Kernel Panics Softraid 7.5

I just set up my M2 Ultra as well, migrated over from M1 Ultra. I have Two Thunderflex8's, both with raids as well as internal SSD drives.

I have not removed/reinstalled Softraid software yet, I have a Pro license for v7.5 and I am up to date on the software but yes, this is from a migration. I've posted prior in the forums, so you'd see my past set up etc if needed. I can't send you a Softraid log because it never gets that far. If I try and start with a drive plugged in, it won't. It just cycle resets. If I plug in the drive when the Mac is on... it freezes, pink screen for a sec, kernal panic and shuts down, entering into the reboot cycle where it never boots until I unplug the Softraid raid.

I'll uninstall, reinstall the Softraid software and cross my fingers. I also don't have my old M1 Ultra to test with. I am def not able to format my raids. I use them for work, they're loaded with data.

I'm really hoping this is a fix that is done sooner than later, as I can't be down for a week or more. Next time it crashes, I'll copy/paste the panic but I am 99% sure it'll be the same as above give or take.

I was editing my last post to include the panic info but I passed the allotted time... so here it is.

My system if it helps with anything. M2 Ultra Mac Studio, 24 core CPU, 76 Core GPU, 32 core neural engine and 192GB ram. 4TB SSD.

panic(cpu 2 caller 0xfffffe002e7a0924): "dart-apciec2 (0xfffffe2999ac0800): DART(DART) error: SID 6 write protect exception on write of DVA 0x100800b7600 (CTE 0x10 SEG 0x40 PTE 0x2d) ERROR_STATUS 0xb0680010 TIME 0xe2cc0e0171 TTE 0 AXI_ID 0" @AppleT8110DART.cpp:1760
Debugger message: panic

The rest of the panic is in the attached txt file.

Hope this helps.

@micheloupatrick 

I cannot say what we have early access to. However, note that OWC's Accelsior was shown at WWDC in the announcement of the new Mac Pro. So you can draw your own conclusions.

I am starting to think this panic is introduced by the latest MacOS release, not the hardware. the 13.5 beta's have this panic. So the new computers may be running a silent version of 13.4.

Its still too early to know for sure.

We have a computer in house now that reproduces this. It should help make progress now.

@jd1 

Yes this is the same panic.

You cannot remove the SoftRAID driver, but you can try "uninstall SoftRAID" and see if it panics without the support files installed.

Can you file a support case? It is easier to track issues like this and we will need some additional data, like a System Diagnose file.

I'm in the same boat, same computer same M2 Ultra chip and having same problems as you. I hope we get a fix soon because we use this machine to work professionally. 

Fingers crossed!

@ajpro 

File a support case, so I can track you better. If you look at the top line of the Panic report (report to Apple), it is a DART panic?

@softraid-support Not sure if I am looking in the right location, but under Softraid / Utilities I have "Uninstall All Softraid Components" as an option.

Uninstall Softraid Driver is greyed out, as is Reinstall Softraid Driver... sorry if this is straightforward, I just want to be sure. Should I select uninstall all softraid components and then possibly reinstall? Deactivate license (under Softraid menu) before hand or no?

Hopefully this helps some..

Like many in this thread, we all set up our Mac Studio M2 Ultra's using a migrated backup based off the Mac Studio M1 Ultra.

When you do a migration, the one crucial thing that needs to also be done which sort of is a common sense thing I guess, but I had forgotten so perhaps others have as well is to make sure your Mac has kernel extensions from identified developers turned on.

Restart and plug in your OWC Raid and see if it works. One of mine did... and yesterday it was simply restarting the entire system. At least now I am not completely dead in the water. 

I tried some of my other OWC products and I think I may have found the issues, or at least something of interest that might have you guys figure this out faster. 

I own two thunderflex 8 Raids.

Flex 1 is a 112TB (8x14tb HDD)

Flex 2 is 88TB (4x18 HDD) and 4x4tb U.2 NVMe SSD.

Inside Flex 2, using the PCIe slot I have a 16tb OWC Accelsior 4m2.

After the security adjustment for start up options, Flex 1 now loads as it should. Inside Softraid 7.5 it shows up, no errors.  All is fine.

Flex 2 on the other hand, when plugged in, the screen goes Pink for a moment, the system resets.

panic(cpu 12 caller 0xfffffe0019618924): "dart-apciec1 (0xfffffe2006575000): DART(DART) error: SID 7 write protect exception on write of DVA 0x100800b7600 (CTE 0x10 SEG 0x40 PTE 0x2d) ERROR_STATUS 0xb0780010 TIME 0x792cba79d8 TTE 0 AXI_ID 0" @AppleT8110DART.cpp:1760
Debugger message: panic

So assuming it is not the ThunderFlex8 itself in regards to HDD drives as my first one loads now, I am guessing it is either the internal 4x4tb U.2 NVMe SSD that came installed by you guys with my ThunderFlex8 or, the internal 16tb OWC Accelsior 4m2 that is causing the panic.

I was about to remove the Accelsior card itself to see if the raid would start without the card (even though I need the card working as well of course) but before doing this.. I unplugged everything and tried to just plug in my 8tb Thunderblade via TB4 cable and exact same thing happened, pink screen, kernel panic.

panic(cpu 0 caller 0xfffffe0018574924): "dart-apciec1 (0xfffffe29a0082000): DART(DART) error: SID 3 write protect exception on write of DVA 0x100800b7600 (CTE 0x10 SEG 0x40 PTE 0x2d) ERROR_STATUS 0xb0380010 TIME 0x1bdd7af779f TTE 0xd7f12e4368801 AXI_ID 0" @AppleT8110DART.cpp:1760
Debugger message: panic

Earlier support had said they wondered if the M2 Mac Studios are running a silent version of 13.4

In the system info, it says Ventura 13.4 (22F2073) (if this helps any).

When I look at the panics back to back, both have the exact same kernel extensions in the backtrace (of course).

Kernel Extensions in backtrace:
com.apple.driver.AppleT8110DART(1.0)[527B7E3B-23E3-3A66-9839-FDD66B25D659]@0xfffffe0019611d40->0xfffffe001961bf2b
dependency: com.apple.driver.AppleARMPlatform(1.0.2)[5C39551D-BD06-385F-AA03-D9B3E5448D90]@0xfffffe0018795590->0xfffffe00187e6333
dependency: com.apple.driver.IODARTFamily(1)[56D2B877-3106-3212-82E8-7E3361BBD2B7]@0xfffffe0019ecd9f0->0xfffffe0019ee1643
com.apple.driver.AppleInterruptControllerV2(1.0d1)[99A769A8-1E62-38A0-8E7B-70FF98F666B3]@0xfffffe00190bd1b0->0xfffffe00190bfa63
dependency: com.apple.driver.AppleARMPlatform(1.0.2)[5C39551D-BD06-385F-AA03-D9B3E5448D90]@0xfffffe0018795590->0xfffffe00187e6333

In the actual panic report they both mention write protect exception. Not sure what that is or if it matters/helps.

DART(DART) error: SID 7 write protect exception on write of DVA 0x100800b7600 (CTE 0x10 SEG 0x40 PTE 0x2d) ERROR_STATUS 0xb0780010

DART(DART) error: SID 3 write protect exception on write of DVA 0x100800b7600 (CTE 0x10 SEG 0x40 PTE 0x2d) ERROR_STATUS 0xb0380010

@jd1 

So far, all panics are on NVMe blades. SO it could be either both your cards, or shuttle in the FLex 8.

You cannot uninstall the SoftRAID driver in 13.3 and later. that is why uninstall driver is grayed out.

Also with 13.3 and later, you no longer need to reduce security for the SoftRAID driver to load. You do, for other thurd party utilities, such as Dock Ejector, or any other app that installs a kernel extension

So far, we see MacOS build with 073 panics, with 063, it does not. We have started testing 13.4.1, but see panics with that also.

After the update to 13.4.1 I'm getting kernel panics again (on an M1 Ultra Studio).

According to the panic log, it's not the DART chip but "LLC bus error from cpu2". Whatever that means.

Such a bummer as it was humming along quite nicely with 13.4 and SoftRAID 7.5 even supporting sleep and all...

@gvansteelant 

Apparently "LLC bus error from cpu2" is a MacOS sleep issue. Here is a comment on the Apple discussion forums:

https://discussions.apple.com/thread/253730722

Seems unrelated to SoftRAID and this current issue. See if their disable sleep suggestion works for you.

Notice: We have found root cause of this bug. While we cannot discuss the particulars, a fix will be coming in the not far future.

Those who are experiencing this panic, will have to either convert their volumes to Apple RAID so you can use your drives now, or else wait for the next MacOS upgrade to be posted.

@softraid-support 

by « convert their volumes » do you mean reformatting using Disk Utility ? Or is there a way to convert a Softraid RAID 0 volume to Apple Raid without deleting the volume ?

@micheloupatrick 
We can only convert HFS volumes at present. We will be adding APFS convert as soon as we can. But it is not days away, probably closer to a couple months.
So for most all users, this means backup (I would just zero the disks), then create R0 as Apple RAID, and restore.

@softraid-support 

OK, thank you.