The upstream Makefile uses DEBUG to control both stripping of debug
information *and* disabling optimisations.
We want to disable stripping of debug information so that (normally)
dh_strip can move it into a separate package later. However we also
want to enable optimisations by default.
Therefore:
- Set the STRIP variable (a command prefix) to "true"
- Set the DEBUG variable (a pseudo-boolean) to "true" if and only if
the "noopt" build option is used
These two are terse by default and don't respect the KBUILD_VERBOSE
environment variable.
(I considered adding V=1 to the MAKEFLAGS we use by default in
debian/rules.real, but this might have different semantics for the few
tools that aren't built using either kbuild or automake.)
This adds support for PREEMPT_RT by adding two new featureset: none and
rt. For now the rt featureset supports armmp only, support for lpae
might be added later. the none featureset supports both flavours armmp
and armmp-lpae.
-----BEGIN PGP SIGNATURE-----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=JUfi
-----END PGP SIGNATURE-----
Merge tag 'debian/4.17.6-2'
Release linux (4.17.6-2).
- Drop the ABI maintenance patch
- Resolve conflict between changes to config file generation in sid
(delete CONFIG_BUILD_SALT) and master (putting them in
arch-dependent packages)
Closes: #872263
- kbuild: Add build salt to the kernel and modules
- [arm64,powerpc,x86] Add build salt to the vDSO
- Set BUILD_SALT equal to the release string
The previous fix ensured that objtool was built for the right host
architecture, but it was still including some UAPI header files
(in particular <asm/bitsperlong.h>) for the build architecture.
Currently we replicate the mapping of Debian architectures to kernel
source architectures here. Use the KERNEL_ARCH variable that is now
passed down by rules.real. We still need to check whether perf is
supported on the architecture, though.
-----BEGIN PGP SIGNATURE-----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=VAyl
-----END PGP SIGNATURE-----
Merge tag 'debian/4.17.6-1'
Release linux (4.17.6-1).
- Drop patches for ABI maintenance or that are already upstream
- Drop ABI reference files
- Introduce linux-bootwrapper-<abiname> package containing boot wrapper
tools for the host architecture
- linux-image: Install symlinks to boot wrapper tools instead of the
native tools built by kbuild
The real problem was not that there were two rules, but that the first
rule didn't work for out-of-tree builds.
After I disabled the second rule, "make man" still succeeded because
of another rule (with no commands) that made all man pages depend on
asciidoc.conf.
This updates the debian changelog for listing changes of this stable
update. It also removes the patches that have been merged upstream.
Signed-off-by: Romain Perier <romain.perier@gmail.com>
objtool is normally built and used during the kernel build process,
in which case it should always be built as a native executable.
But when we build it for linux-tools it should match the target
architecture. We need to override the CC and LD variables to
make this work.
For parallel display outputs like RGB panels on i.MX device
DRM_IMX_PARALLEL_DISPLAY needs to be enabled
Signed-off-by: Sjoerd Simons <sjoerd.simons@collabora.co.uk>
In Linux 4.18, various compiler version and feature tests are invoked
via kconfig rather than via kbuild. This means that we generally
cannot generate kconfig files for foreign architectures.
Move the config files to a new linux-config-<version> package which is
arch-dependent (and also M-A: same).
Make linux-config-<version> and linux-source-<version> recommend each
other.
We generate all config files when building linux-source-<version>,
and this requires running kconfig. kconfig is now built using
bison and flex instead of requiring their output to be already
included in the source tree.
- Various config symbols were removed, renamed or split
- HOTPLUG_PCI_SHPC is now boolean, so set it to built-in
- The stack protector config symbols were changed to two booleans
with different names
The Version class already parses the version into upstream and Debian
revision parts. Match the results of that against two separate
regexps rather than matching the whole version string and potentially
parsing it inconsistently.
As a result we no longer have to explicitly exclude '-' in the
revision_other group.
AM35x use EDMA for McASP audio SoC, resulting in a "missing
SND_EDMA_SOC" message during probe. This enables the support.
Signed-off-by: Romain Perier <romain.perier@gmail.com>
Closes: #898087
We always used to match arbitrary revision suffixes, but in commit
f4dbf1aa82 ".../debian.py: Fix binNMU revision parsing" I excluded
'+' from the revision_other to ensure that it wouldn't match a
binNMU suffix (which is matched separately).
Stop excluding '+' from the revision_other group, and change the
repetition to non-greedy so that any binNMU suffix will still match
the following group.
We didn't check for invalid characters in the upstream or revision
parts. Tighten the regexps for those parts.
Also, failure to match the epoch or revision groups caused
those parts to be treated as part of the upstream version.
Split out the three parts and match them against separate
regexps.
binder.c and binder_alloc.c both define a debug_mask module parameter,
which was fine when they were two separate built-in "modules". Now
that they're grouped together we need to distinguish these parameters,
so rename the one in binder_alloc.c to alloc_debug_mask.
We no longer need to add the crypto-aes or crypto-ecb dependencies
because ext4 uses the common encryption code in fscrypto (and has a
regular symbol dependency on it).
Since upstream added a softdep on "crc32c", we don't actually need to
change ext4 at all now. But let's replace it with "crypto-crc32c",
since that's the module alias the crypto subsystem will actually
request and is consistent with the softdep we add to other
filesystems.
Remove maintainer stanza for the now upstreamed CVE-2018-11506 (included
in 4.16.13) and remove maintainer trailer for the new upstream stable
update import.
I've found Fedora28 tries to improve battery life on laptop.
See https://docs.fedoraproject.org/f28/release-notes/sysadmin/Kernel.html#sect-kernel-battery
There are three points.
1. A new SATA link-powermanagement-policy has been written which mirrors
Windows defaults: med_power_with_dipm, this has been merged for kernel
4.15, as part of this change this new policy will be the default on all
Intel mobile chipsets. This saves aprox. 1.0 - 1.5 Watts of power on an
idle laptop.
2. Enable Intel HDA codec power-saving by default with a 1 second timeout.
This saves aprox. 0.4 Watts of power on an idle laptop.
3. Enable USB autosuspend for USB bluetooth receivers by default. If all
other USB devices on the laptop also have USB auto-suspend enabled
(which typically is true) this saves aprox. 0.4 Watts of power on an idle
laptop.
Add a patch to disable uImage generation to avoid depend on u-boot-tools
Fix typo the EL's flavor names in installer: not same within defines
Malta is never used for r6. (Closes: #898523)
Boston also requires relocation table size >= 0x00121000
In machine-readable format, "or" already mean dual licensed.
Fixes harmless but annoying Lintian warning:
space-in-std-shortname-in-dep5-copyright x11 dual license
Refresh bugfix/all/fs-add-module_softdep-declarations-for-hard-coded-cr.patch
Drop patches applied upstream related to CVE-2018-1093
Cleanup debian/changelog entries
The HiFi2 driver was split up into ACPI and PCI parts, and we need to
select one or both of those instead of SND_SST_ATOM_HIFI2_PLATFORM.
Baed on the help text, it appears that the ACPI part is all we want.
It also gets selected automatically, so this commit doesn't change
the resulting config.
If multiple options in a "choice" group are enabled, kconfig uses the
last (and warns about this). This means that the global default of
CONFIG_PREEMPT_VOLUNTARY=y has always overridden the setting of
CONFIG_PREEMPT_NONE=y here.
Since no-one seems to have complained, assume that the default is OK
and remove this.
Add a new "pkg.linux.nosource" to let users disable building the
linux-source-* package, and allow to set "source: false" to modify
the default behaviour when no rofile is used.
When doing development builds this can save up to 15 minutes of build
time, especially on IO-strapped build workers.
Fully disable building and installing any documentation when the nodoc
build-profile is used.
Among other things this will help reducing build times when doing
development builds, especially on IO-limited build workers.
We don't want to include "-4.9" in them twice. Add a "source_basename"
template variable that excludes any version suffix in the source package
name.
(cherry picked from commit f3c51efdd6e9d0ce32ee5a0f998fdcda930a715c)
For master, nothing is immediately broken without this. Also we have
no longer build a linux-manual package. Change the changelog text
accordingly.
We already had support for disabling the tools build, used by
src:linux-grsec. However in this case, where we're using a different
based version to src:linux, we do still need to build the versioned
tools packages (linux-kbuild-4.9 and linux-perf-4.9). Split the
control template, config setting and rules accordingly.
(cherry picked from commit cb62c945f27ddee476631fa85c6aa67e50ed3bee)
libelf-dev is needed to build objtool when CONFIG_STACK_VALIDATION
and CONFIG_UNWINDER_ORC are enabled, which is currently the case.
Do not mark it as !nopython !pkg.linux.notools otherwise the build
will fail in a minimal chroot (pbuilder/etc).
If building with CONFIG_MODULE_SIG_ALL and CONFIG_DEBUG_INFO the
objcopy call that adds the debuglink has the side-effect of removing
the signature added to the kernel module. Let's explicitly sign the
installed modules again in that case.
Closes: #852715
In VersionLinux, any binNMU version will currently match the
revision_other group in the regexp and therefore never be recognised
as an experimental, security, or backport version. This is probably
harmless in practice because:
- binNMUs don't happen in those suites
- Only debian/bin/gencontrol.py cares about the linux_revision_other
attribute, and it won't be run for a binNMU version
But let's fix it by matching the binNMU suffix separately.
I incorrectly added a requirement of a newline after urgency in the
changelog entry top line as part of
"debian/lib/python/debian_linux/debian.py: Parse bottom lines of
changelog entries". For a binNMU, there will be ", binary-only=yes"
after the urgency. Since we don't currently care about any fields
after urgency, allow either a comma or newline.
- Various ancient SCSI drivers were removed
- BT_HCIBTUART and INFINIBAND_CXGB3_DEBUG were removed
- OMAP_DM_TIMER is now an automatic symbol
- Marvell NAND driver was rewritten, so we enable MTD_NAND_MARVELL
instead of MTD_NAND_PXA3xx
- Various netfilter symbols are now boolean instead of tristate