summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* 00351-cve-2019-20907-fix-infinite-loop-in-tarfile.patchfedora-2.7.18-5Rishi2020-09-294-0/+10
| | | | | | | 00351 # Avoid infinite loop when reading specially crafted TAR files using the tarfile module (CVE-2019-20907). See: https://bugs.python.org/issue39017
* 00289-disable-nis-detection.patchFedora Python maintainers2020-09-291-50/+1
| | | | | | 00289 # Disable automatic detection for the nis module (we handle it it in Setup.dist, see Patch0)
* 00193-enable-loading-sqlite-extensions.patchFedora Python maintainers2020-09-291-1/+1
| | | | | | | | 00193 # Enable loading sqlite extensions. This patch isn't needed for python3.spec, since Python 3 has a configuration option for this. rhbz#1066708 Patch provided by John C. Peterson
* 00191-disable-NOOP.patchFedora Python maintainers2020-09-291-0/+1
| | | | | 00191 # Disabling NOOP test as it fails without internet connection
* 00189-use-rpm-wheels.patchFedora Python maintainers2020-09-291-11/+24
| | | | | | 00189 # Instead of bundled wheels, use our RPM packaged wheels from /usr/share/python-wheels
* 00187-add-RPATH-to-pyexpat.patchFedora Python maintainers2020-09-291-0/+9
| | | | | | | | | 00187 # Add an explicit RPATH to pyexpat.so pointing at the directory containing the system expat (which has the extra XML_SetHashSalt symbol), to avoid an ImportError with a link error if there's an LD_LIBRARY_PATH containing a "vanilla" build of expat (without the symbol)
* 00185-urllib2-honors-noproxy-for-ftp.patchFedora Python maintainers2020-09-291-0/+2
| | | | | | 00185 # Makes urllib2 honor "no_proxy" enviroment variable for "ftp:" URLs when ftp_proxy is set
* 00181-allow-arbitrary-timeout-in-condition-wait.patchFedora Python maintainers2020-09-291-7/+10
| | | | | | | | | | 00181 # Allow arbitrary timeout for Condition.wait, as reported in https://bugzilla.redhat.com/show_bug.cgi?id=917709 Upstream doesn't want this: http://bugs.python.org/issue17748 But we have no better solution downstream yet, and since there is no API breakage, we apply this patch. Doesn't apply to Python 3, where this is fixed otherwise and works.
* 00180-python-add-support-for-ppc64p7.patchFedora Python maintainers2020-09-291-1/+1
| | | | | | 00180 # Enable building on ppc64p7 Not appropriate for upstream, Fedora-specific naming
* 00174-fix-for-usr-move.patchFedora Python maintainers2020-09-291-0/+18
| | | | | | | | | 00174 # Workaround for failure to set up prefix/exec_prefix when running an embededed libpython that sets Py_SetProgramName() to a name not on $PATH when run from the root directory due to https://fedoraproject.org/wiki/Features/UsrMove e.g. cmpi-bindings under systemd (rhbz#817554):
* 00170-gc-assertions.patchFedora Python maintainers2020-09-292-18/+138
| | | | | | | | | | | | 00170 # In debug builds, try to print repr() when a C-level assert fails in the garbage collector (typically indicating a reference-counting error somewhere else e.g in an extension module) Backported to 2.7 from a patch I sent upstream for py3k http://bugs.python.org/issue9263 (rhbz#614680) hiding the proposed new macros/functions within gcmodule.c to avoid exposing them within the extension API. (rhbz#850013)
* 00169-avoid-implicit-usage-of-md5-in-multiprocessing.patchFedora Python maintainers2020-09-291-2/+10
| | | | | | | | | 00169 # Use SHA-256 rather than implicitly using MD5 within the challenge handling in multiprocessing.connection Sent upstream as http://bugs.python.org/issue17258 (rhbz#879695)
* 00167-disable-stack-navigation-tests-when-optimized-in-test_gdb.patchFedora Python maintainers2020-09-291-11/+3
| | | | | | | | | | 00167 # Don't run any of the stack navigation tests in test_gdb when Python is optimized, since there appear to be many different ways in which gdb can fail to read the PyFrameObject* for arbitrary places in the callstack, presumably due to compiler optimization (rhbz#912025) Not yet sent upstream
* 00165-crypt-module-salt-backport.patchFedora Python maintainers2020-09-296-13/+198
| | | | | | | | | | 00165 # Backport to Python 2 from Python 3.3 of improvements to the "crypt" module adding precanned ways of salting a password (rhbz#835021) Based on r88500 patch to py3k from Python 3.3 plus 6482dd1c11ed, 0586c699d467, 62994662676a, 74a1110a3b50, plus edits to docstrings to note that this additional functionality is not standard within 2.7
* 00156-gdb-autoload-safepath.patchFedora Python maintainers2020-09-291-1/+26
| | | | | | | | 00156 # Recent builds of gdb will only auto-load scripts from certain safe locations. Turn off this protection when running test_gdb in the selftest suite to ensure that it can load our -gdb.py script (rhbz#817072): Not yet sent upstream
* 00155-avoid-ctypes-thunks.patchFedora Python maintainers2020-09-291-5/+0
| | | | | | | 00155 # Avoid allocating thunks in ctypes unless absolutely necessary, to avoid generating SELinux denials on "import ctypes" and "import uuid" when embedding Python within httpd (rhbz#814391)
* 00147-add-debug-malloc-stats.patchFedora Python maintainers2020-09-2926-36/+319
| | | | | | | 00147 # Add a sys._debugmallocstats() function Based on patch 202 from RHEL 5's python.spec, with updates from rhbz#737198 Sent upstream as http://bugs.python.org/issue14785
* 00146-hashlib-fips.patchFedora Python maintainers2020-09-295-164/+304
| | | | | | | | | | | | | | | | | | | 00146 # Support OpenSSL FIPS mode (e.g. when OPENSSL_FORCE_FIPS_MODE=1 is set) - handle failures from OpenSSL (e.g. on attempts to use MD5 in a FIPS-enforcing environment) - add a new "usedforsecurity" keyword argument to the various digest algorithms in hashlib so that you can whitelist a callsite with "usedforsecurity=False" (sent upstream for python 3 as http://bugs.python.org/issue9216; this is a backport to python 2.7; see RHEL6 patch 119) - enforce usage of the _hashlib implementation: don't fall back to the _md5 and _sha* modules (leading to clearer error messages if fips selftests fail) - don't build the _md5 and _sha* modules; rely on the _hashlib implementation of hashlib (for example, md5.py will use _hashlib's implementation of MD5, if permitted by the FIPS setting) (rhbz#563986)
* 00144-no-gdbm.patchFedora Python maintainers2020-09-291-1/+1
| | | | | 00144 # (Optionally) disable the gdbm module:
* 00143-tsc-on-ppc.patchFedora Python maintainers2020-09-291-12/+30
| | | | | | | 00143 # Fix the --with-tsc option on ppc64, and rework it on 32-bit ppc to avoid aliasing violations (rhbz#698726) Sent upstream as http://bugs.python.org/issue12872
* 00142-skip-failing-pty-tests-in-rpmbuild.patchFedora Python maintainers2020-09-292-0/+2
| | | | | 00142 # Some pty tests fail when run in mock (rhbz#714627):
* 00140-skip-test_ctypes-known-failure-on-sparc.patchFedora Python maintainers2020-09-291-0/+1
| | | | | | | 00140 # Sparc-specific: skip known failure in test_ctypes: http://bugs.python.org/issue8314 (rhbz#711584) which appears to be a libffi bug
* 00139-skip-test_float-known-failure-on-arm.patchFedora Python maintainers2020-09-291-0/+1
| | | | | | 00139 # ARM-specific: skip known failure in test_float: http://bugs.python.org/issue8265 (rhbz#706253)
* 00138-fix-distutils-tests-in-debug-build.patchFedora Python maintainers2020-09-291-6/+9
| | | | | 00138 # Fixup some tests within distutils to work with how debug builds are set up:
* 00137-skip-distutils-tests-that-fail-in-rpmbuild.patchFedora Python maintainers2020-09-291-0/+1
| | | | | 00137 # Some tests within distutils fail when run in an rpmbuild:
* 00136-skip-tests-of-seeking-stdin-in-rpmbuild.patchFedora Python maintainers2020-09-291-0/+1
| | | | | | 00136 # Some tests try to seek on sys.stdin, but don't work as expected when run within Koji/mock; skip them within the rpm build:
* 00133-skip-test_dl.patchFedora Python maintainers2020-09-291-0/+3
| | | | | 00133 # "dl" is deprecated, and test_dl doesn't work on 64-bit builds:
* 00132-add-rpmbuild-hooks-to-unittest.patchFedora Python maintainers2020-09-292-1/+40
| | | | | | | | | | | | | | | 00132 # Add non-standard hooks to unittest for use in the "check" phase below, when running selftests within the build: @unittest._skipInRpmBuild(reason) for tests that hang or fail intermittently within the build environment, and: @unittest._expectedFailureInRpmBuild for tests that always fail within the build environment The hooks only take effect if WITHIN_PYTHON_RPM_BUILD is set in the environment, which we set manually in the appropriate portion of the "check" phase below (and which potentially other python-* rpms could set, to reuse these unittest hooks in their own "check" phases)
* 00131-disable-tests-in-test_io.patchFedora Python maintainers2020-09-291-0/+1
| | | | | | | | 00131 # The four tests in test_io built on top of check_interrupted_write_retry fail when built in Koji, for ppc and ppc64; for some reason, the SIGALRM handlers are never called, and the call to write runs to completion (rhbz#732998)
* python-2.7.2-add-extension-suffix-to-python-config.patchFedora Python maintainers2020-09-291-1/+3
| | | | | | | | | | | | | | | | | | 00130 # Add "--extension-suffix" option to python-config and python-debug-config (rhbz#732808) This is adapted from 3.2's PEP-3149 support. Fedora's debug build has some non-standard features (see also patch 112 above), though largely shared with Debian/Ubuntu and Windows In particular, SO in the Makefile is currently always just ".so" for our python 2 optimized builds, but for python 2 debug it should be '_d.so', to distinguish the debug vs optimized ABI, following the pattern in the above patch. Not yet sent upstream
* python-2.7.1-fix_test_abc_with_COUNT_ALLOCS.patchFedora Python maintainers2020-09-291-2/+8
| | | | | | | | 00128 # 2.7.1 (in r84230) added a test to test_abc which fails if python is configured with COUNT_ALLOCS, which is the case for our debug build (the COUNT_ALLOCS instrumentation keeps "C" alive). Not yet sent upstream
* 00121-add-Modules-to-build-path.patchFedora Python maintainers2020-09-291-0/+4
| | | | | | | | | | | | | | | | Upstream r79310 removed the "Modules" directory from sys.path when Python is running from the build directory on POSIX to fix a unit test (issue #8205). This seems to have broken the compileall.py done in "make install": it cannot find shared library extension modules at this point in the build (sys.path does not contain DESTDIR/usr/lib(64)/python-2.7/lib-dynload for some reason), leading to the build failing with: Traceback (most recent call last): File "/home/david/rpmbuild/BUILDROOT/python-2.7-0.1.rc2.fc14.x86_64/usr/lib64/python2.7/compileall.py", line 17, in <module> import struct File "/home/david/rpmbuild/BUILDROOT/python-2.7-0.1.rc2.fc14.x86_64/usr/lib64/python2.7/struct.py", line 1, in <module> from _struct import * ImportError: No module named _struct This patch adds the build Modules directory to build path.
* 00114-statvfs-f_flag-constants.patchFedora Python maintainers2020-09-291-0/+37
| | | | | | | 00114 # Add flags for statvfs.f_flag to the constant list in posixmodule (i.e. "os") (rhbz:553020); partially upstream as http://bugs.python.org/issue7647 Not yet sent upstream
* 00113-more-configuration-flags.patchFedora Python maintainers2020-09-292-0/+30
| | | | | | | | | 00113 # Add configure-time support for the COUNT_ALLOCS and CALL_PROFILE options described at http://svn.python.org/projects/python/trunk/Misc/SpecialBuilds.txt so that if they are enabled, they will be in that build's pyconfig.h, so that extension modules will reliably use them Not yet sent upstream
* 00112-2.7.13-debug-build.patchFedora Python maintainers2020-09-299-37/+69
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 00112 # Patch to support building both optimized vs debug stacks DSO ABIs, sharing the same .py and .pyc files, using "_d.so" to signify a debug build of an extension module. Based on Debian's patch for the same, http://patch-tracker.debian.org/patch/series/view/python2.6/2.6.5-2/debug-build.dpatch (which was itself based on the upstream Windows build), but with some changes: * Debian's patch to dynload_shlib.c looks for module_d.so, then module.so, but this can potentially find a module built against the wrong DSO ABI. We instead search for just module_d.so in a debug build * We remove this change from configure.in's build of the Makefile: SO=$DEBUG_EXT.so so that sysconfig.py:customize_compiler stays with shared_lib_extension='.so' on debug builds, so that UnixCCompiler.find_library_file can find system libraries (otherwise "make sharedlibs" fails to find system libraries, erroneously looking e.g. for "libffi_d.so" rather than "libffi.so") * We change Lib/distutils/command/build_ext.py:build_ext.get_ext_filename to add the _d there, when building an extension. This way, "make sharedlibs" can build ctypes, by finding the sysmtem libffi.so (rather than failing to find "libffi_d.so"), and builds the module as _ctypes_d.so * Similarly, update build_ext:get_libraries handling of Py_ENABLE_SHARED by appending "_d" to the python library's name for the debug configuration * We modify Modules/makesetup to add the "_d" to the generated Makefile rules for the various Modules/*.so targets This may introduce issues when building an extension that links directly against another extension (e.g. users of NumPy?), but seems more robust when searching for external libraries * We don't change Lib/distutils/command/build.py: build.build_purelib to embed plat_specifier, leaving it as is, as pure python builds should be unaffected by these differences (we'll be sharing the .py and .pyc files) * We introduce DEBUG_SUFFIX as well as DEBUG_EXT: - DEBUG_EXT is used by ELF files (names and SONAMEs); it will be "_d" for a debug build - DEBUG_SUFFIX is used by filesystem paths; it will be "-debug" for a debug build Both will be empty in an optimized build. "_d" contains characters that are valid ELF metadata, but this leads to various ugly filesystem paths (such as the include path), and DEBUG_SUFFIX allows these paths to have more natural names. Changing this requires changes elsewhere in the distutils code. * We add DEBUG_SUFFIX to PYTHON in the Makefile, so that the two configurations build parallel-installable binaries with different names ("python-debug" vs "python"). * Similarly, we add DEBUG_SUFFIX within python-config and python$(VERSION)-config, so that the two configuration get different paths for these. See also patch 130 below
* 00111-no-static-lib.patchFedora Python maintainers2020-09-291-25/+1
| | | | | | | 00111 # Patch the Makefile.pre.in so that the generated Makefile doesn't try to build a libpythonMAJOR.MINOR.a (bug 550692): Downstream only: not appropriate for upstream
* 00104-lib64-fix-for-test_install.patchFedora Python maintainers2020-09-291-1/+2
| | | | | | 00104 # Only used when "%%{_lib}" == "lib64" Another lib64 fix, for distutils/tests/test_install.py; not upstream:
* python-2.7-lib64-sysconfig.patchFedora Python maintainers2020-09-291-9/+9
| | | | | | | | 00103 # Python 2.7 split out much of the path-handling from distutils/sysconfig.py to a new sysconfig.py (in r77704). We need to make equivalent changes to that new file to ensure that the stdlib and platform-specific code go to /usr/lib64 not /usr/lib, on 64-bit archs:
* 00102-2.7.13-lib64.patchFedora Python maintainers2020-09-298-18/+29
| | | | | | | | Only used when "%%{_lib}" == "lib64" Fixup various paths throughout the build and in distutils from "lib" to "lib64", and add the /usr/lib64/pythonMAJOR.MINOR/site-packages to sitedirs, in front of /usr/lib/pythonMAJOR.MINOR/site-packages Not upstream
* 00055-systemtap.patchFedora Python maintainers2020-09-295-1/+118
| | | | | | | | | 00055 # Systemtap support: add statically-defined probe points Patch based on upstream bug: http://bugs.python.org/issue4111 fixed up by mjw and wcohen for 2.6.2, then fixed up by dmalcolm for 2.6.4 then rewritten by mjw (attachment 390110 of rhbz 545179), then reformatted for 2.7rc1 by dmalcolm:
* python-2.6.4-distutils-rpath.patchFedora Python maintainers2020-09-291-0/+10
| | | | | | | 00017 # Fixup distutils/unixccompiler.py to remove standard library path from rpath: Adapted from Patch0 in ivazquez' python3000 specfile, removing usage of super() as it's an old-style class
* python-2.6-rpath.patchFedora Python maintainers2020-09-291-1/+1
| | | | | | 00016 # Remove an "-rpath $(LIBDIR)" argument from the linkage args in configure.in: FIXME: is this for OSF, not Linux?
* python-2.7rc1-socketmodule-constants2.patchFedora Python maintainers2020-09-291-0/+9
| | | | | | | 00014 # Add various constants to the socketmodule (rhbz#436560): TODO: these patches were added in 2.5.1-22 and 2.5.1-24 but appear not to have been sent upstream yet:
* python-2.7rc1-socketmodule-constants.patchFedora Python maintainers2020-09-291-0/+55
| | | | | | | 00013 # Add various constants to the socketmodule (rhbz#436560): TODO: these patches were added in 2.5.1-22 and 2.5.1-24 but appear not to have been sent upstream yet:
* 00010-2.7.13-binutils-no-dep.patchFedora Python maintainers2020-09-291-1/+5
| | | | | | | | FIXME: Lib/ctypes/util.py posix implementation defines a function _get_soname(f). Upstreams's implementation of this uses objdump to read the SONAME from a library; we avoid this, apparently to minimize space requirements on the live CD: (rhbz:307221)
* python-2.5.1-sqlite-encoding.patchFedora Python maintainers2020-09-292-4/+2
| | | | | | | | | | | | | | | | | | | | | | 00007 # This patch was listed in the changelog as: * Fri Sep 14 2007 Jeremy Katz <katzj@redhat.com> - 2.5.1-11 - fix encoding of sqlite .py files to work around weird encoding problem in Turkish (#283331) A traceback attached to rhbz 244016 shows the problem most clearly: a traceback on attempting to import the sqlite module, with: "SyntaxError: encoding problem: with BOM (__init__.py, line 1)" This seems to come from Parser/tokenizer.c:check_coding_spec Our patch changes two source files within sqlite3, removing the "coding: ISO-8859-1" specs and character E4 = U+00E4 = LATIN SMALL LETTER A WITH DIAERESIS from in ghaering's surname. It may be that the conversion of "ISO-8859-1" to "iso-8859-1" is thwarted by the implementation of "tolower" in the Turkish locale; see: https://bugzilla.redhat.com/show_bug.cgi?id=191096#c9 TODO: Not yet sent upstream, and appears to me (dmalcolm 2010-01-29) that it may be papering over a symptom
* python-2.5.1-plural-fix.patchFedora Python maintainers2020-09-291-0/+2
| | | | | | | | | 00006 # Work around a bug in Python' gettext module relating to the "Plural-Forms" header (rhbz:252136) Related to upstream issues: http://bugs.python.org/issue1448060 and http://bugs.python.org/issue1475523 though the proposed upstream patches are, alas, different
* python-2.5-cflags.patchFedora Python maintainers2020-09-291-1/+1
| | | | | | | 00004 # Add $(CFLAGS) to the linker arguments when linking the "python" binary since some architectures (sparc64) need this (rhbz:199373). Not yet filed upstream
* 00001-pydocnogui.patchFedora Python maintainers2020-09-291-7/+1
| | | | | | | | Removes the "-g" option from "pydoc", for some reason; I believe (dmalcolm 2010-01-29) that this was introduced in this change: - fix pydoc (#68082) in 2.2.1-12 as a response to the -g option needing TkInter installed (Red Hat Linux 8)
* python-2.7.1-config.patchFedora Python maintainers2020-09-291-63/+61
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 00000 # Modules/Setup.dist is ultimately used by the "makesetup" script to construct the Makefile and config.c Upstream leaves many things disabled by default, to try to make it easy as possible to build the code on as many platforms as possible. TODO: many modules can also now be built by setup.py after the python binary has been built; need to assess if we should instead build things there We patch it downstream as follows: - various modules are built by default by upstream as static libraries; we built them as shared libraries - build the "readline" module (appears to also be handled by setup.py now) - build the nis module (which needs the tirpc library since glibc 2.26) - enable the build of the following modules: - array arraymodule.c # array objects - cmath cmathmodule.c # -lm # complex math library functions - math mathmodule.c # -lm # math library functions, e.g. sin() - _struct _struct.c # binary structure packing/unpacking - time timemodule.c # -lm # time operations and variables - operator operator.c # operator.add() and similar goodies - _weakref _weakref.c # basic weak reference support - _testcapi _testcapimodule.c # Python C API test module - _random _randommodule.c # Random number generator - _collections _collectionsmodule.c # Container types - itertools itertoolsmodule.c - strop stropmodule.c - _functools _functoolsmodule.c - _bisect _bisectmodule.c # Bisection algorithms - unicodedata unicodedata.c # static Unicode character database - _locale _localemodule.c - fcntl fcntlmodule.c # fcntl(2) and ioctl(2) - spwd spwdmodule.c # spwd(3) - grp grpmodule.c # grp(3) - select selectmodule.c # select(2); not on ancient System V - mmap mmapmodule.c # Memory-mapped files - _csv _csv.c # CSV file helper - _socket socketmodule.c # Socket module helper for socket(2) - _ssl _ssl.c - crypt cryptmodule.c -lcrypt # crypt(3) - termios termios.c # Steen Lumholt's termios module - resource resource.c # Jeremy Hylton's rlimit interface - audioop audioop.c # Operations on audio samples - imageop imageop.c # Operations on images - _md5 md5module.c md5.c - _sha shamodule.c - _sha256 sha256module.c - _sha512 sha512module.c - linuxaudiodev linuxaudiodev.c - timing timingmodule.c - _tkinter _tkinter.c tkappinit.c - dl dlmodule.c - gdbm gdbmmodule.c - _bsddb _bsddb.c - binascii binascii.c - parser parsermodule.c - cStringIO cStringIO.c - cPickle cPickle.c - zlib zlibmodule.c - _multibytecodec cjkcodecs/multibytecodec.c - _codecs_cn cjkcodecs/_codecs_cn.c - _codecs_hk cjkcodecs/_codecs_hk.c - _codecs_iso2022 cjkcodecs/_codecs_iso2022.c - _codecs_jp cjkcodecs/_codecs_jp.c - _codecs_kr cjkcodecs/_codecs_kr.c - _codecs_tw cjkcodecs/_codecs_tw.c
generated by cgit v0.12 at 2025-04-08 12:21:17 (GMT)