Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
References: <1546318276-18993-1-git-send-email-yong.wu@mediatek.com> <1546318276-18993-3-git-send-email-yong.wu@mediatek.com>
In-Reply-To: <1546318276-18993-3-git-send-email-yong.wu@mediatek.com>
From:   Evan Green <evgreen@chromium.org>
Date:   Mon, 25 Feb 2019 15:53:44 -0800
Message-ID: <CAE=gft7mWs1jc+KQQcPuuhcmxpFRnRBTt35+uC3F20N6R554bQ@mail.gmail.com>
Subject: Re: [PATCH 02/13] driver core: Remove the link if there is no driver
 with AUTO flag
To:     Yong Wu <yong.wu@mediatek.com>
Cc:     Joerg Roedel <joro@8bytes.org>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        Matthias Brugger <matthias.bgg@gmail.com>,
        Rob Herring <robh+dt@kernel.org>,
        Robin Murphy <robin.murphy@arm.com>,
        Tomasz Figa <tfiga@google.com>,
        Will Deacon <will.deacon@arm.com>,
        linux-mediatek@lists.infradead.org, srv_heupstream@mediatek.com,
        "open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS" 
        <devicetree@vger.kernel.org>, LKML <linux-kernel@vger.kernel.org>,
        linux-arm-kernel@lists.infradead.org,
        iommu@lists.linux-foundation.org, Arnd Bergmann <arnd@arndb.de>,
        yingjoe.chen@mediatek.com, youlin.pei@mediatek.com,
        Nicolas Boichat <drinkcat@chromium.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Mon, Dec 31, 2018 at 8:52 PM Yong Wu <yong.wu@mediatek.com> wrote:
>
> DL_FLAG_AUTOREMOVE_CONSUMER/SUPPLIER means "Remove the link
> automatically on consumer/supplier driver unbind", that means we should
> remove whole the device_link when there is no this driver no matter what
> the ref_count of the link is.
>
> CC: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> Signed-off-by: Yong Wu <yong.wu@mediatek.com>
> ---
> The ref_count of our device_link normally is over 1. When the consumer
> device driver is removed, whole the device_link should be removed.
> Thus, I add this patch.
> ---

I will admit to reading about device links for the first time while
reviewing this patch, but I don't really get this. Why use a kref at
all if we're just going to ignore its value? For instance, I see that
if you call device_link_add() with the same supplier and consumer, it
uses the kref to return the same link. That machinery is broken with
your change. Although I don't see any uses of it, you might also
expect a supplier or consumer could do a kref_get() on the link it got
back from device_link_add(), and have a reasonable expectation that
the link wouldn't be freed out from under it. This would also be
broken.

Can you explain why your device_links normally have a reference count
>1, and why those additional references can't be cleaned up in an
orderly fashion?

(To be honest, I don't really understand the case for the AUTOREMOVE
flags at all. Is there some case where the party that set up the link
can't tear it down? Or is this a way to devm_ify the link, where devm
itself doesn't work because the links themselves stall out that
mechanism?)