MP 1,1-5,1 testing TB3 AIC with MP 5,1

[cdf]

It's a four port hub but you're only using 3 ports. Exposing the fourth port is probably more trouble than it's worth - you don't want to expose the problems with the power source (only 3.3V instead of 5V) - and who needs more USB 2.0 ports?

Exactly. Were I designing the hub from scratch, I would use the USB2513B IC (3 ports).

 

[cdf]

I tested the Apple USB-C AV multiport adapters (both the HDMI and VGA versions) with the Titan Ridge card. The video output works as intended. (Interestingly, on a 4K monitor the HDMI version enables the text scaling menu in display preferences, probably because the adapter was never intended to be used on a pre-retina Mac.) The adapters also have a USB-A 2 port, but for some reason that port doesn't work. It has 5V but no data. I wonder if that has to do with how the card handles USB 2...

 

[joevt]

I tested the Apple USB-C AV multiport adapters (both the HDMI and VGA versions) with the Titan Ridge card. The video output works as intended. (Interestingly, on a 4K monitor the HDMI version enables the text scaling menu in display preferences, probably because the adapter was never intended to be used on a pre-retina Mac.) The adapters also have a USB-A 2 port, but for some reason that port doesn't work. It has 5V but no data. I wonder if that has to do with how the card handles USB 2...

The scaling menu probably appears for any display that is 4K or greater. Strange that the USB doesn't work. You say it's working with a USB-C to USB-A adapter with something like a mouse? Did you try both orientations of the USB-C connector for both tests?

 

[aclarubicin]

I tested the Apple USB-C AV multiport adapters (both the HDMI and VGA versions) with the Titan Ridge card. The video output works as intended. (Interestingly, on a 4K monitor the HDMI version enables the text scaling menu in display preferences, probably because the adapter was never intended to be used on a pre-retina Mac.) The adapters also have a USB-A 2 port, but for some reason that port doesn't work. It has 5V but no data. I wonder if that has to do with how the card handles USB 2...

The Titan Ridge needs connection to the USB header for USB2 support. USB 3(.1) support does work properly

 

[cdf]

The scaling menu probably appears for any display that is 4K or greater. Strange that the USB doesn't work. You say it's working with a USB-C to USB-A adapter with something like a mouse? Did you try both orientations of the USB-C connector for both tests?

To test USB 2, I used my display (I don't have a USB-C to USB-A adapter other than the AV ones). The display connects via USB-C and has a built-in USB 2 hub. Devices connected to it work fine. I also tried both orientations of the adapters. Nothing.

The Titan Ridge needs connection to the USB header for USB2 support. USB 3(.1) support does work properly

Take a look at posts #347 and #349

 

[Woof Woof]

I tested the Apple USB-C AV multiport adapters (both the HDMI and VGA versions) with the Titan Ridge card. The video output works as intended. (Interestingly, on a 4K monitor the HDMI version enables the text scaling menu in display preferences, probably because the adapter was never intended to be used on a pre-retina Mac.) The adapters also have a USB-A 2 port, but for some reason that port doesn't work. It has 5V but no data. I wonder if that has to do with how the card handles USB 2...

Doesn't surprise me... I was using a generic USB-C to HDMI/Ethernet/SDCard/USB3 adapter, and not all of the USB ports worked. BTW, on that generic adapter, everything was USB except HDMI. Makes sense... it would then work on both USB-C computers and Thunderbolt 3 computers.

 

[eksu]

Do we think initializing / warming up the TB3 Titan Ridge card would be possible with DriverKit in userspace?

 

[joevt]

Do we think initializing / warming up the TB3 Titan Ridge card would be possible with DriverKit in userspace?

No idea what DriverKit allows. We still have kexts in kernel space. Should start there for now. I see they have DriverKit beta documentation. It looks like it has everything needed for PCI devices. It doesn't appear to have anything to re-enumerate hot-plugged PCI devices and bridges (IOPCIConfigurator). The WWDC video did not mention PCI devices. macOS 10.15 Catalina will continue to allow kexts.

 

[handheldgames]

Has anyone tested titan ridge on the macOS Catalina beta?

 

[atlr]

<My advice was not good, so I am replacing it with this text.>

 

[Woof Woof]

Check out BootCamp 6.2. It shows MacOS volume names. I learned about it in post https://forums.macrumors.com/thread...hyper-threading.2132317/page-88#post-26726229

In Windows 10, locate the directory in the boot camp package containing BootCamp.msi and run the following at a command prompt:
msiexec.exe /i BootCamp.msi

I observed first-hand that MacHALDriver.sys that is part of BootCamp 5.x will cause Windows 10 Build 1903 to bluescreen with a WDF_Violation error. I learned more about that at https://forums.overclockers.co.uk/threads/macbook-air-win-10-1903-wdf_violation.18855372/

Is 6.2 really recent? I downloaded the latest version (I thought it was 6.1) that Brigadier offered for the iMacPro1,1 last week. I'll check again tonight!

 

[alchemistics]

For everyone using brigadier and installs the iMacPro1,1 drivers might need to reinstall the intel thunderbolt drivers. It happened to me that device manager was showing apple inc drivers for the titan ridge card. Therefore no devices were detected before reinstalling the intel ones. FYI.

And I finally managed to get a reliable boot procedure by installing w10 over vmware fusion and winToUsb in legacy bios mode to an ssd attached via usb. Then installed the ssd internally. With the iMacPro1,1 drivers I can boot from w10 directly into macOS and vice versa.

 

[cdf]

And I finally managed to get a reliable boot procedure by installing w10 over vmware fusion and winToUsb in legacy bios mode to an ssd attached via usb. Then installed the ssd internally.

Thanks for the information. Could you explain why you had to install to the SSD externally only to install the SSD internally after?

 

[alchemistics]

Thanks for the information. Could you explain why you had to install to the SSD externally only to install the SSD internally after?

Sure, ya welcome..

The procedure is detailed in this guide:
https://9to5mac.com/2017/08/31/how-windows-10-mac-boot-camp-external-drive-video/

It was quicker for me to mount the SSD over usb in VMware to get windows installed to, than to find a way to mount local drives to the vm (which doesn't seem to work just out of the box..). With a sata to usb adapter it is very simple. I know there are other methods for the installation, but this worked surprisingly well.

As mentioned, the boot-procedure does work in both ways. I even tried to set bootvolume via bless (SIP disabled):

sudo bless -device /dev/disk0s2 -mount /Volumes/"Boot Camp" -setBoot -nextonly

or

sudo bless -mount "/Volumes/Boot Camp" -setBoot -nextonly

.. which resulted in a boot failure (auto shutdown). In the end I wasn't even able to set any bootvolumes anymore through an error in syspref StartupDisk, so I gave up for the moment.

Generally I can live with the noble procedure of "Set Bootvolume via syspref to W10 before macOS shutdown / Cold Boot to W10 -> Powercycle LG5K to get image -> warmboot to macOS".

Honestly with the vega gpu and now functional H264 Hardware Acceleration in FCPX this machine meets my needs for video editing and software development fully. Loyal, stable tower.

This is made possible cuz of the caring sharing of you forum members.

 

[nos1609]

Maybe this would render TB operation without windows preboot stage
Though, needs some correctios for the titanridge 4c used in our case.

 

[joevt]

Maybe this would render TB operation without windows preboot stage
Though, needs some correctios for the titanridge 4c used in our case.

That's only for Hackintosh, having Thunderbolt functionality in the ACPI firmware which probably enables a GPIO to send a Force Power signal to the Thunderbolt controller through the Thunderbolt add-in card header of the motherboard.

Classic Mac Pro doesn't have any Thunderbolt functionality in firmware. We use the jumper pin for force power instead of using a motherboard Thunderbolt add-in card header. Enumerating Thunderbolt devices and enabling PCIe communication to them is done by software in Windows. That's the part that needs to be added to MacOS.

 

[Woof Woof]

Any tips for getting windows to behave better at power on with TB devices connected? I keep getting hangs at the windows logo with devices connected to the TB3 to TB2 adapter. They tend to behave if connected after power on... which is sort of the opposite of what I expected.

Is there a slot that the Titan Ridge behaves behaves better in? Cards that complicate things? (Tempo Sata 3, Amfeltec Squid)

Does any particular version of windows behave better than another? Running 1809 currently... tempted to try 1903.

 

[joevt]

Any tips for getting windows to behave better at power on with TB devices connected? I keep getting hangs at the windows logo with devices connected to the TB3 to TB2 adapter. They tend to behave if connected after power on... which is sort of the opposite of what I expected.

Is there a slot that the Titan Ridge behaves behaves better in? Cards that complicate things? (Tempo Sata 3, Amfeltec Squid)

Does any particular version of windows behave better than another? Running 1809 currently... tempted to try 1903.

I would remove the Tempo Sata 3. It might have some BIOS firmware that affects BIOS booting into Windows.

 

[Woof Woof]

I posted in another thread: Figured out the issue with not being able to boot after selecting a windows volume in Startup Disk (Mac or Windows). Turns out that installing from DVD onto an MBR formatted disk does not guarantee a legacy boot volume. Under the right circumstances, the Windows 10 DVD installer will either refuse to install to the MBR disk, or it will convert the MBR disk to GPT and do a UEFI install.

Had to option boot the DVD from the internal DVD drive and select the Windows partition instead of the default EFI partition and THEN it would install to the MBR disk.

So if the only way you can boot into windows is to either option boot, or remove all Mac volumes... your windows volume is most likely GPT and you are actually booting UEFI.

 

[Woof Woof]

I would remove the Tempo Sata 3. It might have some BIOS firmware that affects BIOS booting into Windows.

After reinstalling Windows and making SURE it is legacy boot this time, things are behaving better. I can cold boot with thunderbolt devices connected. On a cold boot, windows shows the PCIe switch port didn't initialize in device manager, but on warm boot back to windows, that sorts itself out, and the connected TB device shows up.

BTW, do you use your titan ridge card jumpered or not jumpered? (The jumper that forces USB3 to work on cold boot.)

 

[joevt]

BTW, do you use your titan ridge card jumpered or not jumpered? (The jumper that forces USB3 to work on cold boot.)

Jumpered. If PCIe enumeration can be made to happen after boot, then maybe a non-jumpered setup where no devices are connected (until after boot) can be tested.

 

[cdf]

I did some more testing with the Apple USB-C AV multiport adapters (see post #352). On each adapter, the USB-A port is actually connected to a built-in USB 3 hub. Therefore, it seems that the data is not passed to the USB 2 header of the Titan Ridge card, but to the USB 3 circuitry. In fact, USB 3 devices connected to the Titan Ridge card through the adapters work as expected.

 

[joevt]

I did some more testing with the Apple USB-C AV multiport adapters (see post #352). On each adapter, the USB-A port is actually connected to a built-in USB 3 hub. Therefore, it seems that the data is not passed to the USB 2 header of the Titan Ridge card, but to the USB 3 circuitry. In fact, USB 3 devices connected to the Titan Ridge card through the adapters work as expected.

Usually, every USB 3 hub has a corresponding USB 2 hub (USB 3 and USB 2 pins are separate in a USB cable).

Take a look at the example below (I've excluded some buses and devices from the output to make it simpler). It shows a HP Thunderbolt Dock G2 connected in USB mode (using a USB-C to USB-A male cable and a USB-A female to USB-C cable) to a GC-TITAN RIDGE.

The GC-TITAN RIDGE has a USB 3.1 Bus controller (PCI Device ID: 0x15ec) with USB 2.0 and USB 3.1 ports but for some unknown reason, the USB 2.0 lines are source from a different controller through a USB 2.0 header on the GC-TITAN RIDGE. In my case, the USB 2.0 header of the GC-TITAN RIDGE is connected to a HUBBY7 USB 2.0 hub (product ID: 0x0201) which is connected to a USB 2.0 port of a USB 2.0 Bus controller (PCI Device ID: 0x268c) in the MacPro3,1.

Product ID: 0x2820 and 0x0820 are the USB 3.1 gen 2 hub of the HP Thunderbolt Dock G2. Connected to that hub is a USB 3.1 gen 1 hub having Product ID: 0x2807 and 0x5807. Both hubs are internal to the HP Thunderbolt Dock G2. You see that each USB 3 hub has a product ID for the USB 3 connection and a different product ID for the USB 2 connection.

system_profiler SPUSBDataType > systemprofilerusb.txt

USB:

    USB 2.0 Bus:

      Host Controller Driver: AppleUSBEHCIPCI
      PCI Device ID: 0x268c
      PCI Revision ID: 0x0009
      PCI Vendor ID: 0x8086

        USB 2.0 Hub [MTT]:

          Product ID: 0x0201
          Vendor ID: 0x1a40  (TERMINUS TECHNOLOGY INC.)
          Version: 1.00
          Speed: Up to 480 Mb/sec
          Location ID: 0xfd500000 / 1
          Current Available (mA): 500
          Current Required (mA): 100
          Extra Operating Current (mA): 0

            USB2.0 Hub             :

              Product ID: 0x2820
              Vendor ID: 0x2109  (VIA Labs, Inc.)
              Version: 70.13
              Speed: Up to 480 Mb/sec
              Manufacturer: VIA Labs, Inc.       
              Location ID: 0xfd530000 / 6
              Current Available (mA): 500
              Current Required (mA): 0
              Extra Operating Current (mA): 0

                USB2807 Hub:

                  Product ID: 0x2807
                  Vendor ID: 0x0424  (SMSC)
                  Version: 2.04
                  Speed: Up to 480 Mb/sec
                  Manufacturer: Microchip
                  Location ID: 0xfd533000 / 7
                  Current Available (mA): 500
                  Current Required (mA): 0
                  Extra Operating Current (mA): 0

    USB 3.1 Bus:

      Host Controller Driver: AppleUSBXHCIPCI
      PCI Device ID: 0x15ec
      PCI Revision ID: 0x0006
      PCI Vendor ID: 0x8086

        USB3.1 Hub             :

          Product ID: 0x0820
          Vendor ID: 0x2109  (VIA Labs, Inc.)
          Version: 70.13
          Speed: Up to 10 Gb/sec
          Manufacturer: VIA Labs, Inc.       
          Location ID: 0x01300000 / 2
          Current Available (mA): 500
          Current Required (mA): 0
          Extra Operating Current (mA): 0

            USB5807 Hub:

              Product ID: 0x5807
              Vendor ID: 0x0424  (SMSC)
              Version: 2.04
              Speed: Up to 5 Gb/sec
              Manufacturer: Microchip
              Location ID: 0x01330000 / 3
              Current Available (mA): 900
              Current Required (mA): 0
              Extra Operating Current (mA): 0

 

[cdf]

Usually, every USB 3 hub has a corresponding USB 2 hub (USB 3 and USB 2 pins are separate in a USB cable).

This is how an Apple USB-C Digital AV Multiport Adapter is seen on a MacBook9,1:

USB:

    USB 3.0 Bus:

      Host Controller Driver: AppleUSBXHCISPTLP
      PCI Device ID: 0x9d2f
      PCI Revision ID: 0x0021
      PCI Vendor ID: 0x8086

        USB3.0 Hub:

          Product ID: 0x100e
          Vendor ID: 0x05ac (Apple Inc.)
          Version: 45.26
          Speed: Up to 5 Gb/sec
          Manufacturer: Apple Inc.
          Location ID: 0x14500000 / 1
          Current Available (mA): 900
          Current Required (mA): 144
          Extra Operating Current (mA): 0

        USB2.0 Hub:

          Product ID: 0x100f
          Vendor ID: 0x05ac (Apple Inc.)
          Version: 45.26
          Speed: Up to 480 Mb/sec
          Manufacturer: Apple Inc.
          Location ID: 0x14100000 / 2
          Current Available (mA): 500
          Current Required (mA): 100
          Extra Operating Current (mA): 0

            USB-C Digital AV Multiport Adapter:

              Product ID: 0x1460
              Vendor ID: 0x05ac (Apple Inc.)
              Version: 2.35
              Speed: Up to 1.5 Mb/sec
              Manufacturer: Apple Inc.
              Location ID: 0x14120000 / 3
              Current Available (mA): 500
              Current Required (mA): 0
              Extra Operating Current (mA): 0

And this is how it is seen on a MacPro5,1 with the GC-TITAN RIDGE and an SMC USB2514B USB 2.0 hub:

USB:

    USB 2.0 Bus:

      Host Controller Driver: AppleUSBEHCIPCI
      PCI Device ID: 0x3a3c
      PCI Revision ID: 0x0000
      PCI Vendor ID: 0x8086

        Hub:

          Product ID: 0x2514
          Vendor ID: 0x0424  (SMSC)
          Version: b.b3
          Speed: Up to 480 Mb/sec
          Location ID: 0xfa500000 / 1
          Current Available (mA): 500
          Current Required (mA): 2
          Extra Operating Current (mA): 0
          Built-In: Yes

    USB 3.1 Bus:

      Host Controller Driver: AppleUSBXHCITR
      PCI Device ID: 0x15ec
      PCI Revision ID: 0x0006
      PCI Vendor ID: 0x8086

        USB3.0 Hub:

          Product ID: 0x100e
          Vendor ID: 0x05ac (Apple Inc.)
          Version: 45.26
          Speed: Up to 5 Gb/sec
          Manufacturer: Apple Inc.
          Location ID: 0x00300000 / 1
          Current Available (mA): 900
          Current Required (mA): 144
          Extra Operating Current (mA): 0

On the MacBook, the USB 3.0 (Product ID: 0x100e) and 2.0 (Product ID: 0x100f) hubs of the USB-C Digital AV Multiport Adapter appear on the USB 3.0 bus (PCI Device ID: 0x9d2f). Note that the adapter itself appears on the USB 2.0 hub. On the MacPro, only the USB 3.0 hub of the adapter appears on the USB 3.1 bus of the GC-TITAN RIDGE. That is expected. However, the USB 2.0 hub of the adapter does not appear at all, not even on the hub (Product ID: 0x2514) that is connected to the USB header of the GC-TITAN RIDGE.

 

[joevt]

On the MacBook, the USB 3.0 (Product ID: 0x100e) and 2.0 (Product ID: 0x100f) hubs of the USB-C Digital AV Multiport Adapter appear on the USB 3.0 bus (PCI Device ID: 0x9d2f). Note that the adapter itself appears on the USB 2.0 hub. On the MacPro, only the USB 3.0 hub of the adapter appears on the USB 3.1 bus of the GC-TITAN RIDGE. That is expected. However, the USB 2.0 hub of the adapter does not appear at all, not even on the hub (Product ID: 0x2514) that is connected to the USB header of the GC-TITAN RIDGE.

Ok, seems there is a compatibility issue between the SMSC hub and the USB 2.0 hub of the USB-C Digital AV Multiport Adapter.

You said you already tried a powered hub (your display) which worked without problem. Do you have any other unpowered hubs to test such as an Apple keyboard that has USB ports? I don't think being unpowered should be a problem, because the SMSC hub only requires 2 mA and there is 500 mA available. The USB 2.0 hub of the adapter only requires 100 mA, leaving 398 mA for other devices. If there was a power problem, then you would see a "USB Accessory Needs Power" message in macOS.

The USB-C Digital AV Multiport Adapter (product ID: 0x1460) might be a billboard device or a device used for firmware updates of the adapter. USB Prober or ioreg will probably tell you.