Am 27.09.2018 um 08:28 schrieb Noel Grandin:
> On Thu, 27 Sep 2018 at 08:28, Noel Grandin wrote:
>> On Wed, 26 Sep 2018 at 23:58, Jan-Marek Glogowski
>> wrote:
>>
>>> For me it looks like a multi-threaded problem, where something isn't
>>> correct processed in order. I can reproduce crashes when
On Thu, 27 Sep 2018 at 08:28, Noel Grandin wrote:
>
>
> On Wed, 26 Sep 2018 at 23:58, Jan-Marek Glogowski
> wrote:
>
>> For me it looks like a multi-threaded problem, where something isn't
>> correct processed in order. I can reproduce crashes when I run a very
>> parallel make check often, but
On Wed, 26 Sep 2018 at 23:58, Jan-Marek Glogowski wrote:
> For me it looks like a multi-threaded problem, where something isn't
> correct processed in order. I can reproduce crashes when I run a very
> parallel make check often, but never was able to get something in a
> debugger...
>
>
mst
_
Hi Luke,
failures for CppunitTest_vcl_complextext have been around for a few months.
Nobody is able to reproduce them. IMHO they fail most time with loading of the
wrong font.
Looking at the current state of
https://tinderbox.libreoffice.org/MASTER/status.html, the master Jenkins
Windows buil
Over the years, I've probably done several hundred windows builds. Until now I
have never been bitten an intermittent Unit Test failures. This month, I have
started to see fairly regular failure of VclComplexTextTest::testArabic 3 out
of ~15 builds.
Here are 2 separate examples of the Jenkins_W