On Sat, Aug 22, 2026 at 08:40:34PM +0000, Yao Zi wrote: > On Fri, Aug 21, 2026 at 10:20:55AM -0600, Tom Rini wrote: > > On Fri, Aug 21, 2026 at 08:42:14AM +0000, Yao Zi wrote: > > > > > From: Jiaxun Yang <[email protected]> > > > > > > Yet another generic QEMU VIRT machine. > > > > > > LoongArch QEMU splits the main RAM into two parts: the first 512M is > > > placed at 0x0, and the rest is placed at 0x8000000. FDT locates at > > > 0x100000. BIOS is mapped to the first flash mapped to 0x1c000000, where > > > U-Boot starts execution. > > > > > > Signed-off-by: Jiaxun Yang <[email protected]> > > > Signed-off-by: Yao Zi <[email protected]> > > [snip] > > > diff --git a/arch/loongarch/dts/qemu-loongarch64.dts > > > b/arch/loongarch/dts/qemu-loongarch64.dts > > > new file mode 100644 > > > index 000000000000..39d303798fcb > > > --- /dev/null > > > +++ b/arch/loongarch/dts/qemu-loongarch64.dts > > > @@ -0,0 +1,9 @@ > > > +// SPDX-License-Identifier: GPL-2.0+ > > > +/* > > > + * Copyright (C) 2024 Jiaxun Yang <[email protected]> > > > + */ > > > + > > > +/dts-v1/; > > > + > > > +/ { > > > +}; > > > > I thought I had pushed everything so that we do not need to have fake > > empty dts files anymore. > > Do you mean OF_OMIT_DTB? Even with it enabled, a devicetree blob, > specified by CONFIG_dEFAULT_DEVICE_TREE, is always built, but not > integrated in the final u-boot.bin. > > qemu_arm_defconfig also ships an empty devicetree. I'm not sure whether > there's really a way to avoid building a dtb...
There is, but, OK, I didn't post a v2 of the series: https://lore.kernel.org/u-boot/[email protected]/ So, this comment doesn't apply as it's not something you can avoid doing yet. -- Tom
signature.asc
Description: PGP signature
