| Message ID | 20190603090058.qd3tbiffmdgqm34d@gofer.mess.org |
|---|---|
| State | Accepted |
| Headers | show |
| Series | [v2] pwm: bcm2835: improve precision of pwm | expand |
Am 03.06.19 um 11:00 schrieb Sean Young: > If sending IR with carrier of 455kHz using the pwm-ir-tx driver, the > carrier ends up being 476kHz. The clock is set to bcm2835-pwm with a > rate of 10MHz. > > A carrier of 455kHz has a period of 2198ns, but the arithmetic truncates > this to 2100ns rather than 2200ns. So, use DIV_ROUND_CLOSEST() to reduce > rounding errors, and we have a much more accurate carrier of 454.5kHz. > > Reported-by: Andreas Christ <andreas@christ-faesch.ch> > Signed-off-by: Sean Young <sean@mess.org> Acked-by: Stefan Wahren <wahrenst@gmx.net> Thanks
Am 03.06.19 um 11:00 schrieb Sean Young: > If sending IR with carrier of 455kHz using the pwm-ir-tx driver, the > carrier ends up being 476kHz. The clock is set to bcm2835-pwm with a > rate of 10MHz. > > A carrier of 455kHz has a period of 2198ns, but the arithmetic truncates > this to 2100ns rather than 2200ns. So, use DIV_ROUND_CLOSEST() to reduce > rounding errors, and we have a much more accurate carrier of 454.5kHz. > > Reported-by: Andreas Christ <andreas@christ-faesch.ch> > Signed-off-by: Sean Young <sean@mess.org> Acked-by: Stefan Wahren <wahrenst@gmx.de>
Hallo Stefan, On Wed, Jun 19, 2019 at 07:21:43AM +0200, Stefan Wahren wrote: > Am 03.06.19 um 11:00 schrieb Sean Young: > > If sending IR with carrier of 455kHz using the pwm-ir-tx driver, the > > carrier ends up being 476kHz. The clock is set to bcm2835-pwm with a > > rate of 10MHz. > > > > A carrier of 455kHz has a period of 2198ns, but the arithmetic truncates > > this to 2100ns rather than 2200ns. So, use DIV_ROUND_CLOSEST() to reduce > > rounding errors, and we have a much more accurate carrier of 454.5kHz. > > > > Reported-by: Andreas Christ <andreas@christ-faesch.ch> > > Signed-off-by: Sean Young <sean@mess.org> > Acked-by: Stefan Wahren <wahrenst@gmx.de> you already acked that with a slightly different mail address a week ago. Was this intended? Best regards Uwe
Hi Uwe, On 19.06.19 09:22, Uwe Kleine-König wrote: > Hallo Stefan, > > On Wed, Jun 19, 2019 at 07:21:43AM +0200, Stefan Wahren wrote: >> Am 03.06.19 um 11:00 schrieb Sean Young: >>> If sending IR with carrier of 455kHz using the pwm-ir-tx driver, the >>> carrier ends up being 476kHz. The clock is set to bcm2835-pwm with a >>> rate of 10MHz. >>> >>> A carrier of 455kHz has a period of 2198ns, but the arithmetic truncates >>> this to 2100ns rather than 2200ns. So, use DIV_ROUND_CLOSEST() to reduce >>> rounding errors, and we have a much more accurate carrier of 454.5kHz. >>> >>> Reported-by: Andreas Christ <andreas@christ-faesch.ch> >>> Signed-off-by: Sean Young <sean@mess.org> >> Acked-by: Stefan Wahren <wahrenst@gmx.de> > you already acked that with a slightly different mail address a week > ago. Was this intended? thanks. Please ignore this email this was a typo. The correct domain was gmx.net. Stefan > > Best regards > Uwe >
diff --git a/drivers/pwm/pwm-bcm2835.c b/drivers/pwm/pwm-bcm2835.c index 5652f461d994..f6fe0b922e1e 100644 --- a/drivers/pwm/pwm-bcm2835.c +++ b/drivers/pwm/pwm-bcm2835.c @@ -70,7 +70,7 @@ static int bcm2835_pwm_config(struct pwm_chip *chip, struct pwm_device *pwm, return -EINVAL; } - scaler = NSEC_PER_SEC / rate; + scaler = DIV_ROUND_CLOSEST(NSEC_PER_SEC, rate); if (period_ns <= MIN_PERIOD) { dev_err(pc->dev, "period %d not supported, minimum %d\n", @@ -78,8 +78,10 @@ static int bcm2835_pwm_config(struct pwm_chip *chip, struct pwm_device *pwm, return -EINVAL; } - writel(duty_ns / scaler, pc->base + DUTY(pwm->hwpwm)); - writel(period_ns / scaler, pc->base + PERIOD(pwm->hwpwm)); + writel(DIV_ROUND_CLOSEST(duty_ns, scaler), + pc->base + DUTY(pwm->hwpwm)); + writel(DIV_ROUND_CLOSEST(period_ns, scaler), + pc->base + PERIOD(pwm->hwpwm)); return 0; }
If sending IR with carrier of 455kHz using the pwm-ir-tx driver, the carrier ends up being 476kHz. The clock is set to bcm2835-pwm with a rate of 10MHz. A carrier of 455kHz has a period of 2198ns, but the arithmetic truncates this to 2100ns rather than 2200ns. So, use DIV_ROUND_CLOSEST() to reduce rounding errors, and we have a much more accurate carrier of 454.5kHz. Reported-by: Andreas Christ <andreas@christ-faesch.ch> Signed-off-by: Sean Young <sean@mess.org> --- drivers/pwm/pwm-bcm2835.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-)