The package rpms/scipy.git has added or updated architecture specific content in its
spec file (ExclusiveArch/ExcludeArch or %ifarch/%ifnarch) in commit(s):
https://src.fedoraproject.org/cgit/rpms/scipy.git/commit/?id=803fd0655ea3....
Change:
-%ifarch ppc64le
Thanks.
Full change:
============
commit 803fd0655ea31392409c6b9121a612b2bc0ca631
Author: Nikola Forró <nforro(a)redhat.com>
Date: Wed Sep 30 15:34:25 2020 +0200
New upstream release 1.5.2
- resolves: #1853871 and #1840077
diff --git a/.gitignore b/.gitignore
index 821493f..d3a3c93 100644
--- a/.gitignore
+++ b/.gitignore
@@ -29,3 +29,4 @@ scipy-0.7.2.tar.gz
/scipy-1.3.1.tar.gz
/scipy-1.4.1.tar.gz
/scipy-1.5.0.tar.gz
+/scipy-1.5.2.tar.gz
diff --git a/acceptable_failure_rate.patch b/acceptable_failure_rate.patch
deleted file mode 100644
index aacb1f1..0000000
--- a/acceptable_failure_rate.patch
+++ /dev/null
@@ -1,20 +0,0 @@
-diff --git a/scipy/conftest.py b/scipy/conftest.py
-index e5fac23..1a971e8 100644
---- a/scipy/conftest.py
-+++ b/scipy/conftest.py
-@@ -37,3 +38,15 @@ def check_fpu_mode(request):
- warnings.warn("FPU mode changed from {0:#x} to {1:#x} during "
- "the test".format(old_mode, new_mode),
- category=FPUModeChangeWarning, stacklevel=0)
-+
-+
-+ACCEPTABLE_FAILURE_RATE = int(os.environ.get('ACCEPTABLE_FAILURE_RATE', 0))
-+
-+
-+(a)pytest.hookimpl()
-+def pytest_sessionfinish(session, exitstatus):
-+ if exitstatus != 1:
-+ return
-+ failure_rate = (100.0 * session.testsfailed) / session.testscollected
-+ if failure_rate <= ACCEPTABLE_FAILURE_RATE:
-+ session.exitstatus = 0
diff --git a/scipy.spec b/scipy.spec
index 5412207..9ade31b 100644
--- a/scipy.spec
+++ b/scipy.spec
@@ -14,8 +14,8 @@
Summary: Scientific Tools for Python
Name: scipy
-Version: 1.5.0
-Release: 4%{?dist}
+Version: 1.5.2
+Release: 1%{?dist}
# BSD -- whole package except:
# Boost -- scipy/special/cephes/scipy_iv.c
@@ -24,10 +24,8 @@ License: BSD and Boost and Public Domain
Url:
http://www.scipy.org/scipylib/index.html
Source0:
https://github.com/scipy/scipy/releases/download/v%{version}/scipy-%{vers...
-# Previously we ignored the tests results, because they don't always pass
-# Instead of ignoring the results entirely, we allow certain failure rate
-#
https://stackoverflow.com/a/47731333/1839451
-Patch0: acceptable_failure_rate.patch
+#
https://github.com/scipy/scipy/pull/12899
+Patch0: test_nnz_overflow.patch
BuildRequires: fftw-devel, suitesparse-devel
BuildRequires: %{blaslib}-devel
@@ -139,33 +137,24 @@ pathfix.py -pni "%{__python3} %{py3_shbang_opts}"
%{buildroot}%{python3_sitearch
# check against the reference BLAS/LAPACK
export FLEXIBLAS=netlib
-# Skip all tests on s390x because they hangs unexpectedly and randomly
-# and pytest-timeout has no effect. Note that the outcome of the tests
-# was previously ignored anyway so by disabling the test for s390x we
-# are not doing anything more dangerous.
%ifarch s390x
-exit 0
-%endif
-
-%ifarch x86_64
-export ACCEPTABLE_FAILURE_RATE=0
-%else
-# there are usually 10-21 test failing, so we allow 1% failure rate
-# XXX ppc fails 2%+, so we've extended this to 3% for now
-export ACCEPTABLE_FAILURE_RATE=3
-%endif
-
-%ifarch ppc64le
-# test_decomp segfaults on ppc64le
-export k="not test_denormals and not test_decomp"
-%else
-# test_denormals tends to stuck
-export k="not test_denormals"
+# skip failing tests on s390x for now
+export PYTEST_ADDOPTS="-k '\
+ not (TestNoData and test_nodata) and \
+ not test_fortranfile_read_mixed_record and \
+ not test_kde_1d and \
+ not test_kde_1d_weighted and \
+ not test_kde_2d and \
+ not test_kde_2d_weighted and \
+ not test_gaussian_kde_subclassing and \
+ not test_gaussian_kde_covariance_caching and \
+ not test_kde_integer_input and \
+ not test_pdf_logpdf and \
+ not test_pdf_logpdf_weighted'"
%endif
pushd %{buildroot}/%{python3_sitearch}
-# TODO TestIQR.test_scale fails on Python 3.8+ due to some warnings, investigate
-%{pytest} --timeout=500 -k "$k and not (TestIQR and test_scale)" scipy
--numprocesses=auto
+%{pytest} --timeout=300 scipy --numprocesses=auto
# Remove test remnants
rm -rf gram{A,B}
popd
@@ -182,6 +171,10 @@ popd
%endif
%changelog
+* Mon Aug 31 2020 Nikola Forró <nforro(a)redhat.com> - 1.5.2-1
+- New upstream release 1.5.2
+ resolves: #1853871 and #1840077
+
* Sun Aug 16 2020 Iñaki Úcar <iucar(a)fedoraproject.org> - 1.5.0-4
-
https://fedoraproject.org/wiki/Changes/FlexiBLAS_as_BLAS/LAPACK_manager
diff --git a/sources b/sources
index e0a0492..339a005 100644
--- a/sources
+++ b/sources
@@ -1 +1 @@
-SHA512 (scipy-1.5.0.tar.gz) =
4c981a3125a88593cbc5a17417022a5db8f6ccb4c1c24e78afeb9bb26428b81d2d0fe9965caa418753b1bbbda2fa30533fa1307aac6ab168458d68f7b46049be
+SHA512 (scipy-1.5.2.tar.gz) =
45463df30a0f6270d9f4cf52235f31607904a6ae1375e12600e7f1ab2d27b1dc25a6211b49dceb71506be22c756890adaf9f81d9e6be7455def86c9caf0dc923
diff --git a/test_nnz_overflow.patch b/test_nnz_overflow.patch
new file mode 100644
index 0000000..52d5cd3
--- /dev/null
+++ b/test_nnz_overflow.patch
@@ -0,0 +1,30 @@
+From b4da1455615fd7dc6d3554d16008a1e88018fa4d Mon Sep 17 00:00:00 2001
+From: =?UTF-8?q?Nikola=20Forr=C3=B3?= <nforro(a)redhat.com>
+Date: Wed, 30 Sep 2020 11:44:25 +0200
+Subject: [PATCH] TST: skip test_nnz_overflow on 32-bit architectures
+MIME-Version: 1.0
+Content-Type: text/plain; charset=UTF-8
+Content-Transfer-Encoding: 8bit
+
+On a 32-bit architecture, in case check_free_memory() passes, the first
+np.zeros() fails with "ValueError: Maximum allowed dimension exceeded",
+which I believe is expected.
+
+Signed-off-by: Nikola Forró <nforro(a)redhat.com>
+---
+ scipy/sparse/tests/test_sparsetools.py | 2 ++
+ 1 file changed, 2 insertions(+)
+
+diff --git a/scipy/sparse/tests/test_sparsetools.py
b/scipy/sparse/tests/test_sparsetools.py
+index 0c208ef4451..19c7de43fa1 100644
+--- a/scipy/sparse/tests/test_sparsetools.py
++++ b/scipy/sparse/tests/test_sparsetools.py
+@@ -61,6 +61,8 @@ def test_regression_std_vector_dtypes():
+
+
+ @pytest.mark.slow
++(a)pytest.mark.skipif(not (sys.platform.startswith('linux') and
np.dtype(np.intp).itemsize >= 8),
++ reason="test requires 64-bit Linux")
+ def test_nnz_overflow():
+ # Regression test for gh-7230 / gh-7871, checking that coo_todense
+ # with nnz > int32max doesn't overflow.