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: <8dvhtgydaq7tflf8q4rq4fpu.1649846600874@email.android.com>
In-Reply-To: <8dvhtgydaq7tflf8q4rq4fpu.1649846600874@email.android.com>
From:   Geert Uytterhoeven <geert@linux-m68k.org>
Date:   Wed, 13 Apr 2022 13:23:54 +0200
Message-ID: <CAMuHMdXkUKD_+A2B6-x0C5SSva6fr2Ht0N73t9gboOBp5z13xw@mail.gmail.com>
Subject: Re: [PATCH V2] clk: renesas: Fix memory leak of 'cpg'
To:     =?UTF-8?B?55m95rWp5paH?= <baihaowen@meizu.com>
Cc:     Michael Turquette <mturquette@baylibre.com>,
        Stephen Boyd <sboyd@kernel.org>,
        Linux-Renesas <linux-renesas-soc@vger.kernel.org>,
        linux-clk <linux-clk@vger.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 8BIT
Precedence: bulk

Hi Haowen,

On Wed, Apr 13, 2022 at 12:44 PM 白浩文 <baihaowen@meizu.com> wrote:
> But this function __init r8a73a4_cpg_clocks_init will auto free by system after boot

The memory containing the code for the function
r8a73a4_cpg_clocks_init() will indeed be freed.  But the data
structures allocated and prepared by the function will continue to
exist afterwards.

> On Wed, Apr 13, 2022 at 11:24 AM baihaowen <baihaowen@meizu.com> wrote:
> > 在 4/13/22 4:41 PM, Geert Uytterhoeven 写道:
> > > On Wed, Apr 13, 2022 at 10:30 AM Haowen Bai <baihaowen@meizu.com> wrote:
> > >> Fix this issue by freeing the cpg when exiting the function in the
> > >> error/normal path.
> > >>
> > >> Signed-off-by: Haowen Bai <baihaowen@meizu.com>

> > >> --- a/drivers/clk/renesas/clk-r8a73a4.c
> > >> +++ b/drivers/clk/renesas/clk-r8a73a4.c
> > >> @@ -215,7 +215,7 @@ static void __init r8a73a4_cpg_clocks_init(struct device_node *np)
> > >>
> > >>         cpg->reg = of_iomap(np, 0);
> > >>         if (WARN_ON(cpg->reg == NULL))
> > >> -               return;
> > >> +               goto out_free_cpg;
> > > Note that this is a fatal error, i.e. no chance the system will survive this,
> > > so cleaning up is moot.
> > >
> > >>         for (i = 0; i < num_clks; ++i) {
> > >>                 const char *name;
> > >> @@ -233,6 +233,9 @@ static void __init r8a73a4_cpg_clocks_init(struct device_node *np)
> > >>         }
> > >>
> > >>         of_clk_add_provider(np, of_clk_src_onecell_get, &cpg->data);
> > >> +out_free_cpg:
> > >> +       kfree(cpg);
> > >> +       kfree(clks);
> > > Both cpg and clks are still used after returning from this function,
> > > through the registered clocks and clock provider.
> > >
> > >>  }
> > >>  CLK_OF_DECLARE(r8a73a4_cpg_clks, "renesas,r8a73a4-cpg-clocks",
> > >>                r8a73a4_cpg_clocks_init);
> > > NAKed-by: Geert Uytterhoeven <geert+renesas@glider.be>
>
> > Could you show me how and when cpg & clks free ?
>
> They are never freed, as they stay in-use for the lifetime of the system.

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds