -Wextra -Wshorten-64-to-32 -Wsign-compare -Wundef -Wno-format-zero-length -Wpointer-arith -Wno-missing-braces -Wno-missing-field-initializers -pipe -g3 -Wimplicit-fallthrough -O3 -funroll-loops -fPIC -DPIC -c -D_REENTRANT -Iinclude -Iinclude -DJEMALLOC_NO_PRIVATE_NAMESPACE -o src/edata_cache.sym.o src/edata_cache.c
src/background_thread.c:768:6: error: implicit declaration of function 'pthread_create_fptr_init' is invalid in C99 [-Werror,-Wimplicit-function-declaration]
            pthread_create_fptr_init()) {
            ^
src/background_thread.c:768:6: note: did you mean 'pthread_create_wrapper_init'?
src/background_thread.c:34:1: note: 'pthread_create_wrapper_init' declared here
pthread_create_wrapper_init(void) {
^
1 error generated.
make: *** [src/background_thread.sym.o] Error 1
make: *** Waiting for unfinished jobs....
2022-09-07 11:56:41 -07:00
Guangli Dai
ce29b4c3d9 Refactor the remote / cross thread cache bin stats reading
Refactored cache_bin.h so that only one function is racy.
2022-09-06 19:41:19 -07:00
Ivan Zaitsev
36366f3c4c Add double free detection in thread cache for debug build
Add new runtime option `debug_double_free_max_scan` that specifies the max
number of stack entries to scan in the cache bit when trying to detect the
double free bug (currently debug build only).
2022-08-04 16:58:22 -07:00
David Carlier
58478412be OpenBSD build fix. still no cpu affinity.
- enabling pthread_get/pthread_set_name_np api.
- disabling per thread cpu affinity handling, unsupported on this platform.
2022-07-19 13:20:11 -07:00
Qi Wang
a1c7d9c046 Add the missing opt.cache_oblivious handling. 2022-07-14 22:41:27 -07:00
Azat Khuzhin
cb578bbe01 Fix possible "nmalloc >= ndalloc" assertion
In arena_stats_merge() first nmalloc was read, and after ndalloc.

However with this order, it is possible for some thread to incement
ndalloc in between, and then nmalloc < ndalloc, and assertion will fail,
like again found by ClickHouse CI [1] (even after #2234).

  [1]: https://github.com/ClickHouse/ClickHouse/issues/31531

Swap the order to avoid possible assertion.

Cc: @interwq
Follow-up for: #2234
2022-07-11 15:27:51 -07:00
David Carlier
4fc5c4fbac New configure option '--enable-pageid' for Linux
The option makes jemalloc use prctl with PR_SET_VMA to tag memory mappings with
"jemalloc_pg" or "jemalloc_pg_overcommit". This allows to easily identify
jemalloc's mappings in /proc/<pid>/maps. PR_SET_VMA is only available in Linux
5.17 and above.
2022-06-09 18:54:08 -07:00
Alex Lapenkou
5b1f2cc5d7 Implement pvalloc replacement
Despite being an obsolete function, pvalloc is still present in GLIBC and should
work correctly when jemalloc replaces libc allocator.
2022-05-18 17:01:09 -07:00
Qi Wang
cd5aaf308a Improve the failure message upon opt_experimental_infallible_new. 2022-05-17 16:07:40 -07:00
Qi Wang
8cb814629a Make the default option of zero realloc match the system allocator. 2022-05-05 17:11:18 -07:00
Qi Wang
391bad4b95 Avoid abort() in test/integration/cpp/infallible_new_true.
Allow setting the safety check abort hook through mallctl, which avoids abort()
and core dumps.
2022-04-25 11:29:32 -07:00
cuishuang
9a242f16d9 fix some typos
Signed-off-by: cuishuang <imcusg@gmail.com>
2022-04-25 11:29:00 -07:00
Qi Wang
0e29ad4efa Rename zero_realloc option "strict" to "alloc".
With realloc(ptr, 0) being UB per C23, the option name "strict" makes less sense
now.  Rename to "alloc" which describes the behavior.
2022-04-20 10:27:25 -07:00
Alex Lapenkou
a93931537e Do not disable SEC by default for 64k pages platforms
Default SEC max_alloc option value was 32k, disabling SEC for platforms with
lg-page=16. This change enables SEC for all platforms, making minimum max_alloc
value equal to PAGE.
2022-03-24 22:05:35 -07:00
Charles
eaaa368bab Add comments and use meaningful vars in sz_psz2ind. 2022-03-24 16:56:59 -07:00
Alex Lapenkou
5bf03f8ce5 Implement PAGE_FLOOR macro 2022-03-22 17:45:55 -07:00
Alex Lapenkou
52631c90f6 Fix size class calculation for sec
Due to a bug in sec initialization, the number of cached size classes
was equal to 198. The bug caused the creation of more than a hundred of
unused bins, although it didn't affect the caching logic.
2022-03-22 17:45:55 -07:00
Alex Lapenkov
eb65d1b078 Fix FreeBSD system jemalloc TSD cleanup
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.
2022-03-02 10:10:27 -08:00
Azat Khuzhin
78b58379c8 Fix possible "nmalloc >= ndalloc" assertion.
It is possible that ndalloc will be updated before nmalloc, in
arena_large_ralloc_stats_update(), fix this by reorder those calls.

It was found by ClickHouse CI, that periodically hits this assertion [1].

  [1]: https://github.com/ClickHouse/ClickHouse/issues/31531

That issue contains lots of examples, with core dump and some gdb output [2].

  [2]: https://s3.amazonaws.com/clickhouse-test-reports/34951/96390a9263cb5af3d6e42a84988239c9ae87ce32/stress_test__debug__actions_.html

Here you can find binaries for that particular report [3] you need
clickhouse debug build [4].

  [3]: https://s3.amazonaws.com/clickhouse-builds/34951/96390a9263cb5af3d6e42a84988239c9ae87ce32/clickhouse_build_check_(actions)/report.html
  [4]: https://s3.amazonaws.com/clickhouse-builds/34951/96390a9263cb5af3d6e42a84988239c9ae87ce32/package_debug/clickhouse

Brief info from that report:

    2 0x000000002ad6dbfe in arena_stats_merge (tsdn=0x7f2399abdd20, arena=0x7f241ce01080, nthreads=0x7f24e4360958, dss=0x7f24e4360960, dirty_decay_ms=0x7f24e4360968, muzzy_decay_ms=0x7f24e4360970, nactive=0x7f24e4360978, ndirty=0x7f24e43
    e4360988, astats=0x7f24e4360998, bstats=0x7f24e4363310, lstats=0x7f24e4364990, estats=0x7f24e4366e50, hpastats=0x7f24e43693a0, secstats=0x7f24e436a020) at ../contrib/jemalloc/src/arena.c:138
            ndalloc = 226
            nflush = 0
            curlextents = 0
            nmalloc = 225
            nrequests = 0

Here you can see that they differs only by 1.

Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
2022-03-01 12:28:28 -08:00
Alex Lapenkou
ca709c3139 Fix failed assertion due to racy memory access
While calculating the number of stashed pointers, multiple variables
potentially modified by a concurrent thread were used for the
calculation.  This led to some inconsistencies, correctly detected by
the assertions.  The change eliminates some possible inconsistencies by
using unmodified variables and only once a concurrently modified one.
The assertions are omitted for the cases where we acknowledge potential
inconsistencies too.
2022-02-17 09:35:52 -08:00
Qi Wang
8c59c44ffa Add a dependency checking step at the end of malloc_conf_init.
Currently only prof_leak_error and prof_final are checked.
2022-01-26 17:17:48 -08:00
Qi Wang
efc539c040 Initialize prof_leak during prof init.
Otherwise, prof_leak may get set after prof_leak_error, and disagree with each
other.
2022-01-26 17:17:48 -08:00
yunxu
b798fabdf7 Add prof_leak_error option
The option makes the process to exit with error code 1 if a memory leak
is detected. This is useful for implementing automated tools that rely
on leak detection.
2022-01-21 16:24:20 -08:00
Charles
eb196815d6 Avoid calculating size of size class twice & delete sc_data_global. 2022-01-18 11:54:12 -08:00
Qi Wang
ddb170b1d9 Simplify arena_migrate() to take arena_t* instead of indices.
This makes debugging slightly easier and avoids the confusion of "should we
create new arenas" here.
2022-01-11 16:59:22 -08:00
Qi Wang
d66162e032 Fix the extent state checking on the merge error path.
With DSS as primary, the default merge impl will (correctly) decline to merge
when one of the extent is non-dss.  The error path should tolerate the
not-merged extent being in a merging state.
2022-01-11 16:58:47 -08:00
Qi Wang
61978bbe69 Purge all if the last thread migrated away from an arena. 2022-01-06 19:02:26 -08:00
Yuriy Chernyshov
c91e62dd37 #include <features.h> as requested 2022-01-05 18:45:27 -08:00
Yuriy Chernyshov
18510020e7 Fix symbol conflict with musl libc
`__libc` prefixed functions are used by musl libc as non-replaceable malloc stubs.

Fix this conflict by checking if we are linking against glibc.
2022-01-05 18:45:27 -08:00
Qi Wang
f509703af5 Fix two conversion warnings in tcache. 2022-01-04 13:55:06 -08:00
Qi Wang
8b34a788b5 Fix an used-uninitialized warning (false positive). 2021-12-29 14:44:43 -08:00
Qi Wang
e491cef9ab Add stats for stashed bytes in tcache. 2021-12-29 14:44:43 -08:00
Qi Wang
b75822bc6e Implement use-after-free detection using junk and stash.
On deallocation, sampled pointers (specially aligned) get junked and stashed
into tcache (to prevent immediate reuse).  The expected behavior is to have
read-after-free corrupted and stopped by the junk-filling, while
write-after-free is checked when flushing the stashed pointers.
2021-12-29 14:44:43 -08:00
Qi Wang
06aac61c4b Split the core logic of tcache flush into a separate function.
The core function takes a ptr array as input (containing items to be flushed),
which will be reused to flush sanitizer-stashed items.
2021-12-29 14:44:43 -08:00
Qi Wang
d038160f3b Fix shadowed variable usage.
Verified with EXTRA_CFLAGS=-Wshadow.
2021-12-23 10:55:08 -08:00
Qi Wang
60b9637cc0 Only invoke malloc_cpu_count_is_deterministic() when necessary.
Also refactor the handling of the non-deterministic case.  Notably allow the
case with narenas set to proceed w/o warnings, to not affect existing valid use
cases.
2021-12-22 13:52:12 -08:00
Qi Wang
837b37c4ce Fix the time-since computation in HPA.
nstime module guarantees monotonic clock update within a single nstime_t.  This
means, if two separate nstime_t variables are read and updated separately,
nstime_subtract between them may result in underflow.  Fixed by switching to the
time since utility provided by nstime.
2021-12-21 23:37:22 -08:00
Qi Wang
310af725b0 Add nstime_ns_since which obtains the duration since the input time. 2021-12-21 23:37:22 -08:00
Azat Khuzhin
cafe9a3158 Disable percpu arena in case of non deterministic CPU count
Determinitic number of CPUs is important for percpu arena to work
correctly, since it uses cpu index - sched_getcpu(), and if it will
greater then number of CPUs bad thing will happen, or assertion will be
failed in debug build:

    <jemalloc>: ../contrib/jemalloc/src/jemalloc.c:321: Failed assertion: "ind <= narenas_total_get()"
    Aborted (core dumped)

Number of CPUs can be obtained from the following places:
- sched_getaffinity()
- sysconf(_SC_NPROCESSORS_ONLN)
- sysconf(_SC_NPROCESSORS_CONF)

For the sched_getaffinity() you may simply use taskset(1) to run program
on a different cpu, and in case it will be not first, percpu will work
incorrectly, i.e.:

    $ taskset --cpu-list $(( $(getconf _NPROCESSORS_ONLN)-1 )) <your_program>

_SC_NPROCESSORS_ONLN uses /sys/devices/system/cpu/online, LXD/LXC
virtualize /sys/devices/system/cpu/online file [1], and so when you run
container with limited limits.cpus it will bind randomly selected CPU to
it

  [1]: https://github.com/lxc/lxcfs/issues/301

_SC_NPROCESSORS_CONF uses /sys/devices/system/cpu/cpu*, and AFAIK nobody
playing with dentries there.

So if all three of these are equal, percpu arenas should work correctly.

And a small note regardless _SC_NPROCESSORS_ONLN/_SC_NPROCESSORS_CONF,
musl uses sched_getaffinity() for both. So this will also increase the
entropy.

Also note, that you can check is percpu arena really applied using
abort_conf:true.

Refs: https://github.com/jemalloc/jemalloc/pull/1939
Refs: https://github.com/ClickHouse/ClickHouse/issues/32806

v2: move malloc_cpu_count_is_deterministic() into
    malloc_init_hard_recursible() since _SC_NPROCESSORS_CONF does
    allocations for readdir()
v3:
- mark cpu_count_is_deterministic static
- check only if percpu arena is enabled
- check narenas
2021-12-21 11:53:09 -08:00