Message ID | 1456932729-9667-10-git-send-email-lee.jones@linaro.org |
---|---|
State | Superseded |
Headers | show |
On Wed, Mar 02, 2016 at 03:32:07PM +0000, Lee Jones wrote: > Once a PWM Capture has been initiated, the capture call > enables a rising edge detection IRQ, then waits. Once each > of the 3 phase changes have been recorded the thread then > wakes. The remaining part of the call carries out the > relevant calculations and passes back a formatted string to > the caller. > > Signed-off-by: Lee Jones <lee.jones@linaro.org> > --- > drivers/pwm/pwm-sti.c | 72 +++++++++++++++++++++++++++++++++++++++++++++++++++ > 1 file changed, 72 insertions(+) > > diff --git a/drivers/pwm/pwm-sti.c b/drivers/pwm/pwm-sti.c > index 82a69e4..8de9b4a 100644 > --- a/drivers/pwm/pwm-sti.c > +++ b/drivers/pwm/pwm-sti.c > @@ -309,7 +309,79 @@ static void sti_pwm_free(struct pwm_chip *chip, struct pwm_device *pwm) > clear_bit(pwm->hwpwm, &pc->configured); > } > > +static int sti_pwm_capture(struct pwm_chip *chip, struct pwm_device *pwm, > + int channel, char *buf) > +{ > + struct sti_pwm_chip *pc = to_sti_pwmchip(chip); > + struct sti_pwm_compat_data *cdata = pc->cdata; > + struct sti_cpt_data *d = pc->cpt_data[channel]; > + struct device *dev = pc->dev; > + unsigned int f, dc; > + unsigned int high, low; > + bool level; > + int ret; > + > + if (channel > cdata->cpt_num_chan - 1) { > + dev_err(dev, "Channel %d is not valid\n", channel); > + return -EINVAL; > + } > + > + mutex_lock(&d->lock); Should this perhaps reuse the struct pwm_device's ->lock? > + > + /* Prepare capture measurement */ > + d->index = 0; > + regmap_write(pc->regmap, PWM_CPT_EDGE(channel), CPT_EDGE_RISING); > + regmap_field_write(pc->pwm_cpt_int_en, BIT(channel)); > + ret = wait_event_interruptible_timeout(d->wait, d->index > 1, HZ); The timeout here should make sure callers don't hang forever. But maybe you can still make sure that when the PWM gets disabled the wait queue is woken and perhaps return an appropriate error code to let users know that the operation was interrupted. Also, how about letting callers choose the value of the timeout? In some cases they may be interested in long-running signals. In other cases the whole second timeout may be much too long. > + > + /* > + * In case we woke up for another reason than completion > + * make sure to disable the capture. > + */ > + regmap_write(pc->regmap, PWM_CPT_EDGE(channel), CPT_EDGE_DISABLED); The comment here is slightly confusing because it implies that disabling the capture should be done conditionally, whereas it is always disabled. > + > + if (ret == -ERESTARTSYS) > + goto out; > + > + switch (d->index) { > + case 0: > + case 1: > + /* > + * Getting here could mean : > + * - input signal is constant of less than 1Hz > + * - there is no input signal at all > + * > + * In such case the frequency is rounded down to 0 > + * level of the supposed constant signal is reported > + * using duty cycle min and max values. > + */ > + level = gpio_get_value(d->gpio); > + > + ret = sprintf(buf, "0:%u\n", level ? CPT_DC_MAX : 0); > + break; > + case 2: > + /* We have evertying we need */ > + high = d->snapshot[1] - d->snapshot[0]; > + low = d->snapshot[2] - d->snapshot[1]; > + > + /* Calculate frequency in Hz */ > + f = clk_get_rate(pc->cpt_clk) / (1 * (high + low)); The multiplication by 1 is unnecessary. Thierry
On Tue, 12 Apr 2016, Thierry Reding wrote: > On Wed, Mar 02, 2016 at 03:32:07PM +0000, Lee Jones wrote: > > Once a PWM Capture has been initiated, the capture call > > enables a rising edge detection IRQ, then waits. Once each > > of the 3 phase changes have been recorded the thread then > > wakes. The remaining part of the call carries out the > > relevant calculations and passes back a formatted string to > > the caller. > > > > Signed-off-by: Lee Jones <lee.jones@linaro.org> > > --- > > drivers/pwm/pwm-sti.c | 72 +++++++++++++++++++++++++++++++++++++++++++++++++++ > > 1 file changed, 72 insertions(+) > > > > diff --git a/drivers/pwm/pwm-sti.c b/drivers/pwm/pwm-sti.c > > index 82a69e4..8de9b4a 100644 > > --- a/drivers/pwm/pwm-sti.c > > +++ b/drivers/pwm/pwm-sti.c > > @@ -309,7 +309,79 @@ static void sti_pwm_free(struct pwm_chip *chip, struct pwm_device *pwm) > > clear_bit(pwm->hwpwm, &pc->configured); > > } > > > > +static int sti_pwm_capture(struct pwm_chip *chip, struct pwm_device *pwm, > > + int channel, char *buf) > > +{ > > + struct sti_pwm_chip *pc = to_sti_pwmchip(chip); > > + struct sti_pwm_compat_data *cdata = pc->cdata; > > + struct sti_cpt_data *d = pc->cpt_data[channel]; > > + struct device *dev = pc->dev; > > + unsigned int f, dc; > > + unsigned int high, low; > > + bool level; > > + int ret; > > + > > + if (channel > cdata->cpt_num_chan - 1) { > > + dev_err(dev, "Channel %d is not valid\n", channel); > > + return -EINVAL; > > + } > > + > > + mutex_lock(&d->lock); > > Should this perhaps reuse the struct pwm_device's ->lock? > > > + > > + /* Prepare capture measurement */ > > + d->index = 0; > > + regmap_write(pc->regmap, PWM_CPT_EDGE(channel), CPT_EDGE_RISING); > > + regmap_field_write(pc->pwm_cpt_int_en, BIT(channel)); > > + ret = wait_event_interruptible_timeout(d->wait, d->index > 1, HZ); > > The timeout here should make sure callers don't hang forever. But maybe > you can still make sure that when the PWM gets disabled the wait queue > is woken and perhaps return an appropriate error code to let users know > that the operation was interrupted. Sure. I'll look into that. > Also, how about letting callers choose the value of the timeout? In some > cases they may be interested in long-running signals. In other cases the > whole second timeout may be much too long. I'm not opposed to it. How do you suggest we do that? > > + /* > > + * In case we woke up for another reason than completion > > + * make sure to disable the capture. > > + */ > > + regmap_write(pc->regmap, PWM_CPT_EDGE(channel), CPT_EDGE_DISABLED); > > The comment here is slightly confusing because it implies that disabling > the capture should be done conditionally, whereas it is always disabled. Not really. We do it unconditionally for reason explained. It says: "disable the capture just in case X happens" rather than "disable the capture if X happens". Perhaps the language is too subtle. I can reword for clarity.
On Wed, Apr 13, 2016 at 11:25:54AM +0100, Lee Jones wrote: > On Tue, 12 Apr 2016, Thierry Reding wrote: > > > On Wed, Mar 02, 2016 at 03:32:07PM +0000, Lee Jones wrote: > > > Once a PWM Capture has been initiated, the capture call > > > enables a rising edge detection IRQ, then waits. Once each > > > of the 3 phase changes have been recorded the thread then > > > wakes. The remaining part of the call carries out the > > > relevant calculations and passes back a formatted string to > > > the caller. > > > > > > Signed-off-by: Lee Jones <lee.jones@linaro.org> > > > --- > > > drivers/pwm/pwm-sti.c | 72 +++++++++++++++++++++++++++++++++++++++++++++++++++ > > > 1 file changed, 72 insertions(+) > > > > > > diff --git a/drivers/pwm/pwm-sti.c b/drivers/pwm/pwm-sti.c > > > index 82a69e4..8de9b4a 100644 > > > --- a/drivers/pwm/pwm-sti.c > > > +++ b/drivers/pwm/pwm-sti.c > > > @@ -309,7 +309,79 @@ static void sti_pwm_free(struct pwm_chip *chip, struct pwm_device *pwm) > > > clear_bit(pwm->hwpwm, &pc->configured); > > > } > > > > > > +static int sti_pwm_capture(struct pwm_chip *chip, struct pwm_device *pwm, > > > + int channel, char *buf) > > > +{ > > > + struct sti_pwm_chip *pc = to_sti_pwmchip(chip); > > > + struct sti_pwm_compat_data *cdata = pc->cdata; > > > + struct sti_cpt_data *d = pc->cpt_data[channel]; > > > + struct device *dev = pc->dev; > > > + unsigned int f, dc; > > > + unsigned int high, low; > > > + bool level; > > > + int ret; > > > + > > > + if (channel > cdata->cpt_num_chan - 1) { > > > + dev_err(dev, "Channel %d is not valid\n", channel); > > > + return -EINVAL; > > > + } > > > + > > > + mutex_lock(&d->lock); > > > > Should this perhaps reuse the struct pwm_device's ->lock? That was actually a stupid suggestion by me, because that lock is being removed in an unrelated patch series. > > > + /* Prepare capture measurement */ > > > + d->index = 0; > > > + regmap_write(pc->regmap, PWM_CPT_EDGE(channel), CPT_EDGE_RISING); > > > + regmap_field_write(pc->pwm_cpt_int_en, BIT(channel)); > > > + ret = wait_event_interruptible_timeout(d->wait, d->index > 1, HZ); > > > > The timeout here should make sure callers don't hang forever. But maybe > > you can still make sure that when the PWM gets disabled the wait queue > > is woken and perhaps return an appropriate error code to let users know > > that the operation was interrupted. > > Sure. I'll look into that. > > > Also, how about letting callers choose the value of the timeout? In some > > cases they may be interested in long-running signals. In other cases the > > whole second timeout may be much too long. > > I'm not opposed to it. How do you suggest we do that? The easiest would probably be to add an unsigned long timeout parameter to the pwm_capture() function and ->capture() callbacks. But thinking about this further I'm wondering if it might not be easier and more flexible to move the timeout completely outside of this code and into callers. I suspect that the most simple way to do that would be to add a completion to struct pwm_capture that callers can use to wait for completion of a capture. This would make the whole process asynchronous and allow interesting things like making the sysfs capture file pollable, for example. > > > + /* > > > + * In case we woke up for another reason than completion > > > + * make sure to disable the capture. > > > + */ > > > + regmap_write(pc->regmap, PWM_CPT_EDGE(channel), CPT_EDGE_DISABLED); > > > > The comment here is slightly confusing because it implies that disabling > > the capture should be done conditionally, whereas it is always disabled. > > Not really. We do it unconditionally for reason explained. > > It says: > > "disable the capture just in case X happens" > > rather than > > "disable the capture if X happens". > > Perhaps the language is too subtle. I can reword for clarity. I'd be okay with just dropping the comment altogether, it seems rather obvious to me. But clarifying is okay with me, too. Thierry
On Wed, 13 Apr 2016, Thierry Reding wrote: > On Wed, Apr 13, 2016 at 11:25:54AM +0100, Lee Jones wrote: > > On Tue, 12 Apr 2016, Thierry Reding wrote: > > > > > On Wed, Mar 02, 2016 at 03:32:07PM +0000, Lee Jones wrote: > > > > Once a PWM Capture has been initiated, the capture call > > > > enables a rising edge detection IRQ, then waits. Once each > > > > of the 3 phase changes have been recorded the thread then > > > > wakes. The remaining part of the call carries out the > > > > relevant calculations and passes back a formatted string to > > > > the caller. > > > > > > > > Signed-off-by: Lee Jones <lee.jones@linaro.org> > > > > --- > > > > drivers/pwm/pwm-sti.c | 72 +++++++++++++++++++++++++++++++++++++++++++++++++++ > > > > 1 file changed, 72 insertions(+) > > > > > > > > diff --git a/drivers/pwm/pwm-sti.c b/drivers/pwm/pwm-sti.c > > > > index 82a69e4..8de9b4a 100644 > > > > --- a/drivers/pwm/pwm-sti.c > > > > +++ b/drivers/pwm/pwm-sti.c [...] > > > > + /* Prepare capture measurement */ > > > > + d->index = 0; > > > > + regmap_write(pc->regmap, PWM_CPT_EDGE(channel), CPT_EDGE_RISING); > > > > + regmap_field_write(pc->pwm_cpt_int_en, BIT(channel)); > > > > + ret = wait_event_interruptible_timeout(d->wait, d->index > 1, HZ); > > > > > > The timeout here should make sure callers don't hang forever. But maybe > > > you can still make sure that when the PWM gets disabled the wait queue > > > is woken and perhaps return an appropriate error code to let users know > > > that the operation was interrupted. > > > > Sure. I'll look into that. > > > > > Also, how about letting callers choose the value of the timeout? In some > > > cases they may be interested in long-running signals. In other cases the > > > whole second timeout may be much too long. > > > > I'm not opposed to it. How do you suggest we do that? > > The easiest would probably be to add an unsigned long timeout parameter > to the pwm_capture() function and ->capture() callbacks. > > But thinking about this further I'm wondering if it might not be easier > and more flexible to move the timeout completely outside of this code > and into callers. I suspect that the most simple way to do that would be > to add a completion to struct pwm_capture that callers can use to wait > for completion of a capture. This would make the whole process > asynchronous and allow interesting things like making the sysfs capture > file pollable, for example. Okay, so how do you propose we handle this with sysfs? Perhaps another RW file to set it?
On Fri, Apr 15, 2016 at 09:29:00AM +0100, Lee Jones wrote: > On Wed, 13 Apr 2016, Thierry Reding wrote: > > > On Wed, Apr 13, 2016 at 11:25:54AM +0100, Lee Jones wrote: > > > On Tue, 12 Apr 2016, Thierry Reding wrote: > > > > > > > On Wed, Mar 02, 2016 at 03:32:07PM +0000, Lee Jones wrote: > > > > > Once a PWM Capture has been initiated, the capture call > > > > > enables a rising edge detection IRQ, then waits. Once each > > > > > of the 3 phase changes have been recorded the thread then > > > > > wakes. The remaining part of the call carries out the > > > > > relevant calculations and passes back a formatted string to > > > > > the caller. > > > > > > > > > > Signed-off-by: Lee Jones <lee.jones@linaro.org> > > > > > --- > > > > > drivers/pwm/pwm-sti.c | 72 +++++++++++++++++++++++++++++++++++++++++++++++++++ > > > > > 1 file changed, 72 insertions(+) > > > > > > > > > > diff --git a/drivers/pwm/pwm-sti.c b/drivers/pwm/pwm-sti.c > > > > > index 82a69e4..8de9b4a 100644 > > > > > --- a/drivers/pwm/pwm-sti.c > > > > > +++ b/drivers/pwm/pwm-sti.c > > [...] > > > > > > + /* Prepare capture measurement */ > > > > > + d->index = 0; > > > > > + regmap_write(pc->regmap, PWM_CPT_EDGE(channel), CPT_EDGE_RISING); > > > > > + regmap_field_write(pc->pwm_cpt_int_en, BIT(channel)); > > > > > + ret = wait_event_interruptible_timeout(d->wait, d->index > 1, HZ); > > > > > > > > The timeout here should make sure callers don't hang forever. But maybe > > > > you can still make sure that when the PWM gets disabled the wait queue > > > > is woken and perhaps return an appropriate error code to let users know > > > > that the operation was interrupted. > > > > > > Sure. I'll look into that. > > > > > > > Also, how about letting callers choose the value of the timeout? In some > > > > cases they may be interested in long-running signals. In other cases the > > > > whole second timeout may be much too long. > > > > > > I'm not opposed to it. How do you suggest we do that? > > > > The easiest would probably be to add an unsigned long timeout parameter > > to the pwm_capture() function and ->capture() callbacks. > > > > But thinking about this further I'm wondering if it might not be easier > > and more flexible to move the timeout completely outside of this code > > and into callers. I suspect that the most simple way to do that would be > > to add a completion to struct pwm_capture that callers can use to wait > > for completion of a capture. This would make the whole process > > asynchronous and allow interesting things like making the sysfs capture > > file pollable, for example. > > Okay, so how do you propose we handle this with sysfs? Perhaps > another RW file to set it? I'm unfamiliar with how this is done in other drivers, so I'd have to look at them first. I suspect that it would be fine for now to simply redesign the PWM API parts and keep some default timeout in sysfs. It could be extended with some mechanism to override the default timeout in the future. Thierry
diff --git a/drivers/pwm/pwm-sti.c b/drivers/pwm/pwm-sti.c index 82a69e4..8de9b4a 100644 --- a/drivers/pwm/pwm-sti.c +++ b/drivers/pwm/pwm-sti.c @@ -309,7 +309,79 @@ static void sti_pwm_free(struct pwm_chip *chip, struct pwm_device *pwm) clear_bit(pwm->hwpwm, &pc->configured); } +static int sti_pwm_capture(struct pwm_chip *chip, struct pwm_device *pwm, + int channel, char *buf) +{ + struct sti_pwm_chip *pc = to_sti_pwmchip(chip); + struct sti_pwm_compat_data *cdata = pc->cdata; + struct sti_cpt_data *d = pc->cpt_data[channel]; + struct device *dev = pc->dev; + unsigned int f, dc; + unsigned int high, low; + bool level; + int ret; + + if (channel > cdata->cpt_num_chan - 1) { + dev_err(dev, "Channel %d is not valid\n", channel); + return -EINVAL; + } + + mutex_lock(&d->lock); + + /* Prepare capture measurement */ + d->index = 0; + regmap_write(pc->regmap, PWM_CPT_EDGE(channel), CPT_EDGE_RISING); + regmap_field_write(pc->pwm_cpt_int_en, BIT(channel)); + ret = wait_event_interruptible_timeout(d->wait, d->index > 1, HZ); + + /* + * In case we woke up for another reason than completion + * make sure to disable the capture. + */ + regmap_write(pc->regmap, PWM_CPT_EDGE(channel), CPT_EDGE_DISABLED); + + if (ret == -ERESTARTSYS) + goto out; + + switch (d->index) { + case 0: + case 1: + /* + * Getting here could mean : + * - input signal is constant of less than 1Hz + * - there is no input signal at all + * + * In such case the frequency is rounded down to 0 + * level of the supposed constant signal is reported + * using duty cycle min and max values. + */ + level = gpio_get_value(d->gpio); + + ret = sprintf(buf, "0:%u\n", level ? CPT_DC_MAX : 0); + break; + case 2: + /* We have evertying we need */ + high = d->snapshot[1] - d->snapshot[0]; + low = d->snapshot[2] - d->snapshot[1]; + + /* Calculate frequency in Hz */ + f = clk_get_rate(pc->cpt_clk) / (1 * (high + low)); + + /* Calculate the duty cycle */ + dc = CPT_DC_MAX * high / (high + low); + + ret = sprintf(buf, "%u:%u\n", f, dc); + default: + dev_err(dev, "Internal error\n"); + } + +out: + mutex_unlock(&d->lock); + return ret; +} + static const struct pwm_ops sti_pwm_ops = { + .capture = sti_pwm_capture, .config = sti_pwm_config, .enable = sti_pwm_enable, .disable = sti_pwm_disable,
Once a PWM Capture has been initiated, the capture call enables a rising edge detection IRQ, then waits. Once each of the 3 phase changes have been recorded the thread then wakes. The remaining part of the call carries out the relevant calculations and passes back a formatted string to the caller. Signed-off-by: Lee Jones <lee.jones@linaro.org> --- drivers/pwm/pwm-sti.c | 72 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 72 insertions(+)