linux-kernel - Re: [PATCH v5 3/3] mfd: dln2: Send restart command if HW init fails

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <CAPHBK3bruRbdXEO-+R_913LxiaOwNTR2-65T8WD+TAi23OcAFg@mail.gmail.com>
Date:   Thu, 6 Jul 2017 11:01:54 -1000
From:   Jack Andersen <jackoalan@...il.com>
To:     Lee Jones <lee.jones@...aro.org>
Cc:     linux-kernel@...r.kernel.org, linux-iio@...r.kernel.org,
        Jonathan Cameron <jic23@...nel.org>,
        Hartmut Knaack <knaack.h@....de>,
        Lars-Peter Clausen <lars@...afoo.de>,
        Peter Meerwald-Stadler <pmeerw@...erw.net>
Subject: Re: [PATCH v5 3/3] mfd: dln2: Send restart command if HW init fails

The restart command doesn't behave like other commands. It doesn't
return an error code (which indicates a failed transfer in the dln2
module so the probe fails). Attached is a system log snippet to give
you a better idea.

I'm able to consistently reproduce this behavior on 4.12 by
registering an ADC event on the DLN2 then restarting the host without
powering off. Without the dln2_restart patch, it ends on the `dln2:
probe of 1-2:1.0 failed with error -110' line. Shortly after issuing
the restart, the kernel indicates a device disconnect, and it
reconnects about 400ms later, with a usable device state.

On 5 July 2017 at 21:09, Lee Jones <lee.jones@...aro.org> wrote:
> On Wed, 05 Jul 2017, Jack Andersen wrote:
>
>> I have mixed feelings about making this patch. The device appears to
>> lose its ability to send command responses when the driver
>> unexpectedly goes offline (probably overflowing an internal send
>> buffer).
>>
>> The restart command is a send-only way to recover the device.
>> What concerns me is if there's another issue causing init to fail and
>> a restart doesn't fix, this will trigger an endless probe loop. I'm
>> not sure if there's a better way to discern this condition.
>
> Why does dln2_restart() cause an unbind/bind?
>
>> On 5 July 2017 at 11:12, Jack Andersen <jackoalan@...il.com> wrote:
>> > There's an issue with reinitializing DLN2 hardware if ADC channel event
>> > configurations aren't reset before suspending the driver. A physical reconnect
>> > of the device solves the problem, but restarting the machine without a power
>> > cycle leaves the device state inconsistent.
>> >
>> > This patch transmits a restart command in the event of a failed init.
>> > To the kernel's perspective, this is the rough equivalent of a disconnect and
>> > reconnect, so another probe occurs shortly afterwards.
>> >
>> > Signed-off-by: Jack Andersen <jackoalan@...il.com>
>> > ---
>> >  drivers/mfd/dln2.c | 11 ++++++++++-
>> >  1 file changed, 10 insertions(+), 1 deletion(-)
>> >
>> > diff --git a/drivers/mfd/dln2.c b/drivers/mfd/dln2.c
>> > index a22ab8c..1eb0056 100644
>> > --- a/drivers/mfd/dln2.c
>> > +++ b/drivers/mfd/dln2.c
>> > @@ -40,6 +40,7 @@ struct dln2_response {
>> >  #define DLN2_GENERIC_CMD(cmd)          DLN2_CMD(cmd, DLN2_GENERIC_MODULE_ID)
>> >  #define CMD_GET_DEVICE_VER             DLN2_GENERIC_CMD(0x30)
>> >  #define CMD_GET_DEVICE_SN              DLN2_GENERIC_CMD(0x31)
>> > +#define CMD_RESTART                    DLN2_GENERIC_CMD(0x43)
>> >
>> >  #define DLN2_HW_ID                     0x200
>> >  #define DLN2_USB_TIMEOUT               200     /* in ms */
>> > @@ -531,6 +532,12 @@ int dln2_transfer(struct platform_device *pdev, u16 cmd,
>> >  }
>> >  EXPORT_SYMBOL(dln2_transfer);
>> >
>> > +static int dln2_restart(struct dln2_dev *dln2)
>> > +{
>> > +       return _dln2_transfer(dln2, DLN2_HANDLE_CTRL, CMD_RESTART,
>> > +                             NULL, 0, NULL, NULL);
>> > +}
>> > +
>> >  static int dln2_check_hw(struct dln2_dev *dln2)
>> >  {
>> >         int ret;
>> > @@ -782,7 +789,9 @@ static int dln2_probe(struct usb_interface *interface,
>> >
>> >         ret = dln2_hw_init(dln2);
>> >         if (ret < 0) {
>> > -               dev_err(dev, "failed to initialize hardware\n");
>> > +               dev_err(dev,
>> > +                       "failed to initialize hardware; attempting restart\n");
>> > +               dln2_restart(dln2);
>> >                 goto out_stop_rx;
>> >         }
>> >
>> >
>
> --
> Lee Jones
> Linaro STMicroelectronics Landing Team Lead
> Linaro.org │ Open source software for ARM SoCs
> Follow Linaro: Facebook | Twitter | Blog

View attachment "dln2-log.txt" of type "text/plain" (3536 bytes)