Manufacturer-specific data bigger than 53 bytes - removed from D-BUS message

[dmitryk-nca]

Bluez version 5.64 Kernel 6.8.0-40-generic #40~22.04.3-Ubuntu x86_64 BT card is AX201 by Intel

Advertisement message with a manufacturer data longer than 53 bytes - is sent to a D-BUS with the data removed. I use Bleak Python library and it cannot pick the data from D-BUS.

Example of a correct D-BUS message org.freedesktop.DBus.Properties.PropertiesChanged (/org/bluez/hci0/dev_D9_39_4A_AC_6E_5A): ['org.bluez.Device1', {'ManufacturerData': <dbus_fast.signature.Variant ('a{qv}', {****: <dbus_fast.signature.Variant ('ay', bytearray(b'\x02\xfc6\xff\xfc\x00e\x00\x01\x00\x00\x00\x01\n\xa0E\x96\xda\x0cpl\x1c'))>})>}, []]

Example of an incorrect D-BUS message (for a longer data) org.freedesktop.DBus.Properties.PropertiesChanged (/org/bluez/hci0/dev_DA_E3_B0_41_31_20): ['org.bluez.Device1', {'RSSI': <dbus_fast.signature.Variant ('n', -75)>}, []]

Log of a correct message size from btmon and Bleak (sorry, I had to remove the manufacturer ID):

HCI Event: LE Meta Event (0x3e) plen 79 #111 [hci0] 72.242701 LE Extended Advertising Report (0x0d) Num reports: 1 Entry 0 Event type: 0x0001 Props: 0x0001 Connectable Data status: Complete Address type: Random (0x01) Address: DA:E3:B0:41:31:20 (Static) Primary PHY: LE Coded Secondary PHY: LE Coded SID: 0x00 TX power: 127 dBm RSSI: -82 dBm (0xae) Periodic advertising interval: 0.00 msec (0x0000) Direct address type: Public (0x00) Direct address: 00:00:00:00:00:00 (OUI 00-00-00) Data length: 0x35 02 01 06 15 09 41 41 41 5f 61 36 64 36 35 39 38 .....AAA_a6d6598 65 38 30 39 30 37 66 39 30 1b ff 10 00 65 e80907f90..9...e 00 43 02 2a 00 00 00 00 00 00 0a e2 3e 00 00 00 .C.*........>... 00 00 00 0c d0 .....
Flags: 0x06 LE General Discoverable Mode BR/EDR Not Supported Name (complete): AAA_a6d6598e80907f90 Company: **** Data: 1000650043022a0000000000000ae23e0000000000000cd0 @ MGMT Event: Unknown (0x002f) plen 69 {0x0001} [hci0] 72.242763 00 00 20 31 41 b0 e3 da 02 ae 00 00 00 00 35 00 .. 1A.........5. 02 01 06 15 09 41 41 41 5f 61 36 64 36 35 39 38 .....AAA_a6d6598 65 38 30 39 30 37 66 39 30 1b ff 10 00 65 e80907f90..9...e 00 43 02 2a 00 00 00 00 00 00 0a e2 3e 00 00 00 .C.*........>... 00 00 00 0c d0

scan_callback DA:E3:B0:41:31:20: AAA_a6d6598e80907f90 AdvertisementData(local_name='AAA_a6d6598e80907f90', manufacturer_data={***: b'\x10\x00e\x00C\x02\x00\x00\x00\x00\x00\x00\n\xe2>\x00\x00\x00\x00\x00\x00\x0c\xd0'}, rssi=-82)

Example of an incorrect kind of a message:

HCI Event: LE Meta Event (0x3e) plen 89 #95 [hci0] 65.315350 LE Extended Advertising Report (0x0d) Num reports: 1 Entry 0 Event type: 0x0001 Props: 0x0001 Connectable Data status: Complete Address type: Random (0x01) Address: DA:E3:B0:41:31:20 (Static) Primary PHY: LE Coded Secondary PHY: LE Coded SID: 0x00 TX power: 127 dBm RSSI: -75 dBm (0xb5) Periodic advertising interval: 0.00 msec (0x0000) Direct address type: Public (0x00) Direct address: 00:00:00:00:00:00 (OUI 00-00-00) Data length: 0x3f 02 01 06 15 09 41 41 41 5f 61 36 64 36 35 39 38 .....AAA_a6d6598 65 38 30 39 30 37 66 39 30 25 ff 10 00 7a e80907f90%.9...z 00 29 02 2c 00 00 00 00 00 00 0a 99 45 00 16 00 .).,........E... 00 00 00 0c ce 00 00 00 00 00 00 00 00 00 b8 ............... Flags: 0x06 LE General Discoverable Mode BR/EDR Not Supported Name (complete): AAA_a6d6598e80907f90 Company: * Data: 10007a0029022c0000000000000a99450016000000000cce000000000000000000b8 @ MGMT Event: Unknown (0x002f) plen 79 {0x0001} [hci0] 65.315366 00 00 20 31 41 b0 e3 da 02 b5 00 00 00 00 3f 00 .. 1A.........?. 02 01 06 15 09 41 41 41 5f 61 36 64 36 35 39 38 .....AAA_a6d6598 65 38 30 39 30 37 66 39 30 25 ff ** 10 00 7a e80907f90%.9...z 00 29 02 2c 00 00 00 00 00 00 0a 99 45 00 16 00 .).,........E... 00 00 00 0c ce 00 00 00 00 00 00 00 00 00 b8 ...............

scan_callback DA:E3:B0:41:31:20: AAA_a6d6598e80907f90 AdvertisementData(local_name='AAA_a6d6598e80907f90', rssi=-75)

[Vudentz]

@dmitryk-nca if you want to advertise big chunks data then you probably gonna need Extended Advertising feature and make sure this is included.

[dmitryk-nca]

In my case, the advertising device is set to use an extended advertising - as it can be seen from the report:

LE Extended Advertising Report (0x0d)

The issue is on the receiving side (BlueZ) - and looks like it is similar to #832. Honestly - I am a bit struggling to understand in which version it is fixed.

[Vudentz]

It has been merged upstream: https://github.com/bluez/bluez/commit/f40c5857d70a91e417e22da9da20f4193414b8a3

[dmitryk-nca]

Yes, I pulled the 5.78 version and this fix is included.

Unfortunately - it does not solve my issue. Manufacturer-specific data longer than 53 (0x35) bytes shows correctly in btmon, but excluded from D-BUS "PropertyChanged" notification.

On Mon, Oct 7, 2024 at 8:52 PM Luiz Augusto von Dentz < @.***> wrote:

It has been merged upstream: f40c585 https://github.com/bluez/bluez/commit/f40c5857d70a91e417e22da9da20f4193414b8a3

— Reply to this email directly, view it on GitHub https://github.com/bluez/bluez/issues/968#issuecomment-2397542247, or unsubscribe https://github.com/notifications/unsubscribe-auth/BBHKJEKYEPYYWYSCCPXBC7DZ2LC4HAVCNFSM6AAAAABPNR4OC6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGOJXGU2DEMRUG4 . You are receiving this because you were mentioned.Message ID: @.***>

[Vudentz]

Check if the kernel has the following:

https://github.com/bluez/bluetooth-next/commit/db08722fc7d46168fe31d9b8a7b29229dd959f9f https://github.com/bluez/bluetooth-next/commit/112b5090c21905531314fee41f691f0317bbf4f6

[dmitryk-nca]

Thanks, I will check this tomorrow. But if the issue is in the kernel - how comes that btmon shows the data correctly?

On Mon, Oct 7, 2024 at 9:20 PM Luiz Augusto von Dentz < @.***> wrote:

Check if the kernel has the following:

@. https://github.com/bluez/bluetooth-next/commit/db08722fc7d46168fe31d9b8a7b29229dd959f9f @. https://github.com/bluez/bluetooth-next/commit/112b5090c21905531314fee41f691f0317bbf4f6

— Reply to this email directly, view it on GitHub https://github.com/bluez/bluez/issues/968#issuecomment-2397592075, or unsubscribe https://github.com/notifications/unsubscribe-auth/BBHKJEMLFUUQLWI5STFWR3DZ2LGFRAVCNFSM6AAAAABPNR4OC6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGOJXGU4TEMBXGU . You are receiving this because you were mentioned.Message ID: @.***>

[Vudentz]

Thanks, I will check this tomorrow. But if the issue is in the kernel - how comes that btmon shows the data correctly?

It shows correctly over HCI not over MGMT.

[dmitryk-nca]

Not sure I understand. In the log of btmon - MGMT event has same data as the original message:

Data length: 0x3f 02 01 06 15 09 41 41 41 5f 61 36 64 36 35 39 38 .....AAA_a6d6598 65 38 30 39 30 37 66 39 30 25 ff 10 00 7a e80907f90%.9...z 00 29 02 2c 00 00 00 00 00 00 0a 99 45 00 16 00 .).,........E... 00 00 00 0c ce 00 00 00 00 00 00 00 00 00 b8 ............... Flags: 0x06 LE General Discoverable Mode BR/EDR Not Supported Name (complete): AAA_a6d6598e80907f90 Company: * Data: 10007a0029022c0000000000000a99450016000000000cce000000000000000000b8 @ MGMT Event: Unknown (0x002f) plen 79 {0x0001} [hci0] 65.315366 00 00 20 31 41 b0 e3 da 02 b5 00 00 00 00 3f 00 .. 1A.........?. 02 01 06 15 09 41 41 41 5f 61 36 64 36 35 39 38 .....AAA_a6d6598 65 38 30 39 30 37 66 39 30 25 ff ** 10 00 7a e80907f90%.9...z 00 29 02 2c 00 00 00 00 00 00 0a 99 45 00 16 00 .).,........E... 00 00 00 0c ce 00 00 00 00 00 00 00 00 00 b8 ...............

On Mon, Oct 7, 2024 at 10:31 PM Luiz Augusto von Dentz < @.***> wrote:

Thanks, I will check this tomorrow. But if the issue is in the kernel - how comes that btmon shows the data correctly?

It shows correctly over HCI not over MGMT.

— Reply to this email directly, view it on GitHub https://github.com/bluez/bluez/issues/968#issuecomment-2397722590, or unsubscribe https://github.com/notifications/unsubscribe-auth/BBHKJEK2ALMKKMDBKCHLOZ3Z2LOQNAVCNFSM6AAAAABPNR4OC6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGOJXG4ZDENJZGA . You are receiving this because you were mentioned.Message ID: @.***>

[Vudentz]

Not sure I understand. In the log of btmon - MGMT event has same data as the original message: Data length: 0x3f 02 01 06 15 09 41 41 41 5f 61 36 64 36 35 39 38 .....AAA_a6d6598 65 38 30 39 30 37 66 39 30 25 ff 10 00 7a e80907f90%.9...z 00 29 02 2c 00 00 00 00 00 00 0a 99 45 00 16 00 .).,........E... 00 00 00 0c ce 00 00 00 00 00 00 00 00 00 b8 ............... Flags: 0x06 LE General Discoverable Mode BR/EDR Not Supported Name (complete): AAA_a6d6598e80907f90 Company: Data: 10007a0029022c0000000000000a99450016000000000cce000000000000000000b8 @ MGMT Event: Unknown (0x002f) plen 79 {0x0001} [hci0] 65.315366 00 00 20 31 41 b0 e3 da 02 b5 00 00 00 00 3f 00 .. 1A.........?. 02 01 06 15 09 41 41 41 5f 61 36 64 36 35 39 38 .....AAA_a6d6598 65 38 30 39 30 37 66 39 30 25 ff 10 00 7a e80907f90%.9...z 00 29 02 2c 00 00 00 00 00 00 0a 99 45 00 16 00 .).,........E... 00 00 00 0c ce 00 00 00 00 00 00 00 00 00 b8 ............... … On Mon, Oct 7, 2024 at 10:31 PM Luiz Augusto von Dentz < @.> wrote: Thanks, I will check this tomorrow. But if the issue is in the kernel - how comes that btmon shows the data correctly? It shows correctly over HCI not over MGMT. — Reply to this email directly, view it on GitHub <#968 (comment)>, or unsubscribe https://github.com/notifications/unsubscribe-auth/BBHKJEK2ALMKKMDBKCHLOZ3Z2LOQNAVCNFSM6AAAAABPNR4OC6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGOJXG4ZDENJZGA . You are receiving this because you were mentioned.Message ID: @.***>

It is an unknown event it is generating, so it really seems like it is not being processed as a Device Found event, that perhaps explain why the daemon doesn't understand it.

[Vudentz]

@dmitryk-nca can you attack the raw/binariy HCI trace using btmon -w ?