On Sat, Jun 08, 2019 at 03:27:46PM +0800, Hao Xu wrote:
> modify void* to void * for #define inb_p(a) readq((void*)a)
> and #define outb_p(d,a) writeq(d,(void*)a)
>
> Signed-off-by: Hao Xu <haoxu.linuxkernel@xxxxxxxxx>
> ---
> drivers/staging/kpc2000/kpc2000_i2c.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/staging/kpc2000/kpc2000_i2c.c b/drivers/staging/kpc2000/kpc2000_i2c.c
> index a434dd0..de3a0c8 100644
> --- a/drivers/staging/kpc2000/kpc2000_i2c.c
> +++ b/drivers/staging/kpc2000/kpc2000_i2c.c
> @@ -124,9 +124,9 @@ struct i2c_device {
>
> // FIXME!
> #undef inb_p
> -#define inb_p(a) readq((void*)a)
> +#define inb_p(a) readq((void *)a)
> #undef outb_p
> -#define outb_p(d,a) writeq(d,(void*)a)
> +#define outb_p(d,a) writeq(d,(void *)a)
Alternatively to fixing up the style here, did you consider just
removing these two macros altogether and calling [read|write]q
directly throughout the kpc_i2c driver (per the '//FIXME' comment)?
Unless, I'm misunderstanding something, these macros are shadowing the
functions [in|out]b_p, which already exist in io.h. [in|out]b_p are for
8-bit i/o transactions and [read|write]q are for 64-bit transactions, so
shadowing the original [in|out]b_p with something that actually does
64-bit transactions is probably potentially misleading here.
_______________________________________________
devel mailing list
devel@xxxxxxxxxxxxxxxxxxxxxx
http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
