Changes in gpudev_driver.h

d69bb47d - gpudev: expose GPU memory to CPU

Thu, 27 Jan 2022 03:50:28 +0000

gpudev: expose GPU memory to CPUEnable the possibility to expose a GPU memory area and make itaccessible from the CPU.GPU memory has to be allocated via rte_gpu_mem_alloc().This patch allows the gpudev library to map (and unmap),through the GPU driver, a chunk of GPU memory and to returna memory pointer usable by the CPU to access the GPU memory area.Signed-off-by: Elena Agostini List of files: /dpdk/lib/gpudev/gpudev_driver.h

c8557ed4 - gpudev: add alignment for memory allocation

Sat, 08 Jan 2022 00:20:01 +0000

gpudev: add alignment for memory allocationSimilarly to rte_malloc, rte_gpu_mem_alloc accepts asinput the memory alignment size.GPU driver should return GPU memory address alignedwith the input value.Signed-off-by: Elena Agostini List of files: /dpdk/lib/gpudev/gpudev_driver.h

2d61b429 - gpudev: add memory barrier

Mon, 08 Nov 2021 18:58:02 +0000

gpudev: add memory barrierAdd a function for the application to ensure the coherencyof the writes executed by another device into the GPU memory.Signed-off-by: Elena Agostini List of files: /dpdk/lib/gpudev/gpudev_driver.h

e818c4e2 - gpudev: add memory API

Mon, 08 Nov 2021 18:58:01 +0000

gpudev: add memory APIIn heterogeneous computing system, processing is not only in the CPU.Some tasks can be delegated to devices working in parallel.Such workload distribution can be achieved by sharing some memory.As a first step, the features are focused on memory management.A function allows to allocate memory inside the device,or in the main (CPU) memory while making it visible for the device.This memory may be used to save packets or for synchronization data.The next step should focus on GPU processing task control.Signed-off-by: Elena Agostini Signed-off-by: Thomas Monjalon List of files: /dpdk/lib/gpudev/gpudev_driver.h

a9af048a - gpudev: support multi-process

Mon, 08 Nov 2021 18:58:00 +0000

gpudev: support multi-processThe device data shared between processes are moved in a structallocated in a shared memory (a new memzone for all GPUs).The main struct rte_gpu references the shared memoryvia the pointer mpshared.The API function rte_gpu_attach() is added to attach a devicefrom the secondary process.The function rte_gpu_allocate() can be used only by primary process.Signed-off-by: Thomas Monjalon List of files: /dpdk/lib/gpudev/gpudev_driver.h

82e5f6b6 - gpudev: add child device representing a device context

Mon, 08 Nov 2021 18:57:59 +0000

gpudev: add child device representing a device contextThe computing device may operate in some isolated contexts.Memory and processing are isolated in a silo represented bya child device.The context is provided as an opaque by the caller ofrte_gpu_add_child().Signed-off-by: Thomas Monjalon List of files: /dpdk/lib/gpudev/gpudev_driver.h

18cb0756 - gpudev: add event notification

Mon, 08 Nov 2021 18:57:58 +0000

gpudev: add event notificationCallback functions may be registered for a device event.Callback management is per-process and not thread-safe.The events RTE_GPU_EVENT_NEW and RTE_GPU_EVENT_DELare notified respectively after creation and before removalof a device, as part of the library functions.Some future events may be emitted from drivers.Signed-off-by: Thomas Monjalon List of files: /dpdk/lib/gpudev/gpudev_driver.h

8b8036a6 - gpudev: introduce GPU device class library

Mon, 08 Nov 2021 18:57:57 +0000

gpudev: introduce GPU device class libraryIn heterogeneous computing system, processing is not only in the CPU.Some tasks can be delegated to devices working in parallel.The new library gpudev is for dealing with GPGPU computing devicesfrom a DPDK application running on the CPU.The infrastructure is prepared to welcome drivers in drivers/gpu/.Signed-off-by: Elena Agostini Signed-off-by: Thomas Monjalon List of files: /dpdk/lib/gpudev/gpudev_driver.h