Re: Error Compiling LibreOffice Core on Debian 12 - x64

[Rene Engelhard]

Hi again,

sorry, missed a part of a sentence.

Am 04.04.24 um 18:13 schrieb Rene Engelhard:

This is it. gcc and/or ld got killed from the outside.


...by the OOM killer I guess.

Regards,

Rene

Re: Error Compiling LibreOffice Core on Debian 12 - x64

[Rene Engelhard]

Hi,


well..

Am 03.04.24 um 19:45 schrieb Andreas Mantke:

On 2024-03-24 00:17, Andreas Mantke wrote:

I try to compile LibreOffice core on Debian 12 -x64 but the process
ended with an error.

Here the messages in the shell at the end of the build process:

[MOD] unoxml
/builddir/core/filter/source/graphicfilter/icgm/cgmtypes.hxx:108:
warning: type ‘LineType’ violates the C++ One Definition Rule [-Wodr]
   108 | enum LineType   { LT_SOLID = 1, LT_DASH = 2, LT_DOT =
3, LT_DASHDOT = 4, LT_DASHDOTDOT = 5, // Standard
   |
/builddir/core/vcl/inc/regband.hxx:47: note: an enum with different
value name is defined in another translation unit
    47 | enum class LineType { Ascending, Descending };
   |
/builddir/core/filter/source/graphicfilter/icgm/cgmtypes.hxx:108: note:
name ‘LT_SOLID’ differs from name ‘Ascending’ defined in another
translation unit
   108 | enum LineType   { LT_SOLID = 1, LT_DASH = 2, LT_DOT =
3, LT_DASHDOT = 4, LT_DASHDOTDOT = 5, // Standard
   |
/builddir/core/vcl/inc/regband.hxx:47: note: mismatching definition
    47 | enum class LineType { Ascending, Descending };
   |

is only a warning.

g++: fatal error: Getötet signal terminated program lto1
compilation terminated.
lto-wrapper: fatal error: /usr/bin/g++ returned 1 exit status
compilation terminated.
/usr/bin/ld: error: lto-wrapper failed
collect2: error: ld returned 1 exit status


This is it. gcc and/or ld got killed from the outside.

Then there are some options which didn't work my git clone on Debian 12:

'--enable-mergelibs' and '--enable-lto'. Maybe also something to
investigate more in depth once I had some free cycles again.


And that is because of this. mergelibs is too big to link on 32bit.

That is a architectural limitation of 32bit archs which is not going to 
go away. Even if you stick 64GB in there one process can just address 4GB.



If I disable all this options above I could successfully build
LibreOffice again.


Unsurprisingly.


Regards,


Rene

Re: Error Compiling LibreOffice Core on Debian 12 - x64

[Andreas Mantke]

Hi Michael, hi all,

excuse for the late answer, but I had to made a lot of edits and tests
to get to a successful build and later to a successful Docker image. In
addition the build inside a virtual machine took a lot of time for every
new test circle.

Am 25.03.24 um 09:51 schrieb Michael Weghorn:

On 2024-03-24 00:17, Andreas Mantke wrote:

I try to compile LibreOffice core on Debian 12 -x64 but the process
ended with an error.

Here the messages in the shell at the end of the build process:

[MOD] unoxml
/builddir/core/filter/source/graphicfilter/icgm/cgmtypes.hxx:108:
warning: type ‘LineType’ violates the C++ One Definition Rule [-Wodr]
   108 | enum LineType   { LT_SOLID = 1, LT_DASH = 2, LT_DOT =
3, LT_DASHDOT = 4, LT_DASHDOTDOT = 5, // Standard
   |
/builddir/core/vcl/inc/regband.hxx:47: note: an enum with different
value name is defined in another translation unit
    47 | enum class LineType { Ascending, Descending };
   |
/builddir/core/filter/source/graphicfilter/icgm/cgmtypes.hxx:108: note:
name ‘LT_SOLID’ differs from name ‘Ascending’ defined in another
translation unit
   108 | enum LineType   { LT_SOLID = 1, LT_DASH = 2, LT_DOT =
3, LT_DASHDOT = 4, LT_DASHDOTDOT = 5, // Standard
   |
/builddir/core/vcl/inc/regband.hxx:47: note: mismatching definition
    47 | enum class LineType { Ascending, Descending };
   |
g++: fatal error: Getötet signal terminated program lto1
compilation terminated.
lto-wrapper: fatal error: /usr/bin/g++ returned 1 exit status
compilation terminated.
/usr/bin/ld: error: lto-wrapper failed
collect2: error: ld returned 1 exit status
make[1]: *** [/builddir/core/Library_merged.mk:11:
/builddir/core/instdir/program/libmergedlo.so] Fehler 1
make: *** [Makefile:290: build] Fehler 2


I tried in parallel with a build on openSUSE 15.5 -x64 and the build of
LibreOffice  core was successful.

Any hints about a workaround or fix?


Renaming one of the enums could be one approach, e.g. the one defined
in vcl/inc/regband.hxx to something like "RegionBandLineType".  (Just
a first thought when skimming over the file, I don't otherwise know
the involved code.)


I made this edits inside the git clone on the Debian 12 -x64 install and
that fixed that problem. But I got another one and the build was again
not successful.

Thus I created a new Debian 12 environment on Virtualbox and started
again from scratch. I installed the packages mentioned on our wiki page
for building on Debian.
This time the build process finished successful after some hours. I used
no special options for this build process. I run ./autogen.sh and then
make, without any changes on the source code.

I then tested a build with the option '--with-distro=LibreOfficeOnline'
but that didn't work.
I also tried to build with the option '--with-distro=CPLinux-LOKit', but
that gave some errors.
I had to evaluate every option in the latter configuration file
(proposed from Collabora Productivity for a build) and see if the build
got done. I stumbled over some build options that were not available in
LibreOffice core. It seemed they were only implemented in a fork of
LibreOffice but never made it upstream, e.g. '--enable-noto-font',
'--disable-ext-ct2n'. Maybe something to dig into a bit deeper.

Then there are some options which didn't work my git clone on Debian 12:
'--enable-cairo-rgba', '--without-system-cairo', '--disable-gui',
'--enable-mergelibs'  and '--enable-lto'. Maybe also something to
investigate more in depth once I had some free cycles again.

If I disable all this options above I could successfully build
LibreOffice again.

Regards,
Andreas



--
## Free Software Advocate
## Free Online Office (LibreOffice Technology):
## https://github.com/freeonlineoffice/online
## Plone add-on developer
## My blog: http://www.amantke.de/blog

Re: Error Compiling LibreOffice Core on Debian 12 - x64

[Tomaž Vajngerl]

Hi,

On Mon, Mar 25, 2024 at 5:51 PM Michael Weghorn  wrote:
> Renaming one of the enums could be one approach, e.g. the one defined in
> vcl/inc/regband.hxx to something like "RegionBandLineType".  (Just a
> first thought when skimming over the file, I don't otherwise know the
> involved code.)

The types in cgmtypes.hxx could also be changed to enum class and
added into a namespace to prevent conflicts as those are all local to
CGM filter anyway.

Tomaž

Re: Error Compiling LibreOffice Core on Debian 12 - x64

[Michael Weghorn]

On 2024-03-24 00:17, Andreas Mantke wrote:

I try to compile LibreOffice core on Debian 12 -x64 but the process
ended with an error.

Here the messages in the shell at the end of the build process:

[MOD] unoxml
/builddir/core/filter/source/graphicfilter/icgm/cgmtypes.hxx:108:
warning: type ‘LineType’ violates the C++ One Definition Rule [-Wodr]
   108 | enum LineType   { LT_SOLID = 1, LT_DASH = 2, LT_DOT =
3, LT_DASHDOT = 4, LT_DASHDOTDOT = 5, // Standard
   |
/builddir/core/vcl/inc/regband.hxx:47: note: an enum with different
value name is defined in another translation unit
    47 | enum class LineType { Ascending, Descending };
   |
/builddir/core/filter/source/graphicfilter/icgm/cgmtypes.hxx:108: note:
name ‘LT_SOLID’ differs from name ‘Ascending’ defined in another
translation unit
   108 | enum LineType   { LT_SOLID = 1, LT_DASH = 2, LT_DOT =
3, LT_DASHDOT = 4, LT_DASHDOTDOT = 5, // Standard
   |
/builddir/core/vcl/inc/regband.hxx:47: note: mismatching definition
    47 | enum class LineType { Ascending, Descending };
   |
g++: fatal error: Getötet signal terminated program lto1
compilation terminated.
lto-wrapper: fatal error: /usr/bin/g++ returned 1 exit status
compilation terminated.
/usr/bin/ld: error: lto-wrapper failed
collect2: error: ld returned 1 exit status
make[1]: *** [/builddir/core/Library_merged.mk:11:
/builddir/core/instdir/program/libmergedlo.so] Fehler 1
make: *** [Makefile:290: build] Fehler 2


I tried in parallel with a build on openSUSE 15.5 -x64 and the build of
LibreOffice  core was successful.

Any hints about a workaround or fix?


Renaming one of the enums could be one approach, e.g. the one defined in 
vcl/inc/regband.hxx to something like "RegionBandLineType".  (Just a 
first thought when skimming over the file, I don't otherwise know the 
involved code.)


OpenPGP_signature.asc
Description: OpenPGP digital signature

Error Compiling LibreOffice Core on Debian 12 - x64

[Andreas Mantke]

Hello,

I try to compile LibreOffice core on Debian 12 -x64 but the process
ended with an error.

Here the messages in the shell at the end of the build process:

[MOD] unoxml
/builddir/core/filter/source/graphicfilter/icgm/cgmtypes.hxx:108:
warning: type ‘LineType’ violates the C++ One Definition Rule [-Wodr]
  108 | enum LineType   { LT_SOLID = 1, LT_DASH = 2, LT_DOT =
3, LT_DASHDOT = 4, LT_DASHDOTDOT = 5, // Standard
  |
/builddir/core/vcl/inc/regband.hxx:47: note: an enum with different
value name is defined in another translation unit
   47 | enum class LineType { Ascending, Descending };
  |
/builddir/core/filter/source/graphicfilter/icgm/cgmtypes.hxx:108: note:
name ‘LT_SOLID’ differs from name ‘Ascending’ defined in another
translation unit
  108 | enum LineType   { LT_SOLID = 1, LT_DASH = 2, LT_DOT =
3, LT_DASHDOT = 4, LT_DASHDOTDOT = 5, // Standard
  |
/builddir/core/vcl/inc/regband.hxx:47: note: mismatching definition
   47 | enum class LineType { Ascending, Descending };
  |
g++: fatal error: Getötet signal terminated program lto1
compilation terminated.
lto-wrapper: fatal error: /usr/bin/g++ returned 1 exit status
compilation terminated.
/usr/bin/ld: error: lto-wrapper failed
collect2: error: ld returned 1 exit status
make[1]: *** [/builddir/core/Library_merged.mk:11:
/builddir/core/instdir/program/libmergedlo.so] Fehler 1
make: *** [Makefile:290: build] Fehler 2


I tried in parallel with a build on openSUSE 15.5 -x64 and the build of
LibreOffice  core was successful.

Any hints about a workaround or fix?

Regards,
Andreas

--
## Free Software Advocate
## Free Online Office (LibreOffice Technology):
## https://github.com/freeonlineoffice/online
## Plone add-on developer
## My blog: http://www.amantke.de/blog

Re: compiling libreoffice branch distro/collabora/co-2021

[Miklos Vajna]

Hi Danilo,

On Mon, Oct 18, 2021 at 01:10:54PM +0200, Danilo Tomasoni 
 wrote:
> Actually I tried first with master, that exited almost immediately with 
> another error.

What was that error exactly?

> Then I tried this branch that compiled for a long time before giving the 
> error I mentioned.
> 
> Any clue? I'm not cross-compiling. I'm using an aarch64 machine, actually an 
> odroid C2 with his vendor-shipped ubuntu.

There are CODE aarch64 packages at
https://www.collaboraoffice.com/code/linux-packages/, so this is meant
to work. If you have problems, probably best to ask at
https://forum.collaboraonline.com/.

Regards,

Miklos

Re: compiling libreoffice branch distro/collabora/co-2021

[Danilo Tomasoni]

Hello Miklos,
Thank you for your message.
Actually I tried first with master, that exited almost immediately with another 
error.
Then I tried this branch that compiled for a long time before giving the error 
I mentioned.

Any clue? I'm not cross-compiling. I'm using an aarch64 machine, actually an 
odroid C2 with his vendor-shipped ubuntu.


Il 18 ottobre 2021 08:48:09 CEST, Miklos Vajna  ha 
scritto:
>Hi Danilo,
>
>On Thu, Oct 14, 2021 at 07:20:26PM +0200, Danilo Tomasoni 
> wrote:
>> I am trying to build collabora online for aarch64.
>> 
>> I saw that I need to first compile libreoffice branch
>> distro/collabora/co-2021
>> 
>> but when I try to do it I get the following error. Anyone can tell me how
>> can I fix this? Thank you
>
>Does the master branch compile for you with the same autogen.input
>switches?
>
>Regards,
>
>Miklos

Re: compiling libreoffice branch distro/collabora/co-2021

[Miklos Vajna]

Hi Danilo,

On Thu, Oct 14, 2021 at 07:20:26PM +0200, Danilo Tomasoni 
 wrote:
> I am trying to build collabora online for aarch64.
> 
> I saw that I need to first compile libreoffice branch
> distro/collabora/co-2021
> 
> but when I try to do it I get the following error. Anyone can tell me how
> can I fix this? Thank you

Does the master branch compile for you with the same autogen.input
switches?

Regards,

Miklos

compiling libreoffice branch distro/collabora/co-2021

[Danilo Tomasoni]

Hello all,

I am trying to build collabora online for aarch64.

I saw that I need to first compile libreoffice branch 
distro/collabora/co-2021


but when I try to do it I get the following error. Anyone can tell me 
how can I fix this? Thank you


Danilo

---

CMP] filter/source/config/cache/filterconfig1
[MOD] fpicker
[MOD] framework
[ECH] CustomTarget/instsetoo_native/setup/setuprc
[ECH] CustomTarget/instsetoo_native/setup/versionrc
[PRL] CustomTarget/postprocess/images/sorted.lst
[PRL] CustomTarget/postprocess/images/commandimagelist.ilst
[LNK] Library/libunotest.so
Skipping non-existent 
/storage/hardisk/collaboraonline/libreoffice/instdir/share/config/soffice.cfg/modules/BasicIDE/toolbar/fullscreenbar.xml


Skipping non-existent 
/storage/hardisk/collaboraonline/libreoffice/instdir/share/config/soffice.cfg/modules/BasicIDE/toolbar/insertcontrolsbar.xml


Skipping non-existent 
/storage/hardisk/collaboraonline/libreoffice/instdir/share/config/soffice.cfg/modules/BasicIDE/toolbar/formcontrolsbar.xml


Skipping non-existent 
/storage/hardisk/collaboraonline/libreoffice/instdir/share/config/soffice.cfg/modules/BasicIDE/toolbar/macrobar.xml


Skipping non-existent 
/storage/hardisk/collaboraonline/libreoffice/instdir/share/config/soffice.cfg/modules/BasicIDE/toolbar/standardbar.xml


[LOC] top level modules: libreoffice
[LNK] Library/libsfxlo.so
[CMP] connectivity/source/drivers/calc/calc
[CMP] connectivity/source/drivers/dbase/dbase
[CMP] connectivity/source/drivers/flat/flat
[CMP] connectivity/source/drivers/writer/writer
[LNK] Executable/mork_helper
[MOD] cppcanvas
/storage/hardisk/collaboraonline/libreoffice/workdir/CxxObject/unotest/source/cpp/bootstrapfixturebase.o: 
In function `test::BootstrapFixtureBase::tearDown()':
bootstrapfixturebase.cxx:(.text+0x0): undefined reference to 
`StarBASIC::DetachAllDocBasicItems()'

collect2: error: ld returned 1 exit status
/storage/hardisk/collaboraonline/libreoffice/solenv/gbuild/LinkTarget.mk:667: 
recipe for target 
'/storage/hardisk/collaboraonline/libreoffice/workdir/LinkTarget/Library/libunotest.so' 
failed
make[1]: *** 
[/storage/hardisk/collaboraonline/libreoffice/workdir/LinkTarget/Library/libunotest.so] 
Error 1

make[1]: *** Waiting for unfinished jobs
Skipping non-existent 
/storage/hardisk/collaboraonline/libreoffice/instdir/share/config/soffice.cfg/modules/BasicIDE/toolbar/fullscreenbar.xml


Skipping non-existent 
/storage/hardisk/collaboraonline/libreoffice/instdir/share/config/soffice.cfg/modules/BasicIDE/toolbar/insertcontrolsbar.xml


Skipping non-existent 
/storage/hardisk/collaboraonline/libreoffice/instdir/share/config/soffice.cfg/modules/BasicIDE/toolbar/formcontrolsbar.xml


Skipping non-existent 
/storage/hardisk/collaboraonline/libreoffice/instdir/share/config/soffice.cfg/modules/BasicIDE/toolbar/macrobar.xml


Skipping non-existent 
/storage/hardisk/collaboraonline/libreoffice/instdir/share/config/soffice.cfg/modules/BasicIDE/toolbar/standardbar.xml


Skipping non-existent 
/storage/hardisk/collaboraonline/libreoffice/instdir/share/config/soffice.cfg/modules/BasicIDE/menubar/menubar.xml


Makefile:282: recipe for target 'build' failed
make: *** [build] Error 2

---

--
This letter is printed from 100% recycled electrons.
https://danilo.tomasoni.me

0x4E3D0186.asc
Description: application/pgp-keys

Re: Linker error when compiling libreoffice core.

[Michael Stahl]

On 08.09.21 10:26, Caolán McNamara wrote:

On Tue, 2021-09-07 at 16:41 +0200, Mysaa Java wrote:

/sources/libreoffice/ext/core/workdir/CxxObject/writerfilter/source/rtf
tok/rtfdocumentimpl.o: in function
`writerfilter::rtftok::RTFDocumentImpl::handleEmbeddedObject()':
rtfdocumentimpl.cxx:(.text+0x86bc): undefined reference to `non-virtual
thunk to cppu::ImplInheritanceHelper::acquire()'
collect2: error: ld returned 1 exit status
make: ***
[/sources/libreoffice/ext/core/writerfilter/Library_writerfilter.mk:10:
/sources/libreoffice/ext/core/instdir/program/libwriterfilterlo.so]
Error 1
```


I haven't had this exact error before, but we have seen similar in the
past. What's probably relevant is the version of your compiler
(gcc --version). The last time we saw this the workaround deployed was
https://cgit.freedesktop.org/libreoffice/core/commit/?id=c820cb429fc3d3eec9bf6596f0a34995cbd93982

Trying similar for OSeekableInputStreamWrapper might help.


probably it's a bug in the compiler, you could try to upgrade the 
compiler or try if it works if you remove --enable-mergelibs.

Re: Linker error when compiling libreoffice core.

[Caolán McNamara]

On Tue, 2021-09-07 at 16:41 +0200, Mysaa Java wrote:
> /sources/libreoffice/ext/core/workdir/CxxObject/writerfilter/source/rtf
> tok/rtfdocumentimpl.o: in function
> `writerfilter::rtftok::RTFDocumentImpl::handleEmbeddedObject()':
> rtfdocumentimpl.cxx:(.text+0x86bc): undefined reference to `non-virtual
> thunk to cppu::ImplInheritanceHelper com::sun::star::io::XSeekable>::acquire()'
> collect2: error: ld returned 1 exit status
> make: ***
> [/sources/libreoffice/ext/core/writerfilter/Library_writerfilter.mk:10:
> /sources/libreoffice/ext/core/instdir/program/libwriterfilterlo.so]
> Error 1
> ```

I haven't had this exact error before, but we have seen similar in the
past. What's probably relevant is the version of your compiler
(gcc --version). The last time we saw this the workaround deployed was
https://cgit.freedesktop.org/libreoffice/core/commit/?id=c820cb429fc3d3eec9bf6596f0a34995cbd93982

Trying similar for OSeekableInputStreamWrapper might help.

> Steps to Reproduce:
> 1. clone the core repository of libreoffice
> 2. Autogen: ./autogen.sh --without-system-libxml --with-
> distro=LibreOfficeOnline
> 3. Make: make

As an aside, I don't think the linker error is due to the configure
flags used here, but --with-distro=LibreOfficeOnline is an unusual
config choice to pick (and it already includes --without-system-libxml)
and will result in a libreoffice build that doesn't have a gui so if it
does build launching it won't give you something you can interact with.

Linker error when compiling libreoffice core.

[Mysaa Java]

Hello, i hope i am at the right place to report this, i already tried on 
bugzilla and i have been sent here.

I want to compile libreoffice online on my raspberry pi 4 (armv7l) on a custom 
LFS OS. According to the INSTALL file, i have first to build libreoffice core.

But during the `make` command, I get a linker error of type `undefined 
reference to 'non-virtual thunk to ...'`

Here is the complete log of my command:
```
bash-5.0$ make -j 1 -rs -f /sources/libreoffice/ext/core/Makefile.gbuild build
[DEP] LNK:Library/libchartcontrollerlo.so
[LNK] Library/libchartcontrollerlo.so
[CMP] chart2/source/controller/chartcontroller
[BIN] chart2
[MOD] chart2
[DEP] LNK:Library/libcuilo.so
[LNK] Library/libcuilo.so
[CMP] cui/util/cui
[BIN] cui
[MOD] cui
[DEP] LNK:Library/libwriterfilterlo.so
[LNK] Library/libwriterfilterlo.so
/usr/bin/ld: 
/sources/libreoffice/ext/core/workdir/CxxObject/writerfilter/source/rtftok/rtfdocumentimpl.o:
 in function `writerfilter::rtftok::RTFDocumentImpl::handleEmbeddedObject()':
rtfdocumentimpl.cxx:(.text+0x86bc): undefined reference to `non-virtual thunk to 
cppu::ImplInheritanceHelper::acquire()'
collect2: error: ld returned 1 exit status
make: *** 
[/sources/libreoffice/ext/core/writerfilter/Library_writerfilter.mk:10: 
/sources/libreoffice/ext/core/instdir/program/libwriterfilterlo.so] Error 1
```

Steps to Reproduce:
1. clone the core repository of libreoffice
2. Autogen: ./autogen.sh --without-system-libxml --with-distro=LibreOfficeOnline
3. Make: make

Here is the head of ./config.log

./configure --without-system-libxml --disable-cups --disable-dbus 
--disable-dconf --disable-epm --disable-evolution2 --disable-ext-nlpsolver 
--disable-ext-wiki-publisher --disable-firebird-sdbc --disable-gio 
--disable-gstreamer-1-0 --disable-gtk3 --disable-gui --disable-qt5 
--disable-kf5 --disable-odk --disable-online-update --disable-skia 
--enable-pdfimport --disable-postgresql-sdbc --disable-report-builder 
--disable-scripting-beanshell --disable-scripting-javascript --disable-sdremote 
--disable-sdremote-bluetooth --enable-extension-integration --enable-mergelibs 
--enable-python=internal --enable-release-build 
--with-external-dict-dir=/usr/share/hunspell 
--with-external-hyph-dir=/usr/share/hyphen 
--with-external-thes-dir=/usr/share/mythes --with-fonts --with-galleries=no 
--with-lang=ALL --with-linker-hash-style=both --with-system-dicts 
--with-system-zlib --with-theme=colibre --without-branding --without-help 
--without-java --without-junit --with-myspell-dicts --without-package-format 
--without-system-cairo --without-system-jars --without-system-jpeg 
--without-system-libpng --without-system-libxml --without-system-openssl 
--disable-poppler --without-system-postgresql 
--srcdir=/sources/libreoffice/ext/core --enable-option-checking=fata

I also added 
`/sources/libreoffice/ext/core/workdir/UnpackedTarball/libxml2/.libs/` to my 
LD_LIBRARY_PATH to solve another compilation error.

As i am totally unfamilliar with UNO API and with C++ programming, i don't 
think i can do a lot.


Thanks in advance for your time.


Mysaa Java

Re: compiling Libreoffice online android doesn't display document.

[Michael Weghorn]

Hi Miklos and Chetan,

On 17/09/2020 09.07, Miklos Vajna wrote:
> Hi Michael,
> 
> On Wed, Sep 16, 2020 at 01:12:19PM +0200, Michael Weghorn 
>  wrote:
>> (Maybe somebody else can say more on whether including all services into
>> liblo-native-code that are required to support all file formats that
>> LibreOffice supports is fine or would be problematic, e.g. for APK
>> size,...).
> 
> My understanding is that there is no automated way to find out which
> services are needed to support import of exotic file formats, so
> addition / removal of such UNO service ctors have to be considered on a
> case by case basis.
> 
> It's always a good idea to document (in the commit message) what is
> fixed when adding a new one. Removing ctors reduces code size.

thanks for the explanation.

Chetan, I'd personally suggest you start with a simple ODT file and once
your self-built Android app works in general, continue with more
difficult tasks, like looking why specific files are not supported yet.

Regards,
Michael
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: compiling Libreoffice online android doesn't display document.

[Miklos Vajna]

Hi Michael,

On Wed, Sep 16, 2020 at 01:12:19PM +0200, Michael Weghorn  
wrote:
> (Maybe somebody else can say more on whether including all services into
> liblo-native-code that are required to support all file formats that
> LibreOffice supports is fine or would be problematic, e.g. for APK
> size,...).

My understanding is that there is no automated way to find out which
services are needed to support import of exotic file formats, so
addition / removal of such UNO service ctors have to be considered on a
case by case basis.

It's always a good idea to document (in the commit message) what is
fixed when adding a new one. Removing ctors reduces code size.

Regards,

Miklos
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: compiling Libreoffice online android doesn't display document.

[Michael Weghorn]

Hi,

On 14/09/2020 20.12, Chetan Rawat wrote:
> Hi,
> I have been doing my analysis on this  for a while now and after enabling
> few debugging symbols the adb logcat gives following output when opening any
> document:
> 
> 2020-09-14 23:20:44.807 18128-18307/org.libreoffice.androidapp.debug
> D/loolwsd: DBG Mobile-18307 2020-09-14 17:50:44.807636 [ kit_001 ] DBG 
> Calling
> lokit::documentLoad(file:///data/user/0/org.libreoffice.androidapp.debug/cache/LibreOffice2801578603109353115.tmp,
> "Language=en-IN,DeviceFormFactor=mobile").|
> /opt/lo_new/online/kit/Kit.cpp:1171
> 2020-09-14 23:20:45.038 18128-18307/org.libreoffice.androidapp.debug
> W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
> "com_sun_star_comp_Writer_MWAWImportFilter_get_implementation"
> 2020-09-14 23:20:45.042 18128-18307/org.libreoffice.androidapp.debug
> W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
> "org_libreoffice_comp_Writer_EBookImportFilter_get_implementation"
> 2020-09-14 23:20:45.043 18128-18307/org.libreoffice.androidapp.debug
> W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
> "org_libreoffice_comp_Writer_EBookImportFilter_get_implementation"
> 2020-09-14 23:20:45.044 18128-18307/org.libreoffice.androidapp.debug
> W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
> "com_sun_star_comp_Writer_MWAWImportFilter_get_implementation"
> 2020-09-14 23:20:45.046 18128-18307/org.libreoffice.androidapp.debug
> I/chatty: uid=10365(org.libreoffice.androidapp.debug) kit_001 identical 2
> lines
> 2020-09-14 23:20:45.047 18128-18307/org.libreoffice.androidapp.debug
> W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
> "com_sun_star_comp_Writer_MWAWImportFilter_get_implementation"
> 2020-09-14 23:20:45.048 18128-18307/org.libreoffice.androidapp.debug
> W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
> "com_sun_star_comp_Writer_MSWorksImportFilter_get_implementation"
> 2020-09-14 23:20:45.058 18128-18307/org.libreoffice.androidapp.debug
> W/oox.storage: 1:oox/source/helper/zipstorage.cxx:68: ZipStorage::ZipStorage
> exception opening input storage com.sun.star.io.IOException
> 2020-09-14 23:20:45.074 18128-18307/org.libreoffice.androidapp.debug> [...]



> 2020-09-14 23:20:45.830 18128-18186/org.libreoffice.androidapp.debug
> D/GraphicBuffer: register, handle(0x7f437800) (w:812 h:406 s:832 f:0x1
> u:b00) 20200914_233438.mp4
> 
>   
> 2020-09-14 23:20:45.834 18128-18186/org.libreoffice.androidapp.debug
> D/HWUIExtension: MTKProgramCache.generateProgram: 8796098265152
> 
> 
> I guess few required services are missing, could you please assist me with
> this issue.  

I don't have much experience with the online-based Android app, but from
what I've seen with the Android Viewer app, a change similar to commit
https://git.libreoffice.org/core/+/7374fa59b929bea6792f80a6a819b2a06582722f%5E%21/
might make it work (adding the missing services mentioned in your log
output to the same file).

Seeing what services are mentioned in your debug log (like
"com_sun_star_comp_Writer_MSWorksImportFilter_get_implementation"), I'm
wondering what kind of file exactly you're trying to open. Do you have
the same issue when e.g. trying to open an empty ODT file?

(Maybe somebody else can say more on whether including all services into
liblo-native-code that are required to support all file formats that
LibreOffice supports is fine or would be problematic, e.g. for APK
size,...).
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: compiling Libreoffice online android doesn't display document.

[Chetan Rawat]

Hi,
I have been doing my analysis on this  for a while now and after enabling
few debugging symbols the adb logcat gives following output when opening any
document:

2020-09-14 23:20:44.807 18128-18307/org.libreoffice.androidapp.debug
D/loolwsd: DBG Mobile-18307 2020-09-14 17:50:44.807636 [ kit_001 ] DBG 
Calling
lokit::documentLoad(file:///data/user/0/org.libreoffice.androidapp.debug/cache/LibreOffice2801578603109353115.tmp,
"Language=en-IN,DeviceFormFactor=mobile").|
/opt/lo_new/online/kit/Kit.cpp:1171
2020-09-14 23:20:45.038 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"com_sun_star_comp_Writer_MWAWImportFilter_get_implementation"
2020-09-14 23:20:45.042 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"org_libreoffice_comp_Writer_EBookImportFilter_get_implementation"
2020-09-14 23:20:45.043 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"org_libreoffice_comp_Writer_EBookImportFilter_get_implementation"
2020-09-14 23:20:45.044 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"com_sun_star_comp_Writer_MWAWImportFilter_get_implementation"
2020-09-14 23:20:45.046 18128-18307/org.libreoffice.androidapp.debug
I/chatty: uid=10365(org.libreoffice.androidapp.debug) kit_001 identical 2
lines
2020-09-14 23:20:45.047 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"com_sun_star_comp_Writer_MWAWImportFilter_get_implementation"
2020-09-14 23:20:45.048 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"com_sun_star_comp_Writer_MSWorksImportFilter_get_implementation"
2020-09-14 23:20:45.058 18128-18307/org.libreoffice.androidapp.debug
W/oox.storage: 1:oox/source/helper/zipstorage.cxx:68: ZipStorage::ZipStorage
exception opening input storage com.sun.star.io.IOException
2020-09-14 23:20:45.074 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"org_libreoffice_comp_Writer_EBookImportFilter_get_implementation"
2020-09-14 23:20:45.075 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"com_sun_star_comp_Writer_MSWorksImportFilter_get_implementation"
2020-09-14 23:20:45.076 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"com_sun_star_comp_Writer_MWAWImportFilter_get_implementation"
2020-09-14 23:20:45.077 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"org_libreoffice_comp_Writer_EBookImportFilter_get_implementation"
2020-09-14 23:20:45.078 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"org_libreoffice_comp_Writer_PagesImportFilter_get_implementation"
2020-09-14 23:20:45.079 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"com_sun_star_comp_Writer_AbiWordImportFilter_get_implementation"
2020-09-14 23:20:45.082 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"com_sun_star_comp_Impress_MWAWPresentationImportFilter_get_implementation"
2020-09-14 23:20:45.083 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"com_sun_star_comp_Impress_MWAWPresentationImportFilter_get_implementation"
2020-09-14 23:20:45.084 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"org_libreoffice_comp_Impress_KeynoteImportFilter_get_implementation"
2020-09-14 23:20:45.085 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"org_libreoffice_comp_Draw_ZMFImportFilter_get_implementation"
2020-09-14 23:20:45.086 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"org_libreoffice_comp_Draw_QXPImportFilter_get_implementation"
2020-09-14 23:20:45.086 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"org_libreoffice_comp_Draw_PageMakerImportFilter_get_implementation"
2020-09-14 23:20:45.087 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"com_sun_star_comp_Draw_FreehandImportFilter_get_implementation"
2020-09-14 23:20:45.088 18128-18307/org.libreoffice.androidapp.debug
W/cppuhelper: 1:cppuhelper/source/shlib.cxx:288: unknown constructor name
"com_su

Re: compiling Libreoffice online android doesn't display document.

[Chetan Rawat]

For core.git my autogen.input file looks like this.

--build=x86_64-unknown-linux-gnu
--with-android-sdk=/home/ubuntu/Android/Sdk
--with-android-ndk=/home/ubuntu/Android/Sdk/ndk/19.2.5345600
--with-distro=LibreOfficeAndroid
--enable-sal-log

online.git
configured poco 1.10.1 for android with

./configure --config=Android --no-samples --no-tests
--omit=Crypto,NetSSL_OpenSSL,Zip,Data,Data/SQLite,Data/ODBC,Data/MySQL,MongoDB,PDF,CppParser,PageCompiler,JWT

//make
PATH="$PATH":~/Android/Sdk/ndk/19.2.5345600/toolchains/llvm/prebuilt/linux-x86_64/bin
make -j8 ANDROID_ABI=armeabi-v7a CC=armv7a-linux-androideabi21-clang
CXX=armv7a-linux-androideabi21-clang++ SYSLIBS=-static-libstdc++

//install
PATH="$PATH":~/Android/Sdk/ndk/19.2.5345600/toolchains/llvm/prebuilt/linux-x86_64/bin
make -j8 ANDROID_ABI=armeabi-v7a CC=armv7a-linux-androideabi21-clang
CXX=armv7a-linux-androideabi21-clang++ SYSLIBS=-static-libstdc++ install
INSTALLDIR=/opt/poco-android

then in online.git folder
./autogen.sh && ./configure
CC=~/Android/Sdk/ndk/19.2.5345600/toolchains/llvm/prebuilt/linux-x86_64/bin/armv7a-linux-androideabi21-clang
CXX=~/Android/Sdk/ndk/19.2.5345600/toolchains/llvm/prebuilt/linux-x86_64/bin/armv7a-linux-androideabi21-clang++
--host=arm-linux-androideabi --enable-androidapp
--with-lo-builddir=/opt/lo_final/core
--with-poco-includes=/opt/poco-android/include
--with-poco-libs=/opt/poco-android/lib --disable-setcap
--enable-silent-rules --enable-debug

when opening any document adb log show following msgs.

2020-08-18 00:09:24.874 11011-11011/org.libreoffice.androidapp.debug
W/ActivityThread: handleWindowVisibility: no activity for token
android.os.BinderProxy@89a92a8
2020-08-18 00:09:25.099 11011-11106/org.libreoffice.androidapp.debug
W/chromium: [WARNING:dns_config_service_posix.cc(342)] Failed to read
DnsConfig.
2020-08-18 00:09:25.232 11011-11107/org.libreoffice.androidapp.debug
W/ndroidapp.debu: Accessing hidden method
Landroid/media/AudioManager;->getOutputLatency(I)I (greylist, reflection,
allowed)
2020-08-18 00:09:25.236 11011-11107/org.libreoffice.androidapp.debug
W/cr_media: Requires BLUETOOTH permission
2020-08-18 00:09:25.293 11011-11011/org.libreoffice.androidapp.debug
E/GraphicExt: GraphicExtModuleLoader::CreateGraphicExtInstance false
2020-08-18 00:09:25.341 11011-11011/org.libreoffice.androidapp.debug
E/GraphicExt: GraphicExtModuleLoader::CreateGraphicExtInstance false
2020-08-18 00:09:25.982 11011-3/org.libreoffice.androidapp.debug
E/AndroidProtocolHandler: Unable to open asset URL:
file:///android_asset/dist/branding-mobile.css
2020-08-18 00:09:26.139 11011-11140/org.libreoffice.androidapp.debug
W/configmgr: 1:configmgr/source/xcuparser.cxx:904: ignoring modify of
unknown set member node "StarOffice XML (Base) Report" in
"file:///assets/share/registry/res/fcfg_langpack_en-US.xcd"
2020-08-18 00:09:26.139 11011-11140/org.libreoffice.androidapp.debug
W/configmgr: 1:configmgr/source/xcuparser.cxx:904: ignoring modify of
unknown set member node "StarOffice XML (Base) Report Chart" in
"file:///assets/share/registry/res/fcfg_langpack_en-US.xcd"










--
Sent from: 
http://document-foundation-mail-archive.969070.n3.nabble.com/Dev-f1639786.html
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: compiling Libreoffice online android doesn't display document.

[Miklos Vajna]

Hi,

On Sat, Aug 01, 2020 at 12:00:24PM -0700, Chetan Rawat 
 wrote:
> I have Build LOOL for android following instructions from this link
> https://github.com/LibreOffice/online/blob/master/android/README, I have
> used master branches for building both core and online repositories and have
> compiled build for ArmV7 on ubuntu 18.04 machine but when opening any type
> of document it just opens a white screen on the device, also when i try to
> edit the document i can see the cursor but no text is visible as i type and
> after i save the edited document and open it using other app the changes are
> saved in the document.
>  I am not sure if i have missed something in the build process could you
> suggest some guide or solution to this problem.

Can you share your core.git and online.git configure options you used?

Also, any error or warning in the adb log?

Regards,

Miklos
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice

compiling Libreoffice online android doesn't display document.

[Chetan Rawat]

Hi,
I have Build LOOL for android following instructions from this link
https://github.com/LibreOffice/online/blob/master/android/README, I have
used master branches for building both core and online repositories and have
compiled build for ArmV7 on ubuntu 18.04 machine but when opening any type
of document it just opens a white screen on the device, also when i try to
edit the document i can see the cursor but no text is visible as i type and
after i save the edited document and open it using other app the changes are
saved in the document.
 I am not sure if i have missed something in the build process could you
suggest some guide or solution to this problem.



--
Sent from: 
http://document-foundation-mail-archive.969070.n3.nabble.com/Dev-f1639786.html
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Cross compiling LibreOffice - x86_64 to aarch64

[Ray Vine]

On 02/09/19 10:09, Michael Stahl wrote:

On 30.08.19 21:51, Ray Vine wrote:

[CMP] animations/source/animcore/animcore
/w5/build-area/LO-6304/solenv/gbuild/ComponentTarget.mk:50:
*** No LIBFILENAME set at component target:
/w5/build-area/LO-6304/workdir_for_build/ComponentTarget/animations/source/animcore/animcore.component.
  Stop.
make[1]: *** Waiting for unfinished jobs
make: *** [Makefile:295: cross-toolset] Error 2


looks like some library requires "animcore" but the module that contains 
Library_animcore.mk i.e. "animations" is missing in RepositoryModule_build.mk.



Thanks, that set me on the path to a successful build.

Added animations to RepositoryModule_build.mk but then got another error:

| /w5/build-area/LO-6304/solenv/gbuild/LinkTarget.mk:687: *** used LinkTarget 
Library/libopencllo.so not defined.

and decided to try using a sledgehammer to fix it - replaced the modules list 
in RepositoryModule_build.mk with the list from RepositoryModule_host.mk and 
the BUILD system build then completed.

Presuming the purpose of the BUILD system build is to create binaries needed 
for the HOST compilation, this probably creates more than needed, but at least 
gives me a base build that works.
TODO: revisit this to add only the modules required.

Apart then from some issues related to my installation, the HOST build 
continued until getting to coinmp.
What a pita!

'configure' for each sub-directory uses $HOST-pkg-config to generate build data 
[--modversion, --cflags, etc], one use of which is to determine whether the 
source is available:

| checking for COIN-OR package CoinUtils... not given: No package 'coinutils' 
found
| configure: error: Required package CoinUtils not available.
| configure: error: /bin/sh './configure' failed for Osi

It also sets up a PKG_CONFIG_PATH which includes the local path to .pc files in 
the source and that is overwritten by the entry in the wrapper script:

aarch64-linux-gnu-pkg-config fails:
PKG_CONFIG_PATH=/w5/build-area/LO-6304/workdir/UnpackedTarball/coinmp/CoinUtils 
aarch64-linux-gnu-pkg-config --modversion "coinutils"
| Package coinutils was not found in the pkg-config search path.
| Perhaps you should add the directory containing `coinutils.pc`
| to the PKG_CONFIG_PATH environment variable
| No package 'coinutils' found

pkg-config succeeds:
PKG_CONFIG_PATH=/w5/build-area/LO-6304/workdir/UnpackedTarball/coinmp/CoinUtils 
pkg-config --modversion "coinutils"
| 2.9.11

And derived from '--cflags coinutils', the includes path is wrong:
| Making all in Osi
| Making all in src/Osi
| mkdir .libs
|  aarch64-linux-gnu-g++ -DHAVE_CONFIG_H -I. -I. 
-I/sysroot/w5/build-area/LO-6304/workdir/UnpackedTarball/coinmp/CoinUtils/src
 -O3 -pipe -DNDEBUG -pedantic-errors -Wparentheses -Wreturn-type -Wcast-qual -Wall 
-Wpointer-arith -Wwrite-strings -Wconversion -Wno-unknown-pragmas -Wno-long-long 
-DOSI_BUILD -MT OsiAuxInfo.lo -MD -MP -MF .deps/OsiAuxInfo.Tpo -c OsiAuxInfo.cpp  
-fPIC -DPIC -o .libs/OsiAuxInfo.o
| OsiAuxInfo.cpp:10:10: fatal error: CoinPragma.hpp: No such file or directory
|10 | #include "CoinPragma.hpp"
|   |  ^~~~

CoinPragma.hpp is in
/w5/build-area/LO-6304/workdir/UnpackedTarball/coinmp/CoinUtils/src
not
/sysroot/w5/build-area/LO-6304/workdir/UnpackedTarball/coinmp/CoinUtils/src


The eventual work-around I came up with was to modify the wrapper script to 
bypass the PKG_CONFIG_* variables only for the coinmp configure scripts, i.e. 
if PKG_CONFIG_PATH contains CoinUtils:

[[ ! $PKG_CONFIG_PATH == *CoinUtils* ]] && {
PKG_CONFIG_PATH=
PKG_CONFIG_SYSROOT_DIR=/sysroot
PKG_CONFIG_LIBDIR=/sysroot/usr/lib64/pkgconfig
}
exec pkg-config "$@"


That did the trick - the build completed, and so far LO 6.3.0.4 is running OK.


___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Cross compiling LibreOffice - x86_64 to aarch64

[Michael Stahl]

On 30.08.19 21:51, Ray Vine wrote:
  
I would like to cross compile LibreOffice for

aarch64 to run on my RPi3.



/w5/build-area/LO-6304/RepositoryModule_build.mk:12:
*** Module does not exist:
/w5/build-area/LO-6304/mysqlc/Module_mysqlc.mk.  Stop.
make: *** [Makefile:299: cross-toolset] Error 2

That's correct as far as I can see - there is no
mysqlc module, so:
sed -i '/mysqlc/d ' RepositoryModule_build.mk


yes it should have been removed from this file when the library was 
moved to "connectivity". (actually i wonder how it was ever added to 
_build.mk, hopefully no build tool needs it?)



takes me to the next error:

[2]
[CMP] animations/source/animcore/animcore
/w5/build-area/LO-6304/solenv/gbuild/ComponentTarget.mk:50:
*** No LIBFILENAME set at component target:
/w5/build-area/LO-6304/workdir_for_build/ComponentTarget/animations/source/animcore/animcore.component.
  Stop.
make[1]: *** Waiting for unfinished jobs
make: *** [Makefile:295: cross-toolset] Error 2


looks like some library requires "animcore" but the module that contains 
Library_animcore.mk i.e. "animations" is missing in 
RepositoryModule_build.mk.


please fix these problems upload a patch to gerrit :)

https://wiki.documentfoundation.org/Development/gerrit
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice

Cross compiling LibreOffice - x86_64 to aarch64

[Ray Vine]

 
I would like to cross compile LibreOffice for
aarch64 to run on my RPi3.

I've built it natively on x86_64 and am using the
same configure options for the cross compilation,
but getting errors almost immediately when it
starts with the BUILD system build.

As far as I can tell, I'm effectively doing what's
in README.cross|Raspbian.
Building on Slackware with my own cross compiler.

The set up is:

export CONFIGURE_OPTS="
--prefix=/
--disable-postgresql-sdbc
--without-java
--without-junit
--disable-odk
--without-krb5
--without-gssapi
--disable-gtk3
--disable-gtk
--disable-gstreamer-0-10
--disable-gstreamer-1-0
--disable-epm
--enable-release-build
--with-parallelism=6
--with-locales=en-GB
--with-lang=en-GB
--with-help
--with-external-tar=/w5/external_tarballs-6304
--with-system-libatomic_ops
--with-system-nss
--with-system-libpng
--with-theme=colibre"


For a native x86_64 build:
-
./autogen.sh $CONFIGURE_OPTS
make build-nocheck

.. builds successfully


For cross-compiling:
---
export HOST=aarch64-linux-gnu
export SYSROOT=/sysroot
export CC=$HOST-gcc
export CXX=$HOST-g++
mount -B
 $SYSROOT
export LIBDIRSUFFIX=64
export PATH=/bin:$PATH
For HOST configuration:
export PYTHON_CFLAGS=-I$SYSROOT/usr/include/python2.7
export PYTHON_LIBS=-lpython2.7


Set up a pkg-config wrapper script for HOST and
name it $HOST-pkg-config because that's what
configure looks for:

echo $"PKG_CONFIG_PATH='' \\
PKG_CONFIG_SYSROOT_DIR=$SYSROOT \\
PKG_CONFIG_LIBDIR=$SYSROOT/usr/lib$LIBDIRSUFFIX/pkgconfig
\\
exec pkg-config \"\$@\"" >
/bin/$HOST-pkg-config


./autogen.sh $CONFIGURE_OPTS --host=$HOST
--with-build-platform-configure-options="$CONFIGURE_OPTS
--without-help --without-system-dicts"

completes OK and everything looks OK

'make build-nocheck' or 'make' produces two errors
almost immediately:

[1]
make gb_Side=build -j 6 -rs -f
/w5/build-area/LO-6304/Makefile fetch
make gb_Side=build -j 6 -rs -f
/w5/build-area/LO-6304/Makefile.gbuild build-tools
/w5/build-area/LO-6304/RepositoryModule_build.mk:12:
*** Module does not exist:
/w5/build-area/LO-6304/mysqlc/Module_mysqlc.mk.  Stop.
make: *** [Makefile:299: cross-toolset] Error 2

That's correct as far as I can see - there is no
mysqlc module, so:
sed -i '/mysqlc/d ' RepositoryModule_build.mk
takes me to the next error:

[2]
[CMP] animations/source/animcore/animcore
/w5/build-area/LO-6304/solenv/gbuild/ComponentTarget.mk:50:
*** No LIBFILENAME set at component target:
/w5/build-area/LO-6304/workdir_for_build/ComponentTarget/animations/source/animcore/animcore.component.
 Stop.
make[1]: *** Waiting for unfinished jobs
make: *** [Makefile:295: cross-toolset] Error 2

The directory ComponentTarget doesn't exist at
this stage.

This occurs after a minute or so for the cross
compile, whereas it completes for a native x86_64
build, but after about 5 hours into an 8+ hour build.

Running with --with-parallelism=1 takes a lot
longer but eventually gets to the same error.
ComponentTarget does then exist, but doesn't
contain animations:

ls
/w5/build-area/LO-6304/workdir_for_build/ComponentTarget
binaryurp/  comphelper/  io/  remotebridges/  sax/
 stoc/

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: compiling Libreoffice with internationalization

[Regina Henschel]

Hi Adrien,

Adrien Ollier schrieb am 04-Apr-19 um 12:36:

Hello everybody,

I am a French C++ developper and I want to contribute to Libreoffice.
I downloaded the Libreoffice source code and compiled it on my Ubuntu 
desktop computer with is configured in French but all the text in the 
menus is in English. I can't get the User Interface in French as you can 
see in the following screenshot:




You need to add the configure "with-lang" option. For me, German, it is 
--with-lang="de".

More details on configure options use ./autogen.sh --help

Kind regards
Regina
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Compiling libreoffice

[Dmitriy Shilin]

On Sat, Jan 05, 2019 at 10:34:32AM +0530, rajprakash sahoo wrote:
> Hi,
> After I install all the dependencies and start compiling by $make command,
> then my pc(ubuntu) crashes and I have to restart it again.
> So, how can i compile it , please help me out

Hi, you should provide the crash log and more details,
otherwise it is impossible to guess the problem.

Dmitriy.

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice

Compiling libreoffice

[rajprakash sahoo]

Hi,
After I install all the dependencies and start compiling by $make command,
then my pc(ubuntu) crashes and I have to restart it again.
So, how can i compile it , please help me out

Regards,
Rajprakash
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Error compiling LibreOffice git (Xinerama not functional?)

[Clemens Eisserer]

Hi again,

> Look for XineramaIsActive in config.log. That may give you a hint as to
> why the test fails. My guess is a compile error, as, judging from what
> have you written about your other problems, your system doesn't appear
> to be in a good state. (Several of us build on F-27 daily. Just as we
> did on F-26, F-25, etc. before. So a problem like yours would be
> noticed...)

Thanks for the pointer - after some digging I think I've found the issue:

/usr/bin/ld: inkompatibles
/usr/lib/gcc/x86_64-redhat-linux/7/../../../libXinerama.so wird bei
der Suche nach -lXinerama übersprungen
/usr/bin/ld: inkompatibles //lib/libXinerama.so wird bei der Suche
nach -lXinerama übersprungen
/usr/bin/ld: inkompatibles //usr/lib/libXinerama.so wird bei der Suche
nach -lXinerama übersprungen
/usr/bin/ld: -lXinerama kann nicht gefunden werden


[ce@localhost temp]$ file
/usr/lib/gcc/x86_64-redhat-linux/7/../../../libXinerama.so
/usr/lib/gcc/x86_64-redhat-linux/7/../../../libXinerama.so: symbolic
link to libXinerama.so.1.0.0
[ce@localhost temp]$ file
/usr/lib/gcc/x86_64-redhat-linux/7/../../../libXinerama.so.1.0.0
/usr/lib/gcc/x86_64-redhat-linux/7/../../../libXinerama.so.1.0.0: ELF
32-bit LSB shared object, Intel 80386, version 1 (SYSV), dynamically
linked, BuildID[sha1]=6d3e82c64047b7d1bf51fa61af1ae58d8b53a741,
stripped

[ce@localhost temp]$ rpm -qf
/usr/lib/gcc/x86_64-redhat-linux/7/../../../libXinerama.so
libXinerama-devel-1.1.3-9.fc27.i686

So, the package libXinerama-devel.i686 seems to install into
x86_64-redhat-linux and confuses the whole system.

Hopefully I am able to resolv this.

Thanks & br, Clemens
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Error compiling LibreOffice git (Xinerama not functional?)

[David Tardon]

Hi,

On Mon, 2018-02-12 at 12:13 +0100, Clemens Eisserer wrote:
> Hi Stephan,
> 
> > Did you install the relevant -devel packages, libxslt-devel,
> > libXrandr-devel, libXinerama-devel?  Generally, `sudo dnf builddep
> > libreoffice` (as given at
> 
> Sure, I manually instalelled the devel packages - however to make
> sure, I also executed the dnf builddep command you mentioned.
> However, the configure run still fails with the message mentioned
> before - the header is there, but configure does not find
> XineramaIsActive.

Look for XineramaIsActive in config.log. That may give you a hint as to
why the test fails. My guess is a compile error, as, judging from what
have you written about your other problems, your system doesn't appear
to be in a good state. (Several of us build on F-27 daily. Just as we
did on F-26, F-25, etc. before. So a problem like yours would be
noticed...)

D.
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Error compiling LibreOffice git (Xinerama not functional?)

[Stephan Bergmann]

On 12.02.2018 12:13, Clemens Eisserer wrote:

Did you install the relevant -devel packages, libxslt-devel,
libXrandr-devel, libXinerama-devel?  Generally, `sudo dnf builddep
libreoffice` (as given at


Sure, I manually instalelled the devel packages - however to make
sure, I also executed the dnf builddep command you mentioned.
However, the configure run still fails with the message mentioned
before - the header is there, but configure does not find
XineramaIsActive.


Then you'd need to look into config.log why it's failing.
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Error compiling LibreOffice git (Xinerama not functional?)

[Clemens Eisserer]

Hi Stephan,

> Did you install the relevant -devel packages, libxslt-devel,
> libXrandr-devel, libXinerama-devel?  Generally, `sudo dnf builddep
> libreoffice` (as given at

Sure, I manually instalelled the devel packages - however to make
sure, I also executed the dnf builddep command you mentioned.
However, the configure run still fails with the message mentioned
before - the header is there, but configure does not find
XineramaIsActive.

Best regards, Clemens
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Error compiling LibreOffice git (Xinerama not functional?)

[Stephan Bergmann]

On 12.02.2018 12:00, Clemens Eisserer wrote:

I am trying to diagnose some performance issues I am experiencing
running LO on Linux.
Compiling on Fedora-27 seems a little bit troublesome, as I:

* had to set libxslt and libexslt includes and libraries manually
* disable xrandr, despite headers and library installed, it aborted
with "not functional"

however, it seems the feature causing the latest issue can not be
simply disabled via a switch:

checking whether and how to use Xinerama... yes, with dynamic linking
checking X11/extensions/Xinerama.h usability... yes
checking X11/extensions/Xinerama.h presence... yes
checking for X11/extensions/Xinerama.h... yes
checking for XineramaIsActive in -lXinerama... no
configure: error: Xinerama not functional?

Any idea why LibreOffice thinks my Xinerama is not functional?
Can I disable it somehow?


Did you install the relevant -devel packages, libxslt-devel, 
libXrandr-devel, libXinerama-devel?  Generally, `sudo dnf builddep 
libreoffice` (as given at 
) should be a good 
approximation of installing whatever packages will be needed.

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice

Error compiling LibreOffice git (Xinerama not functional?)

[Clemens Eisserer]

Hi,

I am trying to diagnose some performance issues I am experiencing
running LO on Linux.
Compiling on Fedora-27 seems a little bit troublesome, as I:

* had to set libxslt and libexslt includes and libraries manually
* disable xrandr, despite headers and library installed, it aborted
with "not functional"

however, it seems the feature causing the latest issue can not be
simply disabled via a switch:

checking whether and how to use Xinerama... yes, with dynamic linking
checking X11/extensions/Xinerama.h usability... yes
checking X11/extensions/Xinerama.h presence... yes
checking for X11/extensions/Xinerama.h... yes
checking for XineramaIsActive in -lXinerama... no
configure: error: Xinerama not functional?

Any idea why LibreOffice thinks my Xinerama is not functional?
Can I disable it somehow?

Thank you and best regards, Clemens
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Strange error when compiling LibreOffice from master

[Chris Sherlock]

I did some further searching - I think my VM is running out of memory, hence 
this warning. 

That would explain why restarting the compile works!

Thanks for everyone’s feedback. 

Chris

> On 30 Dec 2015, at 11:59 AM, Chris Sherlock  
> wrote:
> 
> Not sure about that - this is an up to date version of 14.10.
> 
> Also not certain that it's a VM issue, I think it's unlikely to be a hardware 
> problem (at least, I hope not!).
> 
> One thing that is consistent is that it always fails on an OpenCL test.
> 
> It is very odd that I can run make after the error presents itself and it 
> continues on as if no problem has occurred.
> 
> The bug suggests getting the preprocessed and submitting to the bcc buglist, 
> I couldn't find info in that - if anyone could point me to the right spot I'd 
> be eternally grateful :-)
> 
> Chris
> 
> On Wednesday, December 30, 2015, jan iversen  > wrote:
> Hi
> 
> I just installed a new ubuntu 14.10 today, and compiled LO without problems.
> 
> Could it be you are running with an old GCC (pointer to doc 4.9 suggest that).
> 
> Or alternatively you need to do a "git pull" to get the latest commits.
> 
> rgds
> jan i.
> 

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Strange error when compiling LibreOffice from master

[Chris Sherlock]

Not sure about that - this is an up to date version of 14.10.

Also not certain that it's a VM issue, I think it's unlikely to be a
hardware problem (at least, I hope not!).

One thing that is consistent is that it always fails on an OpenCL test.

It is very odd that I can run make after the error presents itself and it
continues on as if no problem has occurred.

The bug suggests getting the preprocessed and submitting to the bcc
buglist, I couldn't find info in that - if anyone could point me to the
right spot I'd be eternally grateful :-)

Chris

On Wednesday, December 30, 2015, jan iversen 
wrote:

> Hi
>
> I just installed a new ubuntu 14.10 today, and compiled LO without
> problems.
>
> Could it be you are running with an old GCC (pointer to doc 4.9 suggest
> that).
>
> Or alternatively you need to do a "git pull" to get the latest commits.
>
> rgds
> jan i.
>
>
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Strange error when compiling LibreOffice from master

[jan iversen]

Hi

I just installed a new ubuntu 14.10 today, and compiled LO without problems.

Could it be you are running with an old GCC (pointer to doc 4.9 suggest that).

Or alternatively you need to do a "git pull" to get the latest commits.

rgds
jan i.

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Strange error when compiling LibreOffice from master

[Ashod Nakashian]

On Fri, Dec 25, 2015 at 5:31 PM, Chris Sherlock 
wrote:

>
> Anyone have any ideas why the compiler might be barfing? I run make again
> and it finishes successfully!
>
>
>
I don't have an answer, but I'll leave my thoughts in case they are useful.

I've seen gcc fail with random internal errors when I built within VMs. Not
sure what the reason is, but when I restart the VM things are back to
normal. This happens every so often.

One common explanation would be bad RAM, but I know that's not the case
with me (my dimms are well tested and the system is otherwise extremely
stable).

If this isn't a persistent issue, I'd say restart the VM and try again,
otherwise ignore it. (You can change the VM RAM size as well, make sure
it's not too small nor too large for the host to thrash.)

Another point to keep in mind is that 4.9 hasn't been gcc's best release.
I'd skip it altogether and upgrade. 5.2 works fine (those random occasional
errors notwithstanding).
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Strange error when compiling LibreOffice from master

[Chris Sherlock]

Hi all,

I'm running Ubuntu 14 on a VM via VirtualBox, when I compile from master I
consistently get the following:

[CXX] cppcanvas/qa/extras/emfplus/emfplus.cxx
[CXX] sc/qa/extras/sccondformats.cxx
[CXX] sc/qa/unit/subsequent_filters-test.cxx
[CXX] sc/qa/unit/subsequent_export-test.cxx
[CXX] sc/qa/unit/bugfix-test.cxx
[CXX] sc/qa/extras/htmlexporttest.cxx
[CXX] sc/qa/unit/opencl-test.cxx
[CXX] sc/qa/unit/copy_paste_test.cxx
[CXX] sw/qa/core/uwriter.cxx
[CXX] sw/qa/core/test_ToxLinkProcessor.cxx
/home/chris/repos/libreoffice/sc/qa/unit/opencl-test.cxx: In static member
function ‘static void
ScOpenCLTest::addTestsToSuite(CppUnit::TestSuiteBuilderContextBase&)’:
/home/chris/repos/libreoffice/sc/qa/unit/opencl-test.cxx:302:300: note:
variable tracking size limit exceeded with -fvar-tracking-assignments,
retrying without
 CPPUNIT_TEST_SUITE(ScOpenCLTest);



^
[CXX] sw/qa/core/test_ToxTextGenerator.cxx
[CXX] sw/qa/extras/docbookexport/docbookexport.cxx
[CXX] sw/qa/extras/htmlexport/htmlexport.cxx
[CXX] sw/qa/extras/htmlimport/htmlimport.cxx
[CXX] sw/qa/core/macros-test.cxx
[CXX] sw/qa/extras/ooxmlexport/ooxmlexport.cxx
[CXX] sw/qa/extras/ooxmlexport/ooxmlexport2.cxx
[CXX] sw/qa/extras/ooxmlexport/ooxmlexport3.cxx
/home/chris/repos/libreoffice/sc/qa/unit/ucalc_formula.cxx: In member
function ‘void Test::testFetchVectorRefArray()’:
/home/chris/repos/libreoffice/sc/qa/unit/ucalc_formula.cxx:474:6: note:
variable tracking size limit exceeded with -fvar-tracking-assignments,
retrying without
 void Test::testFetchVectorRefArray()
  ^
{standard input}: Assembler messages:
{standard input}:1265132: Warning: end of file not at end of a line;
newline inserted
{standard input}:1265305: Error: unknown pseudo-op: `.'
g++: internal compiler error: Killed (program cc1plus)
Please submit a full bug report,
with preprocessed source if appropriate.
See  for instructions.
/home/chris/repos/libreoffice/solenv/gbuild/LinkTarget.mk:189: recipe for
target
'/home/chris/repos/libreoffice/workdir/CxxObject/sc/qa/unit/ucalc_formula.o'
failed
make[1]: ***
[/home/chris/repos/libreoffice/workdir/CxxObject/sc/qa/unit/ucalc_formula.o]
Error 4
make[1]: *** Waiting for unfinished jobs

My version of g++ is:

chris@libreoffice-ia64:~/repos/libreoffice$ g++ --version
g++ (Ubuntu 4.9.2-10ubuntu13) 4.9.2
Copyright (C) 2014 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

Anyone have any ideas why the compiler might be barfing? I run make again
and it finishes successfully!

Chris
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Compiling LibreOffice 4.1.2.3 on Solaris

[Michael Stahl]

On 13/11/13 12:32, Jan Holesovsky wrote:
> Eric Bautsch píše v Pá 01. 11. 2013 v 19:57 +:
> 
>> And the output suggests that the versions are OK:
>> bautsche@cressida $ pvs solver/unxsogi.pro/bin/idlc 
>> libm.so.2 (SUNW_1.1);
>> libgcc_s.so.1 (GCC_3.0);
>> libc.so.1 (SUNWprivate_1.1, SUNW_1.1, SUNW_0.9,
>> SUNW_0.7, SYSVABI_1.3);
>> libuno_sal.so.3 (LIBO_UDK_4.1, UDK_3.6, LIBO_UDK_4.0,
>> PRIVATE_1.1, UDK_3
>> _0_0);
>> bautsche@cressida $ pvs
>> solver/unxsogi.pro/lib/libuno_sal.so.3 
>> libgcc_s.so.1 (GCC_3.0);
>> libsocket.so.1 (SUNW_1.1, SUNW_0.7);
>> libnsl.so.1 (SUNW_0.7);
>> libm.so.2 (SUNW_1.1);
>> libpthread.so.1 (SUNW_1.1, SUNW_1.2, SUNW_0.9);
>> libuno_sal.so.3;
>> UDK_3_0_0;
> [...]
> 
> Knowing nothing about Solaris - I wonder why the other libs have eg.
> (SUNW_1.1) or (GCC_3.0), but libuno_sal.so.3 has nothing like that,
> instead the UDK_3_0_0 etc. are each on a separate line as if it was a
> library itself, or what?  Could that be the problem?

iirc the UDK_3_0_0 above is a version that libuno_sal.so.3 itself
exports whereas the things in brackets are versions that libuno_sal.so.3
requires from other (system, hence SUNW_*) libraries, hence the difference.

not obivous to me why it's failing to run anyway.
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Compiling LibreOffice 4.1.2.3 on Solaris

[Eric Bautsch]

No, I haven't played with this any further. I'm intending to, but I was going to 
move the compile over to Solaris 11.1, that being obviously a more stable platform.


I have no idea if the lack of brackets or UDK_3_0_0 on its own is significant, 
but doing a quick pvs on other files (/opt/openoffice.org3/program/*, /usr/bin/* 
and /usr/lib/*.so) it seems that this behaviour is OK for libraries but not 
binaries:


   rpcsec.so:
libgss.so.1 (SUNW_1.2, SUNWprivate_1.1);
libnsl.so.1 (SUNW_0.7, SUNWprivate_1.1);
libc.so.1 (SUNW_1.23);
rpcsec.so.1;
SUNW_1.2;
SUNW_1.1;
SUNWprivate_1.1;
   straddr.so:
libnsl.so.1 (SUNWprivate_1.1, SYSVABI_1.3);
libc.so.1 (SUNW_1.19);
straddr.so.2;
SUNWprivate_2.1;


I'll let you know how I get on though and thanks for your help. Any and all 
pointers always welcome...


Eric


On 13/11/2013 11:32, Jan Holesovsky wrote:

Hi Eric,

Eric Bautsch píše v Pá 01. 11. 2013 v 19:57 +:


Yes, the command is pvs:
http://docs.oracle.com/cd/E19253-01/816-5165/pvs-1/

Did you get anywhere, or still stuck, please? :-)


And the output suggests that the versions are OK:
 bautsche@cressida $ pvs solver/unxsogi.pro/bin/idlc
 libm.so.2 (SUNW_1.1);
 libgcc_s.so.1 (GCC_3.0);
 libc.so.1 (SUNWprivate_1.1, SUNW_1.1, SUNW_0.9,
 SUNW_0.7, SYSVABI_1.3);
 libuno_sal.so.3 (LIBO_UDK_4.1, UDK_3.6, LIBO_UDK_4.0,
 PRIVATE_1.1, UDK_3
 _0_0);
 bautsche@cressida $ pvs
 solver/unxsogi.pro/lib/libuno_sal.so.3
 libgcc_s.so.1 (GCC_3.0);
 libsocket.so.1 (SUNW_1.1, SUNW_0.7);
 libnsl.so.1 (SUNW_0.7);
 libm.so.2 (SUNW_1.1);
 libpthread.so.1 (SUNW_1.1, SUNW_1.2, SUNW_0.9);
 libuno_sal.so.3;
 UDK_3_0_0;

[...]

Knowing nothing about Solaris - I wonder why the other libs have eg.
(SUNW_1.1) or (GCC_3.0), but libuno_sal.so.3 has nothing like that,
instead the UDK_3_0_0 etc. are each on a separate line as if it was a
library itself, or what?  Could that be the problem?

Regards,
Kendy



--
 
  

 /  .   Eric A. Bautsch
/--   __   _____
   / //   /  /
  (_/(___(__/   email: eric.baut...@pobox.com



smime.p7s
Description: S/MIME Cryptographic Signature
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Compiling LibreOffice 4.1.2.3 on Solaris

[Jan Holesovsky]

Hi Eric,

Eric Bautsch píše v Pá 01. 11. 2013 v 19:57 +:

> Yes, the command is pvs:
> http://docs.oracle.com/cd/E19253-01/816-5165/pvs-1/

Did you get anywhere, or still stuck, please? :-)

> And the output suggests that the versions are OK:
> bautsche@cressida $ pvs solver/unxsogi.pro/bin/idlc 
> libm.so.2 (SUNW_1.1);
> libgcc_s.so.1 (GCC_3.0);
> libc.so.1 (SUNWprivate_1.1, SUNW_1.1, SUNW_0.9,
> SUNW_0.7, SYSVABI_1.3);
> libuno_sal.so.3 (LIBO_UDK_4.1, UDK_3.6, LIBO_UDK_4.0,
> PRIVATE_1.1, UDK_3
> _0_0);
> bautsche@cressida $ pvs
> solver/unxsogi.pro/lib/libuno_sal.so.3 
> libgcc_s.so.1 (GCC_3.0);
> libsocket.so.1 (SUNW_1.1, SUNW_0.7);
> libnsl.so.1 (SUNW_0.7);
> libm.so.2 (SUNW_1.1);
> libpthread.so.1 (SUNW_1.1, SUNW_1.2, SUNW_0.9);
> libuno_sal.so.3;
> UDK_3_0_0;
[...]

Knowing nothing about Solaris - I wonder why the other libs have eg.
(SUNW_1.1) or (GCC_3.0), but libuno_sal.so.3 has nothing like that,
instead the UDK_3_0_0 etc. are each on a separate line as if it was a
library itself, or what?  Could that be the problem?

Regards,
Kendy

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Compiling LibreOffice 4.1.2.3 on Solaris

[Eric Bautsch]

Yes, the command is pvs:
http://docs.oracle.com/cd/E19253-01/816-5165/pvs-1/

And the output suggests that the versions are OK:

   bautsche@cressida $ pvs solver/unxsogi.pro/bin/idlc
   libm.so.2 (SUNW_1.1);
   libgcc_s.so.1 (GCC_3.0);
   libc.so.1 (SUNWprivate_1.1, SUNW_1.1, SUNW_0.9, SUNW_0.7, SYSVABI_1.3);
   libuno_sal.so.3 (LIBO_UDK_4.1, UDK_3.6, LIBO_UDK_4.0, PRIVATE_1.1, UDK_3
   _0_0);
   bautsche@cressida $ pvs solver/unxsogi.pro/lib/libuno_sal.so.3
   libgcc_s.so.1 (GCC_3.0);
   libsocket.so.1 (SUNW_1.1, SUNW_0.7);
   libnsl.so.1 (SUNW_0.7);
   libm.so.2 (SUNW_1.1);
   libpthread.so.1 (SUNW_1.1, SUNW_1.2, SUNW_0.9);
   libuno_sal.so.3;
   UDK_3_0_0;
   UDK_3.1;
   UDK_3.2;
   UDK_3.3;
   UDK_3.4;
   UDK_3.5;
   UDK_3.6;
   UDK_3.7;
   UDK_3.8;
   UDK_3.9;
   UDK_3.10;
   UDK_3.11;
   LIBO_UDK_3.5;
   LIBO_UDK_3.6;
   LIBO_UDK_4.0;
   LIBO_UDK_4.1;
   PRIVATE_1.0;
   PRIVATE_1.1;
   PRIVATE_1.2;
   PRIVATE_textenc.1;
   PRIVATE_file.1;
   GLIBCXX_3.4;
   bautsche@cressida $



On 01/11/2013 15:22, Stephan Bergmann wrote:

On 11/01/2013 03:48 PM, Eric Bautsch wrote:

Output from LD_DEBUG=all ldd -r solver/unxsogi.pro/bin/idlc >log 2>&1
attached.


Unfortunately doesn't give a clue why the loader refuses to pick the symobls 
from libuno_sal.so.3.


The last idea I have is that there's something wrong with the symbol versions.


02205: file=libuno_sal.so.3; needed by solver/unxsogi.pro/bin/idlc
02205:
02205: find object=libuno_sal.so.3; searching
02205: search 
path=$ORIGIN/../../ure-link/lib:/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib 
(RUNPATH/RPATH from file solver/unxsogi.pro/bin/idlc)
02205: trying 
path=/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/bin/../../ure-link/lib/libuno_sal.so.3
02205: trying 
path=/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib/libuno_sal.so.3
libuno_sal.so.3 => 
/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib/libuno_sal.so.3
02205: 
file=/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib/libuno_sal.so.3 
[ ELF ]; generating link map

02205: addr: 0xfe4e4000 size: 0x4e38c
02205: lmid: BASE lmco: 0x10
02205:
02205: version needed processing: file=solver/unxsogi.pro/bin/idlc
02205: file version
02205: libuno_sal.so.3 LIBO_UDK_4.1
02205: libuno_sal.so.3 UDK_3.6
02205: libuno_sal.so.3 LIBO_UDK_4.0
02205: libuno_sal.so.3 PRIVATE_1.1
02205: libuno_sal.so.3 UDK_3_0_0


indicates that idlc does expect to see versioned symbols from libuno_sal.so.3, 
but I forgot what the Solaris command is to see what versions the symbols 
exported/required by an object should have (psv? pvs? something like that), 
and whether there's an LD_DEBUG argument in addition to "all" on Solaris to 
make it output information about symbol resolution version mismatch.


Stephan
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice



--
 
  

 /  .   Eric A. Bautsch
/--   __   _____
   / //   /  /
  (_/(___(__/   email: eric.baut...@pobox.com



smime.p7s
Description: S/MIME Cryptographic Signature
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Compiling LibreOffice 4.1.2.3 on Solaris

[Stephan Bergmann]

On 11/01/2013 03:48 PM, Eric Bautsch wrote:

Output from LD_DEBUG=all ldd -r solver/unxsogi.pro/bin/idlc >log 2>&1
attached.


Unfortunately doesn't give a clue why the loader refuses to pick the 
symobls from libuno_sal.so.3.


The last idea I have is that there's something wrong with the symbol 
versions.



02205: file=libuno_sal.so.3;  needed by solver/unxsogi.pro/bin/idlc
02205:
02205: find object=libuno_sal.so.3; searching
02205:  search 
path=$ORIGIN/../../ure-link/lib:/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib
  (RUNPATH/RPATH from file solver/unxsogi.pro/bin/idlc)
02205:  trying 
path=/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/bin/../../ure-link/lib/libuno_sal.so.3
02205:  trying 
path=/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib/libuno_sal.so.3
libuno_sal.so.3 =>
/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib/libuno_sal.so.3
02205: 
file=/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib/libuno_sal.so.3
  [ ELF ]; generating link map
02205: addr: 0xfe4e4000  size:  0x4e38c
02205: lmid:   BASE  lmco: 0x10
02205:
02205: version needed processing: file=solver/unxsogi.pro/bin/idlc
02205: fileversion
02205: libuno_sal.so.3 LIBO_UDK_4.1
02205: libuno_sal.so.3 UDK_3.6
02205: libuno_sal.so.3 LIBO_UDK_4.0
02205: libuno_sal.so.3 PRIVATE_1.1
02205: libuno_sal.so.3 UDK_3_0_0


indicates that idlc does expect to see versioned symbols from 
libuno_sal.so.3, but I forgot what the Solaris command is to see what 
versions the symbols exported/required by an object should have (psv? 
pvs? something like that), and whether there's an LD_DEBUG argument in 
addition to "all" on Solaris to make it output information about symbol 
resolution version mismatch.


Stephan
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Compiling LibreOffice 4.1.2.3 on Solaris

[Stephan Bergmann]

On 11/01/2013 03:24 PM, Eric Bautsch wrote:

On 01/11/2013 13:55, Stephan Bergmann wrote:

Try "ldd -r solver/unxsogi.pro/bin/idlc" to force symbol resolution.



bautsche@cressida $ ldd -r solver/unxsogi.pro/bin/idlc

[...]

 libuno_sal.so.3 =>

/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib/libuno_sal.so.3

[...]

 symbol not found: osl_getProcessInfo
(solver/unxsogi.pro/bin/idlc)



Try "nm -D solver/unxsogi.pro/lib/libuno_sal.so.3" to see whether it
exports rtl_string_new.



bautsche@cressida $ nm -D solver/unxsogi.pro/lib/libuno_sal.so.3

[...]

[336]   |252720|   449|FUNC |GLOB |0|11 |osl_getProcessInfo

[...]

So solver/unxsogi.pro/lib/libuno_sal.so.3 exports osl_getProcessInfo 
etc. but solver/unxsogi.pro/bin/idlc doesn't find it.  Try "LD_DEBUG=all 
-r solver/unxsogi.pro/bin/idlc >log" to get (lots of) detailed output of 
what the loader actually does and grep the generated log e.g. for the 
lines where it says it does not find osl_getProcessInfo.


Stephan
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Compiling LibreOffice 4.1.2.3 on Solaris

[Stephan Bergmann]

On 11/01/2013 02:12 PM, Eric Bautsch wrote:

bautsche@cressida $ ldd solver/unxsogi.pro/bin/idlc
 libnsl.so.1 =>   /lib/libnsl.so.1
 libsocket.so.1 =>/lib/libsocket.so.1
 libreglo.so =>

/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib/libreglo.so
 libuno_sal.so.3 =>

/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib/libuno_sal.so.3
 libuno_salhelpergcc3.so.3 =>

/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib/libuno_salhelpergcc3.so.3
 libstdc++.so.6 =>/usr/lib/libstdc++.so.6
 libm.so.2 => /lib/libm.so.2
 libgcc_s.so.1 => /usr/lib/libgcc_s.so.1
 libc.so.1 => /lib/libc.so.1
 libstorelo.so =>

/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib/libstorelo.so
 libpthread.so.1 =>   /lib/libpthread.so.1
bautsche@cressida $


Try "ldd -r solver/unxsogi.pro/bin/idlc" to force symbol resolution.

Try "nm -D solver/unxsogi.pro/lib/libuno_sal.so.3" to see whether it 
exports rtl_string_new.


Stephan
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Compiling LibreOffice 4.1.2.3 on Solaris

[Eric Bautsch]

Hi Norbert and Stephan.

The LD_LIBRARY_PATH looks fine to me:

   [build IDL] udkapi/com/sun/star/idl
   mkdir -p
   
/export/home/bautsche/libre-office/libreoffice-4.1.2.3/workdir/unxsogi.pro/UnoApiPartTarget/udkapi/com/sun/star/
   && RESPONSEFILE=/tmp/gbuild.LYKziC &&
   
SOLARBINDIR=/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/bin
   
LD_LIBRARY_PATH=${LD_LIBRARY_PATH:+$LD_LIBRARY_PATH:}"/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib:/export/home/bautsche/libre-office/libreoffice-4.1.2.3/instdir/unxsogi.pro/program"
   
/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/bin/idlc
   -I/export/home/bautsche/libre-office/libreoffice-4.1.2.3/udkapi -M
   
/export/home/bautsche/libre-office/libreoffice-4.1.2.3/workdir/unxsogi.pro/Dep/UnoApiPartTarget/udkapi/com/sun/star/
   -O
   
/export/home/bautsche/libre-office/libreoffice-4.1.2.3/workdir/unxsogi.pro/UnoApiPartTarget/udkapi/com/sun/star/
   -verbose @${RESPONSEFILE} > /dev/null && rm -f ${RESPONSEFILE} && touch
   
/export/home/bautsche/libre-office/libreoffice-4.1.2.3/workdir/unxsogi.pro/UnoApiPartTarget/udkapi/com/sun/star/idl.done
   ld.so.1: idlc: fatal: relocation error: file
   
/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/bin/idlc:
   symbol rtl_string_new: referenced symbol not found
   /bin/sh: line 1: 2360: Killed
   make[1]: ***
   
[/export/home/bautsche/libre-office/libreoffice-4.1.2.3/workdir/unxsogi.pro/UnoApiPartTarget/udkapi/com/sun/star/idl.done]
   Killed
   make[1]: Leaving directory
   `/export/home/bautsche/libre-office/libreoffice-4.1.2.3'
   gmake: *** [build] Error 2
   bautsche@cressida $ ls -la
   /export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib
   total 654
   drwxr-xr-x 2 bautsche user 11 Oct 31 12:12 ./
   drwxr-xr-x 4 bautsche user 4 Oct 31 12:12 ../
   -rw-r--r-- 1 bautsche user 66712 Oct 31 12:12 libcodemaker.a
   -rw-r--r-- 1 bautsche user 25514 Oct 31 12:12 libcodemaker_cpp.a
   -rwxr-xr-x 1 bautsche user 133221 Oct 31 12:12 libreglo.so*
   -rwxr-xr-x 1 bautsche user 161724 Oct 31 12:12 libstorelo.so*
   lrwxrwxrwx 1 bautsche user 15 Oct 31 12:12 libuno_sal.so -> libuno_sal.so.3*
   -rwxr-xr-x 1 bautsche user 377081 Oct 31 12:12 libuno_sal.so.3*
   lrwxrwxrwx 1 bautsche user 25 Oct 31 12:12 libuno_salhelpergcc3.so ->
   libuno_salhelpergcc3.so.3*
   -rwxr-xr-x 1 bautsche user 36660 Oct 31 12:12 libuno_salhelpergcc3.so.3*
   -rwxr-xr-x 1 bautsche user 229716 Oct 31 12:12 libunoidllo.so*
   bautsche@cressida $

Also:

   bautsche@cressida $ readelf -d solver/unxsogi.pro/bin/idlc

   Dynamic section at offset 0x3a6a0 contains 29 entries:
   Tag Type Name/Value
   0x0001 (NEEDED) Shared library: [libnsl.so.1]
   0x0001 (NEEDED) Shared library: [libsocket.so.1]
   0x0001 (NEEDED) Shared library: [libreglo.so]
   0x0001 (NEEDED) Shared library: [libuno_sal.so.3]
   0x0001 (NEEDED) Shared library: [libuno_salhelpergcc3.so.3]
   0x0001 (NEEDED) Shared library: [libstdc++.so.6]
   0x0001 (NEEDED) Shared library: [libm.so.2]
   0x0001 (NEEDED) Shared library: [libgcc_s.so.1]
   0x0001 (NEEDED) Shared library: [libc.so.1]
   0x000f (RPATH) Library rpath:
   
[$ORIGIN/../../ure-link/lib:/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib]
   0x000c (INIT) 0x804a140
   0x000d (FINI) 0x8072500
   0x6ef5 (GNU_HASH) 0x80480e8
   0x0005 (STRTAB) 0x8048be0
   0x0006 (SYMTAB) 0x80481b0
   0x000a (STRSZ) 3438 (bytes)
   0x000b (SYMENT) 16 (bytes)
   0x0015 (DEBUG) 0x0
   0x0003 (PLTGOT) 0x80837b0
   0x0002 (PLTRELSZ) 1096 (bytes)
   0x0014 (PLTREL) REL
   0x0017 (JMPREL) 0x8049cec
   0x0011 (REL) 0x8049b94
   0x0012 (RELSZ) 344 (bytes)
   0x0013 (RELENT) 8 (bytes)
   0x6ffe (VERNEED) 0x8049a94
   0x6fff (VERNEEDNUM) 4
   0x6ff0 (VERSYM) 0x804994e
   0x (NULL) 0x0
   bautsche@cressida $ ldd solver/unxsogi.pro/bin/idlc
   libnsl.so.1 => /lib/libnsl.so.1
   libsocket.so.1 => /lib/libsocket.so.1
   libreglo.so =>
   
/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib/libreglo.so
   libuno_sal.so.3 =>
   
/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib/libuno_sal.so.3
   libuno_salhelpergcc3.so.3 =>
   
/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib/libuno_salhelpergcc3.so.3
   libstdc++.so.6 => /usr/lib/libstdc++.so.6
   libm.so.2 => /lib/libm.so.2
   libgcc_s.so.1 => /usr/lib/libgcc_s.so.1
   libc.so.1 => /lib/libc.so.1
   libstorelo.so =>
   
/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/lib/libstorelo.so
   libpthread.so.1 => /lib/libpthread.so.1
   bautsche@cressida $



I wonder whether some of the macro magic that Norbert Thiebaud mentioned (and 
that I have no idea how to make sense of) in sal/rtl/strtmpl.cxx isn't working 
for me on Sola

Re: Compiling LibreOffice 4.1.2.3 on Solaris

[Stephan Bergmann]

On 10/31/2013 01:25 PM, Eric Bautsch wrote:

I'm trying to compile LibreOffice 4.1.2.3 on Solaris 12 (build 26,
currently).

Note: don't get excited, I'm doing this in my spare time and Oracle
are in no way endorsing, supporting, or anything else'ing my efforts.

I've been using this thread for assistance and it's been a great help:
http://comments.gmane.org/gmane.comp.documentfoundation.libreoffice.devel/51422


Note that traditionally OOo had been building/working very well on 
Solaris, but these days there's only very little and seemingly 
occasional initiative to keep that up.  Therefore, Solaris-specific code 
in LO likely keeps rotting over time.


At runtime, the idlc executable needs the libuno_sal.so.3 dynamic 
library to resolve the rtl_string_new symbol (and others).  Therefore, 
idlc contains an RPATH (see "readelf -d solver/unxsogi.pro/bin/idlc") so 
that it finds it *in an installation* ("$ORIGIN/../../ure-link/lib", 
where in an installation idlc is in sdk/bin/ and libuno_sal.so.3 is in 
ure-link/lib/).


However, idlc is also called during the build, where idlc is in 
solver/unxsogi.pro/bin/ and libuno_sal.so.3 is in 
solver/unxsogi.pro/lib/.  To make that work, idlc is called with an 
LD_LIBRARY_PATH that contains solver/unxsogi.pro/lib during the build. 
This is taken care of towards the end of 
solenv/gbuild/platform/com_GCC_defs.mk (gb_Helper_set_ld_path) and 
should work for Solaris.


To see what's going wrong, run "make VERBOSE=t" to see the command line 
of how exactly idlc is called to "[build IDL] udkapi/com/sun/star/idl".


Stephan


I'm currently struggling with getting a working idlc compiled. Here's
what happens:

[build IDL] udkapi/com/sun/star/idl
ld.so.1: idlc: fatal: relocation error: file

/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/bin/idlc:
symbol rtl_string_new: referenced symbol not found
/bin/sh: line 1: 8411: Killed
make[1]: ***

[/export/home/bautsche/libre-office/libreoffice-4.1.2.3/workdir/unxsogi.pro/UnoApiPartTarget/udkapi/com/sun/star/idl.done]
Killed
gmake: *** [build] Error 2
bautsche@cressida $

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: Compiling LibreOffice 4.1.2.3 on Solaris

[Norbert Thiebaud]

On Thu, Oct 31, 2013 at 7:25 AM, Eric Bautsch  wrote:
>
> I think rtl_string_new is defined in include/rtl/string.h, but I can't see
> where the function is actually coming from (running a find -exec grep across
> the code).

see sal/rtl/strtmpl.cxx

This is a hacky way to implement both ustring and string using the
same source and some macro-magic

Norbert
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Compiling LibreOffice 4.1.2.3 on Solaris

[Eric Bautsch]

Hi All.

I'm trying to compile LibreOffice 4.1.2.3 on Solaris 12 (build 26, currently).

   Note: don't get excited, I'm doing this in my spare time and Oracle are in
   no way endorsing, supporting, or anything else'ing my efforts.

I've been using this thread for assistance and it's been a great help: 
http://comments.gmane.org/gmane.comp.documentfoundation.libreoffice.devel/51422


I'm currently struggling with getting a working idlc compiled. Here's what 
happens:

   [build IDL] udkapi/com/sun/star/idl
   ld.so.1: idlc: fatal: relocation error: file
   
/export/home/bautsche/libre-office/libreoffice-4.1.2.3/solver/unxsogi.pro/bin/idlc:
   symbol rtl_string_new: referenced symbol not found
   /bin/sh: line 1: 8411: Killed
   make[1]: ***
   
[/export/home/bautsche/libre-office/libreoffice-4.1.2.3/workdir/unxsogi.pro/UnoApiPartTarget/udkapi/com/sun/star/idl.done]
   Killed
   gmake: *** [build] Error 2
   bautsche@cressida $

I think rtl_string_new is defined in include/rtl/string.h, but I can't see where 
the function is actually coming from (running a find -exec grep across the code).


Here's the compile command for idlc:

   S=/export/home/bautsche/libre-office/libreoffice-4.1.2.3 &&
   O=$S/solver/unxsogi.pro && W=$S/workdir/unxsogi.pro &&  mkdir -p
   $W/LinkTarget/Executable/ && g++ '-Wl,-rpath,$ORIGIN/../../ure-link/lib'
   -Wl,-rpath,$O/lib -Wl,-rpath-link,$O/lib -Wl,-rpath-link,/lib:/usr/lib
   -Wl,-z,combreloc -L$O/lib -L/usr/lib -L/usr/local/bin -L/usr/dt/lib
   -L/usr/openwin/lib  -Wl,--hash-style=gnu -Wl,--dynamic-list-cpp-new
   -Wl,--dynamic-list-cpp-typeinfo -Wl,-Bsymbolic-functions  -Wl,-O1
   $W/CxxObject/idlc/source/idlcmain.o $W/CxxObject/idlc/source/idlc.o
   $W/CxxObject/idlc/source/idlccompile.o
   $W/CxxObject/idlc/source/idlcproduce.o
   $W/CxxObject/idlc/source/errorhandler.o $W/CxxObject/idlc/source/options.o
   $W/CxxObject/idlc/source/fehelper.o
   $W/CxxObject/idlc/source/astdeclaration.o
   $W/CxxObject/idlc/source/astscope.o $W/CxxObject/idlc/source/aststack.o
   $W/CxxObject/idlc/source/astdump.o $W/CxxObject/idlc/source/astinterface.o
   $W/CxxObject/idlc/source/aststruct.o
   $W/CxxObject/idlc/source/aststructinstance.o
   $W/CxxObject/idlc/source/astoperation.o
   $W/CxxObject/idlc/source/astconstant.o $W/CxxObject/idlc/source/astenum.o
   $W/CxxObject/idlc/source/astarray.o $W/CxxObject/idlc/source/astunion.o
   $W/CxxObject/idlc/source/astexpression.o
   $W/CxxObject/idlc/source/astservice.o
   $W/GenCxxObject/YaccTarget/idlc/source/parser.o
   $W/GenCxxObject/LexTarget/idlc/source/scanner.o -Wl,--start-group 
   -Wl,--end-group -Wl,--no-as-needed  -lm -lnsl -lsocket  -lreglo -luno_sal

   -luno_salhelpergcc3 -o $W/LinkTarget/Executable/idlc



I have attached the solenv/gbuild/platform/solaris.mk I am using.

Any help would be greatly appreciated.
Thanks a lot.
Ciao,
Eric

P.S. If this isn't known yet, boost/boost.6139.clang.patch needs a dos2unix 
running on it.


# -*- Mode: makefile-gmake; tab-width: 4; indent-tabs-mode: t -*-
#
# This file is part of the LibreOffice project.
#
# This Source Code Form is subject to the terms of the Mozilla Public
# License, v. 2.0. If a copy of the MPL was not distributed with this
# file, You can obtain one at http://mozilla.org/MPL/2.0/.
#
# This file incorporates work covered by the following license notice:
#
#   Licensed to the Apache Software Foundation (ASF) under one or more
#   contributor license agreements. See the NOTICE file distributed
#   with this work for additional information regarding copyright
#   ownership. The ASF licenses this file to you under the Apache
#   License, Version 2.0 (the "License"); you may not use this file
#   except in compliance with the License. You may obtain a copy of
#   the License at http://www.apache.org/licenses/LICENSE-2.0 .
#

gb_DEVINSTALLROOT := $(DEVINSTALLDIR)/opt

gb_PROGRAMDIRNAME := program

include $(GBUILDDIR)/platform/com_GCC_defs.mk

gb_MKTEMP := mktemp -t gbuild.XX

ifneq ($(origin AR),default)
gb_AR := $(AR)
endif

# use CC/CXX if they are nondefaults
ifneq ($(origin CC),default)
gb_CC := $(CC)
gb_GCCP := $(CC)
endif
ifneq ($(origin CXX),default)
gb_CXX := $(CXX)
endif

ifeq ($(CPUNAME),SPARC)
gb_CPUDEFS := -D__sparcv8plus
endif

gb_OSDEFS := \
-D$(OS) \
-DSYSV \
-DSUN \
-DSUN4 \
-D_REENTRANT \
-D_POSIX_PTHREAD_SEMANTICS \
-D_PTHREADS \
-DUNIX \
-DUNX \
$(PTHREAD_CFLAGS) \

gb_CFLAGS := \
$(gb_CFLAGS_COMMON) \
-fPIC \
-Wdeclaration-after-statement \
-Wshadow \

gb_CXXFLAGS := \
$(gb_CXXFLAGS_COMMON) \
-fPIC \
-Wshadow \
-Woverloaded-virtual \
-Wno-non-virtual-dtor \

# enable debug STL
ifeq ($(gb_ENABLE_DBGUTIL),$(true))
gb_COMPILERDEFS += \
-D_GLIBCXX_DEBUG \

endif

gb_CCVER := $(shell $(gb_CC) -dumpversion | $(gb_AWK) -F. -- '{ print 
$$1*1+$$2*100+$$3 }')
gb_GccLess460 := $(shell expr $(gb_CCVE

Re: core dump when compiling LibreOffice ~master

[Jean-Tiare LE BIGOT]

I had the same problem. The crash is in sot_test_sot. from what I 
remember, there is a test throwing a RuntimeException. An important 
point is that It occurred to me only in Ubuntu 12.10 *alpha*. Moving 
back to the previous stable version "fixed" the problem. Sadly, I've not 
been able to find the exact source of the problem.


Le 2012-08-07 16:56, John Smith a écrit :
On Tue, Aug 7, 2012 at 4:53 PM, John Smith  
wrote:
On Tue, Aug 7, 2012 at 3:37 PM, Caolán McNamara  
wrote:


It's be helpful to find the core.* of the coredump and get a 
backtrace

from it.


I was under the impression that I already attached a trace with my
first email ? Anyway, here it is again.

Please ignore that email; I forgot I created a bug in bugzilla for 
it.

https://bugs.freedesktop.org/show_bug.cgi?id=53155
Please use that instead.

Thanks,


John Smith
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice


--
Jean-Tiare LE BIGOT
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: core dump when compiling LibreOffice ~master

[John Smith]

On Tue, Aug 7, 2012 at 4:53 PM, John Smith  wrote:
> On Tue, Aug 7, 2012 at 3:37 PM, Caolán McNamara  wrote:
>>
>> It's be helpful to find the core.* of the coredump and get a backtrace
>> from it.
>>
> I was under the impression that I already attached a trace with my
> first email ? Anyway, here it is again.
>
Please ignore that email; I forgot I created a bug in bugzilla for it.
https://bugs.freedesktop.org/show_bug.cgi?id=53155
Please use that instead.

Thanks,


John Smith
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: core dump when compiling LibreOffice ~master

[John Smith]

On Tue, Aug 7, 2012 at 3:37 PM, Caolán McNamara  wrote:
>
> It's be helpful to find the core.* of the coredump and get a backtrace
> from it.
>
I was under the impression that I already attached a trace with my
first email ? Anyway, here it is again.


Regards,


John Smith


stacktrace.txt.gz
Description: GNU Zip compressed data
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: core dump when compiling LibreOffice ~master

[Caolán McNamara]

On Mon, 2012-08-06 at 14:48 +0200, John Smith wrote:
> On Mon, Aug 6, 2012 at 10:22 AM, John Smith  wrote:
> > Hi,
> >
> > My configure and make lines were:
> > LDFLAGS+='-fprofile-arcs' CFLAGS+='-fprofile-arcs -ftest-coverage'
> > CXXFLAGS+='-fprofile-arcs -ftest-coverage' CPPFLAGS+='-fprofile-arcs
> > -ftest-coverage' ./configure --enable-debug --with-system-libcmis=no
> > --with-system-saxon=no --with-system-libs
> >  LDFLAGS+='-fprofile-arcs' CFLAGS+='-fprofile-arcs -ftest-coverage'
> > CXXFLAGS+='-fprofile-arcs -ftest-coverage' CPPFLAGS+='-fprofile-arcs
> > -ftest-coverage' make

(core dumped) $O/bin/uno -ro $O/xml/uno_services.rdb -ro
$O/bin/udkapi.rdb -ro $W/UnoApiTarget/bridgetest.rdb -s
com.sun.star.test.bridge.BridgeTest --
com.sun.star.test.bridge.CppTestObject
make[2]: ***

It's be helpful to find the core.* of the coredump and get a backtrace
from it.

Though I presume its the various exciting profiling options which trip
it up.

C.

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: core dump when compiling LibreOffice ~master

[John Smith]

On Mon, Aug 6, 2012 at 10:22 AM, John Smith  wrote:
> Hi,
>
>
> When I try to compile LibreOffice ~master on Linux, I get a build
> failure and a core dump.
>
> I have attached the build_error.log and the stack trace.
>
>
> My configure and make lines were:
> LDFLAGS+='-fprofile-arcs' CFLAGS+='-fprofile-arcs -ftest-coverage'
> CXXFLAGS+='-fprofile-arcs -ftest-coverage' CPPFLAGS+='-fprofile-arcs
> -ftest-coverage' ./configure --enable-debug --with-system-libcmis=no
> --with-system-saxon=no --with-system-libs
>  LDFLAGS+='-fprofile-arcs' CFLAGS+='-fprofile-arcs -ftest-coverage'
> CXXFLAGS+='-fprofile-arcs -ftest-coverage' CPPFLAGS+='-fprofile-arcs
> -ftest-coverage' make
>
>
> Any and all help is appreciated,
>
>
> Regards,
>
>
> John Smith.

I decided to file a bug in LibreOffice bugzilla for this one, since it
is a component of LibreOffice itself
(src/libreoffice/solver/unxlngi6.pro/bin/uno) that is dumping core
here.

https://bugs.freedesktop.org/show_bug.cgi?id=53155


- John Smith
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [Libreoffice] [PATCH] compiling libreoffice-3-4 sal module with dbglevel=2: init test fails

[Lionel Elie Mamane]

On Fri, Aug 05, 2011 at 04:19:49PM +0100, Caolán McNamara wrote:
> On Thu, 2011-08-04 at 12:07 +0200, Lionel Elie Mamane wrote:

>> A two-day-old clone of branch libreoffice-3-4 fails to build module
>> sal with dblevel=2, unittest fails: osl_old_test_file.cxx assertion
>> "#failure 1.1": the failure is that blah/a/.. is not transformed to
>> "blah", but stays "blah/a/.."

>> That's because in my setup symlinks are allowed, and in that setup
>> osl_getAbsoluteFileURL does not touch the last component (the part
>> after the last slash).

> Presumably this is with environmental variable SAL_ALLOW_LINKOO_SYMLINKS
> set to true, i.e. building in an environment where . ./ooenv was
> previously done to set the hacky SAL_ALLOW_LINKOO_SYMLINKS to true

"grep -i symlink LinuxX86-64Env.Set.sh" does not find anything. OTOH,
the C++ source code looks like

oslFileError osl_getAbsoluteFileURL(rtl_uString*  ustrBaseDirURL, rtl_uString* 
ustrRelativeURL, rtl_uString** pustrAbsoluteURL)
{
static char *allow_symlinks = getenv( "SAL_ALLOW_LINKOO_SYMLINKS");

if (!allow_symlinks)
{
  (...)
}
else
{
  // THIS BRANCH IS TAKEN
  (...)
}
solenv/bin/linkoo:export SAL_ALLOW_LINKOO_SYMLINKS=1}

so SAL_ALLOW_LINKOO_SYMLINKS must be set in some way somewhere.

Ah, I found:
solenv/bin/linkoo:export SAL_ALLOW_LINKOO_SYMLINKS=1
sc/source/ui/vba/testvba/runTests.pl: $ENV{"SAL_ALLOW_LINKOO_SYMLINKS"} = "1";
toolkit/workben/layout/TEST:export SAL_ALLOW_LINKOO_SYMLINKS=1


> I worked around this at
> http://cgit.freedesktop.org/libreoffice/ure/commit/?id=784641a19811c22208e7db8efa150e744348d8bf
> in master.

> Don't know, in the light of that, if we still need this proposed patch
> then ?

Your workaround unsets that variable for the test, effectively saying
that the "else" branch above should not be tested at all. From a
general point of view, this looks fishy; if a specific subtest does
not apply to this branch, I'd rather disable just that specific
subtest; I mean the file sal/qa/osl/file/osl_old_test_file.cxx tests
for the SAL_ALLOW_LINKOO_SYMLINKS environment variable, and if it is
set, runs only the subtests that should pass. For example, in

const char * const aSource1[] =
{
"a", "file:///" TEST_VOLUME "bla/a",
///TODO: check if last slash must be omitted in resolved path.
//"a/"   , "file:///" TEST_VOLUME "bla/a",
"../a" , "file:///" TEST_VOLUME "a" ,
"a/.." , "file:///" TEST_VOLUME "bla/",
"a/../b" , "file:///" TEST_VOLUME "bla/b",
".."   , "file:///" TEST_VOLUME "",
"a/b/c/d"   , "file:///" TEST_VOLUME "bla/a/b/c/d",
"a/./c"   , "file:///" TEST_VOLUME "bla/a/c",
"file:///bla/blub", "file:///"  "bla/blub",
0 , 0
}

it would not run two of them, the ones where the first part of the
pair ends with "..".

To get into this specific case, my guess of the semantics that one
wants from osl_getAbsoluteFileURL tells me that these subtests should
pass, too. I mean, if "/flob/foo/bar" is a symlink to qux/bar, do we
want osl_getAbsoluteFileURL("/flob/foo/bar/..") to:

 - CASE 1: effectively refer to the directory "/flob/foo/qux", and
   thus it should remain the string "/flob/foo/bar/.."

 - CASE 2: effectively refer to the directory "/flob/foo", and thus it
   should become the string "/flob/foo"

My guess is for CASE 2. For example, case 2 maintains the invariant
that these two refer the same file or directory:

 - result of osl_getAbsoluteFileURL("/flob/foo/bar/../bral")

 - concatenation of:
   1) result of osl_getAbsoluteFileURL("/flob/foo/bar/..)
   2) the string "/bral"

and that file is /flob/foo/bral. Which is then not the same file as
"/flob/foo/bar/../bral", which is the file "/flob/foo/qux/bral".

On the other hand, CASE 1 maintains the invariant that these two refer
to the same file or directory:

 - the string "/flob/foo/bar/../bral"

 - concatenation of:
   1) result of osl_getAbsoluteFileURL("/flob/foo/bar/..)
   2) the string "/bral"

and that file is "/flob/foo/qux/bral", which is then _not_ the same
file as referred to by the result of
osl_getAbsoluteFileURL("/flob/foo/bar/../bral"), namely
"/flob/foo/bral".


If I'm right in my guess and CASE 2 is true, then we should:

1) revert your patch
2) apply my patch


But if from your better understanding of the codebase you tell me CASE
1 is true, then IMHO we should:

1) revert your patch
2) disable only the subtests that don't apply in the case they don't,
   but still run the others. I'll gladly write that patch something like:


@@
 const char * const aSource1[] =
 {
 "a", "file:///" TEST_VOLUME "bla/a",
 ///TODO: check if last slash must be omitted in resolved path.
 //"a/"   , "file:///" TEST_VOLUME "bla/a",
 "../a" , "file:///" TEST_VOLUME "a" ,
- "a/.." , "file:///" TEST_VOLUME "bla/",
 "a/../b" , "file:///" TEST_VOLUME "bla/b",
- ".."   , "file:///" TEST_VOLUME "",
 "a/b/c/d"   , "file:///" TEST_VOLUME "bla/a/b/c/d",
 "a/./c"   , "file:///" TEST_

Re: [Libreoffice] [PATCH] compiling libreoffice-3-4 sal module with dbglevel=2: init test fails

[Caolán McNamara]

On Thu, 2011-08-04 at 12:07 +0200, Lionel Elie Mamane wrote:
> A two-day-old clone of branch libreoffice-3-4 fails to build module
> sal with dblevel=2, unittest fails: osl_old_test_file.cxx assertion
> "#failure 1.1": the failure is that blah/a/.. is not transformed to
> "blah", but stays "blah/a/.."
> 
> That's because in my setup symlinks are allowed, and in that setup
> osl_getAbsoluteFileURL does not touch the last component (the part
> after the last slash).

Presumably this is with environmental variable SAL_ALLOW_LINKOO_SYMLINKS
set to true, i.e. building in an environment where . ./ooenv was
previously done to set the hacky SAL_ALLOW_LINKOO_SYMLINKS to true

I worked around this at
http://cgit.freedesktop.org/libreoffice/ure/commit/?id=784641a19811c22208e7db8efa150e744348d8bf
in master.

Don't know, in the light of that, if we still need this proposed patch
then ?

C.

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[Libreoffice] [PATCH] compiling libreoffice-3-4 sal module with dbglevel=2: init test fails

[Lionel Elie Mamane]

A two-day-old clone of branch libreoffice-3-4 fails to build module
sal with dblevel=2, unittest fails: osl_old_test_file.cxx assertion
"#failure 1.1": the failure is that blah/a/.. is not transformed to
"blah", but stays "blah/a/.."

That's because in my setup symlinks are allowed, and in that setup
osl_getAbsoluteFileURL does not touch the last component (the part
after the last slash).

The attached patch changes that to _still_ treating the special names
"." and "..": they cannot be symlinks, so they don't have to be
protected against treatment and we want them treated.

-- 
Lionel
>From 9a5d3dddb140912daba08927fb1d58d3516d60f3 Mon Sep 17 00:00:00 2001
From: Lionel Elie Mamane 
Date: Thu, 4 Aug 2011 11:42:29 +0200
Subject: [PATCH] osl_getAbsoluteFileURL symlinks allowed: ".", ".." still treated

---
 sal/osl/unx/file_url.cxx |4 +++-
 1 files changed, 3 insertions(+), 1 deletions(-)

diff --git a/sal/osl/unx/file_url.cxx b/sal/osl/unx/file_url.cxx
index 21fbe37..7caabcd 100644
--- a/sal/osl/unx/file_url.cxx
+++ b/sal/osl/unx/file_url.cxx
@@ -722,7 +722,9 @@ oslFileError osl_getAbsoluteFileURL(rtl_uString*  ustrBaseDirURL, rtl_uString* u
 rtl::OUString base;
 sal_Int32 last_slash = unresolved_path.lastIndexOf( UNICHAR_SLASH );
 
-if (last_slash >= 0 && last_slash + 1 < unresolved_path.getLength())
+if (last_slash >= 0 && last_slash + 1 < unresolved_path.getLength()
+&& ! ( last_slash + 2 == unresolved_path.getLength() && unresolved_path.matchAsciiL(RTL_CONSTASCII_STRINGPARAM("."), last_slash + 1) )
+&& ! ( last_slash + 3 == unresolved_path.getLength() && unresolved_path.matchAsciiL(RTL_CONSTASCII_STRINGPARAM(".."), last_slash + 1) ))
 {
 base = unresolved_path.copy(last_slash+1);
 unresolved_path = unresolved_path.copy(0, last_slash);
-- 
1.7.2.5

___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

[Libreoffice] compiling libreoffice-3-4: copy of libcuuc as dependency of svtools fails (missing source operand)

[Lionel Elie Mamane]

Hi,

A one-day-old clone of branch libreoffice-3-4 fails to build for me.

Directory svtools fails with:

R=/home/master/src/libreoffice && S=$R/libreoffice-3.4 && 
O=$S/solver/340/unxlngx6 && W=$O/workdir &&  mkdir -p $O/lib/ && /bin/cp 
--remove-destination  --force --preserve=timestamps  $O/lib/libicuuc.so 
/bin/cp: missing destination file operand after 
`/home/master/src/libreoffice/libreoffice-3.4/solver/340/unxlngx6/lib/libicuuc.so'
Try `/bin/cp --help' for more information.
make: *** 
[/home/master/src/libreoffice/libreoffice-3.4/solver/340/unxlngx6/lib/libicuuc.so]
 Error 1

So I did:

cd ../icu
build
cd ../svtools
R=/home/master/src/libreoffice && S=$R/libreoffice-3.4 && 
O=$S/solver/340/unxlngx6 && cp -vi --remove-destination  --force 
--preserve=timestamps $S/clone/libs-extern-sys/icu/unxlngx6/lib/libicuuc.so 
$O/lib/libicuuc.so 
make -r

and that worked, at least as far as directory svtools is concerned :)

-- 
Lionel
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice

Re: [Libreoffice] Error compiling LibreOffice for Win32 (WinXP) - module: shell

[Robert Sedak]

 Hi Jesús,

On 6.10.2010 12:22, Jesús Corrius wrote:
> Hi Robert,
>
> On Wed, Oct 6, 2010 at 8:56 AM, Robert Sedak  wrote:
>>
>> Hello to all,
>>
>> I have error in compiling module shell.
>> Please can you see what is the problem?
>>
>> System: Windows XP
>> Compiling via Cygwin
>> autogen parameters:
>> ./autogen.sh  --with-distro=LibreOfficeWin32 --with-git
>> --with-num-cpus=2 --disable-odk  --disable-build-mozilla
>> --with-lang="hr" --with-use-shell=bash --disable-activex --with-win32
>> --with-cl-home="/cygdrive/c/Program Files/Microsoft Visual Studio 9.0/VC"
>> --with-csc-path="/cygdrive/c/WINDOWS/Microsoft.NET/Framework/v3.5"
>> --with-midl-path="/cygdrive/c/Program Files/Microsoft SDKs/Windows/v6.1/Bin"
>> --with-frame-home="/cygdrive/c/Program Files/Microsoft SDKs/Windows/v6.1"
>> --with-ant-home=/apache-ant-1.7.1
>> --with-jdk-home="/cygdrive/c/Program Files/Java/jdk1.6.0_18"
>> --with-nsis-path="/cygdrive/c/Program Files/NSIS"
>> --with-psdk-home="/cygdrive/c/Program Files/Microsoft SDKs/Windows/v6.1"
>> --disable-atl --disable-nss-module --without-junit --disable-fetch-external
>>
>> Entering
>> /cygdrive/f/cygwin/Robert/LibO/build/build/libreoffice-3.2.99.1/shell/source/all/zipfile
>> Making:all_zipfile.dpslo
>> Compiling: shell/source/all/zipfile/cl : Command line error D8038 :
>> invalid argument '-I. -I../../../wntmsci12.pro/inc/zipfile -I../inc
>> -I../../../inc/pch -I../../../inc -I../../../WIN/inc
>> -I../../../wntmsci12.pro/inc -I.
>> -IF:/cygwin/Robert/LibO/build/build/libreoffice-3.2.99.1/solver/330/wntmsci12.pro/inc
>> -IF:/cygwin/Robert/LibO/build/build/libreoffice-3.2.99.1/solver/330/wntmsci12.pro/inc/external
>> -IF:/cygwin/Robert/LibO/build/build/libreoffice-3.2.99.1/solver/330/wntmsci12.pro/inc
>> -IF:/cygwin/Robert/LibO/build/build/libreoffice-3.2.99.1/solenv/wntmsci12/inc
>> -IF:/cygwin/Robert/LibO/build/build/libreoffice-3.2.99.1/solenv/inc
>> -IF:/cygwin/Robert/LibO/build/build/libreoffice-3.2.99.1/res
>> -IF:/cygwin/Robert/LibO/build/build/libreoffice-3.2.99.1/solver/330/wntmsci12.pro/inc
>> -IC:/PROGRA~1/Java/JDK16~1.0_1/include/win32
>> -IC:/PROGRA~1/Java/JDK16~1.0_1/include
>> -IC:/PROGRA~1/MICROS~3/Windows/v6.1/include
>> -IC:/PROGRA~1/MICROS~1.0/VC/include -IC:/PROGRA~1/MIABE2~1/include
>> -IC:/PROGRA~1/MIABE2~1/include
>> -IF:/cygwin/Robert/LibO/build/build/libreoffice-3.2.99.1/solver/330/wntmsci12.pro/inc/offuh
>> -I. -I../../../res -I.'
>> dmake:  Error code 2, while making '../../../wntmsci12.pro/slo/zipfile.obj'
> The errors in the two messages are exactly the same, so I will only
> comment on this one.
>
> I tried to reproduce the the error in my Windows XP (all current
> Service Packs and patches installed) using the same command line
> options you are using and my system works. Can you give us more
> information about your system?
>
> Regards,
Thank you for answer.
I contacted Fridrich via IRC before I got jour message.
He assumed that path lenght is to big.

In this case soution is command :
substr o: f:\cygwin\Robert\LibO\build\
in windows command line.
After that in Cygwin I use path /cygdrive/o.

Best,
Robert Sedak
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice