After 5b42f148e2 "ExternalProject_libxml2 still
needs to build the xmllint executable" (which I had happened to prepare on
macOS, so didn't originally notice the issue addressed here now), Linux builds
like <https://ci.libreoffice.org/job/lo_callgrind_linux/13374/> started to fail
with
> CCLD xmllint
> /opt/rh/devtoolset-11/root/usr/libexec/gcc/x86_64-redhat-linux/11/ld: xmllint-xmllint.o: in function `testSAX':
> /home/tdf/lode/jenkins/workspace/lo_callgrind_linux/workdir/UnpackedTarball/libxml2/xmllint.c:1646: undefined reference to `xmlNewSAXParserCtxt'
> /opt/rh/devtoolset-11/root/usr/libexec/gcc/x86_64-redhat-linux/11/ld: xmllint-xmllint.o: in function `myReallocFunc':
> /home/tdf/lode/jenkins/workspace/lo_callgrind_linux/workdir/UnpackedTarball/libxml2/xmllint.c:357: undefined reference to `xmlMemSize'
> collect2: error: ld returned 1 exit status
> make[2]: *** [Makefile:1007: xmllint] Error 1
> make[1]: *** [/home/tdf/lode/jenkins/workspace/lo_callgrind_linux/external/libxml2/ExternalProject_libxml2.mk:37: /home/tdf/lode/jenkins/workspace/lo_callgrind_linux/workdir/ExternalProject/libxml2/build] Error 1
Turns out that those two functions have been introduced after the previously
used libxml2 2.10.4. No idea how things are supposed to work (given that
workdir/UnpackedTarball/libxml2/libxml2.syms starts off with "Retained for
backward compatibility. Don't add new symbols.", it appears that there should be
some other mechanism at play to make xmllint on Linux find the libxml2.so
symbols it needs, but which doesn't work as intended in our build), but just add
those two symbols to libxml2.syms for now. (With a new version named
LIBXML2_2.11.4, even if the symbols can actually have been introduced in some
other version between 2.10.4 and 2.11.4, but who cares.)
Change-Id: Ib9d5c7901fe94b8014a87b049fc27ef5658fd954
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/152292
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
No idea why fd6cbd983a "upgrade to libxml2-2.11.4"
started to only explicitly make libxml2.la (the commit message doesn't say), but
that caused at least <https://ci.libreoffice.org/job/lo_ubsan/2788/> to fail
with
> make[1]: *** Deleting file '/home/tdf/lode/jenkins/workspace/lo_ubsan/workdir/CustomTarget/sfx2/classification/example.validated'
> [build VAL] CustomTarget/sfx2/classification/example.validated
> /bin/sh: /home/tdf/lode/jenkins/workspace/lo_ubsan/workdir/UnpackedTarball/libxml2//xmllint: No such file or directory
> make[1]: *** [/home/tdf/lode/jenkins/workspace/lo_ubsan/sfx2/CustomTarget_classification.mk:20: /home/tdf/lode/jenkins/workspace/lo_ubsan/workdir/CustomTarget/sfx2/classification/example.validated] Error 1
Change-Id: I7e7aa2bae05cf8674720cfd72c78840cd7a28919
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/152256
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
This is basically ea68de2968 but
for LDFLAGS. A number of external libs cannot use this because
their libtool mishandles -fuse-ld.
Change-Id: Idee379eb0a3afb475b536519ee3de064b4e218f4
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/133639
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
No idea why we just provided the platform flags when cross-
compiling. In the curious case, where the host platform is
detected as x86_64-pc-mingw32 per default and we actually
want to override it with x86_64-pc-cygwin, we don't do a
cross compile, but must override the host platform.
But there is additional special handling needed for the omitted
cross-platform build in the special case of --host=i686-pc-cygwin
and --build=x86_64-pc-cygwin, where we deliberatly ignore cross
building; Windows is already a slow build, so try to keep this
optimization (AMD64 can execute x86 binaries).
There is the theoretical case, where the externals config.guess
would have detected something else and that "magically" even
worked, while the LO detected triplet would fail, but this
should be fixed in the external in any way.
Change-Id: Ib7a9719e0e406fe90334b7611dc3f01b51692bfa
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/129153
Tested-by: Jenkins
Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
When building a static LO with --disable-dynloading on Linux,
--without-system-libs failed for me. And it left me really puzzled:
raptor configure failed and claimed it couldn't link libxml2.
raptor's config.log showed missing math functions. xml2-config of
LO's build is patched and it includes a -lm. The xml2-config in my
chroot doesn't. But we explicitly pass the xml2-config for
non-system-libxml2 build. Reading the configure from raptor didn't
reveal a way, that it could somehow pick up the xml2-config from
the chroot, but that code is autoconf-complex...
When running "sh -x configure", it turned out the configure script
actually picks up the LIBXML_* flags from the environment, which
are set by LO's config_host.mk. These just add -lm for Android.
So this adds a xml2-config.in "dummy", which overwrites the one
from the libxml2 source and just echos LO's LIBXML_* values and
it adds -lm for all DISABLE_DYNLOADING targets.
Change-Id: Ia713cf80c8e7dc989cf23c224e7a0f7ea1210a87
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/116409
Tested-by: Jenkins
Reviewed-by: Jan-Marek Glogowski <glogow@fbihome.de>
See instructions in solenv/gbuild/Trace.mk . This generates a file than
can be viewed e.g. in the Chromium tracing view.
Change-Id: I5f90647c58ca729375525b6daed2d4918adc8188
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/88754
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
... which is, surprisingly enough, required to build the latest libxslt.
Change-Id: Ifbb36ed61b8f68185f9c788f63a8edeb58899f94
Reviewed-on: https://gerrit.libreoffice.org/83311
Tested-by: Jenkins
Reviewed-by: Michael Stahl <michael.stahl@cib.de>
...that had been added with fcb2d8a87a
"external/libxml2: Avoid UBSan nullptr-with-offset"
Change-Id: I7ee234c8c6a868ed825a8ed3fa0dcdc69decb7ba
Reviewed-on: https://gerrit.libreoffice.org/81299
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
...(new with Clang 10 trunk), where adding even an offset of 0 to a null pointer
is UB in C. Seen when building UIConfig_modules/schart:
> [UIL] chart2/uiconfig/ui/3dviewdialog
> xpath.c:14532:5: runtime error: applying zero offset to null pointer
> #0 in xmlXPathTryStreamCompile at workdir/UnpackedTarball/libxml2/xpath.c:14532:5
> #1 in xmlXPathCtxtCompile__internal_alias at workdir/UnpackedTarball/libxml2/xpath.c:14634:12
> #2 in xsltXPathCompileFlags at workdir/UnpackedTarball/libxslt/libxslt/xsltutils.c:2323:11
> #3 in xsltValueOfComp at workdir/UnpackedTarball/libxslt/libxslt/preproc.c:1258:18
> #4 in xsltStylePreCompute at workdir/UnpackedTarball/libxslt/libxslt/preproc.c:2225:6
> #5 in xsltParseTemplateContent at workdir/UnpackedTarball/libxslt/libxslt/xslt.c:4916:13
> #6 in xsltParseStylesheetTemplate at workdir/UnpackedTarball/libxslt/libxslt/xslt.c:5467:5
> #7 in xsltParseStylesheetTop at workdir/UnpackedTarball/libxslt/libxslt/xslt.c:6205:6
> #8 in xsltParseStylesheetProcess at workdir/UnpackedTarball/libxslt/libxslt/xslt.c:6461:2
> #9 in xsltParseStylesheetImportedDoc at workdir/UnpackedTarball/libxslt/libxslt/xslt.c:6675:9
> #10 in xsltParseStylesheetDoc at workdir/UnpackedTarball/libxslt/libxslt/xslt.c:6714:11
> #11 in main at workdir/UnpackedTarball/libxslt/xsltproc/xsltproc.c:888:9
Change-Id: I016ca8d24315385bcfeafca56dda44d9be10f517
Reviewed-on: https://gerrit.libreoffice.org/81285
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
* fixes CVE-2018-14404
* drop one hunk from libxml2-android.patch that was added in commit
6a17d2f2ba7acfec277314b97b50e41532d6b44d; presumably nan() exists now
given that other code is calling it.
Change-Id: I696cc4e1da55536ea1c89a6e0446ce5bc8398ba4
Reviewed-on: https://gerrit.libreoffice.org/66308
Tested-by: Jenkins
Reviewed-by: Michael Stahl <Michael.Stahl@cib.de>
In OOo times, there'd originally been efforts to allow building on Windows with
MinGW. Later, in LO times, this has been shifted to an attempt of cross-
compiling for Windows on Linux. That attempt can be considered abandoned, and
the relevant code rotting.
Due to this heritage, there are now three kinds of MinGW-specific code in LO:
* Code from the original OOo native Windows effort that is no longer relevant
for the LO cross-compilation effort, but has never been removed properly.
* Code from the original OOo native Windows effort that is re-purposed for the
LO cross-compilation effort.
* Code that has been added specifially for the LO cross-compilation effort.
All three kinds of code are removed.
(An unrelated, remaining use of MinGW is for --enable-build-unowinreg, utilizing
--with-mingw-cross-compiler, MINGWCXX, and MINGWSTRIP.)
Change-Id: I49daad8669b4cbe49fa923050c4a4a6ff7dda568
Reviewed-on: https://gerrit.libreoffice.org/34127
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
When Windows build is executed, cscript is used to execute JavaScript files.
This uses cscript from the system to execute *.js files. cscript is not
only capable of executing JavaScript, but also VBScript. Which engine to
run is usually determined by the file extension, except when any installed
program has added a registered association to the used file type. In that
case, the execution of cscript and thus the build fails. This can be prevented
by directly defining the script engine when calling cscript, using the
/e:javascript parameter for *.js targets.
Change-Id: If717b8ae5335acbe4f11c269d3c98a7247a135e6
Reviewed-on: https://gerrit.libreoffice.org/21717
Reviewed-by: Thorsten Behrens <Thorsten.Behrens@CIB.de>
Tested-by: Thorsten Behrens <Thorsten.Behrens@CIB.de>
- drop libxml2-freebsd.patch.1 (upstream libtool 2.4.6 does the same)
- drop libxml2-vc15.patch (fixed upstream)
Change-Id: Ia2f194f39efebd3d2ea924d23a5543ac53e93116
Reviewed-on: https://gerrit.libreoffice.org/20084
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Michael Stahl <mstahl@redhat.com>
...assuming the delayLoadHook in cli_ure/source/native/native_bootstrap.cxx is
no longer necessary and loading of cppuhelper from the program dir cannot fail
regardless in whatever scenario the cli_cppuhelper library itself is loaded.
Change-Id: I13f32b327bca4cce9780864f5e57cdad3860afe5
Caolán McNamara