Hi Tom, Sorry for top posting, am on the road. First: The Allwinner drivers in U-Boot work with mainline kernel DTs for ages, so there are no code changes required. We manually sync the DTs from the kernel for years, but I always do one change:
In kernel v5.13 the Allwinner IRQ controller (not the GIC) DT binding and DTs changed in an incompatible way, and some critical devices now route their IRQ through that controller instead of to the GIC. If an older kernel sees such a new DT, some critical devices fail to probe due to that, and the kernel does not boot. Using the old DT routing fixes that, at the expense of losing some advanced power saving features. So in the interest of better compatibility with older kernels and $fdtcontroladdr (default for UEFI boot), I always kept the old way during the U-Boot sync. This affects all devices with the popular H3, A64 and H6 SoCs, among others. A20 and older or H616s are not affected. So my plan was to migrate the latter group ASAP (after the v6.10 sync), and then maybe the others in the next cycle, which means pulling the plug on original Debian Buster kernels, for instance. 5.13 is three years old by now, so I guess this might be fine now. Does that make sense? Cheers, Andre On 17 July 2024 18:51:53 BST, Tom Rini <tr...@konsulko.com> wrote: >On Wed, Jul 17, 2024 at 03:34:15PM +0100, Andre Przywara wrote: > >> Sync the devicetree files from the official Linux kernel tree, v6.10. >> This is covering Allwinner SoCs with 32-bit and 64-bit ARM cores. >> >> Besides mostly cosmectic changes, this adds cpufreq support to H616 >> boards, Nothing that U-Boot needs for itself, but helpful to pass on >> to kernels. We also get the .dts files for the Tanix TX1 TV box and >> three Anbernic handheld gaming devices. >> >> As before, this omits the non-backwards compatible changes to the R_INTC >> controller, to remain compatible with older kernels. >> >> Signed-off-by: Andre Przywara <andre.przyw...@arm.com> > >This reminded me I needed to see that v6.10-dts is tagged upstream now. >So, for the record, when can we move sunxi to OF_UPSTREAM? Or what's >blocking it? Thanks! > >-- >Tom
