kpriv/pkgsrc
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity
pkgsrc/mk/bsd.pkg.barrier.mk

93 lines
3.1 KiB
Makefile
Raw Normal View History

Rename the install phase to stage-install. Introduce a new install target that defaults to either stage-install or package-install, depending on whether DESTDIR support is active and supported by the package or not.
2010-02-24 23:53:34 +01:00
# $NetBSD: bsd.pkg.barrier.mk,v 1.18 2010/02/24 22:53:34 joerg Exp $
Flatten out recursive makes into a single re-invocation of make by introducing the concept of a "barrier". We separate the user-invokable targets into ones that must happen before the barrier, and ones that must happen after the barrier. The ones that happen after the barrier are run in a sub-make process. In this case, the targets that must be run after the barrier are from the "wrapper" step and beyond. We rewrite the various "flow" targets, e.g. wrapper, configure, build, etc., so that they of the right form to use the barrier target. This now completely removes the concept of PKG_PHASE from pkgsrc. It is replaced with the concept of "before" and "after" the barrier, and this state can be checked by testing for the existence of the barrier cookie file. Because we've removed most of the recursive makes, there is now nowhere to hook the PKG_ERROR_HANDLER.* commands, so remove them for now. As part of this commit, put back the logic that conditionalized the sources for the various cookie files. Because the sources are all "phony" targets, they were always run, regardless of whether or not the cookie file already existed. Now, if a cookie file exists, then that entire phase associated with that cookie file is skipped. Lastly, fix a thinko in configure/bsd.configure.mk where setting NO_CONFIGURE in a package Makefile would manage to skip the "wrapper" step altogether. Fix this by correctly noting "wrapper" and not "patch" as the preceding step to "configure".
2006-07-06 00:21:02 +02:00
Rename <phase>_COOKIE variables to _COOKIE.<phase>. These are private variables so there are no user-visible changes. This change just makes it a little easier to write for loops.
2006-07-07 23:24:27 +02:00
_COOKIE.barrier= ${WRKDIR}/.barrier_cookie
Flatten out recursive makes into a single re-invocation of make by introducing the concept of a "barrier". We separate the user-invokable targets into ones that must happen before the barrier, and ones that must happen after the barrier. The ones that happen after the barrier are run in a sub-make process. In this case, the targets that must be run after the barrier are from the "wrapper" step and beyond. We rewrite the various "flow" targets, e.g. wrapper, configure, build, etc., so that they of the right form to use the barrier target. This now completely removes the concept of PKG_PHASE from pkgsrc. It is replaced with the concept of "before" and "after" the barrier, and this state can be checked by testing for the existence of the barrier cookie file. Because we've removed most of the recursive makes, there is now nowhere to hook the PKG_ERROR_HANDLER.* commands, so remove them for now. As part of this commit, put back the logic that conditionalized the sources for the various cookie files. Because the sources are all "phony" targets, they were always run, regardless of whether or not the cookie file already existed. Now, if a cookie file exists, then that entire phase associated with that cookie file is skipped. Lastly, fix a thinko in configure/bsd.configure.mk where setting NO_CONFIGURE in a package Makefile would manage to skip the "wrapper" step altogether. Fix this by correctly noting "wrapper" and not "patch" as the preceding step to "configure".
2006-07-06 00:21:02 +02:00
Order the _BARRIER_POST_TARGETS so that if more than one is specified on the command-line, then we don't invoke make once for each target, and pass them to the sub-make in a sensible order.
2006-07-06 17:33:19 +02:00
# _BARRIER_PRE_TARGETS is a list of the targets that must be built before
Flatten out recursive makes into a single re-invocation of make by introducing the concept of a "barrier". We separate the user-invokable targets into ones that must happen before the barrier, and ones that must happen after the barrier. The ones that happen after the barrier are run in a sub-make process. In this case, the targets that must be run after the barrier are from the "wrapper" step and beyond. We rewrite the various "flow" targets, e.g. wrapper, configure, build, etc., so that they of the right form to use the barrier target. This now completely removes the concept of PKG_PHASE from pkgsrc. It is replaced with the concept of "before" and "after" the barrier, and this state can be checked by testing for the existence of the barrier cookie file. Because we've removed most of the recursive makes, there is now nowhere to hook the PKG_ERROR_HANDLER.* commands, so remove them for now. As part of this commit, put back the logic that conditionalized the sources for the various cookie files. Because the sources are all "phony" targets, they were always run, regardless of whether or not the cookie file already existed. Now, if a cookie file exists, then that entire phase associated with that cookie file is skipped. Lastly, fix a thinko in configure/bsd.configure.mk where setting NO_CONFIGURE in a package Makefile would manage to skip the "wrapper" step altogether. Fix this by correctly noting "wrapper" and not "patch" as the preceding step to "configure".
2006-07-06 00:21:02 +02:00
# the "barrier" target invokes a new make.
#
Introduce a new target "checksum-phase" which is used as a dependency target by pkgsrc. This new target computes checksums only up until the extract phase is complete. The "checksum" target is now a target that can always be run by the user at any time, regardless of the presence of the work directory. These changes were proposed in PR pkg/36603 by Robert Elz.
2007-08-31 18:30:11 +02:00
_BARRIER_PRE_TARGETS= checksum-phase makedirs depends
Flatten out recursive makes into a single re-invocation of make by introducing the concept of a "barrier". We separate the user-invokable targets into ones that must happen before the barrier, and ones that must happen after the barrier. The ones that happen after the barrier are run in a sub-make process. In this case, the targets that must be run after the barrier are from the "wrapper" step and beyond. We rewrite the various "flow" targets, e.g. wrapper, configure, build, etc., so that they of the right form to use the barrier target. This now completely removes the concept of PKG_PHASE from pkgsrc. It is replaced with the concept of "before" and "after" the barrier, and this state can be checked by testing for the existence of the barrier cookie file. Because we've removed most of the recursive makes, there is now nowhere to hook the PKG_ERROR_HANDLER.* commands, so remove them for now. As part of this commit, put back the logic that conditionalized the sources for the various cookie files. Because the sources are all "phony" targets, they were always run, regardless of whether or not the cookie file already existed. Now, if a cookie file exists, then that entire phase associated with that cookie file is skipped. Lastly, fix a thinko in configure/bsd.configure.mk where setting NO_CONFIGURE in a package Makefile would manage to skip the "wrapper" step altogether. Fix this by correctly noting "wrapper" and not "patch" as the preceding step to "configure".
2006-07-06 00:21:02 +02:00
Order the _BARRIER_POST_TARGETS so that if more than one is specified on the command-line, then we don't invoke make once for each target, and pass them to the sub-make in a sensible order.
2006-07-06 17:33:19 +02:00
# _BARRIER_POST_TARGETS is a list of the targets that must be built after
# the "barrier" target invokes a new make. This list is specially
# ordered so that if more than one is specified on the command-line,
# then pkgsrc will still do the right thing.
Flatten out recursive makes into a single re-invocation of make by introducing the concept of a "barrier". We separate the user-invokable targets into ones that must happen before the barrier, and ones that must happen after the barrier. The ones that happen after the barrier are run in a sub-make process. In this case, the targets that must be run after the barrier are from the "wrapper" step and beyond. We rewrite the various "flow" targets, e.g. wrapper, configure, build, etc., so that they of the right form to use the barrier target. This now completely removes the concept of PKG_PHASE from pkgsrc. It is replaced with the concept of "before" and "after" the barrier, and this state can be checked by testing for the existence of the barrier cookie file. Because we've removed most of the recursive makes, there is now nowhere to hook the PKG_ERROR_HANDLER.* commands, so remove them for now. As part of this commit, put back the logic that conditionalized the sources for the various cookie files. Because the sources are all "phony" targets, they were always run, regardless of whether or not the cookie file already existed. Now, if a cookie file exists, then that entire phase associated with that cookie file is skipped. Lastly, fix a thinko in configure/bsd.configure.mk where setting NO_CONFIGURE in a package Makefile would manage to skip the "wrapper" step altogether. Fix this by correctly noting "wrapper" and not "patch" as the preceding step to "configure".
2006-07-06 00:21:02 +02:00
#
Added a call similar to "make error-check" when the real target fails for some reason. That way, error and warning messages that have long been hidden are now shown. They don't appear very prominently by now, but it's much better than showing them not at all.
2006-10-06 21:04:37 +02:00
# FIXME: why is wrapper before extract?
Move barrier to just after the depends phase. This allows both the "tools" and "wrapper" phases to be able to expect all the dependencies to be installed already.
2006-07-25 20:04:25 +02:00
_BARRIER_POST_TARGETS= tools
_BARRIER_POST_TARGETS+= wrapper
Rearrange sequence so that "wrapper" occurs between "tools" and "extract". This allows all user-definable targets (pre-*, do-*, post-*) to have access to the wrapper scripts.
2006-07-21 16:21:28 +02:00
_BARRIER_POST_TARGETS+= extract
_BARRIER_POST_TARGETS+= patch
Order the _BARRIER_POST_TARGETS so that if more than one is specified on the command-line, then we don't invoke make once for each target, and pass them to the sub-make in a sensible order.
2006-07-06 17:33:19 +02:00
_BARRIER_POST_TARGETS+= configure
Since the "build" target is run inside the pkgsrc barrier, "build-env" gets the same treatment.
2007-11-30 19:54:40 +01:00
_BARRIER_POST_TARGETS+= build build-env
Flatten out recursive makes into a single re-invocation of make by introducing the concept of a "barrier". We separate the user-invokable targets into ones that must happen before the barrier, and ones that must happen after the barrier. The ones that happen after the barrier are run in a sub-make process. In this case, the targets that must be run after the barrier are from the "wrapper" step and beyond. We rewrite the various "flow" targets, e.g. wrapper, configure, build, etc., so that they of the right form to use the barrier target. This now completely removes the concept of PKG_PHASE from pkgsrc. It is replaced with the concept of "before" and "after" the barrier, and this state can be checked by testing for the existence of the barrier cookie file. Because we've removed most of the recursive makes, there is now nowhere to hook the PKG_ERROR_HANDLER.* commands, so remove them for now. As part of this commit, put back the logic that conditionalized the sources for the various cookie files. Because the sources are all "phony" targets, they were always run, regardless of whether or not the cookie file already existed. Now, if a cookie file exists, then that entire phase associated with that cookie file is skipped. Lastly, fix a thinko in configure/bsd.configure.mk where setting NO_CONFIGURE in a package Makefile would manage to skip the "wrapper" step altogether. Fix this by correctly noting "wrapper" and not "patch" as the preceding step to "configure".
2006-07-06 00:21:02 +02:00
_BARRIER_POST_TARGETS+= test
Move inclusion of bsd.pkg.barrier.mk under the .MAIN target because bsd.pkg.barrier.mk uses the "make()" test expression. Also, include "all" as a post-barrier target since it is implicitly the ".MAIN" target when a user just types "make" in a package directory.
2006-07-06 17:43:10 +02:00
_BARRIER_POST_TARGETS+= all
Rename the install phase to stage-install. Introduce a new install target that defaults to either stage-install or package-install, depending on whether DESTDIR support is active and supported by the package or not.
2010-02-24 23:53:34 +01:00
_BARRIER_POST_TARGETS+= stage-install
Order the _BARRIER_POST_TARGETS so that if more than one is specified on the command-line, then we don't invoke make once for each target, and pass them to the sub-make in a sensible order.
2006-07-06 17:33:19 +02:00
_BARRIER_POST_TARGETS+= reinstall
The "replace" target does conceptually the same thing as the "install" target in that it installs the currently-built software into the filesystem. In that case where "replace" is specified as a target on the command line, make "replace" and not "install" be the source target for "package". Also, place the "replace" target between the "install" and "package" targets in _BARRIER_POST_TARGETS as it should be legal to do those steps in that order (but not in another order). These changes make the following work: make replace package In this example, the currently installed package will be replaced and the newly-installed software will be packaged, all within the same make process.
2006-07-07 00:08:32 +02:00
_BARRIER_POST_TARGETS+= replace
Order the _BARRIER_POST_TARGETS so that if more than one is specified on the command-line, then we don't invoke make once for each target, and pass them to the sub-make in a sensible order.
2006-07-06 17:33:19 +02:00
_BARRIER_POST_TARGETS+= package
_BARRIER_POST_TARGETS+= repackage
Add package-install. For non-DESTDIR builds, package and package-install are identical. For DESTDIR builds, the package is not installed to PREFIX as part of the build, so package-install does exactly that after package is done. Change bin-install to call package-install.
2006-11-03 09:01:04 +01:00
_BARRIER_POST_TARGETS+= package-install
Rename the install phase to stage-install. Introduce a new install target that defaults to either stage-install or package-install, depending on whether DESTDIR support is active and supported by the package or not.
2010-02-24 23:53:34 +01:00
_BARRIER_POST_TARGETS+= install
Order the _BARRIER_POST_TARGETS so that if more than one is specified on the command-line, then we don't invoke make once for each target, and pass them to the sub-make in a sensible order.
2006-07-06 17:33:19 +02:00
.for _target_ in ${_BARRIER_POST_TARGETS}
. if make(${_target_})
_BARRIER_CMDLINE_TARGETS+= ${_target_}
. endif
.endfor
Flatten out recursive makes into a single re-invocation of make by introducing the concept of a "barrier". We separate the user-invokable targets into ones that must happen before the barrier, and ones that must happen after the barrier. The ones that happen after the barrier are run in a sub-make process. In this case, the targets that must be run after the barrier are from the "wrapper" step and beyond. We rewrite the various "flow" targets, e.g. wrapper, configure, build, etc., so that they of the right form to use the barrier target. This now completely removes the concept of PKG_PHASE from pkgsrc. It is replaced with the concept of "before" and "after" the barrier, and this state can be checked by testing for the existence of the barrier cookie file. Because we've removed most of the recursive makes, there is now nowhere to hook the PKG_ERROR_HANDLER.* commands, so remove them for now. As part of this commit, put back the logic that conditionalized the sources for the various cookie files. Because the sources are all "phony" targets, they were always run, regardless of whether or not the cookie file already existed. Now, if a cookie file exists, then that entire phase associated with that cookie file is skipped. Lastly, fix a thinko in configure/bsd.configure.mk where setting NO_CONFIGURE in a package Makefile would manage to skip the "wrapper" step altogether. Fix this by correctly noting "wrapper" and not "patch" as the preceding step to "configure".
2006-07-06 00:21:02 +02:00
######################################################################
### barrier (PRIVATE)
######################################################################
### barrier is a helper target that can be used to separate targets
### that should be built in a new make process from being built in
### the current one. The targets that must be built after the "barrier"
### target invokes a new make should be listed in _BARRIER_POST_TARGETS,
### and should be of the form:
###
Modify the barrier so that we always invoke the recursive make process when passing through the barrier. This ensures the PATH (passed via PKGSRC_MAKE_ENV) is correctly set for all phases after the barrier. This fixes a bug in "interactive" pkgsrc use, where if you have no work directory and type "make build && make install", then the "install" step does not have a PATH set to include all the wrapper and tools directories.
2006-07-22 18:31:35 +02:00
### .if !defined(_PKGSRC_BARRIER)
Flatten out recursive makes into a single re-invocation of make by introducing the concept of a "barrier". We separate the user-invokable targets into ones that must happen before the barrier, and ones that must happen after the barrier. The ones that happen after the barrier are run in a sub-make process. In this case, the targets that must be run after the barrier are from the "wrapper" step and beyond. We rewrite the various "flow" targets, e.g. wrapper, configure, build, etc., so that they of the right form to use the barrier target. This now completely removes the concept of PKG_PHASE from pkgsrc. It is replaced with the concept of "before" and "after" the barrier, and this state can be checked by testing for the existence of the barrier cookie file. Because we've removed most of the recursive makes, there is now nowhere to hook the PKG_ERROR_HANDLER.* commands, so remove them for now. As part of this commit, put back the logic that conditionalized the sources for the various cookie files. Because the sources are all "phony" targets, they were always run, regardless of whether or not the cookie file already existed. Now, if a cookie file exists, then that entire phase associated with that cookie file is skipped. Lastly, fix a thinko in configure/bsd.configure.mk where setting NO_CONFIGURE in a package Makefile would manage to skip the "wrapper" step altogether. Fix this by correctly noting "wrapper" and not "patch" as the preceding step to "configure".
2006-07-06 00:21:02 +02:00
### foo: barrier
### .else
### foo: foo's real source dependencies
### .endif
###
### Note that none of foo's real source dependencies should include
### targets that occur before the barrier.
###
Order the _BARRIER_POST_TARGETS so that if more than one is specified on the command-line, then we don't invoke make once for each target, and pass them to the sub-make in a sensible order.
2006-07-06 17:33:19 +02:00
Added a call similar to "make error-check" when the real target fails for some reason. That way, error and warning messages that have long been hidden are now shown. They don't appear very prominently by now, but it's much better than showing them not at all.
2006-10-06 21:04:37 +02:00
.PHONY: barrier-error-check
barrier-error-check: error-check
Flatten out recursive makes into a single re-invocation of make by introducing the concept of a "barrier". We separate the user-invokable targets into ones that must happen before the barrier, and ones that must happen after the barrier. The ones that happen after the barrier are run in a sub-make process. In this case, the targets that must be run after the barrier are from the "wrapper" step and beyond. We rewrite the various "flow" targets, e.g. wrapper, configure, build, etc., so that they of the right form to use the barrier target. This now completely removes the concept of PKG_PHASE from pkgsrc. It is replaced with the concept of "before" and "after" the barrier, and this state can be checked by testing for the existence of the barrier cookie file. Because we've removed most of the recursive makes, there is now nowhere to hook the PKG_ERROR_HANDLER.* commands, so remove them for now. As part of this commit, put back the logic that conditionalized the sources for the various cookie files. Because the sources are all "phony" targets, they were always run, regardless of whether or not the cookie file already existed. Now, if a cookie file exists, then that entire phase associated with that cookie file is skipped. Lastly, fix a thinko in configure/bsd.configure.mk where setting NO_CONFIGURE in a package Makefile would manage to skip the "wrapper" step altogether. Fix this by correctly noting "wrapper" and not "patch" as the preceding step to "configure".
2006-07-06 00:21:02 +02:00
.PHONY: barrier
Rename <phase>_COOKIE variables to _COOKIE.<phase>. These are private variables so there are no user-visible changes. This change just makes it a little easier to write for loops.
2006-07-07 23:24:27 +02:00
barrier: ${_BARRIER_PRE_TARGETS} ${_COOKIE.barrier}
Modify the barrier so that we always invoke the recursive make process when passing through the barrier. This ensures the PATH (passed via PKGSRC_MAKE_ENV) is correctly set for all phases after the barrier. This fixes a bug in "interactive" pkgsrc use, where if you have no work directory and type "make build && make install", then the "install" step does not have a PATH set to include all the wrapper and tools directories.
2006-07-22 18:31:35 +02:00
.if !defined(_PKGSRC_BARRIER)
Order the _BARRIER_POST_TARGETS so that if more than one is specified on the command-line, then we don't invoke make once for each target, and pass them to the sub-make in a sensible order.
2006-07-06 17:33:19 +02:00
. if defined(PKG_VERBOSE)
@${PHASE_MSG} "Invoking \`\`"${_BARRIER_CMDLINE_TARGETS:Q}"'' after barrier for ${PKGNAME}"
. endif
Replaced _PKG_SILENT and _PKG_DEBUG with RUN.
2008-02-07 22:36:13 +01:00
${RUN} \
Don't require pkg_delete or pkg_info -r to list packages in full topological sort for the make update list. This breaks naturally when new dependencies are introduced that are also in the list scheduled for later installation. Reported by David Holland.
2009-05-16 03:27:30 +02:00
if ${TEST} -n "${_PKGSRC_UPDATE_CHECK}" && \
${PKG_INFO} -qe ${PKGNAME}; then \
${PHASE_MSG} "Skipping installation of already handled package"; \
else \
Added a call similar to "make error-check" when the real target fails for some reason. That way, error and warning messages that have long been hidden are now shown. They don't appear very prominently by now, but it's much better than showing them not at all.
2006-10-06 21:04:37 +02:00
cd ${.CURDIR} \
&& ${RECURSIVE_MAKE} ${MAKEFLAGS} _PKGSRC_BARRIER=yes ALLOW_VULNERABLE_PACKAGES= ${_BARRIER_CMDLINE_TARGETS} \
|| { \
exitcode="$$?"; \
${RECURSIVE_MAKE} ${MAKEFLAGS} _PKGSRC_BARRIER=yes barrier-error-check; \
exit "$$exitcode"; \
Don't require pkg_delete or pkg_info -r to list packages in full topological sort for the make update list. This breaks naturally when new dependencies are introduced that are also in the list scheduled for later installation. Reported by David Holland.
2009-05-16 03:27:30 +02:00
}; \
fi
Order the _BARRIER_POST_TARGETS so that if more than one is specified on the command-line, then we don't invoke make once for each target, and pass them to the sub-make in a sensible order.
2006-07-06 17:33:19 +02:00
. if defined(PKG_VERBOSE)
@${PHASE_MSG} "Leaving \`\`"${_BARRIER_CMDLINE_TARGETS:Q}"'' after barrier for ${PKGNAME}"
. endif
Flatten out recursive makes into a single re-invocation of make by introducing the concept of a "barrier". We separate the user-invokable targets into ones that must happen before the barrier, and ones that must happen after the barrier. The ones that happen after the barrier are run in a sub-make process. In this case, the targets that must be run after the barrier are from the "wrapper" step and beyond. We rewrite the various "flow" targets, e.g. wrapper, configure, build, etc., so that they of the right form to use the barrier target. This now completely removes the concept of PKG_PHASE from pkgsrc. It is replaced with the concept of "before" and "after" the barrier, and this state can be checked by testing for the existence of the barrier cookie file. Because we've removed most of the recursive makes, there is now nowhere to hook the PKG_ERROR_HANDLER.* commands, so remove them for now. As part of this commit, put back the logic that conditionalized the sources for the various cookie files. Because the sources are all "phony" targets, they were always run, regardless of whether or not the cookie file already existed. Now, if a cookie file exists, then that entire phase associated with that cookie file is skipped. Lastly, fix a thinko in configure/bsd.configure.mk where setting NO_CONFIGURE in a package Makefile would manage to skip the "wrapper" step altogether. Fix this by correctly noting "wrapper" and not "patch" as the preceding step to "configure".
2006-07-06 00:21:02 +02:00
.endif
######################################################################
### barrier-cookie (PRIVATE)
######################################################################
### barrier-cookie creates the "barrier" cookie file.
###
Rename <phase>_COOKIE variables to _COOKIE.<phase>. These are private variables so there are no user-visible changes. This change just makes it a little easier to write for loops.
2006-07-07 23:24:27 +02:00
${_COOKIE.barrier}:
Replaced _PKG_SILENT and _PKG_DEBUG with RUN.
2008-02-07 22:36:13 +01:00
${RUN}${MKDIR} ${.TARGET:H}
${RUN}${ECHO} ${PKGNAME} > ${.TARGET}
Reference in a new issue Copy permalink