Bluetooth Filter Driver for DS3-compatibility - research notes

Research and Development
Log in to reply
  • nefarius

    While we're already having a working code base for a dedicated Bluetooth host dongle function driver supporting genuine and fake DS3s I wanna also explore the path of a different approach: keeping the default Windows Bluetooth driver stack in place and patching it in a way that it will accept a DS3 connection transparently in parallel to other devices (keyboards, headsets etc.) via filter driver.

    Analysis environment

    • Windows 7 32-Bit
    • Generic Bluetooth Radio (Cambridge Silicon Radio Ltd.)
      • Hardware ID: USB\VID_0A12&PID_0001
    • Genuine Sony DualShock 3 controller

    Tools used

    • DebugView
    • RequestTrace

    Analysis

    On Interrupt Pipe (HCI communication) connection seems to work at least until receiving HCI_Remote_Name_Request_Complete_EV:

    0_1547896876512_71f8b8f8-b2c4-4c35-a290-10d7cb3db73e-image.png

    This corresponds to this

    0_1547896888532_d908db13-e738-4e5c-81c0-18daef84c852-image.png

    in AirBender/WireShock. The HCI_Command_Remote_Name_Request is the last HCI command which has to succeed, which looks like it does. Next is building the L2CAP connection, which seems to get initiated as well a few requests later:

    0_1547896925248_38853936-6843-4d7a-b6ab-8bbec8fd2d78-image.png

    As buffer[6] == 0x01 && buffer[7] == 0x00 represents the L2CAP control channel. So the L2CAP ping-pong should be looked at in greater detail and find the stage where it starts to break.

    2
  • nefarius

    L2CAP_Connection_Request

    00000181	0.82300282	** RT_CompletionRoutineUrbSubmit: Irp: 0x854DC8B8 - IOCTL_INTERNAL_USB_SUBMIT_URB - URB_FUNCTION_BULK_OR_INTERRUPT_TRANSFER - Status STATUS_SUCCESS - Buffer: 0x85703000, MDL: 0x887905F0, Length: 0x10, Pipe:0x85A2137C, Flags: 0x3.	
00000182	0.82301444	** 42 20 0C 00 08 00 01 00  02 01 04 00 11 00 40 7C  B ............@|	
00000183	0.82302463	** 42 20 0C 00 08 00 01 00  02 01 04 00 11 00 40 7C  B ............@|	
00000184	0.82304817	** Irp: 0x854DC8B8 - IOCTL_INTERNAL_USB_SUBMIT_URB - URB_FUNCTION_BULK_OR_INTERRUPT_TRANSFER. Buffer: 0x878B3000, MDL: 0x0, Length: 0x1000, Pipe:0x85A2137C, Flags: 0x3.

    This gets repeated over and over again, so either the correct answer is missing in the trace or not sent at all.

    0
  • nefarius

    This .INF section will add driver as a lower class filter driver for the Bluetooth class (GUID: {e0cbf06c-cd8b-4647-bb8a-263b43f0f974}) and load it on every BTHUSB host device:

    [Filter_Device.NT.HW]
AddReg = Filter_AddReg

[Filter_AddReg]
HKLM,SYSTEM\CurrentControlSet\Control\Class\{e0cbf06c-cd8b-4647-bb8a-263b43f0f974},LowerFilters,0x00010008,Filter

    Section for uninstallation missing.

    EDIT: obsolete, will do this via helper tool instead of INF file.

    0
  • nefarius

    Gotcha! 👮

    2018/12/30-22:27:05.535	TRACE_LEVEL_VERBOSE	>> IOCTL_INTERNAL_USB_SUBMIT_URB
2018/12/30-22:27:05.535	TRACE_LEVEL_VERBOSE	>> >> URB_FUNCTION_BULK_OR_INTERRUPT_TRANSFER (PipeHandle: 854C164C)
2018/12/30-22:27:05.535	TRACE_LEVEL_VERBOSE	Bulk OUT transfer
2018/12/30-22:27:05.535	TRACE_LEVEL_INFORMATION	3C 00 10 00 0C 00 01 00 03 01 08 00 00 00 40 21 02 00 00 00 
2018/12/30-22:27:05.535	TRACE_LEVEL_INFORMATION	L2CAP_Connection_Response (Response: 02)

    Byte 8 is L2CAP_Connection_Response from Windows Bluetooth driver, byte 16 is status:

    /// <summary>
///     Connection refused – PSM not supported.
/// </summary>
L2CAP_ConnectionResponseResult_ConnectionRefusedPsmNotNupported = 0x0002

    This is kinda nice to know because it's the very first L2CAP connection failing because the PSM (Protocol/Service Multiplexer) the DS3 wants to establish a channel for is forbidden in Windows. Let's patch that and see what happens 😏

    0
  • nefarius

    Why this happens is actually documented, especially:

    Some PSMs are reserved for use by Windows:
    SDP: 0x01
    RFCOMM: 0x03
    HID Control: 0x11
    HID Data: 0x13
    BNEP: 0x0F

    It is well-known that the DS3 establishes communication channels via PSM 0x11 and 0x13, which works fine on the PS3 (custom Bluetooth stack firmware) and Linux (no such reservations/limitations). On Windows however this is a pretty big deal and the showstopper for an out-of-the-box working connection of a DS3 with standard Windows Bluetooth stack.

    A pretty neat hack around this limitation was discovered a few years back, but API hooking or binary patches in kernel space are strictly forbidden (mainly for stability and security reasons, even Anti-Virus vendors had to back down from this approach starting with the Vista kernel) so while having this info as a fallback, it's not something we can use in production and therefore a KMDF Bluetooth Profile Driver won't be helpful here.

    Will further go down and explore the lower BTHUSB filter route 😉

    0
  • nefarius

    Hm, even if nothing of this works in the end: I've now accidentally created a Bluetooth sniffer which is nice 😅

    0_1546343450556_4a15c70c-bf43-4bcf-bd12-2593eed2be29-image.png

    0
  • nefarius

    I abandoned the Bluetooth Profile Driver idea to soon; let's see what this skeleton can do for me 🤔

    0_1546352012177_9a0cde3c-1002-4d22-baf5-c52f9f1aba30-image.png

    0
  • nefarius

    Oh my! 😲 This actually worked 👀

    I've currently set up a BTHUSB lower filter driver simply looking for an incoming L2CAP connection request with the dredded PSM 0x11 and patching it to an artificial fixed value (currently 0x5053)

    UrbFunctionBulkInTransferCompleted	2019/01/01-15:26:42.733	TRACE_LEVEL_VERBOSE	UrbFunctionBulkInTransferCompleted Entry
UrbFunctionBulkInTransferCompleted	2019/01/01-15:26:42.733	TRACE_LEVEL_INFORMATION	>> L2CAP_Connection_Request
UrbFunctionBulkInTransferCompleted	2019/01/01-15:26:42.733	TRACE_LEVEL_INFORMATION	++ PSM: 0x11
UrbFunctionBulkInTransferCompleted	2019/01/01-15:26:42.733	TRACE_LEVEL_INFORMATION	++ Patching PSM to 0x5053

    This grants us a way around bthport.sys!BthIsSystemPSM denying the connection right away, so far so good! But that's not enough; how should the Bluetooth stack now know what to do with a non-existent PSM? Well, we need to register a custom Profile Driver for it, of course! 😆

    2019/01/01-15:26:31.001	TRACE_LEVEL_INFORMATION	++ Trying to register PSM 0x5053
2019/01/01-15:26:31.001	TRACE_LEVEL_INFORMATION	++ Got PSM 0x5053
2019/01/01-15:26:31.001	TRACE_LEVEL_VERBOSE	BthPS3RegisterPSM Exit
2019/01/01-15:26:31.001	TRACE_LEVEL_VERBOSE	BthPS3RegisterL2CAPServer Entry
2019/01/01-15:26:31.001	TRACE_LEVEL_VERBOSE	BthPS3IndicationCallback Entry
2019/01/01-15:26:31.001	TRACE_LEVEL_VERBOSE	BthPS3IndicationCallback Exit
2019/01/01-15:26:31.001	TRACE_LEVEL_VERBOSE	BthPS3RegisterL2CAPServer Exit
2019/01/01-15:26:42.763	TRACE_LEVEL_VERBOSE	BthPS3IndicationCallback Entry
2019/01/01-15:26:42.763	TRACE_LEVEL_INFORMATION	BthPS3IndicationCallback ++ IndicationRemoteConnect
2019/01/01-15:26:42.763	TRACE_LEVEL_VERBOSE	BthPS3IndicationCallback Exit

    And we actually receive the connection request in our own profile driver now 😳

    That is... very interesting! Let's see where the road leads me from here on.

    0 B 1 Reply
  • nefarius

    PSMs shall be odd and the least significant bit of the most significant byte shall be zero, hence the following ranges do not contain valid PSMs: 0x0100-0x01FF, 0x0300-0x03FF, 0x0500-0x05FF, 0x0700-0x07FF, 0x0900-0x09FF, 0x0B00-0x0BFF, 0x0D00-0x0DFF, 0x0F00-0x0FFF. All even values are also not valid as PSMs.

    0
  • 414n

    Really interesting, keep up the good work!

    Am I right in assuming these are research notes for an Airbender successor?

    0 nefarius 1 Reply
  • nefarius

    @414n not quite, this is already a successor successor 😆

    The AirBender successor is code-named WireShock and already exists in the lab, has working support for fake DS3s and doesn't require additional software to function (AirBender needs Shibari to actually function).

    However, WireShock is - like AirBender - a USB function driver, which means it needs to replace the entire default Windows Bluetooth driver which means your Bluetooth host gets unusable for anything else that is not a DualShock. That's the same technique ScpToolkit used.

    This research is my attempt to discover "the holy grail"; a solution which keeps the existing Bluetooth drivers in place so headsets, keyboards etc. keep working but allows transparent co-existence with DualShock/Nav/Move 3 devices.

    0
  • nefarius

    Right, now that we've gotten so far to send back a connection response, the second channel pops up and requires a "proxy-PSM" as well:

    UrbFunctionBulkInTransferCompleted	2019/01/05-18:11:16.983	TRACE_LEVEL_INFORMATION	>> L2CAP_Connection_Request
UrbFunctionBulkInTransferCompleted	2019/01/05-18:11:16.983	TRACE_LEVEL_INFORMATION	++ PSM: 0x13

    May I just add that writing a Bluetooth Profile Driver takes forever even if you can steal big chunks of code from the Windows driver samples 😵

    0
  • nefarius

    Oh yeah, registering a proxy PSM for PSM_HID_INTERRUPT as well shoots us even further forward:

    UrbFunctionInterruptInTransferCompleted	2019/01/05-21:13:01.717	TRACE_LEVEL_INFORMATION	HCI_Connection_Complete_EV
UrbFunctionBulkInTransferCompleted	2019/01/05-21:13:01.742	TRACE_LEVEL_VERBOSE	UrbFunctionBulkInTransferCompleted Entry
UrbFunctionBulkInTransferCompleted	2019/01/05-21:13:01.742	TRACE_LEVEL_INFORMATION	>> L2CAP_Connection_Request
UrbFunctionBulkInTransferCompleted	2019/01/05-21:13:01.742	TRACE_LEVEL_INFORMATION	++ PSM: 0x11
UrbFunctionBulkInTransferCompleted	2019/01/05-21:13:01.742	TRACE_LEVEL_INFORMATION	++ Patching PSM to 0x5053
UrbFunctionBulkInTransferCompleted	2019/01/05-21:13:01.763	TRACE_LEVEL_VERBOSE	UrbFunctionBulkInTransferCompleted Entry
UrbFunctionBulkInTransferCompleted	2019/01/05-21:13:01.763	TRACE_LEVEL_INFORMATION	>> L2CAP_Configuration_Request
UrbFunctionBulkInTransferCompleted	2019/01/05-21:13:01.765	TRACE_LEVEL_VERBOSE	UrbFunctionBulkInTransferCompleted Entry
UrbFunctionBulkInTransferCompleted	2019/01/05-21:13:01.765	TRACE_LEVEL_INFORMATION	>> L2CAP_Configuration_Response
UrbFunctionBulkInTransferCompleted	2019/01/05-21:13:01.771	TRACE_LEVEL_VERBOSE	UrbFunctionBulkInTransferCompleted Entry
UrbFunctionBulkInTransferCompleted	2019/01/05-21:13:01.771	TRACE_LEVEL_INFORMATION	>> L2CAP_Connection_Request
UrbFunctionBulkInTransferCompleted	2019/01/05-21:13:01.771	TRACE_LEVEL_INFORMATION	++ PSM: 0x13
UrbFunctionBulkInTransferCompleted	2019/01/05-21:13:01.771	TRACE_LEVEL_INFORMATION	++ Patching PSM to 0x5055
UrbFunctionBulkInTransferCompleted	2019/01/05-21:13:01.781	TRACE_LEVEL_VERBOSE	UrbFunctionBulkInTransferCompleted Entry
UrbFunctionBulkInTransferCompleted	2019/01/05-21:13:01.781	TRACE_LEVEL_INFORMATION	>> L2CAP_Configuration_Request
UrbFunctionBulkInTransferCompleted	2019/01/05-21:13:01.783	TRACE_LEVEL_VERBOSE	UrbFunctionBulkInTransferCompleted Entry
UrbFunctionBulkInTransferCompleted	2019/01/05-21:13:01.783	TRACE_LEVEL_INFORMATION	>> L2CAP_Configuration_Response

    Which generates some more noise in the profile driver:

    2019/01/05-21:13:01.379	TRACE_LEVEL_VERBOSE	BthPS3IndicationCallback Entry
2019/01/05-21:13:01.379	TRACE_LEVEL_INFORMATION	BthPS3IndicationCallback ++ IndicationRemoteConnect
2019/01/05-21:13:01.379	TRACE_LEVEL_VERBOSE	BthPS3SendConnectResponse Entry
2019/01/05-21:13:01.379	TRACE_LEVEL_VERBOSE	BthPS3ConnectionIndicationCallback Entry
2019/01/05-21:13:01.379	TRACE_LEVEL_VERBOSE	BthPS3ConnectionIndicationCallback Exit
2019/01/05-21:13:01.379	TRACE_LEVEL_VERBOSE	BthPS3SendConnectResponse Exit (STATUS_SUCCESS (0x00000000))
2019/01/05-21:13:01.379	TRACE_LEVEL_VERBOSE	BthPS3IndicationCallback Exit
2019/01/05-21:13:01.389	TRACE_LEVEL_VERBOSE	BthPS3RemoteConnectResponseCompletion Entry
2019/01/05-21:13:01.389	TRACE_LEVEL_INFORMATION	Connection completion, status: STATUS_SUCCESS (0x00000000)
2019/01/05-21:13:01.389	TRACE_LEVEL_INFORMATION	Connection established with client
2019/01/05-21:13:01.389	TRACE_LEVEL_INFORMATION	Connection completed, connection: 0xFFFFFA80089CC0C0
2019/01/05-21:13:01.389	TRACE_LEVEL_VERBOSE	BthPS3RemoteConnectResponseCompletion Exit
2019/01/05-21:13:01.395	TRACE_LEVEL_VERBOSE	BthPS3IndicationCallback Entry
2019/01/05-21:13:01.395	TRACE_LEVEL_INFORMATION	BthPS3IndicationCallback ++ IndicationRemoteConnect
2019/01/05-21:13:01.395	TRACE_LEVEL_VERBOSE	BthPS3SendConnectResponse Entry
2019/01/05-21:13:01.395	TRACE_LEVEL_VERBOSE	BthPS3SendConnectResponse Exit (STATUS_SUCCESS (0x00000000))
2019/01/05-21:13:01.395	TRACE_LEVEL_VERBOSE	BthPS3IndicationCallback Exit
2019/01/05-21:13:01.395	TRACE_LEVEL_VERBOSE	BthPS3ConnectionIndicationCallback Entry
2019/01/05-21:13:01.395	TRACE_LEVEL_VERBOSE	BthPS3ConnectionIndicationCallback Exit
2019/01/05-21:13:01.413	TRACE_LEVEL_VERBOSE	BthPS3RemoteConnectResponseCompletion Entry
2019/01/05-21:13:01.413	TRACE_LEVEL_INFORMATION	Connection completion, status: STATUS_DEVICE_NOT_CONNECTED (0xC000009D)
2019/01/05-21:13:01.413	TRACE_LEVEL_VERBOSE	BthPS3RemoteConnectResponseCompletion Exit

    Now ofc. the state machine I ripped off the bthecho sample driver goes crazy here and the connection drops but both channels connection requests actually bubble up to our driver, that's good news 🙂

    0
  • nefarius

    Btw. if you attempt to register one of the "forbidden" PSMs within your profile driver like so:

    DevCtx->Header.ProfileDrvInterface.BthReuseBrb(
        &(DevCtx->RegisterUnregisterBrb),
        BRB_REGISTER_PSM
    );

brb = (struct _BRB_PSM *)
	&(DevCtx->RegisterUnregisterBrb);

brb->Psm = PSM_HID_CONTROL; // LOL, nope

status = BthPS3SendBrbSynchronously(
	DevCtx->Header.IoTarget,
	DevCtx->Header.Request,
	(PBRB)brb,
	sizeof(*brb)
);

    the system crashes 😆 no NT_STATUS failure code, you'll be sent straight to hell. So, uh, don't do that 🙃

    0
  • nefarius

    This is how the profile driver could look like, notice the original host driver still being in place:

    0_1546968648339_773f36f0-6910-4c26-9450-888fe4875590-image.png

    The Hardware ID would be BTHENUM\{CUSTOM_SERVICE_GUID} where in my lab example I simply recycled the one bthecho uses:

    0_1546968684938_a1661077-ee57-467e-af06-1f96fd6e12e1-image.png

    0_1546968857025_ba14d738-13b0-4557-b1ff-52e792f9f6a3-image.png

    0_1546968889258_43c9e941-7715-4309-9141-334c11c140f1-image.png

    The service GUID I registered with a WinAPI call to BluetoothSetLocalServiceInfo:

    DWORD SetBthServiceInfo(
    BOOLEAN bEnabled
    )
{
    DWORD err = ERROR_SUCCESS;
    BLUETOOTH_LOCAL_SERVICE_INFO SvcInfo = {0};
    SvcInfo.Enabled = bEnabled;

    if (FAILED(StringCbCopyW(SvcInfo.szName, sizeof(SvcInfo.szName), BthEchoSampleSvcName)))
    {
        printf("Copying svc name failed\n");
        goto exit;
    }

    if (ERROR_SUCCESS != (err = BluetoothSetLocalServiceInfo(
        NULL, //callee would select the first found radio
        &BTHECHOSAMPLE_SVC_GUID,
        0,
        &SvcInfo
        )))
    {
        printf("BluetoothSetLocalServiceInfo failed, err = %d\n", err);
        goto exit;        
    }
exit:
    return err;    
}

    Again, simply taken from the Microsoft sample. I also noticed, that the bthsrvinst project links against BluetoothApis.lib, which bombs on Windows 7. Link against Bthprops.lib instead and it will work.

    Supported in Windows 8 and later versions of Windows.

    0
  • nefarius

    😲

    2019/01/08-18:50:55.893	TRACE_LEVEL_VERBOSE	BthPS3IndicationCallback Entry
2019/01/08-18:50:55.893	TRACE_LEVEL_INFORMATION	New connection for 5053 from AC7A4D2819AC arrived
2019/01/08-18:50:55.893	TRACE_LEVEL_VERBOSE	L2CAP_PS3_SendConnectResponse Entry
2019/01/08-18:50:55.893	TRACE_LEVEL_VERBOSE	L2CAP_PS3_ConnectResponsePendingCompleted Entry
2019/01/08-18:50:55.893	TRACE_LEVEL_INFORMATION	Connection PENDING completed with status: STATUS_SUCCESS (0x00000000)
2019/01/08-18:50:55.893	TRACE_LEVEL_VERBOSE	L2CAP_PS3_ConnectResponsePendingCompleted Exit
2019/01/08-18:50:55.893	TRACE_LEVEL_VERBOSE	BthPS3ConnectionIndicationCallback Entry
2019/01/08-18:50:55.893	TRACE_LEVEL_VERBOSE	BthPS3ConnectionIndicationCallback Exit
2019/01/08-18:50:55.893	TRACE_LEVEL_VERBOSE	L2CAP_PS3_SendConnectResponse Exit (STATUS_SUCCESS (0x00000000))
2019/01/08-18:50:55.893	TRACE_LEVEL_VERBOSE	BthPS3IndicationCallback Exit
2019/01/08-18:50:55.904	TRACE_LEVEL_VERBOSE	BthPS3RemoteConnectResponseCompletion Entry
2019/01/08-18:50:55.904	TRACE_LEVEL_INFORMATION	Connection completion, status: STATUS_SUCCESS (0x00000000)
2019/01/08-18:50:55.904	TRACE_LEVEL_INFORMATION	Connection established with client
2019/01/08-18:50:55.904	TRACE_LEVEL_INFORMATION	Connection completed, connection: 0xFFFFFA8009F6C480
2019/01/08-18:50:55.904	TRACE_LEVEL_VERBOSE	BthPS3RemoteConnectResponseCompletion Exit
2019/01/08-18:50:55.911	TRACE_LEVEL_VERBOSE	BthPS3IndicationCallback Entry
2019/01/08-18:50:55.911	TRACE_LEVEL_INFORMATION	New connection for 5055 from AC7A4D2819AC arrived
2019/01/08-18:50:55.911	TRACE_LEVEL_VERBOSE	L2CAP_PS3_SendConnectResponse Entry
2019/01/08-18:50:55.911	TRACE_LEVEL_VERBOSE	L2CAP_PS3_SendConnectResponse Exit (STATUS_SUCCESS (0x00000000))
2019/01/08-18:50:55.911	TRACE_LEVEL_VERBOSE	BthPS3IndicationCallback Exit
2019/01/08-18:50:55.911	TRACE_LEVEL_VERBOSE	L2CAP_PS3_ConnectResponsePendingCompleted Entry
2019/01/08-18:50:55.911	TRACE_LEVEL_INFORMATION	Connection PENDING completed with status: STATUS_SUCCESS (0x00000000)
2019/01/08-18:50:55.911	TRACE_LEVEL_VERBOSE	L2CAP_PS3_ConnectResponsePendingCompleted Exit
2019/01/08-18:50:55.911	TRACE_LEVEL_VERBOSE	BthPS3ConnectionIndicationCallback Entry
2019/01/08-18:50:55.911	TRACE_LEVEL_VERBOSE	BthPS3ConnectionIndicationCallback Exit
2019/01/08-18:50:55.927	TRACE_LEVEL_VERBOSE	BthPS3RemoteConnectResponseCompletion Entry

    Right, so the connection handshake works for both Control and Interrupt channel, now I need to copypasta a ton of code to send and respond to the configuration requests. And extend the state machine to handle two channels instead of one like the example code intended. Oh dear 👀

    0
  • nefarius

    Nope, wrong assumption. The profile driver doesn't craft configuration requests, the parent driver does.

    Sniff-log from under BTHUSB:

    2019/01/08-19:06:14.202	TRACE_LEVEL_INFORMATION	>> L2CAP_Connection_Request
2019/01/08-19:06:14.233	TRACE_LEVEL_INFORMATION	<< L2CAP_Connection_Response (Response: 01)
2019/01/08-19:06:14.233	TRACE_LEVEL_INFORMATION	<< L2CAP_Connection_Response (Response: 01)
2019/01/08-19:06:14.233	TRACE_LEVEL_INFORMATION	<< L2CAP_Connection_Response (Response: 00)
2019/01/08-19:06:14.233	TRACE_LEVEL_INFORMATION	<< L2CAP_Configuration_Request
2019/01/08-19:06:14.245	TRACE_LEVEL_INFORMATION	>> L2CAP_Configuration_Request
2019/01/08-19:06:14.245	TRACE_LEVEL_INFORMATION	<< L2CAP_Configuration_Response
2019/01/08-19:06:14.247	TRACE_LEVEL_INFORMATION	>> L2CAP_Configuration_Response
2019/01/08-19:06:14.253	TRACE_LEVEL_INFORMATION	>> L2CAP_Connection_Request
2019/01/08-19:06:14.253	TRACE_LEVEL_INFORMATION	<< L2CAP_Connection_Response (Response: 01)
2019/01/08-19:06:14.253	TRACE_LEVEL_INFORMATION	<< L2CAP_Connection_Response (Response: 01)
2019/01/08-19:06:14.253	TRACE_LEVEL_INFORMATION	<< L2CAP_Connection_Response (Response: 00)
2019/01/08-19:06:14.253	TRACE_LEVEL_INFORMATION	<< L2CAP_Configuration_Request
2019/01/08-19:06:14.266	TRACE_LEVEL_INFORMATION	>> L2CAP_Configuration_Request
2019/01/08-19:06:14.266	TRACE_LEVEL_INFORMATION	<< L2CAP_Configuration_Response
2019/01/08-19:06:14.267	TRACE_LEVEL_INFORMATION	>> L2CAP_Configuration_Response
2019/01/08-19:06:14.275	TRACE_LEVEL_INFORMATION	>> L2CAP_Disconnection_Request
2019/01/08-19:06:14.281	TRACE_LEVEL_INFORMATION	>> L2CAP_Disconnection_Request

    That's both cool and frightening; hopefully the host knows what it's doing 😜

    Also, why is PENDING sent twice per channel 🤔 did I do that or the host...

    EDIT: I did that, no need for it, the parent driver has everything under control 😏

    0
  • nefarius

    Sniffing intensifies 😏

    2019/01/08-20:04:06.318	TRACE_LEVEL_INFORMATION	>> L2CAP_Connection_Request [Code: 0x02, Identifier: 0x01, Length: 4, PSM: 0x5053, SCID: 0x1440]
2019/01/08-20:04:06.349	TRACE_LEVEL_INFORMATION	<< L2CAP_Connection_Response [Code: 0x03, Identifier: 0x01, Length: 8, DCID: 0x0000, SCID: 0x1440, Result: 0x0001, Status: 0x0000]
2019/01/08-20:04:06.349	TRACE_LEVEL_INFORMATION	<< L2CAP_Connection_Response [Code: 0x03, Identifier: 0x01, Length: 8, DCID: 0x0000, SCID: 0x1440, Result: 0x0001, Status: 0x0000]
2019/01/08-20:04:06.349	TRACE_LEVEL_INFORMATION	<< L2CAP_Connection_Response [Code: 0x03, Identifier: 0x01, Length: 8, DCID: 0x0040, SCID: 0x1440, Result: 0x0000, Status: 0x0000]
2019/01/08-20:04:06.349	TRACE_LEVEL_INFORMATION	<< L2CAP_Configuration_Request [Code: 0x04, Identifier: 0x01, Length: 8, DCID: 0x1440, Flags: 0x0000, Options: 0x02A00201]
2019/01/08-20:04:06.360	TRACE_LEVEL_INFORMATION	>> L2CAP_Configuration_Request [Code: 0x04, Identifier: 0x02, Length: 4, DCID: 0x0040, Flags: 0x0000, Options: 0x00001000]
2019/01/08-20:04:06.361	TRACE_LEVEL_INFORMATION	<< L2CAP_Configuration_Response [Code: 0x05, Identifier: 0x02, Length: 6, SCID: 0x1440, Flags: 0x0000, Result: 0x0000, Options: 0x0000]
2019/01/08-20:04:06.362	TRACE_LEVEL_INFORMATION	>> L2CAP_Configuration_Response [Code: 0x05, Identifier: 0x01, Length: 6, SCID: 0x0040, Flags: 0x0000, Result: 0x0000, Options: 0x0052]
2019/01/08-20:04:06.369	TRACE_LEVEL_INFORMATION	>> L2CAP_Connection_Request [Code: 0x02, Identifier: 0x03, Length: 4, PSM: 0x5055, SCID: 0x1481]
2019/01/08-20:04:06.369	TRACE_LEVEL_INFORMATION	<< L2CAP_Connection_Response [Code: 0x03, Identifier: 0x03, Length: 8, DCID: 0x0000, SCID: 0x1481, Result: 0x0001, Status: 0x0000]
2019/01/08-20:04:06.369	TRACE_LEVEL_INFORMATION	<< L2CAP_Connection_Response [Code: 0x03, Identifier: 0x03, Length: 8, DCID: 0x0000, SCID: 0x1481, Result: 0x0001, Status: 0x0000]
2019/01/08-20:04:06.369	TRACE_LEVEL_INFORMATION	<< L2CAP_Connection_Response [Code: 0x03, Identifier: 0x03, Length: 8, DCID: 0x0041, SCID: 0x1481, Result: 0x0000, Status: 0x0000]
2019/01/08-20:04:06.369	TRACE_LEVEL_INFORMATION	<< L2CAP_Configuration_Request [Code: 0x04, Identifier: 0x02, Length: 8, DCID: 0x1481, Flags: 0x0000, Options: 0x02A00201]
2019/01/08-20:04:06.380	TRACE_LEVEL_INFORMATION	>> L2CAP_Configuration_Request [Code: 0x04, Identifier: 0x04, Length: 28, DCID: 0x0041, Flags: 0x0000, Options: 0x02001603]
2019/01/08-20:04:06.381	TRACE_LEVEL_INFORMATION	<< L2CAP_Configuration_Response [Code: 0x05, Identifier: 0x04, Length: 30, SCID: 0x1481, Flags: 0x0000, Result: 0x0001, Options: 0x1603]
2019/01/08-20:04:06.383	TRACE_LEVEL_INFORMATION	>> L2CAP_Configuration_Response [Code: 0x05, Identifier: 0x02, Length: 6, SCID: 0x0041, Flags: 0x0000, Result: 0x0000, Options: 0x0052]
2019/01/08-20:04:06.389	TRACE_LEVEL_INFORMATION	>> L2CAP_Disconnection_Request [Code: 0x06, Identifier: 0x05, Length: 4, DCID: 0x0041, SCID: 0x1481]
2019/01/08-20:04:06.389	TRACE_LEVEL_INFORMATION	<< L2CAP_Disconnection_Response [Code: 0x07, Identifier: 0x05, Length: 4, DCID: 0x0041, SCID: 0x1481]
2019/01/08-20:04:06.399	TRACE_LEVEL_INFORMATION	>> L2CAP_Disconnection_Request [Code: 0x06, Identifier: 0x06, Length: 4, DCID: 0x0040, SCID: 0x1440]
2019/01/08-20:04:06.399	TRACE_LEVEL_INFORMATION	<< L2CAP_Disconnection_Response [Code: 0x07, Identifier: 0x06, Length: 4, DCID: 0x0040, SCID: 0x1440]
    0
  • nefarius

    L2CAP Channel Configuration

    MTU

    A successful agreement on MTU value L2CAP_MAX_MTU:

    << L2CAP_Configuration_Request [Code: 0x04, Identifier: 0x15, Length: 8, DCID: 0x3240, Flags: 0x0000, Options: 0xFFFF0201]
>> L2CAP_Configuration_Request [Code: 0x04, Identifier: 0x02, Length: 4, DCID: 0x0040, Flags: 0x0000, Options: 0x02010000]
<< L2CAP_Configuration_Response [Code: 0x05, Identifier: 0x02, Length: 6, SCID: 0x3240, Flags: 0x0000, Result: 0x0000, Options: 0x0000]
>> L2CAP_Configuration_Response [Code: 0x05, Identifier: 0x15, Length: 10, SCID: 0x0040, Flags: 0x0000, Result: 0x0000, Options: 0x0201]
<< L2CAP_Configuration_Request [Code: 0x04, Identifier: 0x16, Length: 8, DCID: 0x3281, Flags: 0x0000, Options: 0xFFFF0201]
>> L2CAP_Configuration_Response [Code: 0x05, Identifier: 0x16, Length: 10, SCID: 0x0041, Flags: 0x0000, Result: 0x0000, Options: 0x0201]

    QOS

    Controller and host currently can't agree on QUALITY OF SERVICE (QOS) OPTION.

    Option 0x02001603 is translated to: 0x03 (QoS option type) with length of 22 (0x16) and requested option is 0x0200 (0x02 = Service Type as Guaranteed, L2CAP_FLOW_SERVICE_TYPE_GUARANTEED)

    >> L2CAP_Configuration_Request [Code: 0x04, Identifier: 0x04, Length: 28, DCID: 0x0041, Flags: 0x0000, Options: 0x02001603]
<< L2CAP_Configuration_Response [Code: 0x05, Identifier: 0x04, Length: 30, SCID: 0x3281, Flags: 0x0000, Result: 0x0001, Options: 0x1603]
    0
  • nefarius

    Ladies and gentlemen, we got him

    The DS3 is connected to standard Windows Bluetooth Stack 🎉

    I must be dreaming 😂 it actually works, my DS3 is connected to an USB Bluetooth dongle running the standard Windows Bluetooth Stack, the filter and the profile driver! git commit && git push 😆

    Lower filter log

    2019/01/08-21:02:33.479	>> L2CAP_Connection_Request [Code: 0x02, Identifier: 0x01, Length: 4, PSM: 0x5053, SCID: 0x42C0]
2019/01/08-21:02:33.491	<< L2CAP_Connection_Response [Code: 0x03, Identifier: 0x01, Length: 8, DCID: 0x0000, SCID: 0x42C0, Result: 0x0001, Status: 0x0000]
2019/01/08-21:02:33.491	<< L2CAP_Connection_Response [Code: 0x03, Identifier: 0x01, Length: 8, DCID: 0x0040, SCID: 0x42C0, Result: 0x0000, Status: 0x0000]
2019/01/08-21:02:33.491	<< L2CAP_Configuration_Request [Code: 0x04, Identifier: 0x03, Length: 8, DCID: 0x42C0, Flags: 0x0000, Options: 0xFFFF0201]
2019/01/08-21:02:33.497	>> L2CAP_Configuration_Request [Code: 0x04, Identifier: 0x02, Length: 4, DCID: 0x0040, Flags: 0x0000, Options: 0x00010004]
2019/01/08-21:02:33.497	<< L2CAP_Configuration_Response [Code: 0x05, Identifier: 0x02, Length: 6, SCID: 0x42C0, Flags: 0x0000, Result: 0x0000, Options: 0x0000]
2019/01/08-21:02:33.499	>> L2CAP_Configuration_Response [Code: 0x05, Identifier: 0x03, Length: 10, SCID: 0x0040, Flags: 0x0000, Result: 0x0000, Options: 0x0201]
2019/01/08-21:02:33.507	>> L2CAP_Connection_Request [Code: 0x02, Identifier: 0x03, Length: 4, PSM: 0x5055, SCID: 0x4301]
2019/01/08-21:02:33.507	<< L2CAP_Connection_Response [Code: 0x03, Identifier: 0x03, Length: 8, DCID: 0x0000, SCID: 0x4301, Result: 0x0001, Status: 0x0000]
2019/01/08-21:02:33.507	<< L2CAP_Connection_Response [Code: 0x03, Identifier: 0x03, Length: 8, DCID: 0x0041, SCID: 0x4301, Result: 0x0000, Status: 0x0000]
2019/01/08-21:02:33.507	<< L2CAP_Configuration_Request [Code: 0x04, Identifier: 0x04, Length: 8, DCID: 0x4301, Flags: 0x0000, Options: 0xFFFF0201]
2019/01/08-21:02:33.516	>> L2CAP_Configuration_Request [Code: 0x04, Identifier: 0x04, Length: 28, DCID: 0x0041, Flags: 0x0000, Options: 0x02001603]
2019/01/08-21:02:33.516	<< L2CAP_Configuration_Response [Code: 0x05, Identifier: 0x04, Length: 6, SCID: 0x4301, Flags: 0x0000, Result: 0x0000, Options: 0x0000]
2019/01/08-21:02:33.519	>> L2CAP_Configuration_Response [Code: 0x05, Identifier: 0x04, Length: 10, SCID: 0x0041, Flags: 0x0000, Result: 0x0000, Options: 0x0201]

    Profile driver log

    2019/01/08-21:02:33.201	TRACE_LEVEL_VERBOSE	BthPS3IndicationCallback Entry
2019/01/08-21:02:33.201	TRACE_LEVEL_INFORMATION	New connection for PSM 0x5053 from AC7A4D2819AC arrived
2019/01/08-21:02:33.201	TRACE_LEVEL_VERBOSE	L2CAP_PS3_SendConnectResponse Entry
2019/01/08-21:02:33.201	TRACE_LEVEL_VERBOSE	BthPS3ConnectionIndicationCallback Entry
2019/01/08-21:02:33.201	TRACE_LEVEL_VERBOSE	BthPS3ConnectionIndicationCallback Exit
2019/01/08-21:02:33.201	TRACE_LEVEL_VERBOSE	L2CAP_PS3_SendConnectResponse Exit (STATUS_SUCCESS (0x00000000))
2019/01/08-21:02:33.201	TRACE_LEVEL_VERBOSE	BthPS3IndicationCallback Exit
2019/01/08-21:02:33.209	TRACE_LEVEL_VERBOSE	L2CAP_PS3_ConnectResponseCompleted Entry
2019/01/08-21:02:33.209	TRACE_LEVEL_INFORMATION	Connection completion, status: STATUS_SUCCESS (0x00000000)
2019/01/08-21:02:33.209	TRACE_LEVEL_INFORMATION	Connection established with client
2019/01/08-21:02:33.209	TRACE_LEVEL_INFORMATION	Connection completed, connection: 0xFFFFFA8009A7A5E0
2019/01/08-21:02:33.209	TRACE_LEVEL_VERBOSE	L2CAP_PS3_ConnectResponseCompleted Exit
2019/01/08-21:02:33.216	TRACE_LEVEL_VERBOSE	BthPS3IndicationCallback Entry
2019/01/08-21:02:33.216	TRACE_LEVEL_INFORMATION	New connection for PSM 0x5055 from AC7A4D2819AC arrived
2019/01/08-21:02:33.216	TRACE_LEVEL_VERBOSE	L2CAP_PS3_SendConnectResponse Entry
2019/01/08-21:02:33.216	TRACE_LEVEL_VERBOSE	L2CAP_PS3_SendConnectResponse Exit (STATUS_SUCCESS (0x00000000))
2019/01/08-21:02:33.216	TRACE_LEVEL_VERBOSE	BthPS3IndicationCallback Exit
2019/01/08-21:02:33.216	TRACE_LEVEL_VERBOSE	BthPS3ConnectionIndicationCallback Entry
2019/01/08-21:02:33.216	TRACE_LEVEL_VERBOSE	BthPS3ConnectionIndicationCallback Exit
2019/01/08-21:02:33.225	TRACE_LEVEL_VERBOSE	BthPS3ConnectionIndicationCallback Entry
2019/01/08-21:02:33.225	TRACE_LEVEL_INFORMATION	BthPS3ConnectionIndicationCallback ++ IndicationRemoteConfigRequest
2019/01/08-21:02:33.225	TRACE_LEVEL_VERBOSE	BthPS3ConnectionIndicationCallback Exit
2019/01/08-21:02:33.228	TRACE_LEVEL_VERBOSE	L2CAP_PS3_ConnectResponseCompleted Entry
2019/01/08-21:02:33.228	TRACE_LEVEL_INFORMATION	Connection completion, status: STATUS_SUCCESS (0x00000000)
2019/01/08-21:02:33.228	TRACE_LEVEL_INFORMATION	Connection established with client
2019/01/08-21:02:33.228	TRACE_LEVEL_INFORMATION	Connection completed, connection: 0xFFFFFA8009D890C0
2019/01/08-21:02:33.228	TRACE_LEVEL_VERBOSE	L2CAP_PS3_ConnectResponseCompleted Exit
    0