Pull kbuild updates from Michal Marek:
- fix for make headers_install argv explosion with too long path
- scripts/setlocalversion does not call git update-index needlessly
- fix for the src.rpm produced by make rpm-pkg. The new make
image_name can be useful also for other packaging tools.
- scripts/mod/devicetable-offsets.o is not rebuilt during each make run
- make modules_install dependency fix
- scripts/sortextable portability fix
- fix for kbuild to generate the output directory for all object files
in subdirs.
- a couple of minor fixes
* 'kbuild' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild:
kbuild: create directory for dir/file.o
tools/include: use stdint types for user-space byteshift headers
Makefile: Fix install error with make -j option
Fix a build warning in scripts/mod/file2alias.c
improve modalias building
scripts/mod: Spelling s/DEVICEVTABLE/DEVICETABLE/
kbuild: fix error when building from src rpm
scripts/setlocalversion on write-protected source tree
Makefile.lib: align DTB quiet_cmd
kbuild: fix make headers_install when path is too long
The following issue can be reproduced with Linus' tree on
an x86_64 server.
>+ cp /home/user/rpmbuild-test/BUILDROOT/kernel-3.9.2.x86_64/boot/vmlinuz-3.9.2
>cp: missing destination file operand after
>/home/user/rpmbuild-test/BUILDROOT/kernel-3.9.2-1.x86_64/boot/vmlinuz-3.9.2'
>Try `cp --help' for more information.
>error: Bad exit status from /var/tmp/rpm-tmp.R4o0iI (%install)
Here are the commands to reproduce:
make defconfig
make rpm-pkg
Use the resulting src rpm to build as follows:
mkdir ~/rpmbuild-test
cd ~/rpmbuild-test
rpmbuild --rebuild --define "_topdir `pwd`" -vv ~/rpmbuild/SRPMS/kernel-3.10.0_rc1+-1.src.rpm
The issue is because the %install script uses $KBUILD_IMAGE and it hasn't
been set since it is only available in the kbuild system and not in the
%install script.
This patch adds a Makefile target to emit the image_name that can be used
and modifies the mkspec to use the dynamic name in %install.
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
Commit 6501320311 dropped the rpm spec as a
prerequisite for the binrpm-pkg target but forgot to update $< usage,
which causes the rule to break.
This commit fixes that by replacing $< with the spec name.
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
The *pkg targets are always run, so it makes no sense to cache the
generated specfile. This also fixes build errors when the specfile
becomes out of date, without the Makefile noticing it:
$ make rpm
works
$ echo yadadada >localversion-test
$ make rpm
fails, because kernel.spec assumes the old kernel release string
Signed-off-by: Michal Marek <mmarek@suse.cz>
Now that we only package explicitly listed files to the tarball, there
is no need to abuse the parent directory.
Signed-off-by: Michal Marek <mmarek@suse.cz>
The source tree can contain lots of uninteresting data like tag or
cscope files, packaging which slows down make rpm needlessly. It can
also break the build, if the tree contains an unrelated file named
*.spec. The downside of this change is that new subdirectories have to
be added to the KBUILD_ALLDIRS variable in the top-level Makefile. The
upside is that the behavior is more predictable.
Signed-off-by: Michal Marek <mmarek@suse.cz>
When dealing with multiple sub-arches (like 32- and 64-bit on x86, for
example) generating a bunch of kernel tar archives with the same name
but for different sub-arches could get confusing and error-prone. Also,
the build process could overwrite otherwise unrelated builds and you
probably don't want that. So, add the architecture to the archive name
for more clarity and less shoot-yourself-in-the-foot practices.
Signed-off-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Michal Marek <mmarek@suse.cz>
KBUILD_OUTPUT is always empty here, so it is useless to test it. But
while use O=.., objtree and srctree will be different. I compare them
instead.
Signed-off-by: Bin Wang <wbin00@gmail.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
'make rpm-pkg' and 'make binrpm-pkg' fail when the kernel source is
read-only. Specifically, when the RPM spec generated by
scripts/package/mkspec is run, KBUILD_SRC happens to be set to the
source location and thus the invocation of 'make headers_install'
fails when an internal call to 'filechk' tries to write a file into
the source tree.
The fix is to clear KBUILD_SRC for the 'make headers_install'
invocation in the spec file, as is already done for the 'make
modules_install' invocation.
Signed-off-by: David R. Bild <drbild@umich.edu>
Signed-off-by: Michal Marek <mmarek@suse.cz>
Pull kbuild changes from Michal Marek:
"The main part of kbuild for v3.7 contains:
- Fix for scripts/Makefile.modpost to not choke on a '.ko' substring
in the build directory path
- Two warning fixes (modpost and main Makefile)
- __compiletime_error works also with gcc 4.3
- make tar{gz,bz2,xz}-pkg uses default compression settings instead
of saving as many bytes as possible (this should actually be in the
misc branch, I don't know why I applied it here)."
* 'kbuild' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild:
compiler-gcc4.h: correct verion check for __compiletime_error
modpost: Permit .GCC.command.line sections
Kbuild: use normal compression settings for tar*-pkg
scripts/Makefile.modpost: error in finding modules from .mod files.
kbuild: Remove useless warning while appending KCFLAGS
For large kernel configurations (like a distribution kernel)
targz-pkg takes a quite long time to just do the compression.
I clocked it at 15+mins for a SUSE kernel like config on a fast
system. And tarxz and bzip2 are even slower.
The main reason is that the script that is doing the taring sets
the highest compression level (-9). When I change it to just
use the defaults the gzip time for the same kernel goes down
to ~3 mins. I haven't tested xz and bzip, but I expect those
to be much faster too.
I'm not willing to wait that long for a small compression
gain. So just change the script to use the defaults.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Reviewed-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
There were reports of users destroying their Fedora installs by a kernel
tarball that replaces the /lib -> /usr/lib symlink. Let's remove the
toplevel directories from the tarball to prevent this from happening.
Reported-by: Andi Kleen <andi@firstfloor.org>
Suggested-by: Ben Hutchings <ben@decadent.org.uk>
Cc: <stable@kernel.org>
Signed-off-by: Michal Marek <mmarek@suse.cz>
I discovered that make deb-pkg does not add Makefile_32.cpu from
arch/x86 directory when doing i386 kernel build and package build.
Fix it by greedily adding all Makefiles.
Reported-by: Witold Baryluk <baryluk@smp.if.uj.edu.pl>
Signed-off-by: maximilian attems <max@stro.at>
Signed-off-by: Michal Marek <mmarek@suse.cz>
That way they don't file conflict with official firmware package:
trying to overwrite '/lib/firmware/qlogic/1040.bin', which is
also in package firmware-qlogic 0.35
..
Reported-by: Michael Prokop <mika@debian.org>
Signed-off-by: maximilian attems <max@stro.at>
Signed-off-by: Michal Marek <mmarek@suse.cz>
After 303395ac3b, some headers are
autogenerated. Include these autogenerated headers (mainly
unistd_32_ia32.h) in out-of-tree builds to allow DKMS modules to be
built succesfully.
Signed-off-by: Peter Lekensteyn <lekensteyn@gmail.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
The out-of-tree build is broken in 'make deb-pkg'. The
header checks and the header install works on the source and
not on the object tree.
While fixing this also replace the direct 'make' invocations
with the $MAKE variable to be consistent within the script.
Cc: maximilian attems <max@stro.at>
Cc: Ben Hutchings <ben@decadent.org.uk>
Cc: debian-kernel@lists.debian.org
Cc: linux-kbuild@vger.kernel.org
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
The Link to the kernel header files in the debian packages
point to the original build directory. This is a bad choice
if the packages were installed on a different machine. Fix
this in by manually re-creating the link in the builddeb
script.
Cc: maximilian attems <max@stro.at>
Cc: debian-kernel@lists.debian.org
Cc: linux-kbuild@vger.kernel.org
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Reviewed-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Michal Marek <mmarek@suse.cz>
The current use of /tmp for file lists is insecure. Put them under
$objtree/debian instead.
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Cc: stable@vger.kernel.org # 2.6.39+
Acked-by: maximilian attems <max@stro.at>
Signed-off-by: Michal Marek <mmarek@suse.cz>
Specify --git-dir when building perf targets to allow out-of-tree
builds using O=<build-dir>.
The shell command in `git archive' had to be modified to allow proper
file name expansion of the files listed in MANIFEST.
Signed-off-by: matt mooney <mfm@muteddisk.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
* 'packaging' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild-2.6:
kbuild: Create a kernel-headers RPM
rpm-pkg: Fix when current directory is a symlink
Replace '-' in kernel version with '_'
To compile binaries which depend on new kernel interfaces, we need a
kernel-headers RPM
Signed-off-by: Arun Sharma <asharma@fb.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
The better fix would be to stop using the parent directory (principle of
least surprise), but as long as we use it, use it consistently.
Signed-off-by: Michal Marek <mmarek@suse.cz>
