| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
| |
Fixes: cc417b4263e9 ("Allow to override top-level config")
Signed-off-by: Salvatore Bonaccorso <carnil@debian.org>
|
| |
|
|
|
|
|
|
| |
Including kernelarch-X configs is kind of a special case. This config
was dropped if someone ever wanted to override config files for the
top-level config.
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
Setting $DEBIAN_KERNEL_DISABLE_DEBUG or clearing the [build]debug-info
config variable should disable generation of debug info and associated
packages. However the recent refactoring of the handling of such
environment variables resulted in different switches being obeyed in
different places.
Set the build_debug variable based on both switches and then use that
consistently.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The Debian installer can choose a suitable kernel to install based on
hardware information, but image builders can't do that and they need
to know a sensible default kernel for each architecture.
I've resisted defining such a default before now, because there is no
reasonable default on some architectures. But we can still provide a
default for those architectures where there is:
* [alpha,sparc64] default to SMP flavour; both flavours should run on
everything but this seems more useful
* [amd64] default to amd64, which should run on everything
* [arm64] default to arm64, which should run on everything
* [armhf] default to armmp, which should run on everything; non-LPAE
CPUs are very common
* [hppa] default to parisc, which should run on everything
* [i386] default to 686-pae, which requires PAE; non-PAE CPUs are
quite rare
* [ia64] default to itanium, which should run on everything
* [m68k,mips64r6*,ppc64*,riscv64,s390x] default to the only flavour
* [mipsr6*] default to 32-bit flavour; 64-bit systems should use the
64-bit architecture
The architectures with no default are:
* [armel,mips,mips64,mips64el,mipsel,sh4] flavours are platform-specific
* [powerpc] 32-bit flavour won't boot on 64-bit systems and vice versa
Closes: #960209
|
|
|
|
|
|
|
|
|
|
|
| |
Change the code for linux-image and linux-headers metapackages to
look more like that in gencontrol_signed.py:
* The conditions for generating these metapackages are now identical.
Move the code for them into a single if-block.
* Collect the metapackages into a list and assert that there are the
expected number of packages (1 for each template).
|
|
|
|
|
|
|
| |
When buster enters LTS we'll likely want to add a backport of the
bullseye kernel, including signed packages. Correct the definition of
the source_suffix template variable here to allow for that. Also
define source_basename, for consistency with gencontrol.py.
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
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.
|
|
|
|
|
|
|
|
|
|
| |
After a binNMU of linux, we will trigger the signing service to attempt
source uploads of linux-signed-*. But their version strings will end in
"+b<n>", which will be rejected by dpkg-source as invalid.
Replace the "+" with a "." in the linux-signed-* source version. The
binary versions will still include "+b<n>", which is weird for a
non-binNMU but should still be accepted.
|
|
|
|
|
| |
This comment hasn't matched the code since commit 47395473022a
"Write a complete changelog to signing template".
|
|
|
|
|
|
|
|
|
|
| |
Since we got rid of src:linux-latest, there are no longer any
binary dependencies relating src:linux and src:linux-signed-*, and
they can transition to testing separately.
Building linux-headers metapackages from src:linux-signed-* introduces
a new binary dependency that should ensure that all these packages
transition to testing together.
|
| |
|
| |
|
|\
| |
| |
| |
| |
| |
| | |
Release linux (5.3.9-3).
* Drop ABI reference files
* Drop patches already included in 5.4-rc8
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Closes: #943953
Cherry-pick these from upstream:
* arm64: vdso32: Fix broken compat vDSO build warnings
* arm64: Kconfig: Make CONFIG_COMPAT_VDSO a proper Kconfig option
Then update our build config:
* debian/bin/gencontrol.py: Optionally define CROSS_COMPILE_COMPAT make
variable
* Enable COMPAT_VDSO and set $(CROSS_COMPILE_COMPAT) instead of setting
CROSS_COMPILE_COMPAT_VDSO
|
|/
|
|
| |
Signed-off-by: Héctor Orón Martínez <zumbi@debian.org>
|
|
|
|
|
|
|
| |
Commit c5e1ff85ebc2 "gencontrol: Generalise substitution of debhelper
config template" used "image" as the base name for debhelper
configuration templates for all linux-image packages generated from
linux-signed-*. We need to use "image.meta" for the meta-packages.
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Some debhelper config files should be executable, because either:
1. debhelper should execute them (with dh-exec) instead of reading them
directly. Currently we change the permissions after substitution.
2. They are included in signed-template binary packages, and lintian
will complain about them being non-executable scripts. Currently
we override these warnings.
Replace the special case code by copying the file permissions bits
from the template to the output file in substitute_debhelper_config().
|
|
|
|
|
|
|
|
|
| |
dpkg keeps all directories of the old package version (and renames
the old files) until an upgrade is complete. This prevents simply
replacing a directory with a symlink. We have to use
dpkg-maintscript-helper to work around this limitation.
Closes: #942861
|
|
|
|
|
|
|
|
|
|
|
|
| |
Currently we have two copies of the _substitute_file() method, and
lots of somewhat similar invocations of it since different binary
packages need different sets of debhelper configs.
We already name the templates in a fairly consistent way, so we can
replace this with a loop that tries to expand all possible templates
for a package and ignores those that are missing. Add a method for
this to the Gencontrol base class and use it in both subclasses
where we prevously used _substitute_file().
|
|
|
|
|
|
|
|
| |
It's somewhat clearer what we're substituting into a format string
if we name the variables inline, similarly to template files.
In some places we format a mixture of template and other variables.
Leave those unchanged for now.
|
|
|
|
|
|
| |
self.vars only holds the template variables that don't depend on
architecture/featureset/flavour. We should consistently use the
vars parameter for substitutions, where available.
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
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
|
|
|
|
|
|
|
|
| |
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.
|
| |
|
|\
| |
| |
| |
| |
| | |
* Drop ABI reference and ABI maintenance patch
* Drop patches that are already included in 5.2
* Take master's version of the PREEMPT_RT patch set
|
| |\
| | |
| | |
| | |
| | |
| | | |
Release linux (4.19.37-3).
* Drop all the new patches, which are already included in 5.1.3
|
| |\ \
| | | |
| | | |
| | | |
| | | |
| | | | |
* Drop changes to rt patchset
* Refresh "generate pkg-config file for libbpf"
* Drop changes to drivers/firmware/google config
|
| |\ \ \
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Release linux (4.19.28-1).
* Drop patches that are already included in 5.0.1
* Drop ABI bump
|
| | | | | |
|
| |_|_|/
|/| | | |
|
| |_|/
|/| | |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The "recommends" field set in the [image] section for these
configurations overrode the field at the top level. We want
gencontrol.py to concatenate the relations in this section at all
levels.
The ConfigCore.get_merge method supports doing this, but only with
list fields So we need to specify in the config schema that these
fields are comma-separated lists.
|
| | |
| | |
| | |
| | | |
.json file, fixing reproducibility issues.
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Since we don't use the Release and Packages files to verify the
packages we download, it's worth using TLS to reduce the risk of
a man-in-the-middle corrupting them.
ftp.ports.debian.org and security.debian.org don't support TLS
in general, so use deb.debian.org for the ports and security
archives.
|
| |/
|/|
| |
| |
| |
| | |
If the changelog distribution is *-security, fetch from the security
archive. Otherwise, try the main archive, ports, incoming, and
incoming.ports in that order.
|
| |
| |
| |
| |
| | |
We already used it for udebs but not for the regular signed image
package.
|
| |
| |
| |
| |
| | |
We only process the first package found in these templates so fail
if there are more.
|
| | |
|
| |
| |
| |
| |
| | |
Follow the schema change made in
https://salsa.debian.org/ansgar/code-signing/commit/3a07a08a82a8f545461c2604da3bd83175fd3a1c
|
| |
| |
| |
| |
| |
| |
| | |
It is no longer possible to run the "setup" rules without a compiler,
because Kconfig symbols can depend on compiler properties. Add a way
to invoke just the first step of setup, which merges the kconfig files
and overrides together.
|