Apply BOLT optimizations to libpython for shared builds. Most of the C
code is in libpython so it is critical to apply BOLT there fully realize
BOLT benefits.
This change also reworks how BOLT instrumentation is applied. It
effectively removes the readelf based logic added in gh-101525 and
replaces it with a mechanism that saves a copy of the pre-bolt binary
and restores that copy when necessary. This allows us to perform BOLT
optimizations without having to manually delete the output binary to
force a new bolt run.
Also:
- add a clean-bolt target for purging BOLT files and hook that up to the
clean target
- .gitignore BOLT related files
Before and after this refactor, `make` will no-op after a previous run.
Both versions should also share common make DAG deficiencies where
targets fail to trigger as often as they need to or can trigger
prematurely in certain scenarios. e.g. after this change you may need to
`rm profile-bolt-stamp` to force a BOLT run because there aren't
appropriate non-phony targets for BOLT's make target to depend on.
To make it easier to iterate on custom BOLT settings, the flags to pass
to instrumentation and application are now defined in configure and can
be overridden by passing BOLT_INSTRUMENT_FLAGS and BOLT_APPLY_FLAGS.
This merges their code. They're backed by the same single HACL* static library, having them be a single module simplifies maintenance.
This should unbreak the wasm enscripten builds that currently fail due to linking in --whole-archive mode and the HACL* library appearing twice.
Long unnoticed error fixed: _sha512.SHA384Type was doubly assigned and was actually SHA512Type. Nobody depends on those internal names.
Also rename LIBHACL_ make vars to LIBHACL_SHA2_ in preperation for other future HACL things.
Replace the builtin hashlib implementations of SHA2-384 and SHA2-512
originally from LibTomCrypt with formally verified, side-channel resistant
code from the [HACL*](https://github.com/hacl-star/hacl-star/) project.
The builtins remain a fallback only used when OpenSSL does not provide them.
replacing hashlib primitives (for the non-OpenSSL case) with verified implementations from HACL*. This is the first PR in the series, and focuses specifically on SHA2-256 and SHA2-224.
This PR imports Hacl_Streaming_SHA2 into the Python tree. This is the HACL* implementation of SHA2, which combines a core implementation of SHA2 along with a layer of buffer management that allows updating the digest with any number of bytes. This supersedes the previous implementation in the tree.
@franziskuskiefer was kind enough to benchmark the changes: in addition to being verified (thus providing significant safety and security improvements), this implementation also provides a sizeable performance boost!
```
---------------------------------------------------------------
Benchmark Time CPU Iterations
---------------------------------------------------------------
Sha2_256_Streaming 3163 ns 3160 ns 219353 // this PR
LibTomCrypt_Sha2_256 5057 ns 5056 ns 136234 // library used by Python currently
```
The changes in this PR are as follows:
- import the subset of HACL* that covers SHA2-256/224 into `Modules/_hacl`
- rewire sha256module.c to use the HACL* implementation
Co-authored-by: Gregory P. Smith [Google LLC] <greg@krypto.org>
Co-authored-by: Erlend E. Aasland <erlend.aasland@protonmail.com>
Fix the gdbm_compat library detection logic to actually check for
-lgdbm_compat independently of the ndbm detection.
This fixes the build failure with `--with-dbmliborder=gdbm`,
and implicit fallback to ndbm with the default value.
When cross-compiling, the compile/run test for -pthread always fails so -pthread
will never be automatically set without an override from the cache. ac_cv_pthread
can already be overridden, so do the same thing for ac_cv_cxx_thread.
It has had no effect on non-macOS platforms for a long time, and has had
the non-obvious effect of invoking `pkg_config` and not setting
`-DUSING_APPLE_OS_LIBFFI` on macOS since GH-22855.
This got introduced in commit 5884449539
to determine if readline is already linked against curses or tinfo in
the setup.py, which is no longer present.
Added os.setns and os.unshare to easily switch between namespaces
on Linux.
Co-authored-by: Christian Heimes <christian@python.org>
Co-authored-by: CAM Gerlach <CAM.Gerlach@Gerlach.CAM>
Co-authored-by: Victor Stinner <vstinner@python.org>
⚠️⚠️ Note for reviewers, hackers and fellow systems/low-level/compiler engineers ⚠️⚠️
If you have a lot of experience with this kind of shenanigans and want to improve the **first** version, **please make a PR against my branch** or **reach out by email** or **suggest code changes directly on GitHub**.
If you have any **refinements or optimizations** please, wait until the first version is merged before starting hacking or proposing those so we can keep this PR productive.
- support EMSDK tot-upstream and git releases
- allow WASM assents for wasm64-emscripten and WASI. This makes single file distributions on WASI easier.
- decouple WASM assets from browser builds
* Add support for the BOLT post-link binary optimizer
Using [bolt](https://github.com/llvm/llvm-project/tree/main/bolt)
provides a fairly large speedup without any code or functionality
changes. It provides roughly a 1% speedup on pyperformance, and a
4% improvement on the Pyston web macrobenchmarks.
It is gated behind an `--enable-bolt` configure arg because not all
toolchains and environments are supported. It has been tested on a
Linux x86_64 toolchain, using llvm-bolt built from the LLVM 14.0.6
sources (their binary distribution of this version did not include bolt).
Compared to [a previous attempt](https://github.com/faster-cpython/ideas/issues/224),
this commit uses bolt's preferred "instrumentation" approach, as well as adds some non-PIE
flags which enable much better optimizations from bolt.
The effects of this change are a bit more dependent on CPU microarchitecture
than other changes, since it optimizes i-cache behavior which seems
to be a bit more variable between architectures. The 1%/4% numbers
were collected on an Intel Skylake CPU, and on an AMD Zen 3 CPU I
got a slightly larger speedup (2%/4%), and on a c6i.xlarge EC2 instance
I got a slightly lower speedup (1%/3%).
The low speedup on pyperformance is not entirely unexpected, because
BOLT improves i-cache behavior, and the benchmarks in the pyperformance
suite are small and tend to fit in i-cache.
This change uses the existing pgo profiling task (`python -m test --pgo`),
though I was able to measure about a 1% macrobenchmark improvement by
using the macrobenchmarks as the training task. I personally think that
both the PGO and BOLT tasks should be updated to use macrobenchmarks,
but for the sake of splitting up the work this PR uses the existing pgo task.
* Simplify the build flags
* Add a NEWS entry
* Update Makefile.pre.in
Co-authored-by: Dong-hee Na <donghee.na92@gmail.com>
* Update configure.ac
Co-authored-by: Dong-hee Na <donghee.na92@gmail.com>
* Add myself to ACKS
* Add docs
* Other review comments
* fix tab/space issue
* Make it more clear that --enable-bolt is experimental
* Add link to bolt's github page
Co-authored-by: Dong-hee Na <donghee.na92@gmail.com>
Automate WASM build with a new Python script. The script provides
several build profiles with configure flags for Emscripten flavors
and WASI. The script can detect and use Emscripten SDK and WASI SDK from
default locations or env vars.
``configure`` now detects Node arguments and creates HOSTRUNNER
arguments for Node 16. It also sets correct arguments for
``wasm64-emscripten``.
Co-authored-by: Brett Cannon <brett@python.org>
Remove the "configure --with-cxx-main" build option: it didn't work
for many years. Remove the MAINCC variable from configure and
Makefile.
The MAINCC variable was added by the issue gh-42471: commit
0f48d98b74. Previously, --with-cxx-main
was named --with-cxx.
Keep CXX and LDCXXSHARED variables, even if they are no longer used
by Python build system.
- check for ``dup()`` libc function
- handle missing ``F_DUPFD`` in ``dup2()`` replacement function
- add workaround for WASI libc bug in MSG_TRUNC
- ESHUTDOWN is missing, use EPIPE instead
- POLLPRI is missing, define as 0 (no-op)
It makes it easier to look for module states in sysconfig without
special casing suffixes "_CFLAGS", "_DEPS", "_LDFLAGS", "_OBJS",
and "CTYPES_MALLOC_CLOSURE".
- c_longlong and c_longdouble need experimental WASM bigint.
- Skip tests that need threading
- Define ``CTYPES_MAX_ARGCOUNT`` for Emscripten. libffi-emscripten 2022-06-23 supports up to 1000 args.
- Mark more ``umask()`` cases
- ``dup()`` is not supported
- ``/dev/null`` is not available
- document missing features
- mark more modules as not available
This was added for bpo-40514 (gh-84694) to test out a per-interpreter GIL. However, it has since proven unnecessary to keep the experiment in the repo. (It can be done as a branch in a fork like normal.) So here we are removing:
* the configure option
* the macro
* the code enabled by the macro
Python now always use the ``%zu`` and ``%zd`` printf formats to
format a size_t or Py_ssize_t number. Building Python 3.12 requires a
C11 compiler, so these printf formats are now always supported.
* PyObject_Print() and _PyObject_Dump() now use the printf %zd format
to display an object reference count.
* Update PY_FORMAT_SIZE_T comment.
* Remove outdated notes about the %zd format in PyBytes_FromFormat()
and PyUnicode_FromFormat() documentations.
* configure no longer checks for the %zd format and no longer defines
PY_FORMAT_SIZE_T macro in pyconfig.h.
* pymacconfig.h no longer undefines PY_FORMAT_SIZE_T: macOS 10.4 is
no longer supported. Python 3.12 now requires macOS 10.6 (Snow
Leopard) or newer.
This makes macOS gdbm provided by Homebrew not segfault through correct
selection of the linked library (-lgdbm_compat) *AND* the correct ndbm-style
header (gdbm-ndbm.h instead of the invalid ndbm.h).
`HOSTRUNNER` is a program which can be used to run `BUILDPYTHON` for the host platform (for example, `python.js` requires `node`).
Also change depedencies from `build_all` to `all` so that targets which can't build everything (e.g. WASM) can still run `buildbottest` and `pythoninfo`.
cc @tiran
- drop unnecessary ``=1`` suffix from Emscripten flags
- drop unnecessary ``-sWASM`` flag for side modules
- rename ``build_platform`` to ``build_wasm``. I introduced the target
for WASM builds a couple of months ago.
- fix ``--enable-test-modules`` for browser builds
- Remove ``--with-tclk-*`` options from `configure`
- Use pkg-config to detect `_tkinter` dependencies (Tcl/Tk, X11)
- Manual override via environment variables `TCLTK_CFLAGS` and `TCLTK_LIBS`
Shared extension on Emscripten now have suffix
``.cpython-311-wasm32-emscripten.so`` (JS loader) and
``.cpython-311-wasm32-emscripten.wasm`` (WebAssembly code).
- Add requires_fork and requires_subprocess to more tests
- Skip extension import tests if dlopen is not available
- Don't assume that _testcapi is a shared extension
- Skip a lot of socket tests that don't work on Emscripten
- Skip mmap tests, mmap emulation is incomplete
- venv does not work yet
- Cannot get libc from executable
The "entire" test suite is now passing on Emscripten with EMSDK from git head (91 suites are skipped).
In Linux kernel 5.14 one can dynamically request size of altstacksize
based on hardware capabilities with getauxval(AT_MINSIGSTKSZ).
This changes allows for Python extension's request to Linux kernel
to use AMX_TILE instruction set on Sapphire Rapids Xeon processor
to succeed, unblocking use of the ISA in frameworks.
Introduced HAVE_LINUX_AUXVEC_H in configure.ac and pyconfig.h.in
Used cpython_autoconf:269 docker container to generate configure.
The heredoc creation statements use >> to append conftest.c. This can cause
tricky build issues if the file is not correctly removed prior to its
name being reused (such name is reused several times for different
contextual tests during the build). One such result from appending may
cause #include <ac_nonexistent.h> to persist when testing to acquire
PLATFORM_TRIPLET. This can then lead to downstream issues concerning SOABI.
musl libc and gnu libc are not ABI compatible so we need set different
SOABI for musl and not simply assume that all linux is linux-gnu.
Replace linux-gnu with the detected os for the build from config.guess
for linux-musl*.
When Python is built without --enable-shared, the "python" program is
now linked to object files, rather than being linked to the Python
library (libpython.a), to make sure that all symbols are exported.
Previously, the linker omitted some symbols like the Py_FrozenMain()
function.
When Python is configured with --without-static-libpython, the Python
static library (libpython.a) is no longer built.
* Check --without-static-libpython earlier in configure.ac
* Add LINK_PYTHON_OBJS and LINK_PYTHON_DEPS variables to Makefile.
* test_capi now ensures that the "Py_FrozenMain" symbol is exported.
The build system now uses a :program:`_bootstrap_python` interpreter for
freezing and deepfreezing again. To speed up build process the build tools
:program:`_bootstrap_python` and :program:`_freeze_module` are no longer
build with LTO.
Cross building depends on a build Python interpreter, which must have same
version and bytecode as target host Python.
Victor Stinner
Dong-hee Na
T. Wouters
Gregory Szorc
Erlend E. Aasland
Zhang Na
Petr Viktorin
Nikita Sobolev
Matthias Görgens
Jonathan Protzenko
Eli Schwartz
Gregory P. Smith
Eric Snow
Kushal Das
Michał Górny
Dustin Spicuzza
Zachary Ware
Sam James
colorfulappl
Miro Hrončok
Noam Cohen
Sam James
serge-sans-paille
Jelle Zijlstra
Pablo Galindo Salgado
Christian Heimes
Kevin Modzelewski
Illia Volochii
Kumar Aditya
Soumendra Ganguly
Łukasz Langa
Ethan Smith
Thomas Klausner
jonasdlindner
Erlend Egeberg Aasland
Serhiy Storchaka
Hood Chatham
Oleksandr Pavlyk
Brett Cannon
adanhawth
Natanael Copa
Mark Dickinson
Mark Shannon
Ned Deily