Issues loading the driver on Mac Pro 5,1 - If you use this machine, read this to avoid USB connection problems.
Hello.
I'm getting some inconsistent behaviour when loading the driver on Monterey, which I think is related to the launch daemon.
The first time I installed V1.3.1, the interface started up after a reboot, and worked fine until I shut down the machine. Upon restarting, it failed to start up again. I removed the installation manually, and tried to reinstall. V1.3.1 installed fine, but the interface never starts up. I have verified that installer has placed the files in the correct locations, and that the firmware binaries have been downloaded correctly.
V1.3.0 installs fine, the installer reboots the machine, and the interface starts up. With the interface working, if I soft reboot using the Apple menu, the interface starts up correctly. With the interface working, if I completely shut down using the apple menu, then restart, the interface fails to start up. With the interface working, If I unplug and replug the interface, it fails to start up again.
When I attempt to run the MISISPORTFirmwareDownloader manually, it hangs on the message Looking for uninitialised MIDISPORTs with vendor = 0x763.
My best guess is that when the installer grabs the firmware, it is copied to the interface correctly either by the installer or a post install task, but the relevant process for copying in the firmware doesn't run on boot properly.
I am a developer by trade, and experienced with C++ but I have never worked on a macos driver. I'm happy to do some disgnoatics, but I'm not sure where would be best to start.
I will have a dig through the code tonight and see if I can narrow it down a little.
For reference, I'm using a 2010 Mac Pro 5,1 running Monterey using OpenCore, and a MidiSport 2x2 plugged directly into one of the rear USB ports.
I've debugged the launch daemon, and it seems to be detecting the device and loading the firmware correctly. The log from launchd shows that is has seemingly worked...
Downloading firmware
then
Booted MIDISPORT 2x2
But after that, it fails to start up.
If I reboot, it works fine until MIDISPORT is disconnected or the machine is cold booted. Hotplugging the interface simply doesn't work at all.
Ok, disregard that. It's nothing to do with the driver at all, this is a hardware issue.
For reference - if anybody is using this on an upgraded Mac Pro 5,1, there is a USB 1.1 hotplugging issue.
If you are using Sig Sur or later, hot plugging USB 1.1 devices does not work, at all. You therefore cannot plug an older MIDISPORT directly into a USB port on the machine.
You CAN however still use it - plug a USB 2 hub into one of the onboard ports, and plug the MIDISPORT into the hub, and it works perfectly fine.
Hopefully, if anybody is using a Mac Pro upgraded past the maximum supported OS (Catalina I think), this will save you some time.
V1.3.1 is now working beautifully.
Thanks so much for the report, and diagnosis! Interesting finding. I must admit, my test environments have MIDISPORTs connected to Macs through a USB-2 hub, by virtue of needing a USB-A to USB-C adapter. I will however, test on an older MacBook Pro which possesses USB-A ports to verify your diagnosis in the next couple of days when I have access to it.
I'm not sure if this same issue will come up on other machines, but this is apparently a known issue with the Mac Pro 5,1 cheesegrater machines when running unsupported OSs I've never encountered anything like this before at all. Previously, I was running Catalina with the MIDISPORT connected directly to the USB port, with no problems at all, so this is an entirely new one on me.
Thank you for all your work on this driver. I can only speculate as to the number of hours you have saved me now I don't need to rewire my midi devices in chains and reallocate channels to them!
i'm running into the same issue on a mac book pro 2015, running unsupported OS via opencore. Os runs fine, but no matter how I plug in the midisport, it seems to not be recognized, not even getting power at all. Any ideas?