Merging upstream version 6.10.3.

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

1 files changed, 59 insertions, 154 deletions

diff --git a/drivers/gpu/drm/xe/xe_vm.c b/drivers/gpu/drm/xe/xe_vm.c
index 32cd0c978a..4aa3943e6f 100644
--- a/drivers/gpu/drm/xe/xe_vm.c
+++ b/drivers/gpu/drm/xe/xe_vm.c
@@ -21,12 +21,12 @@
 
 #include <generated/xe_wa_oob.h>
 
+#include "regs/xe_gtt_defs.h"
 #include "xe_assert.h"
 #include "xe_bo.h"
 #include "xe_device.h"
 #include "xe_drm_client.h"
 #include "xe_exec_queue.h"
-#include "xe_gt.h"
 #include "xe_gt_pagefault.h"
 #include "xe_gt_tlb_invalidation.h"
 #include "xe_migrate.h"
@@ -38,6 +38,7 @@
 #include "xe_sync.h"
 #include "xe_trace.h"
 #include "xe_wa.h"
+#include "xe_hmm.h"
 
 static struct drm_gem_object *xe_vm_obj(struct xe_vm *vm)
 {
@@ -65,113 +66,14 @@ int xe_vma_userptr_check_repin(struct xe_userptr_vma *uvma)
 
 int xe_vma_userptr_pin_pages(struct xe_userptr_vma *uvma)
 {
-	struct xe_userptr *userptr = &uvma->userptr;
 	struct xe_vma *vma = &uvma->vma;
 	struct xe_vm *vm = xe_vma_vm(vma);
 	struct xe_device *xe = vm->xe;
-	const unsigned long num_pages = xe_vma_size(vma) >> PAGE_SHIFT;
-	struct page **pages;
-	bool in_kthread = !current->mm;
-	unsigned long notifier_seq;
-	int pinned, ret, i;
-	bool read_only = xe_vma_read_only(vma);
 
 	lockdep_assert_held(&vm->lock);
 	xe_assert(xe, xe_vma_is_userptr(vma));
-retry:
-	if (vma->gpuva.flags & XE_VMA_DESTROYED)
-		return 0;
-
-	notifier_seq = mmu_interval_read_begin(&userptr->notifier);
-	if (notifier_seq == userptr->notifier_seq)
-		return 0;
-
-	pages = kvmalloc_array(num_pages, sizeof(*pages), GFP_KERNEL);
-	if (!pages)
-		return -ENOMEM;
-
-	if (userptr->sg) {
-		dma_unmap_sgtable(xe->drm.dev,
-				  userptr->sg,
-				  read_only ? DMA_TO_DEVICE :
-				  DMA_BIDIRECTIONAL, 0);
-		sg_free_table(userptr->sg);
-		userptr->sg = NULL;
-	}
-
-	pinned = ret = 0;
-	if (in_kthread) {
-		if (!mmget_not_zero(userptr->notifier.mm)) {
-			ret = -EFAULT;
-			goto mm_closed;
-		}
-		kthread_use_mm(userptr->notifier.mm);
-	}
-
-	while (pinned < num_pages) {
-		ret = get_user_pages_fast(xe_vma_userptr(vma) +
-					  pinned * PAGE_SIZE,
-					  num_pages - pinned,
-					  read_only ? 0 : FOLL_WRITE,
-					  &pages[pinned]);
-		if (ret < 0)
-			break;
-
-		pinned += ret;
-		ret = 0;
-	}
-
-	if (in_kthread) {
-		kthread_unuse_mm(userptr->notifier.mm);
-		mmput(userptr->notifier.mm);
-	}
-mm_closed:
-	if (ret)
-		goto out;
-
-	ret = sg_alloc_table_from_pages_segment(&userptr->sgt, pages,
-						pinned, 0,
-						(u64)pinned << PAGE_SHIFT,
-						xe_sg_segment_size(xe->drm.dev),
-						GFP_KERNEL);
-	if (ret) {
-		userptr->sg = NULL;
-		goto out;
-	}
-	userptr->sg = &userptr->sgt;
-
-	ret = dma_map_sgtable(xe->drm.dev, userptr->sg,
-			      read_only ? DMA_TO_DEVICE :
-			      DMA_BIDIRECTIONAL,
-			      DMA_ATTR_SKIP_CPU_SYNC |
-			      DMA_ATTR_NO_KERNEL_MAPPING);
-	if (ret) {
-		sg_free_table(userptr->sg);
-		userptr->sg = NULL;
-		goto out;
-	}
-
-	for (i = 0; i < pinned; ++i) {
-		if (!read_only) {
-			lock_page(pages[i]);
-			set_page_dirty(pages[i]);
-			unlock_page(pages[i]);
-		}
-
-		mark_page_accessed(pages[i]);
-	}
-
-out:
-	release_pages(pages, pinned);
-	kvfree(pages);
-
-	if (!(ret < 0)) {
-		userptr->notifier_seq = notifier_seq;
-		if (xe_vma_userptr_check_repin(uvma) == -EAGAIN)
-			goto retry;
-	}
 
-	return ret < 0 ? ret : 0;
+	return xe_hmm_userptr_populate_range(uvma, false);
 }
 
 static bool preempt_fences_waiting(struct xe_vm *vm)
@@ -682,6 +584,10 @@ static bool vma_userptr_invalidate(struct mmu_interval_notifier *mni,
 	if (!mmu_notifier_range_blockable(range))
 		return false;
 
+	vm_dbg(&xe_vma_vm(vma)->xe->drm,
+	       "NOTIFIER: addr=0x%016llx, range=0x%016llx",
+		xe_vma_start(vma), xe_vma_size(vma));
+
 	down_write(&vm->userptr.notifier_lock);
 	mmu_interval_set_seq(mni, cur_seq);
 
@@ -951,8 +857,6 @@ static struct xe_vma *xe_vma_create(struct xe_vm *vm,
 static void xe_vma_destroy_late(struct xe_vma *vma)
 {
 	struct xe_vm *vm = xe_vma_vm(vma);
-	struct xe_device *xe = vm->xe;
-	bool read_only = xe_vma_read_only(vma);
 
 	if (vma->ufence) {
 		xe_sync_ufence_put(vma->ufence);
@@ -960,16 +864,11 @@ static void xe_vma_destroy_late(struct xe_vma *vma)
 	}
 
 	if (xe_vma_is_userptr(vma)) {
-		struct xe_userptr *userptr = &to_userptr_vma(vma)->userptr;
+		struct xe_userptr_vma *uvma = to_userptr_vma(vma);
+		struct xe_userptr *userptr = &uvma->userptr;
 
-		if (userptr->sg) {
-			dma_unmap_sgtable(xe->drm.dev,
-					  userptr->sg,
-					  read_only ? DMA_TO_DEVICE :
-					  DMA_BIDIRECTIONAL, 0);
-			sg_free_table(userptr->sg);
-			userptr->sg = NULL;
-		}
+		if (userptr->sg)
+			xe_hmm_userptr_free_sg(uvma);
 
 		/*
 		 * Since userptr pages are not pinned, we can't remove
@@ -1274,8 +1173,6 @@ static const struct xe_pt_ops xelp_pt_ops = {
 	.pde_encode_bo = xelp_pde_encode_bo,
 };
 
-static void vm_destroy_work_func(struct work_struct *w);
-
 /**
  * xe_vm_create_scratch() - Setup a scratch memory pagetable tree for the
  * given tile and vm.
@@ -1355,8 +1252,6 @@ struct xe_vm *xe_vm_create(struct xe_device *xe, u32 flags)
 	init_rwsem(&vm->userptr.notifier_lock);
 	spin_lock_init(&vm->userptr.invalidated_lock);
 
-	INIT_WORK(&vm->destroy_work, vm_destroy_work_func);
-
 	INIT_LIST_HEAD(&vm->preempt.exec_queues);
 	vm->preempt.min_run_period_ms = 10;	/* FIXME: Wire up to uAPI */
 
@@ -1366,7 +1261,7 @@ struct xe_vm *xe_vm_create(struct xe_device *xe, u32 flags)
 	vm->pt_ops = &xelp_pt_ops;
 
 	if (!(flags & XE_VM_FLAG_MIGRATION))
-		xe_device_mem_access_get(xe);
+		xe_pm_runtime_get_noresume(xe);
 
 	vm_resv_obj = drm_gpuvm_resv_object_alloc(&xe->drm);
 	if (!vm_resv_obj) {
@@ -1411,9 +1306,8 @@ struct xe_vm *xe_vm_create(struct xe_device *xe, u32 flags)
 		vm->batch_invalidate_tlb = true;
 	}
 
-	if (flags & XE_VM_FLAG_LR_MODE) {
+	if (vm->flags & XE_VM_FLAG_LR_MODE) {
 		INIT_WORK(&vm->preempt.rebind_work, preempt_rebind_work_func);
-		vm->flags |= XE_VM_FLAG_LR_MODE;
 		vm->batch_invalidate_tlb = false;
 	}
 
@@ -1477,7 +1371,7 @@ err_no_resv:
 		xe_range_fence_tree_fini(&vm->rftree[id]);
 	kfree(vm);
 	if (!(flags & XE_VM_FLAG_MIGRATION))
-		xe_device_mem_access_put(xe);
+		xe_pm_runtime_put(xe);
 	return ERR_PTR(err);
 }
 
@@ -1595,10 +1489,9 @@ void xe_vm_close_and_put(struct xe_vm *vm)
 	xe_vm_put(vm);
 }
 
-static void vm_destroy_work_func(struct work_struct *w)
+static void xe_vm_free(struct drm_gpuvm *gpuvm)
 {
-	struct xe_vm *vm =
-		container_of(w, struct xe_vm, destroy_work);
+	struct xe_vm *vm = container_of(gpuvm, struct xe_vm, gpuvm);
 	struct xe_device *xe = vm->xe;
 	struct xe_tile *tile;
 	u8 id;
@@ -1612,7 +1505,7 @@ static void vm_destroy_work_func(struct work_struct *w)
 	mutex_destroy(&vm->snap_mutex);
 
 	if (!(vm->flags & XE_VM_FLAG_MIGRATION))
-		xe_device_mem_access_put(xe);
+		xe_pm_runtime_put(xe);
 
 	for_each_tile(tile, xe, id)
 		XE_WARN_ON(vm->pt_root[id]);
@@ -1621,14 +1514,6 @@ static void vm_destroy_work_func(struct work_struct *w)
 	kfree(vm);
 }
 
-static void xe_vm_free(struct drm_gpuvm *gpuvm)
-{
-	struct xe_vm *vm = container_of(gpuvm, struct xe_vm, gpuvm);
-
-	/* To destroy the VM we need to be able to sleep */
-	queue_work(system_unbound_wq, &vm->destroy_work);
-}
-
 struct xe_vm *xe_vm_lookup(struct xe_file *xef, u32 id)
 {
 	struct xe_vm *vm;
@@ -1725,7 +1610,7 @@ next:
 		xe_exec_queue_last_fence_get(wait_exec_queue, vm) : fence;
 	if (last_op) {
 		for (i = 0; i < num_syncs; i++)
-			xe_sync_entry_signal(&syncs[i], NULL, fence);
+			xe_sync_entry_signal(&syncs[i], fence);
 	}
 
 	return fence;
@@ -1799,7 +1684,7 @@ next:
 
 	if (last_op) {
 		for (i = 0; i < num_syncs; i++)
-			xe_sync_entry_signal(&syncs[i], NULL,
+			xe_sync_entry_signal(&syncs[i],
 					     cf ? &cf->base : fence);
 	}
 
@@ -1860,7 +1745,7 @@ static int __xe_vm_bind(struct xe_vm *vm, struct xe_vma *vma,
 		fence = xe_exec_queue_last_fence_get(wait_exec_queue, vm);
 		if (last_op) {
 			for (i = 0; i < num_syncs; i++)
-				xe_sync_entry_signal(&syncs[i], NULL, fence);
+				xe_sync_entry_signal(&syncs[i], fence);
 		}
 	}
 
@@ -2061,7 +1946,7 @@ static int xe_vm_prefetch(struct xe_vm *vm, struct xe_vma *vma,
 	struct xe_exec_queue *wait_exec_queue = to_wait_exec_queue(vm, q);
 	int err;
 
-	xe_assert(vm->xe, region <= ARRAY_SIZE(region_to_mem_type));
+	xe_assert(vm->xe, region < ARRAY_SIZE(region_to_mem_type));
 
 	if (!xe_vma_has_no_bo(vma)) {
 		err = xe_bo_migrate(xe_vma_bo(vma), region_to_mem_type[region]);
@@ -2081,7 +1966,7 @@ static int xe_vm_prefetch(struct xe_vm *vm, struct xe_vma *vma,
 				struct dma_fence *fence =
 					xe_exec_queue_last_fence_get(wait_exec_queue, vm);
 
-				xe_sync_entry_signal(&syncs[i], NULL, fence);
+				xe_sync_entry_signal(&syncs[i], fence);
 				dma_fence_put(fence);
 			}
 		}
@@ -2213,6 +2098,10 @@ vm_bind_ioctl_ops_create(struct xe_vm *vm, struct xe_bo *bo,
 		struct xe_vma_op *op = gpuva_op_to_vma_op(__op);
 
 		if (__op->op == DRM_GPUVA_OP_MAP) {
+			op->map.immediate =
+				flags & DRM_XE_VM_BIND_FLAG_IMMEDIATE;
+			op->map.read_only =
+				flags & DRM_XE_VM_BIND_FLAG_READONLY;
 			op->map.is_null = flags & DRM_XE_VM_BIND_FLAG_NULL;
 			op->map.dumpable = flags & DRM_XE_VM_BIND_FLAG_DUMPABLE;
 			op->map.pat_index = pat_index;
@@ -2407,6 +2296,8 @@ static int vm_bind_ioctl_ops_parse(struct xe_vm *vm, struct xe_exec_queue *q,
 		switch (op->base.op) {
 		case DRM_GPUVA_OP_MAP:
 		{
+			flags |= op->map.read_only ?
+				VMA_CREATE_FLAG_READ_ONLY : 0;
 			flags |= op->map.is_null ?
 				VMA_CREATE_FLAG_IS_NULL : 0;
 			flags |= op->map.dumpable ?
@@ -2551,7 +2442,7 @@ static int op_execute(struct drm_exec *exec, struct xe_vm *vm,
 	case DRM_GPUVA_OP_MAP:
 		err = xe_vm_bind(vm, vma, op->q, xe_vma_bo(vma),
 				 op->syncs, op->num_syncs,
-				 !xe_vm_in_fault_mode(vm),
+				 op->map.immediate || !xe_vm_in_fault_mode(vm),
 				 op->flags & XE_VMA_OP_FIRST,
 				 op->flags & XE_VMA_OP_LAST);
 		break;
@@ -2826,7 +2717,10 @@ static int vm_bind_ioctl_ops_execute(struct xe_vm *vm,
 	return 0;
 }
 
-#define SUPPORTED_FLAGS	(DRM_XE_VM_BIND_FLAG_NULL | \
+#define SUPPORTED_FLAGS	\
+	(DRM_XE_VM_BIND_FLAG_READONLY | \
+	 DRM_XE_VM_BIND_FLAG_IMMEDIATE | \
+	 DRM_XE_VM_BIND_FLAG_NULL | \
 	 DRM_XE_VM_BIND_FLAG_DUMPABLE)
 #define XE_64K_PAGE_MASK 0xffffull
 #define ALL_DRM_XE_SYNCS_FLAGS (DRM_XE_SYNCS_FLAG_WAIT_FOR_OP)
@@ -2959,7 +2853,7 @@ static int vm_bind_ioctl_signal_fences(struct xe_vm *vm,
 		return PTR_ERR(fence);
 
 	for (i = 0; i < num_syncs; i++)
-		xe_sync_entry_signal(&syncs[i], NULL, fence);
+		xe_sync_entry_signal(&syncs[i], fence);
 
 	xe_exec_queue_last_fence_set(to_wait_exec_queue(vm, q), vm,
 				     fence);
@@ -3070,7 +2964,7 @@ int xe_vm_bind_ioctl(struct drm_device *dev, void *data, struct drm_file *file)
 			goto put_obj;
 		}
 
-		if (bos[i]->flags & XE_BO_INTERNAL_64K) {
+		if (bos[i]->flags & XE_BO_FLAG_INTERNAL_64K) {
 			if (XE_IOCTL_DBG(xe, obj_offset &
 					 XE_64K_PAGE_MASK) ||
 			    XE_IOCTL_DBG(xe, addr & XE_64K_PAGE_MASK) ||
@@ -3262,6 +3156,10 @@ int xe_vm_invalidate_vma(struct xe_vma *vma)
 	xe_assert(xe, !xe_vma_is_null(vma));
 	trace_xe_vma_invalidate(vma);
 
+	vm_dbg(&xe_vma_vm(vma)->xe->drm,
+	       "INVALIDATE: addr=0x%016llx, range=0x%016llx",
+		xe_vma_start(vma), xe_vma_size(vma));
+
 	/* Check that we don't race with page-table updates */
 	if (IS_ENABLED(CONFIG_PROVE_LOCKING)) {
 		if (xe_vma_is_userptr(vma)) {
@@ -3380,8 +3278,10 @@ struct xe_vm_snapshot *xe_vm_snapshot_capture(struct xe_vm *vm)
 
 	if (num_snaps)
 		snap = kvzalloc(offsetof(struct xe_vm_snapshot, snap[num_snaps]), GFP_NOWAIT);
-	if (!snap)
+	if (!snap) {
+		snap = num_snaps ? ERR_PTR(-ENOMEM) : ERR_PTR(-ENODEV);
 		goto out_unlock;
+	}
 
 	snap->num_snaps = num_snaps;
 	i = 0;
@@ -3421,6 +3321,9 @@ out_unlock:
 
 void xe_vm_snapshot_capture_delayed(struct xe_vm_snapshot *snap)
 {
+	if (IS_ERR_OR_NULL(snap))
+		return;
+
 	for (int i = 0; i < snap->num_snaps; i++) {
 		struct xe_bo *bo = snap->snap[i].bo;
 		struct iosys_map src;
@@ -3475,13 +3378,21 @@ void xe_vm_snapshot_print(struct xe_vm_snapshot *snap, struct drm_printer *p)
 {
 	unsigned long i, j;
 
-	for (i = 0; i < snap->num_snaps; i++) {
-		if (IS_ERR(snap->snap[i].data))
-			goto uncaptured;
+	if (IS_ERR_OR_NULL(snap)) {
+		drm_printf(p, "[0].error: %li\n", PTR_ERR(snap));
+		return;
+	}
 
+	for (i = 0; i < snap->num_snaps; i++) {
 		drm_printf(p, "[%llx].length: 0x%lx\n", snap->snap[i].ofs, snap->snap[i].len);
-		drm_printf(p, "[%llx].data: ",
-			   snap->snap[i].ofs);
+
+		if (IS_ERR(snap->snap[i].data)) {
+			drm_printf(p, "[%llx].error: %li\n", snap->snap[i].ofs,
+				   PTR_ERR(snap->snap[i].data));
+			continue;
+		}
+
+		drm_printf(p, "[%llx].data: ", snap->snap[i].ofs);
 
 		for (j = 0; j < snap->snap[i].len; j += sizeof(u32)) {
 			u32 *val = snap->snap[i].data + j;
@@ -3491,12 +3402,6 @@ void xe_vm_snapshot_print(struct xe_vm_snapshot *snap, struct drm_printer *p)
 		}
 
 		drm_puts(p, "\n");
-		continue;
-
-uncaptured:
-		drm_printf(p, "Unable to capture range [%llx-%llx]: %li\n",
-			   snap->snap[i].ofs, snap->snap[i].ofs + snap->snap[i].len - 1,
-			   PTR_ERR(snap->snap[i].data));
 	}
 }
 
@@ -3504,7 +3409,7 @@ void xe_vm_snapshot_free(struct xe_vm_snapshot *snap)
 {
 	unsigned long i;
 
-	if (!snap)
+	if (IS_ERR_OR_NULL(snap))
 		return;
 
 	for (i = 0; i < snap->num_snaps; i++) {