This will be a little more efficient since unset CONFIG options are
stripped away from auto.conf, and we can hard-code the path to auto.conf
since it is never overridden.

include/config/kernel.release is generated before %pkg is run.
So, it is guaranteed auto.conf is up-to-date.
Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>

I think is_enabled() and if_enable_echo() in scripts/package/mkdebian
are useful.

builddeb also has many repetitive greps over the kernel config, so I
borrowed the idea to clean it up.
Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>

'$(MAKE) KBUILD_SRC=' changes the working directory back and forth
between objtree and srctree.

It is better to recurse to the top-level Makefile directly.
Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>

Commit 37c8a5fa ("kbuild: consolidate Devicetree dtb build rules")
moved the location of 'dtbs_install' target which caused dtbs to not be
installed when building debian package with 'bindeb-pkg' target. Update
the builddeb script to use the same logic that determines if there's a
'dtbs_install' target which is presence of the arch dts directory. Also,
use CONFIG_OF_EARLY_FLATTREE instead of CONFIG_OF as that's a better
indication of whether we are building dtbs.

This commit will also have the side effect of installing dtbs on any
arch that has dts files. Previously, it was dependent on whether the
arch defined 'dtbs_install'.

Fixes: 37c8a5fa ("kbuild: consolidate Devicetree dtb build rules")
Reported-by: NNuno Gonçalves <nunojpg@gmail.com>
Signed-off-by: NRob Herring <robh@kernel.org>
Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>

Ard Biesheuvel reports bindeb-pkg with O= option is broken in the
following way:

  ...
    LD [M]  sound/soc/rockchip/snd-soc-rk3399-gru-sound.ko
    LD [M]  sound/soc/rockchip/snd-soc-rockchip-pcm.ko
    LD [M]  sound/soc/rockchip/snd-soc-rockchip-rt5645.ko
    LD [M]  sound/soc/rockchip/snd-soc-rockchip-spdif.ko
    LD [M]  sound/soc/sh/rcar/snd-soc-rcar.ko
   fakeroot -u debian/rules binary
  make KERNELRELEASE=4.19.0-12677-g19beffaf7a99-dirty ARCH=arm64 KBUILD_SRC= intdeb-pkg
  /bin/bash /home/ard/linux/scripts/package/builddeb
  Makefile:600: include/config/auto.conf: No such file or directory
  ***
  *** Configuration file ".config" not found!
  ***
  *** Please run some configurator (e.g. "make oldconfig" or
  *** "make menuconfig" or "make xconfig").
  ***
  make[12]: *** [syncconfig] Error 1
  make[11]: *** [syncconfig] Error 2
  make[10]: *** [include/config/auto.conf] Error 2
  make[9]: *** [__sub-make] Error 2
  ...

Prior to commit 80463f1b ("kbuild: add --include-dir flag only
for out-of-tree build"), both srctree and objtree were added to
--include-dir redundantly, and the wrong code '$MAKE image_name'
was working by relying on that. Now, the potential issue that had
previously been hidden just showed up.

'$MAKE image_name' recurses to the generated $(objtree)/Makefile and
ends up with running in srctree, which is incorrect. It should be
invoked with '-f $srctree/Makefile' (or KBUILD_SRC=) to be executed
in objtree.

Fixes: 80463f1b ("kbuild: add --include-dir flag only for out-of-tree build")
Reported-by: NArd Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>
Tested-by: NArd Biesheuvel <ard.biesheuvel@linaro.org>

Move debian/ directory generation out of builddeb to a new script,
mkdebian. The package build commands are kept in builddeb, which
is now an internal command called from debian/rules.

With these changes in place, we can now use dpkg-buildpackage from
deb-pkg and bindeb-pkg removing need for handrolled source/changes
generation.

This patch is based on the criticism of the current state of builddeb
discussed on:

https://patchwork.kernel.org/patch/9656403/Signed-off-by: NRiku Voipio <riku.voipio@linaro.org>
Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>

Since d5d332d3, a couple of links in scripts/dtc/include-prefixes
are additionally required in order to build device trees with the header
package.
Signed-off-by: NJan Kiszka <jan.kiszka@siemens.com>
Reviewed-by: NRiku Voipio <riku.voipio@linaro.org>
Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>

At the end of "make bindeb-pkg" I noticed the following warning:

dpkg-genchanges: warning: unknown substitution variable ${kernel:debarch}

It turns out that since dpkg version 1.19.0 dpkg-genchanges honors
substitution variables in the Description field, while earlier
versions silently left them alone, see https://bugs.debian.org/856547.

The result is an incomplete description of the linux-headers package
in the generated .changes file.  Fix it by passing the kernel:debarch
substitution variable to dpkg-genchanges.
Signed-off-by: NSven Joachim <svenjoac@gmx.de>
Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>

Commit 5620a0d1 ("firmware: delete in-kernel firmware") deleted
in-kernel firmware support, including the firmware install command.

So, the firmware package does not make sense any more.  Remove it.
Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: NRiku Voipio <riku.voipio@linaro.org>
Acked-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Signed-off-by: NAndrew Donnellan <andrew.donnellan@au1.ibm.com>
Signed-off-by: NRiku Voipio <riku.voipio@linaro.org>
Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>

The builddeb script has some hardcoded references to Linux version 2.6
which is ancient. Drop Provides as the virtual packages provided are not
useful anymore. Leave the Provides for linux-kernel-headers, as someone
might still be referring to it.

While at it, updated copyright date and drop Standards-Version: since
the package isn't Debian Standards compliant anyways.

Cc: Timo Sigurdsson <public_timo.s@silentcreek.de>
Signed-off-by: NRiku Voipio <riku.voipio@linaro.org>
Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>

URLs to ftp.kernel.org are still exist though the service is closed [0].
This commit fixes the URLs to use www.kernel.org instead.

[0] https://www.kernel.org/shutting-down-ftp-services.htmlSigned-off-by: NSeongJae Park <sj38.park@gmail.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>

The arch Makefile are fixed to set KBUILD_IMAGE to the full patch, so
the workaround is no longer needed.
Signed-off-by: NMichal Marek <mmarek@suse.com>
Reviewed-by: NRiku Voipio <riku.voipio@linaro.org>
Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>

Both Debian and kernel archs are "arm64" but UTS_MACHINE and gcc say
"aarch64".  Recognizing just the latter should be enough but let's
accept both in case something regresses again or an user sets
UTS_MACHINE=arm64.

Regressed in cfa88c79: arm64: Set UTS_MACHINE in the Makefile.
Signed-off-by: NAdam Borowski <kilobyte@angband.pl>
Acked-by: NRiku Voipio <riku.voipio@linaro.org>
Signed-off-by: NMichal Marek <mmarek@suse.com>

When attempting to build a Debian kernel package, the "scripts/gcc-plugins"
directory does not exist in the output tree unless CONFIG_GCC_PLUGINS=y.
To avoid errors when not defined, this wraps the failing "find" in a config
test.
Reported-by: NFrank Paulsen <frobnic+lkml@gmail.com>
Tested-by: NChristian Kujau <lists@nerdbynature.de>
Signed-off-by: NKees Cook <keescook@chromium.org>
Signed-off-by: NMichal Marek <mmarek@suse.com>

When calling `make deb-pkg` on a system with no codename (for example
Arch Linux), lsb_release sometimes outputs `n/a` as the codename.

This breaks dpkg-parsechangelog, which can't process the changelog
correctly.
Signed-off-by: NMarcin Mielniczuk <marmistrz.dev@gmail.com>
Signed-off-by: NMichal Marek <mmarek@suse.com>

Builddep is not very explicit about file permissions. Actually the file
permissions in the package are largely influenced by the umask of the
user cloning the git and building the package. If that umask does not
set go+r the resulting linux-headers package will prevent non-root users
from building out-of-tree modules. And that is probably just one
unexpected effect.
Being a packaging/install tool builddep should make sure the file
permissions are set correctly and not just derived from a value that is
never checked.

This patch sets ugo read permissions for all packaged files and derives
the executable bit for directories and executables from the file-owner.
Signed-off-by: NHenning Schild <henning.schild@siemens.com>
Signed-off-by: NMichal Marek <mmarek@suse.com>

On May 4th, Bjørn Mork provided patch 697bbc7b ("builddeb: include
objtool binary in headers package"). However, that one only works if
$srctree=$objtree, because the objtool binaries are not written to the
srctree, but to the objtree.
Signed-off-by: NWilfried Klaebe <linux-kernel@lebenslange-mailadresse.de>
Fixes: 697bbc7b ("builddeb: include objtool binary in headers package")
Signed-off-by: NMichal Marek <mmarek@suse.com>

This patch allows to build the whole kernel with GCC plugins. It was ported from
grsecurity/PaX. The infrastructure supports building out-of-tree modules and
building in a separate directory. Cross-compilation is supported too.
Currently the x86, arm, arm64 and uml architectures enable plugins.

The directory of the gcc plugins is scripts/gcc-plugins. You can use a file or a directory
there. The plugins compile with these options:
 * -fno-rtti: gcc is compiled with this option so the plugins must use it too
 * -fno-exceptions: this is inherited from gcc too
 * -fasynchronous-unwind-tables: this is inherited from gcc too
 * -ggdb: it is useful for debugging a plugin (better backtrace on internal
    errors)
 * -Wno-narrowing: to suppress warnings from gcc headers (ipa-utils.h)
 * -Wno-unused-variable: to suppress warnings from gcc headers (gcc_version
    variable, plugin-version.h)

The infrastructure introduces a new Makefile target called gcc-plugins. It
supports all gcc versions from 4.5 to 6.0. The scripts/gcc-plugin.sh script
chooses the proper host compiler (gcc-4.7 can be built by either gcc or g++).
This script also checks the availability of the included headers in
scripts/gcc-plugins/gcc-common.h.

The gcc-common.h header contains frequently included headers for GCC plugins
and it has a compatibility layer for the supported gcc versions.

The gcc-generate-*-pass.h headers automatically generate the registration
structures for GIMPLE, SIMPLE_IPA, IPA and RTL passes.

Note that 'make clean' keeps the *.so files (only the distclean or mrproper
targets clean all) because they are needed for out-of-tree modules.

Based on work created by the PaX Team.
Signed-off-by: NEmese Revfy <re.emese@gmail.com>
Acked-by: NKees Cook <keescook@chromium.org>
Signed-off-by: NMichal Marek <mmarek@suse.com>

The kernel headers package (linux-headers) doesn't include
header files from other architectures required to build
out-of-tree modules.

For e.g. on ARM64, opcodes.h includes the same file from ARM
which causes the following error:
./arch/arm64/include/asm/opcodes.h:1:43: fatal error:
../../arm/include/asm/opcodes.h: No such file or directory
compilation terminated.
Signed-off-by: NAzriel Samson <asamson@codeaurora.org>
Signed-off-by: NMichal Marek <mmarek@suse.com>

"objtool" is required for building external m dules if "Compile-time
stack metadata validation" is enabled.  Otherwise all builds based
on the headers package fail with:

 make[1]: Entering directory '/usr/src/linux-headers-4.6.0-rc6'
 make[2]: *** No rule to make target 'tools/objtool/objtool', needed by 'foo.o'.  Stop.
 Makefile:1598: recipe for target 'foo.ko' failed
 make[1]: *** [foo.ko] Error 2
 make[1]: Leaving directory '/usr/src/linux-headers-4.6.0-rc6'
Signed-off-by: NBjørn Mork <bjorn@mork.no>
Signed-off-by: NMichal Marek <mmarek@suse.com>

The Debian armhf architecture uses the ARM EABI hard-float variant,
whereas armel uses the soft-float variant.  Although the kernel
doesn't use FP itself, CONFIG_VFP must be enabled to support
hard-float userland and will probably be disabled when supporting a
soft-float userland.  So set the architecture to armhf by default when
CONFIG_AEABI and CONFIG_VFP are both enabled.
Signed-off-by: NBen Hutchings <ben@decadent.org.uk>
Acked-by: NIan Campbell <ijc@hellion.org.uk>
Acked-by: NFathi Boudra <fathi.boudra@linaro.org>
Reviewed-by: NRiku Voipio <riku.voipio@linaro.org>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

Commit 3716001b ("deb-pkg: add source package") added the ability to
create a debian changelog file.  This exposed that previously the
builddeb script hasn't cleared debian/files between builds.

As debian/files keeps accumulating entries, the changes file will end up
growing indefinelty.  With outdated entries in debian/files, builddeb
script will exit with failure.  This regression impacts those who use
"make deb-pkg" target to build kernel into a .deb package and never use
"make mrproper" or other means to clean kernel tree from generated
directories.

To fix the regression, remove debian/files before starting build and in
the generated clean rule.

Fixes: 3716001b ("deb-pkg: add source package")
Signed-off-by: NRiku Voipio <riku.voipio@linaro.org>
Reported-by: NDoug Smythies <dsmythies@telus.net>
Tested-by: NDoug Smythies <dsmythies@telus.net>
Tested-by: NKalle Valo <kvalo@codeaurora.org>
Acked-by: NBen Hutchings <ben@decadent.org.uk>
Cc: Michal Marek <mmarek@suse.cz>
Cc: maximilian attems <maks@stro.at>
Cc: Chris J Arges <chris.j.arges@canonical.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Make deb-pkg build both source and binary package like make rpm-pkg does.
For people who only need binary kernel package, there is now bindeb-pkg
target, same target also used to build the .deb files if built from the
source package using dpkg-buildpackage.

Generated source package will build the same kernel .config than what
was available for make deb-pkg. The name of the source package can
be set with KDEB_SOURCENAME enviroment variable.

The source package is useful for GPL compliance, or for feeding to a
automated debian package builder.

Cc: Chris J Arges <chris.j.arges@canonical.com>
Reviewed-by: Nmaximilian attems <maks@stro.at>
Reviewed-by: NBen Hutchings <ben@decadent.org.uk>
Signed-off-by: NRiku Voipio <riku.voipio@linaro.org>
Signed-off-by: NMichal Marek <mmarek@suse.com>

I happened to build a kernel with "make deb-pkg" on a machine with no
network connectivity, but this failed with:

[...]
  INSTALL debian/headertmp/usr/include/asm/ (65 files)
hostname: Name or service not known
../scripts/package/Makefile:90: recipe for target 'deb-pkg' failed
make[2]: *** [deb-pkg] Error 1

In scripts/package/builddeb it tries to construct an email address (that
can be queried in /proc/version later on) but with no network,
the "hostname -f" fails. The following patch falls back to just use the
shortname if we cannot determine our FQDN.
Signed-off-by: NChristian Kujau <lists@nerdbynature.de>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

Every package needs /usr/share/doc/$package_name and
DEBIAN directory, so create them as part of create_package
function.
Signed-off-by: NRiku Voipio <riku.voipio@linaro.org>
Signed-off-by: NMichal Marek <mmarek@suse.com>

When building a package with make deb-pkg (say, for arm), the dtb files are
not added to the package. Given that things are still evolving on arm, it
make sense to have them along with the kernel and modules.
Signed-off-by: NArnaud Patard <arnaud.patard@rtp-net.org>
Reviewed-by: NBen Hutchings <ben@decadent.org.uk>
Acked-by: Nmaximilian attems <max@stro.at>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

If CONFIG_MODULE_SIG_ALL is set, then user expects that all modules are
automatically signed in the result package, as it's for rpm-pkg, binrpm-pkg,
tar, tar-*. For deb-pkg this is correct only if CONFIG_DEBUG_INFO
is NOT set. In that case deb-package contains signed modules.

But if CONFIG_DEBUG_INFO is set, builddeb creates separate package with
debug information. To do that, debug information from all modules
is copied into separate files by objcopy. And loadable kernel modules are
stripped afterwards. Stripping removes previously (during modules_install)
added signatures from loadable kernel modules. Therefore final deb-package
contains unsigned modules despite of set option CONFIG_MODULE_SIG_ALL.

This patch resigns all stripped modules if CONFIG_MODULE_SIG_ALL is set
to solve this problem.
Signed-off-by: NAndrey Skvortsov <andrej.skvortzov@gmail.com>
Acked-by: Nmaximilian attems <max@stro.at>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

create_package() function tries to resolve used architecture
for everry package. Split the setting the architecture to a
new function, set_debarch(), called once on startup.

This allows using debarch from other parts of script as
needed.

v2: Follow Michals suggestion on setting variables at
top scope and also setting the fallback $debarch in the
new function
Signed-off-by: NRiku Voipio <riku.voipio@linaro.org>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

lsb_release command is a good choice to determine the distribution
name for the changelog file in the generated Debian packages [1].
Its installation is no precondition.
In Debian it is still not essential or build-essential.

Ben gave some helpful informations and detailed explanations in [2].
There he also suggested to have an option to explicitly set the
distribution name (see $KDEB_CHANGELOG_DIST variable).

Embedded the improvement as suggested by Thorsten (see [3]):

"This is suboptimal: if KDEB_CHANGELOG_DIST is defined,
lsb_release is not necessary. The following snippet
also omits using its output if it fails but still
produces any:"

Dealing with this issue I learned about "The Colon in the Shell." and
possible pitfalls in this area (see [4,5]). Furthermore, refreshed my
knowledge about redirecting outputs with the echo command (see [5]).
Special thanks to Thorsten, I enjoyed the IRC session with you.

Cooked together the snippets of Ben and Thorsten (see [2,3]).
Tested against Linux v3.19-rc2.

Thanks goes to Alexander, Ben, maximilian and Thorsten for the very
vital help.

[1] https://lkml.org/lkml/2012/4/23/516
[2] http://marc.info/?l=linux-kbuild&m=142022188322321&w=2
[3] http://marc.info/?l=linux-kbuild&m=142023476825460&w=2
[4] http://blog.brlink.eu/index.html#i70
[5] https://www.mirbsd.org/permalinks/wlog-10_e20141209-tg.htm
[6] http://stackoverflow.com/questions/23489934/echo-2-some-text-what-does-it-mean-in-shell-scripting

CC: Alexander Wirt <formorer@debian.org>
Suggested-by: NBen Hutchings <ben@decadent.org.uk>
Suggested-by: NThorsten Glaser <tg@mirbsd.org>
Reviewed-by: NBen Hutchings <ben@decadent.org.uk>
Acked-by: Nmaximilian attems <max@stro.at>
[ dileks: Reviewed his suggested diff in RFC v4 ]
Reviewed-by: NThorsten Glaser <tg@debian.org>
Signed-off-by: NSedat Dilek <sedat.dilek@gmail.com>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

Happy new 2015!

I have combined two patches which I had already sent to linux-kbuild ML.

Today, I prefer "builddeb" as a label for such patches.

[1] http://marc.info/?l=linux-kbuild&m=133521955904706
[2] http://marc.info/?l=linux-kbuild&m=133521955004705

CC: Ben Hutchings <ben@decadent.org.uk>
CC: maximilian attems <max@stro.at>
Signed-off-by: NSedat Dilek <sedat.dilek@gmail.com>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

The -isp option has been deprecated, after it became the default
behaviour back in 2006.

Since dpkg 1.17.11, dpkg-gencontrol emits a warning on -isp usage.

References: https://bugs.debian.org/215233Signed-off-by: NAsbjoern Sloth Toennesen <asbjorn@asbjorn.biz>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

The Debian powerpc little endian architecture is called ppc64el.  This
is the default architecture used by Ubuntu for powerpc.

The below checks the kernel config to see if we are compiling little
endian and sets the Debian arch appropriately.
Signed-off-by: NMichael Neuling <mikey@neuling.org>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

Since the conversion of objtree to use relative pathnames (commit
7e1c0477, "kbuild: Use relative path for $(objtree)"), the debug
info files have been ending up in /debian/dbgtmp/ in the regular
linux-image package instead of the debug files package. Fix up the
paths so that the debug files end up in the -dbg package.

This is based on a similar patch by Darrick.
Reported-and-tested-by: N"Darrick J. Wong" <darrick.wong@oracle.com>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

The Makefiles call the respective interpreter explicitly, but this makes
it easier to use the scripts manually.
Signed-off-by: NMichal Marek <mmarek@suse.cz>

The kernel headers package (linux-headers) doesn't include several
header files required to build out-of-tree modules.

It makes the package unusable on e.g. ARM architecture:
 /usr/src/linux-headers-3.14.0/arch/arm/include/asm/memory.h:24:25:
 fatal error: mach/memory.h: No such file or directory
 #include <mach/memory.h>
 ^
 compilation terminated.
Signed-off-by: NFathi Boudra <fathi.boudra@linaro.org>
Reviewed-by: NBen Hutchings <ben@decadent.org.uk>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

When $srctree or $objtree are relative paths, we cannot change directory
and refer to them in the same subshell. Do the redirection outside of
the subshell to fix this.
Reported-and-tested-by: NIlya Dryomov <idryomov@gmail.com>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

The Debian s390x architecture has 64-bit userland whereas s390 has
32-bit userland.  A 64-bit kernel can be used with either.  Now that
Debian supports multiarch and officially supports s390x, it makes more
sense to assign a 64-bit kernel package to s390x.
Reported-by: NStephen Powell <zlinuxman@wowway.com>
References: https://bugs.debian.org/750925Signed-off-by: NBen Hutchings <ben@decadent.org.uk>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

Signed-off-by: NMasahiro Yamada <yamada.m@jp.panasonic.com>
Signed-off-by: NMichal Marek <mmarek@suse.cz>

Signed-off-by: NFathi Boudra <fathi.boudra@linaro.org>
Reviewed-by: NBen Hutchings <ben@decadent.org.uk>
Signed-off-by: NMichal Marek <mmarek@suse.cz>