Re: [PATCH 19/23] kbuild: support building external modules in a separate build directory

From: Nicolas Schier
Date: Fri Oct 04 2024 - 16:15:18 EST


On Thu, Oct 03, 2024 at 09:47:35PM +0200, Nicolas Schier wrote:
> On Tue, Sep 17, 2024 at 11:16:47PM +0900, Masahiro Yamada wrote:
> > There has been a long-standing request to support building external
> > modules in a separate build directory.
> >
> > This commit introduces a new environment variable, KBUILD_EXTMOD_OUTPUT,
> > and its shorthand Make variable, MO.
> >
> > A simple usage:
> >
> > $ make -C <kernel-dir> M=<module-src-dir> MO=<module-build-dir>
> >
> > Signed-off-by: Masahiro Yamada <masahiroy@xxxxxxxxxx>
> > ---
> >
> > Documentation/kbuild/kbuild.rst | 8 +++++-
> > Documentation/kbuild/modules.rst | 5 +++-
> > Makefile | 44 +++++++++++++++++++++++---------
> > 3 files changed, 43 insertions(+), 14 deletions(-)
> >
> > diff --git a/Documentation/kbuild/kbuild.rst b/Documentation/kbuild/kbuild.rst
> > index 716f6fb70829..66a9dc44ea28 100644
> > --- a/Documentation/kbuild/kbuild.rst
> > +++ b/Documentation/kbuild/kbuild.rst
> > @@ -132,12 +132,18 @@ Specify the output directory when building the kernel.
> > This variable can also be used to point to the kernel output directory when
> > building external modules using kernel build artifacts in a separate build
> > directory. Please note that this does NOT specify the output directory for the
> > -external modules themselves.
> > +external modules themselves. (Use KBUILD_EXTMOD_OUTPUT for that purpose.)
> >
> > The output directory can also be specified using "O=...".
> >
> > Setting "O=..." takes precedence over KBUILD_OUTPUT.
> >
> > +KBUILD_EXTMOD_OUTPUT
> > +--------------------
> > +Specify the output directory for external modules.
> > +
> > +Setting "MO=..." takes precedence over KBUILD_EXTMOD_OUTPUT.
> > +
> > KBUILD_EXTRA_WARN
> > -----------------
> > Specify the extra build checks. The same value can be assigned by passing
> > diff --git a/Documentation/kbuild/modules.rst b/Documentation/kbuild/modules.rst
> > index 3a6e7bdc0889..03347e13eeb5 100644
> > --- a/Documentation/kbuild/modules.rst
> > +++ b/Documentation/kbuild/modules.rst
> > @@ -95,7 +95,7 @@ executed to make module versioning work.
> > of the kernel output directory if the kernel was built in a separate
> > build directory.)
> >
> > - make -C $KDIR M=$PWD
> > + make -C $KDIR M=$PWD [MO=$BUILD_DIR]
> >
> > -C $KDIR
> > The directory that contains the kernel and relevant build
> > @@ -109,6 +109,9 @@ executed to make module versioning work.
> > directory where the external module (kbuild file) is
> > located.
> >
> > + MO=$BUILD_DIR
> > + Speficies a separate output directory for the external module.
>
> s/Speficies/Specifies/
>
> > +
> > 2.3 Targets
> > ===========
> >
> > diff --git a/Makefile b/Makefile
> > index 9fbf7ef6e394..b654baa0763a 100644
> > --- a/Makefile
> > +++ b/Makefile
> > @@ -134,6 +134,10 @@ ifeq ("$(origin M)", "command line")
> > KBUILD_EXTMOD := $(M)
> > endif
> >
> > +ifeq ("$(origin MO)", "command line")
> > + KBUILD_EXTMOD_OUTPUT := $(MO)
> > +endif
> > +
> > $(if $(word 2, $(KBUILD_EXTMOD)), \
> > $(error building multiple external modules is not supported))
>
> Should we also check against multiple output directories?
>
> >
> > @@ -187,7 +191,11 @@ ifdef KBUILD_EXTMOD
> > else
> > objtree := $(CURDIR)
> > endif
> > - output := $(KBUILD_EXTMOD)
> > + output := $(or $(KBUILD_EXTMOD_OUTPUT),$(KBUILD_EXTMOD))
> > + # KBUILD_EXTMOD might be a relative path. Remember its absolute path before
> > + # Make changes the working directory.
> > + export abs_extmodtree := $(realpath $(KBUILD_EXTMOD))
> > + $(if $(abs_extmodtree),,$(error specified external module directory "$(KBUILD_EXTMOD)" does not exist))
> > else
> > objtree := .
> > output := $(KBUILD_OUTPUT)
> > @@ -246,7 +254,6 @@ else # need-sub-make
> > ifeq ($(abs_srctree),$(CURDIR))
> > # building in the source tree
> > srctree := .
> > - building_out_of_srctree :=
> > else
> > ifeq ($(abs_srctree)/,$(dir $(CURDIR)))
> > # building in a subdirectory of the source tree
> > @@ -254,22 +261,23 @@ else
> > else
> > srctree := $(abs_srctree)
> > endif
> > - building_out_of_srctree := 1
> > endif
> >
> > ifneq ($(KBUILD_ABS_SRCTREE),)
> > srctree := $(abs_srctree)
> > endif
> >
> > -VPATH :=
> > +export srctree
> >
> > -ifeq ($(KBUILD_EXTMOD),)
> > -ifdef building_out_of_srctree
> > -VPATH := $(srctree)
> > -endif
> > -endif
> > +_vpath = $(or $(abs_extmodtree),$(srctree))
> >
> > -export building_out_of_srctree srctree VPATH
> > +ifeq ($(realpath $(_vpath)),$(CURDIR))
>
> Just a style consistency question: 'ifeq (,)' with a space after ',' (as a few
> lines above) or without as used here?
>
> > +building_out_of_srctree :=
> > +VPATH :=
> > +else
> > +export building_out_of_srctree := 1
> > +export VPATH := $(_vpath)
> > +endif
> >
> > # To make sure we do not include .config for any of the *config targets
> > # catch them early, and hand them over to scripts/kconfig/Makefile
> > @@ -550,7 +558,7 @@ USERINCLUDE := \
> > LINUXINCLUDE := \
> > -I$(srctree)/arch/$(SRCARCH)/include \
> > -I$(objtree)/arch/$(SRCARCH)/include/generated \
> > - $(if $(building_out_of_srctree),-I$(srctree)/include) \
> > + -I$(srctree)/include \
> > -I$(objtree)/include \
> > $(USERINCLUDE)
> >
> > @@ -640,6 +648,7 @@ quiet_cmd_makefile = GEN Makefile
> > } > Makefile
> >
> > outputmakefile:
> > +ifeq ($(KBUILD_EXTMOD),)
> > @if [ -f $(srctree)/.config -o \
> > -d $(srctree)/include/config -o \
> > -d $(srctree)/arch/$(SRCARCH)/include/generated ]; then \
> > @@ -649,7 +658,16 @@ outputmakefile:
> > echo >&2 "***"; \
> > false; \
> > fi
> > - $(Q)ln -fsn $(srctree) source
> > +else
> > + @if [ -f $(KBUILD_EXTMOD)/modules.order ]; then \

If I do

make O=build
touch fs/btrfs/modules.order
make O=build M=fs/btrfs/ MO=/tmp/btrfs CONFIG_BTRFS_FS=m

the 'modules.order' check does not work correctly. But a consequtive

make -C /tmp/btrfs CONFIG_BTRFS_FS=m

shows the 'The external module source tree is not clean.' message as
expected.

If I replace $(KBUILD_EXTMOD) by $(abs_extmodtree), it works for me also
in the first case.


Kind regards,
Nicolas