Understand the role of drivers in bridging the gap between software and hardware, ensuring smooth hardware functionality.

Drivers Documentation

Posts under Drivers subtopic

Post

Replies

Boosts

Views

Activity

[DriverKit SCSI] SCSI probe stalls for Target ID > 0 with IOUserSCSIParallelInterfaceController
Hello everyone, We are migrating a KEXT storage driver to DriverKit. In our KEXT, we use a "one LUN = one Target" model and successfully create multiple targets in a loop during initialization. We are now trying to replicate this architecture in our DEXT. The issue is that only Target 0 is fully probed and mounted. For Target 1, the lifecycle silently stops after the first TEST UNIT READY command is successfully acknowledged. The macOS SCSI layer never sends any subsequent probe commands (like INQUIRY) to this target. The failure sequence for Target 1, observed from our logs (regardless of whether Target 0 is created), is as follows: AsyncCreateTargetForID(1) -> UserInitializeTargetForID(1) (Succeeds) UserProcessParallelTask(Target: 1, Opcode: TUR) (Succeeds) The DEXT correctly acknowledges the TUR command for Target 1 by returning kSCSITaskStatus_CHECK_CONDITION with UNIT ATTENTION in the Sense Data (Succeeds) <-- Breakpoint --> UserProcessParallelTask(Target: 1, Opcode: INQUIRY) (Never happens) Through log comparison, we have confirmed that the DEXT's response to the TUR command for Target 1 is identical to the successful KEXT's response. We have tried creating only Target 1 (skipping Target 0 entirely), but the behavior is exactly the same -> the probe still stalls after the TUR. We initially suspected a race condition caused by consecutive calls to AsyncCreateTargetForID(). We attempted several methods to ensure that targets are created sequentially, such as trying to build a "creation chain" using OSAction completion handlers. However, these attempts were unsuccessful due to various compilation errors and API misunderstandings. In any case, this "race condition" theory was ultimately disproven by our experiment where creating only Target 1 still resulted in failure. We would like to ask two questions: Is our inability to have a Target ID greater than 0 fully probed by macOS a bug in our own code, or could there be another reason we are unaware of? If we do indeed need a "one-after-another" creation mechanism for AsyncCreateTargetForID, what is the correct way to implement a "chained creation" using OSAction completion handlers in DriverKit? Thank you for any help or guidance. Best Regards, Charles
3
0
1.5k
Jan ’26
Neither macOS 14.7 "Standard" 'AppleUserHIDEventDriver' Matching Driver Nor Custom HIDDriverKit Driver 'IOUserHIDEventService::dispatchDigitizerTouchEvent' API Work for a HID-standard Digitizer Touch Pad Device
I have been working on a multi-platform multi-touch HID-standard digitizer clickpad device. The device uses Bluetooth Low Energy (BLE) as its connectivity transport and advertises HID over GATT. To date, I have the device working successfully on Windows 11 as a multi-touch, gesture-capable click pad with no custom driver or app on Windows. However, I have been having difficulty getting macOS to recognize and react to it as a HID-standard multi-touch click pad digitizer with either the standard Apple HID driver (AppleUserHIDEventDriver) or with a custom-coded driver extension (DEXT) modeled, based on the DTS stylus example and looking at the IOHIDFamily open source driver(s). The trackpad works with full-gesture support on Windows 11 and the descriptors seem to be compliant with the R23 Accessory Guidelines document, §15. With the standard, matching Apple AppleUserHIDEventDriver HID driver, when enumerating using stock-standard HID mouse descriptors, the device works fine on macOS 14.7 "Sonoma" as a relative pointer device with scroll wheel capability (two finger swipe generates a HID scroll report) and a single button. With the standard, matching Apple AppleUserHIDEventDriver HID driver, when enumerating using stock-standard HID digitizer click/touch pad descriptors (those same descriptors used successfully on Windows 11), the device does nothing. No button, no cursor, no gestures, nothing. Looking at ioreg -filtb, all of the key/value pairs for the driver match look correct. Because, even with the Apple open source IOHIDFamily drivers noted above, we could get little visibility into what might be going wrong, I wrote a custom DriverKit/HIDDriverKit driver extension (DEXT) (as noted above, based on the DTS HID stylus example and the open source IOHIDEventDriver. With that custom driver, I can get a single button click from the click pad to work by dispatching button events to dispatchRelativePointerEvent; however, when parsing, processing, and dispatching HID digitizer touch finger (that is, transducer) events via IOUserHIDEventService::dispatchDigitizerTouchEvent, nothing happens. If I log with: % sudo log stream --info --debug --predicate '(subsystem == "com.apple.iohid")' either using the standard AppleUserHIDEventDriver driver or our custom driver, we can see that our input events are tickling the IOHIDNXEventTranslatorSessionFilter HID event filter, so we know HID events are getting from the device into the macOS HID stack. This was further confirmed with the DTS Bluetooth PacketLogger app. Based on these events flowing in and hitting IOHIDNXEventTranslatorSessionFilter, using the standard AppleUserHIDEventDriver driver or our custom driver, clicks or click pad activity will either wake the display or system from sleep and activity will keep the display or system from going to sleep. In short, whether with the stock driver or our custom driver, HID input reports come in over Bluetooth and get processed successfully; however, nothing happens—no pointer movement or gesture recognition. STEPS TO REPRODUCE For the standard AppleUserHIDEventDriver: Pair the device with macOS 14.7 "Sonoma" using the Bluetooth menu. Confirm that it is paired / bonded / connected in the Bluetooth menu. Attempt to click or move one or more fingers on the touchpad surface. Nothing happens. For the our custom driver: Pair the device with macOS 14.7 "Sonoma" using the Bluetooth menu. Confirm that it is paired / bonded / connected in the Bluetooth menu. Attempt to click or move one or more fingers on the touchpad surface. Clicks are correctly registered. With transducer movement, regardless of the number of fingers, nothing happens.
4
0
1.8k
Dec ’25
DEXT (IOUserSCSIParallelInterfaceController): Direct I/O Succeeds, but Buffered I/O Fails with Data Corruption on Large File Copies
Hi all, We are migrating a SCSI HBA driver from KEXT to DriverKit (DEXT), with our DEXT inheriting from IOUserSCSIParallelInterfaceController. We've encountered a data corruption issue that is reliably reproducible under specific conditions and are hoping for some assistance from the community. Hardware and Driver Configuration: Controller: LSI 3108 DEXT Configuration: We are reporting our hardware limitations to the framework via the UserReportHBAConstraints function, with the following key settings: // UserReportHBAConstraints... addConstraint(kIOMaximumSegmentAddressableBitCountKey, 0x20); // 32-bit addConstraint(kIOMaximumSegmentCountWriteKey, 129); addConstraint(kIOMaximumByteCountWriteKey, 0x80000); // 512KB Observed Behavior: Direct I/O vs. Buffered I/O We've observed that the I/O behavior differs drastically depending on whether it goes through the system file cache: 1. Direct I/O (Bypassing System Cache) -> 100% Successful When we use fio with the direct=1 flag, our read/write and data verification tests pass perfectly for all file sizes, including 20GB+. 2. Buffered I/O (Using System Cache) -> 100% Failure at >128MB Whether we use the standard cp command or fio with the direct=1 option removed to simulate buffered I/O, we observe the exact same, clear failure threshold: Test Results: File sizes ≤ 128MB: Success. Data checksums match perfectly. File sizes ≥ 256MB: Failure. Checksums do not match, and the destination file is corrupted. Evidence of failure reproduced with fio (buffered_integrity_test.fio, with direct=1 removed): fio --size=128M buffered_integrity_test.fio -> Test Succeeded (err=0). fio --size=256M buffered_integrity_test.fio -> Test Failed (err=92), reporting the following error, which proves a data mismatch during the verification phase: verify: bad header ... at file ... offset 1048576, length 1048576 fio: ... error=Illegal byte sequence Our Analysis and Hypothesis The phenomenon of "Direct I/O succeeding while Buffered I/O fails" suggests the problem may be related to the cache synchronization mechanism at the end of the I/O process: Our UserProcessParallelTask_Impl function correctly handles READ and WRITE commands. When cp or fio (buffered) runs, the WRITE commands are successfully written to the LSI 3108 controller's onboard DRAM cache, and success is reported up the stack. At the end of the operation, to ensure data is flushed to disk, the macOS file system issues an fsync, which is ultimately translated into a SYNCHRONIZE CACHE SCSI command (Opcode 0x35 or 0x91) and sent to our UserProcessParallelTask_Impl. We hypothesize that our code may not be correctly identifying or handling this SYNCHRONIZE CACHE opcode. It might be reporting "success" up the stack without actually commanding the hardware to flush its cache to the physical disk. The OS receives this "success" status and assumes the operation is safely complete. In reality, however, the last batch of data remains only in the controller's volatile DRAM cache and is eventually lost. This results in an incomplete or incorrect file tail, and while the file size may be correct, the data checksum will inevitably fail. Summary Our DEXT driver performs correctly when handling Direct I/O but consistently fails with data corruption when handling Buffered I/O for files larger than 128MB. We can reliably reproduce this issue using fio with the direct=1 option removed. The root cause is very likely the improper handling of the SYNCHRONIZE CACHE command within our UserProcessParallelTask. P.S. This issue did not exist in the original KEXT version of the driver. We would appreciate any advice or guidance on this issue. Thank you.
13
0
1.8k
Dec ’25
CarPlay not working on iOS 26 beta
Just wanted to check here to see if anyone else is running into the issue of CarPlay not working at all on iOS 26 Beta 1, even with the update on Friday. I plug my phone in (wired) and CarPlay never shows up. I've seen a Reddit thread where other folks are seeing the same thing.
4
1
1.2k
Dec ’25
The total DMA size in DriverKit cannot exceed 2G?
We are developing a DriverKit driver on Apple M1. We use the following code to prepare DMA buffer: IODMACommandSpecification dmaSpecification; bzero(&dmaSpecification, sizeof(dmaSpecification)); dmaSpecification.options = kIODMACommandSpecificationNoOptions; dmaSpecification.maxAddressBits = p_dma_mgr->maxAddressBits; kret = IODMACommand::Create(p_dma_mgr->device, kIODMACommandCreateNoOptions, &dmaSpecification, &impl->dma_cmd ); if (kret != kIOReturnSuccess) { os_log(OS_LOG_DEFAULT, "Error: IODMACommand::Create failed! ret=0x%x\n", kret); impl->user_mem.reset(); IOFree(impl, sizeof(*impl)); return ret; } uint64_t flags = 0; uint32_t segmentsCount = 32; IOAddressSegment segments[32]; kret = impl->dma_cmd->PrepareForDMA(kIODMACommandPrepareForDMANoOptions, impl->user_mem.get(), 0, 0, // 0 for entire memory &flags, &segmentsCount, segments ); if (kret != kIOReturnSuccess) { OSSafeReleaseNULL(impl->dma_cmd); impl->user_mem.reset(); IOFree(impl, sizeof(*impl)); os_log(OS_LOG_DEFAULT, "Error: PrepareForDMA failed! ret=0x%x\n", kret); return kret; } I allocated several 8K BGRA video frames, each with a size of 141557760 bytes, and prepared the DMA according to the method mentioned above. The process was successful when the number of frames was 15 or fewer. However, issues arose when allocating 16 frames: Error: PrepareForDMA failed! ret=0xe00002bd By calculating, I found that the total size of 16 video frames exceeds 2GB. Is there such a limitation in DriverKit that the total DMA size cannot exceed 2GB? Are there any methods that would allow me to bypass this restriction so I can use more video frame buffers?
1
0
860
Dec ’25
Cancel 'Share age range in app'
Hello I'm testing an 'age range sharing' feature using the AgeRangeService API in an app we service. I approved 'Age Sharing' during testing. (For your information, my account is an adult account.) For repeat testing, I would like to delete the our app from 'Apps that requested user age information' or cancel the sharing status. However, there doesn't seem to be such a feature. Is there a way I can't find, or is this a feature that Apple doesn't offer?
1
0
772
Dec ’25
CPActionSheetTemplate not detected in presentedTemplate while CPAlertTemplate is
I'm developing a CarPlay app and encountered an inconsistent behavior with template detection. When I present a CPActionSheetTemplate and then print the presentedTemplate property, it returns nil. However, when I present a CPAlertTemplate, the presentedTemplate property correctly returns the template object. This inconsistency is causing issues in my app where I need to check if there's already a presented template before showing another one to avoid conflicts. Why does CPActionSheetTemplate not get detected in presentedTemplate while CPAlertTemplate does? Is this intended behavior or a bug? Any guidance on how to properly detect if a CPActionSheetTemplate is currently presented would be greatly appreciated.
1
0
559
Dec ’25
Disable ISO15693Tag Popup
Dear Apple CS, I’m working with NFC ISO15693 tags using NFCTagReaderSession / NFCISO15693Tag, and I’d like to read these tags in the background if possible. Is there any way to read this tag type without triggering the system NFC popup that iOS normally shows? Please note it will not be a public app, the app is meant for internal use for our employees only. is there an option to submit a special request for this use case? Thank you in advance!
2
0
967
Dec ’25
INQUIRY command is ILLEGAL REQUEST
I am developing a DriverKit driver with the goal of sending vendor-specific commands to a USB storage device. I have successfully created the DriverKit driver, and when I connect the USB storage device, it appears correctly in IORegistryExplorer. My driver class inherits from IOUserSCSIPeripheralDeviceType00 in the SCSIPeripheralsDriverKit framework. I also created a UserClient class that inherits from IOUserClient, and from its ExternalMethod I tried sending an INQUIRY command as a basic test to confirm that command transmission works. However, the device returns an ILLEGAL REQUEST (Sense Key 0x5 / ASC 0x20). Could someone advise what I might be doing wrong? Below are the logs output from the driver: 2025-11-14 21:00:43.573730+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] Driver - NewUserClient() - Finished. 2025-11-14 21:00:43.573733+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - Start() 2025-11-14 21:00:43.573807+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - Start() - Finished. 2025-11-14 21:00:43.574249+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - ExternalMethod() called 2025-11-14 21:00:43.574258+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - ----- SCSICmdINQUIRY ----- 2025-11-14 21:00:43.574268+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - command.fRequestedByteCountOfTransfer = 512 2025-11-14 21:00:43.575980+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - SCSICmdINQUIRY() UserSendCDB fCompletionStatus = 0x0 2025-11-14 21:00:43.575988+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - SCSICmdINQUIRY() UserSendCDB fServiceResponse = 0x2 2025-11-14 21:00:43.575990+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - SCSICmdINQUIRY() UserSendCDB fSenseDataValid = 0x1 2025-11-14 21:00:43.575992+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - SCSICmdINQUIRY() UserSendCDB VALID_RESPONSE_CODE = 0x70 2025-11-14 21:00:43.575994+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - SCSICmdINQUIRY() UserSendCDB SENSE_KEY = 0x5 2025-11-14 21:00:43.575996+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - SCSICmdINQUIRY() UserSendCDB ADDITIONAL_SENSE_CODE = 0x20 2025-11-14 21:00:43.575998+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - SCSICmdINQUIRY() UserSendCDB ADDITIONAL_SENSE_CODE_QUALIFIER = 0x0 Here is the UserClient class: class SampleDriverKitUserClient: public IOUserClient { public: virtual bool init(void) override; virtual kern_return_t Start(IOService* provider) override; virtual kern_return_t Stop(IOService* provider) override; virtual void free(void) override; virtual kern_return_t ExternalMethod( uint64_t selector, IOUserClientMethodArguments* arguments, const IOUserClientMethodDispatch* dispatch, OSObject* target, void* reference) override; void SCSICmdINQUIRY(SampleDriverKitDriver *driver) LOCALONLY; }; Here is the part that sends the INQUIRY command: void SampleDriverKitUserClient::SCSICmdINQUIRY(SampleDriverKitDriver *driver) { kern_return_t kr = KERN_SUCCESS; SCSIType00OutParameters command = {}; UInt8 dataBuffer[512] = {0}; SCSI_Sense_Data senseData = {0}; Log("----- SCSICmdINQUIRY -----"); SetCommandCDB(&command.fCommandDescriptorBlock, 0x12, 0x00, 0x00, 0x24, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00); command.fLogicalUnitNumber = 0; command.fTimeoutDuration = 10000; // milliseconds command.fRequestedByteCountOfTransfer = sizeof(dataBuffer); Log("command.fRequestedByteCountOfTransfer = %lld", command.fRequestedByteCountOfTransfer); command.fBufferDirection = kIOMemoryDirectionIn; command.fDataTransferDirection = kSCSIDataTransfer_FromTargetToInitiator; command.fDataBufferAddr = reinterpret_cast<uint64_t>(dataBuffer); command.fSenseBufferAddr = reinterpret_cast<uint64_t>(&senseData); command.fSenseLengthRequested = sizeof(senseData); if( driver ) { SCSIType00InParameters response = {}; kr = driver->UserSendCDB(command, &response); if( kr != KERN_SUCCESS ) { Log("SCSICmdINQUIRY() UserSendCDB failed (0x%x)", kr); return; } Log("SCSICmdINQUIRY() UserSendCDB fCompletionStatus = 0x%x", response.fCompletionStatus); Log("SCSICmdINQUIRY() UserSendCDB fServiceResponse = 0x%x", response.fServiceResponse); Log("SCSICmdINQUIRY() UserSendCDB fSenseDataValid = 0x%x", response.fSenseDataValid); Log("SCSICmdINQUIRY() UserSendCDB VALID_RESPONSE_CODE = 0x%x", senseData.VALID_RESPONSE_CODE); Log("SCSICmdINQUIRY() UserSendCDB SENSE_KEY = 0x%x", senseData.SENSE_KEY); Log("SCSICmdINQUIRY() UserSendCDB ADDITIONAL_SENSE_CODE = 0x%x", senseData.ADDITIONAL_SENSE_CODE); Log("SCSICmdINQUIRY() UserSendCDB ADDITIONAL_SENSE_CODE_QUALIFIER = 0x%x", senseData.ADDITIONAL_SENSE_CODE_QUALIFIER); if( response.fServiceResponse == kSCSIServiceResponse_TASK_COMPLETE ) { Log("SCSICmdINQUIRY() UserSendCDB complete success!!"); } for( int i=0; i < 5; i++ ) { Log("data [%04d]=0x%x [%04d]=0x%x [%04d]=0x%x [%04d]=0x%x [%04d]=0x%x [%04d]=0x%x [%04d]=0x%x [%04d]=0x%x", i*8+0, dataBuffer[i*8+0], i*8+1, dataBuffer[i*8+1], i*8+2, dataBuffer[i*8+2], i*8+3, dataBuffer[i*8+3], i*8+4, dataBuffer[i*8+4], i*8+5, dataBuffer[i*8+5], i*8+6, dataBuffer[i*8+6], i*8+7, dataBuffer[i*8+7] ); } char vendorID[9] = {0}; memcpy(vendorID, &dataBuffer[8], 8); Log("vendorID = %s",vendorID); char productID[17] = {0}; memcpy(productID, &dataBuffer[16], 16); Log("productID = %s",productID); } } My environment is: MacBook Pro (M2), macOS 15.6 If anyone has insight into what causes the ILLEGAL REQUEST, or what I am missing when using IOUserSCSIPeripheralDeviceType00 and UserSendCDB, I would greatly appreciate your help. Thank you.
1
0
1.1k
Nov ’25
Provider entitlements check failed
Hello, I'm trying to make a DEXT for a thunderbolt device. I started from the DriverKit template that does a very simple Hello World. I added the DriverKit PCI (development) entitlement in the developer portal. The dext is installed and activated but when I connect my thunderbolt device this is what I see in the console log: kernel DK: mydrv-0x100010a85 waiting for server com.mycompany.mydrvApp.mydrv-100010a85 kernel Driver com.mycompany.mydrvApp.mydrv has crashed 0 time(s) kernelmanagerd Launching dext com.mycompany.mydrvApp.mydrv com.mycompany.mydrvApp.mydrv 0x100010a81 e675cb5ca6b6650163cc231c6af2f7e730b56b0bf394b857ce76f8e3105eb0f1 kernel DK: mydrv-0x100010a89 waiting for server com.mycompany.mydrvApp.mydrv-100010a89 kernelmanagerd Launching driver extension: Dext com.mycompany.mydrvApp.mydrv v1 in executable dext bundle com.mycompany.mydrvApp.mydrv at /Library/SystemExtensions/DC2F3964-043D-445E-A6CF-A9D7C529B39A/com.mycompany.mydrvApp.mydrv.dext default 16:52:31.551867-0500 kernel /Library/SystemExtensions/DC2F3964-043D-445E-A6CF-A9D7C529B39A/com.mycompany.mydrvApp.mydrv.dext/com.mycompany.mydrvApp.mydrv[15788] ==> com.apple.dext kernelmanagerd Found 1 dexts with bundle identifier com.mycompany.mydrvApp.mydrv kernelmanagerd Using unique id e675cb5ca6b6650163cc231c6af2f7e730b56b0bf394b857ce76f8e3105eb0f1 to pick dext matching bundle identifier com.mycompany.mydrvApp.mydrv kernelmanagerd Picked matching dext for bundle identifier com.mycompany.mydrvApp.mydrv: Dext com.mycompany.mydrvApp.mydrv v1 in executable dext bundle com.mycompany.mydrvApp.mydrv at /Library/SystemExtensions/DC2F3964-043D-445E-A6CF-A9D7C529B39A/com.mycompany.mydrvApp.mydrv.dext kernelmanagerd Launching dext com.mycompany.mydrvApp.mydrv com.mycompany.mydrvApp.mydrv 0x100010a85 e675cb5ca6b6650163cc231c6af2f7e730b56b0bf394b857ce76f8e3105eb0f1 kernel DK: com.mycompany.mydrvApp.mydrv[15788] has team identifier L86BQ63GK2 kernelmanagerd Launching driver extension: Dext com.mycompany.mydrvApp.mydrv v1 in executable dext bundle com.mycompany.mydrvApp.mydrv at /Library/SystemExtensions/DC2F3964-043D-445E-A6CF-A9D7C529B39A/com.mycompany.mydrvApp.mydrv.dext kernel kernel DK: mydrv-0x100010a81: provider entitlements check failed kernel DK: IOUserServer(com.mycompany.mydrvApp.mydrv-0x100010a81)-0x100010a8a::exit(Entitlements check failed) What am I missing for the check to pass? Here is my mydrv.entitlements file: I tried adding IOPCIPrimaryMatch with my vendor id in info.plist, but with same result. Developer mode is on and SIP is disabled. Thanks
7
0
819
Nov ’25
Problem with DriverKitUserClientSample
First time working with DEXT, so I'm trying to make "Communicating between a DriverKit extension and a client app" sample project to work and I'm having trouble. I'm following the steps for "Automatically manage signing". It seems that the step “Sign to Run Locally” is wrong now, so I let it to Apple Developer. Now I have a Provisioning Profile error that says it doesn't include the com.apple.developer.driverkit.allow-any-userclient-access entitlement. If I go to Certificates, Identifiers and Profiles for this AppID and go to Capability Requests and select DriverKit Allow Any UserClient Access, this bring me to a page where I select DriverKit Entitlement but Any UserClient Access is not a choice. I tried asking for UserClient Access but the request has been denied saying that the engineering team does not think these entitlements are necessary for testing locally. So I removed com.apple.developer.driverkit.allow-any-userclient-access from NullDriver.entitlements. I can now install the DEXT but when I try "Communicate with Dext" button, I received the error: Failed opening connection to dext with error: 0xe00002e2. Looking in the console, I see this error message: DK: NullDriverUserClient-0x10000ad6a:UC failed userclient-access check, needed bundle ID com.example.apple-samplecode.dext-to-user-client-2-TEST.driver I tried many things, but I can't seem to be able to pass through the check to allow the application to communicate with the dext. What am I missing? Note that I'm focusing only on macOS, not iOS. Thanks.
7
0
704
Nov ’25
Architectural Performance Difference in External Storage I/O Between Intel and Apple Silicon Macs
Hi everyone, We are in the process of migrating a legacy KEXT for our external multi-disk RAID enclosure to the modern DriverKit framework. During the performance validation of our KEXT, we observed a large and consistent maximum throughput difference between Intel-based Macs and Apple Silicon-based Macs. We would like to share our findings and hope to discuss with others in the community to see if you have had similar experiences that could confirm or correct our understanding. The Observation: A Consistent Performance Gap When using the exact same external RAID hardware (an 8-HDD RAID 5 array), driven by our mature KEXT, we see the following results in high-throughput benchmarks (AJA System Test, large sequential writes): On a 2020 Intel-based Mac: We consistently achieve a throughput of ~2500 MB/s. On modern M-series Macs (from M1 to M4): The throughput is consistently capped at ~1500 MB/s. This performance difference of nearly 40% is significant and is present across the entire Apple Silicon product line. Our Hypothesis: A Shift in Architectural Design Philosophy Since the KEXT and external hardware are identical in both tests, we believe this performance difference is not a bug but a fundamental platform architecture distinction. Our hypothesis is as follows: 1. The Intel Mac Era ("Dedicated Throughput") The Intel-based Macs we tested use a dedicated, discrete Intel Thunderbolt controller chip. This chip has its own dedicated PCIe lanes and resources, and its design appears to be singularly focused on maximizing raw, sustained data throughput for external peripherals. 2. The Apple Silicon Era ("Integrated Efficiency") In contrast, M-series Macs use a deeply integrated I/O controller inside the SoC. This controller must share resources, such as the total unified memory bandwidth and the chip's overall power budget, with all other functional units (CPU, GPU, etc.). We speculate that the design priority for this integrated I/O controller has shifted from "maximizing single-task raw throughput" to "maximizing overall system efficiency, multi-task responsiveness, and low latency." As a result, in a pure, single-task storage benchmark, its performance ceiling may be lower than that of the older, dedicated-chip architecture. Our Question to the Community: Is our understanding correct? Have other developers of high-performance storage drivers or peripherals also observed a similar performance ceiling for external storage on Apple Silicon Macs, when compared to high-end Intel Macs? We believe that understanding this as a deliberate architectural trade-off is crucial for setting realistic performance targets for our DEXT. Our current goal has been adjusted to have our DEXT match the KEXT's ~1500 MB/s on the M-series platform. Any insights, confirmations, or corrections from the community or Apple engineers would be greatly appreciated. Thank you very much! Charles
1
0
1.3k
Nov ’25
How to sending capability requests.
I want to create a DriverKit driver and send vendor-specific commands to the storage device. Since UserClient is required to access the DriverKit driver from the app, I am attempting to implement it by referring to the following sample code. [Communicating between a DriverKit extension and a client app] I want to add com.apple.developer.driverkit.allow-any-userclient-access to the driver's Entitlements, submit a Capability Request on the developer site, and create a Provisioning Profile. However, I don't know how to enable com.apple.developer.driverkit.allow-any-userclient-access. I am entering the following information in the “Request a System Extension or DriverKit Entitlement” form. Which entitlement are you applying for? : DriverKit Entitlement Which DriverKit entitlements do you need? : UserClient Access UserClient Bundle IDs : [Bundle ID of MyDriver] Describe your apps and how they’ll use these entitlements. : testing sample code However, even if this request is accepted, I believe only MyDriver will be permitted. How can I grant access to all UserClients?
4
1
340
Nov ’25
How to allocate contiguous memory in DriverKit?
We want to allocate a block of contiguous memory (≤1M) for audio ring DMA usage, but we haven't found any explicit method in the DriverKit documentation for allocating contiguous memory. I'm aware that IOBufferMemoryDescriptor::Create can be used in DriverKit to allocate memory and share it with user space. However, is the allocated memory physically contiguous? Can it guarantee that when I subsequently call PrepareForDMA in IODMACommand, there will be only one segment? Could you please help review this? Thank you!
2
0
627
Oct ’25
How to completely uninstall the old kext driver?
Hi, On macOS 11 and earlier versions, we provided users with the following script to uninstall our kext driver: sudo pkgutil --only-files --files com.magewell.ProCapture | tr '\n' '\0' | xargs -n 1 -0 sudo rm -vf sudo pkgutil --only-dirs --files com.magewell.ProCapture | grep ProCapture[^/]*$ | tr '\n' '\0' | xargs -n 1 -0 sudo rm -rvf sudo pkgutil --forget com.magewell.ProCapture sudo kextcache -system-caches However, this script no longer works on macOS 13 and returns the following error: It looks like you're trying to update the system caches. As of macOS 11, the personality cache is no longer in use for keeping kext matching information up-to-date. For more information, see `man kmutil`. This indicates we can no longer use kextcache -system-caches to clear our driver cache. This creates an issue where even after installing the new dext driver, the dext driver cannot run due to the presence of the old kext driver. We've tried various methods but haven't been able to completely uninstall the old kext driver - after every new system update, the old kext reappears. The specific process is as follows: This is the sequence I followed in my latest test - Device is running macOS 13 Ventura w/ 4247 Pro Capture kext driver installed kmutil inspect | grep -i magewell - this returns references to the kext files in /Library/Extensions, which is expected because I have not yet removed the 4247 kext driver - then I ran the following combination of your removal script and my removal steps: cd / sudo rm -r /Library/Extensions/ProCaptureDriver.kext sudo rm -r /Library/Extensions/ProCaptureEvent.kext sudo rm /System/Volumes/Preboot/*/boot/*/System/Library/Caches/com.apple.kernelcaches/kernelcache.auxkc* sudo pkgutil --only-files --files com.magewell.ProCapture | tr '\n' '\0' | xargs -n 1 -0 sudo rm -vf sudo pkgutil --only-dirs --files com.magewell.ProCapture | grep ProCapture[^/]*$ | tr '\n' '\0' | xargs -n 1 -0 sudo rm -rvf sudo pkgutil --forget com.magewell.ProCapture sudo kextcache --clear-staging sudo kcditto sudo kmutil install --update-preboot sudo shutdown -r now - After this I ran 'kmutil inspect | grep -i magewell' and got no results, which seems good but... - then I ran the upgrade to macOS 15.7 Sequoia - Afterwards I ran 'kmutil inspect | grep -i magewell' and it returned references to the old /Library/Extensions kexts again, although the files no longer exist in /Library/Extensions - I then ran my cleanup process again (slightly different for Sequoia-available commands): sudo rm /System/Volumes/Preboot/*/boot/*/System/Library/Caches/com.apple.kernelcaches/kernelcache.auxkc* sudo kextcache --clear-staging sudo kmutil rebuild sudo kcditto sudo kmutil install --update-preboot sudo shutdown -r now - Then I ran 'kmutil inspect | grep -i magewell' and got no results again - As a next test I ran a minor update to macOS 15.7.1, then ran 'kmutil inspect | grep -i magewell' and the references to the old kexts came back again We have indeed identified a solution to address this issue: kmutil trigger-panic-medic --volume-root /Volumes/<YourVolumeName> However, this method requires booting into Recovery Mode, which is unacceptable for many of our customers. Especially for those who need bulk remote management, having personnel physically operate each machine one by one is simply not feasible. Therefore, is there a method to completely uninstall the kext driver while in normal mode? Thank you!
3
0
743
Oct ’25
Should UserSendCBD work on UAS interfaces?
The device I am trying to develop a firmware updater for is an NVMe drive with a USB4 interface. It can connect in USB4 mode (tunneled NVMe), in USB 3 mode or in USB 2 mode. In USB 2 and USB 3 mode, the device descriptor shows one interface with two alternates. Alternate 0 uses the bulk-only protocol, with one IN and one OUT pipe. Alternate 1 uses the UAS protocol, with two IN and two OUT pipes. I use identical code in my driver to send custom CDBs. I can see using IORegistryExplorer that in USB 2 mode, macOS chooses alternate 0, the bulk-only protocol. My custom CDBs and their accompanying data pay loads are put on the bus, more or less as expected. In USB 3 mode, macOS chooses alternate 1, the UAS protocol. My custom CDB is put on the bus, but no payload data is transferred. Is this expected behavior? If so, is there a way to force the OS to choose alternate 0 even when on USB 3, perhaps with another dext? I'll file a bug about this when Feedback Assistant lets me.
8
0
933
Oct ’25
DriverKit. Data Integrity issue.
Dear Apple engineers and DriverKit developers, We have developed a DriverKit (DEXT) driver for an HBA RAID controller. The RAID controller is connected to hosts through Thunderbolt (PCIe port of the Thunderbolt controller). We use an IO script to verify the developed driver. The test fails after a few (10-12) hours of running with an error: “BUG IN CLIENT OF LIBDISPATCH: dispatch_sync called on queue already owned by current thread”. We inspected the stack trace of the crash report. This error happens in the interrupt handler. Thread 5 Crashed: 0 libdispatch.dylib 0x19671aa8c __DISPATCH_WAIT_FOR_QUEUE__ + 484 1 libdispatch.dylib 0x19671a5d0 _dispatch_sync_f_slow + 152 2 DriverKit 0x195d3fc1c IODispatchQueue::DispatchSync_f(void*, void (*)(void*)) + 296 3 DriverKit 0x195d40860 IOInterruptDispatchSourceThread(void*) + 380 4 libsystem_pthread.dylib 0x1968a3738 _pthread_start + 140 5 libsystem_pthread.dylib 0x1968ac6c8 thread_start + 8 On our side we created 5 DispatchQueue(s) for the interrupt processing and configured 5 interrupt handlers using ConfigureInterrupts(kIOInterruptTypePCIMessagedX, 1, 5). It gives 5 interrupts, as requested and it is not clear what is the reason for the issue. Our code samples are below uint32_t configureInterrupts(uint32_t requested, uint32_t required)  {      const kern_return_t ret = ioPCIdevice->ConfigureInterrupts(kIOInterruptTypePCIMessagedX, required, requested);      if (ret != kIOReturnSuccess)          return 0;      uint64_t interruptType = 0;      uint32_t interruptsCount = 0;      uint32_t interruptIndex = 0;        for ( ;; ++interruptIndex) {          if (IOInterruptDispatchSource::GetInterruptType(ioPCIdevice, interruptIndex, &interruptType) != kIOReturnSuccess)              break;          if ((interruptType & kIOInterruptTypePCIMessagedX) == 0)              continue;          ++interruptsCount;      }      return interruptsCount;  }  ..... // Create DQs     for(int i = 0; i < maxInterrupts; ++i) {          ret = IODispatchQueue::Create(INTERRUPT_DQ_NAME, 0, 0, &ivars->interruptQueue[i]);          if (kIOReturnSuccess != ret || nullptr == ivars->interruptQueue[i]) {              GH_PRINT_ERR("Interrupts queue %d creation failed with error %d", i, ret);              return false;          }      } ..... // Link DQ with interrupt   for(int index = 0; index < maxInterrupts; ++ index) {      kern_return_t ret = CreateActionHandleInterruptRequest(size, &ivars->interruptActions[index]);      if (kIOReturnSuccess != ret) {          GH_PRINT_ERR("Create action for interrupt handler %u failed.", index);          return ;      }      ret = IOInterruptDispatchSource::Create(ivars->PCI_io.dev, index, ivars->interruptQueue[index], &ivars->interruptSources[index]);      if (kIOReturnSuccess != ret || nullptr == ivars->interruptSources[index]) {          GH_PRINT_ERR("Creating interrupt source %u failed for interrupt index %u.", index, index);          return ;      }      ret = ivars->interruptSources[index]->SetHandler(ivars->interruptActions[index]);      if (kIOReturnSuccess != ret) {          GH_PRINT_ERR("Setting the handler for interrupt source %u failed.", index);          return ;      }      ivars->contexts[index] = ivars->interruptActions[index]->GetReference();      }  ..... // definition for interrupt handler void    HandleInterruptRequest ( OSAction *action, uint64_t count, uint64_t time) TYPE (IOInterruptDispatchSource::InterruptOccurred);  Do you have any clue how we can fix this error? Or directions and ways for investigation? Please let us know if you need more details.
7
0
483
Oct ’25
DriverKit. Plug/unplug test leads to MacOS panic
Dear Apple engineers, We have developed a DriverKit (DEXT) driver for an HBA RAID controller. The RAID controller is connected to hosts through Thunderbolt (PCIe port of the Thunderbolt controller). We do plug/unplug tests to verify the developed driver. The test always fails in about 100 cycles with a MacOS crash (panic). The panic contains “LLC Bus error (Unavailable) from cpu0: FAR=0xa40100008 LLC_ERR_STS/ADR/INF=0x80/0x300480a40100008/0x1400000005 addr=0xa40100008 cmd=0x18(ACC_CIFL2C_CMD_RD_LD: request for load miss in E or S state)” At first we assumed that the issue is with hardware. But we did this test on different hosts (MacMini M3 and M4) with different units of our device. The error points to the same physical address FAR=0xa40100008 even if the hosts are different. The 2 full panic logs are attached (one for M4, another one for M3 host). Could you share your understanding of the crash and give any hints on how we can fix it? Please let us know if you need any additional data. Thank you M3 panic: https://drive.google.com/file/d/1GJXd3tTW6ajdrHpFsJxO_tWWYKYIgcMc/view?usp=share_link M4 panic: https://drive.google.com/file/d/1SU-3aBSdhLsyhhxsLknzw9wGvBQ9TbJC/view?usp=share_link
2
0
690
Oct ’25
[DriverKit SCSI] SCSI probe stalls for Target ID > 0 with IOUserSCSIParallelInterfaceController
Hello everyone, We are migrating a KEXT storage driver to DriverKit. In our KEXT, we use a "one LUN = one Target" model and successfully create multiple targets in a loop during initialization. We are now trying to replicate this architecture in our DEXT. The issue is that only Target 0 is fully probed and mounted. For Target 1, the lifecycle silently stops after the first TEST UNIT READY command is successfully acknowledged. The macOS SCSI layer never sends any subsequent probe commands (like INQUIRY) to this target. The failure sequence for Target 1, observed from our logs (regardless of whether Target 0 is created), is as follows: AsyncCreateTargetForID(1) -> UserInitializeTargetForID(1) (Succeeds) UserProcessParallelTask(Target: 1, Opcode: TUR) (Succeeds) The DEXT correctly acknowledges the TUR command for Target 1 by returning kSCSITaskStatus_CHECK_CONDITION with UNIT ATTENTION in the Sense Data (Succeeds) <-- Breakpoint --> UserProcessParallelTask(Target: 1, Opcode: INQUIRY) (Never happens) Through log comparison, we have confirmed that the DEXT's response to the TUR command for Target 1 is identical to the successful KEXT's response. We have tried creating only Target 1 (skipping Target 0 entirely), but the behavior is exactly the same -> the probe still stalls after the TUR. We initially suspected a race condition caused by consecutive calls to AsyncCreateTargetForID(). We attempted several methods to ensure that targets are created sequentially, such as trying to build a "creation chain" using OSAction completion handlers. However, these attempts were unsuccessful due to various compilation errors and API misunderstandings. In any case, this "race condition" theory was ultimately disproven by our experiment where creating only Target 1 still resulted in failure. We would like to ask two questions: Is our inability to have a Target ID greater than 0 fully probed by macOS a bug in our own code, or could there be another reason we are unaware of? If we do indeed need a "one-after-another" creation mechanism for AsyncCreateTargetForID, what is the correct way to implement a "chained creation" using OSAction completion handlers in DriverKit? Thank you for any help or guidance. Best Regards, Charles
Replies
3
Boosts
0
Views
1.5k
Activity
Jan ’26
Neither macOS 14.7 "Standard" 'AppleUserHIDEventDriver' Matching Driver Nor Custom HIDDriverKit Driver 'IOUserHIDEventService::dispatchDigitizerTouchEvent' API Work for a HID-standard Digitizer Touch Pad Device
I have been working on a multi-platform multi-touch HID-standard digitizer clickpad device. The device uses Bluetooth Low Energy (BLE) as its connectivity transport and advertises HID over GATT. To date, I have the device working successfully on Windows 11 as a multi-touch, gesture-capable click pad with no custom driver or app on Windows. However, I have been having difficulty getting macOS to recognize and react to it as a HID-standard multi-touch click pad digitizer with either the standard Apple HID driver (AppleUserHIDEventDriver) or with a custom-coded driver extension (DEXT) modeled, based on the DTS stylus example and looking at the IOHIDFamily open source driver(s). The trackpad works with full-gesture support on Windows 11 and the descriptors seem to be compliant with the R23 Accessory Guidelines document, §15. With the standard, matching Apple AppleUserHIDEventDriver HID driver, when enumerating using stock-standard HID mouse descriptors, the device works fine on macOS 14.7 "Sonoma" as a relative pointer device with scroll wheel capability (two finger swipe generates a HID scroll report) and a single button. With the standard, matching Apple AppleUserHIDEventDriver HID driver, when enumerating using stock-standard HID digitizer click/touch pad descriptors (those same descriptors used successfully on Windows 11), the device does nothing. No button, no cursor, no gestures, nothing. Looking at ioreg -filtb, all of the key/value pairs for the driver match look correct. Because, even with the Apple open source IOHIDFamily drivers noted above, we could get little visibility into what might be going wrong, I wrote a custom DriverKit/HIDDriverKit driver extension (DEXT) (as noted above, based on the DTS HID stylus example and the open source IOHIDEventDriver. With that custom driver, I can get a single button click from the click pad to work by dispatching button events to dispatchRelativePointerEvent; however, when parsing, processing, and dispatching HID digitizer touch finger (that is, transducer) events via IOUserHIDEventService::dispatchDigitizerTouchEvent, nothing happens. If I log with: % sudo log stream --info --debug --predicate '(subsystem == "com.apple.iohid")' either using the standard AppleUserHIDEventDriver driver or our custom driver, we can see that our input events are tickling the IOHIDNXEventTranslatorSessionFilter HID event filter, so we know HID events are getting from the device into the macOS HID stack. This was further confirmed with the DTS Bluetooth PacketLogger app. Based on these events flowing in and hitting IOHIDNXEventTranslatorSessionFilter, using the standard AppleUserHIDEventDriver driver or our custom driver, clicks or click pad activity will either wake the display or system from sleep and activity will keep the display or system from going to sleep. In short, whether with the stock driver or our custom driver, HID input reports come in over Bluetooth and get processed successfully; however, nothing happens—no pointer movement or gesture recognition. STEPS TO REPRODUCE For the standard AppleUserHIDEventDriver: Pair the device with macOS 14.7 "Sonoma" using the Bluetooth menu. Confirm that it is paired / bonded / connected in the Bluetooth menu. Attempt to click or move one or more fingers on the touchpad surface. Nothing happens. For the our custom driver: Pair the device with macOS 14.7 "Sonoma" using the Bluetooth menu. Confirm that it is paired / bonded / connected in the Bluetooth menu. Attempt to click or move one or more fingers on the touchpad surface. Clicks are correctly registered. With transducer movement, regardless of the number of fingers, nothing happens.
Replies
4
Boosts
0
Views
1.8k
Activity
Dec ’25
DEXT (IOUserSCSIParallelInterfaceController): Direct I/O Succeeds, but Buffered I/O Fails with Data Corruption on Large File Copies
Hi all, We are migrating a SCSI HBA driver from KEXT to DriverKit (DEXT), with our DEXT inheriting from IOUserSCSIParallelInterfaceController. We've encountered a data corruption issue that is reliably reproducible under specific conditions and are hoping for some assistance from the community. Hardware and Driver Configuration: Controller: LSI 3108 DEXT Configuration: We are reporting our hardware limitations to the framework via the UserReportHBAConstraints function, with the following key settings: // UserReportHBAConstraints... addConstraint(kIOMaximumSegmentAddressableBitCountKey, 0x20); // 32-bit addConstraint(kIOMaximumSegmentCountWriteKey, 129); addConstraint(kIOMaximumByteCountWriteKey, 0x80000); // 512KB Observed Behavior: Direct I/O vs. Buffered I/O We've observed that the I/O behavior differs drastically depending on whether it goes through the system file cache: 1. Direct I/O (Bypassing System Cache) -> 100% Successful When we use fio with the direct=1 flag, our read/write and data verification tests pass perfectly for all file sizes, including 20GB+. 2. Buffered I/O (Using System Cache) -> 100% Failure at >128MB Whether we use the standard cp command or fio with the direct=1 option removed to simulate buffered I/O, we observe the exact same, clear failure threshold: Test Results: File sizes ≤ 128MB: Success. Data checksums match perfectly. File sizes ≥ 256MB: Failure. Checksums do not match, and the destination file is corrupted. Evidence of failure reproduced with fio (buffered_integrity_test.fio, with direct=1 removed): fio --size=128M buffered_integrity_test.fio -> Test Succeeded (err=0). fio --size=256M buffered_integrity_test.fio -> Test Failed (err=92), reporting the following error, which proves a data mismatch during the verification phase: verify: bad header ... at file ... offset 1048576, length 1048576 fio: ... error=Illegal byte sequence Our Analysis and Hypothesis The phenomenon of "Direct I/O succeeding while Buffered I/O fails" suggests the problem may be related to the cache synchronization mechanism at the end of the I/O process: Our UserProcessParallelTask_Impl function correctly handles READ and WRITE commands. When cp or fio (buffered) runs, the WRITE commands are successfully written to the LSI 3108 controller's onboard DRAM cache, and success is reported up the stack. At the end of the operation, to ensure data is flushed to disk, the macOS file system issues an fsync, which is ultimately translated into a SYNCHRONIZE CACHE SCSI command (Opcode 0x35 or 0x91) and sent to our UserProcessParallelTask_Impl. We hypothesize that our code may not be correctly identifying or handling this SYNCHRONIZE CACHE opcode. It might be reporting "success" up the stack without actually commanding the hardware to flush its cache to the physical disk. The OS receives this "success" status and assumes the operation is safely complete. In reality, however, the last batch of data remains only in the controller's volatile DRAM cache and is eventually lost. This results in an incomplete or incorrect file tail, and while the file size may be correct, the data checksum will inevitably fail. Summary Our DEXT driver performs correctly when handling Direct I/O but consistently fails with data corruption when handling Buffered I/O for files larger than 128MB. We can reliably reproduce this issue using fio with the direct=1 option removed. The root cause is very likely the improper handling of the SYNCHRONIZE CACHE command within our UserProcessParallelTask. P.S. This issue did not exist in the original KEXT version of the driver. We would appreciate any advice or guidance on this issue. Thank you.
Replies
13
Boosts
0
Views
1.8k
Activity
Dec ’25
CarPlay not working on iOS 26 beta
Just wanted to check here to see if anyone else is running into the issue of CarPlay not working at all on iOS 26 Beta 1, even with the update on Friday. I plug my phone in (wired) and CarPlay never shows up. I've seen a Reddit thread where other folks are seeing the same thing.
Replies
4
Boosts
1
Views
1.2k
Activity
Dec ’25
The total DMA size in DriverKit cannot exceed 2G?
We are developing a DriverKit driver on Apple M1. We use the following code to prepare DMA buffer: IODMACommandSpecification dmaSpecification; bzero(&dmaSpecification, sizeof(dmaSpecification)); dmaSpecification.options = kIODMACommandSpecificationNoOptions; dmaSpecification.maxAddressBits = p_dma_mgr->maxAddressBits; kret = IODMACommand::Create(p_dma_mgr->device, kIODMACommandCreateNoOptions, &dmaSpecification, &impl->dma_cmd ); if (kret != kIOReturnSuccess) { os_log(OS_LOG_DEFAULT, "Error: IODMACommand::Create failed! ret=0x%x\n", kret); impl->user_mem.reset(); IOFree(impl, sizeof(*impl)); return ret; } uint64_t flags = 0; uint32_t segmentsCount = 32; IOAddressSegment segments[32]; kret = impl->dma_cmd->PrepareForDMA(kIODMACommandPrepareForDMANoOptions, impl->user_mem.get(), 0, 0, // 0 for entire memory &flags, &segmentsCount, segments ); if (kret != kIOReturnSuccess) { OSSafeReleaseNULL(impl->dma_cmd); impl->user_mem.reset(); IOFree(impl, sizeof(*impl)); os_log(OS_LOG_DEFAULT, "Error: PrepareForDMA failed! ret=0x%x\n", kret); return kret; } I allocated several 8K BGRA video frames, each with a size of 141557760 bytes, and prepared the DMA according to the method mentioned above. The process was successful when the number of frames was 15 or fewer. However, issues arose when allocating 16 frames: Error: PrepareForDMA failed! ret=0xe00002bd By calculating, I found that the total size of 16 video frames exceeds 2GB. Is there such a limitation in DriverKit that the total DMA size cannot exceed 2GB? Are there any methods that would allow me to bypass this restriction so I can use more video frame buffers?
Replies
1
Boosts
0
Views
860
Activity
Dec ’25
Cancel 'Share age range in app'
Hello I'm testing an 'age range sharing' feature using the AgeRangeService API in an app we service. I approved 'Age Sharing' during testing. (For your information, my account is an adult account.) For repeat testing, I would like to delete the our app from 'Apps that requested user age information' or cancel the sharing status. However, there doesn't seem to be such a feature. Is there a way I can't find, or is this a feature that Apple doesn't offer?
Replies
1
Boosts
0
Views
772
Activity
Dec ’25
CPActionSheetTemplate not detected in presentedTemplate while CPAlertTemplate is
I'm developing a CarPlay app and encountered an inconsistent behavior with template detection. When I present a CPActionSheetTemplate and then print the presentedTemplate property, it returns nil. However, when I present a CPAlertTemplate, the presentedTemplate property correctly returns the template object. This inconsistency is causing issues in my app where I need to check if there's already a presented template before showing another one to avoid conflicts. Why does CPActionSheetTemplate not get detected in presentedTemplate while CPAlertTemplate does? Is this intended behavior or a bug? Any guidance on how to properly detect if a CPActionSheetTemplate is currently presented would be greatly appreciated.
Replies
1
Boosts
0
Views
559
Activity
Dec ’25
Disable ISO15693Tag Popup
Dear Apple CS, I’m working with NFC ISO15693 tags using NFCTagReaderSession / NFCISO15693Tag, and I’d like to read these tags in the background if possible. Is there any way to read this tag type without triggering the system NFC popup that iOS normally shows? Please note it will not be a public app, the app is meant for internal use for our employees only. is there an option to submit a special request for this use case? Thank you in advance!
Replies
2
Boosts
0
Views
967
Activity
Dec ’25
INQUIRY command is ILLEGAL REQUEST
I am developing a DriverKit driver with the goal of sending vendor-specific commands to a USB storage device. I have successfully created the DriverKit driver, and when I connect the USB storage device, it appears correctly in IORegistryExplorer. My driver class inherits from IOUserSCSIPeripheralDeviceType00 in the SCSIPeripheralsDriverKit framework. I also created a UserClient class that inherits from IOUserClient, and from its ExternalMethod I tried sending an INQUIRY command as a basic test to confirm that command transmission works. However, the device returns an ILLEGAL REQUEST (Sense Key 0x5 / ASC 0x20). Could someone advise what I might be doing wrong? Below are the logs output from the driver: 2025-11-14 21:00:43.573730+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] Driver - NewUserClient() - Finished. 2025-11-14 21:00:43.573733+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - Start() 2025-11-14 21:00:43.573807+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - Start() - Finished. 2025-11-14 21:00:43.574249+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - ExternalMethod() called 2025-11-14 21:00:43.574258+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - ----- SCSICmdINQUIRY ----- 2025-11-14 21:00:43.574268+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - command.fRequestedByteCountOfTransfer = 512 2025-11-14 21:00:43.575980+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - SCSICmdINQUIRY() UserSendCDB fCompletionStatus = 0x0 2025-11-14 21:00:43.575988+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - SCSICmdINQUIRY() UserSendCDB fServiceResponse = 0x2 2025-11-14 21:00:43.575990+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - SCSICmdINQUIRY() UserSendCDB fSenseDataValid = 0x1 2025-11-14 21:00:43.575992+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - SCSICmdINQUIRY() UserSendCDB VALID_RESPONSE_CODE = 0x70 2025-11-14 21:00:43.575994+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - SCSICmdINQUIRY() UserSendCDB SENSE_KEY = 0x5 2025-11-14 21:00:43.575996+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - SCSICmdINQUIRY() UserSendCDB ADDITIONAL_SENSE_CODE = 0x20 2025-11-14 21:00:43.575998+0900 0x26e9 Default 0x0 0 0 kernel: (SampleDriverKitApp.SampleDriverKitDriver.dext) [DEBUG] UserClient - SCSICmdINQUIRY() UserSendCDB ADDITIONAL_SENSE_CODE_QUALIFIER = 0x0 Here is the UserClient class: class SampleDriverKitUserClient: public IOUserClient { public: virtual bool init(void) override; virtual kern_return_t Start(IOService* provider) override; virtual kern_return_t Stop(IOService* provider) override; virtual void free(void) override; virtual kern_return_t ExternalMethod( uint64_t selector, IOUserClientMethodArguments* arguments, const IOUserClientMethodDispatch* dispatch, OSObject* target, void* reference) override; void SCSICmdINQUIRY(SampleDriverKitDriver *driver) LOCALONLY; }; Here is the part that sends the INQUIRY command: void SampleDriverKitUserClient::SCSICmdINQUIRY(SampleDriverKitDriver *driver) { kern_return_t kr = KERN_SUCCESS; SCSIType00OutParameters command = {}; UInt8 dataBuffer[512] = {0}; SCSI_Sense_Data senseData = {0}; Log("----- SCSICmdINQUIRY -----"); SetCommandCDB(&command.fCommandDescriptorBlock, 0x12, 0x00, 0x00, 0x24, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00); command.fLogicalUnitNumber = 0; command.fTimeoutDuration = 10000; // milliseconds command.fRequestedByteCountOfTransfer = sizeof(dataBuffer); Log("command.fRequestedByteCountOfTransfer = %lld", command.fRequestedByteCountOfTransfer); command.fBufferDirection = kIOMemoryDirectionIn; command.fDataTransferDirection = kSCSIDataTransfer_FromTargetToInitiator; command.fDataBufferAddr = reinterpret_cast<uint64_t>(dataBuffer); command.fSenseBufferAddr = reinterpret_cast<uint64_t>(&senseData); command.fSenseLengthRequested = sizeof(senseData); if( driver ) { SCSIType00InParameters response = {}; kr = driver->UserSendCDB(command, &response); if( kr != KERN_SUCCESS ) { Log("SCSICmdINQUIRY() UserSendCDB failed (0x%x)", kr); return; } Log("SCSICmdINQUIRY() UserSendCDB fCompletionStatus = 0x%x", response.fCompletionStatus); Log("SCSICmdINQUIRY() UserSendCDB fServiceResponse = 0x%x", response.fServiceResponse); Log("SCSICmdINQUIRY() UserSendCDB fSenseDataValid = 0x%x", response.fSenseDataValid); Log("SCSICmdINQUIRY() UserSendCDB VALID_RESPONSE_CODE = 0x%x", senseData.VALID_RESPONSE_CODE); Log("SCSICmdINQUIRY() UserSendCDB SENSE_KEY = 0x%x", senseData.SENSE_KEY); Log("SCSICmdINQUIRY() UserSendCDB ADDITIONAL_SENSE_CODE = 0x%x", senseData.ADDITIONAL_SENSE_CODE); Log("SCSICmdINQUIRY() UserSendCDB ADDITIONAL_SENSE_CODE_QUALIFIER = 0x%x", senseData.ADDITIONAL_SENSE_CODE_QUALIFIER); if( response.fServiceResponse == kSCSIServiceResponse_TASK_COMPLETE ) { Log("SCSICmdINQUIRY() UserSendCDB complete success!!"); } for( int i=0; i < 5; i++ ) { Log("data [%04d]=0x%x [%04d]=0x%x [%04d]=0x%x [%04d]=0x%x [%04d]=0x%x [%04d]=0x%x [%04d]=0x%x [%04d]=0x%x", i*8+0, dataBuffer[i*8+0], i*8+1, dataBuffer[i*8+1], i*8+2, dataBuffer[i*8+2], i*8+3, dataBuffer[i*8+3], i*8+4, dataBuffer[i*8+4], i*8+5, dataBuffer[i*8+5], i*8+6, dataBuffer[i*8+6], i*8+7, dataBuffer[i*8+7] ); } char vendorID[9] = {0}; memcpy(vendorID, &dataBuffer[8], 8); Log("vendorID = %s",vendorID); char productID[17] = {0}; memcpy(productID, &dataBuffer[16], 16); Log("productID = %s",productID); } } My environment is: MacBook Pro (M2), macOS 15.6 If anyone has insight into what causes the ILLEGAL REQUEST, or what I am missing when using IOUserSCSIPeripheralDeviceType00 and UserSendCDB, I would greatly appreciate your help. Thank you.
Replies
1
Boosts
0
Views
1.1k
Activity
Nov ’25
Provider entitlements check failed
Hello, I'm trying to make a DEXT for a thunderbolt device. I started from the DriverKit template that does a very simple Hello World. I added the DriverKit PCI (development) entitlement in the developer portal. The dext is installed and activated but when I connect my thunderbolt device this is what I see in the console log: kernel DK: mydrv-0x100010a85 waiting for server com.mycompany.mydrvApp.mydrv-100010a85 kernel Driver com.mycompany.mydrvApp.mydrv has crashed 0 time(s) kernelmanagerd Launching dext com.mycompany.mydrvApp.mydrv com.mycompany.mydrvApp.mydrv 0x100010a81 e675cb5ca6b6650163cc231c6af2f7e730b56b0bf394b857ce76f8e3105eb0f1 kernel DK: mydrv-0x100010a89 waiting for server com.mycompany.mydrvApp.mydrv-100010a89 kernelmanagerd Launching driver extension: Dext com.mycompany.mydrvApp.mydrv v1 in executable dext bundle com.mycompany.mydrvApp.mydrv at /Library/SystemExtensions/DC2F3964-043D-445E-A6CF-A9D7C529B39A/com.mycompany.mydrvApp.mydrv.dext default 16:52:31.551867-0500 kernel /Library/SystemExtensions/DC2F3964-043D-445E-A6CF-A9D7C529B39A/com.mycompany.mydrvApp.mydrv.dext/com.mycompany.mydrvApp.mydrv[15788] ==> com.apple.dext kernelmanagerd Found 1 dexts with bundle identifier com.mycompany.mydrvApp.mydrv kernelmanagerd Using unique id e675cb5ca6b6650163cc231c6af2f7e730b56b0bf394b857ce76f8e3105eb0f1 to pick dext matching bundle identifier com.mycompany.mydrvApp.mydrv kernelmanagerd Picked matching dext for bundle identifier com.mycompany.mydrvApp.mydrv: Dext com.mycompany.mydrvApp.mydrv v1 in executable dext bundle com.mycompany.mydrvApp.mydrv at /Library/SystemExtensions/DC2F3964-043D-445E-A6CF-A9D7C529B39A/com.mycompany.mydrvApp.mydrv.dext kernelmanagerd Launching dext com.mycompany.mydrvApp.mydrv com.mycompany.mydrvApp.mydrv 0x100010a85 e675cb5ca6b6650163cc231c6af2f7e730b56b0bf394b857ce76f8e3105eb0f1 kernel DK: com.mycompany.mydrvApp.mydrv[15788] has team identifier L86BQ63GK2 kernelmanagerd Launching driver extension: Dext com.mycompany.mydrvApp.mydrv v1 in executable dext bundle com.mycompany.mydrvApp.mydrv at /Library/SystemExtensions/DC2F3964-043D-445E-A6CF-A9D7C529B39A/com.mycompany.mydrvApp.mydrv.dext kernel kernel DK: mydrv-0x100010a81: provider entitlements check failed kernel DK: IOUserServer(com.mycompany.mydrvApp.mydrv-0x100010a81)-0x100010a8a::exit(Entitlements check failed) What am I missing for the check to pass? Here is my mydrv.entitlements file: I tried adding IOPCIPrimaryMatch with my vendor id in info.plist, but with same result. Developer mode is on and SIP is disabled. Thanks
Replies
7
Boosts
0
Views
819
Activity
Nov ’25
Problem with DriverKitUserClientSample
First time working with DEXT, so I'm trying to make "Communicating between a DriverKit extension and a client app" sample project to work and I'm having trouble. I'm following the steps for "Automatically manage signing". It seems that the step “Sign to Run Locally” is wrong now, so I let it to Apple Developer. Now I have a Provisioning Profile error that says it doesn't include the com.apple.developer.driverkit.allow-any-userclient-access entitlement. If I go to Certificates, Identifiers and Profiles for this AppID and go to Capability Requests and select DriverKit Allow Any UserClient Access, this bring me to a page where I select DriverKit Entitlement but Any UserClient Access is not a choice. I tried asking for UserClient Access but the request has been denied saying that the engineering team does not think these entitlements are necessary for testing locally. So I removed com.apple.developer.driverkit.allow-any-userclient-access from NullDriver.entitlements. I can now install the DEXT but when I try "Communicate with Dext" button, I received the error: Failed opening connection to dext with error: 0xe00002e2. Looking in the console, I see this error message: DK: NullDriverUserClient-0x10000ad6a:UC failed userclient-access check, needed bundle ID com.example.apple-samplecode.dext-to-user-client-2-TEST.driver I tried many things, but I can't seem to be able to pass through the check to allow the application to communicate with the dext. What am I missing? Note that I'm focusing only on macOS, not iOS. Thanks.
Replies
7
Boosts
0
Views
704
Activity
Nov ’25
Architectural Performance Difference in External Storage I/O Between Intel and Apple Silicon Macs
Hi everyone, We are in the process of migrating a legacy KEXT for our external multi-disk RAID enclosure to the modern DriverKit framework. During the performance validation of our KEXT, we observed a large and consistent maximum throughput difference between Intel-based Macs and Apple Silicon-based Macs. We would like to share our findings and hope to discuss with others in the community to see if you have had similar experiences that could confirm or correct our understanding. The Observation: A Consistent Performance Gap When using the exact same external RAID hardware (an 8-HDD RAID 5 array), driven by our mature KEXT, we see the following results in high-throughput benchmarks (AJA System Test, large sequential writes): On a 2020 Intel-based Mac: We consistently achieve a throughput of ~2500 MB/s. On modern M-series Macs (from M1 to M4): The throughput is consistently capped at ~1500 MB/s. This performance difference of nearly 40% is significant and is present across the entire Apple Silicon product line. Our Hypothesis: A Shift in Architectural Design Philosophy Since the KEXT and external hardware are identical in both tests, we believe this performance difference is not a bug but a fundamental platform architecture distinction. Our hypothesis is as follows: 1. The Intel Mac Era ("Dedicated Throughput") The Intel-based Macs we tested use a dedicated, discrete Intel Thunderbolt controller chip. This chip has its own dedicated PCIe lanes and resources, and its design appears to be singularly focused on maximizing raw, sustained data throughput for external peripherals. 2. The Apple Silicon Era ("Integrated Efficiency") In contrast, M-series Macs use a deeply integrated I/O controller inside the SoC. This controller must share resources, such as the total unified memory bandwidth and the chip's overall power budget, with all other functional units (CPU, GPU, etc.). We speculate that the design priority for this integrated I/O controller has shifted from "maximizing single-task raw throughput" to "maximizing overall system efficiency, multi-task responsiveness, and low latency." As a result, in a pure, single-task storage benchmark, its performance ceiling may be lower than that of the older, dedicated-chip architecture. Our Question to the Community: Is our understanding correct? Have other developers of high-performance storage drivers or peripherals also observed a similar performance ceiling for external storage on Apple Silicon Macs, when compared to high-end Intel Macs? We believe that understanding this as a deliberate architectural trade-off is crucial for setting realistic performance targets for our DEXT. Our current goal has been adjusted to have our DEXT match the KEXT's ~1500 MB/s on the M-series platform. Any insights, confirmations, or corrections from the community or Apple engineers would be greatly appreciated. Thank you very much! Charles
Replies
1
Boosts
0
Views
1.3k
Activity
Nov ’25
How to sending capability requests.
I want to create a DriverKit driver and send vendor-specific commands to the storage device. Since UserClient is required to access the DriverKit driver from the app, I am attempting to implement it by referring to the following sample code. [Communicating between a DriverKit extension and a client app] I want to add com.apple.developer.driverkit.allow-any-userclient-access to the driver's Entitlements, submit a Capability Request on the developer site, and create a Provisioning Profile. However, I don't know how to enable com.apple.developer.driverkit.allow-any-userclient-access. I am entering the following information in the “Request a System Extension or DriverKit Entitlement” form. Which entitlement are you applying for? : DriverKit Entitlement Which DriverKit entitlements do you need? : UserClient Access UserClient Bundle IDs : [Bundle ID of MyDriver] Describe your apps and how they’ll use these entitlements. : testing sample code However, even if this request is accepted, I believe only MyDriver will be permitted. How can I grant access to all UserClients?
Replies
4
Boosts
1
Views
340
Activity
Nov ’25
How to allocate contiguous memory in DriverKit?
We want to allocate a block of contiguous memory (≤1M) for audio ring DMA usage, but we haven't found any explicit method in the DriverKit documentation for allocating contiguous memory. I'm aware that IOBufferMemoryDescriptor::Create can be used in DriverKit to allocate memory and share it with user space. However, is the allocated memory physically contiguous? Can it guarantee that when I subsequently call PrepareForDMA in IODMACommand, there will be only one segment? Could you please help review this? Thank you!
Replies
2
Boosts
0
Views
627
Activity
Oct ’25
How to completely uninstall the old kext driver?
Hi, On macOS 11 and earlier versions, we provided users with the following script to uninstall our kext driver: sudo pkgutil --only-files --files com.magewell.ProCapture | tr '\n' '\0' | xargs -n 1 -0 sudo rm -vf sudo pkgutil --only-dirs --files com.magewell.ProCapture | grep ProCapture[^/]*$ | tr '\n' '\0' | xargs -n 1 -0 sudo rm -rvf sudo pkgutil --forget com.magewell.ProCapture sudo kextcache -system-caches However, this script no longer works on macOS 13 and returns the following error: It looks like you're trying to update the system caches. As of macOS 11, the personality cache is no longer in use for keeping kext matching information up-to-date. For more information, see `man kmutil`. This indicates we can no longer use kextcache -system-caches to clear our driver cache. This creates an issue where even after installing the new dext driver, the dext driver cannot run due to the presence of the old kext driver. We've tried various methods but haven't been able to completely uninstall the old kext driver - after every new system update, the old kext reappears. The specific process is as follows: This is the sequence I followed in my latest test - Device is running macOS 13 Ventura w/ 4247 Pro Capture kext driver installed kmutil inspect | grep -i magewell - this returns references to the kext files in /Library/Extensions, which is expected because I have not yet removed the 4247 kext driver - then I ran the following combination of your removal script and my removal steps: cd / sudo rm -r /Library/Extensions/ProCaptureDriver.kext sudo rm -r /Library/Extensions/ProCaptureEvent.kext sudo rm /System/Volumes/Preboot/*/boot/*/System/Library/Caches/com.apple.kernelcaches/kernelcache.auxkc* sudo pkgutil --only-files --files com.magewell.ProCapture | tr '\n' '\0' | xargs -n 1 -0 sudo rm -vf sudo pkgutil --only-dirs --files com.magewell.ProCapture | grep ProCapture[^/]*$ | tr '\n' '\0' | xargs -n 1 -0 sudo rm -rvf sudo pkgutil --forget com.magewell.ProCapture sudo kextcache --clear-staging sudo kcditto sudo kmutil install --update-preboot sudo shutdown -r now - After this I ran 'kmutil inspect | grep -i magewell' and got no results, which seems good but... - then I ran the upgrade to macOS 15.7 Sequoia - Afterwards I ran 'kmutil inspect | grep -i magewell' and it returned references to the old /Library/Extensions kexts again, although the files no longer exist in /Library/Extensions - I then ran my cleanup process again (slightly different for Sequoia-available commands): sudo rm /System/Volumes/Preboot/*/boot/*/System/Library/Caches/com.apple.kernelcaches/kernelcache.auxkc* sudo kextcache --clear-staging sudo kmutil rebuild sudo kcditto sudo kmutil install --update-preboot sudo shutdown -r now - Then I ran 'kmutil inspect | grep -i magewell' and got no results again - As a next test I ran a minor update to macOS 15.7.1, then ran 'kmutil inspect | grep -i magewell' and the references to the old kexts came back again We have indeed identified a solution to address this issue: kmutil trigger-panic-medic --volume-root /Volumes/<YourVolumeName> However, this method requires booting into Recovery Mode, which is unacceptable for many of our customers. Especially for those who need bulk remote management, having personnel physically operate each machine one by one is simply not feasible. Therefore, is there a method to completely uninstall the kext driver while in normal mode? Thank you!
Replies
3
Boosts
0
Views
743
Activity
Oct ’25
Should UserSendCBD work on UAS interfaces?
The device I am trying to develop a firmware updater for is an NVMe drive with a USB4 interface. It can connect in USB4 mode (tunneled NVMe), in USB 3 mode or in USB 2 mode. In USB 2 and USB 3 mode, the device descriptor shows one interface with two alternates. Alternate 0 uses the bulk-only protocol, with one IN and one OUT pipe. Alternate 1 uses the UAS protocol, with two IN and two OUT pipes. I use identical code in my driver to send custom CDBs. I can see using IORegistryExplorer that in USB 2 mode, macOS chooses alternate 0, the bulk-only protocol. My custom CDBs and their accompanying data pay loads are put on the bus, more or less as expected. In USB 3 mode, macOS chooses alternate 1, the UAS protocol. My custom CDB is put on the bus, but no payload data is transferred. Is this expected behavior? If so, is there a way to force the OS to choose alternate 0 even when on USB 3, perhaps with another dext? I'll file a bug about this when Feedback Assistant lets me.
Replies
8
Boosts
0
Views
933
Activity
Oct ’25
DriverKit. Data Integrity issue.
Dear Apple engineers and DriverKit developers, We have developed a DriverKit (DEXT) driver for an HBA RAID controller. The RAID controller is connected to hosts through Thunderbolt (PCIe port of the Thunderbolt controller). We use an IO script to verify the developed driver. The test fails after a few (10-12) hours of running with an error: “BUG IN CLIENT OF LIBDISPATCH: dispatch_sync called on queue already owned by current thread”. We inspected the stack trace of the crash report. This error happens in the interrupt handler. Thread 5 Crashed: 0 libdispatch.dylib 0x19671aa8c __DISPATCH_WAIT_FOR_QUEUE__ + 484 1 libdispatch.dylib 0x19671a5d0 _dispatch_sync_f_slow + 152 2 DriverKit 0x195d3fc1c IODispatchQueue::DispatchSync_f(void*, void (*)(void*)) + 296 3 DriverKit 0x195d40860 IOInterruptDispatchSourceThread(void*) + 380 4 libsystem_pthread.dylib 0x1968a3738 _pthread_start + 140 5 libsystem_pthread.dylib 0x1968ac6c8 thread_start + 8 On our side we created 5 DispatchQueue(s) for the interrupt processing and configured 5 interrupt handlers using ConfigureInterrupts(kIOInterruptTypePCIMessagedX, 1, 5). It gives 5 interrupts, as requested and it is not clear what is the reason for the issue. Our code samples are below uint32_t configureInterrupts(uint32_t requested, uint32_t required)  {      const kern_return_t ret = ioPCIdevice->ConfigureInterrupts(kIOInterruptTypePCIMessagedX, required, requested);      if (ret != kIOReturnSuccess)          return 0;      uint64_t interruptType = 0;      uint32_t interruptsCount = 0;      uint32_t interruptIndex = 0;        for ( ;; ++interruptIndex) {          if (IOInterruptDispatchSource::GetInterruptType(ioPCIdevice, interruptIndex, &interruptType) != kIOReturnSuccess)              break;          if ((interruptType & kIOInterruptTypePCIMessagedX) == 0)              continue;          ++interruptsCount;      }      return interruptsCount;  }  ..... // Create DQs     for(int i = 0; i < maxInterrupts; ++i) {          ret = IODispatchQueue::Create(INTERRUPT_DQ_NAME, 0, 0, &ivars->interruptQueue[i]);          if (kIOReturnSuccess != ret || nullptr == ivars->interruptQueue[i]) {              GH_PRINT_ERR("Interrupts queue %d creation failed with error %d", i, ret);              return false;          }      } ..... // Link DQ with interrupt   for(int index = 0; index < maxInterrupts; ++ index) {      kern_return_t ret = CreateActionHandleInterruptRequest(size, &ivars->interruptActions[index]);      if (kIOReturnSuccess != ret) {          GH_PRINT_ERR("Create action for interrupt handler %u failed.", index);          return ;      }      ret = IOInterruptDispatchSource::Create(ivars->PCI_io.dev, index, ivars->interruptQueue[index], &ivars->interruptSources[index]);      if (kIOReturnSuccess != ret || nullptr == ivars->interruptSources[index]) {          GH_PRINT_ERR("Creating interrupt source %u failed for interrupt index %u.", index, index);          return ;      }      ret = ivars->interruptSources[index]->SetHandler(ivars->interruptActions[index]);      if (kIOReturnSuccess != ret) {          GH_PRINT_ERR("Setting the handler for interrupt source %u failed.", index);          return ;      }      ivars->contexts[index] = ivars->interruptActions[index]->GetReference();      }  ..... // definition for interrupt handler void    HandleInterruptRequest ( OSAction *action, uint64_t count, uint64_t time) TYPE (IOInterruptDispatchSource::InterruptOccurred);  Do you have any clue how we can fix this error? Or directions and ways for investigation? Please let us know if you need more details.
Replies
7
Boosts
0
Views
483
Activity
Oct ’25
DriverKit. Plug/unplug test leads to MacOS panic
Dear Apple engineers, We have developed a DriverKit (DEXT) driver for an HBA RAID controller. The RAID controller is connected to hosts through Thunderbolt (PCIe port of the Thunderbolt controller). We do plug/unplug tests to verify the developed driver. The test always fails in about 100 cycles with a MacOS crash (panic). The panic contains “LLC Bus error (Unavailable) from cpu0: FAR=0xa40100008 LLC_ERR_STS/ADR/INF=0x80/0x300480a40100008/0x1400000005 addr=0xa40100008 cmd=0x18(ACC_CIFL2C_CMD_RD_LD: request for load miss in E or S state)” At first we assumed that the issue is with hardware. But we did this test on different hosts (MacMini M3 and M4) with different units of our device. The error points to the same physical address FAR=0xa40100008 even if the hosts are different. The 2 full panic logs are attached (one for M4, another one for M3 host). Could you share your understanding of the crash and give any hints on how we can fix it? Please let us know if you need any additional data. Thank you M3 panic: https://drive.google.com/file/d/1GJXd3tTW6ajdrHpFsJxO_tWWYKYIgcMc/view?usp=share_link M4 panic: https://drive.google.com/file/d/1SU-3aBSdhLsyhhxsLknzw9wGvBQ9TbJC/view?usp=share_link
Replies
2
Boosts
0
Views
690
Activity
Oct ’25