śr., 4 wrz 2024 o 17:56 Guenter Roeck <linux@xxxxxxxxxxxx> napisał(a):
>
> On 9/4/24 08:24, Patryk wrote:
> [ ... ]
> > >>> We'll need to add some code to detect that condition and
> > >>> refrain from clearing the status register if the chip doesn't support
> > >>> writes (or maybe ignore errors from the clear operation). Ignoring the
> > >>> error might be the easiest fix.
> > >>
> > >> I will apply this fix to our codebase then, unless I come up with a better idea.
> > >>
> > >
> > > If yyou don't mind and have the time, it would be great if you can send a
> > > patch to be applied to the upstream kernel. If not, please let me know and
> > > I'll write one myself.
> > >
> > > Thanks,
> > > Guenter
> > >
> >
> > Sure, I'll send a patch.
> >
>
> Thanks. Additional background: The ability to clear status registers was
> added with PMBus version 1.2, so it is not surprising that older chips
> don't support it. The best fix would probably be to read the revision
> register and use its content to decide if the individual status register
> should be cleared or if the clear_faults command should be executed.
> Something like
>
> struct pmbus_data {
> ...
> u8 revision;
> ...
> };
>
> In pmbus.h:
> #define PMBUS_REV_10 0x00
> #define PMBUS_REV_11 0x11
> #define PMBUS_REV_12 0x22
> #definw PMBUS_REV_13 0x33
>
> In pmbus_init_common():
>
> ret = i2c_smbus_read_byte_data(client, PMBUS_REVISION);
> if (!ret)
> data->revision = ret;
>
> In pmbus_show_boolean():
>
> if (regval) {
> if (data->revision >= PMBUS_REV_12)
> ret = _pmbus_write_byte_data(client, page, reg, regval);
> else
> ret = pmbus_clear_fault_page(client, page);
>
> if (ret)
> goto unlock;
> }
Thanks, I'll implement it according to your suggestion, test it (I
have few LTC devices) and if this is working I will send a patch.
BR
Patryk
