[PATCH 1/5] mfplat: Fix asynchronous object creation from URL.

[Derek Lesho]

Signed-off-by: Derek Lesho <dlesho(a)codeweavers.com>
---
 dlls/mfplat/main.c         | 4 +++-
 dlls/mfplat/tests/mfplat.c | 2 --
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/dlls/mfplat/main.c b/dlls/mfplat/main.c
index a6ff97ae04a..8ed18b1b87f 100644
--- a/dlls/mfplat/main.c
+++ b/dlls/mfplat/main.c
@@ -5750,6 +5750,8 @@ static HRESULT resolver_handler_end_create(struct source_resolver *resolver, enu
     if (!(queued_result = heap_alloc_zero(sizeof(*queued_result))))
         return E_OUTOFMEMORY;
 
+    queued_result->origin = origin;
+
     IRtwqAsyncResult_GetObject(inner_result, &handler.handler);
 
     switch (origin)
@@ -5788,8 +5790,8 @@ static HRESULT resolver_handler_end_create(struct source_resolver *resolver, enu
 
         if (SUCCEEDED(RtwqCreateAsyncResult(queued_result->object, data->pCallback, caller_state, &caller_result)))
         {
+            queued_result->inner_result = caller_result;
             RtwqInvokeCallback(caller_result);
-            IRtwqAsyncResult_Release(caller_result);
         }
     }
 
diff --git a/dlls/mfplat/tests/mfplat.c b/dlls/mfplat/tests/mfplat.c
index 55c07adb1fe..a5388840bc5 100644
--- a/dlls/mfplat/tests/mfplat.c
+++ b/dlls/mfplat/tests/mfplat.c
@@ -343,12 +343,10 @@ static HRESULT WINAPI test_create_from_url_callback_Invoke(IMFAsyncCallback *ifa
 
     object = NULL;
     hr = IMFSourceResolver_EndCreateObjectFromURL(resolver, result, &obj_type, &object);
-todo_wine
     ok(hr == S_OK, "Failed to create an object, hr %#x.\n", hr);
 
     hr = IMFAsyncResult_GetObject(result, &object2);
     ok(hr == S_OK, "Failed to get result object, hr %#x.\n", hr);
-todo_wine
     ok(object2 == object, "Unexpected object.\n");
 
     if (object)
-- 
2.28.0

[PATCH 2/5] mfmediaengine: Provide the partial topology to the media session.

[Derek Lesho]

Signed-off-by: Derek Lesho <dlesho(a)codeweavers.com>
---
 dlls/mfmediaengine/main.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/dlls/mfmediaengine/main.c b/dlls/mfmediaengine/main.c
index d93e2b2f38c..d1261e3cd73 100644
--- a/dlls/mfmediaengine/main.c
+++ b/dlls/mfmediaengine/main.c
@@ -418,7 +418,10 @@ static HRESULT media_engine_create_topology(struct media_engine *engine, IMFMedi
             {
                 sd_video = sd;
                 IMFStreamDescriptor_AddRef(sd_video);
+                /* TODO: reintroduce this once we set up video stream nodes */
+#if 0
                 IMFPresentationDescriptor_SelectStream(pd, i);
+#endif
             }
 
             IMFMediaTypeHandler_Release(type_handler);
@@ -480,6 +483,9 @@ static HRESULT media_engine_create_topology(struct media_engine *engine, IMFMedi
             if (audio_src)
                 IMFTopologyNode_Release(audio_src);
         }
+
+        if (SUCCEEDED(hr))
+            hr = IMFMediaSession_SetTopology(engine->session, MFSESSION_SETTOPOLOGY_IMMEDIATE, topology);
     }
 
     if (topology)
-- 
2.28.0

[PATCH 2/5] mfmediaengine: Provide the partial topology to the media session.

[Nikolay Sivov]

On 10/17/20 1:02 AM, Derek Lesho wrote:
> Signed-off-by: Derek Lesho <dlesho(a)codeweavers.com>
> ---
>  dlls/mfmediaengine/main.c | 6 ++++++
>  1 file changed, 6 insertions(+)
>
> diff --git a/dlls/mfmediaengine/main.c b/dlls/mfmediaengine/main.c
> index d93e2b2f38c..d1261e3cd73 100644
> --- a/dlls/mfmediaengine/main.c
> +++ b/dlls/mfmediaengine/main.c
> @@ -418,7 +418,10 @@ static HRESULT media_engine_create_topology(struct media_engine *engine, IMFMedi
>              {
>                  sd_video = sd;
>                  IMFStreamDescriptor_AddRef(sd_video);
> +                /* TODO: reintroduce this once we set up video stream nodes */
> +#if 0
>                  IMFPresentationDescriptor_SelectStream(pd, i);
> +#endif
>              }
> What do we need to set them up?
>
> IMFMediaTypeHandler_Release(type_handler);
> @@ -480,6 +483,9 @@ static HRESULT media_engine_create_topology(struct media_engine *engine, IMFMedi
>              if (audio_src)
>                  IMFTopologyNode_Release(audio_src);
>          }
> +
> +        if (SUCCEEDED(hr))
> +            hr = IMFMediaSession_SetTopology(engine->session, MFSESSION_SETTOPOLOGY_IMMEDIATE, topology);
>      }
> That's fine, either here or returning new topology to set it in the
> callback.
>
> if (topology)

[PATCH 2/5] mfmediaengine: Provide the partial topology to the media session.

[Derek Lesho]

On 10/20/20 6:43 AM, Nikolay Sivov wrote:
> On 10/17/20 1:02 AM, Derek Lesho wrote:
> > Signed-off-by: Derek Lesho <dlesho(a)codeweavers.com>
> > ---
> >   dlls/mfmediaengine/main.c | 6 ++++++
> >   1 file changed, 6 insertions(+)
> >
> > diff --git a/dlls/mfmediaengine/main.c b/dlls/mfmediaengine/main.c
> > index d93e2b2f38c..d1261e3cd73 100644
> > --- a/dlls/mfmediaengine/main.c
> > +++ b/dlls/mfmediaengine/main.c
> > @@ -418,7 +418,10 @@ static HRESULT media_engine_create_topology(struct media_engine *engine, IMFMedi
> >               {
> >                   sd_video = sd;
> >                   IMFStreamDescriptor_AddRef(sd_video);
> > +                /* TODO: reintroduce this once we set up video stream nodes */
> > +#if 0
> >                   IMFPresentationDescriptor_SelectStream(pd, i);
> > +#endif
> >               }
> > What do we need to set them up?

I think we need the sample grabber in frame server mode, and the EVR in 
the normal presentation mode.
> > IMFMediaTypeHandler_Release(type_handler);
> > @@ -480,6 +483,9 @@ static HRESULT media_engine_create_topology(struct media_engine *engine, IMFMedi
> >               if (audio_src)
> >                   IMFTopologyNode_Release(audio_src);
> >           }
> > +
> > +        if (SUCCEEDED(hr))
> > +            hr = IMFMediaSession_SetTopology(engine->session, MFSESSION_SETTOPOLOGY_IMMEDIATE, topology);
> >       }
>
> That's fine, either here or returning new topology to set it in the
> callback.
> Probably less complex to just set it here since we don't access the 
> object later.
> > if (topology)

[PATCH 3/5] mfmediaengine: Issue MF_MEDIA_ENGINE_EVENT_CANPLAY upon MF_TOPOSTATUS_READY.

[Derek Lesho]

Signed-off-by: Derek Lesho <dlesho(a)codeweavers.com>
---
 dlls/mfmediaengine/main.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/dlls/mfmediaengine/main.c b/dlls/mfmediaengine/main.c
index d1261e3cd73..bc89b2f702a 100644
--- a/dlls/mfmediaengine/main.c
+++ b/dlls/mfmediaengine/main.c
@@ -295,6 +295,14 @@ static HRESULT WINAPI media_engine_session_events_Invoke(IMFAsyncCallback *iface
             IMFMediaEngineNotify_EventNotify(engine->callback, event_type == MEBufferingStarted ?
                     MF_MEDIA_ENGINE_EVENT_BUFFERINGSTARTED : MF_MEDIA_ENGINE_EVENT_BUFFERINGENDED, 0, 0);
             break;
+        case MESessionTopologyStatus:
+        {
+            UINT32 topo_status = 0;
+            IMFMediaEvent_GetUINT32(event, &MF_EVENT_TOPOLOGY_STATUS, &topo_status);
+            if (topo_status == MF_TOPOSTATUS_READY)
+                IMFMediaEngineNotify_EventNotify(engine->callback, MF_MEDIA_ENGINE_EVENT_CANPLAY, 0, 0);
+            break;
+        }
     }
 
 failed:
-- 
2.28.0

[PATCH 4/5] mfmediaengine: Issue MF_MEDIA_ENGINE_EVENT_PLAYING upon MESessionStarted.

[Derek Lesho]

Signed-off-by: Derek Lesho <dlesho(a)codeweavers.com>
---
 dlls/mfmediaengine/main.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/dlls/mfmediaengine/main.c b/dlls/mfmediaengine/main.c
index bc89b2f702a..ed57aec7eb1 100644
--- a/dlls/mfmediaengine/main.c
+++ b/dlls/mfmediaengine/main.c
@@ -303,6 +303,10 @@ static HRESULT WINAPI media_engine_session_events_Invoke(IMFAsyncCallback *iface
                 IMFMediaEngineNotify_EventNotify(engine->callback, MF_MEDIA_ENGINE_EVENT_CANPLAY, 0, 0);
             break;
         }
+        case MESessionStarted:
+
+            IMFMediaEngineNotify_EventNotify(engine->callback, MF_MEDIA_ENGINE_EVENT_PLAYING, 0, 0);
+            break;
     }
 
 failed:
-- 
2.28.0

[PATCH 4/5] mfmediaengine: Issue MF_MEDIA_ENGINE_EVENT_PLAYING upon MESessionStarted.

[Marvin]

Hi,

While running your changed tests, I think I found new failures.
Being a bot and all I'm not very good at pattern recognition, so I might be
wrong, but could you please double-check?

Full results can be found at:
https://testbot.winehq.org/JobDetails.pl?Key=80523

Your paranoid android.


=== debiant (32 bit Chinese:China report) ===

mfplat:
mfplat.c:3461: Test failed: Unexpected refcount 1.

=== debiant (32 bit WoW report) ===

mfplat:
Unhandled exception: page fault on execute access to 0x0a2e7823 in 32-bit code (0x0a2e7823).

[PATCH 5/5] mfmediaengine: Issue MF_MEDIA_ENGINE_EVENT_ENDED upon MESessionEnded.

[Derek Lesho]

Signed-off-by: Derek Lesho <dlesho(a)codeweavers.com>
---
 dlls/mfmediaengine/main.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/dlls/mfmediaengine/main.c b/dlls/mfmediaengine/main.c
index ed57aec7eb1..fe9aa81121f 100644
--- a/dlls/mfmediaengine/main.c
+++ b/dlls/mfmediaengine/main.c
@@ -307,6 +307,10 @@ static HRESULT WINAPI media_engine_session_events_Invoke(IMFAsyncCallback *iface
 
             IMFMediaEngineNotify_EventNotify(engine->callback, MF_MEDIA_ENGINE_EVENT_PLAYING, 0, 0);
             break;
+        case MESessionEnded:
+
+            IMFMediaEngineNotify_EventNotify(engine->callback, MF_MEDIA_ENGINE_EVENT_ENDED, 0, 0);
+            break;
     }
 
 failed:
-- 
2.28.0

[PATCH 5/5] mfmediaengine: Issue MF_MEDIA_ENGINE_EVENT_ENDED upon MESessionEnded.

[Marvin]

Hi,

While running your changed tests, I think I found new failures.
Being a bot and all I'm not very good at pattern recognition, so I might be
wrong, but could you please double-check?

Full results can be found at:
https://testbot.winehq.org/JobDetails.pl?Key=80524

Your paranoid android.


=== debiant (32 bit WoW report) ===

mfplat:
Unhandled exception: page fault on execute access to 0x0a2e7823 in 32-bit code (0x0a2e7823).

[Nikolay Sivov]

On 10/17/20 1:02 AM, Derek Lesho wrote:
> Signed-off-by: Derek Lesho <dlesho(a)codeweavers.com>
> ---
>  dlls/mfplat/main.c         | 4 +++-
>  dlls/mfplat/tests/mfplat.c | 2 --
>  2 files changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/dlls/mfplat/main.c b/dlls/mfplat/main.c
> index a6ff97ae04a..8ed18b1b87f 100644
> --- a/dlls/mfplat/main.c
> +++ b/dlls/mfplat/main.c
> @@ -5750,6 +5750,8 @@ static HRESULT resolver_handler_end_create(struct source_resolver *resolver, enu
>      if (!(queued_result = heap_alloc_zero(sizeof(*queued_result))))
>          return E_OUTOFMEMORY;
>
> +    queued_result->origin = origin;
> +
>      IRtwqAsyncResult_GetObject(inner_result, &handler.handler);
> This one is good, only worked before because 0 is occasionally valid too.
>
> switch (origin)
> @@ -5788,8 +5790,8 @@ static HRESULT resolver_handler_end_create(struct source_resolver *resolver, enu
>
> if (SUCCEEDED(RtwqCreateAsyncResult(queued_result->object, data->pCallback, caller_state, &caller_result)))
>          {
> +            queued_result->inner_result = caller_result;
>              RtwqInvokeCallback(caller_result);
> -            IRtwqAsyncResult_Release(caller_result);
>          }
>      }
> It's been a while, could you explain what this part fixes? Maybe we
> could have another test for it.
>
> diff --git a/dlls/mfplat/tests/mfplat.c b/dlls/mfplat/tests/mfplat.c
> index 55c07adb1fe..a5388840bc5 100644
> --- a/dlls/mfplat/tests/mfplat.c
> +++ b/dlls/mfplat/tests/mfplat.c
> @@ -343,12 +343,10 @@ static HRESULT WINAPI test_create_from_url_callback_Invoke(IMFAsyncCallback *ifa
>
> object = NULL;
>      hr = IMFSourceResolver_EndCreateObjectFromURL(resolver, result, &obj_type, &object);
> -todo_wine
>      ok(hr == S_OK, "Failed to create an object, hr %#x.\n", hr);
>
> hr = IMFAsyncResult_GetObject(result, &object2);
>      ok(hr == S_OK, "Failed to get result object, hr %#x.\n", hr);
> -todo_wine
>      ok(object2 == object, "Unexpected object.\n");
> First hunk alone fixes these two tests for me.
>
> if (object)

[Derek Lesho]

On 10/20/20 6:28 AM, Nikolay Sivov wrote:
> On 10/17/20 1:02 AM, Derek Lesho wrote:
> > Signed-off-by: Derek Lesho <dlesho(a)codeweavers.com>
> > ---
> >   dlls/mfplat/main.c         | 4 +++-
> >   dlls/mfplat/tests/mfplat.c | 2 --
> >   2 files changed, 3 insertions(+), 3 deletions(-)
> >
> > diff --git a/dlls/mfplat/main.c b/dlls/mfplat/main.c
> > index a6ff97ae04a..8ed18b1b87f 100644
> > --- a/dlls/mfplat/main.c
> > +++ b/dlls/mfplat/main.c
> > @@ -5750,6 +5750,8 @@ static HRESULT resolver_handler_end_create(struct source_resolver *resolver, enu
> >       if (!(queued_result = heap_alloc_zero(sizeof(*queued_result))))
> >           return E_OUTOFMEMORY;
> >
> > +    queued_result->origin = origin;
> > +
> >       IRtwqAsyncResult_GetObject(inner_result, &handler.handler);
> > This one is good, only worked before because 0 is occasionally valid too.
> >
> > switch (origin)
> > @@ -5788,8 +5790,8 @@ static HRESULT resolver_handler_end_create(struct source_resolver *resolver, enu
> >
> > if (SUCCEEDED(RtwqCreateAsyncResult(queued_result->object, data->pCallback, caller_state, &caller_result)))
> >           {
> > +            queued_result->inner_result = caller_result;
> >               RtwqInvokeCallback(caller_result);
> > -            IRtwqAsyncResult_Release(caller_result);
> >           }
> >       }
> > It's been a while, could you explain what this part fixes? Maybe we
> > could have another test for it.

Ah, good catch, I took a glance at `if (iter->inner_result == result || 
(iter->object == object && iter->origin == origin))` and assumed all 
three fields had to match.  I'll resend without the second hunk.
> > diff --git a/dlls/mfplat/tests/mfplat.c b/dlls/mfplat/tests/mfplat.c
> > index 55c07adb1fe..a5388840bc5 100644
> > --- a/dlls/mfplat/tests/mfplat.c
> > +++ b/dlls/mfplat/tests/mfplat.c
> > @@ -343,12 +343,10 @@ static HRESULT WINAPI test_create_from_url_callback_Invoke(IMFAsyncCallback *ifa
> >
> > object = NULL;
> >       hr = IMFSourceResolver_EndCreateObjectFromURL(resolver, result, &obj_type, &object);
> > -todo_wine
> >       ok(hr == S_OK, "Failed to create an object, hr %#x.\n", hr);
> >
> > hr = IMFAsyncResult_GetObject(result, &object2);
> >       ok(hr == S_OK, "Failed to get result object, hr %#x.\n", hr);
> > -todo_wine
> >       ok(object2 == object, "Unexpected object.\n");
>
> First hunk alone fixes these two tests for me.
> > if (object)