| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
| |
The upstream headers_check target has been a no-op since Linux 5.5,
and appears to have been combined with the headers_install target. It
now produces a warning and is due to be removed entirely after Linux
5.15.
|
| |
|
| |
|
|
|
|
|
| |
* debian/rules.real: Install module.lds under arch directory
* kbuild: Look for module.lds under arch directory too
|
|
|
|
| |
Signed-off-by: Sudip Mukherjee <sudipm.mukherjee@gmail.com>
|
|\
| |
| |
| |
| |
| | |
bpftool: Add bpftool package
See merge request kernel-team/linux!233
|
| | |
|
| | |
|
| | |
|
| | |
|
|/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The debian kernel build switched to use debhelper
compatibility level 12 with this commit:
commit 59a5af36cbf1cc01ef48b91719f999a699d99fab
Author: Ben Hutchings <ben@decadent.org.uk>
debhelper started complaining about level 9, so it's time to upgrade
again.
This change broke the kernel build on the hppa architecture.
Since compat level 12, dh_strip runs the "strip" command on any vmlinuz*
files, which wasn't the case before.
On hppa the 64-bit build needs to run "hppa64-linux-strip" instead of
"strip".
Fix it by adding "vmlinuz" to the exclude list of dh_strip, and thus
behave as compat level 9 did before.
Tested on hppa and x86-64.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
These metapackages were originally introduced to support building
binary packages of out-of-tree modules. They were used by these
source packages:
* linux-modules-{extra,contrib,nonfree}-2.6 build-depended on them,
but were removed before the squeeze release.
* After that, nvidia-graphics-modules build-dependend on them, but
was removed before the stretch release.
* linux-latest build-depended on them for some time, to prevent it
building metapackages before their dependencies. It has now been
merged into this source package.
So it appears that they are now obsolete.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
debhelper started complaining about level 9, so it's time to upgrade
again.
* Build-Depend on debhelper-compat and remove debian/compat
(also in the signed package template)
* hyperv-daemons: Use dh_installsystemd instead of dh_systemd_{enable,start}
* hyperv-daemons: Add "Pre-Depends: ${misc:Pre-Depends}"
I compared the binary packages resulting from this and found the
following differences:
* hyperv-daemons:
- Pre-Depends on init-system-helpers.
- Maintainer scripts restart services after upgrade rather than
stopping before upgrade and starting after.
- Maintainer scripts use deb-systemd-invoke and invoke-rc.d's
--skip-systemd-native option.
- postrm script runs systemctl --system daemon-reload.
This all seems safe and reasonable.
* libcpupower1: shlibs file specifies a versioned dependency of at
least the current upstream version. This shouldn't make any
practical difference because dpkg-shlibdeps actually uses the
symbols files to determine the minimum version.
* libtraceevent1: Same change as libcpupower1.
* linux-image-amd64-signed-template: compat and control files are
updated as intended.
* linux-perf-5.6: 2 of the example source files were compressed,
and now none are. This is probably helpful.
There are some additional cosmetic changes to generated code in
maintainer scripts.
|
|\
| |
| |
| | |
Release linux (5.5.17-1).
|
| |
| |
| |
| | |
Signed-off-by: Sudip Mukherjee <sudipm.mukherjee@gmail.com>
|
|/
|
|
|
|
|
|
|
|
| |
We already include debian/rules.defs, which includes
/usr/share/dpkg/default.mk, which includes
/usr/share/dpkg/architecture.mk. So there's no need to include
it directly as well.
(We were already doing that before I added the direct inclusion, so I
have no idea why I made that change.)
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Most of the work is already done by Sphinx running under Python 3
(since version 4.18~rc4-1~exp1).
However, Documentation/Makefile now invokes another Python script
(scripts/jobserver-exec) using $(PYTHON), which is set to "python" in
the top-level Makefile. None of our build-dependencies depend on
Python 2 any more, so this fails.
Override this by by adding PYTHON=python3 to the recursive make
commands for the document build.
|
|
|
|
|
|
|
| |
Add build rules for libtracevent which will eventually call Makefile
added earlier and will build it and package the shared library.
Signed-off-by: Sudip Mukherjee <sudipm.mukherjee@gmail.com>
|
| |
|
| |
|
|
|
|
|
| |
Since the NEWS file will now be installed in "real" linux-image
packages, rename it to remove the ".meta".
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
debian/{,signing_templates/}rules.real: Add targets for building
metapackages
debian/rules.real: Make metapackage targets dependencies of
binary-arch-arch or binary-indep, as appropriate
debian/bin/gencontrol{,_signed}.py:
* Expand control file templates for metapackages along with the
corresponding "real" package templates
* Implement a "meta" config option under [packages] to allow
suppressing them
* Define the additional variables that these templates use
* Expand the bug-presubj template for linux-image metapackages
debian/bin/gencontrol_signed.py: Add invocation of the install-meta
rule to debian/rules.gen
|
|
|
|
|
|
|
|
| |
Change the -fdebug-prefix-path option's value so the "/" after the
source directory name is deleted rather than prefixed with ".". This
shouldn't make any practical difference, but it makes the debug info
slightly smaller and is consistent with the value upstream sets for
the -fmacro-prefix-path option.
|
|
|
|
|
|
|
|
| |
Kbuild ignores CFLAGS because compiler options for user-space may be
entirely unsuitable for the kernel. However, it does support a
KCFLAGS variable which is used for both built-in and modular code.
Add our options to that instead of both CFLAGS_KERNEL and
CFLAGS_MODULE.
|
|
|
|
|
|
|
|
|
| |
Ideally we would run mkmakefile here, but that embeds its own
absolute filename, which we want to avoid for reproducibility.
However, all that's needed now is a single "include". The top-level
Makefile uses $(CURDIR) and its own filename from $(MAKEFILE_LIST)
to work out the rest.
|
|
|
|
|
| |
gcc 9 and glibc 2.30 dropped support for it, and it's gone from
ports.debian.org.
|
|
|
|
|
|
|
|
| |
libbpf seems to be setting a proper soname and versioning symbols now.
Don't replace upstream's soversion with the kernel version.
Refresh "generate pkg-config file for libbpf" based on the upstream
version.
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Since commit f5f169ba999c "Split build rules for tools to allow
skipping them." debian/rules.d/Makefile is not used and the current
kernel's UAPI headers are not installed. This hasn't caused breakage
yet, probably because many tools have their own workaround using
include/uapi etc. directly, but could break backports builds at some
point.
Move the build of userland headers up into debian/rules.real and
make all tools build targets depend on it.
|
|
|
|
|
|
| |
The REAL_VERSION and PACKAGE_NAME variables aren't defined at the
point where we use them to check whether the package is enabled.
Replace them with their intended expansions to avoid this problem.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The packages we should build are restricted by:
* Package configuration in debian/config (limits which binary packages are
included in debian/control)
* Architecture (specified per package in debian/templates/control.* and
then in debian/control)
* Build profile (specified per package in debian/templates/control.* and
then in debian/control)
The logic for these restrictions is currently repeated in
debian/rules.real, but sometimes it becomes inconsistent with
debian/control (as with my recent changes for libbpf).
dh_listpackages reads debian/control and filters it by the current
host architecture and build profiles, so that it reliably reports
which packages we should build.
Therefore:
* Replace the logic in debian/rules.real with checks for package names
in the output of dh_listpackages
* Remove the redundant flag variables passed by debian/rules and
debian/rules.gen
* Remove the special-casing of stage1 in debian/rules and
debian/rules.gen
|
|
|
|
| |
The architecture filtering needs to be done in debian/rules.real as well.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
On powerpc architectures that may use a bootwrapper, we create a
temporary build_<arch>_<featureset>_<flavour>_bootwrapper directory
for each kernel configuration to hold the related tools which we won't
install for real (because they are always native).
This directory is then matched by the wildcard used in building
linux-config, causing linux-config packages to contain spurious
(empty) kconfig files based on these directories in addition to the
real kconfig files.
Rename the temporary directory to avoid matching that wildcard.
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Split the rules in d/rules.real so that the [un]versioned_tools
knobs can be used to avoid building them.
This is necessary since the build-dependency were moved to be
conditional on those knobs, so the build fails when the
unversioned tools are set to disabled as libpci-dev is not
installed but the tools are built and fail due to it missing.
Signed-off-by: Luca Boccassi <bluca@debian.org>
|
| |
|
|
|
|
|
|
|
| |
CONFIG_DEBUG_INFO and CONFIG_MODULE_SIG are added in gencontrol.py,
so be consistent with that.
This unfortunately requires some ugly escaping of quotes.
|
|
|
|
|
|
| |
instead of editing path
This fixes one of the remaining classes of pycodestyle errors.
|
|\
| |
| |
| |
| |
| |
| |
| | |
Release linux (4.18.10-1).
- Drop new patches that are already included upstream
- Keep ABI number set to "trunk"
- Refresh arm64 APEI workaround patch for 4.19
|
| |
| |
| |
| | |
group specified, to fix reproducibility issues.
|
|\|
| |
| |
| |
| |
| |
| | |
Release linux (4.18.8-1).
- Drop ABI reference files and ABI maintenance patch
- Replace ccp driver patch with upstream version that applies to 4.19
|
| |
| |
| |
| |
| |
| |
| | |
Since the linux-headers-common package was changed to be
architecture-independent, it won't include any binNMU changelog
fragment that should be installed for a per-flavour linux-headers
package.
|
| |
| |
| |
| |
| |
| |
| | |
Fixes lintian warning patch-file-present-but-not-mentioned-in-series.
Also preparation for using dgit, which will remove everything except
the main patch series under debian/patches.
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Currently we don't set all the architecture variables, so direct use
of these makefiles (not under dpkg-architecture or dpkg-buildpackage)
may fail.
In case ARCH and DEB_HOST_ARCH don't match in rules.real, we really
should fail, so add an explicit check for that.
This also fixes lintian warning
debian-rules-sets-dpkg-architecture-variable.
|
| |
| |
| |
| |
| |
| |
| | |
It works better if we specify the correct directories containing
modules. This has been broken since linux-tools 4.0.2-1.
Closes: #908547
|
|/
|
|
|
|
|
|
|
|
|
| |
Merge the configuration and default-configuration directories,
using per-architecture overrides in package-list.
This requires a newer version of kernel-wedge to support
Depends_<arch> properly.
The only immediate change to debian/control is to remove the
different description for nic-modules on sparc64.
|