This is turning messy, so, just for the sake of clarity and resolution, here are the latest updates and some necessary(?) in-line replies.
@Edhawk &
@UtterDisbelief , I really appreciate the time you've taken to reply, but it's frustrating to be met with the same suggestions over and over. Again, this is
not a mapping issue. How do I know?
Well, I managed to flash the 3142 firmware into the ASMedia 2142 controller, and I am now getting 10Gb/s speeds
despite system information still reporting 5Gb/s (see attached image).
The speeds on PSXS,
which, by the way, are the only relevant controller in this thread – not sure why we're still talking about mapping XHCI or the TB USB-C ports – are now at 10gbps whether PXSX (or any other controller) are mapped or not, whether they are configured how @Edhawk is proposing or not. I repeat, the other ports have nothing to do with this, whether correctly or incorrectly mapped or not. Unfortunately The ASMedia controller no longer seems to work in Windows, but I'm ok with that.
Now, for some in-line replies:
Any port served from a physical USB3 (Type-A or Header) should be set with connector type USB3. This goes especially for the virtual USB2 ports served from the USB3 ports/headers, as the virtual ports need to match the parent/physical port characteristics.
This isn't clarified in any documentation or tutorial I've read thus far. I don't get it, but I understand what you're saying. I'll map my ports as such going forward, thank you.
Same goes for Type-C virtual USB2 ports, these need to match the parent/physical port Characteristics, i.e. Type-C with or without switch.
That's already the case, as stated earlier.
The only port(s) that should be set with Connector Type 'USB2' are physical USB2 (Type-A) ports on the rear I/O plate, those with a Black tang.
Any port (BT or case front) served from an internal USB2 header should be set with connector type 'Internal' not USB2.
Weird that USB2 headers should be mapped as internal. Again, I don't get it, but I understand. Thank you, and I'll proceed accordingly.
I did not call you out at all. I just explained why I
think that
@Edhawk explained what he did. It's an easy mistake to make when a thread gets longer and the change of hardware was mentioned many posts ago.
Is it really necessary, considering that I stated the motherboard in question at the very beginning of the thread? We're cluttering the thread with this matter now.
Gosh no (again) I didn't suggest any such thing. All I did was offer my take on the hardware and problems involved. I know nothing about USB after all... You will note I just suggested removing a port conflict to see what affect that had.
There'd be nothing wrong in suggesting such a thing, nor did I mean to imply any offense was taken in such a suggestion. I was seeking to fix an issue here after all, and what was implied might have very well been the source of my issue. I just didn't understand what the proposed solution was, given that I'd already tried many different iterations of USB maps, so I was asking how, exactly, to go about fixing it.
I'm not there yet myself, but I would wonder if ASMedia 2142 drivers are included any more in macOS Sonoma. It was old tech that Apple hasn't used in a while. Hey but it's possible.
I understand, but as I stated, this issue had been happening pre-Sonoma, so it's unlikely to be the source of the issue. Appreciate the thought, though.
Yes and no. You can't simply map a driver into existence. macOS includes Intel drivers and any third-party ones Apple things are relevant. For Sonoma that would be the earliest model MacPro it supports 2019 - because that model could theoretically accept add-on USB controllers. Maybe ASM2142 or more likely ASM3142.
I didn't suggest mapping a driver into existence. I was responding to the implied notion that the incorrect USB mappings could be the source of macOS's confusion in assigning/loading the correct drivers to the relevant controller/ports (PXSX, in this case). As I understand, ASM2142 is natively supported in macOS, so if the mappings were incorrect, that would explain the speed issues. Now with the ASM3142 firmware loaded into the controller, it seems like macOS is a bit happier, even without port mappings, which further supports my conclusion that this hasn't been a mapping issue after all.
I offered a suggestion but you say you've already tried.
Yes, I had been saying that I suspected an issue other than mapping, but folks didnt't seem to be carefully reading/thinking about the question at hand...
If you can boot Windows and get 10GBps then you know it is a macOS problem. There are always drivers for Windows, not so macOS.
Yes, I was finally able to test a drive in windows, and the speeds were indeed 10gbps, so definitely a macOS problem.
Anyway, we can consider this Solved