Re: [FFADO-user] Firewire Audio Setting RME FF400 into Ubuntu 20.10

SourceForge Headquarters 1320 Columbia Street Suite 310 San Diego, CA 92101 +1 (858) 422-6466

Hi Riccardo

On Wed, Nov 25, 2020 at 11:06:19AM +0000, Riccardo Wanke wrote:
>  thank you Jonathan.I don't actually have any other thuderbolt device to
> test, I try to connect a video VGA projector adding to the thunderbolt3-2
> another Apple adapter thunderbolt/VGA and then via VGA-VGA cable, but
> lsusc and lspci responses appear identical to the previous connection with
> RME.

It sounds like the system is not seeing the thunderbolt-2 bus provided by
the adapter.

I should mention at this point that the fact that the adapter shows up on
the USB bus may indicate that this is a USB-thunderbolt2 adapter rather than
a thunderbolt3-thunderbolt2 adapter.  This may or may not explain what
you're seeing.  I am still unsure how USB-thunderbolt2 adapters should look
to the system.

> (by the way, the [Thunderbolt 2- Firewire] apapter I am using to
> connect the RME is an Apple one too).

I gather you don't have any thunderbolt2 ports to plug the
thunderbolt2-firewire adapter in to.

> I compared "sudo dmesg" command with or without connection and I think the
> main difference is here:
>
> with
>    ...
>    [  100.275145] usb 1-1: new low-speed USB device number 5 using xhci_hcd
>    [  100.405215] usb 1-1: No LPM exit latency info found, disabling LPM.
>    [  100.411490] usb 1-1: New USB device found, idVendor=05ac, idProduct=1657, bcdDevice= 3.56
>    [  100.411495] usb 1-1: New USB device strings: Mfr=1, Product=2, SerialNumber=3
>    [  100.411499] usb 1-1: Product: Thunderbolt 3 (USB-C) to Thunderbolt 2 Adapter
>    [  100.411502] usb 1-1: Manufacturer: Apple Inc.
>    [  100.411504] usb 1-1: SerialNumber: DTN0301011GH2YFB3
>    ...
>    [ 1871.329983] usb 1-1: USB disconnect, device number 5
>    [ 1933.670540] usb 1-1: new low-speed USB device number 6 using xhci_hcd
>    [ 1933.800321] usb 1-1: No LPM exit latency info found, disabling LPM.
>    [ 1933.806466] usb 1-1: New USB device found, idVendor=05ac, idProduct=1657, bcdDevice= 3.56
>    [ 1933.806472] usb 1-1: New USB device strings: Mfr=1, Product=2, SerialNumber=3
>    [ 1933.806475] usb 1-1: Product: Thunderbolt 3 (USB-C) to Thunderbolt 2 Adapter
>    [ 1933.806479] usb 1-1: Manufacturer: Apple Inc.
>    [ 1933.806481] usb 1-1: SerialNumber: DTN0301011GH2YFB3

The first block of messages describes the discovery of the adapter as device
number 5.  I'm a little surprised that it's classified as a "low-speed USB
device".  I don't know if that's significant.

The second block shows that device 5 (the usb-thunderbolt2 adapter) is
evidently removed at 1871.329983.  At 1933.670540 (about 62 seconds later)
that same adapter is apparently reconnected.  It is detected identically to
the first sighting at 100.275145, which is expected.

> without
>
>    ...
>    [  100.275145] usb 1-1: new low-speed USB device number 5 using xhci_hcd
>    [  100.405215] usb 1-1: No LPM exit latency info found, disabling LPM.
>    [  100.411490] usb 1-1: New USB device found, idVendor=05ac, idProduct=1657, bcdDevice= 3.56
>    [  100.411495] usb 1-1: New USB device strings: Mfr=1, Product=2, SerialNumber=3
>    [  100.411499] usb 1-1: Product: Thunderbolt 3 (USB-C) to Thunderbolt 2 Adapter
>    [  100.411502] usb 1-1: Manufacturer: Apple Inc.
>    [  100.411504] usb 1-1: SerialNumber: DTN0301011GH2YFB3
>    ...

Rather than being without the adapter connected, I think this block is an
inadvertent duplicate of the first block shown in the "with" case.  Notice
how the timestamps of every single line are identical - something which
would not occur across two different occasions.

> I do not actually understand why it appears "the Apple adapter" even when
> it is unplugged (is there a kind of stored data from previous settings?)

As above, I don't think the "without" output corresponds to a time when the
adapter is not plugged in.  The kernel will only report devices as present
if they really are connected to the USB bus.

> Anyway, I dont really have idea if there is a solution for this.thank you
> for your help,Riccardo

Unfortunately I can't provide many other suggestions - I have no experience
with any version of thunderbolt and can only go on what others report. 
Users have successfully used thunderbolt2-firewire adapters before - in
fact, I think theirs was probably the same Apple one that you have got. 
However, to date I do not recall any posts indicating success with a
thunderbolt3-thunderbolt2-firewire adapter chain.  As to why it's not
working, I'll need to defer to others.  My suspicion at present is that the
first adapter in your chain is really a usb-thunderbolt2 adapter, and I have
no idea if these are expected to operate correctly with Linux.  A kernel
driver of some kind would be needed (which may or may not already exist),
and I don't know how well it would work given the involvement of the USB
bus.

There is some interesting information on

  https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1726299

although it's from a few years ago.  Comment 9 says:

    On non-Macs the firmware is responsible for setting up PCI tunnels and
    DisplayPort tunnels over Thunderbolt.  Apparently it doesn't do that for
    devices behind this adapter, so the firmware needs to be extended.

    Please ask Dell for a BIOS update.  Best case they'll manage to get this
    working in collaboration with Intel engineers, worst case they'll respond
    that they don't support Apple's adapter.

In this context, "firmware" is the firmware of your computer.  It makes
sense that the firmware may have to be involved because tunnelling PCI is
something which requires deep coordination with the chipset.  The tunneling
would make the mainboard expose a new PCI bus to the operating system
through the normal PCI layer.  Comment 11 and others which follow provide
additional information about how all this works technically.

I can't recall if you've mentioned what type of computer you're using.  I'm
assuming it's non-Mac hardware.

Following up comment 13 in the above Ubuntu thread, have you tried booting
your computer with the thunderbolt3-thunderbolt2 adapter connected?  At
least on some hardware it seems that this increases the chances that the
firmware will set up the required PCI tunnels.

Comment 6 (which for me today appears under comment 19) suggests that at
least one person has managed to get your collection of adapters working.  It
operates on the assumption that the thunderbolt-firewire device is not
authorised and is therefore blocked from the operating system by the
computer's hardware.  To make it accessible the following command is
suggested:

  echo 1 > $(dirname $(grep -l "Thunderbolt to FireWire Adapter" /sys/bus/thunderbolt/devices/[0-9]-[0-9]/device_name))/authorized

This assumes your firewire adapter reports its name to be "Thunderbolt to
FireWire Adapter".  Comment 6 goes on to explain what to do if this isn't
the case.  The success was reported on kernel 4.15-rc1.

Further discussion (parts of which are copied in the above launchpad bug
report) can be found at

  https://bugzilla.kernel.org/show_bug.cgi?id=189731

Comment 13 in this thread introduces some pretty serious hardware hacking of
the adapter if you're interested in that kind of thing.

Regards
  jonathan