Messages in this thread | ![/](/images/icornerl.gif) | | Date | Thu, 10 Dec 2015 19:41:19 +0100 | From | Alexandre Belloni <> | Subject | Re: [PATCH v2] RTC: RK808: Work around hardware bug on November 31st |
| |
Hi Julius, Doug,
On 09/12/2015 at 13:32:52 -0800, Julius Werner wrote : > > Thinking about all this: these's actually a totally different > > alternative approach we could take if you wanted. It would fix S5 and > > avoid all the anchor stuff, unless I'm crazy. > > > > Basically totally give up on the RTC time reflecting reality. Add a > > "real time to rk808" and "rk808 time to real time" function. Always > > use it when reading from rk808 and always use it when writing to > > rk808. Choose 2015 as the "truth" year if you want (or pick another > > year). In early 2016 the rk808 always contains 1 day back. In 2017 > > the rk808 always contains 2 days back. Etc, etc, etc. > > > > The firmware would get confused, but ... > > Well... other than that it's crazy and that I'd have to rewrite the > whole patch again, I can't come up with a good argument against this. > In Chromebook firmware the time is only needed for a debug log, so > we'd probably just be willing to accept it being wrong. If U-Boot ever > gets RK808 support, they'll probably just copy the Linux driver > wholesale anyway so they'll conform to the same system. > > So if nobody else raises fundamental objections to this approach, I > guess I'll get started on another patch version. (Further replies > below for reference, but most of that stuff would then become moot.) >
I'll try to review and evaluate both solution by the end of the week (no guarantee though).
-- Alexandre Belloni, Free Electrons Embedded Linux, Kernel and Android engineering http://free-electrons.com
| ![\](/images/icornerr.gif) |