From: Yong Zhao <yong.zhao@amd.com>
Date: Wed, 20 Sep 2017 18:10:18 -0400
Subject: drm/amdkfd: Use VMID bitmap from KGD v2
Git-commit: 44008d7a871ce5a487cbcc4c412a5149145ea442
Patch-mainline: v4.15-rc1
References: FATE#326289 FATE#326079 FATE#326049 FATE#322398 FATE#326166

The hard-coded values related to VMID were removed in KFD, as those
values can be calculated in the KFD initialization function.

v2: remove unnecessary local variable

Signed-off-by: Yong Zhao <yong.zhao@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
Acked-by: Petr Tesarik <ptesarik@suse.com>
---
 drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c                |    9 ++-------
 drivers/gpu/drm/amd/amdkfd/kfd_device.c                |    5 +++++
 drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c  |   13 ++++++-------
 drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h  |    4 ----
 drivers/gpu/drm/amd/amdkfd/kfd_priv.h                  |    7 +++++++
 drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c |    2 +-
 6 files changed, 21 insertions(+), 19 deletions(-)

--- a/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c
@@ -769,13 +769,8 @@ int dbgdev_wave_reset_wavefronts(struct
 	union GRBM_GFX_INDEX_BITS reg_gfx_index;
 	struct kfd_process_device *pdd;
 	struct dbg_wave_control_info wac_info;
-	int temp;
-	int first_vmid_to_scan = 8;
-	int last_vmid_to_scan = 15;
-
-	first_vmid_to_scan = ffs(dev->shared_resources.compute_vmid_bitmap) - 1;
-	temp = dev->shared_resources.compute_vmid_bitmap >> first_vmid_to_scan;
-	last_vmid_to_scan = first_vmid_to_scan + ffz(temp);
+	int first_vmid_to_scan = dev->vm_info.first_vmid_kfd;
+	int last_vmid_to_scan = dev->vm_info.last_vmid_kfd;
 
 	reg_sq_cmd.u32All = 0;
 	status = 0;
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -219,6 +219,11 @@ bool kgd2kfd_device_init(struct kfd_dev
 
 	kfd->shared_resources = *gpu_resources;
 
+	kfd->vm_info.first_vmid_kfd = ffs(gpu_resources->compute_vmid_bitmap)-1;
+	kfd->vm_info.last_vmid_kfd = fls(gpu_resources->compute_vmid_bitmap)-1;
+	kfd->vm_info.vmid_num_kfd = kfd->vm_info.last_vmid_kfd
+			- kfd->vm_info.first_vmid_kfd + 1;
+
 	/* calculate max size of mqds needed for queues */
 	size = max_num_of_queues_per_device *
 			kfd->device_info->mqd_size_aligned;
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
@@ -113,11 +113,11 @@ static int allocate_vmid(struct device_q
 	if (dqm->vmid_bitmap == 0)
 		return -ENOMEM;
 
-	bit = find_first_bit((unsigned long *)&dqm->vmid_bitmap, CIK_VMID_NUM);
+	bit = find_first_bit((unsigned long *)&dqm->vmid_bitmap,
+				dqm->dev->vm_info.vmid_num_kfd);
 	clear_bit(bit, (unsigned long *)&dqm->vmid_bitmap);
 
-	/* Kaveri kfd vmid's starts from vmid 8 */
-	allocated_vmid = bit + KFD_VMID_START_OFFSET;
+	allocated_vmid = bit + dqm->dev->vm_info.first_vmid_kfd;
 	pr_debug("vmid allocation %d\n", allocated_vmid);
 	qpd->vmid = allocated_vmid;
 	q->properties.vmid = allocated_vmid;
@@ -132,7 +132,7 @@ static void deallocate_vmid(struct devic
 				struct qcm_process_device *qpd,
 				struct queue *q)
 {
-	int bit = qpd->vmid - KFD_VMID_START_OFFSET;
+	int bit = qpd->vmid - dqm->dev->vm_info.first_vmid_kfd;
 
 	/* Release the vmid mapping */
 	set_pasid_vmid_mapping(dqm, 0, qpd->vmid);
@@ -507,7 +507,7 @@ static int initialize_nocpsch(struct dev
 				dqm->allocated_queues[pipe] |= 1 << queue;
 	}
 
-	dqm->vmid_bitmap = (1 << VMID_PER_DEVICE) - 1;
+	dqm->vmid_bitmap = (1 << dqm->dev->vm_info.vmid_num_kfd) - 1;
 	dqm->sdma_bitmap = (1 << CIK_SDMA_QUEUES) - 1;
 
 	return 0;
@@ -613,8 +613,7 @@ static int set_sched_resources(struct de
 	int i, mec;
 	struct scheduling_resources res;
 
-	res.vmid_mask = (1 << VMID_PER_DEVICE) - 1;
-	res.vmid_mask <<= KFD_VMID_START_OFFSET;
+	res.vmid_mask = dqm->dev->shared_resources.compute_vmid_bitmap;
 
 	res.queue_mask = 0;
 	for (i = 0; i < KGD_MAX_QUEUES; ++i) {
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
@@ -32,10 +32,6 @@
 #define KFD_UNMAP_LATENCY_MS			(4000)
 #define QUEUE_PREEMPT_DEFAULT_TIMEOUT_MS (2 * KFD_UNMAP_LATENCY_MS + 1000)
 
-#define CIK_VMID_NUM				(8)
-#define KFD_VMID_START_OFFSET			(8)
-#define VMID_PER_DEVICE				CIK_VMID_NUM
-#define KFD_DQM_FIRST_PIPE			(0)
 #define CIK_SDMA_QUEUES				(4)
 #define CIK_SDMA_QUEUES_PER_ENGINE		(2)
 #define CIK_SDMA_ENGINE_NUM			(2)
--- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -141,6 +141,12 @@ struct kfd_mem_obj {
 	uint32_t *cpu_ptr;
 };
 
+struct kfd_vmid_info {
+	uint32_t first_vmid_kfd;
+	uint32_t last_vmid_kfd;
+	uint32_t vmid_num_kfd;
+};
+
 struct kfd_dev {
 	struct kgd_dev *kgd;
 
@@ -162,6 +168,7 @@ struct kfd_dev {
 					   */
 
 	struct kgd2kfd_shared_resources shared_resources;
+	struct kfd_vmid_info vm_info;
 
 	const struct kfd2kgd_calls *kfd2kgd;
 	struct mutex doorbell_mutex;
--- a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
@@ -187,7 +187,7 @@ int pqm_create_queue(struct process_queu
 	case KFD_QUEUE_TYPE_COMPUTE:
 		/* check if there is over subscription */
 		if ((sched_policy == KFD_SCHED_POLICY_HWS_NO_OVERSUBSCRIPTION) &&
-		((dev->dqm->processes_count >= VMID_PER_DEVICE) ||
+		((dev->dqm->processes_count >= dev->vm_info.vmid_num_kfd) ||
 		(dev->dqm->queue_count >= get_queues_num(dev->dqm)))) {
 			pr_err("Over-subscription is not allowed in radeon_kfd.sched_policy == 1\n");
 			retval = -EPERM;