/* number of hw syncs before falling back on blocking */
#define RADEON_NUM_SYNCS 4
-/* number of hw syncs before falling back on blocking */
-#define RADEON_NUM_SYNCS 4
-
/* hardcode those limit for now */
#define RADEON_VA_IB_OFFSET (1 << 20)
#define RADEON_VA_RESERVED_SIZE (8 << 20)
};
struct radeon_fence {
- struct fence base;
+ struct fence base;
- struct radeon_device *rdev;
- uint64_t seq;
+ struct radeon_device *rdev;
+ uint64_t seq;
/* RB, DMA, etc. */
- unsigned ring;
+ unsigned ring;
+ bool is_vm_update;
- wait_queue_t fence_wake;
+ wait_queue_t fence_wake;
};
int radeon_fence_driver_start_ring(struct radeon_device *rdev, int ring);
#endif
};
+struct radeon_bo_list {
+ struct radeon_bo *robj;
+ struct ttm_validate_buffer tv;
+ uint64_t gpu_offset;
+ unsigned prefered_domains;
+ unsigned allowed_domains;
+ uint32_t tiling_flags;
+};
+
/* bo virtual address in a specific vm */
struct radeon_bo_va {
/* protected by bo being reserved */
struct list_head bo_list;
uint32_t flags;
uint64_t addr;
+ struct radeon_fence *last_pt_update;
unsigned ref_count;
/* protected by vm mutex */
struct list_head list;
/* Protected by tbo.reserved */
u32 initial_domain;
- struct ttm_place placements[3];
+ struct ttm_place placements[4];
struct ttm_placement placement;
struct ttm_buffer_object tbo;
struct ttm_bo_kmap_obj kmap;
* Semaphores.
*/
struct radeon_semaphore {
- struct radeon_sa_bo *sa_bo;
- signed waiters;
- uint64_t gpu_addr;
- struct radeon_fence *sync_to[RADEON_NUM_RINGS];
+ struct radeon_sa_bo *sa_bo;
+ signed waiters;
+ uint64_t gpu_addr;
};
int radeon_semaphore_create(struct radeon_device *rdev,
struct radeon_semaphore *semaphore);
bool radeon_semaphore_emit_wait(struct radeon_device *rdev, int ring,
struct radeon_semaphore *semaphore);
-void radeon_semaphore_sync_fence(struct radeon_semaphore *semaphore,
- struct radeon_fence *fence);
-int radeon_semaphore_sync_resv(struct radeon_device *rdev,
- struct radeon_semaphore *semaphore,
- struct reservation_object *resv,
- bool shared);
-int radeon_semaphore_sync_rings(struct radeon_device *rdev,
- struct radeon_semaphore *semaphore,
- int waiting_ring);
void radeon_semaphore_free(struct radeon_device *rdev,
struct radeon_semaphore **semaphore,
struct radeon_fence *fence);
+/*
+ * Synchronization
+ */
+struct radeon_sync {
+ struct radeon_semaphore *semaphores[RADEON_NUM_SYNCS];
+ struct radeon_fence *sync_to[RADEON_NUM_RINGS];
+ struct radeon_fence *last_vm_update;
+};
+
+void radeon_sync_create(struct radeon_sync *sync);
+void radeon_sync_fence(struct radeon_sync *sync,
+ struct radeon_fence *fence);
+int radeon_sync_resv(struct radeon_device *rdev,
+ struct radeon_sync *sync,
+ struct reservation_object *resv,
+ bool shared);
+int radeon_sync_rings(struct radeon_device *rdev,
+ struct radeon_sync *sync,
+ int waiting_ring);
+void radeon_sync_free(struct radeon_device *rdev, struct radeon_sync *sync,
+ struct radeon_fence *fence);
+
/*
* GART structures, functions & helpers
*/
int radeon_doorbell_get(struct radeon_device *rdev, u32 *page);
void radeon_doorbell_free(struct radeon_device *rdev, u32 doorbell);
+void radeon_doorbell_get_kfd_info(struct radeon_device *rdev,
+ phys_addr_t *aperture_base,
+ size_t *aperture_size,
+ size_t *start_offset);
/*
* IRQS.
struct radeon_fence *fence;
struct radeon_vm *vm;
bool is_const_ib;
- struct radeon_semaphore *semaphore;
+ struct radeon_sync sync;
};
struct radeon_ring {
uint64_t addr;
};
+struct radeon_vm_id {
+ unsigned id;
+ uint64_t pd_gpu_addr;
+ /* last flushed PD/PT update */
+ struct radeon_fence *flushed_updates;
+ /* last use of vmid */
+ struct radeon_fence *last_id_use;
+};
+
struct radeon_vm {
- struct rb_root va;
- unsigned id;
+ struct mutex mutex;
+
+ struct rb_root va;
+
+ /* protecting invalidated and freed */
+ spinlock_t status_lock;
/* BOs moved, but not yet updated in the PT */
- struct list_head invalidated;
+ struct list_head invalidated;
/* BOs freed, but not yet updated in the PT */
- struct list_head freed;
+ struct list_head freed;
/* contains the page directory */
- struct radeon_bo *page_directory;
- uint64_t pd_gpu_addr;
- unsigned max_pde_used;
+ struct radeon_bo *page_directory;
+ unsigned max_pde_used;
/* array of page tables, one for each page directory entry */
- struct radeon_vm_pt *page_tables;
+ struct radeon_vm_pt *page_tables;
- struct radeon_bo_va *ib_bo_va;
+ struct radeon_bo_va *ib_bo_va;
- struct mutex mutex;
- /* last fence for cs using this vm */
- struct radeon_fence *fence;
- /* last flush or NULL if we still need to flush */
- struct radeon_fence *last_flush;
- /* last use of vmid */
- struct radeon_fence *last_id_use;
+ /* for id and flush management per ring */
+ struct radeon_vm_id ids[RADEON_NUM_RINGS];
};
struct radeon_vm_manager {
/*
* CS.
*/
-struct radeon_cs_reloc {
- struct drm_gem_object *gobj;
- struct radeon_bo *robj;
- struct ttm_validate_buffer tv;
- uint64_t gpu_offset;
- unsigned prefered_domains;
- unsigned allowed_domains;
- uint32_t tiling_flags;
- uint32_t handle;
-};
-
struct radeon_cs_chunk {
- uint32_t chunk_id;
uint32_t length_dw;
uint32_t *kdata;
void __user *user_ptr;
unsigned idx;
/* relocations */
unsigned nrelocs;
- struct radeon_cs_reloc *relocs;
- struct radeon_cs_reloc **relocs_ptr;
- struct radeon_cs_reloc *vm_bos;
+ struct radeon_bo_list *relocs;
+ struct radeon_bo_list *vm_bos;
struct list_head validated;
unsigned dma_reloc_idx;
/* indices of various chunks */
- int chunk_ib_idx;
- int chunk_relocs_idx;
- int chunk_flags_idx;
- int chunk_const_ib_idx;
+ struct radeon_cs_chunk *chunk_ib;
+ struct radeon_cs_chunk *chunk_relocs;
+ struct radeon_cs_chunk *chunk_flags;
+ struct radeon_cs_chunk *chunk_const_ib;
struct radeon_ib ib;
struct radeon_ib const_ib;
void *track;
static inline u32 radeon_get_ib_value(struct radeon_cs_parser *p, int idx)
{
- struct radeon_cs_chunk *ibc = &p->chunks[p->chunk_ib_idx];
+ struct radeon_cs_chunk *ibc = p->chunk_ib;
if (ibc->kdata)
return ibc->kdata[idx];
u8 t_hyst;
u32 cycle_delay;
u16 t_max;
+ u8 control_mode;
+ u16 default_max_fan_pwm;
+ u16 default_fan_output_sensitivity;
+ u16 fan_output_sensitivity;
bool ucode_fan_control;
};
/* internal thermal controller on rv6xx+ */
enum radeon_int_thermal_type int_thermal_type;
struct device *int_hwmon_dev;
+ /* fan control parameters */
+ bool no_fan;
+ u8 fan_pulses_per_revolution;
+ u8 fan_min_rpm;
+ u8 fan_max_rpm;
/* dpm */
bool dpm_enabled;
struct radeon_dpm dpm;
void (*hdp_flush)(struct radeon_device *rdev, struct radeon_ring *ring);
bool (*emit_semaphore)(struct radeon_device *rdev, struct radeon_ring *cp,
struct radeon_semaphore *semaphore, bool emit_wait);
- void (*vm_flush)(struct radeon_device *rdev, int ridx, struct radeon_vm *vm);
+ void (*vm_flush)(struct radeon_device *rdev, struct radeon_ring *ring,
+ unsigned vm_id, uint64_t pd_addr);
/* testing functions */
int (*ring_test)(struct radeon_device *rdev, struct radeon_ring *cp);
struct radeon_atcs atcs;
/* srbm instance registers */
struct mutex srbm_mutex;
+ /* GRBM index mutex. Protects concurrents access to GRBM index */
+ struct mutex grbm_idx_mutex;
/* clock, powergating flags */
u32 cg_flags;
u32 pg_flags;
u64 vram_pin_size;
u64 gart_pin_size;
+ /* amdkfd interface */
+ struct kfd_dev *kfd;
+ struct radeon_sa_manager kfd_bo;
+
struct mutex mn_lock;
DECLARE_HASHTABLE(mn_hash, 7);
};
#define radeon_ring_ib_execute(rdev, r, ib) (rdev)->asic->ring[(r)]->ib_execute((rdev), (ib))
#define radeon_ring_ib_parse(rdev, r, ib) (rdev)->asic->ring[(r)]->ib_parse((rdev), (ib))
#define radeon_ring_is_lockup(rdev, r, cp) (rdev)->asic->ring[(r)]->is_lockup((rdev), (cp))
-#define radeon_ring_vm_flush(rdev, r, vm) (rdev)->asic->ring[(r)]->vm_flush((rdev), (r), (vm))
+#define radeon_ring_vm_flush(rdev, r, vm_id, pd_addr) (rdev)->asic->ring[(r)->idx]->vm_flush((rdev), (r), (vm_id), (pd_addr))
#define radeon_ring_get_rptr(rdev, r) (rdev)->asic->ring[(r)->idx]->get_rptr((rdev), (r))
#define radeon_ring_get_wptr(rdev, r) (rdev)->asic->ring[(r)->idx]->get_wptr((rdev), (r))
#define radeon_ring_set_wptr(rdev, r) (rdev)->asic->ring[(r)->idx]->set_wptr((rdev), (r))
void radeon_vm_manager_fini(struct radeon_device *rdev);
int radeon_vm_init(struct radeon_device *rdev, struct radeon_vm *vm);
void radeon_vm_fini(struct radeon_device *rdev, struct radeon_vm *vm);
-struct radeon_cs_reloc *radeon_vm_get_bos(struct radeon_device *rdev,
+struct radeon_bo_list *radeon_vm_get_bos(struct radeon_device *rdev,
struct radeon_vm *vm,
struct list_head *head);
struct radeon_fence *radeon_vm_grab_id(struct radeon_device *rdev,
struct radeon_vm *vm, int ring);
void radeon_vm_flush(struct radeon_device *rdev,
struct radeon_vm *vm,
- int ring);
+ int ring, struct radeon_fence *fence);
void radeon_vm_fence(struct radeon_device *rdev,
struct radeon_vm *vm,
struct radeon_fence *fence);
void radeon_cs_dump_packet(struct radeon_cs_parser *p,
struct radeon_cs_packet *pkt);
int radeon_cs_packet_next_reloc(struct radeon_cs_parser *p,
- struct radeon_cs_reloc **cs_reloc,
+ struct radeon_bo_list **cs_reloc,
int nomm);
int r600_cs_common_vline_parse(struct radeon_cs_parser *p,
uint32_t *vline_start_end,
projects / karo-tx-linux.git / blobdiff