Hi Elliot, On 2026-08-07T18:07:09, Elliot Berman <[email protected]> wrote: > x86: qemu: Describe pflash > > QEMU pflash is mapped just below 4GiB. Describe it as a cfi-flash > node so the MTD subsystem can probe it. When a flash device is not > exposed to the guest by QEMU, the address is still backed by ROM, so > we'll just see no flash found rather than faulting. > > Signed-off-by: Elliot Berman <[email protected]> > > arch/x86/dts/pflash.dtsi | 18 ++++++++++++++++++ > arch/x86/dts/qemu-x86_i440fx.dts | 1 + > arch/x86/dts/qemu-x86_q35.dts | 1 + > 3 files changed, 20 insertions(+)
> diff --git a/arch/x86/dts/pflash.dtsi b/arch/x86/dts/pflash.dtsi > @@ -0,0 +1,18 @@ > + * The actual pflash start is driven by QEMU using, based on the size of > U-Boot ROM size. > + * We assume u-boot is loaded as the -pflash/-bios option This sentence does not parse. Please can you reword it, e.g. 'QEMU derives the pflash start from the size of the image passed with -pflash, so this matches only when U-Boot itself is the pflash/bios image'? Also 'u-boot' should be 'U-Boot' in prose. With those fixed: Reviewed-by: Simon Glass <[email protected]> Regards, Simon
