Received-SPF: pass (google.com: domain of linux-kernel+bounces-133037-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.48.161 as permitted sender) client-ip=147.75.48.161;
Precedence: bulk
MIME-Version: 1.0
References: <20240404-dt-cleanup-free-v1-0-c60e6cba8da9@kernel.org>
 <20240404-dt-cleanup-free-v1-3-c60e6cba8da9@kernel.org> <CAGETcx8Wd5OsHWiGSASWkQQtof0D-ScwYsvq9hWizV3DFC27gA@mail.gmail.com>
In-Reply-To: <CAGETcx8Wd5OsHWiGSASWkQQtof0D-ScwYsvq9hWizV3DFC27gA@mail.gmail.com>
From: Rob Herring <robh@kernel.org>
Date: Fri, 5 Apr 2024 08:00:51 -0500
Message-ID: <CAL_JsqK4QpFmLi0WVG0Vpxv82Y1eWxbGS7ESWBhPoKg+WFV5Tw@mail.gmail.com>
Subject: Re: [PATCH 3/3] of: Use scope based of_node_put() cleanups
To: Saravana Kannan <saravanak@google.com>
Cc: Jonathan Cameron <jic23@kernel.org>, devicetree@vger.kernel.org, 
	linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Thu, Apr 4, 2024 at 6:22=E2=80=AFPM Saravana Kannan <saravanak@google.co=
m> wrote:
>
> On Thu, Apr 4, 2024 at 7:15=E2=80=AFAM Rob Herring <robh@kernel.org> wrot=
e:
> >
> > Use the relatively new scope based of_node_put() cleanup to simplify
> > function exit handling. Doing so reduces the chances of forgetting an
> > of_node_put() and simplifies error paths by avoiding the need for goto
> > statements.
> >
> > Signed-off-by: Rob Herring <robh@kernel.org>
> > ---
> >  drivers/of/address.c  | 60 ++++++++++++++++---------------------------=
--------
> >  drivers/of/property.c | 22 ++++++-------------
> >  2 files changed, 26 insertions(+), 56 deletions(-)
> >
> > diff --git a/drivers/of/address.c b/drivers/of/address.c
> > index ae46a3605904..f7b2d535a6d1 100644
> > --- a/drivers/of/address.c
> > +++ b/drivers/of/address.c
> > @@ -491,7 +491,6 @@ static u64 __of_translate_address(struct device_nod=
e *dev,
> >                                   const __be32 *in_addr, const char *rp=
rop,
> >                                   struct device_node **host)
> >  {
> > -       struct device_node *parent =3D NULL;
> >         struct of_bus *bus, *pbus;
> >         __be32 addr[OF_MAX_ADDR_CELLS];
> >         int na, ns, pna, pns;
> > @@ -504,7 +503,7 @@ static u64 __of_translate_address(struct device_nod=
e *dev,
> >
> >         *host =3D NULL;
> >         /* Get parent & match bus type */
> > -       parent =3D get_parent(dev);
> > +       struct device_node *parent __free(device_node) =3D get_parent(d=
ev);
>
> Can we leave the variable definition where it was? We generally define
> all the variables up top. So, defining the one variable in the middle
> feels weird. I at least get when we do this inside for/if blocks. But
> randomly in the middle feels weird.

There's an 'of_node_get(dev);' before this. Ordering wise, we need to
hold the ref on the child before we get its parent. I suppose I can
also convert that to use the cleanups. I'll have to add another local
ptr to do that though.

>
> Similar comments in other places. Since both kfree() and of_put() can
> both handle NULL pointers, I'd be surprised if we HAVE to combine
> these lines.

https://lore.kernel.org/all/CAHk-=3DwgRHiV5VSxtfXA4S6aLUmcQYEuB67u3BJPJPtuE=
Ss1JyA@mail.gmail.com/