Re: [Libreoffice] Fwd: First build resumed but failed on incomplete dependancies
On Fri, May 6, 2011 at 11:54 AM, Christian Lohmaier wrote: > > No, not for "this". > When you force configure to use ccache, but the build doesn't actually > use ccache, the break is what you rightfully get. And of course you also deserve the break when using a version with the argument parsing bug that makes --ccache-skip unusable - detecting this is the whole point of the configure check. Reading the snippets again, It's more likely that not no ccache is used, but an incompatible one. i686-apple-darwin10-g++-4.0.1: objective-c++: No such file or directory is a classical symptom of this... ciao Christian ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] Fwd: First build resumed but failed on incomplete dependancies
Hi Michael, *, On Fri, May 6, 2011 at 11:36 AM, Michael Meeks wrote: > H Peter, > > On Thu, 2011-05-05 at 13:54 -0400, Peter Teeson wrote: > >> > > Hm, did you give --enable-ccache-skip on autogen/configure? It >> > > seems your ccache does not properly eat those arguments. > .. >> > Where do we go from here? > > Thorsten committed a fix for this to master I believe; No, not for "this". When you force configure to use ccache, but the build doesn't actually use ccache, the break is what you rightfully get. > I don't think > you need --enable-ccache-skip if you have a recent ccache (at least so a > recent mail suggested). Yes, that's true, but that's a different problem. Peter's build did not use ccache. All versions of ccache know --ccache-skip, so when it fails because of unknown option, that is passed to real compiler directly, hence ccache not used at all. ciao Christian ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: [Libreoffice] Fwd: First build resumed but failed on incomplete dependancies
H Peter, On Thu, 2011-05-05 at 13:54 -0400, Peter Teeson wrote: > > > Hm, did you give --enable-ccache-skip on autogen/configure? It > > > seems your ccache does not properly eat those arguments. .. > > Where do we go from here? Thorsten committed a fix for this to master I believe; I don't think you need --enable-ccache-skip if you have a recent ccache (at least so a recent mail suggested). So - I would do: $ ./g pull -r # to get the latest code Edit --enable-ccache-skip out of your autogen.lastrun, re-run autogen.sh (no argument necessary it'll just use autogen.lastrun), and re-run make. Hopefully life will be better then, and (hopefully) we'll have fixed the problem for everyone in the meantime too - so good work. Thanks ! Michael. -- michael.me...@novell.com <><, Pseudo Engineer, itinerant idiot ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
[Libreoffice] Fwd: First build resumed but failed on incomplete dependancies
Sorry forgot to cc the list.
Begin forwarded message:
> From: Peter Teeson
> Date: May 5, 2011 1:53:59 PM EDT
> To: Thorsten Behrens
> Subject: Re: [Libreoffice] First build resumed but failed on incomplete
> dependancies
>
> On 2011-05-05, at 11:47 AM, Thorsten Behrens wrote:
>> Peter Teeson wrote:
>>> (0) In fact I have the Build instructions handy as I printed them out for
>>> easy reference.
>>> It's not clear to me from them as to which directory one should be in
>>> to issue them.
>>>
>>> Is it correct to assume git/libo?
>>>
>> Yes
>>
>>> (1) "please re-run build inside each one to isolate the problem"
>>>doesn't make sense to me. Inside what?
>>>
>> Inside those toplevel directories - e.g. "vcl"
>>
>>> (2) looks like it was expecting a file or directory named objective-c++
>>>
>>> (3) log extract for vcl
>>> .
>>> .
>>> Compiling: ucb/source/ucp/tdoc/tdoc_stgelems.cxx
>>> : &&
>>> DYLD_LIBRARY_PATH=/Users/pteeson/git/libo/solver/300/unxmacxi.pro/lib${DYLD_LIBRARY_PATH:+:${DYLD_LIBRARY_PATH}}
>>> xsltproc --nonet --stringparam uri \
>>> 'vnd.sun.star.expand:$OOO_BASE_DIR/program/ucptdoc1.uno.dylib' -o
>>> ../../../unxmacxi.pro/misc/ucptdoc1.component \
>>> /Users/pteeson/git/libo/solenv/bin/createcomponent.xslt
>>> ucptdoc1.component
>>> Compiling: shell/unxmacxi.pro/misc/localebe1.uno_version.c
>>> Compiling: shell/source/backends/macbe/macbecdef.cxx
>>> Compiling: shell/source/backends/macbe/macbackend.cxx
>>> i686-apple-darwin10-g++-4.0.1: objective-c++: No such file or directory
>>> i686-apple-darwin10-g++-4.0.1: language --ccache-skip not recognized
>>> i686-apple-darwin10-g++-4.0.1: language --ccache-skip not recognized
>>>
>> Hm, did you give --enable-ccache-skip on autogen/configure? It seems
>> your ccache does not properly eat those arguments.
>>
>> Cheers,
>>
>> -- Thorsten
>
>
> Yes I did indeed
> And Xcode is 3.2.4 and SDK 10.4 is installed
> My MacPro 2x2.66Ghz Quad-core "Nehalem", 6GB RAM
> Mac OS X 10.6.6 , ... all standard from Apple
>
> Where do we go from here?
>
> respect
>
> Peter
>
> From my log file - NOTE the rm
>
> Gandalf:git pteeson$ rm -dfR libo
> Gandalf:git pteeson$ git clone
> git://anongit.freedesktop.org/libreoffice/bootstrap libo
> Cloning into libo...
> remote: Counting objects: 97475, done.
> remote: Compressing objects: 100% (25906/25906), done.
> remote: Total 97475 (delta 77036), reused 89469 (delta 70540)
> Receiving objects: 100% (97475/97475), 21.17 MiB | 31 KiB/s, done.
> Gandalf:libo pteeson$ ./autogen.sh --with-max-jobs=8 --with-num-cpus=8
> --disable-mozilla --disable-systray --enable-ccache-skip
> writing args to autogen.lastrun
> running ./configure with '--with-max-jobs=8' '--with-num-cpus=8'
> '--disable-mozilla' '--disable-systray' '--enable-ccache-skip'
>
___
LibreOffice mailing list
LibreOffice@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/libreoffice
