snapshots mvebu/cortexa9 possible regression in initramfs boot after acd8e94d20

[Petr Štetiar]

Stijn Segers <foss at volatilesystems.org> [2023-02-06 11:38:40]:

Hi,

> "Petr Štetiar" <ynezz at true.cz> schreef op 6 februari 2023 11:26:35 CET:
> >Hi,
> >
> >I've just noticed, that Omnia fails to boot latest initramfs snapshots[1]:
> >
> >  ## Loading kernel from FIT Image at 01000000 ...
> >     Using 'config-1' configuration
> >     Trying 'kernel-1' kernel subimage
> >       Description:  ARM OpenWrt Linux-5.15.91
> >       ...snip...
> >     Verifying Hash Integrity ... crc32+ sha1+ OK
> >  ## Loading fdt from FIT Image at 01000000 ...
> >     Using 'config-1' configuration
> >     Trying 'fdt-1' fdt subimage
> >       Description:  ARM OpenWrt cznic_turris-omnia device tree blob
> >       ...snip...
> >     Verifying Hash Integrity ... crc32+ sha1+ OK
> >   Booting using the fdt blob at 0x1543900
> >   Uncompressing Kernel Image
> >   Loading Device Tree to 0fff8000, end 0ffffdbe ... OK
> >
> >  Starting kernel ...
> >
> >acd8e94d20 seems to be the last working revision[2], so 5.15.90 booted fine,
> >5.15.91 didn't, so maybe thats the culprit?
> 
> 
> FWIW, a backported 5.15.91 boots fine here on a WRT1200AC with 22.03.

FYI I've reported issue with initramfs image as we don't flash devices during
daily automatic runtime testing. Anyway it seems, that Georgi is observing similar
symptomps on his device[1], so likely 8bc72ea7be3976 might be the real culprit
here.

1. https://github.com/openwrt/openwrt/commit/8bc72ea7be3976711dacc09f0fdab061d6e5152a#commitcomment-99705611