[Bug 51324] Imperium Great Battles of Rome can't play videos

[WineHQ Bugzilla]

https://bugs.winehq.org/show_bug.cgi?id=51324

--- Comment #3 from Lorenzo Ferrillo <lorenzofer at live.it> ---
Created attachment 70190
  --> https://bugs.winehq.org/attachment.cgi?id=70190
amstream repsect renderallstreams flag

Hi Gijs
A demo is not aviable.

The game creqate the multimeda_stream object and then call
IAMMultiMediaStream_Initialize with type = 0, flags= 0x1 (AMMSF_NOGRAPHTHREAD)
and nullptr graph.

Then it call IAMMultiMediaStream_OpenFile  with the movie path and the flag 0x1
(AMMSF_RENDERALLSTREAMS)

Now before commit bdf860033860973bf2b516a1083d5bea86ad8c14 the codepath
involved 
IGraphBuilder_Render(This->graph, This->ipin);
that probably was creating the intermediate filters or renderers. And it
worked.

After the commit it use IFilterGraph2_RenderEx with
AM_RENDEREX_RENDERTOEXISTINGRENDERERS . This fails if no suitable renderer
aren't present.  And it's probably the intended behaviours  of
IAMMultiMediaStream_OpenFile in normal conditions. 

But Imperium use OpenFile with AMMSF_RENDERALLSTREAMS. MSDN state that "The
AMMSF_RENDERALLSTREAMS flag will create default rendering filters for video and
audio if they do not exist"
So it seems that in this specific case the filters are supposed to be created
dynamically, and this application actually require this behaviour to be
respected.

I tested that if I use IFilterGraph2_RenderEx passing 0 instead of
AM_RENDEREX_RENDERTOEXISTINGRENDERERS when AMMSF_RENDERALLSTREAMS is passed the
application works. It also works if it revert to use IGraphBuilder_Render for
this specific case.

Attaching a patch for the first option. I don't know if this is what is
supposed to happen, and testing the exact behaviours with the testbot may not
be trivial.

-- 
Do not reply to this email, post in Bugzilla using the
above URL to reply.
You are receiving this mail because:
You are watching all bug changes.