Output with two wheels:
```
❯ GITHUB_ACTIONS=true ./Tools/build/verify_ensurepip_wheels.py
::error file=/Volumes/RAMDisk/cpython/Lib/ensurepip/_bundled/pip-22.0.4-py3-none-any.whl::Found more than one wheel for package pip.
::error file=/Volumes/RAMDisk/cpython/Lib/ensurepip/_bundled/pip-23.2.1-py3-none-any.whl::Found more than one wheel for package pip.
```
Output without wheels:
```
❯ GITHUB_ACTIONS=true ./Tools/build/verify_ensurepip_wheels.py
::error file=::Could not find a pip wheel on disk.
```
(cherry picked from commit f8a047941f2e4a1848700c21d58a08c9ec6a9c68)
Łukasz Langa [Thu, 24 Aug 2023 10:09:11 +0000 (12:09 +0200)]
[3.9] gh-108342: Make ssl TestPreHandshakeClose more reliable (GH-108370) (#108407)
* In preauth tests of test_ssl, explicitly break reference cycles
invoving SingleConnectionTestServerThread to make sure that the
thread is deleted. Otherwise, the test marks the environment as
altered because the threading module sees a "dangling thread"
(SingleConnectionTestServerThread). This test leak was introduced
by the test added for the fix of issue gh-108310.
* Use support.SHORT_TIMEOUT instead of hardcoded 1.0 or 2.0 seconds
timeout.
* SingleConnectionTestServerThread.run() catchs TimeoutError
* Fix a race condition (missing synchronization) in
test_preauth_data_to_tls_client(): the server now waits until the
client connect() completed in call_after_accept().
* test_https_client_non_tls_response_ignored() calls server.join()
explicitly.
* Replace "localhost" with server.listener.getsockname()[0].
(cherry picked from commit 592bacb6fc0833336c0453e818e9b95016e9fd47)
Co-authored-by: Victor Stinner <vstinner@python.org>
[3.9] gh-108342: Break ref cycle in SSLSocket._create() exc (GH-108344) (#108351)
Explicitly break a reference cycle when SSLSocket._create() raises an
exception. Clear the variable storing the exception, since the
exception traceback contains the variables and so creates a reference
cycle.
Łukasz Langa [Tue, 22 Aug 2023 17:57:10 +0000 (19:57 +0200)]
[3.9] gh-108310: Fix CVE-2023-40217: Check for & avoid the ssl pre-close flaw (#108320)
gh-108310: Fix CVE-2023-40217: Check for & avoid the ssl pre-close flaw
Instances of `ssl.SSLSocket` were vulnerable to a bypass of the TLS handshake
and included protections (like certificate verification) and treating sent
unencrypted data as if it were post-handshake TLS encrypted data.
The vulnerability is caused when a socket is connected, data is sent by the
malicious peer and stored in a buffer, and then the malicious peer closes the
socket within a small timing window before the other peers’ TLS handshake can
begin. After this sequence of events the closed socket will not immediately
attempt a TLS handshake due to not being connected but will also allow the
buffered data to be read as if a successful TLS handshake had occurred.
Co-authored-by: Gregory P. Smith [Google LLC] <greg@krypto.org>
[3.9] gh-103142: Upgrade binary builds and CI to OpenSSL 1.1.1u (GH-105174) (GH-105200) (#105205)
Upgrade builds to OpenSSL 1.1.1u.
Also updates _ssl_data_111.h from OpenSSL 1.1.1u, _ssl_data_300.h from 3.0.9.
Manual edits to the _ssl_data_300.h file prevent it from removing any
existing definitions in case those exist in some peoples builds and were
important (avoiding regressions during backporting).
[3.9] gh-102153: Start stripping C0 control and space chars in `urlsplit` (GH-102508) (GH-104575) (GH-104592) (#104593)
gh-102153: Start stripping C0 control and space chars in `urlsplit` (GH-102508)
`urllib.parse.urlsplit` has already been respecting the WHATWG spec a bit GH-25595.
This adds more sanitizing to respect the "Remove any leading C0 control or space from input" [rule](https://url.spec.whatwg.org/GH-url-parsing:~:text=Remove%20any%20leading%20and%20trailing%20C0%20control%20or%20space%20from%20input.) in response to [CVE-2023-24329](https://nvd.nist.gov/vuln/detail/CVE-2023-24329).
I simplified the docs by eliding the state of the world explanatory
paragraph in this security release only backport. (people will see
that in the mainline /3/ docs)
[3.9] gh-104049: do not expose on-disk location from SimpleHTTPRequestHandler (GH-104067) (#104120)
Do not expose the local server's on-disk location from `SimpleHTTPRequestHandler` when generating a directory index. (unnecessary information disclosure)
[3.9] gh-100001: Omit control characters in http.server stderr logs. (GH-100002) (#100032)
* gh-100001: Omit control characters in http.server stderr logs. (GH-100002)
Replace control characters in http.server.BaseHTTPRequestHandler.log_message with an escaped \xHH sequence to avoid causing problems for the terminal the output is printed to.
(cherry picked from commit d8ab0a4dfa48f881b4ac9ab857d2e9de42f72828)
Co-authored-by: Gregory P. Smith <greg@krypto.org>
* also escape \s (backport of PR #100038).
* add versionadded and remove extra 'to'
Co-authored-by: Gregory P. Smith <greg@krypto.org>
[3.9] gh-97514: Don't use Linux abstract sockets for multiprocessing (GH-98501) (#98504)
Linux abstract sockets are insecure as they lack any form of filesystem
permissions so their use allows anyone on the system to inject code into
the process.
This removes the default preference for abstract sockets in
multiprocessing introduced in Python 3.9+ via
https://github.com/python/cpython/pull/18866 while fixing
https://github.com/python/cpython/issues/84031.
Explicit use of an abstract socket by a user now generates a
RuntimeWarning. If we choose to keep this warning, it should be
backported to the 3.7 and 3.8 branches.
(cherry picked from commit 49f61068f49747164988ffc5a442d2a63874fc17)
Co-authored-by: Gregory P. Smith <greg@krypto.org>
[3.9] gh-98517: Fix buffer overflows in _sha3 module (GH-98519) (#98526)
This is a port of the applicable part of XKCP's fix [1] for
CVE-2022-37454 and avoids the segmentation fault and the infinite
loop in the test cases published in [2].
[3.9] gh-96710: Make the test timing more lenient for the int/str DoS regression test. (GH-96717) (#98196)
gh-96710: Make the test timing more lenient for the int/str DoS regression test. (GH-96717)
A regression would still absolutely fail and even a flaky pass isn't
harmful as it'd fail most of the time across our N system test runs.
Windows has a low resolution timer and CI systems are prone to odd
timing so this just gives more leeway to avoid flakiness.
(cherry picked from commit 11e3548fd1d3445ccde971d613633b58d73c3016)
Co-authored-by: Gregory P. Smith <greg@krypto.org>
[3.9] gh-97897: Prevent os.mkfifo and os.mknod segfaults with macOS 13 SDK (GH-97944) (#97968)
The macOS 13 SDK includes support for the `mkfifoat` and `mknodat` system calls.
Using the `dir_fd` option with either `os.mkfifo` or `os.mknod` could result in a
segfault if cpython is built with the macOS 13 SDK but run on an earlier
version of macOS. Prevent this by adding runtime support for detection of
these system calls ("weaklinking") as is done for other newer syscalls on
macOS.
(cherry picked from commit 6d0a0191a4e5477bd843e62c24d7f3bcad4fd5fc)
Fix command line parsing: reject "-X int_max_str_digits" option with
no value (invalid) when the PYTHONINTMAXSTRDIGITS environment
variable is set to a valid limit.
(cherry picked from commit 41351662bcd21672d8ccfa62fe44d72027e6bcf8)
Co-authored-by: Victor Stinner <vstinner@python.org>
[3.9] gh-97616: list_resize() checks for integer overflow (GH-97617) (GH-97627)
gh-97616: list_resize() checks for integer overflow (GH-97617)
Fix multiplying a list by an integer (list *= int): detect the
integer overflow when the new allocated length is close to the
maximum size. Issue reported by Jordan Limor.
list_resize() now checks for integer overflow before multiplying the
new allocated length by the list item size (sizeof(PyObject*)).
(cherry picked from commit a5f092f3c469b674b8d9ccbd4e4377230c9ac7cf)
Co-authored-by: Victor Stinner <vstinner@python.org>
[3.9] gh-97612: Fix shell injection in get-remote-certificate.py (GH-97613) (GH-97632)
gh-97612: Fix shell injection in get-remote-certificate.py (GH-97613)
Fix a shell code injection vulnerability in the
get-remote-certificate.py example script. The script no longer uses a
shell to run "openssl" commands. Issue reported and initial fix by
Caleb Shortt.
Remove the Windows code path to send "quit" on stdin to the "openssl
s_client" command: use DEVNULL on all platforms instead.
This documents the behavior that has always been the case since timeout
support was introduced in Python 3.3.
(cherry picked from commit b05dd796492160c37c9e15e3882f699f411b3461)
Co-authored-by: Gregory P. Smith <greg@krypto.org>
[3.9] gh-95778: CVE-2020-10735: Prevent DoS by very large int() (#96502)
* Correctly pre-check for int-to-str conversion (#96537)
Converting a large enough `int` to a decimal string raises `ValueError` as expected. However, the raise comes _after_ the quadratic-time base-conversion algorithm has run to completion. For effective DOS prevention, we need some kind of check before entering the quadratic-time loop. Oops! =)
The quick fix: essentially we catch _most_ values that exceed the threshold up front. Those that slip through will still be on the small side (read: sufficiently fast), and will get caught by the existing check so that the limit remains exact.
The justification for the current check. The C code check is:
```c
max_str_digits / (3 * PyLong_SHIFT) <= (size_a - 11) / 10
```
In GitHub markdown math-speak, writing $M$ for `max_str_digits`, $L$ for `PyLong_SHIFT` and $s$ for `size_a`, that check is:
$$\left\lfloor\frac{M}{3L}\right\rfloor \le \left\lfloor\frac{s - 11}{10}\right\rfloor$$
From this it follows that
$$\frac{M}{3L} < \frac{s-1}{10}$$
hence that
$$\frac{L(s-1)}{M} > \frac{10}{3} > \log_2(10).$$
So
$$2^{L(s-1)} > 10^M.$$
But our input integer $a$ satisfies $|a| \ge 2^{L(s-1)}$, so $|a|$ is larger than $10^M$. This shows that we don't accidentally capture anything _below_ the intended limit in the check.
Co-authored-by: Gregory P. Smith [Google LLC] <greg@krypto.org> Co-authored-by: Christian Heimes <christian@python.org> Co-authored-by: Mark Dickinson <dickinsm@gmail.com>
gh-94821: Fix autobind of empty unix domain address (GH-94826) (GH-94875)
When binding a unix socket to an empty address on Linux, the socket is
automatically bound to an available address in the abstract namespace.
>>> s = socket.socket(socket.AF_UNIX, socket.SOCK_STREAM)
>>> s.bind("")
>>> s.getsockname()
b'\x0075499'
Since python 3.9, the socket is bound to the one address:
>>> s.getsockname()
b'\x00'
And trying to bind multiple sockets will fail with:
Traceback (most recent call last):
File "/home/nsoffer/src/cpython/Lib/test/test_socket.py", line 5553, in testAutobind
s2.bind("")
OSError: [Errno 98] Address already in use
Added 2 tests:
- Auto binding empty address on Linux
- Failing to bind an empty address on other platforms
gh-91172: Create a workflow for verifying bundled pip and setuptools (GH-31885) (GH-94123)
Co-authored-by: Hugo van Kemenade <hugovk@users.noreply.github.com> Co-authored-by: Adam Turner <9087854+AA-Turner@users.noreply.github.com>
(cherry picked from commit d36954b7ead06daead3dcf9b0dd9f8002eab508f)
gh-87389: Fix an open redirection vulnerability in http.server. (GH-93879) (GH-94093)
Fix an open redirection vulnerability in the `http.server` module when
an URI path starts with `//` that could produce a 301 Location header
with a misleading target. Vulnerability discovered, and logic fix
proposed, by Hamza Avvan (@hamzaavvan).
gh-91810: Fix regression with writing an XML declaration with encoding='unicode' (GH-93426) (GH-93791)
Suppress writing an XML declaration in open files in ElementTree.write()
with encoding='unicode' and xml_declaration=None.
If file patch is passed to ElementTree.write() with encoding='unicode',
always open a new file in UTF-8.
(cherry picked from commit d7db9dc3cc5b44d0b4ce000571fecf58089a01ec)
The old behavior still raised an error in a similar way, but only
because subsequent calculations happened to fail as well. Better to fail
fast.
This also refactors the tests to split out the `fromtimestamp` and
`utcfromtimestamp` tests, and to get us closer to the actual desired
limits of the functions. As part of this, we also changed the way we
detect platforms where the same limits don't necessarily apply (e.g.
Windows).
As part of refactoring the tests to hit this condition explicitly (even
though the user-facing behvior doesn't change in any way we plan to
guarantee), I noticed that there was a difference in the places that
`datetime.utcfromtimestamp` fails in the C and pure Python versions, which
was fixed by skipping the "probe for fold" logic for UTC specifically —
since UTC doesn't have any folds or gaps, we were never going to find a
fold value anyway. This should prevent some failures in the pure python
`utcfromtimestamp` method on timestamps close to 0001-01-01.
There are two separate news entries for this because one is a
potentially user-facing change, the other is an internal code
correctness change that, if anything, changes some error messages. The
two happen to be coupled because of the test refactoring, but they are
probably best thought of as independent changes.
gh-92530: Fix an issue that occurred after interrupting threading.Condition.notify (GH-92534) (GH-92831)
If Condition.notify() was interrupted just after it released the waiter lock,
but before removing it from the queue, the following calls of notify() failed
with RuntimeError: cannot release un-acquired lock.
(cherry picked from commit 70af994fee7c0850ae859727d9468a5f29375a38)
projects / thirdparty / Python / cpython.git / log
