Re: can/sja1000: potential issue in sja1000.c?

[Marc Kleine-Budde <mkl@pengutronix.de>]

[-- Attachment #1: Type: text/plain, Size: 1204 bytes --]

On 02/10/2012 04:32 PM, Sebastian Haas wrote:
>>> Is 0xff a valid value for REG_SR in the first place? If not we can for
>>> example exit the interrupt handler.
>>
>> Is the access to the address space of an unplugged card defined at all?
>> Or may it even hang the system?
> AFAIK PCMCIA is based on ISA and therefor it is absolutely valid to read
> and even write into this address space. It is up to user to check if the
> device is still there and accessiable.
> 
> For that ems_pcmcia does something like this:
> static irqreturn_t ems_pcmcia_interrupt(int irq, void *dev_id)
> {
> ...
>   /* Card not present */
>   if (readw(card->base_addr) != 0xAA55)
>     return IRQ_HANDLED;
> 
>   Call sja1000_interrupt() ...
> ...
> }

This code just makes the race window smaller, but you still have one.
It's a good idea to limit the number of rounds in the while loop in the
interrupt handler.

Marc

-- 
Pengutronix e.K.                  | Marc Kleine-Budde           |
Industrial Linux Solutions        | Phone: +49-231-2826-924     |
Vertretung West/Dortmund          | Fax:   +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686  | http://www.pengutronix.de   |


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 262 bytes --]