intel_upload_blit_large

intel_upload_blit_large [options]

--help
    Shows the help message and exits.

--size
    Specifies the size (in bytes) of the data to be uploaded and blitted.

--iterations
    Specifies the number of iterations to perform the upload and blit operation.

--dst_x
    Destination X coordinate for the blit.

--dst_y
    Destination Y coordinate for the blit.

--width
    Width of the blit region.

--height
    Height of the blit region.

The `intel_upload_blit_large` command is a utility within the Intel graphics driver ecosystem, specifically used for testing and debugging purposes. It focuses on efficiently transferring and blitting (block transfer) large amounts of data to the GPU's memory. It primarily serves to evaluate the performance and correctness of the underlying memory management and blitting operations within the Intel graphics driver stack, revealing potential bottlenecks or issues in the data upload and rendering pipelines.

This is a tool typically used by developers and testers working on the Intel graphics driver, and is not typically found in general user workflows. It facilitates the creation of specific scenarios that stress the memory subsystem and blitting capabilities of the GPU, allowing for targeted analysis and optimization.

By manipulating parameters like transfer size, transfer alignment, memory configuration, and blitting regions, developers can pinpoint areas where performance can be improved or bugs can be identified.

This command is primarily for internal testing and debugging purposes. It is not intended for general user use and might require specific driver versions or kernel configurations. The specific options available may vary depending on the Intel graphics driver version.

The main purpose is to benchmark and stress test data upload and blitting operations within the Intel GPU driver, not to perform general graphics tasks.

Detailed error reporting may not be available or user-friendly, as the command is designed for developer-level debugging.

intel_gpu_top(1), drm(7)