From d69bb47d21f50bdef6023347b1838dc1f8a5c525 Mon Sep 17 00:00:00 2001 From: Elena Agostini Date: Thu, 27 Jan 2022 03:50:28 +0000 Subject: [PATCH] gpudev: expose GPU memory to CPU Enable the possibility to expose a GPU memory area and make it accessible 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 return a memory pointer usable by the CPU to access the GPU memory area. Signed-off-by: Elena Agostini --- app/test-gpudev/main.c | 63 ++++++++++++++++++++++++++++++++ doc/guides/prog_guide/gpudev.rst | 9 +++++ drivers/gpu/cuda/cuda.c | 2 + lib/gpudev/gpudev.c | 61 +++++++++++++++++++++++++++++++ lib/gpudev/gpudev_driver.h | 6 +++ lib/gpudev/rte_gpudev.h | 49 +++++++++++++++++++++++++ lib/gpudev/version.map | 2 + 7 files changed, 192 insertions(+) diff --git a/app/test-gpudev/main.c b/app/test-gpudev/main.c index 4500a8660b..417f2d78b7 100644 --- a/app/test-gpudev/main.c +++ b/app/test-gpudev/main.c @@ -185,6 +185,68 @@ error: return -1; } +static int +gpu_mem_cpu_map(uint16_t gpu_id) +{ + void *ptr_gpu = NULL; + void *ptr_cpu = NULL; + size_t buf_bytes = 1024; + unsigned int align = 4096; + int ret; + + printf("\n=======> TEST: Map GPU memory for CPU visibility\n\n"); + + /* Alloc memory on GPU 0 with 4kB alignment */ + ptr_gpu = rte_gpu_mem_alloc(gpu_id, buf_bytes, align); + if (ptr_gpu == NULL) { + fprintf(stderr, "rte_gpu_mem_alloc GPU memory returned error\n"); + goto error; + } + printf("GPU memory allocated at 0x%p size is %zd bytes\n", + ptr_gpu, buf_bytes); + + ptr_cpu = rte_gpu_mem_cpu_map(gpu_id, buf_bytes, ptr_gpu); + if (ptr_cpu == NULL) { + fprintf(stderr, "rte_gpu_mem_cpu_map returned error\n"); + goto error; + } + printf("GPU memory mapped for CPU access at 0x%p\n", ptr_cpu); + + ((uint8_t *)ptr_cpu)[0] = 0x4; + ((uint8_t *)ptr_cpu)[1] = 0x5; + ((uint8_t *)ptr_cpu)[2] = 0x6; + + printf("GPU memory first 3 bytes set from CPU: %x %x %x\n", + ((uint8_t *)ptr_cpu)[0], + ((uint8_t *)ptr_cpu)[1], + ((uint8_t *)ptr_cpu)[2]); + + ret = rte_gpu_mem_cpu_unmap(gpu_id, ptr_cpu); + if (ret < 0) { + fprintf(stderr, "rte_gpu_mem_cpu_unmap returned error %d\n", ret); + goto error; + } + printf("GPU memory mapped for CPU access at 0x%p\n", ptr_cpu); + + ret = rte_gpu_mem_free(gpu_id, ptr_gpu); + if (ret < 0) { + fprintf(stderr, "rte_gpu_mem_free returned error %d\n", ret); + goto error; + } + printf("GPU memory 0x%p freed\n", ptr_gpu); + + printf("\n=======> TEST: PASSED\n"); + return 0; + +error: + + rte_gpu_mem_cpu_unmap(gpu_id, ptr_cpu); + rte_gpu_mem_free(gpu_id, ptr_gpu); + + printf("\n=======> TEST: FAILED\n"); + return -1; +} + static int create_update_comm_flag(uint16_t gpu_id) { @@ -402,6 +464,7 @@ main(int argc, char **argv) */ alloc_gpu_memory(gpu_id); register_cpu_memory(gpu_id); + gpu_mem_cpu_map(gpu_id); /** * Communication items test diff --git a/doc/guides/prog_guide/gpudev.rst b/doc/guides/prog_guide/gpudev.rst index ff4626812b..6223207a33 100644 --- a/doc/guides/prog_guide/gpudev.rst +++ b/doc/guides/prog_guide/gpudev.rst @@ -73,6 +73,15 @@ Later, it's also possible to unregister that memory with gpudev. CPU memory registered outside of the gpudev library (e.g. with GPU specific library) cannot be unregistered by the gpudev library. +CPU mapping +~~~~~~~~~~~ + +gpudev can map into the CPU address space a GPU memory address allocated with gpudev. +gpudev returns a pointer the CPU can use to access (ready or write) GPU memory. +Later, it's also possible to unmap that memory with gpudev. +GPU memory CPU mapped outside of the gpudev library (e.g. with GPU specific library) +cannot be unmapped by the gpudev library. + Memory Barrier ~~~~~~~~~~~~~~ diff --git a/drivers/gpu/cuda/cuda.c b/drivers/gpu/cuda/cuda.c index 0ece1bb612..408b659fce 100644 --- a/drivers/gpu/cuda/cuda.c +++ b/drivers/gpu/cuda/cuda.c @@ -1177,6 +1177,8 @@ cuda_gpu_probe(__rte_unused struct rte_pci_driver *pci_drv, struct rte_pci_devic dev->ops.mem_free = cuda_mem_free; dev->ops.mem_register = cuda_mem_register; dev->ops.mem_unregister = cuda_mem_unregister; + dev->ops.mem_cpu_map = NULL; + dev->ops.mem_cpu_unmap = NULL; dev->ops.wmb = cuda_wmb; rte_gpu_complete_new(dev); diff --git a/lib/gpudev/gpudev.c b/lib/gpudev/gpudev.c index 59e2169292..ce92d63257 100644 --- a/lib/gpudev/gpudev.c +++ b/lib/gpudev/gpudev.c @@ -640,6 +640,67 @@ rte_gpu_mem_unregister(int16_t dev_id, void *ptr) return GPU_DRV_RET(dev->ops.mem_unregister(dev, ptr)); } +void * +rte_gpu_mem_cpu_map(int16_t dev_id, size_t size, void *ptr) +{ + struct rte_gpu *dev; + void *ptr_out; + int ret; + + dev = gpu_get_by_id(dev_id); + if (dev == NULL) { + GPU_LOG(ERR, "mem CPU map for invalid device ID %d", dev_id); + rte_errno = ENODEV; + return NULL; + } + + if (dev->ops.mem_cpu_map == NULL) { + GPU_LOG(ERR, "mem CPU map not supported"); + rte_errno = ENOTSUP; + return NULL; + } + + if (ptr == NULL || size == 0) /* dry-run */ + return NULL; + + ret = GPU_DRV_RET(dev->ops.mem_cpu_map(dev, size, ptr, &ptr_out)); + + switch (ret) { + case 0: + return ptr_out; + case -ENOMEM: + case -E2BIG: + rte_errno = -ret; + return NULL; + default: + rte_errno = -EPERM; + return NULL; + } +} + +int +rte_gpu_mem_cpu_unmap(int16_t dev_id, void *ptr) +{ + struct rte_gpu *dev; + + dev = gpu_get_by_id(dev_id); + if (dev == NULL) { + GPU_LOG(ERR, "cpu_unmap mem for invalid device ID %d", dev_id); + rte_errno = ENODEV; + return -rte_errno; + } + + if (dev->ops.mem_cpu_unmap == NULL) { + rte_errno = ENOTSUP; + return -rte_errno; + } + + if (ptr == NULL) /* dry-run */ + return 0; + + return GPU_DRV_RET(dev->ops.mem_cpu_unmap(dev, ptr)); +} + int rte_gpu_wmb(int16_t dev_id) { diff --git a/lib/gpudev/gpudev_driver.h b/lib/gpudev/gpudev_driver.h index 0ed7478e9b..0e55b00bfe 100644 --- a/lib/gpudev/gpudev_driver.h +++ b/lib/gpudev/gpudev_driver.h @@ -31,6 +31,8 @@ typedef int (rte_gpu_mem_alloc_t)(struct rte_gpu *dev, size_t size, unsigned int typedef int (rte_gpu_mem_free_t)(struct rte_gpu *dev, void *ptr); typedef int (rte_gpu_mem_register_t)(struct rte_gpu *dev, size_t size, void *ptr); typedef int (rte_gpu_mem_unregister_t)(struct rte_gpu *dev, void *ptr); +typedef int (rte_gpu_mem_cpu_map_t)(struct rte_gpu *dev, size_t size, void *ptr_in, void **ptr_out); +typedef int (rte_gpu_mem_cpu_unmap_t)(struct rte_gpu *dev, void *ptr); typedef int (rte_gpu_wmb_t)(struct rte_gpu *dev); struct rte_gpu_ops { @@ -46,6 +48,10 @@ struct rte_gpu_ops { rte_gpu_mem_register_t *mem_register; /* Unregister CPU memory from device. */ rte_gpu_mem_unregister_t *mem_unregister; + /* Map GPU memory for CPU visibility. */ + rte_gpu_mem_cpu_map_t *mem_cpu_map; + /* Unmap GPU memory for CPU visibility. */ + rte_gpu_mem_cpu_unmap_t *mem_cpu_unmap; /* Enforce GPU write memory barrier. */ rte_gpu_wmb_t *wmb; }; diff --git a/lib/gpudev/rte_gpudev.h b/lib/gpudev/rte_gpudev.h index ff3ca78c89..5cc4eb5828 100644 --- a/lib/gpudev/rte_gpudev.h +++ b/lib/gpudev/rte_gpudev.h @@ -452,6 +452,55 @@ int rte_gpu_mem_register(int16_t dev_id, size_t size, void *ptr); __rte_experimental int rte_gpu_mem_unregister(int16_t dev_id, void *ptr); +/** + * @warning + * @b EXPERIMENTAL: this API may change without prior notice. + * + * Map a chunk of GPU memory to make it accessible from the CPU + * using the memory pointer returned by the function. + * GPU memory has to be allocated via rte_gpu_mem_alloc(). + * + * @param dev_id + * Device ID requiring mapped memory. + * @param size + * Number of bytes to map. + * Requesting 0 will do nothing. + * @param ptr + * Pointer to the GPU memory area to be mapped. + * NULL is a no-op accepted value. + + * @return + * A pointer to the mapped GPU memory usable by the CPU, otherwise NULL and rte_errno is set: + * - ENODEV if invalid dev_id + * - ENOTSUP if operation not supported by the driver + * - E2BIG if size is higher than limit + * - ENOMEM if out of space + * - EPERM if driver error + */ +__rte_experimental +void *rte_gpu_mem_cpu_map(int16_t dev_id, size_t size, void *ptr); + +/** + * @warning + * @b EXPERIMENTAL: this API may change without prior notice. + * + * Unmap a chunk of GPU memory previously mapped with rte_gpu_mem_cpu_map() + * + * @param dev_id + * Reference device ID. + * @param ptr + * Pointer to the memory area to be unmapped. + * NULL is a no-op accepted value. + * + * @return + * 0 on success, -rte_errno otherwise: + * - ENODEV if invalid dev_id + * - ENOTSUP if operation not supported by the driver + * - EPERM if driver error + */ +__rte_experimental +int rte_gpu_mem_cpu_unmap(int16_t dev_id, void *ptr); + /** * @warning * @b EXPERIMENTAL: this API may change without prior notice. diff --git a/lib/gpudev/version.map b/lib/gpudev/version.map index 2e414c65cc..5bc5d154cd 100644 --- a/lib/gpudev/version.map +++ b/lib/gpudev/version.map @@ -20,8 +20,10 @@ EXPERIMENTAL { rte_gpu_init; rte_gpu_is_valid; rte_gpu_mem_alloc; + rte_gpu_mem_cpu_map; rte_gpu_mem_free; rte_gpu_mem_register; + rte_gpu_mem_cpu_unmap; rte_gpu_mem_unregister; rte_gpu_wmb; }; -- 2.20.1