On 24-11-06 17:06:05, Heikki Krogerus wrote: > That allows the fields in those command data structures to > be easily validated. If an unsupported field is accessed, a > warning is generated. > > This will not force UCSI version checks to be made in every > place where these data structures are accessed, but it will > make it easier to pinpoint issues that are caused by the > unconditional accesses to those fields, and perhaps more > importantly, allow those issues to be noticed immediately. > > Signed-off-by: Heikki Krogerus <heikki.krogerus@xxxxxxxxxxxxxxx> > --- > > Changed since v1: > - The CONCAP macro now reads the correct member. > > I added the version to all the field definitions that need it instead > of adding a complete macro for each of those fields. I think that > should be enough to make the version clear to everybody during use as > well as review. > > --- > drivers/usb/typec/ucsi/psy.c | 28 ++-- > drivers/usb/typec/ucsi/trace.h | 28 ++-- > drivers/usb/typec/ucsi/ucsi.c | 121 +++++++------- > drivers/usb/typec/ucsi/ucsi.h | 252 +++++++++++++++++------------ > drivers/usb/typec/ucsi/ucsi_acpi.c | 7 +- > 5 files changed, 240 insertions(+), 196 deletions(-) > [...] > diff --git a/drivers/usb/typec/ucsi/ucsi.c b/drivers/usb/typec/ucsi/ucsi.c > index e430a0ca4a2b..974a441155e1 100644 > --- a/drivers/usb/typec/ucsi/ucsi.c > +++ b/drivers/usb/typec/ucsi/ucsi.c [...] > @@ -1778,7 +1775,7 @@ static int ucsi_init(struct ucsi *ucsi) > > /* Get PPM capabilities */ > command = UCSI_GET_CAPABILITY; > - ret = ucsi_send_command(ucsi, command, &ucsi->cap, sizeof(ucsi->cap)); > + ret = ucsi_send_command(ucsi, command, &ucsi->cap, UCSI_GET_CAPABILITY_SIZE); This fails for version < UCSI_VERSION_2_0. Size is wrong. > if (ret < 0) > goto err_reset; > [...]
