[med-svn] [Git][med-team/sortmerna][master] 5 commits: Build on big endian 64 architectures
Andreas Tille (@tille)
gitlab at salsa.debian.org
Mon Oct 19 21:09:59 BST 2026
Andreas Tille pushed to branch master at Debian Med / sortmerna
Commits:
6bfbaa0f by Andreas Tille at 2026-10-19T21:42:31+02:00
Build on big endian 64 architectures
Upstream supports AMD64 and ARM64 and the code contains no x86-specific
intrinsics (the bundled sse2neon.h is unused; SIMD is provided by the
system parasail library, which uses NEON on arm64). There is thus no
reason to restrict the package to Intel architectures.
Drop the i386-only -msse2 workaround as well: i386 cannot be built since
libbbhash-dev (needed for BooPHF.h) is not available on that architecture.
- - - - -
c596b832 by Andreas Tille at 2026-10-19T21:44:33+02:00
d/rules: build with CMake directly instead of pybuild. Newer dh-python delegates to CMake, but older versions (as in trixie) run "setup.py clean", which the upstream setup.py does not implement, so the build failed on arm64.
- - - - -
5a750170 by Andreas Tille at 2026-10-19T21:46:01+02:00
d/tests: only copy the test data files that are actually needed.
- - - - -
196e30f1 by Andreas Tille at 2026-10-19T21:47:20+02:00
d/salsa-ci.yml:
- Do not build on i386
- Do not fail on blhc
- - - - -
1a899d27 by Andreas Tille at 2026-10-19T21:50:13+02:00
Upload to unstable
- - - - -
5 changed files:
- debian/changelog
- debian/control
- debian/rules
- debian/salsa-ci.yml
- debian/tests/run-unit-test
Changes:
=====================================
debian/changelog
=====================================
@@ -1,4 +1,4 @@
-sortmerna (7.1.0-1) UNRELEASED; urgency=medium
+sortmerna (7.1.0-1) unstable; urgency=medium
* New upstream version
* Standards-Version: 4.7.4 (routine-update)
@@ -22,8 +22,19 @@ sortmerna (7.1.0-1) UNRELEASED; urgency=medium
upstream tiny/small reference data (FASTA, SAM, BLAST and gzip input) and
drop the "superficial" restriction.
Closes: #1072346
-
- -- Andreas Tille <tille at debian.org> Mon, 19 Oct 2026 14:40:30 +0200
+ * Build on big endian 64 architectures (upstream supports AMD64 and ARM64)
+ * Drop the i386-only -msse2 workaround. i386 cannot be built anyway because
+ libbbhash-dev is not available there.
+ * d/rules: build with CMake directly instead of pybuild. Newer dh-python
+ delegates to CMake, but older versions (as in trixie) run "setup.py
+ clean", which the upstream setup.py does not implement, so the build
+ failed on arm64.
+ * d/tests: only copy the test data files that are actually needed.
+ * d/salsa-ci.yml:
+ - Do not build on i386
+ - Do not fail on blhc
+
+ -- Andreas Tille <tille at debian.org> Mon, 19 Oct 2026 21:47:24 +0200
sortmerna (4.3.7-3) unstable; urgency=medium
=====================================
debian/control
=====================================
@@ -8,6 +8,8 @@ Section: science
Build-Depends:
debhelper-compat (= 14),
dh-sequence-python3,
+ architecture-is-64-bit,
+ architecture-is-little-endian,
cmake,
libbbhash-dev,
libgflags-dev,
@@ -23,7 +25,7 @@ Vcs-Git: https://salsa.debian.org/med-team/sortmerna.git
Homepage: https://github.com/sortmerna/sortmerna
Package: sortmerna
-Architecture: any-amd64 any-i386 x32
+Architecture: any
Depends:
${misc:Depends},
${python3:Depends},
=====================================
debian/rules
=====================================
@@ -7,16 +7,8 @@ export DEB_BUILD_MAINT_OPTIONS=hardening=+all
export CMAKE_PREFIX_PATH=/usr/lib/$(DEB_HOST_MULTIARCH)/cmake/
DTMP=$(CURDIR)/debian/sortmerna
-# In this particular instance, building with sse2 support is not an
-# architecture baseline violation for the i386 release architecture, because
-# the package is supposed to depend on sse2-support.
-ifeq ($(DEB_HOST_ARCH),i386)
-export DEB_CFLAGS_MAINT_APPEND+=-msse2
-export DEB_CXXFLAGS_MAINT_APPEND+=-msse2
-endif
-
%:
- dh $@ --buildsystem=pybuild
+ dh $@ --buildsystem=cmake
execute_after_dh_install:
find debian -name cmake -type d | xargs rm -rvf
=====================================
debian/salsa-ci.yml
=====================================
@@ -1,4 +1,11 @@
---
include:
- - https://salsa.debian.org/salsa-ci-team/pipeline/raw/master/salsa-ci.yml
- - https://salsa.debian.org/salsa-ci-team/pipeline/raw/master/pipeline-jobs.yml
+ - https://salsa.debian.org/salsa-ci-team/pipeline/raw/master/recipes/debian.yml
+
+variables:
+# i386 is explicitly excluded by Build-Depends
+ SALSA_CI_DISABLE_BUILD_PACKAGE_I386: 1
+
+# Permit failure of blhc
+blhc:
+ allow_failure: true
=====================================
debian/tests/run-unit-test
=====================================
@@ -14,8 +14,13 @@ if [ "${AUTOPKGTEST_TMP}" = "" ] ; then
fi
cp -a /usr/share/doc/${pkg}/examples/* "${AUTOPKGTEST_TMP}"
-# test_data is to large to bloat the package examples
-cp -a "test_data" "${AUTOPKGTEST_TMP}"
+# Only copy the data the tests need instead of the whole (large) test_data
+mkdir -p "${AUTOPKGTEST_TMP}/test_data"
+for f in test_ref.fasta test_read.fasta \
+ silva-arc-16s-database-id95.fasta \
+ set7_arc_bac_16S_database_match.fasta ; do
+ cp -a "test_data/${f}" "${AUTOPKGTEST_TMP}/test_data/"
+done
cd "${AUTOPKGTEST_TMP}"
View it on GitLab: https://salsa.debian.org/med-team/sortmerna/-/compare/ac280c2d10ca081b6dda60f8063c691e4dd59403...1a899d271bbc6e3afce1c274994c8a4335e64ab9
--
View it on GitLab: https://salsa.debian.org/med-team/sortmerna/-/compare/ac280c2d10ca081b6dda60f8063c691e4dd59403...1a899d271bbc6e3afce1c274994c8a4335e64ab9
You're receiving this email because of your account on salsa.debian.org. Manage all notifications: https://salsa.debian.org/-/profile/notifications | Help: https://salsa.debian.org/help
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://alioth-lists.debian.net/pipermail/debian-med-commit/attachments/20261019/ed41b550/attachment-0001.htm>
More information about the debian-med-commit
mailing list