diff mbox

gpio: bcm-kona: memory corruption fix

Message ID 1418074473-10811-1-git-send-email-olof@lixom.net
State Accepted
Headers show

Commit Message

Olof Johansson Dec. 8, 2014, 9:34 p.m. UTC
In one instance the base address of the internal controller state
structure is passed into a function doing writel to an offset of
the pointer passed in is used, instead of the register base.

Once I found the bug, I also went back to check for other sparse
warnings in the file, but found none. This one, however, triggered:

drivers/gpio/gpio-bcm-kona.c:552:47: warning: incorrect type in argument 1 (different address spaces)
drivers/gpio/gpio-bcm-kona.c:552:47:    expected void [noderef] <asn:2>*reg_base
drivers/gpio/gpio-bcm-kona.c:552:47:    got struct bcm_kona_gpio *kona_gpio
drivers/gpio/gpio-bcm-kona.c:556:47: warning: incorrect type in argument 1 (different address spaces)
drivers/gpio/gpio-bcm-kona.c:556:47:    expected void [noderef] <asn:2>*reg_base
drivers/gpio/gpio-bcm-kona.c:556:47:    got struct bcm_kona_gpio *kona_gpio

As far as I can tell, this bug has been here for a long time and is
not new, but I found it when hunting down another heisenbug on this
platform.

Not marking for stable since I am unaware of any upstream user of kona
on a product that would benefit from it.

Signed-off-by: Olof Johansson <olof@lixom.net>
---
 drivers/gpio/gpio-bcm-kona.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

Comments

Ray Jui Dec. 8, 2014, 11:35 p.m. UTC | #1
On 12/8/2014 1:34 PM, Olof Johansson wrote:
> In one instance the base address of the internal controller state
> structure is passed into a function doing writel to an offset of
> the pointer passed in is used, instead of the register base.
>
> Once I found the bug, I also went back to check for other sparse
> warnings in the file, but found none. This one, however, triggered:
>
> drivers/gpio/gpio-bcm-kona.c:552:47: warning: incorrect type in argument 1 (different address spaces)
> drivers/gpio/gpio-bcm-kona.c:552:47:    expected void [noderef] <asn:2>*reg_base
> drivers/gpio/gpio-bcm-kona.c:552:47:    got struct bcm_kona_gpio *kona_gpio
> drivers/gpio/gpio-bcm-kona.c:556:47: warning: incorrect type in argument 1 (different address spaces)
> drivers/gpio/gpio-bcm-kona.c:556:47:    expected void [noderef] <asn:2>*reg_base
> drivers/gpio/gpio-bcm-kona.c:556:47:    got struct bcm_kona_gpio *kona_gpio
>
> As far as I can tell, this bug has been here for a long time and is
> not new, but I found it when hunting down another heisenbug on this
> platform.
>
> Not marking for stable since I am unaware of any upstream user of kona
> on a product that would benefit from it.
>
> Signed-off-by: Olof Johansson <olof@lixom.net>

Acked-by: Ray Jui <rjui@broadcom.com>

> ---
>   drivers/gpio/gpio-bcm-kona.c |    4 ++--
>   1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/gpio/gpio-bcm-kona.c b/drivers/gpio/gpio-bcm-kona.c
> index de0801e..d552cca 100644
> --- a/drivers/gpio/gpio-bcm-kona.c
> +++ b/drivers/gpio/gpio-bcm-kona.c
> @@ -549,11 +549,11 @@ static void bcm_kona_gpio_reset(struct bcm_kona_gpio *kona_gpio)
>   	/* disable interrupts and clear status */
>   	for (i = 0; i < kona_gpio->num_bank; i++) {
>   		/* Unlock the entire bank first */
> -		bcm_kona_gpio_write_lock_regs(kona_gpio, i, UNLOCK_CODE);
> +		bcm_kona_gpio_write_lock_regs(kona_gpio->reg_base, i, UNLOCK_CODE);
>   		writel(0xffffffff, reg_base + GPIO_INT_MASK(i));
>   		writel(0xffffffff, reg_base + GPIO_INT_STATUS(i));
>   		/* Now re-lock the bank */
> -		bcm_kona_gpio_write_lock_regs(kona_gpio, i, LOCK_CODE);
> +		bcm_kona_gpio_write_lock_regs(kona_gpio->reg_base, i, LOCK_CODE);
>   	}
>   }
>
>
--
To unsubscribe from this list: send the line "unsubscribe linux-gpio" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Alexandre Courbot Dec. 9, 2014, 3:06 a.m. UTC | #2
On Tue, Dec 9, 2014 at 6:34 AM, Olof Johansson <olof@lixom.net> wrote:
> In one instance the base address of the internal controller state
> structure is passed into a function doing writel to an offset of
> the pointer passed in is used, instead of the register base.
>
> Once I found the bug, I also went back to check for other sparse
> warnings in the file, but found none. This one, however, triggered:
>
> drivers/gpio/gpio-bcm-kona.c:552:47: warning: incorrect type in argument 1 (different address spaces)
> drivers/gpio/gpio-bcm-kona.c:552:47:    expected void [noderef] <asn:2>*reg_base
> drivers/gpio/gpio-bcm-kona.c:552:47:    got struct bcm_kona_gpio *kona_gpio
> drivers/gpio/gpio-bcm-kona.c:556:47: warning: incorrect type in argument 1 (different address spaces)
> drivers/gpio/gpio-bcm-kona.c:556:47:    expected void [noderef] <asn:2>*reg_base
> drivers/gpio/gpio-bcm-kona.c:556:47:    got struct bcm_kona_gpio *kona_gpio
>
> As far as I can tell, this bug has been here for a long time and is
> not new, but I found it when hunting down another heisenbug on this
> platform.

Ouch. Thanks for fixing this.

Reviewed-by: Alexandre Courbot <acourbot@nvidia.com>
--
To unsubscribe from this list: send the line "unsubscribe linux-gpio" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
diff mbox

Patch

diff --git a/drivers/gpio/gpio-bcm-kona.c b/drivers/gpio/gpio-bcm-kona.c
index de0801e..d552cca 100644
--- a/drivers/gpio/gpio-bcm-kona.c
+++ b/drivers/gpio/gpio-bcm-kona.c
@@ -549,11 +549,11 @@  static void bcm_kona_gpio_reset(struct bcm_kona_gpio *kona_gpio)
 	/* disable interrupts and clear status */
 	for (i = 0; i < kona_gpio->num_bank; i++) {
 		/* Unlock the entire bank first */
-		bcm_kona_gpio_write_lock_regs(kona_gpio, i, UNLOCK_CODE);
+		bcm_kona_gpio_write_lock_regs(kona_gpio->reg_base, i, UNLOCK_CODE);
 		writel(0xffffffff, reg_base + GPIO_INT_MASK(i));
 		writel(0xffffffff, reg_base + GPIO_INT_STATUS(i));
 		/* Now re-lock the bank */
-		bcm_kona_gpio_write_lock_regs(kona_gpio, i, LOCK_CODE);
+		bcm_kona_gpio_write_lock_regs(kona_gpio->reg_base, i, LOCK_CODE);
 	}
 }