Re: [PATCH 1/1] watchdog: pcipcwd_show_card_info: wrong format string

[Heinrich Schuchardt]

On 11/05/2016 04:29 PM, Guenter Roeck wrote:
> On 11/05/2016 07:50 AM, Heinrich Schuchardt wrote:
>> fw_rev_major and fw_rev_minor are defined as int.
>> Use %d to print them.
>>
>> Signed-off-by: Heinrich Schuchardt <xypron.glpk@xxxxxx>
>> ---
>>  drivers/watchdog/pcwd_pci.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/watchdog/pcwd_pci.c b/drivers/watchdog/pcwd_pci.c
>> index c0d07ee..e1fbbf6 100644
>> --- a/drivers/watchdog/pcwd_pci.c
>> +++ b/drivers/watchdog/pcwd_pci.c
>> @@ -234,7 +234,7 @@ static void pcipcwd_show_card_info(void)
>>      got_fw_rev = send_command(CMD_GET_FIRMWARE_VERSION, &fw_rev_major,
>>                                  &fw_rev_minor);
>>      if (got_fw_rev)
>> -        sprintf(fw_ver_str, "%u.%02u", fw_rev_major, fw_rev_minor);
>> +        sprintf(fw_ver_str, "%d.%02d", fw_rev_major, fw_rev_minor);
>>      else
>>          sprintf(fw_ver_str, "<card no answer>");
>>
>>
> Hmm ... I don't think that a negative version number makes much sense.
> Turns out inb() returns a char on some architectures, meaning it is signed,
> meaning it _could_ return a negative number if the version number is 128
> or above. I don't want to risk us reporting version number -128.-110 just
> to make compilers happy.

send_command uses inb_p to read single bytes.

The signature of inb_p is
u8 inb_p(unsigned long addr)

So fw_rev_major and fw_rev_minor should always be
in the range of 0..255.

Regards

Heinrich