Re: [PATCH v3] media: si2168: Refactor command setup code

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 12/07/2019 17:47, Brad Love wrote:

> On 04/07/2019 05.33, Marc Gonzalez wrote:
>
>> Refactor the command setup code, and let the compiler determine
>> the size of each command.
>>
>> Reviewed-by: Jonathan Neuschäfer <j.neuschaefer@xxxxxxx>
>> Signed-off-by: Marc Gonzalez <marc.w.gonzalez@xxxxxxx>
>> ---
>> Changes from v1:
>> - Use a real function to populate struct si2168_cmd *cmd, and a trivial
>> macro wrapping it (macro because sizeof).
>> Changes from v2:
>> - Fix header mess
>> - Add Jonathan's tag
>> ---
>>  drivers/media/dvb-frontends/si2168.c | 146 +++++++++------------------
>>  1 file changed, 45 insertions(+), 101 deletions(-)
>>
>> diff --git a/drivers/media/dvb-frontends/si2168.c b/drivers/media/dvb-frontends/si2168.c
>> index c64b360ce6b5..5e81e076369c 100644
>> --- a/drivers/media/dvb-frontends/si2168.c
>> +++ b/drivers/media/dvb-frontends/si2168.c
>> @@ -12,6 +12,16 @@
>>  
>>  static const struct dvb_frontend_ops si2168_ops;
>>  
>> +static void cmd_setup(struct si2168_cmd *cmd, char *args, int wlen, int rlen)
>> +{
>> +	memcpy(cmd->args, args, wlen);
>> +	cmd->wlen = wlen;
>> +	cmd->rlen = rlen;
>> +}
> 
> struct si2168_cmd.args is u8, not char.

Yes, struct si2168_cmd.args is an u8 array.
However, string literals such as "\xa0\x01" are char arrays.
memcpy() ignores the types altogether.

> I also think const should apply to the pointer.

I can do that, even though it's obvious we're not writing
to args in the trivial cmd_setup() body.

>> +#define CMD_SETUP(cmd, args, rlen) \
>> +	cmd_setup(cmd, args, sizeof(args) - 1, rlen)
>> +
> 
> This is only a valid helper if args is a null terminated string.

You say that because of the "-1" arithmetic, I assume?

> It just so happens that every instance in this driver is,

FWIW, there are 2 calls where it is not.
memcpy(cmd.args, &fw->data[(fw->size - remaining) + 1], len);
memcpy(cmd.args, &fw->data[fw->size - remaining], len);

> but that could be a silent pitfall if someone used a u8 array
> with this macro.

Actually, the compiler warns if we pass an u8 array instead of
a char array. IMO, the type is actually a good thing, since it
warns for cases where we don't use a string literal.

> Otherwise I'm ok with the refactoring.

I'll see what I can do...

Regards.



[Index of Archives]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [Linux for Sparc]     [IETF Annouce]     [Security]     [Bugtraq]     [Linux MIPS]     [ECOS]     [Asterisk Internet PBX]     [Linux API]

  Powered by Linux