Add option for stripping modules while installing them.
This function adds support for stripping modules while they are being
installed. CONFIG_DEBUG_KERNEL (which will probably become more
popular as developers use kdump) causes the size of the installed
modules to grow by a factor of 9 or so.
Some kernel package systems solve this problem by stripping the debug
information from /lib/modules after running "make modules_install",
but that may not work for people who are installing directly into
/lib/modules --- root partitions that were sized to handle 16 megs
worth of modules may not be quite so happy with 145 megs of modules,
so the "make modules_install" never succeeds.
This patch allows such users to request modules_install to strip the
modules as they are installed.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Based on the 'headers_install' target, this performs a basic sanity check
on the exported headers -- so far only checking that they do not include
any other headers which aren't selected for import, but easily extendable.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
This adds a make target which exports a subset of headers which contain
definitions which are useful for system libraries and tools. It uses the
BSD 'unifdef' tool to remove instances of #ifdef __KERNEL__, and uses
sed to remove markers like __user.
Based on an original implementation by Arnd Bergmann <arnd@arndb.de>
Hacked about by David Woodhouse <dwmw2@infradead.org>
Reviewed and cleaned up by Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
kbuild does explicitly specify what to do in all cases, and each
time make's built-in rules & variables has been used it has been a bug.
So to speed up things and to avoid the hard-to-debug error situations
ignore the built-in definitions.
If any part of the kernel uses the built-in definitions the build will
just stop there and it should be trivial to fix.
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
The binrpm-pkg target uses KERNELRELEASE when generated its .spec file.
When binrpm-pkg was the first build target run in a tree it generated the
.spec before kernel.release so the Version: tag in the .spec was empty.
I don't know if this is the best fix, but binrpm-pkg works when we
explicitly build kernel.release before descending into package-dir.
Signed-off-by: Zach Brown <zach.brown@oracle.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
If you set KCONFIG_OVERWRITECONFIG in environment, Kconfig will not break
symlinks when .config is a symlink to somewhere else.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
This moves the .kernelrelease file into include/config directory. Remove its
generation from the config step, if the config step doesn't leave a proper
.config behind, it triggers a call to silentoldconfig. Instead its generation
can be done via proper dependencies.
Signed-off-by: Roman Zippel <zippel@linux-m68k.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Now that kconfig can load multiple configurations, it becomes simple to
integrate the split config step, by simply comparing the new .config file with
the old auto.conf (and then saving the new auto.conf). A nice side effect is
that this saves a bit of disk space and cache, as no data needs to be read
from or saved into the splitted config files anymore (e.g. include/config is
now 648KB instead of 5.2MB).
Signed-off-by: Roman Zippel <zippel@linux-m68k.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
This fixes one of the worst kbuild warts left - the broken dependencies used
to check and regenerate the .config file. This was done via an indirect
dependency and the .config itself had an empty command, which can cause make
not to reread the changed .config file.
Instead of this we generate now a new file include/config/auto.conf from
.config, which is used for kbuild and has the proper dependencies. It's also
the main make target now for all files generated during this step (and thus
replaces include/linux/autoconf.h).
This also means we can now relax the syntax requirements for the .config file
and we don't have to rewrite it all the time, i.e. silentoldconfig only
writes .config now when it's necessary to keep it in sync with the Kconfig
files and even this can be suppressed by setting the environment variable
KCONFIG_NOSILENTUPDATE, so the update can (and must) be done manually.
Signed-off-by: Roman Zippel <zippel@linux-m68k.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Change the conditional of the outputmakefile rule to be evaluated entirely
in make, and add a conditional to not touch the generated makefile when e.g.
running 'make install' as root while the build was done as non-root. Also
adjust the comment describing this, and move the message printing and
redirection to mkmakefile.
Signed-off-by: Jan Beulich <jbeulich@novell.com>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Remove *.mod files but not .tmp_versions for external builds
When "make install" is run as root, .tmp_versions is re-created and
becomes owned by root. Subsequent "make" run by user fails because
.tmp_versions cannot be removed.
Signed-off-by: Pavel Roskin <proski@gnu.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
When installing external modules with `make modules_install', the
first thing that happens is a rm -rf of the target directory. This
works only once, and breaks when installing more than one (set of)
external module(s).
With following fix we have the functionality:
- for a in-kernel modules_install the $(MODLIB)/kernel directory will be
deleted before module installation
- for external modules the existing modules will be left as is assuming
one may be building and installign several external modules
Signed-off-by: Andreas Gruenbacher <agruen@suse.de>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
This fixes single targets build so it now works relaiably in
following cases:
- build with mixed kernel source and output files (make single-target)
- build with separate output directory (make O=.. single-target)
- external module with mixed kernel source and output files
(make M='pwd' single-target)
- external module with separate kernel source and output files
(make O=.. M='pwd' single-target)
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
gcc should handle this anyways, and it causes problems when
sprintf is turned into strcpy by gcc behind our backs and
the C fallback version of strcpy is actually defining __builtin_strcpy
Then drop -ffreestanding from the main Makefile because it isn't
needed anymore and implies -fno-builtin, which is wrong now.
(it was only added for x86-64, so dropping it should be safe)
Noticed by Roman Zippel
Cc: Roman Zippel <zippel@linux-m68k.org>
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
As a foundation for reliable stack unwinding, this adds a config option
(available to all architectures except IA64 and those where the module
loader might have problems with the resulting relocations) to enable the
generation of frame unwind information.
Signed-off-by: Jan Beulich <jbeulich@novell.com>
Cc: Miles Bader <uclinux-v850@lsi.nec.co.jp>
Cc: "Luck, Tony" <tony.luck@intel.com>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: Kyle McMartin <kyle@mcmartin.ca>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Paul Mundt <lethal@linux-sh.org>,
Cc: Andi Kleen <ak@muc.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The checks performed by scripts/reference_* has been moved to modpost.
Remove the files and their reference in top-level Makefile.
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
FORCE was not defined => error.
Use kbuild infrastructure to call down to the relevant
Makefile. This enables us to use the FORCE definition from kbuild.
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
I don't see any use case for the CONFIG_CC_ALIGN_* options:
- they are only available if EMBEDDED
- people using EMBEDDED will most likely also enable
CC_OPTIMIZE_FOR_SIZE
- the default for -Os is to disable alignment
In case someone is doing performance comparisons and discovers that the
default settings gcc chooses aren't good, the only sane thing is to discuss
whether it makes sense to change this, not through offering options to change
this locally.
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Rebuilding a previously built tree while using make's -j option from
time to time results in the version.h check running at the same time as
the updating of .kernelrelease, resulting in UTS_RELEASE remaining an
empty string (and as a side effect causing the entire kernel to be
rebuilt).
Signed-Off-By: Jan Beulich <jbeulich@novell.com>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
.PHONY: does not take patterns so use FORCE to achive same effect.
Thanks to "Paul D. Smith" <psmith@gnu.org> for noticing this.
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
The kbuild system takes advantage of an incorrect behavior in GNU make.
Once this behavior is fixed, all files in the kernel rebuild every time,
even if nothing has changed. This patch ensures kbuild works with both
the incorrect and correct behaviors of GNU make.
For more details on the incorrect behavior, see:
http://lists.gnu.org/archive/html/bug-make/2006-03/msg00003.html
Changes in this patch:
- Keep all targets that are to be marked .PHONY in a variable, PHONY.
- Add .PHONY: $(PHONY) to mark them properly.
- Remove any $(PHONY) files from the $? list when determining whether
targets are up-to-date or not.
Signed-off-by: Paul Smith <psmith@gnu.org>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Added a dependency so we do full preparation before trying to build single
file targets. This fixes a case where Andrew Morton did:
make kernel/sched.o
rm include/asm
make kernel/sched.o -> splat
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Rebuilding a previously built tree while using make's -j options from time to
time results in the version.h check running at the same time as the updating
of .kernelrelease, resulting in UTS_RELEASE remaining an empty string (and as
a side effect causing the entire kernel to be rebuilt).
Signed-Off-By: Jan Beulich <jbeulich@novell.com>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Workaround a cscope bug where a trailing ':' in VPATH makes it segfault
and let it build the cross-reference succesfully.
VPATH=/home/mattia/devel/kernel/git/linux-2.6: cscope -b
[1] 17555 segmentation fault VPATH=/home/mattia/devel/kernel/git/linux-2.6: cscope -b
Signed-off-by: Mattia Dongili <malattia@linux.it>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Move $(CC) support functions to Kbuild.include so they are available
in the kbuild files.
In addition the following was done:
o as-option documented in Documentation/kbuild/makefiles.txt
o Moved documentation to new section to match
new scope of functions
o added cc-ifversion used to conditionally select a text string
dependent on actual $(CC) version
o documented cc-ifversion
o change so Kbuild.include is read before the kbuild file
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
To avoid stale modules located in $(MODVERDIR) aka .tmp_versions/
always delete the directory when building an external module.
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
uname -m on MIPS can give a number of results, such as mips64. We
need to add another substitution to the sed call for SUBARCH in the
main Makefile.
Signed-off-by: Martin Michlmayr <tbm@cyrius.com>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Support building individual files when dealing with separate modules.
So say you have a module named "foo" which consist of two .o files bar.o
and fun.o.
You can then do:
make -C $KERNELSRC M=`pwd` bar.o
make -C $KERNELSRC M=`pwd` bar.lst
make -C $KERNELSRC M=`pwd` bar.i
make -C $KERNELSRC M=`pwd` / <= will build all .o files
and link foo.o
make -C $KERNELSRC M=`pwd` foo.ko <= will build the module
and do the modpost step
to create foo.ko
The above will also work if the external module is placed in a
subdirectory using a hirachy of kbuild files.
Thanks to Andreas Gruenbacher <agruen@suse.de> for initial feature
request / bug report.
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Commit 296e0855b0f9a4ec9be17106ac541745a55b2ce1:
"kbuild: fix make -jN with multiple targets with O=..."
causes a ~95% increase in build time for the kernel. Before: 4m21s
after: 8m1.403s. Can we revert this until another approach is found?
Cc: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
.kernelrelease was saved in same directory as kernel source also
with make O=...
Make sure we kick in the normal logic to shift to the output directory
when we build .kernelrelease after executing *config.
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
---
Currently the CPU subtype options are cluttering up arch/sh/Kconfig somewhat.
Given that, this moves all of that in to its own arch/sh/mm/Kconfig. Things
like cache configuration are also moved to this new location.
This also adds support for strict CPU tuning on newer cores, which requires
the addition of as-option.
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The command 'make all modules_install install' would fail
in a virgin tree - pointing at a non-existing directory under
/lib/modules/xxx
KERNELRELEASE is part of MODLIB and we need to create .kernelrelease
before we can properly evaluate KERNELRELEASE,
Changing MODLIB to the recursively expanded flavor let it pick up
the correct KERNELRELEASE value.
Reported by: "Hemmann, Volker Armin" <volker.armin.hemmann@tu-clausthal.de>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Theodore Ts'o