using in mk.conf, especially a bunch of .poison lines for common typos,
including suggestions from aja, espie, naddy and others. Can be used
with .include in mk.conf.
oks from at least william@ phessler@ and several users like it too.
- resolve-lib -needed, less perl invocations (gains a lot)
- show-run-depends, less fancy than run-dir-depends, a bit faster...
- don't create cache for internal targets (external stuff already did it)
define _PERLSCRIPT in pkgpath.mk since getpkgpath should use it...
- expose MISSING_FILES
- treat them as a special kind of IGNORE
- handle that in DPB
that way, ports that are IGNORE'd for other reasons (not correct arch)
stay ignored, and relevant ports with missing distfiles get better error
messages
- try resolving RDEPENDS on tobuild, so that known dependencie are zapped.
- use equivalence info to not reprint same thing twice.
- dump tobuild AND built, since both are usually relevant in case lots of
stuff didn't build
with targets.
lib-depends-check should be defined as ignored along with the rest.
errors-handling should occur last, so that the last block of errors gets
used.
Makefile.inc and bsd.port.arch.mk without infinite recursion.
- use simpler pattern to test for arch stuff, we just need the values, so
.for A B in ${MACHINE_ARCH} ${ARCH}
is enough and avoids the setting of intermediate variables.
- new PROPERTIES variable that's easy to test with e.g. if ${PROPERTIES:Mapm}
- instead of seen/unseen, have an actual constructor. Instead, mark pkgpath
for which we wantinfo/wantbuild.
- only mark EXTRA dependencies as wantinfo. So the devel/haddock,no_deps
temporary error should be gone.
- since we have FLAVOR and SUBPACKAGE available, construct as much info as
we can during vars scanning (see handle_equivalences). This avoids about 150
path rescans during a full bulk. Also, grab the timing and logsizes from
equivalent files, so that most stuff should know show % all the time.
- tweak subdirlist to be a hash, and correctly add pkgpath_and_flavors to it.
That way, we rescan avahi pseudo flavors just once, and not four or five times.
dpb coalesces build dependencies over MULTI_PACKAGES: if we don't substract
from MULTI_PACKAGES, this can lead to bootstrap loops.
Case in point: sysutils/gamin, whose build relies on "no_server" to be
available as a dependency for glib2/gtk+2.
(but IGNORED stuff is properly kept as MULTI_PACKAGES, since it's mostly
intended to avoid strange arch errors)
- if there is no flavor in BUILD_PKGPATH, it's not necessarily the default,
make sure there's an empty flavor by appending a ,
- pass FLAVOR to dump-vars, so that eventually dpb can match "no flavor
specified" to "this is the default flavor", thus getting a bit smarter
(this should speed up the LISTING job by not traversing as many subdirs).
eventually.
- fetch all files
- ignore ignores
- specific builder that doesn't look at existing packages
currently: does not stop when fetch is finished, which is somewhat of the
remaining issue.
Also: change stats to store pid, to make sense of interleaved log files.
This is occasionally useful for pseudo-flavors: these do not get encoded
in the pkgpath, so taking (for instance) sqlports, this generates lines
which are later impossible to exploit based only on the fullpkgpath, as
opposed to fullpkgpath,flavor (which might contain the flavor twice, but
this is not an issue).
Fixes ruby 1.9 build with systrace now that we have these system calls.
The other new *at system calls need to have path restrictions and will
need further work so are still prevented for now (in those cases, the
supplied paths are *relative to a certain FD*, so we can't simply
examine supplied paths).
MP machine, show examples of lines displayed by dpb, document the extra
files produced by fetch. Explain how fetch works (in particular, the *.part
files and the use of ftp -C).
exit if the dependency doesn't match.
then do the actual dependency and the normal tests.
stop looking at full pkgnames list, makes no sense, even configure depends
can use the required ,-subpackage if they want.
This allows stuff such as BUILD_DEPENDS = dir>=5.0:configure
to get out early if the tree is not uptodate, instead of first configuring
then getting out (problem noticed by landry@).
tested by landry@ and jasper@
- if _DEPRUNLIB is empty, there's no need for any computation since the
result will be empty.
- zap the loop on _LIB4* for port-wantlib-args, as run-dir-depends already
grabs them.
- expose wantlib_args and corresponding targets for pkg_create to take
advantage of.
- add a dirty way for developers to waive the fake vs. ports check (PLEASE
use with EXTRA CAUTION).
* cut it into separate targets for readability and better testing
* filter libraries more efficiently
* use internal variable for resolve-lib
create two targets for solving wantlib: one (_fake-wantlib-args) which uses
the information under the fake directory and installed packages, and another
(_port-wantlib-args) which walks packing-lists. The second one can be used
to collect meta-info even when nothing is installed, and thus provide better
package-signature accuracy, or help with lib-depends-check.
The first one uses what's actually in place when a package is built.
For now, we're paranoid and use both when building a package, erroring out
if they don't match, even though _port-wantlib-args is somewhat slower.
there's a safeguard against self-depends anyways.
This prevents ports from getting moved from "built" to "installable"
prematurely.
For instance, devel/libsoup,-gnome doesn't directly depend on
net/glib2-networking, but it depends on devel/libsoup,-main which has
the net/glib2-networking.
Without this patch, dependencies on devel/libsoup,-gnome would often build
prematurely and build net/glib2-networking with them.
- zap heuristic#3, not critical
- don't rebuild the Fetch queue all the time, it's too expensive
- fix bugs in FetchQueue computation, correctly handle 1 distfile case
- create a quick path thru adjust* that handles only new paths.
- use it during LISTING while fetching files, to make more files available
quicker.
- model distfiles as FDEPENDS, DIST, so that they can get dumped.
- add -v option that dumps a bit more stuff in log files.
with these, dpb will fetch all distfiles in a not too surprising order,
and it won't consume 40% cpu while doing so...
it more obvious how much is shared (goal is to unify a large part of these).
In particular, the locking code/stuff already active is glaringly the same...
Polish the distfiles measurements. For now, 3 heuristics that kick in
at various stages of the build (a bit of black magic).
Fix pkgpath->{has} to record both "no dependencies" and "several distfiles
missing" to avoid grabbing small patchfiles which don't lead to valid
builds while heuristic#1 is active.
dpb while checksumming huge tarballs.
Use ftp -C in a systematic way, more complicated logic as to when to
remove temp files, when to keep them: if core exited okay, wrong size
is very bad. If checksum failed on a partial fetch, retry same site with
an empty file...
- clean up clock handling, jobs with timer.
- pass more parameters thru state
- tweak the engine to handle distfiles
- lots of new code in Fetch for new jobs.
should be linked statically as reported by David Talkington. Tested with
all ports showing in bulk build logs as using -all-static.
"that looks ok for our needs" steven@
Deprecate: -t and -T
use -DCONNECTION_TIMEOUT=... -DDISPLAY_TIMEOUT=... instead
New feature: -DSTUCK_TIMEOUT=
kill tasks when they don't show any progress for that long.
