Merge remote-tracking branch 'brahma/fkuehlin/rocm-rel-3.5' into rocm…

…-rel-3.5

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>

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

@@ -1392,7 +1392,7 @@ int amdgpu_amdkfd_gpuvm_free_memory_of_gpu(
 	}
 
 	/* Free the BO*/
-	amdgpu_bo_unref(&mem->bo);
+	drm_gem_object_put_unlocked(&mem->bo->tbo.base);
 	mutex_destroy(&mem->lock);
 	kfree(mem);
 
@@ -1895,7 +1895,8 @@ int amdgpu_amdkfd_gpuvm_import_dmabuf(struct kgd_dev *kgd,
 			| KFD_IOC_ALLOC_MEM_FLAGS_WRITABLE
 			| KFD_IOC_ALLOC_MEM_FLAGS_EXECUTABLE;
 
-	(*mem)->bo = amdgpu_bo_ref(bo);
+	drm_gem_object_get(&bo->tbo.base);
+	(*mem)->bo = bo;
 	(*mem)->va = va;
 	(*mem)->domain = (bo->preferred_domains & AMDGPU_GEM_DOMAIN_VRAM) ?
 		AMDGPU_GEM_DOMAIN_VRAM : AMDGPU_GEM_DOMAIN_GTT;

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

@@ -395,7 +395,9 @@ int amdgpu_xgmi_set_pstate(struct amdgpu_device *adev, int pstate)
 	bool init_low = hive->pstate == AMDGPU_XGMI_PSTATE_UNKNOWN;
 
 	/* fw bug so temporarily disable pstate switching */
-	if (!hive || adev->asic_type == CHIP_VEGA20)
+	return 0;
+
+	if (!hive || adev->asic_type != CHIP_VEGA20)
 		return 0;
 
 	mutex_lock(&hive->hive_lock);

drivers/gpu/drm/amd/amdkfd/Makefile

@@ -53,6 +53,7 @@ AMDKFD_FILES	:= $(AMDKFD_PATH)/kfd_module.o \
 		$(AMDKFD_PATH)/kfd_int_process_v9.o \
 		$(AMDKFD_PATH)/kfd_dbgdev.o \
 		$(AMDKFD_PATH)/kfd_dbgmgr.o \
+		$(AMDKFD_PATH)/kfd_smi_events.o \
 		$(AMDKFD_PATH)/kfd_crat.o \
 		$(AMDKFD_PATH)/kfd_rdma.o \
 		$(AMDKFD_PATH)/kfd_peerdirect.o \

drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c

@@ -24,6 +24,7 @@
 #include "kfd_events.h"
 #include "cik_int.h"
 #include "amdgpu_amdkfd.h"
+#include "kfd_smi_events.h"
 
 static bool cik_event_interrupt_isr(struct kfd_dev *dev,
 					const uint32_t *ih_ring_entry,
@@ -107,6 +108,7 @@ static void cik_event_interrupt_wq(struct kfd_dev *dev,
 		ihre->source_id == CIK_INTSRC_GFX_MEM_PROT_FAULT) {
 		struct kfd_vm_fault_info info;
 
+		kfd_smi_event_update_vmfault(dev, pasid);
 		kfd_process_vm_fault(dev->dqm, pasid);
 
 		memset(&info, 0, sizeof(info));

drivers/gpu/drm/amd/amdkfd/kfd_chardev.c

@@ -47,6 +47,7 @@
 #include "kfd_trace.h"
 
 #include "amdgpu_amdkfd.h"
+#include "kfd_smi_events.h"
 
 static long kfd_ioctl(struct file *, unsigned int, unsigned long);
 static int kfd_open(struct inode *, struct file *);
@@ -2681,7 +2682,7 @@ static int kfd_ioctl_dbg_set_debug_trap(struct file *filep,
 	target = need_proc_create ?
 		kfd_create_process(thread) : kfd_lookup_process_by_pid(pid);
 	if (!target) {
-		pr_err("Cannot find process info info for %i\n",
+		pr_debug("Cannot find process info for %i\n",
 				args->pid);
 		r = -ESRCH;
 		goto out;
@@ -2832,17 +2833,27 @@ static int kfd_ioctl_dbg_set_debug_trap(struct file *filep,
 				data3, /* Grace Period */
 				data1, /* Flags */
 				queue_id_array); /* array of queue ids */
-		if (r)
+
+		if (copy_to_user((void __user *)args->ptr, queue_id_array,
+				sizeof(uint32_t) * data2)) {
+			r = -EFAULT;
 			goto unlock_out;
+		}
+
 		break;
 
 	case KFD_IOC_DBG_TRAP_NODE_RESUME:
 		r = resume_queues(target,
 				data2, /* Number of queues */
 				data1, /* Flags */
 				queue_id_array); /* array of queue ids */
-		if (r)
+
+		if (copy_to_user((void __user *)args->ptr, queue_id_array,
+				sizeof(uint32_t) * data2)) {
+			r = -EFAULT;
 			goto unlock_out;
+		}
+
 		break;
 	case KFD_IOC_DBG_TRAP_QUERY_DEBUG_EVENT:
 		r = kfd_dbg_ev_query_debug_event(pdd, &args->data1,
@@ -2896,6 +2907,20 @@ static int kfd_ioctl_dbg_set_debug_trap(struct file *filep,
 	return r;
 }
 
+/* Handle requests for watching SMI events */
+static int kfd_ioctl_smi_events(struct file *filep,
+				struct kfd_process *p, void *data)
+{
+	struct kfd_ioctl_smi_events_args *args = data;
+	struct kfd_dev *dev;
+
+	dev = kfd_device_by_id(args->gpuid);
+	if (!dev)
+		return -EINVAL;
+
+	return kfd_smi_event_open(dev, &args->anon_fd);
+}
+
 #define AMDKFD_IOCTL_DEF(ioctl, _func, _flags) \
 	[_IOC_NR(ioctl)] = {.cmd = ioctl, .func = _func, .flags = _flags, \
 			    .cmd_drv = 0, .name = #ioctl}
@@ -2992,6 +3017,9 @@ static const struct amdkfd_ioctl_desc amdkfd_ioctls[] = {
 	AMDKFD_IOCTL_DEF(AMDKFD_IOC_ALLOC_QUEUE_GWS,
 			kfd_ioctl_alloc_queue_gws, 0),
 
+	AMDKFD_IOCTL_DEF(AMDKFD_IOC_SMI_EVENTS,
+			kfd_ioctl_smi_events, 0),
+
 	AMDKFD_IOCTL_DEF(AMDKFD_IOC_IPC_IMPORT_HANDLE,
 				kfd_ioctl_ipc_import_handle, 0),
 
@@ -3005,8 +3033,6 @@ static const struct amdkfd_ioctl_desc amdkfd_ioctls[] = {
 			kfd_ioctl_dbg_set_debug_trap, 0),
 };
 
-#define AMDKFD_CORE_IOCTL_COUNT	ARRAY_SIZE(amdkfd_ioctls)
-
 static long kfd_ioctl(struct file *filep, unsigned int cmd, unsigned long arg)
 {
 	struct kfd_process *process;
@@ -3018,10 +3044,8 @@ static long kfd_ioctl(struct file *filep, unsigned int cmd, unsigned long arg)
 	unsigned int usize, asize;
 	int retcode = -EINVAL;
 
-	if (nr >= AMDKFD_CORE_IOCTL_COUNT)
-		goto err_i1;
-
-	if ((nr >= AMDKFD_COMMAND_START) && (nr < AMDKFD_COMMAND_END)) {
+	if (((nr >= AMDKFD_COMMAND_START) && (nr < AMDKFD_COMMAND_END)) ||
+	    ((nr >= AMDKFD_COMMAND_START_2) && (nr < AMDKFD_COMMAND_END_2))) {
 		u32 amdkfd_size;
 
 		ioctl = &amdkfd_ioctls[nr];

drivers/gpu/drm/amd/amdkfd/kfd_debug_events.c

@@ -122,8 +122,6 @@ uint32_t kfd_dbg_get_queue_status_word(struct queue *q, int flags)
 	if (flags & KFD_DBG_EV_FLAG_CLEAR_STATUS)
 		WRITE_ONCE(q->properties.debug_event_type, 0);
 
-	q->properties.is_new = false;
-
 	return queue_status_word;
 }
 
@@ -153,17 +151,22 @@ int kfd_dbg_ev_query_debug_event(struct kfd_process_device *pdd,
 		}
 
 		*event_status = kfd_dbg_get_queue_status_word(q, flags);
-
+		q->properties.is_new = false;
 		goto out;
 	}
 
 	list_for_each_entry(pqn, &pqm->queues, process_queue_list) {
-		unsigned int tmp_status =
-				kfd_dbg_get_queue_status_word(pqn->q, flags);
-		if (pqn->q && (tmp_status & (KFD_DBG_EV_STATUS_TRAP |
-						KFD_DBG_EV_STATUS_VMFAULT))) {
+		unsigned int tmp_status;
+
+		if (!pqn->q)
+			continue;
+
+		tmp_status = kfd_dbg_get_queue_status_word(pqn->q, flags);
+		if (tmp_status & (KFD_DBG_EV_STATUS_TRAP |
+						KFD_DBG_EV_STATUS_VMFAULT)) {
 			*queue_id = pqn->q->properties.queue_id;
 			*event_status = tmp_status;
+			pqn->q->properties.is_new = false;
 			goto out;
 		}
 	}

drivers/gpu/drm/amd/amdkfd/kfd_device.c

@@ -586,6 +586,11 @@ static int kfd_gws_init(struct kfd_dev *kfd)
 	return ret;
 }
 
+static void kfd_smi_init(struct kfd_dev *dev) {
+	INIT_LIST_HEAD(&dev->smi_clients);
+	spin_lock_init(&dev->smi_lock);
+}
+
 bool kgd2kfd_device_init(struct kfd_dev *kfd,
 			 struct drm_device *ddev,
 			 const struct kgd2kfd_shared_resources *gpu_resources)
@@ -700,6 +705,8 @@ bool kgd2kfd_device_init(struct kfd_dev *kfd,
 		goto kfd_topology_add_device_error;
 	}
 
+	kfd_smi_init(kfd);
+
 	kfd->init_complete = true;
 	dev_info(kfd_device, "added device %x:%x\n", kfd->pdev->vendor,
 		 kfd->pdev->device);