linux-kernel - Re: [PATCH] pinctrl: st: Fix Sparse error

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

Message-ID: <54297645.7030208@linaro.org>
Date:	Mon, 29 Sep 2014 16:09:57 +0100
From:	Daniel Thompson <daniel.thompson@...aro.org>
To:	Srinivas Kandagatla <srinivas.kandagatla@...aro.org>,
	Pramod Gurav <pramod.gurav@...rtplayin.com>,
	linux-kernel@...r.kernel.org
CC:	Maxime Coquelin <maxime.coquelin@...com>,
	Patrice Chotard <patrice.chotard@...com>,
	Linus Walleij <linus.walleij@...aro.org>
Subject: Re: [PATCH] pinctrl: st: Fix Sparse error

On 29/09/14 15:46, Srinivas Kandagatla wrote:
> On 29/09/14 14:48, Pramod Gurav wrote:
>> This change fixes below sparse error,
>> drivers/pinctrl/pinctrl-st.c:1515:31: error: incompatible types for
>> operation (>)
>> drivers/pinctrl/pinctrl-st.c:1515:31:    left side has type void
>> [noderef] <asn:2>*irqmux_base
>> drivers/pinctrl/pinctrl-st.c:1515:31:    right side has type int
>>
>> The fix is done by removing a check on info->irqmux_base as
>> info->irqmux_base has already been checked  for error when allocating it.
>> Hence there is no need to redo the check.
>>
>> Cc: Maxime Coquelin <maxime.coquelin@...com>
>> Cc: Patrice Chotard <patrice.chotard@...com>
>> CC: Linus Walleij <linus.walleij@...aro.org>
>> Cc: Srinivas Kandagatla <srinivas.kandagatla@...aro.org>
>> Signed-off-by: Pramod Gurav <pramod.gurav@...rtplayin.com>
>> ---
>>   drivers/pinctrl/pinctrl-st.c |    2 +-
>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/pinctrl/pinctrl-st.c b/drivers/pinctrl/pinctrl-st.c
>> index 5475374..ddeaeda 100644
>> --- a/drivers/pinctrl/pinctrl-st.c
>> +++ b/drivers/pinctrl/pinctrl-st.c
>> @@ -1512,7 +1512,7 @@ static int st_gpiolib_register_bank(struct
>> st_pinctrl *info,
>>                            gpio_irq, st_gpio_irq_handler);
>>       }
>>
>> -    if (info->irqmux_base > 0 || gpio_irq > 0) {
>> +    if (gpio_irq > 0) {
>>           err = gpiochip_irqchip_add(&bank->gpio_chip, &st_gpio_irqchip,
>>                          0, handle_simple_irq,
>>                          IRQ_TYPE_LEVEL_LOW);
> 
> This is not the correct fix. Please see why irqmux_base and gpio_irq are
> used in the driver.
> You are breaking the logic here...
> 
> please read the below comment from the code.
> 
>     /**
>      * GPIO bank can have one of the two possible types of
>      * interrupt-wirings.
>      *
>      * First type is via irqmux, single interrupt is used by multiple
>      * gpio banks. This reduces number of overall interrupts numbers
>      * required. All these banks belong to a single pincontroller.
>      *          _________
>      *         |       |----> [gpio-bank (n)    ]
>      *         |       |----> [gpio-bank (n + 1)]
>      *    [irqN]-- | irq-mux |----> [gpio-bank (n + 2)]
>      *         |       |----> [gpio-bank (...  )]
>      *         |_________|----> [gpio-bank (n + 7)]
>      *
>      * Second type has a dedicated interrupt per each gpio bank.
>      *
>      *    [irqN]----> [gpio-bank (n)]
>      */
> 
> 
> so irqmux_base is first type and gpio_irq is second type.
> if you remove check for irqmux_base here you would end up NOT adding
> irqchip the gpiochip in first type so you break the existing logic here.
> 
> 
> I think the correct fix is:
> 
> diff --git a/drivers/pinctrl/pinctrl-st.c b/drivers/pinctrl/pinctrl-st.c
> index 5475374..4060c30 100644
> --- a/drivers/pinctrl/pinctrl-st.c
> +++ b/drivers/pinctrl/pinctrl-st.c
> @@ -1512,7 +1512,7 @@ static int st_gpiolib_register_bank(struct
> st_pinctrl *info,
>                                              gpio_irq,
> st_gpio_irq_handler);
>         }
> 
> -       if (info->irqmux_base > 0 || gpio_irq > 0) {
> +       if (!IS_ERR(info->irqmux_base) || gpio_irq > 0) {
>                 err = gpiochip_irqchip_add(&bank->gpio_chip,
> &st_gpio_irqchip,
>                                            0, handle_simple_irq,
>                                            IRQ_TYPE_LEVEL_LOW);

IS_ERR() should be irrelavent because the allocation code bombs out on
error. Shouldn't this just be a NULL pointer check?

  if(info->irqmux_base || gpio_irq > 0)


Daniel.

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/