Commit Graph

18 Commits

Author SHA1 Message Date
Yinan Zhang
b549389e4a Correct usize in prof last-N record 2020-09-09 13:31:35 -07:00
David Goldblatt
60993697d8 Prof: Add prof_unbias.
This gives more accurate attribution of bytes and counts to stack traces,
without introducing backwards incompatibilities in heap-profile parsing tools.
We track the ideal reported (to the end user) number of bytes more carefully
inside core jemalloc.  When dumping heap profiles, insteading of outputting our
counts directly, we output counts that will cause parsing tools to give a result
close to the value we want.

We retain the old version as an opt setting, to let users who are tracking
values on a per-component basis to keep their metrics stable until they decide
to switch.
2020-08-05 18:33:55 -07:00
Yinan Zhang
857ebd3daf Make edata pointer on prof recent record an atomic fence 2020-06-09 17:03:05 -07:00
Yinan Zhang
a5ddfa7d91 Use ql for prof last-N list 2020-04-07 17:22:24 -07:00
Qi Wang
d71a145ec1 Chagne prof_accum_t to counter_accum_t for general purpose. 2020-01-29 09:57:55 -08:00
Yinan Zhang
2b604a3016 Record request size in prof recent entries 2020-01-10 12:01:01 -08:00
Yinan Zhang
9a60cf54ec Last-N profiling mode 2019-12-30 15:58:57 -08:00
Yinan Zhang
aa1d71fb7a Rename prof_tctx to alloc_tctx in prof_info_t 2019-12-06 09:47:51 -08:00
Yinan Zhang
b55419f9b9 Restructure profiling
Develop new data structure and code logic for holding profiling
related information stored in the extent that may be needed after the
extent is released, which in particular is the case for the
reallocation code path (e.g. in `rallocx()` and `xallocx()`).  The
data structure is a generalization of `prof_tctx_t`: we previously
only copy out the `prof_tctx` before the extent is released, but we
may be in need of additional fields. Currently the only additional
field is the allocation time field, but there may be more fields in
the future.

The restructuring also resolved a bug: `prof_realloc()` mistakenly
passed the new `ptr` to `prof_free_sampled_object()`, but passing in
the `old_ptr` would crash because it's already been released.  Now
the essential profiling information is collectively copied out early
and safely passed to `prof_free_sampled_object()` after the extent is
released.
2019-11-22 16:31:56 -08:00
Qi Wang
da50d8ce87 Refactor and optimize prof sampling initialization.
Makes the prof sample prng use the tsd prng_state.  This allows us to properly
initialize the sample interval event, without having to create tdata.  As a
result, tdata will be created on demand (when a thread reaches the sample
interval bytes allocated), instead of on the first allocation.
2019-11-11 10:35:37 -08:00
Dave Watson
9ed3bdc848 move bytes until sample to tsd. Fastpath allocation does not need
to load tdata now, avoiding several branches.
2018-10-15 08:24:12 -07:00
David Goldblatt
18ecbfa89e Header refactoring: unify and de-catchall mutex module 2017-05-24 15:27:30 -07:00
David Goldblatt
68da2361d2 Header refactoring: ckh module - remove from the catchall and unify. 2017-04-24 10:33:21 -07:00
David Goldblatt
fa3ad730c4 Header refactoring: prng module - remove from the catchall and unify. 2017-04-24 10:33:21 -07:00
David Goldblatt
63a5cd4cc2 Header refactoring: break out rb.h dependencies 2017-04-11 11:52:30 -07:00
David Goldblatt
30d74db08e Convert accumbytes in prof_accum_t to C11 atomics, when possible 2017-04-05 16:25:37 -07:00
Jason Evans
fa2d64c94b Convert arena->prof_accumbytes synchronization to atomics. 2017-02-16 09:39:46 -08:00
David Goldblatt
77cccac8cd Break up headers into constituent parts
This is part of a broader change to make header files better represent the
dependencies between one another (see
https://github.com/jemalloc/jemalloc/issues/533). It breaks up component headers
into smaller parts that can be made to have a simpler dependency graph.

For the autogenerated headers (smoothstep.h and size_classes.h), no splitting
was necessary, so I didn't add support to emit multiple headers.
2017-01-12 15:43:51 -08:00