Re: [Qemu-devel] [PATCH v2 04/12] i2c: Add a length check to the SMBus write handling

[Corey Minyard <minyard@acm.org>]

On 11/20/18 9:33 AM, Peter Maydell wrote:
> On 15 November 2018 at 19:24,  <minyard@acm.org> wrote:
>> From: Corey Minyard <cminyard@mvista.com>
>>
>> Avoid an overflow.
>>
>> Signed-off-by: Corey Minyard <cminyard@mvista.com>
>> ---
>>   hw/i2c/smbus_slave.c | 6 +++++-
>>   1 file changed, 5 insertions(+), 1 deletion(-)
>>
>> diff --git a/hw/i2c/smbus_slave.c b/hw/i2c/smbus_slave.c
>> index 83ca041b5d..fa988919d8 100644
>> --- a/hw/i2c/smbus_slave.c
>> +++ b/hw/i2c/smbus_slave.c
>> @@ -182,7 +182,11 @@ static int smbus_i2c_send(I2CSlave *s, uint8_t data)
>>       switch (dev->mode) {
>>       case SMBUS_WRITE_DATA:
>>           DPRINTF("Write data %02x\n", data);
>> -        dev->data_buf[dev->data_len++] = data;
>> +        if (dev->data_len >= sizeof(dev->data_buf)) {
>> +            BADF("Too many bytes sent\n");
>> +        } else {
>> +            dev->data_buf[dev->data_len++] = data;
>> +        }
>>           break;
> Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
>
> What happens on a real device in this situation ?

It's device specific.  Some devices (most eeproms, I suspect) will just keep
taking data, wrapping around when you hit the end of the memory. Some
devices (IPMI BMCs, generally) will ignore the extra data.  Some devices
may return a NAK on a byte if it gets too much data.  The specification
says:

    The slave device detects an invalid command or invalid data. In this
    case the slave
    device must NACK the received byte. The master upon detection of
    this condition
    must generate a STOP condition and retry the transaction.

So a NAK may be appropriate here, but it's kind of fuzzy.  Since generating
a NAK is more complicated, I would guess that most devices don't.

-corey


> thanks
> -- PMM