On Sat, Feb 15, 2025 at 02:14:54PM -0500, Alex Lanzano wrote: > On Fri, Feb 14, 2025 at 08:04:41PM -0500, Alex Lanzano wrote: > > On Fri, Feb 14, 2025 at 10:29:29AM +0100, Josef Luštický wrote: > > > Hello Alex, > > > there is a bug in mipi_dbi_hw_reset() function that implements the logic of > > > display reset contrary. > > > It keeps the reset line activated which keeps displays in reset state. > > > > > > I reported the bug to > > > https://gitlab.freedesktop.org/drm/misc/kernel/-/issues/63 > > > > > > Unfortunately, fixing the bug would mean current DTB-ABI breakage and > > > device-trees modification would be needed. > > > I mainly write this email to let you and other people know about it, so > > > hopefully it can be found easier. > > > What are your thoughts? > > Thanks for making me aware. I'll dig into over the weekend and get back > > to you > > Alright so I looked into a bit more. Looks like the MIPI Specification > says that the reset line is active low. So, if we want dt entries to be > correct the logic for setting the reset line in mipi_dbi_hw_reset() > should be flipped. However, like you said, this is going to cause a some > confused developers to break out their oscilloscopes to figure out > why their display isn't working. Would it be easier to flip GPIO flag in DT (and explicitly document that the flag is to be inverted)? -- With best wishes Dmitry
