[Libreoffice] MacOs Build failure in: unit test #2 on library ../../unxmacxi.pro/lib/libswdoctest.dylib

[Norbert Thiebaud]

On Fri, Jan 28, 2011 at 4:26 AM, Caolán McNamara <caolanm at redhat.com> wrote:
> On Fri, 2011-01-28 at 03:59 -0600, Norbert Thiebaud wrote:
>> On Fri, Jan 28, 2011 at 3:39 AM, Caolán McNamara <caolanm at redhat.com> wrote:
>> > On Thu, 2011-01-27 at 15:21 -0600, Norbert Thiebaud wrote:
>> >> ----------------------------------------------------------
>> >> - start unit test #2 on library ../../unxmacxi.pro/lib/libswdoctest.dylib
>> >> ----------------------------------------------------------
>> >> /bin/sh: line 1: 85233 Bus error
>> >
>> > So, can we get any debugging info for that, i.e. get gdb in there
>> >
>> > e.g. export GDBCPPUNITTRACE="gdb --args" and run dmake in the afflicted
>> > dir.
>>
>> (gdb) run
>> Starting program:
>> /Volumes/Raid0/tb/solver/330/unxmacxi.pro/bin/cppunittester
>> ../../unxmacxi.pro/lib/libswdoctest.dylib -headless -invisible
>> -env:UNO_SERVICES=file:///Volumes/Raid0/tb/clone/writer/sw/qa/core/../../unxmacxi.pro/misc/test-bigptrarray/services.rdb
>> -env:UNO_TYPES=file:///Volumes/Raid0/tb/clone/writer/sw/qa/core/../../unxmacxi.pro/misc/test-bigptrarray/types.rdb\
>> file:///Volumes/Raid0/tb/clone/writer/sw/qa/core/../../unxmacxi.pro/misc/test-bigptrarray/udkapi.rdb
>> -env:OOO_BASE_DIR=file:///Volumes/Raid0/tb/clone/writer/sw/qa/core/../../unxmacxi.pro/misc/test-bigptrarray
>> -env:BRAND_BASE_DIR=file:///Volumes/Raid0/tb/clone/writer/sw/qa/core/../../unxmacxi.pro/misc/test-bigptrarray
>> -env:UNO_USER_PACKAGES_CACHE=file:///Volumes/Raid0/tb/clone/writer/sw/qa/core/../../unxmacxi.pro/misc/test-bigptrarray
>> Reading symbols for shared libraries
>> .+++++...........................................................................................
>> done
>> Reading symbols for shared libraries
>> ............................................ done
>> Reading symbols for shared libraries ... done
>> Reading symbols for shared libraries . done
>> Reading symbols for shared libraries . done
>> Reading symbols for shared libraries .. done
>> Reading symbols for shared libraries . done
>>
>> Program received signal EXC_BAD_ACCESS, Could not access memory.
>> Reason: KERN_PROTECTION_FAILURE at address: 0x00000014
>> 0x06c91710 in ResMgr::GetResource ()
>> (gdb) bt
>> #0  0x06c91710 in ResMgr::GetResource ()
>
>> yes it _has_ worked, I'd say a week or so ago
>
> So this has got to be something to do with STAR_RESOURCEPATH, that's
> supposed to set the location where the .res files exist in order to find
> them, but bodge the path.
>
> So stick
>
> echo $(STAR_RESOURCEPATH) before "start unit test" and see what the
> output is.

STAR_RESOURCEPATH =
/Volumes/Raid0/tb/clone/writer/sw/qa/core/../../unxmacxi.pro/bin:/Volumes/Raid0/tb/solver/330/unxmacxi.pro/bin

n_th at tpamac /Volumes/Raid0/tb/sw/unxmacxi.pro/bin $ll
total 24
-rw-------  1 n_th  staff     0 Jan 28 04:34 3BEtea
-rw-------  1 n_th  staff  4025 Jan 28 04:34 EVSwKa
-rw-------  1 n_th  staff  4025 Jan 28 03:56 ErfdWl
-rw-------  1 n_th  staff  4025 Jan 27 15:16 MZqgRE
-rw-------  1 n_th  staff     0 Jan 28 03:56 es41Xg
-rw-------  1 n_th  staff     0 Jan 27 15:16 mGf3xp

that looks weird ?

and solver/330/unxmacxi.pro/bin contains a buttload of .res file...
6864 of them... too many to list them all here :-)

I could not find anything in the git log that attracted attention :-(

I may resort to brute force and just try a bisection :-/

Norbert