Re: Frequent recent killing of runaway CppunitTest_sw_layoutwriter on Jenkins (was: ESC meeting minutes: 2022-04-14)

[Caolán McNamara <caolanm@xxxxxxxxxx>]

On Wed, 2022-04-27 at 16:49 +0200, Noel Grandin wrote:
> On Wed, 27 Apr 2022 at 14:08, Stephan Bergmann <sbergman@xxxxxxxxxx>
> wrote:
> > The take-away probably is that we should make sure to only use
> > fonts in our `make check` tests that are covered by --with-fonts.
> 
> Possibly we can add an assert to that effect somewhere in the font
> substutuaion code?

Playing around, for the fontconfig case in FontCfgWrapper::getFontSet
we have
   addFontSet( FcSetSystem );
   addFontSet( FcSetApplication );
where the FcSetSystem adds the system fonts and the other adds the
application fonts. Omitting the first one would restrict to only our
bundled fonts (but break make check on distro builds that use system
fonts).

Maybe there is some scope for asserting if a font is substituted if
gb_CppunitTest_use_more_fonts is used.