Had another panic (attached)
support file attached. Also found this when searching about DART (fyi)
---
T6000/T6001 are Apple internal names for the M1 Pro and Max SoCs. Although you have a Max (T6001), many of the drivers are going to reference just "T6000" because they're shared between the two.
DART = Device Address Resolution Table. DART is Apple's IOMMU. In Apple Silicon, almost all IO coprocessors have their own private DART, which provides address virtualization and protection services similar to the CPU's MMU.
error: SID 0 PTE invalid exception on write of DVA 0x4939780 (TTBR 0 SEG 0x2 PTE 0x24e)
I interpret this to be a report that a DART caught an illegal write attempt by the device it's guarding.
This doesn't seem like a hardware issue, it's more likely to be a macOS bug. Or, if any of the software you've installed has kernel extensions, it could be in one of them. (You'll know if you installed a KEXT - these days it involves rebooting into Recovery to downgrade the computer's security mode and check a box to permit KEXTs to be loaded.)
---
Taking a look at console, here are some things that seem to be related to USB, dunno if they help at all
AppleUSBXHCIRequest: AppleUSBXHCIRequest::finish: transfer completed with status 0xe00002eb
AppleUSBXHCIStreamingEndpoint: AppleUSBXHCIEndpoint::transferEvent: 3-7 no transfer ring claimed event 0x0000000000000000 1b000000 03078001
AppleT6000USBXHCI@00000000: AppleUSBHostController::disableStreams: pipe <private> completed with 0xe00002f0
AppleT6000USBXHCI@00000000: AppleUSBHostController::createStream: device <private> pipe <private> streamID 32 completed with 0x0
configd@(null): AppleUSBHostUserClient::closeGated: trying to close a non-opened client
This is definitely a MacOS bug. the problem is getting it fixed. We have a Mac Studio in our office that reproduces this 100% of the time with 13.1 and 7.0.1. We reported this in mid December, but I think Holidays had an impact. We hope Apple engineering will be giving this proper attention, as we can set up a debugging environment for them.
I know this is a big ask, but in dealing with apple is there even an idea of timeframe? Would I be better not waiting and use the 64k workaround? I gotta be able to do my job and this is really hampering my ability to do that.
I do understand. We may learn something end of this week, but maybe not. Its like waiting for an elephant to cross a road.
Just wanted to report back. Deleted my RAID 5 and recreated with 64k stripe size. Migrated many terabytes back and forth and zero panics. Expected performance as well.

