mimalloc/readme.md

582 lines
31 KiB
Markdown
Raw Permalink Normal View History

2019-06-19 16:26:12 -07:00
<img align="left" width="100" height="100" src="doc/mimalloc-logo.png"/>
2019-11-21 16:20:35 -08:00
[<img align="right" src="https://dev.azure.com/Daan0324/mimalloc/_apis/build/status/microsoft.mimalloc?branchName=dev"/>](https://dev.azure.com/Daan0324/mimalloc/_build?definitionId=1&_a=summary)
2019-06-19 16:52:58 -07:00
# mimalloc
2019-06-19 16:26:12 -07:00
&nbsp;
2019-06-19 16:52:58 -07:00
mimalloc (pronounced "me-malloc")
2019-06-19 18:22:11 -07:00
is a general purpose allocator with excellent [performance](#performance) characteristics.
2019-06-19 16:26:12 -07:00
Initially developed by Daan Leijen for the run-time systems of the
2020-02-09 18:32:09 -08:00
[Koka](https://github.com/koka-lang/koka) and [Lean](https://github.com/leanprover/lean) languages.
2020-05-05 10:45:10 -07:00
Latest release:`v1.6.3` (2020-05-05).
2019-06-19 16:26:12 -07:00
It 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:
2019-06-19 16:26:12 -07:00
```
> LD_PRELOAD=/usr/bin/libmimalloc.so myprogram
```
2020-01-22 18:16:40 -08:00
It also has an easy way to override the allocator in [Windows](#override_on_windows). Notable aspects of the design include:
2019-06-19 16:26:12 -07:00
2019-10-07 12:08:44 -07:00
- __small and consistent__: the library is about 6k LOC using simple and
2019-06-19 16:26:12 -07:00
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).
- __free list sharding__: the big idea: instead of one big free list (per size class) we have
many smaller lists per memory "page" which both reduces fragmentation
and increases locality --
things that are allocated close in time get allocated close in memory.
2019-06-20 08:01:39 -07:00
(A memory "page" in _mimalloc_ contains blocks of one size class and is
usually 64KiB on a 64-bit system).
2019-06-19 16:26:12 -07:00
- __eager page reset__: when a "page" becomes empty (with increased chance
due to free list sharding) the memory is marked to the OS as unused ("reset" or "purged")
reducing (real) memory pressure and fragmentation, especially in long running
programs.
2019-06-22 17:16:54 -04:00
- __secure__: _mimalloc_ can be built in secure mode, adding guard pages,
2019-06-20 07:58:34 -07:00
randomized allocation, encrypted free lists, etc. to protect against various
2019-11-21 16:17:47 -08:00
heap vulnerabilities. The performance penalty is usually around 10% on average
2019-06-19 17:00:31 -07:00
over our benchmarks.
2019-06-19 18:07:56 -07:00
- __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.
2019-06-19 16:26:12 -07:00
- __bounded__: it does not suffer from _blowup_ \[1\], has bounded worst-case allocation
2019-10-07 12:08:44 -07:00
times (_wcat_), bounded space overhead (~0.2% meta-data, with at most 12.5% waste in allocation sizes),
2019-06-19 18:18:21 -07:00
and has no internal points of contention using only atomic operations.
2019-06-19 18:22:11 -07:00
- __fast__: In our benchmarks (see [below](#performance)),
2020-01-22 18:16:40 -08:00
_mimalloc_ outperforms other leading allocators (_jemalloc_, _tcmalloc_, _Hoard_, etc),
2019-06-19 18:18:21 -07:00
and usually uses less memory (up to 25% more in the worst case). A nice property
2020-02-09 18:32:09 -08:00
is that it does consistently well over a wide range of benchmarks. There is also good huge OS page
2020-01-22 18:16:40 -08:00
support for larger server programs.
2019-06-19 16:26:12 -07:00
2019-06-22 10:40:24 -07:00
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.
2019-06-19 17:00:31 -07:00
2019-06-19 16:52:58 -07:00
Enjoy!
2019-06-19 16:26:12 -07:00
### Releases
2020-05-05 10:45:10 -07:00
* 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 17:59:13 -07:00
* 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 10:19:29 -08:00
* 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 18:53:39 -08:00
* 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 18:32:09 -08:00
* 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,
2020-01-22 18:16:40 -08:00
more eager concurrent free, addition of STL allocator, fixed potential memory leak.
2020-01-16 18:11:29 -08:00
* 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.
2019-12-24 10:38:13 -08:00
* 2019-12-22, `v1.2.2`: stable release 1.2: minor updates.
2019-11-21 16:24:10 -08:00
* 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 12:08:44 -07:00
* 2019-10-07, `v1.1.0`: stable release 1.1.
2019-09-02 10:22:52 -07:00
* 2019-09-01, `v1.0.8`: pre-release 8: more robust windows dynamic overriding, initial huge page support.
2019-10-07 12:08:44 -07:00
* 2019-08-10, `v1.0.6`: pre-release 6: various performance improvements.
2020-04-20 11:37:59 -07:00
Special thanks to:
* Jason Gibson (@jasongibson) for exhaustive testing on large 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.
2019-06-19 16:26:12 -07:00
# Building
## Windows
2019-11-21 16:17:47 -08:00
Open `ide/vs2019/mimalloc.sln` in Visual Studio 2019 and build (or `ide/vs2017/mimalloc.sln`).
2019-06-19 16:26:12 -07:00
The `mimalloc` project builds a static library (in `out/msvc-x64`), while the
`mimalloc-override` project builds a DLL for overriding malloc
in the entire program.
## macOS, Linux, BSD, etc.
2019-06-19 16:26:12 -07:00
We use [`cmake`](https://cmake.org)<sup>1</sup> as the build system:
2019-06-20 09:12:51 -07:00
```
> 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`).
2019-06-19 16:26:12 -07:00
2019-06-20 09:12:51 -07:00
`> sudo make install` (install the library and header files in `/usr/local/lib` and `/usr/local/include`)
2019-06-19 16:26:12 -07:00
You can build the debug version which does many internal checks and
maintains detailed statistics as:
2019-06-20 09:12:51 -07:00
```
> mkdir -p out/debug
> cd out/debug
> cmake -DCMAKE_BUILD_TYPE=Debug ../..
> make
```
This will name the shared library as `libmimalloc-debug.so`.
2019-06-19 16:26:12 -07:00
2019-06-20 07:49:55 -07:00
Finally, you can build a _secure_ version that uses guard pages, encrypted
2019-11-21 16:17:47 -08:00
free lists, etc., as:
2019-06-20 09:12:51 -07:00
```
> mkdir -p out/secure
> cd out/secure
> cmake -DMI_SECURE=ON ../..
2019-06-20 09:12:51 -07:00
> make
```
2019-06-20 07:49:55 -07:00
This will name the shared library as `libmimalloc-secure.so`.
2019-06-19 16:26:12 -07:00
Use `ccmake`<sup>2</sup> instead of `cmake`
to see and customize all the available build options.
Notes:
1. Install CMake: `sudo apt-get install cmake`
2. Install CCMake: `sudo apt-get install cmake-curses-gui`
2019-06-20 07:49:55 -07:00
2019-06-19 16:26:12 -07:00
# 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
2019-06-19 16:26:12 -07:00
```
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:
```
2020-01-16 18:11:29 -08:00
find_package(mimalloc 1.4 REQUIRED)
2019-06-19 16:26:12 -07:00
```
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.
2019-11-21 15:59:54 -08:00
For best performance in C++ programs, it is also recommended to override the
global `new` and `delete` operators. For convience, mimalloc provides
2020-01-16 18:11:29 -08:00
[`mimalloc-new-delete.h`](https://github.com/microsoft/mimalloc/blob/master/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.
2019-06-19 16:26:12 -07:00
You can pass environment variables to print verbose messages (`MIMALLOC_VERBOSE=1`)
and statistics (`MIMALLOC_SHOW_STATS=1`) (in the debug version):
2019-06-19 16:26:12 -07:00
```
> env MIMALLOC_SHOW_STATS=1 ./cfrac 175451865205073170563711388363
2019-06-19 16:26:12 -07:00
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
2020-01-22 18:16:40 -08:00
completely and redirect all calls to the _mimalloc_ library instead .
2019-06-19 16:26:12 -07:00
## Environment Options
2019-06-19 16:26:12 -07:00
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.
- `MIMALLOC_SHOW_ERRORS=1`: show error and warning messages.
2020-04-20 12:37:00 -07:00
- `MIMALLOC_PAGE_RESET=0`: by default, mimalloc will reset (or purge) OS pages that are not in use, to signal to the OS
that the underlying physical memory can be reused. This can reduce memory fragmentation in long running (server)
2020-04-20 12:37:00 -07:00
programs. By setting it to `0` this will no longer be done which can improve performance for batch-like programs.
As an alternative, the `MIMALLOC_RESET_DELAY=`<msecs> can be set higher (100ms by default) to make the page
reset occur less frequently instead of turning it off completely.
2020-04-20 12:37:00 -07:00
- `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_LARGE_OS_PAGES=1`: use large OS pages (2MiB) when available; for some workloads this can significantly
improve performance. Use `MIMALLOC_VERBOSE` to check if the large OS pages are enabled -- usually one needs
2019-09-02 10:48:49 -07:00
to explicitly allow 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
2020-04-20 12:37:00 -07:00
can have fragmented memory (for that reason, we generally recommend to use `MIMALLOC_RESERVE_HUGE_OS_PAGES` instead whenever possible).
2020-01-22 15:21:54 -08:00
<!--
- `MIMALLOC_EAGER_REGION_COMMIT=1`: on Windows, commit large (256MiB) regions eagerly. On Windows, these regions
2019-11-21 15:59:54 -08:00
show in the working set even though usually just a small part is committed to physical memory. This is why it
2020-02-09 18:32:09 -08:00
turned off by default on Windows as it looks not good in the task manager. However, turning it on has no
2020-01-22 15:21:54 -08:00
real drawbacks and may improve performance by a little.
-->
- `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
2019-11-21 15:59:54 -08:00
`MIMALLOC_LARGE_OS_PAGES` in combination with this setting. Just like large OS pages, use with care as reserving
2020-02-09 18:32:09 -08:00
contiguous physical memory can take a long time when memory is fragmented (but reserving the huge pages is done at
2020-01-22 15:21:54 -08:00
startup only once).
2020-04-20 12:37:00 -07:00
Note that we usually need to explicitly enable huge OS pages (as on [Windows][windows-huge] and [Linux][linux-huge])).
With huge OS pages, it may be beneficial to set the setting
`MIMALLOC_EAGER_COMMIT_DELAY=N` (`N` is 1 by default) to delay the initial `N` segments (of 4MiB)
2020-01-16 18:11:29 -08:00
of a thread to not allocate in the huge OS pages; this prevents threads that are short lived
and allocate just a little to take up space in the huge OS page area (which cannot be reset).
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 big 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
2019-06-19 16:26:12 -07:00
## 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),
2020-05-05 10:45:10 -07:00
- 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),
2020-05-05 10:45:10 -07:00
- 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
2020-05-05 10:45:10 -07:00
When _mimalloc_ is built using debug mode, 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.
2020-01-16 18:11:29 -08:00
2019-06-19 16:26:12 -07:00
# Overriding Malloc
Overriding the standard `malloc` can be done either _dynamically_ or _statically_.
## Dynamic override
2019-11-21 15:59:54 -08:00
This is the recommended way to override the standard malloc interface.
2019-06-19 16:26:12 -07:00
2020-01-22 18:16:40 -08:00
### Override on Linux, BSD
2019-06-19 16:26:12 -07:00
On these ELF-based systems we preload the mimalloc shared
2019-06-19 16:26:12 -07:00
library so all calls to the standard `malloc` interface are
resolved to the _mimalloc_ library.
```
> env LD_PRELOAD=/usr/lib/libmimalloc.so myprogram
```
2019-06-19 16:26:12 -07:00
You can set extra environment variables to check that mimalloc is running,
like:
```
> env MIMALLOC_VERBOSE=1 LD_PRELOAD=/usr/lib/libmimalloc.so myprogram
2019-06-19 16:26:12 -07:00
```
or run with the debug version to get detailed statistics:
```
> env MIMALLOC_SHOW_STATS=1 LD_PRELOAD=/usr/lib/libmimalloc-debug.so myprogram
2019-06-19 16:26:12 -07:00
```
2020-01-22 18:16:40 -08:00
### 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_FORCE_FLAT_NAMESPACE=1 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).
2020-02-17 10:19:29 -08:00
(Note: macOS support for dynamic overriding is recent, please report any issues.)
2020-01-22 18:16:40 -08:00
### Override on Windows
2019-06-19 16:26:12 -07:00
2020-04-20 09:56:55 -07:00
<span id="override_on_windows">Overriding on Windows</span> is robust and has the
particular advantage to be able to redirect all malloc/free calls that go through
the (dynamic) C runtime allocator, including those from other DLL's or libraries.
The overriding on Windows requires that you link your program explicitly with
2020-01-16 18:11:29 -08:00
the mimalloc DLL and use the C-runtime library as a DLL (using the `/MD` or `/MDd` switch).
2020-04-20 09:56:55 -07:00
Also, the `mimalloc-redirect.dll` (or `mimalloc-redirect32.dll`) must be available
2019-11-21 15:59:54 -08:00
in the same folder as the main `mimalloc-override.dll` at runtime (as it is a dependency).
The redirection DLL ensures that all calls to the C runtime malloc API get redirected to
mimalloc (in `mimalloc-override.dll`).
2019-09-02 10:48:49 -07:00
To ensure the mimalloc DLL is loaded at run-time it is easiest to insert some
2019-11-21 15:59:54 -08:00
call to the mimalloc API in the `main` function, like `mi_version()`
2019-09-02 10:48:49 -07:00
(or use the `/INCLUDE:mi_version` switch on the linker). See the `mimalloc-override-test` project
2019-11-21 16:17:47 -08:00
for an example on how to use this. For best performance on Windows with C++, it
2020-01-22 18:16:40 -08:00
is also recommended to also override the `new`/`delete` operations (by including
2020-01-16 18:11:29 -08:00
[`mimalloc-new-delete.h`](https://github.com/microsoft/mimalloc/blob/master/include/mimalloc-new-delete.h) a single(!) source file in your project).
2019-06-19 16:26:12 -07:00
2019-09-02 10:48:49 -07:00
The environment variable `MIMALLOC_DISABLE_REDIRECT=1` can be used to disable dynamic
2019-11-21 15:59:54 -08:00
overriding at run-time. Use `MIMALLOC_VERBOSE=1` to check if mimalloc was successfully redirected.
2019-06-19 16:26:12 -07:00
2020-04-20 09:56:55 -07:00
(Note: in principle, it is possible to even patch existing executables without any recompilation
if they are linked with the dynamic C runtime (`ucrtbase.dll`) -- just put the `mimalloc-override.dll`
into the import table (and put `mimalloc-redirect.dll` in the same folder)
2019-11-21 15:59:54 -08:00
Such patching can be done for example with [CFF Explorer](https://ntcore.com/?page_id=388)).
2019-06-19 16:26:12 -07:00
## Static override
On Unix-like systems, you can also statically link with _mimalloc_ to override the standard
2019-06-19 16:26:12 -07:00
malloc interface. The recommended way is to link the final program with the
2019-06-20 07:49:55 -07:00
_mimalloc_ single object file (`mimalloc-override.o`). We use
2019-06-19 16:26:12 -07:00
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-override.o myfile1.c ...
2019-06-19 16:26:12 -07:00
```
2019-11-21 15:59:54 -08:00
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`](https://github.com/microsoft/mimalloc/blob/master/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!
2019-06-19 16:26:12 -07:00
# Performance
2020-01-22 15:05:02 -08:00
Last update: 2020-01-20
2019-06-19 18:07:56 -07:00
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
2020-01-22 15:05:02 -08:00
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.
2019-06-20 07:49:55 -07:00
2020-01-22 15:05:02 -08:00
General memory allocators are interesting as there exists no algorithm that is
2019-06-19 18:07:56 -07:00
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
2020-01-22 15:05:02 -08:00
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 the best
industrial allocators like _jemalloc_ and _tcmalloc_ that are used in some of
the world's largest systems (like Chrome or FreeBSD).
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
2019-06-19 18:07:56 -07:00
as [mimalloc-bench](https://github.com/daanx/mimalloc-bench).
2020-01-22 15:05:02 -08:00
## Benchmark Results on 36-core Intel
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 LibC 2.27 and GCC 7.4.0.
2020-01-22 15:27:47 -08:00
The measured allocators are _mimalloc_ (xmi, tag:v1.4.0, page reset enabled)
2020-01-22 15:05:02 -08:00
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, 2.7.0) (based on _PtMalloc2_).
2020-01-22 15:21:54 -08:00
<img width="90%" src="doc/bench-c5-18xlarge-2020-01-20-a.svg"/>
<img width="90%" src="doc/bench-c5-18xlarge-2020-01-20-b.svg"/>
2020-01-22 15:05:02 -08:00
Any benchmarks ending in `N` run on all processors 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 7%
speedup over _tcmalloc_. This is
2019-06-20 07:49:55 -07:00
quite significant: if Lean spends 20% of its time in the
allocator that means that _mimalloc_ is 1.3&times; faster than _tcmalloc_
2019-06-20 07:58:34 -07:00
here. (This is surprising as that is not measured in a pure
2019-06-20 07:49:55 -07:00
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
2019-06-20 07:58:34 -07:00
in a program as well).
2019-06-20 07:49:55 -07:00
2020-01-22 15:05:02 -08:00
The single threaded _redis_ benchmark again show that most allocators do well on such workloads where _tcmalloc_
did best this time.
2019-06-20 07:49:55 -07:00
2020-01-22 15:05:02 -08:00
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.
2019-06-20 07:49:55 -07:00
2020-01-22 15:05:02 -08:00
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.
2019-06-20 07:49:55 -07:00
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
2020-01-22 15:05:02 -08:00
means that it scales linearly. Here, _tcmalloc_, and
_Hoard_ seem to scale less well and do more than 10% worse on the multi-core version. Even the best industrial
allocators (_tcmalloc_, _jemalloc_, and _tbb_) are more than 10% slower as _mimalloc_ here.
2019-06-20 07:49:55 -07:00
The _sh6bench_ and _sh8bench_ benchmarks are
developed by [MicroQuill](http://www.microquill.com/) as part of SmartHeap.
In _sh6bench_ _mimalloc_ does much
2020-01-22 15:05:02 -08:00
better than the others (more than 1.5&times; faster than _jemalloc_).
2019-06-20 07:49:55 -07:00
We cannot explain this well but believe it is
caused in part by the "reverse" free-ing pattern in _sh6bench_.
2020-01-22 15:05:02 -08:00
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
2019-06-20 07:49:55 -07:00
the _mimalloc_ technique of having non-contended sharded thread free
2020-01-22 15:05:02 -08:00
lists pays off as it outperforms others by a very large margin. Only _rpmalloc_ and _tbb_ also scale well on this benchmark.
2019-06-20 07:49:55 -07:00
2020-01-22 15:05:02 -08:00
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
2019-06-20 07:49:55 -07:00
perform the same, but when running with multiple threads the potential allocator
2020-01-22 15:05:02 -08:00
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
2019-06-20 07:49:55 -07:00
can be avoided with some small implementation changes.
2020-01-22 15:05:02 -08:00
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
2019-06-20 07:49:55 -07:00
how the design of _tbb_ avoids the false cache line sharing.
2020-01-22 15:05:02 -08:00
## On 24-core AMD Epyc
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.
2020-01-22 15:21:54 -08:00
<img width="90%" src="doc/bench-r5a-12xlarge-2020-01-16-a.svg"/>
<img width="90%" src="doc/bench-r5a-12xlarge-2020-01-16-b.svg"/>
2020-01-22 15:05:02 -08:00
## Peak Working Set
The following figure shows the peak working set (rss) of the allocators
on the benchmarks (on the c5.18xlarge instance).
2020-01-22 15:21:54 -08:00
<img width="90%" src="doc/bench-c5-18xlarge-2020-01-20-rss-a.svg"/>
<img width="90%" src="doc/bench-c5-18xlarge-2020-01-20-rss-b.svg"/>
2020-01-22 15:05:02 -08:00
Note that the _xmalloc-testN_ memory usage should be disregarded as it
allocates more the faster the program runs. Similarly, memory usage of
_mstressN_, _rptestN_ and _sh8bench_ can vary depending on scheduling and
speed. Nevertheless, even though _mimalloc_ is fast on these benchmarks we
believe the memory usage is too high and hope to improve.
2019-06-20 07:49:55 -07:00
2019-06-19 16:26:12 -07:00
# 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)
2020-01-22 15:05:02 -08:00
- \[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)
2019-06-19 16:26:12 -07:00
- \[3] D. Grunwald, B. Zorn, and R. Henderson.
_Improving the cache locality of memory allocation_. In R. Cartwright, editor,
2020-01-22 15:05:02 -08:00
Proceedings of the Conference on Programming Language Design and Implementation, pages 177186, New York, NY, USA, June 1993. [pdf](http://citeseer.ist.psu.edu/viewdoc/download?doi=10.1.1.43.6621&rep=rep1&type=pdf)
2019-06-19 16:26:12 -07:00
- \[4] J. Barnes and P. Hut. _A hierarchical O(n*log(n)) force-calculation algorithm_. Nature, 324:446-449, 1986.
2019-06-19 18:07:56 -07:00
- \[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>
2019-06-19 18:07:56 -07:00
2020-01-22 15:05:02 -08:00
- \[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)
2019-06-19 18:07:56 -07:00
- \[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
2020-01-22 15:05:02 -08:00
- \[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,
2019-06-19 18:07:56 -07:00
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, 122135. ACM. 2019.
2020-01-22 15:05:02 -08:00
-->
2019-06-21 09:02:42 -07:00
# 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.