diff options
Diffstat (limited to 'drivers/gpu/drm/i915/i915_vma.c')
| -rw-r--r-- | drivers/gpu/drm/i915/i915_vma.c | 2262 |
1 files changed, 1734 insertions, 528 deletions
diff --git a/drivers/gpu/drm/i915/i915_vma.c b/drivers/gpu/drm/i915/i915_vma.c index 5b4d78cdb4ca..2c0a63664e13 100644 --- a/drivers/gpu/drm/i915/i915_vma.c +++ b/drivers/gpu/drm/i915/i915_vma.c @@ -22,13 +22,53 @@ * */ -#include "i915_vma.h" +#include <linux/sched/mm.h> +#include <linux/dma-fence-array.h> + +#include <drm/drm_gem.h> +#include <drm/drm_print.h> + +#include "display/intel_fb.h" +#include "display/intel_frontbuffer.h" +#include "gem/i915_gem_lmem.h" +#include "gem/i915_gem_object_frontbuffer.h" +#include "gem/i915_gem_tiling.h" +#include "gt/intel_engine.h" +#include "gt/intel_engine_heartbeat.h" +#include "gt/intel_gt.h" +#include "gt/intel_gt_pm.h" +#include "gt/intel_gt_requests.h" +#include "gt/intel_tlb.h" #include "i915_drv.h" -#include "intel_ringbuffer.h" -#include "intel_frontbuffer.h" +#include "i915_gem_evict.h" +#include "i915_sw_fence_work.h" +#include "i915_trace.h" +#include "i915_vma.h" +#include "i915_vma_resource.h" -#include <drm/drm_gem.h> +static inline void assert_vma_held_evict(const struct i915_vma *vma) +{ + /* + * We may be forced to unbind when the vm is dead, to clean it up. + * This is the only exception to the requirement of the object lock + * being held. + */ + if (kref_read(&vma->vm->ref)) + assert_object_held_shared(vma->obj); +} + +static struct kmem_cache *slab_vmas; + +static struct i915_vma *i915_vma_alloc(void) +{ + return kmem_cache_zalloc(slab_vmas, GFP_KERNEL); +} + +static void i915_vma_free(struct i915_vma *vma) +{ + return kmem_cache_free(slab_vmas, vma); +} #if IS_ENABLED(CONFIG_DRM_I915_ERRLOG_GEM) && IS_ENABLED(CONFIG_DRM_DEBUG_MM) @@ -36,23 +76,19 @@ static void vma_print_allocator(struct i915_vma *vma, const char *reason) { - unsigned long entries[12]; - struct stack_trace trace = { - .entries = entries, - .max_entries = ARRAY_SIZE(entries), - }; char buf[512]; if (!vma->node.stack) { - DRM_DEBUG_DRIVER("vma.node [%08llx + %08llx] %s: unknown owner\n", - vma->node.start, vma->node.size, reason); + drm_dbg(vma->obj->base.dev, + "vma.node [%08llx + %08llx] %s: unknown owner\n", + vma->node.start, vma->node.size, reason); return; } - depot_fetch_stack(vma->node.stack, &trace); - snprint_stack_trace(buf, sizeof(buf), &trace, 0); - DRM_DEBUG_DRIVER("vma.node [%08llx + %08llx] %s: inserted at %s\n", - vma->node.start, vma->node.size, reason, buf); + stack_depot_snprint(vma->node.stack, buf, sizeof(buf), 0); + drm_dbg(vma->obj->base.dev, + "vma.node [%08llx + %08llx] %s: inserted at %s\n", + vma->node.start, vma->node.size, reason, buf); } #else @@ -63,97 +99,89 @@ static void vma_print_allocator(struct i915_vma *vma, const char *reason) #endif -struct i915_vma_active { - struct i915_gem_active base; - struct i915_vma *vma; - struct rb_node node; - u64 timeline; -}; - -static void -__i915_vma_retire(struct i915_vma *vma, struct i915_request *rq) +static inline struct i915_vma *active_to_vma(struct i915_active *ref) { - struct drm_i915_gem_object *obj = vma->obj; - - GEM_BUG_ON(!i915_vma_is_active(vma)); - if (--vma->active_count) - return; - - GEM_BUG_ON(!drm_mm_node_allocated(&vma->node)); - list_move_tail(&vma->vm_link, &vma->vm->inactive_list); + return container_of(ref, typeof(struct i915_vma), active); +} - GEM_BUG_ON(!i915_gem_object_is_active(obj)); - if (--obj->active_count) - return; +static int __i915_vma_active(struct i915_active *ref) +{ + struct i915_vma *vma = active_to_vma(ref); - /* Prune the shared fence arrays iff completely idle (inc. external) */ - if (reservation_object_trylock(obj->resv)) { - if (reservation_object_test_signaled_rcu(obj->resv, true)) - reservation_object_add_excl_fence(obj->resv, NULL); - reservation_object_unlock(obj->resv); - } + if (!i915_vma_tryget(vma)) + return -ENOENT; - /* Bump our place on the bound list to keep it roughly in LRU order - * so that we don't steal from recently used but inactive objects - * (unless we are forced to ofc!) + /* + * Exclude global GTT VMA from holding a GT wakeref + * while active, otherwise GPU never goes idle. */ - spin_lock(&rq->i915->mm.obj_lock); - if (obj->bind_count) - list_move_tail(&obj->mm.link, &rq->i915->mm.bound_list); - spin_unlock(&rq->i915->mm.obj_lock); - - obj->mm.dirty = true; /* be paranoid */ - - if (i915_gem_object_has_active_reference(obj)) { - i915_gem_object_clear_active_reference(obj); - i915_gem_object_put(obj); + if (!i915_vma_is_ggtt(vma)) { + /* + * Since we and our _retire() counterpart can be + * called asynchronously, storing a wakeref tracking + * handle inside struct i915_vma is not safe, and + * there is no other good place for that. Hence, + * use untracked variants of intel_gt_pm_get/put(). + */ + intel_gt_pm_get_untracked(vma->vm->gt); } + + return 0; } -static void -i915_vma_retire(struct i915_gem_active *base, struct i915_request *rq) +static void __i915_vma_retire(struct i915_active *ref) { - struct i915_vma_active *active = - container_of(base, typeof(*active), base); + struct i915_vma *vma = active_to_vma(ref); - __i915_vma_retire(active->vma, rq); -} + if (!i915_vma_is_ggtt(vma)) { + /* + * Since we can be called from atomic contexts, + * use an async variant of intel_gt_pm_put(). + */ + intel_gt_pm_put_async_untracked(vma->vm->gt); + } -static void -i915_vma_last_retire(struct i915_gem_active *base, struct i915_request *rq) -{ - __i915_vma_retire(container_of(base, struct i915_vma, last_active), rq); + i915_vma_put(vma); } static struct i915_vma * vma_create(struct drm_i915_gem_object *obj, struct i915_address_space *vm, - const struct i915_ggtt_view *view) + const struct i915_gtt_view *view) { + struct i915_vma *pos = ERR_PTR(-E2BIG); struct i915_vma *vma; struct rb_node *rb, **p; + int err; /* The aliasing_ppgtt should never be used directly! */ - GEM_BUG_ON(vm == &vm->i915->mm.aliasing_ppgtt->vm); + GEM_BUG_ON(vm == &vm->gt->ggtt->alias->vm); - vma = kmem_cache_zalloc(vm->i915->vmas, GFP_KERNEL); + vma = i915_vma_alloc(); if (vma == NULL) return ERR_PTR(-ENOMEM); - vma->active = RB_ROOT; - - init_request_active(&vma->last_active, i915_vma_last_retire); - init_request_active(&vma->last_fence, NULL); - vma->vm = vm; vma->ops = &vm->vma_ops; vma->obj = obj; - vma->resv = obj->resv; vma->size = obj->base.size; vma->display_alignment = I915_GTT_MIN_ALIGNMENT; - if (view && view->type != I915_GGTT_VIEW_NORMAL) { - vma->ggtt_view = *view; - if (view->type == I915_GGTT_VIEW_PARTIAL) { + i915_active_init(&vma->active, __i915_vma_active, __i915_vma_retire, 0); + + /* Declare ourselves safe for use inside shrinkers */ + if (IS_ENABLED(CONFIG_LOCKDEP)) { + fs_reclaim_acquire(GFP_KERNEL); + might_lock(&vma->active.mutex); + fs_reclaim_release(GFP_KERNEL); + } + + INIT_LIST_HEAD(&vma->closed_link); + INIT_LIST_HEAD(&vma->obj_link); + RB_CLEAR_NODE(&vma->obj_node); + + if (view && view->type != I915_GTT_VIEW_NORMAL) { + vma->gtt_view = *view; + if (view->type == I915_GTT_VIEW_PARTIAL) { GEM_BUG_ON(range_overflows_t(u64, view->partial.offset, view->partial.size, @@ -161,9 +189,12 @@ vma_create(struct drm_i915_gem_object *obj, vma->size = view->partial.size; vma->size <<= PAGE_SHIFT; GEM_BUG_ON(vma->size > obj->base.size); - } else if (view->type == I915_GGTT_VIEW_ROTATED) { + } else if (view->type == I915_GTT_VIEW_ROTATED) { vma->size = intel_rotation_info_size(&view->rotated); vma->size <<= PAGE_SHIFT; + } else if (view->type == I915_GTT_VIEW_REMAPPED) { + vma->size = intel_remapped_info_size(&view->remapped); + vma->size <<= PAGE_SHIFT; } } @@ -172,16 +203,26 @@ vma_create(struct drm_i915_gem_object *obj, GEM_BUG_ON(!IS_ALIGNED(vma->size, I915_GTT_PAGE_SIZE)); + err = mutex_lock_interruptible(&vm->mutex); + if (err) { + pos = ERR_PTR(err); + goto err_vma; + } + + vma->vm = vm; + list_add_tail(&vma->vm_link, &vm->unbound_list); + + spin_lock(&obj->vma.lock); if (i915_is_ggtt(vm)) { if (unlikely(overflows_type(vma->size, u32))) - goto err_vma; + goto err_unlock; vma->fence_size = i915_gem_fence_size(vm->i915, vma->size, i915_gem_object_get_tiling(obj), i915_gem_object_get_stride(obj)); if (unlikely(vma->fence_size < vma->size || /* overflow */ vma->fence_size > vm->total)) - goto err_vma; + goto err_unlock; GEM_BUG_ON(!IS_ALIGNED(vma->fence_size, I915_GTT_MIN_ALIGNMENT)); @@ -190,49 +231,66 @@ vma_create(struct drm_i915_gem_object *obj, i915_gem_object_get_stride(obj)); GEM_BUG_ON(!is_power_of_2(vma->fence_alignment)); - /* - * We put the GGTT vma at the start of the vma-list, followed - * by the ppGGTT vma. This allows us to break early when - * iterating over only the GGTT vma for an object, see - * for_each_ggtt_vma() - */ - vma->flags |= I915_VMA_GGTT; - list_add(&vma->obj_link, &obj->vma_list); - } else { - list_add_tail(&vma->obj_link, &obj->vma_list); + __set_bit(I915_VMA_GGTT_BIT, __i915_vma_flags(vma)); } rb = NULL; - p = &obj->vma_tree.rb_node; + p = &obj->vma.tree.rb_node; while (*p) { - struct i915_vma *pos; + long cmp; rb = *p; pos = rb_entry(rb, struct i915_vma, obj_node); - if (i915_vma_compare(pos, vm, view) < 0) + + /* + * If the view already exists in the tree, another thread + * already created a matching vma, so return the older instance + * and dispose of ours. + */ + cmp = i915_vma_compare(pos, vm, view); + if (cmp < 0) p = &rb->rb_right; - else + else if (cmp > 0) p = &rb->rb_left; + else + goto err_unlock; } rb_link_node(&vma->obj_node, rb, p); - rb_insert_color(&vma->obj_node, &obj->vma_tree); - list_add(&vma->vm_link, &vm->unbound_list); + rb_insert_color(&vma->obj_node, &obj->vma.tree); + + if (i915_vma_is_ggtt(vma)) + /* + * We put the GGTT vma at the start of the vma-list, followed + * by the ppGGTT vma. This allows us to break early when + * iterating over only the GGTT vma for an object, see + * for_each_ggtt_vma() + */ + list_add(&vma->obj_link, &obj->vma.list); + else + list_add_tail(&vma->obj_link, &obj->vma.list); + + spin_unlock(&obj->vma.lock); + mutex_unlock(&vm->mutex); return vma; +err_unlock: + spin_unlock(&obj->vma.lock); + list_del_init(&vma->vm_link); + mutex_unlock(&vm->mutex); err_vma: - kmem_cache_free(vm->i915->vmas, vma); - return ERR_PTR(-E2BIG); + i915_vma_free(vma); + return pos; } static struct i915_vma * -vma_lookup(struct drm_i915_gem_object *obj, +i915_vma_lookup(struct drm_i915_gem_object *obj, struct i915_address_space *vm, - const struct i915_ggtt_view *view) + const struct i915_gtt_view *view) { struct rb_node *rb; - rb = obj->vma_tree.rb_node; + rb = obj->vma.tree.rb_node; while (rb) { struct i915_vma *vma = rb_entry(rb, struct i915_vma, obj_node); long cmp; @@ -261,80 +319,264 @@ vma_lookup(struct drm_i915_gem_object *obj, * Once created, the VMA is kept until either the object is freed, or the * address space is closed. * - * Must be called with struct_mutex held. - * * Returns the vma, or an error pointer. */ struct i915_vma * i915_vma_instance(struct drm_i915_gem_object *obj, struct i915_address_space *vm, - const struct i915_ggtt_view *view) + const struct i915_gtt_view *view) { struct i915_vma *vma; - lockdep_assert_held(&obj->base.dev->struct_mutex); - GEM_BUG_ON(view && !i915_is_ggtt(vm)); - GEM_BUG_ON(vm->closed); + GEM_BUG_ON(view && !i915_is_ggtt_or_dpt(vm)); + GEM_BUG_ON(!kref_read(&vm->ref)); - vma = vma_lookup(obj, vm, view); - if (!vma) + spin_lock(&obj->vma.lock); + vma = i915_vma_lookup(obj, vm, view); + spin_unlock(&obj->vma.lock); + + /* vma_create() will resolve the race if another creates the vma */ + if (unlikely(!vma)) vma = vma_create(obj, vm, view); GEM_BUG_ON(!IS_ERR(vma) && i915_vma_compare(vma, vm, view)); - GEM_BUG_ON(!IS_ERR(vma) && vma_lookup(obj, vm, view) != vma); return vma; } +struct i915_vma_work { + struct dma_fence_work base; + struct i915_address_space *vm; + struct i915_vm_pt_stash stash; + struct i915_vma_resource *vma_res; + struct drm_i915_gem_object *obj; + struct i915_sw_dma_fence_cb cb; + unsigned int pat_index; + unsigned int flags; +}; + +static void __vma_bind(struct dma_fence_work *work) +{ + struct i915_vma_work *vw = container_of(work, typeof(*vw), base); + struct i915_vma_resource *vma_res = vw->vma_res; + + /* + * We are about the bind the object, which must mean we have already + * signaled the work to potentially clear/move the pages underneath. If + * something went wrong at that stage then the object should have + * unknown_state set, in which case we need to skip the bind. + */ + if (i915_gem_object_has_unknown_state(vw->obj)) + return; + + vma_res->ops->bind_vma(vma_res->vm, &vw->stash, + vma_res, vw->pat_index, vw->flags); +} + +static void __vma_release(struct dma_fence_work *work) +{ + struct i915_vma_work *vw = container_of(work, typeof(*vw), base); + + if (vw->obj) + i915_gem_object_put(vw->obj); + + i915_vm_free_pt_stash(vw->vm, &vw->stash); + if (vw->vma_res) + i915_vma_resource_put(vw->vma_res); +} + +static const struct dma_fence_work_ops bind_ops = { + .name = "bind", + .work = __vma_bind, + .release = __vma_release, +}; + +struct i915_vma_work *i915_vma_work(void) +{ + struct i915_vma_work *vw; + + vw = kzalloc(sizeof(*vw), GFP_KERNEL); + if (!vw) + return NULL; + + dma_fence_work_init(&vw->base, &bind_ops); + vw->base.dma.error = -EAGAIN; /* disable the worker by default */ + + return vw; +} + +int i915_vma_wait_for_bind(struct i915_vma *vma) +{ + int err = 0; + + if (rcu_access_pointer(vma->active.excl.fence)) { + struct dma_fence *fence; + + rcu_read_lock(); + fence = dma_fence_get_rcu_safe(&vma->active.excl.fence); + rcu_read_unlock(); + if (fence) { + err = dma_fence_wait(fence, true); + dma_fence_put(fence); + } + } + + return err; +} + +#if IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM) +static int i915_vma_verify_bind_complete(struct i915_vma *vma) +{ + struct dma_fence *fence = i915_active_fence_get(&vma->active.excl); + int err; + + if (!fence) + return 0; + + if (dma_fence_is_signaled(fence)) + err = fence->error; + else + err = -EBUSY; + + dma_fence_put(fence); + + return err; +} +#else +#define i915_vma_verify_bind_complete(_vma) 0 +#endif + +I915_SELFTEST_EXPORT void +i915_vma_resource_init_from_vma(struct i915_vma_resource *vma_res, + struct i915_vma *vma) +{ + struct drm_i915_gem_object *obj = vma->obj; + + i915_vma_resource_init(vma_res, vma->vm, vma->pages, &vma->page_sizes, + obj->mm.rsgt, i915_gem_object_is_readonly(obj), + i915_gem_object_is_lmem(obj), obj->mm.region, + vma->ops, vma->private, __i915_vma_offset(vma), + __i915_vma_size(vma), vma->size, vma->guard); +} + /** * i915_vma_bind - Sets up PTEs for an VMA in it's corresponding address space. * @vma: VMA to map - * @cache_level: mapping cache level + * @pat_index: PAT index to set in PTE * @flags: flags like global or local mapping + * @work: preallocated worker for allocating and binding the PTE + * @vma_res: pointer to a preallocated vma resource. The resource is either + * consumed or freed. * * DMA addresses are taken from the scatter-gather table of this object (or of * this VMA in case of non-default GGTT views) and PTE entries set up. * Note that DMA addresses are also the only part of the SG table we care about. */ -int i915_vma_bind(struct i915_vma *vma, enum i915_cache_level cache_level, - u32 flags) +int i915_vma_bind(struct i915_vma *vma, + unsigned int pat_index, + u32 flags, + struct i915_vma_work *work, + struct i915_vma_resource *vma_res) { u32 bind_flags; u32 vma_flags; int ret; + lockdep_assert_held(&vma->vm->mutex); GEM_BUG_ON(!drm_mm_node_allocated(&vma->node)); - GEM_BUG_ON(vma->size > vma->node.size); + GEM_BUG_ON(vma->size > i915_vma_size(vma)); if (GEM_DEBUG_WARN_ON(range_overflows(vma->node.start, vma->node.size, - vma->vm->total))) + vma->vm->total))) { + i915_vma_resource_free(vma_res); return -ENODEV; + } - if (GEM_DEBUG_WARN_ON(!flags)) + if (GEM_DEBUG_WARN_ON(!flags)) { + i915_vma_resource_free(vma_res); return -EINVAL; + } - bind_flags = 0; - if (flags & PIN_GLOBAL) - bind_flags |= I915_VMA_GLOBAL_BIND; - if (flags & PIN_USER) - bind_flags |= I915_VMA_LOCAL_BIND; + bind_flags = flags; + bind_flags &= I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND; - vma_flags = vma->flags & (I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND); - if (flags & PIN_UPDATE) - bind_flags |= vma_flags; - else - bind_flags &= ~vma_flags; - if (bind_flags == 0) + vma_flags = atomic_read(&vma->flags); + vma_flags &= I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND; + + bind_flags &= ~vma_flags; + if (bind_flags == 0) { + i915_vma_resource_free(vma_res); return 0; + } - GEM_BUG_ON(!vma->pages); + GEM_BUG_ON(!atomic_read(&vma->pages_count)); + + /* Wait for or await async unbinds touching our range */ + if (work && bind_flags & vma->vm->bind_async_flags) + ret = i915_vma_resource_bind_dep_await(vma->vm, + &work->base.chain, + vma->node.start, + vma->node.size, + true, + GFP_NOWAIT | + __GFP_RETRY_MAYFAIL | + __GFP_NOWARN); + else + ret = i915_vma_resource_bind_dep_sync(vma->vm, vma->node.start, + vma->node.size, true); + if (ret) { + i915_vma_resource_free(vma_res); + return ret; + } + if (vma->resource || !vma_res) { + /* Rebinding with an additional I915_VMA_*_BIND */ + GEM_WARN_ON(!vma_flags); + i915_vma_resource_free(vma_res); + } else { + i915_vma_resource_init_from_vma(vma_res, vma); + vma->resource = vma_res; + } trace_i915_vma_bind(vma, bind_flags); - ret = vma->ops->bind_vma(vma, cache_level, bind_flags); - if (ret) - return ret; + if (work && bind_flags & vma->vm->bind_async_flags) { + struct dma_fence *prev; + + work->vma_res = i915_vma_resource_get(vma->resource); + work->pat_index = pat_index; + work->flags = bind_flags; - vma->flags |= bind_flags; + /* + * Note we only want to chain up to the migration fence on + * the pages (not the object itself). As we don't track that, + * yet, we have to use the exclusive fence instead. + * + * Also note that we do not want to track the async vma as + * part of the obj->resv->excl_fence as it only affects + * execution and not content or object's backing store lifetime. + */ + prev = i915_active_set_exclusive(&vma->active, &work->base.dma); + if (prev) { + __i915_sw_fence_await_dma_fence(&work->base.chain, + prev, + &work->cb); + dma_fence_put(prev); + } + + work->base.dma.error = 0; /* enable the queue_work() */ + work->obj = i915_gem_object_get(vma->obj); + } else { + ret = i915_gem_object_wait_moving_fence(vma->obj, true); + if (ret) { + i915_vma_resource_free(vma->resource); + vma->resource = NULL; + + return ret; + } + vma->ops->bind_vma(vma->vm, NULL, vma->resource, pat_index, + bind_flags); + } + + atomic_or(bind_flags, &vma->flags); return 0; } @@ -343,29 +585,50 @@ void __iomem *i915_vma_pin_iomap(struct i915_vma *vma) void __iomem *ptr; int err; - /* Access through the GTT requires the device to be awake. */ - assert_rpm_wakelock_held(vma->vm->i915); - - lockdep_assert_held(&vma->vm->i915->drm.struct_mutex); - if (WARN_ON(!i915_vma_is_map_and_fenceable(vma))) { - err = -ENODEV; - goto err; - } + if (WARN_ON_ONCE(vma->obj->flags & I915_BO_ALLOC_GPU_ONLY)) + return IOMEM_ERR_PTR(-EINVAL); GEM_BUG_ON(!i915_vma_is_ggtt(vma)); - GEM_BUG_ON((vma->flags & I915_VMA_GLOBAL_BIND) == 0); + GEM_BUG_ON(!i915_vma_is_bound(vma, I915_VMA_GLOBAL_BIND)); + GEM_BUG_ON(i915_vma_verify_bind_complete(vma)); - ptr = vma->iomap; + ptr = READ_ONCE(vma->iomap); if (ptr == NULL) { - ptr = io_mapping_map_wc(&i915_vm_to_ggtt(vma->vm)->iomap, - vma->node.start, - vma->node.size); + /* + * TODO: consider just using i915_gem_object_pin_map() for lmem + * instead, which already supports mapping non-contiguous chunks + * of pages, that way we can also drop the + * I915_BO_ALLOC_CONTIGUOUS when allocating the object. + */ + if (i915_gem_object_is_lmem(vma->obj)) { + ptr = i915_gem_object_lmem_io_map(vma->obj, 0, + vma->obj->base.size); + } else if (i915_vma_is_map_and_fenceable(vma)) { + ptr = io_mapping_map_wc(&i915_vm_to_ggtt(vma->vm)->iomap, + i915_vma_offset(vma), + i915_vma_size(vma)); + } else { + ptr = (void __iomem *) + i915_gem_object_pin_map(vma->obj, I915_MAP_WC); + if (IS_ERR(ptr)) { + err = PTR_ERR(ptr); + goto err; + } + ptr = page_pack_bits(ptr, 1); + } + if (ptr == NULL) { err = -ENOMEM; goto err; } - vma->iomap = ptr; + if (unlikely(cmpxchg(&vma->iomap, NULL, ptr))) { + if (page_unmask_bits(ptr)) + __i915_gem_object_release_map(vma->obj); + else + io_mapping_unmap(ptr); + ptr = vma->iomap; + } } __i915_vma_pin(vma); @@ -375,30 +638,28 @@ void __iomem *i915_vma_pin_iomap(struct i915_vma *vma) goto err_unpin; i915_vma_set_ggtt_write(vma); - return ptr; + + /* NB Access through the GTT requires the device to be awake. */ + return page_mask_bits(ptr); err_unpin: __i915_vma_unpin(vma); err: - return IO_ERR_PTR(err); + return IOMEM_ERR_PTR(err); } void i915_vma_flush_writes(struct i915_vma *vma) { - if (!i915_vma_has_ggtt_write(vma)) - return; - - i915_gem_flush_ggtt_writes(vma->vm->i915); - - i915_vma_unset_ggtt_write(vma); + if (i915_vma_unset_ggtt_write(vma)) + intel_gt_flush_ggtt_writes(vma->vm->gt); } void i915_vma_unpin_iomap(struct i915_vma *vma) { - lockdep_assert_held(&vma->vm->i915->drm.struct_mutex); - GEM_BUG_ON(vma->iomap == NULL); + /* XXX We keep the mapping until __i915_vma_unbind()/evict() */ + i915_vma_flush_writes(vma); i915_vma_unpin_fence(vma); @@ -418,12 +679,11 @@ void i915_vma_unpin_and_release(struct i915_vma **p_vma, unsigned int flags) GEM_BUG_ON(!obj); i915_vma_unpin(vma); - i915_vma_close(vma); if (flags & I915_VMA_RELEASE_MAP) i915_gem_object_unpin_map(obj); - __i915_gem_object_release_unless_active(obj); + i915_gem_object_put(obj); } bool i915_vma_misplaced(const struct i915_vma *vma, @@ -432,22 +692,29 @@ bool i915_vma_misplaced(const struct i915_vma *vma, if (!drm_mm_node_allocated(&vma->node)) return false; - if (vma->node.size < size) + if (test_bit(I915_VMA_ERROR_BIT, __i915_vma_flags(vma))) + return true; + + if (i915_vma_size(vma) < size) return true; GEM_BUG_ON(alignment && !is_power_of_2(alignment)); - if (alignment && !IS_ALIGNED(vma->node.start, alignment)) + if (alignment && !IS_ALIGNED(i915_vma_offset(vma), alignment)) return true; if (flags & PIN_MAPPABLE && !i915_vma_is_map_and_fenceable(vma)) return true; if (flags & PIN_OFFSET_BIAS && - vma->node.start < (flags & PIN_OFFSET_MASK)) + i915_vma_offset(vma) < (flags & PIN_OFFSET_MASK)) return true; if (flags & PIN_OFFSET_FIXED && - vma->node.start != (flags & PIN_OFFSET_MASK)) + i915_vma_offset(vma) != (flags & PIN_OFFSET_MASK)) + return true; + + if (flags & PIN_OFFSET_GUARD && + vma->guard < (flags & PIN_OFFSET_MASK)) return true; return false; @@ -460,30 +727,19 @@ void __i915_vma_set_map_and_fenceable(struct i915_vma *vma) GEM_BUG_ON(!i915_vma_is_ggtt(vma)); GEM_BUG_ON(!vma->fence_size); - /* - * Explicitly disable for rotated VMA since the display does not - * need the fence and the VMA is not accessible to other users. - */ - if (vma->ggtt_view.type == I915_GGTT_VIEW_ROTATED) - return; - - fenceable = (vma->node.size >= vma->fence_size && - IS_ALIGNED(vma->node.start, vma->fence_alignment)); + fenceable = (i915_vma_size(vma) >= vma->fence_size && + IS_ALIGNED(i915_vma_offset(vma), vma->fence_alignment)); - mappable = vma->node.start + vma->fence_size <= i915_vm_to_ggtt(vma->vm)->mappable_end; + mappable = i915_ggtt_offset(vma) + vma->fence_size <= + i915_vm_to_ggtt(vma->vm)->mappable_end; if (mappable && fenceable) - vma->flags |= I915_VMA_CAN_FENCE; + set_bit(I915_VMA_CAN_FENCE_BIT, __i915_vma_flags(vma)); else - vma->flags &= ~I915_VMA_CAN_FENCE; + clear_bit(I915_VMA_CAN_FENCE_BIT, __i915_vma_flags(vma)); } -static bool color_differs(struct drm_mm_node *node, unsigned long color) -{ - return node->allocated && node->color != color; -} - -bool i915_gem_valid_gtt_space(struct i915_vma *vma, unsigned long cache_level) +bool i915_gem_valid_gtt_space(struct i915_vma *vma, unsigned long color) { struct drm_mm_node *node = &vma->node; struct drm_mm_node *other; @@ -495,7 +751,7 @@ bool i915_gem_valid_gtt_space(struct i915_vma *vma, unsigned long cache_level) * these constraints apply and set the drm_mm.color_adjust * appropriately. */ - if (vma->vm->mm.color_adjust == NULL) + if (!i915_vm_has_cache_coloring(vma->vm)) return true; /* Only valid to be called on an already inserted vma */ @@ -503,56 +759,47 @@ bool i915_gem_valid_gtt_space(struct i915_vma *vma, unsigned long cache_level) GEM_BUG_ON(list_empty(&node->node_list)); other = list_prev_entry(node, node_list); - if (color_differs(other, cache_level) && !drm_mm_hole_follows(other)) + if (i915_node_color_differs(other, color) && + !drm_mm_hole_follows(other)) return false; other = list_next_entry(node, node_list); - if (color_differs(other, cache_level) && !drm_mm_hole_follows(node)) + if (i915_node_color_differs(other, color) && + !drm_mm_hole_follows(node)) return false; return true; } -static void assert_bind_count(const struct drm_i915_gem_object *obj) -{ - /* - * Combine the assertion that the object is bound and that we have - * pinned its pages. But we should never have bound the object - * more than we have pinned its pages. (For complete accuracy, we - * assume that no else is pinning the pages, but as a rough assertion - * that we will not run into problems later, this will do!) - */ - GEM_BUG_ON(atomic_read(&obj->mm.pages_pin_count) < obj->bind_count); -} - /** * i915_vma_insert - finds a slot for the vma in its address space * @vma: the vma + * @ww: An optional struct i915_gem_ww_ctx * @size: requested size in bytes (can be larger than the VMA) * @alignment: required alignment * @flags: mask of PIN_* flags to use * * First we try to allocate some free space that meets the requirements for - * the VMA. Failiing that, if the flags permit, it will evict an old VMA, - * preferrably the oldest idle entry to make room for the new VMA. + * the VMA. Failing that, if the flags permit, it will evict an old VMA, + * preferably the oldest idle entry to make room for the new VMA. * * Returns: * 0 on success, negative error code otherwise. */ static int -i915_vma_insert(struct i915_vma *vma, u64 size, u64 alignment, u64 flags) +i915_vma_insert(struct i915_vma *vma, struct i915_gem_ww_ctx *ww, + u64 size, u64 alignment, u64 flags) { - struct drm_i915_private *dev_priv = vma->vm->i915; - unsigned int cache_level; + unsigned long color, guard; u64 start, end; int ret; - GEM_BUG_ON(i915_vma_is_closed(vma)); - GEM_BUG_ON(vma->flags & (I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND)); + GEM_BUG_ON(i915_vma_is_bound(vma, I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND)); GEM_BUG_ON(drm_mm_node_allocated(&vma->node)); + GEM_BUG_ON(hweight64(flags & (PIN_OFFSET_GUARD | PIN_OFFSET_FIXED | PIN_OFFSET_BIAS)) > 1); size = max(size, vma->size); - alignment = max(alignment, vma->display_alignment); + alignment = max_t(typeof(alignment), alignment, vma->display_alignment); if (flags & PIN_MAPPABLE) { size = max_t(typeof(size), size, vma->fence_size); alignment = max_t(typeof(alignment), @@ -563,67 +810,80 @@ i915_vma_insert(struct i915_vma *vma, u64 size, u64 alignment, u64 flags) GEM_BUG_ON(!IS_ALIGNED(alignment, I915_GTT_MIN_ALIGNMENT)); GEM_BUG_ON(!is_power_of_2(alignment)); + guard = vma->guard; /* retain guard across rebinds */ + if (flags & PIN_OFFSET_GUARD) { + GEM_BUG_ON(overflows_type(flags & PIN_OFFSET_MASK, u32)); + guard = max_t(u32, guard, flags & PIN_OFFSET_MASK); + } + /* + * As we align the node upon insertion, but the hardware gets + * node.start + guard, the easiest way to make that work is + * to make the guard a multiple of the alignment size. + */ + guard = ALIGN(guard, alignment); + start = flags & PIN_OFFSET_BIAS ? flags & PIN_OFFSET_MASK : 0; GEM_BUG_ON(!IS_ALIGNED(start, I915_GTT_PAGE_SIZE)); end = vma->vm->total; if (flags & PIN_MAPPABLE) - end = min_t(u64, end, dev_priv->ggtt.mappable_end); + end = min_t(u64, end, i915_vm_to_ggtt(vma->vm)->mappable_end); if (flags & PIN_ZONE_4G) end = min_t(u64, end, (1ULL << 32) - I915_GTT_PAGE_SIZE); GEM_BUG_ON(!IS_ALIGNED(end, I915_GTT_PAGE_SIZE)); - /* If binding the object/GGTT view requires more space than the entire + alignment = max(alignment, i915_vm_obj_min_alignment(vma->vm, vma->obj)); + + /* + * If binding the object/GGTT view requires more space than the entire * aperture has, reject it early before evicting everything in a vain * attempt to find space. */ - if (size > end) { - DRM_DEBUG("Attempting to bind an object larger than the aperture: request=%llu > %s aperture=%llu\n", - size, flags & PIN_MAPPABLE ? "mappable" : "total", - end); + if (size > end - 2 * guard) { + drm_dbg(vma->obj->base.dev, + "Attempting to bind an object larger than the aperture: request=%llu > %s aperture=%llu\n", + size, flags & PIN_MAPPABLE ? "mappable" : "total", end); return -ENOSPC; } - if (vma->obj) { - ret = i915_gem_object_pin_pages(vma->obj); - if (ret) - return ret; - - cache_level = vma->obj->cache_level; - } else { - cache_level = 0; - } - - GEM_BUG_ON(vma->pages); + color = 0; - ret = vma->ops->set_pages(vma); - if (ret) - goto err_unpin; + if (i915_vm_has_cache_coloring(vma->vm)) + color = vma->obj->pat_index; if (flags & PIN_OFFSET_FIXED) { u64 offset = flags & PIN_OFFSET_MASK; if (!IS_ALIGNED(offset, alignment) || - range_overflows(offset, size, end)) { - ret = -EINVAL; - goto err_clear; - } + range_overflows(offset, size, end)) + return -EINVAL; + /* + * The caller knows not of the guard added by others and + * requests for the offset of the start of its buffer + * to be fixed, which may not be the same as the position + * of the vma->node due to the guard pages. + */ + if (offset < guard || offset + size > end - guard) + return -ENOSPC; - ret = i915_gem_gtt_reserve(vma->vm, &vma->node, - size, offset, cache_level, - flags); + ret = i915_gem_gtt_reserve(vma->vm, ww, &vma->node, + size + 2 * guard, + offset - guard, + color, flags); if (ret) - goto err_clear; + return ret; } else { + size += 2 * guard; /* * We only support huge gtt pages through the 48b PPGTT, * however we also don't want to force any alignment for * objects which need to be tightly packed into the low 32bits. * * Note that we assume that GGTT are limited to 4GiB for the - * forseeable future. See also i915_ggtt_offset(). + * foreseeable future. See also i915_ggtt_offset(). */ if (upper_32_bits(end - 1) && - vma->page_sizes.sg > I915_GTT_PAGE_SIZE) { + vma->page_sizes.sg > I915_GTT_PAGE_SIZE && + !HAS_64K_PAGES(vma->vm->i915)) { /* * We can't mix 64K and 4K PTEs in the same page-table * (2M block), and so to avoid the ugliness and @@ -647,128 +907,830 @@ i915_vma_insert(struct i915_vma *vma, u64 size, u64 alignment, u64 flags) size = round_up(size, I915_GTT_PAGE_SIZE_2M); } - ret = i915_gem_gtt_insert(vma->vm, &vma->node, - size, alignment, cache_level, + ret = i915_gem_gtt_insert(vma->vm, ww, &vma->node, + size, alignment, color, start, end, flags); if (ret) - goto err_clear; + return ret; GEM_BUG_ON(vma->node.start < start); GEM_BUG_ON(vma->node.start + vma->node.size > end); } GEM_BUG_ON(!drm_mm_node_allocated(&vma->node)); - GEM_BUG_ON(!i915_gem_valid_gtt_space(vma, cache_level)); + GEM_BUG_ON(!i915_gem_valid_gtt_space(vma, color)); - list_move_tail(&vma->vm_link, &vma->vm->inactive_list); + list_move_tail(&vma->vm_link, &vma->vm->bound_list); + vma->guard = guard; - if (vma->obj) { - struct drm_i915_gem_object *obj = vma->obj; + return 0; +} - spin_lock(&dev_priv->mm.obj_lock); - list_move_tail(&obj->mm.link, &dev_priv->mm.bound_list); - obj->bind_count++; - spin_unlock(&dev_priv->mm.obj_lock); +static void +i915_vma_detach(struct i915_vma *vma) +{ + GEM_BUG_ON(!drm_mm_node_allocated(&vma->node)); + GEM_BUG_ON(i915_vma_is_bound(vma, I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND)); + + /* + * And finally now the object is completely decoupled from this + * vma, we can drop its hold on the backing storage and allow + * it to be reaped by the shrinker. + */ + list_move_tail(&vma->vm_link, &vma->vm->unbound_list); +} - assert_bind_count(obj); +static bool try_qad_pin(struct i915_vma *vma, unsigned int flags) +{ + unsigned int bound; + + bound = atomic_read(&vma->flags); + + if (flags & PIN_VALIDATE) { + flags &= I915_VMA_BIND_MASK; + + return (flags & bound) == flags; } - return 0; + /* with the lock mandatory for unbind, we don't race here */ + flags &= I915_VMA_BIND_MASK; + do { + if (unlikely(flags & ~bound)) + return false; -err_clear: - vma->ops->clear_pages(vma); -err_unpin: - if (vma->obj) - i915_gem_object_unpin_pages(vma->obj); - return ret; + if (unlikely(bound & (I915_VMA_OVERFLOW | I915_VMA_ERROR))) + return false; + + GEM_BUG_ON(((bound + 1) & I915_VMA_PIN_MASK) == 0); + } while (!atomic_try_cmpxchg(&vma->flags, &bound, bound + 1)); + + return true; } -static void -i915_vma_remove(struct i915_vma *vma) +static struct scatterlist * +rotate_pages(struct drm_i915_gem_object *obj, unsigned int offset, + unsigned int width, unsigned int height, + unsigned int src_stride, unsigned int dst_stride, + struct sg_table *st, struct scatterlist *sg) { - struct drm_i915_private *i915 = vma->vm->i915; + unsigned int column, row; + pgoff_t src_idx; - GEM_BUG_ON(!drm_mm_node_allocated(&vma->node)); - GEM_BUG_ON(vma->flags & (I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND)); + for (column = 0; column < width; column++) { + unsigned int left; - vma->ops->clear_pages(vma); + src_idx = src_stride * (height - 1) + column + offset; + for (row = 0; row < height; row++) { + st->nents++; + /* + * We don't need the pages, but need to initialize + * the entries so the sg list can be happily traversed. + * The only thing we need are DMA addresses. + */ + sg_set_page(sg, NULL, I915_GTT_PAGE_SIZE, 0); + sg_dma_address(sg) = + i915_gem_object_get_dma_address(obj, src_idx); + sg_dma_len(sg) = I915_GTT_PAGE_SIZE; + sg = sg_next(sg); + src_idx -= src_stride; + } - drm_mm_remove_node(&vma->node); - list_move_tail(&vma->vm_link, &vma->vm->unbound_list); + left = (dst_stride - height) * I915_GTT_PAGE_SIZE; - /* - * Since the unbound list is global, only move to that list if - * no more VMAs exist. - */ - if (vma->obj) { - struct drm_i915_gem_object *obj = vma->obj; + if (!left) + continue; - spin_lock(&i915->mm.obj_lock); - if (--obj->bind_count == 0) - list_move_tail(&obj->mm.link, &i915->mm.unbound_list); - spin_unlock(&i915->mm.obj_lock); + st->nents++; /* - * And finally now the object is completely decoupled from this - * vma, we can drop its hold on the backing storage and allow - * it to be reaped by the shrinker. + * The DE ignores the PTEs for the padding tiles, the sg entry + * here is just a convenience to indicate how many padding PTEs + * to insert at this spot. */ - i915_gem_object_unpin_pages(obj); - assert_bind_count(obj); + sg_set_page(sg, NULL, left, 0); + sg_dma_address(sg) = 0; + sg_dma_len(sg) = left; + sg = sg_next(sg); } + + return sg; } -int __i915_vma_do_pin(struct i915_vma *vma, - u64 size, u64 alignment, u64 flags) +static noinline struct sg_table * +intel_rotate_pages(struct intel_rotation_info *rot_info, + struct drm_i915_gem_object *obj) { - const unsigned int bound = vma->flags; - int ret; + unsigned int size = intel_rotation_info_size(rot_info); + struct drm_i915_private *i915 = to_i915(obj->base.dev); + struct sg_table *st; + struct scatterlist *sg; + int ret = -ENOMEM; + int i; + + /* Allocate target SG list. */ + st = kmalloc(sizeof(*st), GFP_KERNEL); + if (!st) + goto err_st_alloc; + + ret = sg_alloc_table(st, size, GFP_KERNEL); + if (ret) + goto err_sg_alloc; + + st->nents = 0; + sg = st->sgl; + + for (i = 0 ; i < ARRAY_SIZE(rot_info->plane); i++) + sg = rotate_pages(obj, rot_info->plane[i].offset, + rot_info->plane[i].width, rot_info->plane[i].height, + rot_info->plane[i].src_stride, + rot_info->plane[i].dst_stride, + st, sg); + + return st; + +err_sg_alloc: + kfree(st); +err_st_alloc: + + drm_dbg(&i915->drm, "Failed to create rotated mapping for object size %zu! (%ux%u tiles, %u pages)\n", + obj->base.size, rot_info->plane[0].width, + rot_info->plane[0].height, size); + + return ERR_PTR(ret); +} + +static struct scatterlist * +add_padding_pages(unsigned int count, + struct sg_table *st, struct scatterlist *sg) +{ + st->nents++; + + /* + * The DE ignores the PTEs for the padding tiles, the sg entry + * here is just a convenience to indicate how many padding PTEs + * to insert at this spot. + */ + sg_set_page(sg, NULL, count * I915_GTT_PAGE_SIZE, 0); + sg_dma_address(sg) = 0; + sg_dma_len(sg) = count * I915_GTT_PAGE_SIZE; + sg = sg_next(sg); + + return sg; +} + +static struct scatterlist * +remap_tiled_color_plane_pages(struct drm_i915_gem_object *obj, + unsigned long offset, unsigned int alignment_pad, + unsigned int width, unsigned int height, + unsigned int src_stride, unsigned int dst_stride, + struct sg_table *st, struct scatterlist *sg, + unsigned int *gtt_offset) +{ + unsigned int row; + + if (!width || !height) + return sg; + + if (alignment_pad) + sg = add_padding_pages(alignment_pad, st, sg); + + for (row = 0; row < height; row++) { + unsigned int left = width * I915_GTT_PAGE_SIZE; + + while (left) { + dma_addr_t addr; + unsigned int length; + + /* + * We don't need the pages, but need to initialize + * the entries so the sg list can be happily traversed. + * The only thing we need are DMA addresses. + */ + + addr = i915_gem_object_get_dma_address_len(obj, offset, &length); + + length = min(left, length); + + st->nents++; + + sg_set_page(sg, NULL, length, 0); + sg_dma_address(sg) = addr; + sg_dma_len(sg) = length; + sg = sg_next(sg); + + offset += length / I915_GTT_PAGE_SIZE; + left -= length; + } + + offset += src_stride - width; + + left = (dst_stride - width) * I915_GTT_PAGE_SIZE; + + if (!left) + continue; + + sg = add_padding_pages(left >> PAGE_SHIFT, st, sg); + } + + *gtt_offset += alignment_pad + dst_stride * height; + + return sg; +} + +static struct scatterlist * +remap_contiguous_pages(struct drm_i915_gem_object *obj, + pgoff_t obj_offset, + unsigned int count, + struct sg_table *st, struct scatterlist *sg) +{ + struct scatterlist *iter; + unsigned int offset; + + iter = i915_gem_object_get_sg_dma(obj, obj_offset, &offset); + GEM_BUG_ON(!iter); + + do { + unsigned int len; + + len = min(sg_dma_len(iter) - (offset << PAGE_SHIFT), + count << PAGE_SHIFT); + sg_set_page(sg, NULL, len, 0); + sg_dma_address(sg) = + sg_dma_address(iter) + (offset << PAGE_SHIFT); + sg_dma_len(sg) = len; + + st->nents++; + count -= len >> PAGE_SHIFT; + if (count == 0) + return sg; + + sg = __sg_next(sg); + iter = __sg_next(iter); + offset = 0; + } while (1); +} + +static struct scatterlist * +remap_linear_color_plane_pages(struct drm_i915_gem_object *obj, + pgoff_t obj_offset, unsigned int alignment_pad, + unsigned int size, + struct sg_table *st, struct scatterlist *sg, + unsigned int *gtt_offset) +{ + if (!size) + return sg; + + if (alignment_pad) + sg = add_padding_pages(alignment_pad, st, sg); - lockdep_assert_held(&vma->vm->i915->drm.struct_mutex); - GEM_BUG_ON((flags & (PIN_GLOBAL | PIN_USER)) == 0); - GEM_BUG_ON((flags & PIN_GLOBAL) && !i915_vma_is_ggtt(vma)); + sg = remap_contiguous_pages(obj, obj_offset, size, st, sg); + sg = sg_next(sg); - if (WARN_ON(bound & I915_VMA_PIN_OVERFLOW)) { - ret = -EBUSY; + *gtt_offset += alignment_pad + size; + + return sg; +} + +static struct scatterlist * +remap_color_plane_pages(const struct intel_remapped_info *rem_info, + struct drm_i915_gem_object *obj, + int color_plane, + struct sg_table *st, struct scatterlist *sg, + unsigned int *gtt_offset) +{ + unsigned int alignment_pad = 0; + + if (rem_info->plane_alignment) + alignment_pad = ALIGN(*gtt_offset, rem_info->plane_alignment) - *gtt_offset; + + if (rem_info->plane[color_plane].linear) + sg = remap_linear_color_plane_pages(obj, + rem_info->plane[color_plane].offset, + alignment_pad, + rem_info->plane[color_plane].size, + st, sg, + gtt_offset); + + else + sg = remap_tiled_color_plane_pages(obj, + rem_info->plane[color_plane].offset, + alignment_pad, + rem_info->plane[color_plane].width, + rem_info->plane[color_plane].height, + rem_info->plane[color_plane].src_stride, + rem_info->plane[color_plane].dst_stride, + st, sg, + gtt_offset); + + return sg; +} + +static noinline struct sg_table * +intel_remap_pages(struct intel_remapped_info *rem_info, + struct drm_i915_gem_object *obj) +{ + unsigned int size = intel_remapped_info_size(rem_info); + struct drm_i915_private *i915 = to_i915(obj->base.dev); + struct sg_table *st; + struct scatterlist *sg; + unsigned int gtt_offset = 0; + int ret = -ENOMEM; + int i; + + /* Allocate target SG list. */ + st = kmalloc(sizeof(*st), GFP_KERNEL); + if (!st) + goto err_st_alloc; + + ret = sg_alloc_table(st, size, GFP_KERNEL); + if (ret) + goto err_sg_alloc; + + st->nents = 0; + sg = st->sgl; + + for (i = 0 ; i < ARRAY_SIZE(rem_info->plane); i++) + sg = remap_color_plane_pages(rem_info, obj, i, st, sg, >t_offset); + + i915_sg_trim(st); + + return st; + +err_sg_alloc: + kfree(st); +err_st_alloc: + + drm_dbg(&i915->drm, "Failed to create remapped mapping for object size %zu! (%ux%u tiles, %u pages)\n", + obj->base.size, rem_info->plane[0].width, + rem_info->plane[0].height, size); + + return ERR_PTR(ret); +} + +static noinline struct sg_table * +intel_partial_pages(const struct i915_gtt_view *view, + struct drm_i915_gem_object *obj) +{ + struct sg_table *st; + struct scatterlist *sg; + unsigned int count = view->partial.size; + int ret = -ENOMEM; + + st = kmalloc(sizeof(*st), GFP_KERNEL); + if (!st) + goto err_st_alloc; + + ret = sg_alloc_table(st, count, GFP_KERNEL); + if (ret) + goto err_sg_alloc; + + st->nents = 0; + + sg = remap_contiguous_pages(obj, view->partial.offset, count, st, st->sgl); + + sg_mark_end(sg); + i915_sg_trim(st); /* Drop any unused tail entries. */ + + return st; + +err_sg_alloc: + kfree(st); +err_st_alloc: + return ERR_PTR(ret); +} + +static int +__i915_vma_get_pages(struct i915_vma *vma) +{ + struct sg_table *pages; + + /* + * The vma->pages are only valid within the lifespan of the borrowed + * obj->mm.pages. When the obj->mm.pages sg_table is regenerated, so + * must be the vma->pages. A simple rule is that vma->pages must only + * be accessed when the obj->mm.pages are pinned. + */ + GEM_BUG_ON(!i915_gem_object_has_pinned_pages(vma->obj)); + + switch (vma->gtt_view.type) { + default: + GEM_BUG_ON(vma->gtt_view.type); + fallthrough; + case I915_GTT_VIEW_NORMAL: + pages = vma->obj->mm.pages; + break; + + case I915_GTT_VIEW_ROTATED: + pages = + intel_rotate_pages(&vma->gtt_view.rotated, vma->obj); + break; + + case I915_GTT_VIEW_REMAPPED: + pages = + intel_remap_pages(&vma->gtt_view.remapped, vma->obj); + break; + + case I915_GTT_VIEW_PARTIAL: + pages = intel_partial_pages(&vma->gtt_view, vma->obj); + break; + } + + if (IS_ERR(pages)) { + drm_err(&vma->vm->i915->drm, + "Failed to get pages for VMA view type %u (%ld)!\n", + vma->gtt_view.type, PTR_ERR(pages)); + return PTR_ERR(pages); + } + + vma->pages = pages; + + return 0; +} + +I915_SELFTEST_EXPORT int i915_vma_get_pages(struct i915_vma *vma) +{ + int err; + + if (atomic_add_unless(&vma->pages_count, 1, 0)) + return 0; + + err = i915_gem_object_pin_pages(vma->obj); + if (err) + return err; + + err = __i915_vma_get_pages(vma); + if (err) goto err_unpin; + + vma->page_sizes = vma->obj->mm.page_sizes; + atomic_inc(&vma->pages_count); + + return 0; + +err_unpin: + __i915_gem_object_unpin_pages(vma->obj); + + return err; +} + +void vma_invalidate_tlb(struct i915_address_space *vm, u32 *tlb) +{ + struct intel_gt *gt; + int id; + + if (!tlb) + return; + + /* + * Before we release the pages that were bound by this vma, we + * must invalidate all the TLBs that may still have a reference + * back to our physical address. It only needs to be done once, + * so after updating the PTE to point away from the pages, record + * the most recent TLB invalidation seqno, and if we have not yet + * flushed the TLBs upon release, perform a full invalidation. + */ + for_each_gt(gt, vm->i915, id) + WRITE_ONCE(tlb[id], + intel_gt_next_invalidate_tlb_full(gt)); +} + +static void __vma_put_pages(struct i915_vma *vma, unsigned int count) +{ + /* We allocate under vma_get_pages, so beware the shrinker */ + GEM_BUG_ON(atomic_read(&vma->pages_count) < count); + + if (atomic_sub_return(count, &vma->pages_count) == 0) { + if (vma->pages != vma->obj->mm.pages) { + sg_free_table(vma->pages); + kfree(vma->pages); + } + vma->pages = NULL; + + i915_gem_object_unpin_pages(vma->obj); } +} - if ((bound & I915_VMA_BIND_MASK) == 0) { - ret = i915_vma_insert(vma, size, alignment, flags); - if (ret) - goto err_unpin; +I915_SELFTEST_EXPORT void i915_vma_put_pages(struct i915_vma *vma) +{ + if (atomic_add_unless(&vma->pages_count, -1, 1)) + return; + + __vma_put_pages(vma, 1); +} + +static void vma_unbind_pages(struct i915_vma *vma) +{ + unsigned int count; + + lockdep_assert_held(&vma->vm->mutex); + + /* The upper portion of pages_count is the number of bindings */ + count = atomic_read(&vma->pages_count); + count >>= I915_VMA_PAGES_BIAS; + GEM_BUG_ON(!count); + + __vma_put_pages(vma, count | count << I915_VMA_PAGES_BIAS); +} + +int i915_vma_pin_ww(struct i915_vma *vma, struct i915_gem_ww_ctx *ww, + u64 size, u64 alignment, u64 flags) +{ + struct i915_vma_work *work = NULL; + struct dma_fence *moving = NULL; + struct i915_vma_resource *vma_res = NULL; + intel_wakeref_t wakeref; + unsigned int bound; + int err; + + assert_vma_held(vma); + GEM_BUG_ON(!ww); + + BUILD_BUG_ON(PIN_GLOBAL != I915_VMA_GLOBAL_BIND); + BUILD_BUG_ON(PIN_USER != I915_VMA_LOCAL_BIND); + + GEM_BUG_ON(!(flags & (PIN_USER | PIN_GLOBAL))); + + /* First try and grab the pin without rebinding the vma */ + if (try_qad_pin(vma, flags)) + return 0; + + err = i915_vma_get_pages(vma); + if (err) + return err; + + /* + * In case of a global GTT, we must hold a runtime-pm wakeref + * while global PTEs are updated. In other cases, we hold + * the rpm reference while the VMA is active. Since runtime + * resume may require allocations, which are forbidden inside + * vm->mutex, get the first rpm wakeref outside of the mutex. + */ + wakeref = intel_runtime_pm_get(&vma->vm->i915->runtime_pm); + + if (flags & vma->vm->bind_async_flags) { + /* lock VM */ + err = i915_vm_lock_objects(vma->vm, ww); + if (err) + goto err_rpm; + + work = i915_vma_work(); + if (!work) { + err = -ENOMEM; + goto err_rpm; + } + + work->vm = vma->vm; + + err = i915_gem_object_get_moving_fence(vma->obj, &moving); + if (err) + goto err_rpm; + + dma_fence_work_chain(&work->base, moving); + + /* Allocate enough page directories to used PTE */ + if (vma->vm->allocate_va_range) { + err = i915_vm_alloc_pt_stash(vma->vm, + &work->stash, + vma->size); + if (err) + goto err_fence; + + err = i915_vm_map_pt_stash(vma->vm, &work->stash); + if (err) + goto err_fence; + } } - GEM_BUG_ON(!drm_mm_node_allocated(&vma->node)); - ret = i915_vma_bind(vma, vma->obj ? vma->obj->cache_level : 0, flags); - if (ret) - goto err_remove; + vma_res = i915_vma_resource_alloc(); + if (IS_ERR(vma_res)) { + err = PTR_ERR(vma_res); + goto err_fence; + } + + /* + * Differentiate between user/kernel vma inside the aliasing-ppgtt. + * + * We conflate the Global GTT with the user's vma when using the + * aliasing-ppgtt, but it is still vitally important to try and + * keep the use cases distinct. For example, userptr objects are + * not allowed inside the Global GTT as that will cause lock + * inversions when we have to evict them the mmu_notifier callbacks - + * but they are allowed to be part of the user ppGTT which can never + * be mapped. As such we try to give the distinct users of the same + * mutex, distinct lockclasses [equivalent to how we keep i915_ggtt + * and i915_ppgtt separate]. + * + * NB this may cause us to mask real lock inversions -- while the + * code is safe today, lockdep may not be able to spot future + * transgressions. + */ + err = mutex_lock_interruptible_nested(&vma->vm->mutex, + !(flags & PIN_GLOBAL)); + if (err) + goto err_vma_res; + + /* No more allocations allowed now we hold vm->mutex */ - GEM_BUG_ON((vma->flags & I915_VMA_BIND_MASK) == 0); + if (unlikely(i915_vma_is_closed(vma))) { + err = -ENOENT; + goto err_unlock; + } + + bound = atomic_read(&vma->flags); + if (unlikely(bound & I915_VMA_ERROR)) { + err = -ENOMEM; + goto err_unlock; + } - if ((bound ^ vma->flags) & I915_VMA_GLOBAL_BIND) - __i915_vma_set_map_and_fenceable(vma); + if (unlikely(!((bound + 1) & I915_VMA_PIN_MASK))) { + err = -EAGAIN; /* pins are meant to be fairly temporary */ + goto err_unlock; + } + + if (unlikely(!(flags & ~bound & I915_VMA_BIND_MASK))) { + if (!(flags & PIN_VALIDATE)) + __i915_vma_pin(vma); + goto err_unlock; + } + + err = i915_active_acquire(&vma->active); + if (err) + goto err_unlock; + + if (!(bound & I915_VMA_BIND_MASK)) { + err = i915_vma_insert(vma, ww, size, alignment, flags); + if (err) + goto err_active; + + if (i915_is_ggtt(vma->vm)) + __i915_vma_set_map_and_fenceable(vma); + } + GEM_BUG_ON(!vma->pages); + err = i915_vma_bind(vma, + vma->obj->pat_index, + flags, work, vma_res); + vma_res = NULL; + if (err) + goto err_remove; + + /* There should only be at most 2 active bindings (user, global) */ + GEM_BUG_ON(bound + I915_VMA_PAGES_ACTIVE < bound); + atomic_add(I915_VMA_PAGES_ACTIVE, &vma->pages_count); + list_move_tail(&vma->vm_link, &vma->vm->bound_list); + + if (!(flags & PIN_VALIDATE)) { + __i915_vma_pin(vma); + GEM_BUG_ON(!i915_vma_is_pinned(vma)); + } + GEM_BUG_ON(!i915_vma_is_bound(vma, flags)); GEM_BUG_ON(i915_vma_misplaced(vma, size, alignment, flags)); - return 0; err_remove: - if ((bound & I915_VMA_BIND_MASK) == 0) { - i915_vma_remove(vma); - GEM_BUG_ON(vma->pages); - GEM_BUG_ON(vma->flags & I915_VMA_BIND_MASK); + if (!i915_vma_is_bound(vma, I915_VMA_BIND_MASK)) { + i915_vma_detach(vma); + drm_mm_remove_node(&vma->node); } -err_unpin: - __i915_vma_unpin(vma); - return ret; +err_active: + i915_active_release(&vma->active); +err_unlock: + mutex_unlock(&vma->vm->mutex); +err_vma_res: + i915_vma_resource_free(vma_res); +err_fence: + if (work) { + /* + * When pinning VMA to GGTT on CHV or BXT with VTD enabled, + * commit VMA binding asynchronously to avoid risk of lock + * inversion among reservation_ww locks held here and + * cpu_hotplug_lock acquired from stop_machine(), which we + * wrap around GGTT updates when running in those environments. + */ + if (i915_vma_is_ggtt(vma) && + intel_vm_no_concurrent_access_wa(vma->vm->i915)) + dma_fence_work_commit(&work->base); + else + dma_fence_work_commit_imm(&work->base); + } +err_rpm: + intel_runtime_pm_put(&vma->vm->i915->runtime_pm, wakeref); + + if (moving) + dma_fence_put(moving); + + i915_vma_put_pages(vma); + return err; } -void i915_vma_close(struct i915_vma *vma) +int i915_vma_pin(struct i915_vma *vma, u64 size, u64 alignment, u64 flags) { - lockdep_assert_held(&vma->vm->i915->drm.struct_mutex); + struct i915_gem_ww_ctx ww; + int err; - GEM_BUG_ON(i915_vma_is_closed(vma)); - vma->flags |= I915_VMA_CLOSED; + i915_gem_ww_ctx_init(&ww, true); +retry: + err = i915_gem_object_lock(vma->obj, &ww); + if (!err) + err = i915_vma_pin_ww(vma, &ww, size, alignment, flags); + if (err == -EDEADLK) { + err = i915_gem_ww_ctx_backoff(&ww); + if (!err) + goto retry; + } + i915_gem_ww_ctx_fini(&ww); + + return err; +} + +static void flush_idle_contexts(struct intel_gt *gt) +{ + struct intel_engine_cs *engine; + enum intel_engine_id id; + + for_each_engine(engine, gt, id) + intel_engine_flush_barriers(engine); + + intel_gt_wait_for_idle(gt, MAX_SCHEDULE_TIMEOUT); +} + +static int __i915_ggtt_pin(struct i915_vma *vma, struct i915_gem_ww_ctx *ww, + u32 align, unsigned int flags) +{ + struct i915_address_space *vm = vma->vm; + struct intel_gt *gt; + struct i915_ggtt *ggtt = i915_vm_to_ggtt(vm); + int err; + + do { + err = i915_vma_pin_ww(vma, ww, 0, align, flags | PIN_GLOBAL); + + if (err != -ENOSPC) { + if (!err) { + err = i915_vma_wait_for_bind(vma); + if (err) + i915_vma_unpin(vma); + } + return err; + } + + /* Unlike i915_vma_pin, we don't take no for an answer! */ + list_for_each_entry(gt, &ggtt->gt_list, ggtt_link) + flush_idle_contexts(gt); + if (mutex_lock_interruptible(&vm->mutex) == 0) { + /* + * We pass NULL ww here, as we don't want to unbind + * locked objects when called from execbuf when pinning + * is removed. This would probably regress badly. + */ + i915_gem_evict_vm(vm, NULL, NULL); + mutex_unlock(&vm->mutex); + } + } while (1); +} + +int i915_ggtt_pin(struct i915_vma *vma, struct i915_gem_ww_ctx *ww, + u32 align, unsigned int flags) +{ + struct i915_gem_ww_ctx _ww; + int err; + + GEM_BUG_ON(!i915_vma_is_ggtt(vma)); + if (ww) + return __i915_ggtt_pin(vma, ww, align, flags); + + lockdep_assert_not_held(&vma->obj->base.resv->lock.base); + + for_i915_gem_ww(&_ww, err, true) { + err = i915_gem_object_lock(vma->obj, &_ww); + if (!err) + err = __i915_ggtt_pin(vma, &_ww, align, flags); + } + + return err; +} + +/** + * i915_ggtt_clear_scanout - Clear scanout flag for all objects ggtt vmas + * @obj: i915 GEM object + * This function clears scanout flags for objects ggtt vmas. These flags are set + * when object is pinned for display use and this function to clear them all is + * targeted to be called by frontbuffer tracking code when the frontbuffer is + * about to be released. + */ +void i915_ggtt_clear_scanout(struct drm_i915_gem_object *obj) +{ + struct i915_vma *vma; + + spin_lock(&obj->vma.lock); + for_each_ggtt_vma(vma, obj) { + i915_vma_clear_scanout(vma); + vma->display_alignment = I915_GTT_MIN_ALIGNMENT; + } + spin_unlock(&obj->vma.lock); +} + +static void __vma_close(struct i915_vma *vma, struct intel_gt *gt) +{ /* * We defer actually closing, unbinding and destroying the VMA until * the next idle point, or if the object is freed in the meantime. By @@ -781,66 +1743,173 @@ void i915_vma_close(struct i915_vma *vma) * causing us to rebind the VMA once more. This ends up being a lot * of wasted work for the steady state. */ - list_add_tail(&vma->closed_link, &vma->vm->i915->gt.closed_vma); + GEM_BUG_ON(i915_vma_is_closed(vma)); + list_add(&vma->closed_link, >->closed_vma); } -void i915_vma_reopen(struct i915_vma *vma) +void i915_vma_close(struct i915_vma *vma) { - lockdep_assert_held(&vma->vm->i915->drm.struct_mutex); + struct intel_gt *gt = vma->vm->gt; + unsigned long flags; + + if (i915_vma_is_ggtt(vma)) + return; - if (vma->flags & I915_VMA_CLOSED) { - vma->flags &= ~I915_VMA_CLOSED; - list_del(&vma->closed_link); + GEM_BUG_ON(!atomic_read(&vma->open_count)); + if (atomic_dec_and_lock_irqsave(&vma->open_count, + >->closed_lock, + flags)) { + __vma_close(vma, gt); + spin_unlock_irqrestore(>->closed_lock, flags); } } -static void __i915_vma_destroy(struct i915_vma *vma) +static void __i915_vma_remove_closed(struct i915_vma *vma) { - struct drm_i915_private *i915 = vma->vm->i915; - struct i915_vma_active *iter, *n; - - GEM_BUG_ON(vma->node.allocated); - GEM_BUG_ON(vma->fence); + list_del_init(&vma->closed_link); +} - GEM_BUG_ON(i915_gem_active_isset(&vma->last_fence)); +void i915_vma_reopen(struct i915_vma *vma) +{ + struct intel_gt *gt = vma->vm->gt; - list_del(&vma->obj_link); - list_del(&vma->vm_link); - if (vma->obj) - rb_erase(&vma->obj_node, &vma->obj->vma_tree); + spin_lock_irq(>->closed_lock); + if (i915_vma_is_closed(vma)) + __i915_vma_remove_closed(vma); + spin_unlock_irq(>->closed_lock); +} - rbtree_postorder_for_each_entry_safe(iter, n, &vma->active, node) { - GEM_BUG_ON(i915_gem_active_isset(&iter->base)); - kfree(iter); - } +static void force_unbind(struct i915_vma *vma) +{ + if (!drm_mm_node_allocated(&vma->node)) + return; - kmem_cache_free(i915->vmas, vma); + atomic_and(~I915_VMA_PIN_MASK, &vma->flags); + WARN_ON(__i915_vma_unbind(vma)); + GEM_BUG_ON(drm_mm_node_allocated(&vma->node)); } -void i915_vma_destroy(struct i915_vma *vma) +static void release_references(struct i915_vma *vma, struct intel_gt *gt, + bool vm_ddestroy) { - lockdep_assert_held(&vma->vm->i915->drm.struct_mutex); + struct drm_i915_gem_object *obj = vma->obj; GEM_BUG_ON(i915_vma_is_active(vma)); - GEM_BUG_ON(i915_vma_is_pinned(vma)); - if (i915_vma_is_closed(vma)) - list_del(&vma->closed_link); + spin_lock(&obj->vma.lock); + list_del(&vma->obj_link); + if (!RB_EMPTY_NODE(&vma->obj_node)) + rb_erase(&vma->obj_node, &obj->vma.tree); + + spin_unlock(&obj->vma.lock); - WARN_ON(i915_vma_unbind(vma)); - __i915_vma_destroy(vma); + spin_lock_irq(>->closed_lock); + __i915_vma_remove_closed(vma); + spin_unlock_irq(>->closed_lock); + + if (vm_ddestroy) + i915_vm_resv_put(vma->vm); + + i915_active_fini(&vma->active); + GEM_WARN_ON(vma->resource); + i915_vma_free(vma); } -void i915_vma_parked(struct drm_i915_private *i915) +/* + * i915_vma_destroy_locked - Remove all weak reference to the vma and put + * the initial reference. + * + * This function should be called when it's decided the vma isn't needed + * anymore. The caller must assure that it doesn't race with another lookup + * plus destroy, typically by taking an appropriate reference. + * + * Current callsites are + * - __i915_gem_object_pages_fini() + * - __i915_vm_close() - Blocks the above function by taking a reference on + * the object. + * - __i915_vma_parked() - Blocks the above functions by taking a reference + * on the vm and a reference on the object. Also takes the object lock so + * destruction from __i915_vma_parked() can be blocked by holding the + * object lock. Since the object lock is only allowed from within i915 with + * an object refcount, holding the object lock also implicitly blocks the + * vma freeing from __i915_gem_object_pages_fini(). + * + * Because of locks taken during destruction, a vma is also guaranteed to + * stay alive while the following locks are held if it was looked up while + * holding one of the locks: + * - vm->mutex + * - obj->vma.lock + * - gt->closed_lock + */ +void i915_vma_destroy_locked(struct i915_vma *vma) +{ + lockdep_assert_held(&vma->vm->mutex); + + force_unbind(vma); + list_del_init(&vma->vm_link); + release_references(vma, vma->vm->gt, false); +} + +void i915_vma_destroy(struct i915_vma *vma) +{ + struct intel_gt *gt; + bool vm_ddestroy; + + mutex_lock(&vma->vm->mutex); + force_unbind(vma); + list_del_init(&vma->vm_link); + vm_ddestroy = vma->vm_ddestroy; + vma->vm_ddestroy = false; + + /* vma->vm may be freed when releasing vma->vm->mutex. */ + gt = vma->vm->gt; + mutex_unlock(&vma->vm->mutex); + release_references(vma, gt, vm_ddestroy); +} + +void i915_vma_parked(struct intel_gt *gt) { struct i915_vma *vma, *next; + LIST_HEAD(closed); - list_for_each_entry_safe(vma, next, &i915->gt.closed_vma, closed_link) { - GEM_BUG_ON(!i915_vma_is_closed(vma)); - i915_vma_destroy(vma); + spin_lock_irq(>->closed_lock); + list_for_each_entry_safe(vma, next, >->closed_vma, closed_link) { + struct drm_i915_gem_object *obj = vma->obj; + struct i915_address_space *vm = vma->vm; + + /* XXX All to avoid keeping a reference on i915_vma itself */ + + if (!kref_get_unless_zero(&obj->base.refcount)) + continue; + + if (!i915_vm_tryget(vm)) { + i915_gem_object_put(obj); + continue; + } + + list_move(&vma->closed_link, &closed); } + spin_unlock_irq(>->closed_lock); - GEM_BUG_ON(!list_empty(&i915->gt.closed_vma)); + /* As the GT is held idle, no vma can be reopened as we destroy them */ + list_for_each_entry_safe(vma, next, &closed, closed_link) { + struct drm_i915_gem_object *obj = vma->obj; + struct i915_address_space *vm = vma->vm; + + if (i915_gem_object_trylock(obj, NULL)) { + INIT_LIST_HEAD(&vma->closed_link); + i915_vma_destroy(vma); + i915_gem_object_unlock(obj); + } else { + /* back you go.. */ + spin_lock_irq(>->closed_lock); + list_add(&vma->closed_link, >->closed_vma); + spin_unlock_irq(>->closed_lock); + } + + i915_gem_object_put(obj); + i915_vm_put(vm); + } } static void __i915_vma_iounmap(struct i915_vma *vma) @@ -850,24 +1919,26 @@ static void __i915_vma_iounmap(struct i915_vma *vma) if (vma->iomap == NULL) return; - io_mapping_unmap(vma->iomap); + if (page_unmask_bits(vma->iomap)) + __i915_gem_object_release_map(vma->obj); + else + io_mapping_unmap(vma->iomap); vma->iomap = NULL; } void i915_vma_revoke_mmap(struct i915_vma *vma) { - struct drm_vma_offset_node *node = &vma->obj->base.vma_node; + struct drm_vma_offset_node *node; u64 vma_offset; - lockdep_assert_held(&vma->vm->i915->drm.struct_mutex); - if (!i915_vma_has_userfault(vma)) return; GEM_BUG_ON(!i915_vma_is_map_and_fenceable(vma)); GEM_BUG_ON(!vma->obj->userfault_count); - vma_offset = vma->ggtt_view.partial.offset << PAGE_SHIFT; + node = &vma->mmo->vma_node; + vma_offset = vma->gtt_view.partial.offset << PAGE_SHIFT; unmap_mapping_range(vma->vm->i915->drm.anon_inode->i_mapping, drm_vma_node_offset_addr(node) + vma_offset, vma->size, @@ -878,251 +1949,386 @@ void i915_vma_revoke_mmap(struct i915_vma *vma) list_del(&vma->obj->userfault_link); } -static void export_fence(struct i915_vma *vma, - struct i915_request *rq, - unsigned int flags) +static int +__i915_request_await_bind(struct i915_request *rq, struct i915_vma *vma) { - struct reservation_object *resv = vma->resv; + return __i915_request_await_exclusive(rq, &vma->active); +} - /* - * Ignore errors from failing to allocate the new fence, we can't - * handle an error right now. Worst case should be missed - * synchronisation leading to rendering corruption. - */ - reservation_object_lock(resv, NULL); - if (flags & EXEC_OBJECT_WRITE) - reservation_object_add_excl_fence(resv, &rq->fence); - else if (reservation_object_reserve_shared(resv, 1) == 0) - reservation_object_add_shared_fence(resv, &rq->fence); - reservation_object_unlock(resv); +static int __i915_vma_move_to_active(struct i915_vma *vma, struct i915_request *rq) +{ + int err; + + /* Wait for the vma to be bound before we start! */ + err = __i915_request_await_bind(rq, vma); + if (err) + return err; + + return i915_active_add_request(&vma->active, rq); } -static struct i915_gem_active *active_instance(struct i915_vma *vma, u64 idx) +int _i915_vma_move_to_active(struct i915_vma *vma, + struct i915_request *rq, + struct dma_fence *fence, + unsigned int flags) { - struct i915_vma_active *active; - struct rb_node **p, *parent; - struct i915_request *old; + struct drm_i915_gem_object *obj = vma->obj; + int err; + + assert_object_held(obj); + + GEM_BUG_ON(!vma->pages); + + if (!(flags & __EXEC_OBJECT_NO_REQUEST_AWAIT)) { + err = i915_request_await_object(rq, vma->obj, flags & EXEC_OBJECT_WRITE); + if (unlikely(err)) + return err; + } + err = __i915_vma_move_to_active(vma, rq); + if (unlikely(err)) + return err; /* - * We track the most recently used timeline to skip a rbtree search - * for the common case, under typical loads we never need the rbtree - * at all. We can reuse the last_active slot if it is empty, that is - * after the previous activity has been retired, or if the active - * matches the current timeline. - * - * Note that we allow the timeline to be active simultaneously in - * the rbtree and the last_active cache. We do this to avoid having - * to search and replace the rbtree element for a new timeline, with - * the cost being that we must be aware that the vma may be retired - * twice for the same timeline (as the older rbtree element will be - * retired before the new request added to last_active). + * Reserve fences slot early to prevent an allocation after preparing + * the workload and associating fences with dma_resv. */ - old = i915_gem_active_raw(&vma->last_active, - &vma->vm->i915->drm.struct_mutex); - if (!old || old->fence.context == idx) - goto out; + if (fence && !(flags & __EXEC_OBJECT_NO_RESERVE)) { + struct dma_fence *curr; + int idx; + + dma_fence_array_for_each(curr, idx, fence) + ; + err = dma_resv_reserve_fences(vma->obj->base.resv, idx); + if (unlikely(err)) + return err; + } - /* Move the currently active fence into the rbtree */ - idx = old->fence.context; + if (flags & EXEC_OBJECT_WRITE) { + struct i915_frontbuffer *front; - parent = NULL; - p = &vma->active.rb_node; - while (*p) { - parent = *p; + front = i915_gem_object_frontbuffer_lookup(obj); + if (unlikely(front)) { + if (intel_frontbuffer_invalidate(&front->base, ORIGIN_CS)) + i915_active_add_request(&front->write, rq); + i915_gem_object_frontbuffer_put(front); + } + } - active = rb_entry(parent, struct i915_vma_active, node); - if (active->timeline == idx) - goto replace; + if (fence) { + struct dma_fence *curr; + enum dma_resv_usage usage; + int idx; + + if (flags & EXEC_OBJECT_WRITE) { + usage = DMA_RESV_USAGE_WRITE; + obj->write_domain = I915_GEM_DOMAIN_RENDER; + obj->read_domains = 0; + } else { + usage = DMA_RESV_USAGE_READ; + obj->write_domain = 0; + } - if (active->timeline < idx) - p = &parent->rb_right; - else - p = &parent->rb_left; + dma_fence_array_for_each(curr, idx, fence) + dma_resv_add_fence(vma->obj->base.resv, curr, usage); } - active = kmalloc(sizeof(*active), GFP_KERNEL); + if (flags & EXEC_OBJECT_NEEDS_FENCE && vma->fence) + i915_active_add_request(&vma->fence->active, rq); - /* kmalloc may retire the vma->last_active request (thanks shrinker)! */ - if (unlikely(!i915_gem_active_raw(&vma->last_active, - &vma->vm->i915->drm.struct_mutex))) { - kfree(active); - goto out; - } + obj->read_domains |= I915_GEM_GPU_DOMAINS; + obj->mm.dirty = true; - if (unlikely(!active)) - return ERR_PTR(-ENOMEM); + GEM_BUG_ON(!i915_vma_is_active(vma)); + return 0; +} + +struct dma_fence *__i915_vma_evict(struct i915_vma *vma, bool async) +{ + struct i915_vma_resource *vma_res = vma->resource; + struct dma_fence *unbind_fence; - init_request_active(&active->base, i915_vma_retire); - active->vma = vma; - active->timeline = idx; + GEM_BUG_ON(i915_vma_is_pinned(vma)); + assert_vma_held_evict(vma); - rb_link_node(&active->node, parent, p); - rb_insert_color(&active->node, &vma->active); + if (i915_vma_is_map_and_fenceable(vma)) { + /* Force a pagefault for domain tracking on next user access */ + i915_vma_revoke_mmap(vma); -replace: - /* - * Overwrite the previous active slot in the rbtree with last_active, - * leaving last_active zeroed. If the previous slot is still active, - * we must be careful as we now only expect to receive one retire - * callback not two, and so much undo the active counting for the - * overwritten slot. - */ - if (i915_gem_active_isset(&active->base)) { - /* Retire ourselves from the old rq->active_list */ - __list_del_entry(&active->base.link); - vma->active_count--; - GEM_BUG_ON(!vma->active_count); + /* + * Check that we have flushed all writes through the GGTT + * before the unbind, other due to non-strict nature of those + * indirect writes they may end up referencing the GGTT PTE + * after the unbind. + * + * Note that we may be concurrently poking at the GGTT_WRITE + * bit from set-domain, as we mark all GGTT vma associated + * with an object. We know this is for another vma, as we + * are currently unbinding this one -- so if this vma will be + * reused, it will be refaulted and have its dirty bit set + * before the next write. + */ + i915_vma_flush_writes(vma); + + /* release the fence reg _after_ flushing */ + i915_vma_revoke_fence(vma); + + clear_bit(I915_VMA_CAN_FENCE_BIT, __i915_vma_flags(vma)); } - GEM_BUG_ON(list_empty(&vma->last_active.link)); - list_replace_init(&vma->last_active.link, &active->base.link); - active->base.request = fetch_and_zero(&vma->last_active.request); -out: - return &vma->last_active; -} + __i915_vma_iounmap(vma); -int i915_vma_move_to_active(struct i915_vma *vma, - struct i915_request *rq, - unsigned int flags) -{ - struct drm_i915_gem_object *obj = vma->obj; - struct i915_gem_active *active; + GEM_BUG_ON(vma->fence); + GEM_BUG_ON(i915_vma_has_userfault(vma)); - lockdep_assert_held(&rq->i915->drm.struct_mutex); - GEM_BUG_ON(!drm_mm_node_allocated(&vma->node)); + /* Object backend must be async capable. */ + GEM_WARN_ON(async && !vma->resource->bi.pages_rsgt); + + /* If vm is not open, unbind is a nop. */ + vma_res->needs_wakeref = i915_vma_is_bound(vma, I915_VMA_GLOBAL_BIND) && + kref_read(&vma->vm->ref); + vma_res->skip_pte_rewrite = !kref_read(&vma->vm->ref) || + vma->vm->skip_pte_rewrite; + trace_i915_vma_unbind(vma); + + if (async) + unbind_fence = i915_vma_resource_unbind(vma_res, + vma->obj->mm.tlb); + else + unbind_fence = i915_vma_resource_unbind(vma_res, NULL); + + vma->resource = NULL; - active = active_instance(vma, rq->fence.context); - if (IS_ERR(active)) - return PTR_ERR(active); + atomic_and(~(I915_VMA_BIND_MASK | I915_VMA_ERROR | I915_VMA_GGTT_WRITE), + &vma->flags); + + i915_vma_detach(vma); + + if (!async) { + if (unbind_fence) { + dma_fence_wait(unbind_fence, false); + dma_fence_put(unbind_fence); + unbind_fence = NULL; + } + vma_invalidate_tlb(vma->vm, vma->obj->mm.tlb); + } /* - * Add a reference if we're newly entering the active list. - * The order in which we add operations to the retirement queue is - * vital here: mark_active adds to the start of the callback list, - * such that subsequent callbacks are called first. Therefore we - * add the active reference first and queue for it to be dropped - * *last*. + * Binding itself may not have completed until the unbind fence signals, + * so don't drop the pages until that happens, unless the resource is + * async_capable. */ - if (!i915_gem_active_isset(active) && !vma->active_count++) { - list_move_tail(&vma->vm_link, &vma->vm->active_list); - obj->active_count++; - } - i915_gem_active_set(active, rq); - GEM_BUG_ON(!i915_vma_is_active(vma)); - GEM_BUG_ON(!obj->active_count); - obj->write_domain = 0; - if (flags & EXEC_OBJECT_WRITE) { - obj->write_domain = I915_GEM_DOMAIN_RENDER; + vma_unbind_pages(vma); + return unbind_fence; +} + +int __i915_vma_unbind(struct i915_vma *vma) +{ + int ret; - if (intel_fb_obj_invalidate(obj, ORIGIN_CS)) - i915_gem_active_set(&obj->frontbuffer_write, rq); + lockdep_assert_held(&vma->vm->mutex); + assert_vma_held_evict(vma); - obj->read_domains = 0; + if (!drm_mm_node_allocated(&vma->node)) + return 0; + + if (i915_vma_is_pinned(vma)) { + vma_print_allocator(vma, "is pinned"); + return -EAGAIN; } - obj->read_domains |= I915_GEM_GPU_DOMAINS; - if (flags & EXEC_OBJECT_NEEDS_FENCE) - i915_gem_active_set(&vma->last_fence, rq); + /* + * After confirming that no one else is pinning this vma, wait for + * any laggards who may have crept in during the wait (through + * a residual pin skipping the vm->mutex) to complete. + */ + ret = i915_vma_sync(vma); + if (ret) + return ret; + + GEM_BUG_ON(i915_vma_is_active(vma)); + __i915_vma_evict(vma, false); - export_fence(vma, rq, flags); + drm_mm_remove_node(&vma->node); /* pairs with i915_vma_release() */ return 0; } -int i915_vma_unbind(struct i915_vma *vma) +static struct dma_fence *__i915_vma_unbind_async(struct i915_vma *vma) { - int ret; + struct dma_fence *fence; - lockdep_assert_held(&vma->vm->i915->drm.struct_mutex); + lockdep_assert_held(&vma->vm->mutex); + + if (!drm_mm_node_allocated(&vma->node)) + return NULL; + + if (i915_vma_is_pinned(vma) || + &vma->obj->mm.rsgt->table != vma->resource->bi.pages) + return ERR_PTR(-EAGAIN); /* - * First wait upon any activity as retiring the request may - * have side-effects such as unpinning or even unbinding this vma. + * We probably need to replace this with awaiting the fences of the + * object's dma_resv when the vma active goes away. When doing that + * we need to be careful to not add the vma_resource unbind fence + * immediately to the object's dma_resv, because then unbinding + * the next vma from the object, in case there are many, will + * actually await the unbinding of the previous vmas, which is + * undesirable. */ - might_sleep(); - if (i915_vma_is_active(vma)) { - struct i915_vma_active *active, *n; + if (i915_sw_fence_await_active(&vma->resource->chain, &vma->active, + I915_ACTIVE_AWAIT_EXCL | + I915_ACTIVE_AWAIT_ACTIVE) < 0) { + return ERR_PTR(-EBUSY); + } - /* - * When a closed VMA is retired, it is unbound - eek. - * In order to prevent it from being recursively closed, - * take a pin on the vma so that the second unbind is - * aborted. - * - * Even more scary is that the retire callback may free - * the object (last active vma). To prevent the explosion - * we defer the actual object free to a worker that can - * only proceed once it acquires the struct_mutex (which - * we currently hold, therefore it cannot free this object - * before we are finished). - */ - __i915_vma_pin(vma); + fence = __i915_vma_evict(vma, true); - ret = i915_gem_active_retire(&vma->last_active, - &vma->vm->i915->drm.struct_mutex); - if (ret) - goto unpin; - - rbtree_postorder_for_each_entry_safe(active, n, - &vma->active, node) { - ret = i915_gem_active_retire(&active->base, - &vma->vm->i915->drm.struct_mutex); - if (ret) - goto unpin; - } + drm_mm_remove_node(&vma->node); /* pairs with i915_vma_release() */ - ret = i915_gem_active_retire(&vma->last_fence, - &vma->vm->i915->drm.struct_mutex); -unpin: - __i915_vma_unpin(vma); - if (ret) - return ret; - } - GEM_BUG_ON(i915_vma_is_active(vma)); + return fence; +} + +int i915_vma_unbind(struct i915_vma *vma) +{ + struct i915_address_space *vm = vma->vm; + intel_wakeref_t wakeref = NULL; + int err; + + assert_object_held_shared(vma->obj); + + /* Optimistic wait before taking the mutex */ + err = i915_vma_sync(vma); + if (err) + return err; + + if (!drm_mm_node_allocated(&vma->node)) + return 0; if (i915_vma_is_pinned(vma)) { vma_print_allocator(vma, "is pinned"); - return -EBUSY; + return -EAGAIN; } + if (i915_vma_is_bound(vma, I915_VMA_GLOBAL_BIND)) + /* XXX not always required: nop_clear_range */ + wakeref = intel_runtime_pm_get(&vm->i915->runtime_pm); + + err = mutex_lock_interruptible_nested(&vma->vm->mutex, !wakeref); + if (err) + goto out_rpm; + + err = __i915_vma_unbind(vma); + mutex_unlock(&vm->mutex); + +out_rpm: + if (wakeref) + intel_runtime_pm_put(&vm->i915->runtime_pm, wakeref); + return err; +} + +int i915_vma_unbind_async(struct i915_vma *vma, bool trylock_vm) +{ + struct drm_i915_gem_object *obj = vma->obj; + struct i915_address_space *vm = vma->vm; + intel_wakeref_t wakeref = NULL; + struct dma_fence *fence; + int err; + + /* + * We need the dma-resv lock since we add the + * unbind fence to the dma-resv object. + */ + assert_object_held(obj); + if (!drm_mm_node_allocated(&vma->node)) return 0; - if (i915_vma_is_map_and_fenceable(vma)) { - /* - * Check that we have flushed all writes through the GGTT - * before the unbind, other due to non-strict nature of those - * indirect writes they may end up referencing the GGTT PTE - * after the unbind. - */ - i915_vma_flush_writes(vma); - GEM_BUG_ON(i915_vma_has_ggtt_write(vma)); + if (i915_vma_is_pinned(vma)) { + vma_print_allocator(vma, "is pinned"); + return -EAGAIN; + } - /* release the fence reg _after_ flushing */ - ret = i915_vma_put_fence(vma); - if (ret) - return ret; + if (!obj->mm.rsgt) + return -EBUSY; - /* Force a pagefault for domain tracking on next user access */ - i915_vma_revoke_mmap(vma); + err = dma_resv_reserve_fences(obj->base.resv, 2); + if (err) + return -EBUSY; - __i915_vma_iounmap(vma); - vma->flags &= ~I915_VMA_CAN_FENCE; + /* + * It would be great if we could grab this wakeref from the + * async unbind work if needed, but we can't because it uses + * kmalloc and it's in the dma-fence signalling critical path. + */ + if (i915_vma_is_bound(vma, I915_VMA_GLOBAL_BIND)) + wakeref = intel_runtime_pm_get(&vm->i915->runtime_pm); + + if (trylock_vm && !mutex_trylock(&vm->mutex)) { + err = -EBUSY; + goto out_rpm; + } else if (!trylock_vm) { + err = mutex_lock_interruptible_nested(&vm->mutex, !wakeref); + if (err) + goto out_rpm; } - GEM_BUG_ON(vma->fence); - GEM_BUG_ON(i915_vma_has_userfault(vma)); - if (likely(!vma->vm->closed)) { - trace_i915_vma_unbind(vma); - vma->ops->unbind_vma(vma); + fence = __i915_vma_unbind_async(vma); + mutex_unlock(&vm->mutex); + if (IS_ERR_OR_NULL(fence)) { + err = PTR_ERR_OR_ZERO(fence); + goto out_rpm; } - vma->flags &= ~(I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND); - i915_vma_remove(vma); + dma_resv_add_fence(obj->base.resv, fence, DMA_RESV_USAGE_READ); + dma_fence_put(fence); - return 0; +out_rpm: + if (wakeref) + intel_runtime_pm_put(&vm->i915->runtime_pm, wakeref); + return err; +} + +int i915_vma_unbind_unlocked(struct i915_vma *vma) +{ + int err; + + i915_gem_object_lock(vma->obj, NULL); + err = i915_vma_unbind(vma); + i915_gem_object_unlock(vma->obj); + + return err; +} + +struct i915_vma *i915_vma_make_unshrinkable(struct i915_vma *vma) +{ + i915_gem_object_make_unshrinkable(vma->obj); + return vma; +} + +void i915_vma_make_shrinkable(struct i915_vma *vma) +{ + i915_gem_object_make_shrinkable(vma->obj); +} + +void i915_vma_make_purgeable(struct i915_vma *vma) +{ + i915_gem_object_make_purgeable(vma->obj); } #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST) #include "selftests/i915_vma.c" #endif + +void i915_vma_module_exit(void) +{ + kmem_cache_destroy(slab_vmas); +} + +int __init i915_vma_module_init(void) +{ + slab_vmas = KMEM_CACHE(i915_vma, SLAB_HWCACHE_ALIGN); + if (!slab_vmas) + return -ENOMEM; + + return 0; +} |