Contribute
Register

<< Solved >> USB-C storage write hangs

Status
Not open for further replies.
@Middleman thanks for the additional insights.

Re SSDT-DTPG, this is a holdover from early guesses about getting TB working. I removed it.

As to SSDT-TB3 or SSDT-TB3HP to accompany the above, I am at a loss. I admit I don't understand this config.

With Big Sur I was able to get TB controller to appear in System Report using a HackinDROM aml by CaseySJ, but never got to test because a Cal Digit TB4 hub I wanted was wayyy backordered. Eventually I punted on whole ideas as USB Type-C was working and that was enough for my simple needs. My TB ambitions were just about completeness.

I also hoped in 2021 when I got this it I could go to RocketLake. But some dreams die.

Re SSDT-UIAC, my quick read on this says this is an alternative, older approach to USB port mapping. What I found says to get a copy of it from USBInjectAll and tailor, it much like using USBmap or Hackintool.

***Is there another angle on using SSDT-UIAC?

I see. Well about the first part, SSDT-TB3 and SSDT-TB3HP are the necessary drivers for Thunderbolt. The TB3 version is the non-hotplug (HP) version. It should be no different to CaseySJ's Hackintdrom file afaik. So long as you have one or the other you should be getting Thunderbolt output of some sort. The SSDT-DTPG - I have seen some users/setups leave both SSDTs in. In some rare cases, SSDT-DTPG may work better with Monterey.

To get the Thunderbolt going (for Titan Ridge), aside from enabling it in BIOS it might be advised to fire up Windows and see if the ports are properly reading using the Intel Thunderbolt app. This is if you haven't already flashed the TB card's chip, in which case it should recognise in macOS right away. On older systems like the Z270 on unflashed ones it was necessary to also update some Intel TBT files to get the card to switch on before rebooting back to macOS to get it to work.

As for the SSDT-UIAC, I found this file was very helpful in getting my Rocket Lake and Z490 systems to boot (along with a USBMap or USBInjectall). Without it it kind of struggled so that was my suggestion to add it in. XHCIPortLimit must be turned off for Monterey.

WRT iGPU, WEG is up to speed with 10900 and only config needed for QS/VTB enable is default Comet Lake aapl,ig-platform-id of 07003B9E and frambuffer-patch-enable, both of which are in my config. The iGPU works fine.

There's a bunch of commented connector config juju in my config. I tried very hard to get an iGPU only config to work to drive a display, but no success. The word from OC state of the union last year was this would prolly never work on z590. I did once see unaccelerated basic video output from the iGPU from the HDMI port, but that was the best I ever got. I think I got that far from a suggestion you gave me a long while ago.

Thanks again for the review and help

Okay. Well while your config seems correct, it is actually directing to the Coffee Lake IGPU values. While it works, it may have slight performance issues, because the Comet Lake IGPU section of the chip isn't activated. Normally I would try to use 00009BC5 for IGPU only and 03009BC5 for dGPU + IGPU, with 9BC50000 as device-id (which worked in Catalina & Big Sur). Both must be added as otherwise Quicksync/Sidecar/3D will not work. Framebuffer values should also be added and framebuffer patch enabled otherwise the system won't register it or have problems with enabling acceleration.

If it helps these are the settings I currently use for Monterey for the 10900K on my Z490 board, which is using SMBIOS iMacPro1,1: Screen Shot 2022-04-18 at 12.21.08 PM.png

With the update of Monterey I noticed there seems to have been some changes made to how the OS sees the IGPUs. The previous Comet Lake IGPU values didn't work so I've had to resort to using these instead.
 
@Middleman thanks for the additional insights.

Re SSDT-DTPG, this is a holdover from early guesses about getting TB working. I removed it.

As to SSDT-TB3 or SSDT-TB3HP to accompany the above, I am at a loss. I admit I don't understand this config.

With Big Sur I was able to get TB controller to appear in System Report using a HackinDROM aml by CaseySJ, but never got to test because a Cal Digit TB4 hub I wanted was wayyy backordered. Eventually I punted on whole ideas as USB Type-C was working and that was enough for my simple needs. My TB ambitions were just about completeness.

I also hoped in 2021 when I got this it I could go to RocketLake. But some dreams die.

Re SSDT-UIAC, my quick read on this says this is an alternative, older approach to USB port mapping. What I found says to get a copy of it from USBInjectAll and tailor, it much like using USBmap or Hackintool.

***Is there another angle on using SSDT-UIAC?

WRT iGPU, WEG is up to speed with 10900 and only config needed for QS/VTB enable is default Comet Lake aapl,ig-platform-id of 07003B9E and frambuffer-patch-enable, both of which are in my config. The iGPU works fine.

There's a bunch of commented connector config juju in my config. I tried very hard to get an iGPU only config to work to drive a display, but no success. The word from OC state of the union last year was this would prolly never work on z590. I did once see unaccelerated basic video output from the iGPU from the HDMI port, but that was the best I ever got. I think I got that far from a suggestion you gave me a long while ago.

Thanks again for the review and help
Normally you would have either:
USBPorts.kext

or:

USBInjectAll.kext and a SSDT
 
***This problem is solved, because it's not a USB problem.

My whole point of view has been a red-herring.

I found another Samsung 970 NVMe and put it in Sabrent NVMe enclosure, formatted as APFS and it works fine attached to both onboard and W5700 Type-C ports.

So there's either something about 980 Pro, or something about having a full install of macOS on the external that's irritating a bug. Oddness is the same drive installed in a NVMe slot works, is bootable and accessible as accessory drive. It suffers from Trim hang up that came about with a Samsung firmware update in Jan. IDK. There are levels of confusion.

Others have been complaining about USB-C external drive access on Apple forums since Monterey came out.

I foolishly assumed it's a USB mapping problem, but it's more complex.

This thread can be closed.
 
***This problem is solved, because it's not a USB problem.

My whole point of view has been a red-herring.

I found another Samsung 970 NVMe and put it in Sabrent NVMe enclosure, formatted as APFS and it works fine attached to both onboard and W5700 Type-C ports.

So there's either something about 980 Pro, or something about having a full install of macOS on the external that's irritating a bug. Oddness is the same drive installed in a NVMe slot works, is bootable and accessible as accessory drive. It suffers from Trim hang up that came about with a Samsung firmware update in Jan. IDK. There are levels of confusion.

Others have been complaining about USB-C external drive access on Apple forums since Monterey came out.

I foolishly assumed it's a USB mapping problem, but it's more complex.

This thread can be closed.
glad you got it sorted

always best to have proper usb ports setup as that can solve a lot of issues
 
***This problem is solved, because it's not a USB problem.

My whole point of view has been a red-herring.

I found another Samsung 970 NVMe and put it in Sabrent NVMe enclosure, formatted as APFS and it works fine attached to both onboard and W5700 Type-C ports.

So there's either something about 980 Pro, or something about having a full install of macOS on the external that's irritating a bug. Oddness is the same drive installed in a NVMe slot works, is bootable and accessible as accessory drive. It suffers from Trim hang up that came about with a Samsung firmware update in Jan. IDK. There are levels of confusion.

Others have been complaining about USB-C external drive access on Apple forums since Monterey came out.

I foolishly assumed it's a USB mapping problem, but it's more complex.

This thread can be closed.
Right. Well about that, vit9696 from Dortania had mentioned this (and Tonymac posted this). > https://www.tonymacx86.com/threads/general-nvme-drive-problems-fatal.316546/
Vit9696 attributed it to problems in Trim support versus Apple's NVMe driver. Some have mentioned it may also be caused by the OpenLinuxBoot driver. The root cause could be to do with how Samsung and Phison controllers work.
It could well be because of Apple's update to anticipate the launch of the Apple Mac Studio line, the drivers for handling Thunderbolt & USB may have changed somewhat (for all we know), since we now also have TB4 & USB4.
 
Holy mackerel!
For some reason, no updates to that thread ever appear in my Activity list.
So I haven't seen months of updates, even though I started the thread.
I'll have to read up :)
 
Status
Not open for further replies.
Back
Top