Age | Commit message (Collapse) | Author |
|
Patches contributed by Mathias Bauer
gnumake4 work variously
http://svn.apache.org/viewvc?view=revision&revision=1394707
http://svn.apache.org/viewvc?view=revision&revision=1394326
http://svn.apache.org/viewvc?view=revision&revision=1397337
http://svn.apache.org/viewvc?view=revision&revision=1397315
http://svn.apache.org/viewvc?view=revision&revision=1396797
Patches contributed by Andre Fischer
Fixed getcsym.awk to handle #-comments that contain special regexp chars.
http://svn.apache.org/viewvc?view=revision&revision=1230971
118778: Added ADDITIONAL_REPOSITORIES environment variable and its
automatic setup in configure.
http://svn.apache.org/viewvc?view=revision&revision=1232004
118160: Added external CoinMP library.
http://svn.apache.org/viewvc?view=revision&revision=1233909
Patches contributed by Herbert Duerr
#i119168# use generic LICENSE file for langpacks and sdks
http://svn.apache.org/viewvc?view=revision&revision=1310178
macosxotoolhelper: need to quote perl regexp if it may contain regexp metachars
http://svn.apache.org/viewvc?view=revision&revision=1183367
allow gbuild with empty sysroot on linux
http://svn.apache.org/viewvc?view=revision&revision=1179186
Patches contributed by Ingo Schmidt
native373: #164472# improvements for msi database
http://svn.apache.org/viewvc?view=revision&revision=1167540
http://svn.apache.org/viewvc?view=revision&revision=1167539
Patches contributed by Jurgen Schmidt
adapt setup package scripts to handle special DS_Store file for
developer snapshot builds
http://svn.apache.org/viewvc?view=revision&revision=1232430
imported patch extensions_i117681.patch
http://svn.apache.org/viewvc?view=revision&revision=1172102
Patches contributed by Michael Stahl
gbuild: RepositoryFixes.mk should be optional
http://svn.apache.org/viewvc?view=revision&revision=1166123
xslt filter: remove the FLA horror
wordml import filter: replace FLA usage with plain XSLT
http://svn.apache.org/viewvc?view=revision&revision=1363727
Patch contributed by Oliver-Rainer Wittmann
i#88652: applied patch, remove unicows deps
http://svn.apache.org/viewvc?view=revision&revision=1177585
Remove lots of OS2 conditionals, re-extract Rhino Java, unwind
cppunit pieces, cleanup Mac image bits, remove coin-mp and
re-package lpsolve, Oxygen & Crystal, fixup qstart bits, expand
MPLv2 subset checking.
Change-Id: Iad5c8a76399620b892671633c0d8c29996db3564
|
|
Change-Id: Id184df89c5c15c55e42f376ed06c83174c486f67
|
|
Change-Id: I83f4c55582d1ff8ec6b468776090b453c4f85b89
|
|
Change-Id: I88f6327c12fa2b782aa5c82e5891006713348477
|
|
Now we build only what we really need for 'build' platform - there is
new build-tools make target.
The list of tools is in solenv/gbuild/extensions/pre_BuildTools.mk.
Also similar is done to some extent for 'host' platform using
gb_Module_add_targets_for_build which is ignored for 'host'.
Change-Id: I6acd1762b16aca366aac1a0688500f27869cfca2
|
|
Change-Id: Ib8d2ddb2efaef26a0a2aba36e9e1c67b1fb019d7
|
|
the gb_Postprocess* foo could also be in gb_Module* as it is
conceptionally close ('do things globally/productwide'). OTOH I see no
obvious reason for e.g. signing not to be done right along with building
a lib/executable anyway instead of in postprocess. The same is likely
true for the other stuff too.
Change-Id: I9c8f569564c056643af7ca39bfe038ed228dcd3d
Reviewed-on: https://gerrit.libreoffice.org/2426
Reviewed-by: Matúš Kukan <matus.kukan@gmail.com>
Reviewed-by: Fridrich Strba <fridrich@documentfoundation.org>
Tested-by: Fridrich Strba <fridrich@documentfoundation.org>
|
|
Change-Id: Ia6c938ef0e45ec86c0c202e95ef28d49fccd3497
|
|
It does not work, anyway.
This reverts commit ee18b30b0391a443a51d7b97060f118478c202af.
Change-Id: Ibe21a2a62ecbfaf8dcd87113bde386050ffde25a
|
|
Change-Id: Ie3248563815c2558c573e525e79024ef083345b7
|
|
Change-Id: Ia556cbfda5c5c3d9e1b2b7b1f488b079ce420367
|
|
... because their result is assigned to variables in various gbuild
classes.
Change-Id: I07c0fc54e10ddd76a6246ac4e9ed56541c9c1257
|
|
Change-Id: I325287cba773df848a8797486c149502abae9c9d
|
|
So that now it's possible to do just
'make CppunitTest_sw_macros_test'
instead of
'make /home/llunak/build/src/l2/workdir/unxlngx6/CppunitTest/sw_macros_test.test'
Change-Id: Ibd1e9ef4fc825043a71bd669b2f5c37ffec68e33
Reviewed-on: https://gerrit.libreoffice.org/1253
Reviewed-by: Peter Foley <jpfoley2@gmail.com>
Reviewed-by: David Ostrovsky <David.Ostrovsky@gmx.de>
Reviewed-by: Luboš Luňák <l.lunak@suse.cz>
Tested-by: Luboš Luňák <l.lunak@suse.cz>
|
|
There can be only one precompiled header with some compilers anyway.
Change-Id: I59e84bee2f57f36753b59b7518d701573affac12
|
|
This is mostly a revert of 992757cc5ebfb1df42636b8d3269cdbc10fc4148
and bc78ac68c2492e63a680f68b524c4a51179ea83e, with tweaks to apply.
Also only MSVC support for now.
Change-Id: Ia926b3c28125b52012156ccfd47932837b09d20a
|
|
This finally blew up spectacularly on Windows because we now have
packages delivering external libraries to the solver, where the
libraries are not defined by Library class, which leads to accidental
inheritance of AUXTARGETS variable and failure to deliver nonexistent
UnpackedTarball/.../libpq/postgresql-sdbc-impl.uno.dll
Remove the custom deliver rule that uses AUXTARGETS, and instead
let those files be handled individually by default Package.mk rules.
This requires a dummy touch rule for each file, and unfortunately a
pattern rule cannot be used here because of silly naming differences
between implibs and dlls on Windows...
AUXTARGETS do not cause problems for clean targets because there is no
interdependency there, so keep them there.
Change-Id: I2138f635334ff4517a2d31ef91cfd03a46745fd6
|
|
Just like CXXFLAGS can be controlled using --enable-selective-debuginfo.
Non-debug libraries can be debug-stripped this way, which can save quite
some disk space with those cppunittests which just include all .o's from sw/sc.
Change-Id: Ib947f76f535161e4eef1421a5539df5e25fa4042
|
|
Always link in gb_STDLIBS, except when the library explicitly opts out
with gb_LinkTarget_disable_standard_system_libs.
Change-Id: I489a99114fbfa46d0421a27cf6c7b899dc268a4a
|
|
add a new gb_LinkTarget_use_system_win32_libs to abstract different
linker options on MSVC and GCC.
Change-Id: Ic9bf2545f59bf7871e6fc06b290c486ddfbec03d
|
|
There are currently 3 different mechanisms being used for frameworks,
which is of course intolerable so we invent a 4th one and standardize on
it: gb_LinkTarget_use_darwin_frameworks
(This doesn't mean using add_libs or externals was wrong, it was just
inconsistent... and i don't see an obvious benefit of using externals here)
Change-Id: I5de9020402c87e7236c6a358c47f02fa56642d3d
|
|
... new gb_LinkTarget_add_standard_system_libs
Change-Id: Ib2bc843098db3d8c6822b45a3d21724e67f57d69
|
|
Change-Id: I023c6b35c2a65da566ff5bb3aea772c99c90dda3
|
|
Change-Id: I8bf119a4ab3fc7c2febfa80176358f668003b7d1
|
|
Change-Id: Ic6645e9943b2445ebb37bb99114f777527c69af9
|
|
Change-Id: I5159f3f41c3e0d6ec775a7d1700cbe1272cdb059
|
|
This reverts commit b8f7cc3c26a95b7dfb41415b24f0e2a3a9f1e33c.
This does not work, because there is no dependency between the
*Object .d files and the *Objects, and so the concat-deps races with
the compiler writing the dep files.
Introducing such a dependency would of course be reverting
8b5a984d45005d3df1c89eae897d6e04612625d8
|
|
* in some scenarios gcc (or ccache) generate relative paths
* thus we need to concat the depfiles in the same makerun as the compiles
* stiil not foolproof:
- cd sw && make and abort with Ctrl-C
- cd .. && make sw
- now we have deps from mixed workdirs
* in the end we should likely always cd to SRCDIR before all compiles
Change-Id: I007da96f38ed2cfe403b692400f447764f6d6988
|
|
Change-Id: I532a6822d6ee023d72a71e012c97929f56c7b05a
|
|
|
|
|
|
The previous commit works but results in a ~0.4 second slowdown on
tail_build, which is completely unacceptable; seems that at the cost of
another ugly global variable in Module.mk we can get the performance back
and then some, so refactor this again...
|
|
so it doesn't use wildcard but instead the actually registered
LinkTargets.
|
|
Change-Id: Ic90495f677dc42c0430492548deaa9f9b10e4d44
|
|
the argument given now can
- be 'all', meaning everything
- start with -, meaning not to enable symbols for that item
- end with /, meaning everything in that toplevel directory
- be a target of the form targettype_name, e.g. Library_swui
Note that there is no ordering and removal takes precedence.
|
|
|
|
Naming convention for gbuild methods:
- "add" is used for stuff that is logically a part of the target
(i.e. not registered at the Module, but defined in the target's makefile)
- "use" is used for stuff that is logically a different target
(i.e. it is registered at the Module, has it's own makefile, may be
in a different module than the target)
|
|
|
|
to be used for generated headers which are not delivered to solver.
|
|
The problem is that make 3.81 does not match pattern rules whose target
ends with '/' against directories.
(regression from 408822b54724f8170a1296ddd67e50d8291c8ee1,
90491a073c5b5faee782ad5eab63276fda2342e6)
So use a fake file ".dir" in the respective directory, and try not to
spawn unnecessary mkdir processes by checking for existence via realpath.
This is all quite ugly and should be reverted once support for make 3.81
is dropped.
Also, fix a pre-existing problem of pattern rules with multiple targets
that used to work by accident in Package.mk.
|
|
Rather create new order-only dependencies on directories where
targets should be delivered.
On cygwin this is much faster.
|
|
It is implemented for Executable, Library and StaticLibrary and works
for C, C++, Objective C and Objective C++ sources.
Please, do not misuse! Better fix your code!
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|