Installation type SCRIPT has beed disabled long ago because of its
overlap with EXE. Remove the dead code around it.
Signed-off-by: Jan Lindemann <jan@janware.com>
jw-pkg supports more than RPM-based package managers, but for
historic reasons, lots of its Makefile variables still have "RPM" in
their names. This is misleading. Replace "RPM" in variable names by
the more generic "PKG" where appropriate.
Signed-off-by: Jan Lindemann <jan@janware.com>
PKG_FORMAT is now more straightforward to get with from the revised
jw-pkg.py distro info --format '%{cascade}', so do that, and do it in
the context where all the other variables are set from the output of
that command.
Signed-off-by: Jan Lindemann <jan@janware.com>
To remove redundancy, get-os.sh needs to be retired in favor of
pkg.py distro info. It's needed in platform.mk, but the only
definiton of JW_PKG_PY is in projects.mk, so move it, along with the
variables essential for the command:
include $(JWBDIR)/make/py-version.mk (defining PYTHON)
JW_PKG_PY
DEVELOPMENT
VERSION_FILE
Signed-off-by: Jan Lindemann <jan@janware.com>
Define Q ?= @, and replace @<command> in recipes by $(Q)<command>.
Meant to be overridden from the environment for debugging as in
Q= make
Signed-off-by: Jan Lindemann <jan@janware.com>
Running "make install" from an arbitrary source directory currently
by default either installs to a user-accessible ENV_PREFIX, or, if
DEVELOPMENT is set to false, tries to install into the system's root
filesystem, but fails over permission errors. This was by design: To
now, I considered trying the latter ill-conceived, because installing
without package manager control bears the risk of leaving unversioned
files in the system.
Actually, thinking again, during development this looks like a valid
use case: Having run pkg-rebuild-reinstall before, installing from a
source directory will leave a trace in the package manager's hash
check output, will be handled during the next clean install, and
might be a useful shortcut for trying things in the root file system.
So make this possible by:
$ DEVELOPMENT=false make install
Signed-off-by: Jan Lindemann <jan@janware.com>
PACKAGE_INSTALL_DIR, true by default, can be used to turn off
target dir creation from a directory to avoid duplication from
other packages
Signed-off-by: Jan Lindemann <jan@janware.com>
jw-build is meant as a generic set of tools for building other projects, so
references to projects being built by jw-build introduce a circular dependency.
Remove those references from defs.mk and rules.mk.
Variables removed by this commit:
BTOOLS_DIR
FEEDFSD_DIR
FEEDFS_DIR
FEEDFS_GUI_DIR
FEEDFS_OBJECTS_DIR
FEEDFS_UTILS_DIR
FEEDFS_WT_DIR
JUX_SHARED_DIR
SITE_DIR
VALDI_DIR
XCHANGE_DIR
Signed-off-by: Jan Lindemann <jan@janware.com>
make carries MAKEFLAGS, PATH, LD_LIBRARY_PATH and umask from the environment to
sub-makes, which sometimes is undesirable. So, instead of make -C,
$(PRISTINE_MAKE) -C will start a sub-make with a minimal environment consisting
of MINIMAL_UMASK, MINIMAL_PATH, MINIMAL_LD_LIBRARY_PATH and MINIMAL_MAKEFLAGS
Signed-off-by: Jan Lindemann <jan@janware.com>
- Remove temporary variable PRJS_DIR
- Add defaults for BUILD_LIBS_PREFIX and BUILD_TOOLS_PREFIX
- Move definitions of varables needed early on from defs.mk
into projects.mk, such as PYTHON and PYTHON_VERSION
Signed-off-by: Jan Lindemann <jan@janware.com>
Predefined PY_INSTALL_DIR was ignored prior to this commit, which shuffles the
py-defs.mk variables around considerably. Still some temporary variables
unnecessarily defined with ?=, but some testing would be good prior to further
changes.
Signed-off-by: Jan Lindemann <jan@janware.com>
It was possible for DEVEL_TARGETS to contain install_PKG_CONFIG without a
package config file being built, fixed that.
Signed-off-by: Jan Lindemann <jan@janware.com>
The contents of the PROJECT variable was misdetected from the directory name in
that everything after a number was discarded. This commit fixes that.
Signed-off-by: Jan Lindemann <jan@janware.com>
- Add support MCU_FLASH_BANK_NAME to support multiple banks
- Prefix binaries with TARGET_PRODUCT, if defined
- Fix flash data for stm32f769i-disco
Signed-off-by: Jan Lindemann <jan@janware.com>
- Group more consistently into overall default choices, template cascade and
compiler flags
- Add definitions for st-nucleo-f103rb (for now)
Signed-off-by: Jan Lindemann <jan@janware.com>
platform.mk is sometimes included before defs.mk, e.g. from
$(TOPDIR)/make/proj.mk, then these definitions are necessary
Signed-off-by: Jan Lindemann <jan@janware.com>
This commit makes it possible to successfully run "make all" against ytools'
again, with TARGET_TUPLE set to i686-ms-w64-mingw. Lots of minor and major
tweaks here and there.
The biggest diff is a move of the architecture-related definitions into
platform.mk. The are needed pretty early on, so that seems reasonable.
Making this work again is part of the larger effort to support cross
buildchains in a more concise way, i.e. without so many if ($(TARGET),mingw))
all over the place. TARGET's relevance should dwindle, until it's finally taken
over by the TARGET_XXX variables extracted from TARGET_TUPLE or TARGET_TRIPLET.
Signed-off-by: Jan Lindemann <jan@janware.com>
Reordered some variables in defs.mk and defs-cpp.mk for clarity. Most notably
FULL_NAME was moved back into defs.mk, which amounts to a bugfix.
Signed-off-by: Jan Lindemann <jan@janware.com>
This commit tries to remove the necessity to call projects.py from $(TOPDIR) to
speed up recursive builds over all projects yet again. This is a major
undertaking. There are two variables which are filled py projects.py in
$(TOPDIR): PREREQ and PREREQ_DIRS. Sadly, the latter is a path relative to
$(TOPDIR)/make, so this is kind of pointless. Unless the cache is maintained in
$(TOPDIR), a thing I tried to avoid. So this commit is only able to cache
$(PREREQ), not $(PREREQ_DIRS), which still is a hassle. Introduced defs-dirs.mk
for that, to make it accessible to make.mk, and modified all the other parts of
the machinery, too.
Signed-off-by: Jan Lindemann <jan@janware.com>
