Commit Graph

3376 Commits

Author SHA1 Message Date
Qi Wang
602edd7566 Enabled -Wstrict-prototypes and fixed warnings. 2023-07-06 12:00:02 -07:00
Kevin Svetlitski
ebd7e99f5c Add a test-case for small profiled allocations
Validate that small allocations (i.e. those with `size <= SC_SMALL_MAXCLASS`)
which are sampled for profiling maintain the expected invariants even
though they now take up less space.
2023-07-03 16:19:06 -07:00
Kevin Svetlitski
5a858c64d6 Reduce the memory overhead of sampled small allocations
Previously, small allocations which were sampled as part of heap
profiling were rounded up to `SC_LARGE_MINCLASS`. This additional memory
usage becomes problematic when the page size is increased, as noted in #2358.

Small allocations are now rounded up to the nearest multiple of `PAGE`
instead, reducing the memory overhead by a factor of 4 in the most
extreme cases.
2023-07-03 16:19:06 -07:00
Kevin Svetlitski
e1338703ef Address compiler warnings in the unit tests 2023-07-03 16:06:35 -07:00
Qi Wang
d131331310 Avoid eager purging on the dedicated oversize arena when using bg thds.
We have observed new workload patterns (namely ML training type) that cycle
through oversized allocations frequently, because 1) the dataset might be sparse
which is faster to go through, and 2) GPU accelerated.  As a result, the eager
purging from the oversize arena becomes a bottleneck.  To offer an easy
solution, allow normal purging of the oversized extents when background threads
are enabled.
2023-06-27 11:57:41 -07:00
Kevin Svetlitski
46e464a26b Fix downloading LLVM in GitHub Action
It turns out LLVM does not include a build for every platform in the
assets for every release, just some of them. As such, I've pinned us to
the latest release version with a corresponding build.
2023-06-23 14:30:49 -07:00
Kevin Svetlitski
f2e00d2fd3 Remove trailing whitespace
Additionally, added a GitHub Action to ensure no more trailing
whitespace will creep in again in the future.

I'm excluding Markdown files from this check, since trailing whitespace
is significant there, and also excluding `build-aux/install-sh` because
there is significant trailing whitespace on the line that sets
`defaultIFS`.
2023-06-23 11:58:18 -07:00
Kevin Svetlitski
05385191d4 Add GitHub action which runs static analysis
Now that all of the various issues that static analysis uncovered have
been fixed (#2431, #2432, #2433, #2436, #2437, #2446), I've added a
GitHub action which will run static analysis for every PR going forward.
When static analysis detects issues with your code, the GitHub action
provides a link to download its findings in a form tailored for human
consumption.

Take a look at [this demonstration of what it looks like when static
analysis issues are
found](https://github.com/Svetlitski/jemalloc/actions/runs/5010245602)
on my fork for an example (make sure to follow the instructions in the
error message to download and inspect the results).
2023-06-23 11:55:43 -07:00
Kevin Svetlitski
bb0333e745 Fix remaining static analysis warnings
Fix or suppress the remaining warnings generated by static analysis.
This is a necessary step before we can incorporate static analysis into
CI. Where possible, I've preferred to modify the code itself instead of
just disabling the warning with a magic comment, so that if we decide to
use different static analysis tools in the future we will be covered
against them raising similar warnings.
2023-06-23 11:50:29 -07:00
Kevin Svetlitski
210f0d0b2b Fix read of uninitialized data in prof_free
In #2433, I inadvertently introduced a regression which causes the use of
uninitialized data. Namely, the control path I added for the safety
check in `arena_prof_info_get` neglected to set `prof_info->alloc_tctx`
when the check fails, resulting in `prof_info.alloc_tctx` being
uninitialized [when it is read at the end of
`prof_free`](90176f8a87/include/jemalloc/internal/prof_inlines.h (L272)).
2023-06-15 18:30:05 -07:00
Kevin Svetlitski
90176f8a87 Fix segfault in rb *_tree_remove
Static analysis flagged this. It's possible to segfault in the
`*_tree_remove` function generated by `rb_gen`, as `nodep` may
still be `NULL` after the initial for loop. I can confirm from reviewing
the fleetwide coredump data that this was in fact being hit in
production, primarily through `tctx_tree_remove`, and much more rarely
through `gctx_tree_remove`.
2023-06-07 14:48:41 -07:00
Qi Wang
86eb49b478 Fix the arena selection for oversized allocations.
Use the per-arena oversize_threshold, instead of the global setting.
2023-06-06 15:03:13 -07:00
Christos Zoulas
5832ef6589 Use a local variable to set the alignment for this particular allocation
instead of changing mmap_flags which makes the change permanent. This was
enforcing large alignments for allocations that did not need it causing
fragmentation. Reported by Andreas Gustafsson.
2023-05-31 14:44:24 -07:00
Kevin Svetlitski
6d4aa33753 Extract the calculation of psset heap assignment for an hpdata into a common function
This is in preparation for upcoming changes I plan to make to this
logic. Extracting it into a common function will make this easier and
less error-prone, and cleans up the existing code regardless.
2023-05-31 11:44:04 -07:00
Arne Welzel
c1d3ad4674 Prune je_malloc_default and do_rallocx in jeprof
Running a simple Ruby and Python execution je_malloc_default and
do_rallocx() in the resulting SVG / text output. Prune these, too.

    MALLOC_CONF='stats_print:true,lg_prof_sample:8,prof:true,prof_final:true' \
        python3 -c '[x for x in range(10000000)]'

    MALLOC_CONF='stats_print:true,lg_prof_sample:8,prof:true,prof_final:true' \
        ruby -e 'puts (0..1000).map{"0"}.join(" ")'
2023-05-31 11:41:09 -07:00
Arne Welzel
d59e30cbc9 Rename fallback_impl to fallbackNewImpl and prune in jeprof
The existing fallback_impl name seemed a bit generic and given
it's static probably okay to rename.

Closes #2451
2023-05-31 11:41:09 -07:00
Qi Wang
d577e9b588 Explicitly cast to unsigned for MALLOCX_ARENA and _TCACHE defines. 2023-05-26 11:52:42 -07:00
Qi Wang
a2259f9fa6 Fix the include path of "jemalloc_internal_overrides.h". 2023-05-25 15:22:02 -07:00
Kevin Svetlitski
9c32689e57 Fix bug where hpa_shard was not being destroyed
It appears that this was a simple mistake where `hpa_shard_disable` was
being called instead of `hpa_shard_destroy`. At present
`hpa_shard_destroy` is not called anywhere at all outside of test-cases,
which further suggests that this is a bug. @davidtgoldblatt noted
however that since HPA is disabled for manual arenas and we don't
support destruction for auto arenas that presently there is no way to
actually trigger this bug. Nonetheless, it should be fixed.
2023-05-18 14:17:38 -07:00
Kevin Svetlitski
4e6f1e9208 Allow overriding LG_PAGE
This is useful for our internal builds where we override the
configuration in the header files generated by autoconf.
2023-05-17 13:55:38 -07:00
Kevin Svetlitski
3e2ba7a651 Remove dead stores detected by static analysis
None of these are harmful, and they are almost certainly optimized
away by the compiler. The motivation for fixing them anyway is that
we'd like to enable static analysis as part of CI, and the first step
towards that is resolving the warnings it produces at present.
2023-05-11 20:27:49 -07:00
Kevin Svetlitski
0288126d9c Fix possible NULL pointer dereference from mallctl("prof.prefix", ...)
Static analysis flagged this issue. Here is a minimal program which
causes a segfault within Jemalloc:
```
#include <jemalloc/jemalloc.h>

const char *malloc_conf = "prof:true";

int main() {
  mallctl("prof.prefix", NULL, NULL, NULL, 0);
}
```

Fixed by checking if `prefix` is `NULL`.
2023-05-11 14:47:50 -07:00
Qi Wang
d4a2b8bab1 Add the prof_sys_thread_name feature in the prof_recent unit test.
This tests the combination of the prof_recent and thread_name features.
Verified that it catches the issue being fixed in this PR.

Also explicitly set thread name in test/unit/prof_recent.  This fixes the name
testing when no default thread name is set (e.g. FreeBSD).
2023-05-11 09:10:57 -07:00
Qi Wang
94ace05832 Fix the prof thread_name reference in prof_recent dump.
As pointed out in #2434, the thread_name in prof_tdata_t was changed in #2407.
This also requires an update for the prof_recent dump, specifically the emitter
expects a "char **" which is fixed in this commit.
2023-05-11 09:10:57 -07:00
Qi Wang
6ea8a7e928 Add config detection for JEMALLOC_HAVE_PTHREAD_SET_NAME_NP.
and use it on the background thread name setting.
2023-05-11 09:10:57 -07:00
auxten
5bac384970 If ptr present check if alloc_ctx.edata == NULL 2023-05-10 17:18:22 -07:00
auxten
019cccc293 Make arenas_lookup_ctl triable 2023-05-10 17:18:22 -07:00
Kevin Svetlitski
dc0a184f8d Fix possible NULL pointer dereference in VERIFY_READ
Static analysis flagged this. Fixed by simply checking `oldlenp`
before dereferencing it.
2023-05-09 10:57:09 -07:00
Kevin Svetlitski
12311fe6c3 Fix segfault in extent_try_coalesce_impl
Static analysis flagged this. `extent_record` was passing `NULL` as the
value for `coalesced` to `extent_try_coalesce`, which in turn passes
that argument to `extent_try_coalesce_impl`, where it is written to
without checking if it is `NULL`. I can confirm from reviewing the
fleetwide coredump data that this was in fact being hit in production.
2023-05-09 10:55:44 -07:00
Kevin Svetlitski
70344a2d38 Make eligible functions static
The codebase is already very disciplined in making any function which
can be `static`, but there are a few that appear to have slipped through
the cracks.
2023-05-08 15:00:02 -07:00
Kevin Svetlitski
6841110bd6 Make edata_cmp_summary_comp 30% faster
`edata_cmp_summary_comp` is one of the very hottest functions, taking up
3% of all time spent inside Jemalloc. I noticed that all existing
callsites rely only on the sign of the value returned by this function,
so I came up with this equivalent branchless implementation which
preserves this property. After empirical measurement, I have found that
this implementation is 30% faster, therefore representing a 1% speed-up
to the allocator as a whole.

At @interwq's suggestion, I've applied the same optimization to
`edata_esnead_comp` in case this function becomes hotter in the future.
2023-05-04 09:59:17 -07:00
Amaury Séchet
f2b28906e6 Some nits in cache_bin.h 2023-05-01 10:21:17 -07:00
Kevin Svetlitski
fc680128e0 Remove errant assert in arena_extent_alloc_large
This codepath may generate deferred work when the HPA is enabled.
See also [@davidtgoldblatt's relevant comment on the PR which
introduced this](https://github.com/jemalloc/jemalloc/pull/2107#discussion_r699770967)
which prevented a similarly incorrect `assert` from being added elsewhere.
2023-05-01 10:00:30 -07:00
Eric Mueller
521970fb2e Check for equality instead of assigning in asserts in hpa_from_pai.
It appears like a simple typo means we're unconditionally overwriting
some fields in hpa_from_pai when asserts are enabled. From hpa_shard_init,
it looks like these fields have these values anyway, so this shouldn't
cause bugs, but if something is wrong it seems better to have these
asserts in place.

See issue #2412.
2023-04-17 20:57:48 -07:00
guangli-dai
5f64ad60cd Remove locked flag set in malloc_mutex_trylock
As a hint flag of the lock, parameter locked should be set only
when the lock is gained or freed.
2023-04-06 10:57:04 -07:00
Qi Wang
434a68e221 Disallow decay during reentrancy.
Decay should not be triggered during reentrant calls (may cause lock order
reversal / deadlocks).  Added a delay_trigger flag to the tickers to bypass
decay when rentrancy_level is not zero.
2023-04-05 10:16:37 -07:00
Qi Wang
e62aa478c7 Rearrange the bools in prof_tdata_t to save some bytes.
This lowered the sizeof(prof_tdata_t) from 200 to 192 which is a round size
class.  Afterwards the tdata_t size remain unchanged with the last commit, which
effectively inlined the storage of thread names for free.
2023-04-05 10:03:12 -07:00
Qi Wang
ce0b7ab6c8 Inline the storage for thread name in prof_tdata_t.
The previous approach managed the thread name in a separate buffer, which causes
races because the thread name update (triggered by new samples) can happen at
the same time as prof dumping (which reads the thread names) -- these two
operations are under separate locks to avoid blocking each other.  Implemented
the thread name storage as part of the tdata struct, which resolves the lifetime
issue and also avoids internal alloc / dalloc during prof_sample.
2023-04-05 10:03:12 -07:00
Qi Wang
6cab460a45 Add a multithreaded test for prof_sys_thread_name.
Verified that this catches the issue being fixed in 5fd5583.
2023-04-05 10:03:12 -07:00
Amaury Séchet
5266152d79 Simplify the logic in ph_remove 2023-03-31 14:35:31 -07:00
Amaury Séchet
be6da4f663 Do not maintain root->prev in ph_remove. 2023-03-31 14:34:57 -07:00
Amaury Séchet
543e2d61e6 Simplify the logic in ph_insert
Also fixes what looks like an off by one error in the lazy aux list
merge part of the code that previously never touched the last node in
the aux list.
2023-03-31 14:34:24 -07:00
guangli-dai
31e01a98f1 Fix the rdtscp detection bug and add prefix for the macro. 2023-03-23 11:16:19 -07:00
Qi Wang
8b64be3441 Explicit arena assignment in test_tcache_max.
Otherwise the associated arena could change with percpu arena enabled.
2023-03-22 15:16:43 -07:00
Qi Wang
8e7353a19b Explicit arena assignment in test_thread_idle.
Otherwise the associated arena could change with percpu arena enabled.
2023-03-22 15:16:43 -07:00
Marvin Schmidt
45249cf5a9 Fix exception specification error for hosts using musl libc
It turns out that the previous commit did not suffice since the
JEMALLOC_SYS_NOTHROW definition also causes the same exception specification
errors as JEMALLOC_USE_CXX_THROW did:
```
x86_64-pc-linux-musl-cc -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 -fvisibility=hidden -Wimplicit-fallthrough -O3 -funroll-loops -march=native -O2 -pipe -c -march=native -O2 -pipe -D_GNU_SOURCE -D_REENTRANT -Iinclude -Iinclude -o src/background_thread.o src/background_thread.c
In file included from src/jemalloc_cpp.cpp:9:
In file included from include/jemalloc/internal/jemalloc_preamble.h:27:
include/jemalloc/internal/../jemalloc.h:254:32: error: exception specification in declaration does not match previous declaration
    void JEMALLOC_SYS_NOTHROW   *je_malloc(size_t size)
                                 ^
include/jemalloc/internal/../jemalloc.h:75:21: note: expanded from macro 'je_malloc'
                    ^
/usr/x86_64-pc-linux-musl/include/stdlib.h:40:7: note: previous declaration is here
void *malloc (size_t);
      ^
```

On systems using the musl C library we have to omit the exception specification
on malloc function family like it's done for MacOS, FreeBSD and OpenBSD.
2023-03-16 12:11:40 -07:00
Marvin Schmidt
aba1645f2d configure: Handle *-linux-musl* hosts properly
This is the same as the `*-*-linux*` case with the two exceptions that
we don't set glibc=1 and don't define JEMALLOC_USE_CXX_THROW
2023-03-16 12:11:40 -07:00
Qi Wang
d503d72129 Add the missing descriptions in AC_DEFINE 2023-03-14 16:47:00 -07:00
Qi Wang
71bc1a3d91 Avoid assuming the arena id in test when percpu_arena is used. 2023-03-13 10:50:10 -07:00
Amaury Séchet
f743690739 Remove unused mutex from hpa_central 2023-03-10 11:25:47 -08:00