Hello Alexandre! Thank you for your review! On Thu, 2024-02-29 at 23:08 +0100, Alexandre Belloni wrote: > > "There is a low probability that some devices will have corruption of the > > registers after the automatic power-on reset if the device is powered up > > with a residual VDD level. It is required that the VDD starts at zero volts ... > Doing this in probe is putting a band-aid on a wooden leg as you are not > guaranteed you will have a probe to catch this case. This should be > rather done in pcf85063_rtc_set_time but it comes with its own set of As I read the datasheet (quoted above) the device has "peculiarities" in Power-On-Reset implementation, namely it's not always detected. In our applications (and probably most other designs) it's the startup of Linux with driver probe, which immediately follows the Power-On event (which had some residue voltage because of large capacitors, whatsoever). It is an issue we really observed and solved in 100% cases for our design. I just thought it might be useful for others because it's also documented by NXP. Maybe not as separate errata document and it's a bit hidden in datasheet but I'd ultimately consider it a POR errata. -- Alexander Sverdlin Siemens AG www.siemens.com
