From patchwork Mon Jun 17 07:18:53 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 10998019 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 7C0F0112C for ; Mon, 17 Jun 2019 07:19:40 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 62BA72884E for ; Mon, 17 Jun 2019 07:19:40 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id 5299B2885E; Mon, 17 Jun 2019 07:19:40 +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 A2F112884E for ; Mon, 17 Jun 2019 07:19:38 +0000 (UTC) Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 92C86891A8; Mon, 17 Jun 2019 07:19:37 +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 87663891AB for ; Mon, 17 Jun 2019 07:19:36 +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 16923038-1500050 for multiple; Mon, 17 Jun 2019 08:19:12 +0100 From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Mon, 17 Jun 2019 08:18:53 +0100 Message-Id: <20190617071912.20256-3-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20190617071912.20256-1-chris@chris-wilson.co.uk> References: <20190617071912.20256-1-chris@chris-wilson.co.uk> MIME-Version: 1.0 Subject: [Intel-gfx] [PATCH 03/22] drm/i915: Skip shrinking already freed pages 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@intel.com, mika.kuoppala@intel.com Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" X-Virus-Scanned: ClamAV using ClamSMTP Previously, we want to shrink the pages of freed objects before they were RCU collected. However, by removing the struct_mutex serialisation around the active reference, we need to acquire an extra reference around the wait. Unfortunately this means that we have to skip objects that are waiting RCU collection. Signed-off-by: Chris Wilson --- drivers/gpu/drm/i915/gem/i915_gem_object.c | 47 +------------------- drivers/gpu/drm/i915/gem/i915_gem_shrinker.c | 5 +++ 2 files changed, 6 insertions(+), 46 deletions(-) diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object.c b/drivers/gpu/drm/i915/gem/i915_gem_object.c index a4047a585c8b..bb5b6e63a2cc 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_object.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_object.c @@ -148,33 +148,6 @@ void i915_gem_close_object(struct drm_gem_object *gem, struct drm_file *file) } } -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 (!i915_gem_object_is_shrinkable(obj)) - return false; - - 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 file_count(obj->base.filp) == 1; -} - static void __i915_gem_free_objects(struct drm_i915_private *i915, struct llist_node *freed) { @@ -224,8 +197,7 @@ static void __i915_gem_free_objects(struct drm_i915_private *i915, 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); + 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)); @@ -323,23 +295,6 @@ 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)) { - struct drm_i915_private *i915 = to_i915(obj->base.dev); - - obj->mm.madv = I915_MADV_DONTNEED; - - if (i915_gem_object_has_pages(obj)) { - unsigned long flags; - - spin_lock_irqsave(&i915->mm.obj_lock, flags); - list_move_tail(&obj->mm.link, &i915->mm.purge_list); - spin_unlock_irqrestore(&i915->mm.obj_lock, flags); - } - } - /* * Before we free the object, make sure any pure RCU-only * read-side critical sections are complete, e.g. diff --git a/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c b/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c index c851c4029597..3a926a8755c6 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c @@ -241,6 +241,9 @@ i915_gem_shrink(struct drm_i915_private *i915, if (!can_release_pages(obj)) continue; + if (!kref_get_unless_zero(&obj->base.refcount)) + continue; + spin_unlock_irqrestore(&i915->mm.obj_lock, flags); if (unsafe_drop_pages(obj)) { @@ -253,7 +256,9 @@ i915_gem_shrink(struct drm_i915_private *i915, } mutex_unlock(&obj->mm.lock); } + scanned += obj->base.size >> PAGE_SHIFT; + i915_gem_object_put(obj); spin_lock_irqsave(&i915->mm.obj_lock, flags); }