On Tue, Jan 8, 2019 at 7:59 PM Lubomir Rintel <lkundrak@xxxxx> wrote:
>
> Move the olpc-ec driver away from the X86 OLPC platform so that it could be
> used by the ARM based laptops too. Notably, the driver for the OLPC battery,
> which is also used on the ARM models, builds on this driver's interface.
>
> It is actually plaform independent: the OLPC EC commands with their argument
> and responses are mostly the same despite the delivery mechanism is
> different.
>
> Signed-off-by: Lubomir Rintel <lkundrak@xxxxx>
> Acked-by: Pavel Machek <pavel@xxxxxx>
> #define OLPC_F_PRESENT 0x01
> #define OLPC_F_DCON 0x02
> -#define OLPC_F_EC_WIDE_SCI 0x04
Still same concern, i.e. if the initially added definitions are
coupled together by some reason, why we split them?
As I told either move all, or none or put in commit message how
decoupling would benefit.
> + /* EC version 0x5f adds support for wide SCI mask */
> + if (ec->version >= 0x5f) {
> + __be16 ec_word = cpu_to_be16(bits);
> +
> + return olpc_ec_cmd(EC_WRITE_EXT_SCI_MASK, (void *) &ec_word, 2, NULL, 0);
No space is needed in "(void *)&ec_word".
> + } else {
> + unsigned char ec_byte = bits & 0xff;
Shouldn't be u8?
> + return olpc_ec_cmd(EC_WRITE_SCI_MASK, &ec_byte, 1, NULL, 0);
> + }
> +}
> +EXPORT_SYMBOL_GPL(olpc_ec_mask_write);
> +int olpc_ec_sci_query(u16 *sci_value)
> +{
> + struct olpc_ec_priv *ec = ec_priv;
> + int ret;
> +
> + if (WARN_ON(!ec))
> + return -ENODEV;
> +
> + /* EC version 0x5f adds support for wide SCI mask */
> + if (ec->version >= 0x5f) {
> + __be16 ec_word;
> +
> + ret = olpc_ec_cmd(EC_EXT_SCI_QUERY, NULL, 0, (void *)&ec_word, 2);
> + if (ret == 0)
> + *sci_value = be16_to_cpu(ec_word);
> + } else {
> + unsigned char ec_byte;
u8?
> +
> + ret = olpc_ec_cmd(EC_SCI_QUERY, NULL, 0, &ec_byte, 1);
> + if (ret == 0)
> + *sci_value = ec_byte;
> + }
> +
> + return ret;
> +}
> +EXPORT_SYMBOL_GPL(olpc_ec_sci_query);
--
With Best Regards,
Andy Shevchenko
