intel/compute-runtime
1
0
Fork 0
mirror of https://github.com/intel/compute-runtime.git synced 2025-12-30 01:35:20 +08:00
Code Issues Packages Projects Releases Wiki Activity

feature: add L0 API to wait for completion of all submissions to given device

Browse Source 
Related-To: NEO-14560
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
This commit is contained in:
Mateusz Jablonski
2025-05-07 13:20:39 +00:00
committed by Compute-Runtime-Automation
parent ba85f7417d
commit 1b42ebf7fb
6 changed files with 136 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
level_zero/include/level_zero/ze_intel_gpu.h
View File

@@ -316,6 +316,22 @@ uint32_t ZE_APICALL zerDeviceTranslateToIdentifier(ze_device_handle_t hDevice);
/// - device handle associated with the identifier
ze_device_handle_t ZE_APICALL zerIdentifierTranslateToDeviceHandle(uint32_t identifier); ///< [in] integer identifier of the device
/// @brief Global device synchronization
///
/// @details
/// - The application may call this function from simultaneous threads.
/// - The implementation of this function should be lock-free.
/// - Ensures that everything that was submitted to the device is completed.
/// - Ensures that all submissions in all queues on device are completed.
/// - It is not allowed to call this function while some command list are in graph capture mode.
/// - Returns error if error is detected during execution on device.
/// - Hangs indefinitely if GPU execution is blocked on non signaled event.
///
/// @returns
/// - ::ZE_RESULT_SUCCESS
/// - ::ZE_RESULT_ERROR_DEVICE_LOST
ze_result_t ZE_APICALL zeDeviceSynchronize(ze_device_handle_t hDevice); ///> [in] handle of the device
#if defined(__cplusplus)
} // extern "C"
#endif