mpage [Thu, 21 Nov 2024 19:22:21 +0000 (11:22 -0800)]
gh-115999: Specialize `LOAD_GLOBAL` in free-threaded builds (#126607)
Enable specialization of LOAD_GLOBAL in free-threaded builds.
Thread-safety of specialization in free-threaded builds is provided by the following:
A critical section is held on both the globals and builtins objects during specialization. This ensures we get an atomic view of both builtins and globals during specialization.
Generation of new keys versions is made atomic in free-threaded builds.
Existing helpers are used to atomically modify the opcode.
Thread-safety of specialized instructions in free-threaded builds is provided by the following:
Relaxed atomics are used when loading and storing dict keys versions. This avoids potential data races as the dict keys versions are read without holding the dictionary's per-object lock in version guards.
Dicts keys objects are passed from keys version guards to the downstream uops. This ensures that we are loading from the correct offset in the keys object. Once a unicode key has been stored in a keys object for a combined dictionary in free-threaded builds, the offset that it is stored in will never be reused for a different key. Once the version guard passes, we know that we are reading from the correct offset.
The dictionary read fast-path is used to read values from the dictionary once we know the correct offset.
Eric Snow [Thu, 21 Nov 2024 18:08:38 +0000 (11:08 -0700)]
gh-114940: Add _Py_FOR_EACH_TSTATE_UNLOCKED(), and Friends (gh-127077)
This is a precursor to the actual fix for gh-114940, where we will change these macros to use the new lock. This change is almost entirely mechanical; the exceptions are the loops in codeobject.c and ceval.c, which now hold the "head" lock. Note that almost all of the uses of _Py_FOR_EACH_TSTATE_UNLOCKED() here will change to _Py_FOR_EACH_TSTATE_BEGIN() once we add the new per-interpreter lock.
Serhiy Storchaka [Thu, 21 Nov 2024 11:15:12 +0000 (13:15 +0200)]
gh-126997: Fix support of non-ASCII strings in pickletools (GH-127062)
* Fix support of STRING and GLOBAL opcodes with non-ASCII arguments.
* dis() now outputs non-ASCII bytes in STRING, BINSTRING and
SHORT_BINSTRING arguments as escaped (\xXX).
Cody Maloney [Thu, 21 Nov 2024 09:33:12 +0000 (01:33 -0800)]
gh-127076: Disable strace tests under LD_PRELOAD (#127086)
Distribution tooling (ex. sandbox on Gentoo and fakeroot on Debian) uses
LD_PRELOAD to intercept system calls and potentially modify them when
building. These tools can change the set of system calls, so disable
system call testing under these cases.
mpage [Wed, 20 Nov 2024 22:54:48 +0000 (14:54 -0800)]
gh-115999: Don't take a reason in unspecialize (#127030)
Don't take a reason in unspecialize
We only want to compute the reason if stats are enabled. Optimizing
compilers should optimize this away for us (gcc and clang do), but
it's better to be safe than sorry.
Gregory P. Smith [Wed, 20 Nov 2024 16:18:58 +0000 (08:18 -0800)]
gh-97514: Authenticate the forkserver control socket. (GH-99309)
This adds authentication to the forkserver control socket. In the past only filesystem permissions protected this socket from code injection into the forkserver process by limiting access to the same UID, which didn't exist when Linux abstract namespace sockets were used (see issue) meaning that any process in the same system network namespace could inject code. We've since stopped using abstract namespace sockets by default, but protecting our control sockets regardless of type is a good idea.
This reuses the HMAC based shared key auth already used by `multiprocessing.connection` sockets for other purposes.
Doing this is useful so that filesystem permissions are not relied upon and trust isn't implied by default between all processes running as the same UID with access to the unix socket.
### pyperformance benchmarks
No significant changes. Including `concurrent_imap` which exercises `multiprocessing.Pool.imap` in that suite.
### Microbenchmarks
This does _slightly_ slow down forkserver use. How much so appears to depend on the platform. Modern platforms and simple platforms are less impacted. This PR adds additional IPC round trips to the control socket to tell forkserver to spawn a new process. Systems with potentially high latency IPC are naturally impacted more.
Typically a 1-4% slowdown on a very targeted process creation microbenchmark, with a worst case overloaded system slowdown of 20%. No evidence that these slowdowns appear in practical sense. See the PR for details.
Barney Gale [Tue, 19 Nov 2024 21:19:30 +0000 (21:19 +0000)]
GH-85168: Use filesystem encoding when converting to/from `file` URIs (#126852)
Adjust `urllib.request.url2pathname()` and `pathname2url()` to use the
filesystem encoding when quoting and unquoting file URIs, rather than
forcing use of UTF-8.
No changes are needed in the `nturl2path` module because Windows always
uses UTF-8, per PEP 529.
Eric Snow [Tue, 19 Nov 2024 19:59:19 +0000 (12:59 -0700)]
gh-126914: Store the Preallocated Thread State's Pointer in a PyInterpreterState Field (gh-126989)
This approach eliminates the originally reported race. It also gets rid of the deadlock reported in gh-96071, so we can remove the workaround added then.
Serhiy Storchaka [Mon, 18 Nov 2024 17:45:25 +0000 (19:45 +0200)]
gh-123803: Support arbitrary code page encodings on Windows (GH-123804)
If the cpXXX encoding is not directly implemented in Python, fall back
to use the Windows-specific API codecs.code_page_encode() and
codecs.code_page_decode().
Stan U. [Mon, 18 Nov 2024 15:29:14 +0000 (15:29 +0000)]
gh-126911: Update credits output (#126913)
Co-authored-by: Petr Viktorin <encukou@gmail.com> Co-authored-by: blurb-it[bot] <43283697+blurb-it[bot]@users.noreply.github.com> Co-authored-by: Terry Jan Reedy <tjreedy@udel.edu> Co-authored-by: Hugo van Kemenade <1324225+hugovk@users.noreply.github.com>
Michał Górny [Mon, 18 Nov 2024 13:05:55 +0000 (13:05 +0000)]
gh-126909: Fix running xattr tests on systems with lower limits (#126930)
Modify the extended attribute tests to write fewer and smaller extended
attributes, in order to fit within filesystems with total xattr limit
of 1 KiB (e.g. ext4 with 1 KiB blocks). Previously, the test would
write over 2 KiB, making it fail with ENOSPC on such systems.
Serhiy Storchaka [Mon, 18 Nov 2024 09:53:45 +0000 (11:53 +0200)]
gh-67877: Fix memory leaks in terminated RE matching (GH-126840)
If SRE(match) function terminates abruptly, either because of a signal
or because memory allocation fails, allocated SRE_REPEAT blocks might
be never released.
RanKKI [Sat, 16 Nov 2024 23:01:52 +0000 (10:01 +1100)]
gh-124452: Fix header mismatches when folding/unfolding with email message (#125919)
The header-folder of the new email API has a long standing known buglet where
if the first token is longer than max_line_length, it puts that token on the next
line. It turns out there is also a *parsing* bug when parsing such a header:
the space prefixing that first, non-empty line gets preserved and tacked on to
the start of the header value, which is not the expected behavior per the RFCs.
The bug arises from the fact that the parser assumed that there would be at
least one token on the line with the header, which is going to be true for
probably every email producer other than the python email library with its
folding buglet. Clearly, though, this is a case that needs to be handled
correctly. The fix is simple: strip the blanks off the start of the whole
value, not just the first physical line of the value.
This unifies the code for nodejs and the code for the browser. After this
commit, the browser example doesn't work; this will be fixed in a
subsequent update.
Alex Gaynor [Fri, 15 Nov 2024 23:09:05 +0000 (18:09 -0500)]
Added a warning to the urljoin docs, indicating that it is not safe to use with attacker controlled URLs (GH-126659)
This was flagged to me at a party today by someone who works in red-teaming as a frequently encountered footgun. Documenting the potentially unexpected behavior seemed like a good place to start.
Petr Viktorin [Fri, 15 Nov 2024 12:48:57 +0000 (13:48 +0100)]
gh-125063: marshal: Add version 5, improve documentation (GH-126829)
* Document that slices can be marshalled
* Deduplicate and organize the list of supported types
in docs
* Organize the type code list in marshal.c, to make
it more obvious that this is a versioned format
* Back-fill some historical info
Co-authored-by: Michael Droettboom <mdboom@gmail.com>
For dlsym(), a return value of NULL does not necessarily indicate
an error [1].
Therefore, to avoid using stale (or NULL) dlerror() values, we must:
1. clear the previous error state by calling dlerror()
2. call dlsym()
3. call dlerror()
If the return value of dlerror() is not NULL, an error occured.
In ctypes we choose to treat a NULL return value from dlsym()
as a "not found" error. This is the same as the fallback
message we use on Windows, Cygwin or when getting/formatting
the error reason fails.
Discard two leading slashes from the beginning of a `file:` URI if they
introduce an empty authority section. As a result, file URIs like
`///etc/hosts` are correctly parsed as `/etc/hosts`.
Petr Viktorin [Thu, 14 Nov 2024 08:31:14 +0000 (09:31 +0100)]
gh-123832: Adjust `socket.getaddrinfo` docs for better POSIX compliance (GH-126182)
* gh-123832: Adjust `socket.getaddrinfo` docs for better POSIX compliance
This changes nothing changes for CPython supported platforms,
but hints how to deal with platforms that stick to the letter of
the spec.
It also marks `socket.getaddrinfo` as a wrapper around `getaddrinfo(3)`;
specifically, workarounds to make the function work consistently across
platforms are out of scope in its code.
Include wording similar to the POSIX's “by providing options and by
limiting the returned information”, which IMO suggests that the
hints limit the resulting list compared to the defaults, *but* can
be interpreted differently. Details are added in a note.
Specifically say that this wraps the underlying C function. So, the
details are in OS docs. The “full range of results” bit goes away.
Use `AF_UNSPEC` rather than zero for the *family* default, although
I don't think a system where it's nonzero would be very usable.
Suggest setting proto and/or type (with examples, as the appropriate
values aren't obvious). Say why you probably want to do that that
on all systems; mention the behavior on the “letter of the spec”
systems.
Suggest that the results should be tried in order, which is,
AFAIK best practice -- see RFC 6724 section 2, and its predecessor
from 2003 (which are specific to IP, but indicate how people use this):
> Well-behaved applications SHOULD iterate through the list of
> addresses returned from `getaddrinfo()` until they find a working address.
Co-authored-by: Carol Willing <carolcode@willingconsulting.com>
One of the most common reasons I see the old `pipes` module still in use
when porting to Python 3.13 is for the undocumented `pipes.quote`
function, which can easily be replaced with `shlex.quote`. I think it's
worth specifically calling this out, since being directed to the
`subprocess` module would be confusing in this case.