On Mon 2019-06-17 09:16:25, Matthias Kaehlcke wrote: > Hi Pavel, > > On Sun, Jun 16, 2019 at 05:41:43PM +0200, Pavel Machek wrote: > > Hi! > > > > > This reverts commit 288ceb85b505c19abe1895df068dda5ed20cf482. > > > > > > According to the commit message the AUO B101EAN01 panel on minnie > > > requires a PWM delay of 200 ms, however this is not what the > > > datasheet says. The datasheet mentions a *max* delay of 200 ms > > > for T2 ("delay from LCDVDD to black video generation") and T3 > > > ("delay from LCDVDD to HPD high"), which aren't related to the > > > PWM. The backlight power sequence does not specify min/max > > > constraints for T15 (time from PWM on to BL enable) or T16 > > > (time from BL disable to PWM off). > > > > > > Signed-off-by: Matthias Kaehlcke > > > --- > > > Enric, if you think I misinterpreted the datasheet please holler! > > > > Was this tested? > > I performed limited manually testing. > > minnie ships with the Chrome OS 3.14 downstream, which doesn't include > this delay, to my knowledge there are no open display related bugs for > minnie. One could argue that a the configuration without the delay was > widely field tested > > > Does patch being reverted actually break anything? > > To my knowledge it doesn't really break anything, however there is a > short user perceptible delay between switching on the LCD and > switching on the backlight. It's not the end of the world, but if it's > not actually needed better avoid it. > > > If so, cc stable? > > I guess this is an edge case, were you could go either way. I'm fine > with respinning and cc-ing stable. Ok, if it is just a small delay, stable probably does not need to be involved. Thanks, Pavel -- (english) http://www.livejournal.com/~pavelmachek (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html