Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <20220726175043.1027731-1-robdclark@gmail.com> <20220726175043.1027731-10-robdclark@gmail.com>
 <49fe9ecf-b1bd-a21b-8d8c-e4a33e3fa821@collabora.com>
In-Reply-To: <49fe9ecf-b1bd-a21b-8d8c-e4a33e3fa821@collabora.com>
From:   Rob Clark <robdclark@gmail.com>
Date:   Fri, 29 Jul 2022 08:40:24 -0700
Message-ID: <CAF6AEGuKU839m6TiARN3EwjPToo-qpdZR5cGD+BdJeiObjeY4A@mail.gmail.com>
Subject: Re: [PATCH v3 09/15] drm/gem: Add LRU/shrinker helper
To:     Dmitry Osipenko <dmitry.osipenko@collabora.com>
Cc:     dri-devel@lists.freedesktop.org, freedreno@lists.freedesktop.org,
        linux-arm-msm@vger.kernel.org, Rob Clark <robdclark@chromium.org>,
        Daniel Vetter <daniel@ffwll.ch>,
        Thomas Zimmermann <tzimmermann@suse.de>,
        Maarten Lankhorst <maarten.lankhorst@linux.intel.com>,
        Maxime Ripard <mripard@kernel.org>,
        David Airlie <airlied@linux.ie>,
        open list <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Fri, Jul 29, 2022 at 8:27 AM Dmitry Osipenko
<dmitry.osipenko@collabora.com> wrote:
>
> On 7/26/22 20:50, Rob Clark wrote:
> > +/**
> > + * drm_gem_lru_move_tail_locked - move the object to the tail of the LRU
> > + *
> > + * If the object is already in this LRU it will be moved to the
> > + * tail.  Otherwise it will be removed from whichever other LRU
> > + * it is in (if any) and moved into this LRU.
> > + *
> > + * Call with LRU lock held.
> > + *
> > + * @lru: The LRU to move the object into.
> > + * @obj: The GEM object to move into this LRU
> > + */
> > +void
> > +drm_gem_lru_move_tail_locked(struct drm_gem_lru *lru, struct drm_gem_object *obj)
> > +{
> > +     lockdep_assert_held_once(lru->lock);
> > +
> > +     if (obj->lru)
> > +             lru_remove(obj);
>
> The obj->lru also needs to be locked if lru != obj->lru, isn't it? And
> then we should add lockdep_assert_held_once(obj->lru->lock).
>

It is expected (mentioned in comment on drm_gem_lru::lock) that all
lru's are sharing the same lock.  Possibly that could be made more
obvious?  Having per-lru locks wouldn't really work for accessing the
single drm_gem_object::lru_node.

BR,
-R