kata-containers/regorus
1
0
Fork 0
mirror of https://github.com/microsoft/regorus.git synced 2026-08-05 02:16:11 +00:00
Code Issues Packages Projects Releases Wiki Activity

feat(memory): Allocator-backed global memory limits (#544)

Browse Source 
Policy evaluation at scale needs to be able to set memory limits
so that a bad policy does not hog memory or to ensure that
policy evaluation itself does not use too much memory which could
cause other components to suffer.

This PR introduces capability to set and enforce global memory limits.
It also lays the groundwork for enabling per evaluation limits in future.

Once a global memory limit is set, Regorus maintains per thread counters
to track memory activity (allocation, deallocation) of a thread.
These counters are periodically flushed to global memory counters.
Per thread counters avoid the contention that updating global counters
on each alloc/free would cause.

Policy evaluation periodically checks these counters and raises errors
if allocated memory has exceeded the configured limit.

Currently memory limit capability is exposed only to FFI and C#.

Also update mimalloc to v2.2.6

Signed-off-by: Anand Krishnamoorthi <anakrish@microsoft.com>
This commit is contained in:
Anand Krishnamoorthi
2026-01-24 07:08:54 +05:30
committed by GitHub
parent 80686d6ed1
commit fd59bb5a91
93 changed files with 9772 additions and 3779 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
mimalloc/mimalloc-sys/mimalloc/LICENSE
View File

@@ -1,6 +1,6 @@
MIT License
Copyright (c) 2018-2021 Microsoft Corporation, Daan Leijen
Copyright (c) 2018-2025 Microsoft Corporation, Daan Leijen
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal

950
mimalloc/mimalloc-sys/mimalloc/README.md
View File

@@ -1,3 +1,949 @@
# Mimalloc
The `src`/`include` folders contain the [mimalloc](https://github.com/microsoft/mimalloc) source from [`v2.1.2`](https://github.com/microsoft/mimalloc/tree/v2.1.2)(`43ce4bd7fd34bcc730c1c7471c99995597415488`).
<img align="left" width="100" height="100" src="doc/mimalloc-logo.png"/>
[<img align="right" src="https://dev.azure.com/Daan0324/mimalloc/_apis/build/status/microsoft.mimalloc?branchName=dev3"/>](https://dev.azure.com/Daan0324/mimalloc/_build?definitionId=1&_a=summary)
# mimalloc
&nbsp;
mimalloc (pronounced "me-malloc")
is a general purpose allocator with excellent [performance](#performance) characteristics.
Initially developed by Daan Leijen for the runtime systems of the
[Koka](https://koka-lang.github.io) and [Lean](https://github.com/leanprover/lean) languages.
Latest release : `v3.2.6` (2026-01-08) release candidate 1, please report any issues.
Latest v2 release: `v2.2.6` (2026-01-08).
Latest v1 release: `v1.9.6` (2026-01-08).
mimalloc is a drop-in replacement for `malloc` and can be used in other programs
without code changes, for example, on dynamically linked ELF-based systems (Linux, BSD, etc.) you can use it as:
```
> LD_PRELOAD=/usr/lib/libmimalloc.so myprogram
```
It also includes a way to dynamically override the default allocator in [Windows](#override_on_windows).
Notable aspects of the design include:
- __small and consistent__: the library is about 10k LOC using simple and
consistent data structures. This makes it very suitable
to integrate and adapt in other projects. For runtime systems it
provides hooks for a monotonic _heartbeat_ and deferred freeing (for
bounded worst-case times with reference counting).
Partly due to its simplicity, mimalloc has been ported to many systems (Windows, macOS,
Linux, WASM, various BSD's, Haiku, MUSL, etc) and has excellent support for dynamic overriding.
At the same time, it is an industrial strength allocator that runs (very) large scale
distributed services on thousands of machines with excellent worst case latencies.
- __free list sharding__: instead of one big free list (per size class) we have
many smaller lists per "mimalloc page" which reduces fragmentation and
increases locality --
things that are allocated close in time get allocated close in memory.
(A mimalloc page contains blocks of one size class and is usually 64KiB on a 64-bit system).
- __free list multi-sharding__: the big idea! Not only do we shard the free list
per mimalloc page, but for each page we have multiple free lists. In particular, there
is one list for thread-local `free` operations, and another one for concurrent `free`
operations. Free-ing from another thread can now be a single CAS without needing
sophisticated coordination between threads. Since there will be
thousands of separate free lists, contention is naturally distributed over the heap,
and the chance of contending on a single location will be low -- this is quite
similar to randomized algorithms like skip lists where adding
a random oracle removes the need for a more complex algorithm.
- __eager page purging__: when a "page" becomes empty (with increased chance
due to free list sharding) the memory is marked to the OS as unused (reset or decommitted)
reducing (real) memory pressure and fragmentation, especially in long running
programs.
- __secure__: _mimalloc_ can be built in secure mode, adding guard pages,
randomized allocation, encrypted free lists, etc. to protect against various
heap vulnerabilities. The performance penalty is usually around 10% on average
over our benchmarks.
- __first-class heaps__: efficiently create and use multiple heaps to allocate across different regions.
A heap can be destroyed at once instead of deallocating each object separately.
New: v3 has true first-class heaps where one can allocate in a heap from any thread.
- __bounded__: it does not suffer from _blowup_ \[1\], has bounded worst-case allocation
times (_wcat_) (upto OS primitives), bounded space overhead (~0.2% meta-data, with low
internal fragmentation), and has no internal points of contention using only atomic operations.
- __fast__: In our benchmarks (see [below](#performance)),
_mimalloc_ outperforms other leading allocators (_jemalloc_, _tcmalloc_, _Hoard_, etc),
and often uses less memory. A nice property is that it does consistently well over a wide range
of benchmarks. There is also good huge OS page support for larger server programs.
The [documentation](https://microsoft.github.io/mimalloc) gives a full overview of the API.
You can read more on the design of _mimalloc_ in the [technical report](https://www.microsoft.com/en-us/research/publication/mimalloc-free-list-sharding-in-action) which also has detailed benchmark results.
Enjoy!
### Branches
* `main`: latest stable release (still based on `dev2`).
* `dev`: development branch for mimalloc v1. Use this branch for submitting PR's.
* `dev2`: development branch for mimalloc v2. This branch is downstream of `dev`
(and is essentially equal to `dev` except for `src/segment.c`). Uses larger sliced segments to manage
mimalloc pages that can reduce fragmentation.
* `dev3`: development branch for mimalloc v3 rc1. This branch is downstream of `dev`. This version
simplifies the lock-free ownership of previous versions, and improves sharing of memory between
threads. On certain large workloads this version may use (much) less memory.
Also support true first-class heaps and more efficient heap-walking.
### Releases
* 2026-01-08, `v1.9.6`, `v2.2.6`, `v3.2.6` (rc1) : Important bug fixes. Many improvements to v3 including
true first-class heaps where one can allocate in heap from any thread, and track statistics per heap as well.
Added `MIMALLOC_ALLOW_THP` option. This is by default enabled except on Android. When THP is detected on v3,
mimalloc will set the `MIMALLOC_MINIMAL_PURGE_SIZE` to 2MiB to avoid breaking up potential THP huge pages.
v3 uses faster TLS access on Windows, and has improved performance for `mi_calloc` and aligned allocations.
Fixed rare race condition on older v3, fixed potential buffer overflow in debug statistics, add API for returning
allocated sizes on allocation and free.
* 2025-06-09, `v1.9.4`, `v2.2.4`, `v3.1.4` (beta) : Some important bug fixes, including a case where OS memory
was not always fully released. Improved v3 performance, build on XBox, fix build on Android, support interpose
for older macOS versions, use MADV_FREE_REUSABLE on macOS, always check commit success, better support for Windows
fixed TLS offset, etc.
* 2025-03-28, `v1.9.3`, `v2.2.3`, `v3.0.3` (beta) : Various small bug and build fixes, including:
fix arm32 pre v7 builds, fix mingw build, get runtime statistics, improve statistic commit counts,
fix execution on non BMI1 x64 systems.
* 2025-03-06, `v1.9.2`, `v2.2.2`, `v3.0.2-beta`: Various small bug and build fixes.
Add `mi_options_print`, `mi_arenas_print`, and the experimental `mi_stat_get` and `mi_stat_get_json`.
Add `mi_thread_set_in_threadpool` and `mi_heap_set_numa_affinity` (v3 only). Add vcpkg portfile.
Upgrade mimalloc-redirect to v1.3.2. `MI_OPT_ARCH` is off by default now but still assumes armv8.1-a on arm64
for fast atomic operations. Add QNX support.
* 2025-01-03, `v1.8.9`, `v2.1.9`, `v3.0.1-alpha`: Interim release. Support Windows arm64. New [guarded](#guarded) build that can place OS
guard pages behind objects to catch buffer overflows as they occur.
Many small fixes: build on Windows arm64, cygwin, riscV, and dragonfly; fix Windows static library initialization to account for
thread local destructors (in Rust/C++); macOS tag change; macOS TLS slot fix; improve stats;
consistent `mimalloc.dll` on Windows (instead of `mimalloc-override.dll`); fix mimalloc-redirect on Win11 H2;
add 0-byte to canary; upstream CPython fixes; reduce .bss size; allow fixed TLS slot on Windows for improved performance.
* 2024-05-21, `v1.8.7`, `v2.1.7`: Fix build issues on less common platforms. Started upstreaming patches
from the CPython [integration](https://github.com/python/cpython/issues/113141#issuecomment-2119255217). Upstream `vcpkg` patches.
* 2024-05-13, `v1.8.6`, `v2.1.6`: Fix build errors on various (older) platforms. Refactored aligned allocation.
* 2024-04-22, `v1.8.4`, `v2.1.4`: Fixes various bugs and build issues. Add `MI_LIBC_MUSL` cmake flag for musl builds.
Free-ing code is refactored into a separate module (`free.c`). Mimalloc page info is simplified with the block size
directly available (and new `block_size_shift` to improve aligned block free-ing).
New approach to collection of abandoned segments: When
a thread terminates the segments it owns are abandoned (containing still live objects) and these can be
reclaimed by other threads. We no longer use a list of abandoned segments but this is now done using bitmaps in arena's
which is more concurrent (and more aggressive). Abandoned memory can now also be reclaimed if a thread frees an object in
an abandoned page (which can be disabled using `mi_option_abandoned_reclaim_on_free`). The option `mi_option_max_segment_reclaim`
gives a maximum percentage of abandoned segments that can be reclaimed per try (=10%).
* 2023-04-24, `v1.8.2`, `v2.1.2`: Fixes build issues on freeBSD, musl, and C17 (UE 5.1.1). Reduce code size/complexity
by removing regions and segment-cache's and only use arenas with improved memory purging -- this may improve memory
usage as well for larger services. Renamed options for consistency. Improved Valgrind and ASAN checking.
* 2023-04-03, `v1.8.1`, `v2.1.1`: Fixes build issues on some platforms.
* 2023-03-29, `v1.8.0`, `v2.1.0`: Improved support dynamic overriding on Windows 11. Improved tracing precision
with [asan](#asan) and [Valgrind](#valgrind), and added Windows event tracing [ETW](#ETW) (contributed by Xinglong He). Created an OS
abstraction layer to make it easier to port and separate platform dependent code (in `src/prim`). Fixed C++ STL compilation on older Microsoft C++ compilers, and various small bug fixes.
* 2022-12-23, `v1.7.9`, `v2.0.9`: Supports building with [asan](#asan) and improved [Valgrind](#valgrind) support.
Support arbitrary large alignments (in particular for `std::pmr` pools).
Added C++ STL allocators attached to a specific heap (thanks @vmarkovtsev).
Heap walks now visit all object (including huge objects). Support Windows nano server containers (by Johannes Schindelin,@dscho).
Various small bug fixes.
* 2022-11-03, `v1.7.7`, `v2.0.7`: Initial support for [Valgrind](#valgrind) for leak testing and heap block overflow
detection. Initial
support for attaching heaps to a specific memory area (only in v2). Fix `realloc` behavior for zero size blocks, remove restriction to integral multiple of the alignment in `alloc_align`, improved aligned allocation performance, reduced contention with many threads on few processors (thank you @dposluns!), vs2022 support, support `pkg-config`, .
* 2022-04-14, `v1.7.6`, `v2.0.6`: fix fallback path for aligned OS allocation on Windows, improve Windows aligned allocation
even when compiling with older SDK's, fix dynamic overriding on macOS Monterey, fix MSVC C++ dynamic overriding, fix
warnings under Clang 14, improve performance if many OS threads are created and destroyed, fix statistics for large object
allocations, using MIMALLOC_VERBOSE=1 has no maximum on the number of error messages, various small fixes.
* 2022-02-14, `v1.7.5`, `v2.0.5` (alpha): fix malloc override on
Windows 11, fix compilation with musl, potentially reduced
committed memory, add `bin/minject` for Windows,
improved wasm support, faster aligned allocation,
various small fixes.
* [Older release notes](#older-release-notes)
Special thanks to:
* [David Carlier](https://devnexen.blogspot.com/) (@devnexen) for his many contributions, and making
mimalloc work better on many less common operating systems, like Haiku, Dragonfly, etc.
* Mary Feofanova (@mary3000), Evgeniy Moiseenko, and Manuel Pöter (@mpoeter) for making mimalloc TSAN checkable, and finding
memory model bugs using the [genMC] model checker.
* Weipeng Liu (@pongba), Zhuowei Li, Junhua Wang, and Jakub Szymanski, for their early support of mimalloc and deployment
at large scale services, leading to many improvements in the mimalloc algorithms for large workloads.
* Jason Gibson (@jasongibson) for exhaustive testing on large scale workloads and server environments, and finding complex bugs
in (early versions of) `mimalloc`.
* Manuel Pöter (@mpoeter) and Sam Gross(@colesbury) for finding an ABA concurrency issue in abandoned segment reclamation. Sam also created the [no GIL](https://github.com/colesbury/nogil) Python fork which
uses mimalloc internally.
[genMC]: https://plv.mpi-sws.org/genmc/
### Usage
mimalloc is used in various large scale low-latency services and programs, for example:
<a href="https://www.bing.com"><img height="50" align="left" src="https://upload.wikimedia.org/wikipedia/commons/e/e9/Bing_logo.svg"></a>
<a href="https://azure.microsoft.com/"><img height="50" align="left" src="https://upload.wikimedia.org/wikipedia/commons/a/a8/Microsoft_Azure_Logo.svg"></a>
<a href="https://deathstrandingpc.505games.com"><img height="100" src="doc/ds-logo.png"></a>
<a href="https://docs.unrealengine.com/4.26/en-US/WhatsNew/Builds/ReleaseNotes/4_25/"><img height="100" src="doc/unreal-logo.svg"></a>
<a href="https://cab.spbu.ru/software/spades/"><img height="100" src="doc/spades-logo.png"></a>
# Building
## Windows
Open `ide/vs2022/mimalloc.sln` in Visual Studio 2022 and build.
The `mimalloc-lib` project builds a static library (in `out/msvc-x64`), while the
`mimalloc-override-dll` project builds a DLL for overriding malloc
in the entire program.
## Linux, macOS, BSD, etc.
We use [`cmake`](https://cmake.org) as the build system:
```
> mkdir -p out/release
> cd out/release
> cmake ../..
> make
```
This builds the library as a shared (dynamic)
library (`.so` or `.dylib`), a static library (`.a`), and
as a single object file (`.o`).
`> sudo make install` (install the library and header files in `/usr/local/lib` and `/usr/local/include`)
You can build the debug version which does many internal checks and
maintains detailed statistics as:
```
> mkdir -p out/debug
> cd out/debug
> cmake -DCMAKE_BUILD_TYPE=Debug ../..
> make
```
This will name the shared library as `libmimalloc-debug.so`.
Finally, you can build a _secure_ version that uses guard pages, encrypted free lists, etc., as:
```
> mkdir -p out/secure
> cd out/secure
> cmake -DMI_SECURE=ON ../..
> make
```
This will name the shared library as `libmimalloc-secure.so`.
Use `cmake ../.. -LH` to see all the available build options.
The examples use the default compiler. If you like to use another, use:
```
> CC=clang CXX=clang++ cmake ../..
```
## Cmake with Visual Studio
You can also use cmake on Windows. Open a Visual Studio 2022 development prompt
and invoke `cmake` with the right [generator](https://cmake.org/cmake/help/latest/generator/Visual%20Studio%2017%202022.html)
and architecture, like:
```
> cmake ..\.. -G "Visual Studio 17 2022" -A x64 -DMI_OVERRIDE=ON
```
The cmake build type is specified when actually building, for example:
```
> cmake --build . --config=Release
```
You can also install the [LLVM toolset](https://learn.microsoft.com/en-us/cpp/build/clang-support-msbuild?view=msvc-170#install-1)
on Windows to build with the `clang-cl` compiler directly:
```
> cmake ../.. -G "Visual Studio 17 2022" -T ClangCl
```
## Single Source
You can also directly build the single `src/static.c` file as part of your project without
needing `cmake` at all. Make sure to also add the mimalloc `include` directory to the include path.
# Using the Library
The preferred usage is including `<mimalloc.h>`, linking with
the shared- or static library, and using the `mi_malloc` API exclusively for allocation. For example,
```
> gcc -o myprogram -lmimalloc myfile.c
```
mimalloc uses only safe OS calls (`mmap` and `VirtualAlloc`) and can co-exist
with other allocators linked to the same program.
If you use `cmake`, you can simply use:
```
find_package(mimalloc 1.8 REQUIRED)
```
in your `CMakeLists.txt` to find a locally installed mimalloc. Then use either:
```
target_link_libraries(myapp PUBLIC mimalloc)
```
to link with the shared (dynamic) library, or:
```
target_link_libraries(myapp PUBLIC mimalloc-static)
```
to link with the static library. See `test\CMakeLists.txt` for an example.
For best performance in C++ programs, it is also recommended to override the
global `new` and `delete` operators. For convenience, mimalloc provides
[`mimalloc-new-delete.h`](include/mimalloc-new-delete.h) which does this for you -- just include it in a single(!) source file in your project.
In C++, mimalloc also provides the `mi_stl_allocator` struct which implements the `std::allocator`
interface.
You can pass environment variables to print verbose messages (`MIMALLOC_VERBOSE=1`)
and statistics (`MIMALLOC_SHOW_STATS=1`) (in the debug version):
```
> env MIMALLOC_SHOW_STATS=1 ./cfrac 175451865205073170563711388363
175451865205073170563711388363 = 374456281610909315237213 * 468551
heap stats: peak total freed unit
normal 2: 16.4 kb 17.5 mb 17.5 mb 16 b ok
normal 3: 16.3 kb 15.2 mb 15.2 mb 24 b ok
normal 4: 64 b 4.6 kb 4.6 kb 32 b ok
normal 5: 80 b 118.4 kb 118.4 kb 40 b ok
normal 6: 48 b 48 b 48 b 48 b ok
normal 17: 960 b 960 b 960 b 320 b ok
heap stats: peak total freed unit
normal: 33.9 kb 32.8 mb 32.8 mb 1 b ok
huge: 0 b 0 b 0 b 1 b ok
total: 33.9 kb 32.8 mb 32.8 mb 1 b ok
malloc requested: 32.8 mb
committed: 58.2 kb 58.2 kb 58.2 kb 1 b ok
reserved: 2.0 mb 2.0 mb 2.0 mb 1 b ok
reset: 0 b 0 b 0 b 1 b ok
segments: 1 1 1
-abandoned: 0
pages: 6 6 6
-abandoned: 0
mmaps: 3
mmap fast: 0
mmap slow: 1
threads: 0
elapsed: 2.022s
process: user: 1.781s, system: 0.016s, faults: 756, reclaims: 0, rss: 2.7 mb
```
The above model of using the `mi_` prefixed API is not always possible
though in existing programs that already use the standard malloc interface,
and another option is to override the standard malloc interface
completely and redirect all calls to the _mimalloc_ library instead .
## Environment Options
You can set further options either programmatically (using [`mi_option_set`](https://microsoft.github.io/mimalloc/group__options.html)), or via environment variables:
- `MIMALLOC_SHOW_STATS=1`: show statistics when the program terminates.
- `MIMALLOC_VERBOSE=1`: show verbose messages (including statistics).
- `MIMALLOC_SHOW_ERRORS=1`: show error and warning messages.
Advanced options:
- `MIMALLOC_ARENA_EAGER_COMMIT=2`: turns on eager commit for the large arenas (usually 1GiB) from which mimalloc
allocates segments and pages. Set this to 2 (default) to
only enable this on overcommit systems (e.g. Linux). Set this to 1 to enable explicitly on other systems
as well (like Windows or macOS) which may improve performance (as the whole arena is committed at once).
Note that eager commit only increases the commit but not the actual the peak resident set
(rss) so it is generally ok to enable this.
- `MIMALLOC_PURGE_DELAY=N`: the delay in `N` milli-seconds (by default `1000` in v3) after which mimalloc will purge
OS pages that are not in use. This signals to the OS that the underlying physical memory can be reused which
can reduce memory fragmentation especially in long running (server) programs. Setting `N` to `0` purges immediately when
a page becomes unused which can improve memory usage but also decreases performance.
Setting it to `-1` disables purging completely.
- `MIMALLOC_PURGE_DECOMMITS=1`: By default "purging" memory means unused memory is decommitted (`MEM_DECOMMIT` on Windows,
`MADV_DONTNEED` (which decresease rss immediately) on `mmap` systems). Set this to 0 to instead "reset" unused
memory on a purge (`MEM_RESET` on Windows, generally `MADV_FREE` (which does not decrease rss immediately) on `mmap` systems).
Mimalloc generally does not "free" OS memory but only "purges" OS memory, in other words, it tries to keep virtual
address ranges and decommits within those ranges (to make the underlying physical memory available to other processes).
Further options for large workloads and services:
- `MIMALLOC_ALLOW_THP=1`: By default always allow transparent huge pages (THP) on Linux systems. On Android only this is
by default off. When set to `0`, THP is disabled for the process that mimalloc runs in. If enabled, mimalloc also sets
the `MIMALLOC_MINIMAL_PURGE_SIZE` in v3 to 2MiB to avoid potentially breaking up transparent huge pages.
- `MIMALLOC_USE_NUMA_NODES=N`: pretend there are at most `N` NUMA nodes. If not set, the actual NUMA nodes are detected
at runtime. Setting `N` to 1 may avoid problems in some virtual environments. Also, setting it to a lower number than
the actual NUMA nodes is fine and will only cause threads to potentially allocate more memory across actual NUMA
nodes (but this can happen in any case as NUMA local allocation is always a best effort but not guaranteed).
- `MIMALLOC_ALLOW_LARGE_OS_PAGES=0`: Set to 1 to use large OS pages (2 or 4MiB) when available; for some workloads this can
significantly improve performance. However, large OS pages cannot be purged or shared with other processes so may lead
to increased memory usage in some cases.
Use `MIMALLOC_VERBOSE` to check if the large OS pages are enabled -- usually one needs
to explicitly give permissions for large OS pages (as on [Windows][windows-huge] and [Linux][linux-huge]). However, sometimes
the OS is very slow to reserve contiguous physical memory for large OS pages so use with care on systems that
can have fragmented memory (for that reason, we generally recommend to use `MIMALLOC_RESERVE_HUGE_OS_PAGES` instead whenever possible).
- `MIMALLOC_RESERVE_HUGE_OS_PAGES=N`: where `N` is the number of 1GiB _huge_ OS pages. This reserves the huge pages at
startup and sometimes this can give a large (latency) performance improvement on big workloads.
Usually it is better to not use `MIMALLOC_ALLOW_LARGE_OS_PAGES=1` in combination with this setting. Just like large
OS pages, use with care as reserving
contiguous physical memory can take a long time when memory is fragmented (but reserving the huge pages is done at
startup only once).
Note that we usually need to explicitly give permission for huge OS pages (as on [Windows][windows-huge] and [Linux][linux-huge])).
The huge pages are usually allocated evenly among NUMA nodes.
We can use `MIMALLOC_RESERVE_HUGE_OS_PAGES_AT=N` where `N` is the numa node (starting at 0) to allocate all
the huge pages at a specific numa node instead.
Use caution when using `fork` in combination with either large or huge OS pages: on a fork, the OS uses copy-on-write
for all pages in the original process including the huge OS pages. When any memory is now written in that area, the
OS will copy the entire 1GiB huge page (or 2MiB large page) which can cause the memory usage to grow in large increments.
[linux-huge]: https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/5/html/tuning_and_optimizing_red_hat_enterprise_linux_for_oracle_9i_and_10g_databases/sect-oracle_9i_and_10g_tuning_guide-large_memory_optimization_big_pages_and_huge_pages-configuring_huge_pages_in_red_hat_enterprise_linux_4_or_5
[windows-huge]: https://docs.microsoft.com/en-us/sql/database-engine/configure-windows/enable-the-lock-pages-in-memory-option-windows?view=sql-server-2017
## Secure Mode
_mimalloc_ can be build in secure mode by using the `-DMI_SECURE=ON` flags in `cmake`. This build enables various mitigations
to make mimalloc more robust against exploits. In particular:
- All internal mimalloc pages are surrounded by guard pages and the heap metadata is behind a guard page as well (so a buffer overflow
exploit cannot reach into the metadata).
- All free list pointers are
[encoded](https://github.com/microsoft/mimalloc/blob/783e3377f79ee82af43a0793910a9f2d01ac7863/include/mimalloc-internal.h#L396)
with per-page keys which is used both to prevent overwrites with a known pointer, as well as to detect heap corruption.
- Double free's are detected (and ignored).
- The free lists are initialized in a random order and allocation randomly chooses between extension and reuse within a page to
mitigate against attacks that rely on a predicable allocation order. Similarly, the larger heap blocks allocated by mimalloc
from the OS are also address randomized.
As always, evaluate with care as part of an overall security strategy as all of the above are mitigations but not guarantees.
## Debug Mode
When _mimalloc_ is built using debug mode, (`-DCMAKE_BUILD_TYPE=Debug`),
various checks are done at runtime to catch development errors.
- Statistics are maintained in detail for each object size. They can be shown using `MIMALLOC_SHOW_STATS=1` at runtime.
- All objects have padding at the end to detect (byte precise) heap block overflows.
- Double free's, and freeing invalid heap pointers are detected.
- Corrupted free-lists and some forms of use-after-free are detected.
## Guarded Mode
<span id="guarded">_mimalloc_ can be build in guarded mode using the `-DMI_GUARDED=ON` flags in `cmake`.</span>
This enables placing OS guard pages behind certain object allocations to catch buffer overflows as they occur.
This can be invaluable to catch buffer-overflow bugs in large programs. However, it also means that any object
allocated with a guard page takes at least 8 KiB memory for the guard page and its alignment. As such, allocating
a guard page for every allocation may be too expensive both in terms of memory, and in terms of performance with
many system calls. Therefore, there are various environment variables (and options) to tune this:
- `MIMALLOC_GUARDED_SAMPLE_RATE=N`: Set the sample rate to `N` (by default 4000). This mode places a guard page
behind every `N` suitable object allocations (per thread). Since the performance in guarded mode without placing
guard pages is close to release mode, this can be used to enable guard pages even in production to catch latent
buffer overflow bugs. Set the sample rate to `1` to guard every object, and to `0` to place no guard pages at all.
- `MIMALLOC_GUARDED_SAMPLE_SEED=N`: Start sampling at `N` (by default random). Can be used to reproduce a buffer
overflow if needed.
- `MIMALLOC_GUARDED_MIN=N`, `MIMALLOC_GUARDED_MAX=N`: Minimal and maximal _rounded_ object sizes for which a guard
page is considered (`0` and `1GiB` respectively). If you suspect a buffer overflow occurs with an object of size
141, set the minimum and maximum to `148` and the sample rate to `1` to have all of those guarded.
- `MIMALLOC_GUARDED_PRECISE=1`: If we have an object of size 13, we would usually place it an aligned 16 bytes in
front of the guard page. Using `MIMALLOC_GUARDED_PRECISE` places it exactly 13 bytes before a page so that even
a 1 byte overflow is detected. This violates the C/C++ minimal alignment guarantees though so use with care.
# Overriding Standard Malloc
Overriding the standard `malloc` (and `new`) can be done either _dynamically_ or _statically_.
## Dynamic override
This is the recommended way to override the standard malloc interface.
### Dynamic Override on Linux, BSD
On these ELF-based systems we preload the mimalloc shared
library so all calls to the standard `malloc` interface are
resolved to the _mimalloc_ library.
```
> env LD_PRELOAD=/usr/lib/libmimalloc.so myprogram
```
You can set extra environment variables to check that mimalloc is running,
like:
```
> env MIMALLOC_VERBOSE=1 LD_PRELOAD=/usr/lib/libmimalloc.so myprogram
```
or run with the debug version to get detailed statistics:
```
> env MIMALLOC_SHOW_STATS=1 LD_PRELOAD=/usr/lib/libmimalloc-debug.so myprogram
```
### Dynamic Override on MacOS
On macOS we can also preload the mimalloc shared
library so all calls to the standard `malloc` interface are
resolved to the _mimalloc_ library.
```
> env DYLD_INSERT_LIBRARIES=/usr/lib/libmimalloc.dylib myprogram
```
Note that certain security restrictions may apply when doing this from
the [shell](https://stackoverflow.com/questions/43941322/dyld-insert-libraries-ignored-when-calling-application-through-bash).
### Dynamic Override on Windows
<span id="override_on_windows">We use a separate redirection DLL to override mimalloc on Windows</span>
such that we redirect all malloc/free calls that go through the (dynamic) C runtime allocator,
including those from other DLL's or libraries. As it intercepts all allocation calls on a low level,
it can be used on large programs that include other 3rd party components.
There are four requirements to make the overriding work well:
1. Use the C-runtime library as a DLL (using the `/MD` or `/MDd` switch).
2. Link your program explicitly with the `mimalloc.dll.lib` export library for the `mimalloc.dll`.
(which must be compiled with `-DMI_OVERRIDE=ON`, which is the default though).
To ensure the `mimalloc.dll` is actually loaded at run-time it is easiest
to insert some call to the mimalloc API in the `main` function, like `mi_version()`
(or use the `/include:mi_version` switch on the linker command, or
similarly, `#pragma comment(linker, "/include:mi_version")` in some source file).
See the `mimalloc-test-override` project for an example on how to use this.
3. The `mimalloc-redirect.dll` must be put in the same directory as the main
`mimalloc.dll` at runtime (as it is a dependency of that DLL).
The redirection DLL ensures that all calls to the C runtime malloc API get
redirected to mimalloc functions (which reside in `mimalloc.dll`).
4. Ensure the `mimalloc.dll` comes as early as possible in the import
list of the final executable (so it can intercept all potential allocations).
You can use `minject -l <exe>` to check this if needed.
For best performance on Windows with C++, it is also recommended to also override
the `new`/`delete` operations (by including [`mimalloc-new-delete.h`](include/mimalloc-new-delete.h)
a single(!) source file in your project).
The environment variable `MIMALLOC_DISABLE_REDIRECT=1` can be used to disable dynamic
overriding at run-time. Use `MIMALLOC_VERBOSE=1` to check if mimalloc was successfully
redirected.
For different platforms than x64, you may need a specific [redirection dll](bin).
Furthermore, we cannot always re-link an executable or ensure `mimalloc.dll` comes
first in the import table. In such cases the [`minject`](bin) tool can be used
to patch the executable's import tables.
## Static override
On Unix-like systems, you can also statically link with _mimalloc_ to override the standard
malloc interface. The recommended way is to link the final program with the
_mimalloc_ single object file (`mimalloc.o`). We use
an object file instead of a library file as linkers give preference to
that over archives to resolve symbols. To ensure that the standard
malloc interface resolves to the _mimalloc_ library, link it as the first
object file. For example:
```
> gcc -o myprogram mimalloc.o myfile1.c ...
```
Another way to override statically that works on all platforms, is to
link statically to mimalloc (as shown in the introduction) and include a
header file in each source file that re-defines `malloc` etc. to `mi_malloc`.
This is provided by [`mimalloc-override.h`](include/mimalloc-override.h). This only works
reliably though if all sources are
under your control or otherwise mixing of pointers from different heaps may occur!
# Tools
Generally, we recommend using the standard allocator with memory tracking tools, but mimalloc
can also be build to support the [address sanitizer][asan] or the excellent [Valgrind] tool.
Moreover, it can be build to support Windows event tracing ([ETW]).
This has a small performance overhead but does allow detecting memory leaks and byte-precise
buffer overflows directly on final executables. See also the `test/test-wrong.c` file to test with various tools.
## Valgrind
To build with [valgrind] support, use the `MI_TRACK_VALGRIND=ON` cmake option:
```
> cmake ../.. -DMI_TRACK_VALGRIND=ON
```
This can also be combined with secure mode or debug mode.
You can then run your programs directly under valgrind:
```
> valgrind <myprogram>
```
If you rely on overriding `malloc`/`free` by mimalloc (instead of using the `mi_malloc`/`mi_free` API directly),
you also need to tell `valgrind` to not intercept those calls itself, and use:
```
> MIMALLOC_SHOW_STATS=1 valgrind --soname-synonyms=somalloc=*mimalloc* -- <myprogram>
```
By setting the `MIMALLOC_SHOW_STATS` environment variable you can check that mimalloc is indeed
used and not the standard allocator. Even though the [Valgrind option][valgrind-soname]
is called `--soname-synonyms`, this also works when overriding with a static library or object file.
To dynamically override mimalloc using `LD_PRELOAD` together with `valgrind`, use:
```
> valgrind --trace-children=yes --soname-synonyms=somalloc=*mimalloc* /usr/bin/env LD_PRELOAD=/usr/lib/libmimalloc.so -- <myprogram>
```
See also the `test/test-wrong.c` file to test with `valgrind`.
Valgrind support is in its initial development -- please report any issues.
[Valgrind]: https://valgrind.org/
[valgrind-soname]: https://valgrind.org/docs/manual/manual-core.html#opt.soname-synonyms
## ASAN
To build with the address sanitizer, use the `-DMI_TRACK_ASAN=ON` cmake option:
```
> cmake ../.. -DMI_TRACK_ASAN=ON
```
This can also be combined with secure mode or debug mode.
You can then run your programs as:'
```
> ASAN_OPTIONS=verbosity=1 <myprogram>
```
When you link a program with an address sanitizer build of mimalloc, you should
generally compile that program too with the address sanitizer enabled.
For example, assuming you build mimalloc in `out/debug`:
```
clang -g -o test-wrong -Iinclude test/test-wrong.c out/debug/libmimalloc-asan-debug.a -lpthread -fsanitize=address -fsanitize-recover=address
```
Since the address sanitizer redirects the standard allocation functions, on some platforms (macOSX for example)
it is required to compile mimalloc with `-DMI_OVERRIDE=OFF`.
Address sanitizer support is in its initial development -- please report any issues.
[asan]: https://github.com/google/sanitizers/wiki/AddressSanitizer
## ETW
Event tracing for Windows ([ETW]) provides a high performance way to capture all allocations though
mimalloc and analyze them later. To build with ETW support, use the `-DMI_TRACK_ETW=ON` cmake option.
You can then capture an allocation trace using the Windows performance recorder (WPR), using the
`src/prim/windows/etw-mimalloc.wprp` profile. In an admin prompt, you can use:
```
> wpr -start src\prim\windows\etw-mimalloc.wprp -filemode
> <my_mimalloc_program>
> wpr -stop <my_mimalloc_program>.etl
```
and then open `<my_mimalloc_program>.etl` in the Windows Performance Analyzer (WPA), or
use a tool like [TraceControl] that is specialized for analyzing mimalloc traces.
[ETW]: https://learn.microsoft.com/en-us/windows-hardware/test/wpt/event-tracing-for-windows
[TraceControl]: https://github.com/xinglonghe/TraceControl
# Performance
Last update: 2021-01-30
We tested _mimalloc_ against many other top allocators over a wide
range of benchmarks, ranging from various real world programs to
synthetic benchmarks that see how the allocator behaves under more
extreme circumstances. In our benchmark suite, _mimalloc_ outperforms other leading
allocators (_jemalloc_, _tcmalloc_, _Hoard_, etc), and has a similar memory footprint. A nice property is that it
does consistently well over the wide range of benchmarks.
General memory allocators are interesting as there exists no algorithm that is
optimal -- for a given allocator one can usually construct a workload
where it does not do so well. The goal is thus to find an allocation
strategy that performs well over a wide range of benchmarks without
suffering from (too much) underperformance in less common situations.
As always, interpret these results with care since some benchmarks test synthetic
or uncommon situations that may never apply to your workloads. For example, most
allocators do not do well on `xmalloc-testN` but that includes even the best
industrial allocators like _jemalloc_ and _tcmalloc_ that are used in some of
the world's largest systems (like Chrome or FreeBSD).
Also, the benchmarks here do not measure the behaviour on very large and long-running server workloads,
or worst-case latencies of allocation. Much work has gone into `mimalloc` to work well on such
workloads (for example, to reduce virtual memory fragmentation on long-running services)
but such optimizations are not always reflected in the current benchmark suite.
We show here only an overview -- for
more specific details and further benchmarks we refer to the
[technical report](https://www.microsoft.com/en-us/research/publication/mimalloc-free-list-sharding-in-action).
The benchmark suite is automated and available separately
as [mimalloc-bench](https://github.com/daanx/mimalloc-bench).
## Benchmark Results on a 16-core AMD 5950x (Zen3)
Testing on the 16-core AMD 5950x processor at 3.4Ghz (4.9Ghz boost), with
with 32GiB memory at 3600Mhz, running Ubuntu 20.04 with glibc 2.31 and GCC 9.3.0.
We measure three versions of _mimalloc_: the main version `mi` (tag:v1.7.0),
the new v2.0 beta version as `xmi` (tag:v2.0.0), and the main version in secure mode as `smi` (tag:v1.7.0).
The other allocators are
Google's [_tcmalloc_](https://github.com/gperftools/gperftools) (`tc`, tag:gperftools-2.8.1) used in Chrome,
Facebook's [_jemalloc_](https://github.com/jemalloc/jemalloc) (`je`, tag:5.2.1) by Jason Evans used in Firefox and FreeBSD,
the Intel thread building blocks [allocator](https://github.com/intel/tbb) (`tbb`, tag:v2020.3),
[rpmalloc](https://github.com/mjansson/rpmalloc) (`rp`,tag:1.4.1) by Mattias Jansson,
the original scalable [_Hoard_](https://github.com/emeryberger/Hoard) (git:d880f72) allocator by Emery Berger \[1],
the memory compacting [_Mesh_](https://github.com/plasma-umass/Mesh) (git:67ff31a) allocator by
Bobby Powers _et al_ \[8],
and finally the default system allocator (`glibc`, 2.31) (based on _PtMalloc2_).
<img width="90%" src="doc/bench-2021/bench-amd5950x-2021-01-30-a.svg"/>
<img width="90%" src="doc/bench-2021/bench-amd5950x-2021-01-30-b.svg"/>
Any benchmarks ending in `N` run on all 32 logical cores in parallel.
Results are averaged over 10 runs and reported relative
to mimalloc (where 1.2 means it took 1.2&times; longer to run).
The legend also contains the _overall relative score_ between the
allocators where 100 points is the maximum if an allocator is fastest on
all benchmarks.
The single threaded _cfrac_ benchmark by Dave Barrett is an implementation of
continued fraction factorization which uses many small short-lived allocations.
All allocators do well on such common usage, where _mimalloc_ is just a tad
faster than _tcmalloc_ and
_jemalloc_.
The _leanN_ program is interesting as a large realistic and
concurrent workload of the [Lean](https://github.com/leanprover/lean)
theorem prover compiling its own standard library, and there is a 13%
speedup over _tcmalloc_. This is
quite significant: if Lean spends 20% of its time in the
allocator that means that _mimalloc_ is 1.6&times; faster than _tcmalloc_
here. (This is surprising as that is not measured in a pure
allocation benchmark like _alloc-test_. We conjecture that we see this
outsized improvement here because _mimalloc_ has better locality in
the allocation which improves performance for the *other* computations
in a program as well).
The single threaded _redis_ benchmark again show that most allocators do well on such workloads.
The _larsonN_ server benchmark by Larson and Krishnan \[2] allocates and frees between threads. They observed this
behavior (which they call _bleeding_) in actual server applications, and the benchmark simulates this.
Here, _mimalloc_ is quite a bit faster than _tcmalloc_ and _jemalloc_ probably due to the object migration between different threads.
The _mstressN_ workload performs many allocations and re-allocations,
and migrates objects between threads (as in _larsonN_). However, it also
creates and destroys the _N_ worker threads a few times keeping some objects
alive beyond the life time of the allocating thread. We observed this
behavior in many larger server applications.
The [_rptestN_](https://github.com/mjansson/rpmalloc-benchmark) benchmark
by Mattias Jansson is a allocator test originally designed
for _rpmalloc_, and tries to simulate realistic allocation patterns over
multiple threads. Here the differences between allocators become more apparent.
The second benchmark set tests specific aspects of the allocators and
shows even more extreme differences between them.
The _alloc-test_, by
[OLogN Technologies AG](http://ithare.com/testing-memory-allocators-ptmalloc2-tcmalloc-hoard-jemalloc-while-trying-to-simulate-real-world-loads/), is a very allocation intensive benchmark doing millions of
allocations in various size classes. The test is scaled such that when an
allocator performs almost identically on _alloc-test1_ as _alloc-testN_ it
means that it scales linearly.
The _sh6bench_ and _sh8bench_ benchmarks are
developed by [MicroQuill](http://www.microquill.com/) as part of SmartHeap.
In _sh6bench_ _mimalloc_ does much
better than the others (more than 2.5&times; faster than _jemalloc_).
We cannot explain this well but believe it is
caused in part by the "reverse" free-ing pattern in _sh6bench_.
The _sh8bench_ is a variation with object migration
between threads; whereas _tcmalloc_ did well on _sh6bench_, the addition of object migration causes it to be 10&times; slower than before.
The _xmalloc-testN_ benchmark by Lever and Boreham \[5] and Christian Eder, simulates an asymmetric workload where
some threads only allocate, and others only free -- they observed this pattern in
larger server applications. Here we see that
the _mimalloc_ technique of having non-contended sharded thread free
lists pays off as it outperforms others by a very large margin. Only _rpmalloc_, _tbb_, and _glibc_ also scale well on this benchmark.
The _cache-scratch_ benchmark by Emery Berger \[1], and introduced with
the Hoard allocator to test for _passive-false_ sharing of cache lines.
With a single thread they all
perform the same, but when running with multiple threads the potential allocator
induced false sharing of the cache lines can cause large run-time differences.
Crundal \[6] describes in detail why the false cache line sharing occurs in the _tcmalloc_ design, and also discusses how this
can be avoided with some small implementation changes.
Only the _tbb_, _rpmalloc_ and _mesh_ allocators also avoid the
cache line sharing completely, while _Hoard_ and _glibc_ seem to mitigate
the effects. Kukanov and Voss \[7] describe in detail
how the design of _tbb_ avoids the false cache line sharing.
## On a 36-core Intel Xeon
For completeness, here are the results on a big Amazon
[c5.18xlarge](https://aws.amazon.com/ec2/instance-types/#Compute_Optimized) instance
consisting of a 2&times;18-core Intel Xeon (Cascade Lake) at 3.4GHz (boost 3.5GHz)
with 144GiB ECC memory, running Ubuntu 20.04 with glibc 2.31, GCC 9.3.0, and
Clang 10.0.0. This time, the mimalloc allocators (mi, xmi, and smi) were
compiled with the Clang compiler instead of GCC.
The results are similar to the AMD results but it is interesting to
see the differences in the _larsonN_, _mstressN_, and _xmalloc-testN_ benchmarks.
<img width="90%" src="doc/bench-2021/bench-c5-18xlarge-2021-01-30-a.svg"/>
<img width="90%" src="doc/bench-2021/bench-c5-18xlarge-2021-01-30-b.svg"/>
## Peak Working Set
The following figure shows the peak working set (rss) of the allocators
on the benchmarks (on the c5.18xlarge instance).
<img width="90%" src="doc/bench-2021/bench-c5-18xlarge-2021-01-30-rss-a.svg"/>
<img width="90%" src="doc/bench-2021/bench-c5-18xlarge-2021-01-30-rss-b.svg"/>
Note that the _xmalloc-testN_ memory usage should be disregarded as it
allocates more the faster the program runs. Similarly, memory usage of
_larsonN_, _mstressN_, _rptestN_ and _sh8bench_ can vary depending on scheduling and
speed. Nevertheless, we hope to improve the memory usage on _mstressN_
and _rptestN_ (just as _cfrac_, _larsonN_ and _sh8bench_ have a small working set which skews the results).
<!--
# Previous Benchmarks
Todo: should we create a separate page for this?
## Benchmark Results on 36-core Intel: 2020-01-20
Testing on a big Amazon EC2 compute instance
([c5.18xlarge](https://aws.amazon.com/ec2/instance-types/#Compute_Optimized))
consisting of a 72 processor Intel Xeon at 3GHz
with 144GiB ECC memory, running Ubuntu 18.04.1 with glibc 2.27 and GCC 7.4.0.
The measured allocators are _mimalloc_ (xmi, tag:v1.4.0, page reset enabled)
and its secure build as _smi_,
Google's [_tcmalloc_](https://github.com/gperftools/gperftools) (tc, tag:gperftools-2.7) used in Chrome,
Facebook's [_jemalloc_](https://github.com/jemalloc/jemalloc) (je, tag:5.2.1) by Jason Evans used in Firefox and FreeBSD,
the Intel thread building blocks [allocator](https://github.com/intel/tbb) (tbb, tag:2020),
[rpmalloc](https://github.com/mjansson/rpmalloc) (rp,tag:1.4.0) by Mattias Jansson,
the original scalable [_Hoard_](https://github.com/emeryberger/Hoard) (tag:3.13) allocator by Emery Berger \[1],
the memory compacting [_Mesh_](https://github.com/plasma-umass/Mesh) (git:51222e7) allocator by
Bobby Powers _et al_ \[8],
and finally the default system allocator (glibc, 2.27) (based on _PtMalloc2_).
<img width="90%" src="doc/bench-2020/bench-c5-18xlarge-2020-01-20-a.svg"/>
<img width="90%" src="doc/bench-2020/bench-c5-18xlarge-2020-01-20-b.svg"/>
The following figure shows the peak working set (rss) of the allocators
on the benchmarks (on the c5.18xlarge instance).
<img width="90%" src="doc/bench-2020/bench-c5-18xlarge-2020-01-20-rss-a.svg"/>
<img width="90%" src="doc/bench-2020/bench-c5-18xlarge-2020-01-20-rss-b.svg"/>
## On 24-core AMD Epyc, 2020-01-16
For completeness, here are the results on a
[r5a.12xlarge](https://aws.amazon.com/ec2/instance-types/#Memory_Optimized) instance
having a 48 processor AMD Epyc 7000 at 2.5GHz with 384GiB of memory.
The results are similar to the Intel results but it is interesting to
see the differences in the _larsonN_, _mstressN_, and _xmalloc-testN_ benchmarks.
<img width="90%" src="doc/bench-2020/bench-r5a-12xlarge-2020-01-16-a.svg"/>
<img width="90%" src="doc/bench-2020/bench-r5a-12xlarge-2020-01-16-b.svg"/>
-->
# References
- \[1] Emery D. Berger, Kathryn S. McKinley, Robert D. Blumofe, and Paul R. Wilson.
_Hoard: A Scalable Memory Allocator for Multithreaded Applications_
the Ninth International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS-IX). Cambridge, MA, November 2000.
[pdf](http://www.cs.utexas.edu/users/mckinley/papers/asplos-2000.pdf)
- \[2] P. Larson and M. Krishnan. _Memory allocation for long-running server applications_.
In ISMM, Vancouver, B.C., Canada, 1998. [pdf](http://citeseer.ist.psu.edu/viewdoc/download?doi=10.1.1.45.1947&rep=rep1&type=pdf)
- \[3] D. Grunwald, B. Zorn, and R. Henderson.
_Improving the cache locality of memory allocation_. In R. Cartwright, editor,
Proceedings of the Conference on Programming Language Design and Implementation, pages 177–186, New York, NY, USA, June 1993. [pdf](http://citeseer.ist.psu.edu/viewdoc/download?doi=10.1.1.43.6621&rep=rep1&type=pdf)
- \[4] J. Barnes and P. Hut. _A hierarchical O(n*log(n)) force-calculation algorithm_. Nature, 324:446-449, 1986.
- \[5] C. Lever, and D. Boreham. _Malloc() Performance in a Multithreaded Linux Environment._
In USENIX Annual Technical Conference, Freenix Session. San Diego, CA. Jun. 2000.
Available at <https://github.com/kuszmaul/SuperMalloc/tree/master/tests>
- \[6] Timothy Crundal. _Reducing Active-False Sharing in TCMalloc_. 2016. CS16S1 project at the Australian National University. [pdf](http://courses.cecs.anu.edu.au/courses/CSPROJECTS/16S1/Reports/Timothy_Crundal_Report.pdf)
- \[7] Alexey Kukanov, and Michael J Voss.
_The Foundations for Scalable Multi-Core Software in Intel Threading Building Blocks._
Intel Technology Journal 11 (4). 2007
- \[8] Bobby Powers, David Tench, Emery D. Berger, and Andrew McGregor.
_Mesh: Compacting Memory Management for C/C++_
In Proceedings of the 40th ACM SIGPLAN Conference on Programming Language Design and Implementation (PLDI'19), June 2019, pages 333-–346.
<!--
- \[9] Paul Liétar, Theodore Butler, Sylvan Clebsch, Sophia Drossopoulou, Juliana Franco, Matthew J Parkinson,
Alex Shamis, Christoph M Wintersteiger, and David Chisnall.
_Snmalloc: A Message Passing Allocator._
In Proceedings of the 2019 ACM SIGPLAN International Symposium on Memory Management, 122–135. ACM. 2019.
-->
# Contributing
This project welcomes contributions and suggestions. Most contributions require you to agree to a
Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us
the rights to use your contribution. For details, visit https://cla.microsoft.com.
When you submit a pull request, a CLA-bot will automatically determine whether you need to provide
a CLA and decorate the PR appropriately (e.g., label, comment). Simply follow the instructions
provided by the bot. You will only need to do this once across all repos using our CLA.
# Older Release Notes
* 2021-11-14, `v1.7.3`, `v2.0.3` (beta): improved WASM support, improved macOS support and performance (including
M1), improved performance for v2 for large objects, Python integration improvements, more standard
installation directories, various small fixes.
* 2021-06-17, `v1.7.2`, `v2.0.2` (beta): support M1, better installation layout on Linux, fix
thread_id on Android, prefer 2-6TiB area for aligned allocation to work better on pre-windows 8, various small fixes.
* 2021-04-06, `v1.7.1`, `v2.0.1` (beta): fix bug in arena allocation for huge pages, improved aslr on large allocations, initial M1 support (still experimental).
* 2021-01-31, `v2.0.0`: beta release 2.0: new slice algorithm for managing internal mimalloc pages.
* 2021-01-31, `v1.7.0`: stable release 1.7: support explicit user provided memory regions, more precise statistics,
improve macOS overriding, initial support for Apple M1, improved DragonFly support, faster memcpy on Windows, various small fixes.
* 2020-09-24, `v1.6.7`: stable release 1.6: using standard C atomics, passing tsan testing, improved
handling of failing to commit on Windows, add [`mi_process_info`](https://github.com/microsoft/mimalloc/blob/master/include/mimalloc.h#L156) api call.
* 2020-08-06, `v1.6.4`: stable release 1.6: improved error recovery in low-memory situations,
support for IllumOS and Haiku, NUMA support for Vista/XP, improved NUMA detection for AMD Ryzen, ubsan support.
* 2020-05-05, `v1.6.3`: stable release 1.6: improved behavior in out-of-memory situations, improved malloc zones on macOS,
build PIC static libraries by default, add option to abort on out-of-memory, line buffered statistics.
* 2020-04-20, `v1.6.2`: stable release 1.6: fix compilation on Android, MingW, Raspberry, and Conda,
stability fix for Windows 7, fix multiple mimalloc instances in one executable, fix `strnlen` overload,
fix aligned debug padding.
* 2020-02-17, `v1.6.1`: stable release 1.6: minor updates (build with clang-cl, fix alignment issue for small objects).
* 2020-02-09, `v1.6.0`: stable release 1.6: fixed potential memory leak, improved overriding
and thread local support on FreeBSD, NetBSD, DragonFly, and macOSX. New byte-precise
heap block overflow detection in debug mode (besides the double-free detection and free-list
corruption detection). Add `nodiscard` attribute to most allocation functions.
Enable `MIMALLOC_PAGE_RESET` by default. New reclamation strategy for abandoned heap pages
for better memory footprint.
* 2020-02-09, `v1.5.0`: stable release 1.5: improved free performance, small bug fixes.
* 2020-01-22, `v1.4.0`: stable release 1.4: improved performance for delayed OS page reset,
more eager concurrent free, addition of STL allocator, fixed potential memory leak.
* 2020-01-15, `v1.3.0`: stable release 1.3: bug fixes, improved randomness and [stronger
free list encoding](https://github.com/microsoft/mimalloc/blob/783e3377f79ee82af43a0793910a9f2d01ac7863/include/mimalloc-internal.h#L396) in secure mode.
* 2019-12-22, `v1.2.2`: stable release 1.2: minor updates.
* 2019-11-22, `v1.2.0`: stable release 1.2: bug fixes, improved secure mode (free list corruption checks, double free mitigation). Improved dynamic overriding on Windows.
* 2019-10-07, `v1.1.0`: stable release 1.1.
* 2019-09-01, `v1.0.8`: pre-release 8: more robust windows dynamic overriding, initial huge page support.
* 2019-08-10, `v1.0.6`: pre-release 6: various performance improvements.

3
mimalloc/mimalloc-sys/mimalloc/include/mimalloc-override.h
View File

@@ -24,7 +24,7 @@ not accidentally mix pointers from different allocators).
#define free(p) mi_free(p)
#define strdup(s) mi_strdup(s)
#define strndup(s,n) mi_strndup(s,n)
#define strndup(s,n) mi_strndup(s,n)
#define realpath(f,n) mi_realpath(f,n)
// Microsoft extensions
@@ -43,6 +43,7 @@ not accidentally mix pointers from different allocators).
#define reallocf(p,n) mi_reallocf(p,n)
#define malloc_size(p) mi_usable_size(p)
#define malloc_usable_size(p) mi_usable_size(p)
#define malloc_good_size(sz) mi_malloc_good_size(sz)
#define cfree(p) mi_free(p)
#define valloc(n) mi_valloc(n)

104
mimalloc/mimalloc-sys/mimalloc/include/mimalloc-stats.h Normal file
View File

@@ -0,0 +1,104 @@
/* ----------------------------------------------------------------------------
Copyright (c) 2018-2025, Microsoft Research, Daan Leijen
This is free software; you can redistribute it and/or modify it under the
terms of the MIT license. A copy of the license can be found in the file
"LICENSE" at the root of this distribution.
-----------------------------------------------------------------------------*/
#pragma once
#ifndef MIMALLOC_STATS_H
#define MIMALLOC_STATS_H
#include <mimalloc.h>
#include <stdint.h>
#define MI_STAT_VERSION 3 // increased on every backward incompatible change
// count allocation over time
typedef struct mi_stat_count_s {
int64_t total; // total allocated
int64_t peak; // peak allocation
int64_t current; // current allocation
} mi_stat_count_t;
// counters only increase
typedef struct mi_stat_counter_s {
int64_t total; // total count
} mi_stat_counter_t;
#define MI_STAT_FIELDS() \
MI_STAT_COUNT(pages) /* count of mimalloc pages */ \
MI_STAT_COUNT(reserved) /* reserved memory bytes */ \
MI_STAT_COUNT(committed) /* committed bytes */ \
MI_STAT_COUNTER(reset) /* reset bytes */ \
MI_STAT_COUNTER(purged) /* purged bytes */ \
MI_STAT_COUNT(page_committed) /* committed memory inside pages */ \
MI_STAT_COUNT(pages_abandoned) /* abandonded pages count */ \
MI_STAT_COUNT(threads) /* number of threads */ \
MI_STAT_COUNT(malloc_normal) /* allocated bytes <= MI_LARGE_OBJ_SIZE_MAX */ \
MI_STAT_COUNT(malloc_huge) /* allocated bytes in huge pages */ \
MI_STAT_COUNT(malloc_requested) /* malloc requested bytes */ \
\
MI_STAT_COUNTER(mmap_calls) \
MI_STAT_COUNTER(commit_calls) \
MI_STAT_COUNTER(reset_calls) \
MI_STAT_COUNTER(purge_calls) \
MI_STAT_COUNTER(arena_count) /* number of memory arena's */ \
MI_STAT_COUNTER(malloc_normal_count) /* number of blocks <= MI_LARGE_OBJ_SIZE_MAX */ \
MI_STAT_COUNTER(malloc_huge_count) /* number of huge bloks */ \
MI_STAT_COUNTER(malloc_guarded_count) /* number of allocations with guard pages */ \
\
/* internal statistics */ \
MI_STAT_COUNTER(arena_rollback_count) \
MI_STAT_COUNTER(arena_purges) \
MI_STAT_COUNTER(pages_extended) /* number of page extensions */ \
MI_STAT_COUNTER(pages_retire) /* number of pages that are retired */ \
MI_STAT_COUNTER(page_searches) /* total pages searched for a fresh page */ \
MI_STAT_COUNTER(page_searches_count) /* searched count for a fresh page */ \
/* only on v1 and v2 */ \
MI_STAT_COUNT(segments) \
MI_STAT_COUNT(segments_abandoned) \
MI_STAT_COUNT(segments_cache) \
MI_STAT_COUNT(_segments_reserved) \
/* only on v3 */ \
MI_STAT_COUNTER(pages_reclaim_on_alloc) \
MI_STAT_COUNTER(pages_reclaim_on_free) \
MI_STAT_COUNTER(pages_reabandon_full) \
MI_STAT_COUNTER(pages_unabandon_busy_wait) \
// Define the statistics structure
#define MI_BIN_HUGE (73U) // see types.h
#define MI_STAT_COUNT(stat) mi_stat_count_t stat;
#define MI_STAT_COUNTER(stat) mi_stat_counter_t stat;
typedef struct mi_stats_s
{
int version;
MI_STAT_FIELDS()
// future extension
mi_stat_count_t _stat_reserved[4];
mi_stat_counter_t _stat_counter_reserved[4];
// size segregated statistics
mi_stat_count_t malloc_bins[MI_BIN_HUGE+1]; // allocation per size bin
mi_stat_count_t page_bins[MI_BIN_HUGE+1]; // pages allocated per size bin
} mi_stats_t;
#undef MI_STAT_COUNT
#undef MI_STAT_COUNTER
// Exported definitions
#ifdef __cplusplus
extern "C" {
#endif
mi_decl_export void mi_stats_get( size_t stats_size, mi_stats_t* stats ) mi_attr_noexcept;
mi_decl_export char* mi_stats_get_json( size_t buf_size, char* buf ) mi_attr_noexcept; // use mi_free to free the result if the input buf == NULL
#ifdef __cplusplus
}
#endif
#endif // MIMALLOC_STATS_H

148
mimalloc/mimalloc-sys/mimalloc/include/mimalloc.h
View File

@@ -1,5 +1,5 @@
/* ----------------------------------------------------------------------------
Copyright (c) 2018-2023, Microsoft Research, Daan Leijen
Copyright (c) 2018-2026, Microsoft Research, Daan Leijen
This is free software; you can redistribute it and/or modify it under the
terms of the MIT license. A copy of the license can be found in the file
"LICENSE" at the root of this distribution.
@@ -8,7 +8,7 @@ terms of the MIT license. A copy of the license can be found in the file
#ifndef MIMALLOC_H
#define MIMALLOC_H
#define MI_MALLOC_VERSION 212 // major + 2 digits minor
#define MI_MALLOC_VERSION 226 // major + 2 digits minor
// ------------------------------------------------------
// Compiler specific attributes
@@ -154,16 +154,21 @@ mi_decl_export void mi_stats_reset(void) mi_attr_noexcept;
mi_decl_export void mi_stats_merge(void) mi_attr_noexcept;
mi_decl_export void mi_stats_print(void* out) mi_attr_noexcept; // backward compatibility: `out` is ignored and should be NULL
mi_decl_export void mi_stats_print_out(mi_output_fun* out, void* arg) mi_attr_noexcept;
mi_decl_export void mi_process_init(void) mi_attr_noexcept;
mi_decl_export void mi_thread_init(void) mi_attr_noexcept;
mi_decl_export void mi_thread_done(void) mi_attr_noexcept;
mi_decl_export void mi_thread_stats_print_out(mi_output_fun* out, void* arg) mi_attr_noexcept;
mi_decl_export void mi_options_print(void) mi_attr_noexcept;
mi_decl_export void mi_process_info(size_t* elapsed_msecs, size_t* user_msecs, size_t* system_msecs,
size_t* current_rss, size_t* peak_rss,
size_t* current_commit, size_t* peak_commit, size_t* page_faults) mi_attr_noexcept;
// Generally do not use the following as these are usually called automatically
mi_decl_export void mi_process_init(void) mi_attr_noexcept;
mi_decl_export void mi_cdecl mi_process_done(void) mi_attr_noexcept;
mi_decl_export void mi_thread_init(void) mi_attr_noexcept;
mi_decl_export void mi_thread_done(void) mi_attr_noexcept;
// -------------------------------------------------------------------------------------
// Aligned allocation
// Note that `alignment` always follows `size` for consistency with unaligned
@@ -180,6 +185,22 @@ mi_decl_nodiscard mi_decl_export void* mi_realloc_aligned(void* p, size_t newsiz
mi_decl_nodiscard mi_decl_export void* mi_realloc_aligned_at(void* p, size_t newsize, size_t alignment, size_t offset) mi_attr_noexcept mi_attr_alloc_size(2);
// -----------------------------------------------------------------
// Return allocated block size (if the return value is not NULL)
// -----------------------------------------------------------------
mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_umalloc(size_t size, size_t* block_size) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(1);
mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_ucalloc(size_t count, size_t size, size_t* block_size) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size2(1,2);
mi_decl_nodiscard mi_decl_export void* mi_urealloc(void* p, size_t newsize, size_t* block_size_pre, size_t* block_size_post) mi_attr_noexcept mi_attr_alloc_size(2);
mi_decl_export void mi_ufree(void* p, size_t* block_size) mi_attr_noexcept;
mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_umalloc_aligned(size_t size, size_t alignment, size_t* block_size) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(1) mi_attr_alloc_align(2);
mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_uzalloc_aligned(size_t size, size_t alignment, size_t* block_size) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(1) mi_attr_alloc_align(2);
mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_umalloc_small(size_t size, size_t* block_size) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(1);
mi_decl_nodiscard mi_decl_export mi_decl_restrict void* mi_uzalloc_small(size_t size, size_t* block_size) mi_attr_noexcept mi_attr_malloc mi_attr_alloc_size(1);
// -------------------------------------------------------------------------------------
// Heaps: first-class, but can only allocate from the same thread that created it.
// -------------------------------------------------------------------------------------
@@ -259,23 +280,25 @@ typedef struct mi_heap_area_s {
size_t used; // number of allocated blocks
size_t block_size; // size in bytes of each block
size_t full_block_size; // size in bytes of a full block including padding and metadata.
int heap_tag; // heap tag associated with this area
} mi_heap_area_t;
typedef bool (mi_cdecl mi_block_visit_fun)(const mi_heap_t* heap, const mi_heap_area_t* area, void* block, size_t block_size, void* arg);
mi_decl_export bool mi_heap_visit_blocks(const mi_heap_t* heap, bool visit_all_blocks, mi_block_visit_fun* visitor, void* arg);
mi_decl_export bool mi_heap_visit_blocks(const mi_heap_t* heap, bool visit_blocks, mi_block_visit_fun* visitor, void* arg);
// Experimental
mi_decl_nodiscard mi_decl_export bool mi_is_in_heap_region(const void* p) mi_attr_noexcept;
mi_decl_nodiscard mi_decl_export bool mi_is_redirected(void) mi_attr_noexcept;
mi_decl_export int mi_reserve_huge_os_pages_interleave(size_t pages, size_t numa_nodes, size_t timeout_msecs) mi_attr_noexcept;
mi_decl_export int mi_reserve_huge_os_pages_at(size_t pages, int numa_node, size_t timeout_msecs) mi_attr_noexcept;
mi_decl_export int mi_reserve_huge_os_pages_interleave(size_t pages, size_t numa_nodes, size_t timeout_msecs) mi_attr_noexcept;
mi_decl_export int mi_reserve_huge_os_pages_at(size_t pages, int numa_node, size_t timeout_msecs) mi_attr_noexcept;
mi_decl_export int mi_reserve_os_memory(size_t size, bool commit, bool allow_large) mi_attr_noexcept;
mi_decl_export bool mi_manage_os_memory(void* start, size_t size, bool is_committed, bool is_large, bool is_zero, int numa_node) mi_attr_noexcept;
mi_decl_export int mi_reserve_os_memory(size_t size, bool commit, bool allow_large) mi_attr_noexcept;
mi_decl_export bool mi_manage_os_memory(void* start, size_t size, bool is_committed, bool is_large, bool is_zero, int numa_node) mi_attr_noexcept;
mi_decl_export void mi_debug_show_arenas(void) mi_attr_noexcept;
mi_decl_export void mi_debug_show_arenas(void) mi_attr_noexcept;
mi_decl_export void mi_arenas_print(void) mi_attr_noexcept;
// Experimental: heaps associated with specific memory arena's
typedef int mi_arena_id_t;
@@ -289,8 +312,36 @@ mi_decl_export bool mi_manage_os_memory_ex(void* start, size_t size, bool is_co
mi_decl_nodiscard mi_decl_export mi_heap_t* mi_heap_new_in_arena(mi_arena_id_t arena_id);
#endif
// Experimental: allow sub-processes whose memory areas stay separated (and no reclamation between them)
// Used for example for separate interpreters in one process.
typedef void* mi_subproc_id_t;
mi_decl_export mi_subproc_id_t mi_subproc_main(void);
mi_decl_export mi_subproc_id_t mi_subproc_new(void);
mi_decl_export void mi_subproc_delete(mi_subproc_id_t subproc);
mi_decl_export void mi_subproc_add_current_thread(mi_subproc_id_t subproc); // this should be called right after a thread is created (and no allocation has taken place yet)
// Experimental: visit abandoned heap areas (that are not owned by a specific heap)
mi_decl_export bool mi_abandoned_visit_blocks(mi_subproc_id_t subproc_id, int heap_tag, bool visit_blocks, mi_block_visit_fun* visitor, void* arg);
// Experimental: objects followed by a guard page.
// A sample rate of 0 disables guarded objects, while 1 uses a guard page for every object.
// A seed of 0 uses a random start point. Only objects within the size bound are eligable for guard pages.
mi_decl_export void mi_heap_guarded_set_sample_rate(mi_heap_t* heap, size_t sample_rate, size_t seed);
mi_decl_export void mi_heap_guarded_set_size_bound(mi_heap_t* heap, size_t min, size_t max);
// Experimental: communicate that the thread is part of a threadpool
mi_decl_export void mi_thread_set_in_threadpool(void) mi_attr_noexcept;
// Experimental: create a new heap with a specified heap tag. Set `allow_destroy` to false to allow the thread
// to reclaim abandoned memory (with a compatible heap_tag and arena_id) but in that case `mi_heap_destroy` will
// fall back to `mi_heap_delete`.
mi_decl_nodiscard mi_decl_export mi_heap_t* mi_heap_new_ex(int heap_tag, bool allow_destroy, mi_arena_id_t arena_id);
// deprecated
mi_decl_export int mi_reserve_huge_os_pages(size_t pages, double max_secs, size_t* pages_reserved) mi_attr_noexcept;
mi_decl_export int mi_reserve_huge_os_pages(size_t pages, double max_secs, size_t* pages_reserved) mi_attr_noexcept;
mi_decl_export void mi_collect_reduce(size_t target_thread_owned) mi_attr_noexcept;
// ------------------------------------------------------
@@ -318,40 +369,53 @@ mi_decl_export int mi_reserve_huge_os_pages(size_t pages, double max_secs, size
typedef enum mi_option_e {
// stable options
mi_option_show_errors, // print error messages
mi_option_show_stats, // print statistics on termination
mi_option_verbose, // print verbose messages
// the following options are experimental (see src/options.h)
mi_option_eager_commit, // eager commit segments? (after `eager_commit_delay` segments) (=1)
mi_option_arena_eager_commit, // eager commit arenas? Use 2 to enable just on overcommit systems (=2)
mi_option_purge_decommits, // should a memory purge decommit (or only reset) (=1)
mi_option_allow_large_os_pages, // allow large (2MiB) OS pages, implies eager commit
mi_option_reserve_huge_os_pages, // reserve N huge OS pages (1GiB/page) at startup
mi_option_reserve_huge_os_pages_at, // reserve huge OS pages at a specific NUMA node
mi_option_reserve_os_memory, // reserve specified amount of OS memory in an arena at startup
mi_option_show_errors, // print error messages
mi_option_show_stats, // print statistics on termination
mi_option_verbose, // print verbose messages
// advanced options
mi_option_eager_commit, // eager commit segments? (after `eager_commit_delay` segments) (=1)
mi_option_arena_eager_commit, // eager commit arenas? Use 2 to enable just on overcommit systems (=2)
mi_option_purge_decommits, // should a memory purge decommit? (=1). Set to 0 to use memory reset on a purge (instead of decommit)
mi_option_allow_large_os_pages, // allow use of large (2 or 4 MiB) OS pages, implies eager commit.
mi_option_reserve_huge_os_pages, // reserve N huge OS pages (1GiB pages) at startup
mi_option_reserve_huge_os_pages_at, // reserve huge OS pages at a specific NUMA node
mi_option_reserve_os_memory, // reserve specified amount of OS memory in an arena at startup (internally, this value is in KiB; use `mi_option_get_size`)
mi_option_deprecated_segment_cache,
mi_option_deprecated_page_reset,
mi_option_abandoned_page_purge, // immediately purge delayed purges on thread termination
mi_option_deprecated_segment_reset,
mi_option_eager_commit_delay,
mi_option_purge_delay, // memory purging is delayed by N milli seconds; use 0 for immediate purging or -1 for no purging at all.
mi_option_use_numa_nodes, // 0 = use all available numa nodes, otherwise use at most N nodes.
mi_option_limit_os_alloc, // 1 = do not use OS memory for allocation (but only programmatically reserved arenas)
mi_option_os_tag, // tag used for OS logging (macOS only for now)
mi_option_max_errors, // issue at most N error messages
mi_option_max_warnings, // issue at most N warning messages
mi_option_max_segment_reclaim,
mi_option_destroy_on_exit, // if set, release all memory on exit; sometimes used for dynamic unloading but can be unsafe.
mi_option_arena_reserve, // initial memory size in KiB for arena reservation (1GiB on 64-bit)
mi_option_arena_purge_mult,
mi_option_abandoned_page_purge, // immediately purge delayed purges on thread termination
mi_option_deprecated_segment_reset,
mi_option_eager_commit_delay, // the first N segments per thread are not eagerly committed (but per page in the segment on demand)
mi_option_purge_delay, // memory purging is delayed by N milli seconds; use 0 for immediate purging or -1 for no purging at all. (=10)
mi_option_use_numa_nodes, // 0 = use all available numa nodes, otherwise use at most N nodes.
mi_option_disallow_os_alloc, // 1 = do not use OS memory for allocation (but only programmatically reserved arenas)
mi_option_os_tag, // tag used for OS logging (macOS only for now) (=100)
mi_option_max_errors, // issue at most N error messages
mi_option_max_warnings, // issue at most N warning messages
mi_option_max_segment_reclaim, // max. percentage of the abandoned segments can be reclaimed per try (=10%)
mi_option_destroy_on_exit, // if set, release all memory on exit; sometimes used for dynamic unloading but can be unsafe
mi_option_arena_reserve, // initial memory size for arena reservation (= 1 GiB on 64-bit) (internally, this value is in KiB; use `mi_option_get_size`)
mi_option_arena_purge_mult, // multiplier for `purge_delay` for the purging delay for arenas (=10)
mi_option_purge_extend_delay,
mi_option_abandoned_reclaim_on_free, // allow to reclaim an abandoned segment on a free (=1)
mi_option_disallow_arena_alloc, // 1 = do not use arena's for allocation (except if using specific arena id's)
mi_option_retry_on_oom, // retry on out-of-memory for N milli seconds (=400), set to 0 to disable retries. (only on windows)
mi_option_visit_abandoned, // allow visiting heap blocks from abandoned threads (=0)
mi_option_guarded_min, // only used when building with MI_GUARDED: minimal rounded object size for guarded objects (=0)
mi_option_guarded_max, // only used when building with MI_GUARDED: maximal rounded object size for guarded objects (=0)
mi_option_guarded_precise, // disregard minimal alignment requirement to always place guarded blocks exactly in front of a guard page (=0)
mi_option_guarded_sample_rate, // 1 out of N allocations in the min/max range will be guarded (=1000)
mi_option_guarded_sample_seed, // can be set to allow for a (more) deterministic re-execution when a guard page is triggered (=0)
mi_option_target_segments_per_thread, // experimental (=0)
mi_option_generic_collect, // collect heaps every N (=10000) generic allocation calls
mi_option_allow_thp, // allow transparent huge pages? (=1) (on Android =0 by default). Set to 0 to disable THP for the process.
_mi_option_last,
// legacy option names
mi_option_large_os_pages = mi_option_allow_large_os_pages,
mi_option_eager_region_commit = mi_option_arena_eager_commit,
mi_option_reset_decommits = mi_option_purge_decommits,
mi_option_reset_delay = mi_option_purge_delay,
mi_option_abandoned_page_reset = mi_option_abandoned_page_purge
mi_option_abandoned_page_reset = mi_option_abandoned_page_purge,
mi_option_limit_os_alloc = mi_option_disallow_os_alloc
} mi_option_t;
@@ -494,7 +558,7 @@ template<class T, bool _mi_destroy> struct _mi_heap_stl_allocator_common : publi
using typename _mi_stl_allocator_common<T>::value_type;
using typename _mi_stl_allocator_common<T>::pointer;
_mi_heap_stl_allocator_common(mi_heap_t* hp) : heap(hp) { } /* will not delete nor destroy the passed in heap */
_mi_heap_stl_allocator_common(mi_heap_t* hp) : heap(hp, [](mi_heap_t*) {}) {} /* will not delete nor destroy the passed in heap */
#if (__cplusplus >= 201703L) // C++17
mi_decl_nodiscard T* allocate(size_type count) { return static_cast<T*>(mi_heap_alloc_new_n(this->heap.get(), count, sizeof(T))); }
@@ -513,7 +577,7 @@ template<class T, bool _mi_destroy> struct _mi_heap_stl_allocator_common : publi
protected:
std::shared_ptr<mi_heap_t> heap;
template<class U, bool D> friend struct _mi_heap_stl_allocator_common;
_mi_heap_stl_allocator_common() {
mi_heap_t* hp = mi_heap_new();
this->heap.reset(hp, (_mi_destroy ? &heap_destroy : &heap_delete)); /* calls heap_delete/destroy when the refcount drops to zero */
@@ -530,7 +594,7 @@ private:
template<class T> struct mi_heap_stl_allocator : public _mi_heap_stl_allocator_common<T, false> {
using typename _mi_heap_stl_allocator_common<T, false>::size_type;
mi_heap_stl_allocator() : _mi_heap_stl_allocator_common<T, false>() { } // creates fresh heap that is deleted when the destructor is called
mi_heap_stl_allocator(mi_heap_t* hp) : _mi_heap_stl_allocator_common<T, false>(hp) { } // no delete nor destroy on the passed in heap
mi_heap_stl_allocator(mi_heap_t* hp) : _mi_heap_stl_allocator_common<T, false>(hp) { } // no delete nor destroy on the passed in heap
template<class U> mi_heap_stl_allocator(const mi_heap_stl_allocator<U>& x) mi_attr_noexcept : _mi_heap_stl_allocator_common<T, false>(x) { }
mi_heap_stl_allocator select_on_container_copy_construction() const { return *this; }
@@ -547,7 +611,7 @@ template<class T1, class T2> bool operator!=(const mi_heap_stl_allocator<T1>& x,
template<class T> struct mi_heap_destroy_stl_allocator : public _mi_heap_stl_allocator_common<T, true> {
using typename _mi_heap_stl_allocator_common<T, true>::size_type;
mi_heap_destroy_stl_allocator() : _mi_heap_stl_allocator_common<T, true>() { } // creates fresh heap that is destroyed when the destructor is called
mi_heap_destroy_stl_allocator(mi_heap_t* hp) : _mi_heap_stl_allocator_common<T, true>(hp) { } // no delete nor destroy on the passed in heap
mi_heap_destroy_stl_allocator(mi_heap_t* hp) : _mi_heap_stl_allocator_common<T, true>(hp) { } // no delete nor destroy on the passed in heap
template<class U> mi_heap_destroy_stl_allocator(const mi_heap_destroy_stl_allocator<U>& x) mi_attr_noexcept : _mi_heap_stl_allocator_common<T, true>(x) { }
mi_heap_destroy_stl_allocator select_on_container_copy_construction() const { return *this; }

234
mimalloc/mimalloc-sys/mimalloc/include/mimalloc/atomic.h
View File

@@ -1,5 +1,5 @@
/* ----------------------------------------------------------------------------
Copyright (c) 2018-2023 Microsoft Research, Daan Leijen
Copyright (c) 2018-2024 Microsoft Research, Daan Leijen
This is free software; you can redistribute it and/or modify it under the
terms of the MIT license. A copy of the license can be found in the file
"LICENSE" at the root of this distribution.
@@ -8,10 +8,21 @@ terms of the MIT license. A copy of the license can be found in the file
#ifndef MIMALLOC_ATOMIC_H
#define MIMALLOC_ATOMIC_H
// include windows.h or pthreads.h
#if defined(_WIN32)
#ifndef WIN32_LEAN_AND_MEAN
#define WIN32_LEAN_AND_MEAN
#endif
#include <windows.h>
#elif !defined(__wasi__) && (!defined(__EMSCRIPTEN__) || defined(__EMSCRIPTEN_PTHREADS__))
#define MI_USE_PTHREADS
#include <pthread.h>
#endif
// --------------------------------------------------------------------------------------------
// Atomics
// We need to be portable between C, C++, and MSVC.
// We base the primitives on the C/C++ atomics and create a mimimal wrapper for MSVC in C compilation mode.
// We base the primitives on the C/C++ atomics and create a minimal wrapper for MSVC in C compilation mode.
// This is why we try to use only `uintptr_t` and `<type>*` as atomic types.
// To gain better insight in the range of used atomics, we use explicitly named memory order operations
// instead of passing the memory order as a parameter.
@@ -20,29 +31,33 @@ terms of the MIT license. A copy of the license can be found in the file
#if defined(__cplusplus)
// Use C++ atomics
#include <atomic>
#define _Atomic(tp) std::atomic<tp>
#define mi_atomic(name) std::atomic_##name
#define mi_memory_order(name) std::memory_order_##name
#if !defined(ATOMIC_VAR_INIT) || (__cplusplus >= 202002L) // c++20, see issue #571
#define MI_ATOMIC_VAR_INIT(x) x
#define _Atomic(tp) std::atomic<tp>
#define mi_atomic(name) std::atomic_##name
#define mi_memory_order(name) std::memory_order_##name
#if (__cplusplus >= 202002L) // c++20, see issue #571
#define MI_ATOMIC_VAR_INIT(x) x
#elif !defined(ATOMIC_VAR_INIT)
#define MI_ATOMIC_VAR_INIT(x) x
#else
#define MI_ATOMIC_VAR_INIT(x) ATOMIC_VAR_INIT(x)
#define MI_ATOMIC_VAR_INIT(x) ATOMIC_VAR_INIT(x)
#endif
#elif defined(_MSC_VER)
// Use MSVC C wrapper for C11 atomics
#define _Atomic(tp) tp
#define MI_ATOMIC_VAR_INIT(x) x
#define mi_atomic(name) mi_atomic_##name
#define mi_memory_order(name) mi_memory_order_##name
#define _Atomic(tp) tp
#define MI_ATOMIC_VAR_INIT(x) x
#define mi_atomic(name) mi_atomic_##name
#define mi_memory_order(name) mi_memory_order_##name
#else
// Use C11 atomics
#include <stdatomic.h>
#define mi_atomic(name) atomic_##name
#define mi_memory_order(name) memory_order_##name
#if !defined(ATOMIC_VAR_INIT) || (__STDC_VERSION__ >= 201710L) // c17, see issue #735
#define MI_ATOMIC_VAR_INIT(x) x
#define mi_atomic(name) atomic_##name
#define mi_memory_order(name) memory_order_##name
#if (__STDC_VERSION__ >= 201710L) // c17, see issue #735
#define MI_ATOMIC_VAR_INIT(x) x
#elif !defined(ATOMIC_VAR_INIT)
#define MI_ATOMIC_VAR_INIT(x) x
#else
#define MI_ATOMIC_VAR_INIT(x) ATOMIC_VAR_INIT(x)
#define MI_ATOMIC_VAR_INIT(x) ATOMIC_VAR_INIT(x)
#endif
#endif
@@ -57,6 +72,7 @@ terms of the MIT license. A copy of the license can be found in the file
#define mi_atomic_load_relaxed(p) mi_atomic(load_explicit)(p,mi_memory_order(relaxed))
#define mi_atomic_store_release(p,x) mi_atomic(store_explicit)(p,x,mi_memory_order(release))
#define mi_atomic_store_relaxed(p,x) mi_atomic(store_explicit)(p,x,mi_memory_order(relaxed))
#define mi_atomic_exchange_relaxed(p,x) mi_atomic(exchange_explicit)(p,x,mi_memory_order(relaxed))
#define mi_atomic_exchange_release(p,x) mi_atomic(exchange_explicit)(p,x,mi_memory_order(release))
#define mi_atomic_exchange_acq_rel(p,x) mi_atomic(exchange_explicit)(p,x,mi_memory_order(acq_rel))
#define mi_atomic_cas_weak_release(p,exp,des) mi_atomic_cas_weak(p,exp,des,mi_memory_order(release),mi_memory_order(relaxed))
@@ -95,6 +111,8 @@ static inline intptr_t mi_atomic_subi(_Atomic(intptr_t)*p, intptr_t sub);
#define mi_atomic_cas_ptr_weak_release(tp,p,exp,des) mi_atomic_cas_weak_release(p,exp,(tp*)des)
#define mi_atomic_cas_ptr_weak_acq_rel(tp,p,exp,des) mi_atomic_cas_weak_acq_rel(p,exp,(tp*)des)
#define mi_atomic_cas_ptr_strong_release(tp,p,exp,des) mi_atomic_cas_strong_release(p,exp,(tp*)des)
#define mi_atomic_cas_ptr_strong_acq_rel(tp,p,exp,des) mi_atomic_cas_strong_acq_rel(p,exp,(tp*)des)
#define mi_atomic_exchange_ptr_relaxed(tp,p,x) mi_atomic_exchange_relaxed(p,(tp*)x)
#define mi_atomic_exchange_ptr_release(tp,p,x) mi_atomic_exchange_release(p,(tp*)x)
#define mi_atomic_exchange_ptr_acq_rel(tp,p,x) mi_atomic_exchange_acq_rel(p,(tp*)x)
#else
@@ -103,6 +121,8 @@ static inline intptr_t mi_atomic_subi(_Atomic(intptr_t)*p, intptr_t sub);
#define mi_atomic_cas_ptr_weak_release(tp,p,exp,des) mi_atomic_cas_weak_release(p,exp,des)
#define mi_atomic_cas_ptr_weak_acq_rel(tp,p,exp,des) mi_atomic_cas_weak_acq_rel(p,exp,des)
#define mi_atomic_cas_ptr_strong_release(tp,p,exp,des) mi_atomic_cas_strong_release(p,exp,des)
#define mi_atomic_cas_ptr_strong_acq_rel(tp,p,exp,des) mi_atomic_cas_strong_acq_rel(p,exp,des)
#define mi_atomic_exchange_ptr_relaxed(tp,p,x) mi_atomic_exchange_relaxed(p,x)
#define mi_atomic_exchange_ptr_release(tp,p,x) mi_atomic_exchange_release(p,x)
#define mi_atomic_exchange_ptr_acq_rel(tp,p,x) mi_atomic_exchange_acq_rel(p,x)
#endif
@@ -111,6 +131,12 @@ static inline intptr_t mi_atomic_subi(_Atomic(intptr_t)*p, intptr_t sub);
static inline int64_t mi_atomic_addi64_relaxed(volatile int64_t* p, int64_t add) {
return mi_atomic(fetch_add_explicit)((_Atomic(int64_t)*)p, add, mi_memory_order(relaxed));
}
static inline void mi_atomic_void_addi64_relaxed(volatile int64_t* p, const volatile int64_t* padd) {
const int64_t add = mi_atomic_load_relaxed((_Atomic(int64_t)*)padd);
if (add != 0) {
mi_atomic(fetch_add_explicit)((_Atomic(int64_t)*)p, add, mi_memory_order(relaxed));
}
}
static inline void mi_atomic_maxi64_relaxed(volatile int64_t* p, int64_t x) {
int64_t current = mi_atomic_load_relaxed((_Atomic(int64_t)*)p);
while (current < x && !mi_atomic_cas_weak_release((_Atomic(int64_t)*)p, &current, x)) { /* nothing */ };
@@ -128,9 +154,7 @@ static inline void mi_atomic_maxi64_relaxed(volatile int64_t* p, int64_t x) {
#elif defined(_MSC_VER)
// MSVC C compilation wrapper that uses Interlocked operations to model C11 atomics.
#define WIN32_LEAN_AND_MEAN
#include <windows.h>
// Legacy MSVC plain C compilation wrapper that uses Interlocked operations to model C11 atomics.
#include <intrin.h>
#ifdef _WIN64
typedef LONG64 msc_intptr_t;
@@ -195,7 +219,7 @@ static inline uintptr_t mi_atomic_load_explicit(_Atomic(uintptr_t) const* p, mi_
#else
uintptr_t x = *p;
if (mo > mi_memory_order_relaxed) {
while (!mi_atomic_compare_exchange_weak_explicit(p, &x, x, mo, mi_memory_order_relaxed)) { /* nothing */ };
while (!mi_atomic_compare_exchange_weak_explicit((_Atomic(uintptr_t)*)p, &x, x, mo, mi_memory_order_relaxed)) { /* nothing */ };
}
return x;
#endif
@@ -244,6 +268,13 @@ static inline int64_t mi_atomic_addi64_relaxed(volatile _Atomic(int64_t)*p, int6
return current;
#endif
}
static inline void mi_atomic_void_addi64_relaxed(volatile int64_t* p, const volatile int64_t* padd) {
const int64_t add = *padd;
if (add != 0) {
mi_atomic_addi64_relaxed((volatile _Atomic(int64_t)*)p, add);
}
}
static inline void mi_atomic_maxi64_relaxed(volatile _Atomic(int64_t)*p, int64_t x) {
int64_t current;
do {
@@ -274,6 +305,8 @@ static inline bool mi_atomic_casi64_strong_acq_rel(volatile _Atomic(int64_t*)p,
#define mi_atomic_cas_ptr_weak_release(tp,p,exp,des) mi_atomic_cas_weak_release((_Atomic(uintptr_t)*)(p),(uintptr_t*)exp,(uintptr_t)des)
#define mi_atomic_cas_ptr_weak_acq_rel(tp,p,exp,des) mi_atomic_cas_weak_acq_rel((_Atomic(uintptr_t)*)(p),(uintptr_t*)exp,(uintptr_t)des)
#define mi_atomic_cas_ptr_strong_release(tp,p,exp,des) mi_atomic_cas_strong_release((_Atomic(uintptr_t)*)(p),(uintptr_t*)exp,(uintptr_t)des)
#define mi_atomic_cas_ptr_strong_acq_rel(tp,p,exp,des) mi_atomic_cas_strong_acq_rel((_Atomic(uintptr_t)*)(p),(uintptr_t*)exp,(uintptr_t)des)
#define mi_atomic_exchange_ptr_relaxed(tp,p,x) (tp*)mi_atomic_exchange_relaxed((_Atomic(uintptr_t)*)(p),(uintptr_t)x)
#define mi_atomic_exchange_ptr_release(tp,p,x) (tp*)mi_atomic_exchange_release((_Atomic(uintptr_t)*)(p),(uintptr_t)x)
#define mi_atomic_exchange_ptr_acq_rel(tp,p,x) (tp*)mi_atomic_exchange_acq_rel((_Atomic(uintptr_t)*)(p),(uintptr_t)x)
@@ -296,11 +329,16 @@ static inline intptr_t mi_atomic_subi(_Atomic(intptr_t)*p, intptr_t sub) {
return (intptr_t)mi_atomic_addi(p, -sub);
}
// ----------------------------------------------------------------------
// Once and Guard
// ----------------------------------------------------------------------
typedef _Atomic(uintptr_t) mi_atomic_once_t;
// Returns true only on the first invocation
static inline bool mi_atomic_once( mi_atomic_once_t* once ) {
if (mi_atomic_load_relaxed(once) != 0) return false; // quick test
if (mi_atomic_load_relaxed(once) != 0) return false; // quick test
uintptr_t expected = 0;
return mi_atomic_cas_strong_acq_rel(once, &expected, (uintptr_t)1); // try to set to 1
}
@@ -316,15 +354,16 @@ typedef _Atomic(uintptr_t) mi_atomic_guard_t;
// ----------------------------------------------------------------------
// Yield
// ----------------------------------------------------------------------
#if defined(__cplusplus)
#include <thread>
static inline void mi_atomic_yield(void) {
std::this_thread::yield();
}
#elif defined(_WIN32)
#define WIN32_LEAN_AND_MEAN
#include <windows.h>
static inline void mi_atomic_yield(void) {
YieldProcessor();
}
@@ -334,8 +373,9 @@ static inline void mi_atomic_yield(void) {
_mm_pause();
}
#elif (defined(__GNUC__) || defined(__clang__)) && \
(defined(__x86_64__) || defined(__i386__) || defined(__arm__) || defined(__armel__) || defined(__ARMEL__) || \
defined(__aarch64__) || defined(__powerpc__) || defined(__ppc__) || defined(__PPC__)) || defined(__POWERPC__)
(defined(__x86_64__) || defined(__i386__) || \
defined(__aarch64__) || defined(__arm__) || \
defined(__powerpc__) || defined(__ppc__) || defined(__PPC__) || defined(__POWERPC__))
#if defined(__x86_64__) || defined(__i386__)
static inline void mi_atomic_yield(void) {
__asm__ volatile ("pause" ::: "memory");
@@ -344,10 +384,16 @@ static inline void mi_atomic_yield(void) {
static inline void mi_atomic_yield(void) {
__asm__ volatile("wfe");
}
#elif (defined(__arm__) && __ARM_ARCH__ >= 7)
#elif defined(__arm__)
#if __ARM_ARCH >= 7
static inline void mi_atomic_yield(void) {
__asm__ volatile("yield" ::: "memory");
}
#else
static inline void mi_atomic_yield(void) {
__asm__ volatile ("nop" ::: "memory");
}
#endif
#elif defined(__powerpc__) || defined(__ppc__) || defined(__PPC__) || defined(__POWERPC__)
#ifdef __APPLE__
static inline void mi_atomic_yield(void) {
@@ -358,10 +404,6 @@ static inline void mi_atomic_yield(void) {
__asm__ __volatile__ ("or 27,27,27" ::: "memory");
}
#endif
#elif defined(__armel__) || defined(__ARMEL__)
static inline void mi_atomic_yield(void) {
__asm__ volatile ("nop" ::: "memory");
}
#endif
#elif defined(__sun)
// Fallback for other archs
@@ -382,4 +424,134 @@ static inline void mi_atomic_yield(void) {
#endif
// ----------------------------------------------------------------------
// Locks
// These do not have to be recursive and should be light-weight
// in-process only locks. Only used for reserving arena's and to
// maintain the abandoned list.
// ----------------------------------------------------------------------
#if _MSC_VER
#pragma warning(disable:26110) // unlock with holding lock
#endif
#define mi_lock(lock) for(bool _go = (mi_lock_acquire(lock),true); _go; (mi_lock_release(lock), _go=false) )
#if defined(_WIN32)
#if 1
#define mi_lock_t SRWLOCK // slim reader-writer lock
static inline bool mi_lock_try_acquire(mi_lock_t* lock) {
return TryAcquireSRWLockExclusive(lock);
}
static inline void mi_lock_acquire(mi_lock_t* lock) {
AcquireSRWLockExclusive(lock);
}
static inline void mi_lock_release(mi_lock_t* lock) {
ReleaseSRWLockExclusive(lock);
}
static inline void mi_lock_init(mi_lock_t* lock) {
InitializeSRWLock(lock);
}
static inline void mi_lock_done(mi_lock_t* lock) {
(void)(lock);
}
#else
#define mi_lock_t CRITICAL_SECTION
static inline bool mi_lock_try_acquire(mi_lock_t* lock) {
return TryEnterCriticalSection(lock);
}
static inline void mi_lock_acquire(mi_lock_t* lock) {
EnterCriticalSection(lock);
}
static inline void mi_lock_release(mi_lock_t* lock) {
LeaveCriticalSection(lock);
}
static inline void mi_lock_init(mi_lock_t* lock) {
InitializeCriticalSection(lock);
}
static inline void mi_lock_done(mi_lock_t* lock) {
DeleteCriticalSection(lock);
}
#endif
#elif defined(MI_USE_PTHREADS)
void _mi_error_message(int err, const char* fmt, ...);
#define mi_lock_t pthread_mutex_t
static inline bool mi_lock_try_acquire(mi_lock_t* lock) {
return (pthread_mutex_trylock(lock) == 0);
}
static inline void mi_lock_acquire(mi_lock_t* lock) {
const int err = pthread_mutex_lock(lock);
if (err != 0) {
_mi_error_message(err, "internal error: lock cannot be acquired\n");
}
}
static inline void mi_lock_release(mi_lock_t* lock) {
pthread_mutex_unlock(lock);
}
static inline void mi_lock_init(mi_lock_t* lock) {
pthread_mutex_init(lock, NULL);
}
static inline void mi_lock_done(mi_lock_t* lock) {
pthread_mutex_destroy(lock);
}
#elif defined(__cplusplus)
#include <mutex>
#define mi_lock_t std::mutex
static inline bool mi_lock_try_acquire(mi_lock_t* lock) {
return lock->try_lock();
}
static inline void mi_lock_acquire(mi_lock_t* lock) {
lock->lock();
}
static inline void mi_lock_release(mi_lock_t* lock) {
lock->unlock();
}
static inline void mi_lock_init(mi_lock_t* lock) {
(void)(lock);
}
static inline void mi_lock_done(mi_lock_t* lock) {
(void)(lock);
}
#else
// fall back to poor man's locks.
// this should only be the case in a single-threaded environment (like __wasi__)
#define mi_lock_t _Atomic(uintptr_t)
static inline bool mi_lock_try_acquire(mi_lock_t* lock) {
uintptr_t expected = 0;
return mi_atomic_cas_strong_acq_rel(lock, &expected, (uintptr_t)1);
}
static inline void mi_lock_acquire(mi_lock_t* lock) {
for (int i = 0; i < 1000; i++) { // for at most 1000 tries?
if (mi_lock_try_acquire(lock)) return;
mi_atomic_yield();
}
}
static inline void mi_lock_release(mi_lock_t* lock) {
mi_atomic_store_release(lock, (uintptr_t)0);
}
static inline void mi_lock_init(mi_lock_t* lock) {
mi_lock_release(lock);
}
static inline void mi_lock_done(mi_lock_t* lock) {
(void)(lock);
}
#endif
#endif // __MIMALLOC_ATOMIC_H

742
mimalloc/mimalloc-sys/mimalloc/include/mimalloc/internal.h
View File

@@ -8,14 +8,18 @@ terms of the MIT license. A copy of the license can be found in the file
#ifndef MIMALLOC_INTERNAL_H
#define MIMALLOC_INTERNAL_H
// --------------------------------------------------------------------------
// This file contains the interal API's of mimalloc and various utility
// This file contains the internal API's of mimalloc and various utility
// functions and macros.
// --------------------------------------------------------------------------
#include "mimalloc/types.h"
#include "mimalloc/track.h"
#include "types.h"
#include "track.h"
// --------------------------------------------------------------------------
// Compiler defines
// --------------------------------------------------------------------------
#if (MI_DEBUG>0)
#define mi_trace_message(...) _mi_trace_message(__VA_ARGS__)
@@ -23,193 +27,48 @@ terms of the MIT license. A copy of the license can be found in the file
#define mi_trace_message(...)
#endif
#define MI_CACHE_LINE 64
#define mi_decl_cache_align mi_decl_align(64)
#if defined(_MSC_VER)
#pragma warning(disable:4127) // suppress constant conditional warning (due to MI_SECURE paths)
#pragma warning(disable:26812) // unscoped enum warning
#define mi_decl_noinline __declspec(noinline)
#define mi_decl_thread __declspec(thread)
#define mi_decl_cache_align __declspec(align(MI_CACHE_LINE))
#define mi_decl_align(a) __declspec(align(a))
#define mi_decl_noreturn __declspec(noreturn)
#define mi_decl_weak
#define mi_decl_hidden
#define mi_decl_cold
#elif (defined(__GNUC__) && (__GNUC__ >= 3)) || defined(__clang__) // includes clang and icc
#define mi_decl_noinline __attribute__((noinline))
#define mi_decl_thread __thread
#define mi_decl_cache_align __attribute__((aligned(MI_CACHE_LINE)))
#define mi_decl_align(a) __attribute__((aligned(a)))
#define mi_decl_noreturn __attribute__((noreturn))
#define mi_decl_weak __attribute__((weak))
#define mi_decl_hidden __attribute__((visibility("hidden")))
#if (__GNUC__ >= 4) || defined(__clang__)
#define mi_decl_cold __attribute__((cold))
#else
#define mi_decl_cold
#endif
#elif __cplusplus >= 201103L // c++11
#define mi_decl_noinline
#define mi_decl_thread thread_local
#define mi_decl_align(a) alignas(a)
#define mi_decl_noreturn [[noreturn]]
#define mi_decl_weak
#define mi_decl_hidden
#define mi_decl_cold
#else
#define mi_decl_noinline
#define mi_decl_thread __thread // hope for the best :-)
#define mi_decl_cache_align
#define mi_decl_align(a)
#define mi_decl_noreturn
#define mi_decl_weak
#define mi_decl_hidden
#define mi_decl_cold
#endif
#if defined(__EMSCRIPTEN__) && !defined(__wasi__)
#define __wasi__
#endif
#if defined(__cplusplus)
#define mi_decl_externc extern "C"
#else
#define mi_decl_externc
#endif
// pthreads
#if !defined(_WIN32) && !defined(__wasi__)
#define MI_USE_PTHREADS
#include <pthread.h>
#endif
// "options.c"
void _mi_fputs(mi_output_fun* out, void* arg, const char* prefix, const char* message);
void _mi_fprintf(mi_output_fun* out, void* arg, const char* fmt, ...);
void _mi_warning_message(const char* fmt, ...);
void _mi_verbose_message(const char* fmt, ...);
void _mi_trace_message(const char* fmt, ...);
void _mi_options_init(void);
void _mi_error_message(int err, const char* fmt, ...);
// random.c
void _mi_random_init(mi_random_ctx_t* ctx);
void _mi_random_init_weak(mi_random_ctx_t* ctx);
void _mi_random_reinit_if_weak(mi_random_ctx_t * ctx);
void _mi_random_split(mi_random_ctx_t* ctx, mi_random_ctx_t* new_ctx);
uintptr_t _mi_random_next(mi_random_ctx_t* ctx);
uintptr_t _mi_heap_random_next(mi_heap_t* heap);
uintptr_t _mi_os_random_weak(uintptr_t extra_seed);
static inline uintptr_t _mi_random_shuffle(uintptr_t x);
// init.c
extern mi_decl_cache_align mi_stats_t _mi_stats_main;
extern mi_decl_cache_align const mi_page_t _mi_page_empty;
bool _mi_is_main_thread(void);
size_t _mi_current_thread_count(void);
bool _mi_preloading(void); // true while the C runtime is not initialized yet
mi_threadid_t _mi_thread_id(void) mi_attr_noexcept;
mi_heap_t* _mi_heap_main_get(void); // statically allocated main backing heap
void _mi_thread_done(mi_heap_t* heap);
void _mi_thread_data_collect(void);
// os.c
void _mi_os_init(void); // called from process init
void* _mi_os_alloc(size_t size, mi_memid_t* memid, mi_stats_t* stats);
void _mi_os_free(void* p, size_t size, mi_memid_t memid, mi_stats_t* stats);
void _mi_os_free_ex(void* p, size_t size, bool still_committed, mi_memid_t memid, mi_stats_t* stats);
size_t _mi_os_page_size(void);
size_t _mi_os_good_alloc_size(size_t size);
bool _mi_os_has_overcommit(void);
bool _mi_os_has_virtual_reserve(void);
bool _mi_os_purge(void* p, size_t size, mi_stats_t* stats);
bool _mi_os_reset(void* addr, size_t size, mi_stats_t* tld_stats);
bool _mi_os_commit(void* p, size_t size, bool* is_zero, mi_stats_t* stats);
bool _mi_os_decommit(void* addr, size_t size, mi_stats_t* stats);
bool _mi_os_protect(void* addr, size_t size);
bool _mi_os_unprotect(void* addr, size_t size);
bool _mi_os_purge(void* p, size_t size, mi_stats_t* stats);
bool _mi_os_purge_ex(void* p, size_t size, bool allow_reset, mi_stats_t* stats);
void* _mi_os_alloc_aligned(size_t size, size_t alignment, bool commit, bool allow_large, mi_memid_t* memid, mi_stats_t* stats);
void* _mi_os_alloc_aligned_at_offset(size_t size, size_t alignment, size_t align_offset, bool commit, bool allow_large, mi_memid_t* memid, mi_stats_t* tld_stats);
void* _mi_os_get_aligned_hint(size_t try_alignment, size_t size);
bool _mi_os_use_large_page(size_t size, size_t alignment);
size_t _mi_os_large_page_size(void);
void* _mi_os_alloc_huge_os_pages(size_t pages, int numa_node, mi_msecs_t max_secs, size_t* pages_reserved, size_t* psize, mi_memid_t* memid);
// arena.c
mi_arena_id_t _mi_arena_id_none(void);
void _mi_arena_free(void* p, size_t size, size_t still_committed_size, mi_memid_t memid, mi_stats_t* stats);
void* _mi_arena_alloc(size_t size, bool commit, bool allow_large, mi_arena_id_t req_arena_id, mi_memid_t* memid, mi_os_tld_t* tld);
void* _mi_arena_alloc_aligned(size_t size, size_t alignment, size_t align_offset, bool commit, bool allow_large, mi_arena_id_t req_arena_id, mi_memid_t* memid, mi_os_tld_t* tld);
bool _mi_arena_memid_is_suitable(mi_memid_t memid, mi_arena_id_t request_arena_id);
bool _mi_arena_contains(const void* p);
void _mi_arena_collect(bool force_purge, mi_stats_t* stats);
void _mi_arena_unsafe_destroy_all(mi_stats_t* stats);
// "segment-map.c"
void _mi_segment_map_allocated_at(const mi_segment_t* segment);
void _mi_segment_map_freed_at(const mi_segment_t* segment);
// "segment.c"
mi_page_t* _mi_segment_page_alloc(mi_heap_t* heap, size_t block_size, size_t page_alignment, mi_segments_tld_t* tld, mi_os_tld_t* os_tld);
void _mi_segment_page_free(mi_page_t* page, bool force, mi_segments_tld_t* tld);
void _mi_segment_page_abandon(mi_page_t* page, mi_segments_tld_t* tld);
bool _mi_segment_try_reclaim_abandoned( mi_heap_t* heap, bool try_all, mi_segments_tld_t* tld);
void _mi_segment_thread_collect(mi_segments_tld_t* tld);
#if MI_HUGE_PAGE_ABANDON
void _mi_segment_huge_page_free(mi_segment_t* segment, mi_page_t* page, mi_block_t* block);
#else
void _mi_segment_huge_page_reset(mi_segment_t* segment, mi_page_t* page, mi_block_t* block);
#endif
uint8_t* _mi_segment_page_start(const mi_segment_t* segment, const mi_page_t* page, size_t* page_size); // page start for any page
void _mi_abandoned_reclaim_all(mi_heap_t* heap, mi_segments_tld_t* tld);
void _mi_abandoned_await_readers(void);
void _mi_abandoned_collect(mi_heap_t* heap, bool force, mi_segments_tld_t* tld);
// "page.c"
void* _mi_malloc_generic(mi_heap_t* heap, size_t size, bool zero, size_t huge_alignment) mi_attr_noexcept mi_attr_malloc;
void _mi_page_retire(mi_page_t* page) mi_attr_noexcept; // free the page if there are no other pages with many free blocks
void _mi_page_unfull(mi_page_t* page);
void _mi_page_free(mi_page_t* page, mi_page_queue_t* pq, bool force); // free the page
void _mi_page_abandon(mi_page_t* page, mi_page_queue_t* pq); // abandon the page, to be picked up by another thread...
void _mi_heap_delayed_free_all(mi_heap_t* heap);
bool _mi_heap_delayed_free_partial(mi_heap_t* heap);
void _mi_heap_collect_retired(mi_heap_t* heap, bool force);
void _mi_page_use_delayed_free(mi_page_t* page, mi_delayed_t delay, bool override_never);
bool _mi_page_try_use_delayed_free(mi_page_t* page, mi_delayed_t delay, bool override_never);
size_t _mi_page_queue_append(mi_heap_t* heap, mi_page_queue_t* pq, mi_page_queue_t* append);
void _mi_deferred_free(mi_heap_t* heap, bool force);
void _mi_page_free_collect(mi_page_t* page,bool force);
void _mi_page_reclaim(mi_heap_t* heap, mi_page_t* page); // callback from segments
size_t _mi_bin_size(uint8_t bin); // for stats
uint8_t _mi_bin(size_t size); // for stats
// "heap.c"
void _mi_heap_destroy_pages(mi_heap_t* heap);
void _mi_heap_collect_abandon(mi_heap_t* heap);
void _mi_heap_set_default_direct(mi_heap_t* heap);
bool _mi_heap_memid_is_suitable(mi_heap_t* heap, mi_memid_t memid);
void _mi_heap_unsafe_destroy_all(void);
// "stats.c"
void _mi_stats_done(mi_stats_t* stats);
mi_msecs_t _mi_clock_now(void);
mi_msecs_t _mi_clock_end(mi_msecs_t start);
mi_msecs_t _mi_clock_start(void);
// "alloc.c"
void* _mi_page_malloc(mi_heap_t* heap, mi_page_t* page, size_t size, bool zero) mi_attr_noexcept; // called from `_mi_malloc_generic`
void* _mi_heap_malloc_zero(mi_heap_t* heap, size_t size, bool zero) mi_attr_noexcept;
void* _mi_heap_malloc_zero_ex(mi_heap_t* heap, size_t size, bool zero, size_t huge_alignment) mi_attr_noexcept; // called from `_mi_heap_malloc_aligned`
void* _mi_heap_realloc_zero(mi_heap_t* heap, void* p, size_t newsize, bool zero) mi_attr_noexcept;
mi_block_t* _mi_page_ptr_unalign(const mi_segment_t* segment, const mi_page_t* page, const void* p);
bool _mi_free_delayed_block(mi_block_t* block);
void _mi_free_generic(const mi_segment_t* segment, mi_page_t* page, bool is_local, void* p) mi_attr_noexcept; // for runtime integration
void _mi_padding_shrink(const mi_page_t* page, const mi_block_t* block, const size_t min_size);
// option.c, c primitives
char _mi_toupper(char c);
int _mi_strnicmp(const char* s, const char* t, size_t n);
void _mi_strlcpy(char* dest, const char* src, size_t dest_size);
void _mi_strlcat(char* dest, const char* src, size_t dest_size);
size_t _mi_strlen(const char* s);
size_t _mi_strnlen(const char* s, size_t max_len);
#if MI_DEBUG>1
bool _mi_page_is_valid(mi_page_t* page);
#endif
// ------------------------------------------------------
// Branches
// ------------------------------------------------------
#if defined(__GNUC__) || defined(__clang__)
#define mi_unlikely(x) (__builtin_expect(!!(x),false))
#define mi_likely(x) (__builtin_expect(!!(x),true))
@@ -222,7 +81,219 @@ bool _mi_page_is_valid(mi_page_t* page);
#endif
#ifndef __has_builtin
#define __has_builtin(x) 0
#define __has_builtin(x) 0
#endif
#if defined(__cplusplus)
#define mi_decl_externc extern "C"
#else
#define mi_decl_externc
#endif
#if defined(__EMSCRIPTEN__) && !defined(__wasi__)
#define __wasi__
#endif
// --------------------------------------------------------------------------
// Internal functions
// --------------------------------------------------------------------------
// "libc.c"
#include <stdarg.h>
int _mi_vsnprintf(char* buf, size_t bufsize, const char* fmt, va_list args);
int _mi_snprintf(char* buf, size_t buflen, const char* fmt, ...);
char _mi_toupper(char c);
int _mi_strnicmp(const char* s, const char* t, size_t n);
void _mi_strlcpy(char* dest, const char* src, size_t dest_size);
void _mi_strlcat(char* dest, const char* src, size_t dest_size);
size_t _mi_strlen(const char* s);
size_t _mi_strnlen(const char* s, size_t max_len);
bool _mi_getenv(const char* name, char* result, size_t result_size);
// "options.c"
void _mi_fputs(mi_output_fun* out, void* arg, const char* prefix, const char* message);
void _mi_fprintf(mi_output_fun* out, void* arg, const char* fmt, ...);
void _mi_message(const char* fmt, ...);
void _mi_warning_message(const char* fmt, ...);
void _mi_verbose_message(const char* fmt, ...);
void _mi_trace_message(const char* fmt, ...);
void _mi_options_init(void);
long _mi_option_get_fast(mi_option_t option);
void _mi_error_message(int err, const char* fmt, ...);
// random.c
void _mi_random_init(mi_random_ctx_t* ctx);
void _mi_random_init_weak(mi_random_ctx_t* ctx);
void _mi_random_reinit_if_weak(mi_random_ctx_t * ctx);
void _mi_random_split(mi_random_ctx_t* ctx, mi_random_ctx_t* new_ctx);
uintptr_t _mi_random_next(mi_random_ctx_t* ctx);
uintptr_t _mi_heap_random_next(mi_heap_t* heap);
uintptr_t _mi_os_random_weak(uintptr_t extra_seed);
static inline uintptr_t _mi_random_shuffle(uintptr_t x);
// init.c
extern mi_decl_hidden mi_decl_cache_align mi_stats_t _mi_stats_main;
extern mi_decl_hidden mi_decl_cache_align const mi_page_t _mi_page_empty;
void _mi_auto_process_init(void);
void mi_cdecl _mi_auto_process_done(void) mi_attr_noexcept;
bool _mi_is_redirected(void);
bool _mi_allocator_init(const char** message);
void _mi_allocator_done(void);
bool _mi_is_main_thread(void);
size_t _mi_current_thread_count(void);
bool _mi_preloading(void); // true while the C runtime is not initialized yet
void _mi_thread_done(mi_heap_t* heap);
void _mi_thread_data_collect(void);
void _mi_tld_init(mi_tld_t* tld, mi_heap_t* bheap);
mi_threadid_t _mi_thread_id(void) mi_attr_noexcept;
mi_heap_t* _mi_heap_main_get(void); // statically allocated main backing heap
mi_subproc_t* _mi_subproc_from_id(mi_subproc_id_t subproc_id);
void _mi_heap_guarded_init(mi_heap_t* heap);
// os.c
void _mi_os_init(void); // called from process init
void* _mi_os_alloc(size_t size, mi_memid_t* memid);
void* _mi_os_zalloc(size_t size, mi_memid_t* memid);
void _mi_os_free(void* p, size_t size, mi_memid_t memid);
void _mi_os_free_ex(void* p, size_t size, bool still_committed, mi_memid_t memid);
size_t _mi_os_page_size(void);
size_t _mi_os_good_alloc_size(size_t size);
bool _mi_os_has_overcommit(void);
bool _mi_os_has_virtual_reserve(void);
bool _mi_os_reset(void* addr, size_t size);
bool _mi_os_decommit(void* addr, size_t size);
bool _mi_os_unprotect(void* addr, size_t size);
bool _mi_os_purge(void* p, size_t size);
bool _mi_os_purge_ex(void* p, size_t size, bool allow_reset, size_t stat_size);
void _mi_os_reuse(void* p, size_t size);
mi_decl_nodiscard bool _mi_os_commit(void* p, size_t size, bool* is_zero);
mi_decl_nodiscard bool _mi_os_commit_ex(void* addr, size_t size, bool* is_zero, size_t stat_size);
bool _mi_os_protect(void* addr, size_t size);
void* _mi_os_alloc_aligned(size_t size, size_t alignment, bool commit, bool allow_large, mi_memid_t* memid);
void* _mi_os_alloc_aligned_at_offset(size_t size, size_t alignment, size_t align_offset, bool commit, bool allow_large, mi_memid_t* memid);
void* _mi_os_get_aligned_hint(size_t try_alignment, size_t size);
bool _mi_os_canuse_large_page(size_t size, size_t alignment);
size_t _mi_os_large_page_size(void);
void* _mi_os_alloc_huge_os_pages(size_t pages, int numa_node, mi_msecs_t max_secs, size_t* pages_reserved, size_t* psize, mi_memid_t* memid);
int _mi_os_numa_node_count(void);
int _mi_os_numa_node(void);
// arena.c
mi_arena_id_t _mi_arena_id_none(void);
void _mi_arena_free(void* p, size_t size, size_t still_committed_size, mi_memid_t memid);
void* _mi_arena_alloc(size_t size, bool commit, bool allow_large, mi_arena_id_t req_arena_id, mi_memid_t* memid);
void* _mi_arena_alloc_aligned(size_t size, size_t alignment, size_t align_offset, bool commit, bool allow_large, mi_arena_id_t req_arena_id, mi_memid_t* memid);
bool _mi_arena_memid_is_suitable(mi_memid_t memid, mi_arena_id_t request_arena_id);
bool _mi_arena_contains(const void* p);
void _mi_arenas_collect(bool force_purge);
void _mi_arena_unsafe_destroy_all(void);
bool _mi_arena_segment_clear_abandoned(mi_segment_t* segment);
void _mi_arena_segment_mark_abandoned(mi_segment_t* segment);
void* _mi_arena_meta_zalloc(size_t size, mi_memid_t* memid);
void _mi_arena_meta_free(void* p, mi_memid_t memid, size_t size);
typedef struct mi_arena_field_cursor_s { // abstract struct
size_t os_list_count; // max entries to visit in the OS abandoned list
size_t start; // start arena idx (may need to be wrapped)
size_t end; // end arena idx (exclusive, may need to be wrapped)
size_t bitmap_idx; // current bit idx for an arena
mi_subproc_t* subproc; // only visit blocks in this sub-process
bool visit_all; // ensure all abandoned blocks are seen (blocking)
bool hold_visit_lock; // if the subproc->abandoned_os_visit_lock is held
} mi_arena_field_cursor_t;
void _mi_arena_field_cursor_init(mi_heap_t* heap, mi_subproc_t* subproc, bool visit_all, mi_arena_field_cursor_t* current);
mi_segment_t* _mi_arena_segment_clear_abandoned_next(mi_arena_field_cursor_t* previous);
void _mi_arena_field_cursor_done(mi_arena_field_cursor_t* current);
// "segment-map.c"
void _mi_segment_map_allocated_at(const mi_segment_t* segment);
void _mi_segment_map_freed_at(const mi_segment_t* segment);
void _mi_segment_map_unsafe_destroy(void);
// "segment.c"
mi_page_t* _mi_segment_page_alloc(mi_heap_t* heap, size_t block_size, size_t page_alignment, mi_segments_tld_t* tld);
void _mi_segment_page_free(mi_page_t* page, bool force, mi_segments_tld_t* tld);
void _mi_segment_page_abandon(mi_page_t* page, mi_segments_tld_t* tld);
bool _mi_segment_try_reclaim_abandoned( mi_heap_t* heap, bool try_all, mi_segments_tld_t* tld);
void _mi_segment_collect(mi_segment_t* segment, bool force);
#if MI_HUGE_PAGE_ABANDON
void _mi_segment_huge_page_free(mi_segment_t* segment, mi_page_t* page, mi_block_t* block);
#else
void _mi_segment_huge_page_reset(mi_segment_t* segment, mi_page_t* page, mi_block_t* block);
#endif
uint8_t* _mi_segment_page_start(const mi_segment_t* segment, const mi_page_t* page, size_t* page_size); // page start for any page
void _mi_abandoned_reclaim_all(mi_heap_t* heap, mi_segments_tld_t* tld);
void _mi_abandoned_collect(mi_heap_t* heap, bool force, mi_segments_tld_t* tld);
bool _mi_segment_attempt_reclaim(mi_heap_t* heap, mi_segment_t* segment);
bool _mi_segment_visit_blocks(mi_segment_t* segment, int heap_tag, bool visit_blocks, mi_block_visit_fun* visitor, void* arg);
// "page.c"
void* _mi_malloc_generic(mi_heap_t* heap, size_t size, bool zero, size_t huge_alignment, size_t* usable) mi_attr_noexcept mi_attr_malloc;
void _mi_page_retire(mi_page_t* page) mi_attr_noexcept; // free the page if there are no other pages with many free blocks
void _mi_page_unfull(mi_page_t* page);
void _mi_page_free(mi_page_t* page, mi_page_queue_t* pq, bool force); // free the page
void _mi_page_abandon(mi_page_t* page, mi_page_queue_t* pq); // abandon the page, to be picked up by another thread...
void _mi_page_force_abandon(mi_page_t* page);
void _mi_heap_delayed_free_all(mi_heap_t* heap);
bool _mi_heap_delayed_free_partial(mi_heap_t* heap);
void _mi_heap_collect_retired(mi_heap_t* heap, bool force);
void _mi_page_use_delayed_free(mi_page_t* page, mi_delayed_t delay, bool override_never);
bool _mi_page_try_use_delayed_free(mi_page_t* page, mi_delayed_t delay, bool override_never);
size_t _mi_page_queue_append(mi_heap_t* heap, mi_page_queue_t* pq, mi_page_queue_t* append);
void _mi_deferred_free(mi_heap_t* heap, bool force);
void _mi_page_free_collect(mi_page_t* page,bool force);
void _mi_page_reclaim(mi_heap_t* heap, mi_page_t* page); // callback from segments
size_t _mi_page_stats_bin(const mi_page_t* page); // for stats
size_t _mi_bin_size(size_t bin); // for stats
size_t _mi_bin(size_t size); // for stats
// "heap.c"
void _mi_heap_init(mi_heap_t* heap, mi_tld_t* tld, mi_arena_id_t arena_id, bool noreclaim, uint8_t tag);
void _mi_heap_destroy_pages(mi_heap_t* heap);
void _mi_heap_collect_abandon(mi_heap_t* heap);
void _mi_heap_set_default_direct(mi_heap_t* heap);
bool _mi_heap_memid_is_suitable(mi_heap_t* heap, mi_memid_t memid);
void _mi_heap_unsafe_destroy_all(mi_heap_t* heap);
mi_heap_t* _mi_heap_by_tag(mi_heap_t* heap, uint8_t tag);
void _mi_heap_area_init(mi_heap_area_t* area, mi_page_t* page);
bool _mi_heap_area_visit_blocks(const mi_heap_area_t* area, mi_page_t* page, mi_block_visit_fun* visitor, void* arg);
// "stats.c"
void _mi_stats_done(mi_stats_t* stats);
void _mi_stats_merge_thread(mi_tld_t* tld);
mi_msecs_t _mi_clock_now(void);
mi_msecs_t _mi_clock_end(mi_msecs_t start);
mi_msecs_t _mi_clock_start(void);
// "alloc.c"
void* _mi_page_malloc_zero(mi_heap_t* heap, mi_page_t* page, size_t size, bool zero, size_t* usable) mi_attr_noexcept; // called from `_mi_malloc_generic`
void* _mi_page_malloc(mi_heap_t* heap, mi_page_t* page, size_t size) mi_attr_noexcept; // called from `_mi_heap_malloc_aligned`
void* _mi_page_malloc_zeroed(mi_heap_t* heap, mi_page_t* page, size_t size) mi_attr_noexcept; // called from `_mi_heap_malloc_aligned`
void* _mi_heap_malloc_zero(mi_heap_t* heap, size_t size, bool zero) mi_attr_noexcept;
void* _mi_heap_malloc_zero_ex(mi_heap_t* heap, size_t size, bool zero, size_t huge_alignment, size_t* usable) mi_attr_noexcept; // called from `_mi_heap_malloc_aligned`
void* _mi_heap_realloc_zero(mi_heap_t* heap, void* p, size_t newsize, bool zero, size_t* usable_pre, size_t* usable_post) mi_attr_noexcept;
mi_block_t* _mi_page_ptr_unalign(const mi_page_t* page, const void* p);
bool _mi_free_delayed_block(mi_block_t* block);
void _mi_free_generic(mi_segment_t* segment, mi_page_t* page, bool is_local, void* p) mi_attr_noexcept; // for runtime integration
void _mi_padding_shrink(const mi_page_t* page, const mi_block_t* block, const size_t min_size);
#if MI_DEBUG>1
bool _mi_page_is_valid(mi_page_t* page);
#endif
@@ -250,6 +321,32 @@ bool _mi_page_is_valid(mi_page_t* page);
#endif
// ------------------------------------------------------
// Assertions
// ------------------------------------------------------
#if (MI_DEBUG)
// use our own assertion to print without memory allocation
mi_decl_noreturn mi_decl_cold void _mi_assert_fail(const char* assertion, const char* fname, unsigned int line, const char* func) mi_attr_noexcept;
#define mi_assert(expr) ((expr) ? (void)0 : _mi_assert_fail(#expr,__FILE__,__LINE__,__func__))
#else
#define mi_assert(x)
#endif
#if (MI_DEBUG>1)
#define mi_assert_internal mi_assert
#else
#define mi_assert_internal(x)
#endif
#if (MI_DEBUG>2)
#define mi_assert_expensive mi_assert
#else
#define mi_assert_expensive(x)
#endif
/* -----------------------------------------------------------
Inlined definitions
----------------------------------------------------------- */
@@ -267,7 +364,7 @@ bool _mi_page_is_valid(mi_page_t* page);
#define MI_INIT64(x) MI_INIT32(x),MI_INIT32(x)
#define MI_INIT128(x) MI_INIT64(x),MI_INIT64(x)
#define MI_INIT256(x) MI_INIT128(x),MI_INIT128(x)
#define MI_INIT74(x) MI_INIT64(x),MI_INIT8(x),x(),x()
#include <string.h>
// initialize a local variable to zero; use memset as compilers optimize constant sized memset's
@@ -308,12 +405,31 @@ static inline uintptr_t _mi_align_down(uintptr_t sz, size_t alignment) {
}
}
// Align a pointer upwards
static inline void* mi_align_up_ptr(void* p, size_t alignment) {
return (void*)_mi_align_up((uintptr_t)p, alignment);
}
// Align a pointer downwards
static inline void* mi_align_down_ptr(void* p, size_t alignment) {
return (void*)_mi_align_down((uintptr_t)p, alignment);
}
// Divide upwards: `s <= _mi_divide_up(s,d)*d < s+d`.
static inline uintptr_t _mi_divide_up(uintptr_t size, size_t divider) {
mi_assert_internal(divider != 0);
return (divider == 0 ? size : ((size + divider - 1) / divider));
}
// clamp an integer
static inline size_t _mi_clamp(size_t sz, size_t min, size_t max) {
if (sz < min) return min;
else if (sz > max) return max;
else return sz;
}
// Is memory zero initialized?
static inline bool mi_mem_is_zero(const void* p, size_t size) {
for (size_t i = 0; i < size; i++) {
@@ -347,10 +463,10 @@ static inline bool mi_mul_overflow(size_t count, size_t size, size_t* total) {
}
#else /* __builtin_umul_overflow is unavailable */
static inline bool mi_mul_overflow(size_t count, size_t size, size_t* total) {
#define MI_MUL_NO_OVERFLOW ((size_t)1 << (4*sizeof(size_t))) // sqrt(SIZE_MAX)
#define MI_MUL_COULD_OVERFLOW ((size_t)1 << (4*sizeof(size_t))) // sqrt(SIZE_MAX)
*total = count * size;
// note: gcc/clang optimize this to directly check the overflow flag
return ((size >= MI_MUL_NO_OVERFLOW || count >= MI_MUL_NO_OVERFLOW) && size > 0 && (SIZE_MAX / size) < count);
return ((size >= MI_MUL_COULD_OVERFLOW || count >= MI_MUL_COULD_OVERFLOW) && size > 0 && (SIZE_MAX / size) < count);
}
#endif
@@ -375,7 +491,7 @@ static inline bool mi_count_size_overflow(size_t count, size_t size, size_t* tot
Heap functions
------------------------------------------------------------------------------------------- */
extern const mi_heap_t _mi_heap_empty; // read-only empty heap, initial value of the thread local default heap
extern mi_decl_hidden const mi_heap_t _mi_heap_empty; // read-only empty heap, initial value of the thread local default heap
static inline bool mi_heap_is_backing(const mi_heap_t* heap) {
return (heap->tld->heap_backing == heap);
@@ -383,11 +499,11 @@ static inline bool mi_heap_is_backing(const mi_heap_t* heap) {
static inline bool mi_heap_is_initialized(mi_heap_t* heap) {
mi_assert_internal(heap != NULL);
return (heap != &_mi_heap_empty);
return (heap != NULL && heap != &_mi_heap_empty);
}
static inline uintptr_t _mi_ptr_cookie(const void* p) {
extern mi_heap_t _mi_heap_main;
extern mi_decl_hidden mi_heap_t _mi_heap_main;
mi_assert_internal(_mi_heap_main.cookie != 0);
return ((uintptr_t)p ^ _mi_heap_main.cookie);
}
@@ -407,9 +523,14 @@ static inline mi_page_t* _mi_heap_get_free_small_page(mi_heap_t* heap, size_t si
// Large aligned blocks may be aligned at N*MI_SEGMENT_SIZE (inside a huge segment > MI_SEGMENT_SIZE),
// and we need align "down" to the segment info which is `MI_SEGMENT_SIZE` bytes before it;
// therefore we align one byte before `p`.
// We check for NULL afterwards on 64-bit systems to improve codegen for `mi_free`.
static inline mi_segment_t* _mi_ptr_segment(const void* p) {
mi_assert_internal(p != NULL);
return (mi_segment_t*)(((uintptr_t)p - 1) & ~MI_SEGMENT_MASK);
mi_segment_t* const segment = (mi_segment_t*)(((uintptr_t)p - 1) & ~MI_SEGMENT_MASK);
#if MI_INTPTR_SIZE <= 4
return (p==NULL ? NULL : segment);
#else
return ((intptr_t)segment <= 0 ? NULL : segment);
#endif
}
static inline mi_page_t* mi_slice_to_page(mi_slice_t* s) {
@@ -424,7 +545,8 @@ static inline mi_slice_t* mi_page_to_slice(mi_page_t* p) {
// Segment belonging to a page
static inline mi_segment_t* _mi_page_segment(const mi_page_t* page) {
mi_segment_t* segment = _mi_ptr_segment(page);
mi_assert_internal(page!=NULL);
mi_segment_t* segment = _mi_ptr_segment(page);
mi_assert_internal(segment == NULL || ((mi_slice_t*)page >= segment->slices && (mi_slice_t*)page < segment->slices + segment->slice_entries));
return segment;
}
@@ -452,31 +574,28 @@ static inline mi_page_t* _mi_segment_page_of(const mi_segment_t* segment, const
}
// Quick page start for initialized pages
static inline uint8_t* _mi_page_start(const mi_segment_t* segment, const mi_page_t* page, size_t* page_size) {
return _mi_segment_page_start(segment, page, page_size);
static inline uint8_t* mi_page_start(const mi_page_t* page) {
mi_assert_internal(page->page_start != NULL);
mi_assert_expensive(_mi_segment_page_start(_mi_page_segment(page),page,NULL) == page->page_start);
return page->page_start;
}
// Get the page containing the pointer
static inline mi_page_t* _mi_ptr_page(void* p) {
mi_assert_internal(p!=NULL);
return _mi_segment_page_of(_mi_ptr_segment(p), p);
}
// Get the block size of a page (special case for huge objects)
static inline size_t mi_page_block_size(const mi_page_t* page) {
const size_t bsize = page->xblock_size;
mi_assert_internal(bsize > 0);
if mi_likely(bsize < MI_HUGE_BLOCK_SIZE) {
return bsize;
}
else {
size_t psize;
_mi_segment_page_start(_mi_page_segment(page), page, &psize);
return psize;
}
mi_assert_internal(page->block_size > 0);
return page->block_size;
}
static inline bool mi_page_is_huge(const mi_page_t* page) {
return (_mi_page_segment(page)->kind == MI_SEGMENT_HUGE);
mi_assert_internal((page->is_huge && _mi_page_segment(page)->kind == MI_SEGMENT_HUGE) ||
(!page->is_huge && _mi_page_segment(page)->kind != MI_SEGMENT_HUGE));
return page->is_huge;
}
// Get the usable block size of a page without fixed padding.
@@ -511,6 +630,7 @@ static inline mi_heap_t* mi_page_heap(const mi_page_t* page) {
static inline void mi_page_set_heap(mi_page_t* page, mi_heap_t* heap) {
mi_assert_internal(mi_page_thread_free_flag(page) != MI_DELAYED_FREEING);
mi_atomic_store_release(&page->xheap,(uintptr_t)heap);
if (heap != NULL) { page->heap_tag = heap->tag; }
}
// Thread free flag helpers
@@ -550,7 +670,7 @@ static inline bool mi_page_immediate_available(const mi_page_t* page) {
}
// is more than 7/8th of a page in use?
static inline bool mi_page_mostly_used(const mi_page_t* page) {
static inline bool mi_page_is_mostly_used(const mi_page_t* page) {
if (page==NULL) return true;
uint16_t frac = page->reserved / 8U;
return (page->reserved - page->used <= frac);
@@ -581,6 +701,39 @@ static inline void mi_page_set_has_aligned(mi_page_t* page, bool has_aligned) {
page->flags.x.has_aligned = has_aligned;
}
/* -------------------------------------------------------------------
Guarded objects
------------------------------------------------------------------- */
#if MI_GUARDED
static inline bool mi_block_ptr_is_guarded(const mi_block_t* block, const void* p) {
const ptrdiff_t offset = (uint8_t*)p - (uint8_t*)block;
return (offset >= (ptrdiff_t)(sizeof(mi_block_t)) && block->next == MI_BLOCK_TAG_GUARDED);
}
static inline bool mi_heap_malloc_use_guarded(mi_heap_t* heap, size_t size) {
// this code is written to result in fast assembly as it is on the hot path for allocation
const size_t count = heap->guarded_sample_count - 1; // if the rate was 0, this will underflow and count for a long time..
if mi_likely(count != 0) {
// no sample
heap->guarded_sample_count = count;
return false;
}
else if (size >= heap->guarded_size_min && size <= heap->guarded_size_max) {
// use guarded allocation
heap->guarded_sample_count = heap->guarded_sample_rate; // reset
return (heap->guarded_sample_rate != 0);
}
else {
// failed size criteria, rewind count (but don't write to an empty heap)
if (heap->guarded_sample_rate != 0) { heap->guarded_sample_count = 1; }
return false;
}
}
mi_decl_restrict void* _mi_heap_malloc_guarded(mi_heap_t* heap, size_t size, bool zero) mi_attr_noexcept;
#endif
/* -------------------------------------------------------------------
Encoding/Decoding the free list next pointers
@@ -640,6 +793,16 @@ static inline mi_encoded_t mi_ptr_encode(const void* null, const void* p, const
return mi_rotl(x ^ keys[1], keys[0]) + keys[0];
}
static inline uint32_t mi_ptr_encode_canary(const void* null, const void* p, const uintptr_t* keys) {
const uint32_t x = (uint32_t)(mi_ptr_encode(null,p,keys));
// make the lowest byte 0 to prevent spurious read overflows which could be a security issue (issue #951)
#ifdef MI_BIG_ENDIAN
return (x & 0x00FFFFFF);
#else
return (x & 0xFFFFFF00);
#endif
}
static inline mi_block_t* mi_block_nextx( const void* null, const mi_block_t* block, const uintptr_t* keys ) {
mi_track_mem_defined(block,sizeof(mi_block_t));
mi_block_t* next;
@@ -726,12 +889,12 @@ size_t _mi_commit_mask_next_run(const mi_commit_mask_t* cm, size_t* idx);
#define mi_commit_mask_foreach(cm,idx,count) \
idx = 0; \
while ((count = _mi_commit_mask_next_run(cm,&idx)) > 0) {
while ((count = _mi_commit_mask_next_run(cm,&idx)) > 0) {
#define mi_commit_mask_foreach_end() \
idx += count; \
}
/* -----------------------------------------------------------
@@ -749,8 +912,10 @@ static inline mi_memid_t _mi_memid_none(void) {
return _mi_memid_create(MI_MEM_NONE);
}
static inline mi_memid_t _mi_memid_create_os(bool committed, bool is_zero, bool is_large) {
static inline mi_memid_t _mi_memid_create_os(void* base, size_t size, bool committed, bool is_zero, bool is_large) {
mi_memid_t memid = _mi_memid_create(MI_MEM_OS);
memid.mem.os.base = base;
memid.mem.os.size = size;
memid.initially_committed = committed;
memid.initially_zero = is_zero;
memid.is_pinned = is_large;
@@ -764,7 +929,7 @@ static inline mi_memid_t _mi_memid_create_os(bool committed, bool is_zero, bool
static inline uintptr_t _mi_random_shuffle(uintptr_t x) {
if (x==0) { x = 17; } // ensure we don't get stuck in generating zeros
#if (MI_INTPTR_SIZE==8)
#if (MI_INTPTR_SIZE>=8)
// by Sebastiano Vigna, see: <http://xoshiro.di.unimi.it/splitmix64.c>
x ^= x >> 30;
x *= 0xbf58476d1ce4e5b9UL;
@@ -782,24 +947,6 @@ static inline uintptr_t _mi_random_shuffle(uintptr_t x) {
return x;
}
// -------------------------------------------------------------------
// Optimize numa node access for the common case (= one node)
// -------------------------------------------------------------------
int _mi_os_numa_node_get(mi_os_tld_t* tld);
size_t _mi_os_numa_node_count_get(void);
extern _Atomic(size_t) _mi_numa_node_count;
static inline int _mi_os_numa_node(mi_os_tld_t* tld) {
if mi_likely(mi_atomic_load_relaxed(&_mi_numa_node_count) == 1) { return 0; }
else return _mi_os_numa_node_get(tld);
}
static inline size_t _mi_os_numa_node_count(void) {
const size_t count = mi_atomic_load_relaxed(&_mi_numa_node_count);
if mi_likely(count > 0) { return count; }
else return _mi_os_numa_node_count_get();
}
// -----------------------------------------------------------------------
@@ -810,21 +957,21 @@ static inline size_t _mi_os_numa_node_count(void) {
#include <limits.h> // LONG_MAX
#define MI_HAVE_FAST_BITSCAN
static inline size_t mi_clz(uintptr_t x) {
if (x==0) return MI_INTPTR_BITS;
#if (INTPTR_MAX == LONG_MAX)
return __builtin_clzl(x);
#else
return __builtin_clzll(x);
#endif
static inline size_t mi_clz(size_t x) {
if (x==0) return MI_SIZE_BITS;
#if (SIZE_MAX == ULONG_MAX)
return __builtin_clzl(x);
#else
return __builtin_clzll(x);
#endif
}
static inline size_t mi_ctz(uintptr_t x) {
if (x==0) return MI_INTPTR_BITS;
#if (INTPTR_MAX == LONG_MAX)
return __builtin_ctzl(x);
#else
return __builtin_ctzll(x);
#endif
static inline size_t mi_ctz(size_t x) {
if (x==0) return MI_SIZE_BITS;
#if (SIZE_MAX == ULONG_MAX)
return __builtin_ctzl(x);
#else
return __builtin_ctzll(x);
#endif
}
#elif defined(_MSC_VER)
@@ -832,38 +979,40 @@ static inline size_t mi_ctz(uintptr_t x) {
#include <limits.h> // LONG_MAX
#include <intrin.h> // BitScanReverse64
#define MI_HAVE_FAST_BITSCAN
static inline size_t mi_clz(uintptr_t x) {
if (x==0) return MI_INTPTR_BITS;
static inline size_t mi_clz(size_t x) {
if (x==0) return MI_SIZE_BITS;
unsigned long idx;
#if (INTPTR_MAX == LONG_MAX)
_BitScanReverse(&idx, x);
#else
_BitScanReverse64(&idx, x);
#endif
return ((MI_INTPTR_BITS - 1) - idx);
#if (SIZE_MAX == ULONG_MAX)
_BitScanReverse(&idx, x);
#else
_BitScanReverse64(&idx, x);
#endif
return ((MI_SIZE_BITS - 1) - (size_t)idx);
}
static inline size_t mi_ctz(uintptr_t x) {
if (x==0) return MI_INTPTR_BITS;
static inline size_t mi_ctz(size_t x) {
if (x==0) return MI_SIZE_BITS;
unsigned long idx;
#if (INTPTR_MAX == LONG_MAX)
_BitScanForward(&idx, x);
#else
_BitScanForward64(&idx, x);
#endif
return idx;
#if (SIZE_MAX == ULONG_MAX)
_BitScanForward(&idx, x);
#else
_BitScanForward64(&idx, x);
#endif
return (size_t)idx;
}
#else
static inline size_t mi_ctz32(uint32_t x) {
static inline size_t mi_ctz_generic32(uint32_t x) {
// de Bruijn multiplication, see <http://supertech.csail.mit.edu/papers/debruijn.pdf>
static const unsigned char debruijn[32] = {
static const uint8_t debruijn[32] = {
0, 1, 28, 2, 29, 14, 24, 3, 30, 22, 20, 15, 25, 17, 4, 8,
31, 27, 13, 23, 21, 19, 16, 7, 26, 12, 18, 6, 11, 5, 10, 9
};
if (x==0) return 32;
return debruijn[((x & -(int32_t)x) * 0x077CB531UL) >> 27];
return debruijn[(uint32_t)((x & -(int32_t)x) * (uint32_t)(0x077CB531U)) >> 27];
}
static inline size_t mi_clz32(uint32_t x) {
static inline size_t mi_clz_generic32(uint32_t x) {
// de Bruijn multiplication, see <http://supertech.csail.mit.edu/papers/debruijn.pdf>
static const uint8_t debruijn[32] = {
31, 22, 30, 21, 18, 10, 29, 2, 20, 17, 15, 13, 9, 6, 28, 1,
@@ -875,37 +1024,61 @@ static inline size_t mi_clz32(uint32_t x) {
x |= x >> 4;
x |= x >> 8;
x |= x >> 16;
return debruijn[(uint32_t)(x * 0x07C4ACDDUL) >> 27];
return debruijn[(uint32_t)(x * (uint32_t)(0x07C4ACDDU)) >> 27];
}
static inline size_t mi_clz(uintptr_t x) {
if (x==0) return MI_INTPTR_BITS;
#if (MI_INTPTR_BITS <= 32)
return mi_clz32((uint32_t)x);
#else
size_t count = mi_clz32((uint32_t)(x >> 32));
if (count < 32) return count;
return (32 + mi_clz32((uint32_t)x));
#endif
static inline size_t mi_ctz(size_t x) {
if (x==0) return MI_SIZE_BITS;
#if (MI_SIZE_BITS <= 32)
return mi_ctz_generic32((uint32_t)x);
#else
const uint32_t lo = (uint32_t)x;
if (lo != 0) {
return mi_ctz_generic32(lo);
}
else {
return (32 + mi_ctz_generic32((uint32_t)(x>>32)));
}
#endif
}
static inline size_t mi_ctz(uintptr_t x) {
if (x==0) return MI_INTPTR_BITS;
#if (MI_INTPTR_BITS <= 32)
return mi_ctz32((uint32_t)x);
#else
size_t count = mi_ctz32((uint32_t)x);
if (count < 32) return count;
return (32 + mi_ctz32((uint32_t)(x>>32)));
#endif
static inline size_t mi_clz(size_t x) {
if (x==0) return MI_SIZE_BITS;
#if (MI_SIZE_BITS <= 32)
return mi_clz_generic32((uint32_t)x);
#else
const uint32_t hi = (uint32_t)(x>>32);
if (hi != 0) {
return mi_clz_generic32(hi);
}
else {
return 32 + mi_clz_generic32((uint32_t)x);
}
#endif
}
#endif
// "bit scan reverse": Return index of the highest bit (or MI_INTPTR_BITS if `x` is zero)
static inline size_t mi_bsr(uintptr_t x) {
return (x==0 ? MI_INTPTR_BITS : MI_INTPTR_BITS - 1 - mi_clz(x));
// "bit scan reverse": Return index of the highest bit (or MI_SIZE_BITS if `x` is zero)
static inline size_t mi_bsr(size_t x) {
return (x==0 ? MI_SIZE_BITS : MI_SIZE_BITS - 1 - mi_clz(x));
}
size_t _mi_popcount_generic(size_t x);
static inline size_t mi_popcount(size_t x) {
if (x<=1) return x;
if (x==SIZE_MAX) return MI_SIZE_BITS;
#if defined(__GNUC__)
#if (SIZE_MAX == ULONG_MAX)
return __builtin_popcountl(x);
#else
return __builtin_popcountll(x);
#endif
#else
return _mi_popcount_generic(x);
#endif
}
// ---------------------------------------------------------------------------------
// Provide our own `_mi_memcpy` for potential performance optimizations.
@@ -917,9 +1090,10 @@ static inline size_t mi_bsr(uintptr_t x) {
#if !MI_TRACK_ENABLED && defined(_WIN32) && (defined(_M_IX86) || defined(_M_X64))
#include <intrin.h>
extern bool _mi_cpu_has_fsrm;
extern mi_decl_hidden bool _mi_cpu_has_fsrm;
extern mi_decl_hidden bool _mi_cpu_has_erms;
static inline void _mi_memcpy(void* dst, const void* src, size_t n) {
if (_mi_cpu_has_fsrm) {
if ((_mi_cpu_has_fsrm && n <= 128) || (_mi_cpu_has_erms && n > 128)) {
__movsb((unsigned char*)dst, (const unsigned char*)src, n);
}
else {
@@ -927,7 +1101,7 @@ static inline void _mi_memcpy(void* dst, const void* src, size_t n) {
}
}
static inline void _mi_memzero(void* dst, size_t n) {
if (_mi_cpu_has_fsrm) {
if ((_mi_cpu_has_fsrm && n <= 128) || (_mi_cpu_has_erms && n > 128)) {
__stosb((unsigned char*)dst, 0, n);
}
else {

200
mimalloc/mimalloc-sys/mimalloc/include/mimalloc/prim.h
View File

@@ -1,5 +1,5 @@
/* ----------------------------------------------------------------------------
Copyright (c) 2018-2023, Microsoft Research, Daan Leijen
Copyright (c) 2018-2024, Microsoft Research, Daan Leijen
This is free software; you can redistribute it and/or modify it under the
terms of the MIT license. A copy of the license can be found in the file
"LICENSE" at the root of this distribution.
@@ -7,27 +7,29 @@ terms of the MIT license. A copy of the license can be found in the file
#pragma once
#ifndef MIMALLOC_PRIM_H
#define MIMALLOC_PRIM_H
#include "internal.h" // mi_decl_hidden
// --------------------------------------------------------------------------
// This file specifies the primitive portability API.
// Each OS/host needs to implement these primitives, see `src/prim`
// for implementations on Window, macOS, WASI, and Linux/Unix.
//
// note: on all primitive functions, we always have result parameters != NUL, and:
// note: on all primitive functions, we always have result parameters != NULL, and:
// addr != NULL and page aligned
// size > 0 and page aligned
// return value is an error code an int where 0 is success.
// the return value is an error code as an `int` where 0 is success
// --------------------------------------------------------------------------
// OS memory configuration
typedef struct mi_os_mem_config_s {
size_t page_size; // 4KiB
size_t large_page_size; // 2MiB
size_t alloc_granularity; // smallest allocation size (on Windows 64KiB)
bool has_overcommit; // can we reserve more memory than can be actually committed?
bool must_free_whole; // must allocated blocks be freed as a whole (false for mmap, true for VirtualAlloc)
bool has_virtual_reserve; // supports virtual address space reservation? (if true we can reserve virtual address space without using commit or physical memory)
size_t page_size; // default to 4KiB
size_t large_page_size; // 0 if not supported, usually 2MiB (4MiB on Windows)
size_t alloc_granularity; // smallest allocation size (usually 4KiB, on Windows 64KiB)
size_t physical_memory_in_kib; // physical memory size in KiB
size_t virtual_address_bits; // usually 48 or 56 bits on 64-bit systems. (used to determine secure randomization)
bool has_overcommit; // can we reserve more memory than can be actually committed?
bool has_partial_free; // can allocated blocks be freed partially? (true for mmap, false for VirtualAlloc)
bool has_virtual_reserve; // supports virtual address space reservation? (if true we can reserve virtual address space without using commit or physical memory)
} mi_os_mem_config_t;
// Initialize
@@ -35,15 +37,16 @@ void _mi_prim_mem_init( mi_os_mem_config_t* config );
// Free OS memory
int _mi_prim_free(void* addr, size_t size );
// Allocate OS memory. Return NULL on error.
// The `try_alignment` is just a hint and the returned pointer does not have to be aligned.
// If `commit` is false, the virtual memory range only needs to be reserved (with no access)
// If `commit` is false, the virtual memory range only needs to be reserved (with no access)
// which will later be committed explicitly using `_mi_prim_commit`.
// `is_zero` is set to true if the memory was zero initialized (as on most OS's)
// The `hint_addr` address is either `NULL` or a preferred allocation address but can be ignored.
// pre: !commit => !allow_large
// try_alignment >= _mi_os_page_size() and a power of 2
int _mi_prim_alloc(size_t size, size_t try_alignment, bool commit, bool allow_large, bool* is_large, bool* is_zero, void** addr);
int _mi_prim_alloc(void* hint_addr, size_t size, size_t try_alignment, bool commit, bool allow_large, bool* is_large, bool* is_zero, void** addr);
// Commit memory. Returns error code or 0 on success.
// For example, on Linux this would make the memory PROT_READ|PROT_WRITE.
@@ -56,10 +59,15 @@ int _mi_prim_commit(void* addr, size_t size, bool* is_zero);
// pre: needs_recommit != NULL
int _mi_prim_decommit(void* addr, size_t size, bool* needs_recommit);
// Reset memory. The range keeps being accessible but the content might be reset.
// Reset memory. The range keeps being accessible but the content might be reset to zero at any moment.
// Returns error code or 0 on success.
int _mi_prim_reset(void* addr, size_t size);
// Reuse memory. This is called for memory that is already committed but
// may have been reset (`_mi_prim_reset`) or decommitted (`_mi_prim_decommit`) where `needs_recommit` was false.
// Returns error code or 0 on success. On most platforms this is a no-op.
int _mi_prim_reuse(void* addr, size_t size);
// Protect memory. Returns error code or 0 on success.
int _mi_prim_protect(void* addr, size_t size, bool protect);
@@ -82,11 +90,11 @@ mi_msecs_t _mi_prim_clock_now(void);
typedef struct mi_process_info_s {
mi_msecs_t elapsed;
mi_msecs_t utime;
mi_msecs_t stime;
size_t current_rss;
size_t peak_rss;
mi_msecs_t stime;
size_t current_rss;
size_t peak_rss;
size_t current_commit;
size_t peak_commit;
size_t peak_commit;
size_t page_faults;
} mi_process_info_t;
@@ -116,42 +124,30 @@ void _mi_prim_thread_associate_default_heap(mi_heap_t* heap);
//-------------------------------------------------------------------
// Thread id: `_mi_prim_thread_id()`
//
// Getting the thread id should be performant as it is called in the
// fast path of `_mi_free` and we specialize for various platforms as
// inlined definitions. Regular code should call `init.c:_mi_thread_id()`.
// We only require _mi_prim_thread_id() to return a unique id
// for each thread (unequal to zero).
// Access to TLS (thread local storage) slots.
// We need fast access to both a unique thread id (in `free.c:mi_free`) and
// to a thread-local heap pointer (in `alloc.c:mi_malloc`).
// To achieve this we use specialized code for various platforms.
//-------------------------------------------------------------------
// defined in `init.c`; do not use these directly
extern mi_decl_thread mi_heap_t* _mi_heap_default; // default heap to allocate from
extern bool _mi_process_is_initialized; // has mi_process_init been called?
static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept;
#if defined(_WIN32)
#define WIN32_LEAN_AND_MEAN
#include <windows.h>
static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept {
// Windows: works on Intel and ARM in both 32- and 64-bit
return (uintptr_t)NtCurrentTeb();
}
// We use assembly for a fast thread id on the main platforms. The TLS layout depends on
// both the OS and libc implementation so we use specific tests for each main platform.
// On some libc + platform combinations we can directly access a thread-local storage (TLS) slot.
// The TLS layout depends on both the OS and libc implementation so we use specific tests for each main platform.
// If you test on another platform and it works please send a PR :-)
// see also https://akkadia.org/drepper/tls.pdf for more info on the TLS register.
#elif defined(__GNUC__) && ( \
(defined(__GLIBC__) && (defined(__x86_64__) || defined(__i386__) || defined(__arm__) || defined(__aarch64__))) \
|| (defined(__APPLE__) && (defined(__x86_64__) || defined(__aarch64__))) \
|| (defined(__BIONIC__) && (defined(__x86_64__) || defined(__i386__) || defined(__arm__) || defined(__aarch64__))) \
//
// Note: we would like to prefer `__builtin_thread_pointer()` nowadays instead of using assembly,
// but unfortunately we can not detect support reliably (see issue #883)
// We also use it on Apple OS as we use a TLS slot for the default heap there.
#if defined(__GNUC__) && ( \
(defined(__GLIBC__) && (defined(__x86_64__) || defined(__i386__) || (defined(__arm__) && __ARM_ARCH >= 7) || defined(__aarch64__))) \
|| (defined(__APPLE__) && (defined(__x86_64__) || defined(__aarch64__) || defined(__POWERPC__))) \
|| (defined(__BIONIC__) && (defined(__x86_64__) || defined(__i386__) || (defined(__arm__) && __ARM_ARCH >= 7) || defined(__aarch64__))) \
|| (defined(__FreeBSD__) && (defined(__x86_64__) || defined(__i386__) || defined(__aarch64__))) \
|| (defined(__OpenBSD__) && (defined(__x86_64__) || defined(__i386__) || defined(__aarch64__))) \
)
#define MI_HAS_TLS_SLOT 1
static inline void* mi_prim_tls_slot(size_t slot) mi_attr_noexcept {
void* res;
const size_t ofs = (slot*sizeof(void*));
@@ -175,6 +171,9 @@ static inline void* mi_prim_tls_slot(size_t slot) mi_attr_noexcept {
__asm__ volatile ("mrs %0, tpidr_el0" : "=r" (tcb));
#endif
res = tcb[slot];
#elif defined(__APPLE__) && defined(__POWERPC__) // ppc, issue #781
MI_UNUSED(ofs);
res = pthread_getspecific(slot);
#endif
return res;
}
@@ -202,9 +201,104 @@ static inline void mi_prim_tls_slot_set(size_t slot, void* value) mi_attr_noexce
__asm__ volatile ("mrs %0, tpidr_el0" : "=r" (tcb));
#endif
tcb[slot] = value;
#elif defined(__APPLE__) && defined(__POWERPC__) // ppc, issue #781
MI_UNUSED(ofs);
pthread_setspecific(slot, value);
#endif
}
#elif _WIN32 && MI_WIN_USE_FIXED_TLS && !defined(MI_WIN_USE_FLS)
// On windows we can store the thread-local heap at a fixed TLS slot to avoid
// thread-local initialization checks in the fast path.
// We allocate a user TLS slot at process initialization (see `windows/prim.c`)
// and store the offset `_mi_win_tls_offset`.
#define MI_HAS_TLS_SLOT 1 // 2 = we can reliably initialize the slot (saving a test on each malloc)
extern mi_decl_hidden size_t _mi_win_tls_offset;
#if MI_WIN_USE_FIXED_TLS > 1
#define MI_TLS_SLOT (MI_WIN_USE_FIXED_TLS)
#elif MI_SIZE_SIZE == 4
#define MI_TLS_SLOT (0x0E10 + _mi_win_tls_offset) // User TLS slots <https://en.wikipedia.org/wiki/Win32_Thread_Information_Block>
#else
#define MI_TLS_SLOT (0x1480 + _mi_win_tls_offset) // User TLS slots <https://en.wikipedia.org/wiki/Win32_Thread_Information_Block>
#endif
static inline void* mi_prim_tls_slot(size_t slot) mi_attr_noexcept {
#if (_M_X64 || _M_AMD64) && !defined(_M_ARM64EC)
return (void*)__readgsqword((unsigned long)slot); // direct load at offset from gs
#elif _M_IX86 && !defined(_M_ARM64EC)
return (void*)__readfsdword((unsigned long)slot); // direct load at offset from fs
#else
return ((void**)NtCurrentTeb())[slot / sizeof(void*)];
#endif
}
static inline void mi_prim_tls_slot_set(size_t slot, void* value) mi_attr_noexcept {
((void**)NtCurrentTeb())[slot / sizeof(void*)] = value;
}
#endif
//-------------------------------------------------------------------
// Get a fast unique thread id.
//
// Getting the thread id should be performant as it is called in the
// fast path of `_mi_free` and we specialize for various platforms as
// inlined definitions. Regular code should call `init.c:_mi_thread_id()`.
// We only require _mi_prim_thread_id() to return a unique id
// for each thread (unequal to zero).
//-------------------------------------------------------------------
// Do we have __builtin_thread_pointer? This would be the preferred way to get a unique thread id
// but unfortunately, it seems we cannot test for this reliably at this time (see issue #883)
// Nevertheless, it seems needed on older graviton platforms (see issue #851).
// For now, we only enable this for specific platforms.
#if !defined(__APPLE__) /* on apple (M1) the wrong register is read (tpidr_el0 instead of tpidrro_el0) so fall back to TLS slot assembly (<https://github.com/microsoft/mimalloc/issues/343#issuecomment-763272369>)*/ \
&& !defined(__CYGWIN__) \
&& !defined(MI_LIBC_MUSL) \
&& (!defined(__clang_major__) || __clang_major__ >= 14) /* older clang versions emit bad code; fall back to using the TLS slot (<https://lore.kernel.org/linux-arm-kernel/202110280952.352F66D8@keescook/T/>) */
#if (defined(__GNUC__) && (__GNUC__ >= 7) && defined(__aarch64__)) /* aarch64 for older gcc versions (issue #851) */ \
|| (defined(__GNUC__) && (__GNUC__ >= 11) && defined(__x86_64__)) \
|| (defined(__clang_major__) && (__clang_major__ >= 14) && (defined(__aarch64__) || defined(__x86_64__)))
#define MI_USE_BUILTIN_THREAD_POINTER 1
#endif
#endif
// defined in `init.c`; do not use these directly
extern mi_decl_hidden mi_decl_thread mi_heap_t* _mi_heap_default; // default heap to allocate from
extern mi_decl_hidden bool _mi_process_is_initialized; // has mi_process_init been called?
static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept;
// Get a unique id for the current thread.
#if defined(MI_PRIM_THREAD_ID)
static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept {
return MI_PRIM_THREAD_ID(); // used for example by CPython for a free threaded build (see python/cpython#115488)
}
#elif defined(_WIN32)
static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept {
// Windows: works on Intel and ARM in both 32- and 64-bit
return (uintptr_t)NtCurrentTeb();
}
#elif MI_USE_BUILTIN_THREAD_POINTER
static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept {
// Works on most Unix based platforms with recent compilers
return (uintptr_t)__builtin_thread_pointer();
}
#elif MI_HAS_TLS_SLOT
static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept {
#if defined(__BIONIC__)
// issue #384, #495: on the Bionic libc (Android), slot 1 is the thread id
@@ -230,7 +324,8 @@ static inline mi_threadid_t _mi_prim_thread_id(void) mi_attr_noexcept {
/* ----------------------------------------------------------------------------------------
The thread local default heap: `_mi_prim_get_default_heap()`
Get the thread local default heap: `_mi_prim_get_default_heap()`
This is inlined here as it is on the fast path for allocation functions.
On most platforms (Windows, Linux, FreeBSD, NetBSD, etc), this just returns a
@@ -251,7 +346,6 @@ static inline mi_heap_t* mi_prim_get_default_heap(void);
#if defined(MI_MALLOC_OVERRIDE)
#if defined(__APPLE__) // macOS
#define MI_TLS_SLOT 89 // seems unused?
// #define MI_TLS_RECURSE_GUARD 1
// other possible unused ones are 9, 29, __PTK_FRAMEWORK_JAVASCRIPTCORE_KEY4 (94), __PTK_FRAMEWORK_GC_KEY9 (112) and __PTK_FRAMEWORK_OLDGC_KEY9 (89)
// see <https://github.com/rweichler/substrate/blob/master/include/pthread_machdep.h>
#elif defined(__OpenBSD__)
@@ -268,16 +362,21 @@ static inline mi_heap_t* mi_prim_get_default_heap(void);
#endif
#if defined(MI_TLS_SLOT)
#if MI_TLS_SLOT
# if !defined(MI_HAS_TLS_SLOT)
# error "trying to use a TLS slot for the default heap, but the mi_prim_tls_slot primitives are not defined"
# endif
static inline mi_heap_t* mi_prim_get_default_heap(void) {
mi_heap_t* heap = (mi_heap_t*)mi_prim_tls_slot(MI_TLS_SLOT);
#if MI_HAS_TLS_SLOT == 1 // check if the TLS slot is initialized
if mi_unlikely(heap == NULL) {
#ifdef __GNUC__
__asm(""); // prevent conditional load of the address of _mi_heap_empty
#endif
heap = (mi_heap_t*)&_mi_heap_empty;
}
#endif
return heap;
}
@@ -301,7 +400,7 @@ static inline mi_heap_t* mi_prim_get_default_heap(void) {
#elif defined(MI_TLS_PTHREAD)
extern pthread_key_t _mi_heap_default_key;
extern mi_decl_hidden pthread_key_t _mi_heap_default_key;
static inline mi_heap_t* mi_prim_get_default_heap(void) {
mi_heap_t* heap = (mi_unlikely(_mi_heap_default_key == (pthread_key_t)(-1)) ? _mi_heap_main_get() : (mi_heap_t*)pthread_getspecific(_mi_heap_default_key));
return (mi_unlikely(heap == NULL) ? (mi_heap_t*)&_mi_heap_empty : heap);
@@ -319,5 +418,4 @@ static inline mi_heap_t* mi_prim_get_default_heap(void) {
#endif // mi_prim_get_default_heap()
#endif // MIMALLOC_PRIM_H

6
mimalloc/mimalloc-sys/mimalloc/include/mimalloc/track.h
View File

@@ -34,7 +34,7 @@ The corresponding `mi_track_free` still uses the block start pointer and origina
The `mi_track_resize` is currently unused but could be called on reallocations within a block.
`mi_track_init` is called at program start.
The following macros are for tools like asan and valgrind to track whether memory is
The following macros are for tools like asan and valgrind to track whether memory is
defined, undefined, or not accessible at all:
#define mi_track_mem_defined(p,size)
@@ -82,8 +82,6 @@ defined, undefined, or not accessible at all:
#define MI_TRACK_HEAP_DESTROY 1
#define MI_TRACK_TOOL "ETW"
#define WIN32_LEAN_AND_MEAN
#include <windows.h>
#include "../src/prim/windows/etw.h"
#define mi_track_init() EventRegistermicrosoft_windows_mimalloc();
@@ -94,7 +92,7 @@ defined, undefined, or not accessible at all:
// no tracking
#define MI_TRACK_ENABLED 0
#define MI_TRACK_HEAP_DESTROY 0
#define MI_TRACK_HEAP_DESTROY 0
#define MI_TRACK_TOOL "none"
#define mi_track_malloc_size(p,reqsize,size,zero)

392
mimalloc/mimalloc-sys/mimalloc/include/mimalloc/types.h
View File

@@ -1,5 +1,5 @@
/* ----------------------------------------------------------------------------
Copyright (c) 2018-2023, Microsoft Research, Daan Leijen
Copyright (c) 2018-2024, Microsoft Research, Daan Leijen
This is free software; you can redistribute it and/or modify it under the
terms of the MIT license. A copy of the license can be found in the file
"LICENSE" at the root of this distribution.
@@ -13,15 +13,20 @@ terms of the MIT license. A copy of the license can be found in the file
// mi_heap_t : all data for a thread-local heap, contains
// lists of all managed heap pages.
// mi_segment_t : a larger chunk of memory (32GiB) from where pages
// are allocated.
// mi_page_t : a mimalloc page (usually 64KiB or 512KiB) from
// are allocated. A segment is divided in slices (64KiB) from
// which pages are allocated.
// mi_page_t : a "mimalloc" page (usually 64KiB or 512KiB) from
// where objects are allocated.
// Note: we write "OS page" for OS memory pages while
// using plain "page" for mimalloc pages (`mi_page_t`).
// --------------------------------------------------------------------------
#include <mimalloc-stats.h>
#include <stddef.h> // ptrdiff_t
#include <stdint.h> // uintptr_t, uint16_t, etc
#include "mimalloc/atomic.h" // _Atomic
#include <stdbool.h> // bool
#include "atomic.h" // _Atomic
#ifdef _MSC_VER
#pragma warning(disable:4214) // bitfield is not int
@@ -63,13 +68,20 @@ terms of the MIT license. A copy of the license can be found in the file
// #define MI_DEBUG 2 // + internal assertion checks
// #define MI_DEBUG 3 // + extensive internal invariant checking (cmake -DMI_DEBUG_FULL=ON)
#if !defined(MI_DEBUG)
#if !defined(NDEBUG) || defined(_DEBUG)
#define MI_DEBUG 2
#else
#if defined(MI_BUILD_RELEASE) || defined(NDEBUG)
#define MI_DEBUG 0
#else
#define MI_DEBUG 2
#endif
#endif
// Use guard pages behind objects of a certain size (set by the MIMALLOC_DEBUG_GUARDED_MIN/MAX options)
// Padding should be disabled when using guard pages
// #define MI_GUARDED 1
#if defined(MI_GUARDED)
#define MI_PADDING 0
#endif
// Reserve extra padding at the end of each block to be more resilient against heap block overflows.
// The padding can detect buffer overflow on free.
#if !defined(MI_PADDING) && (MI_SECURE>=3 || MI_DEBUG>=1 || (MI_TRACK_VALGRIND || MI_TRACK_ASAN || MI_TRACK_ETW))
@@ -89,10 +101,11 @@ terms of the MIT license. A copy of the license can be found in the file
#endif
// We used to abandon huge pages but to eagerly deallocate if freed from another thread,
// but that makes it not possible to visit them during a heap walk or include them in a
// `mi_heap_destroy`. We therefore instead reset/decommit the huge blocks if freed from
// another thread so most memory is available until it gets properly freed by the owning thread.
// We used to abandon huge pages in order to eagerly deallocate it if freed from another thread.
// Unfortunately, that makes it not possible to visit them during a heap walk or include them in a
// `mi_heap_destroy`. We therefore instead reset/decommit the huge blocks nowadays if freed from
// another thread so the memory becomes "virtually" available (and eventually gets properly freed by
// the owning thread).
// #define MI_HUGE_PAGE_ABANDON 1
@@ -157,17 +170,24 @@ typedef int32_t mi_ssize_t;
// Main tuning parameters for segment and page sizes
// Sizes for 64-bit (usually divide by two for 32-bit)
#ifndef MI_SEGMENT_SLICE_SHIFT
#define MI_SEGMENT_SLICE_SHIFT (13 + MI_INTPTR_SHIFT) // 64KiB (32KiB on 32-bit)
#endif
#ifndef MI_SEGMENT_SHIFT
#if MI_INTPTR_SIZE > 4
#define MI_SEGMENT_SHIFT ( 9 + MI_SEGMENT_SLICE_SHIFT) // 32MiB
#else
#define MI_SEGMENT_SHIFT ( 7 + MI_SEGMENT_SLICE_SHIFT) // 4MiB on 32-bit
#endif
#endif
#ifndef MI_SMALL_PAGE_SHIFT
#define MI_SMALL_PAGE_SHIFT (MI_SEGMENT_SLICE_SHIFT) // 64KiB
#endif
#ifndef MI_MEDIUM_PAGE_SHIFT
#define MI_MEDIUM_PAGE_SHIFT ( 3 + MI_SMALL_PAGE_SHIFT) // 512KiB
#endif
// Derived constants
#define MI_SEGMENT_SIZE (MI_ZU(1)<<MI_SEGMENT_SHIFT)
@@ -179,30 +199,37 @@ typedef int32_t mi_ssize_t;
#define MI_SMALL_PAGE_SIZE (MI_ZU(1)<<MI_SMALL_PAGE_SHIFT)
#define MI_MEDIUM_PAGE_SIZE (MI_ZU(1)<<MI_MEDIUM_PAGE_SHIFT)
#define MI_SMALL_OBJ_SIZE_MAX (MI_SMALL_PAGE_SIZE/4) // 8KiB on 64-bit
#define MI_MEDIUM_OBJ_SIZE_MAX (MI_MEDIUM_PAGE_SIZE/4) // 128KiB on 64-bit
#define MI_MEDIUM_OBJ_WSIZE_MAX (MI_MEDIUM_OBJ_SIZE_MAX/MI_INTPTR_SIZE)
#define MI_LARGE_OBJ_SIZE_MAX (MI_SEGMENT_SIZE/2) // 32MiB on 64-bit
#define MI_SMALL_OBJ_SIZE_MAX (MI_SMALL_PAGE_SIZE/8) // 8 KiB on 64-bit
#define MI_MEDIUM_OBJ_SIZE_MAX (MI_MEDIUM_PAGE_SIZE/8) // 64 KiB on 64-bit
#define MI_MEDIUM_OBJ_WSIZE_MAX (MI_MEDIUM_OBJ_SIZE_MAX/MI_INTPTR_SIZE)
#define MI_LARGE_OBJ_SIZE_MAX (MI_SEGMENT_SIZE/2) // 16 MiB on 64-bit
#define MI_LARGE_OBJ_WSIZE_MAX (MI_LARGE_OBJ_SIZE_MAX/MI_INTPTR_SIZE)
// Maximum number of size classes. (spaced exponentially in 12.5% increments)
#define MI_BIN_HUGE (73U)
#if MI_BIN_HUGE != 73U
#error "mimalloc internal: expecting 73 bins"
#endif
#if (MI_MEDIUM_OBJ_WSIZE_MAX >= 655360)
#error "mimalloc internal: define more bins"
#endif
// Maximum slice offset (15)
#define MI_MAX_SLICE_OFFSET ((MI_ALIGNMENT_MAX / MI_SEGMENT_SLICE_SIZE) - 1)
// Maximum block size for which blocks are guaranteed to be block size aligned. (see `segment.c:_mi_segment_page_start`)
#define MI_MAX_ALIGN_GUARANTEE (MI_MEDIUM_OBJ_SIZE_MAX)
// Used as a special value to encode block sizes in 32 bits.
#define MI_HUGE_BLOCK_SIZE ((uint32_t)(2*MI_GiB))
// Alignments over MI_BLOCK_ALIGNMENT_MAX are allocated in dedicated huge page segments
#define MI_BLOCK_ALIGNMENT_MAX (MI_SEGMENT_SIZE >> 1)
// blocks up to this size are always allocated aligned
#define MI_MAX_ALIGN_GUARANTEE (8*MI_MAX_ALIGN_SIZE)
// Maximum slice count (255) for which we can find the page for interior pointers
#define MI_MAX_SLICE_OFFSET_COUNT ((MI_BLOCK_ALIGNMENT_MAX / MI_SEGMENT_SLICE_SIZE) - 1)
// Alignments over MI_ALIGNMENT_MAX are allocated in dedicated huge page segments
#define MI_ALIGNMENT_MAX (MI_SEGMENT_SIZE >> 1)
// we never allocate more than PTRDIFF_MAX (see also <https://sourceware.org/ml/libc-announce/2019/msg00001.html>)
// on 64-bit+ systems we also limit the maximum allocation size such that the slice count fits in 32-bits. (issue #877)
#if (PTRDIFF_MAX > INT32_MAX) && (PTRDIFF_MAX >= (MI_SEGMENT_SLIZE_SIZE * UINT32_MAX))
#define MI_MAX_ALLOC_SIZE (MI_SEGMENT_SLICE_SIZE * (UINT32_MAX-1))
#else
#define MI_MAX_ALLOC_SIZE PTRDIFF_MAX
#endif
// ------------------------------------------------------
@@ -221,13 +248,20 @@ typedef struct mi_block_s {
mi_encoded_t next;
} mi_block_t;
#if MI_GUARDED
// we always align guarded pointers in a block at an offset
// the block `next` field is then used as a tag to distinguish regular offset aligned blocks from guarded ones
#define MI_BLOCK_TAG_ALIGNED ((mi_encoded_t)(0))
#define MI_BLOCK_TAG_GUARDED (~MI_BLOCK_TAG_ALIGNED)
#endif
// The delayed flags are used for efficient multi-threaded free-ing
typedef enum mi_delayed_e {
MI_USE_DELAYED_FREE = 0, // push on the owning heap thread delayed list
MI_DELAYED_FREEING = 1, // temporary: another thread is accessing the owning heap
MI_NO_DELAYED_FREE = 2, // optimize: push on page local thread free queue if another block is already in the heap thread delayed free list
MI_NEVER_DELAYED_FREE = 3 // sticky, only resets on page reclaim
MI_NEVER_DELAYED_FREE = 3 // sticky: used for abandoned pages without a owning heap; this only resets on page reclaim
} mi_delayed_t;
@@ -244,7 +278,7 @@ typedef union mi_page_flags_s {
#else
// under thread sanitizer, use a byte for each flag to suppress warning, issue #130
typedef union mi_page_flags_s {
uint16_t full_aligned;
uint32_t full_aligned;
struct {
uint8_t in_full;
uint8_t has_aligned;
@@ -266,7 +300,6 @@ typedef uintptr_t mi_thread_free_t;
// implement a monotonic heartbeat. The `thread_free` list is needed for
// avoiding atomic operations in the common case.
//
//
// `used - |thread_free|` == actual blocks that are in use (alive)
// `used - |thread_free| + |free| + |local_free| == capacity`
//
@@ -274,16 +307,13 @@ typedef uintptr_t mi_thread_free_t;
// the number of memory accesses in the `mi_page_all_free` function(s).
//
// Notes:
// - Access is optimized for `mi_free` and `mi_page_alloc` (in `alloc.c`)
// - Access is optimized for `free.c:mi_free` and `alloc.c:mi_page_alloc`
// - Using `uint16_t` does not seem to slow things down
// - The size is 8 words on 64-bit which helps the page index calculations
// (and 10 words on 32-bit, and encoded free lists add 2 words. Sizes 10
// and 12 are still good for address calculation)
// - To limit the structure size, the `xblock_size` is 32-bits only; for
// blocks > MI_HUGE_BLOCK_SIZE the size is determined from the segment page size
// - `thread_free` uses the bottom bits as a delayed-free flags to optimize
// - The size is 12 words on 64-bit which helps the page index calculations
// (and 14 words on 32-bit, and encoded free lists add 2 words)
// - `xthread_free` uses the bottom bits as a delayed-free flags to optimize
// concurrent frees where only the first concurrent free adds to the owning
// heap `thread_delayed_free` list (see `alloc.c:mi_free_block_mt`).
// heap `thread_delayed_free` list (see `free.c:mi_free_block_mt`).
// The invariant is that no-delayed-free is only set if there is
// at least one block that will be added, or as already been added, to
// the owning heap `thread_delayed_free` list. This guarantees that pages
@@ -291,21 +321,26 @@ typedef uintptr_t mi_thread_free_t;
typedef struct mi_page_s {
// "owned" by the segment
uint32_t slice_count; // slices in this page (0 if not a page)
uint32_t slice_offset; // distance from the actual page data slice (0 if a page)
uint8_t is_committed : 1; // `true` if the page virtual memory is committed
uint8_t is_zero_init : 1; // `true` if the page was initially zero initialized
uint32_t slice_offset; // distance from the actual page data slice (0 if a page)
uint8_t is_committed:1; // `true` if the page virtual memory is committed
uint8_t is_zero_init:1; // `true` if the page was initially zero initialized
uint8_t is_huge:1; // `true` if the page is in a huge segment (`segment->kind == MI_SEGMENT_HUGE`)
// padding
// layout like this to optimize access in `mi_malloc` and `mi_free`
uint16_t capacity; // number of blocks committed, must be the first field, see `segment.c:page_clear`
uint16_t reserved; // number of blocks reserved in memory
mi_page_flags_t flags; // `in_full` and `has_aligned` flags (8 bits)
uint8_t free_is_zero : 1; // `true` if the blocks in the free list are zero initialized
uint8_t retire_expire : 7; // expiration count for retired blocks
uint8_t free_is_zero:1; // `true` if the blocks in the free list are zero initialized
uint8_t retire_expire:7; // expiration count for retired blocks
mi_block_t* free; // list of available free blocks (`malloc` allocates from this list)
uint32_t used; // number of blocks in use (including blocks in `local_free` and `thread_free`)
uint32_t xblock_size; // size available in each block (always `>0`)
mi_block_t* local_free; // list of deferred free blocks by this thread (migrates to `free`)
uint16_t used; // number of blocks in use (including blocks in `thread_free`)
uint8_t block_size_shift; // if not zero, then `(1 << block_size_shift) == block_size` (only used for fast path in `free.c:_mi_page_ptr_unalign`)
uint8_t heap_tag; // tag of the owning heap, used to separate heaps by object type
// padding
size_t block_size; // size available in each block (always `>0`)
uint8_t* page_start; // start of the page area containing the blocks
#if (MI_ENCODE_FREELIST || MI_PADDING)
uintptr_t keys[2]; // two random keys to encode the free lists (see `_mi_block_next`) or padding canary
@@ -317,10 +352,8 @@ typedef struct mi_page_s {
struct mi_page_s* next; // next page owned by this thread with the same `block_size`
struct mi_page_s* prev; // previous page owned by this thread with the same `block_size`
// 64-bit 9 words, 32-bit 12 words, (+2 for secure)
#if MI_INTPTR_SIZE==8
uintptr_t padding[1];
#endif
// 64-bit 11 words, 32-bit 13 words, (+2 for secure)
void* padding[1];
} mi_page_t;
@@ -331,21 +364,22 @@ typedef struct mi_page_s {
typedef enum mi_page_kind_e {
MI_PAGE_SMALL, // small blocks go into 64KiB pages inside a segment
MI_PAGE_MEDIUM, // medium blocks go into medium pages inside a segment
MI_PAGE_LARGE, // larger blocks go into a page of just one block
MI_PAGE_HUGE, // huge blocks (> 16 MiB) are put into a single page in a single segment.
MI_PAGE_MEDIUM, // medium blocks go into 512KiB pages inside a segment
MI_PAGE_LARGE, // larger blocks go into a single page spanning a whole segment
MI_PAGE_HUGE // a huge page is a single page in a segment of variable size
// used for blocks `> MI_LARGE_OBJ_SIZE_MAX` or an aligment `> MI_BLOCK_ALIGNMENT_MAX`.
} mi_page_kind_t;
typedef enum mi_segment_kind_e {
MI_SEGMENT_NORMAL, // MI_SEGMENT_SIZE size with pages inside.
MI_SEGMENT_HUGE, // > MI_LARGE_SIZE_MAX segment with just one huge page inside.
MI_SEGMENT_HUGE, // segment with just one huge page inside.
} mi_segment_kind_t;
// ------------------------------------------------------
// A segment holds a commit mask where a bit is set if
// the corresponding MI_COMMIT_SIZE area is committed.
// The MI_COMMIT_SIZE must be a multiple of the slice
// size. If it is equal we have the most fine grained
// size. If it is equal we have the most fine grained
// decommit (but setting it higher can be more efficient).
// The MI_MINIMAL_COMMIT_SIZE is the minimal amount that will
// be committed in one go which can be set higher than
@@ -353,9 +387,9 @@ typedef enum mi_segment_kind_e {
// is still tracked in fine-grained MI_COMMIT_SIZE chunks)
// ------------------------------------------------------
#define MI_MINIMAL_COMMIT_SIZE (1*MI_SEGMENT_SLICE_SIZE)
#define MI_MINIMAL_COMMIT_SIZE (1*MI_SEGMENT_SLICE_SIZE)
#define MI_COMMIT_SIZE (MI_SEGMENT_SLICE_SIZE) // 64KiB
#define MI_COMMIT_MASK_BITS (MI_SEGMENT_SIZE / MI_COMMIT_SIZE)
#define MI_COMMIT_MASK_BITS (MI_SEGMENT_SIZE / MI_COMMIT_SIZE)
#define MI_COMMIT_MASK_FIELD_BITS MI_SIZE_BITS
#define MI_COMMIT_MASK_FIELD_COUNT (MI_COMMIT_MASK_BITS / MI_COMMIT_MASK_FIELD_BITS)
@@ -371,13 +405,17 @@ typedef mi_page_t mi_slice_t;
typedef int64_t mi_msecs_t;
// ---------------------------------------------------------------
// a memory id tracks the provenance of arena/OS allocated memory
// ---------------------------------------------------------------
// Memory can reside in arena's, direct OS allocated, or statically allocated. The memid keeps track of this.
typedef enum mi_memkind_e {
MI_MEM_NONE, // not allocated
MI_MEM_EXTERNAL, // not owned by mimalloc but provided externally (via `mi_manage_os_memory` for example)
MI_MEM_STATIC, // allocated in a static area and should not be freed (for arena meta data for example)
MI_MEM_OS, // allocated from the OS
MI_MEM_OS_HUGE, // allocated as huge os pages
MI_MEM_OS_HUGE, // allocated as huge OS pages (usually 1GiB, pinned to physical memory)
MI_MEM_OS_REMAP, // allocated in a remapable area (i.e. using `mremap`)
MI_MEM_ARENA // allocated from an arena (the usual case)
} mi_memkind_t;
@@ -388,13 +426,13 @@ static inline bool mi_memkind_is_os(mi_memkind_t memkind) {
typedef struct mi_memid_os_info {
void* base; // actual base address of the block (used for offset aligned allocations)
size_t alignment; // alignment at allocation
size_t size; // full allocation size
} mi_memid_os_info_t;
typedef struct mi_memid_arena_info {
size_t block_index; // index in the arena
mi_arena_id_t id; // arena id (>= 1)
bool is_exclusive; // the arena can only be used for specific arena allocations
bool is_exclusive; // this arena can only be used for specific arena allocations
} mi_memid_arena_info_t;
typedef struct mi_memid_s {
@@ -402,47 +440,64 @@ typedef struct mi_memid_s {
mi_memid_os_info_t os; // only used for MI_MEM_OS
mi_memid_arena_info_t arena; // only used for MI_MEM_ARENA
} mem;
bool is_pinned; // `true` if we cannot decommit/reset/protect in this memory (e.g. when allocated using large OS pages)
bool is_pinned; // `true` if we cannot decommit/reset/protect in this memory (e.g. when allocated using large (2Mib) or huge (1GiB) OS pages)
bool initially_committed;// `true` if the memory was originally allocated as committed
bool initially_zero; // `true` if the memory was originally zero initialized
mi_memkind_t memkind;
} mi_memid_t;
// Segments are large allocated memory blocks (8mb on 64 bit) from
// the OS. Inside segments we allocated fixed size _pages_ that
// contain blocks.
// -----------------------------------------------------------------------------------------
// Segments are large allocated memory blocks (32mb on 64 bit) from arenas or the OS.
//
// Inside segments we allocated fixed size mimalloc pages (`mi_page_t`) that contain blocks.
// The start of a segment is this structure with a fixed number of slice entries (`slices`)
// usually followed by a guard OS page and the actual allocation area with pages.
// While a page is not allocated, we view it's data as a `mi_slice_t` (instead of a `mi_page_t`).
// Of any free area, the first slice has the info and `slice_offset == 0`; for any subsequent
// slices part of the area, the `slice_offset` is the byte offset back to the first slice
// (so we can quickly find the page info on a free, `internal.h:_mi_segment_page_of`).
// For slices, the `block_size` field is repurposed to signify if a slice is used (`1`) or not (`0`).
// Small and medium pages use a fixed amount of slices to reduce slice fragmentation, while
// large and huge pages span a variable amount of slices.
typedef struct mi_subproc_s mi_subproc_t;
typedef struct mi_segment_s {
// constant fields
mi_memid_t memid; // memory id for arena allocation
bool allow_decommit;
bool allow_purge;
mi_memid_t memid; // memory id for arena/OS allocation
bool allow_decommit; // can we decommmit the memory
bool allow_purge; // can we purge the memory (reset or decommit)
size_t segment_size;
mi_subproc_t* subproc; // segment belongs to sub process
// segment fields
mi_msecs_t purge_expire;
mi_commit_mask_t purge_mask;
mi_commit_mask_t commit_mask;
_Atomic(struct mi_segment_s*) abandoned_next;
mi_msecs_t purge_expire; // purge slices in the `purge_mask` after this time
mi_commit_mask_t purge_mask; // slices that can be purged
mi_commit_mask_t commit_mask; // slices that are currently committed
// from here is zero initialized
struct mi_segment_s* next; // the list of freed segments in the cache (must be first field, see `segment.c:mi_segment_init`)
bool was_reclaimed; // true if it was reclaimed (used to limit on-free reclamation)
bool dont_free; // can be temporarily true to ensure the segment is not freed
size_t abandoned; // abandoned pages (i.e. the original owning thread stopped) (`abandoned <= used`)
size_t abandoned_visits; // count how often this segment is visited in the abandoned list (to force reclaim it it is too long)
size_t abandoned_visits; // count how often this segment is visited during abondoned reclamation (to force reclaim if it takes too long)
size_t used; // count of pages in use
uintptr_t cookie; // verify addresses in debug mode: `mi_ptr_cookie(segment) == segment->cookie`
uintptr_t cookie; // verify addresses in debug mode: `mi_ptr_cookie(segment) == segment->cookie`
struct mi_segment_s* abandoned_os_next; // only used for abandoned segments outside arena's, and only if `mi_option_visit_abandoned` is enabled
struct mi_segment_s* abandoned_os_prev;
size_t segment_slices; // for huge segments this may be different from `MI_SLICES_PER_SEGMENT`
size_t segment_info_slices; // initial slices we are using segment info and possible guard pages.
size_t segment_info_slices; // initial count of slices that we are using for segment info and possible guard pages.
// layout like this to optimize access in `mi_free`
mi_segment_kind_t kind;
size_t slice_entries; // entries in the `slices` array, at most `MI_SLICES_PER_SEGMENT`
_Atomic(mi_threadid_t) thread_id; // unique id of the thread owning this segment
mi_slice_t slices[MI_SLICES_PER_SEGMENT+1]; // one more for huge blocks with large alignment
mi_slice_t slices[MI_SLICES_PER_SEGMENT+1]; // one extra final entry for huge blocks with large alignment
} mi_segment_t;
@@ -499,22 +554,84 @@ typedef struct mi_padding_s {
// A heap owns a set of pages.
struct mi_heap_s {
mi_tld_t* tld;
mi_page_t* pages_free_direct[MI_PAGES_DIRECT]; // optimize: array where every entry points a page with possibly free blocks in the corresponding queue for that size.
mi_page_queue_t pages[MI_BIN_FULL + 1]; // queue of pages for each size class (or "bin")
_Atomic(mi_block_t*) thread_delayed_free;
mi_threadid_t thread_id; // thread this heap belongs too
mi_arena_id_t arena_id; // arena id if the heap belongs to a specific arena (or 0)
mi_arena_id_t arena_id; // arena id if the heap belongs to a specific arena (or 0)
uintptr_t cookie; // random cookie to verify pointers (see `_mi_ptr_cookie`)
uintptr_t keys[2]; // two random keys used to encode the `thread_delayed_free` list
mi_random_ctx_t random; // random number context used for secure allocation
size_t page_count; // total number of pages in the `pages` queues.
size_t page_retired_min; // smallest retired index (retired pages are fully free, but still in the page queues)
size_t page_retired_max; // largest retired index into the `pages` array.
long generic_count; // how often is `_mi_malloc_generic` called?
long generic_collect_count; // how often is `_mi_malloc_generic` called without collecting?
mi_heap_t* next; // list of heaps per thread
bool no_reclaim; // `true` if this heap should not reclaim abandoned pages
uint8_t tag; // custom tag, can be used for separating heaps based on the object types
#if MI_GUARDED
size_t guarded_size_min; // minimal size for guarded objects
size_t guarded_size_max; // maximal size for guarded objects
size_t guarded_sample_rate; // sample rate (set to 0 to disable guarded pages)
size_t guarded_sample_count; // current sample count (counting down to 0)
#endif
mi_page_t* pages_free_direct[MI_PAGES_DIRECT]; // optimize: array where every entry points a page with possibly free blocks in the corresponding queue for that size.
mi_page_queue_t pages[MI_BIN_FULL + 1]; // queue of pages for each size class (or "bin")
};
// ------------------------------------------------------
// Sub processes do not reclaim or visit segments
// from other sub processes. These are essentially the
// static variables of a process.
// ------------------------------------------------------
struct mi_subproc_s {
_Atomic(size_t) abandoned_count; // count of abandoned segments for this sub-process
_Atomic(size_t) abandoned_os_list_count; // count of abandoned segments in the os-list
mi_lock_t abandoned_os_lock; // lock for the abandoned os segment list (outside of arena's) (this lock protect list operations)
mi_lock_t abandoned_os_visit_lock; // ensure only one thread per subproc visits the abandoned os list
mi_segment_t* abandoned_os_list; // doubly-linked list of abandoned segments outside of arena's (in OS allocated memory)
mi_segment_t* abandoned_os_list_tail; // the tail-end of the list
mi_memid_t memid; // provenance of this memory block
};
// ------------------------------------------------------
// Thread Local data
// ------------------------------------------------------
// A "span" is is an available range of slices. The span queues keep
// track of slice spans of at most the given `slice_count` (but more than the previous size class).
typedef struct mi_span_queue_s {
mi_slice_t* first;
mi_slice_t* last;
size_t slice_count;
} mi_span_queue_t;
#define MI_SEGMENT_BIN_MAX (35) // 35 == mi_segment_bin(MI_SLICES_PER_SEGMENT)
// Segments thread local data
typedef struct mi_segments_tld_s {
mi_span_queue_t spans[MI_SEGMENT_BIN_MAX+1]; // free slice spans inside segments
size_t count; // current number of segments;
size_t peak_count; // peak number of segments
size_t current_size; // current size of all segments
size_t peak_size; // peak size of all segments
size_t reclaim_count;// number of reclaimed (abandoned) segments
mi_subproc_t* subproc; // sub-process this thread belongs to.
mi_stats_t* stats; // points to tld stats
} mi_segments_tld_t;
// Thread local data
struct mi_tld_s {
unsigned long long heartbeat; // monotonic heartbeat count
bool recurse; // true if deferred was called; used to prevent infinite recursion.
mi_heap_t* heap_backing; // backing heap of this thread (cannot be deleted)
mi_heap_t* heaps; // list of heaps in this thread (so we can abandon all when the thread terminates)
mi_segments_tld_t segments; // segment tld
mi_stats_t stats; // statistics
};
// ------------------------------------------------------
// Debug
@@ -530,30 +647,10 @@ struct mi_heap_s {
#define MI_DEBUG_PADDING (0xDE)
#endif
#if (MI_DEBUG)
// use our own assertion to print without memory allocation
void _mi_assert_fail(const char* assertion, const char* fname, unsigned int line, const char* func );
#define mi_assert(expr) ((expr) ? (void)0 : _mi_assert_fail(#expr,__FILE__,__LINE__,__func__))
#else
#define mi_assert(x)
#endif
#if (MI_DEBUG>1)
#define mi_assert_internal mi_assert
#else
#define mi_assert_internal(x)
#endif
#if (MI_DEBUG>2)
#define mi_assert_expensive mi_assert
#else
#define mi_assert_expensive(x)
#endif
// ------------------------------------------------------
// Statistics
// ------------------------------------------------------
#ifndef MI_STAT
#if (MI_DEBUG>0)
#define MI_STAT 2
@@ -562,109 +659,28 @@ void _mi_assert_fail(const char* assertion, const char* fname, unsigned int line
#endif
#endif
typedef struct mi_stat_count_s {
int64_t allocated;
int64_t freed;
int64_t peak;
int64_t current;
} mi_stat_count_t;
typedef struct mi_stat_counter_s {
int64_t total;
int64_t count;
} mi_stat_counter_t;
typedef struct mi_stats_s {
mi_stat_count_t segments;
mi_stat_count_t pages;
mi_stat_count_t reserved;
mi_stat_count_t committed;
mi_stat_count_t reset;
mi_stat_count_t purged;
mi_stat_count_t page_committed;
mi_stat_count_t segments_abandoned;
mi_stat_count_t pages_abandoned;
mi_stat_count_t threads;
mi_stat_count_t normal;
mi_stat_count_t huge;
mi_stat_count_t large;
mi_stat_count_t malloc;
mi_stat_count_t segments_cache;
mi_stat_counter_t pages_extended;
mi_stat_counter_t mmap_calls;
mi_stat_counter_t commit_calls;
mi_stat_counter_t reset_calls;
mi_stat_counter_t purge_calls;
mi_stat_counter_t page_no_retire;
mi_stat_counter_t searches;
mi_stat_counter_t normal_count;
mi_stat_counter_t huge_count;
mi_stat_counter_t large_count;
#if MI_STAT>1
mi_stat_count_t normal_bins[MI_BIN_HUGE+1];
#endif
} mi_stats_t;
// add to stat keeping track of the peak
void _mi_stat_increase(mi_stat_count_t* stat, size_t amount);
void _mi_stat_decrease(mi_stat_count_t* stat, size_t amount);
void _mi_stat_adjust_decrease(mi_stat_count_t* stat, size_t amount);
// counters can just be increased
void _mi_stat_counter_increase(mi_stat_counter_t* stat, size_t amount);
#if (MI_STAT)
#define mi_stat_increase(stat,amount) _mi_stat_increase( &(stat), amount)
#define mi_stat_decrease(stat,amount) _mi_stat_decrease( &(stat), amount)
#define mi_stat_adjust_decrease(stat,amount) _mi_stat_adjust_decrease( &(stat), amount)
#define mi_stat_counter_increase(stat,amount) _mi_stat_counter_increase( &(stat), amount)
#else
#define mi_stat_increase(stat,amount) (void)0
#define mi_stat_decrease(stat,amount) (void)0
#define mi_stat_counter_increase(stat,amount) (void)0
#define mi_stat_increase(stat,amount) ((void)0)
#define mi_stat_decrease(stat,amount) ((void)0)
#define mi_stat_adjust_decrease(stat,amount) ((void)0)
#define mi_stat_counter_increase(stat,amount) ((void)0)
#endif
#define mi_heap_stat_counter_increase(heap,stat,amount) mi_stat_counter_increase( (heap)->tld->stats.stat, amount)
#define mi_heap_stat_increase(heap,stat,amount) mi_stat_increase( (heap)->tld->stats.stat, amount)
#define mi_heap_stat_decrease(heap,stat,amount) mi_stat_decrease( (heap)->tld->stats.stat, amount)
// ------------------------------------------------------
// Thread Local data
// ------------------------------------------------------
// A "span" is is an available range of slices. The span queues keep
// track of slice spans of at most the given `slice_count` (but more than the previous size class).
typedef struct mi_span_queue_s {
mi_slice_t* first;
mi_slice_t* last;
size_t slice_count;
} mi_span_queue_t;
#define MI_SEGMENT_BIN_MAX (35) // 35 == mi_segment_bin(MI_SLICES_PER_SEGMENT)
// OS thread local data
typedef struct mi_os_tld_s {
size_t region_idx; // start point for next allocation
mi_stats_t* stats; // points to tld stats
} mi_os_tld_t;
// Segments thread local data
typedef struct mi_segments_tld_s {
mi_span_queue_t spans[MI_SEGMENT_BIN_MAX+1]; // free slice spans inside segments
size_t count; // current number of segments;
size_t peak_count; // peak number of segments
size_t current_size; // current size of all segments
size_t peak_size; // peak size of all segments
mi_stats_t* stats; // points to tld stats
mi_os_tld_t* os; // points to os stats
} mi_segments_tld_t;
// Thread local data
struct mi_tld_s {
unsigned long long heartbeat; // monotonic heartbeat count
bool recurse; // true if deferred was called; used to prevent infinite recursion.
mi_heap_t* heap_backing; // backing heap of this thread (cannot be deleted)
mi_heap_t* heaps; // list of heaps in this thread (so we can abandon all when the thread terminates)
mi_segments_tld_t segments; // segment tld
mi_os_tld_t os; // os tld
mi_stats_t stats; // statistics
};
#define mi_heap_stat_adjust_decrease(heap,stat,amount) mi_stat_adjust_decrease( (heap)->tld->stats.stat, amount)
#endif

215
mimalloc/mimalloc-sys/mimalloc/src/alloc-aligned.c
View File

@@ -15,68 +15,107 @@ terms of the MIT license. A copy of the license can be found in the file
// Aligned Allocation
// ------------------------------------------------------
// Fallback primitive aligned allocation -- split out for better codegen
static mi_decl_noinline void* mi_heap_malloc_zero_aligned_at_fallback(mi_heap_t* const heap, const size_t size, const size_t alignment, const size_t offset, const bool zero) mi_attr_noexcept
static bool mi_malloc_is_naturally_aligned( size_t size, size_t alignment ) {
// objects up to `MI_MAX_ALIGN_GUARANTEE` are allocated aligned to their size (see `segment.c:_mi_segment_page_start`).
mi_assert_internal(_mi_is_power_of_two(alignment) && (alignment > 0));
if (alignment > size) return false;
if (alignment <= MI_MAX_ALIGN_SIZE) return true;
const size_t bsize = mi_good_size(size);
return (bsize <= MI_MAX_ALIGN_GUARANTEE && (bsize & (alignment-1)) == 0);
}
#if MI_GUARDED
static mi_decl_restrict void* mi_heap_malloc_guarded_aligned(mi_heap_t* heap, size_t size, size_t alignment, bool zero) mi_attr_noexcept {
// use over allocation for guarded blocksl
mi_assert_internal(alignment > 0 && alignment < MI_BLOCK_ALIGNMENT_MAX);
const size_t oversize = size + alignment - 1;
void* base = _mi_heap_malloc_guarded(heap, oversize, zero);
void* p = mi_align_up_ptr(base, alignment);
mi_track_align(base, p, (uint8_t*)p - (uint8_t*)base, size);
mi_assert_internal(mi_usable_size(p) >= size);
mi_assert_internal(_mi_is_aligned(p, alignment));
return p;
}
static void* mi_heap_malloc_zero_no_guarded(mi_heap_t* heap, size_t size, bool zero, size_t* usable) {
const size_t rate = heap->guarded_sample_rate;
// only write if `rate!=0` so we don't write to the constant `_mi_heap_empty`
if (rate != 0) { heap->guarded_sample_rate = 0; }
void* p = _mi_heap_malloc_zero_ex(heap, size, zero, 0, usable);
if (rate != 0) { heap->guarded_sample_rate = rate; }
return p;
}
#else
static void* mi_heap_malloc_zero_no_guarded(mi_heap_t* heap, size_t size, bool zero, size_t* usable) {
return _mi_heap_malloc_zero_ex(heap, size, zero, 0, usable);
}
#endif
// Fallback aligned allocation that over-allocates -- split out for better codegen
static mi_decl_noinline void* mi_heap_malloc_zero_aligned_at_overalloc(mi_heap_t* const heap, const size_t size, const size_t alignment, const size_t offset, const bool zero, size_t* usable) mi_attr_noexcept
{
mi_assert_internal(size <= PTRDIFF_MAX);
mi_assert_internal(size <= (MI_MAX_ALLOC_SIZE - MI_PADDING_SIZE));
mi_assert_internal(alignment != 0 && _mi_is_power_of_two(alignment));
const uintptr_t align_mask = alignment - 1; // for any x, `(x & align_mask) == (x % alignment)`
const size_t padsize = size + MI_PADDING_SIZE;
// use regular allocation if it is guaranteed to fit the alignment constraints
if (offset==0 && alignment<=padsize && padsize<=MI_MAX_ALIGN_GUARANTEE && (padsize&align_mask)==0) {
void* p = _mi_heap_malloc_zero(heap, size, zero);
mi_assert_internal(p == NULL || ((uintptr_t)p % alignment) == 0);
return p;
}
void* p;
size_t oversize;
if mi_unlikely(alignment > MI_ALIGNMENT_MAX) {
if mi_unlikely(alignment > MI_BLOCK_ALIGNMENT_MAX) {
// use OS allocation for very large alignment and allocate inside a huge page (dedicated segment with 1 page)
// This can support alignments >= MI_SEGMENT_SIZE by ensuring the object can be aligned at a point in the
// first (and single) page such that the segment info is `MI_SEGMENT_SIZE` bytes before it (so it can be found by aligning the pointer down)
if mi_unlikely(offset != 0) {
// todo: cannot support offset alignment for very large alignments yet
#if MI_DEBUG > 0
#if MI_DEBUG > 0
_mi_error_message(EOVERFLOW, "aligned allocation with a very large alignment cannot be used with an alignment offset (size %zu, alignment %zu, offset %zu)\n", size, alignment, offset);
#endif
#endif
return NULL;
}
oversize = (size <= MI_SMALL_SIZE_MAX ? MI_SMALL_SIZE_MAX + 1 /* ensure we use generic malloc path */ : size);
p = _mi_heap_malloc_zero_ex(heap, oversize, false, alignment); // the page block size should be large enough to align in the single huge page block
// note: no guarded as alignment > 0
p = _mi_heap_malloc_zero_ex(heap, oversize, false, alignment, usable); // the page block size should be large enough to align in the single huge page block
// zero afterwards as only the area from the aligned_p may be committed!
if (p == NULL) return NULL;
if (p == NULL) return NULL;
}
else {
// otherwise over-allocate
oversize = size + alignment - 1;
p = _mi_heap_malloc_zero(heap, oversize, zero);
oversize = (size < MI_MAX_ALIGN_SIZE ? MI_MAX_ALIGN_SIZE : size) + alignment - 1; // adjust for size <= 16; with size 0 and aligment 64k, we would allocate a 64k block and pointing just beyond that.
p = mi_heap_malloc_zero_no_guarded(heap, oversize, zero, usable);
if (p == NULL) return NULL;
}
mi_page_t* page = _mi_ptr_page(p);
// .. and align within the allocation
const uintptr_t align_mask = alignment - 1; // for any x, `(x & align_mask) == (x % alignment)`
const uintptr_t poffset = ((uintptr_t)p + offset) & align_mask;
const uintptr_t adjust = (poffset == 0 ? 0 : alignment - poffset);
mi_assert_internal(adjust < alignment);
void* aligned_p = (void*)((uintptr_t)p + adjust);
if (aligned_p != p) {
mi_page_t* page = _mi_ptr_page(p);
mi_page_set_has_aligned(page, true);
#if MI_GUARDED
// set tag to aligned so mi_usable_size works with guard pages
if (adjust >= sizeof(mi_block_t)) {
mi_block_t* const block = (mi_block_t*)p;
block->next = MI_BLOCK_TAG_ALIGNED;
}
#endif
_mi_padding_shrink(page, (mi_block_t*)p, adjust + size);
}
// todo: expand padding if overallocated ?
mi_assert_internal(mi_page_usable_block_size(_mi_ptr_page(p)) >= adjust + size);
mi_assert_internal(p == _mi_page_ptr_unalign(_mi_ptr_segment(aligned_p), _mi_ptr_page(aligned_p), aligned_p));
mi_assert_internal(mi_page_usable_block_size(page) >= adjust + size);
mi_assert_internal(((uintptr_t)aligned_p + offset) % alignment == 0);
mi_assert_internal(mi_usable_size(aligned_p)>=size);
mi_assert_internal(mi_usable_size(p) == mi_usable_size(aligned_p)+adjust);
#if MI_DEBUG > 1
mi_page_t* const apage = _mi_ptr_page(aligned_p);
void* unalign_p = _mi_page_ptr_unalign(apage, aligned_p);
mi_assert_internal(p == unalign_p);
#endif
// now zero the block if needed
if (alignment > MI_ALIGNMENT_MAX) {
// for the tracker, on huge aligned allocations only from the start of the large block is defined
if (alignment > MI_BLOCK_ALIGNMENT_MAX) {
// for the tracker, on huge aligned allocations only the memory from the start of the large block is defined
mi_track_mem_undefined(aligned_p, size);
if (zero) {
_mi_memzero_aligned(aligned_p, mi_usable_size(aligned_p));
@@ -85,12 +124,51 @@ static mi_decl_noinline void* mi_heap_malloc_zero_aligned_at_fallback(mi_heap_t*
if (p != aligned_p) {
mi_track_align(p,aligned_p,adjust,mi_usable_size(aligned_p));
}
#if MI_GUARDED
mi_track_mem_defined(p, sizeof(mi_block_t));
#endif
}
return aligned_p;
}
// Generic primitive aligned allocation -- split out for better codegen
static mi_decl_noinline void* mi_heap_malloc_zero_aligned_at_generic(mi_heap_t* const heap, const size_t size, const size_t alignment, const size_t offset, const bool zero, size_t* usable) mi_attr_noexcept
{
mi_assert_internal(alignment != 0 && _mi_is_power_of_two(alignment));
// we don't allocate more than MI_MAX_ALLOC_SIZE (see <https://sourceware.org/ml/libc-announce/2019/msg00001.html>)
if mi_unlikely(size > (MI_MAX_ALLOC_SIZE - MI_PADDING_SIZE)) {
#if MI_DEBUG > 0
_mi_error_message(EOVERFLOW, "aligned allocation request is too large (size %zu, alignment %zu)\n", size, alignment);
#endif
return NULL;
}
// use regular allocation if it is guaranteed to fit the alignment constraints.
// this is important to try as the fast path in `mi_heap_malloc_zero_aligned` only works when there exist
// a page with the right block size, and if we always use the over-alloc fallback that would never happen.
if (offset == 0 && mi_malloc_is_naturally_aligned(size,alignment)) {
void* p = mi_heap_malloc_zero_no_guarded(heap, size, zero, usable);
mi_assert_internal(p == NULL || ((uintptr_t)p % alignment) == 0);
const bool is_aligned_or_null = (((uintptr_t)p) & (alignment-1))==0;
if mi_likely(is_aligned_or_null) {
return p;
}
else {
// this should never happen if the `mi_malloc_is_naturally_aligned` check is correct..
mi_assert(false);
mi_free(p);
}
}
// fall back to over-allocation
return mi_heap_malloc_zero_aligned_at_overalloc(heap,size,alignment,offset,zero,usable);
}
// Primitive aligned allocation
static void* mi_heap_malloc_zero_aligned_at(mi_heap_t* const heap, const size_t size, const size_t alignment, const size_t offset, const bool zero) mi_attr_noexcept
static void* mi_heap_malloc_zero_aligned_at(mi_heap_t* const heap, const size_t size,
const size_t alignment, const size_t offset, const bool zero,
size_t* usable) mi_attr_noexcept
{
// note: we don't require `size > offset`, we just guarantee that the address at offset is aligned regardless of the allocated size.
if mi_unlikely(alignment == 0 || !_mi_is_power_of_two(alignment)) { // require power-of-two (see <https://en.cppreference.com/w/c/memory/aligned_alloc>)
@@ -100,33 +178,33 @@ static void* mi_heap_malloc_zero_aligned_at(mi_heap_t* const heap, const size_t
return NULL;
}
if mi_unlikely(size > PTRDIFF_MAX) { // we don't allocate more than PTRDIFF_MAX (see <https://sourceware.org/ml/libc-announce/2019/msg00001.html>)
#if MI_DEBUG > 0
_mi_error_message(EOVERFLOW, "aligned allocation request is too large (size %zu, alignment %zu)\n", size, alignment);
#endif
return NULL;
#if MI_GUARDED
if (offset==0 && alignment < MI_BLOCK_ALIGNMENT_MAX && mi_heap_malloc_use_guarded(heap,size)) {
return mi_heap_malloc_guarded_aligned(heap, size, alignment, zero);
}
const uintptr_t align_mask = alignment-1; // for any x, `(x & align_mask) == (x % alignment)`
const size_t padsize = size + MI_PADDING_SIZE; // note: cannot overflow due to earlier size > PTRDIFF_MAX check
#endif
// try first if there happens to be a small block available with just the right alignment
if mi_likely(padsize <= MI_SMALL_SIZE_MAX && alignment <= padsize) {
if mi_likely(size <= MI_SMALL_SIZE_MAX && alignment <= size) {
const uintptr_t align_mask = alignment-1; // for any x, `(x & align_mask) == (x % alignment)`
const size_t padsize = size + MI_PADDING_SIZE;
mi_page_t* page = _mi_heap_get_free_small_page(heap, padsize);
const bool is_aligned = (((uintptr_t)page->free+offset) & align_mask)==0;
if mi_likely(page->free != NULL && is_aligned)
{
#if MI_STAT>1
mi_heap_stat_increase(heap, malloc, size);
#endif
void* p = _mi_page_malloc(heap, page, padsize, zero); // TODO: inline _mi_page_malloc
mi_assert_internal(p != NULL);
mi_assert_internal(((uintptr_t)p + offset) % alignment == 0);
mi_track_malloc(p,size,zero);
return p;
if mi_likely(page->free != NULL) {
const bool is_aligned = (((uintptr_t)page->free + offset) & align_mask)==0;
if mi_likely(is_aligned)
{
if (usable!=NULL) { *usable = mi_page_usable_block_size(page); }
void* p = (zero ? _mi_page_malloc_zeroed(heap,page,padsize) : _mi_page_malloc(heap,page,padsize)); // call specific page malloc for better codegen
mi_assert_internal(p != NULL);
mi_assert_internal(((uintptr_t)p + offset) % alignment == 0);
mi_track_malloc(p,size,zero);
return p;
}
}
}
// fallback
return mi_heap_malloc_zero_aligned_at_fallback(heap, size, alignment, offset, zero);
// fallback to generic aligned allocation
return mi_heap_malloc_zero_aligned_at_generic(heap, size, alignment, offset, zero, usable);
}
@@ -135,31 +213,16 @@ static void* mi_heap_malloc_zero_aligned_at(mi_heap_t* const heap, const size_t
// ------------------------------------------------------
mi_decl_nodiscard mi_decl_restrict void* mi_heap_malloc_aligned_at(mi_heap_t* heap, size_t size, size_t alignment, size_t offset) mi_attr_noexcept {
return mi_heap_malloc_zero_aligned_at(heap, size, alignment, offset, false);
return mi_heap_malloc_zero_aligned_at(heap, size, alignment, offset, false, NULL);
}
mi_decl_nodiscard mi_decl_restrict void* mi_heap_malloc_aligned(mi_heap_t* heap, size_t size, size_t alignment) mi_attr_noexcept {
if mi_unlikely(alignment == 0 || !_mi_is_power_of_two(alignment)) return NULL;
#if !MI_PADDING
// without padding, any small sized allocation is naturally aligned (see also `_mi_segment_page_start`)
if mi_likely(_mi_is_power_of_two(size) && size >= alignment && size <= MI_SMALL_SIZE_MAX)
#else
// with padding, we can only guarantee this for fixed alignments
if mi_likely((alignment == sizeof(void*) || (alignment == MI_MAX_ALIGN_SIZE && size > (MI_MAX_ALIGN_SIZE/2)))
&& size <= MI_SMALL_SIZE_MAX)
#endif
{
// fast path for common alignment and size
return mi_heap_malloc_small(heap, size);
}
else {
return mi_heap_malloc_aligned_at(heap, size, alignment, 0);
}
return mi_heap_malloc_aligned_at(heap, size, alignment, 0);
}
// ensure a definition is emitted
#if defined(__cplusplus)
static void* _mi_heap_malloc_aligned = (void*)&mi_heap_malloc_aligned;
void* _mi_extern_heap_malloc_aligned = (void*)&mi_heap_malloc_aligned;
#endif
// ------------------------------------------------------
@@ -167,7 +230,7 @@ static void* _mi_heap_malloc_aligned = (void*)&mi_heap_malloc_aligned;
// ------------------------------------------------------
mi_decl_nodiscard mi_decl_restrict void* mi_heap_zalloc_aligned_at(mi_heap_t* heap, size_t size, size_t alignment, size_t offset) mi_attr_noexcept {
return mi_heap_malloc_zero_aligned_at(heap, size, alignment, offset, true);
return mi_heap_malloc_zero_aligned_at(heap, size, alignment, offset, true, NULL);
}
mi_decl_nodiscard mi_decl_restrict void* mi_heap_zalloc_aligned(mi_heap_t* heap, size_t size, size_t alignment) mi_attr_noexcept {
@@ -192,6 +255,10 @@ mi_decl_nodiscard mi_decl_restrict void* mi_malloc_aligned(size_t size, size_t a
return mi_heap_malloc_aligned(mi_prim_get_default_heap(), size, alignment);
}
mi_decl_nodiscard mi_decl_restrict void* mi_umalloc_aligned(size_t size, size_t alignment, size_t* block_size) mi_attr_noexcept {
return mi_heap_malloc_zero_aligned_at(mi_prim_get_default_heap(), size, alignment, 0, false, block_size);
}
mi_decl_nodiscard mi_decl_restrict void* mi_zalloc_aligned_at(size_t size, size_t alignment, size_t offset) mi_attr_noexcept {
return mi_heap_zalloc_aligned_at(mi_prim_get_default_heap(), size, alignment, offset);
}
@@ -200,6 +267,10 @@ mi_decl_nodiscard mi_decl_restrict void* mi_zalloc_aligned(size_t size, size_t a
return mi_heap_zalloc_aligned(mi_prim_get_default_heap(), size, alignment);
}
mi_decl_nodiscard mi_decl_restrict void* mi_uzalloc_aligned(size_t size, size_t alignment, size_t* block_size) mi_attr_noexcept {
return mi_heap_malloc_zero_aligned_at(mi_prim_get_default_heap(), size, alignment, 0, true, block_size);
}
mi_decl_nodiscard mi_decl_restrict void* mi_calloc_aligned_at(size_t count, size_t size, size_t alignment, size_t offset) mi_attr_noexcept {
return mi_heap_calloc_aligned_at(mi_prim_get_default_heap(), count, size, alignment, offset);
}
@@ -215,8 +286,8 @@ mi_decl_nodiscard mi_decl_restrict void* mi_calloc_aligned(size_t count, size_t
static void* mi_heap_realloc_zero_aligned_at(mi_heap_t* heap, void* p, size_t newsize, size_t alignment, size_t offset, bool zero) mi_attr_noexcept {
mi_assert(alignment > 0);
if (alignment <= sizeof(uintptr_t)) return _mi_heap_realloc_zero(heap,p,newsize,zero);
if (p == NULL) return mi_heap_malloc_zero_aligned_at(heap,newsize,alignment,offset,zero);
if (alignment <= sizeof(uintptr_t)) return _mi_heap_realloc_zero(heap,p,newsize,zero,NULL,NULL);
if (p == NULL) return mi_heap_malloc_zero_aligned_at(heap,newsize,alignment,offset,zero,NULL);
size_t size = mi_usable_size(p);
if (newsize <= size && newsize >= (size - (size / 2))
&& (((uintptr_t)p + offset) % alignment) == 0) {
@@ -240,7 +311,7 @@ static void* mi_heap_realloc_zero_aligned_at(mi_heap_t* heap, void* p, size_t ne
static void* mi_heap_realloc_zero_aligned(mi_heap_t* heap, void* p, size_t newsize, size_t alignment, bool zero) mi_attr_noexcept {
mi_assert(alignment > 0);
if (alignment <= sizeof(uintptr_t)) return _mi_heap_realloc_zero(heap,p,newsize,zero);
if (alignment <= sizeof(uintptr_t)) return _mi_heap_realloc_zero(heap,p,newsize,zero,NULL,NULL);
size_t offset = ((uintptr_t)p % alignment); // use offset of previous allocation (p can be NULL)
return mi_heap_realloc_zero_aligned_at(heap,p,newsize,alignment,offset,zero);
}
@@ -296,3 +367,5 @@ mi_decl_nodiscard void* mi_recalloc_aligned_at(void* p, size_t newcount, size_t
mi_decl_nodiscard void* mi_recalloc_aligned(void* p, size_t newcount, size_t size, size_t alignment) mi_attr_noexcept {
return mi_heap_recalloc_aligned(mi_prim_get_default_heap(), p, newcount, size, alignment);
}

45
mimalloc/mimalloc-sys/mimalloc/src/alloc-override.c
View File

@@ -23,7 +23,7 @@ mi_decl_externc size_t malloc_good_size(size_t size);
#endif
// helper definition for C override of C++ new
typedef struct mi_nothrow_s { int _tag; } mi_nothrow_t;
typedef void* mi_nothrow_t;
// ------------------------------------------------------
// Override system malloc
@@ -71,22 +71,20 @@ typedef struct mi_nothrow_s { int _tag; } mi_nothrow_t;
#define MI_INTERPOSE_FUN(oldfun,newfun) { (const void*)&newfun, (const void*)&oldfun }
#define MI_INTERPOSE_MI(fun) MI_INTERPOSE_FUN(fun,mi_##fun)
__attribute__((used)) static struct mi_interpose_s _mi_interposes[] __attribute__((section("__DATA, __interpose"))) =
#define MI_INTERPOSE_DECLS(name) __attribute__((used)) static struct mi_interpose_s name[] __attribute__((section("__DATA, __interpose")))
MI_INTERPOSE_DECLS(_mi_interposes) =
{
MI_INTERPOSE_MI(malloc),
MI_INTERPOSE_MI(calloc),
MI_INTERPOSE_MI(realloc),
MI_INTERPOSE_MI(strdup),
MI_INTERPOSE_MI(strndup),
MI_INTERPOSE_MI(realpath),
MI_INTERPOSE_MI(posix_memalign),
MI_INTERPOSE_MI(reallocf),
MI_INTERPOSE_MI(valloc),
MI_INTERPOSE_FUN(malloc_size,mi_malloc_size_checked),
MI_INTERPOSE_MI(malloc_good_size),
#if defined(MAC_OS_X_VERSION_10_15) && MAC_OS_X_VERSION_MAX_ALLOWED >= MAC_OS_X_VERSION_10_15
MI_INTERPOSE_MI(aligned_alloc),
#endif
#ifdef MI_OSX_ZONE
// we interpose malloc_default_zone in alloc-override-osx.c so we can use mi_free safely
MI_INTERPOSE_MI(free),
@@ -97,6 +95,12 @@ typedef struct mi_nothrow_s { int _tag; } mi_nothrow_t;
MI_INTERPOSE_FUN(vfree,mi_cfree),
#endif
};
MI_INTERPOSE_DECLS(_mi_interposes_10_7) __OSX_AVAILABLE(10.7) = {
MI_INTERPOSE_MI(strndup),
};
MI_INTERPOSE_DECLS(_mi_interposes_10_15) __OSX_AVAILABLE(10.15) = {
MI_INTERPOSE_MI(aligned_alloc),
};
#ifdef __cplusplus
extern "C" {
@@ -128,11 +132,19 @@ typedef struct mi_nothrow_s { int _tag; } mi_nothrow_t;
// cannot override malloc unless using a dll.
// we just override new/delete which does work in a static library.
#else
// On all other systems forward to our API
// On all other systems forward allocation primitives to our API
mi_decl_export void* malloc(size_t size) MI_FORWARD1(mi_malloc, size)
mi_decl_export void* calloc(size_t size, size_t n) MI_FORWARD2(mi_calloc, size, n)
mi_decl_export void* realloc(void* p, size_t newsize) MI_FORWARD2(mi_realloc, p, newsize)
mi_decl_export void free(void* p) MI_FORWARD0(mi_free, p)
mi_decl_export void free(void* p) MI_FORWARD0(mi_free, p)
// In principle we do not need to forward `strdup`/`strndup` but on some systems these do not use `malloc` internally (but a more primitive call)
// We only override if `strdup` is not a macro (as on some older libc's, see issue #885)
#if !defined(strdup)
mi_decl_export char* strdup(const char* str) MI_FORWARD1(mi_strdup, str)
#endif
#if !defined(strndup) && (!defined(__APPLE__) || (defined(MAC_OS_X_VERSION_10_7) && MAC_OS_X_VERSION_MAX_ALLOWED >= MAC_OS_X_VERSION_10_7))
mi_decl_export char* strndup(const char* str, size_t n) MI_FORWARD2(mi_strndup, str, n)
#endif
#endif
#if (defined(__GNUC__) || defined(__clang__)) && !defined(__APPLE__)
@@ -192,11 +204,17 @@ typedef struct mi_nothrow_s { int _tag; } mi_nothrow_t;
void _ZdaPv(void* p) MI_FORWARD0(mi_free,p) // delete[]
void _ZdlPvm(void* p, size_t n) MI_FORWARD02(mi_free_size,p,n)
void _ZdaPvm(void* p, size_t n) MI_FORWARD02(mi_free_size,p,n)
void _ZdlPvSt11align_val_t(void* p, size_t al) { mi_free_aligned(p,al); }
void _ZdaPvSt11align_val_t(void* p, size_t al) { mi_free_aligned(p,al); }
void _ZdlPvmSt11align_val_t(void* p, size_t n, size_t al) { mi_free_size_aligned(p,n,al); }
void _ZdaPvmSt11align_val_t(void* p, size_t n, size_t al) { mi_free_size_aligned(p,n,al); }
void _ZdlPvRKSt9nothrow_t(void* p, mi_nothrow_t tag) { MI_UNUSED(tag); mi_free(p); } // operator delete(void*, std::nothrow_t const&)
void _ZdaPvRKSt9nothrow_t(void* p, mi_nothrow_t tag) { MI_UNUSED(tag); mi_free(p); } // operator delete[](void*, std::nothrow_t const&)
void _ZdlPvSt11align_val_tRKSt9nothrow_t(void* p, size_t al, mi_nothrow_t tag) { MI_UNUSED(tag); mi_free_aligned(p,al); } // operator delete(void*, std::align_val_t, std::nothrow_t const&)
void _ZdaPvSt11align_val_tRKSt9nothrow_t(void* p, size_t al, mi_nothrow_t tag) { MI_UNUSED(tag); mi_free_aligned(p,al); } // operator delete[](void*, std::align_val_t, std::nothrow_t const&)
#if (MI_INTPTR_SIZE==8)
void* _Znwm(size_t n) MI_FORWARD1(mi_new,n) // new 64-bit
void* _Znam(size_t n) MI_FORWARD1(mi_new,n) // new[] 64-bit
@@ -232,7 +250,7 @@ extern "C" {
// Forward Posix/Unix calls as well
void* reallocf(void* p, size_t newsize) MI_FORWARD2(mi_reallocf,p,newsize)
size_t malloc_size(const void* p) MI_FORWARD1(mi_usable_size,p)
#if !defined(__ANDROID__) && !defined(__FreeBSD__)
#if !defined(__ANDROID__) && !defined(__FreeBSD__) && !defined(__DragonFly__)
size_t malloc_usable_size(void *p) MI_FORWARD1(mi_usable_size,p)
#else
size_t malloc_usable_size(const void *p) MI_FORWARD1(mi_usable_size,p)
@@ -259,10 +277,11 @@ extern "C" {
// no forwarding here due to aliasing/name mangling issues
void cfree(void* p) { mi_free(p); }
void* pvalloc(size_t size) { return mi_pvalloc(size); }
void* reallocarray(void* p, size_t count, size_t size) { return mi_reallocarray(p, count, size); }
int reallocarr(void* p, size_t count, size_t size) { return mi_reallocarr(p, count, size); }
void* memalign(size_t alignment, size_t size) { return mi_memalign(alignment, size); }
void* _aligned_malloc(size_t alignment, size_t size) { return mi_aligned_alloc(alignment, size); }
void* reallocarray(void* p, size_t count, size_t size) { return mi_reallocarray(p, count, size); }
// some systems define reallocarr so mark it as a weak symbol (#751)
mi_decl_weak int reallocarr(void* p, size_t count, size_t size) { return mi_reallocarr(p, count, size); }
#if defined(__wasi__)
// forward __libc interface (see PR #667)
@@ -272,8 +291,8 @@ void* _aligned_malloc(size_t alignment, size_t size) { return mi_aligned_allo
void __libc_free(void* p) MI_FORWARD0(mi_free, p)
void* __libc_memalign(size_t alignment, size_t size) { return mi_memalign(alignment, size); }
#elif defined(__GLIBC__) && defined(__linux__)
// forward __libc interface (needed for glibc-based Linux distributions)
#elif defined(__linux__)
// forward __libc interface (needed for glibc-based and musl-based Linux distributions)
void* __libc_malloc(size_t size) MI_FORWARD1(mi_malloc,size)
void* __libc_calloc(size_t count, size_t size) MI_FORWARD2(mi_calloc,count,size)
void* __libc_realloc(void* p, size_t size) MI_FORWARD2(mi_realloc,p,size)

787
mimalloc/mimalloc-sys/mimalloc/src/alloc.c
View File

@@ -1,5 +1,5 @@
/* ----------------------------------------------------------------------------
Copyright (c) 2018-2022, Microsoft Research, Daan Leijen
Copyright (c) 2018-2024, Microsoft Research, Daan Leijen
This is free software; you can redistribute it and/or modify it under the
terms of the MIT license. A copy of the license can be found in the file
"LICENSE" at the root of this distribution.
@@ -18,6 +18,7 @@ terms of the MIT license. A copy of the license can be found in the file
#define MI_IN_ALLOC_C
#include "alloc-override.c"
#include "free.c"
#undef MI_IN_ALLOC_C
// ------------------------------------------------------
@@ -26,19 +27,27 @@ terms of the MIT license. A copy of the license can be found in the file
// Fast allocation in a page: just pop from the free list.
// Fall back to generic allocation only if the list is empty.
extern inline void* _mi_page_malloc(mi_heap_t* heap, mi_page_t* page, size_t size, bool zero) mi_attr_noexcept {
mi_assert_internal(page->xblock_size==0||mi_page_block_size(page) >= size);
// Note: in release mode the (inlined) routine is about 7 instructions with a single test.
extern inline void* _mi_page_malloc_zero(mi_heap_t* heap, mi_page_t* page, size_t size, bool zero, size_t* usable) mi_attr_noexcept
{
mi_assert_internal(size >= MI_PADDING_SIZE);
mi_assert_internal(page->block_size == 0 /* empty heap */ || mi_page_block_size(page) >= size);
// check the free list
mi_block_t* const block = page->free;
if mi_unlikely(block == NULL) {
return _mi_malloc_generic(heap, size, zero, 0);
return _mi_malloc_generic(heap, size, zero, 0, usable);
}
mi_assert_internal(block != NULL && _mi_ptr_page(block) == page);
if (usable != NULL) { *usable = mi_page_usable_block_size(page); };
// pop from the free list
page->used++;
page->free = mi_block_next(page, block);
page->used++;
mi_assert_internal(page->free == NULL || _mi_ptr_page(page->free) == page);
mi_assert_internal(page->block_size < MI_MAX_ALIGN_SIZE || _mi_is_aligned(block, MI_MAX_ALIGN_SIZE));
#if MI_DEBUG>3
if (page->free_is_zero) {
if (page->free_is_zero && size > sizeof(*block)) {
mi_assert_expensive(mi_mem_is_zero(block+1,size - sizeof(*block)));
}
#endif
@@ -50,77 +59,95 @@ extern inline void* _mi_page_malloc(mi_heap_t* heap, mi_page_t* page, size_t siz
// zero the block? note: we need to zero the full block size (issue #63)
if mi_unlikely(zero) {
mi_assert_internal(page->xblock_size != 0); // do not call with zero'ing for huge blocks (see _mi_malloc_generic)
mi_assert_internal(page->xblock_size >= MI_PADDING_SIZE);
mi_assert_internal(page->block_size != 0); // do not call with zero'ing for huge blocks (see _mi_malloc_generic)
mi_assert_internal(!mi_page_is_huge(page));
#if MI_PADDING
mi_assert_internal(page->block_size >= MI_PADDING_SIZE);
#endif
if (page->free_is_zero) {
block->next = 0;
mi_track_mem_defined(block, page->xblock_size - MI_PADDING_SIZE);
mi_track_mem_defined(block, page->block_size - MI_PADDING_SIZE);
}
else {
_mi_memzero_aligned(block, page->xblock_size - MI_PADDING_SIZE);
}
_mi_memzero_aligned(block, page->block_size - MI_PADDING_SIZE);
}
}
#if (MI_DEBUG>0) && !MI_TRACK_ENABLED && !MI_TSAN
#if (MI_DEBUG>0) && !MI_TRACK_ENABLED && !MI_TSAN
if (!zero && !mi_page_is_huge(page)) {
memset(block, MI_DEBUG_UNINIT, mi_page_usable_block_size(page));
}
#elif (MI_SECURE!=0)
#elif (MI_SECURE!=0)
if (!zero) { block->next = 0; } // don't leak internal data
#endif
#endif
#if (MI_STAT>0)
#if (MI_STAT>0)
const size_t bsize = mi_page_usable_block_size(page);
if (bsize <= MI_MEDIUM_OBJ_SIZE_MAX) {
mi_heap_stat_increase(heap, normal, bsize);
mi_heap_stat_counter_increase(heap, normal_count, 1);
#if (MI_STAT>1)
mi_heap_stat_increase(heap, malloc_normal, bsize);
mi_heap_stat_counter_increase(heap, malloc_normal_count, 1);
#if (MI_STAT>1)
const size_t bin = _mi_bin(bsize);
mi_heap_stat_increase(heap, normal_bins[bin], 1);
#endif
mi_heap_stat_increase(heap, malloc_bins[bin], 1);
mi_heap_stat_increase(heap, malloc_requested, size - MI_PADDING_SIZE);
#endif
}
#endif
#endif
#if MI_PADDING // && !MI_TRACK_ENABLED
mi_padding_t* const padding = (mi_padding_t*)((uint8_t*)block + mi_page_usable_block_size(page));
ptrdiff_t delta = ((uint8_t*)padding - (uint8_t*)block - (size - MI_PADDING_SIZE));
#if (MI_DEBUG>=2)
mi_assert_internal(delta >= 0 && mi_page_usable_block_size(page) >= (size - MI_PADDING_SIZE + delta));
#if MI_PADDING // && !MI_TRACK_ENABLED
mi_padding_t* const padding = (mi_padding_t*)((uint8_t*)block + mi_page_usable_block_size(page));
ptrdiff_t delta = ((uint8_t*)padding - (uint8_t*)block - (size - MI_PADDING_SIZE));
#if (MI_DEBUG>=2)
mi_assert_internal(delta >= 0 && mi_page_usable_block_size(page) >= (size - MI_PADDING_SIZE + delta));
#endif
mi_track_mem_defined(padding,sizeof(mi_padding_t)); // note: re-enable since mi_page_usable_block_size may set noaccess
padding->canary = mi_ptr_encode_canary(page,block,page->keys);
padding->delta = (uint32_t)(delta);
#if MI_PADDING_CHECK
if (!mi_page_is_huge(page)) {
uint8_t* fill = (uint8_t*)padding - delta;
const size_t maxpad = (delta > MI_MAX_ALIGN_SIZE ? MI_MAX_ALIGN_SIZE : delta); // set at most N initial padding bytes
for (size_t i = 0; i < maxpad; i++) { fill[i] = MI_DEBUG_PADDING; }
}
#endif
#endif
mi_track_mem_defined(padding,sizeof(mi_padding_t)); // note: re-enable since mi_page_usable_block_size may set noaccess
padding->canary = (uint32_t)(mi_ptr_encode(page,block,page->keys));
padding->delta = (uint32_t)(delta);
#if MI_PADDING_CHECK
if (!mi_page_is_huge(page)) {
uint8_t* fill = (uint8_t*)padding - delta;
const size_t maxpad = (delta > MI_MAX_ALIGN_SIZE ? MI_MAX_ALIGN_SIZE : delta); // set at most N initial padding bytes
for (size_t i = 0; i < maxpad; i++) { fill[i] = MI_DEBUG_PADDING; }
}
#endif
#endif
return block;
}
static inline mi_decl_restrict void* mi_heap_malloc_small_zero(mi_heap_t* heap, size_t size, bool zero) mi_attr_noexcept {
// extra entries for improved efficiency in `alloc-aligned.c`.
extern void* _mi_page_malloc(mi_heap_t* heap, mi_page_t* page, size_t size) mi_attr_noexcept {
return _mi_page_malloc_zero(heap,page,size,false,NULL);
}
extern void* _mi_page_malloc_zeroed(mi_heap_t* heap, mi_page_t* page, size_t size) mi_attr_noexcept {
return _mi_page_malloc_zero(heap,page,size,true,NULL);
}
#if MI_GUARDED
mi_decl_restrict void* _mi_heap_malloc_guarded(mi_heap_t* heap, size_t size, bool zero) mi_attr_noexcept;
#endif
static inline mi_decl_restrict void* mi_heap_malloc_small_zero(mi_heap_t* heap, size_t size, bool zero, size_t* usable) mi_attr_noexcept {
mi_assert(heap != NULL);
mi_assert(size <= MI_SMALL_SIZE_MAX);
#if MI_DEBUG
const uintptr_t tid = _mi_thread_id();
mi_assert(heap->thread_id == 0 || heap->thread_id == tid); // heaps are thread local
#endif
mi_assert(size <= MI_SMALL_SIZE_MAX);
#if (MI_PADDING)
#if (MI_PADDING || MI_GUARDED)
if (size == 0) { size = sizeof(void*); }
#endif
mi_page_t* page = _mi_heap_get_free_small_page(heap, size + MI_PADDING_SIZE);
void* const p = _mi_page_malloc(heap, page, size + MI_PADDING_SIZE, zero);
mi_track_malloc(p,size,zero);
#if MI_STAT>1
if (p != NULL) {
if (!mi_heap_is_initialized(heap)) { heap = mi_prim_get_default_heap(); }
mi_heap_stat_increase(heap, malloc, mi_usable_size(p));
#if MI_GUARDED
if (mi_heap_malloc_use_guarded(heap,size)) {
return _mi_heap_malloc_guarded(heap, size, zero);
}
#endif
// get page in constant time, and allocate from it
mi_page_t* page = _mi_heap_get_free_small_page(heap, size + MI_PADDING_SIZE);
void* const p = _mi_page_malloc_zero(heap, page, size + MI_PADDING_SIZE, zero, usable);
mi_track_malloc(p,size,zero);
#if MI_DEBUG>3
if (p != NULL && zero) {
mi_assert_expensive(mi_mem_is_zero(p, size));
@@ -131,7 +158,7 @@ static inline mi_decl_restrict void* mi_heap_malloc_small_zero(mi_heap_t* heap,
// allocate a small block
mi_decl_nodiscard extern inline mi_decl_restrict void* mi_heap_malloc_small(mi_heap_t* heap, size_t size) mi_attr_noexcept {
return mi_heap_malloc_small_zero(heap, size, false);
return mi_heap_malloc_small_zero(heap, size, false, NULL);
}
mi_decl_nodiscard extern inline mi_decl_restrict void* mi_malloc_small(size_t size) mi_attr_noexcept {
@@ -139,22 +166,24 @@ mi_decl_nodiscard extern inline mi_decl_restrict void* mi_malloc_small(size_t si
}
// The main allocation function
extern inline void* _mi_heap_malloc_zero_ex(mi_heap_t* heap, size_t size, bool zero, size_t huge_alignment) mi_attr_noexcept {
extern inline void* _mi_heap_malloc_zero_ex(mi_heap_t* heap, size_t size, bool zero, size_t huge_alignment, size_t* usable) mi_attr_noexcept {
// fast path for small objects
if mi_likely(size <= MI_SMALL_SIZE_MAX) {
mi_assert_internal(huge_alignment == 0);
return mi_heap_malloc_small_zero(heap, size, zero);
return mi_heap_malloc_small_zero(heap, size, zero, usable);
}
#if MI_GUARDED
else if (huge_alignment==0 && mi_heap_malloc_use_guarded(heap,size)) {
return _mi_heap_malloc_guarded(heap, size, zero);
}
#endif
else {
// regular allocation
mi_assert(heap!=NULL);
mi_assert(heap->thread_id == 0 || heap->thread_id == _mi_thread_id()); // heaps are thread local
void* const p = _mi_malloc_generic(heap, size + MI_PADDING_SIZE, zero, huge_alignment); // note: size can overflow but it is detected in malloc_generic
void* const p = _mi_malloc_generic(heap, size + MI_PADDING_SIZE, zero, huge_alignment, usable); // note: size can overflow but it is detected in malloc_generic
mi_track_malloc(p,size,zero);
#if MI_STAT>1
if (p != NULL) {
if (!mi_heap_is_initialized(heap)) { heap = mi_prim_get_default_heap(); }
mi_heap_stat_increase(heap, malloc, mi_usable_size(p));
}
#endif
#if MI_DEBUG>3
if (p != NULL && zero) {
mi_assert_expensive(mi_mem_is_zero(p, size));
@@ -165,7 +194,7 @@ extern inline void* _mi_heap_malloc_zero_ex(mi_heap_t* heap, size_t size, bool z
}
extern inline void* _mi_heap_malloc_zero(mi_heap_t* heap, size_t size, bool zero) mi_attr_noexcept {
return _mi_heap_malloc_zero_ex(heap, size, zero, 0);
return _mi_heap_malloc_zero_ex(heap, size, zero, 0, NULL);
}
mi_decl_nodiscard extern inline mi_decl_restrict void* mi_heap_malloc(mi_heap_t* heap, size_t size) mi_attr_noexcept {
@@ -178,7 +207,7 @@ mi_decl_nodiscard extern inline mi_decl_restrict void* mi_malloc(size_t size) mi
// zero initialized small block
mi_decl_nodiscard mi_decl_restrict void* mi_zalloc_small(size_t size) mi_attr_noexcept {
return mi_heap_malloc_small_zero(mi_prim_get_default_heap(), size, true);
return mi_heap_malloc_small_zero(mi_prim_get_default_heap(), size, true, NULL);
}
mi_decl_nodiscard extern inline mi_decl_restrict void* mi_heap_zalloc(mi_heap_t* heap, size_t size) mi_attr_noexcept {
@@ -190,484 +219,6 @@ mi_decl_nodiscard mi_decl_restrict void* mi_zalloc(size_t size) mi_attr_noexcept
}
// ------------------------------------------------------
// Check for double free in secure and debug mode
// This is somewhat expensive so only enabled for secure mode 4
// ------------------------------------------------------
#if (MI_ENCODE_FREELIST && (MI_SECURE>=4 || MI_DEBUG!=0))
// linear check if the free list contains a specific element
static bool mi_list_contains(const mi_page_t* page, const mi_block_t* list, const mi_block_t* elem) {
while (list != NULL) {
if (elem==list) return true;
list = mi_block_next(page, list);
}
return false;
}
static mi_decl_noinline bool mi_check_is_double_freex(const mi_page_t* page, const mi_block_t* block) {
// The decoded value is in the same page (or NULL).
// Walk the free lists to verify positively if it is already freed
if (mi_list_contains(page, page->free, block) ||
mi_list_contains(page, page->local_free, block) ||
mi_list_contains(page, mi_page_thread_free(page), block))
{
_mi_error_message(EAGAIN, "double free detected of block %p with size %zu\n", block, mi_page_block_size(page));
return true;
}
return false;
}
#define mi_track_page(page,access) { size_t psize; void* pstart = _mi_page_start(_mi_page_segment(page),page,&psize); mi_track_mem_##access( pstart, psize); }
static inline bool mi_check_is_double_free(const mi_page_t* page, const mi_block_t* block) {
bool is_double_free = false;
mi_block_t* n = mi_block_nextx(page, block, page->keys); // pretend it is freed, and get the decoded first field
if (((uintptr_t)n & (MI_INTPTR_SIZE-1))==0 && // quick check: aligned pointer?
(n==NULL || mi_is_in_same_page(block, n))) // quick check: in same page or NULL?
{
// Suspicous: decoded value a in block is in the same page (or NULL) -- maybe a double free?
// (continue in separate function to improve code generation)
is_double_free = mi_check_is_double_freex(page, block);
}
return is_double_free;
}
#else
static inline bool mi_check_is_double_free(const mi_page_t* page, const mi_block_t* block) {
MI_UNUSED(page);
MI_UNUSED(block);
return false;
}
#endif
// ---------------------------------------------------------------------------
// Check for heap block overflow by setting up padding at the end of the block
// ---------------------------------------------------------------------------
#if MI_PADDING // && !MI_TRACK_ENABLED
static bool mi_page_decode_padding(const mi_page_t* page, const mi_block_t* block, size_t* delta, size_t* bsize) {
*bsize = mi_page_usable_block_size(page);
const mi_padding_t* const padding = (mi_padding_t*)((uint8_t*)block + *bsize);
mi_track_mem_defined(padding,sizeof(mi_padding_t));
*delta = padding->delta;
uint32_t canary = padding->canary;
uintptr_t keys[2];
keys[0] = page->keys[0];
keys[1] = page->keys[1];
bool ok = ((uint32_t)mi_ptr_encode(page,block,keys) == canary && *delta <= *bsize);
mi_track_mem_noaccess(padding,sizeof(mi_padding_t));
return ok;
}
// Return the exact usable size of a block.
static size_t mi_page_usable_size_of(const mi_page_t* page, const mi_block_t* block) {
size_t bsize;
size_t delta;
bool ok = mi_page_decode_padding(page, block, &delta, &bsize);
mi_assert_internal(ok); mi_assert_internal(delta <= bsize);
return (ok ? bsize - delta : 0);
}
// When a non-thread-local block is freed, it becomes part of the thread delayed free
// list that is freed later by the owning heap. If the exact usable size is too small to
// contain the pointer for the delayed list, then shrink the padding (by decreasing delta)
// so it will later not trigger an overflow error in `mi_free_block`.
void _mi_padding_shrink(const mi_page_t* page, const mi_block_t* block, const size_t min_size) {
size_t bsize;
size_t delta;
bool ok = mi_page_decode_padding(page, block, &delta, &bsize);
mi_assert_internal(ok);
if (!ok || (bsize - delta) >= min_size) return; // usually already enough space
mi_assert_internal(bsize >= min_size);
if (bsize < min_size) return; // should never happen
size_t new_delta = (bsize - min_size);
mi_assert_internal(new_delta < bsize);
mi_padding_t* padding = (mi_padding_t*)((uint8_t*)block + bsize);
mi_track_mem_defined(padding,sizeof(mi_padding_t));
padding->delta = (uint32_t)new_delta;
mi_track_mem_noaccess(padding,sizeof(mi_padding_t));
}
#else
static size_t mi_page_usable_size_of(const mi_page_t* page, const mi_block_t* block) {
MI_UNUSED(block);
return mi_page_usable_block_size(page);
}
void _mi_padding_shrink(const mi_page_t* page, const mi_block_t* block, const size_t min_size) {
MI_UNUSED(page);
MI_UNUSED(block);
MI_UNUSED(min_size);
}
#endif
#if MI_PADDING && MI_PADDING_CHECK
static bool mi_verify_padding(const mi_page_t* page, const mi_block_t* block, size_t* size, size_t* wrong) {
size_t bsize;
size_t delta;
bool ok = mi_page_decode_padding(page, block, &delta, &bsize);
*size = *wrong = bsize;
if (!ok) return false;
mi_assert_internal(bsize >= delta);
*size = bsize - delta;
if (!mi_page_is_huge(page)) {
uint8_t* fill = (uint8_t*)block + bsize - delta;
const size_t maxpad = (delta > MI_MAX_ALIGN_SIZE ? MI_MAX_ALIGN_SIZE : delta); // check at most the first N padding bytes
mi_track_mem_defined(fill, maxpad);
for (size_t i = 0; i < maxpad; i++) {
if (fill[i] != MI_DEBUG_PADDING) {
*wrong = bsize - delta + i;
ok = false;
break;
}
}
mi_track_mem_noaccess(fill, maxpad);
}
return ok;
}
static void mi_check_padding(const mi_page_t* page, const mi_block_t* block) {
size_t size;
size_t wrong;
if (!mi_verify_padding(page,block,&size,&wrong)) {
_mi_error_message(EFAULT, "buffer overflow in heap block %p of size %zu: write after %zu bytes\n", block, size, wrong );
}
}
#else
static void mi_check_padding(const mi_page_t* page, const mi_block_t* block) {
MI_UNUSED(page);
MI_UNUSED(block);
}
#endif
// only maintain stats for smaller objects if requested
#if (MI_STAT>0)
static void mi_stat_free(const mi_page_t* page, const mi_block_t* block) {
#if (MI_STAT < 2)
MI_UNUSED(block);
#endif
mi_heap_t* const heap = mi_heap_get_default();
const size_t bsize = mi_page_usable_block_size(page);
#if (MI_STAT>1)
const size_t usize = mi_page_usable_size_of(page, block);
mi_heap_stat_decrease(heap, malloc, usize);
#endif
if (bsize <= MI_MEDIUM_OBJ_SIZE_MAX) {
mi_heap_stat_decrease(heap, normal, bsize);
#if (MI_STAT > 1)
mi_heap_stat_decrease(heap, normal_bins[_mi_bin(bsize)], 1);
#endif
}
else if (bsize <= MI_LARGE_OBJ_SIZE_MAX) {
mi_heap_stat_decrease(heap, large, bsize);
}
else {
mi_heap_stat_decrease(heap, huge, bsize);
}
}
#else
static void mi_stat_free(const mi_page_t* page, const mi_block_t* block) {
MI_UNUSED(page); MI_UNUSED(block);
}
#endif
#if MI_HUGE_PAGE_ABANDON
#if (MI_STAT>0)
// maintain stats for huge objects
static void mi_stat_huge_free(const mi_page_t* page) {
mi_heap_t* const heap = mi_heap_get_default();
const size_t bsize = mi_page_block_size(page); // to match stats in `page.c:mi_page_huge_alloc`
if (bsize <= MI_LARGE_OBJ_SIZE_MAX) {
mi_heap_stat_decrease(heap, large, bsize);
}
else {
mi_heap_stat_decrease(heap, huge, bsize);
}
}
#else
static void mi_stat_huge_free(const mi_page_t* page) {
MI_UNUSED(page);
}
#endif
#endif
// ------------------------------------------------------
// Free
// ------------------------------------------------------
// multi-threaded free (or free in huge block if compiled with MI_HUGE_PAGE_ABANDON)
static mi_decl_noinline void _mi_free_block_mt(mi_page_t* page, mi_block_t* block)
{
// The padding check may access the non-thread-owned page for the key values.
// that is safe as these are constant and the page won't be freed (as the block is not freed yet).
mi_check_padding(page, block);
_mi_padding_shrink(page, block, sizeof(mi_block_t)); // for small size, ensure we can fit the delayed thread pointers without triggering overflow detection
// huge page segments are always abandoned and can be freed immediately
mi_segment_t* segment = _mi_page_segment(page);
if (segment->kind == MI_SEGMENT_HUGE) {
#if MI_HUGE_PAGE_ABANDON
// huge page segments are always abandoned and can be freed immediately
mi_stat_huge_free(page);
_mi_segment_huge_page_free(segment, page, block);
return;
#else
// huge pages are special as they occupy the entire segment
// as these are large we reset the memory occupied by the page so it is available to other threads
// (as the owning thread needs to actually free the memory later).
_mi_segment_huge_page_reset(segment, page, block);
#endif
}
#if (MI_DEBUG>0) && !MI_TRACK_ENABLED && !MI_TSAN // note: when tracking, cannot use mi_usable_size with multi-threading
if (segment->kind != MI_SEGMENT_HUGE) { // not for huge segments as we just reset the content
memset(block, MI_DEBUG_FREED, mi_usable_size(block));
}
#endif
// Try to put the block on either the page-local thread free list, or the heap delayed free list.
mi_thread_free_t tfreex;
bool use_delayed;
mi_thread_free_t tfree = mi_atomic_load_relaxed(&page->xthread_free);
do {
use_delayed = (mi_tf_delayed(tfree) == MI_USE_DELAYED_FREE);
if mi_unlikely(use_delayed) {
// unlikely: this only happens on the first concurrent free in a page that is in the full list
tfreex = mi_tf_set_delayed(tfree,MI_DELAYED_FREEING);
}
else {
// usual: directly add to page thread_free list
mi_block_set_next(page, block, mi_tf_block(tfree));
tfreex = mi_tf_set_block(tfree,block);
}
} while (!mi_atomic_cas_weak_release(&page->xthread_free, &tfree, tfreex));
if mi_unlikely(use_delayed) {
// racy read on `heap`, but ok because MI_DELAYED_FREEING is set (see `mi_heap_delete` and `mi_heap_collect_abandon`)
mi_heap_t* const heap = (mi_heap_t*)(mi_atomic_load_acquire(&page->xheap)); //mi_page_heap(page);
mi_assert_internal(heap != NULL);
if (heap != NULL) {
// add to the delayed free list of this heap. (do this atomically as the lock only protects heap memory validity)
mi_block_t* dfree = mi_atomic_load_ptr_relaxed(mi_block_t, &heap->thread_delayed_free);
do {
mi_block_set_nextx(heap,block,dfree, heap->keys);
} while (!mi_atomic_cas_ptr_weak_release(mi_block_t,&heap->thread_delayed_free, &dfree, block));
}
// and reset the MI_DELAYED_FREEING flag
tfree = mi_atomic_load_relaxed(&page->xthread_free);
do {
tfreex = tfree;
mi_assert_internal(mi_tf_delayed(tfree) == MI_DELAYED_FREEING);
tfreex = mi_tf_set_delayed(tfree,MI_NO_DELAYED_FREE);
} while (!mi_atomic_cas_weak_release(&page->xthread_free, &tfree, tfreex));
}
}
// regular free
static inline void _mi_free_block(mi_page_t* page, bool local, mi_block_t* block)
{
// and push it on the free list
//const size_t bsize = mi_page_block_size(page);
if mi_likely(local) {
// owning thread can free a block directly
if mi_unlikely(mi_check_is_double_free(page, block)) return;
mi_check_padding(page, block);
#if (MI_DEBUG>0) && !MI_TRACK_ENABLED && !MI_TSAN
if (!mi_page_is_huge(page)) { // huge page content may be already decommitted
memset(block, MI_DEBUG_FREED, mi_page_block_size(page));
}
#endif
mi_block_set_next(page, block, page->local_free);
page->local_free = block;
page->used--;
if mi_unlikely(mi_page_all_free(page)) {
_mi_page_retire(page);
}
else if mi_unlikely(mi_page_is_in_full(page)) {
_mi_page_unfull(page);
}
}
else {
_mi_free_block_mt(page,block);
}
}
// Adjust a block that was allocated aligned, to the actual start of the block in the page.
mi_block_t* _mi_page_ptr_unalign(const mi_segment_t* segment, const mi_page_t* page, const void* p) {
mi_assert_internal(page!=NULL && p!=NULL);
const size_t diff = (uint8_t*)p - _mi_page_start(segment, page, NULL);
const size_t adjust = (diff % mi_page_block_size(page));
return (mi_block_t*)((uintptr_t)p - adjust);
}
void mi_decl_noinline _mi_free_generic(const mi_segment_t* segment, mi_page_t* page, bool is_local, void* p) mi_attr_noexcept {
mi_block_t* const block = (mi_page_has_aligned(page) ? _mi_page_ptr_unalign(segment, page, p) : (mi_block_t*)p);
mi_stat_free(page, block); // stat_free may access the padding
mi_track_free_size(block, mi_page_usable_size_of(page,block));
_mi_free_block(page, is_local, block);
}
// Get the segment data belonging to a pointer
// This is just a single `and` in assembly but does further checks in debug mode
// (and secure mode) if this was a valid pointer.
static inline mi_segment_t* mi_checked_ptr_segment(const void* p, const char* msg)
{
MI_UNUSED(msg);
mi_assert(p != NULL);
#if (MI_DEBUG>0)
if mi_unlikely(((uintptr_t)p & (MI_INTPTR_SIZE - 1)) != 0) {
_mi_error_message(EINVAL, "%s: invalid (unaligned) pointer: %p\n", msg, p);
return NULL;
}
#endif
mi_segment_t* const segment = _mi_ptr_segment(p);
mi_assert_internal(segment != NULL);
#if (MI_DEBUG>0)
if mi_unlikely(!mi_is_in_heap_region(p)) {
#if (MI_INTPTR_SIZE == 8 && defined(__linux__))
if (((uintptr_t)p >> 40) != 0x7F) { // linux tends to align large blocks above 0x7F000000000 (issue #640)
#else
{
#endif
_mi_warning_message("%s: pointer might not point to a valid heap region: %p\n"
"(this may still be a valid very large allocation (over 64MiB))\n", msg, p);
if mi_likely(_mi_ptr_cookie(segment) == segment->cookie) {
_mi_warning_message("(yes, the previous pointer %p was valid after all)\n", p);
}
}
}
#endif
#if (MI_DEBUG>0 || MI_SECURE>=4)
if mi_unlikely(_mi_ptr_cookie(segment) != segment->cookie) {
_mi_error_message(EINVAL, "%s: pointer does not point to a valid heap space: %p\n", msg, p);
return NULL;
}
#endif
return segment;
}
// Free a block
// fast path written carefully to prevent spilling on the stack
void mi_free(void* p) mi_attr_noexcept
{
if mi_unlikely(p == NULL) return;
mi_segment_t* const segment = mi_checked_ptr_segment(p,"mi_free");
const bool is_local= (_mi_prim_thread_id() == mi_atomic_load_relaxed(&segment->thread_id));
mi_page_t* const page = _mi_segment_page_of(segment, p);
if mi_likely(is_local) { // thread-local free?
if mi_likely(page->flags.full_aligned == 0) // and it is not a full page (full pages need to move from the full bin), nor has aligned blocks (aligned blocks need to be unaligned)
{
mi_block_t* const block = (mi_block_t*)p;
if mi_unlikely(mi_check_is_double_free(page, block)) return;
mi_check_padding(page, block);
mi_stat_free(page, block);
#if (MI_DEBUG>0) && !MI_TRACK_ENABLED && !MI_TSAN
memset(block, MI_DEBUG_FREED, mi_page_block_size(page));
#endif
mi_track_free_size(p, mi_page_usable_size_of(page,block)); // faster then mi_usable_size as we already know the page and that p is unaligned
mi_block_set_next(page, block, page->local_free);
page->local_free = block;
if mi_unlikely(--page->used == 0) { // using this expression generates better code than: page->used--; if (mi_page_all_free(page))
_mi_page_retire(page);
}
}
else {
// page is full or contains (inner) aligned blocks; use generic path
_mi_free_generic(segment, page, true, p);
}
}
else {
// not thread-local; use generic path
_mi_free_generic(segment, page, false, p);
}
}
// return true if successful
bool _mi_free_delayed_block(mi_block_t* block) {
// get segment and page
const mi_segment_t* const segment = _mi_ptr_segment(block);
mi_assert_internal(_mi_ptr_cookie(segment) == segment->cookie);
mi_assert_internal(_mi_thread_id() == segment->thread_id);
mi_page_t* const page = _mi_segment_page_of(segment, block);
// Clear the no-delayed flag so delayed freeing is used again for this page.
// This must be done before collecting the free lists on this page -- otherwise
// some blocks may end up in the page `thread_free` list with no blocks in the
// heap `thread_delayed_free` list which may cause the page to be never freed!
// (it would only be freed if we happen to scan it in `mi_page_queue_find_free_ex`)
if (!_mi_page_try_use_delayed_free(page, MI_USE_DELAYED_FREE, false /* dont overwrite never delayed */)) {
return false;
}
// collect all other non-local frees to ensure up-to-date `used` count
_mi_page_free_collect(page, false);
// and free the block (possibly freeing the page as well since used is updated)
_mi_free_block(page, true, block);
return true;
}
// Bytes available in a block
mi_decl_noinline static size_t mi_page_usable_aligned_size_of(const mi_segment_t* segment, const mi_page_t* page, const void* p) mi_attr_noexcept {
const mi_block_t* block = _mi_page_ptr_unalign(segment, page, p);
const size_t size = mi_page_usable_size_of(page, block);
const ptrdiff_t adjust = (uint8_t*)p - (uint8_t*)block;
mi_assert_internal(adjust >= 0 && (size_t)adjust <= size);
return (size - adjust);
}
static inline size_t _mi_usable_size(const void* p, const char* msg) mi_attr_noexcept {
if (p == NULL) return 0;
const mi_segment_t* const segment = mi_checked_ptr_segment(p, msg);
const mi_page_t* const page = _mi_segment_page_of(segment, p);
if mi_likely(!mi_page_has_aligned(page)) {
const mi_block_t* block = (const mi_block_t*)p;
return mi_page_usable_size_of(page, block);
}
else {
// split out to separate routine for improved code generation
return mi_page_usable_aligned_size_of(segment, page, p);
}
}
mi_decl_nodiscard size_t mi_usable_size(const void* p) mi_attr_noexcept {
return _mi_usable_size(p, "mi_usable_size");
}
// ------------------------------------------------------
// Allocation extensions
// ------------------------------------------------------
void mi_free_size(void* p, size_t size) mi_attr_noexcept {
MI_UNUSED_RELEASE(size);
mi_assert(p == NULL || size <= _mi_usable_size(p,"mi_free_size"));
mi_free(p);
}
void mi_free_size_aligned(void* p, size_t size, size_t alignment) mi_attr_noexcept {
MI_UNUSED_RELEASE(alignment);
mi_assert(((uintptr_t)p % alignment) == 0);
mi_free_size(p,size);
}
void mi_free_aligned(void* p, size_t alignment) mi_attr_noexcept {
MI_UNUSED_RELEASE(alignment);
mi_assert(((uintptr_t)p % alignment) == 0);
mi_free(p);
}
mi_decl_nodiscard extern inline mi_decl_restrict void* mi_heap_calloc(mi_heap_t* heap, size_t count, size_t size) mi_attr_noexcept {
size_t total;
if (mi_count_size_overflow(count,size,&total)) return NULL;
@@ -678,6 +229,29 @@ mi_decl_nodiscard mi_decl_restrict void* mi_calloc(size_t count, size_t size) mi
return mi_heap_calloc(mi_prim_get_default_heap(),count,size);
}
// Return usable size
mi_decl_nodiscard mi_decl_restrict void* mi_umalloc_small(size_t size, size_t* usable) mi_attr_noexcept {
return mi_heap_malloc_small_zero(mi_prim_get_default_heap(), size, false, usable);
}
mi_decl_nodiscard mi_decl_restrict void* mi_heap_umalloc(mi_heap_t* heap, size_t size, size_t* usable) mi_attr_noexcept {
return _mi_heap_malloc_zero_ex(heap, size, false, 0, usable);
}
mi_decl_nodiscard mi_decl_restrict void* mi_umalloc(size_t size, size_t* usable) mi_attr_noexcept {
return mi_heap_umalloc(mi_prim_get_default_heap(), size, usable);
}
mi_decl_nodiscard mi_decl_restrict void* mi_uzalloc(size_t size, size_t* usable) mi_attr_noexcept {
return _mi_heap_malloc_zero_ex(mi_prim_get_default_heap(), size, true, 0, usable);
}
mi_decl_nodiscard mi_decl_restrict void* mi_ucalloc(size_t count, size_t size, size_t* usable) mi_attr_noexcept {
size_t total;
if (mi_count_size_overflow(count,size,&total)) return NULL;
return mi_uzalloc(total, usable);
}
// Uninitialized `calloc`
mi_decl_nodiscard extern mi_decl_restrict void* mi_heap_mallocn(mi_heap_t* heap, size_t count, size_t size) mi_attr_noexcept {
size_t total;
@@ -697,25 +271,38 @@ void* mi_expand(void* p, size_t newsize) mi_attr_noexcept {
return NULL;
#else
if (p == NULL) return NULL;
const size_t size = _mi_usable_size(p,"mi_expand");
const mi_page_t* const page = mi_validate_ptr_page(p,"mi_expand");
const size_t size = _mi_usable_size(p,page);
if (newsize > size) return NULL;
return p; // it fits
#endif
}
void* _mi_heap_realloc_zero(mi_heap_t* heap, void* p, size_t newsize, bool zero) mi_attr_noexcept {
void* _mi_heap_realloc_zero(mi_heap_t* heap, void* p, size_t newsize, bool zero, size_t* usable_pre, size_t* usable_post) mi_attr_noexcept {
// if p == NULL then behave as malloc.
// else if size == 0 then reallocate to a zero-sized block (and don't return NULL, just as mi_malloc(0)).
// (this means that returning NULL always indicates an error, and `p` will not have been freed in that case.)
const size_t size = _mi_usable_size(p,"mi_realloc"); // also works if p == NULL (with size 0)
const mi_page_t* page;
size_t size;
if (p==NULL) {
page = NULL;
size = 0;
if (usable_pre!=NULL) { *usable_pre = 0; }
}
else {
page = mi_validate_ptr_page(p,"mi_realloc");
size = _mi_usable_size(p,page);
if (usable_pre!=NULL) { *usable_pre = mi_page_usable_block_size(page); }
}
if mi_unlikely(newsize <= size && newsize >= (size / 2) && newsize > 0) { // note: newsize must be > 0 or otherwise we return NULL for realloc(NULL,0)
mi_assert_internal(p!=NULL);
// todo: do not track as the usable size is still the same in the free; adjust potential padding?
// mi_track_resize(p,size,newsize)
// if (newsize < size) { mi_track_mem_noaccess((uint8_t*)p + newsize, size - newsize); }
if (usable_post!=NULL) { *usable_post = mi_page_usable_block_size(page); }
return p; // reallocation still fits and not more than 50% waste
}
void* newp = mi_heap_malloc(heap,newsize);
void* newp = mi_heap_umalloc(heap,newsize,usable_post);
if mi_likely(newp != NULL) {
if (zero && newsize > size) {
// also set last word in the previous allocation to zero to ensure any padding is zero-initialized
@@ -736,7 +323,7 @@ void* _mi_heap_realloc_zero(mi_heap_t* heap, void* p, size_t newsize, bool zero)
}
mi_decl_nodiscard void* mi_heap_realloc(mi_heap_t* heap, void* p, size_t newsize) mi_attr_noexcept {
return _mi_heap_realloc_zero(heap, p, newsize, false);
return _mi_heap_realloc_zero(heap, p, newsize, false, NULL, NULL);
}
mi_decl_nodiscard void* mi_heap_reallocn(mi_heap_t* heap, void* p, size_t count, size_t size) mi_attr_noexcept {
@@ -754,7 +341,7 @@ mi_decl_nodiscard void* mi_heap_reallocf(mi_heap_t* heap, void* p, size_t newsiz
}
mi_decl_nodiscard void* mi_heap_rezalloc(mi_heap_t* heap, void* p, size_t newsize) mi_attr_noexcept {
return _mi_heap_realloc_zero(heap, p, newsize, true);
return _mi_heap_realloc_zero(heap, p, newsize, true, NULL, NULL);
}
mi_decl_nodiscard void* mi_heap_recalloc(mi_heap_t* heap, void* p, size_t count, size_t size) mi_attr_noexcept {
@@ -772,6 +359,10 @@ mi_decl_nodiscard void* mi_reallocn(void* p, size_t count, size_t size) mi_attr_
return mi_heap_reallocn(mi_prim_get_default_heap(),p,count,size);
}
mi_decl_nodiscard void* mi_urealloc(void* p, size_t newsize, size_t* usable_pre, size_t* usable_post) mi_attr_noexcept {
return _mi_heap_realloc_zero(mi_prim_get_default_heap(),p,newsize, false, usable_pre, usable_post);
}
// Reallocate but free `p` on errors
mi_decl_nodiscard void* mi_reallocf(void* p, size_t newsize) mi_attr_noexcept {
return mi_heap_reallocf(mi_prim_get_default_heap(),p,newsize);
@@ -794,11 +385,11 @@ mi_decl_nodiscard void* mi_recalloc(void* p, size_t count, size_t size) mi_attr_
// `strdup` using mi_malloc
mi_decl_nodiscard mi_decl_restrict char* mi_heap_strdup(mi_heap_t* heap, const char* s) mi_attr_noexcept {
if (s == NULL) return NULL;
size_t n = strlen(s);
char* t = (char*)mi_heap_malloc(heap,n+1);
size_t len = _mi_strlen(s);
char* t = (char*)mi_heap_malloc(heap,len+1);
if (t == NULL) return NULL;
_mi_memcpy(t, s, n);
t[n] = 0;
_mi_memcpy(t, s, len);
t[len] = 0;
return t;
}
@@ -809,13 +400,11 @@ mi_decl_nodiscard mi_decl_restrict char* mi_strdup(const char* s) mi_attr_noexce
// `strndup` using mi_malloc
mi_decl_nodiscard mi_decl_restrict char* mi_heap_strndup(mi_heap_t* heap, const char* s, size_t n) mi_attr_noexcept {
if (s == NULL) return NULL;
const char* end = (const char*)memchr(s, 0, n); // find end of string in the first `n` characters (returns NULL if not found)
const size_t m = (end != NULL ? (size_t)(end - s) : n); // `m` is the minimum of `n` or the end-of-string
mi_assert_internal(m <= n);
char* t = (char*)mi_heap_malloc(heap, m+1);
const size_t len = _mi_strnlen(s,n); // len <= n
char* t = (char*)mi_heap_malloc(heap, len+1);
if (t == NULL) return NULL;
_mi_memcpy(t, s, m);
t[m] = 0;
_mi_memcpy(t, s, len);
t[len] = 0;
return t;
}
@@ -829,7 +418,7 @@ mi_decl_nodiscard mi_decl_restrict char* mi_strndup(const char* s, size_t n) mi_
#ifndef PATH_MAX
#define PATH_MAX MAX_PATH
#endif
#include <windows.h>
mi_decl_nodiscard mi_decl_restrict char* mi_heap_realpath(mi_heap_t* heap, const char* fname, char* resolved_name) mi_attr_noexcept {
// todo: use GetFullPathNameW to allow longer file names
char buf[PATH_MAX];
@@ -869,7 +458,8 @@ char* mi_heap_realpath(mi_heap_t* heap, const char* fname, char* resolved_name)
char* rname = realpath(fname, NULL);
if (rname == NULL) return NULL;
char* result = mi_heap_strdup(heap, rname);
free(rname); // use regular free! (which may be redirected to our free but that's ok)
mi_cfree(rname); // use checked free (which may be redirected to our free but that's ok)
// note: with ASAN realpath is intercepted and mi_cfree may leak the returned pointer :-(
return result;
}
/*
@@ -913,9 +503,13 @@ static bool mi_try_new_handler(bool nothrow) {
#endif
if (h==NULL) {
_mi_error_message(ENOMEM, "out of memory in 'new'");
#if defined(_CPPUNWIND) || defined(__cpp_exceptions) // exceptions are not always enabled
if (!nothrow) {
throw std::bad_alloc();
}
#else
MI_UNUSED(nothrow);
#endif
return false;
}
else {
@@ -992,7 +586,7 @@ mi_decl_nodiscard mi_decl_restrict void* mi_heap_alloc_new_n(mi_heap_t* heap, si
}
mi_decl_nodiscard mi_decl_restrict void* mi_new_n(size_t count, size_t size) {
return mi_heap_alloc_new_n(mi_prim_get_default_heap(), size, count);
return mi_heap_alloc_new_n(mi_prim_get_default_heap(), count, size);
}
@@ -1039,6 +633,86 @@ mi_decl_nodiscard void* mi_new_reallocn(void* p, size_t newcount, size_t size) {
}
}
#if MI_GUARDED
// We always allocate a guarded allocation at an offset (`mi_page_has_aligned` will be true).
// We then set the first word of the block to `0` for regular offset aligned allocations (in `alloc-aligned.c`)
// and the first word to `~0` for guarded allocations to have a correct `mi_usable_size`
static void* mi_block_ptr_set_guarded(mi_block_t* block, size_t obj_size) {
// TODO: we can still make padding work by moving it out of the guard page area
mi_page_t* const page = _mi_ptr_page(block);
mi_page_set_has_aligned(page, true);
block->next = MI_BLOCK_TAG_GUARDED;
// set guard page at the end of the block
mi_segment_t* const segment = _mi_page_segment(page);
const size_t block_size = mi_page_block_size(page); // must use `block_size` to match `mi_free_local`
const size_t os_page_size = _mi_os_page_size();
mi_assert_internal(block_size >= obj_size + os_page_size + sizeof(mi_block_t));
if (block_size < obj_size + os_page_size + sizeof(mi_block_t)) {
// should never happen
mi_free(block);
return NULL;
}
uint8_t* guard_page = (uint8_t*)block + block_size - os_page_size;
mi_assert_internal(_mi_is_aligned(guard_page, os_page_size));
if mi_likely(segment->allow_decommit && _mi_is_aligned(guard_page, os_page_size)) {
const bool ok = _mi_os_protect(guard_page, os_page_size);
if mi_unlikely(!ok) {
_mi_warning_message("failed to set a guard page behind an object (object %p of size %zu)\n", block, block_size);
}
}
else {
_mi_warning_message("unable to set a guard page behind an object due to pinned memory (large OS pages?) (object %p of size %zu)\n", block, block_size);
}
// align pointer just in front of the guard page
size_t offset = block_size - os_page_size - obj_size;
mi_assert_internal(offset > sizeof(mi_block_t));
if (offset > MI_BLOCK_ALIGNMENT_MAX) {
// give up to place it right in front of the guard page if the offset is too large for unalignment
offset = MI_BLOCK_ALIGNMENT_MAX;
}
void* p = (uint8_t*)block + offset;
mi_track_align(block, p, offset, obj_size);
mi_track_mem_defined(block, sizeof(mi_block_t));
return p;
}
mi_decl_restrict void* _mi_heap_malloc_guarded(mi_heap_t* heap, size_t size, bool zero) mi_attr_noexcept
{
#if defined(MI_PADDING_SIZE)
mi_assert(MI_PADDING_SIZE==0);
#endif
// allocate multiple of page size ending in a guard page
// ensure minimal alignment requirement?
const size_t os_page_size = _mi_os_page_size();
const size_t obj_size = (mi_option_is_enabled(mi_option_guarded_precise) ? size : _mi_align_up(size, MI_MAX_ALIGN_SIZE));
const size_t bsize = _mi_align_up(_mi_align_up(obj_size, MI_MAX_ALIGN_SIZE) + sizeof(mi_block_t), MI_MAX_ALIGN_SIZE);
const size_t req_size = _mi_align_up(bsize + os_page_size, os_page_size);
mi_block_t* const block = (mi_block_t*)_mi_malloc_generic(heap, req_size, zero, 0 /* huge_alignment */, NULL);
if (block==NULL) return NULL;
void* const p = mi_block_ptr_set_guarded(block, obj_size);
// stats
mi_track_malloc(p, size, zero);
if (p != NULL) {
if (!mi_heap_is_initialized(heap)) { heap = mi_prim_get_default_heap(); }
#if MI_STAT>1
mi_heap_stat_adjust_decrease(heap, malloc_requested, req_size);
mi_heap_stat_increase(heap, malloc_requested, size);
#endif
_mi_stat_counter_increase(&heap->tld->stats.malloc_guarded_count, 1);
}
#if MI_DEBUG>3
if (p != NULL && zero) {
mi_assert_expensive(mi_mem_is_zero(p, size));
}
#endif
return p;
}
#endif
// ------------------------------------------------------
// ensure explicit external inline definitions are emitted!
// ------------------------------------------------------
@@ -1046,6 +720,7 @@ mi_decl_nodiscard void* mi_new_reallocn(void* p, size_t newcount, size_t size) {
#ifdef __cplusplus
void* _mi_externs[] = {
(void*)&_mi_page_malloc,
(void*)&_mi_page_malloc_zero,
(void*)&_mi_heap_malloc_zero,
(void*)&_mi_heap_malloc_zero_ex,
(void*)&mi_malloc,

346
mimalloc/mimalloc-sys/mimalloc/src/arena-abandon.c Normal file
View File

@@ -0,0 +1,346 @@
/* ----------------------------------------------------------------------------
Copyright (c) 2019-2024, Microsoft Research, Daan Leijen
This is free software; you can redistribute it and/or modify it under the
terms of the MIT license. A copy of the license can be found in the file
"LICENSE" at the root of this distribution.
-----------------------------------------------------------------------------*/
#if !defined(MI_IN_ARENA_C)
#error "this file should be included from 'arena.c' (so mi_arena_t is visible)"
// add includes help an IDE
#include "mimalloc.h"
#include "mimalloc/internal.h"
#include "bitmap.h"
#endif
// Minimal exports for arena-abandoned.
size_t mi_arena_id_index(mi_arena_id_t id);
mi_arena_t* mi_arena_from_index(size_t idx);
size_t mi_arena_get_count(void);
void* mi_arena_block_start(mi_arena_t* arena, mi_bitmap_index_t bindex);
bool mi_arena_memid_indices(mi_memid_t memid, size_t* arena_index, mi_bitmap_index_t* bitmap_index);
/* -----------------------------------------------------------
Abandoned blocks/segments:
_mi_arena_segment_clear_abandoned
_mi_arena_segment_mark_abandoned
This is used to atomically abandon/reclaim segments
(and crosses the arena API but it is convenient to have here).
Abandoned segments still have live blocks; they get reclaimed
when a thread frees a block in it, or when a thread needs a fresh
segment.
Abandoned segments are atomically marked in the `block_abandoned`
bitmap of arenas. Any segments allocated outside arenas are put
in the sub-process `abandoned_os_list`. This list is accessed
using locks but this should be uncommon and generally uncontended.
Reclaim and visiting either scan through the `block_abandoned`
bitmaps of the arena's, or visit the `abandoned_os_list`
A potentially nicer design is to use arena's for everything
and perhaps have virtual arena's to map OS allocated memory
but this would lack the "density" of our current arena's. TBC.
----------------------------------------------------------- */
// reclaim a specific OS abandoned segment; `true` on success.
// sets the thread_id.
static bool mi_arena_segment_os_clear_abandoned(mi_segment_t* segment, bool take_lock) {
mi_assert(segment->memid.memkind != MI_MEM_ARENA);
// not in an arena, remove from list of abandoned os segments
mi_subproc_t* const subproc = segment->subproc;
if (take_lock && !mi_lock_try_acquire(&subproc->abandoned_os_lock)) {
return false; // failed to acquire the lock, we just give up
}
// remove atomically from the abandoned os list (if possible!)
bool reclaimed = false;
mi_segment_t* const next = segment->abandoned_os_next;
mi_segment_t* const prev = segment->abandoned_os_prev;
if (next != NULL || prev != NULL || subproc->abandoned_os_list == segment) {
#if MI_DEBUG>3
// find ourselves in the abandoned list (and check the count)
bool found = false;
size_t count = 0;
for (mi_segment_t* current = subproc->abandoned_os_list; current != NULL; current = current->abandoned_os_next) {
if (current == segment) { found = true; }
count++;
}
mi_assert_internal(found);
mi_assert_internal(count == mi_atomic_load_relaxed(&subproc->abandoned_os_list_count));
#endif
// remove (atomically) from the list and reclaim
if (prev != NULL) { prev->abandoned_os_next = next; }
else { subproc->abandoned_os_list = next; }
if (next != NULL) { next->abandoned_os_prev = prev; }
else { subproc->abandoned_os_list_tail = prev; }
segment->abandoned_os_next = NULL;
segment->abandoned_os_prev = NULL;
mi_atomic_decrement_relaxed(&subproc->abandoned_count);
mi_atomic_decrement_relaxed(&subproc->abandoned_os_list_count);
if (take_lock) { // don't reset the thread_id when iterating
mi_atomic_store_release(&segment->thread_id, _mi_thread_id());
}
reclaimed = true;
}
if (take_lock) { mi_lock_release(&segment->subproc->abandoned_os_lock); }
return reclaimed;
}
// reclaim a specific abandoned segment; `true` on success.
// sets the thread_id.
bool _mi_arena_segment_clear_abandoned(mi_segment_t* segment) {
if mi_unlikely(segment->memid.memkind != MI_MEM_ARENA) {
return mi_arena_segment_os_clear_abandoned(segment, true /* take lock */);
}
// arena segment: use the blocks_abandoned bitmap.
size_t arena_idx;
size_t bitmap_idx;
mi_arena_memid_indices(segment->memid, &arena_idx, &bitmap_idx);
mi_arena_t* arena = mi_arena_from_index(arena_idx);
mi_assert_internal(arena != NULL);
// reclaim atomically
bool was_marked = _mi_bitmap_unclaim(arena->blocks_abandoned, arena->field_count, 1, bitmap_idx);
if (was_marked) {
mi_assert_internal(mi_atomic_load_acquire(&segment->thread_id) == 0);
mi_atomic_decrement_relaxed(&segment->subproc->abandoned_count);
mi_atomic_store_release(&segment->thread_id, _mi_thread_id());
}
// mi_assert_internal(was_marked);
mi_assert_internal(!was_marked || _mi_bitmap_is_claimed(arena->blocks_inuse, arena->field_count, 1, bitmap_idx));
//mi_assert_internal(arena->blocks_committed == NULL || _mi_bitmap_is_claimed(arena->blocks_committed, arena->field_count, 1, bitmap_idx));
return was_marked;
}
// mark a specific OS segment as abandoned
static void mi_arena_segment_os_mark_abandoned(mi_segment_t* segment) {
mi_assert(segment->memid.memkind != MI_MEM_ARENA);
// not in an arena; we use a list of abandoned segments
mi_subproc_t* const subproc = segment->subproc;
mi_lock(&subproc->abandoned_os_lock) {
// push on the tail of the list (important for the visitor)
mi_segment_t* prev = subproc->abandoned_os_list_tail;
mi_assert_internal(prev == NULL || prev->abandoned_os_next == NULL);
mi_assert_internal(segment->abandoned_os_prev == NULL);
mi_assert_internal(segment->abandoned_os_next == NULL);
if (prev != NULL) { prev->abandoned_os_next = segment; }
else { subproc->abandoned_os_list = segment; }
subproc->abandoned_os_list_tail = segment;
segment->abandoned_os_prev = prev;
segment->abandoned_os_next = NULL;
mi_atomic_increment_relaxed(&subproc->abandoned_os_list_count);
mi_atomic_increment_relaxed(&subproc->abandoned_count);
// and release the lock
}
return;
}
// mark a specific segment as abandoned
// clears the thread_id.
void _mi_arena_segment_mark_abandoned(mi_segment_t* segment)
{
mi_assert_internal(segment->used == segment->abandoned);
mi_atomic_store_release(&segment->thread_id, (uintptr_t)0); // mark as abandoned for multi-thread free's
if mi_unlikely(segment->memid.memkind != MI_MEM_ARENA) {
mi_arena_segment_os_mark_abandoned(segment);
return;
}
// segment is in an arena, mark it in the arena `blocks_abandoned` bitmap
size_t arena_idx;
size_t bitmap_idx;
mi_arena_memid_indices(segment->memid, &arena_idx, &bitmap_idx);
mi_arena_t* arena = mi_arena_from_index(arena_idx);
mi_assert_internal(arena != NULL);
// set abandonment atomically
mi_subproc_t* const subproc = segment->subproc; // don't access the segment after setting it abandoned
const bool was_unmarked = _mi_bitmap_claim(arena->blocks_abandoned, arena->field_count, 1, bitmap_idx, NULL);
if (was_unmarked) { mi_atomic_increment_relaxed(&subproc->abandoned_count); }
mi_assert_internal(was_unmarked);
mi_assert_internal(_mi_bitmap_is_claimed(arena->blocks_inuse, arena->field_count, 1, bitmap_idx));
}
/* -----------------------------------------------------------
Iterate through the abandoned blocks/segments using a cursor.
This is used for reclaiming and abandoned block visiting.
----------------------------------------------------------- */
// start a cursor at a randomized arena
void _mi_arena_field_cursor_init(mi_heap_t* heap, mi_subproc_t* subproc, bool visit_all, mi_arena_field_cursor_t* current) {
mi_assert_internal(heap == NULL || heap->tld->segments.subproc == subproc);
current->bitmap_idx = 0;
current->subproc = subproc;
current->visit_all = visit_all;
current->hold_visit_lock = false;
const size_t abandoned_count = mi_atomic_load_relaxed(&subproc->abandoned_count);
const size_t abandoned_list_count = mi_atomic_load_relaxed(&subproc->abandoned_os_list_count);
const size_t max_arena = mi_arena_get_count();
if (heap != NULL && heap->arena_id != _mi_arena_id_none()) {
// for a heap that is bound to one arena, only visit that arena
current->start = mi_arena_id_index(heap->arena_id);
current->end = current->start + 1;
current->os_list_count = 0;
}
else {
// otherwise visit all starting at a random location
if (abandoned_count > abandoned_list_count && max_arena > 0) {
current->start = (heap == NULL || max_arena == 0 ? 0 : (mi_arena_id_t)(_mi_heap_random_next(heap) % max_arena));
current->end = current->start + max_arena;
}
else {
current->start = 0;
current->end = 0;
}
current->os_list_count = abandoned_list_count; // max entries to visit in the os abandoned list
}
mi_assert_internal(current->start <= max_arena);
}
void _mi_arena_field_cursor_done(mi_arena_field_cursor_t* current) {
if (current->hold_visit_lock) {
mi_lock_release(&current->subproc->abandoned_os_visit_lock);
current->hold_visit_lock = false;
}
}
static mi_segment_t* mi_arena_segment_clear_abandoned_at(mi_arena_t* arena, mi_subproc_t* subproc, mi_bitmap_index_t bitmap_idx) {
// try to reclaim an abandoned segment in the arena atomically
if (!_mi_bitmap_unclaim(arena->blocks_abandoned, arena->field_count, 1, bitmap_idx)) return NULL;
mi_assert_internal(_mi_bitmap_is_claimed(arena->blocks_inuse, arena->field_count, 1, bitmap_idx));
mi_segment_t* segment = (mi_segment_t*)mi_arena_block_start(arena, bitmap_idx);
mi_assert_internal(mi_atomic_load_relaxed(&segment->thread_id) == 0);
// check that the segment belongs to our sub-process
// note: this is the reason we need the `abandoned_visit` lock in the case abandoned visiting is enabled.
// without the lock an abandoned visit may otherwise fail to visit all abandoned segments in the sub-process.
// for regular reclaim it is fine to miss one sometimes so without abandoned visiting we don't need the `abandoned_visit` lock.
if (segment->subproc != subproc) {
// it is from another sub-process, re-mark it and continue searching
const bool was_zero = _mi_bitmap_claim(arena->blocks_abandoned, arena->field_count, 1, bitmap_idx, NULL);
mi_assert_internal(was_zero); MI_UNUSED(was_zero);
return NULL;
}
else {
// success, we unabandoned a segment in our sub-process
mi_atomic_decrement_relaxed(&subproc->abandoned_count);
return segment;
}
}
static mi_segment_t* mi_arena_segment_clear_abandoned_next_field(mi_arena_field_cursor_t* previous) {
const size_t max_arena = mi_arena_get_count();
size_t field_idx = mi_bitmap_index_field(previous->bitmap_idx);
size_t bit_idx = mi_bitmap_index_bit_in_field(previous->bitmap_idx);
// visit arena's (from the previous cursor)
for (; previous->start < previous->end; previous->start++, field_idx = 0, bit_idx = 0) {
// index wraps around
size_t arena_idx = (previous->start >= max_arena ? previous->start % max_arena : previous->start);
mi_arena_t* arena = mi_arena_from_index(arena_idx);
if (arena != NULL) {
bool has_lock = false;
// visit the abandoned fields (starting at previous_idx)
for (; field_idx < arena->field_count; field_idx++, bit_idx = 0) {
size_t field = mi_atomic_load_relaxed(&arena->blocks_abandoned[field_idx]);
if mi_unlikely(field != 0) { // skip zero fields quickly
// we only take the arena lock if there are actually abandoned segments present
if (!has_lock && mi_option_is_enabled(mi_option_visit_abandoned)) {
has_lock = (previous->visit_all ? (mi_lock_acquire(&arena->abandoned_visit_lock),true) : mi_lock_try_acquire(&arena->abandoned_visit_lock));
if (!has_lock) {
if (previous->visit_all) {
_mi_error_message(EFAULT, "internal error: failed to visit all abandoned segments due to failure to acquire the visitor lock");
}
// skip to next arena
break;
}
}
mi_assert_internal(has_lock || !mi_option_is_enabled(mi_option_visit_abandoned));
// visit each set bit in the field (todo: maybe use `ctz` here?)
for (; bit_idx < MI_BITMAP_FIELD_BITS; bit_idx++) {
// pre-check if the bit is set
size_t mask = ((size_t)1 << bit_idx);
if mi_unlikely((field & mask) == mask) {
mi_bitmap_index_t bitmap_idx = mi_bitmap_index_create(field_idx, bit_idx);
mi_segment_t* const segment = mi_arena_segment_clear_abandoned_at(arena, previous->subproc, bitmap_idx);
if (segment != NULL) {
//mi_assert_internal(arena->blocks_committed == NULL || _mi_bitmap_is_claimed(arena->blocks_committed, arena->field_count, 1, bitmap_idx));
if (has_lock) { mi_lock_release(&arena->abandoned_visit_lock); }
previous->bitmap_idx = mi_bitmap_index_create_ex(field_idx, bit_idx + 1); // start at next one for the next iteration
return segment;
}
}
}
}
}
if (has_lock) { mi_lock_release(&arena->abandoned_visit_lock); }
}
}
return NULL;
}
static mi_segment_t* mi_arena_segment_clear_abandoned_next_list(mi_arena_field_cursor_t* previous) {
// go through the abandoned_os_list
// we only allow one thread per sub-process to do to visit guarded by the `abandoned_os_visit_lock`.
// The lock is released when the cursor is released.
if (!previous->hold_visit_lock) {
previous->hold_visit_lock = (previous->visit_all ? (mi_lock_acquire(&previous->subproc->abandoned_os_visit_lock),true)
: mi_lock_try_acquire(&previous->subproc->abandoned_os_visit_lock));
if (!previous->hold_visit_lock) {
if (previous->visit_all) {
_mi_error_message(EFAULT, "internal error: failed to visit all abandoned segments due to failure to acquire the OS visitor lock");
}
return NULL; // we cannot get the lock, give up
}
}
// One list entry at a time
while (previous->os_list_count > 0) {
previous->os_list_count--;
mi_lock_acquire(&previous->subproc->abandoned_os_lock); // this could contend with concurrent OS block abandonment and reclaim from `free`
mi_segment_t* segment = previous->subproc->abandoned_os_list;
// pop from head of the list, a subsequent mark will push at the end (and thus we iterate through os_list_count entries)
if (segment == NULL || mi_arena_segment_os_clear_abandoned(segment, false /* we already have the lock */)) {
mi_lock_release(&previous->subproc->abandoned_os_lock);
return segment;
}
// already abandoned, try again
mi_lock_release(&previous->subproc->abandoned_os_lock);
}
// done
mi_assert_internal(previous->os_list_count == 0);
return NULL;
}
// reclaim abandoned segments
// this does not set the thread id (so it appears as still abandoned)
mi_segment_t* _mi_arena_segment_clear_abandoned_next(mi_arena_field_cursor_t* previous) {
if (previous->start < previous->end) {
// walk the arena
mi_segment_t* segment = mi_arena_segment_clear_abandoned_next_field(previous);
if (segment != NULL) { return segment; }
}
// no entries in the arena's anymore, walk the abandoned OS list
mi_assert_internal(previous->start == previous->end);
return mi_arena_segment_clear_abandoned_next_list(previous);
}
bool mi_abandoned_visit_blocks(mi_subproc_id_t subproc_id, int heap_tag, bool visit_blocks, mi_block_visit_fun* visitor, void* arg) {
// (unfortunately) the visit_abandoned option must be enabled from the start.
// This is to avoid taking locks if abandoned list visiting is not required (as for most programs)
if (!mi_option_is_enabled(mi_option_visit_abandoned)) {
_mi_error_message(EFAULT, "internal error: can only visit abandoned blocks when MIMALLOC_VISIT_ABANDONED=ON");
return false;
}
mi_arena_field_cursor_t current;
_mi_arena_field_cursor_init(NULL, _mi_subproc_from_id(subproc_id), true /* visit all (blocking) */, &current);
mi_segment_t* segment;
bool ok = true;
while (ok && (segment = _mi_arena_segment_clear_abandoned_next(&current)) != NULL) {
ok = _mi_segment_visit_blocks(segment, heap_tag, visit_blocks, visitor, arg);
_mi_arena_segment_mark_abandoned(segment);
}
_mi_arena_field_cursor_done(&current);
return ok;
}

537
mimalloc/mimalloc-sys/mimalloc/src/arena.c
View File

File diff suppressed because it is too large Load Diff

43
mimalloc/mimalloc-sys/mimalloc/src/bitmap.c
View File

@@ -7,7 +7,7 @@ terms of the MIT license. A copy of the license can be found in the file
/* ----------------------------------------------------------------------------
Concurrent bitmap that can set/reset sequences of bits atomically,
represeted as an array of fields where each field is a machine word (`size_t`)
represented as an array of fields where each field is a machine word (`size_t`)
There are two api's; the standard one cannot have sequences that cross
between the bitmap fields (and a sequence must be <= MI_BITMAP_FIELD_BITS).
@@ -81,7 +81,7 @@ inline bool _mi_bitmap_try_find_claim_field(mi_bitmap_t bitmap, size_t idx, cons
// on to the next bit range
#ifdef MI_HAVE_FAST_BITSCAN
mi_assert_internal(mapm != 0);
const size_t shift = (count == 1 ? 1 : (MI_INTPTR_BITS - mi_clz(mapm) - bitidx));
const size_t shift = (count == 1 ? 1 : (MI_SIZE_BITS - mi_clz(mapm) - bitidx));
mi_assert_internal(shift > 0 && shift <= count);
#else
const size_t shift = 1;
@@ -164,7 +164,7 @@ static bool mi_bitmap_is_claimedx(mi_bitmap_t bitmap, size_t bitmap_fields, size
return ((field & mask) == mask);
}
// Try to set `count` bits at `bitmap_idx` from 0 to 1 atomically.
// Try to set `count` bits at `bitmap_idx` from 0 to 1 atomically.
// Returns `true` if successful when all previous `count` bits were 0.
bool _mi_bitmap_try_claim(mi_bitmap_t bitmap, size_t bitmap_fields, size_t count, mi_bitmap_index_t bitmap_idx) {
const size_t idx = mi_bitmap_index_field(bitmap_idx);
@@ -172,9 +172,9 @@ bool _mi_bitmap_try_claim(mi_bitmap_t bitmap, size_t bitmap_fields, size_t count
const size_t mask = mi_bitmap_mask_(count, bitidx);
mi_assert_internal(bitmap_fields > idx); MI_UNUSED(bitmap_fields);
size_t expected = mi_atomic_load_relaxed(&bitmap[idx]);
do {
do {
if ((expected & mask) != 0) return false;
}
}
while (!mi_atomic_cas_strong_acq_rel(&bitmap[idx], &expected, expected | mask));
mi_assert_internal((expected & mask) == 0);
return true;
@@ -212,7 +212,7 @@ static bool mi_bitmap_try_find_claim_field_across(mi_bitmap_t bitmap, size_t bit
if (initial == 0) return false;
if (initial >= count) return _mi_bitmap_try_find_claim_field(bitmap, idx, count, bitmap_idx); // no need to cross fields (this case won't happen for us)
if (_mi_divide_up(count - initial, MI_BITMAP_FIELD_BITS) >= (bitmap_fields - idx)) return false; // not enough entries
// scan ahead
size_t found = initial;
size_t mask = 0; // mask bits for the final field
@@ -279,6 +279,7 @@ rollback:
newmap = (map & ~initial_mask);
} while (!mi_atomic_cas_strong_acq_rel(field, &map, newmap));
}
mi_stat_counter_increase(_mi_stats_main.arena_rollback_count,1);
// retry? (we make a recursive call instead of goto to be able to use const declarations)
if (retries <= 2) {
return mi_bitmap_try_find_claim_field_across(bitmap, bitmap_fields, idx, count, retries+1, bitmap_idx);
@@ -303,11 +304,13 @@ bool _mi_bitmap_try_find_from_claim_across(mi_bitmap_t bitmap, const size_t bitm
for (size_t visited = 0; visited < bitmap_fields; visited++, idx++) {
if (idx >= bitmap_fields) { idx = 0; } // wrap
// first try to claim inside a field
/*
if (count <= MI_BITMAP_FIELD_BITS) {
if (_mi_bitmap_try_find_claim_field(bitmap, idx, count, bitmap_idx)) {
return true;
}
}
*/
// if that fails, then try to claim across fields
if (mi_bitmap_try_find_claim_field_across(bitmap, bitmap_fields, idx, count, 0, bitmap_idx)) {
return true;
@@ -366,7 +369,7 @@ bool _mi_bitmap_unclaim_across(mi_bitmap_t bitmap, size_t bitmap_fields, size_t
// Set `count` bits at `bitmap_idx` to 1 atomically
// Returns `true` if all `count` bits were 0 previously. `any_zero` is `true` if there was at least one zero bit.
bool _mi_bitmap_claim_across(mi_bitmap_t bitmap, size_t bitmap_fields, size_t count, mi_bitmap_index_t bitmap_idx, bool* pany_zero) {
bool _mi_bitmap_claim_across(mi_bitmap_t bitmap, size_t bitmap_fields, size_t count, mi_bitmap_index_t bitmap_idx, bool* pany_zero, size_t* already_set) {
size_t idx = mi_bitmap_index_field(bitmap_idx);
size_t pre_mask;
size_t mid_mask;
@@ -374,28 +377,31 @@ bool _mi_bitmap_claim_across(mi_bitmap_t bitmap, size_t bitmap_fields, size_t co
size_t mid_count = mi_bitmap_mask_across(bitmap_idx, bitmap_fields, count, &pre_mask, &mid_mask, &post_mask);
bool all_zero = true;
bool any_zero = false;
size_t one_count = 0;
_Atomic(size_t)*field = &bitmap[idx];
size_t prev = mi_atomic_or_acq_rel(field++, pre_mask);
if ((prev & pre_mask) != 0) all_zero = false;
if ((prev & pre_mask) != 0) { all_zero = false; one_count += mi_popcount(prev & pre_mask); }
if ((prev & pre_mask) != pre_mask) any_zero = true;
while (mid_count-- > 0) {
prev = mi_atomic_or_acq_rel(field++, mid_mask);
if ((prev & mid_mask) != 0) all_zero = false;
if ((prev & mid_mask) != 0) { all_zero = false; one_count += mi_popcount(prev & mid_mask); }
if ((prev & mid_mask) != mid_mask) any_zero = true;
}
if (post_mask!=0) {
prev = mi_atomic_or_acq_rel(field, post_mask);
if ((prev & post_mask) != 0) all_zero = false;
if ((prev & post_mask) != 0) { all_zero = false; one_count += mi_popcount(prev & post_mask); }
if ((prev & post_mask) != post_mask) any_zero = true;
}
if (pany_zero != NULL) { *pany_zero = any_zero; }
if (already_set != NULL) { *already_set = one_count; };
mi_assert_internal(all_zero ? one_count == 0 : one_count <= count);
return all_zero;
}
// Returns `true` if all `count` bits were 1.
// `any_ones` is `true` if there was at least one bit set to one.
static bool mi_bitmap_is_claimedx_across(mi_bitmap_t bitmap, size_t bitmap_fields, size_t count, mi_bitmap_index_t bitmap_idx, bool* pany_ones) {
static bool mi_bitmap_is_claimedx_across(mi_bitmap_t bitmap, size_t bitmap_fields, size_t count, mi_bitmap_index_t bitmap_idx, bool* pany_ones, size_t* already_set) {
size_t idx = mi_bitmap_index_field(bitmap_idx);
size_t pre_mask;
size_t mid_mask;
@@ -403,30 +409,33 @@ static bool mi_bitmap_is_claimedx_across(mi_bitmap_t bitmap, size_t bitmap_field
size_t mid_count = mi_bitmap_mask_across(bitmap_idx, bitmap_fields, count, &pre_mask, &mid_mask, &post_mask);
bool all_ones = true;
bool any_ones = false;
size_t one_count = 0;
mi_bitmap_field_t* field = &bitmap[idx];
size_t prev = mi_atomic_load_relaxed(field++);
if ((prev & pre_mask) != pre_mask) all_ones = false;
if ((prev & pre_mask) != 0) any_ones = true;
if ((prev & pre_mask) != 0) { any_ones = true; one_count += mi_popcount(prev & pre_mask); }
while (mid_count-- > 0) {
prev = mi_atomic_load_relaxed(field++);
if ((prev & mid_mask) != mid_mask) all_ones = false;
if ((prev & mid_mask) != 0) any_ones = true;
if ((prev & mid_mask) != 0) { any_ones = true; one_count += mi_popcount(prev & mid_mask); }
}
if (post_mask!=0) {
prev = mi_atomic_load_relaxed(field);
if ((prev & post_mask) != post_mask) all_ones = false;
if ((prev & post_mask) != 0) any_ones = true;
if ((prev & post_mask) != 0) { any_ones = true; one_count += mi_popcount(prev & post_mask); }
}
if (pany_ones != NULL) { *pany_ones = any_ones; }
if (already_set != NULL) { *already_set = one_count; }
mi_assert_internal(all_ones ? one_count == count : one_count < count);
return all_ones;
}
bool _mi_bitmap_is_claimed_across(mi_bitmap_t bitmap, size_t bitmap_fields, size_t count, mi_bitmap_index_t bitmap_idx) {
return mi_bitmap_is_claimedx_across(bitmap, bitmap_fields, count, bitmap_idx, NULL);
bool _mi_bitmap_is_claimed_across(mi_bitmap_t bitmap, size_t bitmap_fields, size_t count, mi_bitmap_index_t bitmap_idx, size_t* already_set) {
return mi_bitmap_is_claimedx_across(bitmap, bitmap_fields, count, bitmap_idx, NULL, already_set);
}
bool _mi_bitmap_is_any_claimed_across(mi_bitmap_t bitmap, size_t bitmap_fields, size_t count, mi_bitmap_index_t bitmap_idx) {
bool any_ones;
mi_bitmap_is_claimedx_across(bitmap, bitmap_fields, count, bitmap_idx, &any_ones);
mi_bitmap_is_claimedx_across(bitmap, bitmap_fields, count, bitmap_idx, &any_ones, NULL);
return any_ones;
}

12
mimalloc/mimalloc-sys/mimalloc/src/bitmap.h
View File

@@ -7,7 +7,7 @@ terms of the MIT license. A copy of the license can be found in the file
/* ----------------------------------------------------------------------------
Concurrent bitmap that can set/reset sequences of bits atomically,
represeted as an array of fields where each field is a machine word (`size_t`)
represented as an array of fields where each field is a machine word (`size_t`)
There are two api's; the standard one cannot have sequences that cross
between the bitmap fields (and a sequence must be <= MI_BITMAP_FIELD_BITS).
@@ -35,9 +35,13 @@ typedef mi_bitmap_field_t* mi_bitmap_t;
typedef size_t mi_bitmap_index_t;
// Create a bit index.
static inline mi_bitmap_index_t mi_bitmap_index_create_ex(size_t idx, size_t bitidx) {
mi_assert_internal(bitidx <= MI_BITMAP_FIELD_BITS);
return (idx*MI_BITMAP_FIELD_BITS) + bitidx;
}
static inline mi_bitmap_index_t mi_bitmap_index_create(size_t idx, size_t bitidx) {
mi_assert_internal(bitidx < MI_BITMAP_FIELD_BITS);
return (idx*MI_BITMAP_FIELD_BITS) + bitidx;
return mi_bitmap_index_create_ex(idx,bitidx);
}
// Create a bit index.
@@ -107,9 +111,9 @@ bool _mi_bitmap_unclaim_across(mi_bitmap_t bitmap, size_t bitmap_fields, size_t
// Set `count` bits at `bitmap_idx` to 1 atomically
// Returns `true` if all `count` bits were 0 previously. `any_zero` is `true` if there was at least one zero bit.
bool _mi_bitmap_claim_across(mi_bitmap_t bitmap, size_t bitmap_fields, size_t count, mi_bitmap_index_t bitmap_idx, bool* pany_zero);
bool _mi_bitmap_claim_across(mi_bitmap_t bitmap, size_t bitmap_fields, size_t count, mi_bitmap_index_t bitmap_idx, bool* pany_zero, size_t* already_set);
bool _mi_bitmap_is_claimed_across(mi_bitmap_t bitmap, size_t bitmap_fields, size_t count, mi_bitmap_index_t bitmap_idx);
bool _mi_bitmap_is_claimed_across(mi_bitmap_t bitmap, size_t bitmap_fields, size_t count, mi_bitmap_index_t bitmap_idx, size_t* already_set);
bool _mi_bitmap_is_any_claimed_across(mi_bitmap_t bitmap, size_t bitmap_fields, size_t count, mi_bitmap_index_t bitmap_idx);
#endif

588
mimalloc/mimalloc-sys/mimalloc/src/free.c Normal file
View File

@@ -0,0 +1,588 @@
/* ----------------------------------------------------------------------------
Copyright (c) 2018-2024, Microsoft Research, Daan Leijen
This is free software; you can redistribute it and/or modify it under the
terms of the MIT license. A copy of the license can be found in the file
"LICENSE" at the root of this distribution.
-----------------------------------------------------------------------------*/
#if !defined(MI_IN_ALLOC_C)
#error "this file should be included from 'alloc.c' (so aliases can work from alloc-override)"
// add includes help an IDE
#include "mimalloc.h"
#include "mimalloc/internal.h"
#include "mimalloc/prim.h" // _mi_prim_thread_id()
#endif
// forward declarations
static void mi_check_padding(const mi_page_t* page, const mi_block_t* block);
static bool mi_check_is_double_free(const mi_page_t* page, const mi_block_t* block);
static size_t mi_page_usable_size_of(const mi_page_t* page, const mi_block_t* block);
static void mi_stat_free(const mi_page_t* page, const mi_block_t* block);
// ------------------------------------------------------
// Free
// ------------------------------------------------------
// forward declaration of multi-threaded free (`_mt`) (or free in huge block if compiled with MI_HUGE_PAGE_ABANDON)
static mi_decl_noinline void mi_free_block_mt(mi_page_t* page, mi_segment_t* segment, mi_block_t* block);
// regular free of a (thread local) block pointer
// fast path written carefully to prevent spilling on the stack
static inline void mi_free_block_local(mi_page_t* page, mi_block_t* block, bool track_stats, bool check_full)
{
// checks
if mi_unlikely(mi_check_is_double_free(page, block)) return;
mi_check_padding(page, block);
if (track_stats) { mi_stat_free(page, block); }
#if (MI_DEBUG>0) && !MI_TRACK_ENABLED && !MI_TSAN && !MI_GUARDED
if (!mi_page_is_huge(page)) { // huge page content may be already decommitted
memset(block, MI_DEBUG_FREED, mi_page_block_size(page));
}
#endif
if (track_stats) { mi_track_free_size(block, mi_page_usable_size_of(page, block)); } // faster then mi_usable_size as we already know the page and that p is unaligned
// actual free: push on the local free list
mi_block_set_next(page, block, page->local_free);
page->local_free = block;
if mi_unlikely(--page->used == 0) {
_mi_page_retire(page);
}
else if mi_unlikely(check_full && mi_page_is_in_full(page)) {
_mi_page_unfull(page);
}
}
// Adjust a block that was allocated aligned, to the actual start of the block in the page.
// note: this can be called from `mi_free_generic_mt` where a non-owning thread accesses the
// `page_start` and `block_size` fields; however these are constant and the page won't be
// deallocated (as the block we are freeing keeps it alive) and thus safe to read concurrently.
mi_block_t* _mi_page_ptr_unalign(const mi_page_t* page, const void* p) {
mi_assert_internal(page!=NULL && p!=NULL);
size_t diff = (uint8_t*)p - page->page_start;
size_t adjust;
if mi_likely(page->block_size_shift != 0) {
adjust = diff & (((size_t)1 << page->block_size_shift) - 1);
}
else {
adjust = diff % mi_page_block_size(page);
}
return (mi_block_t*)((uintptr_t)p - adjust);
}
// forward declaration for a MI_GUARDED build
#if MI_GUARDED
static void mi_block_unguard(mi_page_t* page, mi_block_t* block, void* p); // forward declaration
static inline void mi_block_check_unguard(mi_page_t* page, mi_block_t* block, void* p) {
if (mi_block_ptr_is_guarded(block, p)) { mi_block_unguard(page, block, p); }
}
#else
static inline void mi_block_check_unguard(mi_page_t* page, mi_block_t* block, void* p) {
MI_UNUSED(page); MI_UNUSED(block); MI_UNUSED(p);
}
#endif
// free a local pointer (page parameter comes first for better codegen)
static void mi_decl_noinline mi_free_generic_local(mi_page_t* page, mi_segment_t* segment, void* p) mi_attr_noexcept {
MI_UNUSED(segment);
mi_block_t* const block = (mi_page_has_aligned(page) ? _mi_page_ptr_unalign(page, p) : (mi_block_t*)p);
mi_block_check_unguard(page, block, p);
mi_free_block_local(page, block, true /* track stats */, true /* check for a full page */);
}
// free a pointer owned by another thread (page parameter comes first for better codegen)
static void mi_decl_noinline mi_free_generic_mt(mi_page_t* page, mi_segment_t* segment, void* p) mi_attr_noexcept {
mi_block_t* const block = _mi_page_ptr_unalign(page, p); // don't check `has_aligned` flag to avoid a race (issue #865)
mi_block_check_unguard(page, block, p);
mi_free_block_mt(page, segment, block);
}
// generic free (for runtime integration)
void mi_decl_noinline _mi_free_generic(mi_segment_t* segment, mi_page_t* page, bool is_local, void* p) mi_attr_noexcept {
if (is_local) mi_free_generic_local(page,segment,p);
else mi_free_generic_mt(page,segment,p);
}
// Get the segment data belonging to a pointer
// This is just a single `and` in release mode but does further checks in debug mode
// (and secure mode) to see if this was a valid pointer.
static inline mi_segment_t* mi_checked_ptr_segment(const void* p, const char* msg)
{
MI_UNUSED(msg);
#if (MI_DEBUG>0)
if mi_unlikely(((uintptr_t)p & (MI_INTPTR_SIZE - 1)) != 0 && !mi_option_is_enabled(mi_option_guarded_precise)) {
_mi_error_message(EINVAL, "%s: invalid (unaligned) pointer: %p\n", msg, p);
return NULL;
}
#endif
mi_segment_t* const segment = _mi_ptr_segment(p);
if mi_unlikely(segment==NULL) return segment;
#if (MI_DEBUG>0)
if mi_unlikely(!mi_is_in_heap_region(p)) {
#if (MI_INTPTR_SIZE == 8 && defined(__linux__))
if (((uintptr_t)p >> 40) != 0x7F) { // linux tends to align large blocks above 0x7F000000000 (issue #640)
#else
{
#endif
_mi_warning_message("%s: pointer might not point to a valid heap region: %p\n"
"(this may still be a valid very large allocation (over 64MiB))\n", msg, p);
if mi_likely(_mi_ptr_cookie(segment) == segment->cookie) {
_mi_warning_message("(yes, the previous pointer %p was valid after all)\n", p);
}
}
}
#endif
#if (MI_DEBUG>0 || MI_SECURE>=4)
if mi_unlikely(_mi_ptr_cookie(segment) != segment->cookie) {
_mi_error_message(EINVAL, "%s: pointer does not point to a valid heap space: %p\n", msg, p);
return NULL;
}
#endif
return segment;
}
// Free a block
// Fast path written carefully to prevent register spilling on the stack
static inline void mi_free_ex(void* p, size_t* usable) mi_attr_noexcept
{
mi_segment_t* const segment = mi_checked_ptr_segment(p,"mi_free");
if mi_unlikely(segment==NULL) return;
const bool is_local = (_mi_prim_thread_id() == mi_atomic_load_relaxed(&segment->thread_id));
mi_page_t* const page = _mi_segment_page_of(segment, p);
if (usable!=NULL) { *usable = mi_page_usable_block_size(page); }
if mi_likely(is_local) { // thread-local free?
if mi_likely(page->flags.full_aligned == 0) { // and it is not a full page (full pages need to move from the full bin), nor has aligned blocks (aligned blocks need to be unaligned)
// thread-local, aligned, and not a full page
mi_block_t* const block = (mi_block_t*)p;
mi_free_block_local(page, block, true /* track stats */, false /* no need to check if the page is full */);
}
else {
// page is full or contains (inner) aligned blocks; use generic path
mi_free_generic_local(page, segment, p);
}
}
else {
// not thread-local; use generic path
mi_free_generic_mt(page, segment, p);
}
}
void mi_free(void* p) mi_attr_noexcept {
mi_free_ex(p,NULL);
}
void mi_ufree(void* p, size_t* usable) mi_attr_noexcept {
mi_free_ex(p,usable);
}
// return true if successful
bool _mi_free_delayed_block(mi_block_t* block) {
// get segment and page
mi_assert_internal(block!=NULL);
const mi_segment_t* const segment = _mi_ptr_segment(block);
mi_assert_internal(_mi_ptr_cookie(segment) == segment->cookie);
mi_assert_internal(_mi_thread_id() == segment->thread_id);
mi_page_t* const page = _mi_segment_page_of(segment, block);
// Clear the no-delayed flag so delayed freeing is used again for this page.
// This must be done before collecting the free lists on this page -- otherwise
// some blocks may end up in the page `thread_free` list with no blocks in the
// heap `thread_delayed_free` list which may cause the page to be never freed!
// (it would only be freed if we happen to scan it in `mi_page_queue_find_free_ex`)
if (!_mi_page_try_use_delayed_free(page, MI_USE_DELAYED_FREE, false /* dont overwrite never delayed */)) {
return false;
}
// collect all other non-local frees (move from `thread_free` to `free`) to ensure up-to-date `used` count
_mi_page_free_collect(page, false);
// and free the block (possibly freeing the page as well since `used` is updated)
mi_free_block_local(page, block, false /* stats have already been adjusted */, true /* check for a full page */);
return true;
}
// ------------------------------------------------------
// Multi-threaded Free (`_mt`)
// ------------------------------------------------------
// Push a block that is owned by another thread on its page-local thread free
// list or it's heap delayed free list. Such blocks are later collected by
// the owning thread in `_mi_free_delayed_block`.
static void mi_decl_noinline mi_free_block_delayed_mt( mi_page_t* page, mi_block_t* block )
{
// Try to put the block on either the page-local thread free list,
// or the heap delayed free list (if this is the first non-local free in that page)
mi_thread_free_t tfreex;
bool use_delayed;
mi_thread_free_t tfree = mi_atomic_load_relaxed(&page->xthread_free);
do {
use_delayed = (mi_tf_delayed(tfree) == MI_USE_DELAYED_FREE);
if mi_unlikely(use_delayed) {
// unlikely: this only happens on the first concurrent free in a page that is in the full list
tfreex = mi_tf_set_delayed(tfree,MI_DELAYED_FREEING);
}
else {
// usual: directly add to page thread_free list
mi_block_set_next(page, block, mi_tf_block(tfree));
tfreex = mi_tf_set_block(tfree,block);
}
} while (!mi_atomic_cas_weak_release(&page->xthread_free, &tfree, tfreex));
// If this was the first non-local free, we need to push it on the heap delayed free list instead
if mi_unlikely(use_delayed) {
// racy read on `heap`, but ok because MI_DELAYED_FREEING is set (see `mi_heap_delete` and `mi_heap_collect_abandon`)
mi_heap_t* const heap = (mi_heap_t*)(mi_atomic_load_acquire(&page->xheap)); //mi_page_heap(page);
mi_assert_internal(heap != NULL);
if (heap != NULL) {
// add to the delayed free list of this heap. (do this atomically as the lock only protects heap memory validity)
mi_block_t* dfree = mi_atomic_load_ptr_relaxed(mi_block_t, &heap->thread_delayed_free);
do {
mi_block_set_nextx(heap,block,dfree, heap->keys);
} while (!mi_atomic_cas_ptr_weak_release(mi_block_t,&heap->thread_delayed_free, &dfree, block));
}
// and reset the MI_DELAYED_FREEING flag
tfree = mi_atomic_load_relaxed(&page->xthread_free);
do {
tfreex = tfree;
mi_assert_internal(mi_tf_delayed(tfree) == MI_DELAYED_FREEING);
tfreex = mi_tf_set_delayed(tfree,MI_NO_DELAYED_FREE);
} while (!mi_atomic_cas_weak_release(&page->xthread_free, &tfree, tfreex));
}
}
// Multi-threaded free (`_mt`) (or free in huge block if compiled with MI_HUGE_PAGE_ABANDON)
static void mi_decl_noinline mi_free_block_mt(mi_page_t* page, mi_segment_t* segment, mi_block_t* block)
{
// first see if the segment was abandoned and if we can reclaim it into our thread
if (_mi_option_get_fast(mi_option_abandoned_reclaim_on_free) != 0 &&
#if MI_HUGE_PAGE_ABANDON
segment->page_kind != MI_PAGE_HUGE &&
#endif
mi_atomic_load_relaxed(&segment->thread_id) == 0 && // segment is abandoned?
mi_prim_get_default_heap() != (mi_heap_t*)&_mi_heap_empty) // and we did not already exit this thread (without this check, a fresh heap will be initalized (issue #944))
{
// the segment is abandoned, try to reclaim it into our heap
if (_mi_segment_attempt_reclaim(mi_heap_get_default(), segment)) {
mi_assert_internal(_mi_thread_id() == mi_atomic_load_relaxed(&segment->thread_id));
mi_assert_internal(mi_heap_get_default()->tld->segments.subproc == segment->subproc);
mi_free(block); // recursively free as now it will be a local free in our heap
return;
}
}
// The padding check may access the non-thread-owned page for the key values.
// that is safe as these are constant and the page won't be freed (as the block is not freed yet).
mi_check_padding(page, block);
// adjust stats (after padding check and potentially recursive `mi_free` above)
mi_stat_free(page, block); // stat_free may access the padding
mi_track_free_size(block, mi_page_usable_size_of(page,block));
// for small size, ensure we can fit the delayed thread pointers without triggering overflow detection
_mi_padding_shrink(page, block, sizeof(mi_block_t));
if (segment->kind == MI_SEGMENT_HUGE) {
#if MI_HUGE_PAGE_ABANDON
// huge page segments are always abandoned and can be freed immediately
_mi_segment_huge_page_free(segment, page, block);
return;
#else
// huge pages are special as they occupy the entire segment
// as these are large we reset the memory occupied by the page so it is available to other threads
// (as the owning thread needs to actually free the memory later).
_mi_segment_huge_page_reset(segment, page, block);
#endif
}
else {
#if (MI_DEBUG>0) && !MI_TRACK_ENABLED && !MI_TSAN // note: when tracking, cannot use mi_usable_size with multi-threading
memset(block, MI_DEBUG_FREED, mi_usable_size(block));
#endif
}
// and finally free the actual block by pushing it on the owning heap
// thread_delayed free list (or heap delayed free list)
mi_free_block_delayed_mt(page,block);
}
// ------------------------------------------------------
// Usable size
// ------------------------------------------------------
// Bytes available in a block
static size_t mi_decl_noinline mi_page_usable_aligned_size_of(const mi_page_t* page, const void* p) mi_attr_noexcept {
const mi_block_t* block = _mi_page_ptr_unalign(page, p);
const size_t size = mi_page_usable_size_of(page, block);
const ptrdiff_t adjust = (uint8_t*)p - (uint8_t*)block;
mi_assert_internal(adjust >= 0 && (size_t)adjust <= size);
const size_t aligned_size = (size - adjust);
#if MI_GUARDED
if (mi_block_ptr_is_guarded(block, p)) {
return aligned_size - _mi_os_page_size();
}
#endif
return aligned_size;
}
static inline mi_page_t* mi_validate_ptr_page(const void* p, const char* msg) {
const mi_segment_t* const segment = mi_checked_ptr_segment(p, msg);
if mi_unlikely(segment==NULL) return NULL;
mi_page_t* const page = _mi_segment_page_of(segment, p);
return page;
}
static inline size_t _mi_usable_size(const void* p, const mi_page_t* page) mi_attr_noexcept {
if mi_unlikely(page==NULL) return 0;
if mi_likely(!mi_page_has_aligned(page)) {
const mi_block_t* block = (const mi_block_t*)p;
return mi_page_usable_size_of(page, block);
}
else {
// split out to separate routine for improved code generation
return mi_page_usable_aligned_size_of(page, p);
}
}
mi_decl_nodiscard size_t mi_usable_size(const void* p) mi_attr_noexcept {
const mi_page_t* const page = mi_validate_ptr_page(p,"mi_usable_size");
return _mi_usable_size(p,page);
}
// ------------------------------------------------------
// Free variants
// ------------------------------------------------------
void mi_free_size(void* p, size_t size) mi_attr_noexcept {
MI_UNUSED_RELEASE(size);
#if MI_DEBUG
const mi_page_t* const page = mi_validate_ptr_page(p,"mi_free_size");
const size_t available = _mi_usable_size(p,page);
mi_assert(p == NULL || size <= available || available == 0 /* invalid pointer */ );
#endif
mi_free(p);
}
void mi_free_size_aligned(void* p, size_t size, size_t alignment) mi_attr_noexcept {
MI_UNUSED_RELEASE(alignment);
mi_assert(((uintptr_t)p % alignment) == 0);
mi_free_size(p,size);
}
void mi_free_aligned(void* p, size_t alignment) mi_attr_noexcept {
MI_UNUSED_RELEASE(alignment);
mi_assert(((uintptr_t)p % alignment) == 0);
mi_free(p);
}
// ------------------------------------------------------
// Check for double free in secure and debug mode
// This is somewhat expensive so only enabled for secure mode 4
// ------------------------------------------------------
#if (MI_ENCODE_FREELIST && (MI_SECURE>=4 || MI_DEBUG!=0))
// linear check if the free list contains a specific element
static bool mi_list_contains(const mi_page_t* page, const mi_block_t* list, const mi_block_t* elem) {
while (list != NULL) {
if (elem==list) return true;
list = mi_block_next(page, list);
}
return false;
}
static mi_decl_noinline bool mi_check_is_double_freex(const mi_page_t* page, const mi_block_t* block) {
// The decoded value is in the same page (or NULL).
// Walk the free lists to verify positively if it is already freed
if (mi_list_contains(page, page->free, block) ||
mi_list_contains(page, page->local_free, block) ||
mi_list_contains(page, mi_page_thread_free(page), block))
{
_mi_error_message(EAGAIN, "double free detected of block %p with size %zu\n", block, mi_page_block_size(page));
return true;
}
return false;
}
#define mi_track_page(page,access) { size_t psize; void* pstart = _mi_page_start(_mi_page_segment(page),page,&psize); mi_track_mem_##access( pstart, psize); }
static inline bool mi_check_is_double_free(const mi_page_t* page, const mi_block_t* block) {
bool is_double_free = false;
mi_block_t* n = mi_block_nextx(page, block, page->keys); // pretend it is freed, and get the decoded first field
if (((uintptr_t)n & (MI_INTPTR_SIZE-1))==0 && // quick check: aligned pointer?
(n==NULL || mi_is_in_same_page(block, n))) // quick check: in same page or NULL?
{
// Suspicious: decoded value a in block is in the same page (or NULL) -- maybe a double free?
// (continue in separate function to improve code generation)
is_double_free = mi_check_is_double_freex(page, block);
}
return is_double_free;
}
#else
static inline bool mi_check_is_double_free(const mi_page_t* page, const mi_block_t* block) {
MI_UNUSED(page);
MI_UNUSED(block);
return false;
}
#endif
// ---------------------------------------------------------------------------
// Check for heap block overflow by setting up padding at the end of the block
// ---------------------------------------------------------------------------
#if MI_PADDING // && !MI_TRACK_ENABLED
static bool mi_page_decode_padding(const mi_page_t* page, const mi_block_t* block, size_t* delta, size_t* bsize) {
*bsize = mi_page_usable_block_size(page);
const mi_padding_t* const padding = (mi_padding_t*)((uint8_t*)block + *bsize);
mi_track_mem_defined(padding,sizeof(mi_padding_t));
*delta = padding->delta;
uint32_t canary = padding->canary;
uintptr_t keys[2];
keys[0] = page->keys[0];
keys[1] = page->keys[1];
bool ok = (mi_ptr_encode_canary(page,block,keys) == canary && *delta <= *bsize);
mi_track_mem_noaccess(padding,sizeof(mi_padding_t));
return ok;
}
// Return the exact usable size of a block.
static size_t mi_page_usable_size_of(const mi_page_t* page, const mi_block_t* block) {
size_t bsize;
size_t delta;
bool ok = mi_page_decode_padding(page, block, &delta, &bsize);
mi_assert_internal(ok); mi_assert_internal(delta <= bsize);
return (ok ? bsize - delta : 0);
}
// When a non-thread-local block is freed, it becomes part of the thread delayed free
// list that is freed later by the owning heap. If the exact usable size is too small to
// contain the pointer for the delayed list, then shrink the padding (by decreasing delta)
// so it will later not trigger an overflow error in `mi_free_block`.
void _mi_padding_shrink(const mi_page_t* page, const mi_block_t* block, const size_t min_size) {
size_t bsize;
size_t delta;
bool ok = mi_page_decode_padding(page, block, &delta, &bsize);
mi_assert_internal(ok);
if (!ok || (bsize - delta) >= min_size) return; // usually already enough space
mi_assert_internal(bsize >= min_size);
if (bsize < min_size) return; // should never happen
size_t new_delta = (bsize - min_size);
mi_assert_internal(new_delta < bsize);
mi_padding_t* padding = (mi_padding_t*)((uint8_t*)block + bsize);
mi_track_mem_defined(padding,sizeof(mi_padding_t));
padding->delta = (uint32_t)new_delta;
mi_track_mem_noaccess(padding,sizeof(mi_padding_t));
}
#else
static size_t mi_page_usable_size_of(const mi_page_t* page, const mi_block_t* block) {
MI_UNUSED(block);
return mi_page_usable_block_size(page);
}
void _mi_padding_shrink(const mi_page_t* page, const mi_block_t* block, const size_t min_size) {
MI_UNUSED(page);
MI_UNUSED(block);
MI_UNUSED(min_size);
}
#endif
#if MI_PADDING && MI_PADDING_CHECK
static bool mi_verify_padding(const mi_page_t* page, const mi_block_t* block, size_t* size, size_t* wrong) {
size_t bsize;
size_t delta;
bool ok = mi_page_decode_padding(page, block, &delta, &bsize);
*size = *wrong = bsize;
if (!ok) return false;
mi_assert_internal(bsize >= delta);
*size = bsize - delta;
if (!mi_page_is_huge(page)) {
uint8_t* fill = (uint8_t*)block + bsize - delta;
const size_t maxpad = (delta > MI_MAX_ALIGN_SIZE ? MI_MAX_ALIGN_SIZE : delta); // check at most the first N padding bytes
mi_track_mem_defined(fill, maxpad);
for (size_t i = 0; i < maxpad; i++) {
if (fill[i] != MI_DEBUG_PADDING) {
*wrong = bsize - delta + i;
ok = false;
break;
}
}
mi_track_mem_noaccess(fill, maxpad);
}
return ok;
}
static void mi_check_padding(const mi_page_t* page, const mi_block_t* block) {
size_t size;
size_t wrong;
if (!mi_verify_padding(page,block,&size,&wrong)) {
_mi_error_message(EFAULT, "buffer overflow in heap block %p of size %zu: write after %zu bytes\n", block, size, wrong );
}
}
#else
static void mi_check_padding(const mi_page_t* page, const mi_block_t* block) {
MI_UNUSED(page);
MI_UNUSED(block);
}
#endif
// only maintain stats for smaller objects if requested
#if (MI_STAT>0)
static void mi_stat_free(const mi_page_t* page, const mi_block_t* block) {
MI_UNUSED(block);
mi_heap_t* const heap = mi_heap_get_default();
const size_t bsize = mi_page_usable_block_size(page);
// #if (MI_STAT>1)
// const size_t usize = mi_page_usable_size_of(page, block);
// mi_heap_stat_decrease(heap, malloc_requested, usize);
// #endif
if (bsize <= MI_MEDIUM_OBJ_SIZE_MAX) {
mi_heap_stat_decrease(heap, malloc_normal, bsize);
#if (MI_STAT > 1)
mi_heap_stat_decrease(heap, malloc_bins[_mi_bin(bsize)], 1);
#endif
}
//else if (bsize <= MI_LARGE_OBJ_SIZE_MAX) {
// mi_heap_stat_decrease(heap, malloc_large, bsize);
//}
else {
mi_heap_stat_decrease(heap, malloc_huge, bsize);
}
}
#else
static void mi_stat_free(const mi_page_t* page, const mi_block_t* block) {
MI_UNUSED(page); MI_UNUSED(block);
}
#endif
// Remove guard page when building with MI_GUARDED
#if MI_GUARDED
static void mi_block_unguard(mi_page_t* page, mi_block_t* block, void* p) {
MI_UNUSED(p);
mi_assert_internal(mi_block_ptr_is_guarded(block, p));
mi_assert_internal(mi_page_has_aligned(page));
mi_assert_internal((uint8_t*)p - (uint8_t*)block >= (ptrdiff_t)sizeof(mi_block_t));
mi_assert_internal(block->next == MI_BLOCK_TAG_GUARDED);
const size_t bsize = mi_page_block_size(page);
const size_t psize = _mi_os_page_size();
mi_assert_internal(bsize > psize);
mi_assert_internal(_mi_page_segment(page)->allow_decommit);
void* gpage = (uint8_t*)block + bsize - psize;
mi_assert_internal(_mi_is_aligned(gpage, psize));
_mi_os_unprotect(gpage, psize);
}
#endif

291
mimalloc/mimalloc-sys/mimalloc/src/heap.c
View File

@@ -32,7 +32,7 @@ static bool mi_heap_visit_pages(mi_heap_t* heap, heap_page_visitor_fun* fn, void
#if MI_DEBUG>1
size_t total = heap->page_count;
size_t count = 0;
#endif
#endif
for (size_t i = 0; i <= MI_BIN_FULL; i++) {
mi_page_queue_t* pq = &heap->pages[i];
@@ -59,7 +59,7 @@ static bool mi_heap_page_is_valid(mi_heap_t* heap, mi_page_queue_t* pq, mi_page_
MI_UNUSED(pq);
mi_assert_internal(mi_page_heap(page) == heap);
mi_segment_t* segment = _mi_page_segment(page);
mi_assert_internal(segment->thread_id == heap->thread_id);
mi_assert_internal(mi_atomic_load_relaxed(&segment->thread_id) == heap->thread_id);
mi_assert_expensive(_mi_page_is_valid(page));
return true;
}
@@ -95,6 +95,11 @@ static bool mi_heap_page_collect(mi_heap_t* heap, mi_page_queue_t* pq, mi_page_t
mi_assert_internal(mi_heap_page_is_valid(heap, pq, page, NULL, NULL));
mi_collect_t collect = *((mi_collect_t*)arg_collect);
_mi_page_free_collect(page, collect >= MI_FORCE);
if (collect == MI_FORCE) {
// note: call before a potential `_mi_page_free` as the segment may be freed if this was the last used page in that segment.
mi_segment_t* segment = _mi_page_segment(page);
_mi_segment_collect(segment, true /* force? */);
}
if (mi_page_all_free(page)) {
// no more used blocks, free the page.
// note: this will free retired pages as well.
@@ -120,21 +125,25 @@ static void mi_heap_collect_ex(mi_heap_t* heap, mi_collect_t collect)
{
if (heap==NULL || !mi_heap_is_initialized(heap)) return;
const bool force = collect >= MI_FORCE;
const bool force = (collect >= MI_FORCE);
_mi_deferred_free(heap, force);
// note: never reclaim on collect but leave it to threads that need storage to reclaim
const bool force_main =
// python/cpython#112532: we may be called from a thread that is not the owner of the heap
const bool is_main_thread = (_mi_is_main_thread() && heap->thread_id == _mi_thread_id());
// note: never reclaim on collect but leave it to threads that need storage to reclaim
const bool force_main =
#ifdef NDEBUG
collect == MI_FORCE
#else
collect >= MI_FORCE
#endif
&& _mi_is_main_thread() && mi_heap_is_backing(heap) && !heap->no_reclaim;
&& is_main_thread && mi_heap_is_backing(heap) && !heap->no_reclaim;
if (force_main) {
// the main thread is abandoned (end-of-program), try to reclaim all abandoned segments.
// if all memory is freed by now, all segments should be freed.
// note: this only collects in the current subprocess
_mi_abandoned_reclaim_all(heap, &heap->tld->segments);
}
@@ -158,16 +167,16 @@ static void mi_heap_collect_ex(mi_heap_t* heap, mi_collect_t collect)
// note: forced purge can be quite expensive if many threads are created/destroyed so we do not force on abandonment
_mi_abandoned_collect(heap, collect == MI_FORCE /* force? */, &heap->tld->segments);
// collect segment local caches
if (force) {
_mi_segment_thread_collect(&heap->tld->segments);
// if forced, collect thread data cache on program-exit (or shared library unload)
if (force && is_main_thread && mi_heap_is_backing(heap)) {
_mi_thread_data_collect(); // collect thread data cache
}
// collect regions on program-exit (or shared library unload)
if (force && _mi_is_main_thread() && mi_heap_is_backing(heap)) {
_mi_thread_data_collect(); // collect thread data cache
_mi_arena_collect(true /* force purge */, &heap->tld->stats);
}
// collect arenas (this is program wide so don't force purges on abandonment of threads)
_mi_arenas_collect(collect == MI_FORCE /* force purge? */);
// merge statistics
if (collect <= MI_FORCE) { _mi_stats_merge_thread(heap->tld); }
}
void _mi_heap_collect_abandon(mi_heap_t* heap) {
@@ -206,27 +215,48 @@ mi_heap_t* mi_heap_get_backing(void) {
return bheap;
}
mi_decl_nodiscard mi_heap_t* mi_heap_new_in_arena(mi_arena_id_t arena_id) {
mi_heap_t* bheap = mi_heap_get_backing();
mi_heap_t* heap = mi_heap_malloc_tp(bheap, mi_heap_t); // todo: OS allocate in secure mode?
if (heap == NULL) return NULL;
void _mi_heap_init(mi_heap_t* heap, mi_tld_t* tld, mi_arena_id_t arena_id, bool noreclaim, uint8_t tag) {
_mi_memcpy_aligned(heap, &_mi_heap_empty, sizeof(mi_heap_t));
heap->tld = bheap->tld;
heap->thread_id = _mi_thread_id();
heap->arena_id = arena_id;
_mi_random_split(&bheap->random, &heap->random);
heap->cookie = _mi_heap_random_next(heap) | 1;
heap->tld = tld;
heap->thread_id = _mi_thread_id();
heap->arena_id = arena_id;
heap->no_reclaim = noreclaim;
heap->tag = tag;
if (heap == tld->heap_backing) {
#if defined(_WIN32) && !defined(MI_SHARED_LIB)
_mi_random_init_weak(&heap->random); // prevent allocation failure during bcrypt dll initialization with static linking (issue #1185)
#else
_mi_random_init(&heap->random);
#endif
}
else {
_mi_random_split(&tld->heap_backing->random, &heap->random);
}
heap->cookie = _mi_heap_random_next(heap) | 1;
heap->keys[0] = _mi_heap_random_next(heap);
heap->keys[1] = _mi_heap_random_next(heap);
heap->no_reclaim = true; // don't reclaim abandoned pages or otherwise destroy is unsafe
_mi_heap_guarded_init(heap);
// push on the thread local heaps list
heap->next = heap->tld->heaps;
heap->tld->heaps = heap;
}
mi_decl_nodiscard mi_heap_t* mi_heap_new_ex(int heap_tag, bool allow_destroy, mi_arena_id_t arena_id) {
mi_heap_t* bheap = mi_heap_get_backing();
mi_heap_t* heap = mi_heap_malloc_tp(bheap, mi_heap_t); // todo: OS allocate in secure mode?
if (heap == NULL) return NULL;
mi_assert(heap_tag >= 0 && heap_tag < 256);
_mi_heap_init(heap, bheap->tld, arena_id, allow_destroy /* no reclaim? */, (uint8_t)heap_tag /* heap tag */);
return heap;
}
mi_decl_nodiscard mi_heap_t* mi_heap_new_in_arena(mi_arena_id_t arena_id) {
return mi_heap_new_ex(0 /* default heap tag */, false /* don't allow `mi_heap_destroy` */, arena_id);
}
mi_decl_nodiscard mi_heap_t* mi_heap_new(void) {
return mi_heap_new_in_arena(_mi_arena_id_none());
// don't reclaim abandoned memory or otherwise destroy is unsafe
return mi_heap_new_ex(0 /* default heap tag */, true /* no reclaim */, _mi_arena_id_none());
}
bool _mi_heap_memid_is_suitable(mi_heap_t* heap, mi_memid_t memid) {
@@ -279,6 +309,18 @@ static void mi_heap_free(mi_heap_t* heap) {
mi_free(heap);
}
// return a heap on the same thread as `heap` specialized for the specified tag (if it exists)
mi_heap_t* _mi_heap_by_tag(mi_heap_t* heap, uint8_t tag) {
if (heap->tag == tag) {
return heap;
}
for (mi_heap_t *curr = heap->tld->heaps; curr != NULL; curr = curr->next) {
if (curr->tag == tag) {
return curr;
}
}
return NULL;
}
/* -----------------------------------------------------------
Heap destroy
@@ -296,24 +338,25 @@ static bool _mi_heap_page_destroy(mi_heap_t* heap, mi_page_queue_t* pq, mi_page_
// stats
const size_t bsize = mi_page_block_size(page);
if (bsize > MI_MEDIUM_OBJ_SIZE_MAX) {
if (bsize <= MI_LARGE_OBJ_SIZE_MAX) {
mi_heap_stat_decrease(heap, large, bsize);
}
else {
mi_heap_stat_decrease(heap, huge, bsize);
//if (bsize <= MI_LARGE_OBJ_SIZE_MAX) {
// mi_heap_stat_decrease(heap, malloc_large, bsize);
//}
//else
{
mi_heap_stat_decrease(heap, malloc_huge, bsize);
}
}
#if (MI_STAT)
#if (MI_STAT>0)
_mi_page_free_collect(page, false); // update used count
const size_t inuse = page->used;
if (bsize <= MI_LARGE_OBJ_SIZE_MAX) {
mi_heap_stat_decrease(heap, normal, bsize * inuse);
#if (MI_STAT>1)
mi_heap_stat_decrease(heap, normal_bins[_mi_bin(bsize)], inuse);
#endif
mi_heap_stat_decrease(heap, malloc_normal, bsize * inuse);
#if (MI_STAT>1)
mi_heap_stat_decrease(heap, malloc_bins[_mi_bin(bsize)], inuse);
#endif
}
mi_heap_stat_decrease(heap, malloc, bsize * inuse); // todo: off for aligned blocks...
#endif
// mi_heap_stat_decrease(heap, malloc_requested, bsize * inuse); // todo: off for aligned blocks...
#endif
/// pretend it is all free now
mi_assert_internal(mi_page_thread_free(page) == NULL);
@@ -347,7 +390,13 @@ void mi_heap_destroy(mi_heap_t* heap) {
mi_assert(heap->no_reclaim);
mi_assert_expensive(mi_heap_is_valid(heap));
if (heap==NULL || !mi_heap_is_initialized(heap)) return;
#if MI_GUARDED
// _mi_warning_message("'mi_heap_destroy' called but MI_GUARDED is enabled -- using `mi_heap_delete` instead (heap at %p)\n", heap);
mi_heap_delete(heap);
return;
#else
if (!heap->no_reclaim) {
_mi_warning_message("'mi_heap_destroy' called but ignored as the heap was not created with 'allow_destroy' (heap at %p)\n", heap);
// don't free in case it may contain reclaimed pages
mi_heap_delete(heap);
}
@@ -360,12 +409,14 @@ void mi_heap_destroy(mi_heap_t* heap) {
_mi_heap_destroy_pages(heap);
mi_heap_free(heap);
}
#endif
}
// forcefully destroy all heaps in the current thread
void _mi_heap_unsafe_destroy_all(void) {
mi_heap_t* bheap = mi_heap_get_backing();
mi_heap_t* curr = bheap->tld->heaps;
void _mi_heap_unsafe_destroy_all(mi_heap_t* heap) {
mi_assert_internal(heap != NULL);
if (heap == NULL) return;
mi_heap_t* curr = heap->tld->heaps;
while (curr != NULL) {
mi_heap_t* next = curr->next;
if (curr->no_reclaim) {
@@ -416,6 +467,12 @@ static void mi_heap_absorb(mi_heap_t* heap, mi_heap_t* from) {
mi_heap_reset_pages(from);
}
// are two heaps compatible with respect to heap-tag, exclusive arena etc.
static bool mi_heaps_are_compatible(mi_heap_t* heap1, mi_heap_t* heap2) {
return (heap1->tag == heap2->tag && // store same kind of objects
heap1->arena_id == heap2->arena_id); // same arena preference
}
// Safe delete a heap without freeing any still allocated blocks in that heap.
void mi_heap_delete(mi_heap_t* heap)
{
@@ -424,9 +481,10 @@ void mi_heap_delete(mi_heap_t* heap)
mi_assert_expensive(mi_heap_is_valid(heap));
if (heap==NULL || !mi_heap_is_initialized(heap)) return;
if (!mi_heap_is_backing(heap)) {
// tranfer still used pages to the backing heap
mi_heap_absorb(heap->tld->heap_backing, heap);
mi_heap_t* bheap = heap->tld->heap_backing;
if (bheap != heap && mi_heaps_are_compatible(bheap,heap)) {
// transfer still used pages to the backing heap
mi_heap_absorb(bheap, heap);
}
else {
// the backing heap abandons its pages
@@ -474,8 +532,7 @@ static bool mi_heap_page_check_owned(mi_heap_t* heap, mi_page_queue_t* pq, mi_pa
MI_UNUSED(heap);
MI_UNUSED(pq);
bool* found = (bool*)vfound;
mi_segment_t* segment = _mi_page_segment(page);
void* start = _mi_page_start(segment, page, NULL);
void* start = mi_page_start(page);
void* end = (uint8_t*)start + (page->capacity * mi_page_block_size(page));
*found = (p >= start && p < end);
return (!*found); // continue if not found
@@ -500,54 +557,97 @@ bool mi_check_owned(const void* p) {
enable visiting all blocks of all heaps across threads
----------------------------------------------------------- */
// Separate struct to keep `mi_page_t` out of the public interface
typedef struct mi_heap_area_ex_s {
mi_heap_area_t area;
mi_page_t* page;
} mi_heap_area_ex_t;
void _mi_heap_area_init(mi_heap_area_t* area, mi_page_t* page) {
const size_t bsize = mi_page_block_size(page);
const size_t ubsize = mi_page_usable_block_size(page);
area->reserved = page->reserved * bsize;
area->committed = page->capacity * bsize;
area->blocks = mi_page_start(page);
area->used = page->used; // number of blocks in use (#553)
area->block_size = ubsize;
area->full_block_size = bsize;
area->heap_tag = page->heap_tag;
}
static bool mi_heap_area_visit_blocks(const mi_heap_area_ex_t* xarea, mi_block_visit_fun* visitor, void* arg) {
mi_assert(xarea != NULL);
if (xarea==NULL) return true;
const mi_heap_area_t* area = &xarea->area;
mi_page_t* page = xarea->page;
static void mi_get_fast_divisor(size_t divisor, uint64_t* magic, size_t* shift) {
mi_assert_internal(divisor > 0 && divisor <= UINT32_MAX);
*shift = MI_SIZE_BITS - mi_clz(divisor - 1);
*magic = ((((uint64_t)1 << 32) * (((uint64_t)1 << *shift) - divisor)) / divisor + 1);
}
static size_t mi_fast_divide(size_t n, uint64_t magic, size_t shift) {
mi_assert_internal(n <= UINT32_MAX);
const uint64_t hi = ((uint64_t)n * magic) >> 32;
return (size_t)((hi + n) >> shift);
}
bool _mi_heap_area_visit_blocks(const mi_heap_area_t* area, mi_page_t* page, mi_block_visit_fun* visitor, void* arg) {
mi_assert(area != NULL);
if (area==NULL) return true;
mi_assert(page != NULL);
if (page == NULL) return true;
_mi_page_free_collect(page,true);
_mi_page_free_collect(page,true); // collect both thread_delayed and local_free
mi_assert_internal(page->local_free == NULL);
if (page->used == 0) return true;
const size_t bsize = mi_page_block_size(page);
const size_t ubsize = mi_page_usable_block_size(page); // without padding
size_t psize;
uint8_t* pstart = _mi_page_start(_mi_page_segment(page), page, &psize);
size_t psize;
uint8_t* const pstart = _mi_segment_page_start(_mi_page_segment(page), page, &psize);
mi_heap_t* const heap = mi_page_heap(page);
const size_t bsize = mi_page_block_size(page);
const size_t ubsize = mi_page_usable_block_size(page); // without padding
// optimize page with one block
if (page->capacity == 1) {
// optimize page with one block
mi_assert_internal(page->used == 1 && page->free == NULL);
return visitor(mi_page_heap(page), area, pstart, ubsize, arg);
}
mi_assert(bsize <= UINT32_MAX);
// optimize full pages
if (page->used == page->capacity) {
uint8_t* block = pstart;
for (size_t i = 0; i < page->capacity; i++) {
if (!visitor(heap, area, block, ubsize, arg)) return false;
block += bsize;
}
return true;
}
// create a bitmap of free blocks.
#define MI_MAX_BLOCKS (MI_SMALL_PAGE_SIZE / sizeof(void*))
uintptr_t free_map[MI_MAX_BLOCKS / sizeof(uintptr_t)];
memset(free_map, 0, sizeof(free_map));
uintptr_t free_map[MI_MAX_BLOCKS / MI_INTPTR_BITS];
const uintptr_t bmapsize = _mi_divide_up(page->capacity, MI_INTPTR_BITS);
memset(free_map, 0, bmapsize * sizeof(intptr_t));
if (page->capacity % MI_INTPTR_BITS != 0) {
// mark left-over bits at the end as free
size_t shift = (page->capacity % MI_INTPTR_BITS);
uintptr_t mask = (UINTPTR_MAX << shift);
free_map[bmapsize - 1] = mask;
}
// fast repeated division by the block size
uint64_t magic;
size_t shift;
mi_get_fast_divisor(bsize, &magic, &shift);
#if MI_DEBUG>1
size_t free_count = 0;
#endif
for (mi_block_t* block = page->free; block != NULL; block = mi_block_next(page,block)) {
for (mi_block_t* block = page->free; block != NULL; block = mi_block_next(page, block)) {
#if MI_DEBUG>1
free_count++;
#endif
mi_assert_internal((uint8_t*)block >= pstart && (uint8_t*)block < (pstart + psize));
size_t offset = (uint8_t*)block - pstart;
mi_assert_internal(offset % bsize == 0);
size_t blockidx = offset / bsize; // Todo: avoid division?
mi_assert_internal( blockidx < MI_MAX_BLOCKS);
size_t bitidx = (blockidx / sizeof(uintptr_t));
size_t bit = blockidx - (bitidx * sizeof(uintptr_t));
mi_assert_internal(offset <= UINT32_MAX);
size_t blockidx = mi_fast_divide(offset, magic, shift);
mi_assert_internal(blockidx == offset / bsize);
mi_assert_internal(blockidx < MI_MAX_BLOCKS);
size_t bitidx = (blockidx / MI_INTPTR_BITS);
size_t bit = blockidx - (bitidx * MI_INTPTR_BITS);
free_map[bitidx] |= ((uintptr_t)1 << bit);
}
mi_assert_internal(page->capacity == (free_count + page->used));
@@ -556,42 +656,53 @@ static bool mi_heap_area_visit_blocks(const mi_heap_area_ex_t* xarea, mi_block_v
#if MI_DEBUG>1
size_t used_count = 0;
#endif
for (size_t i = 0; i < page->capacity; i++) {
size_t bitidx = (i / sizeof(uintptr_t));
size_t bit = i - (bitidx * sizeof(uintptr_t));
uintptr_t m = free_map[bitidx];
if (bit == 0 && m == UINTPTR_MAX) {
i += (sizeof(uintptr_t) - 1); // skip a run of free blocks
uint8_t* block = pstart;
for (size_t i = 0; i < bmapsize; i++) {
if (free_map[i] == 0) {
// every block is in use
for (size_t j = 0; j < MI_INTPTR_BITS; j++) {
#if MI_DEBUG>1
used_count++;
#endif
if (!visitor(heap, area, block, ubsize, arg)) return false;
block += bsize;
}
}
else if ((m & ((uintptr_t)1 << bit)) == 0) {
#if MI_DEBUG>1
used_count++;
#endif
uint8_t* block = pstart + (i * bsize);
if (!visitor(mi_page_heap(page), area, block, ubsize, arg)) return false;
else {
// visit the used blocks in the mask
uintptr_t m = ~free_map[i];
while (m != 0) {
#if MI_DEBUG>1
used_count++;
#endif
size_t bitidx = mi_ctz(m);
if (!visitor(heap, area, block + (bitidx * bsize), ubsize, arg)) return false;
m &= m - 1; // clear least significant bit
}
block += bsize * MI_INTPTR_BITS;
}
}
mi_assert_internal(page->used == used_count);
return true;
}
typedef bool (mi_heap_area_visit_fun)(const mi_heap_t* heap, const mi_heap_area_ex_t* area, void* arg);
// Separate struct to keep `mi_page_t` out of the public interface
typedef struct mi_heap_area_ex_s {
mi_heap_area_t area;
mi_page_t* page;
} mi_heap_area_ex_t;
typedef bool (mi_heap_area_visit_fun)(const mi_heap_t* heap, const mi_heap_area_ex_t* area, void* arg);
static bool mi_heap_visit_areas_page(mi_heap_t* heap, mi_page_queue_t* pq, mi_page_t* page, void* vfun, void* arg) {
MI_UNUSED(heap);
MI_UNUSED(pq);
mi_heap_area_visit_fun* fun = (mi_heap_area_visit_fun*)vfun;
mi_heap_area_ex_t xarea;
const size_t bsize = mi_page_block_size(page);
const size_t ubsize = mi_page_usable_block_size(page);
xarea.page = page;
xarea.area.reserved = page->reserved * bsize;
xarea.area.committed = page->capacity * bsize;
xarea.area.blocks = _mi_page_start(_mi_page_segment(page), page, NULL);
xarea.area.used = page->used; // number of blocks in use (#553)
xarea.area.block_size = ubsize;
xarea.area.full_block_size = bsize;
_mi_heap_area_init(&xarea.area, page);
return fun(heap, &xarea, arg);
}
@@ -612,7 +723,7 @@ static bool mi_heap_area_visitor(const mi_heap_t* heap, const mi_heap_area_ex_t*
mi_visit_blocks_args_t* args = (mi_visit_blocks_args_t*)arg;
if (!args->visitor(heap, &xarea->area, NULL, xarea->area.block_size, args->arg)) return false;
if (args->visit_blocks) {
return mi_heap_area_visit_blocks(xarea, args->visitor, args->arg);
return _mi_heap_area_visit_blocks(&xarea->area, xarea->page, args->visitor, args->arg);
}
else {
return true;

414
mimalloc/mimalloc-sys/mimalloc/src/init.c
View File

@@ -14,25 +14,27 @@ terms of the MIT license. A copy of the license can be found in the file
// Empty page used to initialize the small free pages array
const mi_page_t _mi_page_empty = {
0, false, false, false,
0,
false, false, false, false,
0, // capacity
0, // reserved capacity
{ 0 }, // flags
false, // is_zero
0, // retire_expire
NULL, // free
0, // used
0, // xblock_size
NULL, // local_free
0, // used
0, // block size shift
0, // heap tag
0, // block_size
NULL, // page_start
#if (MI_PADDING || MI_ENCODE_FREELIST)
{ 0, 0 },
#endif
MI_ATOMIC_VAR_INIT(0), // xthread_free
MI_ATOMIC_VAR_INIT(0), // xheap
NULL, NULL
#if MI_INTPTR_SIZE==8
, { 0 } // padding
#endif
};
#define MI_PAGE_EMPTY() ((mi_page_t*)&_mi_page_empty)
@@ -65,27 +67,26 @@ const mi_page_t _mi_page_empty = {
QNULL(MI_MEDIUM_OBJ_WSIZE_MAX + 1 /* 655360, Huge queue */), \
QNULL(MI_MEDIUM_OBJ_WSIZE_MAX + 2) /* Full queue */ }
#define MI_STAT_COUNT_NULL() {0,0,0,0}
#define MI_STAT_COUNT_NULL() {0,0,0}
// Empty statistics
#if MI_STAT>1
#define MI_STAT_COUNT_END_NULL() , { MI_STAT_COUNT_NULL(), MI_INIT32(MI_STAT_COUNT_NULL) }
#else
#define MI_STAT_COUNT_END_NULL()
#endif
#define MI_STATS_NULL \
MI_STAT_COUNT_NULL(), MI_STAT_COUNT_NULL(), \
MI_STAT_COUNT_NULL(), MI_STAT_COUNT_NULL(), \
MI_STAT_COUNT_NULL(), MI_STAT_COUNT_NULL(), \
MI_STAT_COUNT_NULL(), MI_STAT_COUNT_NULL(), \
MI_STAT_COUNT_NULL(), MI_STAT_COUNT_NULL(), \
MI_STAT_COUNT_NULL(), MI_STAT_COUNT_NULL(), \
MI_STAT_COUNT_NULL(), MI_STAT_COUNT_NULL(), \
MI_STAT_COUNT_NULL(), \
{ 0, 0 }, { 0, 0 }, { 0, 0 }, { 0, 0 }, \
{ 0, 0 }, { 0, 0 }, { 0, 0 }, { 0, 0 }, { 0, 0 }, { 0, 0 } \
MI_STAT_COUNT_END_NULL()
MI_STAT_COUNT_NULL(), MI_STAT_COUNT_NULL(), MI_STAT_COUNT_NULL(), \
{ 0 }, { 0 }, \
MI_STAT_COUNT_NULL(), MI_STAT_COUNT_NULL(), MI_STAT_COUNT_NULL(), \
MI_STAT_COUNT_NULL(), MI_STAT_COUNT_NULL(), MI_STAT_COUNT_NULL(), \
{ 0 }, { 0 }, { 0 }, { 0 }, \
{ 0 }, { 0 }, { 0 }, { 0 }, \
\
{ 0 }, { 0 }, { 0 }, { 0 }, { 0 }, { 0 }, \
MI_INIT4(MI_STAT_COUNT_NULL), \
{ 0 }, { 0 }, { 0 }, { 0 }, \
\
{ MI_INIT4(MI_STAT_COUNT_NULL) }, \
{ { 0 }, { 0 }, { 0 }, { 0 } }, \
\
{ MI_INIT74(MI_STAT_COUNT_NULL) }, \
{ MI_INIT74(MI_STAT_COUNT_NULL) }
// Empty slice span queues for every bin
@@ -110,8 +111,6 @@ const mi_page_t _mi_page_empty = {
mi_decl_cache_align const mi_heap_t _mi_heap_empty = {
NULL,
MI_SMALL_PAGES_EMPTY,
MI_PAGE_QUEUES_EMPTY,
MI_ATOMIC_VAR_INIT(NULL),
0, // tid
0, // cookie
@@ -120,20 +119,27 @@ mi_decl_cache_align const mi_heap_t _mi_heap_empty = {
{ {0}, {0}, 0, true }, // random
0, // page count
MI_BIN_FULL, 0, // page retired min/max
0, 0, // generic count
NULL, // next
false
false, // can reclaim
0, // tag
#if MI_GUARDED
0, 0, 0, 1, // count is 1 so we never write to it (see `internal.h:mi_heap_malloc_use_guarded`)
#endif
MI_SMALL_PAGES_EMPTY,
MI_PAGE_QUEUES_EMPTY
};
static mi_decl_cache_align mi_subproc_t mi_subproc_default;
#define tld_empty_stats ((mi_stats_t*)((uint8_t*)&tld_empty + offsetof(mi_tld_t,stats)))
#define tld_empty_os ((mi_os_tld_t*)((uint8_t*)&tld_empty + offsetof(mi_tld_t,os)))
mi_decl_cache_align static const mi_tld_t tld_empty = {
0,
false,
NULL, NULL,
{ MI_SEGMENT_SPAN_QUEUES_EMPTY, 0, 0, 0, 0, tld_empty_stats, tld_empty_os }, // segments
{ 0, tld_empty_stats }, // os
{ MI_STATS_NULL } // stats
{ MI_SEGMENT_SPAN_QUEUES_EMPTY, 0, 0, 0, 0, 0, &mi_subproc_default, tld_empty_stats }, // segments
{ MI_STAT_VERSION, MI_STATS_NULL } // stats
};
mi_threadid_t _mi_thread_id(void) mi_attr_noexcept {
@@ -143,20 +149,17 @@ mi_threadid_t _mi_thread_id(void) mi_attr_noexcept {
// the thread-local default heap for allocation
mi_decl_thread mi_heap_t* _mi_heap_default = (mi_heap_t*)&_mi_heap_empty;
extern mi_heap_t _mi_heap_main;
extern mi_decl_hidden mi_heap_t _mi_heap_main;
static mi_tld_t tld_main = {
static mi_decl_cache_align mi_tld_t tld_main = {
0, false,
&_mi_heap_main, & _mi_heap_main,
{ MI_SEGMENT_SPAN_QUEUES_EMPTY, 0, 0, 0, 0, &tld_main.stats, &tld_main.os }, // segments
{ 0, &tld_main.stats }, // os
{ MI_STATS_NULL } // stats
{ MI_SEGMENT_SPAN_QUEUES_EMPTY, 0, 0, 0, 0, 0, &mi_subproc_default, &tld_main.stats }, // segments
{ MI_STAT_VERSION, MI_STATS_NULL } // stats
};
mi_heap_t _mi_heap_main = {
mi_decl_cache_align mi_heap_t _mi_heap_main = {
&tld_main,
MI_SMALL_PAGES_EMPTY,
MI_PAGE_QUEUES_EMPTY,
MI_ATOMIC_VAR_INIT(NULL),
0, // thread id
0, // initial cookie
@@ -165,13 +168,58 @@ mi_heap_t _mi_heap_main = {
{ {0x846ca68b}, {0}, 0, true }, // random
0, // page count
MI_BIN_FULL, 0, // page retired min/max
0, 0, // generic count
NULL, // next heap
false // can reclaim
false, // can reclaim
0, // tag
#if MI_GUARDED
0, 0, 0, 0,
#endif
MI_SMALL_PAGES_EMPTY,
MI_PAGE_QUEUES_EMPTY
};
bool _mi_process_is_initialized = false; // set to `true` in `mi_process_init`.
mi_stats_t _mi_stats_main = { MI_STATS_NULL };
mi_stats_t _mi_stats_main = { MI_STAT_VERSION, MI_STATS_NULL };
#if MI_GUARDED
mi_decl_export void mi_heap_guarded_set_sample_rate(mi_heap_t* heap, size_t sample_rate, size_t seed) {
heap->guarded_sample_rate = sample_rate;
heap->guarded_sample_count = sample_rate; // count down samples
if (heap->guarded_sample_rate > 1) {
if (seed == 0) {
seed = _mi_heap_random_next(heap);
}
heap->guarded_sample_count = (seed % heap->guarded_sample_rate) + 1; // start at random count between 1 and `sample_rate`
}
}
mi_decl_export void mi_heap_guarded_set_size_bound(mi_heap_t* heap, size_t min, size_t max) {
heap->guarded_size_min = min;
heap->guarded_size_max = (min > max ? min : max);
}
void _mi_heap_guarded_init(mi_heap_t* heap) {
mi_heap_guarded_set_sample_rate(heap,
(size_t)mi_option_get_clamp(mi_option_guarded_sample_rate, 0, LONG_MAX),
(size_t)mi_option_get(mi_option_guarded_sample_seed));
mi_heap_guarded_set_size_bound(heap,
(size_t)mi_option_get_clamp(mi_option_guarded_min, 0, LONG_MAX),
(size_t)mi_option_get_clamp(mi_option_guarded_max, 0, LONG_MAX) );
}
#else
mi_decl_export void mi_heap_guarded_set_sample_rate(mi_heap_t* heap, size_t sample_rate, size_t seed) {
MI_UNUSED(heap); MI_UNUSED(sample_rate); MI_UNUSED(seed);
}
mi_decl_export void mi_heap_guarded_set_size_bound(mi_heap_t* heap, size_t min, size_t max) {
MI_UNUSED(heap); MI_UNUSED(min); MI_UNUSED(max);
}
void _mi_heap_guarded_init(mi_heap_t* heap) {
MI_UNUSED(heap);
}
#endif
static void mi_heap_main_init(void) {
@@ -186,6 +234,9 @@ static void mi_heap_main_init(void) {
_mi_heap_main.cookie = _mi_heap_random_next(&_mi_heap_main);
_mi_heap_main.keys[0] = _mi_heap_random_next(&_mi_heap_main);
_mi_heap_main.keys[1] = _mi_heap_random_next(&_mi_heap_main);
mi_lock_init(&mi_subproc_default.abandoned_os_lock);
mi_lock_init(&mi_subproc_default.abandoned_os_visit_lock);
_mi_heap_guarded_init(&_mi_heap_main);
}
}
@@ -194,6 +245,56 @@ mi_heap_t* _mi_heap_main_get(void) {
return &_mi_heap_main;
}
/* -----------------------------------------------------------
Sub process
----------------------------------------------------------- */
mi_subproc_id_t mi_subproc_main(void) {
return NULL;
}
mi_subproc_id_t mi_subproc_new(void) {
mi_memid_t memid = _mi_memid_none();
mi_subproc_t* subproc = (mi_subproc_t*)_mi_arena_meta_zalloc(sizeof(mi_subproc_t), &memid);
if (subproc == NULL) return NULL;
subproc->memid = memid;
subproc->abandoned_os_list = NULL;
mi_lock_init(&subproc->abandoned_os_lock);
mi_lock_init(&subproc->abandoned_os_visit_lock);
return subproc;
}
mi_subproc_t* _mi_subproc_from_id(mi_subproc_id_t subproc_id) {
return (subproc_id == NULL ? &mi_subproc_default : (mi_subproc_t*)subproc_id);
}
void mi_subproc_delete(mi_subproc_id_t subproc_id) {
if (subproc_id == NULL) return;
mi_subproc_t* subproc = _mi_subproc_from_id(subproc_id);
// check if there are no abandoned segments still..
bool safe_to_delete = false;
mi_lock(&subproc->abandoned_os_lock) {
if (subproc->abandoned_os_list == NULL) {
safe_to_delete = true;
}
}
if (!safe_to_delete) return;
// safe to release
// todo: should we refcount subprocesses?
mi_lock_done(&subproc->abandoned_os_lock);
mi_lock_done(&subproc->abandoned_os_visit_lock);
_mi_arena_meta_free(subproc, subproc->memid, sizeof(mi_subproc_t));
}
void mi_subproc_add_current_thread(mi_subproc_id_t subproc_id) {
mi_heap_t* heap = mi_heap_get_default();
if (heap == NULL) return;
mi_assert(heap->tld->segments.subproc == &mi_subproc_default);
if (heap->tld->segments.subproc != &mi_subproc_default) return;
heap->tld->segments.subproc = _mi_subproc_from_id(subproc_id);
}
/* -----------------------------------------------------------
Initialization and freeing of the thread local heaps
@@ -201,9 +302,9 @@ mi_heap_t* _mi_heap_main_get(void) {
// note: in x64 in release build `sizeof(mi_thread_data_t)` is under 4KiB (= OS page size).
typedef struct mi_thread_data_s {
mi_heap_t heap; // must come first due to cast in `_mi_heap_done`
mi_heap_t heap; // must come first due to cast in `_mi_heap_done`
mi_tld_t tld;
mi_memid_t memid;
mi_memid_t memid; // must come last due to zero'ing
} mi_thread_data_t;
@@ -212,12 +313,11 @@ typedef struct mi_thread_data_s {
// destroy many OS threads, this may causes too much overhead
// per thread so we maintain a small cache of recently freed metadata.
#define TD_CACHE_SIZE (16)
#define TD_CACHE_SIZE (32)
static _Atomic(mi_thread_data_t*) td_cache[TD_CACHE_SIZE];
static mi_thread_data_t* mi_thread_data_zalloc(void) {
// try to find thread metadata in the cache
bool is_zero = false;
mi_thread_data_t* td = NULL;
for (int i = 0; i < TD_CACHE_SIZE; i++) {
td = mi_atomic_load_ptr_relaxed(mi_thread_data_t, &td_cache[i]);
@@ -225,32 +325,25 @@ static mi_thread_data_t* mi_thread_data_zalloc(void) {
// found cached allocation, try use it
td = mi_atomic_exchange_ptr_acq_rel(mi_thread_data_t, &td_cache[i], NULL);
if (td != NULL) {
break;
_mi_memzero(td, offsetof(mi_thread_data_t,memid));
return td;
}
}
}
// if that fails, allocate as meta data
mi_memid_t memid;
td = (mi_thread_data_t*)_mi_os_zalloc(sizeof(mi_thread_data_t), &memid);
if (td == NULL) {
mi_memid_t memid;
td = (mi_thread_data_t*)_mi_os_alloc(sizeof(mi_thread_data_t), &memid, &_mi_stats_main);
// if this fails, try once more. (issue #257)
td = (mi_thread_data_t*)_mi_os_zalloc(sizeof(mi_thread_data_t), &memid);
if (td == NULL) {
// if this fails, try once more. (issue #257)
td = (mi_thread_data_t*)_mi_os_alloc(sizeof(mi_thread_data_t), &memid, &_mi_stats_main);
if (td == NULL) {
// really out of memory
_mi_error_message(ENOMEM, "unable to allocate thread local heap metadata (%zu bytes)\n", sizeof(mi_thread_data_t));
}
}
if (td != NULL) {
td->memid = memid;
is_zero = memid.initially_zero;
// really out of memory
_mi_error_message(ENOMEM, "unable to allocate thread local heap metadata (%zu bytes)\n", sizeof(mi_thread_data_t));
return NULL;
}
}
if (td != NULL && !is_zero) {
_mi_memzero_aligned(td, sizeof(*td));
}
td->memid = memid;
return td;
}
@@ -266,7 +359,7 @@ static void mi_thread_data_free( mi_thread_data_t* tdfree ) {
}
}
// if that fails, just free it directly
_mi_os_free(tdfree, sizeof(mi_thread_data_t), tdfree->memid, &_mi_stats_main);
_mi_os_free(tdfree, sizeof(mi_thread_data_t), tdfree->memid);
}
void _mi_thread_data_collect(void) {
@@ -276,14 +369,14 @@ void _mi_thread_data_collect(void) {
if (td != NULL) {
td = mi_atomic_exchange_ptr_acq_rel(mi_thread_data_t, &td_cache[i], NULL);
if (td != NULL) {
_mi_os_free(td, sizeof(mi_thread_data_t), td->memid, &_mi_stats_main);
_mi_os_free(td, sizeof(mi_thread_data_t), td->memid);
}
}
}
}
// Initialize the thread local default heap, called from `mi_thread_init`
static bool _mi_heap_init(void) {
static bool _mi_thread_heap_init(void) {
if (mi_heap_is_initialized(mi_prim_get_default_heap())) return true;
if (_mi_is_main_thread()) {
// mi_assert_internal(_mi_heap_main.thread_id != 0); // can happen on freeBSD where alloc is called before any initialization
@@ -299,26 +392,24 @@ static bool _mi_heap_init(void) {
mi_tld_t* tld = &td->tld;
mi_heap_t* heap = &td->heap;
_mi_memcpy_aligned(tld, &tld_empty, sizeof(*tld));
_mi_memcpy_aligned(heap, &_mi_heap_empty, sizeof(*heap));
heap->thread_id = _mi_thread_id();
_mi_random_init(&heap->random);
heap->cookie = _mi_heap_random_next(heap) | 1;
heap->keys[0] = _mi_heap_random_next(heap);
heap->keys[1] = _mi_heap_random_next(heap);
heap->tld = tld;
tld->heap_backing = heap;
tld->heaps = heap;
tld->segments.stats = &tld->stats;
tld->segments.os = &tld->os;
tld->os.stats = &tld->stats;
_mi_tld_init(tld, heap); // must be before `_mi_heap_init`
_mi_heap_init(heap, tld, _mi_arena_id_none(), false /* can reclaim */, 0 /* default tag */);
_mi_heap_set_default_direct(heap);
}
return false;
}
// initialize thread local data
void _mi_tld_init(mi_tld_t* tld, mi_heap_t* bheap) {
_mi_memcpy_aligned(tld, &tld_empty, sizeof(mi_tld_t));
tld->heap_backing = bheap;
tld->heaps = NULL;
tld->segments.subproc = &mi_subproc_default;
tld->segments.stats = &tld->stats;
}
// Free the thread local default heap (called from `mi_thread_done`)
static bool _mi_heap_done(mi_heap_t* heap) {
static bool _mi_thread_heap_done(mi_heap_t* heap) {
if (!mi_heap_is_initialized(heap)) return true;
// reset default heap
@@ -415,7 +506,7 @@ void mi_thread_init(void) mi_attr_noexcept
// initialize the thread local default heap
// (this will call `_mi_heap_set_default_direct` and thus set the
// fiber/pthread key to a non-zero value, ensuring `_mi_thread_done` is called)
if (_mi_heap_init()) return; // returns true if already initialized
if (_mi_thread_heap_init()) return; // returns true if already initialized
_mi_stat_increase(&_mi_stats_main.threads, 1);
mi_atomic_increment_relaxed(&thread_count);
@@ -426,28 +517,28 @@ void mi_thread_done(void) mi_attr_noexcept {
_mi_thread_done(NULL);
}
void _mi_thread_done(mi_heap_t* heap)
void _mi_thread_done(mi_heap_t* heap)
{
// calling with NULL implies using the default heap
if (heap == NULL) {
heap = mi_prim_get_default_heap();
if (heap == NULL) {
heap = mi_prim_get_default_heap();
if (heap == NULL) return;
}
// prevent re-entrancy through heap_done/heap_set_default_direct (issue #699)
if (!mi_heap_is_initialized(heap)) {
return;
return;
}
// adjust stats
mi_atomic_decrement_relaxed(&thread_count);
_mi_stat_decrease(&_mi_stats_main.threads, 1);
// check thread-id as on Windows shutdown with FLS the main (exit) thread may call this on thread-local heaps...
if (heap->thread_id != _mi_thread_id()) return;
// abandon the thread local heap
if (_mi_heap_done(heap)) return; // returns true if already ran
if (_mi_thread_heap_done(heap)) return; // returns true if already ran
}
void _mi_heap_set_default_direct(mi_heap_t* heap) {
@@ -455,7 +546,7 @@ void _mi_heap_set_default_direct(mi_heap_t* heap) {
#if defined(MI_TLS_SLOT)
mi_prim_tls_slot_set(MI_TLS_SLOT,heap);
#elif defined(MI_TLS_PTHREAD_SLOT_OFS)
*mi_tls_pthread_heap_slot() = heap;
*mi_prim_tls_pthread_heap_slot() = heap;
#elif defined(MI_TLS_PTHREAD)
// we use _mi_heap_default_key
#else
@@ -464,61 +555,30 @@ void _mi_heap_set_default_direct(mi_heap_t* heap) {
// ensure the default heap is passed to `_mi_thread_done`
// setting to a non-NULL value also ensures `mi_thread_done` is called.
_mi_prim_thread_associate_default_heap(heap);
_mi_prim_thread_associate_default_heap(heap);
}
void mi_thread_set_in_threadpool(void) mi_attr_noexcept {
// nothing
}
// --------------------------------------------------------
// Run functions on process init/done, and thread init/done
// --------------------------------------------------------
static void mi_cdecl mi_process_done(void);
static bool os_preloading = true; // true until this module is initialized
static bool mi_redirected = false; // true if malloc redirects to mi_malloc
// Returns true if this module has not been initialized; Don't use C runtime routines until it returns false.
bool mi_decl_noinline _mi_preloading(void) {
return os_preloading;
}
// Returns true if mimalloc was redirected
mi_decl_nodiscard bool mi_is_redirected(void) mi_attr_noexcept {
return mi_redirected;
return _mi_is_redirected();
}
// Communicate with the redirection module on Windows
#if defined(_WIN32) && defined(MI_SHARED_LIB) && !defined(MI_WIN_NOREDIRECT)
#ifdef __cplusplus
extern "C" {
#endif
mi_decl_export void _mi_redirect_entry(DWORD reason) {
// called on redirection; careful as this may be called before DllMain
if (reason == DLL_PROCESS_ATTACH) {
mi_redirected = true;
}
else if (reason == DLL_PROCESS_DETACH) {
mi_redirected = false;
}
else if (reason == DLL_THREAD_DETACH) {
mi_thread_done();
}
}
__declspec(dllimport) bool mi_cdecl mi_allocator_init(const char** message);
__declspec(dllimport) void mi_cdecl mi_allocator_done(void);
#ifdef __cplusplus
}
#endif
#else
static bool mi_allocator_init(const char** message) {
if (message != NULL) *message = NULL;
return true;
}
static void mi_allocator_done(void) {
// nothing to do
}
#endif
// Called once by the process loader
static void mi_process_load(void) {
// Called once by the process loader from `src/prim/prim.c`
void _mi_auto_process_init(void) {
mi_heap_main_init();
#if defined(__APPLE__) || defined(MI_TLS_RECURSE_GUARD)
volatile mi_heap_t* dummy = _mi_heap_default; // access TLS to allocate it before setting tls_initialized to true;
@@ -526,17 +586,14 @@ static void mi_process_load(void) {
#endif
os_preloading = false;
mi_assert_internal(_mi_is_main_thread());
#if !(defined(_WIN32) && defined(MI_SHARED_LIB)) // use Dll process detach (see below) instead of atexit (issue #521)
atexit(&mi_process_done);
#endif
_mi_options_init();
mi_process_setup_auto_thread_done();
mi_process_init();
if (mi_redirected) _mi_verbose_message("malloc is redirected.\n");
if (_mi_is_redirected()) _mi_verbose_message("malloc is redirected.\n");
// show message from the redirector (if present)
const char* msg = NULL;
mi_allocator_init(&msg);
_mi_allocator_init(&msg);
if (msg != NULL && (mi_option_is_enabled(mi_option_verbose) || mi_option_is_enabled(mi_option_show_errors))) {
_mi_fputs(NULL,NULL,NULL,msg);
}
@@ -548,12 +605,15 @@ static void mi_process_load(void) {
#if defined(_WIN32) && (defined(_M_IX86) || defined(_M_X64))
#include <intrin.h>
mi_decl_cache_align bool _mi_cpu_has_fsrm = false;
mi_decl_cache_align bool _mi_cpu_has_erms = false;
static void mi_detect_cpu_features(void) {
// FSRM for fast rep movsb support (AMD Zen3+ (~2020) or Intel Ice Lake+ (~2017))
// FSRM for fast short rep movsb/stosb support (AMD Zen3+ (~2020) or Intel Ice Lake+ (~2017))
// EMRS for fast enhanced rep movsb/stosb support
int32_t cpu_info[4];
__cpuid(cpu_info, 7);
_mi_cpu_has_fsrm = ((cpu_info[3] & (1 << 4)) != 0); // bit 4 of EDX : see <https://en.wikipedia.org/wiki/CPUID#EAX=7,_ECX=0:_Extended_Features>
_mi_cpu_has_erms = ((cpu_info[1] & (1 << 9)) != 0); // bit 9 of EBX : see <https://en.wikipedia.org/wiki/CPUID#EAX=7,_ECX=0:_Extended_Features>
}
#else
static void mi_detect_cpu_features(void) {
@@ -576,14 +636,6 @@ void mi_process_init(void) mi_attr_noexcept {
mi_detect_cpu_features();
_mi_os_init();
mi_heap_main_init();
#if MI_DEBUG
_mi_verbose_message("debug level : %d\n", MI_DEBUG);
#endif
_mi_verbose_message("secure level: %d\n", MI_SECURE);
_mi_verbose_message("mem tracking: %s\n", MI_TRACK_TOOL);
#if MI_TSAN
_mi_verbose_message("thread santizer enabled\n");
#endif
mi_thread_init();
#if defined(_WIN32)
@@ -598,7 +650,7 @@ void mi_process_init(void) mi_attr_noexcept {
if (mi_option_is_enabled(mi_option_reserve_huge_os_pages)) {
size_t pages = mi_option_get_clamp(mi_option_reserve_huge_os_pages, 0, 128*1024);
long reserve_at = mi_option_get(mi_option_reserve_huge_os_pages_at);
int reserve_at = (int)mi_option_get_clamp(mi_option_reserve_huge_os_pages_at, -1, INT_MAX);
if (reserve_at != -1) {
mi_reserve_huge_os_pages_at(pages, reserve_at, pages*500);
} else {
@@ -613,8 +665,8 @@ void mi_process_init(void) mi_attr_noexcept {
}
}
// Called when the process is done (through `at_exit`)
static void mi_cdecl mi_process_done(void) {
// Called when the process is done (cdecl as it is used with `at_exit` on some platforms)
void mi_cdecl mi_process_done(void) mi_attr_noexcept {
// only shutdown if we were initialized
if (!_mi_process_is_initialized) return;
// ensure we are called once
@@ -622,15 +674,20 @@ static void mi_cdecl mi_process_done(void) {
if (process_done) return;
process_done = true;
// get the default heap so we don't need to acces thread locals anymore
mi_heap_t* heap = mi_prim_get_default_heap(); // use prim to not initialize any heap
mi_assert_internal(heap != NULL);
// release any thread specific resources and ensure _mi_thread_done is called on all but the main thread
_mi_prim_thread_done_auto_done();
#ifndef MI_SKIP_COLLECT_ON_EXIT
#if (MI_DEBUG || !defined(MI_SHARED_LIB))
// free all memory if possible on process exit. This is not needed for a stand-alone process
// but should be done if mimalloc is statically linked into another shared library which
// is repeatedly loaded/unloaded, see issue #281.
mi_collect(true /* force */ );
mi_heap_collect(heap, true /* force */ );
#endif
#endif
@@ -638,72 +695,21 @@ static void mi_cdecl mi_process_done(void) {
// since after process_done there might still be other code running that calls `free` (like at_exit routines,
// or C-runtime termination code.
if (mi_option_is_enabled(mi_option_destroy_on_exit)) {
mi_collect(true /* force */);
_mi_heap_unsafe_destroy_all(); // forcefully release all memory held by all heaps (of this thread only!)
_mi_arena_unsafe_destroy_all(& _mi_heap_main_get()->tld->stats);
mi_heap_collect(heap, true /* force */);
_mi_heap_unsafe_destroy_all(heap); // forcefully release all memory held by all heaps (of this thread only!)
_mi_arena_unsafe_destroy_all();
_mi_segment_map_unsafe_destroy();
}
if (mi_option_is_enabled(mi_option_show_stats) || mi_option_is_enabled(mi_option_verbose)) {
mi_stats_print(NULL);
}
mi_allocator_done();
_mi_allocator_done();
_mi_verbose_message("process done: 0x%zx\n", _mi_heap_main.thread_id);
os_preloading = true; // don't call the C runtime anymore
}
#if defined(_WIN32) && defined(MI_SHARED_LIB)
// Windows DLL: easy to hook into process_init and thread_done
__declspec(dllexport) BOOL WINAPI DllMain(HINSTANCE inst, DWORD reason, LPVOID reserved) {
MI_UNUSED(reserved);
MI_UNUSED(inst);
if (reason==DLL_PROCESS_ATTACH) {
mi_process_load();
}
else if (reason==DLL_PROCESS_DETACH) {
mi_process_done();
}
else if (reason==DLL_THREAD_DETACH) {
if (!mi_is_redirected()) {
mi_thread_done();
}
}
return TRUE;
}
#elif defined(_MSC_VER)
// MSVC: use data section magic for static libraries
// See <https://www.codeguru.com/cpp/misc/misc/applicationcontrol/article.php/c6945/Running-Code-Before-and-After-Main.htm>
static int _mi_process_init(void) {
mi_process_load();
return 0;
}
typedef int(*_mi_crt_callback_t)(void);
#if defined(_M_X64) || defined(_M_ARM64)
__pragma(comment(linker, "/include:" "_mi_msvc_initu"))
#pragma section(".CRT$XIU", long, read)
#else
__pragma(comment(linker, "/include:" "__mi_msvc_initu"))
#endif
#pragma data_seg(".CRT$XIU")
mi_decl_externc _mi_crt_callback_t _mi_msvc_initu[] = { &_mi_process_init };
#pragma data_seg()
#elif defined(__cplusplus)
// C++: use static initialization to detect process start
static bool _mi_process_init(void) {
mi_process_load();
return (_mi_heap_main.thread_id != 0);
}
static bool mi_initialized = _mi_process_init();
#elif defined(__GNUC__) || defined(__clang__)
// GCC,Clang: use the constructor attribute
static void __attribute__((constructor)) _mi_process_init(void) {
mi_process_load();
}
#else
#pragma message("define a way to call mi_process_load on your platform")
#endif
void mi_cdecl _mi_auto_process_done(void) mi_attr_noexcept {
if (_mi_option_get_fast(mi_option_destroy_on_exit)>1) return;
mi_process_done();
}

334
mimalloc/mimalloc-sys/mimalloc/src/libc.c Normal file
View File

@@ -0,0 +1,334 @@
/* ----------------------------------------------------------------------------
Copyright (c) 2018-2023, Microsoft Research, Daan Leijen
This is free software; you can redistribute it and/or modify it under the
terms of the MIT license. A copy of the license can be found in the file
"LICENSE" at the root of this distribution.
-----------------------------------------------------------------------------*/
// --------------------------------------------------------
// This module defines various std libc functions to reduce
// the dependency on libc, and also prevent errors caused
// by some libc implementations when called before `main`
// executes (due to malloc redirection)
// --------------------------------------------------------
#include "mimalloc.h"
#include "mimalloc/internal.h"
#include "mimalloc/prim.h" // mi_prim_getenv
char _mi_toupper(char c) {
if (c >= 'a' && c <= 'z') return (c - 'a' + 'A');
else return c;
}
int _mi_strnicmp(const char* s, const char* t, size_t n) {
if (n == 0) return 0;
for (; *s != 0 && *t != 0 && n > 0; s++, t++, n--) {
if (_mi_toupper(*s) != _mi_toupper(*t)) break;
}
return (n == 0 ? 0 : *s - *t);
}
void _mi_strlcpy(char* dest, const char* src, size_t dest_size) {
if (dest==NULL || src==NULL || dest_size == 0) return;
// copy until end of src, or when dest is (almost) full
while (*src != 0 && dest_size > 1) {
*dest++ = *src++;
dest_size--;
}
// always zero terminate
*dest = 0;
}
void _mi_strlcat(char* dest, const char* src, size_t dest_size) {
if (dest==NULL || src==NULL || dest_size == 0) return;
// find end of string in the dest buffer
while (*dest != 0 && dest_size > 1) {
dest++;
dest_size--;
}
// and catenate
_mi_strlcpy(dest, src, dest_size);
}
size_t _mi_strlen(const char* s) {
if (s==NULL) return 0;
size_t len = 0;
while(s[len] != 0) { len++; }
return len;
}
size_t _mi_strnlen(const char* s, size_t max_len) {
if (s==NULL) return 0;
size_t len = 0;
while(s[len] != 0 && len < max_len) { len++; }
return len;
}
#ifdef MI_NO_GETENV
bool _mi_getenv(const char* name, char* result, size_t result_size) {
MI_UNUSED(name);
MI_UNUSED(result);
MI_UNUSED(result_size);
return false;
}
#else
bool _mi_getenv(const char* name, char* result, size_t result_size) {
if (name==NULL || result == NULL || result_size < 64) return false;
return _mi_prim_getenv(name,result,result_size);
}
#endif
// --------------------------------------------------------
// Define our own limited `_mi_vsnprintf` and `_mi_snprintf`
// This is mostly to avoid calling these when libc is not yet
// initialized (and to reduce dependencies)
//
// format: d i, p x u, s
// prec: z l ll L
// width: 10
// align-left: -
// fill: 0
// plus: +
// --------------------------------------------------------
static void mi_outc(char c, char** out, char* end) {
char* p = *out;
if (p >= end) return;
*p = c;
*out = p + 1;
}
static void mi_outs(const char* s, char** out, char* end) {
if (s == NULL) return;
char* p = *out;
while (*s != 0 && p < end) {
*p++ = *s++;
}
*out = p;
}
static void mi_out_fill(char fill, size_t len, char** out, char* end) {
char* p = *out;
for (size_t i = 0; i < len && p < end; i++) {
*p++ = fill;
}
*out = p;
}
static void mi_out_alignright(char fill, char* start, size_t len, size_t extra, char* end) {
if (len == 0 || extra == 0) return;
if (start + len + extra >= end) return;
// move `len` characters to the right (in reverse since it can overlap)
for (size_t i = 1; i <= len; i++) {
start[len + extra - i] = start[len - i];
}
// and fill the start
for (size_t i = 0; i < extra; i++) {
start[i] = fill;
}
}
static void mi_out_num(uintmax_t x, size_t base, char prefix, char** out, char* end)
{
if (x == 0 || base == 0 || base > 16) {
if (prefix != 0) { mi_outc(prefix, out, end); }
mi_outc('0',out,end);
}
else {
// output digits in reverse
char* start = *out;
while (x > 0) {
char digit = (char)(x % base);
mi_outc((digit <= 9 ? '0' + digit : 'A' + digit - 10),out,end);
x = x / base;
}
if (prefix != 0) {
mi_outc(prefix, out, end);
}
size_t len = *out - start;
// and reverse in-place
for (size_t i = 0; i < (len / 2); i++) {
char c = start[len - i - 1];
start[len - i - 1] = start[i];
start[i] = c;
}
}
}
#define MI_NEXTC() c = *in; if (c==0) break; in++;
int _mi_vsnprintf(char* buf, size_t bufsize, const char* fmt, va_list args) {
if (buf == NULL || bufsize == 0 || fmt == NULL) return 0;
buf[bufsize - 1] = 0;
char* const end = buf + (bufsize - 1);
const char* in = fmt;
char* out = buf;
while (true) {
if (out >= end) break;
char c;
MI_NEXTC();
if (c != '%') {
if ((c >= ' ' && c <= '~') || c=='\n' || c=='\r' || c=='\t') { // output visible ascii or standard control only
mi_outc(c, &out, end);
}
}
else {
MI_NEXTC();
char fill = ' ';
size_t width = 0;
char numtype = 'd';
char numplus = 0;
bool alignright = true;
if (c == '+' || c == ' ') { numplus = c; MI_NEXTC(); }
if (c == '-') { alignright = false; MI_NEXTC(); }
if (c == '0') { fill = '0'; MI_NEXTC(); }
if (c >= '1' && c <= '9') {
width = (c - '0'); MI_NEXTC();
while (c >= '0' && c <= '9') {
width = (10 * width) + (c - '0'); MI_NEXTC();
}
if (c == 0) break; // extra check due to while
}
if (c == 'z' || c == 't' || c == 'L') { numtype = c; MI_NEXTC(); }
else if (c == 'l') {
numtype = c; MI_NEXTC();
if (c == 'l') { numtype = 'L'; MI_NEXTC(); }
}
char* start = out;
if (c == 's') {
// string
const char* s = va_arg(args, const char*);
mi_outs(s, &out, end);
}
else if (c == 'p' || c == 'x' || c == 'u') {
// unsigned
uintmax_t x = 0;
if (c == 'x' || c == 'u') {
if (numtype == 'z') x = va_arg(args, size_t);
else if (numtype == 't') x = va_arg(args, uintptr_t); // unsigned ptrdiff_t
else if (numtype == 'L') x = va_arg(args, unsigned long long);
else if (numtype == 'l') x = va_arg(args, unsigned long);
else x = va_arg(args, unsigned int);
}
else if (c == 'p') {
x = va_arg(args, uintptr_t);
mi_outs("0x", &out, end);
start = out;
width = (width >= 2 ? width - 2 : 0);
}
if (width == 0 && (c == 'x' || c == 'p')) {
if (c == 'p') { width = 2 * (x <= UINT32_MAX ? 4 : ((x >> 16) <= UINT32_MAX ? 6 : sizeof(void*))); }
if (width == 0) { width = 2; }
fill = '0';
}
mi_out_num(x, (c == 'x' || c == 'p' ? 16 : 10), numplus, &out, end);
}
else if (c == 'i' || c == 'd') {
// signed
intmax_t x = 0;
if (numtype == 'z') x = va_arg(args, intptr_t );
else if (numtype == 't') x = va_arg(args, ptrdiff_t);
else if (numtype == 'L') x = va_arg(args, long long);
else if (numtype == 'l') x = va_arg(args, long);
else x = va_arg(args, int);
char pre = 0;
if (x < 0) {
pre = '-';
if (x > INTMAX_MIN) { x = -x; }
}
else if (numplus != 0) {
pre = numplus;
}
mi_out_num((uintmax_t)x, 10, pre, &out, end);
}
else if (c >= ' ' && c <= '~') {
// unknown format
mi_outc('%', &out, end);
mi_outc(c, &out, end);
}
// fill & align
mi_assert_internal(out <= end);
mi_assert_internal(out >= start);
const size_t len = out - start;
if (len < width) {
mi_out_fill(fill, width - len, &out, end);
if (alignright && out <= end) {
mi_out_alignright(fill, start, len, width - len, end);
}
}
}
}
mi_assert_internal(out <= end);
*out = 0;
return (int)(out - buf);
}
int _mi_snprintf(char* buf, size_t buflen, const char* fmt, ...) {
va_list args;
va_start(args, fmt);
const int written = _mi_vsnprintf(buf, buflen, fmt, args);
va_end(args);
return written;
}
#if MI_SIZE_SIZE == 4
#define mi_mask_even_bits32 (0x55555555)
#define mi_mask_even_pairs32 (0x33333333)
#define mi_mask_even_nibbles32 (0x0F0F0F0F)
// sum of all the bytes in `x` if it is guaranteed that the sum < 256!
static size_t mi_byte_sum32(uint32_t x) {
// perform `x * 0x01010101`: the highest byte contains the sum of all bytes.
x += (x << 8);
x += (x << 16);
return (size_t)(x >> 24);
}
static size_t mi_popcount_generic32(uint32_t x) {
// first count each 2-bit group `a`, where: a==0b00 -> 00, a==0b01 -> 01, a==0b10 -> 01, a==0b11 -> 10
// in other words, `a - (a>>1)`; to do this in parallel, we need to mask to prevent spilling a bit pair
// into the lower bit-pair:
x = x - ((x >> 1) & mi_mask_even_bits32);
// add the 2-bit pair results
x = (x & mi_mask_even_pairs32) + ((x >> 2) & mi_mask_even_pairs32);
// add the 4-bit nibble results
x = (x + (x >> 4)) & mi_mask_even_nibbles32;
// each byte now has a count of its bits, we can sum them now:
return mi_byte_sum32(x);
}
mi_decl_noinline size_t _mi_popcount_generic(size_t x) {
return mi_popcount_generic32(x);
}
#else
#define mi_mask_even_bits64 (0x5555555555555555)
#define mi_mask_even_pairs64 (0x3333333333333333)
#define mi_mask_even_nibbles64 (0x0F0F0F0F0F0F0F0F)
// sum of all the bytes in `x` if it is guaranteed that the sum < 256!
static size_t mi_byte_sum64(uint64_t x) {
x += (x << 8);
x += (x << 16);
x += (x << 32);
return (size_t)(x >> 56);
}
static size_t mi_popcount_generic64(uint64_t x) {
x = x - ((x >> 1) & mi_mask_even_bits64);
x = (x & mi_mask_even_pairs64) + ((x >> 2) & mi_mask_even_pairs64);
x = (x + (x >> 4)) & mi_mask_even_nibbles64;
return mi_byte_sum64(x);
}
mi_decl_noinline size_t _mi_popcount_generic(size_t x) {
return mi_popcount_generic64(x);
}
#endif

319
mimalloc/mimalloc-sys/mimalloc/src/options.c
View File

@@ -9,9 +9,9 @@ terms of the MIT license. A copy of the license can be found in the file
#include "mimalloc/atomic.h"
#include "mimalloc/prim.h" // mi_prim_out_stderr
#include <stdio.h> // FILE
#include <stdio.h> // stdin/stdout
#include <stdlib.h> // abort
#include <stdarg.h>
static long mi_max_error_count = 16; // stop outputting errors after this (use < 0 for no limit)
@@ -47,6 +47,67 @@ typedef struct mi_option_desc_s {
#define MI_OPTION(opt) mi_option_##opt, #opt, NULL
#define MI_OPTION_LEGACY(opt,legacy) mi_option_##opt, #opt, #legacy
// Some options can be set at build time for statically linked libraries
// (use `-DMI_EXTRA_CPPDEFS="opt1=val1;opt2=val2"`)
//
// This is useful if we cannot pass them as environment variables
// (and setting them programmatically would be too late)
#ifndef MI_DEFAULT_VERBOSE
#define MI_DEFAULT_VERBOSE 0
#endif
#ifndef MI_DEFAULT_EAGER_COMMIT
#define MI_DEFAULT_EAGER_COMMIT 1
#endif
#ifndef MI_DEFAULT_ARENA_EAGER_COMMIT
#define MI_DEFAULT_ARENA_EAGER_COMMIT 2
#endif
// in KiB
#ifndef MI_DEFAULT_ARENA_RESERVE
#if (MI_INTPTR_SIZE>4)
#define MI_DEFAULT_ARENA_RESERVE 1024L*1024L
#else
#define MI_DEFAULT_ARENA_RESERVE 128L*1024L
#endif
#endif
#ifndef MI_DEFAULT_DISALLOW_ARENA_ALLOC
#define MI_DEFAULT_DISALLOW_ARENA_ALLOC 0
#endif
#ifndef MI_DEFAULT_ALLOW_LARGE_OS_PAGES
#define MI_DEFAULT_ALLOW_LARGE_OS_PAGES 0
#endif
#ifndef MI_DEFAULT_RESERVE_HUGE_OS_PAGES
#define MI_DEFAULT_RESERVE_HUGE_OS_PAGES 0
#endif
#ifndef MI_DEFAULT_RESERVE_OS_MEMORY
#define MI_DEFAULT_RESERVE_OS_MEMORY 0
#endif
#ifndef MI_DEFAULT_GUARDED_SAMPLE_RATE
#if MI_GUARDED
#define MI_DEFAULT_GUARDED_SAMPLE_RATE 4000
#else
#define MI_DEFAULT_GUARDED_SAMPLE_RATE 0
#endif
#endif
#ifndef MI_DEFAULT_ALLOW_THP
#if defined(__ANDROID__)
#define MI_DEFAULT_ALLOW_THP 0
#else
#define MI_DEFAULT_ALLOW_THP 1
#endif
#endif
// Static options
static mi_option_desc_t options[_mi_option_last] =
{
// stable options
@@ -56,16 +117,21 @@ static mi_option_desc_t options[_mi_option_last] =
{ 0, UNINIT, MI_OPTION(show_errors) },
#endif
{ 0, UNINIT, MI_OPTION(show_stats) },
{ 0, UNINIT, MI_OPTION(verbose) },
{ MI_DEFAULT_VERBOSE, UNINIT, MI_OPTION(verbose) },
// the following options are experimental and not all combinations make sense.
{ 1, UNINIT, MI_OPTION(eager_commit) }, // commit per segment directly (4MiB) (but see also `eager_commit_delay`)
{ 2, UNINIT, MI_OPTION_LEGACY(arena_eager_commit,eager_region_commit) }, // eager commit arena's? 2 is used to enable this only on an OS that has overcommit (i.e. linux)
// some of the following options are experimental and not all combinations are allowed.
{ MI_DEFAULT_EAGER_COMMIT,
UNINIT, MI_OPTION(eager_commit) }, // commit per segment directly (4MiB) (but see also `eager_commit_delay`)
{ MI_DEFAULT_ARENA_EAGER_COMMIT,
UNINIT, MI_OPTION_LEGACY(arena_eager_commit,eager_region_commit) }, // eager commit arena's? 2 is used to enable this only on an OS that has overcommit (i.e. linux)
{ 1, UNINIT, MI_OPTION_LEGACY(purge_decommits,reset_decommits) }, // purge decommits memory (instead of reset) (note: on linux this uses MADV_DONTNEED for decommit)
{ 0, UNINIT, MI_OPTION_LEGACY(allow_large_os_pages,large_os_pages) }, // use large OS pages, use only with eager commit to prevent fragmentation of VMA's
{ 0, UNINIT, MI_OPTION(reserve_huge_os_pages) }, // per 1GiB huge pages
{ MI_DEFAULT_ALLOW_LARGE_OS_PAGES,
UNINIT, MI_OPTION_LEGACY(allow_large_os_pages,large_os_pages) }, // use large OS pages, use only with eager commit to prevent fragmentation of VMA's
{ MI_DEFAULT_RESERVE_HUGE_OS_PAGES,
UNINIT, MI_OPTION(reserve_huge_os_pages) }, // per 1GiB huge pages
{-1, UNINIT, MI_OPTION(reserve_huge_os_pages_at) }, // reserve huge pages at node N
{ 0, UNINIT, MI_OPTION(reserve_os_memory) },
{ MI_DEFAULT_RESERVE_OS_MEMORY,
UNINIT, MI_OPTION(reserve_os_memory) }, // reserve N KiB OS memory in advance (use `option_get_size`)
{ 0, UNINIT, MI_OPTION(deprecated_segment_cache) }, // cache N segments per thread
{ 0, UNINIT, MI_OPTION(deprecated_page_reset) }, // reset page memory on free
{ 0, UNINIT, MI_OPTION_LEGACY(abandoned_page_purge,abandoned_page_reset) }, // reset free page memory when a thread terminates
@@ -77,40 +143,113 @@ static mi_option_desc_t options[_mi_option_last] =
#endif
{ 10, UNINIT, MI_OPTION_LEGACY(purge_delay,reset_delay) }, // purge delay in milli-seconds
{ 0, UNINIT, MI_OPTION(use_numa_nodes) }, // 0 = use available numa nodes, otherwise use at most N nodes.
{ 0, UNINIT, MI_OPTION(limit_os_alloc) }, // 1 = do not use OS memory for allocation (but only reserved arenas)
{ 0, UNINIT, MI_OPTION_LEGACY(disallow_os_alloc,limit_os_alloc) }, // 1 = do not use OS memory for allocation (but only reserved arenas)
{ 100, UNINIT, MI_OPTION(os_tag) }, // only apple specific for now but might serve more or less related purpose
{ 16, UNINIT, MI_OPTION(max_errors) }, // maximum errors that are output
{ 16, UNINIT, MI_OPTION(max_warnings) }, // maximum warnings that are output
{ 8, UNINIT, MI_OPTION(max_segment_reclaim)}, // max. number of segment reclaims from the abandoned segments per try.
{ 32, UNINIT, MI_OPTION(max_errors) }, // maximum errors that are output
{ 32, UNINIT, MI_OPTION(max_warnings) }, // maximum warnings that are output
{ 10, UNINIT, MI_OPTION(max_segment_reclaim)}, // max. percentage of the abandoned segments to be reclaimed per try.
{ 0, UNINIT, MI_OPTION(destroy_on_exit)}, // release all OS memory on process exit; careful with dangling pointer or after-exit frees!
#if (MI_INTPTR_SIZE>4)
{ 1024L * 1024L, UNINIT, MI_OPTION(arena_reserve) }, // reserve memory N KiB at a time
#else
{ 128L * 1024L, UNINIT, MI_OPTION(arena_reserve) },
#endif
{ 10, UNINIT, MI_OPTION(arena_purge_mult) }, // purge delay multiplier for arena's
{ MI_DEFAULT_ARENA_RESERVE, UNINIT, MI_OPTION(arena_reserve) }, // reserve memory N KiB at a time (=1GiB) (use `option_get_size`)
{ 10, UNINIT, MI_OPTION(arena_purge_mult) }, // purge delay multiplier for arena's
{ 1, UNINIT, MI_OPTION_LEGACY(purge_extend_delay, decommit_extend_delay) },
{ 0, UNINIT, MI_OPTION(abandoned_reclaim_on_free) },// reclaim an abandoned segment on a free
{ MI_DEFAULT_DISALLOW_ARENA_ALLOC, UNINIT, MI_OPTION(disallow_arena_alloc) }, // 1 = do not use arena's for allocation (except if using specific arena id's)
{ 400, UNINIT, MI_OPTION(retry_on_oom) }, // windows only: retry on out-of-memory for N milli seconds (=400), set to 0 to disable retries.
#if defined(MI_VISIT_ABANDONED)
{ 1, INITIALIZED, MI_OPTION(visit_abandoned) }, // allow visiting heap blocks in abandoned segments; requires taking locks during reclaim.
#else
{ 0, UNINIT, MI_OPTION(visit_abandoned) },
#endif
{ 0, UNINIT, MI_OPTION(guarded_min) }, // only used when building with MI_GUARDED: minimal rounded object size for guarded objects
{ MI_GiB, UNINIT, MI_OPTION(guarded_max) }, // only used when building with MI_GUARDED: maximal rounded object size for guarded objects
{ 0, UNINIT, MI_OPTION(guarded_precise) }, // disregard minimal alignment requirement to always place guarded blocks exactly in front of a guard page (=0)
{ MI_DEFAULT_GUARDED_SAMPLE_RATE,
UNINIT, MI_OPTION(guarded_sample_rate)}, // 1 out of N allocations in the min/max range will be guarded (=4000)
{ 0, UNINIT, MI_OPTION(guarded_sample_seed)},
{ 0, UNINIT, MI_OPTION(target_segments_per_thread) }, // abandon segments beyond this point, or 0 to disable.
{ 10000, UNINIT, MI_OPTION(generic_collect) }, // collect heaps every N (=10000) generic allocation calls
{ MI_DEFAULT_ALLOW_THP,
UNINIT, MI_OPTION(allow_thp) } // allow transparent huge pages?
};
static void mi_option_init(mi_option_desc_t* desc);
static bool mi_option_has_size_in_kib(mi_option_t option) {
return (option == mi_option_reserve_os_memory || option == mi_option_arena_reserve);
}
void _mi_options_init(void) {
// called on process load; should not be called before the CRT is initialized!
// (e.g. do not call this from process_init as that may run before CRT initialization)
// called on process load
mi_add_stderr_output(); // now it safe to use stderr for output
for(int i = 0; i < _mi_option_last; i++ ) {
mi_option_t option = (mi_option_t)i;
long l = mi_option_get(option); MI_UNUSED(l); // initialize
// if (option != mi_option_verbose)
{
mi_option_desc_t* desc = &options[option];
_mi_verbose_message("option '%s': %ld\n", desc->name, desc->value);
}
}
mi_max_error_count = mi_option_get(mi_option_max_errors);
mi_max_warning_count = mi_option_get(mi_option_max_warnings);
#if MI_GUARDED
if (mi_option_get(mi_option_guarded_sample_rate) > 0) {
if (mi_option_is_enabled(mi_option_allow_large_os_pages)) {
mi_option_disable(mi_option_allow_large_os_pages);
_mi_warning_message("option 'allow_large_os_pages' is disabled to allow for guarded objects\n");
}
}
#endif
if (mi_option_is_enabled(mi_option_verbose)) { mi_options_print(); }
}
#define mi_stringifyx(str) #str // and stringify
#define mi_stringify(str) mi_stringifyx(str) // expand
void mi_options_print(void) mi_attr_noexcept
{
// show version
const int vermajor = MI_MALLOC_VERSION/100;
const int verminor = (MI_MALLOC_VERSION%100)/10;
const int verpatch = (MI_MALLOC_VERSION%10);
_mi_message("v%i.%i.%i%s%s (built on %s, %s)\n", vermajor, verminor, verpatch,
#if defined(MI_CMAKE_BUILD_TYPE)
", " mi_stringify(MI_CMAKE_BUILD_TYPE)
#else
""
#endif
,
#if defined(MI_GIT_DESCRIBE)
", git " mi_stringify(MI_GIT_DESCRIBE)
#else
""
#endif
, __DATE__, __TIME__);
// show options
for (int i = 0; i < _mi_option_last; i++) {
mi_option_t option = (mi_option_t)i;
long l = mi_option_get(option); MI_UNUSED(l); // possibly initialize
mi_option_desc_t* desc = &options[option];
_mi_message("option '%s': %ld %s\n", desc->name, desc->value, (mi_option_has_size_in_kib(option) ? "KiB" : ""));
}
// show build configuration
_mi_message("debug level : %d\n", MI_DEBUG );
_mi_message("secure level: %d\n", MI_SECURE );
_mi_message("mem tracking: %s\n", MI_TRACK_TOOL);
#if MI_GUARDED
_mi_message("guarded build: %s\n", mi_option_get(mi_option_guarded_sample_rate) != 0 ? "enabled" : "disabled");
#endif
#if MI_TSAN
_mi_message("thread santizer enabled\n");
#endif
}
long _mi_option_get_fast(mi_option_t option) {
mi_assert(option >= 0 && option < _mi_option_last);
mi_option_desc_t* desc = &options[option];
mi_assert(desc->option == option); // index should match the option
//mi_assert(desc->init != UNINIT);
return desc->value;
}
mi_decl_nodiscard long mi_option_get(mi_option_t option) {
mi_assert(option >= 0 && option < _mi_option_last);
if (option < 0 || option >= _mi_option_last) return 0;
@@ -128,9 +267,12 @@ mi_decl_nodiscard long mi_option_get_clamp(mi_option_t option, long min, long ma
}
mi_decl_nodiscard size_t mi_option_get_size(mi_option_t option) {
mi_assert_internal(option == mi_option_reserve_os_memory || option == mi_option_arena_reserve);
long x = mi_option_get(option);
return (x < 0 ? 0 : (size_t)x * MI_KiB);
const long x = mi_option_get(option);
size_t size = (x < 0 ? 0 : (size_t)x);
if (mi_option_has_size_in_kib(option)) {
size *= MI_KiB;
}
return size;
}
void mi_option_set(mi_option_t option, long value) {
@@ -140,6 +282,13 @@ void mi_option_set(mi_option_t option, long value) {
mi_assert(desc->option == option); // index should match the option
desc->value = value;
desc->init = INITIALIZED;
// ensure min/max range; be careful to not recurse.
if (desc->option == mi_option_guarded_min && _mi_option_get_fast(mi_option_guarded_max) < value) {
mi_option_set(mi_option_guarded_max, value);
}
else if (desc->option == mi_option_guarded_max && _mi_option_get_fast(mi_option_guarded_min) > value) {
mi_option_set(mi_option_guarded_min, value);
}
}
void mi_option_set_default(mi_option_t option, long value) {
@@ -183,7 +332,7 @@ static void mi_cdecl mi_out_stderr(const char* msg, void* arg) {
// an output function is registered it is called immediately with
// the output up to that point.
#ifndef MI_MAX_DELAY_OUTPUT
#define MI_MAX_DELAY_OUTPUT ((size_t)(32*1024))
#define MI_MAX_DELAY_OUTPUT ((size_t)(16*1024))
#endif
static char out_buf[MI_MAX_DELAY_OUTPUT+1];
static _Atomic(size_t) out_len;
@@ -269,7 +418,7 @@ static _Atomic(size_t) warning_count; // = 0; // when >= max_warning_count stop
// (recursively) invoke malloc again to allocate space for the thread local
// variables on demand. This is why we use a _mi_preloading test on such
// platforms. However, C code generator may move the initial thread local address
// load before the `if` and we therefore split it out in a separate funcion.
// load before the `if` and we therefore split it out in a separate function.
static mi_decl_thread bool recurse = false;
static mi_decl_noinline bool mi_recurse_enter_prim(void) {
@@ -283,14 +432,14 @@ static mi_decl_noinline void mi_recurse_exit_prim(void) {
}
static bool mi_recurse_enter(void) {
#if defined(__APPLE__) || defined(MI_TLS_RECURSE_GUARD)
#if defined(__APPLE__) || defined(__ANDROID__) || defined(MI_TLS_RECURSE_GUARD)
if (_mi_preloading()) return false;
#endif
return mi_recurse_enter_prim();
}
static void mi_recurse_exit(void) {
#if defined(__APPLE__) || defined(MI_TLS_RECURSE_GUARD)
#if defined(__APPLE__) || defined(__ANDROID__) || defined(MI_TLS_RECURSE_GUARD)
if (_mi_preloading()) return;
#endif
mi_recurse_exit_prim();
@@ -311,12 +460,12 @@ void _mi_fputs(mi_output_fun* out, void* arg, const char* prefix, const char* me
}
// Define our own limited `fprintf` that avoids memory allocation.
// We do this using `snprintf` with a limited buffer.
// We do this using `_mi_vsnprintf` with a limited buffer.
static void mi_vfprintf( mi_output_fun* out, void* arg, const char* prefix, const char* fmt, va_list args ) {
char buf[512];
if (fmt==NULL) return;
if (!mi_recurse_enter()) return;
vsnprintf(buf,sizeof(buf)-1,fmt,args);
_mi_vsnprintf(buf, sizeof(buf)-1, fmt, args);
mi_recurse_exit();
_mi_fputs(out,arg,prefix,buf);
}
@@ -331,7 +480,7 @@ void _mi_fprintf( mi_output_fun* out, void* arg, const char* fmt, ... ) {
static void mi_vfprintf_thread(mi_output_fun* out, void* arg, const char* prefix, const char* fmt, va_list args) {
if (prefix != NULL && _mi_strnlen(prefix,33) <= 32 && !_mi_is_main_thread()) {
char tprefix[64];
snprintf(tprefix, sizeof(tprefix), "%sthread 0x%llx: ", prefix, (unsigned long long)_mi_thread_id());
_mi_snprintf(tprefix, sizeof(tprefix), "%sthread 0x%tx: ", prefix, (uintptr_t)_mi_thread_id());
mi_vfprintf(out, arg, tprefix, fmt, args);
}
else {
@@ -339,6 +488,13 @@ static void mi_vfprintf_thread(mi_output_fun* out, void* arg, const char* prefix
}
}
void _mi_message(const char* fmt, ...) {
va_list args;
va_start(args, fmt);
mi_vfprintf_thread(NULL, NULL, "mimalloc: ", fmt, args);
va_end(args);
}
void _mi_trace_message(const char* fmt, ...) {
if (mi_option_get(mi_option_verbose) <= 1) return; // only with verbose level 2 or higher
va_list args;
@@ -376,7 +532,7 @@ void _mi_warning_message(const char* fmt, ...) {
#if MI_DEBUG
void _mi_assert_fail(const char* assertion, const char* fname, unsigned line, const char* func ) {
mi_decl_noreturn mi_decl_cold void _mi_assert_fail(const char* assertion, const char* fname, unsigned line, const char* func ) mi_attr_noexcept {
_mi_fprintf(NULL, NULL, "mimalloc: assertion failed: at \"%s\":%u, %s\n assertion: \"%s\"\n", fname, line, (func==NULL?"":func), assertion);
abort();
}
@@ -434,68 +590,6 @@ void _mi_error_message(int err, const char* fmt, ...) {
// --------------------------------------------------------
// Initialize options by checking the environment
// --------------------------------------------------------
char _mi_toupper(char c) {
if (c >= 'a' && c <= 'z') return (c - 'a' + 'A');
else return c;
}
int _mi_strnicmp(const char* s, const char* t, size_t n) {
if (n == 0) return 0;
for (; *s != 0 && *t != 0 && n > 0; s++, t++, n--) {
if (_mi_toupper(*s) != _mi_toupper(*t)) break;
}
return (n == 0 ? 0 : *s - *t);
}
void _mi_strlcpy(char* dest, const char* src, size_t dest_size) {
if (dest==NULL || src==NULL || dest_size == 0) return;
// copy until end of src, or when dest is (almost) full
while (*src != 0 && dest_size > 1) {
*dest++ = *src++;
dest_size--;
}
// always zero terminate
*dest = 0;
}
void _mi_strlcat(char* dest, const char* src, size_t dest_size) {
if (dest==NULL || src==NULL || dest_size == 0) return;
// find end of string in the dest buffer
while (*dest != 0 && dest_size > 1) {
dest++;
dest_size--;
}
// and catenate
_mi_strlcpy(dest, src, dest_size);
}
size_t _mi_strlen(const char* s) {
if (s==NULL) return 0;
size_t len = 0;
while(s[len] != 0) { len++; }
return len;
}
size_t _mi_strnlen(const char* s, size_t max_len) {
if (s==NULL) return 0;
size_t len = 0;
while(s[len] != 0 && len < max_len) { len++; }
return len;
}
#ifdef MI_NO_GETENV
static bool mi_getenv(const char* name, char* result, size_t result_size) {
MI_UNUSED(name);
MI_UNUSED(result);
MI_UNUSED(result_size);
return false;
}
#else
static bool mi_getenv(const char* name, char* result, size_t result_size) {
if (name==NULL || result == NULL || result_size < 64) return false;
return _mi_prim_getenv(name,result,result_size);
}
#endif
// TODO: implement ourselves to reduce dependencies on the C runtime
#include <stdlib.h> // strtol
@@ -508,11 +602,11 @@ static void mi_option_init(mi_option_desc_t* desc) {
char buf[64+1];
_mi_strlcpy(buf, "mimalloc_", sizeof(buf));
_mi_strlcat(buf, desc->name, sizeof(buf));
bool found = mi_getenv(buf, s, sizeof(s));
bool found = _mi_getenv(buf, s, sizeof(s));
if (!found && desc->legacy_name != NULL) {
_mi_strlcpy(buf, "mimalloc_", sizeof(buf));
_mi_strlcat(buf, desc->legacy_name, sizeof(buf));
found = mi_getenv(buf, s, sizeof(s));
found = _mi_getenv(buf, s, sizeof(s));
if (found) {
_mi_warning_message("environment option \"mimalloc_%s\" is deprecated -- use \"mimalloc_%s\" instead.\n", desc->legacy_name, desc->name);
}
@@ -535,18 +629,23 @@ static void mi_option_init(mi_option_desc_t* desc) {
else {
char* end = buf;
long value = strtol(buf, &end, 10);
if (desc->option == mi_option_reserve_os_memory || desc->option == mi_option_arena_reserve) {
// this option is interpreted in KiB to prevent overflow of `long`
if (mi_option_has_size_in_kib(desc->option)) {
// this option is interpreted in KiB to prevent overflow of `long` for large allocations
// (long is 32-bit on 64-bit windows, which allows for 4TiB max.)
size_t size = (value < 0 ? 0 : (size_t)value);
bool overflow = false;
if (*end == 'K') { end++; }
else if (*end == 'M') { value *= MI_KiB; end++; }
else if (*end == 'G') { value *= MI_MiB; end++; }
else { value = (value + MI_KiB - 1) / MI_KiB; }
if (end[0] == 'I' && end[1] == 'B') { end += 2; }
else if (*end == 'B') { end++; }
else if (*end == 'M') { overflow = mi_mul_overflow(size,MI_KiB,&size); end++; }
else if (*end == 'G') { overflow = mi_mul_overflow(size,MI_MiB,&size); end++; }
else if (*end == 'T') { overflow = mi_mul_overflow(size,MI_GiB,&size); end++; }
else { size = (size + MI_KiB - 1) / MI_KiB; }
if (end[0] == 'I' && end[1] == 'B') { end += 2; } // KiB, MiB, GiB, TiB
else if (*end == 'B') { end++; } // Kb, Mb, Gb, Tb
if (overflow || size > MI_MAX_ALLOC_SIZE) { size = (MI_MAX_ALLOC_SIZE / MI_KiB); }
value = (size > LONG_MAX ? LONG_MAX : (long)size);
}
if (*end == 0) {
desc->value = value;
desc->init = INITIALIZED;
mi_option_set(desc->option, value);
}
else {
// set `init` first to avoid recursion through _mi_warning_message on mimalloc_verbose.

363
mimalloc/mimalloc-sys/mimalloc/src/os.c
View File

@@ -1,5 +1,5 @@
/* ----------------------------------------------------------------------------
Copyright (c) 2018-2023, Microsoft Research, Daan Leijen
Copyright (c) 2018-2025, Microsoft Research, Daan Leijen
This is free software; you can redistribute it and/or modify it under the
terms of the MIT license. A copy of the license can be found in the file
"LICENSE" at the root of this distribution.
@@ -9,27 +9,45 @@ terms of the MIT license. A copy of the license can be found in the file
#include "mimalloc/atomic.h"
#include "mimalloc/prim.h"
#define mi_os_stat_increase(stat,amount) _mi_stat_increase(&_mi_stats_main.stat, amount)
#define mi_os_stat_decrease(stat,amount) _mi_stat_decrease(&_mi_stats_main.stat, amount)
#define mi_os_stat_counter_increase(stat,inc) _mi_stat_counter_increase(&_mi_stats_main.stat, inc)
/* -----------------------------------------------------------
Initialization.
On windows initializes support for aligned allocation and
large OS pages (if MIMALLOC_LARGE_OS_PAGES is true).
----------------------------------------------------------- */
#ifndef MI_DEFAULT_VIRTUAL_ADDRESS_BITS
#if MI_INTPTR_SIZE < 8
#define MI_DEFAULT_VIRTUAL_ADDRESS_BITS 32
#else
#define MI_DEFAULT_VIRTUAL_ADDRESS_BITS 48
#endif
#endif
#ifndef MI_DEFAULT_PHYSICAL_MEMORY_IN_KIB
#if MI_INTPTR_SIZE < 8
#define MI_DEFAULT_PHYSICAL_MEMORY_IN_KIB 4*MI_MiB // 4 GiB
#else
#define MI_DEFAULT_PHYSICAL_MEMORY_IN_KIB 32*MI_MiB // 32 GiB
#endif
#endif
static mi_os_mem_config_t mi_os_mem_config = {
4096, // page size
0, // large page size (usually 2MiB)
4096, // allocation granularity
true, // has overcommit? (if true we use MAP_NORESERVE on mmap systems)
false, // must free whole? (on mmap systems we can free anywhere in a mapped range, but on Windows we must free the entire span)
true // has virtual reserve? (if true we can reserve virtual address space without using commit or physical memory)
4096, // page size
0, // large page size (usually 2MiB)
4096, // allocation granularity
MI_DEFAULT_PHYSICAL_MEMORY_IN_KIB,
MI_DEFAULT_VIRTUAL_ADDRESS_BITS,
true, // has overcommit? (if true we use MAP_NORESERVE on mmap systems)
false, // can we partially free allocated blocks? (on mmap systems we can free anywhere in a mapped range, but on Windows we must free the entire span)
true // has virtual reserve? (if true we can reserve virtual address space without using commit or physical memory)
};
bool _mi_os_has_overcommit(void) {
return mi_os_mem_config.has_overcommit;
}
bool _mi_os_has_virtual_reserve(void) {
bool _mi_os_has_virtual_reserve(void) {
return mi_os_mem_config.has_virtual_reserve;
}
@@ -44,9 +62,9 @@ size_t _mi_os_large_page_size(void) {
return (mi_os_mem_config.large_page_size != 0 ? mi_os_mem_config.large_page_size : _mi_os_page_size());
}
bool _mi_os_use_large_page(size_t size, size_t alignment) {
bool _mi_os_canuse_large_page(size_t size, size_t alignment) {
// if we have access, check the size and alignment requirements
if (mi_os_mem_config.large_page_size == 0 || !mi_option_is_enabled(mi_option_allow_large_os_pages)) return false;
if (mi_os_mem_config.large_page_size == 0) return false;
return ((size % mi_os_mem_config.large_page_size) == 0 && (alignment % mi_os_mem_config.large_page_size) == 0);
}
@@ -70,25 +88,18 @@ void _mi_os_init(void) {
/* -----------------------------------------------------------
Util
-------------------------------------------------------------- */
bool _mi_os_decommit(void* addr, size_t size, mi_stats_t* stats);
bool _mi_os_commit(void* addr, size_t size, bool* is_zero, mi_stats_t* tld_stats);
static void* mi_align_up_ptr(void* p, size_t alignment) {
return (void*)_mi_align_up((uintptr_t)p, alignment);
}
static void* mi_align_down_ptr(void* p, size_t alignment) {
return (void*)_mi_align_down((uintptr_t)p, alignment);
}
bool _mi_os_decommit(void* addr, size_t size);
bool _mi_os_commit(void* addr, size_t size, bool* is_zero);
/* -----------------------------------------------------------
aligned hinting
-------------------------------------------------------------- */
// On 64-bit systems, we can do efficient aligned allocation by using
// the 2TiB to 30TiB area to allocate those.
#if (MI_INTPTR_SIZE >= 8)
// On systems with enough virtual address bits, we can do efficient aligned allocation by using
// the 2TiB to 30TiB area to allocate those. If we have at least 46 bits of virtual address
// space (64TiB) we use this technique. (but see issue #939)
#if (MI_INTPTR_SIZE >= 8) && !defined(MI_NO_ALIGNED_HINT)
static mi_decl_cache_align _Atomic(uintptr_t)aligned_base;
// Return a MI_SEGMENT_SIZE aligned address that is probably available.
@@ -105,6 +116,7 @@ static mi_decl_cache_align _Atomic(uintptr_t)aligned_base;
void* _mi_os_get_aligned_hint(size_t try_alignment, size_t size)
{
if (try_alignment <= 1 || try_alignment > MI_SEGMENT_SIZE) return NULL;
if (mi_os_mem_config.virtual_address_bits < 46) return NULL; // < 64TiB virtual address space
size = _mi_align_up(size, MI_SEGMENT_SIZE);
if (size > 1*MI_GiB) return NULL; // guarantee the chance of fixed valid address is at most 1/(MI_HINT_AREA / 1<<30) = 1/4096.
#if (MI_SECURE>0)
@@ -132,54 +144,61 @@ void* _mi_os_get_aligned_hint(size_t try_alignment, size_t size) {
}
#endif
/* -----------------------------------------------------------
Free memory
-------------------------------------------------------------- */
static void mi_os_free_huge_os_pages(void* p, size_t size, mi_stats_t* stats);
static void mi_os_free_huge_os_pages(void* p, size_t size);
static void mi_os_prim_free(void* addr, size_t size, bool still_committed, mi_stats_t* tld_stats) {
MI_UNUSED(tld_stats);
static void mi_os_prim_free(void* addr, size_t size, size_t commit_size) {
mi_assert_internal((size % _mi_os_page_size()) == 0);
if (addr == NULL || size == 0) return; // || _mi_os_is_huge_reserved(addr)
int err = _mi_prim_free(addr, size);
if (addr == NULL) return; // || _mi_os_is_huge_reserved(addr)
int err = _mi_prim_free(addr, size); // allow size==0 (issue #1041)
if (err != 0) {
_mi_warning_message("unable to free OS memory (error: %d (0x%x), size: 0x%zx bytes, address: %p)\n", err, err, size, addr);
}
mi_stats_t* stats = &_mi_stats_main;
if (still_committed) { _mi_stat_decrease(&stats->committed, size); }
_mi_stat_decrease(&stats->reserved, size);
if (commit_size > 0) {
mi_os_stat_decrease(committed, commit_size);
}
mi_os_stat_decrease(reserved, size);
}
void _mi_os_free_ex(void* addr, size_t size, bool still_committed, mi_memid_t memid, mi_stats_t* tld_stats) {
void _mi_os_free_ex(void* addr, size_t size, bool still_committed, mi_memid_t memid) {
if (mi_memkind_is_os(memid.memkind)) {
size_t csize = _mi_os_good_alloc_size(size);
size_t csize = memid.mem.os.size;
if (csize==0) { csize = _mi_os_good_alloc_size(size); }
mi_assert_internal(csize >= size);
size_t commit_size = (still_committed ? csize : 0);
void* base = addr;
// different base? (due to alignment)
if (memid.mem.os.base != NULL) {
if (memid.mem.os.base != base) {
mi_assert(memid.mem.os.base <= addr);
mi_assert((uint8_t*)memid.mem.os.base + memid.mem.os.alignment >= (uint8_t*)addr);
base = memid.mem.os.base;
csize += ((uint8_t*)addr - (uint8_t*)memid.mem.os.base);
const size_t diff = (uint8_t*)addr - (uint8_t*)memid.mem.os.base;
if (memid.mem.os.size==0) {
csize += diff;
}
if (still_committed) {
commit_size -= diff; // the (addr-base) part was already un-committed
}
}
// free it
if (memid.memkind == MI_MEM_OS_HUGE) {
mi_assert(memid.is_pinned);
mi_os_free_huge_os_pages(base, csize, tld_stats);
mi_os_free_huge_os_pages(base, csize);
}
else {
mi_os_prim_free(base, csize, still_committed, tld_stats);
mi_os_prim_free(base, csize, (still_committed ? commit_size : 0));
}
}
else {
// nothing to do
// nothing to do
mi_assert(memid.memkind < MI_MEM_OS);
}
}
void _mi_os_free(void* p, size_t size, mi_memid_t memid, mi_stats_t* tld_stats) {
_mi_os_free_ex(p, size, true, memid, tld_stats);
void _mi_os_free(void* p, size_t size, mi_memid_t memid) {
_mi_os_free_ex(p, size, true, memid);
}
@@ -188,39 +207,46 @@ void _mi_os_free(void* p, size_t size, mi_memid_t memid, mi_stats_t* tld_stats)
-------------------------------------------------------------- */
// Note: the `try_alignment` is just a hint and the returned pointer is not guaranteed to be aligned.
static void* mi_os_prim_alloc(size_t size, size_t try_alignment, bool commit, bool allow_large, bool* is_large, bool* is_zero, mi_stats_t* stats) {
// Also `hint_addr` is a hint and may be ignored.
static void* mi_os_prim_alloc_at(void* hint_addr, size_t size, size_t try_alignment, bool commit, bool allow_large, bool* is_large, bool* is_zero) {
mi_assert_internal(size > 0 && (size % _mi_os_page_size()) == 0);
mi_assert_internal(is_zero != NULL);
mi_assert_internal(is_large != NULL);
if (size == 0) return NULL;
if (!commit) { allow_large = false; }
if (try_alignment == 0) { try_alignment = 1; } // avoid 0 to ensure there will be no divide by zero when aligning
*is_zero = false;
void* p = NULL;
int err = _mi_prim_alloc(size, try_alignment, commit, allow_large, is_large, is_zero, &p);
void* p = NULL;
int err = _mi_prim_alloc(hint_addr, size, try_alignment, commit, allow_large, is_large, is_zero, &p);
if (err != 0) {
_mi_warning_message("unable to allocate OS memory (error: %d (0x%x), size: 0x%zx bytes, align: 0x%zx, commit: %d, allow large: %d)\n", err, err, size, try_alignment, commit, allow_large);
_mi_warning_message("unable to allocate OS memory (error: %d (0x%x), addr: %p, size: 0x%zx bytes, align: 0x%zx, commit: %d, allow large: %d)\n", err, err, hint_addr, size, try_alignment, commit, allow_large);
}
mi_stat_counter_increase(stats->mmap_calls, 1);
mi_os_stat_counter_increase(mmap_calls, 1);
if (p != NULL) {
_mi_stat_increase(&stats->reserved, size);
if (commit) {
_mi_stat_increase(&stats->committed, size);
mi_os_stat_increase(reserved, size);
if (commit) {
mi_os_stat_increase(committed, size);
// seems needed for asan (or `mimalloc-test-api` fails)
#ifdef MI_TRACK_ASAN
if (*is_zero) { mi_track_mem_defined(p,size); }
else { mi_track_mem_undefined(p,size); }
#endif
}
}
}
return p;
}
static void* mi_os_prim_alloc(size_t size, size_t try_alignment, bool commit, bool allow_large, bool* is_large, bool* is_zero) {
return mi_os_prim_alloc_at(NULL, size, try_alignment, commit, allow_large, is_large, is_zero);
}
// Primitive aligned allocation from the OS.
// This function guarantees the allocated memory is aligned.
static void* mi_os_prim_alloc_aligned(size_t size, size_t alignment, bool commit, bool allow_large, bool* is_large, bool* is_zero, void** base, mi_stats_t* stats) {
static void* mi_os_prim_alloc_aligned(size_t size, size_t alignment, bool commit, bool allow_large, bool* is_large, bool* is_zero, void** base) {
mi_assert_internal(alignment >= _mi_os_page_size() && ((alignment & (alignment - 1)) == 0));
mi_assert_internal(size > 0 && (size % _mi_os_page_size()) == 0);
mi_assert_internal(is_large != NULL);
@@ -230,8 +256,8 @@ static void* mi_os_prim_alloc_aligned(size_t size, size_t alignment, bool commit
if (!(alignment >= _mi_os_page_size() && ((alignment & (alignment - 1)) == 0))) return NULL;
size = _mi_align_up(size, _mi_os_page_size());
// try first with a hint (this will be aligned directly on Win 10+ or BSD)
void* p = mi_os_prim_alloc(size, alignment, commit, allow_large, is_large, is_zero, stats);
// try first with a requested alignment hint (this will usually be aligned directly on Win 10+ or BSD)
void* p = mi_os_prim_alloc(size, alignment, commit, allow_large, is_large, is_zero);
if (p == NULL) return NULL;
// aligned already?
@@ -240,16 +266,18 @@ static void* mi_os_prim_alloc_aligned(size_t size, size_t alignment, bool commit
}
else {
// if not aligned, free it, overallocate, and unmap around it
#if !MI_TRACK_ASAN
_mi_warning_message("unable to allocate aligned OS memory directly, fall back to over-allocation (size: 0x%zx bytes, address: %p, alignment: 0x%zx, commit: %d)\n", size, p, alignment, commit);
mi_os_prim_free(p, size, commit, stats);
#endif
if (p != NULL) { mi_os_prim_free(p, size, (commit ? size : 0)); }
if (size >= (SIZE_MAX - alignment)) return NULL; // overflow
const size_t over_size = size + alignment;
if (mi_os_mem_config.must_free_whole) { // win32 virtualAlloc cannot free parts of an allocate block
if (!mi_os_mem_config.has_partial_free) { // win32 virtualAlloc cannot free parts of an allocated block
// over-allocate uncommitted (virtual) memory
p = mi_os_prim_alloc(over_size, 1 /*alignment*/, false /* commit? */, false /* allow_large */, is_large, is_zero, stats);
p = mi_os_prim_alloc(over_size, 1 /*alignment*/, false /* commit? */, false /* allow_large */, is_large, is_zero);
if (p == NULL) return NULL;
// set p to the aligned part in the full region
// note: this is dangerous on Windows as VirtualFree needs the actual base pointer
// this is handled though by having the `base` field in the memid's
@@ -258,25 +286,28 @@ static void* mi_os_prim_alloc_aligned(size_t size, size_t alignment, bool commit
// explicitly commit only the aligned part
if (commit) {
_mi_os_commit(p, size, NULL, stats);
if (!_mi_os_commit(p, size, NULL)) {
mi_os_prim_free(*base, over_size, 0);
return NULL;
}
}
}
else { // mmap can free inside an allocation
// overallocate...
p = mi_os_prim_alloc(over_size, 1, commit, false, is_large, is_zero, stats);
p = mi_os_prim_alloc(over_size, 1, commit, false, is_large, is_zero);
if (p == NULL) return NULL;
// and selectively unmap parts around the over-allocated area. (noop on sbrk)
// and selectively unmap parts around the over-allocated area.
void* aligned_p = mi_align_up_ptr(p, alignment);
size_t pre_size = (uint8_t*)aligned_p - (uint8_t*)p;
size_t mid_size = _mi_align_up(size, _mi_os_page_size());
size_t post_size = over_size - pre_size - mid_size;
mi_assert_internal(pre_size < over_size&& post_size < over_size&& mid_size >= size);
if (pre_size > 0) { mi_os_prim_free(p, pre_size, commit, stats); }
if (post_size > 0) { mi_os_prim_free((uint8_t*)aligned_p + mid_size, post_size, commit, stats); }
// we can return the aligned pointer on `mmap` (and sbrk) systems
if (pre_size > 0) { mi_os_prim_free(p, pre_size, (commit ? pre_size : 0)); }
if (post_size > 0) { mi_os_prim_free((uint8_t*)aligned_p + mid_size, post_size, (commit ? post_size : 0)); }
// we can return the aligned pointer on `mmap` systems
p = aligned_p;
*base = aligned_p; // since we freed the pre part, `*base == p`.
*base = aligned_p; // since we freed the pre part, `*base == p`.
}
}
@@ -289,51 +320,78 @@ static void* mi_os_prim_alloc_aligned(size_t size, size_t alignment, bool commit
OS API: alloc and alloc_aligned
----------------------------------------------------------- */
void* _mi_os_alloc(size_t size, mi_memid_t* memid, mi_stats_t* tld_stats) {
MI_UNUSED(tld_stats);
void* _mi_os_alloc(size_t size, mi_memid_t* memid) {
*memid = _mi_memid_none();
mi_stats_t* stats = &_mi_stats_main;
if (size == 0) return NULL;
size = _mi_os_good_alloc_size(size);
bool os_is_large = false;
bool os_is_zero = false;
void* p = mi_os_prim_alloc(size, 0, true, false, &os_is_large, &os_is_zero, stats);
if (p != NULL) {
*memid = _mi_memid_create_os(true, os_is_zero, os_is_large);
}
void* p = mi_os_prim_alloc(size, 0, true, false, &os_is_large, &os_is_zero);
if (p == NULL) return NULL;
*memid = _mi_memid_create_os(p, size, true, os_is_zero, os_is_large);
mi_assert_internal(memid->mem.os.size >= size);
mi_assert_internal(memid->initially_committed);
return p;
}
void* _mi_os_alloc_aligned(size_t size, size_t alignment, bool commit, bool allow_large, mi_memid_t* memid, mi_stats_t* tld_stats)
void* _mi_os_alloc_aligned(size_t size, size_t alignment, bool commit, bool allow_large, mi_memid_t* memid)
{
MI_UNUSED(&_mi_os_get_aligned_hint); // suppress unused warnings
MI_UNUSED(tld_stats);
*memid = _mi_memid_none();
if (size == 0) return NULL;
size = _mi_os_good_alloc_size(size);
alignment = _mi_align_up(alignment, _mi_os_page_size());
bool os_is_large = false;
bool os_is_zero = false;
void* os_base = NULL;
void* p = mi_os_prim_alloc_aligned(size, alignment, commit, allow_large, &os_is_large, &os_is_zero, &os_base, &_mi_stats_main /*tld->stats*/ );
if (p != NULL) {
*memid = _mi_memid_create_os(commit, os_is_zero, os_is_large);
memid->mem.os.base = os_base;
memid->mem.os.alignment = alignment;
}
void* p = mi_os_prim_alloc_aligned(size, alignment, commit, allow_large, &os_is_large, &os_is_zero, &os_base );
if (p == NULL) return NULL;
*memid = _mi_memid_create_os(p, size, commit, os_is_zero, os_is_large);
memid->mem.os.base = os_base;
memid->mem.os.size += ((uint8_t*)p - (uint8_t*)os_base); // todo: return from prim_alloc_aligned?
mi_assert_internal(memid->mem.os.size >= size);
mi_assert_internal(_mi_is_aligned(p,alignment));
if (commit) { mi_assert_internal(memid->initially_committed); }
return p;
}
mi_decl_nodiscard static void* mi_os_ensure_zero(void* p, size_t size, mi_memid_t* memid) {
if (p==NULL || size==0) return p;
// ensure committed
if (!memid->initially_committed) {
bool is_zero = false;
if (!_mi_os_commit(p, size, &is_zero)) {
_mi_os_free(p, size, *memid);
return NULL;
}
memid->initially_committed = true;
}
// ensure zero'd
if (memid->initially_zero) return p;
_mi_memzero_aligned(p,size);
memid->initially_zero = true;
return p;
}
void* _mi_os_zalloc(size_t size, mi_memid_t* memid) {
void* p = _mi_os_alloc(size,memid);
return mi_os_ensure_zero(p, size, memid);
}
/* -----------------------------------------------------------
OS aligned allocation with an offset. This is used
for large alignments > MI_ALIGNMENT_MAX. We use a large mimalloc
for large alignments > MI_BLOCK_ALIGNMENT_MAX. We use a large mimalloc
page where the object can be aligned at an offset from the start of the segment.
As we may need to overallocate, we need to free such pointers using `mi_free_aligned`
to use the actual start of the memory region.
----------------------------------------------------------- */
void* _mi_os_alloc_aligned_at_offset(size_t size, size_t alignment, size_t offset, bool commit, bool allow_large, mi_memid_t* memid, mi_stats_t* tld_stats) {
void* _mi_os_alloc_aligned_at_offset(size_t size, size_t alignment, size_t offset, bool commit, bool allow_large, mi_memid_t* memid) {
mi_assert(offset <= MI_SEGMENT_SIZE);
mi_assert(offset <= size);
mi_assert((alignment % _mi_os_page_size()) == 0);
@@ -341,20 +399,20 @@ void* _mi_os_alloc_aligned_at_offset(size_t size, size_t alignment, size_t offse
if (offset > MI_SEGMENT_SIZE) return NULL;
if (offset == 0) {
// regular aligned allocation
return _mi_os_alloc_aligned(size, alignment, commit, allow_large, memid, tld_stats);
return _mi_os_alloc_aligned(size, alignment, commit, allow_large, memid);
}
else {
// overallocate to align at an offset
const size_t extra = _mi_align_up(offset, alignment) - offset;
const size_t oversize = size + extra;
void* const start = _mi_os_alloc_aligned(oversize, alignment, commit, allow_large, memid, tld_stats);
void* const start = _mi_os_alloc_aligned(oversize, alignment, commit, allow_large, memid);
if (start == NULL) return NULL;
void* const p = (uint8_t*)start + extra;
mi_assert(_mi_is_aligned((uint8_t*)p + offset, alignment));
// decommit the overallocation at the start
if (commit && extra > _mi_os_page_size()) {
_mi_os_decommit(start, extra, tld_stats);
_mi_os_decommit(start, extra);
}
return p;
}
@@ -388,51 +446,51 @@ static void* mi_os_page_align_area_conservative(void* addr, size_t size, size_t*
return mi_os_page_align_areax(true, addr, size, newsize);
}
bool _mi_os_commit(void* addr, size_t size, bool* is_zero, mi_stats_t* tld_stats) {
MI_UNUSED(tld_stats);
mi_stats_t* stats = &_mi_stats_main;
bool _mi_os_commit_ex(void* addr, size_t size, bool* is_zero, size_t stat_size) {
if (is_zero != NULL) { *is_zero = false; }
_mi_stat_increase(&stats->committed, size); // use size for precise commit vs. decommit
_mi_stat_counter_increase(&stats->commit_calls, 1);
mi_os_stat_counter_increase(commit_calls, 1);
// page align range
size_t csize;
void* start = mi_os_page_align_areax(false /* conservative? */, addr, size, &csize);
if (csize == 0) return true;
// commit
// commit
bool os_is_zero = false;
int err = _mi_prim_commit(start, csize, &os_is_zero);
int err = _mi_prim_commit(start, csize, &os_is_zero);
if (err != 0) {
_mi_warning_message("cannot commit OS memory (error: %d (0x%x), address: %p, size: 0x%zx bytes)\n", err, err, start, csize);
return false;
}
if (os_is_zero && is_zero != NULL) {
if (os_is_zero && is_zero != NULL) {
*is_zero = true;
mi_assert_expensive(mi_mem_is_zero(start, csize));
}
// note: the following seems required for asan (otherwise `mimalloc-test-stress` fails)
#ifdef MI_TRACK_ASAN
if (os_is_zero) { mi_track_mem_defined(start,csize); }
else { mi_track_mem_undefined(start,csize); }
else { mi_track_mem_undefined(start,csize); }
#endif
mi_os_stat_increase(committed, stat_size); // use size for precise commit vs. decommit
return true;
}
static bool mi_os_decommit_ex(void* addr, size_t size, bool* needs_recommit, mi_stats_t* tld_stats) {
MI_UNUSED(tld_stats);
mi_stats_t* stats = &_mi_stats_main;
bool _mi_os_commit(void* addr, size_t size, bool* is_zero) {
return _mi_os_commit_ex(addr, size, is_zero, size);
}
static bool mi_os_decommit_ex(void* addr, size_t size, bool* needs_recommit, size_t stat_size) {
mi_assert_internal(needs_recommit!=NULL);
_mi_stat_decrease(&stats->committed, size);
mi_os_stat_decrease(committed, stat_size);
// page align
size_t csize;
void* start = mi_os_page_align_area_conservative(addr, size, &csize);
if (csize == 0) return true;
if (csize == 0) return true;
// decommit
*needs_recommit = true;
int err = _mi_prim_decommit(start,csize,needs_recommit);
int err = _mi_prim_decommit(start,csize,needs_recommit);
if (err != 0) {
_mi_warning_message("cannot decommit OS memory (error: %d (0x%x), address: %p, size: 0x%zx bytes)\n", err, err, start, csize);
}
@@ -440,9 +498,9 @@ static bool mi_os_decommit_ex(void* addr, size_t size, bool* needs_recommit, mi_
return (err == 0);
}
bool _mi_os_decommit(void* addr, size_t size, mi_stats_t* tld_stats) {
bool _mi_os_decommit(void* addr, size_t size) {
bool needs_recommit;
return mi_os_decommit_ex(addr, size, &needs_recommit, tld_stats);
return mi_os_decommit_ex(addr, size, &needs_recommit, size);
}
@@ -450,13 +508,13 @@ bool _mi_os_decommit(void* addr, size_t size, mi_stats_t* tld_stats) {
// but may be used later again. This will release physical memory
// pages and reduce swapping while keeping the memory committed.
// We page align to a conservative area inside the range to reset.
bool _mi_os_reset(void* addr, size_t size, mi_stats_t* stats) {
bool _mi_os_reset(void* addr, size_t size) {
// page align conservatively within the range
size_t csize;
void* start = mi_os_page_align_area_conservative(addr, size, &csize);
if (csize == 0) return true; // || _mi_os_is_huge_reserved(addr)
_mi_stat_increase(&stats->reset, csize);
_mi_stat_counter_increase(&stats->reset_calls, 1);
mi_os_stat_counter_increase(reset, csize);
mi_os_stat_counter_increase(reset_calls, 1);
#if (MI_DEBUG>1) && !MI_SECURE && !MI_TRACK_ENABLED // && !MI_TSAN
memset(start, 0, csize); // pretend it is eagerly reset
@@ -470,33 +528,44 @@ bool _mi_os_reset(void* addr, size_t size, mi_stats_t* stats) {
}
// either resets or decommits memory, returns true if the memory needs
void _mi_os_reuse( void* addr, size_t size ) {
// page align conservatively within the range
size_t csize = 0;
void* const start = mi_os_page_align_area_conservative(addr, size, &csize);
if (csize == 0) return;
const int err = _mi_prim_reuse(start, csize);
if (err != 0) {
_mi_warning_message("cannot reuse OS memory (error: %d (0x%x), address: %p, size: 0x%zx bytes)\n", err, err, start, csize);
}
}
// either resets or decommits memory, returns true if the memory needs
// to be recommitted if it is to be re-used later on.
bool _mi_os_purge_ex(void* p, size_t size, bool allow_reset, mi_stats_t* stats)
bool _mi_os_purge_ex(void* p, size_t size, bool allow_reset, size_t stat_size)
{
if (mi_option_get(mi_option_purge_delay) < 0) return false; // is purging allowed?
_mi_stat_counter_increase(&stats->purge_calls, 1);
_mi_stat_increase(&stats->purged, size);
mi_os_stat_counter_increase(purge_calls, 1);
mi_os_stat_counter_increase(purged, size);
if (mi_option_is_enabled(mi_option_purge_decommits) && // should decommit?
!_mi_preloading()) // don't decommit during preloading (unsafe)
{
bool needs_recommit = true;
mi_os_decommit_ex(p, size, &needs_recommit, stats);
return needs_recommit;
mi_os_decommit_ex(p, size, &needs_recommit, stat_size);
return needs_recommit;
}
else {
if (allow_reset) { // this can sometimes be not allowed if the range is not fully committed
_mi_os_reset(p, size, stats);
_mi_os_reset(p, size);
}
return false; // needs no recommit
}
}
// either resets or decommits memory, returns true if the memory needs
// either resets or decommits memory, returns true if the memory needs
// to be recommitted if it is to be re-used later on.
bool _mi_os_purge(void* p, size_t size, mi_stats_t * stats) {
return _mi_os_purge_ex(p, size, true, stats);
bool _mi_os_purge(void* p, size_t size) {
return _mi_os_purge_ex(p, size, true, size);
}
// Protect a region in memory to be not accessible.
@@ -577,7 +646,7 @@ void* _mi_os_alloc_huge_os_pages(size_t pages, int numa_node, mi_msecs_t max_mse
if (psize != NULL) *psize = 0;
if (pages_reserved != NULL) *pages_reserved = 0;
size_t size = 0;
uint8_t* start = mi_os_claim_huge_pages(pages, &size);
uint8_t* const start = mi_os_claim_huge_pages(pages, &size);
if (start == NULL) return NULL; // or 32-bit systems
// Allocate one page at the time but try to place them contiguously
@@ -603,15 +672,15 @@ void* _mi_os_alloc_huge_os_pages(size_t pages, int numa_node, mi_msecs_t max_mse
// no success, issue a warning and break
if (p != NULL) {
_mi_warning_message("could not allocate contiguous huge OS page %zu at %p\n", page, addr);
mi_os_prim_free(p, MI_HUGE_OS_PAGE_SIZE, true, &_mi_stats_main);
mi_os_prim_free(p, MI_HUGE_OS_PAGE_SIZE, MI_HUGE_OS_PAGE_SIZE);
}
break;
}
// success, record it
page++; // increase before timeout check (see issue #711)
_mi_stat_increase(&_mi_stats_main.committed, MI_HUGE_OS_PAGE_SIZE);
_mi_stat_increase(&_mi_stats_main.reserved, MI_HUGE_OS_PAGE_SIZE);
mi_os_stat_increase(committed, MI_HUGE_OS_PAGE_SIZE);
mi_os_stat_increase(reserved, MI_HUGE_OS_PAGE_SIZE);
// check for timeout
if (max_msecs > 0) {
@@ -633,7 +702,7 @@ void* _mi_os_alloc_huge_os_pages(size_t pages, int numa_node, mi_msecs_t max_mse
if (psize != NULL) { *psize = page * MI_HUGE_OS_PAGE_SIZE; }
if (page != 0) {
mi_assert(start != NULL);
*memid = _mi_memid_create_os(true /* is committed */, all_zero, true /* is_large */);
*memid = _mi_memid_create_os(start, size, true /* is committed */, all_zero, true /* is_large */);
memid->memkind = MI_MEM_OS_HUGE;
mi_assert(memid->is_pinned);
#ifdef MI_TRACK_ASAN
@@ -645,45 +714,57 @@ void* _mi_os_alloc_huge_os_pages(size_t pages, int numa_node, mi_msecs_t max_mse
// free every huge page in a range individually (as we allocated per page)
// note: needed with VirtualAlloc but could potentially be done in one go on mmap'd systems.
static void mi_os_free_huge_os_pages(void* p, size_t size, mi_stats_t* stats) {
static void mi_os_free_huge_os_pages(void* p, size_t size) {
if (p==NULL || size==0) return;
uint8_t* base = (uint8_t*)p;
while (size >= MI_HUGE_OS_PAGE_SIZE) {
mi_os_prim_free(base, MI_HUGE_OS_PAGE_SIZE, true, stats);
mi_os_prim_free(base, MI_HUGE_OS_PAGE_SIZE, MI_HUGE_OS_PAGE_SIZE);
size -= MI_HUGE_OS_PAGE_SIZE;
base += MI_HUGE_OS_PAGE_SIZE;
}
}
/* ----------------------------------------------------------------------------
Support NUMA aware allocation
-----------------------------------------------------------------------------*/
_Atomic(size_t) _mi_numa_node_count; // = 0 // cache the node count
static _Atomic(size_t) mi_numa_node_count; // = 0 // cache the node count
size_t _mi_os_numa_node_count_get(void) {
size_t count = mi_atomic_load_acquire(&_mi_numa_node_count);
if (count <= 0) {
int _mi_os_numa_node_count(void) {
size_t count = mi_atomic_load_acquire(&mi_numa_node_count);
if mi_unlikely(count == 0) {
long ncount = mi_option_get(mi_option_use_numa_nodes); // given explicitly?
if (ncount > 0) {
if (ncount > 0 && ncount < INT_MAX) {
count = (size_t)ncount;
}
else {
count = _mi_prim_numa_node_count(); // or detect dynamically
if (count == 0) count = 1;
const size_t n = _mi_prim_numa_node_count(); // or detect dynamically
if (n == 0 || n > INT_MAX) { count = 1; }
else { count = n; }
}
mi_atomic_store_release(&_mi_numa_node_count, count); // save it
mi_atomic_store_release(&mi_numa_node_count, count); // save it
_mi_verbose_message("using %zd numa regions\n", count);
}
return count;
mi_assert_internal(count > 0 && count <= INT_MAX);
return (int)count;
}
int _mi_os_numa_node_get(mi_os_tld_t* tld) {
MI_UNUSED(tld);
size_t numa_count = _mi_os_numa_node_count();
static int mi_os_numa_node_get(void) {
int numa_count = _mi_os_numa_node_count();
if (numa_count<=1) return 0; // optimize on single numa node systems: always node 0
// never more than the node count and >= 0
size_t numa_node = _mi_prim_numa_node();
const size_t n = _mi_prim_numa_node();
int numa_node = (n < INT_MAX ? (int)n : 0);
if (numa_node >= numa_count) { numa_node = numa_node % numa_count; }
return (int)numa_node;
return numa_node;
}
int _mi_os_numa_node(void) {
if mi_likely(mi_atomic_load_relaxed(&mi_numa_node_count) == 1) {
return 0;
}
else {
return mi_os_numa_node_get();
}
}

181
mimalloc/mimalloc-sys/mimalloc/src/page-queue.c
View File

@@ -1,5 +1,5 @@
/*----------------------------------------------------------------------------
Copyright (c) 2018-2020, Microsoft Research, Daan Leijen
Copyright (c) 2018-2024, Microsoft Research, Daan Leijen
This is free software; you can redistribute it and/or modify it under the
terms of the MIT license. A copy of the license can be found in the file
"LICENSE" at the root of this distribution.
@@ -11,6 +11,10 @@ terms of the MIT license. A copy of the license can be found in the file
#ifndef MI_IN_PAGE_C
#error "this file should be included from 'page.c'"
// include to help an IDE
#include "mimalloc.h"
#include "mimalloc/internal.h"
#include "mimalloc/atomic.h"
#endif
/* -----------------------------------------------------------
@@ -53,27 +57,23 @@ static inline bool mi_page_queue_is_special(const mi_page_queue_t* pq) {
// Returns MI_BIN_HUGE if the size is too large.
// We use `wsize` for the size in "machine word sizes",
// i.e. byte size == `wsize*sizeof(void*)`.
static inline uint8_t mi_bin(size_t size) {
static inline size_t mi_bin(size_t size) {
size_t wsize = _mi_wsize_from_size(size);
uint8_t bin;
if (wsize <= 1) {
bin = 1;
#if defined(MI_ALIGN4W)
if mi_likely(wsize <= 4) {
return (wsize <= 1 ? 1 : (wsize+1)&~1); // round to double word sizes
}
#if defined(MI_ALIGN4W)
else if (wsize <= 4) {
bin = (uint8_t)((wsize+1)&~1); // round to double word sizes
#elif defined(MI_ALIGN2W)
if mi_likely(wsize <= 8) {
return (wsize <= 1 ? 1 : (wsize+1)&~1); // round to double word sizes
}
#elif defined(MI_ALIGN2W)
else if (wsize <= 8) {
bin = (uint8_t)((wsize+1)&~1); // round to double word sizes
#else
if mi_likely(wsize <= 8) {
return (wsize == 0 ? 1 : wsize);
}
#else
else if (wsize <= 8) {
bin = (uint8_t)wsize;
}
#endif
else if (wsize > MI_MEDIUM_OBJ_WSIZE_MAX) {
bin = MI_BIN_HUGE;
#endif
else if mi_unlikely(wsize > MI_MEDIUM_OBJ_WSIZE_MAX) {
return MI_BIN_HUGE;
}
else {
#if defined(MI_ALIGN4W)
@@ -81,15 +81,14 @@ static inline uint8_t mi_bin(size_t size) {
#endif
wsize--;
// find the highest bit
uint8_t b = (uint8_t)mi_bsr(wsize); // note: wsize != 0
const size_t b = (MI_SIZE_BITS - 1 - mi_clz(wsize)); // note: wsize != 0
// and use the top 3 bits to determine the bin (~12.5% worst internal fragmentation).
// - adjust with 3 because we use do not round the first 8 sizes
// which each get an exact bin
bin = ((b << 2) + (uint8_t)((wsize >> (b - 2)) & 0x03)) - 3;
mi_assert_internal(bin < MI_BIN_HUGE);
const size_t bin = ((b << 2) + ((wsize >> (b - 2)) & 0x03)) - 3;
mi_assert_internal(bin > 0 && bin < MI_BIN_HUGE);
return bin;
}
mi_assert_internal(bin > 0 && bin <= MI_BIN_HUGE);
return bin;
}
@@ -98,21 +97,21 @@ static inline uint8_t mi_bin(size_t size) {
Queue of pages with free blocks
----------------------------------------------------------- */
uint8_t _mi_bin(size_t size) {
size_t _mi_bin(size_t size) {
return mi_bin(size);
}
size_t _mi_bin_size(uint8_t bin) {
size_t _mi_bin_size(size_t bin) {
return _mi_heap_empty.pages[bin].block_size;
}
// Good size for allocation
size_t mi_good_size(size_t size) mi_attr_noexcept {
if (size <= MI_MEDIUM_OBJ_SIZE_MAX) {
return _mi_bin_size(mi_bin(size));
return _mi_bin_size(mi_bin(size + MI_PADDING_SIZE));
}
else {
return _mi_align_up(size,_mi_os_page_size());
return _mi_align_up(size + MI_PADDING_SIZE,_mi_os_page_size());
}
}
@@ -137,21 +136,37 @@ static bool mi_heap_contains_queue(const mi_heap_t* heap, const mi_page_queue_t*
}
#endif
static mi_page_queue_t* mi_page_queue_of(const mi_page_t* page) {
uint8_t bin = (mi_page_is_in_full(page) ? MI_BIN_FULL : mi_bin(page->xblock_size));
mi_heap_t* heap = mi_page_heap(page);
mi_assert_internal(heap != NULL && bin <= MI_BIN_FULL);
mi_page_queue_t* pq = &heap->pages[bin];
mi_assert_internal(bin >= MI_BIN_HUGE || page->xblock_size == pq->block_size);
mi_assert_expensive(mi_page_queue_contains(pq, page));
return pq;
static inline bool mi_page_is_large_or_huge(const mi_page_t* page) {
return (mi_page_block_size(page) > MI_MEDIUM_OBJ_SIZE_MAX || mi_page_is_huge(page));
}
static size_t mi_page_bin(const mi_page_t* page) {
const size_t bin = (mi_page_is_in_full(page) ? MI_BIN_FULL : (mi_page_is_huge(page) ? MI_BIN_HUGE : mi_bin(mi_page_block_size(page))));
mi_assert_internal(bin <= MI_BIN_FULL);
return bin;
}
// returns the page bin without using MI_BIN_FULL for statistics
size_t _mi_page_stats_bin(const mi_page_t* page) {
const size_t bin = (mi_page_is_huge(page) ? MI_BIN_HUGE : mi_bin(mi_page_block_size(page)));
mi_assert_internal(bin <= MI_BIN_HUGE);
return bin;
}
static mi_page_queue_t* mi_heap_page_queue_of(mi_heap_t* heap, const mi_page_t* page) {
uint8_t bin = (mi_page_is_in_full(page) ? MI_BIN_FULL : mi_bin(page->xblock_size));
mi_assert_internal(bin <= MI_BIN_FULL);
mi_assert_internal(heap!=NULL);
const size_t bin = mi_page_bin(page);
mi_page_queue_t* pq = &heap->pages[bin];
mi_assert_internal(mi_page_is_in_full(page) || page->xblock_size == pq->block_size);
mi_assert_internal((mi_page_block_size(page) == pq->block_size) ||
(mi_page_is_large_or_huge(page) && mi_page_queue_is_huge(pq)) ||
(mi_page_is_in_full(page) && mi_page_queue_is_full(pq)));
return pq;
}
static mi_page_queue_t* mi_page_queue_of(const mi_page_t* page) {
mi_heap_t* heap = mi_page_heap(page);
mi_page_queue_t* pq = mi_heap_page_queue_of(heap, page);
mi_assert_expensive(mi_page_queue_contains(pq, page));
return pq;
}
@@ -181,7 +196,7 @@ static inline void mi_heap_queue_first_update(mi_heap_t* heap, const mi_page_que
}
else {
// find previous size; due to minimal alignment upto 3 previous bins may need to be skipped
uint8_t bin = mi_bin(size);
size_t bin = mi_bin(size);
const mi_page_queue_t* prev = pq - 1;
while( bin == mi_bin(prev->block_size) && prev > &heap->pages[0]) {
prev--;
@@ -206,7 +221,9 @@ static bool mi_page_queue_is_empty(mi_page_queue_t* queue) {
static void mi_page_queue_remove(mi_page_queue_t* queue, mi_page_t* page) {
mi_assert_internal(page != NULL);
mi_assert_expensive(mi_page_queue_contains(queue, page));
mi_assert_internal(page->xblock_size == queue->block_size || (page->xblock_size > MI_MEDIUM_OBJ_SIZE_MAX && mi_page_queue_is_huge(queue)) || (mi_page_is_in_full(page) && mi_page_queue_is_full(queue)));
mi_assert_internal(mi_page_block_size(page) == queue->block_size ||
(mi_page_is_large_or_huge(page) && mi_page_queue_is_huge(queue)) ||
(mi_page_is_in_full(page) && mi_page_queue_is_full(queue)));
mi_heap_t* heap = mi_page_heap(page);
if (page->prev != NULL) page->prev->next = page->next;
@@ -232,8 +249,8 @@ static void mi_page_queue_push(mi_heap_t* heap, mi_page_queue_t* queue, mi_page_
#if MI_HUGE_PAGE_ABANDON
mi_assert_internal(_mi_page_segment(page)->kind != MI_SEGMENT_HUGE);
#endif
mi_assert_internal(page->xblock_size == queue->block_size ||
(page->xblock_size > MI_MEDIUM_OBJ_SIZE_MAX) ||
mi_assert_internal(mi_page_block_size(page) == queue->block_size ||
(mi_page_is_large_or_huge(page) && mi_page_queue_is_huge(queue)) ||
(mi_page_is_in_full(page) && mi_page_queue_is_full(queue)));
mi_page_set_in_full(page, mi_page_queue_is_full(queue));
@@ -254,19 +271,30 @@ static void mi_page_queue_push(mi_heap_t* heap, mi_page_queue_t* queue, mi_page_
heap->page_count++;
}
static void mi_page_queue_move_to_front(mi_heap_t* heap, mi_page_queue_t* queue, mi_page_t* page) {
mi_assert_internal(mi_page_heap(page) == heap);
mi_assert_internal(mi_page_queue_contains(queue, page));
if (queue->first == page) return;
mi_page_queue_remove(queue, page);
mi_page_queue_push(heap, queue, page);
mi_assert_internal(queue->first == page);
}
static void mi_page_queue_enqueue_from(mi_page_queue_t* to, mi_page_queue_t* from, mi_page_t* page) {
static void mi_page_queue_enqueue_from_ex(mi_page_queue_t* to, mi_page_queue_t* from, bool enqueue_at_end, mi_page_t* page) {
mi_assert_internal(page != NULL);
mi_assert_expensive(mi_page_queue_contains(from, page));
mi_assert_expensive(!mi_page_queue_contains(to, page));
mi_assert_internal((page->xblock_size == to->block_size && page->xblock_size == from->block_size) ||
(page->xblock_size == to->block_size && mi_page_queue_is_full(from)) ||
(page->xblock_size == from->block_size && mi_page_queue_is_full(to)) ||
(page->xblock_size > MI_LARGE_OBJ_SIZE_MAX && mi_page_queue_is_huge(to)) ||
(page->xblock_size > MI_LARGE_OBJ_SIZE_MAX && mi_page_queue_is_full(to)));
const size_t bsize = mi_page_block_size(page);
MI_UNUSED(bsize);
mi_assert_internal((bsize == to->block_size && bsize == from->block_size) ||
(bsize == to->block_size && mi_page_queue_is_full(from)) ||
(bsize == from->block_size && mi_page_queue_is_full(to)) ||
(mi_page_is_large_or_huge(page) && mi_page_queue_is_huge(to)) ||
(mi_page_is_large_or_huge(page) && mi_page_queue_is_full(to)));
mi_heap_t* heap = mi_page_heap(page);
// delete from `from`
if (page->prev != NULL) page->prev->next = page->next;
if (page->next != NULL) page->next->prev = page->prev;
if (page == from->last) from->last = page->prev;
@@ -277,22 +305,59 @@ static void mi_page_queue_enqueue_from(mi_page_queue_t* to, mi_page_queue_t* fro
mi_heap_queue_first_update(heap, from);
}
page->prev = to->last;
page->next = NULL;
if (to->last != NULL) {
mi_assert_internal(heap == mi_page_heap(to->last));
to->last->next = page;
to->last = page;
// insert into `to`
if (enqueue_at_end) {
// enqueue at the end
page->prev = to->last;
page->next = NULL;
if (to->last != NULL) {
mi_assert_internal(heap == mi_page_heap(to->last));
to->last->next = page;
to->last = page;
}
else {
to->first = page;
to->last = page;
mi_heap_queue_first_update(heap, to);
}
}
else {
to->first = page;
to->last = page;
mi_heap_queue_first_update(heap, to);
if (to->first != NULL) {
// enqueue at 2nd place
mi_assert_internal(heap == mi_page_heap(to->first));
mi_page_t* next = to->first->next;
page->prev = to->first;
page->next = next;
to->first->next = page;
if (next != NULL) {
next->prev = page;
}
else {
to->last = page;
}
}
else {
// enqueue at the head (singleton list)
page->prev = NULL;
page->next = NULL;
to->first = page;
to->last = page;
mi_heap_queue_first_update(heap, to);
}
}
mi_page_set_in_full(page, mi_page_queue_is_full(to));
}
static void mi_page_queue_enqueue_from(mi_page_queue_t* to, mi_page_queue_t* from, mi_page_t* page) {
mi_page_queue_enqueue_from_ex(to, from, true /* enqueue at the end */, page);
}
static void mi_page_queue_enqueue_from_full(mi_page_queue_t* to, mi_page_queue_t* from, mi_page_t* page) {
// note: we could insert at the front to increase reuse, but it slows down certain benchmarks (like `alloc-test`)
mi_page_queue_enqueue_from_ex(to, from, true /* enqueue at the end of the `to` queue? */, page);
}
// Only called from `mi_heap_absorb`.
size_t _mi_page_queue_append(mi_heap_t* heap, mi_page_queue_t* pq, mi_page_queue_t* append) {
mi_assert_internal(mi_heap_contains_queue(heap,pq));

299
mimalloc/mimalloc-sys/mimalloc/src/page.c
View File

@@ -1,5 +1,5 @@
/*----------------------------------------------------------------------------
Copyright (c) 2018-2020, Microsoft Research, Daan Leijen
Copyright (c) 2018-2024, Microsoft Research, Daan Leijen
This is free software; you can redistribute it and/or modify it under the
terms of the MIT license. A copy of the license can be found in the file
"LICENSE" at the root of this distribution.
@@ -37,7 +37,7 @@ static inline mi_block_t* mi_page_block_at(const mi_page_t* page, void* page_sta
}
static void mi_page_init(mi_heap_t* heap, mi_page_t* page, size_t size, mi_tld_t* tld);
static void mi_page_extend_free(mi_heap_t* heap, mi_page_t* page, mi_tld_t* tld);
static bool mi_page_extend_free(mi_heap_t* heap, mi_page_t* page, mi_tld_t* tld);
#if (MI_DEBUG>=3)
static size_t mi_page_list_count(mi_page_t* page, mi_block_t* head) {
@@ -59,7 +59,7 @@ static inline uint8_t* mi_page_area(const mi_page_t* page) {
static bool mi_page_list_is_valid(mi_page_t* page, mi_block_t* p) {
size_t psize;
uint8_t* page_area = _mi_page_start(_mi_page_segment(page), page, &psize);
uint8_t* page_area = _mi_segment_page_start(_mi_page_segment(page), page, &psize);
mi_block_t* start = (mi_block_t*)page_area;
mi_block_t* end = (mi_block_t*)(page_area + psize);
while(p != NULL) {
@@ -78,14 +78,13 @@ static bool mi_page_list_is_valid(mi_page_t* page, mi_block_t* p) {
}
static bool mi_page_is_valid_init(mi_page_t* page) {
mi_assert_internal(page->xblock_size > 0);
mi_assert_internal(mi_page_block_size(page) > 0);
mi_assert_internal(page->used <= page->capacity);
mi_assert_internal(page->capacity <= page->reserved);
mi_segment_t* segment = _mi_page_segment(page);
uint8_t* start = _mi_page_start(segment,page,NULL);
mi_assert_internal(start == _mi_segment_page_start(segment,page,NULL));
//const size_t bsize = mi_page_block_size(page);
uint8_t* start = mi_page_start(page);
mi_assert_internal(start == _mi_segment_page_start(_mi_page_segment(page), page, NULL));
mi_assert_internal(page->is_huge == (_mi_page_segment(page)->kind == MI_SEGMENT_HUGE));
//mi_assert_internal(start + page->capacity*page->block_size == page->top);
mi_assert_internal(mi_page_list_is_valid(page,page->free));
@@ -113,7 +112,7 @@ static bool mi_page_is_valid_init(mi_page_t* page) {
return true;
}
extern bool _mi_process_is_initialized; // has mi_process_init been called?
extern mi_decl_hidden bool _mi_process_is_initialized; // has mi_process_init been called?
bool _mi_page_is_valid(mi_page_t* page) {
mi_assert_internal(mi_page_is_valid_init(page));
@@ -125,9 +124,9 @@ bool _mi_page_is_valid(mi_page_t* page) {
mi_assert_internal(!_mi_process_is_initialized || segment->thread_id==0 || segment->thread_id == mi_page_heap(page)->thread_id);
#if MI_HUGE_PAGE_ABANDON
if (segment->kind != MI_SEGMENT_HUGE)
if (segment->kind != MI_SEGMENT_HUGE)
#endif
{
{
mi_page_queue_t* pq = mi_page_queue_of(page);
mi_assert_internal(mi_page_queue_contains(pq, page));
mi_assert_internal(pq->block_size==mi_page_block_size(page) || mi_page_block_size(page) > MI_MEDIUM_OBJ_SIZE_MAX || mi_page_is_in_full(page));
@@ -193,8 +192,8 @@ static void _mi_page_thread_free_collect(mi_page_t* page)
if (head == NULL) return;
// find the tail -- also to get a proper count (without data races)
uint32_t max_count = page->capacity; // cannot collect more than capacity
uint32_t count = 1;
size_t max_count = page->capacity; // cannot collect more than capacity
size_t count = 1;
mi_block_t* tail = head;
mi_block_t* next;
while ((next = mi_block_next(page,tail)) != NULL && count <= max_count) {
@@ -212,7 +211,7 @@ static void _mi_page_thread_free_collect(mi_page_t* page)
page->local_free = head;
// update counts now
page->used -= count;
page->used -= (uint16_t)count;
}
void _mi_page_free_collect(mi_page_t* page, bool force) {
@@ -263,7 +262,7 @@ void _mi_page_reclaim(mi_heap_t* heap, mi_page_t* page) {
#if MI_HUGE_PAGE_ABANDON
mi_assert_internal(_mi_page_segment(page)->kind != MI_SEGMENT_HUGE);
#endif
// TODO: push on full queue immediately if it is full?
mi_page_queue_t* pq = mi_page_queue(heap, mi_page_block_size(page));
mi_page_queue_push(heap, pq, page);
@@ -277,19 +276,22 @@ static mi_page_t* mi_page_fresh_alloc(mi_heap_t* heap, mi_page_queue_t* pq, size
mi_assert_internal(mi_heap_contains_queue(heap, pq));
mi_assert_internal(page_alignment > 0 || block_size > MI_MEDIUM_OBJ_SIZE_MAX || block_size == pq->block_size);
#endif
mi_page_t* page = _mi_segment_page_alloc(heap, block_size, page_alignment, &heap->tld->segments, &heap->tld->os);
mi_page_t* page = _mi_segment_page_alloc(heap, block_size, page_alignment, &heap->tld->segments);
if (page == NULL) {
// this may be out-of-memory, or an abandoned page was reclaimed (and in our queue)
return NULL;
}
#if MI_HUGE_PAGE_ABANDON
mi_assert_internal(pq==NULL || _mi_page_segment(page)->page_kind != MI_PAGE_HUGE);
#endif
mi_assert_internal(page_alignment >0 || block_size > MI_MEDIUM_OBJ_SIZE_MAX || _mi_page_segment(page)->kind != MI_SEGMENT_HUGE);
mi_assert_internal(pq!=NULL || page->xblock_size != 0);
mi_assert_internal(pq!=NULL || mi_page_block_size(page) >= block_size);
// a fresh page was found, initialize it
const size_t full_block_size = ((pq == NULL || mi_page_queue_is_huge(pq)) ? mi_page_block_size(page) : block_size); // see also: mi_segment_huge_page_alloc
const size_t full_block_size = (pq == NULL || mi_page_is_huge(page) ? mi_page_block_size(page) : block_size); // see also: mi_segment_huge_page_alloc
mi_assert_internal(full_block_size >= block_size);
mi_page_init(heap, page, full_block_size, heap->tld);
mi_heap_stat_increase(heap, pages, 1);
mi_heap_stat_increase(heap, page_bins[_mi_page_stats_bin(page)], 1);
if (pq != NULL) { mi_page_queue_push(heap, pq, page); }
mi_assert_expensive(_mi_page_is_valid(page));
return page;
@@ -357,7 +359,7 @@ void _mi_page_unfull(mi_page_t* page) {
mi_page_set_in_full(page, false); // to get the right queue
mi_page_queue_t* pq = mi_heap_page_queue_of(heap, page);
mi_page_set_in_full(page, true);
mi_page_queue_enqueue_from(pq, pqfull, page);
mi_page_queue_enqueue_from_full(pq, pqfull, page);
}
static void mi_page_to_full(mi_page_t* page, mi_page_queue_t* pq) {
@@ -403,6 +405,28 @@ void _mi_page_abandon(mi_page_t* page, mi_page_queue_t* pq) {
_mi_segment_page_abandon(page,segments_tld);
}
// force abandon a page
void _mi_page_force_abandon(mi_page_t* page) {
mi_heap_t* heap = mi_page_heap(page);
// mark page as not using delayed free
_mi_page_use_delayed_free(page, MI_NEVER_DELAYED_FREE, false);
// ensure this page is no longer in the heap delayed free list
_mi_heap_delayed_free_all(heap);
// We can still access the page meta-info even if it is freed as we ensure
// in `mi_segment_force_abandon` that the segment is not freed (yet)
if (page->capacity == 0) return; // it may have been freed now
// and now unlink it from the page queue and abandon (or free)
mi_page_queue_t* pq = mi_heap_page_queue_of(heap, page);
if (mi_page_all_free(page)) {
_mi_page_free(page, pq, false);
}
else {
_mi_page_abandon(page, pq);
}
}
// Free a page with no more free blocks
void _mi_page_free(mi_page_t* page, mi_page_queue_t* pq, bool force) {
@@ -415,20 +439,18 @@ void _mi_page_free(mi_page_t* page, mi_page_queue_t* pq, bool force) {
// no more aligned blocks in here
mi_page_set_has_aligned(page, false);
mi_heap_t* heap = mi_page_heap(page);
// remove from the page list
// (no need to do _mi_heap_delayed_free first as all blocks are already free)
mi_heap_t* heap = mi_page_heap(page);
mi_segments_tld_t* segments_tld = &heap->tld->segments;
mi_page_queue_remove(pq, page);
// and free it
// and free it
mi_page_set_heap(page,NULL);
_mi_segment_page_free(page, force, segments_tld);
}
// Retire parameters
#define MI_MAX_RETIRE_SIZE (MI_MEDIUM_OBJ_SIZE_MAX)
#define MI_MAX_RETIRE_SIZE MI_MEDIUM_OBJ_SIZE_MAX // should be less than size for MI_BIN_HUGE
#define MI_RETIRE_CYCLES (16)
// Retire a page with no more used blocks
@@ -441,7 +463,7 @@ void _mi_page_retire(mi_page_t* page) mi_attr_noexcept {
mi_assert_internal(page != NULL);
mi_assert_expensive(_mi_page_is_valid(page));
mi_assert_internal(mi_page_all_free(page));
mi_page_set_has_aligned(page, false);
// don't retire too often..
@@ -451,10 +473,12 @@ void _mi_page_retire(mi_page_t* page) mi_attr_noexcept {
// how to check this efficiently though...
// for now, we don't retire if it is the only page left of this size class.
mi_page_queue_t* pq = mi_page_queue_of(page);
if mi_likely(page->xblock_size <= MI_MAX_RETIRE_SIZE && !mi_page_queue_is_special(pq)) { // not too large && not full or huge queue?
#if MI_RETIRE_CYCLES > 0
const size_t bsize = mi_page_block_size(page);
if mi_likely( /* bsize < MI_MAX_RETIRE_SIZE && */ !mi_page_queue_is_special(pq)) { // not full or huge queue?
if (pq->last==page && pq->first==page) { // the only page in the queue?
mi_stat_counter_increase(_mi_stats_main.page_no_retire,1);
page->retire_expire = 1 + (page->xblock_size <= MI_SMALL_OBJ_SIZE_MAX ? MI_RETIRE_CYCLES : MI_RETIRE_CYCLES/4);
mi_stat_counter_increase(_mi_stats_main.pages_retire,1);
page->retire_expire = (bsize <= MI_SMALL_OBJ_SIZE_MAX ? MI_RETIRE_CYCLES : MI_RETIRE_CYCLES/4);
mi_heap_t* heap = mi_page_heap(page);
mi_assert_internal(pq >= heap->pages);
const size_t index = pq - heap->pages;
@@ -462,9 +486,10 @@ void _mi_page_retire(mi_page_t* page) mi_attr_noexcept {
if (index < heap->page_retired_min) heap->page_retired_min = index;
if (index > heap->page_retired_max) heap->page_retired_max = index;
mi_assert_internal(mi_page_all_free(page));
return; // dont't free after all
return; // don't free after all
}
}
#endif
_mi_page_free(page, pq, false);
}
@@ -516,7 +541,7 @@ static void mi_page_free_list_extend_secure(mi_heap_t* const heap, mi_page_t* co
#endif
mi_assert_internal(page->capacity + extend <= page->reserved);
mi_assert_internal(bsize == mi_page_block_size(page));
void* const page_area = _mi_page_start(_mi_page_segment(page), page, NULL);
void* const page_area = mi_page_start(page);
// initialize a randomized free list
// set up `slice_count` slices to alternate between
@@ -574,7 +599,7 @@ static mi_decl_noinline void mi_page_free_list_extend( mi_page_t* const page, co
#endif
mi_assert_internal(page->capacity + extend <= page->reserved);
mi_assert_internal(bsize == mi_page_block_size(page));
void* const page_area = _mi_page_start(_mi_page_segment(page), page, NULL );
void* const page_area = mi_page_start(page);
mi_block_t* const start = mi_page_block_at(page, page_area, bsize, page->capacity);
@@ -607,26 +632,23 @@ static mi_decl_noinline void mi_page_free_list_extend( mi_page_t* const page, co
// Note: we also experimented with "bump" allocation on the first
// allocations but this did not speed up any benchmark (due to an
// extra test in malloc? or cache effects?)
static void mi_page_extend_free(mi_heap_t* heap, mi_page_t* page, mi_tld_t* tld) {
MI_UNUSED(tld);
static bool mi_page_extend_free(mi_heap_t* heap, mi_page_t* page, mi_tld_t* tld) {
mi_assert_expensive(mi_page_is_valid_init(page));
#if (MI_SECURE<=2)
mi_assert(page->free == NULL);
mi_assert(page->local_free == NULL);
if (page->free != NULL) return;
if (page->free != NULL) return true;
#endif
if (page->capacity >= page->reserved) return;
if (page->capacity >= page->reserved) return true;
size_t page_size;
_mi_page_start(_mi_page_segment(page), page, &page_size);
mi_stat_counter_increase(tld->stats.pages_extended, 1);
// calculate the extend count
const size_t bsize = (page->xblock_size < MI_HUGE_BLOCK_SIZE ? page->xblock_size : page_size);
const size_t bsize = mi_page_block_size(page);
size_t extend = page->reserved - page->capacity;
mi_assert_internal(extend > 0);
size_t max_extend = (bsize >= MI_MAX_EXTEND_SIZE ? MI_MIN_EXTEND : MI_MAX_EXTEND_SIZE/(uint32_t)bsize);
size_t max_extend = (bsize >= MI_MAX_EXTEND_SIZE ? MI_MIN_EXTEND : MI_MAX_EXTEND_SIZE/bsize);
if (max_extend < MI_MIN_EXTEND) { max_extend = MI_MIN_EXTEND; }
mi_assert_internal(max_extend > 0);
@@ -650,6 +672,7 @@ static void mi_page_extend_free(mi_heap_t* heap, mi_page_t* page, mi_tld_t* tld)
page->capacity += (uint16_t)extend;
mi_stat_increase(tld->stats.page_committed, extend * bsize);
mi_assert_expensive(mi_page_is_valid_init(page));
return true;
}
// Initialize a fresh page
@@ -660,11 +683,10 @@ static void mi_page_init(mi_heap_t* heap, mi_page_t* page, size_t block_size, mi
mi_assert_internal(block_size > 0);
// set fields
mi_page_set_heap(page, heap);
page->xblock_size = (block_size < MI_HUGE_BLOCK_SIZE ? (uint32_t)block_size : MI_HUGE_BLOCK_SIZE); // initialize before _mi_segment_page_start
page->block_size = block_size;
size_t page_size;
const void* page_start = _mi_segment_page_start(segment, page, &page_size);
MI_UNUSED(page_start);
mi_track_mem_noaccess(page_start,page_size);
page->page_start = _mi_segment_page_start(segment, page, &page_size);
mi_track_mem_noaccess(page->page_start,page_size);
mi_assert_internal(mi_page_block_size(page) <= page_size);
mi_assert_internal(page_size <= page->slice_count*MI_SEGMENT_SLICE_SIZE);
mi_assert_internal(page_size / block_size < (1L<<16));
@@ -677,12 +699,18 @@ static void mi_page_init(mi_heap_t* heap, mi_page_t* page, size_t block_size, mi
page->free_is_zero = page->is_zero_init;
#if MI_DEBUG>2
if (page->is_zero_init) {
mi_track_mem_defined(page_start, page_size);
mi_assert_expensive(mi_mem_is_zero(page_start, page_size));
mi_track_mem_defined(page->page_start, page_size);
mi_assert_expensive(mi_mem_is_zero(page->page_start, page_size));
}
#endif
mi_assert_internal(page->is_committed);
if (block_size > 0 && _mi_is_power_of_two(block_size)) {
page->block_size_shift = (uint8_t)(mi_ctz((uintptr_t)block_size));
}
else {
page->block_size_shift = 0;
}
mi_assert_internal(page->capacity == 0);
mi_assert_internal(page->free == NULL);
mi_assert_internal(page->used == 0);
@@ -695,11 +723,14 @@ static void mi_page_init(mi_heap_t* heap, mi_page_t* page, size_t block_size, mi
mi_assert_internal(page->keys[0] != 0);
mi_assert_internal(page->keys[1] != 0);
#endif
mi_assert_internal(page->block_size_shift == 0 || (block_size == ((size_t)1 << page->block_size_shift)));
mi_assert_expensive(mi_page_is_valid_init(page));
// initialize an initial free list
mi_page_extend_free(heap,page,tld);
mi_assert(mi_page_immediate_available(page));
if (mi_page_extend_free(heap,page,tld)) {
mi_assert(mi_page_immediate_available(page));
}
return;
}
@@ -707,6 +738,17 @@ static void mi_page_init(mi_heap_t* heap, mi_page_t* page, size_t block_size, mi
Find pages with free blocks
-------------------------------------------------------------*/
// search for a best next page to use for at most N pages (often cut short if immediate blocks are available)
#define MI_MAX_CANDIDATE_SEARCH (4)
// is the page not yet used up to its reserved space?
static bool mi_page_is_expandable(const mi_page_t* page) {
mi_assert_internal(page != NULL);
mi_assert_internal(page->capacity <= page->reserved);
return (page->capacity < page->reserved);
}
// Find a page with free blocks of `page->block_size`.
static mi_page_t* mi_page_queue_find_free_ex(mi_heap_t* heap, mi_page_queue_t* pq, bool first_try)
{
@@ -714,38 +756,82 @@ static mi_page_t* mi_page_queue_find_free_ex(mi_heap_t* heap, mi_page_queue_t* p
#if MI_STAT
size_t count = 0;
#endif
size_t candidate_count = 0; // we reset this on the first candidate to limit the search
mi_page_t* page_candidate = NULL; // a page with free space
mi_page_t* page = pq->first;
while (page != NULL)
{
mi_page_t* next = page->next; // remember next
#if MI_STAT
#if MI_STAT
count++;
#endif
candidate_count++;
// 0. collect freed blocks by us and other threads
// collect freed blocks by us and other threads
_mi_page_free_collect(page, false);
// 1. if the page contains free blocks, we are done
if (mi_page_immediate_available(page)) {
#if MI_MAX_CANDIDATE_SEARCH > 1
// search up to N pages for a best candidate
// is the local free list non-empty?
const bool immediate_available = mi_page_immediate_available(page);
// if the page is completely full, move it to the `mi_pages_full`
// queue so we don't visit long-lived pages too often.
if (!immediate_available && !mi_page_is_expandable(page)) {
mi_assert_internal(!mi_page_is_in_full(page) && !mi_page_immediate_available(page));
mi_page_to_full(page, pq);
}
else {
// the page has free space, make it a candidate
// we prefer non-expandable pages with high usage as candidates (to reduce commit, and increase chances of free-ing up pages)
if (page_candidate == NULL) {
page_candidate = page;
candidate_count = 0;
}
// prefer to reuse fuller pages (in the hope the less used page gets freed)
else if (page->used >= page_candidate->used && !mi_page_is_mostly_used(page) && !mi_page_is_expandable(page)) {
page_candidate = page;
}
// if we find a non-expandable candidate, or searched for N pages, return with the best candidate
if (immediate_available || candidate_count > MI_MAX_CANDIDATE_SEARCH) {
mi_assert_internal(page_candidate!=NULL);
break;
}
}
#else
// first-fit algorithm
// If the page contains free blocks, we are done
if (mi_page_immediate_available(page) || mi_page_is_expandable(page)) {
break; // pick this one
}
// 2. Try to extend
if (page->capacity < page->reserved) {
mi_page_extend_free(heap, page, heap->tld);
mi_assert_internal(mi_page_immediate_available(page));
break;
}
// 3. If the page is completely full, move it to the `mi_pages_full`
// If the page is completely full, move it to the `mi_pages_full`
// queue so we don't visit long-lived pages too often.
mi_assert_internal(!mi_page_is_in_full(page) && !mi_page_immediate_available(page));
mi_page_to_full(page, pq);
#endif
page = next;
} // for each page
mi_heap_stat_counter_increase(heap, searches, count);
mi_heap_stat_counter_increase(heap, page_searches, count);
mi_heap_stat_counter_increase(heap, page_searches_count, 1);
// set the page to the best candidate
if (page_candidate != NULL) {
page = page_candidate;
}
if (page != NULL) {
if (!mi_page_immediate_available(page)) {
mi_assert_internal(mi_page_is_expandable(page));
if (!mi_page_extend_free(heap, page, heap->tld)) {
page = NULL; // failed to extend
}
}
mi_assert_internal(page == NULL || mi_page_immediate_available(page));
}
if (page == NULL) {
_mi_heap_collect_retired(heap, false); // perhaps make a page available?
@@ -756,10 +842,14 @@ static mi_page_t* mi_page_queue_find_free_ex(mi_heap_t* heap, mi_page_queue_t* p
}
}
else {
mi_assert(pq->first == page);
// move the page to the front of the queue
mi_page_queue_move_to_front(heap, pq, page);
page->retire_expire = 0;
// _mi_heap_collect_retired(heap, false); // update retire counts; note: increases rss on MemoryLoad bench so don't do this
}
mi_assert_internal(page == NULL || mi_page_immediate_available(page));
return page;
}
@@ -767,7 +857,9 @@ static mi_page_t* mi_page_queue_find_free_ex(mi_heap_t* heap, mi_page_queue_t* p
// Find a page with free blocks of `size`.
static inline mi_page_t* mi_find_free_page(mi_heap_t* heap, size_t size) {
mi_page_queue_t* pq = mi_page_queue(heap,size);
mi_page_queue_t* pq = mi_page_queue(heap, size);
// check the first page: we even do this with candidate search or otherwise we re-search every time
mi_page_t* page = pq->first;
if (page != NULL) {
#if (MI_SECURE>=3) // in secure mode, we extend half the time to increase randomness
@@ -786,6 +878,7 @@ static inline mi_page_t* mi_find_free_page(mi_heap_t* heap, size_t size) {
return page; // fast path
}
}
return mi_page_queue_find_free_ex(heap, pq, true);
}
@@ -820,11 +913,9 @@ void mi_register_deferred_free(mi_deferred_free_fun* fn, void* arg) mi_attr_noex
----------------------------------------------------------- */
// Large and huge page allocation.
// Huge pages are allocated directly without being in a queue.
// Because huge pages contain just one block, and the segment contains
// just that page, we always treat them as abandoned and any thread
// that frees the block can free the whole page and segment directly.
// Huge pages are also use if the requested alignment is very large (> MI_ALIGNMENT_MAX).
// Huge pages contain just one block, and the segment contains just that page (as `MI_SEGMENT_HUGE`).
// Huge pages are also use if the requested alignment is very large (> MI_BLOCK_ALIGNMENT_MAX)
// so their size is not always `> MI_LARGE_OBJ_SIZE_MAX`.
static mi_page_t* mi_large_huge_page_alloc(mi_heap_t* heap, size_t size, size_t page_alignment) {
size_t block_size = _mi_os_good_alloc_size(size);
mi_assert_internal(mi_bin(block_size) == MI_BIN_HUGE || page_alignment > 0);
@@ -832,33 +923,35 @@ static mi_page_t* mi_large_huge_page_alloc(mi_heap_t* heap, size_t size, size_t
#if MI_HUGE_PAGE_ABANDON
mi_page_queue_t* pq = (is_huge ? NULL : mi_page_queue(heap, block_size));
#else
mi_page_queue_t* pq = mi_page_queue(heap, is_huge ? MI_HUGE_BLOCK_SIZE : block_size); // not block_size as that can be low if the page_alignment > 0
mi_page_queue_t* pq = mi_page_queue(heap, is_huge ? MI_LARGE_OBJ_SIZE_MAX+1 : block_size);
mi_assert_internal(!is_huge || mi_page_queue_is_huge(pq));
#endif
mi_page_t* page = mi_page_fresh_alloc(heap, pq, block_size, page_alignment);
if (page != NULL) {
mi_assert_internal(mi_page_immediate_available(page));
if (is_huge) {
mi_assert_internal(mi_page_is_huge(page));
mi_assert_internal(_mi_page_segment(page)->kind == MI_SEGMENT_HUGE);
mi_assert_internal(_mi_page_segment(page)->used==1);
#if MI_HUGE_PAGE_ABANDON
mi_assert_internal(_mi_page_segment(page)->thread_id==0); // abandoned, not in the huge queue
mi_page_set_heap(page, NULL);
#endif
#endif
}
else {
mi_assert_internal(_mi_page_segment(page)->kind != MI_SEGMENT_HUGE);
mi_assert_internal(!mi_page_is_huge(page));
}
const size_t bsize = mi_page_usable_block_size(page); // note: not `mi_page_block_size` to account for padding
if (bsize <= MI_LARGE_OBJ_SIZE_MAX) {
mi_heap_stat_increase(heap, large, bsize);
mi_heap_stat_counter_increase(heap, large_count, 1);
/*if (bsize <= MI_LARGE_OBJ_SIZE_MAX) {
mi_heap_stat_increase(heap, malloc_large, bsize);
mi_heap_stat_counter_increase(heap, malloc_large_count, 1);
}
else {
mi_heap_stat_increase(heap, huge, bsize);
mi_heap_stat_counter_increase(heap, huge_count, 1);
else */
{
_mi_stat_increase(&heap->tld->stats.malloc_huge, bsize);
_mi_stat_counter_increase(&heap->tld->stats.malloc_huge_count, 1);
}
}
return page;
@@ -869,9 +962,9 @@ static mi_page_t* mi_large_huge_page_alloc(mi_heap_t* heap, size_t size, size_t
// Note: in debug mode the size includes MI_PADDING_SIZE and might have overflowed.
static mi_page_t* mi_find_page(mi_heap_t* heap, size_t size, size_t huge_alignment) mi_attr_noexcept {
// huge allocation?
const size_t req_size = size - MI_PADDING_SIZE; // correct for padding_size in case of an overflow on `size`
const size_t req_size = size - MI_PADDING_SIZE; // correct for padding_size in case of an overflow on `size`
if mi_unlikely(req_size > (MI_MEDIUM_OBJ_SIZE_MAX - MI_PADDING_SIZE) || huge_alignment > 0) {
if mi_unlikely(req_size > PTRDIFF_MAX) { // we don't allocate more than PTRDIFF_MAX (see <https://sourceware.org/ml/libc-announce/2019/msg00001.html>)
if mi_unlikely(req_size > MI_MAX_ALLOC_SIZE) {
_mi_error_message(EOVERFLOW, "allocation request is too large (%zu bytes)\n", req_size);
return NULL;
}
@@ -882,7 +975,7 @@ static mi_page_t* mi_find_page(mi_heap_t* heap, size_t size, size_t huge_alignme
else {
// otherwise find a page with free blocks in our size segregated queues
#if MI_PADDING
mi_assert_internal(size >= MI_PADDING_SIZE);
mi_assert_internal(size >= MI_PADDING_SIZE);
#endif
return mi_find_free_page(heap, size);
}
@@ -890,24 +983,36 @@ static mi_page_t* mi_find_page(mi_heap_t* heap, size_t size, size_t huge_alignme
// Generic allocation routine if the fast path (`alloc.c:mi_page_malloc`) does not succeed.
// Note: in debug mode the size includes MI_PADDING_SIZE and might have overflowed.
// The `huge_alignment` is normally 0 but is set to a multiple of MI_SEGMENT_SIZE for
// very large requested alignments in which case we use a huge segment.
void* _mi_malloc_generic(mi_heap_t* heap, size_t size, bool zero, size_t huge_alignment) mi_attr_noexcept
// The `huge_alignment` is normally 0 but is set to a multiple of MI_SLICE_SIZE for
// very large requested alignments in which case we use a huge singleton page.
void* _mi_malloc_generic(mi_heap_t* heap, size_t size, bool zero, size_t huge_alignment, size_t* usable) mi_attr_noexcept
{
mi_assert_internal(heap != NULL);
// initialize if necessary
if mi_unlikely(!mi_heap_is_initialized(heap)) {
heap = mi_heap_get_default(); // calls mi_thread_init
heap = mi_heap_get_default(); // calls mi_thread_init
if mi_unlikely(!mi_heap_is_initialized(heap)) { return NULL; }
}
mi_assert_internal(mi_heap_is_initialized(heap));
// call potential deferred free routines
_mi_deferred_free(heap, false);
// do administrative tasks every N generic mallocs
if mi_unlikely(++heap->generic_count >= 100) {
heap->generic_collect_count += heap->generic_count;
heap->generic_count = 0;
// call potential deferred free routines
_mi_deferred_free(heap, false);
// free delayed frees from other threads (but skip contended ones)
_mi_heap_delayed_free_partial(heap);
// free delayed frees from other threads (but skip contended ones)
_mi_heap_delayed_free_partial(heap);
// collect every once in a while (10000 by default)
const long generic_collect = mi_option_get_clamp(mi_option_generic_collect, 1, 1000000L);
if (heap->generic_collect_count >= generic_collect) {
heap->generic_collect_count = 0;
mi_heap_collect(heap, false /* force? */);
}
}
// find (or allocate) a page of the right size
mi_page_t* page = mi_find_page(heap, size, huge_alignment);
@@ -926,14 +1031,20 @@ void* _mi_malloc_generic(mi_heap_t* heap, size_t size, bool zero, size_t huge_al
mi_assert_internal(mi_page_block_size(page) >= size);
// and try again, this time succeeding! (i.e. this should never recurse through _mi_page_malloc)
if mi_unlikely(zero && page->xblock_size == 0) {
void* p;
if mi_unlikely(zero && mi_page_is_huge(page)) {
// note: we cannot call _mi_page_malloc with zeroing for huge blocks; we zero it afterwards in that case.
void* p = _mi_page_malloc(heap, page, size, false);
p = _mi_page_malloc_zero(heap, page, size, false, usable);
mi_assert_internal(p != NULL);
_mi_memzero_aligned(p, mi_page_usable_block_size(page));
return p;
}
else {
return _mi_page_malloc(heap, page, size, zero);
p = _mi_page_malloc_zero(heap, page, size, zero, usable);
mi_assert_internal(p != NULL);
}
// move singleton pages to the full queue
if (page->reserved == page->used) {
mi_page_to_full(page, mi_page_queue_of(page));
}
return p;
}

249
mimalloc/mimalloc-sys/mimalloc/src/prim/emscripten/prim.c Normal file
View File

@@ -0,0 +1,249 @@
/* ----------------------------------------------------------------------------
Copyright (c) 2018-2025, Microsoft Research, Daan Leijen, Alon Zakai
This is free software; you can redistribute it and/or modify it under the
terms of the MIT license. A copy of the license can be found in the file
"LICENSE" at the root of this distribution.
-----------------------------------------------------------------------------*/
// This file is included in `src/prim/prim.c`
#include "mimalloc.h"
#include "mimalloc/internal.h"
#include "mimalloc/atomic.h"
#include "mimalloc/prim.h"
// Design
// ======
//
// mimalloc is built on top of emmalloc. emmalloc is a minimal allocator on top
// of sbrk. The reason for having three layers here is that we want mimalloc to
// be able to allocate and release system memory properly, the same way it would
// when using VirtualAlloc on Windows or mmap on POSIX, and sbrk is too limited.
// Specifically, sbrk can only go up and down, and not "skip" over regions, and
// so we end up either never freeing memory to the system, or we can get stuck
// with holes.
//
// Atm wasm generally does *not* free memory back the system: once grown, we do
// not shrink back down (https://github.com/WebAssembly/design/issues/1397).
// However, that is expected to improve
// (https://github.com/WebAssembly/memory-control/blob/main/proposals/memory-control/Overview.md)
// and so we do not want to bake those limitations in here.
//
// Even without that issue, we want our system allocator to handle holes, that
// is, it should merge freed regions and allow allocating new content there of
// the full size, etc., so that we do not waste space. That means that the
// system allocator really does need to handle the general problem of allocating
// and freeing variable-sized chunks of memory in a random order, like malloc/
// free do. And so it makes sense to layer mimalloc on top of such an
// implementation.
//
// emmalloc makes sense for the lower level because it is small and simple while
// still fully handling merging of holes etc. It is not the most efficient
// allocator, but our assumption is that mimalloc needs to be fast while the
// system allocator underneath it is called much less frequently.
//
//---------------------------------------------
// init
//---------------------------------------------
void _mi_prim_mem_init( mi_os_mem_config_t* config) {
config->page_size = 64*MI_KiB; // WebAssembly has a fixed page size: 64KiB
config->alloc_granularity = 16;
config->has_overcommit = false;
config->has_partial_free = false;
config->has_virtual_reserve = false;
}
extern void emmalloc_free(void*);
int _mi_prim_free(void* addr, size_t size) {
if (size==0) return 0;
emmalloc_free(addr);
return 0;
}
//---------------------------------------------
// Allocation
//---------------------------------------------
extern void* emmalloc_memalign(size_t alignment, size_t size);
// Note: the `try_alignment` is just a hint and the returned pointer is not guaranteed to be aligned.
int _mi_prim_alloc(void* hint_addr, size_t size, size_t try_alignment, bool commit, bool allow_large, bool* is_large, bool* is_zero, void** addr) {
MI_UNUSED(try_alignment); MI_UNUSED(allow_large); MI_UNUSED(commit); MI_UNUSED(hint_addr);
*is_large = false;
// TODO: Track the highest address ever seen; first uses of it are zeroes.
// That assumes no one else uses sbrk but us (they could go up,
// scribble, and then down), but we could assert on that perhaps.
*is_zero = false;
// emmalloc has a minimum alignment size.
#define MIN_EMMALLOC_ALIGN 8
if (try_alignment < MIN_EMMALLOC_ALIGN) {
try_alignment = MIN_EMMALLOC_ALIGN;
}
void* p = emmalloc_memalign(try_alignment, size);
*addr = p;
if (p == 0) {
return ENOMEM;
}
return 0;
}
//---------------------------------------------
// Commit/Reset
//---------------------------------------------
int _mi_prim_commit(void* addr, size_t size, bool* is_zero) {
MI_UNUSED(addr); MI_UNUSED(size);
// See TODO above.
*is_zero = false;
return 0;
}
int _mi_prim_decommit(void* addr, size_t size, bool* needs_recommit) {
MI_UNUSED(addr); MI_UNUSED(size);
*needs_recommit = false;
return 0;
}
int _mi_prim_reset(void* addr, size_t size) {
MI_UNUSED(addr); MI_UNUSED(size);
return 0;
}
int _mi_prim_reuse(void* addr, size_t size) {
MI_UNUSED(addr); MI_UNUSED(size);
return 0;
}
int _mi_prim_protect(void* addr, size_t size, bool protect) {
MI_UNUSED(addr); MI_UNUSED(size); MI_UNUSED(protect);
return 0;
}
//---------------------------------------------
// Huge pages and NUMA nodes
//---------------------------------------------
int _mi_prim_alloc_huge_os_pages(void* hint_addr, size_t size, int numa_node, bool* is_zero, void** addr) {
MI_UNUSED(hint_addr); MI_UNUSED(size); MI_UNUSED(numa_node);
*is_zero = true;
*addr = NULL;
return ENOSYS;
}
size_t _mi_prim_numa_node(void) {
return 0;
}
size_t _mi_prim_numa_node_count(void) {
return 1;
}
//----------------------------------------------------------------
// Clock
//----------------------------------------------------------------
#include <emscripten/html5.h>
mi_msecs_t _mi_prim_clock_now(void) {
return emscripten_date_now();
}
//----------------------------------------------------------------
// Process info
//----------------------------------------------------------------
void _mi_prim_process_info(mi_process_info_t* pinfo)
{
// use defaults
MI_UNUSED(pinfo);
}
//----------------------------------------------------------------
// Output
//----------------------------------------------------------------
#include <emscripten/console.h>
void _mi_prim_out_stderr( const char* msg) {
emscripten_console_error(msg);
}
//----------------------------------------------------------------
// Environment
//----------------------------------------------------------------
bool _mi_prim_getenv(const char* name, char* result, size_t result_size) {
// For code size reasons, do not support environ customization for now.
MI_UNUSED(name);
MI_UNUSED(result);
MI_UNUSED(result_size);
return false;
}
//----------------------------------------------------------------
// Random
//----------------------------------------------------------------
bool _mi_prim_random_buf(void* buf, size_t buf_len) {
int err = getentropy(buf, buf_len);
return !err;
}
//----------------------------------------------------------------
// Thread init/done
//----------------------------------------------------------------
#if defined(MI_USE_PTHREADS)
// use pthread local storage keys to detect thread ending
// (and used with MI_TLS_PTHREADS for the default heap)
pthread_key_t _mi_heap_default_key = (pthread_key_t)(-1);
static void mi_pthread_done(void* value) {
if (value!=NULL) {
_mi_thread_done((mi_heap_t*)value);
}
}
void _mi_prim_thread_init_auto_done(void) {
mi_assert_internal(_mi_heap_default_key == (pthread_key_t)(-1));
pthread_key_create(&_mi_heap_default_key, &mi_pthread_done);
}
void _mi_prim_thread_done_auto_done(void) {
// nothing to do
}
void _mi_prim_thread_associate_default_heap(mi_heap_t* heap) {
if (_mi_heap_default_key != (pthread_key_t)(-1)) { // can happen during recursive invocation on freeBSD
pthread_setspecific(_mi_heap_default_key, heap);
}
}
#else
void _mi_prim_thread_init_auto_done(void) {
// nothing
}
void _mi_prim_thread_done_auto_done(void) {
// nothing
}
void _mi_prim_thread_associate_default_heap(mi_heap_t* heap) {
MI_UNUSED(heap);
}
#endif

9
mimalloc/mimalloc-sys/mimalloc/src/prim/osx/alloc-override-zone.c
View File

@@ -83,7 +83,7 @@ static void zone_destroy(malloc_zone_t* zone) {
}
static unsigned zone_batch_malloc(malloc_zone_t* zone, size_t size, void** ps, unsigned count) {
size_t i;
unsigned i;
for (i = 0; i < count; i++) {
ps[i] = zone_malloc(zone, size);
if (ps[i] == NULL) break;
@@ -225,7 +225,9 @@ static malloc_zone_t mi_malloc_zone = {
// switch to version 9+ on OSX 10.6 to support memalign.
.memalign = &zone_memalign,
.free_definite_size = &zone_free_definite_size,
#if defined(MAC_OS_X_VERSION_10_7) && (MAC_OS_X_VERSION_MAX_ALLOWED >= MAC_OS_X_VERSION_10_7)
.pressure_relief = &zone_pressure_relief,
#endif
#if defined(MAC_OS_X_VERSION_10_14) && (MAC_OS_X_VERSION_MAX_ALLOWED >= MAC_OS_X_VERSION_10_14)
.claimed_address = &zone_claimed_address,
#endif
@@ -416,10 +418,11 @@ static inline malloc_zone_t* mi_get_default_zone(void)
}
#if defined(__clang__)
__attribute__((constructor(0)))
__attribute__((constructor(101))) // highest priority
#else
__attribute__((constructor)) // seems not supported by g++-11 on the M1
__attribute__((constructor)) // priority level is not supported by gcc
#endif
__attribute__((used))
static void _mi_macos_override_malloc(void) {
malloc_zone_t* purgeable_zone = NULL;

52
mimalloc/mimalloc-sys/mimalloc/src/prim/prim.c
View File

@@ -18,7 +18,59 @@ terms of the MIT license. A copy of the license can be found in the file
#define MI_USE_SBRK
#include "wasi/prim.c" // memory-grow or sbrk (Wasm)
#elif defined(__EMSCRIPTEN__)
#include "emscripten/prim.c" // emmalloc_*, + pthread support
#else
#include "unix/prim.c" // mmap() (Linux, macOSX, BSD, Illumnos, Haiku, DragonFly, etc.)
#endif
// Generic process initialization
#ifndef MI_PRIM_HAS_PROCESS_ATTACH
#if defined(__GNUC__) || defined(__clang__)
// gcc,clang: use the constructor/destructor attribute
// which for both seem to run before regular constructors/destructors
#if defined(__clang__)
#define mi_attr_constructor __attribute__((constructor(101)))
#define mi_attr_destructor __attribute__((destructor(101)))
#else
#define mi_attr_constructor __attribute__((constructor))
#define mi_attr_destructor __attribute__((destructor))
#endif
static void mi_attr_constructor mi_process_attach(void) {
_mi_auto_process_init();
}
static void mi_attr_destructor mi_process_detach(void) {
_mi_auto_process_done();
}
#elif defined(__cplusplus)
// C++: use static initialization to detect process start/end
// This is not guaranteed to be first/last but the best we can generally do?
struct mi_init_done_t {
mi_init_done_t() {
_mi_auto_process_init();
}
~mi_init_done_t() {
_mi_auto_process_done();
}
};
static mi_init_done_t mi_init_done;
#else
#pragma message("define a way to call _mi_auto_process_init/done on your platform")
#endif
#endif
// Generic allocator init/done callback
#ifndef MI_PRIM_HAS_ALLOCATOR_INIT
bool _mi_is_redirected(void) {
return false;
}
bool _mi_allocator_init(const char** message) {
if (message != NULL) { *message = NULL; }
return true;
}
void _mi_allocator_done(void) {
// nothing to do
}
#endif

347
mimalloc/mimalloc-sys/mimalloc/src/prim/unix/prim.c
View File

@@ -1,5 +1,5 @@
/* ----------------------------------------------------------------------------
Copyright (c) 2018-2023, Microsoft Research, Daan Leijen
Copyright (c) 2018-2025, Microsoft Research, Daan Leijen
This is free software; you can redistribute it and/or modify it under the
terms of the MIT license. A copy of the license can be found in the file
"LICENSE" at the root of this distribution.
@@ -22,25 +22,35 @@ terms of the MIT license. A copy of the license can be found in the file
#include "mimalloc.h"
#include "mimalloc/internal.h"
#include "mimalloc/atomic.h"
#include "mimalloc/prim.h"
#include <sys/mman.h> // mmap
#include <unistd.h> // sysconf
#include <fcntl.h> // open, close, read, access
#include <stdlib.h> // getenv, arc4random_buf
#if defined(__linux__)
#include <features.h>
#include <fcntl.h>
#include <sys/prctl.h> // THP disable, PR_SET_VMA
#include <sys/sysinfo.h> // sysinfo
#if defined(__GLIBC__) && !defined(PR_SET_VMA)
#include <linux/prctl.h>
#endif
#if defined(__GLIBC__)
#include <linux/mman.h> // linux mmap flags
#include <linux/mman.h> // linux mmap flags
#else
#include <sys/mman.h>
#endif
#elif defined(__APPLE__)
#include <AvailabilityMacros.h>
#include <TargetConditionals.h>
#if !TARGET_IOS_IPHONE && !TARGET_IOS_SIMULATOR
#include <mach/vm_statistics.h>
#if !defined(TARGET_OS_OSX) || TARGET_OS_OSX // see issue #879, used to be (!TARGET_IOS_IPHONE && !TARGET_IOS_SIMULATOR)
#include <mach/vm_statistics.h> // VM_MAKE_TAG, VM_FLAGS_SUPERPAGE_SIZE_2MB, etc.
#endif
#if !defined(MAC_OS_X_VERSION_10_7)
#define MAC_OS_X_VERSION_10_7 1070
#endif
#include <sys/sysctl.h>
#elif defined(__FreeBSD__) || defined(__DragonFly__)
#include <sys/param.h>
#if __FreeBSD_version >= 1200000
@@ -50,44 +60,54 @@ terms of the MIT license. A copy of the license can be found in the file
#include <sys/sysctl.h>
#endif
#if !defined(__HAIKU__) && !defined(__APPLE__) && !defined(__CYGWIN__)
#if (defined(__linux__) && !defined(__ANDROID__)) || defined(__FreeBSD__)
#define MI_HAS_SYSCALL_H
#include <sys/syscall.h>
#endif
#if !defined(MADV_DONTNEED) && defined(POSIX_MADV_DONTNEED) // QNX
#define MADV_DONTNEED POSIX_MADV_DONTNEED
#endif
#if !defined(MADV_FREE) && defined(POSIX_MADV_FREE) // QNX
#define MADV_FREE POSIX_MADV_FREE
#endif
#define MI_UNIX_LARGE_PAGE_SIZE (2*MI_MiB) // TODO: can we query the OS for this?
//------------------------------------------------------------------------------------
// Use syscalls for some primitives to allow for libraries that override open/read/close etc.
// and do allocation themselves; using syscalls prevents recursion when mimalloc is
// and do allocation themselves; using syscalls prevents recursion when mimalloc is
// still initializing (issue #713)
// Declare inline to avoid unused function warnings.
//------------------------------------------------------------------------------------
#if defined(MI_HAS_SYSCALL_H) && defined(SYS_open) && defined(SYS_close) && defined(SYS_read) && defined(SYS_access)
static int mi_prim_open(const char* fpath, int open_flags) {
static inline int mi_prim_open(const char* fpath, int open_flags) {
return syscall(SYS_open,fpath,open_flags,0);
}
static ssize_t mi_prim_read(int fd, void* buf, size_t bufsize) {
static inline ssize_t mi_prim_read(int fd, void* buf, size_t bufsize) {
return syscall(SYS_read,fd,buf,bufsize);
}
static int mi_prim_close(int fd) {
static inline int mi_prim_close(int fd) {
return syscall(SYS_close,fd);
}
static int mi_prim_access(const char *fpath, int mode) {
static inline int mi_prim_access(const char *fpath, int mode) {
return syscall(SYS_access,fpath,mode);
}
#elif !defined(__APPLE__) // avoid unused warnings
#else
static int mi_prim_open(const char* fpath, int open_flags) {
static inline int mi_prim_open(const char* fpath, int open_flags) {
return open(fpath,open_flags);
}
static ssize_t mi_prim_read(int fd, void* buf, size_t bufsize) {
static inline ssize_t mi_prim_read(int fd, void* buf, size_t bufsize) {
return read(fd,buf,bufsize);
}
static int mi_prim_close(int fd) {
static inline int mi_prim_close(int fd) {
return close(fd);
}
static int mi_prim_access(const char *fpath, int mode) {
static inline int mi_prim_access(const char *fpath, int mode) {
return access(fpath,mode);
}
@@ -101,40 +121,94 @@ static int mi_prim_access(const char *fpath, int mode) {
static bool unix_detect_overcommit(void) {
bool os_overcommit = true;
#if defined(__linux__)
int fd = mi_prim_open("/proc/sys/vm/overcommit_memory", O_RDONLY);
if (fd >= 0) {
char buf[32];
ssize_t nread = mi_prim_read(fd, &buf, sizeof(buf));
mi_prim_close(fd);
// <https://www.kernel.org/doc/Documentation/vm/overcommit-accounting>
// 0: heuristic overcommit, 1: always overcommit, 2: never overcommit (ignore NORESERVE)
if (nread >= 1) {
os_overcommit = (buf[0] == '0' || buf[0] == '1');
#if defined(__linux__)
int fd = mi_prim_open("/proc/sys/vm/overcommit_memory", O_RDONLY);
if (fd >= 0) {
char buf[32];
ssize_t nread = mi_prim_read(fd, &buf, sizeof(buf));
mi_prim_close(fd);
// <https://www.kernel.org/doc/Documentation/vm/overcommit-accounting>
// 0: heuristic overcommit, 1: always overcommit, 2: never overcommit (ignore NORESERVE)
if (nread >= 1) {
os_overcommit = (buf[0] == '0' || buf[0] == '1');
}
}
}
#elif defined(__FreeBSD__)
int val = 0;
size_t olen = sizeof(val);
if (sysctlbyname("vm.overcommit", &val, &olen, NULL, 0) == 0) {
os_overcommit = (val != 0);
}
#else
// default: overcommit is true
#endif
#elif defined(__FreeBSD__)
int val = 0;
size_t olen = sizeof(val);
if (sysctlbyname("vm.overcommit", &val, &olen, NULL, 0) == 0) {
os_overcommit = (val != 0);
}
#else
// default: overcommit is true
#endif
return os_overcommit;
}
void _mi_prim_mem_init( mi_os_mem_config_t* config ) {
// try to detect the physical memory dynamically (if possible)
static void unix_detect_physical_memory( size_t page_size, size_t* physical_memory_in_kib ) {
#if defined(CTL_HW) && (defined(HW_PHYSMEM64) || defined(HW_MEMSIZE)) // freeBSD, macOS
MI_UNUSED(page_size);
int64_t physical_memory = 0;
size_t length = sizeof(int64_t);
#if defined(HW_PHYSMEM64)
int mib[2] = { CTL_HW, HW_PHYSMEM64 };
#else
int mib[2] = { CTL_HW, HW_MEMSIZE };
#endif
const int err = sysctl(mib, 2, &physical_memory, &length, NULL, 0);
if (err==0 && physical_memory > 0) {
const int64_t phys_in_kib = physical_memory / MI_KiB;
if (phys_in_kib > 0 && (uint64_t)phys_in_kib <= SIZE_MAX) {
*physical_memory_in_kib = (size_t)phys_in_kib;
}
}
#elif defined(__linux__)
MI_UNUSED(page_size);
struct sysinfo info; _mi_memzero_var(info);
const int err = sysinfo(&info);
if (err==0 && info.totalram > 0 && info.totalram <= SIZE_MAX) {
*physical_memory_in_kib = (size_t)info.totalram / MI_KiB;
}
#elif defined(_SC_PHYS_PAGES) // do not use by default as it might cause allocation (by using `fopen` to parse /proc/meminfo) (issue #1100)
const long pphys = sysconf(_SC_PHYS_PAGES);
const size_t psize_in_kib = page_size / MI_KiB;
if (psize_in_kib > 0 && pphys > 0 && (unsigned long)pphys <= SIZE_MAX && (size_t)pphys <= (SIZE_MAX/psize_in_kib)) {
*physical_memory_in_kib = (size_t)pphys * psize_in_kib;
}
#endif
}
void _mi_prim_mem_init( mi_os_mem_config_t* config )
{
long psize = sysconf(_SC_PAGESIZE);
if (psize > 0) {
if (psize > 0 && (unsigned long)psize < SIZE_MAX) {
config->page_size = (size_t)psize;
config->alloc_granularity = (size_t)psize;
unix_detect_physical_memory(config->page_size, &config->physical_memory_in_kib);
}
config->large_page_size = 2*MI_MiB; // TODO: can we query the OS for this?
config->large_page_size = MI_UNIX_LARGE_PAGE_SIZE;
config->has_overcommit = unix_detect_overcommit();
config->must_free_whole = false; // mmap can free in parts
config->has_partial_free = true; // mmap can free in parts
config->has_virtual_reserve = true; // todo: check if this true for NetBSD? (for anonymous mmap with PROT_NONE)
// disable transparent huge pages for this process?
#if (defined(__linux__) || defined(__ANDROID__)) && defined(PR_GET_THP_DISABLE)
#if defined(MI_NO_THP)
if (true)
#else
if (!mi_option_is_enabled(mi_option_allow_thp)) // disable THP if requested through an option
#endif
{
int val = 0;
if (prctl(PR_GET_THP_DISABLE, &val, 0, 0, 0) != 0) {
// Most likely since distros often come with always/madvise settings.
val = 1;
// Disabling only for mimalloc process rather than touching system wide settings
(void)prctl(PR_SET_THP_DISABLE, &val, 0, 0, 0);
}
}
#endif
}
@@ -143,6 +217,7 @@ void _mi_prim_mem_init( mi_os_mem_config_t* config ) {
//---------------------------------------------
int _mi_prim_free(void* addr, size_t size ) {
if (size==0) return 0;
bool err = (munmap(addr, size) == -1);
return (err ? errno : 0);
}
@@ -154,31 +229,44 @@ int _mi_prim_free(void* addr, size_t size ) {
static int unix_madvise(void* addr, size_t size, int advice) {
#if defined(__sun)
return madvise((caddr_t)addr, size, advice); // Solaris needs cast (issue #520)
int res = madvise((caddr_t)addr, size, advice); // Solaris needs cast (issue #520)
#elif defined(__QNX__)
int res = posix_madvise(addr, size, advice);
#else
return madvise(addr, size, advice);
int res = madvise(addr, size, advice);
#endif
return (res==0 ? 0 : errno);
}
static void* unix_mmap_prim(void* addr, size_t size, size_t try_alignment, int protect_flags, int flags, int fd) {
static void* unix_mmap_prim(void* addr, size_t size, int protect_flags, int flags, int fd) {
void* p = mmap(addr, size, protect_flags, flags, fd, 0 /* offset */);
#if defined(__linux__) && defined(PR_SET_VMA)
if (p!=MAP_FAILED && p!=NULL) {
prctl(PR_SET_VMA, PR_SET_VMA_ANON_NAME, p, size, "mimalloc");
}
#endif
return p;
}
static void* unix_mmap_prim_aligned(void* addr, size_t size, size_t try_alignment, int protect_flags, int flags, int fd) {
MI_UNUSED(try_alignment);
void* p = NULL;
#if defined(MAP_ALIGNED) // BSD
if (addr == NULL && try_alignment > 1 && (try_alignment % _mi_os_page_size()) == 0) {
size_t n = mi_bsr(try_alignment);
if (((size_t)1 << n) == try_alignment && n >= 12 && n <= 30) { // alignment is a power of 2 and 4096 <= alignment <= 1GiB
p = mmap(addr, size, protect_flags, flags | MAP_ALIGNED(n), fd, 0);
if (p==MAP_FAILED || !_mi_is_aligned(p,try_alignment)) {
p = unix_mmap_prim(addr, size, protect_flags, flags | MAP_ALIGNED(n), fd);
if (p==MAP_FAILED || !_mi_is_aligned(p,try_alignment)) {
int err = errno;
_mi_warning_message("unable to directly request aligned OS memory (error: %d (0x%x), size: 0x%zx bytes, alignment: 0x%zx, hint address: %p)\n", err, err, size, try_alignment, addr);
_mi_trace_message("unable to directly request aligned OS memory (error: %d (0x%x), size: 0x%zx bytes, alignment: 0x%zx, hint address: %p)\n", err, err, size, try_alignment, addr);
}
if (p!=MAP_FAILED) return p;
// fall back to regular mmap
// fall back to regular mmap
}
}
#elif defined(MAP_ALIGN) // Solaris
if (addr == NULL && try_alignment > 1 && (try_alignment % _mi_os_page_size()) == 0) {
p = mmap((void*)try_alignment, size, protect_flags, flags | MAP_ALIGN, fd, 0); // addr parameter is the required alignment
p = unix_mmap_prim((void*)try_alignment, size, protect_flags, flags | MAP_ALIGN, fd); // addr parameter is the required alignment
if (p!=MAP_FAILED) return p;
// fall back to regular mmap
}
@@ -188,22 +276,22 @@ static void* unix_mmap_prim(void* addr, size_t size, size_t try_alignment, int p
if (addr == NULL) {
void* hint = _mi_os_get_aligned_hint(try_alignment, size);
if (hint != NULL) {
p = mmap(hint, size, protect_flags, flags, fd, 0);
if (p==MAP_FAILED || !_mi_is_aligned(p,try_alignment)) {
p = unix_mmap_prim(hint, size, protect_flags, flags, fd);
if (p==MAP_FAILED || !_mi_is_aligned(p,try_alignment)) {
#if MI_TRACK_ENABLED // asan sometimes does not instrument errno correctly?
int err = 0;
#else
int err = errno;
#endif
_mi_warning_message("unable to directly request hinted aligned OS memory (error: %d (0x%x), size: 0x%zx bytes, alignment: 0x%zx, hint address: %p)\n", err, err, size, try_alignment, hint);
_mi_trace_message("unable to directly request hinted aligned OS memory (error: %d (0x%x), size: 0x%zx bytes, alignment: 0x%zx, hint address: %p)\n", err, err, size, try_alignment, hint);
}
if (p!=MAP_FAILED) return p;
// fall back to regular mmap
// fall back to regular mmap
}
}
#endif
// regular mmap
p = mmap(addr, size, protect_flags, flags, fd, 0);
p = unix_mmap_prim(addr, size, protect_flags, flags, fd);
if (p!=MAP_FAILED) return p;
// failed to allocate
return NULL;
@@ -213,7 +301,7 @@ static int unix_mmap_fd(void) {
#if defined(VM_MAKE_TAG)
// macOS: tracking anonymous page with a specific ID. (All up to 98 are taken officially but LLVM sanitizers had taken 99)
int os_tag = (int)mi_option_get(mi_option_os_tag);
if (os_tag < 100 || os_tag > 255) { os_tag = 100; }
if (os_tag < 100 || os_tag > 255) { os_tag = 254; }
return VM_MAKE_TAG(os_tag);
#else
return -1;
@@ -237,7 +325,7 @@ static void* unix_mmap(void* addr, size_t size, size_t try_alignment, int protec
protect_flags |= PROT_MAX(PROT_READ | PROT_WRITE); // BSD
#endif
// huge page allocation
if ((large_only || _mi_os_use_large_page(size, try_alignment)) && allow_large) {
if (allow_large && (large_only || (_mi_os_canuse_large_page(size, try_alignment) && mi_option_is_enabled(mi_option_allow_large_os_pages)))) {
static _Atomic(size_t) large_page_try_ok; // = 0;
size_t try_ok = mi_atomic_load_acquire(&large_page_try_ok);
if (!large_only && try_ok > 0) {
@@ -258,7 +346,7 @@ static void* unix_mmap(void* addr, size_t size, size_t try_alignment, int protec
#endif
#ifdef MAP_HUGE_1GB
static bool mi_huge_pages_available = true;
if ((size % MI_GiB) == 0 && mi_huge_pages_available) {
if (large_only && (size % MI_GiB) == 0 && mi_huge_pages_available) {
lflags |= MAP_HUGE_1GB;
}
else
@@ -274,13 +362,15 @@ static void* unix_mmap(void* addr, size_t size, size_t try_alignment, int protec
if (large_only || lflags != flags) {
// try large OS page allocation
*is_large = true;
p = unix_mmap_prim(addr, size, try_alignment, protect_flags, lflags, lfd);
p = unix_mmap_prim_aligned(addr, size, try_alignment, protect_flags, lflags, lfd);
#ifdef MAP_HUGE_1GB
if (p == NULL && (lflags & MAP_HUGE_1GB) != 0) {
if (p == NULL && (lflags & MAP_HUGE_1GB) == MAP_HUGE_1GB) {
mi_huge_pages_available = false; // don't try huge 1GiB pages again
_mi_warning_message("unable to allocate huge (1GiB) page, trying large (2MiB) pages instead (errno: %i)\n", errno);
if (large_only) {
_mi_warning_message("unable to allocate huge (1GiB) page, trying large (2MiB) pages instead (errno: %i)\n", errno);
}
lflags = ((lflags & ~MAP_HUGE_1GB) | MAP_HUGE_2MB);
p = unix_mmap_prim(addr, size, try_alignment, protect_flags, lflags, lfd);
p = unix_mmap_prim_aligned(addr, size, try_alignment, protect_flags, lflags, lfd);
}
#endif
if (large_only) return p;
@@ -293,8 +383,9 @@ static void* unix_mmap(void* addr, size_t size, size_t try_alignment, int protec
// regular allocation
if (p == NULL) {
*is_large = false;
p = unix_mmap_prim(addr, size, try_alignment, protect_flags, flags, fd);
if (p != NULL) {
p = unix_mmap_prim_aligned(addr, size, try_alignment, protect_flags, flags, fd);
#if !defined(MI_NO_THP)
if (p != NULL && allow_large && mi_option_is_enabled(mi_option_allow_thp) && _mi_os_canuse_large_page(size, try_alignment)) {
#if defined(MADV_HUGEPAGE)
// Many Linux systems don't allow MAP_HUGETLB but they support instead
// transparent huge pages (THP). Generally, it is not required to call `madvise` with MADV_HUGE
@@ -302,35 +393,35 @@ static void* unix_mmap(void* addr, size_t size, size_t try_alignment, int protec
// in that case -- in particular for our large regions (in `memory.c`).
// However, some systems only allow THP if called with explicit `madvise`, so
// when large OS pages are enabled for mimalloc, we call `madvise` anyways.
if (allow_large && _mi_os_use_large_page(size, try_alignment)) {
if (unix_madvise(p, size, MADV_HUGEPAGE) == 0) {
*is_large = true; // possibly
};
}
if (unix_madvise(p, size, MADV_HUGEPAGE) == 0) {
// *is_large = true; // possibly
};
#elif defined(__sun)
if (allow_large && _mi_os_use_large_page(size, try_alignment)) {
struct memcntl_mha cmd = {0};
cmd.mha_pagesize = large_os_page_size;
cmd.mha_cmd = MHA_MAPSIZE_VA;
if (memcntl((caddr_t)p, size, MC_HAT_ADVISE, (caddr_t)&cmd, 0, 0) == 0) {
*is_large = true;
}
struct memcntl_mha cmd = {0};
cmd.mha_pagesize = _mi_os_large_page_size();
cmd.mha_cmd = MHA_MAPSIZE_VA;
if (memcntl((caddr_t)p, size, MC_HAT_ADVISE, (caddr_t)&cmd, 0, 0) == 0) {
// *is_large = true; // possibly
}
#endif
}
#endif
}
return p;
}
// Note: the `try_alignment` is just a hint and the returned pointer is not guaranteed to be aligned.
int _mi_prim_alloc(size_t size, size_t try_alignment, bool commit, bool allow_large, bool* is_large, bool* is_zero, void** addr) {
int _mi_prim_alloc(void* hint_addr, size_t size, size_t try_alignment, bool commit, bool allow_large, bool* is_large, bool* is_zero, void** addr) {
mi_assert_internal(size > 0 && (size % _mi_os_page_size()) == 0);
mi_assert_internal(commit || !allow_large);
mi_assert_internal(try_alignment > 0);
if (hint_addr == NULL && size >= 8*MI_UNIX_LARGE_PAGE_SIZE && try_alignment > 1 && _mi_is_power_of_two(try_alignment) && try_alignment < MI_UNIX_LARGE_PAGE_SIZE) {
try_alignment = MI_UNIX_LARGE_PAGE_SIZE; // try to align along large page size for larger allocations
}
*is_zero = true;
int protect_flags = (commit ? (PROT_WRITE | PROT_READ) : PROT_NONE);
*addr = unix_mmap(NULL, size, try_alignment, protect_flags, false, allow_large, is_large);
int protect_flags = (commit ? (PROT_WRITE | PROT_READ) : PROT_NONE);
*addr = unix_mmap(hint_addr, size, try_alignment, protect_flags, false, allow_large, is_large);
return (*addr != NULL ? 0 : errno);
}
@@ -357,20 +448,34 @@ int _mi_prim_commit(void* start, size_t size, bool* is_zero) {
// was either from mmap PROT_NONE, or from decommit MADV_DONTNEED, but
// we sometimes call commit on a range with still partially committed
// memory and `mprotect` does not zero the range.
*is_zero = false;
*is_zero = false;
int err = mprotect(start, size, (PROT_READ | PROT_WRITE));
if (err != 0) {
err = errno;
if (err != 0) {
err = errno;
unix_mprotect_hint(err);
}
return err;
}
int _mi_prim_reuse(void* start, size_t size) {
MI_UNUSED(start); MI_UNUSED(size);
#if defined(__APPLE__) && defined(MADV_FREE_REUSE)
return unix_madvise(start, size, MADV_FREE_REUSE);
#endif
return 0;
}
int _mi_prim_decommit(void* start, size_t size, bool* needs_recommit) {
int err = 0;
// decommit: use MADV_DONTNEED as it decreases rss immediately (unlike MADV_FREE)
err = unix_madvise(start, size, MADV_DONTNEED);
#if !MI_DEBUG && !MI_SECURE
int err = 0;
#if defined(__APPLE__) && defined(MADV_FREE_REUSABLE)
// decommit on macOS: use MADV_FREE_REUSABLE as it does immediate rss accounting (issue #1097)
err = unix_madvise(start, size, MADV_FREE_REUSABLE);
if (err) { err = unix_madvise(start, size, MADV_DONTNEED); }
#else
// decommit: use MADV_DONTNEED as it decreases rss immediately (unlike MADV_FREE)
err = unix_madvise(start, size, MADV_DONTNEED);
#endif
#if !MI_DEBUG && MI_SECURE<=2
*needs_recommit = false;
#else
*needs_recommit = true;
@@ -381,20 +486,28 @@ int _mi_prim_decommit(void* start, size_t size, bool* needs_recommit) {
*needs_recommit = true;
const int fd = unix_mmap_fd();
void* p = mmap(start, size, PROT_NONE, (MAP_FIXED | MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE), fd, 0);
if (p != start) { err = errno; }
if (p != start) { err = errno; }
*/
return err;
}
int _mi_prim_reset(void* start, size_t size) {
// We try to use `MADV_FREE` as that is the fastest. A drawback though is that it
// will not reduce the `rss` stats in tools like `top` even though the memory is available
// to other processes. With the default `MIMALLOC_PURGE_DECOMMITS=1` we ensure that by
// default `MADV_DONTNEED` is used though.
int err = 0;
// on macOS can use MADV_FREE_REUSABLE (but we disable this for now as it seems slower)
#if 0 && defined(__APPLE__) && defined(MADV_FREE_REUSABLE)
err = unix_madvise(start, size, MADV_FREE_REUSABLE);
if (err==0) return 0;
// fall through
#endif
#if defined(MADV_FREE)
// Otherwise, we try to use `MADV_FREE` as that is the fastest. A drawback though is that it
// will not reduce the `rss` stats in tools like `top` even though the memory is available
// to other processes. With the default `MIMALLOC_PURGE_DECOMMITS=1` we ensure that by
// default `MADV_DONTNEED` is used though.
static _Atomic(size_t) advice = MI_ATOMIC_VAR_INIT(MADV_FREE);
int oadvice = (int)mi_atomic_load_relaxed(&advice);
int err;
while ((err = unix_madvise(start, size, oadvice)) != 0 && errno == EAGAIN) { errno = 0; };
if (err != 0 && errno == EINVAL && oadvice == MADV_FREE) {
// if MADV_FREE is not supported, fall back to MADV_DONTNEED from now on
@@ -402,14 +515,14 @@ int _mi_prim_reset(void* start, size_t size) {
err = unix_madvise(start, size, MADV_DONTNEED);
}
#else
int err = unix_madvise(start, size, MADV_DONTNEED);
err = unix_madvise(start, size, MADV_DONTNEED);
#endif
return err;
}
int _mi_prim_protect(void* start, size_t size, bool protect) {
int err = mprotect(start, size, protect ? PROT_NONE : (PROT_READ | PROT_WRITE));
if (err != 0) { err = errno; }
if (err != 0) { err = errno; }
unix_mprotect_hint(err);
return err;
}
@@ -449,8 +562,8 @@ int _mi_prim_alloc_huge_os_pages(void* hint_addr, size_t size, int numa_node, bo
long err = mi_prim_mbind(*addr, size, MPOL_PREFERRED, &numa_mask, 8*MI_INTPTR_SIZE, 0);
if (err != 0) {
err = errno;
_mi_warning_message("failed to bind huge (1GiB) pages to numa node %d (error: %ld (0x%lx))\n", numa_node, err, err);
}
_mi_warning_message("failed to bind huge (1GiB) pages to numa node %d (error: %d (0x%x))\n", numa_node, err, err);
}
}
return (*addr != NULL ? 0 : errno);
}
@@ -472,8 +585,6 @@ int _mi_prim_alloc_huge_os_pages(void* hint_addr, size_t size, int numa_node, bo
#if defined(__linux__)
#include <stdio.h> // snprintf
size_t _mi_prim_numa_node(void) {
#if defined(MI_HAS_SYSCALL_H) && defined(SYS_getcpu)
unsigned long node = 0;
@@ -491,7 +602,7 @@ size_t _mi_prim_numa_node_count(void) {
unsigned node = 0;
for(node = 0; node < 256; node++) {
// enumerate node entries -- todo: it there a more efficient way to do this? (but ensure there is no allocation)
snprintf(buf, sizeof(buf), "/sys/devices/system/node/node%u", node + 1);
_mi_snprintf(buf, 127, "/sys/devices/system/node/node%u", node + 1);
if (mi_prim_access(buf,R_OK) != 0) break;
}
return (node+1);
@@ -567,9 +678,9 @@ mi_msecs_t _mi_prim_clock_now(void) {
// low resolution timer
mi_msecs_t _mi_prim_clock_now(void) {
#if !defined(CLOCKS_PER_SEC) || (CLOCKS_PER_SEC == 1000) || (CLOCKS_PER_SEC == 0)
return (mi_msecs_t)clock();
return (mi_msecs_t)clock();
#elif (CLOCKS_PER_SEC < 1000)
return (mi_msecs_t)clock() * (1000 / (mi_msecs_t)CLOCKS_PER_SEC);
return (mi_msecs_t)clock() * (1000 / (mi_msecs_t)CLOCKS_PER_SEC);
#else
return (mi_msecs_t)clock() / ((mi_msecs_t)CLOCKS_PER_SEC / 1000);
#endif
@@ -609,7 +720,7 @@ void _mi_prim_process_info(mi_process_info_t* pinfo)
pinfo->stime = timeval_secs(&rusage.ru_stime);
#if !defined(__HAIKU__)
pinfo->page_faults = rusage.ru_majflt;
#endif
#endif
#if defined(__HAIKU__)
// Haiku does not have (yet?) a way to
// get these stats per process
@@ -729,40 +840,30 @@ bool _mi_prim_getenv(const char* name, char* result, size_t result_size) {
// Random
//----------------------------------------------------------------
#if defined(__APPLE__)
#include <AvailabilityMacros.h>
#if defined(MAC_OS_X_VERSION_10_10) && MAC_OS_X_VERSION_MAX_ALLOWED >= MAC_OS_X_VERSION_10_10
#if defined(__APPLE__) && defined(MAC_OS_X_VERSION_10_15) && (MAC_OS_X_VERSION_MIN_REQUIRED >= MAC_OS_X_VERSION_10_15)
#include <CommonCrypto/CommonCryptoError.h>
#include <CommonCrypto/CommonRandom.h>
#endif
bool _mi_prim_random_buf(void* buf, size_t buf_len) {
#if defined(MAC_OS_X_VERSION_10_15) && MAC_OS_X_VERSION_MAX_ALLOWED >= MAC_OS_X_VERSION_10_15
// We prefere CCRandomGenerateBytes as it returns an error code while arc4random_buf
// may fail silently on macOS. See PR #390, and <https://opensource.apple.com/source/Libc/Libc-1439.40.11/gen/FreeBSD/arc4random.c.auto.html>
return (CCRandomGenerateBytes(buf, buf_len) == kCCSuccess);
#else
// fall back on older macOS
arc4random_buf(buf, buf_len);
return true;
#endif
// We prefer CCRandomGenerateBytes as it returns an error code while arc4random_buf
// may fail silently on macOS. See PR #390, and <https://opensource.apple.com/source/Libc/Libc-1439.40.11/gen/FreeBSD/arc4random.c.auto.html>
return (CCRandomGenerateBytes(buf, buf_len) == kCCSuccess);
}
#elif defined(__ANDROID__) || defined(__DragonFly__) || \
defined(__FreeBSD__) || defined(__NetBSD__) || defined(__OpenBSD__) || \
defined(__sun)
defined(__sun) || \
(defined(__APPLE__) && (MAC_OS_X_VERSION_MIN_REQUIRED >= MAC_OS_X_VERSION_10_7))
#include <stdlib.h>
bool _mi_prim_random_buf(void* buf, size_t buf_len) {
arc4random_buf(buf, buf_len);
return true;
}
#elif defined(__linux__) || defined(__HAIKU__)
#elif defined(__APPLE__) || defined(__linux__) || defined(__HAIKU__) // also for old apple versions < 10.7 (issue #829)
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <errno.h>
bool _mi_prim_random_buf(void* buf, size_t buf_len) {
@@ -833,7 +934,9 @@ void _mi_prim_thread_init_auto_done(void) {
}
void _mi_prim_thread_done_auto_done(void) {
// nothing to do
if (_mi_heap_default_key != (pthread_key_t)(-1)) { // do not leak the key, see issue #809
pthread_key_delete(_mi_heap_default_key);
}
}
void _mi_prim_thread_associate_default_heap(mi_heap_t* heap) {
@@ -842,7 +945,7 @@ void _mi_prim_thread_associate_default_heap(mi_heap_t* heap) {
}
}
#else
#else
void _mi_prim_thread_init_auto_done(void) {
// nothing

25
mimalloc/mimalloc-sys/mimalloc/src/prim/wasi/prim.c
View File

@@ -9,9 +9,11 @@ terms of the MIT license. A copy of the license can be found in the file
#include "mimalloc.h"
#include "mimalloc/internal.h"
#include "mimalloc/atomic.h"
#include "mimalloc/prim.h"
#include <stdio.h> // fputs
#include <stdlib.h> // getenv
//---------------------------------------------
// Initialize
//---------------------------------------------
@@ -19,8 +21,8 @@ terms of the MIT license. A copy of the license can be found in the file
void _mi_prim_mem_init( mi_os_mem_config_t* config ) {
config->page_size = 64*MI_KiB; // WebAssembly has a fixed page size: 64KiB
config->alloc_granularity = 16;
config->has_overcommit = false;
config->must_free_whole = true;
config->has_overcommit = false;
config->has_partial_free = false;
config->has_virtual_reserve = false;
}
@@ -40,6 +42,8 @@ int _mi_prim_free(void* addr, size_t size ) {
//---------------------------------------------
#if defined(MI_USE_SBRK)
#include <unistd.h> // for sbrk
static void* mi_memory_grow( size_t size ) {
void* p = sbrk(size);
if (p == (void*)(-1)) return NULL;
@@ -115,8 +119,8 @@ static void* mi_prim_mem_grow(size_t size, size_t try_alignment) {
}
// Note: the `try_alignment` is just a hint and the returned pointer is not guaranteed to be aligned.
int _mi_prim_alloc(size_t size, size_t try_alignment, bool commit, bool allow_large, bool* is_large, bool* is_zero, void** addr) {
MI_UNUSED(allow_large); MI_UNUSED(commit);
int _mi_prim_alloc(void* hint_addr, size_t size, size_t try_alignment, bool commit, bool allow_large, bool* is_large, bool* is_zero, void** addr) {
MI_UNUSED(allow_large); MI_UNUSED(commit); MI_UNUSED(hint_addr);
*is_large = false;
*is_zero = false;
*addr = mi_prim_mem_grow(size, try_alignment);
@@ -129,7 +133,7 @@ int _mi_prim_alloc(size_t size, size_t try_alignment, bool commit, bool allow_la
//---------------------------------------------
int _mi_prim_commit(void* addr, size_t size, bool* is_zero) {
MI_UNUSED(addr); MI_UNUSED(size);
MI_UNUSED(addr); MI_UNUSED(size);
*is_zero = false;
return 0;
}
@@ -145,6 +149,11 @@ int _mi_prim_reset(void* addr, size_t size) {
return 0;
}
int _mi_prim_reuse(void* addr, size_t size) {
MI_UNUSED(addr); MI_UNUSED(size);
return 0;
}
int _mi_prim_protect(void* addr, size_t size, bool protect) {
MI_UNUSED(addr); MI_UNUSED(size); MI_UNUSED(protect);
return 0;
@@ -194,9 +203,9 @@ mi_msecs_t _mi_prim_clock_now(void) {
// low resolution timer
mi_msecs_t _mi_prim_clock_now(void) {
#if !defined(CLOCKS_PER_SEC) || (CLOCKS_PER_SEC == 1000) || (CLOCKS_PER_SEC == 0)
return (mi_msecs_t)clock();
return (mi_msecs_t)clock();
#elif (CLOCKS_PER_SEC < 1000)
return (mi_msecs_t)clock() * (1000 / (mi_msecs_t)CLOCKS_PER_SEC);
return (mi_msecs_t)clock() * (1000 / (mi_msecs_t)CLOCKS_PER_SEC);
#else
return (mi_msecs_t)clock() / ((mi_msecs_t)CLOCKS_PER_SEC / 1000);
#endif

393
mimalloc/mimalloc-sys/mimalloc/src/prim/windows/prim.c
View File

@@ -9,10 +9,13 @@ terms of the MIT license. A copy of the license can be found in the file
#include "mimalloc.h"
#include "mimalloc/internal.h"
#include "mimalloc/atomic.h"
#include "mimalloc/prim.h"
#include <stdio.h> // fputs, stderr
// xbox has no console IO
#if !defined(WINAPI_FAMILY_PARTITION) || WINAPI_FAMILY_PARTITION(WINAPI_PARTITION_APP | WINAPI_PARTITION_SYSTEM)
#define MI_HAS_CONSOLE_IO
#endif
//---------------------------------------------
// Dynamically bind Windows API points for portability
@@ -46,22 +49,33 @@ typedef struct MI_MEM_ADDRESS_REQUIREMENTS_S {
#define MI_MEM_EXTENDED_PARAMETER_NONPAGED_HUGE 0x00000010
#include <winternl.h>
typedef PVOID (__stdcall *PVirtualAlloc2)(HANDLE, PVOID, SIZE_T, ULONG, ULONG, MI_MEM_EXTENDED_PARAMETER*, ULONG);
typedef NTSTATUS (__stdcall *PNtAllocateVirtualMemoryEx)(HANDLE, PVOID*, SIZE_T*, ULONG, ULONG, MI_MEM_EXTENDED_PARAMETER*, ULONG);
typedef PVOID (__stdcall *PVirtualAlloc2)(HANDLE, PVOID, SIZE_T, ULONG, ULONG, MI_MEM_EXTENDED_PARAMETER*, ULONG);
typedef LONG (__stdcall *PNtAllocateVirtualMemoryEx)(HANDLE, PVOID*, SIZE_T*, ULONG, ULONG, MI_MEM_EXTENDED_PARAMETER*, ULONG); // avoid NTSTATUS as it is not defined on xbox (pr #1084)
static PVirtualAlloc2 pVirtualAlloc2 = NULL;
static PNtAllocateVirtualMemoryEx pNtAllocateVirtualMemoryEx = NULL;
// Similarly, GetNumaProcesorNodeEx is only supported since Windows 7
// Similarly, GetNumaProcessorNodeEx is only supported since Windows 7 (and GetNumaNodeProcessorMask is not supported on xbox)
typedef struct MI_PROCESSOR_NUMBER_S { WORD Group; BYTE Number; BYTE Reserved; } MI_PROCESSOR_NUMBER;
typedef VOID (__stdcall *PGetCurrentProcessorNumberEx)(MI_PROCESSOR_NUMBER* ProcNumber);
typedef BOOL (__stdcall *PGetNumaProcessorNodeEx)(MI_PROCESSOR_NUMBER* Processor, PUSHORT NodeNumber);
typedef BOOL (__stdcall* PGetNumaNodeProcessorMaskEx)(USHORT Node, PGROUP_AFFINITY ProcessorMask);
typedef BOOL (__stdcall *PGetNumaProcessorNode)(UCHAR Processor, PUCHAR NodeNumber);
typedef BOOL (__stdcall* PGetNumaNodeProcessorMask)(UCHAR Node, PULONGLONG ProcessorMask);
typedef BOOL (__stdcall* PGetNumaHighestNodeNumber)(PULONG Node);
static PGetCurrentProcessorNumberEx pGetCurrentProcessorNumberEx = NULL;
static PGetNumaProcessorNodeEx pGetNumaProcessorNodeEx = NULL;
static PGetNumaNodeProcessorMaskEx pGetNumaNodeProcessorMaskEx = NULL;
static PGetNumaProcessorNode pGetNumaProcessorNode = NULL;
static PGetNumaNodeProcessorMask pGetNumaNodeProcessorMask = NULL;
static PGetNumaHighestNodeNumber pGetNumaHighestNodeNumber = NULL;
// Not available on xbox
typedef SIZE_T(__stdcall* PGetLargePageMinimum)(VOID);
static PGetLargePageMinimum pGetLargePageMinimum = NULL;
// Available after Windows XP
typedef BOOL (__stdcall *PGetPhysicallyInstalledSystemMemory)( PULONGLONG TotalMemoryInKilobytes );
//---------------------------------------------
// Enable large page support dynamically (if possible)
@@ -72,6 +86,7 @@ static bool win_enable_large_os_pages(size_t* large_page_size)
static bool large_initialized = false;
if (large_initialized) return (_mi_os_large_page_size() > 0);
large_initialized = true;
if (pGetLargePageMinimum==NULL) return false; // no large page support (xbox etc.)
// Try to see if large OS pages are supported
// To use large pages on Windows, we first need access permission
@@ -90,8 +105,8 @@ static bool win_enable_large_os_pages(size_t* large_page_size)
if (ok) {
err = GetLastError();
ok = (err == ERROR_SUCCESS);
if (ok && large_page_size != NULL) {
*large_page_size = GetLargePageMinimum();
if (ok && large_page_size != NULL && pGetLargePageMinimum != NULL) {
*large_page_size = (*pGetLargePageMinimum)();
}
}
}
@@ -112,13 +127,19 @@ static bool win_enable_large_os_pages(size_t* large_page_size)
void _mi_prim_mem_init( mi_os_mem_config_t* config )
{
config->has_overcommit = false;
config->must_free_whole = true;
config->has_partial_free = false;
config->has_virtual_reserve = true;
// get the page size
SYSTEM_INFO si;
GetSystemInfo(&si);
if (si.dwPageSize > 0) { config->page_size = si.dwPageSize; }
if (si.dwAllocationGranularity > 0) { config->alloc_granularity = si.dwAllocationGranularity; }
// get virtual address bits
if ((uintptr_t)si.lpMaximumApplicationAddress > 0) {
const size_t vbits = MI_SIZE_BITS - mi_clz((uintptr_t)si.lpMaximumApplicationAddress);
config->virtual_address_bits = vbits;
}
// get the VirtualAlloc2 function
HINSTANCE hDll;
hDll = LoadLibrary(TEXT("kernelbase.dll"));
@@ -141,8 +162,22 @@ void _mi_prim_mem_init( mi_os_mem_config_t* config )
pGetNumaProcessorNodeEx = (PGetNumaProcessorNodeEx)(void (*)(void))GetProcAddress(hDll, "GetNumaProcessorNodeEx");
pGetNumaNodeProcessorMaskEx = (PGetNumaNodeProcessorMaskEx)(void (*)(void))GetProcAddress(hDll, "GetNumaNodeProcessorMaskEx");
pGetNumaProcessorNode = (PGetNumaProcessorNode)(void (*)(void))GetProcAddress(hDll, "GetNumaProcessorNode");
pGetNumaNodeProcessorMask = (PGetNumaNodeProcessorMask)(void (*)(void))GetProcAddress(hDll, "GetNumaNodeProcessorMask");
pGetNumaHighestNodeNumber = (PGetNumaHighestNodeNumber)(void (*)(void))GetProcAddress(hDll, "GetNumaHighestNodeNumber");
pGetLargePageMinimum = (PGetLargePageMinimum)(void (*)(void))GetProcAddress(hDll, "GetLargePageMinimum");
// Get physical memory (not available on XP, so check dynamically)
PGetPhysicallyInstalledSystemMemory pGetPhysicallyInstalledSystemMemory = (PGetPhysicallyInstalledSystemMemory)(void (*)(void))GetProcAddress(hDll,"GetPhysicallyInstalledSystemMemory");
if (pGetPhysicallyInstalledSystemMemory != NULL) {
ULONGLONG memInKiB = 0;
if ((*pGetPhysicallyInstalledSystemMemory)(&memInKiB)) {
if (memInKiB > 0 && memInKiB <= SIZE_MAX) {
config->physical_memory_in_kib = (size_t)memInKiB;
}
}
}
FreeLibrary(hDll);
}
// Enable large/huge OS page support?
if (mi_option_is_enabled(mi_option_allow_large_os_pages) || mi_option_is_enabled(mi_option_reserve_huge_os_pages)) {
win_enable_large_os_pages(&config->large_page_size);
}
@@ -162,7 +197,7 @@ int _mi_prim_free(void* addr, size_t size ) {
// In mi_os_mem_alloc_aligned the fallback path may have returned a pointer inside
// the memory region returned by VirtualAlloc; in that case we need to free using
// the start of the region.
MEMORY_BASIC_INFORMATION info = { 0 };
MEMORY_BASIC_INFORMATION info; _mi_memzero_var(info);
VirtualQuery(addr, &info, sizeof(info));
if (info.AllocationBase < addr && ((uint8_t*)addr - (uint8_t*)info.AllocationBase) < (ptrdiff_t)MI_SEGMENT_SIZE) {
errcode = 0;
@@ -178,7 +213,7 @@ int _mi_prim_free(void* addr, size_t size ) {
// VirtualAlloc
//---------------------------------------------
static void* win_virtual_alloc_prim(void* addr, size_t size, size_t try_alignment, DWORD flags) {
static void* win_virtual_alloc_prim_once(void* addr, size_t size, size_t try_alignment, DWORD flags) {
#if (MI_INTPTR_SIZE >= 8)
// on 64-bit systems, try to use the virtual address area after 2TiB for 4MiB aligned allocations
if (addr == NULL) {
@@ -192,7 +227,7 @@ static void* win_virtual_alloc_prim(void* addr, size_t size, size_t try_alignmen
}
#endif
// on modern Windows try use VirtualAlloc2 for aligned allocation
if (try_alignment > 1 && (try_alignment % _mi_os_page_size()) == 0 && pVirtualAlloc2 != NULL) {
if (addr == NULL && try_alignment > 1 && (try_alignment % _mi_os_page_size()) == 0 && pVirtualAlloc2 != NULL) {
MI_MEM_ADDRESS_REQUIREMENTS reqs = { 0, 0, 0 };
reqs.Alignment = try_alignment;
MI_MEM_EXTENDED_PARAMETER param = { {0, 0}, {0} };
@@ -200,20 +235,61 @@ static void* win_virtual_alloc_prim(void* addr, size_t size, size_t try_alignmen
param.Arg.Pointer = &reqs;
void* p = (*pVirtualAlloc2)(GetCurrentProcess(), addr, size, flags, PAGE_READWRITE, &param, 1);
if (p != NULL) return p;
_mi_warning_message("unable to allocate aligned OS memory (%zu bytes, error code: 0x%x, address: %p, alignment: %zu, flags: 0x%x)\n", size, GetLastError(), addr, try_alignment, flags);
_mi_warning_message("unable to allocate aligned OS memory (0x%zx bytes, error code: 0x%x, address: %p, alignment: 0x%zx, flags: 0x%x)\n", size, GetLastError(), addr, try_alignment, flags);
// fall through on error
}
// last resort
return VirtualAlloc(addr, size, flags, PAGE_READWRITE);
}
static bool win_is_out_of_memory_error(DWORD err) {
switch (err) {
case ERROR_COMMITMENT_MINIMUM:
case ERROR_COMMITMENT_LIMIT:
case ERROR_PAGEFILE_QUOTA:
case ERROR_NOT_ENOUGH_MEMORY:
return true;
default:
return false;
}
}
static void* win_virtual_alloc_prim(void* addr, size_t size, size_t try_alignment, DWORD flags) {
long max_retry_msecs = mi_option_get_clamp(mi_option_retry_on_oom, 0, 2000); // at most 2 seconds
if (max_retry_msecs == 1) { max_retry_msecs = 100; } // if one sets the option to "true"
for (long tries = 1; tries <= 10; tries++) { // try at most 10 times (=2200ms)
void* p = win_virtual_alloc_prim_once(addr, size, try_alignment, flags);
if (p != NULL) {
// success, return the address
return p;
}
else if (max_retry_msecs > 0 && (try_alignment <= 2*MI_SEGMENT_ALIGN) &&
(flags&MEM_COMMIT) != 0 && (flags&MEM_LARGE_PAGES) == 0 &&
win_is_out_of_memory_error(GetLastError())) {
// if committing regular memory and being out-of-memory,
// keep trying for a bit in case memory frees up after all. See issue #894
_mi_warning_message("out-of-memory on OS allocation, try again... (attempt %lu, 0x%zx bytes, error code: 0x%x, address: %p, alignment: 0x%zx, flags: 0x%x)\n", tries, size, GetLastError(), addr, try_alignment, flags);
long sleep_msecs = tries*40; // increasing waits
if (sleep_msecs > max_retry_msecs) { sleep_msecs = max_retry_msecs; }
max_retry_msecs -= sleep_msecs;
Sleep(sleep_msecs);
}
else {
// otherwise return with an error
break;
}
}
return NULL;
}
static void* win_virtual_alloc(void* addr, size_t size, size_t try_alignment, DWORD flags, bool large_only, bool allow_large, bool* is_large) {
mi_assert_internal(!(large_only && !allow_large));
static _Atomic(size_t) large_page_try_ok; // = 0;
void* p = NULL;
// Try to allocate large OS pages (2MiB) if allowed or required.
if ((large_only || _mi_os_use_large_page(size, try_alignment))
&& allow_large && (flags&MEM_COMMIT)!=0 && (flags&MEM_RESERVE)!=0) {
if ((large_only || (_mi_os_canuse_large_page(size, try_alignment) && mi_option_is_enabled(mi_option_allow_large_os_pages)))
&& allow_large && (flags&MEM_COMMIT)!=0 && (flags&MEM_RESERVE)!=0)
{
size_t try_ok = mi_atomic_load_acquire(&large_page_try_ok);
if (!large_only && try_ok > 0) {
// if a large page allocation fails, it seems the calls to VirtualAlloc get very expensive.
@@ -240,14 +316,14 @@ static void* win_virtual_alloc(void* addr, size_t size, size_t try_alignment, DW
return p;
}
int _mi_prim_alloc(size_t size, size_t try_alignment, bool commit, bool allow_large, bool* is_large, bool* is_zero, void** addr) {
int _mi_prim_alloc(void* hint_addr, size_t size, size_t try_alignment, bool commit, bool allow_large, bool* is_large, bool* is_zero, void** addr) {
mi_assert_internal(size > 0 && (size % _mi_os_page_size()) == 0);
mi_assert_internal(commit || !allow_large);
mi_assert_internal(try_alignment > 0);
*is_zero = true;
int flags = MEM_RESERVE;
if (commit) { flags |= MEM_COMMIT; }
*addr = win_virtual_alloc(NULL, size, try_alignment, flags, false, allow_large, is_large);
*addr = win_virtual_alloc(hint_addr, size, try_alignment, flags, false, allow_large, is_large);
return (*addr != NULL ? 0 : (int)GetLastError());
}
@@ -276,7 +352,7 @@ int _mi_prim_commit(void* addr, size_t size, bool* is_zero) {
return 0;
}
int _mi_prim_decommit(void* addr, size_t size, bool* needs_recommit) {
int _mi_prim_decommit(void* addr, size_t size, bool* needs_recommit) {
BOOL ok = VirtualFree(addr, size, MEM_DECOMMIT);
*needs_recommit = true; // for safety, assume always decommitted even in the case of an error.
return (ok ? 0 : (int)GetLastError());
@@ -293,6 +369,11 @@ int _mi_prim_reset(void* addr, size_t size) {
return (p != NULL ? 0 : (int)GetLastError());
}
int _mi_prim_reuse(void* addr, size_t size) {
MI_UNUSED(addr); MI_UNUSED(size);
return 0;
}
int _mi_prim_protect(void* addr, size_t size, bool protect) {
DWORD oldprotect = 0;
BOOL ok = VirtualProtect(addr, size, protect ? PAGE_NOACCESS : PAGE_READWRITE, &oldprotect);
@@ -324,7 +405,7 @@ static void* _mi_prim_alloc_huge_os_pagesx(void* hint_addr, size_t size, int num
}
SIZE_T psize = size;
void* base = hint_addr;
NTSTATUS err = (*pNtAllocateVirtualMemoryEx)(GetCurrentProcess(), &base, &psize, flags, PAGE_READWRITE, params, param_count);
LONG err = (*pNtAllocateVirtualMemoryEx)(GetCurrentProcess(), &base, &psize, flags, PAGE_READWRITE, params, param_count);
if (err == 0 && base != NULL) {
return base;
}
@@ -378,9 +459,11 @@ size_t _mi_prim_numa_node(void) {
size_t _mi_prim_numa_node_count(void) {
ULONG numa_max = 0;
GetNumaHighestNodeNumber(&numa_max);
if (pGetNumaHighestNodeNumber!=NULL) {
(*pGetNumaHighestNodeNumber)(&numa_max);
}
// find the highest node number that has actual processors assigned to it. Issue #282
while(numa_max > 0) {
while (numa_max > 0) {
if (pGetNumaNodeProcessorMaskEx != NULL) {
// Extended API is supported
GROUP_AFFINITY affinity;
@@ -391,8 +474,10 @@ size_t _mi_prim_numa_node_count(void) {
else {
// Vista or earlier, use older API that is limited to 64 processors.
ULONGLONG mask;
if (GetNumaNodeProcessorMask((UCHAR)numa_max, &mask)) {
if (mask != 0) break; // found the maximum non-empty node
if (pGetNumaNodeProcessorMask != NULL) {
if ((*pGetNumaNodeProcessorMask)((UCHAR)numa_max, &mask)) {
if (mask != 0) break; // found the maximum non-empty node
}
};
}
// max node was invalid or had no processor assigned, try again
@@ -428,7 +513,6 @@ mi_msecs_t _mi_prim_clock_now(void) {
// Process Info
//----------------------------------------------------------------
#include <windows.h>
#include <psapi.h>
static mi_msecs_t filetime_msecs(const FILETIME* ftime) {
@@ -451,7 +535,7 @@ void _mi_prim_process_info(mi_process_info_t* pinfo)
GetProcessTimes(GetCurrentProcess(), &ct, &et, &st, &ut);
pinfo->utime = filetime_msecs(&ut);
pinfo->stime = filetime_msecs(&st);
// load psapi on demand
if (pGetProcessMemoryInfo == NULL) {
HINSTANCE hDll = LoadLibrary(TEXT("psapi.dll"));
@@ -461,11 +545,10 @@ void _mi_prim_process_info(mi_process_info_t* pinfo)
}
// get process info
PROCESS_MEMORY_COUNTERS info;
memset(&info, 0, sizeof(info));
PROCESS_MEMORY_COUNTERS info; _mi_memzero_var(info);
if (pGetProcessMemoryInfo != NULL) {
pGetProcessMemoryInfo(GetCurrentProcess(), &info, sizeof(info));
}
}
pinfo->current_rss = (size_t)info.WorkingSetSize;
pinfo->peak_rss = (size_t)info.PeakWorkingSetSize;
pinfo->current_commit = (size_t)info.PagefileUsage;
@@ -477,24 +560,28 @@ void _mi_prim_process_info(mi_process_info_t* pinfo)
// Output
//----------------------------------------------------------------
void _mi_prim_out_stderr( const char* msg )
void _mi_prim_out_stderr( const char* msg )
{
// on windows with redirection, the C runtime cannot handle locale dependent output
// after the main thread closes so we use direct console output.
if (!_mi_preloading()) {
// _cputs(msg); // _cputs cannot be used at is aborts if it fails to lock the console
// _cputs(msg); // _cputs cannot be used as it aborts when failing to lock the console
static HANDLE hcon = INVALID_HANDLE_VALUE;
static bool hconIsConsole;
static bool hconIsConsole = false;
if (hcon == INVALID_HANDLE_VALUE) {
CONSOLE_SCREEN_BUFFER_INFO sbi;
hcon = GetStdHandle(STD_ERROR_HANDLE);
#ifdef MI_HAS_CONSOLE_IO
CONSOLE_SCREEN_BUFFER_INFO sbi;
hconIsConsole = ((hcon != INVALID_HANDLE_VALUE) && GetConsoleScreenBufferInfo(hcon, &sbi));
#endif
}
const size_t len = _mi_strlen(msg);
if (len > 0 && len < UINT32_MAX) {
DWORD written = 0;
if (hconIsConsole) {
#ifdef MI_HAS_CONSOLE_IO
WriteConsoleA(hcon, msg, (DWORD)len, &written, NULL);
#endif
}
else if (hcon != INVALID_HANDLE_VALUE) {
// use direct write if stderr was redirected
@@ -524,7 +611,6 @@ bool _mi_prim_getenv(const char* name, char* result, size_t result_size) {
}
//----------------------------------------------------------------
// Random
//----------------------------------------------------------------
@@ -560,63 +646,234 @@ bool _mi_prim_random_buf(void* buf, size_t buf_len) {
}
if (pBCryptGenRandom == NULL) return false;
}
return (pBCryptGenRandom(NULL, (PUCHAR)buf, (ULONG)buf_len, BCRYPT_USE_SYSTEM_PREFERRED_RNG) >= 0);
return (pBCryptGenRandom(NULL, (PUCHAR)buf, (ULONG)buf_len, BCRYPT_USE_SYSTEM_PREFERRED_RNG) >= 0);
}
#endif // MI_USE_RTLGENRANDOM
//----------------------------------------------------------------
// Thread init/done
// Process & Thread Init/Done
//----------------------------------------------------------------
#if !defined(MI_SHARED_LIB)
// use thread local storage keys to detect thread ending
#include <fibersapi.h>
#if (_WIN32_WINNT < 0x600) // before Windows Vista
WINBASEAPI DWORD WINAPI FlsAlloc( _In_opt_ PFLS_CALLBACK_FUNCTION lpCallback );
WINBASEAPI PVOID WINAPI FlsGetValue( _In_ DWORD dwFlsIndex );
WINBASEAPI BOOL WINAPI FlsSetValue( _In_ DWORD dwFlsIndex, _In_opt_ PVOID lpFlsData );
WINBASEAPI BOOL WINAPI FlsFree(_In_ DWORD dwFlsIndex);
#if MI_WIN_USE_FIXED_TLS==1
mi_decl_cache_align size_t _mi_win_tls_offset = 0;
#endif
static DWORD mi_fls_key = (DWORD)(-1);
//static void mi_debug_out(const char* s) {
// HANDLE h = GetStdHandle(STD_ERROR_HANDLE);
// WriteConsole(h, s, (DWORD)_mi_strlen(s), NULL, NULL);
//}
static void NTAPI mi_fls_done(PVOID value) {
mi_heap_t* heap = (mi_heap_t*)value;
if (heap != NULL) {
_mi_thread_done(heap);
FlsSetValue(mi_fls_key, NULL); // prevent recursion as _mi_thread_done may set it back to the main heap, issue #672
static void mi_win_tls_init(DWORD reason) {
if (reason==DLL_PROCESS_ATTACH || reason==DLL_THREAD_ATTACH) {
#if MI_WIN_USE_FIXED_TLS==1 // we must allocate a TLS slot dynamically
if (_mi_win_tls_offset == 0 && reason == DLL_PROCESS_ATTACH) {
const DWORD tls_slot = TlsAlloc(); // usually returns slot 1
if (tls_slot == TLS_OUT_OF_INDEXES) {
_mi_error_message(EFAULT, "unable to allocate the a TLS slot (rebuild without MI_WIN_USE_FIXED_TLS?)\n");
}
_mi_win_tls_offset = (size_t)tls_slot * sizeof(void*);
}
#endif
#if MI_HAS_TLS_SLOT >= 2 // we must initialize the TLS slot before any allocation
if (mi_prim_get_default_heap() == NULL) {
_mi_heap_set_default_direct((mi_heap_t*)&_mi_heap_empty);
#if MI_DEBUG && MI_WIN_USE_FIXED_TLS==1
void* const p = TlsGetValue((DWORD)(_mi_win_tls_offset / sizeof(void*)));
mi_assert_internal(p == (void*)&_mi_heap_empty);
#endif
}
#endif
}
}
void _mi_prim_thread_init_auto_done(void) {
mi_fls_key = FlsAlloc(&mi_fls_done);
static void NTAPI mi_win_main(PVOID module, DWORD reason, LPVOID reserved) {
MI_UNUSED(reserved);
MI_UNUSED(module);
mi_win_tls_init(reason);
if (reason==DLL_PROCESS_ATTACH) {
_mi_auto_process_init();
}
else if (reason==DLL_PROCESS_DETACH) {
_mi_auto_process_done();
}
else if (reason==DLL_THREAD_DETACH && !_mi_is_redirected()) {
_mi_thread_done(NULL);
}
}
void _mi_prim_thread_done_auto_done(void) {
// call thread-done on all threads (except the main thread) to prevent
// dangling callback pointer if statically linked with a DLL; Issue #208
FlsFree(mi_fls_key);
}
void _mi_prim_thread_associate_default_heap(mi_heap_t* heap) {
mi_assert_internal(mi_fls_key != (DWORD)(-1));
FlsSetValue(mi_fls_key, heap);
}
#if defined(MI_SHARED_LIB)
#define MI_PRIM_HAS_PROCESS_ATTACH 1
#else
// Windows DLL: easy to hook into process_init and thread_done
BOOL WINAPI DllMain(HINSTANCE inst, DWORD reason, LPVOID reserved) {
mi_win_main((PVOID)inst,reason,reserved);
return TRUE;
}
// Dll; nothing to do as in that case thread_done is handled through the DLL_THREAD_DETACH event.
// nothing to do since `_mi_thread_done` is handled through the DLL_THREAD_DETACH event.
void _mi_prim_thread_init_auto_done(void) { }
void _mi_prim_thread_done_auto_done(void) { }
void _mi_prim_thread_associate_default_heap(mi_heap_t* heap) {
MI_UNUSED(heap);
}
void _mi_prim_thread_init_auto_done(void) {
}
#elif !defined(MI_WIN_USE_FLS)
#define MI_PRIM_HAS_PROCESS_ATTACH 1
void _mi_prim_thread_done_auto_done(void) {
}
static void NTAPI mi_win_main_attach(PVOID module, DWORD reason, LPVOID reserved) {
if (reason == DLL_PROCESS_ATTACH || reason == DLL_THREAD_ATTACH) {
mi_win_main(module, reason, reserved);
}
}
static void NTAPI mi_win_main_detach(PVOID module, DWORD reason, LPVOID reserved) {
if (reason == DLL_PROCESS_DETACH || reason == DLL_THREAD_DETACH) {
mi_win_main(module, reason, reserved);
}
}
void _mi_prim_thread_associate_default_heap(mi_heap_t* heap) {
MI_UNUSED(heap);
}
// Set up TLS callbacks in a statically linked library by using special data sections.
// See <https://stackoverflow.com/questions/14538159/tls-callback-in-windows>
// We use 2 entries to ensure we call attach events before constructors
// are called, and detach events after destructors are called.
#if defined(__cplusplus)
extern "C" {
#endif
#if defined(_WIN64)
#pragma comment(linker, "/INCLUDE:_tls_used")
#pragma comment(linker, "/INCLUDE:_mi_tls_callback_pre")
#pragma comment(linker, "/INCLUDE:_mi_tls_callback_post")
#pragma const_seg(".CRT$XLB")
extern const PIMAGE_TLS_CALLBACK _mi_tls_callback_pre[];
const PIMAGE_TLS_CALLBACK _mi_tls_callback_pre[] = { &mi_win_main_attach };
#pragma const_seg()
#pragma const_seg(".CRT$XLY")
extern const PIMAGE_TLS_CALLBACK _mi_tls_callback_post[];
const PIMAGE_TLS_CALLBACK _mi_tls_callback_post[] = { &mi_win_main_detach };
#pragma const_seg()
#else
#pragma comment(linker, "/INCLUDE:__tls_used")
#pragma comment(linker, "/INCLUDE:__mi_tls_callback_pre")
#pragma comment(linker, "/INCLUDE:__mi_tls_callback_post")
#pragma data_seg(".CRT$XLB")
PIMAGE_TLS_CALLBACK _mi_tls_callback_pre[] = { &mi_win_main_attach };
#pragma data_seg()
#pragma data_seg(".CRT$XLY")
PIMAGE_TLS_CALLBACK _mi_tls_callback_post[] = { &mi_win_main_detach };
#pragma data_seg()
#endif
#if defined(__cplusplus)
}
#endif
// nothing to do since `_mi_thread_done` is handled through the DLL_THREAD_DETACH event.
void _mi_prim_thread_init_auto_done(void) { }
void _mi_prim_thread_done_auto_done(void) { }
void _mi_prim_thread_associate_default_heap(mi_heap_t* heap) {
MI_UNUSED(heap);
}
#else // deprecated: statically linked, use fiber api
#if defined(_MSC_VER) // on clang/gcc use the constructor attribute (in `src/prim/prim.c`)
// MSVC: use data section magic for static libraries
// See <https://www.codeguru.com/cpp/misc/misc/applicationcontrol/article.php/c6945/Running-Code-Before-and-After-Main.htm>
#define MI_PRIM_HAS_PROCESS_ATTACH 1
static int mi_process_attach(void) {
mi_win_main(NULL,DLL_PROCESS_ATTACH,NULL);
atexit(&_mi_auto_process_done);
return 0;
}
typedef int(*mi_crt_callback_t)(void);
#if defined(_WIN64)
#pragma comment(linker, "/INCLUDE:_mi_tls_callback")
#pragma section(".CRT$XIU", long, read)
#else
#pragma comment(linker, "/INCLUDE:__mi_tls_callback")
#endif
#pragma data_seg(".CRT$XIU")
mi_decl_externc mi_crt_callback_t _mi_tls_callback[] = { &mi_process_attach };
#pragma data_seg()
#endif
// use the fiber api for calling `_mi_thread_done`.
#include <fibersapi.h>
#if (_WIN32_WINNT < 0x600) // before Windows Vista
WINBASEAPI DWORD WINAPI FlsAlloc( _In_opt_ PFLS_CALLBACK_FUNCTION lpCallback );
WINBASEAPI PVOID WINAPI FlsGetValue( _In_ DWORD dwFlsIndex );
WINBASEAPI BOOL WINAPI FlsSetValue( _In_ DWORD dwFlsIndex, _In_opt_ PVOID lpFlsData );
WINBASEAPI BOOL WINAPI FlsFree(_In_ DWORD dwFlsIndex);
#endif
static DWORD mi_fls_key = (DWORD)(-1);
static void NTAPI mi_fls_done(PVOID value) {
mi_heap_t* heap = (mi_heap_t*)value;
if (heap != NULL) {
_mi_thread_done(heap);
FlsSetValue(mi_fls_key, NULL); // prevent recursion as _mi_thread_done may set it back to the main heap, issue #672
}
}
void _mi_prim_thread_init_auto_done(void) {
mi_fls_key = FlsAlloc(&mi_fls_done);
}
void _mi_prim_thread_done_auto_done(void) {
// call thread-done on all threads (except the main thread) to prevent
// dangling callback pointer if statically linked with a DLL; Issue #208
FlsFree(mi_fls_key);
}
void _mi_prim_thread_associate_default_heap(mi_heap_t* heap) {
mi_assert_internal(mi_fls_key != (DWORD)(-1));
FlsSetValue(mi_fls_key, heap);
}
#endif
// ----------------------------------------------------
// Communicate with the redirection module on Windows
// ----------------------------------------------------
#if defined(MI_SHARED_LIB) && !defined(MI_WIN_NOREDIRECT)
#define MI_PRIM_HAS_ALLOCATOR_INIT 1
static bool mi_redirected = false; // true if malloc redirects to mi_malloc
bool _mi_is_redirected(void) {
return mi_redirected;
}
#ifdef __cplusplus
extern "C" {
#endif
mi_decl_export void _mi_redirect_entry(DWORD reason) {
// called on redirection; careful as this may be called before DllMain
mi_win_tls_init(reason);
if (reason == DLL_PROCESS_ATTACH) {
mi_redirected = true;
}
else if (reason == DLL_PROCESS_DETACH) {
mi_redirected = false;
}
else if (reason == DLL_THREAD_DETACH) {
_mi_thread_done(NULL);
}
}
__declspec(dllimport) bool mi_cdecl mi_allocator_init(const char** message);
__declspec(dllimport) void mi_cdecl mi_allocator_done(void);
#ifdef __cplusplus
}
#endif
bool _mi_allocator_init(const char** message) {
return mi_allocator_init(message);
}
void _mi_allocator_done(void) {
mi_allocator_done();
}
#endif

22
mimalloc/mimalloc-sys/mimalloc/src/random.c
View File

@@ -143,13 +143,17 @@ void _mi_random_split(mi_random_ctx_t* ctx, mi_random_ctx_t* ctx_new) {
uintptr_t _mi_random_next(mi_random_ctx_t* ctx) {
mi_assert_internal(mi_random_is_initialized(ctx));
#if MI_INTPTR_SIZE <= 4
return chacha_next32(ctx);
#elif MI_INTPTR_SIZE == 8
return (((uintptr_t)chacha_next32(ctx) << 32) | chacha_next32(ctx));
#else
# error "define mi_random_next for this platform"
#endif
uintptr_t r;
do {
#if MI_INTPTR_SIZE <= 4
r = chacha_next32(ctx);
#elif MI_INTPTR_SIZE == 8
r = (((uintptr_t)chacha_next32(ctx) << 32) | chacha_next32(ctx));
#else
# error "define mi_random_next for this platform"
#endif
} while (r==0);
return r;
}
@@ -163,7 +167,7 @@ uintptr_t _mi_os_random_weak(uintptr_t extra_seed) {
x ^= _mi_prim_clock_now();
// and do a few randomization steps
uintptr_t max = ((x ^ (x >> 17)) & 0x0F) + 1;
for (uintptr_t i = 0; i < max; i++) {
for (uintptr_t i = 0; i < max || x==0; i++, x++) {
x = _mi_random_shuffle(x);
}
mi_assert_internal(x != 0);
@@ -179,7 +183,7 @@ static void mi_random_init_ex(mi_random_ctx_t* ctx, bool use_weak) {
if (!use_weak) { _mi_warning_message("unable to use secure randomness\n"); }
#endif
uintptr_t x = _mi_os_random_weak(0);
for (size_t i = 0; i < 8; i++) { // key is eight 32-bit words.
for (size_t i = 0; i < 8; i++, x++) { // key is eight 32-bit words.
x = _mi_random_shuffle(x);
((uint32_t*)key)[i] = (uint32_t)x;
}

177
mimalloc/mimalloc-sys/mimalloc/src/segment-map.c
View File

@@ -16,138 +16,127 @@ terms of the MIT license. A copy of the license can be found in the file
#include "mimalloc/internal.h"
#include "mimalloc/atomic.h"
#if (MI_INTPTR_SIZE==8)
#define MI_MAX_ADDRESS ((size_t)40 << 40) // 40TB (to include huge page areas)
// Reduce total address space to reduce .bss (due to the `mi_segment_map`)
#if (MI_INTPTR_SIZE > 4) && MI_TRACK_ASAN
#define MI_SEGMENT_MAP_MAX_ADDRESS (128*1024ULL*MI_GiB) // 128 TiB (see issue #881)
#elif (MI_INTPTR_SIZE > 4)
#define MI_SEGMENT_MAP_MAX_ADDRESS (48*1024ULL*MI_GiB) // 48 TiB
#else
#define MI_MAX_ADDRESS ((size_t)2 << 30) // 2Gb
#define MI_SEGMENT_MAP_MAX_ADDRESS (UINT32_MAX)
#endif
#define MI_SEGMENT_MAP_BITS (MI_MAX_ADDRESS / MI_SEGMENT_SIZE)
#define MI_SEGMENT_MAP_SIZE (MI_SEGMENT_MAP_BITS / 8)
#define MI_SEGMENT_MAP_WSIZE (MI_SEGMENT_MAP_SIZE / MI_INTPTR_SIZE)
#define MI_SEGMENT_MAP_PART_SIZE (MI_INTPTR_SIZE*MI_KiB - 128) // 128 > sizeof(mi_memid_t) !
#define MI_SEGMENT_MAP_PART_BITS (8*MI_SEGMENT_MAP_PART_SIZE)
#define MI_SEGMENT_MAP_PART_ENTRIES (MI_SEGMENT_MAP_PART_SIZE / MI_INTPTR_SIZE)
#define MI_SEGMENT_MAP_PART_BIT_SPAN (MI_SEGMENT_ALIGN) // memory area covered by 1 bit
static _Atomic(uintptr_t) mi_segment_map[MI_SEGMENT_MAP_WSIZE + 1]; // 2KiB per TB with 64MiB segments
#if (MI_SEGMENT_MAP_PART_BITS < (MI_SEGMENT_MAP_MAX_ADDRESS / MI_SEGMENT_MAP_PART_BIT_SPAN)) // prevent overflow on 32-bit (issue #1017)
#define MI_SEGMENT_MAP_PART_SPAN (MI_SEGMENT_MAP_PART_BITS * MI_SEGMENT_MAP_PART_BIT_SPAN)
#else
#define MI_SEGMENT_MAP_PART_SPAN MI_SEGMENT_MAP_MAX_ADDRESS
#endif
static size_t mi_segment_map_index_of(const mi_segment_t* segment, size_t* bitidx) {
#define MI_SEGMENT_MAP_MAX_PARTS ((MI_SEGMENT_MAP_MAX_ADDRESS / MI_SEGMENT_MAP_PART_SPAN) + 1)
// A part of the segment map.
typedef struct mi_segmap_part_s {
mi_memid_t memid;
_Atomic(uintptr_t) map[MI_SEGMENT_MAP_PART_ENTRIES];
} mi_segmap_part_t;
// Allocate parts on-demand to reduce .bss footprint
static _Atomic(mi_segmap_part_t*) mi_segment_map[MI_SEGMENT_MAP_MAX_PARTS]; // = { NULL, .. }
static mi_segmap_part_t* mi_segment_map_index_of(const mi_segment_t* segment, bool create_on_demand, size_t* idx, size_t* bitidx) {
// note: segment can be invalid or NULL.
mi_assert_internal(_mi_ptr_segment(segment + 1) == segment); // is it aligned on MI_SEGMENT_SIZE?
if ((uintptr_t)segment >= MI_MAX_ADDRESS) {
*bitidx = 0;
return MI_SEGMENT_MAP_WSIZE;
}
else {
const uintptr_t segindex = ((uintptr_t)segment) / MI_SEGMENT_SIZE;
*bitidx = segindex % MI_INTPTR_BITS;
const size_t mapindex = segindex / MI_INTPTR_BITS;
mi_assert_internal(mapindex < MI_SEGMENT_MAP_WSIZE);
return mapindex;
*idx = 0;
*bitidx = 0;
if ((uintptr_t)segment >= MI_SEGMENT_MAP_MAX_ADDRESS) return NULL;
const uintptr_t segindex = ((uintptr_t)segment) / MI_SEGMENT_MAP_PART_SPAN;
if (segindex >= MI_SEGMENT_MAP_MAX_PARTS) return NULL;
mi_segmap_part_t* part = mi_atomic_load_ptr_relaxed(mi_segmap_part_t, &mi_segment_map[segindex]);
// allocate on demand to reduce .bss footprint
if mi_unlikely(part == NULL) {
if (!create_on_demand) return NULL;
mi_memid_t memid;
part = (mi_segmap_part_t*)_mi_os_zalloc(sizeof(mi_segmap_part_t), &memid);
if (part == NULL) return NULL;
part->memid = memid;
mi_segmap_part_t* expected = NULL;
if (!mi_atomic_cas_ptr_strong_release(mi_segmap_part_t, &mi_segment_map[segindex], &expected, part)) {
_mi_os_free(part, sizeof(mi_segmap_part_t), memid);
part = expected;
if (part == NULL) return NULL;
}
}
mi_assert(part != NULL);
const uintptr_t offset = ((uintptr_t)segment) % MI_SEGMENT_MAP_PART_SPAN;
const uintptr_t bitofs = offset / MI_SEGMENT_MAP_PART_BIT_SPAN;
*idx = bitofs / MI_INTPTR_BITS;
*bitidx = bitofs % MI_INTPTR_BITS;
return part;
}
void _mi_segment_map_allocated_at(const mi_segment_t* segment) {
if (segment->memid.memkind == MI_MEM_ARENA) return; // we lookup segments first in the arena's and don't need the segment map
size_t index;
size_t bitidx;
size_t index = mi_segment_map_index_of(segment, &bitidx);
mi_assert_internal(index <= MI_SEGMENT_MAP_WSIZE);
if (index==MI_SEGMENT_MAP_WSIZE) return;
uintptr_t mask = mi_atomic_load_relaxed(&mi_segment_map[index]);
mi_segmap_part_t* part = mi_segment_map_index_of(segment, true /* alloc map if needed */, &index, &bitidx);
if (part == NULL) return; // outside our address range..
uintptr_t mask = mi_atomic_load_relaxed(&part->map[index]);
uintptr_t newmask;
do {
newmask = (mask | ((uintptr_t)1 << bitidx));
} while (!mi_atomic_cas_weak_release(&mi_segment_map[index], &mask, newmask));
} while (!mi_atomic_cas_weak_release(&part->map[index], &mask, newmask));
}
void _mi_segment_map_freed_at(const mi_segment_t* segment) {
if (segment->memid.memkind == MI_MEM_ARENA) return;
size_t index;
size_t bitidx;
size_t index = mi_segment_map_index_of(segment, &bitidx);
mi_assert_internal(index <= MI_SEGMENT_MAP_WSIZE);
if (index == MI_SEGMENT_MAP_WSIZE) return;
uintptr_t mask = mi_atomic_load_relaxed(&mi_segment_map[index]);
mi_segmap_part_t* part = mi_segment_map_index_of(segment, false /* don't alloc if not present */, &index, &bitidx);
if (part == NULL) return; // outside our address range..
uintptr_t mask = mi_atomic_load_relaxed(&part->map[index]);
uintptr_t newmask;
do {
newmask = (mask & ~((uintptr_t)1 << bitidx));
} while (!mi_atomic_cas_weak_release(&mi_segment_map[index], &mask, newmask));
} while (!mi_atomic_cas_weak_release(&part->map[index], &mask, newmask));
}
// Determine the segment belonging to a pointer or NULL if it is not in a valid segment.
static mi_segment_t* _mi_segment_of(const void* p) {
if (p == NULL) return NULL;
mi_segment_t* segment = _mi_ptr_segment(p);
mi_assert_internal(segment != NULL);
mi_segment_t* segment = _mi_ptr_segment(p); // segment can be NULL
size_t index;
size_t bitidx;
size_t index = mi_segment_map_index_of(segment, &bitidx);
// fast path: for any pointer to valid small/medium/large object or first MI_SEGMENT_SIZE in huge
const uintptr_t mask = mi_atomic_load_relaxed(&mi_segment_map[index]);
mi_segmap_part_t* part = mi_segment_map_index_of(segment, false /* dont alloc if not present */, &index, &bitidx);
if (part == NULL) return NULL;
const uintptr_t mask = mi_atomic_load_relaxed(&part->map[index]);
if mi_likely((mask & ((uintptr_t)1 << bitidx)) != 0) {
bool cookie_ok = (_mi_ptr_cookie(segment) == segment->cookie);
mi_assert_internal(cookie_ok); MI_UNUSED(cookie_ok);
return segment; // yes, allocated by us
}
if (index==MI_SEGMENT_MAP_WSIZE) return NULL;
// TODO: maintain max/min allocated range for efficiency for more efficient rejection of invalid pointers?
// search downwards for the first segment in case it is an interior pointer
// could be slow but searches in MI_INTPTR_SIZE * MI_SEGMENT_SIZE (512MiB) steps trough
// valid huge objects
// note: we could maintain a lowest index to speed up the path for invalid pointers?
size_t lobitidx;
size_t loindex;
uintptr_t lobits = mask & (((uintptr_t)1 << bitidx) - 1);
if (lobits != 0) {
loindex = index;
lobitidx = mi_bsr(lobits); // lobits != 0
}
else if (index == 0) {
return NULL;
}
else {
mi_assert_internal(index > 0);
uintptr_t lomask = mask;
loindex = index;
do {
loindex--;
lomask = mi_atomic_load_relaxed(&mi_segment_map[loindex]);
} while (lomask != 0 && loindex > 0);
if (lomask == 0) return NULL;
lobitidx = mi_bsr(lomask); // lomask != 0
}
mi_assert_internal(loindex < MI_SEGMENT_MAP_WSIZE);
// take difference as the addresses could be larger than the MAX_ADDRESS space.
size_t diff = (((index - loindex) * (8*MI_INTPTR_SIZE)) + bitidx - lobitidx) * MI_SEGMENT_SIZE;
segment = (mi_segment_t*)((uint8_t*)segment - diff);
if (segment == NULL) return NULL;
mi_assert_internal((void*)segment < p);
bool cookie_ok = (_mi_ptr_cookie(segment) == segment->cookie);
mi_assert_internal(cookie_ok);
if mi_unlikely(!cookie_ok) return NULL;
if (((uint8_t*)segment + mi_segment_size(segment)) <= (uint8_t*)p) return NULL; // outside the range
mi_assert_internal(p >= (void*)segment && (uint8_t*)p < (uint8_t*)segment + mi_segment_size(segment));
return segment;
return NULL;
}
// Is this a valid pointer in our heap?
static bool mi_is_valid_pointer(const void* p) {
return ((_mi_segment_of(p) != NULL) || (_mi_arena_contains(p)));
static bool mi_is_valid_pointer(const void* p) {
// first check if it is in an arena, then check if it is OS allocated
return (_mi_arena_contains(p) || _mi_segment_of(p) != NULL);
}
mi_decl_nodiscard mi_decl_export bool mi_is_in_heap_region(const void* p) mi_attr_noexcept {
return mi_is_valid_pointer(p);
}
/*
// Return the full segment range belonging to a pointer
static void* mi_segment_range_of(const void* p, size_t* size) {
mi_segment_t* segment = _mi_segment_of(p);
if (segment == NULL) {
if (size != NULL) *size = 0;
return NULL;
void _mi_segment_map_unsafe_destroy(void) {
for (size_t i = 0; i < MI_SEGMENT_MAP_MAX_PARTS; i++) {
mi_segmap_part_t* part = mi_atomic_exchange_ptr_relaxed(mi_segmap_part_t, &mi_segment_map[i], NULL);
if (part != NULL) {
_mi_os_free(part, sizeof(mi_segmap_part_t), part->memid);
}
}
else {
if (size != NULL) *size = segment->segment_size;
return segment;
}
mi_assert_expensive(page == NULL || mi_segment_is_valid(_mi_page_segment(page),tld));
mi_assert_internal(page == NULL || (mi_segment_page_size(_mi_page_segment(page)) - (MI_SECURE == 0 ? 0 : _mi_os_page_size())) >= block_size);
mi_reset_delayed(tld);
mi_assert_internal(page == NULL || mi_page_not_in_queue(page, tld));
return page;
}
*/

849
mimalloc/mimalloc-sys/mimalloc/src/segment.c
View File

File diff suppressed because it is too large Load Diff

3
mimalloc/mimalloc-sys/mimalloc/src/static.c
View File

@@ -27,10 +27,11 @@ terms of the MIT license. A copy of the license can be found in the file
#include "bitmap.c"
#include "heap.c"
#include "init.c"
#include "libc.c"
#include "options.c"
#include "os.c"
#include "page.c" // includes page-queue.c
#include "random.c"
#include "random.c"
#include "segment.c"
#include "segment-map.c"
#include "stats.c"

400
mimalloc/mimalloc-sys/mimalloc/src/stats.c
View File

@@ -9,7 +9,6 @@ terms of the MIT license. A copy of the license can be found in the file
#include "mimalloc/atomic.h"
#include "mimalloc/prim.h"
#include <stdio.h> // snprintf
#include <string.h> // memset
#if defined(_MSC_VER) && (_MSC_VER < 1920)
@@ -27,38 +26,29 @@ static bool mi_is_in_main(void* stat) {
static void mi_stat_update(mi_stat_count_t* stat, int64_t amount) {
if (amount == 0) return;
if (mi_is_in_main(stat))
if mi_unlikely(mi_is_in_main(stat))
{
// add atomically (for abandoned pages)
int64_t current = mi_atomic_addi64_relaxed(&stat->current, amount);
// if (stat == &_mi_stats_main.committed) { mi_assert_internal(current + amount >= 0); };
mi_atomic_maxi64_relaxed(&stat->peak, current + amount);
if (amount > 0) {
mi_atomic_addi64_relaxed(&stat->allocated,amount);
}
else {
mi_atomic_addi64_relaxed(&stat->freed, -amount);
mi_atomic_addi64_relaxed(&stat->total,amount);
}
}
else {
// add thread local
stat->current += amount;
if (stat->current > stat->peak) stat->peak = stat->current;
if (amount > 0) {
stat->allocated += amount;
}
else {
stat->freed += -amount;
}
if (stat->current > stat->peak) { stat->peak = stat->current; }
if (amount > 0) { stat->total += amount; }
}
}
void _mi_stat_counter_increase(mi_stat_counter_t* stat, size_t amount) {
if (mi_is_in_main(stat)) {
mi_atomic_addi64_relaxed( &stat->count, 1 );
mi_atomic_addi64_relaxed( &stat->total, (int64_t)amount );
}
else {
stat->count++;
stat->total += amount;
}
}
@@ -71,64 +61,70 @@ void _mi_stat_decrease(mi_stat_count_t* stat, size_t amount) {
mi_stat_update(stat, -((int64_t)amount));
}
// must be thread safe as it is called from stats_merge
static void mi_stat_add(mi_stat_count_t* stat, const mi_stat_count_t* src, int64_t unit) {
if (stat==src) return;
if (src->allocated==0 && src->freed==0) return;
mi_atomic_addi64_relaxed( &stat->allocated, src->allocated * unit);
mi_atomic_addi64_relaxed( &stat->current, src->current * unit);
mi_atomic_addi64_relaxed( &stat->freed, src->freed * unit);
// peak scores do not work across threads..
mi_atomic_addi64_relaxed( &stat->peak, src->peak * unit);
static void mi_stat_adjust(mi_stat_count_t* stat, int64_t amount) {
if (amount == 0) return;
if mi_unlikely(mi_is_in_main(stat))
{
// adjust atomically
mi_atomic_addi64_relaxed(&stat->current, amount);
mi_atomic_addi64_relaxed(&stat->total,amount);
}
else {
// adjust local
stat->current += amount;
stat->total += amount;
}
}
static void mi_stat_counter_add(mi_stat_counter_t* stat, const mi_stat_counter_t* src, int64_t unit) {
if (stat==src) return;
mi_atomic_addi64_relaxed( &stat->total, src->total * unit);
mi_atomic_addi64_relaxed( &stat->count, src->count * unit);
void _mi_stat_adjust_decrease(mi_stat_count_t* stat, size_t amount) {
mi_stat_adjust(stat, -((int64_t)amount));
}
// must be thread safe as it is called from stats_merge
static void mi_stat_count_add_mt(mi_stat_count_t* stat, const mi_stat_count_t* src) {
if (stat==src) return;
mi_atomic_void_addi64_relaxed(&stat->total, &src->total);
const int64_t prev_current = mi_atomic_addi64_relaxed(&stat->current, src->current);
// Global current plus thread peak approximates new global peak
// note: peak scores do really not work across threads.
// we used to just add them together but that often overestimates in practice.
// similarly, max does not seem to work well. The current approach
// by Artem Kharytoniuk (@artem-lunarg) seems to work better, see PR#1112
// for a longer description.
mi_atomic_maxi64_relaxed(&stat->peak, prev_current + src->peak);
}
static void mi_stat_counter_add_mt(mi_stat_counter_t* stat, const mi_stat_counter_t* src) {
if (stat==src) return;
mi_atomic_void_addi64_relaxed(&stat->total, &src->total);
}
#define MI_STAT_COUNT(stat) mi_stat_count_add_mt(&stats->stat, &src->stat);
#define MI_STAT_COUNTER(stat) mi_stat_counter_add_mt(&stats->stat, &src->stat);
// must be thread safe as it is called from stats_merge
static void mi_stats_add(mi_stats_t* stats, const mi_stats_t* src) {
if (stats==src) return;
mi_stat_add(&stats->segments, &src->segments,1);
mi_stat_add(&stats->pages, &src->pages,1);
mi_stat_add(&stats->reserved, &src->reserved, 1);
mi_stat_add(&stats->committed, &src->committed, 1);
mi_stat_add(&stats->reset, &src->reset, 1);
mi_stat_add(&stats->purged, &src->purged, 1);
mi_stat_add(&stats->page_committed, &src->page_committed, 1);
mi_stat_add(&stats->pages_abandoned, &src->pages_abandoned, 1);
mi_stat_add(&stats->segments_abandoned, &src->segments_abandoned, 1);
mi_stat_add(&stats->threads, &src->threads, 1);
// copy all fields
MI_STAT_FIELDS()
mi_stat_add(&stats->malloc, &src->malloc, 1);
mi_stat_add(&stats->segments_cache, &src->segments_cache, 1);
mi_stat_add(&stats->normal, &src->normal, 1);
mi_stat_add(&stats->huge, &src->huge, 1);
mi_stat_add(&stats->large, &src->large, 1);
mi_stat_counter_add(&stats->pages_extended, &src->pages_extended, 1);
mi_stat_counter_add(&stats->mmap_calls, &src->mmap_calls, 1);
mi_stat_counter_add(&stats->commit_calls, &src->commit_calls, 1);
mi_stat_counter_add(&stats->reset_calls, &src->reset_calls, 1);
mi_stat_counter_add(&stats->purge_calls, &src->purge_calls, 1);
mi_stat_counter_add(&stats->page_no_retire, &src->page_no_retire, 1);
mi_stat_counter_add(&stats->searches, &src->searches, 1);
mi_stat_counter_add(&stats->normal_count, &src->normal_count, 1);
mi_stat_counter_add(&stats->huge_count, &src->huge_count, 1);
mi_stat_counter_add(&stats->large_count, &src->large_count, 1);
#if MI_STAT>1
#if MI_STAT>1
for (size_t i = 0; i <= MI_BIN_HUGE; i++) {
if (src->normal_bins[i].allocated > 0 || src->normal_bins[i].freed > 0) {
mi_stat_add(&stats->normal_bins[i], &src->normal_bins[i], 1);
}
mi_stat_count_add_mt(&stats->malloc_bins[i], &src->malloc_bins[i]);
}
#endif
for (size_t i = 0; i <= MI_BIN_HUGE; i++) {
mi_stat_count_add_mt(&stats->page_bins[i], &src->page_bins[i]);
}
#endif
}
#undef MI_STAT_COUNT
#undef MI_STAT_COUNTER
/* -----------------------------------------------------------
Display statistics
----------------------------------------------------------- */
@@ -146,7 +142,7 @@ static void mi_printf_amount(int64_t n, int64_t unit, mi_output_fun* out, void*
const int64_t pos = (n < 0 ? -n : n);
if (pos < base) {
if (n!=1 || suffix[0] != 'B') { // skip printing 1 B for the unit column
snprintf(buf, len, "%d %-3s", (int)n, (n==0 ? "" : suffix));
_mi_snprintf(buf, len, "%lld %-3s", (long long)n, (n==0 ? "" : suffix));
}
}
else {
@@ -158,8 +154,8 @@ static void mi_printf_amount(int64_t n, int64_t unit, mi_output_fun* out, void*
const long whole = (long)(tens/10);
const long frac1 = (long)(tens%10);
char unitdesc[8];
snprintf(unitdesc, 8, "%s%s%s", magnitude, (base==1024 ? "i" : ""), suffix);
snprintf(buf, len, "%ld.%ld %-3s", whole, (frac1 < 0 ? -frac1 : frac1), unitdesc);
_mi_snprintf(unitdesc, 8, "%s%s%s", magnitude, (base==1024 ? "i" : ""), suffix);
_mi_snprintf(buf, len, "%ld.%ld %-3s", whole, (frac1 < 0 ? -frac1 : frac1), unitdesc);
}
_mi_fprintf(out, arg, (fmt==NULL ? "%12s" : fmt), buf);
}
@@ -176,14 +172,29 @@ static void mi_print_count(int64_t n, int64_t unit, mi_output_fun* out, void* ar
static void mi_stat_print_ex(const mi_stat_count_t* stat, const char* msg, int64_t unit, mi_output_fun* out, void* arg, const char* notok ) {
_mi_fprintf(out, arg,"%10s:", msg);
if (unit > 0) {
mi_print_amount(stat->peak, unit, out, arg);
mi_print_amount(stat->allocated, unit, out, arg);
mi_print_amount(stat->freed, unit, out, arg);
mi_print_amount(stat->current, unit, out, arg);
mi_print_amount(unit, 1, out, arg);
mi_print_count(stat->allocated, unit, out, arg);
if (stat->allocated > stat->freed) {
if (unit != 0) {
if (unit > 0) {
mi_print_amount(stat->peak, unit, out, arg);
mi_print_amount(stat->total, unit, out, arg);
// mi_print_amount(stat->freed, unit, out, arg);
mi_print_amount(stat->current, unit, out, arg);
mi_print_amount(unit, 1, out, arg);
mi_print_count(stat->total, unit, out, arg);
}
else {
mi_print_amount(stat->peak, -1, out, arg);
mi_print_amount(stat->total, -1, out, arg);
// mi_print_amount(stat->freed, -1, out, arg);
mi_print_amount(stat->current, -1, out, arg);
if (unit == -1) {
_mi_fprintf(out, arg, "%24s", "");
}
else {
mi_print_amount(-unit, 1, out, arg);
mi_print_count((stat->total / -unit), 0, out, arg);
}
}
if (stat->current != 0) {
_mi_fprintf(out, arg, " ");
_mi_fprintf(out, arg, (notok == NULL ? "not all freed" : notok));
_mi_fprintf(out, arg, "\n");
@@ -192,26 +203,9 @@ static void mi_stat_print_ex(const mi_stat_count_t* stat, const char* msg, int64
_mi_fprintf(out, arg, " ok\n");
}
}
else if (unit<0) {
mi_print_amount(stat->peak, -1, out, arg);
mi_print_amount(stat->allocated, -1, out, arg);
mi_print_amount(stat->freed, -1, out, arg);
mi_print_amount(stat->current, -1, out, arg);
if (unit==-1) {
_mi_fprintf(out, arg, "%24s", "");
}
else {
mi_print_amount(-unit, 1, out, arg);
mi_print_count((stat->allocated / -unit), 0, out, arg);
}
if (stat->allocated > stat->freed)
_mi_fprintf(out, arg, " not all freed!\n");
else
_mi_fprintf(out, arg, " ok\n");
}
else {
mi_print_amount(stat->peak, 1, out, arg);
mi_print_amount(stat->allocated, 1, out, arg);
mi_print_amount(stat->total, 1, out, arg);
_mi_fprintf(out, arg, "%11s", " "); // no freed
mi_print_amount(stat->current, 1, out, arg);
_mi_fprintf(out, arg, "\n");
@@ -222,11 +216,14 @@ static void mi_stat_print(const mi_stat_count_t* stat, const char* msg, int64_t
mi_stat_print_ex(stat, msg, unit, out, arg, NULL);
}
static void mi_stat_peak_print(const mi_stat_count_t* stat, const char* msg, int64_t unit, mi_output_fun* out, void* arg) {
#if MI_STAT>1
static void mi_stat_total_print(const mi_stat_count_t* stat, const char* msg, int64_t unit, mi_output_fun* out, void* arg) {
_mi_fprintf(out, arg, "%10s:", msg);
mi_print_amount(stat->peak, unit, out, arg);
_mi_fprintf(out, arg, "%12s", " "); // no peak
mi_print_amount(stat->total, unit, out, arg);
_mi_fprintf(out, arg, "\n");
}
#endif
static void mi_stat_counter_print(const mi_stat_counter_t* stat, const char* msg, mi_output_fun* out, void* arg ) {
_mi_fprintf(out, arg, "%10s:", msg);
@@ -235,8 +232,8 @@ static void mi_stat_counter_print(const mi_stat_counter_t* stat, const char* msg
}
static void mi_stat_counter_print_avg(const mi_stat_counter_t* stat, const char* msg, mi_output_fun* out, void* arg) {
const int64_t avg_tens = (stat->count == 0 ? 0 : (stat->total*10 / stat->count));
static void mi_stat_average_print(size_t count, size_t total, const char* msg, mi_output_fun* out, void* arg) {
const int64_t avg_tens = (count == 0 ? 0 : (total*10 / count));
const long avg_whole = (long)(avg_tens/10);
const long avg_frac1 = (long)(avg_tens%10);
_mi_fprintf(out, arg, "%10s: %5ld.%ld avg\n", msg, avg_whole, avg_frac1);
@@ -244,7 +241,7 @@ static void mi_stat_counter_print_avg(const mi_stat_counter_t* stat, const char*
static void mi_print_header(mi_output_fun* out, void* arg ) {
_mi_fprintf(out, arg, "%10s: %11s %11s %11s %11s %11s %11s\n", "heap stats", "peak ", "total ", "freed ", "current ", "unit ", "count ");
_mi_fprintf(out, arg, "%10s: %11s %11s %11s %11s %11s\n", "heap stats", "peak ", "total ", "current ", "block ", "total# ");
}
#if MI_STAT>1
@@ -252,10 +249,10 @@ static void mi_stats_print_bins(const mi_stat_count_t* bins, size_t max, const c
bool found = false;
char buf[64];
for (size_t i = 0; i <= max; i++) {
if (bins[i].allocated > 0) {
if (bins[i].total > 0) {
found = true;
int64_t unit = _mi_bin_size((uint8_t)i);
snprintf(buf, 64, "%s %3lu", fmt, (long)i);
_mi_snprintf(buf, 64, "%s %3lu", fmt, (long)i);
mi_stat_print(&bins[i], buf, unit, out, arg);
}
}
@@ -313,41 +310,45 @@ static void _mi_stats_print(mi_stats_t* stats, mi_output_fun* out0, void* arg0)
// and print using that
mi_print_header(out,arg);
#if MI_STAT>1
mi_stats_print_bins(stats->normal_bins, MI_BIN_HUGE, "normal",out,arg);
mi_stats_print_bins(stats->malloc_bins, MI_BIN_HUGE, "bin",out,arg);
#endif
#if MI_STAT
mi_stat_print(&stats->normal, "normal", (stats->normal_count.count == 0 ? 1 : -(stats->normal.allocated / stats->normal_count.count)), out, arg);
mi_stat_print(&stats->large, "large", (stats->large_count.count == 0 ? 1 : -(stats->large.allocated / stats->large_count.count)), out, arg);
mi_stat_print(&stats->huge, "huge", (stats->huge_count.count == 0 ? 1 : -(stats->huge.allocated / stats->huge_count.count)), out, arg);
mi_stat_count_t total = { 0,0,0,0 };
mi_stat_add(&total, &stats->normal, 1);
mi_stat_add(&total, &stats->large, 1);
mi_stat_add(&total, &stats->huge, 1);
mi_stat_print(&total, "total", 1, out, arg);
mi_stat_print(&stats->malloc_normal, "binned", (stats->malloc_normal_count.total == 0 ? 1 : -1), out, arg);
// mi_stat_print(&stats->malloc_large, "large", (stats->malloc_large_count.total == 0 ? 1 : -1), out, arg);
mi_stat_print(&stats->malloc_huge, "huge", (stats->malloc_huge_count.total == 0 ? 1 : -1), out, arg);
mi_stat_count_t total = { 0,0,0 };
mi_stat_count_add_mt(&total, &stats->malloc_normal);
// mi_stat_count_add(&total, &stats->malloc_large);
mi_stat_count_add_mt(&total, &stats->malloc_huge);
mi_stat_print_ex(&total, "total", 1, out, arg, "");
#endif
#if MI_STAT>1
mi_stat_print(&stats->malloc, "malloc req", 1, out, arg);
mi_stat_total_print(&stats->malloc_requested, "malloc req", 1, out, arg);
_mi_fprintf(out, arg, "\n");
#endif
mi_stat_print_ex(&stats->reserved, "reserved", 1, out, arg, "");
mi_stat_print_ex(&stats->committed, "committed", 1, out, arg, "");
mi_stat_peak_print(&stats->reset, "reset", 1, out, arg );
mi_stat_peak_print(&stats->purged, "purged", 1, out, arg );
mi_stat_print(&stats->page_committed, "touched", 1, out, arg);
mi_stat_counter_print(&stats->reset, "reset", out, arg );
mi_stat_counter_print(&stats->purged, "purged", out, arg );
mi_stat_print_ex(&stats->page_committed, "touched", 1, out, arg, "");
mi_stat_print(&stats->segments, "segments", -1, out, arg);
mi_stat_print(&stats->segments_abandoned, "-abandoned", -1, out, arg);
mi_stat_print(&stats->segments_cache, "-cached", -1, out, arg);
mi_stat_print(&stats->pages, "pages", -1, out, arg);
mi_stat_print(&stats->pages_abandoned, "-abandoned", -1, out, arg);
mi_stat_counter_print(&stats->pages_extended, "-extended", out, arg);
mi_stat_counter_print(&stats->page_no_retire, "-noretire", out, arg);
mi_stat_counter_print(&stats->pages_retire, "-retire", out, arg);
mi_stat_counter_print(&stats->arena_count, "arenas", out, arg);
// mi_stat_counter_print(&stats->arena_crossover_count, "-crossover", out, arg);
mi_stat_counter_print(&stats->arena_rollback_count, "-rollback", out, arg);
mi_stat_counter_print(&stats->mmap_calls, "mmaps", out, arg);
mi_stat_counter_print(&stats->commit_calls, "commits", out, arg);
mi_stat_counter_print(&stats->reset_calls, "resets", out, arg);
mi_stat_counter_print(&stats->purge_calls, "purges", out, arg);
mi_stat_counter_print(&stats->malloc_guarded_count, "guarded", out, arg);
mi_stat_print(&stats->threads, "threads", -1, out, arg);
mi_stat_counter_print_avg(&stats->searches, "searches", out, arg);
_mi_fprintf(out, arg, "%10s: %5zu\n", "numa nodes", _mi_os_numa_node_count());
mi_stat_average_print(stats->page_searches_count.total, stats->page_searches.total, "searches", out, arg);
_mi_fprintf(out, arg, "%10s: %5i\n", "numa nodes", _mi_os_numa_node_count());
size_t elapsed;
size_t user_time;
@@ -358,12 +359,12 @@ static void _mi_stats_print(mi_stats_t* stats, mi_output_fun* out0, void* arg0)
size_t peak_commit;
size_t page_faults;
mi_process_info(&elapsed, &user_time, &sys_time, &current_rss, &peak_rss, &current_commit, &peak_commit, &page_faults);
_mi_fprintf(out, arg, "%10s: %5ld.%03ld s\n", "elapsed", elapsed/1000, elapsed%1000);
_mi_fprintf(out, arg, "%10s: user: %ld.%03ld s, system: %ld.%03ld s, faults: %lu, rss: ", "process",
user_time/1000, user_time%1000, sys_time/1000, sys_time%1000, (unsigned long)page_faults );
_mi_fprintf(out, arg, "%10s: %5zu.%03zu s\n", "elapsed", elapsed/1000, elapsed%1000);
_mi_fprintf(out, arg, "%10s: user: %zu.%03zu s, system: %zu.%03zu s, faults: %zu, peak rss: ", "process",
user_time/1000, user_time%1000, sys_time/1000, sys_time%1000, page_faults );
mi_printf_amount((int64_t)peak_rss, 1, out, arg, "%s");
if (peak_commit > 0) {
_mi_fprintf(out, arg, ", commit: ");
_mi_fprintf(out, arg, ", peak commit: ");
mi_printf_amount((int64_t)peak_commit, 1, out, arg, "%s");
}
_mi_fprintf(out, arg, "\n");
@@ -394,6 +395,10 @@ void mi_stats_merge(void) mi_attr_noexcept {
mi_stats_merge_from( mi_stats_get_default() );
}
void _mi_stats_merge_thread(mi_tld_t* tld) {
mi_stats_merge_from( &tld->stats );
}
void _mi_stats_done(mi_stats_t* stats) { // called from `mi_thread_done`
mi_stats_merge_from(stats);
}
@@ -455,7 +460,7 @@ mi_decl_export void mi_process_info(size_t* elapsed_msecs, size_t* user_msecs, s
pinfo.page_faults = 0;
_mi_prim_process_info(&pinfo);
if (elapsed_msecs!=NULL) *elapsed_msecs = (pinfo.elapsed < 0 ? 0 : (pinfo.elapsed < (mi_msecs_t)PTRDIFF_MAX ? (size_t)pinfo.elapsed : PTRDIFF_MAX));
if (user_msecs!=NULL) *user_msecs = (pinfo.utime < 0 ? 0 : (pinfo.utime < (mi_msecs_t)PTRDIFF_MAX ? (size_t)pinfo.utime : PTRDIFF_MAX));
if (system_msecs!=NULL) *system_msecs = (pinfo.stime < 0 ? 0 : (pinfo.stime < (mi_msecs_t)PTRDIFF_MAX ? (size_t)pinfo.stime : PTRDIFF_MAX));
@@ -465,3 +470,164 @@ mi_decl_export void mi_process_info(size_t* elapsed_msecs, size_t* user_msecs, s
if (peak_commit!=NULL) *peak_commit = pinfo.peak_commit;
if (page_faults!=NULL) *page_faults = pinfo.page_faults;
}
// --------------------------------------------------------
// Return statistics
// --------------------------------------------------------
void mi_stats_get(size_t stats_size, mi_stats_t* stats) mi_attr_noexcept {
if (stats == NULL || stats_size == 0) return;
_mi_memzero(stats, stats_size);
const size_t size = (stats_size > sizeof(mi_stats_t) ? sizeof(mi_stats_t) : stats_size);
_mi_memcpy(stats, &_mi_stats_main, size);
stats->version = MI_STAT_VERSION;
}
// --------------------------------------------------------
// Statics in json format
// --------------------------------------------------------
typedef struct mi_heap_buf_s {
char* buf;
size_t size;
size_t used;
bool can_realloc;
} mi_heap_buf_t;
static bool mi_heap_buf_expand(mi_heap_buf_t* hbuf) {
if (hbuf==NULL) return false;
if (hbuf->buf != NULL && hbuf->size>0) {
hbuf->buf[hbuf->size-1] = 0;
}
if (hbuf->size > SIZE_MAX/2 || !hbuf->can_realloc) return false;
const size_t newsize = (hbuf->size == 0 ? mi_good_size(12*MI_KiB) : 2*hbuf->size);
char* const newbuf = (char*)mi_rezalloc(hbuf->buf, newsize);
if (newbuf == NULL) return false;
hbuf->buf = newbuf;
hbuf->size = newsize;
return true;
}
static void mi_heap_buf_print(mi_heap_buf_t* hbuf, const char* msg) {
if (msg==NULL || hbuf==NULL) return;
if (hbuf->used + 1 >= hbuf->size && !hbuf->can_realloc) return;
for (const char* src = msg; *src != 0; src++) {
char c = *src;
if (hbuf->used + 1 >= hbuf->size) {
if (!mi_heap_buf_expand(hbuf)) return;
}
mi_assert_internal(hbuf->used < hbuf->size);
hbuf->buf[hbuf->used++] = c;
}
mi_assert_internal(hbuf->used < hbuf->size);
hbuf->buf[hbuf->used] = 0;
}
static void mi_heap_buf_print_count_bin(mi_heap_buf_t* hbuf, const char* prefix, mi_stat_count_t* stat, size_t bin, bool add_comma) {
const size_t binsize = _mi_bin_size(bin);
const size_t pagesize = (binsize <= MI_SMALL_OBJ_SIZE_MAX ? MI_SMALL_PAGE_SIZE :
(binsize <= MI_MEDIUM_OBJ_SIZE_MAX ? MI_MEDIUM_PAGE_SIZE :
#if MI_LARGE_PAGE_SIZE
(binsize <= MI_LARGE_OBJ_SIZE_MAX ? MI_LARGE_PAGE_SIZE : 0)
#else
0
#endif
));
char buf[128];
_mi_snprintf(buf, 128, "%s{ \"total\": %lld, \"peak\": %lld, \"current\": %lld, \"block_size\": %zu, \"page_size\": %zu }%s\n", prefix, stat->total, stat->peak, stat->current, binsize, pagesize, (add_comma ? "," : ""));
buf[127] = 0;
mi_heap_buf_print(hbuf, buf);
}
static void mi_heap_buf_print_count(mi_heap_buf_t* hbuf, const char* prefix, mi_stat_count_t* stat, bool add_comma) {
char buf[128];
_mi_snprintf(buf, 128, "%s{ \"total\": %lld, \"peak\": %lld, \"current\": %lld }%s\n", prefix, stat->total, stat->peak, stat->current, (add_comma ? "," : ""));
buf[127] = 0;
mi_heap_buf_print(hbuf, buf);
}
static void mi_heap_buf_print_count_value(mi_heap_buf_t* hbuf, const char* name, mi_stat_count_t* stat) {
char buf[128];
_mi_snprintf(buf, 128, " \"%s\": ", name);
buf[127] = 0;
mi_heap_buf_print(hbuf, buf);
mi_heap_buf_print_count(hbuf, "", stat, true);
}
static void mi_heap_buf_print_value(mi_heap_buf_t* hbuf, const char* name, int64_t val) {
char buf[128];
_mi_snprintf(buf, 128, " \"%s\": %lld,\n", name, val);
buf[127] = 0;
mi_heap_buf_print(hbuf, buf);
}
static void mi_heap_buf_print_size(mi_heap_buf_t* hbuf, const char* name, size_t val, bool add_comma) {
char buf[128];
_mi_snprintf(buf, 128, " \"%s\": %zu%s\n", name, val, (add_comma ? "," : ""));
buf[127] = 0;
mi_heap_buf_print(hbuf, buf);
}
static void mi_heap_buf_print_counter_value(mi_heap_buf_t* hbuf, const char* name, mi_stat_counter_t* stat) {
mi_heap_buf_print_value(hbuf, name, stat->total);
}
#define MI_STAT_COUNT(stat) mi_heap_buf_print_count_value(&hbuf, #stat, &stats->stat);
#define MI_STAT_COUNTER(stat) mi_heap_buf_print_counter_value(&hbuf, #stat, &stats->stat);
char* mi_stats_get_json(size_t output_size, char* output_buf) mi_attr_noexcept {
mi_heap_buf_t hbuf = { NULL, 0, 0, true };
if (output_size > 0 && output_buf != NULL) {
_mi_memzero(output_buf, output_size);
hbuf.buf = output_buf;
hbuf.size = output_size;
hbuf.can_realloc = false;
}
else {
if (!mi_heap_buf_expand(&hbuf)) return NULL;
}
mi_heap_buf_print(&hbuf, "{\n");
mi_heap_buf_print_value(&hbuf, "version", MI_STAT_VERSION);
mi_heap_buf_print_value(&hbuf, "mimalloc_version", MI_MALLOC_VERSION);
// process info
mi_heap_buf_print(&hbuf, " \"process\": {\n");
size_t elapsed;
size_t user_time;
size_t sys_time;
size_t current_rss;
size_t peak_rss;
size_t current_commit;
size_t peak_commit;
size_t page_faults;
mi_process_info(&elapsed, &user_time, &sys_time, &current_rss, &peak_rss, &current_commit, &peak_commit, &page_faults);
mi_heap_buf_print_size(&hbuf, "elapsed_msecs", elapsed, true);
mi_heap_buf_print_size(&hbuf, "user_msecs", user_time, true);
mi_heap_buf_print_size(&hbuf, "system_msecs", sys_time, true);
mi_heap_buf_print_size(&hbuf, "page_faults", page_faults, true);
mi_heap_buf_print_size(&hbuf, "rss_current", current_rss, true);
mi_heap_buf_print_size(&hbuf, "rss_peak", peak_rss, true);
mi_heap_buf_print_size(&hbuf, "commit_current", current_commit, true);
mi_heap_buf_print_size(&hbuf, "commit_peak", peak_commit, false);
mi_heap_buf_print(&hbuf, " },\n");
// statistics
mi_stats_t* stats = &_mi_stats_main;
MI_STAT_FIELDS()
// size bins
mi_heap_buf_print(&hbuf, " \"malloc_bins\": [\n");
for (size_t i = 0; i <= MI_BIN_HUGE; i++) {
mi_heap_buf_print_count_bin(&hbuf, " ", &stats->malloc_bins[i], i, i!=MI_BIN_HUGE);
}
mi_heap_buf_print(&hbuf, " ],\n");
mi_heap_buf_print(&hbuf, " \"page_bins\": [\n");
for (size_t i = 0; i <= MI_BIN_HUGE; i++) {
mi_heap_buf_print_count_bin(&hbuf, " ", &stats->page_bins[i], i, i!=MI_BIN_HUGE);
}
mi_heap_buf_print(&hbuf, " ]\n");
mi_heap_buf_print(&hbuf, "}\n");
return hbuf.buf;
}