1 files changed, 54 insertions, 27 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index 2fc58e658986..668939a14206 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -98,6 +98,9 @@ extern int amdgpu_sched_hw_submission;
 #define AMDGPU_MAX_COMPUTE_RINGS                8
 #define AMDGPU_MAX_VCE_RINGS                    2
 
+/* max number of IP instances */
+#define AMDGPU_MAX_SDMA_INSTANCES               2
+
 /* number of hw syncs before falling back on blocking */
 #define AMDGPU_NUM_SYNCS                        4
 
@@ -183,6 +186,7 @@ struct amdgpu_vm;
 struct amdgpu_ring;
 struct amdgpu_semaphore;
 struct amdgpu_cs_parser;
+struct amdgpu_job;
 struct amdgpu_irq_src;
 struct amdgpu_fpriv;
 
@@ -246,7 +250,7 @@ struct amdgpu_buffer_funcs {
         unsigned        copy_num_dw;
 
         /* used for buffer migration */
-        void (*emit_copy_buffer)(struct amdgpu_ring *ring,
+        void (*emit_copy_buffer)(struct amdgpu_ib *ib,
                                  /* src addr in bytes */
                                  uint64_t src_offset,
                                  /* dst addr in bytes */
@@ -261,7 +265,7 @@ struct amdgpu_buffer_funcs {
         unsigned        fill_num_dw;
 
         /* used for buffer clearing */
-        void (*emit_fill_buffer)(struct amdgpu_ring *ring,
+        void (*emit_fill_buffer)(struct amdgpu_ib *ib,
                                  /* value to write to memory */
                                  uint32_t src_data,
                                  /* dst addr in bytes */
@@ -339,6 +343,8 @@ struct amdgpu_ring_funcs {
         int (*test_ring)(struct amdgpu_ring *ring);
         int (*test_ib)(struct amdgpu_ring *ring);
         bool (*is_lockup)(struct amdgpu_ring *ring);
+        /* insert NOP packets */
+        void (*insert_nop)(struct amdgpu_ring *ring, uint32_t count);
 };
 
 /*
@@ -440,8 +446,10 @@ int amdgpu_fence_wait_empty(struct amdgpu_ring *ring);
 unsigned amdgpu_fence_count_emitted(struct amdgpu_ring *ring);
 
 signed long amdgpu_fence_wait_any(struct amdgpu_device *adev,
-                          struct amdgpu_fence **fences,
-                          bool intr, long t);
+                                  struct fence **array,
+                                  uint32_t count,
+                                  bool intr,
+                                  signed long t);
 struct amdgpu_fence *amdgpu_fence_ref(struct amdgpu_fence *fence);
 void amdgpu_fence_unref(struct amdgpu_fence **fence);
 
@@ -514,7 +522,7 @@ int amdgpu_copy_buffer(struct amdgpu_ring *ring,
                        uint64_t dst_offset,
                        uint32_t byte_count,
                        struct reservation_object *resv,
-                       struct amdgpu_fence **fence);
+                       struct fence **fence);
 int amdgpu_mmap(struct file *filp, struct vm_area_struct *vma);
 
 struct amdgpu_bo_list_entry {
@@ -650,7 +658,7 @@ struct amdgpu_sa_bo {
         struct amdgpu_sa_manager        *manager;
         unsigned                        soffset;
         unsigned                        eoffset;
-        struct amdgpu_fence             *fence;
+        struct fence                    *fence;
 };
 
 /*
@@ -692,7 +700,7 @@ bool amdgpu_semaphore_emit_wait(struct amdgpu_ring *ring,
                                 struct amdgpu_semaphore *semaphore);
 void amdgpu_semaphore_free(struct amdgpu_device *adev,
                            struct amdgpu_semaphore **semaphore,
-                           struct amdgpu_fence *fence);
+                           struct fence *fence);
 
 /*
  * Synchronization
@@ -700,7 +708,8 @@ void amdgpu_semaphore_free(struct amdgpu_device *adev,
 struct amdgpu_sync {
         struct amdgpu_semaphore *semaphores[AMDGPU_NUM_SYNCS];
         struct amdgpu_fence     *sync_to[AMDGPU_MAX_RINGS];
-        struct amdgpu_fence     *last_vm_update;
+        DECLARE_HASHTABLE(fences, 4);
+        struct fence            *last_vm_update;
 };
 
 void amdgpu_sync_create(struct amdgpu_sync *sync);
@@ -712,8 +721,10 @@ int amdgpu_sync_resv(struct amdgpu_device *adev,
                      void *owner);
 int amdgpu_sync_rings(struct amdgpu_sync *sync,
                       struct amdgpu_ring *ring);
+struct fence *amdgpu_sync_get_fence(struct amdgpu_sync *sync);
+int amdgpu_sync_wait(struct amdgpu_sync *sync);
 void amdgpu_sync_free(struct amdgpu_device *adev, struct amdgpu_sync *sync,
-                      struct amdgpu_fence *fence);
+                      struct fence *fence);
 
 /*
  * GART structures, functions & helpers
@@ -871,7 +882,7 @@ int amdgpu_sched_ib_submit_kernel_helper(struct amdgpu_device *adev,
                                          struct amdgpu_ring *ring,
                                          struct amdgpu_ib *ibs,
                                          unsigned num_ibs,
-                                         int (*free_job)(struct amdgpu_cs_parser *),
+                                         int (*free_job)(struct amdgpu_job *),
                                          void *owner,
                                          struct fence **fence);
 
@@ -957,7 +968,7 @@ struct amdgpu_vm_id {
         unsigned                id;
         uint64_t                pd_gpu_addr;
         /* last flushed PD/PT update */
-        struct amdgpu_fence     *flushed_updates;
+        struct fence            *flushed_updates;
         /* last use of vmid */
         struct amdgpu_fence     *last_id_use;
 };
@@ -1042,7 +1053,7 @@ struct amdgpu_ctx *amdgpu_ctx_get(struct amdgpu_fpriv *fpriv, uint32_t id);
 int amdgpu_ctx_put(struct amdgpu_ctx *ctx);
 
 uint64_t amdgpu_ctx_add_fence(struct amdgpu_ctx *ctx, struct amdgpu_ring *ring,
-                              struct fence *fence, uint64_t queued_seq);
+                              struct fence *fence);
 struct fence *amdgpu_ctx_get_fence(struct amdgpu_ctx *ctx,
                                    struct amdgpu_ring *ring, uint64_t seq);
 
@@ -1078,8 +1089,6 @@ struct amdgpu_bo_list {
 };
 
 struct amdgpu_bo_list *
-amdgpu_bo_list_clone(struct amdgpu_bo_list *list);
-struct amdgpu_bo_list *
 amdgpu_bo_list_get(struct amdgpu_fpriv *fpriv, int id);
 void amdgpu_bo_list_put(struct amdgpu_bo_list *list);
 void amdgpu_bo_list_free(struct amdgpu_bo_list *list);
@@ -1210,6 +1219,7 @@ int amdgpu_ib_ring_tests(struct amdgpu_device *adev);
 void amdgpu_ring_free_size(struct amdgpu_ring *ring);
 int amdgpu_ring_alloc(struct amdgpu_ring *ring, unsigned ndw);
 int amdgpu_ring_lock(struct amdgpu_ring *ring, unsigned ndw);
+void amdgpu_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count);
 void amdgpu_ring_commit(struct amdgpu_ring *ring);
 void amdgpu_ring_unlock_commit(struct amdgpu_ring *ring);
 void amdgpu_ring_undo(struct amdgpu_ring *ring);
@@ -1255,14 +1265,16 @@ struct amdgpu_cs_parser {
 
         /* user fence */
         struct amdgpu_user_fence uf;
+};
 
-        struct amdgpu_ring *ring;
-        struct mutex job_lock;
-        struct work_struct job_work;
-        int (*prepare_job)(struct amdgpu_cs_parser *sched_job);
-        int (*run_job)(struct amdgpu_cs_parser *sched_job);
-        int (*free_job)(struct amdgpu_cs_parser *sched_job);
-        struct amd_sched_fence *s_fence;
+struct amdgpu_job {
+        struct amd_sched_job    base;
+        struct amdgpu_device    *adev;
+        struct amdgpu_ib        *ibs;
+        uint32_t                num_ibs;
+        struct mutex            job_lock;
+        struct amdgpu_user_fence uf;
+        int (*free_job)(struct amdgpu_job *sched_job);
 };
 
 static inline u32 amdgpu_get_ib_value(struct amdgpu_cs_parser *p, uint32_t ib_idx, int idx)
@@ -1659,7 +1671,6 @@ struct amdgpu_uvd {
         struct amdgpu_bo        *vcpu_bo;
         void                    *cpu_addr;
         uint64_t                gpu_addr;
-        void                    *saved_bo;
         atomic_t                handles[AMDGPU_MAX_UVD_HANDLES];
         struct drm_file         *filp[AMDGPU_MAX_UVD_HANDLES];
         struct delayed_work     idle_work;
@@ -1703,6 +1714,7 @@ struct amdgpu_sdma {
         uint32_t                feature_version;
 
         struct amdgpu_ring      ring;
+        bool                    burst_nop;
 };
 
 /*
@@ -2051,7 +2063,7 @@ struct amdgpu_device {
         struct amdgpu_gfx               gfx;
 
         /* sdma */
-        struct amdgpu_sdma              sdma[2];
+        struct amdgpu_sdma              sdma[AMDGPU_MAX_SDMA_INSTANCES];
         struct amdgpu_irq_src           sdma_trap_irq;
         struct amdgpu_irq_src           sdma_illegal_inst_irq;
 
@@ -2190,6 +2202,21 @@ static inline void amdgpu_ring_write(struct amdgpu_ring *ring, uint32_t v)
         ring->ring_free_dw--;
 }
 
+static inline struct amdgpu_sdma * amdgpu_get_sdma_instance(struct amdgpu_ring *ring)
+{
+        struct amdgpu_device *adev = ring->adev;
+        int i;
+
+        for (i = 0; i < AMDGPU_MAX_SDMA_INSTANCES; i++)
+                if (&adev->sdma[i].ring == ring)
+                        break;
+
+        if (i < AMDGPU_MAX_SDMA_INSTANCES)
+                return &adev->sdma[i];
+        else
+                return NULL;
+}
+
 /*
  * ASICs macro.
  */
@@ -2241,8 +2268,8 @@ static inline void amdgpu_ring_write(struct amdgpu_ring *ring, uint32_t v)
 #define amdgpu_display_add_connector(adev, ci, sd, ct, ib, coi, h, r) (adev)->mode_info.funcs->add_connector((adev), (ci), (sd), (ct), (ib), (coi), (h), (r))
 #define amdgpu_display_stop_mc_access(adev, s) (adev)->mode_info.funcs->stop_mc_access((adev), (s))
 #define amdgpu_display_resume_mc_access(adev, s) (adev)->mode_info.funcs->resume_mc_access((adev), (s))
-#define amdgpu_emit_copy_buffer(adev, r, s, d, b) (adev)->mman.buffer_funcs->emit_copy_buffer((r), (s), (d), (b))
-#define amdgpu_emit_fill_buffer(adev, r, s, d, b) (adev)->mman.buffer_funcs->emit_fill_buffer((r), (s), (d), (b))
+#define amdgpu_emit_copy_buffer(adev, ib, s, d, b) (adev)->mman.buffer_funcs->emit_copy_buffer((ib),  (s), (d), (b))
+#define amdgpu_emit_fill_buffer(adev, ib, s, d, b) (adev)->mman.buffer_funcs->emit_fill_buffer((ib), (s), (d), (b))
 #define amdgpu_dpm_get_temperature(adev) (adev)->pm.funcs->get_temperature((adev))
 #define amdgpu_dpm_pre_set_power_state(adev) (adev)->pm.funcs->pre_set_power_state((adev))
 #define amdgpu_dpm_set_power_state(adev) (adev)->pm.funcs->set_power_state((adev))
@@ -2343,7 +2370,7 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring,
                       struct amdgpu_sync *sync);
 void amdgpu_vm_flush(struct amdgpu_ring *ring,
                      struct amdgpu_vm *vm,
-                     struct amdgpu_fence *updates);
+                     struct fence *updates);
 void amdgpu_vm_fence(struct amdgpu_device *adev,
                      struct amdgpu_vm *vm,
                      struct amdgpu_fence *fence);
@@ -2373,7 +2400,7 @@ int amdgpu_vm_bo_unmap(struct amdgpu_device *adev,
                        uint64_t addr);
 void amdgpu_vm_bo_rmv(struct amdgpu_device *adev,
                       struct amdgpu_bo_va *bo_va);
-
+int amdgpu_vm_free_job(struct amdgpu_job *job);
 /*
  * functions used by amdgpu_encoder.c
  */