Bug#1076540: virt-install: --cloud-init option is ignored when used in conjunction with --location
Package: virt-manager Version: 1:4.1.0-4 Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu oracular ubuntu-patch Dear Maintainer, When running virt-install with both --location and --cloud-init options, the latter is silently ignored. Although the --cloud-init option is properly honored when used with --cdrom instead of --location ; but --cdrom is incompatible with --extra-args ; which is useful as well. There is an upstream bug report [1] with an open PR [2]. I have added the patch from the PR as a quilt patch (refreshed) in the attached debdiff. [1] https://github.com/virt-manager/virt-manager/issues/469 [2] https://github.com/virt-manager/virt-manager/pull/485 In Ubuntu, the attached patch was applied to achieve the following: * d/p/virtinstall-Fix-location-and-cloud-init-conflict.patch: do not ignore virt-install's --cloud-init option when used in conjunction with --location. (LP: #2073461) Thanks for considering the patch. -- System Information: Debian Release: trixie/sid APT prefers noble-updates APT policy: (500, 'noble-updates'), (500, 'noble-security'), (500, 'noble'), (100, 'noble-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.8.0-38-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled Versions of packages virt-manager is related to: ii libvirt-clients 10.0.0-2ubuntu8.2 ii libvirt-daemon 10.0.0-2ubuntu8.2 ii libvirt0 10.0.0-2ubuntu8.2 ii osinfo-db0.20240523-0ubuntu0.24.04.1 diff -Nru virt-manager-4.1.0/debian/patches/series virt-manager-4.1.0/debian/patches/series --- virt-manager-4.1.0/debian/patches/series2024-05-09 09:45:01.0 +0200 +++ virt-manager-4.1.0/debian/patches/series2024-07-18 09:54:44.0 +0200 @@ -3,3 +3,4 @@ upstream_tests-Fix-host-copy-XML-with-libvirt-10.1.0.patch upstream_build-use-super.patch upstream_build-use-setuptools.command.build-when-available.patch +virtinstall-Fix-location-and-cloud-init-conflict.patch diff -Nru virt-manager-4.1.0/debian/patches/virtinstall-Fix-location-and-cloud-init-conflict.patch virt-manager-4.1.0/debian/patches/virtinstall-Fix-location-and-cloud-init-conflict.patch --- virt-manager-4.1.0/debian/patches/virtinstall-Fix-location-and-cloud-init-conflict.patch 1970-01-01 01:00:00.0 +0100 +++ virt-manager-4.1.0/debian/patches/virtinstall-Fix-location-and-cloud-init-conflict.patch 2024-07-18 09:54:44.0 +0200 @@ -0,0 +1,208 @@ +Description: virtinstall: Fix --location and --cloud-init conflict + So far, --cloud-init is ignored when --location is specified. + . + Resolves: virt-manager#469 +Author: Patrice Lacouture +Origin: upstream, https://github.com/virt-manager/virt-manager/pull/485 +Bug: https://github.com/virt-manager/virt-manager/issues/469 +Bug-Ubuntu: https://launchpad.net/bugs/2073461 +Applied-Upstream: no +Last-Update: 2024-07-18 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +--- /dev/null b/tests/data/cli/compare/virt-install-location-iso-and-cloud-init.xml +@@ -0,0 +1,172 @@ ++ ++ fedora17 ++ ---- ++ ++http://libosinfo.org/xmlns/libvirt/domain/1.0;> ++ http://fedoraproject.org/fedora/17"/> ++ ++ ++ 65536 ++ 65536 ++ 2 ++ ++hvm ++/VIRTINST-TESTSUITE/vmlinuz ++/VIRTINST-TESTSUITE/initrd.img ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++/usr/bin/qemu-system-x86_64 ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ /dev/urandom ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ds=nocloud ++ ++ ++ destroy ++ ++ ++ fedora17 ++ ---- ++ ++http://libosinfo.org/xmlns/libvirt/domain/1.0;> ++ http://fedoraproject.org/fedora/17"/> ++ ++ ++ 65536 ++ 65536 ++ 2 ++ ++hvm ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++/usr/bin/qemu-system-x86_64 ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ /dev/urandom ++ ++ ++ ++ ++ ++ ++ +--- a/tests/test_cli.py
Bug#1075720: authbind does not honor CFLAGS/LDFLAGS set by the maintainer
Package: authbind Version: 2.1.3 Followup-For: Bug #1075720 User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu oracular ubuntu-patch Control: tags -1 patch Dear Maintainer, To avoid modifying the upstream makefile, I decided to pass the CFLAGS & LDFLAGS variables on the CLI at the make invocation level. Attaching the new debdiff. Thanks, Olivier diff -Nru authbind-2.1.3build2/debian/control authbind-2.1.3ubuntu1~ppa4/debian/control --- authbind-2.1.3build2/debian/control 2024-04-08 17:54:08.0 +0200 +++ authbind-2.1.3ubuntu1~ppa4/debian/control 2024-07-03 17:42:33.0 +0200 @@ -1,8 +1,7 @@ Source: authbind Section: utils Priority: optional -Maintainer: Ubuntu Developers -XSBC-Original-Maintainer: Ian Jackson +Maintainer: Ian Jackson Standards-Version: 3.9.1 Package: authbind diff -Nru authbind-2.1.3build2/debian/rules authbind-2.1.3ubuntu1~ppa4/debian/rules --- authbind-2.1.3build2/debian/rules 2021-12-17 19:34:02.0 +0100 +++ authbind-2.1.3ubuntu1~ppa4/debian/rules 2024-07-03 17:42:33.0 +0200 @@ -41,12 +41,8 @@ INSTALL_SCRIPT = $(INSTALL) -p-o root -g root -m 755 INSTALL_DIR = $(INSTALL) -p -d -o root -g root -m 755 -CFLAGS = -O2 -Wall -LDFLAGS = - -CFLAGS += -g -LDFLAGS += -g +include /usr/share/dpkg/buildflags.mk ifeq (,$(findstring nostrip,$(DEB_BUILD_OPTIONS))) STRIP=$(TOOL_PREFIX)strip @@ -55,8 +51,6 @@ STRIP=: endif -export CFLAGS -export LDFLAGS export INSTALL export INSTALL_FILE export INSTALL_PROGRAM @@ -68,7 +62,7 @@ udp=debian/tmp/usr/share/doc/$(package) build-arch: - $(MAKE) prefix=/usr CC='$(CC)' LD='$(TOOL_PREFIX)ld' + $(MAKE) prefix=/usr CC='$(CC)' LD='$(TOOL_PREFIX)ld' CFLAGS='$(CFLAGS)' LDFLAGS='$(LDFLAGS)' binary-arch: checkroot build rm -rf debian/tmp
Bug#1075720: authbind does not honor CFLAGS/LDFLAGS set by the maintainer
Hello, I just noticed that my quilt patch does not automatically apply because authbind does not declare d/source/format. As a consequence, the patch is ignored and the build still runs without the expected CFLAGS / LDFLAGS. I will rework the patch and update. Thanks, Olivier
Bug#1075720: authbind does not honor CFLAGS/LDFLAGS set by the maintainer
Package: authbind Version: 2.1.3 Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu oracular ubuntu-patch Dear Maintainer, When building authbind, CFLAGS and LDFLAGS are overwritten by the upstream makefile (and are also hardcoded in debian/rules). Therefore, flags specified via dpkg-buildflags are ignored. In Ubuntu, the attached patch was applied to enable frame-pointers and enable LTO. Thanks for considering the patch. -- System Information: Debian Release: trixie/sid APT prefers noble-updates APT policy: (500, 'noble-updates'), (500, 'noble-security'), (500, 'noble'), (100, 'noble-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.8.0-36-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru authbind-2.1.3build2/debian/control authbind-2.1.3ubuntu1~ppa2/debian/control --- authbind-2.1.3build2/debian/control 2024-04-08 17:54:08.0 +0200 +++ authbind-2.1.3ubuntu1~ppa2/debian/control 2024-07-03 17:42:33.0 +0200 @@ -1,8 +1,7 @@ Source: authbind Section: utils Priority: optional -Maintainer: Ubuntu Developers -XSBC-Original-Maintainer: Ian Jackson +Maintainer: Ian Jackson Standards-Version: 3.9.1 Package: authbind diff -Nru authbind-2.1.3build2/debian/patches/makefile-honor-cflags-ldflags.patch authbind-2.1.3ubuntu1~ppa2/debian/patches/makefile-honor-cflags-ldflags.patch --- authbind-2.1.3build2/debian/patches/makefile-honor-cflags-ldflags.patch 1970-01-01 01:00:00.0 +0100 +++ authbind-2.1.3ubuntu1~ppa2/debian/patches/makefile-honor-cflags-ldflags.patch 2024-07-03 17:42:33.0 +0200 @@ -0,0 +1,28 @@ +Description: Honor CFLAGS and LDFLAGS set by the maintainer + The upstream makefile used to override any CFLAGS / LDFLAGS set by the + maintainer when building authbind. This prevents us from controlling the + presence of frame pointers when building. It also prevents us from enabling + LTO. + Fixed by dropping directives that overwrite CFLAGS & LDFLAGS ; and only adding + to those variables when relevant. +Author: Olivier Gayot +Bug-Ubuntu: https://launchpad.net/bugs/2071841 +Forwarded: no +Last-Update: 2024-07-03 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +--- a/Makefile b/Makefile +@@ -34,11 +34,8 @@ + INSTALL_DIR ?= install -o root -g root -m 755 -d + STRIP ?= strip + +-OPTIMISE= -O2 +-LDFLAGS= -g + LIBS= -ldl -lc +-CFLAGS= -g $(OPTIMISE) \ +- -Wall -Wwrite-strings -Wpointer-arith -Wimplicit \ ++CFLAGS+= -Wall -Wwrite-strings -Wpointer-arith -Wimplicit \ + -Wnested-externs -Wmissing-prototypes -Wstrict-prototypes + CPPFLAGS= -DMAJOR_VER='"$(MAJOR)"' -DMINOR_VER='"$(MINOR)"' \ + -DLIBAUTHBIND='"$(lib_dir)/$(LIBCANON)"' \ diff -Nru authbind-2.1.3build2/debian/patches/series authbind-2.1.3ubuntu1~ppa2/debian/patches/series --- authbind-2.1.3build2/debian/patches/series 1970-01-01 01:00:00.0 +0100 +++ authbind-2.1.3ubuntu1~ppa2/debian/patches/series2024-07-03 17:42:33.0 +0200 @@ -0,0 +1 @@ +makefile-honor-cflags-ldflags.patch diff -Nru authbind-2.1.3build2/debian/rules authbind-2.1.3ubuntu1~ppa2/debian/rules --- authbind-2.1.3build2/debian/rules 2021-12-17 19:34:02.0 +0100 +++ authbind-2.1.3ubuntu1~ppa2/debian/rules 2024-07-03 17:42:33.0 +0200 @@ -41,12 +41,10 @@ INSTALL_SCRIPT = $(INSTALL) -p-o root -g root -m 755 INSTALL_DIR = $(INSTALL) -p -d -o root -g root -m 755 -CFLAGS = -O2 -Wall -LDFLAGS = +DPKG_EXPORT_BUILDFLAGS := 1 -CFLAGS += -g -LDFLAGS += -g +include /usr/share/dpkg/buildflags.mk ifeq (,$(findstring nostrip,$(DEB_BUILD_OPTIONS))) STRIP=$(TOOL_PREFIX)strip @@ -55,8 +53,6 @@ STRIP=: endif -export CFLAGS -export LDFLAGS export INSTALL export INSTALL_FILE export INSTALL_PROGRAM
Bug#1067665: cvise 2.10.0-1 still build-depends on python-pytest-flake8 (broken with recent flake8)
Package: cvise Version: 2.10.0-1 Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu noble ubuntu-patch Dear Maintainer, In version 2.10.0, upstream dropped the dependency on pytest-flake8 because it is no longer maintained [1] and broken with modern flake8. However, in 2.10.0-1 in Debian, the build dependency on python3-pytest-flake8 is still present, albeit unused during the build process. This is preventing removal of python-pytest-flake8. In Ubuntu, the attached patch was applied to achieve the following: * Drop dependency on python-pytest-flake8 (broken with modern flake8) which was already dropped upstream in cvise 2.10 and is now unused (LP: #2058917) Thanks for considering the patch. -- System Information: Debian Release: trixie/sid APT prefers noble APT policy: (500, 'noble') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.8.0-11-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru cvise-2.10.0/debian/control cvise-2.10.0/debian/control --- cvise-2.10.0/debian/control 2024-03-12 16:49:12.0 +0100 +++ cvise-2.10.0/debian/control 2024-03-25 10:45:07.0 +0100 @@ -13,7 +13,6 @@ python3-pebble, python3-psutil, python3-pytest , - python3-pytest-flake8 , llvm-18-dev, libclang-18-dev, clang-18, clang-format-18, # clang-tools-18, clang-tidy-18, clangd-18, unifdef,
Bug#1067466: black autopkgtest fails because python3-typed-ast no longer in archive
Package: black Version: 24.2.0-1 Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu noble ubuntu-patch Dear Maintainer, Since the removal of python3-typed-ast from the archive [1], black cannot install its test dependencies anymore and therefore fails autopkg tests. https://tracker.debian.org/news/1469662/python3-typed-ast-removed-from-testing/ The AST library is provided by the Python standard library since Python 3.8 so the dependency has been unused for some time and can be safely dropped. In src:black 22.10.0-2, the dependency on python3-typed-ast was dropped from d/control but not from d/test/control: https://metadata.ftp-master.debian.org/changelogs//main/b/black/black_24.2.0-1_changelog In Ubuntu, the attached patch was applied to achieve the following: * d/test/control: drop unused dependency on python3-typed-ast, which was dropped from the archive. (LP: #2058692) Thanks for considering the patch. -- System Information: Debian Release: trixie/sid APT prefers noble APT policy: (500, 'noble') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.8.0-11-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru black-24.2.0/debian/tests/control black-24.2.0/debian/tests/control --- black-24.2.0/debian/tests/control 2024-02-15 10:47:43.0 +0100 +++ black-24.2.0/debian/tests/control 2024-03-21 23:04:40.0 +0100 @@ -9,7 +9,6 @@ python3-pytest, python3-pathspec, python3-regex, - python3-typed-ast, python3-aiohttp, python3-tomli, python3-mypy-extensions,
Bug#1067168: urwid raises when rendering an empty Pile or Columns as a flow widget
Package: urwid Version: 2.6.4-1 Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu noble ubuntu-patch Dear Maintainer, With the version of urwid 2.6.4-1 currently in trixie, the following code fails with an exception: ```python from urwid import Pile Pile([ ("pack", Pile([])), ]).render((10,)) ``` File "/usr/lib/python3/dist-packages/urwid/widget/widget.py", line 112, in cached_render canv = fn(self, size, focus=focus) ^^^ File "/usr/lib/python3/dist-packages/urwid/widget/pile.py", line 822, in render _widths, heights, size_args = self.get_rows_sizes(size, focus) File "/usr/lib/python3/dist-packages/urwid/widget/pile.py", line 730, in get_rows_sizes heights.append(w.pack(w_h_arg, item_focus)[1]) ^^^ File "/usr/lib/python3/dist-packages/urwid/widget/pile.py", line 744, in pack return super().pack(size, focus) ^ File "/usr/lib/python3/dist-packages/urwid/widget/widget.py", line 401, in pack raise WidgetError(f"Cannot pack (maxcol,) size, this is not a flow widget: {self!r}") urwid.widget.widget.WidgetError: Cannot pack (maxcol,) size, this is not a flow widget: The same code used to work with urwid 2.1.2-4 in bookworm. I applied a fix from upstream [1] that was included in urwid 2.6.5. I think the proper way forward would be to take a more recent upstream version of urwid. That said, we are in feature freeze downstream in Ubuntu so I skipped the refactoring bits. In Ubuntu, the attached patch was applied to achieve the following: * Apply upstream patch to fix a crash when rendering an empty Pile or an empty Columns as a flow widget. (LP: #2058388) + d/patches/fix-crash-empty-pile.patch Thanks for considering the patch. -- System Information: Debian Release: trixie/sid APT prefers noble APT policy: (500, 'noble'), (100, 'noble-proposed') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.8.0-11-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled [1] https://github.com/urwid/urwid/commit/83c278b53de431a9b41d7ddadf5f318914246593 diff -Nru urwid-2.6.4/debian/patches/fix-crash-empty-pile.patch urwid-2.6.4/debian/patches/fix-crash-empty-pile.patch --- urwid-2.6.4/debian/patches/fix-crash-empty-pile.patch 1970-01-01 01:00:00.0 +0100 +++ urwid-2.6.4/debian/patches/fix-crash-empty-pile.patch 2024-03-19 14:23:31.0 +0100 @@ -0,0 +1,158 @@ +Description: Fix crash when rendering empty Pile or Columns as a flow widget + Special case: in case of `Columns`/`Pile` empty - use fallback sizing (#843) + . + * Extend `repr` to provide brief info about contents + . +Author: Aleksei Stepanov +Origin: upstream, https://github.com/urwid/urwid/commit/83c278b53de431a9b41d7ddadf5f318914246593 +Bug-Ubuntu: https://launchpad.net/bugs/2058388 +Last-Update: 2024-03-19 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +diff --git a/urwid/widget/columns.py b/urwid/widget/columns.py +index e3fa134..5a3421f 100644 +--- a/urwid/widget/columns.py b/urwid/widget/columns.py +@@ -66,40 +66,47 @@ class Columns(Widget, WidgetContainerMixin, WidgetContainerListContentsMixin): + + # BOX-only widget + >>> Columns((SolidFill("#"),)) +- ++ + + # BOX-only widget with "get height from max" + >>> Columns((SolidFill("#"),), box_columns=(0,)) +- ++ + + # FLOW-only + >>> Columns((Edit(),)) +- ++ + + # FLOW allowed by "box_columns" + >>> Columns((Edit(), SolidFill("#")), box_columns=(1,)) +- ++ + + # FLOW/FIXED + >>> Columns((Text("T"),)) +- ++ + + # GIVEN BOX only -> BOX only + >>> Columns(((5, SolidFill("#")),), box_columns=(0,)) +- ++ + + # No FLOW - BOX only + >>> Columns(((5, SolidFill("#")), SolidFill("*")), box_columns=(0, 1)) +- ++ + + # FIXED only -> FIXED + >>> Columns(((WHSettings.PACK, BigText("1", font)),)) +- ++ + + # Invalid sizing combination -> use fallback settings (and produce warning) + >>> Columns(((WHSettings.PACK, SolidFill("#")),)) +- ++ ++ ++# Special case: empty columns widget sizing is impossible to calculate ++>>> Columns(()) ++ + """ ++if not self.contents: ++return frozenset((urwid.BOX, urwid.FLOW)) ++ + strict_box = False + has_flow = False + +@@ -280,6 +287,13 @@ class
Bug#1058392: python-oauth2client: FTBFS: failed tests
Package: python-oauth2client Followup-For: Bug #1058392 User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu noble ubuntu-patch Control: tags -1 patch Dear Maintainer, Python 3.12 dropped support in unittest.TestCase for assertRaisesRegexp and assertEquals. These two functions were provided as a compability layer since their deprecation in Python 3.2. In Ubuntu, the attached patch was applied to achieve the following: * Replace obsolete assert* calls (dropped from Python 3.12) with their correct alternative to fix FTBFS. - debian/patches/python312-support.patch Thanks for considering the patch. -- System Information: Debian Release: trixie/sid APT prefers mantic-updates APT policy: (500, 'mantic-updates'), (500, 'mantic-security'), (500, 'mantic'), (100, 'mantic-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru python-oauth2client-4.1.3/debian/patches/python312-support.patch python-oauth2client-4.1.3/debian/patches/python312-support.patch --- python-oauth2client-4.1.3/debian/patches/python312-support.patch 1970-01-01 01:00:00.0 +0100 +++ python-oauth2client-4.1.3/debian/patches/python312-support.patch 2024-02-16 10:51:37.0 +0100 @@ -0,0 +1,252 @@ +Description: Fix use of obsolete assert statements dropped in Python 3.12 + Since Python 3.2, assertEquals and assertRaisesRegexp (+ others) have been + deprecated in favor of assertEqual and assertRaisesRegex. + In Python 3.12, the deprecated names got dropped, causing oauth2client to + FTBFS. + . + The patch is not meant for upstream inclusion since the project upstream is + deprecated (in favor of python-google-auth) and the repository is read only. +Author: Olivier Gayot +Bug-Debian: https://bugs.debian.org/1058392 +Forwarded: not-needed +Last-Update: 2024-02-16 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +--- a/tests/contrib/appengine/test_appengine.py b/tests/contrib/appengine/test_appengine.py +@@ -906,7 +906,7 @@ + self.decorator = decorator + self.test_required() + http = self.decorator.http() +-self.assertEquals('foo_access_token', ++self.assertEqual('foo_access_token', + http.request.credentials.access_token) + + # revoke_uri is not required +--- a/tests/contrib/test_devshell.py b/tests/contrib/test_devshell.py +@@ -205,7 +205,7 @@ + def test_no_refresh_token(self): + with _AuthReferenceServer(): + creds = devshell.DevshellCredentials() +-self.assertEquals(None, creds.refresh_token) ++self.assertIsNone(creds.refresh_token) + + @mock.patch('oauth2client.client._UTCNOW') + def test_reads_credentials(self, utcnow): +--- a/tests/contrib/test_keyring_storage.py b/tests/contrib/test_keyring_storage.py +@@ -104,7 +104,7 @@ +autospec=True) as get_password: + store = keyring_storage.Storage('my_unit_test', 'me') + credentials = store.get() +-self.assertEquals(None, credentials) ++self.assertIsNone(credentials) + get_password.assert_called_once_with('my_unit_test', 'me') + + def test_get_with_malformed_json_credentials_stored(self): +@@ -113,7 +113,7 @@ +autospec=True) as get_password: + store = keyring_storage.Storage('my_unit_test', 'me') + credentials = store.get() +-self.assertEquals(None, credentials) ++self.assertIsNone(credentials) + get_password.assert_called_once_with('my_unit_test', 'me') + + def test_get_and_set_with_json_credentials_stored(self): +@@ -139,7 +139,7 @@ +return_value=None, +autospec=True) as set_password: + store = keyring_storage.Storage('my_unit_test', 'me') +-self.assertEquals(None, store.get()) ++self.assertIsNone(store.get()) + + store.put(credentials) + +--- a/tests/test_client.py b/tests/test_client.py +@@ -436,7 +436,7 @@ + 'File {0} \(pointed by {1} environment variable\) does not ' + 'exist!'.format( + nonexistent_file, client.GOOGLE_APPLICATION_CREDENTIALS)) +-with self.assertRaisesRegexp(client.ApplicationDefaultCredentialsError, ++with self.assertRaisesRegex(client.ApplicationDefaultCredentialsError, + expected_err_msg): + client._get_environment_variable_file() + +@@ -541,7 +541,7 @@ + "'{1}' values\)&qu
Bug#1063782: python-oauth2client: FTBFS with flask 3
Package: python-oauth2client Followup-For: Bug #1063782 User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu noble ubuntu-patch Control: tags -1 patch Dear Maintainer, Flask 3.0 dropped support for _app_ctx_stack and _request_ctx_stack. Those attributes were deprecated since Flask 2.2. In Ubuntu, the attached patch was applied to achieve the following: * Replace use of flask._app_ctx_stack (dropped from flask 3) by flask.g to fix FTBFS. (LP: #2052771) - debian/patches/flask3-use-g-instead-of-app_ctx_stack.patch Thanks for considering the patch. -- System Information: Debian Release: trixie/sid APT prefers mantic-updates APT policy: (500, 'mantic-updates'), (500, 'mantic-security'), (500, 'mantic'), (100, 'mantic-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru python-oauth2client-4.1.3/debian/patches/flask3-use-g-instead-of-app_ctx_stack.patch python-oauth2client-4.1.3/debian/patches/flask3-use-g-instead-of-app_ctx_stack.patch --- python-oauth2client-4.1.3/debian/patches/flask3-use-g-instead-of-app_ctx_stack.patch 1970-01-01 01:00:00.0 +0100 +++ python-oauth2client-4.1.3/debian/patches/flask3-use-g-instead-of-app_ctx_stack.patch 2024-02-16 10:51:37.0 +0100 @@ -0,0 +1,43 @@ +Description: Use flask.g instead of flask._app_ctx_stack + In flask 2.2, _app_ctx_stack and _request_ctx_stack got deprecated in favor of + using flask.g. The compatibility layer was dropped from flask 3, which now raises: + E ImportError: cannot import name '_app_ctx_stack' from 'flask' + Replaced by the use of flask.g as mentioned in the documentation. + . + The patch is not meant for upstream inclusion since the project upstream is + deprecated (in favor of python-google-auth) and the repository is read only. +Author: Olivier Gayot +Bug-Ubuntu: https://launchpad.net/bugs/2052771 +Bug-Debian: https://bugs.debian.org/1063782 +Forwarded: not-needed +Last-Update: 2024-02-16 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +--- a/oauth2client/contrib/flask_util.py b/oauth2client/contrib/flask_util.py +@@ -170,8 +170,8 @@ + + try: + from flask import Blueprint +-from flask import _app_ctx_stack + from flask import current_app ++from flask import g as flask_g + from flask import redirect + from flask import request + from flask import session +@@ -434,12 +434,10 @@ + @property + def credentials(self): + """The credentials for the current user or None if unavailable.""" +-ctx = _app_ctx_stack.top ++if not hasattr(flask_g, _CREDENTIALS_KEY): ++flask_g.google_oauth2_credentials = self.storage.get() + +-if not hasattr(ctx, _CREDENTIALS_KEY): +-ctx.google_oauth2_credentials = self.storage.get() +- +-return ctx.google_oauth2_credentials ++return flask_g.google_oauth2_credentials + + def has_credentials(self): + """Returns True if there are valid credentials for the current user.""" diff -Nru python-oauth2client-4.1.3/debian/patches/series python-oauth2client-4.1.3/debian/patches/series --- python-oauth2client-4.1.3/debian/patches/series 2023-08-17 18:22:35.0 +0200 +++ python-oauth2client-4.1.3/debian/patches/series 2024-02-16 10:51:37.0 +0100 @@ -1,3 +1,5 @@ +flask3-use-g-instead-of-app_ctx_stack.patch skip-network-doing-unit-test.patch remove-broken-tests.patch fix-hmac.new-call-in-py3.8.patch
Bug#1064017: python-requests-unixsocket: Please run python-requests-unixsocket test-suite as part of autopkgtest
Package: python-requests-unixsocket Severity: wishlist Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu noble ubuntu-patch Dear Maintainer, The recent migration of urllib3 and src:requests broke python-requests-unixsocket (see bug 1063892 [1]). This could have been easily noticed if request_unisocket's test-suite had run as an autopkg test. python-requests-unixsocket does not use pybuild so I came up with a small d/t/control + d/t/python3-requests-unixsocket to make the test-suite run during autopkgtest execution. In Ubuntu, the attached patch was applied to achieve the following: * Execute test-suite as part of autopkgtest (LP: #2053267) Thanks for considering the patch. [1] https://bugs.debian.org/1063892 -- System Information: Debian Release: trixie/sid APT prefers mantic-updates APT policy: (500, 'mantic-updates'), (500, 'mantic-security'), (500, 'mantic'), (100, 'mantic-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru python-requests-unixsocket-0.3.0/debian/tests/control python-requests-unixsocket-0.3.0/debian/tests/control --- python-requests-unixsocket-0.3.0/debian/tests/control 1970-01-01 01:00:00.0 +0100 +++ python-requests-unixsocket-0.3.0/debian/tests/control 2024-02-15 19:59:04.0 +0100 @@ -0,0 +1,2 @@ +Tests: python3-requests-unixsocket +Depends: @, @builddeps@ diff -Nru python-requests-unixsocket-0.3.0/debian/tests/python3-requests-unixsocket python-requests-unixsocket-0.3.0/debian/tests/python3-requests-unixsocket --- python-requests-unixsocket-0.3.0/debian/tests/python3-requests-unixsocket 1970-01-01 01:00:00.0 +0100 +++ python-requests-unixsocket-0.3.0/debian/tests/python3-requests-unixsocket 2024-02-15 19:59:02.0 +0100 @@ -0,0 +1,13 @@ +#!/bin/bash + +set -efu + +python3_all="$(py3versions -s 2>/dev/null)" + +cp -r requests_unixsocket/tests "$AUTOPKGTEST_TMP/" +cd "$AUTOPKGTEST_TMP" + +for py in $python3_all; do +echo "=== $py ===" +$py -m pytest --verbose 2>&1 +done
Bug#1063892: python-requests-unixsocket: HTTPConnection.request() got an unexpected keyword argument 'chunked' with requests 2.29+
Package: python-requests-unixsocket Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu noble ubuntu-patch Dear Maintainer, When using python-requests-unixsocket 0.3.0-3 against requests 2.29+ (currently 2.31.0+dfsg-1 is in testing), the following code produces an exception: ```python3 import requests_unixsocket with requests_unixsocket.Session() as session: session.get('http+unix://%2Ftmp/') ``` TypeError: HTTPConnection.request() got an unexpected keyword argument 'chunked' There is a bug upstream [1] for it. And I think the regression was introduced by PR 6226 [2] in the underlying library (requests). I applied a patch that was proposed upstream but not yet merged. In Ubuntu, the attached patch was applied to achieve the following: * Fix incompatibility with requests 2.29 (LP: #2053016). - d/p/0001-Inherit-HTTPConnection-through-urllib3.connection-no.patch Thanks for considering the patch. [1] https://github.com/msabramo/requests-unixsocket/issues/70 [2] https://github.com/psf/requests/pull/6226 -- System Information: Debian Release: trixie/sid APT prefers mantic-updates APT policy: (500, 'mantic-updates'), (500, 'mantic-security'), (500, 'mantic'), (100, 'mantic-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru python-requests-unixsocket-0.3.0/debian/patches/0001-Inherit-HTTPConnection-through-urllib3.connection-no.patch python-requests-unixsocket-0.3.0/debian/patches/0001-Inherit-HTTPConnection-through-urllib3.connection-no.patch --- python-requests-unixsocket-0.3.0/debian/patches/0001-Inherit-HTTPConnection-through-urllib3.connection-no.patch 1970-01-01 01:00:00.0 +0100 +++ python-requests-unixsocket-0.3.0/debian/patches/0001-Inherit-HTTPConnection-through-urllib3.connection-no.patch 2024-02-14 11:16:52.0 +0100 @@ -0,0 +1,47 @@ +Description: Inherit HTTPConnection through urllib3.connection, not httplib + By inheriting from `urllib3.connection.HTTPConnection` (that inherits + from `httplib.HTTPConnection` itself), we can adapt to the internal + changes in urllib3 2.0 that added a `request()` method that is + incompatible with httplib.HTTPConnection.request. + . + This fixes the incompatibility between urllib3 2.0 and requests 1.26+, + which was the first version that stopped vendoring urllib3. + . + Reference: https://github.com/docker/docker-py/issues/3113#issuecomment-1531570788 + Fixes: #70 +Author: Martin Roukala +Origin: upstream, https://github.com/msabramo/requests-unixsocket/pull/69 +Bug: https://github.com/msabramo/requests-unixsocket/issues/70 +Bug-Ubuntu: https://launchpad.net/bugs/2053016 +Applied-Upstream: no +Last-Update: 2024-02-14 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +diff --git a/requests_unixsocket/adapters.py b/requests_unixsocket/adapters.py +index 83e1400..513c243 100644 +--- a/requests_unixsocket/adapters.py b/requests_unixsocket/adapters.py +@@ -3,11 +3,6 @@ import socket + from requests.adapters import HTTPAdapter + from requests.compat import urlparse, unquote + +-try: +-import http.client as httplib +-except ImportError: +-import httplib +- + try: + from requests.packages import urllib3 + except ImportError: +@@ -16,7 +11,7 @@ except ImportError: + + # The following was adapted from some code from docker-py + # https://github.com/docker/docker-py/blob/master/docker/transport/unixconn.py +-class UnixHTTPConnection(httplib.HTTPConnection, object): ++class UnixHTTPConnection(urllib3.connection.HTTPConnection, object): + + def __init__(self, unix_socket_url, timeout=60): + """Create an HTTP connection to a unix domain socket +-- +2.40.1 + diff -Nru python-requests-unixsocket-0.3.0/debian/patches/series python-requests-unixsocket-0.3.0/debian/patches/series --- python-requests-unixsocket-0.3.0/debian/patches/series 1970-01-01 01:00:00.0 +0100 +++ python-requests-unixsocket-0.3.0/debian/patches/series 2024-02-14 11:12:51.0 +0100 @@ -0,0 +1 @@ +0001-Inherit-HTTPConnection-through-urllib3.connection-no.patch
Bug#1063782: python-oauth2client: FTBFS with flask 3
Source: python-oauth2client Severity: important Dear Maintainer, Building (or running autopkgtest) for python-oauth2client against flask 3.0 leads to an error when running the test-suite: Traceback: oauth2client/contrib/flask_util.py:173: in from flask import _app_ctx_stack E ImportError: cannot import name '_app_ctx_stack' from 'flask' (/usr/lib/python3/dist-packages/flask/__init__.py) With previous versions of flask, the same code would produce a warning instead: oauth2client/contrib/flask_util.py:173: DeprecationWarning: '_app_ctx_stack' is deprecated and will be removed in Flask 2.3. Flask decided to remove the deprecated _app_ctx_stack [1] in 3.0.0. The recommended approach is to "use 'g' to store data instead." Quote from Version 2.2.0 changelog [2]: > The app and request contexts are managed using Python context vars directly > rather than Werkzeug’s LocalStack. This should result in better performance > and memory use. #4682 > > Extension maintainers, be aware that _app_ctx_stack.top and > _request_ctx_stack.top are deprecated. Store data on g instead using a unique > prefix, like g._extension_name_attr. > Nowadays oauth2client is marked deprecated [3] and the upstream repository [4] has been read only for four years. NOTE: When building with python3.12, there are other errors but they have already been reported as part of bug 1058392 [5] Thank you [1] https://github.com/pallets/flask/pull/5223 [2] https://flask.palletsprojects.com/en/2.3.x/changes/#version-2-2-0 [3] https://google-auth.readthedocs.io/en/latest/oauth2client-deprecation.html [4] https://github.com/googleapis/oauth2client [5] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1058392 -- System Information: Debian Release: trixie/sid APT prefers mantic-updates APT policy: (500, 'mantic-updates'), (500, 'mantic-security'), (500, 'mantic'), (100, 'mantic-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled
Bug#1057530: nvme-stas: autopkgtest fail when runners have esoteric network interfaces
Package: nvme-stas Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu noble ubuntu-patch Dear Maintainer, When an autopkgtest runner has esoteric network interfaces configured, the test-suite (i.e., test-udev.py and/or test-iputil.py) is likely to fail. The following types of interfaces are known to cause failures on autopkgtest runners in Ubuntu: * vlan * gre * dummy * veth I am not sure if Debian autopkgtest runners would be affected but the test suite is a bit fragile at the moment, so I think the patches would make sense. NOTE: Ubuntu is ahead of Debian by two patches, so the autopkgtest failures might not unravel until the following are addressed: * https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1054533 * https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1057031 In Ubuntu, the attached patch was applied to achieve the following: * Skip mac2iface test for esoteric network interfaces (LP: #2045690) * Add upstream patch to fix udev test for esoteric network interfaces Thanks for considering the patch. -- System Information: Debian Release: trixie/sid APT prefers mantic-updates APT policy: (500, 'mantic-updates'), (500, 'mantic-security'), (500, 'mantic') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru nvme-stas-2.3/debian/patches/fix-test-udev-with-dummy-interfaces.patch nvme-stas-2.3/debian/patches/fix-test-udev-with-dummy-interfaces.patch --- nvme-stas-2.3/debian/patches/fix-test-udev-with-dummy-interfaces.patch 1970-01-01 01:00:00.0 +0100 +++ nvme-stas-2.3/debian/patches/fix-test-udev-with-dummy-interfaces.patch 2023-12-05 21:04:32.0 +0100 @@ -0,0 +1,26 @@ +Description: Fix test-udev with dummy interfaces +Author: Martin Belanger +Origin: upstream, https://github.com/linux-nvme/nvme-stas/commit/371c1e875a9a660adca241265e4cd460ee7e5e5c +Bug: https://github.com/linux-nvme/nvme-stas/issues/407 +Last-Update: 2023-12-05 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +diff --git a/test/test-udev.py b/test/test-udev.py +index ba484e0..9621edc 100755 +--- a/test/test-udev.py b/test/test-udev.py +@@ -295,6 +295,11 @@ class Test(unittest.TestCase): + def test__cid_matches_tid(self): + ifaces = iputil.net_if_addrs() + for ifname, addrs in self.ifaces.items(): ++# contains a subset of the interfaces found in . ++# So, let's make sure that we only test with the interfaces found in both. ++if ifname not in ifaces: ++continue ++ + ## + # IPV4 + +-- +2.40.1 + diff -Nru nvme-stas-2.3/debian/patches/series nvme-stas-2.3/debian/patches/series --- nvme-stas-2.3/debian/patches/series 2023-11-28 11:01:46.0 +0100 +++ nvme-stas-2.3/debian/patches/series 2023-12-05 21:04:32.0 +0100 @@ -1,3 +1,5 @@ fix-test-libnvme-version.patch fix-test-udev-failing-multiple-IPv6.patch fix-iputil-endianness-issue.patch +skip-mac2iface-test-for-esoteric-interfaces.patch +fix-test-udev-with-dummy-interfaces.patch diff -Nru nvme-stas-2.3/debian/patches/skip-mac2iface-test-for-esoteric-interfaces.patch nvme-stas-2.3/debian/patches/skip-mac2iface-test-for-esoteric-interfaces.patch --- nvme-stas-2.3/debian/patches/skip-mac2iface-test-for-esoteric-interfaces.patch 1970-01-01 01:00:00.0 +0100 +++ nvme-stas-2.3/debian/patches/skip-mac2iface-test-for-esoteric-interfaces.patch 2023-12-05 21:04:32.0 +0100 @@ -0,0 +1,70 @@ +Description: Skip mac2iface test for esoteric interfaces + mac2iface takes a MAC address as argument and returns the corresponding + interface (if any). + The mac2iface tests will however invoke mac2iface with invalid MAC addresses + when esoteric network interfaces are present on the system. As an example, + armhf autopkgtest runners in Ubuntu have gre interfaces configured so the + test-suite fails. + . + We now ensure that the test-suite calls mac2iface with only valid MAC + addresses. + . + Furthermore, sometimes the same MAC address is assigned to more than one + interface on the system (this is true for VLAN interfaces for instance). This + confuses mac2iface, which returns only the first match. This scenario is + possibly more of a nvme-stas bug than a test-suite bug, but for now we will + just skip the interfaces that have a duplicate MAC address. +Author: Olivier Gayot +Bug-Ubuntu: https://launchpad.net/bugs/2045690 +Forwarded: https://github.com/linux-nvme/nvme-stas/pull/411 +Last-Update: 2023-12-05 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +--- a/test
Bug#1057467: lomiri-download-manager: Use of GetConnectionAppArmorSecurityContext dropped from DBus years ago
Package: lomiri-download-manager Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu noble ubuntu-patch Dear Maintainer, During runtime, lomiri-download-manager still tries to consume the GetConnectionAppArmorSecurityContext DBus method, which was added in Ubuntu around 10 years ago and never made it upstream. Instead, upstream DBus added another way to access the AppArmor security context through GetConnectionCredentials [1]. The GetConnectionAppArmorSecurityContext method does not exist anymore in Debian ; so there is no point trying to use it. I also submitted the patch upstream [2]. In Ubuntu, the attached patch was applied to achieve the following: * Replace deprecated calls to GetConnectionAppArmorSecurityContext by calls to GetConnectionCredentials (LP: #1489489). Thanks for considering the patch. [1] https://dbus.freedesktop.org/doc/dbus-specification.html#bus-messages-get-connection-credentials [2] https://gitlab.com/ubports/development/core/lomiri-download-manager/-/merge_requests/24 -- System Information: Debian Release: trixie/sid APT prefers mantic-updates APT policy: (500, 'mantic-updates'), (500, 'mantic-security'), (500, 'mantic') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru lomiri-download-manager-0.1.2/debian/patches/1001-drop-deprecated-GetConnectionAppArmorSecurityContext.patch lomiri-download-manager-0.1.2/debian/patches/1001-drop-deprecated-GetConnectionAppArmorSecurityContext.patch --- lomiri-download-manager-0.1.2/debian/patches/1001-drop-deprecated-GetConnectionAppArmorSecurityContext.patch 1970-01-01 01:00:00.0 +0100 +++ lomiri-download-manager-0.1.2/debian/patches/1001-drop-deprecated-GetConnectionAppArmorSecurityContext.patch 2023-12-05 10:20:56.0 +0100 @@ -0,0 +1,386 @@ +Description: Drop deprecated calls to GetConnectionAppArmorSecurityContext + Around 10 years ago, we introduced GetConnectionAppArmorSecurityContext() in + DBus in Ubuntu. Upstream pushed back and instead exposed the AppArmor's + context in org.freedesktop.DBus.GetConnectionCredentials(). + So far, we have maintained GetconnectionAppArmorSecurityContext() as a delta + in Ubuntu so that existing software do not break. + However, the recommended way (and the only portable way) is to use + GetConnectionCredentials() instead. Furthermore, it is about time we drop + support for the former in Ubuntu. + . + The patch is inspired from 4c4d7961261b41ac41f24c8ce75563ab12f6b74c from the + https://gitlab.com/ubports/development/core/lomiri-thumbnailer/ repository. +Author: Olivier Gayot +Bug-Ubuntu: https://launchpad.net/bugs/1489489 +Forwarded: https://gitlab.com/ubports/development/core/lomiri-download-manager/-/merge_requests/24 +Last-Update: 2023-12-05 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +--- a/src/common/priv/lomiri/transfers/system/apparmor.cpp b/src/common/priv/lomiri/transfers/system/apparmor.cpp +@@ -19,6 +19,7 @@ + #include + #include + #include ++#include + #include + #include + #include +@@ -37,6 +38,8 @@ + namespace System { + + QString AppArmor::UNCONFINED_ID = "unconfined"; ++QString AppArmor::LINUX_SECURITY_LABEL = "LinuxSecurityLabel"; ++ + + AppArmor::AppArmor(DBusConnection* connection, + QObject* parent) +@@ -68,15 +71,38 @@ + } + + QString ++AppArmor::labelFromCredentials(const QVariantMap ) ++{ ++// The contents of this map are described in the specification here: ++// http://dbus.freedesktop.org/doc/dbus-specification.html#bus-messages-get-connection-credentials ++QByteArray label = map.value(LINUX_SECURITY_LABEL).value(); ++ ++if (label.size() == 0) { ++return ""; ++} ++ ++// The label is null terminated. ++assert(label[label.size()-1] == '\0'); ++label.truncate(label.size() - 1); ++// Trim the mode off the end of the label. ++int pos = label.lastIndexOf(' '); ++if (pos > 0 && label.endsWith(')') && label[pos+1] == '(') ++{ ++label.truncate(pos); // LCOV_EXCL_LINE ++} ++return QString::fromUtf8(label.constData(), label.size()); ++} ++ ++QString + AppArmor::appId(QString caller) { +-QScopedPointer > reply( +-_dbus->GetConnectionAppArmorSecurityContext(caller)); ++QScopedPointer > reply( ++_dbus->GetConnectionCredentials(caller)); + // blocking but should be ok for now + reply->waitForFinished(); + if (reply->isError()) { + return ""; + } +-return reply->value(); ++return labelFro
Bug#1056533: testresources autopkg tests fail with Python 3.12
Package: testresources Followup-For: Bug #1056533 User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu noble ubuntu-patch Control: tags -1 patch Dear Maintainer, The testresources autopkgtest test-suite currently fails with Python 3.12 because it still uses unittest.TestCase.failIf. failIf was a deprecated alias for assertFalse and was dropped from Python 3.12. Applying a patch from upstream allows the test-suite to pass with Python 3.12. In Ubuntu, the attached patch was applied to achieve the following: * Fix test-suite so it can run with Python 3.12. (LP: #2045302) Thanks for considering the patch. -- System Information: Debian Release: trixie/sid APT prefers mantic-updates APT policy: (500, 'mantic-updates'), (500, 'mantic-security'), (500, 'mantic') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru testresources-2.0.1/debian/patches/python3.12.patch testresources-2.0.1/debian/patches/python3.12.patch --- testresources-2.0.1/debian/patches/python3.12.patch 1970-01-01 01:00:00.0 +0100 +++ testresources-2.0.1/debian/patches/python3.12.patch 2023-11-30 18:59:39.0 +0100 @@ -0,0 +1,35 @@ +Description: Refactor failIf to assertFalse for Python 3.12 Compatibility + This commit replaces deprecated failIf calls with assertFalse in the + test_resourced_test_case.py file. The failIf method was removed in + Python 3.12 [1-3]. + . + [1] https://docs.python.org/3.12/whatsnew/3.12.html#removed + [2] https://github.com/python/cpython/issues/89325 + [3] https://github.com/python/cpython/pull/28268 +Author: Petr Vaněk +Origin: upstream, https://github.com/testing-cabal/testresources/pull/15 +Bug: https://launchpad.net/bugs/2045302 +Bug-Ubuntu: https://launchpad.net/bugs/2045302 +Bug-Debian: https://bugs.debian.org/1056533 +Applied-Upstream: https://github.com/testing-cabal/testresources/commit/7bb62a13fa1d28717c10f3152b5e8ea479c8e9d2 +Last-Update: 2023-11-30 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +--- a/testresources/tests/test_resourced_test_case.py b/testresources/tests/test_resourced_test_case.py +@@ -129,7 +129,7 @@ + self.resourced_case.resources = [("foo", self.resource_manager)] + self.resourced_case.setUpResources() + self.resourced_case.tearDownResources() +-self.failIf(hasattr(self.resourced_case, "foo")) ++self.assertFalse(hasattr(self.resourced_case, "foo")) + + def testTearDownResourcesStopsUsingResource(self): + # tearDownResources records that there is one less use of each +@@ -158,5 +158,5 @@ + self.assertEqual(self.resourced_case.foo, self.resource) + self.assertEqual(self.resource_manager._uses, 1) + self.resourced_case.tearDown() +-self.failIf(hasattr(self.resourced_case, "foo")) ++self.assertFalse(hasattr(self.resourced_case, "foo")) + self.assertEqual(self.resource_manager._uses, 0) diff -Nru testresources-2.0.1/debian/patches/series testresources-2.0.1/debian/patches/series --- testresources-2.0.1/debian/patches/series 2022-10-17 12:22:49.0 +0200 +++ testresources-2.0.1/debian/patches/series 2023-11-30 18:59:07.0 +0100 @@ -1,2 +1,3 @@ remove-non-deterministic-ftbfs-testBasicSortTests.patch python3.10.patch +python3.12.patch
Bug#1055571: cracklib2 tests fail with Python 3.12
Package: cracklib2 Followup-For: Bug #1055571 User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu noble ubuntu-patch Control: tags -1 patch Dear Maintainer, When attempting to build src:cracklib2 against Python 3.12, the build-time test-suite fails, making the package FTBFS. The test-suite still uses unittest.TestCase.assertEquals() which was a deprecated alias for unittest.TestCase.assertEqual() (without the s). The alias was dropped from Python 3.12. In Ubuntu, the attached patch was applied to achieve the following: * Fix build against Python 3.12. (LP: #2045290) Thanks for considering the patch. -- System Information: Debian Release: trixie/sid APT prefers mantic-updates APT policy: (500, 'mantic-updates'), (500, 'mantic-security'), (500, 'mantic') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru cracklib2-2.9.6/debian/patches/python3.12-support.patch cracklib2-2.9.6/debian/patches/python3.12-support.patch --- cracklib2-2.9.6/debian/patches/python3.12-support.patch 1970-01-01 01:00:00.0 +0100 +++ cracklib2-2.9.6/debian/patches/python3.12-support.patch 2023-11-30 17:46:13.0 +0100 @@ -0,0 +1,77 @@ +Description: Fix build against Python 3.12 + The build-time test suite still used the obsolete assertEquals function - + which was obsolete and got dropped from Python 3.12. Use the replacement + instead so we can build with Python 3.12. +Author: Olivier Gayot +Bug-Ubuntu: https://launchpad.net/bugs/2045290 +Bug-Debian: https://bugs.debian.org/1055571 +Forwarded: https://github.com/cracklib/cracklib/pull/75 +Last-Update: 2023-11-30 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +--- a/python/test_cracklib.py b/python/test_cracklib.py +@@ -76,41 +76,41 @@ + def test_simple_lower(self): + for passwd in ['t' * i for i in range( + cracklib.MIN_LENGTH - cracklib.LOW_CREDIT)]: +-self.assertEquals( ++self.assertEqual( + 1, cracklib.simple(passwd), + 'password {0} should be detected as too simple'.format( + passwd)) +-self.assertEquals(0, cracklib.simple( ++self.assertEqual(0, cracklib.simple( + 't' * (cracklib.MIN_LENGTH - cracklib.LOW_CREDIT))) + + def test_simple_upper(self): + for passwd in ['T' * i for i in range( + cracklib.MIN_LENGTH - cracklib.UP_CREDIT)]: +-self.assertEquals( ++self.assertEqual( + 1, cracklib.simple(passwd), + 'password {0} should be detected as too simple'.format( + passwd)) +-self.assertEquals(0, cracklib.simple( ++self.assertEqual(0, cracklib.simple( + 'T' * (cracklib.MIN_LENGTH - cracklib.UP_CREDIT))) + + def test_simple_digit(self): + for passwd in ['1' * i for i in range( + cracklib.MIN_LENGTH - cracklib.DIG_CREDIT)]: +-self.assertEquals( ++self.assertEqual( + 1, cracklib.simple(passwd), + 'password {0} should be detected as too simple'.format( + passwd)) +-self.assertEquals(0, cracklib.simple( ++self.assertEqual(0, cracklib.simple( + '1' * (cracklib.MIN_LENGTH - cracklib.DIG_CREDIT))) + + def test_simple_other(self): + for passwd in ['#' * i for i in range( + cracklib.MIN_LENGTH - cracklib.OTH_CREDIT)]: +-self.assertEquals( ++self.assertEqual( + 1, cracklib.simple(passwd), + 'password {0} should be detected as too simple'.format( + passwd)) +-self.assertEquals(0, cracklib.simple( ++self.assertEqual(0, cracklib.simple( + '#' * (cracklib.MIN_LENGTH - cracklib.OTH_CREDIT))) + + def test_simple_combinations(self): +@@ -119,11 +119,11 @@ + cracklib.MIN_LENGTH - + cracklib.LOW_CREDIT - + cracklib.OTH_CREDIT)]: +-self.assertEquals( ++self.assertEqual( + 1, cracklib.simple(passwd), + 'password {0} should be detected as too simple'.format( + passwd)) +-self.assertEquals(0, cracklib.simple( ++self.assertEqual(0, cracklib.simple( + testset[:(cracklib.MIN_LENGTH - cracklib.LOW_CREDIT - + cracklib.OTH_CREDIT)])) + diff -Nru cracklib2-2.9.6/debian/patches/series cracklib2-2.9.6/debian/patches/series --- cracklib2-2.9.6/debian/patches/series 2022-11-29 08:26:33.0 +0100 +++ cracklib2
Bug#1055687: khmer ftbfs with Python 3.12
Hi, Thank you for uploading 3.0.0~a3+dfsg-6! I was about to request a sync in Ubuntu. Sadly, I noticed that I forgot to forward one change that was done as an intermediate upload in Ubuntu. The change is required for Python 3.12 support. Sorry about that! I've just opened https://salsa.debian.org/med-team/khmer/-/merge_requests/1 with the missing bits to correct the mistake. Best regards, Olivier
Bug#1055687: khmer ftbfs with Python 3.12
Hi Andreas, On Wed, 29 Nov 2023 14:26:01 +0100 Andreas Tille wrote: When looking at the first patch > it applies to a series file containing the patches > > refresh_cython > find_object_files_at_right_loc.patch > > at the end. I'd love to profit from all those patches. Where > can I find these? > You can find all the patches that are currently used for khmer 3.0.0~a3+dfsg-5 in the khmer tree from Debian Med: https://salsa.debian.org/med-team/khmer/-/tree/master/debian/patches Thanks, Olivier
Bug#1057031: nvme-stas: autopkgtest hanging on s390x
Hello, nvme-stas v2.3.1 is out now :) https://github.com/linux-nvme/nvme-stas/releases/tag/v2.3.1 Best regards, Olivier
Bug#1057031: nvme-stas: autopkgtest hanging on s390x
Thanks Daniel! FYI Martin Bélanger (from nvme-stas upstream) will be preparing a 2.3.1 release of nvme-stas. It should contain some of the patches needed so if we are not in a hurry, we might as well wait for 2.3.1 to be out. Thanks, Olivier On 11/28/23 11:38, Daniel Baumann wrote: > Hi Olivier, > > thanks you, that is much apperciated. I've been mostly away/VAC the last > two weeks, but I'll take care about this and the other patch later today. > > Regards, > Daniel
Bug#1057031: nvme-stas: autopkgtest hanging on s390x
Package: nvme-stas Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu noble ubuntu-patch Dear Maintainer, After applying fixes from bug 1054533 [1], autopkgtest for nvme-stas hangs on s390x and ends up failing on timeout. Excerpt from an autopkgtest run on Ubuntu [2]. > 96s test__data_matches_ip (test-iputil.Test.test__data_matches_ip) ... ok > 596s test_get_interface (test-iputil.Test.test_get_interface) > 10600s Check that get_interface() returns the right info ... autopkgtest > [02:09:30]: ERROR: timed out on command "su -s /bin/bash ubuntu -c set -e; > [...] This is caused by an endianness issue in staslib.iputil when packing/unpacking netlink data structures. It can be fixed by using host byte ordering instead of "hard-coding" little endian. I forwarded the fix upstream [3]. In Ubuntu, the attached patch was applied to achieve the following: * Fix endianness issues in staslib.iputil causing hangs on s390x. (LP: #2045000) Thanks for considering the patch. [1] https://bugs.debian.org/1054533 [2] https://autopkgtest.ubuntu.com/results/autopkgtest-noble/noble/s390x/n/nvme-stas/20231128_021002_94626@/log.gz [3] https://github.com/linux-nvme/nvme-stas/pull/406 -- System Information: Debian Release: trixie/sid APT prefers mantic-updates APT policy: (500, 'mantic-updates'), (500, 'mantic-security'), (500, 'mantic') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru nvme-stas-2.3/debian/control nvme-stas-2.3/debian/control --- nvme-stas-2.3/debian/control2023-11-27 17:45:04.0 +0100 +++ nvme-stas-2.3/debian/control2023-11-28 11:01:46.0 +0100 @@ -1,8 +1,7 @@ Source: nvme-stas Section: net Priority: optional -Maintainer: Ubuntu Developers -XSBC-Original-Maintainer: Daniel Baumann +Maintainer: Daniel Baumann Uploaders: Benjamin Drung , Build-Depends: diff -Nru nvme-stas-2.3/debian/patches/fix-iputil-endianness-issue.patch nvme-stas-2.3/debian/patches/fix-iputil-endianness-issue.patch --- nvme-stas-2.3/debian/patches/fix-iputil-endianness-issue.patch 1970-01-01 01:00:00.0 +0100 +++ nvme-stas-2.3/debian/patches/fix-iputil-endianness-issue.patch 2023-11-28 11:01:46.0 +0100 @@ -0,0 +1,103 @@ +Description: iputil: pack and unpack using native byte ordering + On big-endian architectures (such as s390x), tests defined in + test-iputil.py would hang: + . + > 596s test_get_interface (test-iputil.Test.test_get_interface) + > 10600s Check that get_interface() returns the right info ... + . + Indeed, running the following hangs as well: + . + > from staslib import iputil + > + > iputil.net_if_addrs() + . + This is an endianness issue that can be fixed by using native + byte-ordering when packing and unpacking netlink data structures such as + ifaddrmsg. +Author: Olivier Gayot +Bug-Ubuntu: https://launchpad.net/bugs/2045000 +Forwarded: https://github.com/linux-nvme/nvme-stas/pull/406 +Last-Update: 2023-11-28 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +--- a/staslib/iputil.py b/staslib/iputil.py +@@ -42,7 +42,7 @@ + __u32 nlmsg_pid; /* Sending process port ID */ + }; + ''' +-return struct.pack('
Bug#1054533: nvme-stas: autopkgtest fails on all architectures
Package: nvme-stas Followup-For: Bug #1054533 User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu noble ubuntu-patch Control: tags -1 patch Dear Maintainer, Several test cases in nvme-stas 2.3-1 were making the autopkgtest fail. In Ubuntu, the attached patch was applied to achieve the following: * Add Test-Depends on nvme-cli to fix the absence of /etc/nvme/host{nqn,id} when running the test suite. (LP: #2043792) - This is only a workaround for autopkgtests. stafd can still fail with the same error at runtime outside of autopkgtests when configured with an appropriate "controller =" directive. * Fix failing test from mocked libnvme version. (LP: #2043792) * Fix legacy Udev test failing when multiples IPv6 addresses are set. (LP: #2043792) Thanks for considering the patch. -- System Information: Debian Release: trixie/sid APT prefers mantic-updates APT policy: (500, 'mantic-updates'), (500, 'mantic-security'), (500, 'mantic') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru nvme-stas-2.3/debian/patches/fix-test-libnvme-version.patch nvme-stas-2.3/debian/patches/fix-test-libnvme-version.patch --- nvme-stas-2.3/debian/patches/fix-test-libnvme-version.patch 1970-01-01 01:00:00.0 +0100 +++ nvme-stas-2.3/debian/patches/fix-test-libnvme-version.patch 2023-11-21 09:29:41.0 +0100 @@ -0,0 +1,52 @@ +Description: Fix mock libnvme test so it works in GitHub actions & autopkgtest + Different Python test frameworks manage Python processes differently + when running tests. When running `python3 -m unittest` for instance, it + looks like the same process executes all the tests. Therefore when one + test module T1 imports a module, the module is not re-imported if needed + by T2. + . + This causes issues with test-defs.py which tries to mock the libnvme + module. Indeed, when it is previously imported by staslib, the mocked + libnvme module does not get re-imported. + . + Fixed by removing staslib and staslib.defs from the imported modules + before executing the test. +Author: Olivier Gayot +Origin: upstream, https://github.com/linux-nvme/nvme-stas/commit/2efebdb4dec6d5a6341a577273a1ce7b57fe488c +Bug-Ubuntu: https://launchpad.net/bugs/2043792 +Bug-Debian: https://bugs.debian.org/1054533 +Last-Update: 2023-11-27 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +--- a/test/test-defs.py b/test/test-defs.py +@@ -1,4 +1,5 @@ + #!/usr/bin/python3 ++import contextlib + import os + import sys + import unittest +@@ -9,13 +10,17 @@ + '''Testing defs.py by mocking the libnvme package''' + + def test_libnvme_version(self): +-# For unknown reasons, this test does +-# not work when run from GitHub Actions. +-if not os.getenv('GITHUB_ACTIONS'): +-from staslib import defs ++# Ensure that we re-import staslib & staslib.defs if the current Python ++# process has them already imported. ++with contextlib.suppress(KeyError): ++sys.modules.pop('staslib.defs') ++with contextlib.suppress(KeyError): ++sys.modules.pop('staslib') + +-libnvme_ver = defs.LIBNVME_VERSION +-self.assertEqual(libnvme_ver, '?.?') ++from staslib import defs ++ ++libnvme_ver = defs.LIBNVME_VERSION ++self.assertEqual(libnvme_ver, '?.?') + + @classmethod + def setUpClass(cls): # called once before all the tests diff -Nru nvme-stas-2.3/debian/patches/fix-test-udev-failing-multiple-IPv6.patch nvme-stas-2.3/debian/patches/fix-test-udev-failing-multiple-IPv6.patch --- nvme-stas-2.3/debian/patches/fix-test-udev-failing-multiple-IPv6.patch 1970-01-01 01:00:00.0 +0100 +++ nvme-stas-2.3/debian/patches/fix-test-udev-failing-multiple-IPv6.patch 2023-11-21 09:29:41.0 +0100 @@ -0,0 +1,29 @@ +Description: Fix test-udev failing when multiple IPv6 addresses are used + test-udev was failing when an interface had more than one IPv6 + addresses assigned. This was due to the test checking that the + number of assigned IPv6 addresses was exactly 1 (== 1) instead of + checking greater-equal to 1 (>= 1). +Author: Martin Belanger +Origin: upstream, https://github.com/linux-nvme/nvme-stas/commit/66b42bdd83bb6a83b30da3aed5fea9ed4da882f6 +Bug: https://github.com/linux-nvme/nvme-stas/issues/403 +Bug-Ubuntu: https://launchpad.net/bugs/2043792 +Bug-Debian: https://bugs.debian.org/1054533 +Last-Update: 2023-11-27 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +diff --git a/test/test-udev.py b/test/test-udev.py +index be257d9..ba484e0 100755 +--
Bug#1055687: khmer ftbfs with Python 3.12
Package: khmer Followup-For: Bug #1055687 User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu noble ubuntu-patch Control: tags -1 patch Dear Maintainer, My previous patch was unfortunately very incomplete. I am submitting another patch that fixes the remaining issues when building with Python 3.12. Both debdiffs should be applied for the build to succeed. Additional fixes that were needed for the build to succeed: * Python 3.12 dropped the "imp" module. Updated to use importlib instead. * Python 3.12 is much less forgiving when a script opens a file for writing and forgets to close it. Most Python scripts in scripts/ failed to do so or did so inconsistently. This resulted in the test suite either hanging or failing. I went through all invocations of open() / get_file_writer() and ensured that the resources are cleaned up. The resulting patch is sadly difficult to read though. I submitted all changes upstream, although I doubt somebody will pick them up: https://github.com/dib-lab/khmer/pull/1922 In Ubuntu, the attached patch was applied to achieve the following: * Fix build against Python 3.12 (LP: #2044383). Thanks for considering the patch. -- System Information: Debian Release: trixie/sid APT prefers mantic-updates APT policy: (500, 'mantic-updates'), (500, 'mantic-security'), (500, 'mantic') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru khmer-3.0.0~a3+dfsg/debian/control khmer-3.0.0~a3+dfsg/debian/control --- khmer-3.0.0~a3+dfsg/debian/control 2023-11-25 17:44:28.0 +0100 +++ khmer-3.0.0~a3+dfsg/debian/control 2023-11-26 02:28:32.0 +0100 @@ -1,6 +1,5 @@ Source: khmer -Maintainer: Ubuntu Developers -XSBC-Original-Maintainer: Debian Med Packaging Team +Maintainer: Debian Med Packaging Team Uploaders: Michael R. Crusoe , Kevin Murray Section: science diff -Nru khmer-3.0.0~a3+dfsg/debian/patches/close-opened-files.patch khmer-3.0.0~a3+dfsg/debian/patches/close-opened-files.patch --- khmer-3.0.0~a3+dfsg/debian/patches/close-opened-files.patch 1970-01-01 01:00:00.0 +0100 +++ khmer-3.0.0~a3+dfsg/debian/patches/close-opened-files.patch 2023-11-26 02:28:32.0 +0100 @@ -0,0 +1,1124 @@ +Description: ensure that Python scripts close files that they open for writing + Python scripts under scripts/ in the source tree do not consistently close + files that they open for writing. While some of the scripts use context + managers, most of them do not (or do so inconsistently). + In previous releases of Ubuntu, this apparently was not much of a concern. + However, Python 3.12 seems to be much less forgiving when files are not + properly closed. When running the test suite, many of the files that are not + explicitly closed appear truncated. This leads to various tests failing or + hanging and causing FTBFS when the test suite runs at build time. + . + Furthermore, khmer defines the get_file_writer() function, but it cannot be + consistently used as a context manager because it sometimes closes the + underlying file descriptor ; and sometimes does not depending on the + arguments. + . + Fixed by defining a new FileWriter context manager and ensuring that + each call to open() / get_file_writer() frees up resources properly. +Author: Olivier Gayot +Bug-Ubuntu: https://launchpad.net/bugs/2044383 +Bug-Debian: https://bugs.debian.org/1055687 +Forwarded: https://github.com/dib-lab/khmer/pull/1922 +Last-Update: 2023-11-26 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +Index: khmer-3.0.0~a3+dfsg/scripts/abundance-dist.py +=== +--- khmer-3.0.0~a3+dfsg.orig/scripts/abundance-dist.py 2023-11-26 20:23:39.915485717 +0100 khmer-3.0.0~a3+dfsg/scripts/abundance-dist.py 2023-11-26 20:23:39.911485747 +0100 +@@ -42,6 +42,7 @@ + Use '-h' for parameter help. + """ + ++import contextlib + import sys + import csv + import khmer +@@ -143,26 +144,28 @@ + sys.exit(1) + + if args.output_histogram_filename in ('-', '/dev/stdout'): +-countgraph_fp = sys.stdout ++countgraph_ctx = contextlib.nullcontext(enter_result=sys.stdout) + else: +-countgraph_fp = open(args.output_histogram_filename, 'w') +-countgraph_fp_csv = csv.writer(countgraph_fp) +-# write headers: +-countgraph_fp_csv.writerow(['abundance', 'count', 'cumulative', +-'cumulative_fraction']) +- +-sofar = 0 +-for _, i in enumerate(abundances): +-if i == 0 and not args.output_zero: +-continue ++countgraph_ctx = open(args.outp
Bug#1056591: acedb FTBFS against glibc 2.38
Package: acedb Severity: important Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu noble ubuntu-patch Dear Maintainer, Trying to build with glibc 2.38 fails with the following error: ../wh/utils.h:34:7: error: conflicting types for ‘strcasestr’; have ‘char *(char *, char *)’ 34 | char *strcasestr (char *str1, char *str2); | ^~ In file included from ../wh/mystdlib.h:227, from ../wh/regular.h:51, from utils.c:37: /usr/include/string.h:380:14: note: previous declaration of ‘strcasestr’ with type ‘char *(const char *, const char *)’ Before glibc 2.38, the strcasestr function was only made available when building with _GNU_SOURCE. Starting with glibc 2.38, the function is available by default. In Ubuntu, the attached patch was applied to achieve the following: * Fix build against glibc 2.38 (LP: #2044385) Thanks for considering the patch. -- System Information: Debian Release: trixie/sid APT prefers mantic-updates APT policy: (500, 'mantic-updates'), (500, 'mantic-security'), (500, 'mantic') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru acedb-4.9.39+dfsg.02/debian/patches/glibc-2.38.patch acedb-4.9.39+dfsg.02/debian/patches/glibc-2.38.patch --- acedb-4.9.39+dfsg.02/debian/patches/glibc-2.38.patch1970-01-01 01:00:00.0 +0100 +++ acedb-4.9.39+dfsg.02/debian/patches/glibc-2.38.patch2023-11-23 16:07:13.0 +0100 @@ -0,0 +1,47 @@ +Description: Fix build against glibc 2.38 + In previous glibc versions, strcasestr would only be available if building + with _GNU_SOURCE. + Starting with glibc 2.38, strcasestr is now available by default. + Ensure that acedb does not redefine the function if we're building with a modern glibc. +Author: Olivier Gayot +Bug: +Bug-Ubuntu: https://launchpad.net/bugs/2044385 +Forwarded: no +Last-Update: 2023-11-23 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +Index: acedb-4.9.39+dfsg.02/w1/utils.c +=== +--- acedb-4.9.39+dfsg.02.orig/w1/utils.c 2005-10-04 15:15:00.0 +0200 acedb-4.9.39+dfsg.02/w1/utils.c2023-11-23 16:48:20.118800952 +0100 +@@ -774,7 +774,7 @@ + // + /* case-insensitive version of strstr */ + +-#ifndef DARWIN ++#ifndef HAS_STRCASESTR + char *strcasestr(char *str1, char *str2) + { + g_strup(str1); +Index: acedb-4.9.39+dfsg.02/wh/utils.h +=== +--- acedb-4.9.39+dfsg.02.orig/wh/utils.h 2023-11-23 16:19:02.454527757 +0100 acedb-4.9.39+dfsg.02/wh/utils.h2023-11-23 16:47:40.099066905 +0100 +@@ -29,7 +29,16 @@ + *--- + */ + +-#ifndef DARWIN ++/* Ensure that features.h or an equivalent is included. */ ++#include ++ ++#if defined(DARWIN) || defined(_GNU_SOURCE) ++# define HAS_STRCASESTR ++#elif __GLIBC__ >= 2 && __GLIBC_MINOR__ >= 38 ++# define HAS_STRCASESTR ++#endif ++ ++#ifndef HAS_STRCASESTR + /* case-insensitive version of strstr */ + char *strcasestr (char *str1, char *str2); + #endif diff -Nru acedb-4.9.39+dfsg.02/debian/patches/series acedb-4.9.39+dfsg.02/debian/patches/series --- acedb-4.9.39+dfsg.02/debian/patches/series 2023-08-15 16:23:13.0 +0200 +++ acedb-4.9.39+dfsg.02/debian/patches/series 2023-11-23 16:07:13.0 +0100 @@ -13,3 +13,4 @@ glibc2.32.patch # drop_gtk2.patch just_build_efetch.patch +glibc-2.38.patch
Bug#1055687: khmer ftbfs with Python 3.12
Package: khmer Followup-For: Bug #1055687 User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu noble ubuntu-patch Control: tags -1 patch Dear Maintainer, In Ubuntu, the attached patch was applied to achieve the following: * Fix build against Python 3.12 (LP: #2044383). Thanks for considering the patch. -- System Information: Debian Release: trixie/sid APT prefers mantic-updates APT policy: (500, 'mantic-updates'), (500, 'mantic-security'), (500, 'mantic') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru khmer-3.0.0~a3+dfsg/debian/patches/python3.12-support.patch khmer-3.0.0~a3+dfsg/debian/patches/python3.12-support.patch --- khmer-3.0.0~a3+dfsg/debian/patches/python3.12-support.patch 1970-01-01 01:00:00.0 +0100 +++ khmer-3.0.0~a3+dfsg/debian/patches/python3.12-support.patch 2023-11-23 15:24:51.0 +0100 @@ -0,0 +1,24 @@ +Description: Add support for Python 3.12 + Ever since Python 3.2, configparser.SafeConfigParser has been deprecated in + favor of configparser.ConfigParser. An alias existed for backward compability + but the alias was dropped from Python 3.12. +Author: Olivier Gayot +Bug-Ubuntu: https://launchpad.net/bugs/2044383 +Bug-Debian: https://bugs.debian.org/1055687 +Forwarded: https://github.com/dib-lab/khmer/pull/1922 +Last-Update: 2023-11-23 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +Index: khmer-3.0.0~a3+dfsg/versioneer.py +=== +--- khmer-3.0.0~a3+dfsg.orig/versioneer.py 2019-03-13 14:42:12.0 +0100 khmer-3.0.0~a3+dfsg/versioneer.py 2023-11-23 15:19:50.025827413 +0100 +@@ -339,7 +339,7 @@ + # configparser.NoOptionError (if it lacks "VCS="). See the docstring at + # the top of versioneer.py for instructions on writing your setup.cfg . + setup_cfg = os.path.join(root, "setup.cfg") +-parser = configparser.SafeConfigParser() ++parser = configparser.ConfigParser() + with open(setup_cfg, "r") as f: + parser.readfp(f) + VCS = parser.get("versioneer", "VCS") # mandatory diff -Nru khmer-3.0.0~a3+dfsg/debian/patches/series khmer-3.0.0~a3+dfsg/debian/patches/series --- khmer-3.0.0~a3+dfsg/debian/patches/series 2023-09-09 08:30:33.0 +0200 +++ khmer-3.0.0~a3+dfsg/debian/patches/series 2023-11-23 15:19:31.0 +0100 @@ -17,3 +17,4 @@ stringio-buffer.patch refresh_cython find_object_files_at_right_loc.patch +python3.12-support.patch
Bug#1054533: nvme-stas: autopkgtest fails on all architectures
Although this issue appears as a regression in nvme-stas 2.3-1, it looks like it was already affecting the usability of the nvme-stas version in testing (i.e., 2.2.2-2) - outside of autopkgtests. Here with nvme-stas 2.2.2-2 (after configuring a controller in /etc/stas/stafd.conf): Nov 17 11:14:05 usable-whippet systemd[1]: Reloaded stafd.service - STorage Appliance Finder (STAF). Nov 17 11:14:07 usable-whippet stafd[1268]: Error reading mandatory Host NQN (see stasadm --help): [Errno 2] No such file or directory: '/etc/nvme/hostnqn' Nov 17 11:14:07 usable-whippet systemd[1]: stafd.service: Main process exited, code=exited, status=1/FAILURE Nov 17 11:14:07 usable-whippet systemd[1]: stafd.service: Failed with result 'exit-code'. To make the autopkgtest pass, I believe we can either: * add a Test-Depends on nvme-cli; or * running those commands as part of the autopkgtest: * mkdir --parents /etc/nvme * systemctl start stas-config@hostnqn * systemctl start stas-config@hostid That said, I suggest we look for a solution that also resolves the issue outside of autopkgtests. Would introducing a Depends on nvme-cli be a bad thing? Kind regards, Olivier
Bug#1042449: libprotocol-http2-perl: Test-suite fails if OpenSSL configured with seclevel 2
On 7/28/23 13:48, gregor herrmann wrote: > On Fri, 28 Jul 2023 12:30:38 +0200, Olivier Gayot wrote: > >> Debian is currently unaffected (I assume the security level is set to 1 >> at build-time) but in the future OpenSSL 3.1 will reject TLSv1 at >> security level 1. > > Thanks for forwarding before this hits Debian! Anytime, thanks for applying the fix so quickly! Olivier
Bug#1042449: libprotocol-http2-perl: Test-suite fails if OpenSSL configured with seclevel 2
Package: libprotocol-http2-perl Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu mantic ubuntu-patch Dear Maintainer, The package uses the hardcoded tlsv1 value in its test-suite. When OpenSSL has been built with security level 2 (or is set to level 2 at runtime), the TLSv1 protocol is rejected. This makes the libprotocol-http2-perl build / autopkgtest fail. There is an upstream bug report: https://github.com/vlet/p5-Protocol-HTTP2/issues/15 And a PR was opened usptream: https://github.com/vlet/p5-Protocol-HTTP2/pull/16 Debian is currently unaffected (I assume the security level is set to 1 at build-time) but in the future OpenSSL 3.1 will reject TLSv1 at security level 1. In Ubuntu, the attached patch was applied to achieve the following: * Do not hardcode the test-suite TLS version to tlsv1 - which is disabled by OpenSSL seclevel 2 on Ubuntu (LP: #2023586). Thanks for considering the patch. -- System Information: Debian Release: bookworm/sid APT prefers lunar-updates APT policy: (500, 'lunar-updates'), (500, 'lunar-security'), (500, 'lunar'), (100, 'lunar-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru libprotocol-http2-perl-1.10/debian/patches/no-tlsv1.patch libprotocol-http2-perl-1.10/debian/patches/no-tlsv1.patch --- libprotocol-http2-perl-1.10/debian/patches/no-tlsv1.patch 1970-01-01 01:00:00.0 +0100 +++ libprotocol-http2-perl-1.10/debian/patches/no-tlsv1.patch 2023-07-28 11:43:40.0 +0200 @@ -0,0 +1,32 @@ +Description: Remove hardcoded tlsv1 protocol version + The test-suite of libprotocol-http2-perl uses a hardcoded value of tlsv1 - + which is disabled in Ubuntu by means of OpenSSL seclevel. Specifying another + version like tlsv1_2 would work but it seems sensible to leave that up to the + system decide. +Author: Olivier Gayot +Bug-Ubuntu: https://launchpad.net/bugs/2023586 +Forwarded: https://github.com/vlet/p5-Protocol-HTTP2/pull/16 +Last-Update: 2023-07-28 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +Index: b/t/lib/PH2ClientServerTest.pm +=== +--- a/t/lib/PH2ClientServerTest.pm 2023-07-28 11:35:33.957861624 +0200 b/t/lib/PH2ClientServerTest.pm 2023-07-28 11:43:04.843734902 +0200 +@@ -43,7 +43,6 @@ + if ( !$h{upgrade} && ( $h{npn} || $h{alpn} ) ) { + eval { + $tls = AnyEvent::TLS->new( +-method=> 'tlsv1', + cert_file => $tls_crt, + key_file => $tls_key, + ); +@@ -122,7 +121,7 @@ + } + elsif ( $h{npn} || $h{alpn} ) { + eval { +-$tls = AnyEvent::TLS->new( method => 'tlsv1', ); ++$tls = AnyEvent::TLS->new(); + + if ( delete $h{npn} ) { + diff -Nru libprotocol-http2-perl-1.10/debian/patches/series libprotocol-http2-perl-1.10/debian/patches/series --- libprotocol-http2-perl-1.10/debian/patches/series 1970-01-01 01:00:00.0 +0100 +++ libprotocol-http2-perl-1.10/debian/patches/series 2023-07-28 11:43:11.0 +0200 @@ -0,0 +1 @@ +no-tlsv1.patch
Bug#1042079: onednn 2.7.4-1 FTBFS on arm64
Package: onednn Version: 2.7.4-1 Followup-For: Bug #1042079 User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu mantic ubuntu-patch Control: tags -1 patch Dear Maintainer, In Ubuntu, the attached patch was applied to achieve the following: * Apply upstream patches to fix build on arm64 (LP: #2028759) + d/patches/lp2028759/cpu-aarch64-update-xbyak_aarch64-into-the-lastes-ver(1).patch + d/patches/lp2028759/cpu-aarch64-fix-getting-cache-sizes-on-macOS.patch + d/patches/lp2028759/cpu-aarch64-update-xbyak_aarch64-into-the-lastes-ver(2).patch The second patch (macOS specific) is not essential but allows the third patch to apply with fewer modifications. Thanks for considering the patch. -- System Information: Debian Release: bookworm/sid APT prefers lunar-updates APT policy: (500, 'lunar-updates'), (500, 'lunar-security'), (500, 'lunar'), (100, 'lunar-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru onednn-2.7.4/debian/patches/lp2028759/cpu-aarch64-fix-getting-cache-sizes-on-macOS.patch onednn-2.7.4/debian/patches/lp2028759/cpu-aarch64-fix-getting-cache-sizes-on-macOS.patch --- onednn-2.7.4/debian/patches/lp2028759/cpu-aarch64-fix-getting-cache-sizes-on-macOS.patch 1970-01-01 01:00:00.0 +0100 +++ onednn-2.7.4/debian/patches/lp2028759/cpu-aarch64-fix-getting-cache-sizes-on-macOS.patch 2023-07-26 12:14:38.0 +0200 @@ -0,0 +1,49 @@ +Description: cpu: aarch64: fix getting cache sizes on macOS + Notes from ogayot: This is not strictly needed for Ubuntu but allows the + subsequent patch (i.e., + cpu-aarch64-update-xbyak_aarch64-into-the-latest-ver(2).patch) to apply without + too many modifications. +Author: Denis Samoilov +Bug-Ubuntu: https://launchpad.net/bugs/2028759 +Bug-Debian: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1042079 +Forwarded: not-needed +Applied-Upstream: https://github.com/oneapi-src/oneDNN/commit/ba91a536cb59558b3f3880951def160fe24b2820 +Last-Update: 2023-07-26 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +diff --git a/src/cpu/aarch64/xbyak_aarch64/src/util_impl.cpp b/src/cpu/aarch64/xbyak_aarch64/src/util_impl.cpp +index 59f712c7b..cb800b250 100644 +--- a/src/cpu/aarch64/xbyak_aarch64/src/util_impl.cpp b/src/cpu/aarch64/xbyak_aarch64/src/util_impl.cpp +@@ -52,17 +52,25 @@ void Cpu::setCacheHierarchy() { + } + } else { + /** +- * @ToDo Get chache information by `sysconf` ++ * @ToDo Get cache information by `sysconf` + * for the case thd dictionary is unavailable. + */ ++ ++// _SC_LEVEL_DCACHE_SIZE macros are not defined on macOS. ++#if defined(__APPLE__) ++#define GET_CACHE_SIZE(ID) 0 ++#else ++#define GET_CACHE_SIZE(ID) sysconf(ID) ++#endif ++ + #define XBYAK_AARCH64_CACHE_SIZE(LEVEL, SIZE, ID, CORES, VAL) \ +- cache_size = sysconf(ID); \ ++ cache_size = GET_CACHE_SIZE(ID); \ + VAL[LEVEL] = cache_size ? (cache_size / (CORES)) : ((SIZE) / (CORES)); + + uint32_t cache_size; + + /* If `sysconf` returns zero as cache sizes, 32KiB, 1MiB, 0 and 0 is set as +- 1st, 2nd, 3rd and 4th level cache sizes. 2nd cahce is assumed as sharing cache. */ ++ 1st, 2nd, 3rd and 4th level cache sizes. 2nd cache is assumed as sharing cache. */ + XBYAK_AARCH64_CACHE_SIZE(0, 1024 * 32, _SC_LEVEL1_DCACHE_SIZE, 1, coresSharingDataCache_); + XBYAK_AARCH64_CACHE_SIZE(1, 1024 * 1024, _SC_LEVEL2_CACHE_SIZE, 1, coresSharingDataCache_); + XBYAK_AARCH64_CACHE_SIZE(2, 0, _SC_LEVEL3_CACHE_SIZE, 1, coresSharingDataCache_); +-- +2.39.2 + diff -Nru onednn-2.7.4/debian/patches/lp2028759/cpu-aarch64-update-xbyak_aarch64-into-the-latest-ver(1).patch onednn-2.7.4/debian/patches/lp2028759/cpu-aarch64-update-xbyak_aarch64-into-the-latest-ver(1).patch --- onednn-2.7.4/debian/patches/lp2028759/cpu-aarch64-update-xbyak_aarch64-into-the-latest-ver(1).patch
Bug#1042079: onednn 2.7.4-1 FTBFS on arm64
Source: onednn Version: 2.7.4-1 Severity: normal Dear Maintainer, Building onednn 2.7.4-1 on arm64 fails with many of the following errors: In file included from /<>/src/cpu/aarch64/xbyak_aarch64/src/xbyak_aarch64_impl.cpp:17: /<>/src/cpu/aarch64/xbyak_aarch64/xbyak_aarch64/xbyak_aarch64.h:70:7: error: ‘uint64_t’ does not name a type 70 | const uint64_t SP_IDX = 31; | ^~~~ /<>/src/cpu/aarch64/xbyak_aarch64/xbyak_aarch64/xbyak_aarch64.h:56:1: note: ‘uint64_t’ is defined in header ‘’; did you forget to ‘#include ’? 55 | #include +++ |+#include 56 | /<>/src/cpu/aarch64/xbyak_aarch64/xbyak_aarch64/xbyak_aarch64.h:71:7: error: ‘uint64_t’ does not name a type 71 | const uint64_t NUM_VREG_BYTES = 16; | ^~~~ /<>/src/cpu/aarch64/xbyak_aarch64/xbyak_aarch64/xbyak_aarch64.h:71:7: note: ‘uint64_t’ is defined in header ‘’; did you forget to ‘#include ’? /<>/src/cpu/aarch64/xbyak_aarch64/xbyak_aarch64/xbyak_aarch64.h:72:7: error: ‘uint64_t’ does not name a type 72 | const uint64_t NUM_ZREG_BYTES = 64; | ^~~~ /<>/src/cpu/aarch64/xbyak_aarch64/xbyak_aarch64/xbyak_aarch64.h:72:7: note: ‘uint64_t’ is defined in header ‘’; did you forget to ‘#include ’? In file included from /<>/src/cpu/aarch64/xbyak_aarch64/xbyak_aarch64/xbyak_aarch64_adr.h:19, from /<>/src/cpu/aarch64/xbyak_aarch64/xbyak_aarch64/xbyak_aarch64_gen.h:21, from /<>/src/cpu/aarch64/xbyak_aarch64/xbyak_aarch64/xbyak_aarch64.h:73: /<>/src/cpu/aarch64/xbyak_aarch64/xbyak_aarch64/xbyak_aarch64_reg.h:96:3: error: ‘uint32_t’ does not name a type 96 | uint32_t bit_; | ^~~~ /<>/src/cpu/aarch64/xbyak_aarch64/xbyak_aarch64/xbyak_aarch64_reg.h:19:1: note: ‘uint32_t’ is defined in header ‘’; did you forget to ‘#include ’? Build log: https://buildd.debian.org/status/fetch.php?pkg=onednn=arm64=2.7.4-1=1689851490=0 There is an existing bug report upstream and the issue seems to have been fixed in onednn 3.1: https://github.com/oneapi-src/oneDNN/issues/1600 Thanks, Oilvier -- System Information: Debian Release: bookworm/sid APT prefers lunar-updates APT policy: (500, 'lunar-updates'), (500, 'lunar-security'), (500, 'lunar'), (100, 'lunar-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled
Bug#1037454: zulucrypt autopkgtests fail
Package: zulucrypt Followup-For: Bug #1037454 User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu mantic ubuntu-patch Control: tags -1 patch Dear Maintainer, The test-suite seems to fail on virt-servers that use sudo to run as root. For some reason, zuluCrypt reads the SUDO_UID variable and calls seteuid(SUDO_UID) if it exists - before attemping to open the test luks file (that was created as root earlier). I opened an upstream bug report: https://github.com/mhogomchungu/zuluCrypt/issues/209 Unsetting the SUDO_UID variable helps making the autopkgtest succeeds - although it does not prevent somebody from manually invoking $ sudo /usr/bin/zuluCrypt-cli --test on the target system In Ubuntu, the attached patch was applied to achieve the following: * Unset SUDO_UID when running the test suite - to make autopkgtest succeed on all virt-servers (LP: #2023614). Thanks for considering the patch. -- System Information: Debian Release: bookworm/sid APT prefers lunar-updates APT policy: (500, 'lunar-updates'), (500, 'lunar-security'), (500, 'lunar'), (100, 'lunar-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru zulucrypt-6.2.0/debian/tests/control zulucrypt-6.2.0/debian/tests/control --- zulucrypt-6.2.0/debian/tests/control2022-09-19 04:02:31.0 +0200 +++ zulucrypt-6.2.0/debian/tests/control2023-07-25 09:44:49.0 +0200 @@ -1,2 +1,2 @@ -Test-Command: zuluCrypt-cli --test +Test-Command: env -u SUDO_UID zuluCrypt-cli --test Restrictions: allow-stderr,isolation-machine,needs-root
Bug#1040243: php-doctrine-cache: test-suite fails when using php-cache-integration-tests 0.17.0-4 (experimental)
Package: php-doctrine-cache Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu mantic ubuntu-patch Dear Maintainer, php-doctrine-cache requires php-cache-tag-interop which used to be provided by php-cache-integration-tests. Starting with php-cache-integration-tests 0.17.0-4, the dependency on php-cache-tag-interop was dropped so the dependencies of php-doctrine-cache are no longer satisfied. In Ubuntu, the attached patch was applied to achieve the following: * Merge with Debian unstable (LP: #2025673). Remaining changes: - d/control: Add b-d for php-cache-tag-interop removed from php-cache-integration-tests. - d/t/control: Add test dependency on php-cache-tag-interop removed from php-cache-integration-tests. Thanks for considering the patch. -- System Information: Debian Release: bookworm/sid APT prefers lunar-updates APT policy: (500, 'lunar-updates'), (500, 'lunar-security'), (500, 'lunar'), (100, 'lunar-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru php-doctrine-cache-2.2.0/debian/control php-doctrine-cache-2.2.0/debian/control --- php-doctrine-cache-2.2.0/debian/control 2023-06-25 10:20:40.0 +0200 +++ php-doctrine-cache-2.2.0/debian/control 2023-07-03 19:47:31.0 +0200 @@ -7,6 +7,7 @@ memcached, php-apcu, php-cache-integration-tests, + php-cache-tag-interop, php-memcache, php-memcached, php-psr-cache, diff -Nru php-doctrine-cache-2.2.0/debian/tests/control php-doctrine-cache-2.2.0/debian/tests/control --- php-doctrine-cache-2.2.0/debian/tests/control 2022-10-17 08:08:20.0 +0200 +++ php-doctrine-cache-2.2.0/debian/tests/control 2023-07-03 19:47:31.0 +0200 @@ -3,6 +3,7 @@ Depends: memcached, php-apcu, php-cache-integration-tests, + php-cache-tag-interop, php-memcache, php-memcached, php-psr-cache,
Bug#1038890: pyparted: Testsuite fails against parted 3.6
Package: pyparted Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu mantic ubuntu-patch Dear Maintainer, Running the pyparted test-suite against parted 3.6 leads to the following exception: 399s == 399s FAIL: runTest (tests.test__ped_disktype.DiskTypeStrTestCase.runTest) 399s -- 399s Traceback (most recent call last): 399s File "/tmp/autopkgtest.kQxzZO/build.bFs/src/tests/test__ped_disktype.py", line 78, in runTest 399s self.assertEqual(str(self.disktype['msdos']), '_ped.DiskType instance --\n name: msdos features: 1') 399s AssertionError: '_ped.DiskType instance --\n name: msdos features: 5' != '_ped.DiskType instance --\n name: msdos features: 1' 399s _ped.DiskType instance -- 399s - name: msdos features: 5? ^ 399s + name: msdos features: 1? ^ The failure is visible in the following autopkgtest report. https://ci.debian.net/data/autopkgtest/unstable/amd64/p/pyparted/34699773/log.gz A patch exists upstream: https://github.com/dcantrell/pyparted/pull/100 In Ubuntu, the attached patch was applied to achieve the following: * Apply upstream patch to fix test-suite against parted 3.6 (LP: #2024678) Thanks for considering the patch. -- System Information: Debian Release: bookworm/sid APT prefers lunar-updates APT policy: (500, 'lunar-updates'), (500, 'lunar-security'), (500, 'lunar'), (100, 'lunar-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru pyparted-3.12.0/debian/patches/0001-tests-Remove-feature-values-from-DiskTypeStrTestCase.patch pyparted-3.12.0/debian/patches/0001-tests-Remove-feature-values-from-DiskTypeStrTestCase.patch --- pyparted-3.12.0/debian/patches/0001-tests-Remove-feature-values-from-DiskTypeStrTestCase.patch 1970-01-01 01:00:00.0 +0100 +++ pyparted-3.12.0/debian/patches/0001-tests-Remove-feature-values-from-DiskTypeStrTestCase.patch 2023-06-22 18:08:47.0 +0200 @@ -0,0 +1,43 @@ +Description: tests: Remove feature values from DiskTypeStrTestCase + These inevitably cause pyparted to fail when built against a new version + of parted. They don't provide any additional validation, other tests + already check for the expected features. +Author: "Brian C. Lane" +Origin: upstream, https://github.com/dcantrell/pyparted/pull/100 +Bug-Ubuntu: https://launchpad.net/bugs/2024678 +Last-Update: 2023-06-22 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +Index: pyparted-3.12.0/tests/test__ped_disktype.py +=== +--- pyparted-3.12.0.orig/tests/test__ped_disktype.py 2023-06-22 18:02:36.965289725 +0200 pyparted-3.12.0/tests/test__ped_disktype.py2023-06-22 18:05:14.716398913 +0200 +@@ -75,13 +75,18 @@ + + class DiskTypeStrTestCase(RequiresDiskTypes): + def runTest(self): +-self.assertEqual(str(self.disktype['msdos']), '_ped.DiskType instance --\n name: msdos features: 1') +-self.assertEqual(str(self.disktype['aix']), '_ped.DiskType instance --\n name: aix features: 0') +-self.assertEqual(str(self.disktype['sun']), '_ped.DiskType instance --\n name: sun features: 0') +-self.assertEqual(str(self.disktype['amiga']), '_ped.DiskType instance --\n name: amiga features: 2') +-self.assertEqual(str(self.disktype['gpt']), '_ped.DiskType instance --\n name: gpt features: 2') +-self.assertEqual(str(self.disktype['mac']), '_ped.DiskType instance --\n name: mac features: 2') +-self.assertEqual(str(self.disktype['bsd']), '_ped.DiskType instance --\n name: bsd features: 0') +-self.assertEqual(str(self.disktype['pc98']), '_ped.DiskType instance --\n name: pc98 features: 2') +-self.assertEqual(str(self.disktype['loop']), '_ped.DiskType instance --\n name: loop features: 0') +-self.assertEqual(str(self.disktype['dvh']), '_ped.DiskType instance --\n name: dvh features: 3') ++types = { ++"msdos": '_ped.DiskType instance --\n name: msdos features:', ++'aix': '_ped.DiskType instance --\n name: aix features: ', ++'sun': '_ped.DiskType instance --\n name: sun features: ', ++'amiga': '_ped.DiskType instance --\n name: amiga features: ', ++'gpt': '_ped.DiskType instance --\n name: gpt features: ', ++'mac': '_ped.DiskType instance --\n name: mac features: ', ++'bsd': '_ped.DiskType instance --\n name: bsd features: ', ++
Bug#1035965: python-ansible-pygments: Autopkgtest fails against pygments 2.15
Package: python-ansible-pygments Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu mantic ubuntu-patch Dear Maintainer, Since the move to pygments 2.15, the autopkgtest testsuite fails with the following error: Traceback (most recent call last): File "/home/olivier/dev/canonical/ansible-pygments/tests/lexer_test.py", line 131, in test_ansible_output_lexer() File "/home/olivier/dev/canonical/ansible-pygments/tests/lexer_test.py", line 78, in test_ansible_output_lexer assert result == R"""ok: [windows] = { ^^ AssertionError The bug was fixed upstream by https://github.com/ansible-community/ansible-pygments/pull/34. In Ubuntu, the attached patch was applied to achieve the following: * Add patch from upstream to fix test suite against pygments 2.15 Added patch: debian/patches/Make-lexer-test-compare-tokens.patch (LP: #2019237) Thanks for considering the patch. -- System Information: Debian Release: bookworm/sid APT prefers lunar-updates APT policy: (500, 'lunar-updates'), (500, 'lunar-security'), (500, 'lunar') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru python-ansible-pygments-0.1.1/debian/patches/Make-lexer-test-compare-tokens.patch python-ansible-pygments-0.1.1/debian/patches/Make-lexer-test-compare-tokens.patch --- python-ansible-pygments-0.1.1/debian/patches/Make-lexer-test-compare-tokens.patch 1970-01-01 01:00:00.0 +0100 +++ python-ansible-pygments-0.1.1/debian/patches/Make-lexer-test-compare-tokens.patch 2023-05-11 21:30:42.0 +0200 @@ -0,0 +1,342 @@ +Description: Make lexer test compare tokens + With new releases of `pygments`, this test was getting broken again + and again because it originally relied on HTML the formatted text + output. Hence, it was prone to external influence for no good reason. + . + This patch rewrites it to solely rely on the underlying tokens that + our lexer produces, therefore reducing the dependence on unrelated + changes in the `pygments` library. +Author: Sviatoslav Sydorenko +Origin: upstream, https://github.com/ansible-community/ansible-pygments/pull/34 +Bug: https://github.com/ansible-community/ansible-pygments/issues/33 +Bug-Ubuntu: https://bugs.launchpad.net/bugs/2019237 +Applied-Upstream: https://github.com/ansible-community/ansible-pygments/commit/179c74e9f1a5dc870dec6d4db6cab60d2dca1ed2 +Last-Update: 2023-05-11 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +diff --git a/tests/lexer_test.py b/tests/lexer_test.py +index 5b700a9..a842a7c 100644 +--- a/tests/lexer_test.py b/tests/lexer_test.py +@@ -1,26 +1,24 @@ + # Author: Felix Fontein + # License: BSD-2-Clause + # Copyright: Felix Fontein , 2021 +-"""Tests for Pygments lexers.""" ++"""Tests for Pygments lexers. + +-from pygments import highlight +-# pylint: disable=no-name-in-module +-# Ref: https://github.com/PyCQA/pylint/issues/491 +-from pygments.formatters import HtmlFormatter +- +-from ansible_pygments.lexers import AnsibleOutputLexer ++They rely on token comparison for stability reasons. Relying on ++additional style formatting is known to break with updates to ++the pygments library itself. ++""" + ++from pygments import __version__ as _pygments_version ++from pygments.lexers import get_lexer_by_name as _get_lexer_by_name ++from pygments.token import Token + +-def run_test(data, lexer): +-"""Format the data snippet as HTML using a given lexer.""" +-formatter = HtmlFormatter() +-result = highlight(data, lexer, formatter) +-return formatter.get_style_defs('.highlight'), result ++PYGMENTS_VERSION_INFO = tuple(map(int, _pygments_version.split('.'))) ++IS_OLD_PYGMENTS_PRE_2_14 = PYGMENTS_VERSION_INFO <= (2, 14, 0) + + + def test_ansible_output_lexer(): +-"""Test that AnsibleOutputLexer produces expected HTML output.""" +-data = R""" ++"""Test that ``AnsibleOutputLexer`` produces expected tokens.""" ++ansible_play_output_example = R""" + ok: [windows] => { + "account": { + "account_name": "vagrant-domain", +@@ -71,58 +69,226 @@ Sunday 11 November 2018 20:19:25 +0100 (0:00:00.607) 0:10:36.974 *** + + changed: [localhost] + """ +-_, result = run_test(data, AnsibleOutputLexer()) +-print(result) +- +-# pylint: disable=line-too-long +-assert result == R"""ok: [windows] = { +-account: { +-account_name: vagrant-domain, +-
Bug#1034485: os-prober expects to run in a new private mount namespace, but new namespace is not private
Dear maintainer, I opened the following PR: https://salsa.debian.org/installer-team/os-prober/-/merge_requests/20 Best regards, Olivier
Bug#1034485: os-prober expects to run in a new private mount namespace, but new namespace is not private
Package: os-prober Version: 1.81 Severity: normal Dear Maintainer, During execution of os-prober, other processes on the system can see the temporary mounts to /var/lib/os-prober/mount even though os-prober runs in a separate mount namespace. In order to run os-prober in a more isolated mode, we introduced the newns.c source file a while ago. We build it to a binary and ship it in os-prober and os-prober-udeb. The original idea was to run os-prober in a private mount namespace. Sadly, calling the unshare(CLONE_NEWNS) system call is only enough to create a new mount namespace. But it is not enough to make the new namespace private. While we can patch newns.c to make the new mount namespace private, relying on unshare(1) from util-linux (which is an essential package) seems like a more viable option. I will open a PR with a potential fix. Thanks, Olivier See also: https://github.com/util-linux/util-linux/commit/f0f22e9c6f109f8c1234caa3173368ef43b023eb -- System Information: Debian Release: bookworm/sid APT prefers lunar APT policy: (500, 'lunar') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 6.1.0-16-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled Versions of packages os-prober depends on: ii grub-common 2.06-2ubuntu16 ii libc62.37-0ubuntu2 ii mount2.38.1-4ubuntu1 os-prober recommends no packages. os-prober suggests no packages. -- no debconf information
Bug#1030305: django-rich: FTBFS when building with TERM=unknown
Package: django-rich Followup-For: Bug #1030305 User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu lunar ubuntu-patch Control: tags -1 patch Hi, The patch was merged upstream after receiving some modifications. I'm attaching an updated debdiff which backports the patch applied upstream. Best regards, diff -Nru django-rich-1.4.0/debian/patches/0001-Fix-failing-tests-if-the-TERM-variable-is-set-to-unk.patch django-rich-1.4.0/debian/patches/0001-Fix-failing-tests-if-the-TERM-variable-is-set-to-unk.patch --- django-rich-1.4.0/debian/patches/0001-Fix-failing-tests-if-the-TERM-variable-is-set-to-unk.patch 1970-01-01 01:00:00.0 +0100 +++ django-rich-1.4.0/debian/patches/0001-Fix-failing-tests-if-the-TERM-variable-is-set-to-unk.patch 2023-02-02 12:05:10.0 +0100 @@ -0,0 +1,57 @@ +Description: Fix failing tests if the TERM variable is set to unknown +Origin: upstream, https://github.com/adamchainz/django-rich/commit/d437e9248322540b976d25f8a78a413e581c5dac +Bug-Debian: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1030305 +Bug-Ubuntu: https://bugs.launchpad.net/ubuntu/+source/django-rich/+bug/2004553 +Applied-Upstream: django-rich >> 1.4.0 +Last-Update: 2023-02-03 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +Index: django-rich-1.4.0/tests/test_management.py +=== +--- django-rich-1.4.0.orig/tests/test_management.py2023-02-03 09:25:32.460627256 +0100 django-rich-1.4.0/tests/test_management.py 2023-02-03 09:28:26.431930919 +0100 +@@ -1,12 +1,16 @@ + from __future__ import annotations + ++import os + from functools import partial + from inspect import Parameter, Signature, signature + from io import StringIO ++from typing import Any + from unittest import mock + + import pytest +-from django.core.management import BaseCommand, CommandError, call_command ++from django.core.management import BaseCommand ++from django.core.management import call_command as base_call_command ++from django.core.management import CommandError + from django.test import SimpleTestCase + from rich.console import Console + +@@ -28,6 +32,12 @@ + return True + + ++def call_command(*args: str, **kwargs: Any) -> None: ++# Ensure rich uses colouring and consistent width ++with mock.patch.dict(os.environ, TERM="", COLUMNS="80"): ++base_call_command(*args, **kwargs) ++ ++ + class RichCommandTests(SimpleTestCase): + def test_init_signature(self): + rc_signature = strip_annotations(signature(RichCommand.__init__)) +Index: django-rich-1.4.0/tests/test_test.py +=== +--- django-rich-1.4.0.orig/tests/test_test.py 2023-02-03 09:25:32.460627256 +0100 django-rich-1.4.0/tests/test_test.py 2023-02-03 09:25:32.460627256 +0100 +@@ -88,6 +88,9 @@ + **os.environ, + "DJANGO_SETTINGS_MODULE": "tests.settings", + "COVERAGE_PROCESS_START": str(SETUP_CFG_PATH), ++# Ensure rich uses colouring and consistent width ++"TERM": "", ++"COLUMNS": "80", + }, + **kwargs, + ) diff -Nru django-rich-1.4.0/debian/patches/series django-rich-1.4.0/debian/patches/series --- django-rich-1.4.0/debian/patches/series 2023-01-15 08:16:57.0 +0100 +++ django-rich-1.4.0/debian/patches/series 2023-02-02 12:05:10.0 +0100 @@ -1,3 +1,4 @@ Use-python3-within-test-call.patch tests-Drop-various-python-version-handling.patch tests-Another-bunch-of-adoptions-for-Python-3.11.patch +0001-Fix-failing-tests-if-the-TERM-variable-is-set-to-unk.patch
Bug#1013395: mpdecimal: FTBFS with Sphinx 5.0, docutils 0.18: make[1]: *** [debian/rules:60: override_dh_sphinxdoc] Error 255
Hi, > This happens because mpdecimal does not rebuild documentation from > source, but installs pre-built files provided by upstream. > I recommend to rebuild the documentation from source. I had a look but could not find the sources of the documentation. The upstream tarball only contains the pre-built version. Have we ever asked upstream if they would be okay to provide the sources of the doc? I would check but I can't find a link to archives of the libmpdec mailing lists [1]. Also, the website suggests that the mailing lists are not operational since Jan 2021 so I don't know how likely we will find out. [1] https://www.bytereef.org/mpdecimal/mailinglists.html Olivier
Bug#1030305: django-rich: FTBFS when building with TERM=unknown
Package: django-rich Followup-For: Bug #1030305 User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu lunar ubuntu-patch Control: tags -1 patch Adding a revised debdiff. This version applies the patch that I submitted upstream: https://github.com/adamchainz/django-rich/pull/103 The patch modifies the upstream test-suite rather than overriding the TERM variable at dh_auto_test time. Thanks! -- System Information: Debian Release: bookworm/sid APT prefers kinetic-updates APT policy: (500, 'kinetic-updates'), (500, 'kinetic-security'), (500, 'kinetic'), (100, 'kinetic-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 5.19.0-29-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru django-rich-1.4.0/debian/patches/0001-Fix-failing-tests-if-the-TERM-variable-is-set-to-unk.patch django-rich-1.4.0/debian/patches/0001-Fix-failing-tests-if-the-TERM-variable-is-set-to-unk.patch --- django-rich-1.4.0/debian/patches/0001-Fix-failing-tests-if-the-TERM-variable-is-set-to-unk.patch 1970-01-01 01:00:00.0 +0100 +++ django-rich-1.4.0/debian/patches/0001-Fix-failing-tests-if-the-TERM-variable-is-set-to-unk.patch 2023-02-02 12:05:10.0 +0100 @@ -0,0 +1,41 @@ +Description: Fix failing tests if the TERM variable is set to unknown + When running the test-suite with TERM=unknown, the tests fail because + the generated output is not "rich". + . + This happens because python rich will disable color and style if the TERM + variable is set to either "unknown" or "dumb" [1] + . + This patch monkey-patches the environment to make the test suite + insensitive to the TERM variable. + . + [1] https://rich.readthedocs.io/en/stable/console.html?highlight=unknown#environment-variables +Author: Olivier Gayot +Origin: upstream, https://github.com/adamchainz/django-rich/pull/103/commits/ac19f46572fb0550a9b7a688114c298dd7b6c872 +Bug-Debian: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1030305 +Bug-Ubuntu: https://bugs.launchpad.net/ubuntu/+source/django-rich/+bug/2004553 +Forwarded: https://github.com/adamchainz/django-rich/pull/103 +Last-Update: 2023-02-02 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +Index: django-rich-1.4.0/tests/test_management.py +=== +--- django-rich-1.4.0.orig/tests/test_management.py2023-02-02 15:06:49.343335475 +0100 django-rich-1.4.0/tests/test_management.py 2023-02-02 15:06:49.343335475 +0100 +@@ -3,6 +3,7 @@ + from functools import partial + from inspect import Parameter, Signature, signature + from io import StringIO ++import os + from unittest import mock + + import pytest +@@ -28,6 +29,9 @@ + return True + + ++# python rich will disable color/style if TERM is set to "unknown" or "dumb" ++# https://rich.readthedocs.io/en/stable/console.html?highlight=unknown#environment-variables ++@mock.patch.dict(os.environ, TERM="") + class RichCommandTests(SimpleTestCase): + def test_init_signature(self): + rc_signature = strip_annotations(signature(RichCommand.__init__)) diff -Nru django-rich-1.4.0/debian/patches/clear-TERM-variable.patch django-rich-1.4.0/debian/patches/clear-TERM-variable.patch --- django-rich-1.4.0/debian/patches/clear-TERM-variable.patch 1970-01-01 01:00:00.0 +0100 +++ django-rich-1.4.0/debian/patches/clear-TERM-variable.patch 2023-02-02 12:05:10.0 +0100 @@ -0,0 +1,20 @@ +Index: django-rich-1.4.0/tests/test_management.py +=== +--- django-rich-1.4.0.orig/tests/test_management.py2022-06-05 17:27:00.0 +0200 django-rich-1.4.0/tests/test_management.py 2023-02-02 14:50:44.876420552 +0100 +@@ -3,6 +3,7 @@ + from functools import partial + from inspect import Parameter, Signature, signature + from io import StringIO ++import os + from unittest import mock + + import pytest +@@ -28,6 +29,7 @@ + return True + + ++@mock.patch.dict(os.environ, TERM="") + class RichCommandTests(SimpleTestCase): + def test_init_signature(self): + rc_signature = strip_annotations(signature(RichCommand.__init__)) diff -Nru django-rich-1.4.0/debian/patches/series django-rich-1.4.0/debian/patches/series --- django-rich-1.4.0/debian/patches/series 2023-01-15 08:16:57.0 +0100 +++ django-rich-1.4.0/debian/patches/series 2023-02-02 12:05:10.0 +0100 @@ -1,3 +1,4 @@ Use-python3-within-test-call.patch tests-Drop-various-python-version-handling.patch tests-Another-bunch-of-adoptions-for-Python-3.11.patch +0001-Fix-failing-tests-if-the-TERM-variable-is-set-to-unk.patch
Bug#1030305: django-rich: FTBFS when building with TERM=unknown
Package: django-rich Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu lunar ubuntu-patch Dear Maintainer, When building on Ubuntu, the build-time test suite fails with the following error: self = def test_output_force_color(self): stdout = StringIO() call_command("example", "--force-color", stdout=stdout) > assert stdout.getvalue() == "\x1b[1;31mAlert!\x1b[0m\n" E AssertionError: assert 'Alert!\n' == '\x1b[1;31mAlert!\x1b[0m\n' E - [1;31mAlert![0m E + Alert! This happens because the underlying python rich library disables color/style when the TERM variable is set to unknown (which is what the Ubuntu builders have): > Setting the environment variable TERM to "dumb" or "unknown" will > disable color/style and some features that require moving the cursor, > such as progress bars. https://rich.readthedocs.io/en/stable/console.html?highlight=unknown#environment-variables Although currently Debian is not affected (I believe builders have a different value for TERM), it feels wrong that the build-time test-suite would depend on the TERM environment variable. Therefore, I think the patch still makes sense. In Ubuntu, the attached patch was applied to achieve the following: * clear TERM variable when running build test suite since python-rich disables color/style when TERM=unknown (LP: #2004553) Thanks for considering the patch. -- System Information: Debian Release: bookworm/sid APT prefers kinetic-updates APT policy: (500, 'kinetic-updates'), (500, 'kinetic-security'), (500, 'kinetic'), (100, 'kinetic-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 5.19.0-29-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru django-rich-1.4.0/debian/rules django-rich-1.4.0/debian/rules --- django-rich-1.4.0/debian/rules 2023-01-15 08:19:23.0 +0100 +++ django-rich-1.4.0/debian/rules 2023-02-02 12:05:10.0 +0100 @@ -12,4 +12,4 @@ # Upstream uses a more agressive call, wont work here # as at least the package we build isn't installed so # using a modified pytest call. - dh_auto_test -- --system=custom --test-args="{interpreter} -m coverage run -m pytest tests" + TERM= dh_auto_test -- --system=custom --test-args="{interpreter} -m coverage run -m pytest tests"
Bug#1030238: kitty: crypto test failure at build time when RTLIM_MEMLOCK is too low
Package: kitty Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu lunar ubuntu-patch Dear Maintainer, When building the package for ppc64el on Ubuntu, the crypto test fails with the following error: OSError: [Errno 12] Cannot allocate memory This happens because the ppc64el builders have their RLIMIT_MEMLOCK value set to the pagesize (64k) ; so only a single call to mlock(2) can succeed. Although the build does not fail in Debian (I believe the RLIMIT_MEMLOCK is set to a more sensible value), I think the patch could still make sense. There is an upstream bug report: https://github.com/kovidgoyal/kitty/issues/5466 In Ubuntu, the attached patch was applied to achieve the following: * disable crypto test if RLIMIT_MEMLOCK is too low (LP: #2004435) Thanks for considering the patch. -- System Information: Debian Release: bookworm/sid APT prefers kinetic-updates APT policy: (500, 'kinetic-updates'), (500, 'kinetic-security'), (500, 'kinetic'), (100, 'kinetic-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 5.19.0-29-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru kitty-0.26.5/debian/control kitty-0.26.5/debian/control --- kitty-0.26.5/debian/control 2023-01-30 20:32:11.0 +0100 +++ kitty-0.26.5/debian/control 2023-02-01 12:31:03.0 +0100 @@ -1,7 +1,6 @@ Source: kitty Priority: optional -Maintainer: Ubuntu Developers -XSBC-Original-Maintainer: James McCoy +Maintainer: James McCoy Section: x11 Standards-Version: 4.6.2 Rules-Requires-Root: no diff -Nru kitty-0.26.5/debian/patches/conditionally-skip-crypto-test.patch kitty-0.26.5/debian/patches/conditionally-skip-crypto-test.patch --- kitty-0.26.5/debian/patches/conditionally-skip-crypto-test.patch 1970-01-01 01:00:00.0 +0100 +++ kitty-0.26.5/debian/patches/conditionally-skip-crypto-test.patch 2023-02-01 12:31:03.0 +0100 @@ -0,0 +1,36 @@ +Description: skip crypto test if RLIMIT_MEMLOCK is too low + When building on ppc64el in Ubuntu, RLIMIT_MEMLOCK is set to the same value as + the pagesize. This means only a single call to mlock() can succeed. + Since the crypto test makes multiple calls to mlock(), the test fails in the + described environment. + Make sure we skip the test if the RLIMIT_MEMLOCK is too low. +Author: Olivier Gayot +Bug-Ubuntu: https://bugs.launchpad.net/bugs/2004435 +Forwarded: https://github.com/kovidgoyal/kitty/issues/5466#issuecomment-1412004797 +Last-Update: 2023-02-01 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +Index: kitty-0.26.5/kitty_tests/crypto.py +=== +--- kitty-0.26.5.orig/kitty_tests/crypto.py2023-02-01 13:17:33.662880021 +0100 kitty-0.26.5/kitty_tests/crypto.py 2023-02-01 13:25:53.667978172 +0100 +@@ -3,12 +3,19 @@ + + + import os ++import resource ++import unittest + + from . import BaseTest + + ++memlock_limit, _ = resource.getrlimit(resource.RLIMIT_MEMLOCK) ++pagesize = resource.getpagesize() ++ ++ + class TestCrypto(BaseTest): + ++@unittest.skipIf(memlock_limit <= pagesize, "RLIMIT_MEMLOCK is too low") + def test_elliptic_curve_data_exchange(self): + from kitty.fast_data_types import ( + AES256GCMDecrypt, AES256GCMEncrypt, CryptoError, EllipticCurveKey diff -Nru kitty-0.26.5/debian/patches/series kitty-0.26.5/debian/patches/series --- kitty-0.26.5/debian/patches/series 2023-01-04 12:14:15.0 +0100 +++ kitty-0.26.5/debian/patches/series 2023-02-01 12:31:03.0 +0100 @@ -1,3 +1,4 @@ bash-integration-fix-clone-in-kitty-not-.patch define-singlekey-bitfields-according-to-.patch try-to-fix-tests-failing-on-python-3.11.patch +conditionally-skip-crypto-test.patch
Bug#1028586: sysvinit: Missing 50-ubuntu-logging file when building on Ubuntu
I opened a pull-request to fix the issue: https://salsa.debian.org/debian/sysvinit/-/merge_requests/8
Bug#1028586: sysvinit: Missing 50-ubuntu-logging file when building on Ubuntu
Source: sysvinit Severity: normal Dear Maintainer, When building on Ubuntu, the sysvinit-utils package is supposed to include the lib/lsb/init-functions.d/50-ubuntu-logging file. The file used to be shipped with lsb-base on Ubuntu and should now be shipped with sysvinit-utils (only on Ubuntu) following the move of init-functions from lsb-base to sysvinit-utils. -- System Information: Debian Release: bookworm/sid APT prefers kinetic-updates APT policy: (500, 'kinetic-updates'), (500, 'kinetic-security'), (500, 'kinetic'), (100, 'kinetic-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 5.19.0-28-generic (SMP w/8 CPU threads; PREEMPT) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled
Bug#1020081: aft: FTBFS: make: *** [debian/rules:47: binary-indep] Error 25
Package: aft Followup-For: Bug #1020081 User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu kinetic ubuntu-patch Control: tags -1 patch Dear Maintainer, The last upload of aft changed the format of a doc file from HTML to RTF. The packaging was not updated accordingly. This prevents the package from building. In Ubuntu, the attached patch was applied to achieve the following: * Fix name of installed doc file (LP: #1992672) Thanks for considering the patch. -- System Information: Debian Release: bookworm/sid APT prefers jammy-updates APT policy: (500, 'jammy-updates'), (500, 'jammy-security'), (500, 'jammy'), (100, 'jammy-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 5.15.0-48-generic (SMP w/8 CPU threads) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru aft-5.098/debian/docs aft-5.098/debian/docs --- aft-5.098/debian/docs 2022-08-25 17:21:28.0 +0200 +++ aft-5.098/debian/docs 2022-10-12 17:41:56.0 +0200 @@ -1,5 +1,5 @@ aft-refman.html -aft2rtf-doc.html +aft2rtf-doc.rtf aft.gif debian/vim debian/jed
Bug#1021544: libcamera: FTBFS on ppc64el - is not a constant expression
Package: libcamera Version: 0~git20211108+1b30992b623e-4 Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu kinetic ubuntu-patch Dear Maintainer, libcamera currently fails to build from source on the ppc64el architecture. The latest snapshot (i.e., git20211108+1b30992b623e) contains long double expressions that GCC fails to determine at compile time on ppc64el. These expressions were not present in the previous snapshot. Tweaking the "guilty" expressions was enough to make the package build. This was the recommended solution on RHEL in a similar issue for a different package: https://bugzilla.redhat.com/show_bug.cgi?id=1538817#c6 Another option is to build with the -mlong-double-64 flag. In Ubuntu, the attached patch was applied to achieve the following: * Fix FTBFS on ppc64el with GCC (LP: #1992331) Thanks for considering the patch. -- System Information: Debian Release: bookworm/sid APT prefers jammy-updates APT policy: (500, 'jammy-updates'), (500, 'jammy-security'), (500, 'jammy'), (100, 'jammy-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 5.15.0-48-generic (SMP w/8 CPU threads) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru libcamera-0~git20211108+1b30992b623e/debian/patches/ppc64el-fix-ftbfs-gcc.patch libcamera-0~git20211108+1b30992b623e/debian/patches/ppc64el-fix-ftbfs-gcc.patch --- libcamera-0~git20211108+1b30992b623e/debian/patches/ppc64el-fix-ftbfs-gcc.patch 1970-01-01 01:00:00.0 +0100 +++ libcamera-0~git20211108+1b30992b623e/debian/patches/ppc64el-fix-ftbfs-gcc.patch 2022-10-10 10:23:29.0 +0200 @@ -0,0 +1,34 @@ +Description: fix expressions not considered constant on ppc64el with GCC + On ppc64el, GCC seems to be unable to determine the result of some long double + expressions at compile time. This makes libcamera fail to build from source on + ppc64el. + Make sure we only use constexpr on expressions that are considered constant on + all architectures. +Author: Olivier Gayot +Bug-Ubuntu: https://bugs.launchpad.net/ubuntu/+source/libcamera/+bug/1992331 +Forwarded: no +Last-Update: 2022-10-10 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +Index: libcamera-0~git20211108+1b30992b623e/src/ipa/raspberrypi/raspberrypi.cpp +=== +--- libcamera-0~git20211108+1b30992b623e.orig/src/ipa/raspberrypi/raspberrypi.cpp 2022-10-10 08:24:22.091329072 + libcamera-0~git20211108+1b30992b623e/src/ipa/raspberrypi/raspberrypi.cpp 2022-10-10 08:45:31.686454143 + +@@ -61,7 +61,7 @@ + /* Configure the sensor with these values initially. */ + constexpr double defaultAnalogueGain = 1.0; + constexpr Duration defaultExposureTime = 20.0ms; +-constexpr Duration defaultMinFrameDuration = 1.0s / 30.0; ++constexpr Duration defaultMinFrameDuration = 1s / 30.0; + constexpr Duration defaultMaxFrameDuration = 250.0s; + + /* +@@ -70,7 +70,7 @@ + * we rate-limit the controller Prepare() and Process() calls to lower than or + * equal to this rate. + */ +-constexpr Duration controllerMinFrameDuration = 1.0s / 60.0; ++constexpr Duration controllerMinFrameDuration = 1s / 60.0; + + LOG_DEFINE_CATEGORY(IPARPI) + diff -Nru libcamera-0~git20211108+1b30992b623e/debian/patches/series libcamera-0~git20211108+1b30992b623e/debian/patches/series --- libcamera-0~git20211108+1b30992b623e/debian/patches/series 1970-01-01 01:00:00.0 +0100 +++ libcamera-0~git20211108+1b30992b623e/debian/patches/series 2022-10-10 10:23:29.0 +0200 @@ -0,0 +1 @@ +ppc64el-fix-ftbfs-gcc.patch
Bug#1020493: user-setup: add sgx group to reserved-usernames (added as system group by udev.postint)
Package: user-setup Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu kinetic ubuntu-patch Dear Maintainer, The current version of udev.postinst in Debian creates four system groups: * kvm * input * render * sgx kvm, input and render were added to reserved-usernames when user-setup 1.83 was released. sgx however, was added to udev.postinst later with systemd 248-1 so it did not make it to reserved-usernames at the time. * Add sgx group to reserved-usernames; udev.postinst (src:systemd) adds it as a system group. Thanks for considering the patch. -- System Information: Debian Release: bookworm/sid APT prefers jammy-updates APT policy: (500, 'jammy-updates'), (500, 'jammy-security'), (500, 'jammy'), (100, 'jammy-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 5.15.0-47-generic (SMP w/8 CPU threads) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru user-setup-1.90ubuntu1/debian/control user-setup-1.90ubuntu2/debian/control --- user-setup-1.90ubuntu1/debian/control 2022-08-22 14:21:02.0 +0200 +++ user-setup-1.90ubuntu2/debian/control 2022-09-22 09:28:39.0 +0200 @@ -1,8 +1,7 @@ Source: user-setup Section: debian-installer Priority: optional -Maintainer: Ubuntu Developers -XSBC-Original-Maintainer: Debian Install System Team +Maintainer: Debian Install System Team Uploaders: Colin Watson , Steve McIntyre <93...@debian.org> Build-Depends: debhelper-compat (= 13), dpkg-dev (>= 1.7.0), po-debconf (>= 0.5.0) Standards-Version: 3.9.5 diff -Nru user-setup-1.90ubuntu1/reserved-usernames user-setup-1.90ubuntu2/reserved-usernames --- user-setup-1.90ubuntu1/reserved-usernames 2022-08-22 14:21:02.0 +0200 +++ user-setup-1.90ubuntu2/reserved-usernames 2022-09-22 09:28:39.0 +0200 @@ -111,6 +111,7 @@ saned sbuild scanner +sgx slocate ssh sshd
Bug#1004579: mplayer: FTBFS with ffmpeg 5.0
Package: mplayer Version: 2:1.4+ds1-3 Followup-For: Bug #1004579 User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu kinetic ubuntu-patch Control: tags -1 patch Dear Maintainer, The upstream release 1.5 supports ffmpeg 5 and would solve this issue. In the meantime, I have attempted to backport the upstream patches on top of the 1.4 version. In Ubuntu, the attached patch was applied to achieve the following: * Backport upstream patches to build against ffmpeg 5 (LP: #1987114) Thanks for considering the patch. -- System Information: Debian Release: bookworm/sid APT prefers jammy-updates APT policy: (500, 'jammy-updates'), (500, 'jammy-security'), (500, 'jammy'), (100, 'jammy-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 5.15.0-46-generic (SMP w/8 CPU threads) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru mplayer-1.4+ds1/debian/patches/r38215.patch mplayer-1.4+ds1/debian/patches/r38215.patch --- mplayer-1.4+ds1/debian/patches/r38215.patch 1970-01-01 01:00:00.0 +0100 +++ mplayer-1.4+ds1/debian/patches/r38215.patch 2022-08-19 17:10:25.0 +0200 @@ -0,0 +1,61 @@ +Description: av_helpers: switch to new lavc audio encode API. +Origin: upstream, commit: r38215 +Bug-Debian: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1004579 +Forwarded: no +Last-Update: 2022-08-19 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +Index: b/av_helpers.c +=== +--- a/av_helpers.c 2022-08-19 16:52:03.979666019 +0200 b/av_helpers.c 2022-08-19 16:52:37.0 +0200 +@@ -98,7 +98,6 @@ + if (!avcodec_initialized) { + show_av_version(MSGT_DECVIDEO, "libavcodec", LIBAVCODEC_VERSION_INT, + avcodec_version(), avcodec_configuration()); +-avcodec_register_all(); + avcodec_initialized = 1; + av_log_set_callback(mp_msp_av_log_callback); + } +@@ -109,7 +108,6 @@ + if (!avformat_initialized) { + show_av_version(MSGT_DEMUX, "libavformat", LIBAVFORMAT_VERSION_INT, + avformat_version(), avformat_configuration()); +-av_register_all(); + avformat_initialized = 1; + av_log_set_callback(mp_msp_av_log_callback); + } +@@ -132,8 +130,6 @@ + ctx->channels, + src_len / bps, bps); + } +-pkt.data = dst; +-pkt.size = dst_len; + frame->nb_samples = src_len / ctx->channels / bps; + if (planar) { + // TODO: this is horribly inefficient. +@@ -150,11 +146,22 @@ + } + } + } ++frame->format = ctx->sample_fmt; ++frame->channels = ctx->channels; + n = avcodec_fill_audio_frame(frame, ctx->channels, ctx->sample_fmt, src, src_len, 1); + if (n < 0) return 0; +-n = avcodec_encode_audio2(ctx, , frame, ); ++n = avcodec_send_frame(ctx, frame); ++av_init_packet(); ++got = avcodec_receive_packet(ctx, ); + av_frame_free(); + if (planar) av_free(src); + if (n < 0) return n; +-return got ? pkt.size : 0; ++if (got >= 0) { ++int size = pkt.size; ++if (size > dst_len) return -1; ++memcpy(dst, pkt.data, size); ++av_packet_unref(); ++return size; ++} ++return 0; + } diff -Nru mplayer-1.4+ds1/debian/patches/r38216.patch mplayer-1.4+ds1/debian/patches/r38216.patch --- mplayer-1.4+ds1/debian/patches/r38216.patch 1970-01-01 01:00:00.0 +0100 +++ mplayer-1.4+ds1/debian/patches/r38216.patch 2022-08-19 17:09:46.0 +0200 @@ -0,0 +1,86 @@ +Description: ad_ffmpeg: switch to new lavc API. +Origin: upstream, commit: r38216 +Bug-Debian: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1004579 +Forwarded: no +Last-Update: 2022-08-19 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +Index: b/libmpcodecs/ad_ffmpeg.c +=== +--- a/libmpcodecs/ad_ffmpeg.c 2022-08-19 16:52:04.011665773 +0200 b/libmpcodecs/ad_ffmpeg.c 2022-08-19 16:52:04.003665834 +0200 +@@ -310,34 +310,41 @@ + + static int decode_audio(sh_audio_t *sh_audio,unsigned char *buf,int minlen,int maxlen) + { ++int draining_started = 0; + unsigned char *start=NULL; +-int y,len=-1, got_frame; ++int y,len=-1; + AVFrame *frame = av_frame_alloc(); + + if (!frame) + return AVERROR(ENOMEM); + + while(lencontext, frame); ++ if (y == AVERROR(EAGAIN) || y == AVERROR_EOF) { ++ AVPacket pkt; + double pts; + int x=ds_get_packet_pts(sh_audio->ds,, ); + if(x<=0) { + start = NULL; + x = 0; + ds_parse(sh_audio->ds, , , MP_NOPTS_VALUE, 0); +-
Bug#1016539: libregexp-wildcards-perl: We want to make sure that the test-suite that libregexp-wildcards-perl has runs at autopkgtest time.
Package: libregexp-wildcards-perl Version: 1.05-2 Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu kinetic ubuntu-patch Dear Maintainer, *** Reporter, please consider answering these questions, where appropriate *** As part of a main inclusion request in Ubuntu, we want to ensure that the test suites runs at autopkgtest time. *** End of the template - remove these template lines *** *** /tmp/tmpi21cmu5u/bug_body In Ubuntu, the attached patch was applied to achieve the following: * Make sure the test-suite runs at autopkgtest time (LP: #1983397). Thanks for considering the patch. -- System Information: Debian Release: bookworm/sid APT prefers jammy-updates APT policy: (500, 'jammy-updates'), (500, 'jammy-security'), (500, 'jammy'), (100, 'jammy-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 5.15.0-43-generic (SMP w/8 CPU threads) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru libregexp-wildcards-perl-1.05/debian/control libregexp-wildcards-perl-1.05/debian/control --- libregexp-wildcards-perl-1.05/debian/control2018-11-10 15:44:55.0 +0100 +++ libregexp-wildcards-perl-1.05/debian/control2022-08-01 11:32:32.0 +0200 @@ -5,6 +5,7 @@ Vcs-Browser: https://git.in-ulm.de/cbiedl/libregexp-wildcards-perl Vcs-Git: https://git.in-ulm.de/cbiedl/libregexp-wildcards-perl.git Build-Depends: debhelper (>= 10~), +Testsuite: autopkgtest-pkg-perl Priority: optional Section: perl
Bug#1012606: keyutils: key.dns_resolver sets unlimited lifetime for cached records
Package: keyutils Version: 1.6.1-3 Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu kinetic ubuntu-patch Dear Maintainer, * What led up to the situation? Connect to a file share server hosted in the cloud using cifs. At some point the SMB server moves to a different cluster (with a set of IP addresses) and its DNS records are updated accordingly. * What was the outcome of this action? The kernel tries to reconnect to the old IP that was cached for unlimited time. * What outcome did you expect instead? After _some_ delay, the kernel should perform DNS resolution again and pick-up the new IP ; so it can reconnect to the server successfully. *** /tmp/tmprouv6pz2/bug_body In Ubuntu, the attached patch was applied to achieve the following: * Apply upstream patch that applies a configurable default TTL for DNS records. Thanks for considering the patch. -- System Information: Debian Release: bookworm/sid APT prefers jammy-updates APT policy: (500, 'jammy-updates'), (500, 'jammy-security'), (500, 'jammy'), (100, 'jammy-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 5.15.0-35-generic (SMP w/4 CPU threads) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru keyutils-1.6.1/debian/patches/apply-default-ttl-to-records.patch keyutils-1.6.1/debian/patches/apply-default-ttl-to-records.patch --- keyutils-1.6.1/debian/patches/apply-default-ttl-to-records.patch 1970-01-01 01:00:00.0 +0100 +++ keyutils-1.6.1/debian/patches/apply-default-ttl-to-records.patch 2022-02-28 10:43:38.0 +0100 @@ -0,0 +1,522 @@ +From 75e7568dc516db698093b33ea273e1b4a30b70be Mon Sep 17 00:00:00 2001 +From: David Howells +Date: Tue, 14 Apr 2020 16:07:26 +0100 +Subject: dns: Apply a default TTL to records obtained from getaddrinfo() + Address records obtained from getaddrinfo() don't come with any TTL + information, even if they're obtained from the DNS, with the result that + key.dns_resolver upcall program doesn't set an expiry time on dns_resolver + records unless they include a component obtained directly from the DNS, + such as an SRV or AFSDB record. + . + Fix this to apply a default TTL of 10mins in the event that we haven't got + one. This can be configured in /etc/keyutils/key.dns_resolver.conf by + adding the line: + . + default_ttl = + . + to the file. + . + Signed-off-by: David Howells + Reviewed-by: Ben Boeckel + Reviewed-by: Jeff Layton +Origin: upstream, https://git.kernel.org/pub/scm/linux/kernel/git/dhowells/keyutils.git/commit/?id=75e7568dc516db698093b33ea273e1b4a30b70be +Bug-Ubuntu: https://bugs.launchpad.net/ubuntu/+source/keyutils/+bug/1962453 +Last-Updated: 2022-02-28 + +--- + Makefile| 1 + + dns.afsdb.c | 16 ++-- + key.dns.h | 4 + + key.dns_resolver.c | 208 +--- + man/key.dns_resolver.8 | 25 -- + man/key.dns_resolver.conf.5 | 48 ++ + 6 files changed, 277 insertions(+), 25 deletions(-) + create mode 100644 man/key.dns_resolver.conf.5 + +--- a/Makefile b/Makefile +@@ -204,6 +204,7 @@ + $(INSTALL) -D key.dns_resolver $(DESTDIR)$(SBINDIR)/key.dns_resolver + $(INSTALL) -D -m 0644 request-key.conf $(DESTDIR)$(ETCDIR)/request-key.conf + mkdir -p $(DESTDIR)$(ETCDIR)/request-key.d ++ mkdir -p $(DESTDIR)$(ETCDIR)/keyutils + mkdir -p $(DESTDIR)$(MAN1) + $(INSTALL) -m 0644 $(wildcard man/*.1) $(DESTDIR)$(MAN1) + mkdir -p $(DESTDIR)$(MAN3) +--- a/dns.afsdb.c b/dns.afsdb.c +@@ -37,8 +37,6 @@ + */ + #include "key.dns.h" + +-static unsigned long afs_ttl = ULONG_MAX; +- + /* + * + */ +@@ -114,8 +112,8 @@ + } + } + +- afs_ttl = ttl; +- info("ttl: %u", ttl); ++ key_expiry = ttl; ++ info("ttl: %u", key_expiry); + } + + /* +@@ -203,8 +201,8 @@ + } + } + +- afs_ttl = ttl; +- info("ttl: %u", ttl); ++ key_expiry = ttl; ++ info("ttl: %u", key_expiry); + } + + /* +@@ -240,7 +238,7 @@ + /* look up the hostnames we've obtained to get the actual addresses */ + afsdb_hosts_to_addrs(handle, ns_s_an); + +- info("DNS query AFSDB RR results:%u ttl:%lu", payload_index, afs_ttl); ++ info("DNS query AFSDB RR results:%u ttl:%u", payload_index, key_expiry); + return 0; + } + +@@ -279,7 +277,7 @@ + /* look up the hostnames we've obtained to get the actual addresses */ + srv_hosts_to_addrs(handle, ns_s_an); + +- info("DNS query VL SRV RR results:%u ttl:%lu", payload_index, afs_ttl); ++ info("DNS query VL SRV RR results:%u ttl:%u", payload_index, key_expiry); + return 0; + } + +@@ -293,7 +291,7 @@ + + /* set
Bug#1012263: curl: Multi-line header support is broken
Package: curl Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu kinetic ubuntu-patch Dear Maintainer, In version 7.83.1, the support for multi-line headers is broken. This causes autopkgtest regressions in python-tornado: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1012224 In Ubuntu, the attached patch was applied to achieve the following: * Apply upstream patch to fix multi-line header support (LP: #1976619) Thanks for considering the patch. -- System Information: Debian Release: bookworm/sid APT prefers jammy-updates APT policy: (500, 'jammy-updates'), (500, 'jammy-security'), (500, 'jammy'), (100, 'jammy-backports') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 5.15.0-33-generic (SMP w/4 CPU threads) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru curl-7.83.1/debian/patches/12-http-restore-header-folding-behavior.patch curl-7.83.1/debian/patches/12-http-restore-header-folding-behavior.patch --- curl-7.83.1/debian/patches/12-http-restore-header-folding-behavior.patch 1970-01-01 01:00:00.0 +0100 +++ curl-7.83.1/debian/patches/12-http-restore-header-folding-behavior.patch 2022-06-02 13:44:50.0 +0200 @@ -0,0 +1,268 @@ +Description: restore header folding behavior + Folded header lines will now get passed through like before. The headers + API is adapted and will get and provide the content "un-folded". + Added test 1274 and extended test 1940 to verify. +Origin: upstream, https://github.com/curl/curl/pull/8899 +Bug-Ubuntu: https://bugs.launchpad.net/ubuntu/+source/curl/+bug/1976619 +Applied-Upstream: https://github.com/curl/curl/commit/c9b60f005358a364cbcddbebd8d12593acffdd84 +Last-Update: 2022-06-02 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +Index: b/docs/libcurl/curl_easy_header.3 +=== +--- a/docs/libcurl/curl_easy_header.3 2022-06-02 13:39:23.698800691 +0200 b/docs/libcurl/curl_easy_header.3 2022-06-02 13:39:23.694800445 +0200 +@@ -95,7 +95,9 @@ + + The data \fBvalue\fP field points to, comes exactly as delivered over the + network but with leading and trailing whitespace and newlines stripped +-off. The `value` data is nul-terminated. ++off. The `value` data is nul-terminated. For legacy HTTP/1 "folded headers", ++this API provides the full single value in an unfolded manner with a single ++whitespace between the lines. + + \fBamount\fP is how many headers using this name that exist, within the origin + and request scope asked for. +Index: b/lib/headers.c +=== +--- a/lib/headers.c2022-06-02 13:39:23.698800691 +0200 b/lib/headers.c2022-06-02 13:39:23.694800445 +0200 +@@ -216,6 +216,54 @@ + return CURLE_OK; + } + ++static CURLcode append_value(struct Curl_easy *data, const char *value, ++ size_t vlen) /* length of the incoming header */ ++{ ++ struct Curl_header_store *hs; ++ struct Curl_header_store *newhs; ++ size_t olen; /* length of the old value */ ++ size_t offset; ++ DEBUGASSERT(data->state.prevhead); ++ hs = data->state.prevhead; ++ olen = strlen(hs->value); ++ offset = hs->value - hs->buffer; ++ ++ /* skip all trailing space letters */ ++ while(vlen && ISSPACE(value[vlen - 1])) ++vlen--; ++ ++ /* save only one leading space */ ++ while((vlen > 1) && ISSPACE(value[0]) && ISSPACE(value[1])) { ++vlen--; ++value++; ++ } ++ ++ /* since this header block might move in the realloc below, it needs to ++ first be unlinked from the list and then re-added again after the ++ realloc */ ++ Curl_llist_remove(>state.httphdrs, >node, NULL); ++ ++ newhs = Curl_saferealloc(hs, sizeof(*hs) + vlen + olen + 1); ++ if(!newhs) ++return CURLE_OUT_OF_MEMORY; ++ /* ->name' and ->value point into ->buffer (to keep the header allocation ++ in a single memory block), which now potentially have moved. Adjust ++ them. */ ++ newhs->name = newhs->buffer; ++ newhs->value = >buffer[offset]; ++ ++ /* put the data at the end of the previous data, not the newline */ ++ memcpy(>value[olen], value, vlen); ++ newhs->value[olen + vlen] = 0; /* zero terminate at newline */ ++ ++ /* insert this node into the list of headers */ ++ Curl_llist_insert_next(>state.httphdrs, data->state.httphdrs.tail, ++ newhs, >node); ++ data->state.prevhead = newhs; ++ return CURLE_OK; ++} ++ ++ + /* + * Curl_headers_push() gets passed a full HTTP header to store. It gets called + * immediately before the header callback. The header is CRLF terminated. +@@ -242,6 +290,10 @@ + } + hlen = end - header + 1; + ++ if((header[0] == ' ') || (header[0] ==
Bug#1011276: requests: Fix autopkgtest when http_proxy, https_proxy or no_proxy is set
Package: requests Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu kinetic ubuntu-patch Dear Maintainer, When moving from 2.25.1+dfsg to 2.27.1+dfsg, we enabled the pytest test-suite in autopkgtest. Some of the tests fail when HTTP proxy variables are set: FAILED tests/test_requests.py::TestRequests::test_errors[http://doesnotexist.google.com-ConnectionError] FAILED tests/test_requests.py::TestRequests::test_respect_proxy_env_on_send_self_prepared_request FAILED tests/test_requests.py::TestRequests::test_respect_proxy_env_on_send_session_prepared_request FAILED tests/test_requests.py::TestRequests::test_respect_proxy_env_on_send_with_redirects FAILED tests/test_requests.py::TestRequests::test_respect_proxy_env_on_get - ... FAILED tests/test_requests.py::TestRequests::test_respect_proxy_env_on_request *** /tmp/tmp877wny2v/bug_body In Ubuntu, the attached patch was applied to achieve the following: * Fix autopkgtest when http_proxy, https_proxy or no_proxy variable is set (LP: #1974182) Thanks for considering the patch. -- System Information: Debian Release: bookworm/sid APT prefers jammy-updates APT policy: (500, 'jammy-updates'), (500, 'jammy-security'), (500, 'jammy') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 5.15.0-30-generic (SMP w/4 CPU threads) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru requests-2.27.1+dfsg/debian/patches/0002-Fix-tests-with-HTTP-proxy.patch requests-2.27.1+dfsg/debian/patches/0002-Fix-tests-with-HTTP-proxy.patch --- requests-2.27.1+dfsg/debian/patches/0002-Fix-tests-with-HTTP-proxy.patch 1970-01-01 01:00:00.0 +0100 +++ requests-2.27.1+dfsg/debian/patches/0002-Fix-tests-with-HTTP-proxy.patch 2022-05-19 14:14:07.0 +0200 @@ -0,0 +1,83 @@ +Description: Fix autopkgtest when HTTP/HTTPS proxy is set + The pytest suite does not expect the http_proxy, https_proxy and no_proxy + variables to be present in the environment. They make pytest fail and + therefore autopkgtest fail as well. +Author: Olivier Gayot +Bug: +Bug-Ubuntu: https://bugs.launchpad.net/ubuntu/+source/requests/+bug/1974182 +Forwarded: no +Last-Update: 2022-05-19 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +Index: b/tests/test_requests.py +=== +--- a/tests/test_requests.py 2022-01-05 16:35:04.0 +0100 b/tests/test_requests.py 2022-05-19 14:13:08.454208318 +0200 +@@ -535,8 +535,9 @@ + ('http://fe80::5054:ff:fe5a:fc0', InvalidURL) + )) + def test_errors(self, url, exception): +-with pytest.raises(exception): +-requests.get(url, timeout=1) ++with override_environ(http_proxy=None, https_proxy=None): ++with pytest.raises(exception): ++requests.get(url, timeout=1) + + def test_proxy_error(self): + # any proxy related error (address resolution, no route to host, etc) should result in a ProxyError +@@ -557,14 +558,14 @@ + requests.get(httpbin(), proxies={'http': 'http:///example.com:8080'}) + + def test_respect_proxy_env_on_send_self_prepared_request(self, httpbin): +-with override_environ(http_proxy=INVALID_PROXY): ++with override_environ(no_proxy=None, http_proxy=INVALID_PROXY): + with pytest.raises(ProxyError): + session = requests.Session() + request = requests.Request('GET', httpbin()) + session.send(request.prepare()) + + def test_respect_proxy_env_on_send_session_prepared_request(self, httpbin): +-with override_environ(http_proxy=INVALID_PROXY): ++with override_environ(no_proxy=None, http_proxy=INVALID_PROXY): + with pytest.raises(ProxyError): + session = requests.Session() + request = requests.Request('GET', httpbin()) +@@ -572,7 +573,7 @@ + session.send(prepared) + + def test_respect_proxy_env_on_send_with_redirects(self, httpbin): +-with override_environ(http_proxy=INVALID_PROXY): ++with override_environ(no_proxy=None, http_proxy=INVALID_PROXY): + with pytest.raises(ProxyError): + session = requests.Session() + url = httpbin('redirect/1') +@@ -581,13 +582,13 @@ + session.send(request.prepare()) + + def test_respect_proxy_env_on_get(self, httpbin): +-with override_environ(http_proxy=INVALID_PROXY): ++with override_environ(no_proxy=None, http_proxy=INVALID_PROXY): + with pytest.raises(ProxyError): + session = requests.Session() + session.get(httpbin()) + + def
Bug#1010891: libcloud: autopkgtest failure on big-endian architectures
Package: libcloud Version: 3.4.1-2 Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu kinetic ubuntu-patch Dear Maintainer, Since the move to upstream 3.4.1, a new test breaks autopkgtest on big-endian architectures: chan.recv.side_effect = ['臘'.encode('utf-32'), 'a', 'b'] stdout = client._consume_stdout(chan).getvalue() > self.assertEqual('\x00\x00&\x01\x00ab', stdout) E AssertionError: '\x00\x00&\x01\x00ab' != '\x00\x00\x00\x01' test/compute/test_ssh_client.py:563: AssertionError In Ubuntu, the attached patch was applied to achieve the following: * Fix autopkgtest failure on s390x (big-endian) (LP: #1973158) * The bug also affects Debian as shown in this log: https://ci.debian.net/data/autopkgtest/testing/s390x/libc/libcloud/21309362/log.gz Thanks for considering the patch. -- System Information: Debian Release: bookworm/sid APT prefers jammy-updates APT policy: (500, 'jammy-updates'), (500, 'jammy-security'), (500, 'jammy') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 5.15.0-27-generic (SMP w/4 CPU threads) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru libcloud-3.4.1/debian/control libcloud-3.4.1/debian/control --- libcloud-3.4.1/debian/control 2022-05-11 15:51:43.0 +0200 +++ libcloud-3.4.1/debian/control 2022-05-11 15:51:43.0 +0200 @@ -1,8 +1,7 @@ Source: libcloud Section: python Priority: optional -Maintainer: Ubuntu Developers -XSBC-Original-Maintainer: Debian Python Team +Maintainer: Debian Python Team Uploaders: Soren Hansen , Hans-Christoph Steiner , diff -Nru libcloud-3.4.1/debian/patches/df060de6ea325cba2b6411780728705252543bc3.patch libcloud-3.4.1/debian/patches/df060de6ea325cba2b6411780728705252543bc3.patch --- libcloud-3.4.1/debian/patches/df060de6ea325cba2b6411780728705252543bc3.patch 1970-01-01 01:00:00.0 +0100 +++ libcloud-3.4.1/debian/patches/df060de6ea325cba2b6411780728705252543bc3.patch 2022-05-11 15:51:43.0 +0200 @@ -0,0 +1,45 @@ +Description: test: fix test_ssh_client on big-endian architectures + The tests test_consume_*_chunk_contains_no_utf8_character are meant to + ensure support for non-UTF-8 characters. To do so, we use UTF-8 + characters and compare their binary representation after being fed to + paramiko. + + Unfortunately, UTF-32 is affected by endianness, resulting in failed + tests on big-endian architectures. + + Fixed by checking the endianness and comparing the UTF-32 characters to + their respective binary representation. +Author: Olivier Gayot +Bug-Ubuntu: https://bugs.launchpad.net/ubuntu/+source/libcloud/+bug/1973158 +Forwarded: https://github.com/apache/libcloud/pull/1693 +Last-Update: 2022-05-12 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +Index: libcloud-3.4.1/libcloud/test/compute/test_ssh_client.py +=== +--- libcloud-3.4.1.orig/libcloud/test/compute/test_ssh_client.py 2022-05-12 12:36:47.037706112 +0200 libcloud-3.4.1/libcloud/test/compute/test_ssh_client.py2022-05-12 12:37:34.529747766 +0200 +@@ -560,7 +560,10 @@ + chan.recv.side_effect = ['臘'.encode('utf-32'), 'a', 'b'] + + stdout = client._consume_stdout(chan).getvalue() +-self.assertEqual('\x00\x00&\x01\x00ab', stdout) ++if sys.byteorder == "little": ++self.assertEqual("\x00\x00&\x01\x00ab", stdout) ++else: ++self.assertEqual("\x00\x00\x00\x01", stdout) + self.assertEqual(len(stdout), 7) + + def test_consume_stderr_chunk_contains_non_utf8_character(self): +@@ -574,7 +577,10 @@ + chan.recv_stderr.side_effect = ['臘'.encode('utf-32'), 'a', 'b'] + + stderr = client._consume_stderr(chan).getvalue() +-self.assertEqual('\x00\x00&\x01\x00ab', stderr) ++if sys.byteorder == "little": ++self.assertEqual("\x00\x00&\x01\x00ab", stderr) ++else: ++self.assertEqual("\x00\x00\x00\x01", stderr) + self.assertEqual(len(stderr), 7) + + def test_keep_alive_and_compression(self): diff -Nru libcloud-3.4.1/debian/patches/series libcloud-3.4.1/debian/patches/series --- libcloud-3.4.1/debian/patches/series2022-05-11 15:51:43.0 +0200 +++ libcloud-3.4.1/debian/patches/series2022-05-11 15:51:43.0 +0200 @@ -3,3 +3,4 @@ 43b71091ebd159570302b695f0a244f4d0a7e497.patch 83bd562337abdda4b1ba6706999b13aa358bdd28.patch disable-online-test.patch +df060de6ea325cba2b6411780728705252543bc3.patch
Bug#1010881: libcloud: Autopkgtest failure when no network or https_proxy/http_proxy set
Package: libcloud Version: 3.4.1-2 Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu kinetic ubuntu-patch Dear Maintainer, Since the move to 3.4.1, libcloud fails autopkgtest when either: * Internet access is not available * http_proxy or https_proxy is set In Ubuntu, the attached patch was applied to achieve the following: * Disable test that requires access to the Internet. * Applied patches forwarded upstream to fix failures when HTTP proxy variables are set (LP: #1972888) Thanks for considering the patch. -- System Information: Debian Release: bookworm/sid APT prefers jammy-updates APT policy: (500, 'jammy-updates'), (500, 'jammy-security'), (500, 'jammy') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 5.15.0-27-generic (SMP w/4 CPU threads) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru libcloud-3.4.1/debian/patches/43b71091ebd159570302b695f0a244f4d0a7e497.patch libcloud-3.4.1/debian/patches/43b71091ebd159570302b695f0a244f4d0a7e497.patch --- libcloud-3.4.1/debian/patches/43b71091ebd159570302b695f0a244f4d0a7e497.patch 1970-01-01 01:00:00.0 +0100 +++ libcloud-3.4.1/debian/patches/43b71091ebd159570302b695f0a244f4d0a7e497.patch 2022-05-11 15:51:43.0 +0200 @@ -0,0 +1,49 @@ +Description: stop HTTP server (thread) properly + test/test_http.py: stop HTTP server (thread) properly + + The HttpLayerTestCase test class creates a HTTPServer instance and runs + it in a separate thread. + + After running all the test-cases, we attempt to close the server and the + thread by killing the thread. Unfortunately, the code that does that is + unreachable because the tear-down member function is called tearDownCls + intead of tearDownClass. + + Moreover, there is no threading.Thread.kill function. This was undetected + because the code was unreachable. + + The proper way to clean things up is to: + + 1. Stop the HTTP server using HTTPServer.shutdown() + 2. Release the resources using HTTPServer.server_close() + 3. Join the thread using threading.Thread.join() + + The original patch from upstream was missing the call to + HTTPServer.server_close(), which was added later. + + This patch is not strictly necessary to fix the tests but subsequent patches + depend on tearDownClass being spelled properly. +Author: Olivier Gayot +Bug-Ubuntu: https://bugs.launchpad.net/ubuntu/+source/libcloud/+bug/1972888 +Forwarded: https://github.com/apache/libcloud/pull/1690 +Applied-Upstream: https://github.com/apache/libcloud/pull/1690/commits/43b71091ebd159570302b695f0a244f4d0a7e497 +Last-Update: 2022-05-11 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +Index: b/libcloud/test/test_http.py +=== +--- a/libcloud/test/test_http.py 2022-05-11 17:11:36.733859740 +0200 b/libcloud/test/test_http.py 2022-05-11 17:11:36.729859732 +0200 +@@ -116,8 +116,10 @@ + cls.mock_server_thread.start() + + @classmethod +-def tearDownCls(cls): +-cls.mock_server_thread.kill() ++def tearDownClass(cls): ++cls.mock_server.shutdown() ++cls.mock_server.server_close() ++cls.mock_server_thread.join() + + def test_prepared_request_empty_body_chunked_encoding_not_used(self): + connection = LibcloudConnection(host=self.listen_host, port=self.listen_port) diff -Nru libcloud-3.4.1/debian/patches/83bd562337abdda4b1ba6706999b13aa358bdd28.patch libcloud-3.4.1/debian/patches/83bd562337abdda4b1ba6706999b13aa358bdd28.patch --- libcloud-3.4.1/debian/patches/83bd562337abdda4b1ba6706999b13aa358bdd28.patch 1970-01-01 01:00:00.0 +0100 +++ libcloud-3.4.1/debian/patches/83bd562337abdda4b1ba6706999b13aa358bdd28.patch 2022-05-11 15:51:43.0 +0200 @@ -0,0 +1,46 @@ +Description: test/test_http.py: also backup/restore HTTP proxy settings + + The HttpLayerTestCase class relies on the LibcloudConnection class - + which honors https_proxy and http_proxy variables. + + Since the tests expect connections to the loopback interface (e.g., + 127.0.0.1), we need to clear the http_proxy and https_proxy + variables before running the tests. Later on, we need to restore the + variables just like we did in test/test_connection.py +Author: Olivier Gayot +Bug-Ubuntu: https://bugs.launchpad.net/ubuntu/+source/libcloud/+bug/1972888 +Forwarded: https://github.com/apache/libcloud/pull/1692 +Applied-Upstream: https://github.com/apache/libcloud/commit/83bd562337abdda4b1ba6706999b13aa358bdd28 +Last-Update: 2022-05-11 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +Index: b/libcloud/test/test_http.py
Bug#1010803: python-fluids: Unpatched tests from .pc directory are executed as normal tests
Package: python-fluids Version: 1.0.9-1 Severity: normal Tags: patch User: ubuntu-de...@lists.ubuntu.com Usertags: origin-ubuntu kinetic ubuntu-patch Dear Maintainer, * Some tests (run by autopkgtest through pytest) require network. * The existing patch d/patches/fix-online-tests.patch is meant to modify the tests to get rid of the requirement to be online. * Unfortunately, after the patch is applied by quilt, pytest runs twice over each modified test, e.g.: * tests/test_design_climate.py * .pc/fix-online-tests.patch/tests/test_design_climate.py * Tests under the .pc directory are unpatched, resulting in autopkgtest failure if network is not available. In Ubuntu, the attached patch was applied to achieve the following: * Prevent autopkgtest failure (system under test does not have access to the Internet in Ubuntu) * Although autopkgtest does not fail in Debian currently (does the system under test have access to the Internet?), autopkgtest logs [1] show that pytest executes tests from under the .pc directory. This could lead to unexpected side-effects: .pc/fix-online-tests.patch/tests/test_design_climate.py::test_heating_degree_days PASSED [ 0%] .pc/fix-online-tests.patch/tests/test_design_climate.py::test_cooling_degree_days PASSED [ 0%] .pc/fix-online-tests.patch/tests/test_design_climate.py::test_month_average_temperature PASSED [ 0%] .pc/fix-online-tests.patch/tests/test_design_climate.py::test_IntegratedSurfaceDatabaseStation PASSED [ 0%] .pc/fix-online-tests.patch/tests/test_design_climate.py::test_data PASSED [ 0%] .pc/fix-online-tests.patch/tests/test_design_climate.py::test_correct_WBAN PASSED [ 0%] .pc/fix-online-tests.patch/tests/test_design_climate.py::test_get_closest_station PASSED [ 0%] .pc/fix-online-tests.patch/tests/test_design_climate.py::test_get_station_year_text PASSED [ 0%] * Exclude unpatched tests (in .pc/ directory) from being executed by pytest (LP: #1972805) [1] https://ci.debian.net/data/autopkgtest/unstable/amd64/p/python-fluids/19996479/log.gz Thanks for considering the patch. -- System Information: Debian Release: bookworm/sid APT prefers jammy-updates APT policy: (500, 'jammy-updates'), (500, 'jammy-security'), (500, 'jammy') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 5.15.0-27-generic (SMP w/4 CPU threads) Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system) LSM: AppArmor: enabled diff -Nru python-fluids-1.0.9/debian/patches/pytest-exclude-quilt-dir.patch python-fluids-1.0.9/debian/patches/pytest-exclude-quilt-dir.patch --- python-fluids-1.0.9/debian/patches/pytest-exclude-quilt-dir.patch 1970-01-01 01:00:00.0 +0100 +++ python-fluids-1.0.9/debian/patches/pytest-exclude-quilt-dir.patch 2022-05-10 10:32:08.0 +0200 @@ -0,0 +1,23 @@ +Description: Exclude .pc/ directory when executing pytest + We have debian patches (managed by quilt) that modify the tests that are run + with pytest. Unfortunately, pytest also runs over unpatched tests that quilt + stores in the .pc/ directory. This results in autopkgtest failures. + This patch instructs pytest to exclude the .pc/ directory. +Author: Olivier Gayot +Bug-Ubuntu: https://bugs.launchpad.net/ubuntu/+source/python-fluids/+bug/1972805 +Last-Update: 2022-05-10 +--- +This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ +Index: python-fluids-1.0.9ubuntu1/pytest.ini +=== +--- python-fluids-1.0.9ubuntu1.orig/pytest.ini 2022-05-10 10:31:18.887277881 +0200 python-fluids-1.0.9ubuntu1/pytest.ini 2022-05-10 10:31:41.219339940 +0200 +@@ -1,6 +1,6 @@ + [pytest] + addopts = --doctest-glob='*.rst' --ignore='setup.py' --ignore='manual_runner.py' --ignore='make_test_stubs.py' +-norecursedirs = .cache .git htmlcov notebooks dist ++norecursedirs = .cache .git .pc htmlcov notebooks dist + doctest_optionflags= NORMALIZE_WHITESPACE + + markers = diff -Nru python-fluids-1.0.9/debian/patches/series python-fluids-1.0.9/debian/patches/series --- python-fluids-1.0.9/debian/patches/series 2022-01-25 19:52:10.0 +0100 +++ python-fluids-1.0.9/debian/patches/series 2022-05-10 10:31:27.0 +0200 @@ -2,3 +2,4 @@ remove-shields-dot-io-images.patch use-debian-mathjax.patch fix-online-tests.patch +pytest-exclude-quilt-dir.patch