1 /* 2 * Copyright 2021 Advanced Micro Devices, Inc. 3 * 4 * Permission is hereby granted, free of charge, to any person obtaining a 5 * copy of this software and associated documentation files (the "Software"), 6 * to deal in the Software without restriction, including without limitation 7 * the rights to use, copy, modify, merge, publish, distribute, sublicense, 8 * and/or sell copies of the Software, and to permit persons to whom the 9 * Software is furnished to do so, subject to the following conditions: 10 * 11 * The above copyright notice and this permission notice shall be included in 12 * all copies or substantial portions of the Software. 13 * 14 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 15 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 16 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL 17 * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR 18 * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, 19 * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR 20 * OTHER DEALINGS IN THE SOFTWARE. 21 * 22 */ 23 24 #ifndef __AMDGPU_RESET_H__ 25 #define __AMDGPU_RESET_H__ 26 27 #include "amdgpu.h" 28 29 enum AMDGPU_RESET_FLAGS { 30 31 AMDGPU_NEED_FULL_RESET = 0, 32 AMDGPU_SKIP_HW_RESET = 1, 33 }; 34 35 struct amdgpu_reset_context { 36 enum amd_reset_method method; 37 struct amdgpu_device *reset_req_dev; 38 struct amdgpu_job *job; 39 struct amdgpu_hive_info *hive; 40 struct list_head *reset_device_list; 41 unsigned long flags; 42 }; 43 44 struct amdgpu_reset_handler { 45 enum amd_reset_method reset_method; 46 struct list_head handler_list; 47 int (*prepare_env)(struct amdgpu_reset_control *reset_ctl, 48 struct amdgpu_reset_context *context); 49 int (*prepare_hwcontext)(struct amdgpu_reset_control *reset_ctl, 50 struct amdgpu_reset_context *context); 51 int (*perform_reset)(struct amdgpu_reset_control *reset_ctl, 52 struct amdgpu_reset_context *context); 53 int (*restore_hwcontext)(struct amdgpu_reset_control *reset_ctl, 54 struct amdgpu_reset_context *context); 55 int (*restore_env)(struct amdgpu_reset_control *reset_ctl, 56 struct amdgpu_reset_context *context); 57 58 int (*do_reset)(struct amdgpu_device *adev); 59 }; 60 61 struct amdgpu_reset_control { 62 void *handle; 63 struct work_struct reset_work; 64 struct mutex reset_lock; 65 struct list_head reset_handlers; 66 atomic_t in_reset; 67 enum amd_reset_method active_reset; 68 struct amdgpu_reset_handler *(*get_reset_handler)( 69 struct amdgpu_reset_control *reset_ctl, 70 struct amdgpu_reset_context *context); 71 void (*async_reset)(struct work_struct *work); 72 }; 73 74 75 enum amdgpu_reset_domain_type { 76 SINGLE_DEVICE, 77 XGMI_HIVE 78 }; 79 80 struct amdgpu_reset_domain { 81 struct kref refcount; 82 struct workqueue_struct *wq; 83 enum amdgpu_reset_domain_type type; 84 struct rw_semaphore sem; 85 atomic_t in_gpu_reset; 86 atomic_t reset_res; 87 }; 88 89 90 int amdgpu_reset_init(struct amdgpu_device *adev); 91 int amdgpu_reset_fini(struct amdgpu_device *adev); 92 93 int amdgpu_reset_prepare_hwcontext(struct amdgpu_device *adev, 94 struct amdgpu_reset_context *reset_context); 95 96 int amdgpu_reset_perform_reset(struct amdgpu_device *adev, 97 struct amdgpu_reset_context *reset_context); 98 99 int amdgpu_reset_add_handler(struct amdgpu_reset_control *reset_ctl, 100 struct amdgpu_reset_handler *handler); 101 102 struct amdgpu_reset_domain *amdgpu_reset_create_reset_domain(enum amdgpu_reset_domain_type type, 103 char *wq_name); 104 105 void amdgpu_reset_destroy_reset_domain(struct kref *ref); 106 107 static inline bool amdgpu_reset_get_reset_domain(struct amdgpu_reset_domain *domain) 108 { 109 return kref_get_unless_zero(&domain->refcount) != 0; 110 } 111 112 static inline void amdgpu_reset_put_reset_domain(struct amdgpu_reset_domain *domain) 113 { 114 kref_put(&domain->refcount, amdgpu_reset_destroy_reset_domain); 115 } 116 117 static inline bool amdgpu_reset_domain_schedule(struct amdgpu_reset_domain *domain, 118 struct work_struct *work) 119 { 120 return queue_work(domain->wq, work); 121 } 122 123 void amdgpu_device_lock_reset_domain(struct amdgpu_reset_domain *reset_domain); 124 125 void amdgpu_device_unlock_reset_domain(struct amdgpu_reset_domain *reset_domain); 126 127 #endif 128