On Thu, Oct 6, 2016 at 12:24 PM, Maxime Ripard <maxime.ripard@xxxxxxxxxxxxxxxxxx> wrote: > Hi David, > > On Thu, Oct 06, 2016 at 08:47:56PM +1100, David Gibson wrote: >> On Fri, Sep 30, 2016 at 03:57:13PM +0200, Maxime Ripard wrote: >> > Hi, >> > >> > The device tree overlays are a great solution to the issue raised by >> > the bunch expandable boards we find everywhere these days, like the >> > Beaglebone, Raspberry Pi or CHIP. >> > >> > Although for now Linux is the only available tool that can deal with >> > overlays, some other components like bootloaders or user-space tools >> > might need to apply these overlays on top of a base device tree. >> > >> > To address these use cases, we introduce a new function to the libfdt, >> > fdt_overlay_apply, that does just that. >> > >> > You can find a test program here: http://code.bulix.org/792zum-99476?raw >> > >> > This is the last patches sent to U-boot, with the modifications >> > suggested by David, and the additional test cases. >> > >> > Let me know what you think! >> > Maxime >> >> Looking pretty good. There's a minor error in 3/4 which I've pointed >> out, it can be fixed in a followup. >> >> An error in the conditionals means it's trying to run some of the >> tests which require an overlay-aware dtc on even when that's not >> available - I adjusted 4/4 to correct that. > > While chasing that, I found that it was actually a bug in a condition > that was stricter than it should, I'll send follow-up with that and > the fix you asked. > >> It also hit build failures on Travis, because the compiler version >> there complains about local variables named 'index' shadowing the >> standard library function. I've corrected that with a followup patch. >> >> With that done, the whole lot has been committed to master. > > Thanks a lot, > Maxime Now that upstream libfdt can resolve and apply overlays we need a way of generating them. I have a set of three patches from Pantelis that add "plugin" support to dtc, together with two downstream Raspberry Pi patches that fix an uninitialised memory reference and restrict __local_fixups__ generation to overlays. Is there already a plan to get these upstreamed? If so, can we get the downstream fixes included? Phil -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
