Age | Commit message (Collapse) | Author |
|
We are lucky, saxparser(expwrap) is only needed to build localedata* libs
which are only runtime dependencies, and can be built *after* libmerged.
So, simplify and move runtime deps to CppunitTest.
This reverts commits 59cea45ec247df1acb691308c940ff97673e4c48
and partially 5e45637568ceefee21fe329a5254881963fecd7a
Change-Id: I1a1fb4a48fd1a22a9b3a48f0eb0f123649c113b0
|
|
Change-Id: Ied595f97c2a6501a7e80a378a474689c43ad60d9
|
|
It was most probably not working anyway and is something different to
libmerged, which intends to merge only core set of libraries.
To build everything into one library, approach similar to what we do for
Android could be better idea.
Change-Id: Id5b8c2fa2b26e8faaaeb022ea55eb2e39909e875
|
|
Change-Id: I8e36ad989ec54117f85105c24bc1c1442e0a493b
Reviewed-on: https://gerrit.libreoffice.org/15454
Reviewed-by: Andras Timar <andras.timar@collabora.com>
Tested-by: Andras Timar <andras.timar@collabora.com>
|
|
...after f29e959c827a5858e3baeea6e748aa597ac9fb94 "enable -Woverloaded-virtual
gcc warning" had already enabled it for unxgcc.mk a long time ago.
Change-Id: Ic33e026a3d7200c779cec0cd9f23058debc0c74e
|
|
expwrap links against sax and can't link against libmerged because
it is needed to build libmerged :-/ (saxparser in i18npool)
Change-Id: Ie4dca11e8e1bb16a2506cf7cdb7d69b804d98cc0
Reviewed-on: https://gerrit.libreoffice.org/15447
Tested-by: Jenkins <ci@libreoffice.org>
Tested-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
Reviewed-by: Christian Lohmaier <lohmaier+LibreOffice@googlemail.com>
|
|
as those are fake, just textfiles with dll extension that signtool
doesn't like.
Also made the text more descriptive "invalid" is ambiguous,
"invalid - merged lib" states the reason why the file is a dummy.
Change-Id: I31801fd0c3aa593549fac5e6275189e18bbc0010
Reviewed-on: https://gerrit.libreoffice.org/15444
Reviewed-by: Michael Stahl <mstahl@redhat.com>
Tested-by: Jenkins <ci@libreoffice.org>
|
|
Yes, you read that right, if there's no space in front of the first
argument, cl.exe will *ignore* it.
Change-Id: I73d9bf026705538e53b5d91091a4c96eb15739b2
|
|
Change-Id: Ie077fc273601607cba72099d650c6774f08240fb
|
|
Change-Id: I6211762b03bd85a1388027906858020a767f12f6
|
|
Change-Id: Ica21cefb705f4140b9f298e09be04b4950407b78
|
|
Change-Id: I79a357a51a3d98774745001d5f99b045ef964255
Reviewed-on: https://gerrit.libreoffice.org/15218
Reviewed-by: Andras Timar <andras.timar@collabora.com>
Tested-by: Andras Timar <andras.timar@collabora.com>
|
|
Change-Id: Ic4d59228f3295e8400bcacf637ff4def7669b775
|
|
Change-Id: Ibd74c58efa7f1a24de409820655fb98b4fc13df3
|
|
...though that requires switching of C++ -Wmissing-braces for Clang and for old
GCC (new GCC already dropped it from C++ -Wall, see
<https://gcc.gnu.org/bugzilla/show_bug.cgi?id=25137#c17>).
Change-Id: I92acc4c529d634c4190e0ee4b65d1fbe5b1b521f
|
|
- otherwise colored gbuild output is broken, so this is really important
Change-Id: I6affbafe8f7f841a4061e2a42f6c0858b3f4675b
|
|
On a regular build, gbuild print over and over
the 'loaded modules' which is a long landry list
that serve little purpsose but to make it hard
to read the output
If you _really_ want to still see that use verbose=t
Change-Id: I0950f471b68e84c0cbd2c0afd2721394f70da398
|
|
Change-Id: I4fb99c446f1ea0892d9210885d9d93f3dfe24ceb
|
|
Change-Id: I15492fcc687a21b740dc2d97a47485e2e760055f
|
|
|
|
|
|
|
|
Add constructors which are no longer part of sd_component_getFactory()
as of commit 4dafecfe4a72204dd994a3f66c5d886ffcfc965a (sd: convert
sd.component to constructor usage, 2015-02-28).
Change-Id: I209fdd9bc8c8b81cb7e69796a2afb248df856fc9
|
|
Change-Id: Ib1d4746d222c0a59a2ace0a97ce0fb1bac18f4ef
|
|
Change-Id: I992773c84eaf2dff63814e6223323fe99ae5a415
|
|
Change-Id: I7566dcf21aae6b800adb58d8c94e350bb0da720c
|
|
Its code is in the vcl library nowadays.
Change-Id: Idb659e541226724004660102f6641c38a2312c27
|
|
Change-Id: I7df8716e90e9097336919d2a9208a549b8f7d3d0
|
|
Change-Id: Ice8504041f22e00f2e5010813d9dff1d2987c8d6
|
|
Change-Id: Ibaf6d48ad2593664d81d7557fd470f6cdf1d0a4f
|
|
Change-Id: I74dc843b73fac6b4ffe0ece7b9fa7247c06ccfce
|
|
Change-Id: Ib43ba7f5ccf3c8858dc9cd75bb46cfbefd7bc84a
|
|
Change-Id: I6d92bb282ed13ce9a9bb14829b04f9ceae41ae61
|
|
Change-Id: Iede89154d0422425c392b03932b46ac88a5ccb35
|
|
Change-Id: Ife39c5114f946244691c72c9defa8854fdf8a1b6
|
|
To avoid losing functionality (components missing at run-time), we need to add
the corrsponding new-style functions instead of the old-style one that was
removed. In this case it was just one.
(In those cases where one old-style is replaces with several new-style ones,
ideally one can then avoid some of the replacements in the iOS and Android
case, and save on code size.)
Change-Id: I6d4a90e84600debdf62b47408264c439a6f0e312
|
|
Change-Id: Idcdebf7ae7f38a11672eae441af981d385b392c0
|
|
... makes it hard to clean up these problems.
But the warnings on almost every std::copy and std::transform are
obnoxious, so disable them with _SCL_SECURE_NO_WARNINGS
Change-Id: Ieea611175fb16b78ff972d194767d3d192992de5
|
|
Since with the 8.x SDK it is not possible to target Windows XP anyway
according to MSDN, just set _WIN32_WINNT to _WIN32_WINNT_VISTA so that
we can at least work on cleaning up API calls deprecated in SDK 8.1.
Change-Id: Ic433caf454f57187b91ef6fdec3f3bcde3f1af79
|
|
... and remove the definitions in makefiles, except in fpicker, which
defines a different value for the Vista file picker.
The WINVER value is derived automatically from _WIN32_WINNT by SDK
header sdkddkver.h.
Change-Id: I73358a9bf5d070f71821654cae2a29335a754a21
|
|
Change-Id: Ib6d9233e8c322fd7165fce8763bb2a4eee9a7850
|
|
Change-Id: Id45dde81bb699e8c6614b70a815588b884d9a313
|
|
Change-Id: Ib5238d44d7bce6651ddbdc4f843f5aaab4eb97c2
|
|
The sc_subsequent_filters_test was failing because of a lock file
because it did not use the unittest configuration.
Refactor gb_CppunitTest_use_configuration so it uses both the instdir
and unittest configuration to prevent such errors.
In case there ever is a test that does not work with the unittest
configuration it should call gb_CppunitTest_use_instdir_configuration.
Change-Id: Ibc00d42f8b6102d50d922f51173120798fa45c6e
|
|
As discussed in b4f6b26b5a1a78fecfa95ec2eb7ac8b80495d8aa "SAL_DLLPUBLIC_RTTI for
proper RTTI visibility for LLVM," RTTI-based -fsanitize= checks with Clang on
Linux need special precautions to make RTTI symbols visible across DSOs. The
approach taken there, as well as in 598d8194b0ea1a64e0ebba28a86c128bafa57c7c
"Visible function type RTTI for Clang -fsanitize=function," was to add explicit
SAL_DLLPUBLIC_RTTI annontations to relevant type definitions. However, for
-fsanitize=vptr that would have required many more of those, so it appears
easier to "misuse" -fsanitize-ms-compat in that case, which happens to give all
RTTI symbols default visibility (while otherwise still honoring our
SAL_DLLPUBLIC/PRIVATE annotations).
The SAL_DLLPUBLIC_RTTI annotations from 598d8194b0ea1a64e0ebba28a86c128bafa57c7c
"Visible function type RTTI for Clang -fsanitize=function" can likely be removed
again.
Change-Id: Ibeff7ab8c908111a7dc66ff0677204f112b24db8
|
|
Change-Id: Icf872e269c7e427ba1287ccd0082974c9426449e
|
|
Change-Id: I957bf1d5bd88b15539e95dfa4aa87a8e63a96439
|
|
Michael Stahl <mst@openoffice.org> Mar 30 2011
solaris11: #i117606#: solaris.mk: add -Bdirect -z defs to LDFLAGS
Change-Id: Iab16960010810185a3b1f7c617c44db3d9822fb5
Reviewed-on: https://gerrit.libreoffice.org/14579
Tested-by: Jenkins <ci@libreoffice.org>
Tested-by: Michael Stahl <mstahl@redhat.com>
Reviewed-by: Michael Stahl <mstahl@redhat.com>
|
|
With this, ODF_text_reference_v3.odt does not report any broken services, but
the app still crashes anyway - tdf#89642.
Change-Id: If73835534184ffbac1c91b2c5bca33bded50c41f
|
|
Change-Id: I902fcb3fef7100279c34494e95cacfd3e5caf153
|