Merge branch 'drm-next-4.15' of git://people.freedesktop.org/~agd5f/l…

…inux into drm-next

More new stuff for 4.15. Highlights:
- Add clock query interface for raven
- Add new FENCE_TO_HANDLE ioctl
- UVD video encode ring support on polaris
- transparent huge page DMA support
- deadlock fixes
- compute pipe lru tweaks
- powerplay cleanups and regression fixes
- fix duplicate symbol issue with radeon and amdgpu
- misc bug fixes

* 'drm-next-4.15' of git://people.freedesktop.org/~agd5f/linux: (72 commits)
  drm/radeon/dp: make radeon_dp_get_dp_link_config static
  drm/radeon: move ci_send_msg_to_smc to where it's used
  drm/amd/sched: fix deadlock caused by unsignaled fences of deleted jobs
  drm/amd/sched: NULL out the s_fence field after run_job
  drm/amd/sched: move adding finish callback to amd_sched_job_begin
  drm/amd/sched: fix an outdated comment
  drm/amd/sched: rename amd_sched_entity_pop_job
  drm/amdgpu: minor coding style fix
  drm/ttm: add transparent huge page support for DMA allocations v2
  drm/ttm: add support for different pool sizes
  drm/ttm: remove unsued options from ttm_mem_global_alloc_page
  drm/amdgpu: add uvd enc irq
  drm/amdgpu: add uvd enc ib test
  drm/amdgpu: add uvd enc ring test
  drm/amdgpu: add uvd enc vm functions (v2)
  drm/amdgpu: add uvd enc into run queue
  drm/amdgpu: add uvd enc rings
  drm/amdgpu: add new uvd enc ring methods
  drm/amdgpu: add uvd enc command in header
  drm/amdgpu: add uvd enc registers in header
  ...

drivers/gpu/drm/amd/amdgpu/amdgpu.h

@@ -121,6 +121,7 @@ extern int amdgpu_cntl_sb_buf_per_se;
 extern int amdgpu_param_buf_per_se;
 extern int amdgpu_job_hang_limit;
 extern int amdgpu_lbpw;
+extern int amdgpu_compute_multipipe;
 
 #ifdef CONFIG_DRM_AMDGPU_SI
 extern int amdgpu_si_support;
@@ -1310,6 +1311,8 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
 int amdgpu_gem_op_ioctl(struct drm_device *dev, void *data,
 			struct drm_file *filp);
 int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp);
+int amdgpu_cs_fence_to_handle_ioctl(struct drm_device *dev, void *data,
+				    struct drm_file *filp);
 int amdgpu_cs_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *filp);
 int amdgpu_cs_wait_fences_ioctl(struct drm_device *dev, void *data,
 				struct drm_file *filp);
@@ -1524,7 +1527,6 @@ struct amdgpu_device {
 
 	/* powerplay */
 	struct amd_powerplay		powerplay;
-	bool				pp_enabled;
 	bool				pp_force_state_enabled;
 
 	/* dpm */

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c

@@ -338,6 +338,7 @@ static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
 	struct cik_mqd *m;
 	uint32_t *mqd_hqd;
 	uint32_t reg, wptr_val, data;
+	bool valid_wptr = false;
 
 	m = get_mqd(mqd);
 
@@ -356,7 +357,14 @@ static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
 			     CP_HQD_PQ_DOORBELL_CONTROL, DOORBELL_EN, 1);
 	WREG32(mmCP_HQD_PQ_DOORBELL_CONTROL, data);
 
-	if (read_user_wptr(mm, wptr, wptr_val))
+	/* read_user_ptr may take the mm->mmap_sem.
+	 * release srbm_mutex to avoid circular dependency between
+	 * srbm_mutex->mm_sem->reservation_ww_class_mutex->srbm_mutex.
+	 */
+	release_queue(kgd);
+	valid_wptr = read_user_wptr(mm, wptr, wptr_val);
+	acquire_queue(kgd, pipe_id, queue_id);
+	if (valid_wptr)
 		WREG32(mmCP_HQD_PQ_WPTR, (wptr_val << wptr_shift) & wptr_mask);
 
 	data = REG_SET_FIELD(m->cp_hqd_active, CP_HQD_ACTIVE, ACTIVE, 1);

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c

@@ -292,6 +292,7 @@ static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
 	struct vi_mqd *m;
 	uint32_t *mqd_hqd;
 	uint32_t reg, wptr_val, data;
+	bool valid_wptr = false;
 
 	m = get_mqd(mqd);
 
@@ -339,7 +340,14 @@ static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
 			     CP_HQD_PQ_DOORBELL_CONTROL, DOORBELL_EN, 1);
 	WREG32(mmCP_HQD_PQ_DOORBELL_CONTROL, data);
 
-	if (read_user_wptr(mm, wptr, wptr_val))
+	/* read_user_ptr may take the mm->mmap_sem.
+	 * release srbm_mutex to avoid circular dependency between
+	 * srbm_mutex->mm_sem->reservation_ww_class_mutex->srbm_mutex.
+	 */
+	release_queue(kgd);
+	valid_wptr = read_user_wptr(mm, wptr, wptr_val);
+	acquire_queue(kgd, pipe_id, queue_id);
+	if (valid_wptr)
 		WREG32(mmCP_HQD_PQ_WPTR, (wptr_val << wptr_shift) & wptr_mask);
 
 	data = REG_SET_FIELD(m->cp_hqd_active, CP_HQD_ACTIVE, ACTIVE, 1);

drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c

@@ -42,6 +42,28 @@ struct amdgpu_cgs_device {
 	struct amdgpu_device *adev =					\
 		((struct amdgpu_cgs_device *)cgs_device)->adev
 
+static void *amdgpu_cgs_register_pp_handle(struct cgs_device *cgs_device,
+			int (*call_back_func)(struct amd_pp_init *, void **))
+{
+	CGS_FUNC_ADEV;
+	struct amd_pp_init pp_init;
+	struct amd_powerplay *amd_pp;
+
+	if (call_back_func == NULL)
+		return NULL;
+
+	amd_pp = &(adev->powerplay);
+	pp_init.chip_family = adev->family;
+	pp_init.chip_id = adev->asic_type;
+	pp_init.pm_en = (amdgpu_dpm != 0 && !amdgpu_sriov_vf(adev)) ? true : false;
+	pp_init.feature_mask = amdgpu_pp_feature_mask;
+	pp_init.device = cgs_device;
+	if (call_back_func(&pp_init, &(amd_pp->pp_handle)))
+		return NULL;
+
+	return adev->powerplay.pp_handle;
+}
+
 static int amdgpu_cgs_alloc_gpu_mem(struct cgs_device *cgs_device,
 				    enum cgs_gpu_mem_type type,
 				    uint64_t size, uint64_t align,
@@ -1179,6 +1201,7 @@ static const struct cgs_ops amdgpu_cgs_ops = {
 	.is_virtualization_enabled = amdgpu_cgs_is_virtualization_enabled,
 	.enter_safe_mode = amdgpu_cgs_enter_safe_mode,
 	.lock_grbm_idx = amdgpu_cgs_lock_grbm_idx,
+	.register_pp_handle = amdgpu_cgs_register_pp_handle,
 };
 
 static const struct cgs_os_ops amdgpu_cgs_os_ops = {

drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c

@@ -25,6 +25,7 @@
  *    Jerome Glisse <glisse@freedesktop.org>
  */
 #include <linux/pagemap.h>
+#include <linux/sync_file.h>
 #include <drm/drmP.h>
 #include <drm/amdgpu_drm.h>
 #include <drm/drm_syncobj.h>
@@ -1330,6 +1331,66 @@ static struct dma_fence *amdgpu_cs_get_fence(struct amdgpu_device *adev,
 	return fence;
 }
 
+int amdgpu_cs_fence_to_handle_ioctl(struct drm_device *dev, void *data,
+				    struct drm_file *filp)
+{
+	struct amdgpu_device *adev = dev->dev_private;
+	struct amdgpu_fpriv *fpriv = filp->driver_priv;
+	union drm_amdgpu_fence_to_handle *info = data;
+	struct dma_fence *fence;
+	struct drm_syncobj *syncobj;
+	struct sync_file *sync_file;
+	int fd, r;
+
+	if (amdgpu_kms_vram_lost(adev, fpriv))
+		return -ENODEV;
+
+	fence = amdgpu_cs_get_fence(adev, filp, &info->in.fence);
+	if (IS_ERR(fence))
+		return PTR_ERR(fence);
+
+	switch (info->in.what) {
+	case AMDGPU_FENCE_TO_HANDLE_GET_SYNCOBJ:
+		r = drm_syncobj_create(&syncobj, 0, fence);
+		dma_fence_put(fence);
+		if (r)
+			return r;
+		r = drm_syncobj_get_handle(filp, syncobj, &info->out.handle);
+		drm_syncobj_put(syncobj);
+		return r;
+
+	case AMDGPU_FENCE_TO_HANDLE_GET_SYNCOBJ_FD:
+		r = drm_syncobj_create(&syncobj, 0, fence);
+		dma_fence_put(fence);
+		if (r)
+			return r;
+		r = drm_syncobj_get_fd(syncobj, (int*)&info->out.handle);
+		drm_syncobj_put(syncobj);
+		return r;
+
+	case AMDGPU_FENCE_TO_HANDLE_GET_SYNC_FILE_FD:
+		fd = get_unused_fd_flags(O_CLOEXEC);
+		if (fd < 0) {
+			dma_fence_put(fence);
+			return fd;
+		}
+
+		sync_file = sync_file_create(fence);
+		dma_fence_put(fence);
+		if (!sync_file) {
+			put_unused_fd(fd);
+			return -ENOMEM;
+		}
+
+		fd_install(fd, sync_file->file);
+		info->out.handle = fd;
+		return 0;
+
+	default:
+		return -EINVAL;
+	}
+}
+
 /**
  * amdgpu_cs_wait_all_fence - wait on all fences to signal
  *

drivers/gpu/drm/amd/amdgpu/amdgpu_device.c

@@ -56,6 +56,7 @@
 #include "amdgpu_vf_error.h"
 
 #include "amdgpu_amdkfd.h"
+#include "amdgpu_pm.h"
 
 MODULE_FIRMWARE("amdgpu/vega10_gpu_info.bin");
 MODULE_FIRMWARE("amdgpu/raven_gpu_info.bin");
@@ -1603,6 +1604,7 @@ static int amdgpu_init(struct amdgpu_device *adev)
 			return r;
 		}
 		adev->ip_blocks[i].status.sw = true;
+
 		/* need to do gmc hw init early so we can allocate gpu mem */
 		if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_GMC) {
 			r = amdgpu_vram_scratch_init(adev);
@@ -1633,6 +1635,11 @@ static int amdgpu_init(struct amdgpu_device *adev)
 		}
 	}
 
+	mutex_lock(&adev->firmware.mutex);
+	if (amdgpu_ucode_init_bo(adev))
+		adev->firmware.load_type = AMDGPU_FW_LOAD_DIRECT;
+	mutex_unlock(&adev->firmware.mutex);
+
 	for (i = 0; i < adev->num_ip_blocks; i++) {
 		if (!adev->ip_blocks[i].status.sw)
 			continue;
@@ -1768,6 +1775,8 @@ static int amdgpu_fini(struct amdgpu_device *adev)
 
 		adev->ip_blocks[i].status.hw = false;
 	}
+	if (adev->firmware.load_type != AMDGPU_FW_LOAD_DIRECT)
+		amdgpu_ucode_fini_bo(adev);
 
 	for (i = adev->num_ip_blocks - 1; i >= 0; i--) {
 		if (!adev->ip_blocks[i].status.sw)
@@ -2040,6 +2049,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	mutex_init(&adev->srbm_mutex);
 	mutex_init(&adev->grbm_idx_mutex);
 	mutex_init(&adev->mn_lock);
+	mutex_init(&adev->virt.vf_errors.lock);
 	hash_init(adev->mn_hash);
 
 	amdgpu_check_arguments(adev);
@@ -2125,7 +2135,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	r = amdgpu_atombios_init(adev);
 	if (r) {
 		dev_err(adev->dev, "amdgpu_atombios_init failed\n");
-		amdgpu_vf_error_put(AMDGIM_ERROR_VF_ATOMBIOS_INIT_FAIL, 0, 0);
+		amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_ATOMBIOS_INIT_FAIL, 0, 0);
 		goto failed;
 	}
 
@@ -2136,15 +2146,15 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	if (amdgpu_vpost_needed(adev)) {
 		if (!adev->bios) {
 			dev_err(adev->dev, "no vBIOS found\n");
-			amdgpu_vf_error_put(AMDGIM_ERROR_VF_NO_VBIOS, 0, 0);
+			amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_NO_VBIOS, 0, 0);
 			r = -EINVAL;
 			goto failed;
 		}
 		DRM_INFO("GPU posting now...\n");
 		r = amdgpu_atom_asic_init(adev->mode_info.atom_context);
 		if (r) {
 			dev_err(adev->dev, "gpu post error!\n");
-			amdgpu_vf_error_put(AMDGIM_ERROR_VF_GPU_POST_ERROR, 0, 0);
+			amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_GPU_POST_ERROR, 0, 0);
 			goto failed;
 		}
 	} else {
@@ -2156,15 +2166,15 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 		r = amdgpu_atomfirmware_get_clock_info(adev);
 		if (r) {
 			dev_err(adev->dev, "amdgpu_atomfirmware_get_clock_info failed\n");
-			amdgpu_vf_error_put(AMDGIM_ERROR_VF_ATOMBIOS_GET_CLOCK_FAIL, 0, 0);
+			amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_ATOMBIOS_GET_CLOCK_FAIL, 0, 0);
 			goto failed;
 		}
 	} else {
 		/* Initialize clocks */
 		r = amdgpu_atombios_get_clock_info(adev);
 		if (r) {
 			dev_err(adev->dev, "amdgpu_atombios_get_clock_info failed\n");
-			amdgpu_vf_error_put(AMDGIM_ERROR_VF_ATOMBIOS_GET_CLOCK_FAIL, 0, 0);
+			amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_ATOMBIOS_GET_CLOCK_FAIL, 0, 0);
 			goto failed;
 		}
 		/* init i2c buses */
@@ -2175,7 +2185,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	r = amdgpu_fence_driver_init(adev);
 	if (r) {
 		dev_err(adev->dev, "amdgpu_fence_driver_init failed\n");
-		amdgpu_vf_error_put(AMDGIM_ERROR_VF_FENCE_INIT_FAIL, 0, 0);
+		amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_FENCE_INIT_FAIL, 0, 0);
 		goto failed;
 	}
 
@@ -2185,7 +2195,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	r = amdgpu_init(adev);
 	if (r) {
 		dev_err(adev->dev, "amdgpu_init failed\n");
-		amdgpu_vf_error_put(AMDGIM_ERROR_VF_AMDGPU_INIT_FAIL, 0, 0);
+		amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_AMDGPU_INIT_FAIL, 0, 0);
 		amdgpu_fini(adev);
 		goto failed;
 	}
@@ -2205,7 +2215,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	r = amdgpu_ib_pool_init(adev);
 	if (r) {
 		dev_err(adev->dev, "IB initialization failed (%d).\n", r);
-		amdgpu_vf_error_put(AMDGIM_ERROR_VF_IB_INIT_FAIL, 0, r);
+		amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_IB_INIT_FAIL, 0, r);
 		goto failed;
 	}
 
@@ -2215,6 +2225,10 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 
 	amdgpu_fbdev_init(adev);
 
+	r = amdgpu_pm_sysfs_init(adev);
+	if (r)
+		DRM_ERROR("registering pm debugfs failed (%d).\n", r);
+
 	r = amdgpu_gem_debugfs_init(adev);
 	if (r)
 		DRM_ERROR("registering gem debugfs failed (%d).\n", r);
@@ -2254,7 +2268,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	r = amdgpu_late_init(adev);
 	if (r) {
 		dev_err(adev->dev, "amdgpu_late_init failed\n");
-		amdgpu_vf_error_put(AMDGIM_ERROR_VF_AMDGPU_LATE_INIT_FAIL, 0, r);
+		amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_AMDGPU_LATE_INIT_FAIL, 0, r);
 		goto failed;
 	}
 
@@ -2311,6 +2325,7 @@ void amdgpu_device_fini(struct amdgpu_device *adev)
 	iounmap(adev->rmmio);
 	adev->rmmio = NULL;
 	amdgpu_doorbell_fini(adev);
+	amdgpu_pm_sysfs_fini(adev);
 	amdgpu_debugfs_regs_cleanup(adev);
 }
 
@@ -2936,7 +2951,7 @@ int amdgpu_gpu_reset(struct amdgpu_device *adev)
 		}
 	} else {
 		dev_err(adev->dev, "asic resume failed (%d).\n", r);
-		amdgpu_vf_error_put(AMDGIM_ERROR_VF_ASIC_RESUME_FAIL, 0, r);
+		amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_ASIC_RESUME_FAIL, 0, r);
 		for (i = 0; i < AMDGPU_MAX_RINGS; ++i) {
 			if (adev->rings[i] && adev->rings[i]->sched.thread) {
 				kthread_unpark(adev->rings[i]->sched.thread);
@@ -2950,7 +2965,7 @@ int amdgpu_gpu_reset(struct amdgpu_device *adev)
 	if (r) {
 		/* bad news, how to tell it to userspace ? */
 		dev_info(adev->dev, "GPU reset failed\n");
-		amdgpu_vf_error_put(AMDGIM_ERROR_VF_GPU_RESET_FAIL, 0, r);
+		amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_GPU_RESET_FAIL, 0, r);
 	}
 	else {
 		dev_info(adev->dev, "GPU reset successed!\n");

drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h

@@ -356,6 +356,10 @@ enum amdgpu_pcie_gen {
 		((adev)->powerplay.pp_funcs->switch_power_profile(\
 			(adev)->powerplay.pp_handle, type))
 
+#define amdgpu_dpm_set_clockgating_by_smu(adev, msg_id) \
+		((adev)->powerplay.pp_funcs->set_clockgating_by_smu(\
+			(adev)->powerplay.pp_handle, msg_id))
+
 struct amdgpu_dpm {
 	struct amdgpu_ps        *ps;
 	/* number of valid power states */

drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c

@@ -70,9 +70,10 @@
  * - 3.18.0 - Export gpu always on cu bitmap
  * - 3.19.0 - Add support for UVD MJPEG decode
  * - 3.20.0 - Add support for local BOs
+ * - 3.21.0 - Add DRM_AMDGPU_FENCE_TO_HANDLE ioctl
  */
 #define KMS_DRIVER_MAJOR	3
-#define KMS_DRIVER_MINOR	20
+#define KMS_DRIVER_MINOR	21
 #define KMS_DRIVER_PATCHLEVEL	0
 
 int amdgpu_vram_limit = 0;
@@ -122,6 +123,7 @@ int amdgpu_cntl_sb_buf_per_se = 0;
 int amdgpu_param_buf_per_se = 0;
 int amdgpu_job_hang_limit = 0;
 int amdgpu_lbpw = -1;
+int amdgpu_compute_multipipe = -1;
 
 MODULE_PARM_DESC(vramlimit, "Restrict VRAM for testing, in megabytes");
 module_param_named(vramlimit, amdgpu_vram_limit, int, 0600);
@@ -265,6 +267,9 @@ module_param_named(job_hang_limit, amdgpu_job_hang_limit, int ,0444);
 MODULE_PARM_DESC(lbpw, "Load Balancing Per Watt (LBPW) support (1 = enable, 0 = disable, -1 = auto)");
 module_param_named(lbpw, amdgpu_lbpw, int, 0444);
 
+MODULE_PARM_DESC(compute_multipipe, "Force compute queues to be spread across pipes (1 = enable, 0 = disable, -1 = auto)");
+module_param_named(compute_multipipe, amdgpu_compute_multipipe, int, 0444);
+
 #ifdef CONFIG_DRM_AMDGPU_SI
 
 #if defined(CONFIG_DRM_RADEON) || defined(CONFIG_DRM_RADEON_MODULE)