| Message ID | 1422028354-31891-4-git-send-email-lars@metafoo.de (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
Em Fri, 23 Jan 2015 16:52:22 +0100 Lars-Peter Clausen <lars@metafoo.de> escreveu: > Use a inline function instead of a macro for the container_of helper for > getting the driver's state struct from a control. A inline function has the > advantage that it is more typesafe and nicer in general. I don't see any advantage on this. See: container_of is already a macro, and it is written in a way that, if you use it with inconsistent values, the compilation will break. Also, there's the risk that, for whatever reason, gcc to decide to not inline this. So, this doesn't sound a good idea. Regards, Mauro > > Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> > Acked-by: Hans Verkuil <hans.verkuil@cisco.com> > --- > drivers/media/i2c/adv7180.c | 13 +++++++------ > 1 file changed, 7 insertions(+), 6 deletions(-) > > diff --git a/drivers/media/i2c/adv7180.c b/drivers/media/i2c/adv7180.c > index f424a4d..f2508abe 100644 > --- a/drivers/media/i2c/adv7180.c > +++ b/drivers/media/i2c/adv7180.c > @@ -130,9 +130,11 @@ struct adv7180_state { > bool powered; > u8 input; > }; > -#define to_adv7180_sd(_ctrl) (&container_of(_ctrl->handler, \ > - struct adv7180_state, \ > - ctrl_hdl)->sd) > + > +static struct adv7180_state *ctrl_to_adv7180(struct v4l2_ctrl *ctrl) > +{ > + return container_of(ctrl->handler, struct adv7180_state, ctrl_hdl); > +} > > static v4l2_std_id adv7180_std_to_v4l2(u8 status1) > { > @@ -345,9 +347,8 @@ static int adv7180_s_power(struct v4l2_subdev *sd, int on) > > static int adv7180_s_ctrl(struct v4l2_ctrl *ctrl) > { > - struct v4l2_subdev *sd = to_adv7180_sd(ctrl); > - struct adv7180_state *state = to_state(sd); > - struct i2c_client *client = v4l2_get_subdevdata(sd); > + struct adv7180_state *state = ctrl_to_adv7180(ctrl); > + struct i2c_client *client = v4l2_get_subdevdata(&state->sd); > int ret = mutex_lock_interruptible(&state->mutex); > int val; > -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 02/02/2015 02:36 PM, Mauro Carvalho Chehab wrote: > Em Fri, 23 Jan 2015 16:52:22 +0100 > Lars-Peter Clausen <lars@metafoo.de> escreveu: > >> Use a inline function instead of a macro for the container_of helper for >> getting the driver's state struct from a control. A inline function has the >> advantage that it is more typesafe and nicer in general. > > I don't see any advantage on this. > > See: container_of is already a macro, and it is written in a way that, if > you use it with inconsistent values, the compilation will break. > > Also, there's the risk that, for whatever reason, gcc to decide to not > inline this. For the record: I disagree with this. I think a static inline is more readable than a macro. It is also consistent with other existing i2c subdev drivers since they all use a static inline as well. And if in rare cases gcc might decide not to inline this, then so what? You really won't notice any difference. It's an i2c device, so it's slow as molasses anyway. Readability is much more important IMHO. On the other hand, it's not critical enough for me to make a big deal out of it if this patch is dropped. Regards, Hans > > So, this doesn't sound a good idea. > > Regards, > Mauro > >> >> Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> >> Acked-by: Hans Verkuil <hans.verkuil@cisco.com> >> --- >> drivers/media/i2c/adv7180.c | 13 +++++++------ >> 1 file changed, 7 insertions(+), 6 deletions(-) >> >> diff --git a/drivers/media/i2c/adv7180.c b/drivers/media/i2c/adv7180.c >> index f424a4d..f2508abe 100644 >> --- a/drivers/media/i2c/adv7180.c >> +++ b/drivers/media/i2c/adv7180.c >> @@ -130,9 +130,11 @@ struct adv7180_state { >> bool powered; >> u8 input; >> }; >> -#define to_adv7180_sd(_ctrl) (&container_of(_ctrl->handler, \ >> - struct adv7180_state, \ >> - ctrl_hdl)->sd) >> + >> +static struct adv7180_state *ctrl_to_adv7180(struct v4l2_ctrl *ctrl) >> +{ >> + return container_of(ctrl->handler, struct adv7180_state, ctrl_hdl); >> +} >> >> static v4l2_std_id adv7180_std_to_v4l2(u8 status1) >> { >> @@ -345,9 +347,8 @@ static int adv7180_s_power(struct v4l2_subdev *sd, int on) >> >> static int adv7180_s_ctrl(struct v4l2_ctrl *ctrl) >> { >> - struct v4l2_subdev *sd = to_adv7180_sd(ctrl); >> - struct adv7180_state *state = to_state(sd); >> - struct i2c_client *client = v4l2_get_subdevdata(sd); >> + struct adv7180_state *state = ctrl_to_adv7180(ctrl); >> + struct i2c_client *client = v4l2_get_subdevdata(&state->sd); >> int ret = mutex_lock_interruptible(&state->mutex); >> int val; >> -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 02/02/2015 02:36 PM, Mauro Carvalho Chehab wrote: > Em Fri, 23 Jan 2015 16:52:22 +0100 > Lars-Peter Clausen <lars@metafoo.de> escreveu: > >> Use a inline function instead of a macro for the container_of helper for >> getting the driver's state struct from a control. A inline function has the >> advantage that it is more typesafe and nicer in general. > > I don't see any advantage on this. > > See: container_of is already a macro, and it is written in a way that, if > you use it with inconsistent values, the compilation will break. Yes, container_of is a macro, because it needs to be a macro. Compilation will also not always break if you pass in a incorrect type, it might succeed with even generating a warning. Furthermore if compilation breaks the error message is completely incomprehensible. Using a function instead makes sure that the error message you get is in the style of "passing argument of wrong type to function, expected typeX, got typeY". > > Also, there's the risk that, for whatever reason, gcc to decide to not > inline this. If the compiler does not inline this it probably has a good reason to do so. Not inlining this will not break the functionality, so it is not a problem. - Lars -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/drivers/media/i2c/adv7180.c b/drivers/media/i2c/adv7180.c index f424a4d..f2508abe 100644 --- a/drivers/media/i2c/adv7180.c +++ b/drivers/media/i2c/adv7180.c @@ -130,9 +130,11 @@ struct adv7180_state { bool powered; u8 input; }; -#define to_adv7180_sd(_ctrl) (&container_of(_ctrl->handler, \ - struct adv7180_state, \ - ctrl_hdl)->sd) + +static struct adv7180_state *ctrl_to_adv7180(struct v4l2_ctrl *ctrl) +{ + return container_of(ctrl->handler, struct adv7180_state, ctrl_hdl); +} static v4l2_std_id adv7180_std_to_v4l2(u8 status1) { @@ -345,9 +347,8 @@ static int adv7180_s_power(struct v4l2_subdev *sd, int on) static int adv7180_s_ctrl(struct v4l2_ctrl *ctrl) { - struct v4l2_subdev *sd = to_adv7180_sd(ctrl); - struct adv7180_state *state = to_state(sd); - struct i2c_client *client = v4l2_get_subdevdata(sd); + struct adv7180_state *state = ctrl_to_adv7180(ctrl); + struct i2c_client *client = v4l2_get_subdevdata(&state->sd); int ret = mutex_lock_interruptible(&state->mutex); int val;