diff options
Diffstat (limited to 'drivers/gpu/drm/amd/amdkfd/kfd_device.c')
| -rw-r--r-- | drivers/gpu/drm/amd/amdkfd/kfd_device.c | 68 |
1 files changed, 46 insertions, 22 deletions
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c index 62aa6c9d5123d..f1a225a20719b 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c @@ -53,6 +53,7 @@ extern const struct kfd2kgd_calls arcturus_kfd2kgd; extern const struct kfd2kgd_calls aldebaran_kfd2kgd; extern const struct kfd2kgd_calls gfx_v10_kfd2kgd; extern const struct kfd2kgd_calls gfx_v10_3_kfd2kgd; +extern const struct kfd2kgd_calls gfx_v11_kfd2kgd; static int kfd_gtt_sa_init(struct kfd_dev *kfd, unsigned int buf_size, unsigned int chunk_size); @@ -60,7 +61,7 @@ static void kfd_gtt_sa_fini(struct kfd_dev *kfd); static int kfd_resume(struct kfd_dev *kfd); -static void kfd_device_info_set_sdma_queue_num(struct kfd_dev *kfd) +static void kfd_device_info_set_sdma_info(struct kfd_dev *kfd) { uint32_t sdma_version = kfd->adev->ip_versions[SDMA0_HWIP][0]; @@ -72,6 +73,7 @@ static void kfd_device_info_set_sdma_queue_num(struct kfd_dev *kfd) case IP_VERSION(4, 1, 2):/* RENOIR */ case IP_VERSION(5, 2, 1):/* VANGOGH */ case IP_VERSION(5, 2, 3):/* YELLOW_CARP */ + case IP_VERSION(6, 0, 1): kfd->device_info.num_sdma_queues_per_engine = 2; break; case IP_VERSION(4, 2, 0):/* VEGA20 */ @@ -85,6 +87,8 @@ static void kfd_device_info_set_sdma_queue_num(struct kfd_dev *kfd) case IP_VERSION(5, 2, 2):/* NAVY_FLOUNDER */ case IP_VERSION(5, 2, 4):/* DIMGREY_CAVEFISH */ case IP_VERSION(5, 2, 5):/* BEIGE_GOBY */ + case IP_VERSION(6, 0, 0): + case IP_VERSION(6, 0, 2): kfd->device_info.num_sdma_queues_per_engine = 8; break; default: @@ -93,6 +97,19 @@ static void kfd_device_info_set_sdma_queue_num(struct kfd_dev *kfd) sdma_version); kfd->device_info.num_sdma_queues_per_engine = 8; } + + switch (sdma_version) { + case IP_VERSION(6, 0, 0): + case IP_VERSION(6, 0, 1): + case IP_VERSION(6, 0, 2): + /* Reserve 1 for paging and 1 for gfx */ + kfd->device_info.num_reserved_sdma_queues_per_engine = 2; + /* BIT(0)=engine-0 queue-0; BIT(1)=engine-1 queue-0; BIT(2)=engine-0 queue-1; ... */ + kfd->device_info.reserved_sdma_queues_bitmap = 0xFULL; + break; + default: + break; + } } static void kfd_device_info_set_event_interrupt_class(struct kfd_dev *kfd) @@ -121,6 +138,11 @@ static void kfd_device_info_set_event_interrupt_class(struct kfd_dev *kfd) case IP_VERSION(10, 3, 5): /* BEIGE_GOBY */ kfd->device_info.event_interrupt_class = &event_interrupt_class_v9; break; + case IP_VERSION(11, 0, 0): + case IP_VERSION(11, 0, 1): + case IP_VERSION(11, 0, 2): + kfd->device_info.event_interrupt_class = &event_interrupt_class_v11; + break; default: dev_warn(kfd_device, "v9 event interrupt handler is set due to " "mismatch of gc ip block(GC_HWIP:0x%x).\n", gc_version); @@ -145,7 +167,7 @@ static void kfd_device_info_init(struct kfd_dev *kfd, kfd->device_info.ih_ring_entry_size = 8 * sizeof(uint32_t); kfd->device_info.supports_cwsr = true; - kfd_device_info_set_sdma_queue_num(kfd); + kfd_device_info_set_sdma_info(kfd); kfd_device_info_set_event_interrupt_class(kfd); @@ -346,6 +368,18 @@ struct kfd_dev *kgd2kfd_probe(struct amdgpu_device *adev, bool vf) if (!vf) f2g = &gfx_v10_3_kfd2kgd; break; + case IP_VERSION(11, 0, 0): + gfx_target_version = 110000; + f2g = &gfx_v11_kfd2kgd; + break; + case IP_VERSION(11, 0, 1): + gfx_target_version = 110003; + f2g = &gfx_v11_kfd2kgd; + break; + case IP_VERSION(11, 0, 2): + gfx_target_version = 110002; + f2g = &gfx_v11_kfd2kgd; + break; default: break; } @@ -575,6 +609,8 @@ bool kgd2kfd_device_init(struct kfd_dev *kfd, if (kfd_resume(kfd)) goto kfd_resume_error; + amdgpu_amdkfd_get_local_mem_info(kfd->adev, &kfd->local_mem_info); + if (kfd_topology_add_device(kfd)) { dev_err(kfd_device, "Error adding device to topology\n"); goto kfd_topology_add_device_error; @@ -873,8 +909,6 @@ out: static int kfd_gtt_sa_init(struct kfd_dev *kfd, unsigned int buf_size, unsigned int chunk_size) { - unsigned int num_of_longs; - if (WARN_ON(buf_size < chunk_size)) return -EINVAL; if (WARN_ON(buf_size == 0)) @@ -885,11 +919,8 @@ static int kfd_gtt_sa_init(struct kfd_dev *kfd, unsigned int buf_size, kfd->gtt_sa_chunk_size = chunk_size; kfd->gtt_sa_num_of_chunks = buf_size / chunk_size; - num_of_longs = (kfd->gtt_sa_num_of_chunks + BITS_PER_LONG - 1) / - BITS_PER_LONG; - - kfd->gtt_sa_bitmap = kcalloc(num_of_longs, sizeof(long), GFP_KERNEL); - + kfd->gtt_sa_bitmap = bitmap_zalloc(kfd->gtt_sa_num_of_chunks, + GFP_KERNEL); if (!kfd->gtt_sa_bitmap) return -ENOMEM; @@ -899,13 +930,12 @@ static int kfd_gtt_sa_init(struct kfd_dev *kfd, unsigned int buf_size, mutex_init(&kfd->gtt_sa_lock); return 0; - } static void kfd_gtt_sa_fini(struct kfd_dev *kfd) { mutex_destroy(&kfd->gtt_sa_lock); - kfree(kfd->gtt_sa_bitmap); + bitmap_free(kfd->gtt_sa_bitmap); } static inline uint64_t kfd_gtt_sa_calc_gpu_addr(uint64_t start_addr, @@ -973,7 +1003,7 @@ kfd_gtt_restart_search: /* If we need only one chunk, mark it as allocated and get out */ if (size <= kfd->gtt_sa_chunk_size) { pr_debug("Single bit\n"); - set_bit(found, kfd->gtt_sa_bitmap); + __set_bit(found, kfd->gtt_sa_bitmap); goto kfd_gtt_out; } @@ -1011,10 +1041,8 @@ kfd_gtt_restart_search: (*mem_obj)->range_start, (*mem_obj)->range_end); /* Mark the chunks as allocated */ - for (found = (*mem_obj)->range_start; - found <= (*mem_obj)->range_end; - found++) - set_bit(found, kfd->gtt_sa_bitmap); + bitmap_set(kfd->gtt_sa_bitmap, (*mem_obj)->range_start, + (*mem_obj)->range_end - (*mem_obj)->range_start + 1); kfd_gtt_out: mutex_unlock(&kfd->gtt_sa_lock); @@ -1029,8 +1057,6 @@ kfd_gtt_no_free_chunk: int kfd_gtt_sa_free(struct kfd_dev *kfd, struct kfd_mem_obj *mem_obj) { - unsigned int bit; - /* Act like kfree when trying to free a NULL object */ if (!mem_obj) return 0; @@ -1041,10 +1067,8 @@ int kfd_gtt_sa_free(struct kfd_dev *kfd, struct kfd_mem_obj *mem_obj) mutex_lock(&kfd->gtt_sa_lock); /* Mark the chunks as free */ - for (bit = mem_obj->range_start; - bit <= mem_obj->range_end; - bit++) - clear_bit(bit, kfd->gtt_sa_bitmap); + bitmap_clear(kfd->gtt_sa_bitmap, mem_obj->range_start, + mem_obj->range_end - mem_obj->range_start + 1); mutex_unlock(&kfd->gtt_sa_lock); |