On 06/14/2017 03:32 PM, Adam Ford wrote: > On Wed, Jun 14, 2017 at 7:47 AM, Adam Ford <aford173@xxxxxxxxx> wrote: >> On Wed, Jun 14, 2017 at 6:16 AM, Tony Lindgren <tony@xxxxxxxxxxx> wrote: >>> * Adam Ford <aford173@xxxxxxxxx> [170614 03:53]: >>>> I have a DM3730 with the dmtimer (timer10) controlling the pwm >>>> backlight for my LCD. >>>> >>>> pwms = <&pwm10 0 5000000 0>; >>>> brightness-levels = <0 10 20 30 40 50 60 70 80 90 100>; >>>> default-brightness-level = <7>; >>>> >>>> If I cat /dev/urandom > /dev/fb0 to setup a test pattern then start to >>>> change the brightness in user-space, there is a momentary flicker on >>>> the backlight when the backlight changes intensity. Screens that are >>>> predominantly black are not as noticeable. >>>> >>>> We also have a Win CE implementation on the same board which does not >>>> do this, so I believe it to be software related and not hardware. Has >>>> anyone else noticed this? If so, might you have any suggestions to >>>> reducing this? I checked 4.9.y and 4.11.y and both seem to exhibit >>>> the same behavior. I don't have other pwm options, so I cannot >>>> determine if this is a problem with dmtimer or the pwm-bl driver. >>> >>> Maybe some issue changing the timer rate? Could it be that we're >>> missing stopping of the dmtimer in the pwm driver during the rate >>> change? >>> >> >> I think we do stop the timer during the chage. >> >> Looking at the code in pwm_omap_dmtimer_config: >> >> /* >> * We MUST stop the associated dual-mode timer before attempting to >> * write its registers, but calls to omap_dm_timer_start/stop must >> * be balanced so check if timer is active before calling timer_stop. >> */ >> timer_active = pm_runtime_active(&omap->dm_timer_pdev->dev); >> >> if (timer_active) >> omap->pdata->stop(omap->dm_timer); >> >> [ Change a bunch of stuff] >> >> /* If config was called while timer was running it must be reenabled. */ >> if (timer_active) >> pwm_omap_dmtimer_start(omap); >> >> I am wondering if we need to somehow sync the stop with the pwm when >> the duty cycle is in the off position so it doesn't stay on while we >> change the registers. I am not really sure how to go about that. Some >> LED's may be active high while others are active low. >> > > Reading the DM3730 Technical Reference manual, it reads: The timer > counter register (GPTi.TCRR) can be loaded when stopped or on-the-fly > (while counting). > > I removed the protection code (what shown above) around the changing > of the registers, and the flicker went away. I am hoping Neil might > have some comment on this. If removing this is deemed acceptable, I > can push a patch, but I don't want to arbitrarily assume it would > work. > Hi, Maybe we should not stop it only when the counter register is to be changed. This mean we should store the current load/match/polarity in the local context. Neil > >> adam >> >>> Regards, >>> >>> Tony -- To unsubscribe from this list: send the line "unsubscribe linux-omap" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
