[OE-core] [PATCH 3/3] gtk+3: update 3.24.5 -> 3.24.8

[Richard Purdie]

On Mon, 2019-06-10 at 13:32 +0100, Richard Purdie wrote:
> On Mon, 2019-06-10 at 12:47 +0100, Richard Purdie wrote:
> > On Sat, 2019-06-08 at 10:05 +0100, Richard Purdie wrote:
> > > On Fri, 2019-06-07 at 14:19 +0200, Alexander Kanavin wrote:
> > > > Rebase 0003-Add-disable-opengl-configure-option.patch
> > > > 
> > > > Signed-off-by: Alexander Kanavin <alex.kanavin at gmail.com>
> > > > ---
> > > >  ...-Add-disable-opengl-configure-option.patch | 41 ++++++++++-
> > > > --
> > > > --
> > > > --
> > > > --
> > > >  .../gtk+/{gtk+3_3.24.5.bb => gtk+3_3.24.8.bb} |  4 +-
> > > >  2 files changed, 23 insertions(+), 22 deletions(-)
> > > >  rename meta/recipes-gnome/gtk+/{gtk+3_3.24.5.bb =>
> > > > gtk+3_3.24.8.bb}
> > > > (84%)
> > > 
> > > I retested with the date/time on the workers fixed and this
> > > series
> > > present. We saw:
> > > 
> > > https://autobuilder.yoctoproject.org/typhoon/#/builders/47/builds/693
> > > 
> > > so I think there might be something not quite right with opengl
> > > disabled (the main difference to nodistro I can think of?)
> > 
> > Its now in master:
> > 
> > https://autobuilder.yoctoproject.org/typhoon/#/builders/76/builds/687
> > 
> > so I think there is some kind of parallel make race in the new
> > gtk+?
> > 
> > The autobuilder is proving very tricky to read at the moment and
> > intermittent bugs are making it into master :(
> 
> I've realised the error wasn't:
> 
> Failed to open file “./gdk.gresource.xml”: No such file or directory
> 
> but
> 
> > glib-compile-resources --target=bloatpad-gresources.c --
> sourcedir=../../../gtk+-3.24.5/examples/bp --generate-source
> ../../../gtk+-3.24.5/examples/bp/bloatpad.gresources.xml
> > (glib-compile-resources:91029): GLib-ERROR **: 02:42:13.583:
> > creating
> thread 'gmain': Error creating thread: Resource temporarily
> unavailable
> > make[3]: *** [bloatpad-gresources.c] Trace/breakpoint trap
> 
> which sounds like a parallelism issue :/. Question is which resource
> we're running out of...

Looking at that machine, max user processes is 4096 which I think we've
had issues with in the past.

Cc'ing Michael.

Cheers,

Richard