Commit Graph

191 Commits

Author SHA1 Message Date
Qi Wang
ec532e2c5c Implement per-CPU arena.
The new feature, opt.percpu_arena, determines thread-arena association
dynamically based CPU id. Three modes are supported: "percpu", "phycpu"
and disabled.

"percpu" uses the current core id (with help from sched_getcpu())
directly as the arena index, while "phycpu" will assign threads on the
same physical CPU to the same arena. In other words, "percpu" means # of
arenas == # of CPUs, while "phycpu" has # of arenas == 1/2 * (# of
CPUs). Note that no runtime check on whether hyper threading is enabled
is added yet.

When enabled, threads will be migrated between arenas when a CPU change
is detected. In the current design, to reduce overhead from reading CPU
id, each arena tracks the thread accessed most recently. When a new
thread comes in, we will read CPU id and update arena if necessary.
2017-03-08 23:19:01 -08:00
David Goldblatt
424e3428b1 Make type abbreviations consistent: ssize_t is zd everywhere 2017-03-06 18:49:19 -08:00
David Goldblatt
e9852b5776 Disentangle assert and util
This is the first header refactoring diff, #533.  It splits the assert and util
components into separate, hermetic, header files.  In the process, it splits out
two of the large sub-components of util (the stdio.h replacement, and bit
manipulation routines) into their own components (malloc_io.h and bit_util.h).
This is mostly to break up cyclic dependencies, but it also breaks off a good
chunk of the catch-all-ness of util, which is nice.
2017-03-06 15:08:43 -08:00
David Goldblatt
d4ac7582f3 Introduce a backport of C11 atomics
This introduces a backport of C11 atomics.  It has four implementations; ranked
in order of preference, they are:
- GCC/Clang __atomic builtins
- GCC/Clang __sync builtins
- MSVC _Interlocked builtins
- C11 atomics, from <stdatomic.h>

The primary advantages are:
- Close adherence to the standard API gives us a defined memory model.
- Type safety: atomic objects are now separate types from non-atomic ones, so
  that it's impossible to mix up atomic and non-atomic updates (which is
  undefined behavior that compilers are starting to take advantage of).
- Efficiency: we can specify ordering for operations, avoiding fences and
  atomic operations on strongly ordered architectures (example:
  `atomic_write_u32(ptr, val);` involves a CAS loop, whereas
  `atomic_store(ptr, val, ATOMIC_RELEASE);` is a plain store.

This diff leaves in the current atomics API (implementing them in terms of the
backport).  This lets us transition uses over piecemeal.

Testing:
This is by nature hard to test. I've manually tested the first three options on
Linux on gcc by futzing with the #defines manually, on freebsd with gcc and
clang, on MSVC, and on OS X with clang.  All of these were x86 machines though,
and we don't have any test infrastructure set up for non-x86 platforms.
2017-03-03 13:40:59 -08:00
Jason Evans
6b5cba4191 Convert arena->stats synchronization to atomics. 2017-02-16 09:39:46 -08:00
Jason Evans
fa2d64c94b Convert arena->prof_accumbytes synchronization to atomics. 2017-02-16 09:39:46 -08:00
Jason Evans
d27f29b468 Disentangle arena and extent locking.
Refactor arena and extent locking protocols such that arena and
extent locks are never held when calling into the extent_*_wrapper()
API.  This requires extra care during purging since the arena lock no
longer protects the inner purging logic.  It also requires extra care to
protect extents from being merged with adjacent extents.

Convert extent_t's 'active' flag to an enumerated 'state', so that
retained extents are explicitly marked as such, rather than depending on
ring linkage state.

Refactor the extent collections (and their synchronization) for cached
and retained extents into extents_t.  Incorporate LRU functionality to
support purging.  Incorporate page count accounting, which replaces
arena->ndirty and arena->stats.retained.

Assert that no core locks are held when entering any internal
[de]allocation functions.  This is in addition to existing assertions
that no locks are held when entering external [de]allocation functions.

Audit and document synchronization protocols for all arena_t fields.

This fixes a potential deadlock due to recursive allocation during
gdump, in a similar fashion to b49c649bc1
(Fix lock order reversal during gdump.), but with a necessarily much
broader code impact.
2017-02-01 16:43:46 -08:00
Jason Evans
c0cc5db871 Replace tabs following #define with spaces.
This resolves #564.
2017-01-20 21:45:53 -08:00
Jason Evans
f408643a4c Remove extraneous parens around return arguments.
This resolves #540.
2017-01-20 21:43:07 -08:00
Jason Evans
c4c2592c83 Update brace style.
Add braces around single-line blocks, and remove line breaks before
function-opening braces.

This resolves #537.
2017-01-20 21:43:07 -08:00
Mike Hommey
0f7376eb62 Don't rely on OSX SDK malloc/malloc.h for malloc_zone struct definitions
The SDK jemalloc is built against might be not be the latest for various
reasons, but the resulting binary ought to work on newer versions of
OSX.

In order to ensure this, we need the fullest definitions possible, so
copy what we need from the latest version of malloc/malloc.h available
on opensource.apple.com.
2017-01-17 20:13:28 -08:00
Jason Evans
ffbb7dac3d Remove leading blank lines from function bodies.
This resolves #535.
2017-01-13 14:49:24 -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
David Goldblatt
94c5d22a4d Remove mb.h, which is unused 2017-01-11 13:24:30 -08:00
Jason Evans
edf1bafb2b Implement arena.<i>.destroy .
Add MALLCTL_ARENAS_DESTROYED for accessing destroyed arena stats as an
analogue to MALLCTL_ARENAS_ALL.

This resolves #382.
2017-01-06 18:58:46 -08:00
Jason Evans
d778dd2afc Refactor ctl_stats_t.
Refactor ctl_stats_t to be a demand-zeroed non-growing data structure.
To keep the size from being onerous (~60 MiB) on 32-bit systems, convert
the arenas field to contain pointers rather than directly embedded
ctl_arena_stats_t elements.
2017-01-06 18:58:45 -08:00
Jason Evans
a0dd3a4483 Implement per arena base allocators.
Add/rename related mallctls:
- Add stats.arenas.<i>.base .
- Rename stats.arenas.<i>.metadata to stats.arenas.<i>.internal .
- Add stats.arenas.<i>.resident .

Modify the arenas.extend mallctl to take an optional (extent_hooks_t *)
argument so that it is possible for all base allocations to be serviced
by the specified extent hooks.

This resolves #463.
2016-12-26 18:08:28 -08:00
Jason Evans
c1baa0a9b7 Add huge page configuration and pages_[no}huge().
Add the --with-lg-hugepage configure option, but automatically configure
LG_HUGEPAGE even if it isn't specified.

Add the pages_[no]huge() functions, which toggle huge page state via
madvise(..., MADV_[NO]HUGEPAGE) calls.
2016-12-26 17:59:34 -08:00
Dave Watson
2319152d9f jemalloc cpp new/delete bindings
Adds cpp bindings for jemalloc, along with necessary autoconf settings.
This is mostly to add sized deallocation support, which can't be added
from C directly.  Sized deallocation is ~10% microbench improvement.

* Import ax_cxx_compile_stdcxx.m4 from the autoconf repo, seems like the
  easiest way to get c++14 detection.
* Adds various other changes, like CXXFLAGS, to configure.ac.
* Adds new rules to Makefile.in for src/jemalloc-cpp.cpp, and a basic
  unittest.
* Both new and delete are overridden, to ensure jemalloc is used for
  both.
* TODO future enhancement of avoiding extra PLT thunks for new and
  delete - sdallocx and malloc are publicly exported jemalloc symbols,
  using an alias would link them directly.  Unfortunately, was having
  trouble getting it to play nice with jemalloc's namespace support.

Testing:
Tested gcc 4.8, gcc 5, gcc 5.2, clang 4.0.  Only gcc >= 5 has sized
deallocation support, verified that the rest build correctly.

Tested mac osx and Centos.

Tested --with-jemalloc-prefix and --without-export.

This resolves #202.
2016-12-12 18:36:06 -08:00
Jason Evans
ea9961acdb Fix psz/pind edge cases.
Add an "over-size" extent heap in which to store extents which exceed
the maximum size class (plus cache-oblivious padding, if enabled).
Remove psz2ind_clamp() and use psz2ind() instead so that trying to
allocate the maximum size class can in principle succeed.  In practice,
this allows assertions to hold so that OOM errors can be successfully
generated.
2016-11-03 22:33:34 -07:00
Jason Evans
48d4adfbeb Avoid negation of unsigned numbers.
Rather than relying on two's complement negation for alignment mask
generation, use bitwise not and addition.  This dodges warnings from
MSVC, and should be strength-reduced by compiler optimization anyway.
2016-10-27 21:26:33 -07:00
Jason Evans
b54d160dc4 Do not (recursively) allocate within tsd_fetch().
Refactor tsd so that tsdn_fetch() does not trigger allocation, since
allocation could cause infinite recursion.

This resolves #458.
2016-10-20 23:59:12 -07:00
Jason Evans
e5effef428 Add/use adaptive spinning.
Add spin_t and spin_{init,adaptive}(), which provide a simple
abstraction for adaptive spinning.

Adaptively spin during busy waits in bootstrapping and rtree node
initialization.
2016-10-13 14:55:39 -07:00
Jason Evans
9acd5cf178 Remove all vestiges of chunks.
Remove mallctls:
- opt.lg_chunk
- stats.cactive

This resolves #464.
2016-10-12 11:55:43 -07:00
Jason Evans
871a9498e1 Fix size class overflow bugs.
Avoid calling s2u() on raw extent sizes in extent_recycle().

Clamp psz2ind() (implemented as psz2ind_clamp()) when inserting/removing
into/from size-segregated extent heaps.
2016-10-03 14:18:55 -07:00
Eric Le Bihan
df0d273a07 Fix LG_QUANTUM definition for sparc64
GCC 4.9.3 cross-compiled for sparc64 defines __sparc_v9__, not
__sparc64__ nor __sparcv9. This prevents LG_QUANTUM from being defined
properly. Adding this new value to the check solves the issue.
2016-09-26 15:13:07 -07:00
Elliot Ronaghan
8a1a794b0c Don't use compact red-black trees with the pgi compiler
Some bug (either in the red-black tree code, or in the pgi compiler) seems to
cause red-black trees to become unbalanced. This issue seems to go away if we
don't use compact red-black trees. Since red-black trees don't seem to be used
much anymore, I opted for what seems to be an easy fix here instead of digging
in and trying to find the root cause of the bug.

Some context in case it's helpful:

I experienced a ton of segfaults while using pgi as Chapel's target compiler
with jemalloc 4.0.4. The little bit of debugging I did pointed me somewhere
deep in red-black tree manipulation, but I didn't get a chance to investigate
further. It looks like 4.2.0 replaced most uses of red-black trees with
pairing-heaps, which seems to avoid whatever bug I was hitting.

However, `make check_unit` was still failing on the rb test, so I figured the
core issue was just being masked. Here's the `make check_unit` failure:

```sh
=== test/unit/rb ===
test_rb_empty: pass
tree_recurse:test/unit/rb.c:90: Failed assertion: (((_Bool) (((uintptr_t) (left_node)->link.rbn_right_red) & ((size_t)1)))) == (false) --> true != false: Node should be black
test_rb_random:test/unit/rb.c:274: Failed assertion: (imbalances) == (0) --> 1 != 0: Tree is unbalanced
tree_recurse:test/unit/rb.c:90: Failed assertion: (((_Bool) (((uintptr_t) (left_node)->link.rbn_right_red) & ((size_t)1)))) == (false) --> true != false: Node should be black
test_rb_random:test/unit/rb.c:274: Failed assertion: (imbalances) == (0) --> 1 != 0: Tree is unbalanced
node_remove:test/unit/rb.c:190: Failed assertion: (imbalances) == (0) --> 2 != 0: Tree is unbalanced
<jemalloc>: test/unit/rb.c:43: Failed assertion: "pathp[-1].cmp < 0"
test/test.sh: line 22: 12926 Aborted
Test harness error
```

While starting to debug I saw the RB_COMPACT option and decided to check if
turning that off resolved the bug. It seems to have fixed it (`make check_unit`
passes and the segfaults under Chapel are gone) so it seems like on okay
work-around. I'd imagine this has performance implications for red-black trees
under pgi, but if they're not going to be used much anymore it's probably not a
big deal.
2016-06-08 14:48:55 -07:00
Jason Evans
dd752c1ffd Fix potential VM map fragmentation regression.
Revert 245ae6036c (Support --with-lg-page
values larger than actual page size.), because it could cause VM map
fragmentation if the kernel grows mmap()ed memory downward.

This resolves #391.
2016-06-07 14:15:49 -07:00
Jason Evans
7be2ebc23f Make tsd cleanup functions optional, remove noop cleanup functions. 2016-06-05 20:42:24 -07:00
Jason Evans
751f2c332d Remove obsolete stats.arenas.<i>.metadata.mapped mallctl.
Rename stats.arenas.<i>.metadata.allocated mallctl to
stats.arenas.<i>.metadata .
2016-06-05 20:42:24 -07:00
Jason Evans
03eea4fb8b Better document --enable-ivsalloc. 2016-06-05 20:42:24 -07:00
Jason Evans
0c4932eb1e s/chunk_lookup/extent_lookup/g, s/chunks_rtree/extents_rtree/g 2016-06-05 20:42:23 -07:00
Jason Evans
7d63fed0fd Rename huge to large. 2016-06-05 20:42:23 -07:00
Jason Evans
ed2c2427a7 Use huge size class infrastructure for large size classes. 2016-06-05 20:42:18 -07:00
Jason Evans
4731cd47f7 Allow chunks to not be naturally aligned.
Precisely size extents for huge size classes that aren't multiples of
chunksize.
2016-06-03 12:27:41 -07:00
Jason Evans
d78846c989 Replace extent_achunk_[gs]et() with extent_slab_[gs]et(). 2016-06-03 12:27:41 -07:00
Jason Evans
fae8344098 Add extent_active_[gs]et().
Always initialize extents' runs_dirty and chunks_cache linkage.
2016-06-03 12:27:41 -07:00
Jason Evans
6f71844659 Move *PAGE* definitions to pages.h. 2016-06-03 12:27:41 -07:00
Jason Evans
8c9be3e837 Refactor rtree to always use base_alloc() for node allocation. 2016-06-03 12:27:41 -07:00
Jason Evans
db72272bef Use rtree-based chunk lookups rather than pointer bit twiddling.
Look up chunk metadata via the radix tree, rather than using
CHUNK_ADDR2BASE().

Propagate pointer's containing extent.

Minimize extent lookups by doing a single lookup (e.g. in free()) and
propagating the pointer's extent into nearly all the functions that may
need it.
2016-06-03 12:27:41 -07:00
Jason Evans
a7a6f5bc96 Rename extent_node_t to extent_t. 2016-05-16 12:21:28 -07:00
Jason Evans
7bb00ae9d6 Refactor runs_avail.
Use pszind_t size classes rather than szind_t size classes, and always
reserve space for NPSIZES elements.  This removes unused heaps that are
not multiples of the page size, and adds (currently) unused heaps for
all huge size classes, with the immediate benefit that the size of
arena_t allocations is constant (no longer dependent on chunk size).
2016-05-16 12:21:21 -07:00
Jason Evans
226c446979 Implement pz2ind(), pind2sz(), and psz2u().
These compute size classes and indices similarly to size2index(),
index2size() and s2u(), respectively, but using the subset of size
classes that are multiples of the page size.  Note that pszind_t and
szind_t are not interchangeable.
2016-05-13 10:31:54 -07:00
Jason Evans
17c021c177 Remove redzone support.
This resolves #369.
2016-05-13 10:27:33 -07:00
Jason Evans
ba5c709517 Remove quarantine support. 2016-05-13 10:25:05 -07:00
Jason Evans
9a8add1510 Remove Valgrind support. 2016-05-13 09:56:18 -07:00
Jason Evans
73d3d58dc2 Optimize witness fast path.
Short-circuit commonly called witness functions so that they only
execute in debug builds, and remove equivalent guards from mutex
functions.  This avoids pointless code execution in
witness_assert_lockless(), which is typically called twice per
allocation/deallocation function invocation.

Inline commonly called witness functions so that optimized builds can
completely remove calls as dead code.
2016-05-11 15:38:06 -07:00
Jason Evans
c1e00ef2a6 Resolve bootstrapping issues when embedded in FreeBSD libc.
b2c0d6322d (Add witness, a simple online
locking validator.) caused a broad propagation of tsd throughout the
internal API, but tsd_fetch() was designed to fail prior to tsd
bootstrapping.  Fix this by splitting tsd_t into non-nullable tsd_t and
nullable tsdn_t, and modifying all internal APIs that do not critically
rely on tsd to take nullable pointers.  Furthermore, add the
tsd_booted_get() function so that tsdn_fetch() can probe whether tsd
bootstrapping is complete and return NULL if not.  All dangerous
conversions of nullable pointers are tsdn_tsd() calls that assert-fail
on invalid conversion.
2016-05-10 22:51:33 -07:00
Jason Evans
919e4a0ea9 Add LG_QUANTUM definition for the RISC-V architecture. 2016-05-06 17:15:32 -07:00
Jason Evans
3ef51d7f73 Optimize the fast paths of calloc() and [m,d,sd]allocx().
This is a broader application of optimizations to malloc() and free() in
f4a0f32d34 (Fast-path improvement:
reduce # of branches and unnecessary operations.).

This resolves #321.
2016-05-06 14:37:39 -07:00