On Thu, 2014-11-27 at 17:18 +0100, Luis de Bethencourt wrote:
> Fixing 80 character limit warnings in octeon/ethernet-rx.c

Hello again Luis.

Another thing you might consider is to  align
multiple line statements to the appropriate
open parenthesis.

> diff --git a/drivers/staging/octeon/ethernet-rx.c 
> b/drivers/staging/octeon/ethernet-rx.c
[]
> @@ -126,13 +127,15 @@ static inline int cvm_oct_check_rcv_error(cvmx_wqe_t 
> *work)
>  
>                       if (*ptr == 0xd5) {
>                               /*
> -                               printk_ratelimited("Port %d received 0xd5 
> preamble\n", work->ipprt);
> +                               printk_ratelimited("Port %d received 0xd5 
> preamble\n",
> +                                       work->ipprt);
>                                */

This is in a commented out block, but this
would look better like:

                printk_ratelimited("format",
                                   args...);


> @@ -212,17 +215,20 @@ static int cvm_oct_napi_poll(struct napi_struct *napi, 
> int budget)
>               did_work_request = 0;
>               if (work == NULL) {
>                       union cvmx_pow_wq_int wq_int;
> +
>                       wq_int.u64 = 0;
>                       wq_int.s.iq_dis = 1 << pow_receive_group;
>                       wq_int.s.wq_int = 1 << pow_receive_group;
>                       cvmx_write_csr(CVMX_POW_WQ_INT, wq_int.u64);
>                       break;
>               }
> -             pskb = (struct sk_buff 
> **)(cvm_oct_get_buffer_ptr(work->packet_ptr) - sizeof(void *));
> +             pskb = (struct sk_buff 
> **)(cvm_oct_get_buffer_ptr(work->packet_ptr) -
> +                     sizeof(void *));

cvm_oct_get_buffer_ptr returns a void pointer
so it doesn't need a cast.

a possible fix is just to remove the cast

                pskb = cvm_oct_get_buffer_ptr(work->packet_ptr) - sizeof(void 
*);

>               prefetch(pskb);

If the cast type above to a ** is to be believed,
maybe this should be prefetch(*pskb)

If you do these, they should be separate patches.

_______________________________________________
devel mailing list
de...@linuxdriverproject.org
http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel

Reply via email to