From patchwork Thu Apr 25 09:19:46 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 10916453 Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 70B5C14DB for ; Thu, 25 Apr 2019 09:22:15 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 620A528C55 for ; Thu, 25 Apr 2019 09:22:15 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id 53A0828C74; Thu, 25 Apr 2019 09:22:15 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-5.2 required=2.0 tests=BAYES_00,MAILING_LIST_MULTI, RCVD_IN_DNSWL_MED autolearn=ham version=3.3.1 Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher DHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.wl.linuxfoundation.org (Postfix) with ESMTPS id 245C728C71 for ; Thu, 25 Apr 2019 09:22:12 +0000 (UTC) Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id AF78B89249; Thu, 25 Apr 2019 09:22:08 +0000 (UTC) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from fireflyinternet.com (mail.fireflyinternet.com [109.228.58.192]) by gabe.freedesktop.org (Postfix) with ESMTPS id BBF2789248 for ; Thu, 25 Apr 2019 09:22:03 +0000 (UTC) X-Default-Received-SPF: pass (skip=forwardok (res=PASS)) x-ip-name=78.156.65.138; Received: from haswell.alporthouse.com (unverified [78.156.65.138]) by fireflyinternet.com (Firefly Internet (M1)) with ESMTP id 16355009-1500050 for multiple; Thu, 25 Apr 2019 10:20:09 +0100 From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Thu, 25 Apr 2019 10:19:46 +0100 Message-Id: <20190425092004.9995-27-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20190425092004.9995-1-chris@chris-wilson.co.uk> References: <20190425092004.9995-1-chris@chris-wilson.co.uk> MIME-Version: 1.0 Subject: [Intel-gfx] [PATCH 27/45] drm/i915: Move shmem object setup to its own file X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: Matthew Auld Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" X-Virus-Scanned: ClamAV using ClamSMTP Split the plain old shmem object into its own file to start decluttering i915_gem.c v2: Lose the confusing, hysterical raisins, suffix of _gtt. Signed-off-by: Chris Wilson Reviewed-by: Matthew Auld --- drivers/gpu/drm/i915/Makefile | 3 +- drivers/gpu/drm/i915/gem/i915_gem_object.c | 298 +++++++ drivers/gpu/drm/i915/gem/i915_gem_object.h | 41 +- drivers/gpu/drm/i915/gem/i915_gem_shmem.c | 512 +++++++++++ drivers/gpu/drm/i915/gt/intel_lrc.c | 4 +- drivers/gpu/drm/i915/gt/intel_ringbuffer.c | 2 +- drivers/gpu/drm/i915/gvt/cmd_parser.c | 21 +- drivers/gpu/drm/i915/i915_drv.h | 10 - drivers/gpu/drm/i915/i915_gem.c | 826 +----------------- drivers/gpu/drm/i915/i915_perf.c | 2 +- drivers/gpu/drm/i915/intel_fbdev.c | 2 +- drivers/gpu/drm/i915/intel_guc.c | 2 +- drivers/gpu/drm/i915/intel_uc_fw.c | 3 +- drivers/gpu/drm/i915/selftests/huge_pages.c | 6 +- .../gpu/drm/i915/selftests/i915_gem_dmabuf.c | 8 +- .../gpu/drm/i915/selftests/i915_gem_object.c | 4 +- 16 files changed, 883 insertions(+), 861 deletions(-) create mode 100644 drivers/gpu/drm/i915/gem/i915_gem_shmem.c diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile index deac659ddef0..447143fadb05 100644 --- a/drivers/gpu/drm/i915/Makefile +++ b/drivers/gpu/drm/i915/Makefile @@ -87,7 +87,8 @@ i915-y += $(gt-y) # GEM (Graphics Execution Management) code obj-y += gem/ gem-y += \ - gem/i915_gem_object.o + gem/i915_gem_object.o \ + gem/i915_gem_shmem.o i915-y += \ $(gem-y) \ i915_active.o \ diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object.c b/drivers/gpu/drm/i915/gem/i915_gem_object.c index 8179252bb39b..86e7e88817af 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_object.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_object.c @@ -26,6 +26,7 @@ #include "../i915_drv.h" #include "../i915_globals.h" +#include "../intel_frontbuffer.h" static struct i915_global_object { struct i915_global base; @@ -42,6 +43,64 @@ void i915_gem_object_free(struct drm_i915_gem_object *obj) return kmem_cache_free(global.slab_objects, obj); } +/* some bookkeeping */ +static void i915_gem_info_add_obj(struct drm_i915_private *i915, + u64 size) +{ + spin_lock(&i915->mm.object_stat_lock); + i915->mm.object_count++; + i915->mm.object_memory += size; + spin_unlock(&i915->mm.object_stat_lock); +} + +static void i915_gem_info_remove_obj(struct drm_i915_private *i915, + u64 size) +{ + spin_lock(&i915->mm.object_stat_lock); + i915->mm.object_count--; + i915->mm.object_memory -= size; + spin_unlock(&i915->mm.object_stat_lock); +} + +static void +frontbuffer_retire(struct i915_active_request *active, + struct i915_request *request) +{ + struct drm_i915_gem_object *obj = + container_of(active, typeof(*obj), frontbuffer_write); + + intel_fb_obj_flush(obj, ORIGIN_CS); +} + +void i915_gem_object_init(struct drm_i915_gem_object *obj, + const struct drm_i915_gem_object_ops *ops) +{ + mutex_init(&obj->mm.lock); + + spin_lock_init(&obj->vma.lock); + INIT_LIST_HEAD(&obj->vma.list); + + INIT_LIST_HEAD(&obj->lut_list); + INIT_LIST_HEAD(&obj->batch_pool_link); + + init_rcu_head(&obj->rcu); + + obj->ops = ops; + + reservation_object_init(&obj->__builtin_resv); + obj->resv = &obj->__builtin_resv; + + obj->frontbuffer_ggtt_origin = ORIGIN_GTT; + i915_active_request_init(&obj->frontbuffer_write, + NULL, frontbuffer_retire); + + obj->mm.madv = I915_MADV_WILLNEED; + INIT_RADIX_TREE(&obj->mm.get_page.radix, GFP_KERNEL | __GFP_NOWARN); + mutex_init(&obj->mm.get_page.lock); + + i915_gem_info_add_obj(to_i915(obj->base.dev), obj->base.size); +} + /** * Mark up the object's coherency levels for a given cache_level * @obj: #drm_i915_gem_object @@ -64,6 +123,245 @@ void i915_gem_object_set_cache_coherency(struct drm_i915_gem_object *obj, !(obj->cache_coherent & I915_BO_CACHE_COHERENT_FOR_WRITE); } +void i915_gem_close_object(struct drm_gem_object *gem, struct drm_file *file) +{ + struct drm_i915_private *i915 = to_i915(gem->dev); + struct drm_i915_gem_object *obj = to_intel_bo(gem); + struct drm_i915_file_private *fpriv = file->driver_priv; + struct i915_lut_handle *lut, *ln; + + mutex_lock(&i915->drm.struct_mutex); + + list_for_each_entry_safe(lut, ln, &obj->lut_list, obj_link) { + struct i915_gem_context *ctx = lut->ctx; + struct i915_vma *vma; + + GEM_BUG_ON(ctx->file_priv == ERR_PTR(-EBADF)); + if (ctx->file_priv != fpriv) + continue; + + vma = radix_tree_delete(&ctx->handles_vma, lut->handle); + GEM_BUG_ON(vma->obj != obj); + + /* We allow the process to have multiple handles to the same + * vma, in the same fd namespace, by virtue of flink/open. + */ + GEM_BUG_ON(!vma->open_count); + if (!--vma->open_count && !i915_vma_is_ggtt(vma)) + i915_vma_close(vma); + + list_del(&lut->obj_link); + list_del(&lut->ctx_link); + + i915_lut_handle_free(lut); + __i915_gem_object_release_unless_active(obj); + } + + mutex_unlock(&i915->drm.struct_mutex); +} + +static bool discard_backing_storage(struct drm_i915_gem_object *obj) +{ + /* If we are the last user of the backing storage (be it shmemfs + * pages or stolen etc), we know that the pages are going to be + * immediately released. In this case, we can then skip copying + * back the contents from the GPU. + */ + + if (obj->mm.madv != I915_MADV_WILLNEED) + return false; + + if (!obj->base.filp) + return true; + + /* At first glance, this looks racy, but then again so would be + * userspace racing mmap against close. However, the first external + * reference to the filp can only be obtained through the + * i915_gem_mmap_ioctl() which safeguards us against the user + * acquiring such a reference whilst we are in the middle of + * freeing the object. + */ + return atomic_long_read(&obj->base.filp->f_count) == 1; +} + +static void __i915_gem_free_objects(struct drm_i915_private *i915, + struct llist_node *freed) +{ + struct drm_i915_gem_object *obj, *on; + intel_wakeref_t wakeref; + + wakeref = intel_runtime_pm_get(i915); + llist_for_each_entry_safe(obj, on, freed, freed) { + struct i915_vma *vma, *vn; + + trace_i915_gem_object_destroy(obj); + + mutex_lock(&i915->drm.struct_mutex); + + GEM_BUG_ON(i915_gem_object_is_active(obj)); + list_for_each_entry_safe(vma, vn, &obj->vma.list, obj_link) { + GEM_BUG_ON(i915_vma_is_active(vma)); + vma->flags &= ~I915_VMA_PIN_MASK; + i915_vma_destroy(vma); + } + GEM_BUG_ON(!list_empty(&obj->vma.list)); + GEM_BUG_ON(!RB_EMPTY_ROOT(&obj->vma.tree)); + + /* This serializes freeing with the shrinker. Since the free + * is delayed, first by RCU then by the workqueue, we want the + * shrinker to be able to free pages of unreferenced objects, + * or else we may oom whilst there are plenty of deferred + * freed objects. + */ + if (i915_gem_object_has_pages(obj)) { + spin_lock(&i915->mm.obj_lock); + list_del_init(&obj->mm.link); + spin_unlock(&i915->mm.obj_lock); + } + + mutex_unlock(&i915->drm.struct_mutex); + + GEM_BUG_ON(obj->bind_count); + GEM_BUG_ON(obj->userfault_count); + GEM_BUG_ON(atomic_read(&obj->frontbuffer_bits)); + GEM_BUG_ON(!list_empty(&obj->lut_list)); + + if (obj->ops->release) + obj->ops->release(obj); + + if (WARN_ON(i915_gem_object_has_pinned_pages(obj))) + atomic_set(&obj->mm.pages_pin_count, 0); + __i915_gem_object_put_pages(obj, I915_MM_NORMAL); + GEM_BUG_ON(i915_gem_object_has_pages(obj)); + + if (obj->base.import_attach) + drm_prime_gem_destroy(&obj->base, NULL); + + reservation_object_fini(&obj->__builtin_resv); + drm_gem_object_release(&obj->base); + i915_gem_info_remove_obj(i915, obj->base.size); + + bitmap_free(obj->bit_17); + i915_gem_object_free(obj); + + GEM_BUG_ON(!atomic_read(&i915->mm.free_count)); + atomic_dec(&i915->mm.free_count); + + if (on) + cond_resched(); + } + intel_runtime_pm_put(i915, wakeref); +} + +void i915_gem_flush_free_objects(struct drm_i915_private *i915) +{ + struct llist_node *freed; + + /* Free the oldest, most stale object to keep the free_list short */ + freed = NULL; + if (!llist_empty(&i915->mm.free_list)) { /* quick test for hotpath */ + /* Only one consumer of llist_del_first() allowed */ + spin_lock(&i915->mm.free_lock); + freed = llist_del_first(&i915->mm.free_list); + spin_unlock(&i915->mm.free_lock); + } + if (unlikely(freed)) { + freed->next = NULL; + __i915_gem_free_objects(i915, freed); + } +} + +static void __i915_gem_free_work(struct work_struct *work) +{ + struct drm_i915_private *i915 = + container_of(work, struct drm_i915_private, mm.free_work); + struct llist_node *freed; + + /* + * All file-owned VMA should have been released by this point through + * i915_gem_close_object(), or earlier by i915_gem_context_close(). + * However, the object may also be bound into the global GTT (e.g. + * older GPUs without per-process support, or for direct access through + * the GTT either for the user or for scanout). Those VMA still need to + * unbound now. + */ + + spin_lock(&i915->mm.free_lock); + while ((freed = llist_del_all(&i915->mm.free_list))) { + spin_unlock(&i915->mm.free_lock); + + __i915_gem_free_objects(i915, freed); + if (need_resched()) + return; + + spin_lock(&i915->mm.free_lock); + } + spin_unlock(&i915->mm.free_lock); +} + +static void __i915_gem_free_object_rcu(struct rcu_head *head) +{ + struct drm_i915_gem_object *obj = + container_of(head, typeof(*obj), rcu); + struct drm_i915_private *i915 = to_i915(obj->base.dev); + + /* + * We reuse obj->rcu for the freed list, so we had better not treat + * it like a rcu_head from this point forwards. And we expect all + * objects to be freed via this path. + */ + destroy_rcu_head(&obj->rcu); + + /* + * Since we require blocking on struct_mutex to unbind the freed + * object from the GPU before releasing resources back to the + * system, we can not do that directly from the RCU callback (which may + * be a softirq context), but must instead then defer that work onto a + * kthread. We use the RCU callback rather than move the freed object + * directly onto the work queue so that we can mix between using the + * worker and performing frees directly from subsequent allocations for + * crude but effective memory throttling. + */ + if (llist_add(&obj->freed, &i915->mm.free_list)) + queue_work(i915->wq, &i915->mm.free_work); +} + +void i915_gem_free_object(struct drm_gem_object *gem_obj) +{ + struct drm_i915_gem_object *obj = to_intel_bo(gem_obj); + + if (obj->mm.quirked) + __i915_gem_object_unpin_pages(obj); + + if (discard_backing_storage(obj)) + obj->mm.madv = I915_MADV_DONTNEED; + + /* + * Before we free the object, make sure any pure RCU-only + * read-side critical sections are complete, e.g. + * i915_gem_busy_ioctl(). For the corresponding synchronized + * lookup see i915_gem_object_lookup_rcu(). + */ + atomic_inc(&to_i915(obj->base.dev)->mm.free_count); + call_rcu(&obj->rcu, __i915_gem_free_object_rcu); +} + +void __i915_gem_object_release_unless_active(struct drm_i915_gem_object *obj) +{ + lockdep_assert_held(&obj->base.dev->struct_mutex); + + if (!i915_gem_object_has_active_reference(obj) && + i915_gem_object_is_active(obj)) + i915_gem_object_set_active_reference(obj); + else + i915_gem_object_put(obj); +} + +void i915_gem_init__objects(struct drm_i915_private *i915) +{ + INIT_WORK(&i915->mm.free_work, __i915_gem_free_work); +} + static void i915_global_objects_shrink(void) { kmem_cache_shrink(global.slab_objects); diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object.h b/drivers/gpu/drm/i915/gem/i915_gem_object.h index 061b20c3da5b..ad82303f741a 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_object.h +++ b/drivers/gpu/drm/i915/gem/i915_gem_object.h @@ -15,9 +15,29 @@ #include "i915_gem_object_types.h" +void i915_gem_init__objects(struct drm_i915_private *i915); + struct drm_i915_gem_object *i915_gem_object_alloc(void); void i915_gem_object_free(struct drm_i915_gem_object *obj); +void i915_gem_object_init(struct drm_i915_gem_object *obj, + const struct drm_i915_gem_object_ops *ops); +struct drm_i915_gem_object * +i915_gem_object_create_shmem(struct drm_i915_private *i915, u64 size); +struct drm_i915_gem_object * +i915_gem_object_create_shmem_from_data(struct drm_i915_private *i915, + const void *data, size_t size); + +extern const struct drm_i915_gem_object_ops i915_gem_shmem_ops; +void __i915_gem_object_release_shmem(struct drm_i915_gem_object *obj, + struct sg_table *pages, + bool needs_clflush); + +void i915_gem_close_object(struct drm_gem_object *gem, struct drm_file *file); +void i915_gem_free_object(struct drm_gem_object *obj); + +void i915_gem_flush_free_objects(struct drm_i915_private *i915); + /** * i915_gem_object_lookup_rcu - look up a temporary GEM object from its handle * @filp: DRM file private date @@ -338,8 +358,23 @@ void i915_gem_object_set_cache_coherency(struct drm_i915_gem_object *obj, unsigned int cache_level); void i915_gem_object_flush_if_display(struct drm_i915_gem_object *obj); -void __i915_gem_object_release_shmem(struct drm_i915_gem_object *obj, - struct sg_table *pages, - bool needs_clflush); +static inline bool cpu_write_needs_clflush(struct drm_i915_gem_object *obj) +{ + if (obj->cache_dirty) + return false; + + if (!(obj->cache_coherent & I915_BO_CACHE_COHERENT_FOR_WRITE)) + return true; + + return obj->pin_global; /* currently in use by HW, keep flushed */ +} + +static inline void __start_cpu_write(struct drm_i915_gem_object *obj) +{ + obj->read_domains = I915_GEM_DOMAIN_CPU; + obj->write_domain = I915_GEM_DOMAIN_CPU; + if (cpu_write_needs_clflush(obj)) + obj->cache_dirty = true; +} #endif diff --git a/drivers/gpu/drm/i915/gem/i915_gem_shmem.c b/drivers/gpu/drm/i915/gem/i915_gem_shmem.c new file mode 100644 index 000000000000..9d885b5e6643 --- /dev/null +++ b/drivers/gpu/drm/i915/gem/i915_gem_shmem.c @@ -0,0 +1,512 @@ +/* + * SPDX-License-Identifier: MIT + * + * Copyright © 2014-2016 Intel Corporation + */ + +#include +#include + +#include "i915_gem_object.h" + +#include "../i915_drv.h" + +/* + * Move pages to appropriate lru and release the pagevec, decrementing the + * ref count of those pages. + */ +static void check_release_pagevec(struct pagevec *pvec) +{ + check_move_unevictable_pages(pvec); + __pagevec_release(pvec); + cond_resched(); +} + +static int shmem_get_pages(struct drm_i915_gem_object *obj) +{ + struct drm_i915_private *i915 = to_i915(obj->base.dev); + const unsigned long page_count = obj->base.size / PAGE_SIZE; + unsigned long i; + struct address_space *mapping; + struct sg_table *st; + struct scatterlist *sg; + struct sgt_iter sgt_iter; + struct page *page; + unsigned long last_pfn = 0; /* suppress gcc warning */ + unsigned int max_segment = i915_sg_segment_size(); + unsigned int sg_page_sizes; + struct pagevec pvec; + gfp_t noreclaim; + int ret; + + /* + * Assert that the object is not currently in any GPU domain. As it + * wasn't in the GTT, there shouldn't be any way it could have been in + * a GPU cache + */ + GEM_BUG_ON(obj->read_domains & I915_GEM_GPU_DOMAINS); + GEM_BUG_ON(obj->write_domain & I915_GEM_GPU_DOMAINS); + + /* + * If there's no chance of allocating enough pages for the whole + * object, bail early. + */ + if (page_count > totalram_pages()) + return -ENOMEM; + + st = kmalloc(sizeof(*st), GFP_KERNEL); + if (!st) + return -ENOMEM; + +rebuild_st: + if (sg_alloc_table(st, page_count, GFP_KERNEL)) { + kfree(st); + return -ENOMEM; + } + + /* + * Get the list of pages out of our struct file. They'll be pinned + * at this point until we release them. + * + * Fail silently without starting the shrinker + */ + mapping = obj->base.filp->f_mapping; + mapping_set_unevictable(mapping); + noreclaim = mapping_gfp_constraint(mapping, ~__GFP_RECLAIM); + noreclaim |= __GFP_NORETRY | __GFP_NOWARN; + + sg = st->sgl; + st->nents = 0; + sg_page_sizes = 0; + for (i = 0; i < page_count; i++) { + const unsigned int shrink[] = { + (I915_SHRINK_BOUND | + I915_SHRINK_UNBOUND | + I915_SHRINK_PURGEABLE), + 0, + }, *s = shrink; + gfp_t gfp = noreclaim; + + do { + cond_resched(); + page = shmem_read_mapping_page_gfp(mapping, i, gfp); + if (!IS_ERR(page)) + break; + + if (!*s) { + ret = PTR_ERR(page); + goto err_sg; + } + + i915_gem_shrink(i915, 2 * page_count, NULL, *s++); + + /* + * We've tried hard to allocate the memory by reaping + * our own buffer, now let the real VM do its job and + * go down in flames if truly OOM. + * + * However, since graphics tend to be disposable, + * defer the oom here by reporting the ENOMEM back + * to userspace. + */ + if (!*s) { + /* reclaim and warn, but no oom */ + gfp = mapping_gfp_mask(mapping); + + /* + * Our bo are always dirty and so we require + * kswapd to reclaim our pages (direct reclaim + * does not effectively begin pageout of our + * buffers on its own). However, direct reclaim + * only waits for kswapd when under allocation + * congestion. So as a result __GFP_RECLAIM is + * unreliable and fails to actually reclaim our + * dirty pages -- unless you try over and over + * again with !__GFP_NORETRY. However, we still + * want to fail this allocation rather than + * trigger the out-of-memory killer and for + * this we want __GFP_RETRY_MAYFAIL. + */ + gfp |= __GFP_RETRY_MAYFAIL; + } + } while (1); + + if (!i || + sg->length >= max_segment || + page_to_pfn(page) != last_pfn + 1) { + if (i) { + sg_page_sizes |= sg->length; + sg = sg_next(sg); + } + st->nents++; + sg_set_page(sg, page, PAGE_SIZE, 0); + } else { + sg->length += PAGE_SIZE; + } + last_pfn = page_to_pfn(page); + + /* Check that the i965g/gm workaround works. */ + WARN_ON((gfp & __GFP_DMA32) && (last_pfn >= 0x00100000UL)); + } + if (sg) { /* loop terminated early; short sg table */ + sg_page_sizes |= sg->length; + sg_mark_end(sg); + } + + /* Trim unused sg entries to avoid wasting memory. */ + i915_sg_trim(st); + + ret = i915_gem_gtt_prepare_pages(obj, st); + if (ret) { + /* + * DMA remapping failed? One possible cause is that + * it could not reserve enough large entries, asking + * for PAGE_SIZE chunks instead may be helpful. + */ + if (max_segment > PAGE_SIZE) { + for_each_sgt_page(page, sgt_iter, st) + put_page(page); + sg_free_table(st); + + max_segment = PAGE_SIZE; + goto rebuild_st; + } else { + dev_warn(&i915->drm.pdev->dev, + "Failed to DMA remap %lu pages\n", + page_count); + goto err_pages; + } + } + + if (i915_gem_object_needs_bit17_swizzle(obj)) + i915_gem_object_do_bit_17_swizzle(obj, st); + + __i915_gem_object_set_pages(obj, st, sg_page_sizes); + + return 0; + +err_sg: + sg_mark_end(sg); +err_pages: + mapping_clear_unevictable(mapping); + pagevec_init(&pvec); + for_each_sgt_page(page, sgt_iter, st) { + if (!pagevec_add(&pvec, page)) + check_release_pagevec(&pvec); + } + if (pagevec_count(&pvec)) + check_release_pagevec(&pvec); + sg_free_table(st); + kfree(st); + + /* + * shmemfs first checks if there is enough memory to allocate the page + * and reports ENOSPC should there be insufficient, along with the usual + * ENOMEM for a genuine allocation failure. + * + * We use ENOSPC in our driver to mean that we have run out of aperture + * space and so want to translate the error from shmemfs back to our + * usual understanding of ENOMEM. + */ + if (ret == -ENOSPC) + ret = -ENOMEM; + + return ret; +} + +void +__i915_gem_object_release_shmem(struct drm_i915_gem_object *obj, + struct sg_table *pages, + bool needs_clflush) +{ + GEM_BUG_ON(obj->mm.madv == __I915_MADV_PURGED); + + if (obj->mm.madv == I915_MADV_DONTNEED) + obj->mm.dirty = false; + + if (needs_clflush && + (obj->read_domains & I915_GEM_DOMAIN_CPU) == 0 && + !(obj->cache_coherent & I915_BO_CACHE_COHERENT_FOR_READ)) + drm_clflush_sg(pages); + + __start_cpu_write(obj); +} + +static void +shmem_put_pages(struct drm_i915_gem_object *obj, struct sg_table *pages) +{ + struct sgt_iter sgt_iter; + struct pagevec pvec; + struct page *page; + + __i915_gem_object_release_shmem(obj, pages, true); + + i915_gem_gtt_finish_pages(obj, pages); + + if (i915_gem_object_needs_bit17_swizzle(obj)) + i915_gem_object_save_bit_17_swizzle(obj, pages); + + mapping_clear_unevictable(file_inode(obj->base.filp)->i_mapping); + + pagevec_init(&pvec); + for_each_sgt_page(page, sgt_iter, pages) { + if (obj->mm.dirty) + set_page_dirty(page); + + if (obj->mm.madv == I915_MADV_WILLNEED) + mark_page_accessed(page); + + if (!pagevec_add(&pvec, page)) + check_release_pagevec(&pvec); + } + if (pagevec_count(&pvec)) + check_release_pagevec(&pvec); + obj->mm.dirty = false; + + sg_free_table(pages); + kfree(pages); +} + +static int +shmem_pwrite(struct drm_i915_gem_object *obj, + const struct drm_i915_gem_pwrite *arg) +{ + struct address_space *mapping = obj->base.filp->f_mapping; + char __user *user_data = u64_to_user_ptr(arg->data_ptr); + u64 remain, offset; + unsigned int pg; + + /* Caller already validated user args */ + GEM_BUG_ON(!access_ok(user_data, arg->size)); + + /* + * Before we instantiate/pin the backing store for our use, we + * can prepopulate the shmemfs filp efficiently using a write into + * the pagecache. We avoid the penalty of instantiating all the + * pages, important if the user is just writing to a few and never + * uses the object on the GPU, and using a direct write into shmemfs + * allows it to avoid the cost of retrieving a page (either swapin + * or clearing-before-use) before it is overwritten. + */ + if (i915_gem_object_has_pages(obj)) + return -ENODEV; + + if (obj->mm.madv != I915_MADV_WILLNEED) + return -EFAULT; + + /* + * Before the pages are instantiated the object is treated as being + * in the CPU domain. The pages will be clflushed as required before + * use, and we can freely write into the pages directly. If userspace + * races pwrite with any other operation; corruption will ensue - + * that is userspace's prerogative! + */ + + remain = arg->size; + offset = arg->offset; + pg = offset_in_page(offset); + + do { + unsigned int len, unwritten; + struct page *page; + void *data, *vaddr; + int err; + char c; + + len = PAGE_SIZE - pg; + if (len > remain) + len = remain; + + /* Prefault the user page to reduce potential recursion */ + err = __get_user(c, user_data); + if (err) + return err; + + err = __get_user(c, user_data + len - 1); + if (err) + return err; + + err = pagecache_write_begin(obj->base.filp, mapping, + offset, len, 0, + &page, &data); + if (err < 0) + return err; + + vaddr = kmap_atomic(page); + unwritten = __copy_from_user_inatomic(vaddr + pg, + user_data, + len); + kunmap_atomic(vaddr); + + err = pagecache_write_end(obj->base.filp, mapping, + offset, len, len - unwritten, + page, data); + if (err < 0) + return err; + + /* We don't handle -EFAULT, leave it to the caller to check */ + if (unwritten) + return -ENODEV; + + remain -= len; + user_data += len; + offset += len; + pg = 0; + } while (remain); + + return 0; +} + +const struct drm_i915_gem_object_ops i915_gem_shmem_ops = { + .flags = I915_GEM_OBJECT_HAS_STRUCT_PAGE | + I915_GEM_OBJECT_IS_SHRINKABLE, + + .get_pages = shmem_get_pages, + .put_pages = shmem_put_pages, + + .pwrite = shmem_pwrite, +}; + +static int create_shmem(struct drm_i915_private *i915, + struct drm_gem_object *obj, + size_t size) +{ + unsigned long flags = VM_NORESERVE; + struct file *filp; + + drm_gem_private_object_init(&i915->drm, obj, size); + + if (i915->mm.gemfs) + filp = shmem_file_setup_with_mnt(i915->mm.gemfs, "i915", size, + flags); + else + filp = shmem_file_setup("i915", size, flags); + if (IS_ERR(filp)) + return PTR_ERR(filp); + + obj->filp = filp; + return 0; +} + +struct drm_i915_gem_object * +i915_gem_object_create_shmem(struct drm_i915_private *i915, u64 size) +{ + struct drm_i915_gem_object *obj; + struct address_space *mapping; + unsigned int cache_level; + gfp_t mask; + int ret; + + /* There is a prevalence of the assumption that we fit the object's + * page count inside a 32bit _signed_ variable. Let's document this and + * catch if we ever need to fix it. In the meantime, if you do spot + * such a local variable, please consider fixing! + */ + if (size >> PAGE_SHIFT > INT_MAX) + return ERR_PTR(-E2BIG); + + if (overflows_type(size, obj->base.size)) + return ERR_PTR(-E2BIG); + + obj = i915_gem_object_alloc(); + if (!obj) + return ERR_PTR(-ENOMEM); + + ret = create_shmem(i915, &obj->base, size); + if (ret) + goto fail; + + mask = GFP_HIGHUSER | __GFP_RECLAIMABLE; + if (IS_I965GM(i915) || IS_I965G(i915)) { + /* 965gm cannot relocate objects above 4GiB. */ + mask &= ~__GFP_HIGHMEM; + mask |= __GFP_DMA32; + } + + mapping = obj->base.filp->f_mapping; + mapping_set_gfp_mask(mapping, mask); + GEM_BUG_ON(!(mapping_gfp_mask(mapping) & __GFP_RECLAIM)); + + i915_gem_object_init(obj, &i915_gem_shmem_ops); + + obj->write_domain = I915_GEM_DOMAIN_CPU; + obj->read_domains = I915_GEM_DOMAIN_CPU; + + if (HAS_LLC(i915)) + /* On some devices, we can have the GPU use the LLC (the CPU + * cache) for about a 10% performance improvement + * compared to uncached. Graphics requests other than + * display scanout are coherent with the CPU in + * accessing this cache. This means in this mode we + * don't need to clflush on the CPU side, and on the + * GPU side we only need to flush internal caches to + * get data visible to the CPU. + * + * However, we maintain the display planes as UC, and so + * need to rebind when first used as such. + */ + cache_level = I915_CACHE_LLC; + else + cache_level = I915_CACHE_NONE; + + i915_gem_object_set_cache_coherency(obj, cache_level); + + trace_i915_gem_object_create(obj); + + return obj; + +fail: + i915_gem_object_free(obj); + return ERR_PTR(ret); +} + +/* Allocate a new GEM object and fill it with the supplied data */ +struct drm_i915_gem_object * +i915_gem_object_create_shmem_from_data(struct drm_i915_private *dev_priv, + const void *data, size_t size) +{ + struct drm_i915_gem_object *obj; + struct file *file; + size_t offset; + int err; + + obj = i915_gem_object_create_shmem(dev_priv, round_up(size, PAGE_SIZE)); + if (IS_ERR(obj)) + return obj; + + GEM_BUG_ON(obj->write_domain != I915_GEM_DOMAIN_CPU); + + file = obj->base.filp; + offset = 0; + do { + unsigned int len = min_t(typeof(size), size, PAGE_SIZE); + struct page *page; + void *pgdata, *vaddr; + + err = pagecache_write_begin(file, file->f_mapping, + offset, len, 0, + &page, &pgdata); + if (err < 0) + goto fail; + + vaddr = kmap(page); + memcpy(vaddr, data, len); + kunmap(page); + + err = pagecache_write_end(file, file->f_mapping, + offset, len, len, + page, pgdata); + if (err < 0) + goto fail; + + size -= len; + data += len; + offset += len; + } while (size); + + return obj; + +fail: + i915_gem_object_put(obj); + return ERR_PTR(err); +} diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c index d793d976402d..a833b6c14516 100644 --- a/drivers/gpu/drm/i915/gt/intel_lrc.c +++ b/drivers/gpu/drm/i915/gt/intel_lrc.c @@ -1938,7 +1938,7 @@ static int lrc_setup_wa_ctx(struct intel_engine_cs *engine) struct i915_vma *vma; int err; - obj = i915_gem_object_create(engine->i915, CTX_WA_BB_OBJ_SIZE); + obj = i915_gem_object_create_shmem(engine->i915, CTX_WA_BB_OBJ_SIZE); if (IS_ERR(obj)) return PTR_ERR(obj); @@ -3035,7 +3035,7 @@ static int execlists_context_deferred_alloc(struct intel_context *ce, */ context_size += LRC_HEADER_PAGES * PAGE_SIZE; - ctx_obj = i915_gem_object_create(engine->i915, context_size); + ctx_obj = i915_gem_object_create_shmem(engine->i915, context_size); if (IS_ERR(ctx_obj)) return PTR_ERR(ctx_obj); diff --git a/drivers/gpu/drm/i915/gt/intel_ringbuffer.c b/drivers/gpu/drm/i915/gt/intel_ringbuffer.c index 2edeedb748ed..0b503b82db0f 100644 --- a/drivers/gpu/drm/i915/gt/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/gt/intel_ringbuffer.c @@ -1410,7 +1410,7 @@ alloc_context_vma(struct intel_engine_cs *engine) struct i915_vma *vma; int err; - obj = i915_gem_object_create(i915, engine->context_size); + obj = i915_gem_object_create_shmem(i915, engine->context_size); if (IS_ERR(obj)) return ERR_CAST(obj); diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.c b/drivers/gpu/drm/i915/gvt/cmd_parser.c index ab002cfd3cab..fe66285f0236 100644 --- a/drivers/gpu/drm/i915/gvt/cmd_parser.c +++ b/drivers/gpu/drm/i915/gvt/cmd_parser.c @@ -1721,7 +1721,7 @@ static int perform_bb_shadow(struct parser_exec_state *s) int ret = 0; struct intel_vgpu_mm *mm = (s->buf_addr_type == GTT_BUFFER) ? s->vgpu->gtt.ggtt_mm : s->workload->shadow_mm; - unsigned long gma_start_offset = 0; + unsigned long start_offset = 0; /* get the start gm address of the batch buffer */ gma = get_gma_bb_from_cmd(s, 1); @@ -1738,7 +1738,7 @@ static int perform_bb_shadow(struct parser_exec_state *s) bb->ppgtt = (s->buf_addr_type == GTT_BUFFER) ? false : true; - /* the gma_start_offset stores the batch buffer's start gma's + /* the start_offset stores the batch buffer's start gma's * offset relative to page boundary. so for non-privileged batch * buffer, the shadowed gem object holds exactly the same page * layout as original gem object. This is for the convience of @@ -1750,10 +1750,11 @@ static int perform_bb_shadow(struct parser_exec_state *s) * that of shadowed page. */ if (bb->ppgtt) - gma_start_offset = gma & ~I915_GTT_PAGE_MASK; + start_offset = gma & ~I915_GTT_PAGE_MASK; - bb->obj = i915_gem_object_create(s->vgpu->gvt->dev_priv, - roundup(bb_size + gma_start_offset, PAGE_SIZE)); + bb->obj = i915_gem_object_create_shmem(s->vgpu->gvt->dev_priv, + round_up(bb_size + start_offset, + PAGE_SIZE)); if (IS_ERR(bb->obj)) { ret = PTR_ERR(bb->obj); goto err_free_bb; @@ -1776,7 +1777,7 @@ static int perform_bb_shadow(struct parser_exec_state *s) ret = copy_gma_to_hva(s->vgpu, mm, gma, gma + bb_size, - bb->va + gma_start_offset); + bb->va + start_offset); if (ret < 0) { gvt_vgpu_err("fail to copy guest ring buffer\n"); ret = -EFAULT; @@ -1802,7 +1803,7 @@ static int perform_bb_shadow(struct parser_exec_state *s) * buffer's gma in pair. After all, we don't want to pin the shadow * buffer here (too early). */ - s->ip_va = bb->va + gma_start_offset; + s->ip_va = bb->va + start_offset; s->ip_gma = gma; return 0; err_unmap: @@ -2825,9 +2826,9 @@ static int shadow_indirect_ctx(struct intel_shadow_wa_ctx *wa_ctx) int ret = 0; void *map; - obj = i915_gem_object_create(workload->vgpu->gvt->dev_priv, - roundup(ctx_size + CACHELINE_BYTES, - PAGE_SIZE)); + obj = i915_gem_object_create_shmem(workload->vgpu->gvt->dev_priv, + roundup(ctx_size + CACHELINE_BYTES, + PAGE_SIZE)); if (IS_ERR(obj)) return PTR_ERR(obj); diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 86ca50ea9b5d..007334e892a3 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -2833,16 +2833,6 @@ void i915_gem_load_init_fences(struct drm_i915_private *dev_priv); int i915_gem_freeze(struct drm_i915_private *dev_priv); int i915_gem_freeze_late(struct drm_i915_private *dev_priv); -void i915_gem_object_init(struct drm_i915_gem_object *obj, - const struct drm_i915_gem_object_ops *ops); -struct drm_i915_gem_object * -i915_gem_object_create(struct drm_i915_private *dev_priv, u64 size); -struct drm_i915_gem_object * -i915_gem_object_create_from_data(struct drm_i915_private *dev_priv, - const void *data, size_t size); -void i915_gem_close_object(struct drm_gem_object *gem, struct drm_file *file); -void i915_gem_free_object(struct drm_gem_object *obj); - static inline void i915_gem_drain_freed_objects(struct drm_i915_private *i915) { if (!atomic_read(&i915->mm.free_count)) diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 78ca69dcfaa6..4c5fba0c7bf6 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -57,19 +57,6 @@ #include "intel_frontbuffer.h" #include "intel_pm.h" -static void i915_gem_flush_free_objects(struct drm_i915_private *i915); - -static bool cpu_write_needs_clflush(struct drm_i915_gem_object *obj) -{ - if (obj->cache_dirty) - return false; - - if (!(obj->cache_coherent & I915_BO_CACHE_COHERENT_FOR_WRITE)) - return true; - - return obj->pin_global; /* currently in use by HW, keep flushed */ -} - static int insert_mappable_node(struct i915_ggtt *ggtt, struct drm_mm_node *node, u32 size) @@ -87,25 +74,6 @@ remove_mappable_node(struct drm_mm_node *node) drm_mm_remove_node(node); } -/* some bookkeeping */ -static void i915_gem_info_add_obj(struct drm_i915_private *dev_priv, - u64 size) -{ - spin_lock(&dev_priv->mm.object_stat_lock); - dev_priv->mm.object_count++; - dev_priv->mm.object_memory += size; - spin_unlock(&dev_priv->mm.object_stat_lock); -} - -static void i915_gem_info_remove_obj(struct drm_i915_private *dev_priv, - u64 size) -{ - spin_lock(&dev_priv->mm.object_stat_lock); - dev_priv->mm.object_count--; - dev_priv->mm.object_memory -= size; - spin_unlock(&dev_priv->mm.object_stat_lock); -} - int i915_gem_get_aperture_ioctl(struct drm_device *dev, void *data, struct drm_file *file) @@ -206,32 +174,6 @@ static int i915_gem_object_get_pages_phys(struct drm_i915_gem_object *obj) return err; } -static void __start_cpu_write(struct drm_i915_gem_object *obj) -{ - obj->read_domains = I915_GEM_DOMAIN_CPU; - obj->write_domain = I915_GEM_DOMAIN_CPU; - if (cpu_write_needs_clflush(obj)) - obj->cache_dirty = true; -} - -void -__i915_gem_object_release_shmem(struct drm_i915_gem_object *obj, - struct sg_table *pages, - bool needs_clflush) -{ - GEM_BUG_ON(obj->mm.madv == __I915_MADV_PURGED); - - if (obj->mm.madv == I915_MADV_DONTNEED) - obj->mm.dirty = false; - - if (needs_clflush && - (obj->read_domains & I915_GEM_DOMAIN_CPU) == 0 && - !(obj->cache_coherent & I915_BO_CACHE_COHERENT_FOR_READ)) - drm_clflush_sg(pages); - - __start_cpu_write(obj); -} - static void i915_gem_object_put_pages_phys(struct drm_i915_gem_object *obj, struct sg_table *pages) @@ -283,8 +225,6 @@ static const struct drm_i915_gem_object_ops i915_gem_phys_ops = { .release = i915_gem_object_release_phys, }; -static const struct drm_i915_gem_object_ops i915_gem_object_ops; - int i915_gem_object_unbind(struct drm_i915_gem_object *obj) { struct i915_vma *vma; @@ -541,7 +481,7 @@ i915_gem_create(struct drm_file *file, return -EINVAL; /* Allocate the new object */ - obj = i915_gem_object_create(dev_priv, size); + obj = i915_gem_object_create_shmem(dev_priv, size); if (IS_ERR(obj)) return PTR_ERR(obj); @@ -2065,52 +2005,6 @@ void __i915_gem_object_truncate(struct drm_i915_gem_object *obj) obj->mm.pages = ERR_PTR(-EFAULT); } -/* - * Move pages to appropriate lru and release the pagevec, decrementing the - * ref count of those pages. - */ -static void check_release_pagevec(struct pagevec *pvec) -{ - check_move_unevictable_pages(pvec); - __pagevec_release(pvec); - cond_resched(); -} - -static void -i915_gem_object_put_pages_gtt(struct drm_i915_gem_object *obj, - struct sg_table *pages) -{ - struct sgt_iter sgt_iter; - struct pagevec pvec; - struct page *page; - - __i915_gem_object_release_shmem(obj, pages, true); - i915_gem_gtt_finish_pages(obj, pages); - - if (i915_gem_object_needs_bit17_swizzle(obj)) - i915_gem_object_save_bit_17_swizzle(obj, pages); - - mapping_clear_unevictable(file_inode(obj->base.filp)->i_mapping); - - pagevec_init(&pvec); - for_each_sgt_page(page, sgt_iter, pages) { - if (obj->mm.dirty) - set_page_dirty(page); - - if (obj->mm.madv == I915_MADV_WILLNEED) - mark_page_accessed(page); - - if (!pagevec_add(&pvec, page)) - check_release_pagevec(&pvec); - } - if (pagevec_count(&pvec)) - check_release_pagevec(&pvec); - obj->mm.dirty = false; - - sg_free_table(pages); - kfree(pages); -} - static void __i915_gem_object_reset_page_iter(struct drm_i915_gem_object *obj) { struct radix_tree_iter iter; @@ -2226,196 +2120,6 @@ bool i915_sg_trim(struct sg_table *orig_st) return true; } -static int i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj) -{ - struct drm_i915_private *dev_priv = to_i915(obj->base.dev); - const unsigned long page_count = obj->base.size / PAGE_SIZE; - unsigned long i; - struct address_space *mapping; - struct sg_table *st; - struct scatterlist *sg; - struct sgt_iter sgt_iter; - struct page *page; - unsigned long last_pfn = 0; /* suppress gcc warning */ - unsigned int max_segment = i915_sg_segment_size(); - unsigned int sg_page_sizes; - struct pagevec pvec; - gfp_t noreclaim; - int ret; - - /* - * Assert that the object is not currently in any GPU domain. As it - * wasn't in the GTT, there shouldn't be any way it could have been in - * a GPU cache - */ - GEM_BUG_ON(obj->read_domains & I915_GEM_GPU_DOMAINS); - GEM_BUG_ON(obj->write_domain & I915_GEM_GPU_DOMAINS); - - /* - * If there's no chance of allocating enough pages for the whole - * object, bail early. - */ - if (page_count > totalram_pages()) - return -ENOMEM; - - st = kmalloc(sizeof(*st), GFP_KERNEL); - if (st == NULL) - return -ENOMEM; - -rebuild_st: - if (sg_alloc_table(st, page_count, GFP_KERNEL)) { - kfree(st); - return -ENOMEM; - } - - /* - * Get the list of pages out of our struct file. They'll be pinned - * at this point until we release them. - * - * Fail silently without starting the shrinker - */ - mapping = obj->base.filp->f_mapping; - mapping_set_unevictable(mapping); - noreclaim = mapping_gfp_constraint(mapping, ~__GFP_RECLAIM); - noreclaim |= __GFP_NORETRY | __GFP_NOWARN; - - sg = st->sgl; - st->nents = 0; - sg_page_sizes = 0; - for (i = 0; i < page_count; i++) { - const unsigned int shrink[] = { - I915_SHRINK_BOUND | I915_SHRINK_UNBOUND | I915_SHRINK_PURGEABLE, - 0, - }, *s = shrink; - gfp_t gfp = noreclaim; - - do { - cond_resched(); - page = shmem_read_mapping_page_gfp(mapping, i, gfp); - if (!IS_ERR(page)) - break; - - if (!*s) { - ret = PTR_ERR(page); - goto err_sg; - } - - i915_gem_shrink(dev_priv, 2 * page_count, NULL, *s++); - - /* - * We've tried hard to allocate the memory by reaping - * our own buffer, now let the real VM do its job and - * go down in flames if truly OOM. - * - * However, since graphics tend to be disposable, - * defer the oom here by reporting the ENOMEM back - * to userspace. - */ - if (!*s) { - /* reclaim and warn, but no oom */ - gfp = mapping_gfp_mask(mapping); - - /* - * Our bo are always dirty and so we require - * kswapd to reclaim our pages (direct reclaim - * does not effectively begin pageout of our - * buffers on its own). However, direct reclaim - * only waits for kswapd when under allocation - * congestion. So as a result __GFP_RECLAIM is - * unreliable and fails to actually reclaim our - * dirty pages -- unless you try over and over - * again with !__GFP_NORETRY. However, we still - * want to fail this allocation rather than - * trigger the out-of-memory killer and for - * this we want __GFP_RETRY_MAYFAIL. - */ - gfp |= __GFP_RETRY_MAYFAIL; - } - } while (1); - - if (!i || - sg->length >= max_segment || - page_to_pfn(page) != last_pfn + 1) { - if (i) { - sg_page_sizes |= sg->length; - sg = sg_next(sg); - } - st->nents++; - sg_set_page(sg, page, PAGE_SIZE, 0); - } else { - sg->length += PAGE_SIZE; - } - last_pfn = page_to_pfn(page); - - /* Check that the i965g/gm workaround works. */ - WARN_ON((gfp & __GFP_DMA32) && (last_pfn >= 0x00100000UL)); - } - if (sg) { /* loop terminated early; short sg table */ - sg_page_sizes |= sg->length; - sg_mark_end(sg); - } - - /* Trim unused sg entries to avoid wasting memory. */ - i915_sg_trim(st); - - ret = i915_gem_gtt_prepare_pages(obj, st); - if (ret) { - /* - * DMA remapping failed? One possible cause is that - * it could not reserve enough large entries, asking - * for PAGE_SIZE chunks instead may be helpful. - */ - if (max_segment > PAGE_SIZE) { - for_each_sgt_page(page, sgt_iter, st) - put_page(page); - sg_free_table(st); - - max_segment = PAGE_SIZE; - goto rebuild_st; - } else { - dev_warn(&dev_priv->drm.pdev->dev, - "Failed to DMA remap %lu pages\n", - page_count); - goto err_pages; - } - } - - if (i915_gem_object_needs_bit17_swizzle(obj)) - i915_gem_object_do_bit_17_swizzle(obj, st); - - __i915_gem_object_set_pages(obj, st, sg_page_sizes); - - return 0; - -err_sg: - sg_mark_end(sg); -err_pages: - mapping_clear_unevictable(mapping); - pagevec_init(&pvec); - for_each_sgt_page(page, sgt_iter, st) { - if (!pagevec_add(&pvec, page)) - check_release_pagevec(&pvec); - } - if (pagevec_count(&pvec)) - check_release_pagevec(&pvec); - sg_free_table(st); - kfree(st); - - /* - * shmemfs first checks if there is enough memory to allocate the page - * and reports ENOSPC should there be insufficient, along with the usual - * ENOMEM for a genuine allocation failure. - * - * We use ENOSPC in our driver to mean that we have run out of aperture - * space and so want to translate the error from shmemfs back to our - * usual understanding of ENOMEM. - */ - if (ret == -ENOSPC) - ret = -ENOMEM; - - return ret; -} - void __i915_gem_object_set_pages(struct drm_i915_gem_object *obj, struct sg_table *pages, unsigned int sg_page_sizes) @@ -2662,133 +2366,6 @@ void __i915_gem_object_flush_map(struct drm_i915_gem_object *obj, } } -static int -i915_gem_object_pwrite_gtt(struct drm_i915_gem_object *obj, - const struct drm_i915_gem_pwrite *arg) -{ - struct address_space *mapping = obj->base.filp->f_mapping; - char __user *user_data = u64_to_user_ptr(arg->data_ptr); - u64 remain, offset; - unsigned int pg; - - /* Caller already validated user args */ - GEM_BUG_ON(!access_ok(user_data, arg->size)); - - /* - * Before we instantiate/pin the backing store for our use, we - * can prepopulate the shmemfs filp efficiently using a write into - * the pagecache. We avoid the penalty of instantiating all the - * pages, important if the user is just writing to a few and never - * uses the object on the GPU, and using a direct write into shmemfs - * allows it to avoid the cost of retrieving a page (either swapin - * or clearing-before-use) before it is overwritten. - */ - if (i915_gem_object_has_pages(obj)) - return -ENODEV; - - if (obj->mm.madv != I915_MADV_WILLNEED) - return -EFAULT; - - /* - * Before the pages are instantiated the object is treated as being - * in the CPU domain. The pages will be clflushed as required before - * use, and we can freely write into the pages directly. If userspace - * races pwrite with any other operation; corruption will ensue - - * that is userspace's prerogative! - */ - - remain = arg->size; - offset = arg->offset; - pg = offset_in_page(offset); - - do { - unsigned int len, unwritten; - struct page *page; - void *data, *vaddr; - int err; - char c; - - len = PAGE_SIZE - pg; - if (len > remain) - len = remain; - - /* Prefault the user page to reduce potential recursion */ - err = __get_user(c, user_data); - if (err) - return err; - - err = __get_user(c, user_data + len - 1); - if (err) - return err; - - err = pagecache_write_begin(obj->base.filp, mapping, - offset, len, 0, - &page, &data); - if (err < 0) - return err; - - vaddr = kmap_atomic(page); - unwritten = __copy_from_user_inatomic(vaddr + pg, - user_data, - len); - kunmap_atomic(vaddr); - - err = pagecache_write_end(obj->base.filp, mapping, - offset, len, len - unwritten, - page, data); - if (err < 0) - return err; - - /* We don't handle -EFAULT, leave it to the caller to check */ - if (unwritten) - return -ENODEV; - - remain -= len; - user_data += len; - offset += len; - pg = 0; - } while (remain); - - return 0; -} - -void i915_gem_close_object(struct drm_gem_object *gem, struct drm_file *file) -{ - struct drm_i915_private *i915 = to_i915(gem->dev); - struct drm_i915_gem_object *obj = to_intel_bo(gem); - struct drm_i915_file_private *fpriv = file->driver_priv; - struct i915_lut_handle *lut, *ln; - - mutex_lock(&i915->drm.struct_mutex); - - list_for_each_entry_safe(lut, ln, &obj->lut_list, obj_link) { - struct i915_gem_context *ctx = lut->ctx; - struct i915_vma *vma; - - GEM_BUG_ON(ctx->file_priv == ERR_PTR(-EBADF)); - if (ctx->file_priv != fpriv) - continue; - - vma = radix_tree_delete(&ctx->handles_vma, lut->handle); - GEM_BUG_ON(vma->obj != obj); - - /* We allow the process to have multiple handles to the same - * vma, in the same fd namespace, by virtue of flink/open. - */ - GEM_BUG_ON(!vma->open_count); - if (!--vma->open_count && !i915_vma_is_ggtt(vma)) - i915_vma_close(vma); - - list_del(&lut->obj_link); - list_del(&lut->ctx_link); - - i915_lut_handle_free(lut); - __i915_gem_object_release_unless_active(obj); - } - - mutex_unlock(&i915->drm.struct_mutex); -} - static unsigned long to_wait_timeout(s64 timeout_ns) { if (timeout_ns < 0) @@ -3787,348 +3364,6 @@ i915_gem_madvise_ioctl(struct drm_device *dev, void *data, return err; } -static void -frontbuffer_retire(struct i915_active_request *active, - struct i915_request *request) -{ - struct drm_i915_gem_object *obj = - container_of(active, typeof(*obj), frontbuffer_write); - - intel_fb_obj_flush(obj, ORIGIN_CS); -} - -void i915_gem_object_init(struct drm_i915_gem_object *obj, - const struct drm_i915_gem_object_ops *ops) -{ - mutex_init(&obj->mm.lock); - - spin_lock_init(&obj->vma.lock); - INIT_LIST_HEAD(&obj->vma.list); - - INIT_LIST_HEAD(&obj->lut_list); - INIT_LIST_HEAD(&obj->batch_pool_link); - - init_rcu_head(&obj->rcu); - - obj->ops = ops; - - reservation_object_init(&obj->__builtin_resv); - obj->resv = &obj->__builtin_resv; - - obj->frontbuffer_ggtt_origin = ORIGIN_GTT; - i915_active_request_init(&obj->frontbuffer_write, - NULL, frontbuffer_retire); - - obj->mm.madv = I915_MADV_WILLNEED; - INIT_RADIX_TREE(&obj->mm.get_page.radix, GFP_KERNEL | __GFP_NOWARN); - mutex_init(&obj->mm.get_page.lock); - - i915_gem_info_add_obj(to_i915(obj->base.dev), obj->base.size); -} - -static const struct drm_i915_gem_object_ops i915_gem_object_ops = { - .flags = I915_GEM_OBJECT_HAS_STRUCT_PAGE | - I915_GEM_OBJECT_IS_SHRINKABLE, - - .get_pages = i915_gem_object_get_pages_gtt, - .put_pages = i915_gem_object_put_pages_gtt, - - .pwrite = i915_gem_object_pwrite_gtt, -}; - -static int i915_gem_object_create_shmem(struct drm_device *dev, - struct drm_gem_object *obj, - size_t size) -{ - struct drm_i915_private *i915 = to_i915(dev); - unsigned long flags = VM_NORESERVE; - struct file *filp; - - drm_gem_private_object_init(dev, obj, size); - - if (i915->mm.gemfs) - filp = shmem_file_setup_with_mnt(i915->mm.gemfs, "i915", size, - flags); - else - filp = shmem_file_setup("i915", size, flags); - - if (IS_ERR(filp)) - return PTR_ERR(filp); - - obj->filp = filp; - - return 0; -} - -struct drm_i915_gem_object * -i915_gem_object_create(struct drm_i915_private *dev_priv, u64 size) -{ - struct drm_i915_gem_object *obj; - struct address_space *mapping; - unsigned int cache_level; - gfp_t mask; - int ret; - - /* There is a prevalence of the assumption that we fit the object's - * page count inside a 32bit _signed_ variable. Let's document this and - * catch if we ever need to fix it. In the meantime, if you do spot - * such a local variable, please consider fixing! - */ - if (size >> PAGE_SHIFT > INT_MAX) - return ERR_PTR(-E2BIG); - - if (overflows_type(size, obj->base.size)) - return ERR_PTR(-E2BIG); - - obj = i915_gem_object_alloc(); - if (obj == NULL) - return ERR_PTR(-ENOMEM); - - ret = i915_gem_object_create_shmem(&dev_priv->drm, &obj->base, size); - if (ret) - goto fail; - - mask = GFP_HIGHUSER | __GFP_RECLAIMABLE; - if (IS_I965GM(dev_priv) || IS_I965G(dev_priv)) { - /* 965gm cannot relocate objects above 4GiB. */ - mask &= ~__GFP_HIGHMEM; - mask |= __GFP_DMA32; - } - - mapping = obj->base.filp->f_mapping; - mapping_set_gfp_mask(mapping, mask); - GEM_BUG_ON(!(mapping_gfp_mask(mapping) & __GFP_RECLAIM)); - - i915_gem_object_init(obj, &i915_gem_object_ops); - - obj->write_domain = I915_GEM_DOMAIN_CPU; - obj->read_domains = I915_GEM_DOMAIN_CPU; - - if (HAS_LLC(dev_priv)) - /* On some devices, we can have the GPU use the LLC (the CPU - * cache) for about a 10% performance improvement - * compared to uncached. Graphics requests other than - * display scanout are coherent with the CPU in - * accessing this cache. This means in this mode we - * don't need to clflush on the CPU side, and on the - * GPU side we only need to flush internal caches to - * get data visible to the CPU. - * - * However, we maintain the display planes as UC, and so - * need to rebind when first used as such. - */ - cache_level = I915_CACHE_LLC; - else - cache_level = I915_CACHE_NONE; - - i915_gem_object_set_cache_coherency(obj, cache_level); - - trace_i915_gem_object_create(obj); - - return obj; - -fail: - i915_gem_object_free(obj); - return ERR_PTR(ret); -} - -static bool discard_backing_storage(struct drm_i915_gem_object *obj) -{ - /* If we are the last user of the backing storage (be it shmemfs - * pages or stolen etc), we know that the pages are going to be - * immediately released. In this case, we can then skip copying - * back the contents from the GPU. - */ - - if (obj->mm.madv != I915_MADV_WILLNEED) - return false; - - if (obj->base.filp == NULL) - return true; - - /* At first glance, this looks racy, but then again so would be - * userspace racing mmap against close. However, the first external - * reference to the filp can only be obtained through the - * i915_gem_mmap_ioctl() which safeguards us against the user - * acquiring such a reference whilst we are in the middle of - * freeing the object. - */ - return atomic_long_read(&obj->base.filp->f_count) == 1; -} - -static void __i915_gem_free_objects(struct drm_i915_private *i915, - struct llist_node *freed) -{ - struct drm_i915_gem_object *obj, *on; - intel_wakeref_t wakeref; - - wakeref = intel_runtime_pm_get(i915); - llist_for_each_entry_safe(obj, on, freed, freed) { - struct i915_vma *vma, *vn; - - trace_i915_gem_object_destroy(obj); - - mutex_lock(&i915->drm.struct_mutex); - - GEM_BUG_ON(i915_gem_object_is_active(obj)); - list_for_each_entry_safe(vma, vn, &obj->vma.list, obj_link) { - GEM_BUG_ON(i915_vma_is_active(vma)); - vma->flags &= ~I915_VMA_PIN_MASK; - i915_vma_destroy(vma); - } - GEM_BUG_ON(!list_empty(&obj->vma.list)); - GEM_BUG_ON(!RB_EMPTY_ROOT(&obj->vma.tree)); - - /* This serializes freeing with the shrinker. Since the free - * is delayed, first by RCU then by the workqueue, we want the - * shrinker to be able to free pages of unreferenced objects, - * or else we may oom whilst there are plenty of deferred - * freed objects. - */ - if (i915_gem_object_has_pages(obj)) { - spin_lock(&i915->mm.obj_lock); - list_del_init(&obj->mm.link); - spin_unlock(&i915->mm.obj_lock); - } - - mutex_unlock(&i915->drm.struct_mutex); - - GEM_BUG_ON(obj->bind_count); - GEM_BUG_ON(obj->userfault_count); - GEM_BUG_ON(atomic_read(&obj->frontbuffer_bits)); - GEM_BUG_ON(!list_empty(&obj->lut_list)); - - if (obj->ops->release) - obj->ops->release(obj); - - if (WARN_ON(i915_gem_object_has_pinned_pages(obj))) - atomic_set(&obj->mm.pages_pin_count, 0); - __i915_gem_object_put_pages(obj, I915_MM_NORMAL); - GEM_BUG_ON(i915_gem_object_has_pages(obj)); - - if (obj->base.import_attach) - drm_prime_gem_destroy(&obj->base, NULL); - - reservation_object_fini(&obj->__builtin_resv); - drm_gem_object_release(&obj->base); - i915_gem_info_remove_obj(i915, obj->base.size); - - bitmap_free(obj->bit_17); - i915_gem_object_free(obj); - - GEM_BUG_ON(!atomic_read(&i915->mm.free_count)); - atomic_dec(&i915->mm.free_count); - - if (on) - cond_resched(); - } - intel_runtime_pm_put(i915, wakeref); -} - -static void i915_gem_flush_free_objects(struct drm_i915_private *i915) -{ - struct llist_node *freed; - - /* Free the oldest, most stale object to keep the free_list short */ - freed = NULL; - if (!llist_empty(&i915->mm.free_list)) { /* quick test for hotpath */ - /* Only one consumer of llist_del_first() allowed */ - spin_lock(&i915->mm.free_lock); - freed = llist_del_first(&i915->mm.free_list); - spin_unlock(&i915->mm.free_lock); - } - if (unlikely(freed)) { - freed->next = NULL; - __i915_gem_free_objects(i915, freed); - } -} - -static void __i915_gem_free_work(struct work_struct *work) -{ - struct drm_i915_private *i915 = - container_of(work, struct drm_i915_private, mm.free_work); - struct llist_node *freed; - - /* - * All file-owned VMA should have been released by this point through - * i915_gem_close_object(), or earlier by i915_gem_context_close(). - * However, the object may also be bound into the global GTT (e.g. - * older GPUs without per-process support, or for direct access through - * the GTT either for the user or for scanout). Those VMA still need to - * unbound now. - */ - - spin_lock(&i915->mm.free_lock); - while ((freed = llist_del_all(&i915->mm.free_list))) { - spin_unlock(&i915->mm.free_lock); - - __i915_gem_free_objects(i915, freed); - if (need_resched()) - return; - - spin_lock(&i915->mm.free_lock); - } - spin_unlock(&i915->mm.free_lock); -} - -static void __i915_gem_free_object_rcu(struct rcu_head *head) -{ - struct drm_i915_gem_object *obj = - container_of(head, typeof(*obj), rcu); - struct drm_i915_private *i915 = to_i915(obj->base.dev); - - /* - * We reuse obj->rcu for the freed list, so we had better not treat - * it like a rcu_head from this point forwards. And we expect all - * objects to be freed via this path. - */ - destroy_rcu_head(&obj->rcu); - - /* - * Since we require blocking on struct_mutex to unbind the freed - * object from the GPU before releasing resources back to the - * system, we can not do that directly from the RCU callback (which may - * be a softirq context), but must instead then defer that work onto a - * kthread. We use the RCU callback rather than move the freed object - * directly onto the work queue so that we can mix between using the - * worker and performing frees directly from subsequent allocations for - * crude but effective memory throttling. - */ - if (llist_add(&obj->freed, &i915->mm.free_list)) - queue_work(i915->wq, &i915->mm.free_work); -} - -void i915_gem_free_object(struct drm_gem_object *gem_obj) -{ - struct drm_i915_gem_object *obj = to_intel_bo(gem_obj); - - if (obj->mm.quirked) - __i915_gem_object_unpin_pages(obj); - - if (discard_backing_storage(obj)) - obj->mm.madv = I915_MADV_DONTNEED; - - /* - * Before we free the object, make sure any pure RCU-only - * read-side critical sections are complete, e.g. - * i915_gem_busy_ioctl(). For the corresponding synchronized - * lookup see i915_gem_object_lookup_rcu(). - */ - atomic_inc(&to_i915(obj->base.dev)->mm.free_count); - call_rcu(&obj->rcu, __i915_gem_free_object_rcu); -} - -void __i915_gem_object_release_unless_active(struct drm_i915_gem_object *obj) -{ - lockdep_assert_held(&obj->base.dev->struct_mutex); - - if (!i915_gem_object_has_active_reference(obj) && - i915_gem_object_is_active(obj)) - i915_gem_object_set_active_reference(obj); - else - i915_gem_object_put(obj); -} - void i915_gem_sanitize(struct drm_i915_private *i915) { intel_wakeref_t wakeref; @@ -4725,7 +3960,7 @@ static void i915_gem_init__mm(struct drm_i915_private *i915) INIT_LIST_HEAD(&i915->mm.fence_list); INIT_LIST_HEAD(&i915->mm.userfault_list); - INIT_WORK(&i915->mm.free_work, __i915_gem_free_work); + i915_gem_init__objects(i915); } int i915_gem_init_early(struct drm_i915_private *dev_priv) @@ -4891,57 +4126,6 @@ void i915_gem_track_fb(struct drm_i915_gem_object *old, } } -/* Allocate a new GEM object and fill it with the supplied data */ -struct drm_i915_gem_object * -i915_gem_object_create_from_data(struct drm_i915_private *dev_priv, - const void *data, size_t size) -{ - struct drm_i915_gem_object *obj; - struct file *file; - size_t offset; - int err; - - obj = i915_gem_object_create(dev_priv, round_up(size, PAGE_SIZE)); - if (IS_ERR(obj)) - return obj; - - GEM_BUG_ON(obj->write_domain != I915_GEM_DOMAIN_CPU); - - file = obj->base.filp; - offset = 0; - do { - unsigned int len = min_t(typeof(size), size, PAGE_SIZE); - struct page *page; - void *pgdata, *vaddr; - - err = pagecache_write_begin(file, file->f_mapping, - offset, len, 0, - &page, &pgdata); - if (err < 0) - goto fail; - - vaddr = kmap(page); - memcpy(vaddr, data, len); - kunmap(page); - - err = pagecache_write_end(file, file->f_mapping, - offset, len, len, - page, pgdata); - if (err < 0) - goto fail; - - size -= len; - data += len; - offset += len; - } while (size); - - return obj; - -fail: - i915_gem_object_put(obj); - return ERR_PTR(err); -} - struct scatterlist * i915_gem_object_get_sg(struct drm_i915_gem_object *obj, unsigned int n, @@ -5103,7 +4287,7 @@ int i915_gem_object_attach_phys(struct drm_i915_gem_object *obj, int align) if (obj->ops == &i915_gem_phys_ops) return 0; - if (obj->ops != &i915_gem_object_ops) + if (obj->ops != &i915_gem_shmem_ops) return -EINVAL; err = i915_gem_object_unbind(obj); @@ -5139,12 +4323,12 @@ int i915_gem_object_attach_phys(struct drm_i915_gem_object *obj, int align) __i915_gem_object_pin_pages(obj); if (!IS_ERR_OR_NULL(pages)) - i915_gem_object_ops.put_pages(obj, pages); + i915_gem_shmem_ops.put_pages(obj, pages); mutex_unlock(&obj->mm.lock); return 0; err_xfer: - obj->ops = &i915_gem_object_ops; + obj->ops = &i915_gem_shmem_ops; if (!IS_ERR_OR_NULL(pages)) { unsigned int sg_page_sizes = i915_sg_page_sizes(pages->sgl); diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c index c4995d5a16d2..379fd89a180f 100644 --- a/drivers/gpu/drm/i915/i915_perf.c +++ b/drivers/gpu/drm/i915/i915_perf.c @@ -1510,7 +1510,7 @@ static int alloc_oa_buffer(struct drm_i915_private *dev_priv) BUILD_BUG_ON_NOT_POWER_OF_2(OA_BUFFER_SIZE); BUILD_BUG_ON(OA_BUFFER_SIZE < SZ_128K || OA_BUFFER_SIZE > SZ_16M); - bo = i915_gem_object_create(dev_priv, OA_BUFFER_SIZE); + bo = i915_gem_object_create_shmem(dev_priv, OA_BUFFER_SIZE); if (IS_ERR(bo)) { DRM_ERROR("Failed to allocate OA buffer\n"); ret = PTR_ERR(bo); diff --git a/drivers/gpu/drm/i915/intel_fbdev.c b/drivers/gpu/drm/i915/intel_fbdev.c index 89db71996148..0d3a6fa674e6 100644 --- a/drivers/gpu/drm/i915/intel_fbdev.c +++ b/drivers/gpu/drm/i915/intel_fbdev.c @@ -144,7 +144,7 @@ static int intelfb_alloc(struct drm_fb_helper *helper, if (size * 2 < dev_priv->stolen_usable_size) obj = i915_gem_object_create_stolen(dev_priv, size); if (obj == NULL) - obj = i915_gem_object_create(dev_priv, size); + obj = i915_gem_object_create_shmem(dev_priv, size); if (IS_ERR(obj)) { DRM_ERROR("failed to allocate framebuffer\n"); ret = PTR_ERR(obj); diff --git a/drivers/gpu/drm/i915/intel_guc.c b/drivers/gpu/drm/i915/intel_guc.c index c4ac29309fcc..98e45a500a15 100644 --- a/drivers/gpu/drm/i915/intel_guc.c +++ b/drivers/gpu/drm/i915/intel_guc.c @@ -690,7 +690,7 @@ struct i915_vma *intel_guc_allocate_vma(struct intel_guc *guc, u32 size) u64 flags; int ret; - obj = i915_gem_object_create(dev_priv, size); + obj = i915_gem_object_create_shmem(dev_priv, size); if (IS_ERR(obj)) return ERR_CAST(obj); diff --git a/drivers/gpu/drm/i915/intel_uc_fw.c b/drivers/gpu/drm/i915/intel_uc_fw.c index b9cb6fea9332..3257a054cb0b 100644 --- a/drivers/gpu/drm/i915/intel_uc_fw.c +++ b/drivers/gpu/drm/i915/intel_uc_fw.c @@ -159,7 +159,8 @@ void intel_uc_fw_fetch(struct drm_i915_private *dev_priv, goto fail; } - obj = i915_gem_object_create_from_data(dev_priv, fw->data, fw->size); + obj = i915_gem_object_create_shmem_from_data(dev_priv, + fw->data, fw->size); if (IS_ERR(obj)) { err = PTR_ERR(obj); DRM_DEBUG_DRIVER("%s fw object_create err=%d\n", diff --git a/drivers/gpu/drm/i915/selftests/huge_pages.c b/drivers/gpu/drm/i915/selftests/huge_pages.c index 1e1f83326a96..ce4ec87698f6 100644 --- a/drivers/gpu/drm/i915/selftests/huge_pages.c +++ b/drivers/gpu/drm/i915/selftests/huge_pages.c @@ -1390,7 +1390,7 @@ static int igt_ppgtt_gemfs_huge(void *arg) for (i = 0; i < ARRAY_SIZE(sizes); ++i) { unsigned int size = sizes[i]; - obj = i915_gem_object_create(i915, size); + obj = i915_gem_object_create_shmem(i915, size); if (IS_ERR(obj)) return PTR_ERR(obj); @@ -1568,7 +1568,7 @@ static int igt_tmpfs_fallback(void *arg) i915->mm.gemfs = NULL; - obj = i915_gem_object_create(i915, PAGE_SIZE); + obj = i915_gem_object_create_shmem(i915, PAGE_SIZE); if (IS_ERR(obj)) { err = PTR_ERR(obj); goto out_restore; @@ -1628,7 +1628,7 @@ static int igt_shrink_thp(void *arg) return 0; } - obj = i915_gem_object_create(i915, SZ_2M); + obj = i915_gem_object_create_shmem(i915, SZ_2M); if (IS_ERR(obj)) return PTR_ERR(obj); diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_dmabuf.c b/drivers/gpu/drm/i915/selftests/i915_gem_dmabuf.c index 2b943ee246c9..cc65a503e2f0 100644 --- a/drivers/gpu/drm/i915/selftests/i915_gem_dmabuf.c +++ b/drivers/gpu/drm/i915/selftests/i915_gem_dmabuf.c @@ -33,7 +33,7 @@ static int igt_dmabuf_export(void *arg) struct drm_i915_gem_object *obj; struct dma_buf *dmabuf; - obj = i915_gem_object_create(i915, PAGE_SIZE); + obj = i915_gem_object_create_shmem(i915, PAGE_SIZE); if (IS_ERR(obj)) return PTR_ERR(obj); @@ -57,7 +57,7 @@ static int igt_dmabuf_import_self(void *arg) struct dma_buf *dmabuf; int err; - obj = i915_gem_object_create(i915, PAGE_SIZE); + obj = i915_gem_object_create_shmem(i915, PAGE_SIZE); if (IS_ERR(obj)) return PTR_ERR(obj); @@ -232,7 +232,7 @@ static int igt_dmabuf_export_vmap(void *arg) void *ptr; int err; - obj = i915_gem_object_create(i915, PAGE_SIZE); + obj = i915_gem_object_create_shmem(i915, PAGE_SIZE); if (IS_ERR(obj)) return PTR_ERR(obj); @@ -279,7 +279,7 @@ static int igt_dmabuf_export_kmap(void *arg) void *ptr; int err; - obj = i915_gem_object_create(i915, 2*PAGE_SIZE); + obj = i915_gem_object_create_shmem(i915, 2 * PAGE_SIZE); if (IS_ERR(obj)) return PTR_ERR(obj); diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_object.c b/drivers/gpu/drm/i915/selftests/i915_gem_object.c index 088b2aa05dcd..ff31fe0cecec 100644 --- a/drivers/gpu/drm/i915/selftests/i915_gem_object.c +++ b/drivers/gpu/drm/i915/selftests/i915_gem_object.c @@ -36,7 +36,7 @@ static int igt_gem_object(void *arg) /* Basic test to ensure we can create an object */ - obj = i915_gem_object_create(i915, PAGE_SIZE); + obj = i915_gem_object_create_shmem(i915, PAGE_SIZE); if (IS_ERR(obj)) { err = PTR_ERR(obj); pr_err("i915_gem_object_create failed, err=%d\n", err); @@ -59,7 +59,7 @@ static int igt_phys_object(void *arg) * i.e. exercise the i915_gem_object_phys API. */ - obj = i915_gem_object_create(i915, PAGE_SIZE); + obj = i915_gem_object_create_shmem(i915, PAGE_SIZE); if (IS_ERR(obj)) { err = PTR_ERR(obj); pr_err("i915_gem_object_create failed, err=%d\n", err);