Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <20210725094246.pkdpvl5aaaftur3a@pengutronix.de> <20210725103630.23864-1-paskripkin@gmail.com>
In-Reply-To: <20210725103630.23864-1-paskripkin@gmail.com>
From:   Yasushi SHOJI <yasushi.shoji@gmail.com>
Date:   Sun, 25 Jul 2021 22:27:37 +0900
Message-ID: <CAELBRWKfyOBanMBteO=LpL9R1QMp97zTYtKY689jeR2gDOa_Gw@mail.gmail.com>
Subject: Re: [PATCH] net: can: add missing urb->transfer_dma initialization
To:     Pavel Skripkin <paskripkin@gmail.com>
Cc:     mkl@pengutronix.de, wg@grandegger.com, linux-can@vger.kernel.org,
        linux-kernel@vger.kernel.org, netdev@vger.kernel.org,
        Yasushi SHOJI <yashi@spacecubics.com>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

Hi Pavel,

I've tested this patch on top of v5.14-rc2.  All good.

Tested-by: Yasushi SHOJI <yashi@spacecubics.com>

Some nitpicks.

On Sun, Jul 25, 2021 at 7:36 PM Pavel Skripkin <paskripkin@gmail.com> wrote:
>
> Yasushi reported, that his Microchip CAN Analyzer stopped working since
> commit 91c02557174b ("can: mcba_usb: fix memory leak in mcba_usb").
> The problem was in missing urb->transfer_dma initialization.
>
> In my previous patch to this driver I refactored mcba_usb_start() code to
> avoid leaking usb coherent buffers. To achive it, I passed local stack

achieve

> variable to usb_alloc_coherent() and then saved it to private array to
> correctly free all coherent buffers on ->close() call. But I forgot to
> inialize urb->transfer_dma with variable passed to usb_alloc_coherent().

initialize

> All of this was causing device to not work, since dma addr 0 is not valid
> and following log can be found on bug report page, which points exactly to
> problem described above.
>
> [   33.862175] DMAR: [DMA Write] Request device [00:14.0] PASID ffffffff fault addr 0 [fault reason 05] PTE Write access is not set
>
> Bug report: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=990850
>
> Reported-by: Yasushi SHOJI <yasushi.shoji@gmail.com>
> Fixes: 91c02557174b ("can: mcba_usb: fix memory leak in mcba_usb")
> Signed-off-by: Pavel Skripkin <paskripkin@gmail.com>
> ---
>  drivers/net/can/usb/mcba_usb.c | 2 ++
>  1 file changed, 2 insertions(+)
>
> diff --git a/drivers/net/can/usb/mcba_usb.c b/drivers/net/can/usb/mcba_usb.c
> index a45865bd7254..a1a154c08b7f 100644
> --- a/drivers/net/can/usb/mcba_usb.c
> +++ b/drivers/net/can/usb/mcba_usb.c
> @@ -653,6 +653,8 @@ static int mcba_usb_start(struct mcba_priv *priv)
>                         break;
>                 }
>
> +               urb->transfer_dma = buf_dma;
> +
>                 usb_fill_bulk_urb(urb, priv->udev,
>                                   usb_rcvbulkpipe(priv->udev, MCBA_USB_EP_IN),
>                                   buf, MCBA_USB_RX_BUFF_SIZE,
> --
> 2.32.0

Pavel, thanks again for your quick fix. :-)

Best,
--
               yashi