Eric Snow [Mon, 8 Jul 2024 21:10:00 +0000 (15:10 -0600)]
gh-121110: Fix Extension Module Tests Under Py_TRACE_REFS Builds (gh-121503)
The change in gh-118157 (b2cd54a) should have also updated clear_singlephase_extension() but didn't. We fix that here. Note that clear_singlephase_extension() (AKA _PyImport_ClearExtension()) is only used in tests.
Sam Gross [Mon, 8 Jul 2024 18:52:07 +0000 (14:52 -0400)]
gh-121368: Fix seq lock memory ordering in _PyType_Lookup (#121388)
The `_PySeqLock_EndRead` function needs an acquire fence to ensure that
the load of the sequence happens after any loads within the read side
critical section. The missing fence can trigger bugs on macOS arm64.
Additionally, we need a release fence in `_PySeqLock_LockWrite` to
ensure that the sequence update is visible before any modifications to
the cache entry.
Barney Gale [Sat, 6 Jul 2024 16:18:39 +0000 (17:18 +0100)]
GH-73991: Support preserving metadata in `pathlib.Path.copy()` (#120806)
Add *preserve_metadata* keyword-only argument to `pathlib.Path.copy()`, defaulting to false. When set to true, we copy timestamps, permissions, extended attributes and flags where available, like `shutil.copystat()`. The argument has no effect on Windows, where metadata is always copied.
Internally (in the pathlib ABCs), path types gain `_readable_metadata` and `_writable_metadata` attributes. These sets of strings describe what kinds of metadata can be retrieved and stored. We take an intersection of `source._readable_metadata` and `target._writable_metadata` to minimise reads/writes. A new `_read_metadata()` method accepts a set of metadata keys and returns a dict with those keys, and a new `_write_metadata()` method accepts a dict of metadata. We *might* make these public in future, but it's hard to justify while the ABCs are still private.
Update example of str.split, bytes.split (#121287)
In `{str,bytes}.strip(chars)`, multiple characters are not treated as a
prefix/suffix, but as individual characters. This may make users confuse
whether `split` has similar behavior.
Users may incorrectly expect that
`'Good morning, John.'.split(', .') == ['Good', 'morning', 'John']`
gh-121288: Make error message for index() methods consistent (GH-121395)
Make error message for index() methods consistent
Remove the repr of the searched value (which can be arbitrary large)
from ValueError messages for list.index(), range.index(), deque.index(),
deque.remove() and ShareableList.index(). Make the error messages
consistent with error messages for other index() and remove()
methods.
The tracemalloc_tracebacks hash table has traceback keys and NULL
values, but its destructors do not reflect this -- key_destroy_func is
NULL while value_destroy_func is raw_free. Swap these to free the
traceback keys instead.
gh-120754: Update estimated_size in C truncate (#121357)
Sometimes a large file is truncated (test_largefile). While
estimated_size is used as a estimate (the read will stil get the number
of bytes in the file), that it is much larger than the actual size of
data can result in a significant over allocation and sometimes lead to
a MemoryError / running out of memory.
This brings the C implementation to match the Python _pyio
implementation.
gh-120754: Reduce system calls in full-file FileIO.readall() case (#120755)
This reduces the system call count of a simple program[0] that reads all
the `.rst` files in Doc by over 10% (5706 -> 4734 system calls on my
linux system, 5813 -> 4875 on my macOS)
This reduces the number of `fstat()` calls always and seek calls most
the time. Stat was always called twice, once at open (to error early on
directories), and a second time to get the size of the file to be able
to read the whole file in one read. Now the size is cached with the
first call.
The code keeps an optimization that if the user had previously read a
lot of data, the current position is subtracted from the number of bytes
to read. That is somewhat expensive so only do it on larger files,
otherwise just try and read the extra bytes and resize the PyBytes as
needeed.
I built a little test program to validate the behavior + assumptions
around relative costs and then ran it under `strace` to get a log of the
system calls. Full samples below[1].
After the changes, this is everything in one `filename.read_text()`:
This does make some tradeoffs
1. If the file size changes between open() and readall(), this will
still get all the data but might have more read calls.
2. I experimented with avoiding the stat + cached result for small files
in general, but on my dev workstation at least that tended to reduce
performance compared to using the fstat().
[0]
```python3
from pathlib import Path
nlines = []
for filename in Path("cpython/Doc").glob("**/*.rst"):
nlines.append(len(filename.read_text()))
```
gh-117983: Defer import of threading for lazy module loading (#120233)
As noted in gh-117983, the import importlib.util can be triggered at
interpreter startup under some circumstances, so adding threading makes
it a potentially obligatory load.
Lazy loading is not used in the stdlib, so this removes an unnecessary
load for the majority of users and slightly increases the cost of the
first lazily loaded module.
An obligatory threading load breaks gevent, which monkeypatches the
stdlib. Although unsupported, there doesn't seem to be an offsetting
benefit to breaking their use case.
For reference, here are benchmarks for the current main branch:
```
❯ hyperfine -w 8 './python -c "import importlib.util"'
Benchmark 1: ./python -c "import importlib.util"
Time (mean ± σ): 9.7 ms ± 0.7 ms [User: 7.7 ms, System: 1.8 ms]
Range (min … max): 8.4 ms … 13.1 ms 313 runs
```
And with this patch:
```
❯ hyperfine -w 8 './python -c "import importlib.util"'
Benchmark 1: ./python -c "import importlib.util"
Time (mean ± σ): 8.4 ms ± 0.7 ms [User: 6.8 ms, System: 1.4 ms]
Range (min … max): 7.2 ms … 11.7 ms 352 runs
```
Compare to:
```
❯ hyperfine -w 8 './python -c pass'
Benchmark 1: ./python -c pass
Time (mean ± σ): 7.6 ms ± 0.6 ms [User: 5.9 ms, System: 1.6 ms]
Range (min … max): 6.7 ms … 11.3 ms 390 runs
```
This roughly halves the import time of importlib.util.
gh-61103: Support float and long double complex types in ctypes module (#121248)
This amends 6988ff02a5: memory allocation for
stginfo->ffi_type_pointer.elements in PyCSimpleType_init() should be
more generic (perhaps someday fmt->pffi_type->elements will be not a
two-elements array).
It should finally resolve #61103.
Co-authored-by: Victor Stinner <vstinner@python.org> Co-authored-by: Bénédikt Tran <10796600+picnixz@users.noreply.github.com>
Barney Gale [Wed, 3 Jul 2024 03:30:29 +0000 (04:30 +0100)]
GH-73991: Support copying directory symlinks on older Windows (#120807)
Check for `ERROR_INVALID_PARAMETER` when calling `_winapi.CopyFile2()` and
raise `UnsupportedOperation`. In `Path.copy()`, handle this exception and
fall back to the `PathBase.copy()` implementation.
Victor Stinner [Tue, 2 Jul 2024 08:34:13 +0000 (10:34 +0200)]
Move get_signal_name() to test.support (#121251)
* Move get_signal_name() from test.libregrtest to test.support.
* Use get_signal_name() in support.script_helper.
* support.script_helper now decodes stdout and stderr from UTF-8,
instead of ASCII, if a command failed.
Victor Stinner [Mon, 1 Jul 2024 08:30:33 +0000 (10:30 +0200)]
gh-121188: Sanitize invalid XML characters in regrtest (#121195)
When creating the JUnit XML file, regrtest now escapes characters
which are invalid in XML, such as the chr(27) control character used
in ANSI escape sequences.
gh-113565: Improve and harden detection of curses dependencies (#119816)
1. Use pkg-config to check for ncursesw/panelw. If that fails, use
pkg-config to check for ncurses/panel.
2. Regardless of pkg-config output, search for curses/panel headers, so
we're sure we have all defines in pyconfig.h.
3. Regardless of pkg-config output, check if libncurses or libncursesw
contains the 'initscr' symbol; if it does _and_ pkg-config failed
earlier, add the resulting -llib linker option to CURSES_LIBS.
Ditto for 'update_panels' and PANEL_LIBS.
4. Wrap the rest of the checks with WITH_SAVE_ENV and make sure we're
using updated LIBS and CPPFLAGS for those.
gh-87744: fix waitpid race while calling send_signal in asyncio (#121126)
asyncio earlier relied on subprocess module to send signals to the process, this has some drawbacks one being that subprocess module unnecessarily calls waitpid on child processes and hence it races with asyncio implementation which internally uses child watchers. To mitigate this, now asyncio sends signals directly to the process without going through the subprocess on non windows systems. On Windows it fallbacks to subprocess module handling but on windows there are no child watchers so this issue doesn't exists altogether.
Barney Gale [Sat, 29 Jun 2024 15:09:47 +0000 (16:09 +0100)]
GH-119054: Add "Expanding and resolving paths" section to pathlib docs. (#120970)
Add dedicated subsection for `home()`, `expanduser()`, `cwd()`,
`absolute()`, `resolve()` and `readlink()`. The position of this section
keeps all the `Path` constructors (`Path()`, `Path.from_uri()`,
`Path.home()` and `Path.cwd()`) near the top. Within the section, closely
related methods are kept adjacent. Specifically:
-.`home()` and `expanduser()` (the former calls the latter)
- `cwd()` and `absolute()` (the former calls the latter)
- `absolute()` and `resolve()` (both make paths absolute)
- `resolve()` and `readlink()` (both read symlink targets)
- Ditto `cwd()` and `absolute()`
- Ditto `absolute()` and `resolve()`
gh-119372: Recover inf's and zeros in _Py_c_quot (GH-119457)
In some cases, previously computed as (nan+nanj), we could
recover meaningful component values in the result, see
e.g. the C11, Annex G.5.2, routine _Cdivd().
projects / thirdparty / Python / cpython.git / log
