rewrite of %check: fine-grained test exclusions
* Sat Sep 10 2011 David Malcolm <dmalcolm@redhat.com> - 3.2.2-2 - rewrite of "check", introducing downstream-only hooks for skipping specific cases in an rpmbuild (patch 132), and fixing/skipping failing tests in a more fine-grained manner than before; (patches 106, 133-142 sparsely, moving patches for consistency with python.spec: 128 to 134, 126 to 135, 127 to 141)
This commit is contained in:
parent
06b3c5b484
commit
0c8875f3f6
9 changed files with 227 additions and 87 deletions
188
python3.spec
188
python3.spec
|
|
@ -118,7 +118,7 @@
|
|||
Summary: Version 3 of the Python programming language aka Python 3000
|
||||
Name: python3
|
||||
Version: %{pybasever}.2
|
||||
Release: 1%{?dist}
|
||||
Release: 2%{?dist}
|
||||
License: Python
|
||||
Group: Development/Languages
|
||||
|
||||
|
|
@ -234,6 +234,10 @@ Patch104: python-3.1.2-more-configuration-flags.patch
|
|||
# (rhbz:553020); partially upstream as http://bugs.python.org/issue7647
|
||||
Patch105: python-3.2a1-statvfs-f_flag-constants.patch
|
||||
|
||||
# Only used when "%{_lib}" == "lib64"
|
||||
# Another lib64 fix, for distutils/tests/test_install.py; not upstream:
|
||||
Patch106: 00106-lib64-fix-for-test_install.patch
|
||||
|
||||
# COUNT_ALLOCS is useful for debugging, but the upstream behaviour of always
|
||||
# emitting debug info to stdout on exit is too verbose and makes it harder to
|
||||
# use the debug build. Add a "PYTHONDUMPCOUNTS" environment variable which
|
||||
|
|
@ -241,20 +245,6 @@ Patch105: python-3.2a1-statvfs-f_flag-constants.patch
|
|||
# Not yet sent upstream:
|
||||
Patch125: less-verbose-COUNT_ALLOCS.patch
|
||||
|
||||
# test_weakref's test_callback_in_cycle_resurrection doesn't work with
|
||||
# COUNT_ALLOCS, as the metrics keep "C" alive. Work around this for our
|
||||
# debug build:
|
||||
# Not yet sent upstream
|
||||
Patch126: python-3.2b2-test-weakref-COUNT_ALLOCS_fix.patch
|
||||
|
||||
# Similar COUNT_ALLOCS fixes for test_gc
|
||||
# Not yet sent upstream
|
||||
Patch127: python-3.2b2-fix-test-gc-COUNT_ALLOCS.patch
|
||||
|
||||
# Similar COUNT_ALLOCS fixes for test_sys
|
||||
# Not yet sent upstream
|
||||
Patch128: python-3.2b2-test_sys-COUNT_ALLOCS.patch
|
||||
|
||||
# In my koji builds, /root/bin is in the PATH for some reason
|
||||
# This leads to test_subprocess.py failing, due to "test_leaking_fds_on_error"
|
||||
# trying every dir in PATH for "nonexisting_i_hope", which leads to it raising
|
||||
|
|
@ -277,6 +267,74 @@ Patch130: python-2.7.2-tsc-on-ppc.patch
|
|||
# (rhbz#732998)
|
||||
Patch131: python-2.7.2-disable-tests-in-test_io.patch
|
||||
|
||||
# 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)
|
||||
Patch132: 00132-add-rpmbuild-hooks-to-unittest.patch
|
||||
|
||||
# 00133-skip-test_dl.patch is not relevant for python3: the "dl" module no
|
||||
# longer exists
|
||||
|
||||
# Fix a failure in test_sys.py when configured with COUNT_ALLOCS enabled
|
||||
# Not yet sent upstream
|
||||
Patch134: 00134-fix-COUNT_ALLOCS-failure-in-test_sys.patch
|
||||
|
||||
# test_weakref's test_callback_in_cycle_resurrection doesn't work with
|
||||
# COUNT_ALLOCS, as the metrics keep "C" alive. Work around this for our
|
||||
# debug build:
|
||||
# Not yet sent upstream
|
||||
Patch135: 00135-fix-test-within-test_weakref-in-debug-build.patch
|
||||
|
||||
# Patch136: 00136-skip-tests-of-seeking-stdin-in-rpmbuild.patch does not seem
|
||||
# to be needed by python3
|
||||
|
||||
# Some tests within distutils fail when run in an rpmbuild:
|
||||
Patch137: 00137-skip-distutils-tests-that-fail-in-rpmbuild.patch
|
||||
|
||||
# Patch138: 00138-fix-distutils-tests-in-debug-build.patch is not relevant for
|
||||
# python3
|
||||
|
||||
# ARM-specific: skip known failure in test_float:
|
||||
# http://bugs.python.org/issue8265 (rhbz#706253)
|
||||
Patch139: 00139-skip-test_float-known-failure-on-arm.patch
|
||||
|
||||
# Patch140: 00140-skip-test_ctypes-known-failure-on-sparc.patch does not appear
|
||||
# to be relevant for python3
|
||||
|
||||
# Fix test_gc's test_newinstance case when configured with COUNT_ALLOCS:
|
||||
Patch141: 00141-fix-test_gc_with_COUNT_ALLOCS.patch
|
||||
|
||||
# Some pty tests fail when run in mock (rhbz#714627):
|
||||
Patch142: 00142-skip-failing-pty-tests-in-rpmbuild.patch
|
||||
|
||||
# (New patches go here ^^^)
|
||||
#
|
||||
# When adding new patches to "python" and "python3" in Fedora 17 onwards,
|
||||
# please try to keep the patch numbers in-sync between the two specfiles:
|
||||
#
|
||||
# - use the same patch number across both specfiles for conceptually-equivalent
|
||||
# fixes, ideally with the same name
|
||||
#
|
||||
# - when a patch is relevan to both specfiles, use the same introductory
|
||||
# comment in both specfiles where possible (to improve "diff" output when
|
||||
# comparing them)
|
||||
#
|
||||
# - when a patch is only relevant for one of the two specfiles, leave a gap
|
||||
# in the patch numbering in the other specfile, adding a comment when
|
||||
# omitting a patch, both in the manifest section here, and in the "prep"
|
||||
# phase below
|
||||
#
|
||||
# Hopefully this will make it easier to ensure that all relevant fixes are
|
||||
# applied to both versions.
|
||||
|
||||
# This is the generated patch to "configure"; see the description of
|
||||
# %{regenerate_autotooling_patch}
|
||||
# above:
|
||||
|
|
@ -424,6 +482,7 @@ rm -r Modules/zlib || exit 1
|
|||
|
||||
%if "%{_lib}" == "lib64"
|
||||
%patch102 -p1
|
||||
%patch106 -p1
|
||||
%endif
|
||||
|
||||
%patch104 -p1 -b .more-configuration-flags
|
||||
|
|
@ -431,9 +490,7 @@ rm -r Modules/zlib || exit 1
|
|||
%patch105 -p1 -b .statvfs-f-flag-constants
|
||||
|
||||
%patch125 -p1 -b .less-verbose-COUNT_ALLOCS
|
||||
%patch126 -p1
|
||||
%patch127 -p1
|
||||
%patch128 -p1
|
||||
|
||||
%patch129 -p1
|
||||
%patch130 -p1 -b .tsc-on-ppc
|
||||
|
||||
|
|
@ -441,6 +498,20 @@ rm -r Modules/zlib || exit 1
|
|||
%patch131 -p1
|
||||
%endif
|
||||
|
||||
%patch132 -p1
|
||||
# 00133: not for python3
|
||||
%patch134 -p1
|
||||
%patch135 -p1
|
||||
# 00136: not for python3
|
||||
%patch137 -p1
|
||||
# 00138: not for python3
|
||||
%ifarch %{arm}
|
||||
%patch139 -p1
|
||||
%endif
|
||||
# 00140: not for python3
|
||||
%patch141 -p1
|
||||
%patch142 -p1
|
||||
|
||||
# Currently (2010-01-15), http://docs.python.org/library is for 2.6, and there
|
||||
# are many differences between 2.6 and the Python 3 library.
|
||||
#
|
||||
|
|
@ -849,79 +920,16 @@ CheckPython() {
|
|||
|
||||
echo STARTING: CHECKING OF PYTHON FOR CONFIGURATION: $ConfName
|
||||
|
||||
# Notes about disabled tests:
|
||||
#
|
||||
# test_distutils.py
|
||||
# This one tries to build an RPM inside the rpmbuild; I'll simply
|
||||
# let this one fail for now (has trouble linking against -lpython3.1; perhaps
|
||||
# LD_LIBRARY_PATH is being discarded somewhere?)
|
||||
#
|
||||
# test_http*
|
||||
# I've seen occasional hangs in some http tests when running the test suite
|
||||
# inside Koji on Python 3. For that reason I exclude them
|
||||
#
|
||||
# test_openpty:
|
||||
# Fails in Koji, possibly due to a mock issue (rhbz#714627)
|
||||
# test test_openpty failed -- Traceback (most recent call last):
|
||||
# File "/builddir/build/BUILD/Python-3.2/Lib/test/test_openpty.py", line 12, in test
|
||||
# master, slave = os.openpty()
|
||||
# OSError: [Errno 2] No such file or directory
|
||||
#
|
||||
# test_pty:
|
||||
# Fails in Koji, possibly due to a mock issue (rhbz#714627)
|
||||
# test test_pty failed -- Traceback (most recent call last):
|
||||
# File "/builddir/build/BUILD/Python-3.2/Lib/test/test_pty.py", line 114, in test_fork
|
||||
# pid, master_fd = pty.fork()
|
||||
# File "/builddir/build/BUILD/Python-3.2/Lib/pty.py", line 107, in fork
|
||||
# master_fd, slave_fd = openpty()
|
||||
# File "/builddir/build/BUILD/Python-3.2/Lib/pty.py", line 29, in openpty
|
||||
# master_fd, slave_name = _open_terminal()
|
||||
# File "/builddir/build/BUILD/Python-3.2/Lib/pty.py", line 70, in _open_terminal
|
||||
# raise os.error('out of pty devices')
|
||||
# OSError: out of pty devices
|
||||
#
|
||||
# test_socket:
|
||||
# test_socket.py:testSockName can fail here if DNS isn't properly set up:
|
||||
# my_ip_addr = socket.gethostbyname(socket.gethostname())
|
||||
# socket.gaierror: [Errno -2] Name or service not known
|
||||
#
|
||||
# test_telnet:
|
||||
# can get a "socket.error: [Errno 104] Connection reset by peer"
|
||||
#
|
||||
# Some additional tests fail when running the test suite as non-root outside of
|
||||
# the build, due to permissions issues.
|
||||
|
||||
%ifarch %{sparc}
|
||||
EXCLUDED_TESTS="\
|
||||
test_distutils \
|
||||
test_httplib \
|
||||
test_http_cookies \
|
||||
test_openpty \
|
||||
test_pty.py \
|
||||
test_socket \
|
||||
test_telnet \
|
||||
test_ctypes \
|
||||
test_openpty \
|
||||
test_pty \
|
||||
%{nil}"
|
||||
%else
|
||||
EXCLUDED_TESTS="\
|
||||
test_distutils \
|
||||
test_httplib \
|
||||
test_http_cookies \
|
||||
test_openpty \
|
||||
test_pty.py \
|
||||
test_socket \
|
||||
test_telnet \
|
||||
%{nil}"
|
||||
%endif
|
||||
# Note that we're running the tests using the version of the code in the builddir,
|
||||
# not in the buildroot.
|
||||
|
||||
# Run the upstream test suite
|
||||
# Run the upstream test suite, setting "WITHIN_PYTHON_RPM_BUILD" so that the
|
||||
# our non-standard decorators take effect on the relevant tests:
|
||||
# @unittest._skipInRpmBuild(reason)
|
||||
# @unittest._expectedFailureInRpmBuild
|
||||
WITHIN_PYTHON_RPM_BUILD= \
|
||||
LD_LIBRARY_PATH=$ConfDir $ConfDir/python -m test.regrtest \
|
||||
--verbose --findleaks \
|
||||
-x $EXCLUDED_TESTS
|
||||
--verbose --findleaks
|
||||
|
||||
echo FINISHED: CHECKING OF PYTHON FOR CONFIGURATION: $ConfName
|
||||
|
||||
|
|
@ -1308,6 +1316,12 @@ rm -fr %{buildroot}
|
|||
# ======================================================
|
||||
|
||||
%changelog
|
||||
* Sat Sep 10 2011 David Malcolm <dmalcolm@redhat.com> - 3.2.2-2
|
||||
- rewrite of "check", introducing downstream-only hooks for skipping specific
|
||||
cases in an rpmbuild (patch 132), and fixing/skipping failing tests in a more
|
||||
fine-grained manner than before; (patches 106, 133-142 sparsely, moving
|
||||
patches for consistency with python.spec: 128 to 134, 126 to 135, 127 to 141)
|
||||
|
||||
* Tue Sep 6 2011 David Malcolm <dmalcolm@redhat.com> - 3.2.2-1
|
||||
- 3.2.2
|
||||
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue