[wdmaudiodev] Re: Need help in KS Topology Test - WinBlue, Wave test- Certification

  • From: Matthew van Eerde <Matthew.van.Eerde@xxxxxxxxxxxxx>
  • To: "wdmaudiodev@xxxxxxxxxxxxx" <wdmaudiodev@xxxxxxxxxxxxx>
  • Date: Fri, 28 Aug 2015 21:05:02 +0000

You can see how it looks when it’s working by installing the SysVad driver and
running the test.

https://github.com/Microsoft/Windows-driver-samples/tree/master/audio/sysvad

One thing I left out last time: you should expect GetModes() to be called
twice. The first time, portcls.sys wants to know how many modes you support;
set *NumSignalProcessingModes = 1. The second time, portcls.sys will hand you a
buffer of just the right size (you should validate this); you should copy all
the modes you support (all one of them) into this buffer.

From: wdmaudiodev-bounce@xxxxxxxxxxxxx
[mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx] On Behalf Of Parinitha Kashyap
Sent: Friday, August 28, 2015 1:37 PM
To: wdmaudiodev
Cc: Girish Pattabiraman
Subject: [wdmaudiodev] Re: Need help in KS Topology Test - WinBlue, Wave test-
Certification

Does the pin instantiated with RAW mode need to be of a specific wave format?
Like the sampling rate to be 48kHZ?

Warm regards,
Parinitha Kashyap

On Wed, Aug 26, 2015 at 5:53 PM, Parinitha Kashyap
<parinitha.kashyap@xxxxxxxxx<mailto:parinitha.kashyap@xxxxxxxxx>> wrote:
Hi Matthew,

So, I tried putting logs in all the different places in the MiniPort::NewStream
method and also the Stream::Init call. After every stage, I have a
NT_SUCCESS(ntStatus) call. It returns true every time. I tried removing small
pieces of code from the interface implementation and when I remove the memcpy_s
call, I get the audio back. I thought there must be something going wrong with
this copy call. I printed out the exact hex data that is being copied to
SignalProcessingMode and it is exactly similar to the
AUDIO_SIGNALPROCESSINGMODE_RAW hex value, but still no audio. I might be
missing some part of the code elsewhere. Is there any data format that I need
to handle? Any place where I have to implement the audio signal processing
attribute or pass it as a parameter in any function?? Is the StreamCreated call
in CMiniportWaveRTStream::Init method necessary?

Warm regards,
Parinitha Kashyap

On Tue, Aug 25, 2015 at 12:04 PM, Matthew van Eerde
<Matthew.van.Eerde@xxxxxxxxxxxxx<mailto:Matthew.van.Eerde@xxxxxxxxxxxxx>> wrote:

Has it got anything to do with the failure

Definitely.

'Driver supports processing modes but the preferred formats donot have the
attribute flag set, so setting it

The test is complaining that the driver returned a “preferred format” that does
not have any attributes. This is fine.

From: wdmaudiodev-bounce@xxxxxxxxxxxxx<mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx>
[mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx<mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx>]
On Behalf Of Parinitha Kashyap
Sent: Tuesday, August 25, 2015 11:26 AM

To: wdmaudiodev
Cc: Girish Pattabiraman
Subject: [wdmaudiodev] Re: Need help in KS Topology Test - WinBlue, Wave test-
Certification

Ok. But just a thought. This is happening after I implemented the audio signal
processing interface. Has it got anything to do with the failure? In KS
Topology test, it says that 'Driver supports processing modes but the preferred
formats donot have the attribute flag set, so setting it [hr = E_INVALIDARG
(0x80070057)]'. What format is it talking about here? Since KS pin creation
returns ERROR_NO_MATCH, I was wondering if it was missing any format.

Warm regards,
Parinitha Kashyap

On Tue, Aug 25, 2015 at 8:34 AM, Matthew van Eerde
<Matthew.van.Eerde@xxxxxxxxxxxxx<mailto:Matthew.van.Eerde@xxxxxxxxxxxxx>> wrote:
OK. The error in your screenshot means that KsCreatePin is failing. Sometimes
this fails because your NewStream method returns a failing NTSTATUS; sometimes
it fails without calling into your NewStream method.

From: wdmaudiodev-bounce@xxxxxxxxxxxxx<mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx>
[mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx<mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx>]
On Behalf Of Parinitha Kashyap
Sent: Monday, August 24, 2015 5:04 PM

To: wdmaudiodev
Cc: Girish Pattabiraman
Subject: [wdmaudiodev] Re: Need help in KS Topology Test - WinBlue, Wave test-
Certification

Yes, new stream is being called. I put dbgprint logs at some places. It looks
like the pin initialization is a success too.

Warm regards,
Parinitha Kashyap

On Mon, Aug 24, 2015 at 8:16 AM, Matthew van Eerde
<Matthew.van.Eerde@xxxxxxxxxxxxx<mailto:Matthew.van.Eerde@xxxxxxxxxxxxx>> wrote:
Is your NewStream function actually being called?

From: wdmaudiodev-bounce@xxxxxxxxxxxxx<mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx>
[mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx<mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx>]
On Behalf Of Parinitha Kashyap
Sent: Friday, August 21, 2015 6:50 PM
To: wdmaudiodev
Cc: Girish Pattabiraman

Subject: [wdmaudiodev] Re: Need help in KS Topology Test - WinBlue, Wave test-
Certification

Ok, so I did all the validation and the rest of the steps as you said. The
'Verify audio effects discovery' test passed after making this change but the
audio is not coming out of the driver. KS Topology test and wave test show
failures in pin creation. The exact error is shown in the screenshots attached.
The maximum no. of instances that our streaming pin allows is 1. If the pin is
already allocated, we return STATUS_INVALID_DEVICE_REQUEST. I do not know if
this is creating a problem. KSCreatePin returns 1169. When I looked in MSDN, it
said that if any data format is missing it might throw this error, It also said
that pin creation checks if appropriate attributes are supported. So, I
implemented extracting the data format attributes in the new stream and
checking if buffers are allocated correctly for the attribute
KSATTRIBUTE_AUDIOSIGNALPROCESSING_MODE (GetAttributesFromAttributeList), but
the pin creation still fails.

Warm regards,
Parinitha Kashyap

On Thu, Aug 20, 2015 at 8:24 AM, Matthew van Eerde
<Matthew.van.Eerde@xxxxxxxxxxxxx<mailto:Matthew.van.Eerde@xxxxxxxxxxxxx>> wrote:
The sample implementation of GetModes is here:

https://github.com/Microsoft/Windows-driver-samples/blob/master/audio/sysvad/EndpointsCommon/minwavert.cpp

I see you have a comment that says to do validation, but you’re not actually
doing the validation.

The way GetModes works is:

1) Do the validation as it says in the comment

2) Verify that *NumSignalProcessingModes is big enough (on input, this is
set to the size of the buffer, in GUIDs)

3) Write the appropriate number of mode GUIDs to the memory

4) Set *NumSignalProcessingModes to the number of mode GUIDs that you wrote

From: wdmaudiodev-bounce@xxxxxxxxxxxxx<mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx>
[mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx<mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx>]
On Behalf Of Parinitha Kashyap
Sent: Wednesday, August 19, 2015 3:32 PM
To: Girish Pattabiraman
Cc: wdmaudiodev

Subject: [wdmaudiodev] Re: Need help in KS Topology Test - WinBlue, Wave test-
Certification

I have attached the screenshot as well.

Warm regards,
Parinitha Kashyap

On Wed, Aug 19, 2015 at 3:00 PM, Parinitha Kashyap
<parinitha.kashyap@xxxxxxxxx<mailto:parinitha.kashyap@xxxxxxxxx>> wrote:
yes, I am doing that.

Warm regards,
Parinitha Kashyap

On Wed, Aug 19, 2015 at 2:53 PM, Girish Pattabiraman
<girish.pattabiraman@xxxxxxxxx<mailto:girish.pattabiraman@xxxxxxxxx>> wrote:

Are you handling IID_IMiniportAudioSignalProcessing in
NonDelegatingQueryInterface?



Sent from Mail<http://go.microsoft.com/fwlink/?LinkId=550986> for Windows 10





From: Parinitha Kashyap
Sent: Wednesday, August 19, 2015 2:48 PM
To: wdmaudiodev
Subject: [wdmaudiodev] Re: Need help in KS Topology Test - WinBlue, Wave test-
Certification


Hello,

I implemented the IMiniportAudioSignalProcessing interface on my wavecyclic
driver as follows since 'raw' mode needs to be reported:
STDMETHODIMP_(NTSTATUS) CPhorusMiniport::GetModes
(
_In_ ULONG Pin,
_Out_writes_opt_(*NumSignalProcessingModes) GUID* SignalProcessingModes,
_Inout_ ULONG*
NumSignalProcessingModes
)
/*

1. If Pin is not a valid pin number,
return STATUS_INVALID_PARAMETER.

2. If Pin is a valid pin number and it supports n modes (n>0),
init out-parameters and return STATUS_SUCCESS.

3. Else this pin doesn't support any mode,
return STATUS_NOT_SUPPORTED.
example: bridge pins or another mode-not-aware pins.

*/
{
DPF_ENTER_PAGED();

//DPF_ENTER(("[CMiniportWaveRT::GetModes]"));

NTSTATUS ntStatus = STATUS_SUCCESS;
errno_t err;
GUID mode = AUDIO_SIGNALPROCESSINGMODE_RAW;

// copy supported processing modes
err = memcpy_s(
SignalProcessingModes,
sizeof(GUID),
&mode,
sizeof(GUID)
);

if(err) {
ntStatus = STATUS_UNSUCCESSFUL;
}

return ntStatus;
}

After running the test, I still get the error 'ERROR: Pin[00] :
KSPROPERTY_AUDIOSIGNALPROCESSING_MODES is required but not supported' in the
task 'KS Topology Filters\KS Pins\Check Processing Modes' along with 'FAIL:
Pin[00] : DeviceIoControl failed to get basic support for pin property
{4F67B528-30C9-40DE-B2FB-859DDD1F3470}/0, returned error code 31'. I get the
same deviceIOControl error in the task 'KS Topology Filters\KS Pins\Check Pin
Instance Count' as well. In the method 'PropertyHandlerBasic', we do check if
the property request type is KSPROPERTY_TYPE_BASICSUPPORT and assign the
relevant valuesize. Anything else required to be done in GetModes??

Warm regards,
Parinitha Kashyap

On Wed, Aug 5, 2015 at 4:31 PM, Matthew van Eerde
<Matthew.van.Eerde@xxxxxxxxxxxxx<mailto:Matthew.van.Eerde@xxxxxxxxxxxxx>> wrote:
KSPROPERTY_AUDIOSIGNALPROCESSING_MODES is a new property for all audio drivers
that was introduced in Windows 8.1.

portcls miniports can implement IMiniportAudioSignalProcessing to have
portcls.sys support the property on their behalf.

The SysVad sample driver shows how to implement it:
https://github.com/Microsoft/Windows-driver-samples/tree/master/audio/sysvad

From: wdmaudiodev-bounce@xxxxxxxxxxxxx<mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx>
[mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx<mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx>]
On Behalf Of Parinitha Kashyap
Sent: Wednesday, August 5, 2015 4:22 PM
To: wdmaudiodev@xxxxxxxxxxxxx<mailto:wdmaudiodev@xxxxxxxxxxxxx>
Subject: [wdmaudiodev] Need help in KS Topology Test - WinBlue, Wave test -
Certification

Hi,

My driver is a virtual sound driver based on wavecyclic. I am running tests on
windows 10 RTM using the latest HLK software (v10.0.26624.0) and the latest HLK
filters. I am getting a failure in KS Topology Test - WinBlue in the sub task
'KS Topology Filters\KS Pins\Check Processing Modes'. The error is 'ERROR:
Pin[00] : KSPROPERTY_AUDIOSIGNALPROCESSING_MODES is required but not
supported'. When I ran this test on win 8 on the same driver, it passed. Will
Microsoft release filters to pass this test? Or is this an extra requirement
for the wavecyclic drivers on win 10 to have the new DDI for audio signal
processing modes implemented?

The same error occurs in wave test for the sub tests
'WDMAudio::WaveTest::TAEF_Render_VerifyRawOrDefaultProcessingSupported#4' and
'WDMAudio::WaveTest::TAEF_Render_VerifyRawOrDefaultProcessingSupported#5'. The
exact error text for wave test is as follows: FAIL: The pin 0 doesn't support
KSPROPERTY_AUDIOSIGNALPROCESSING_MODES. As in KS Topology test, the wave test
passed on win 8 with the same driver.

I have attached a screenshot for the wave test.

Pls help.

Thanks,
Warm regards,
Parinitha Kashyap










Other related posts:

  1. Home
  2. wdmaudiodev
  3. Archive
  4. 08-2015