cpet/openbsd-ports
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
71,509 Commits 1 Branch 0 Tags 554 MiB
c71b9ad066
Commit Graph

144 Commits

Author SHA1 Message Date
espie
c71b9ad066 expose interface to "refresh" display after some disturbing operation. 2011-12-10 14:47:36 +00:00
espie
3081c4ac9a remove "needed" debug log now that things work fine. 2011-12-05 21:27:53 +00:00
espie
03c4aa3410 ouch, don't forget to lock at front of prepare (okay, we can avoid it 
if no-depends but that's complicated)

also, two distinct critical sections: we don't need to stay locked at
the end of show-prepare-result.

Even though that's not really important, simplify logic a wee little bit
 2011-12-05 21:18:55 +00:00
espie
b10178048d move uninstall behavior to uninstall, unbreak !-J 
as noticed by fgs@
 2011-12-05 18:29:39 +00:00
espie
abb8ddf57d turn -J into an option taking a number. 
always run show-prepare-results, and always serialize
depends/prepare/show-prepare-results(/junk) on a single host.
 2011-12-05 16:10:00 +00:00
espie
a38035bd67 clean-up job running a bit: add methods to Grabber and PortBuilder objects 
to access more stuff they need.

Remove distantshell->make, ask state instead.
Provide state with the program we run, and the parameters we pass it.

- same make for distant and local hosts
- pass BUILD_ONCE=Yes if -a.
 2011-12-04 12:05:41 +00:00
espie
3858760dd5 cut up big method into smaller parts 2011-12-04 10:57:46 +00:00
espie
69df96b71c properly forget old info when rescanning... 2011-12-04 10:39:52 +00:00
espie
1674edcbbd open needed just once. 2011-12-03 11:10:48 +00:00
espie
c1a46f71e8 faster NFS: if the missing file suddenly reappears, we don't need to rescan 
the path, just put it back in the queue, and forget it ever happened.
 2011-12-03 11:03:07 +00:00
espie
0037a6f50e little message tweak (to be changed anyways) 2011-12-03 09:29:41 +00:00
espie
4d52ff4e54 new -J option 
- refactor redirect a bit to store multi-line results to a temporary file
- new show-prepare-results stage using that
- prepare and show-prepare-results shouldn't be NoTime, since they're
predictable.
- lock host thru depends -> junk in the junk case.
acquiring the lock is done thru polling in the child, so that it doesn't block
dpb proper.
- new "junk" stage that gathers dependencies and deletes them.
 2011-12-02 22:37:36 +00:00
espie
2cfd4c615d new junk option (to remove packages as they're no longer needed) to pass 
around thru JobBuilder (along with locks since we need to write stuff into
that file)
 2011-12-02 22:33:46 +00:00
espie
b5c447a355 don't buffer printing to locks, to be able to get dependencies early 
gather needed=* lines from locks (may want to cache that info later,
but I don't think it has any actual performance impact)
 2011-12-02 22:32:07 +00:00
espie
770bcd5b5a methods required to be able to lock per-host 2011-12-02 22:29:28 +00:00
espie
18a4d6f4bf remove dangerous optimization: have paths explicitly pass thru the depends 
stages, even if the package is already there.

This could break EXTRA depends where the package would already be there,
and thus its dependency tree not really solved, and thus depends could
easily miss packages...

problem reported by nigel@... and not reported by ajacoutot@ :)
 2011-12-02 11:52:06 +00:00
espie
f44c50300c EXTRA should (partly) be handled like DEPENDS/BDEPENDS 2011-12-02 11:40:25 +00:00
espie
ac5bbeeeae rpath without libs, that's a fun one 2011-12-01 11:13:25 +00:00
espie
2a68fab763 thinko 2011-12-01 11:11:23 +00:00
espie
34c5369453 unconfuse naming 2011-11-27 17:31:54 +00:00
espie
e0c5221e97 add a bit of debug code that logs everything that's going on... 2011-11-27 16:15:52 +00:00
espie
bc6cc54b9b move binary scanner into its own module 2011-11-27 14:49:42 +00:00
espie
5b1516ea0b unify ldd/objdump runs somewhat 
tweak FileSource to return its directory separately from the filename,
so that running objdump does a chdir
adjust filename to be a relative path.
 2011-11-27 11:29:33 +00:00
espie
60bc66c627 use PREPARE_CHECK_ONLY to weed out bugs in dpb. 
also, missed passing dontclean to one instance of add_normal_tasks, oops
 2011-11-22 16:48:01 +00:00
espie
4b801b7017 shorter lock line: user doesn't really care that multiple pkgpaths ran into 
the same lock, rather that it's a lock, and what the lockname is.
 2011-11-22 16:46:44 +00:00
espie
fdc7a9c9fd subtle effect of equivalence: sometimes, we do not get back the subdir 
we asked for.

This can be a problem, as rescan relies on it to reenter the path in the
various queues.

So, when we build the equivalence hash, give special treatment to extra
paths that match the actual subdirs list.

Problem reported by naddy@, who sees sporadic errors linked to NFS lag,
which mean that paths sometimes get flagged as errors, then later the error
vanishes automatically when the package appears...

well, this makes darn sure the correct pkgpath does not vanish.
 2011-11-22 16:44:53 +00:00
jasper
f3ebafe1ff Output an error message if we're supposed to link with a non-existent library, 
instead of puking perl errors.

ok espie@
 2011-11-14 22:14:38 +00:00
jasper
3dcf344837 Make output slightly more similar like gnu libtool in various 
modes. Simplify Exec->command() a bit too while here as suggested by espie@

ok espie@
 2011-11-14 22:12:08 +00:00
espie
2f6111d127 explain to the engine how to install stuff when it's ready (no error 
checks though).

new -I option to do that.
change -C option to be more useful.
 2011-11-14 21:57:47 +00:00
espie
5ffb1d6f17 allow check for local cores 2011-11-14 21:56:04 +00:00
espie
c5a81668f4 show parents when we recurse thru subdirs. a bit noisy, but useful when 
you want to figure out why some big stuff you wanted to avoid is building
 2011-11-14 19:03:41 +00:00
espie
aba2721637 simplify lock code so that lock/unlock actually works, even if I lose 
small-grained locks for fullpkgpath
also fix a stupid parenting bug...
 2011-11-13 22:18:04 +00:00
espie
2ab4b2c7e7 better FETCH_MANUALLY handling: 
- 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
 2011-11-13 10:34:35 +00:00
espie
e15fa3097d oops, fix FETCH_MANUALLY bug. 
Prepare for MISSING_FILES
 2011-11-13 10:28:32 +00:00
espie
8f77bf5ec5 make stuff slightly more verbose all the time 2011-11-12 13:19:26 +00:00
espie
9efee931ed smarter dump at end of build: 
- 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
 2011-11-09 08:28:55 +00:00
espie
b0e6afe5e4 split DEPENDS into not done/done to help prepare work correctly 2011-11-08 10:26:38 +00:00
espie
cb47853a4e backout shrinkage that breaks prepare 2011-11-08 09:07:24 +00:00
espie
fb0691e921 share MULTI_PACKAGES. 
simplify the default case (SUBPACKAGE=-main).
 2011-11-07 16:03:19 +00:00
espie
b016e09750 modify pkgpath to store info in a more compact way (in particular, 
get rid of sawflavor, and just use the existence of the key to mark the
presence of flavors).

also remove yet more info once we used it.
 2011-11-07 13:23:09 +00:00
espie
8bba5b93fa remove a bit more data 2011-11-07 13:21:46 +00:00
espie
3915fb3951 don't auto-vivify info. 2011-11-06 17:56:53 +00:00
espie
7238b18cc3 make sure print_parent gets the correct path 2011-11-06 16:56:23 +00:00
espie
ebe71b8762 slightly better error handling and reporting: 
- log to pkgpath whenever we can
- log FETCH_MANUALLY issues as errors, so we can fix them
 2011-11-06 12:23:28 +00:00
espie
50ac27a1e8 use print_parent 2011-11-06 12:22:17 +00:00
espie
63910d84bb document why it's broken. Link more parent stuff. 2011-11-06 12:21:47 +00:00
espie
3f5146bbc7 make it simpler to print parent. 
zap copy_flavors which I no longer use
 2011-11-06 12:21:07 +00:00
espie
0ea853e6e2 convenience function to log to correct path more easily 2011-11-06 12:20:32 +00:00
espie
1ccfe6a255 propagate IGNOREd status to dependent ports. 
display directly into engine.log.
replace all info with a stub "ignored already".
This shrinks memory usage some more.
 2011-11-05 18:27:13 +00:00
espie
561e2aa812 shrink a bit: the reason for copy_flavors use is long gone 
remove the debug scaffolding that duplicates depends list
 2011-11-05 18:25:36 +00:00