Re: Panic on boot for armv7 on rpi2

On Thu, Jan 08, 2026 at 09:44:57PM -0800, Mark Millard wrote:
> On 1/8/26 20:05, Mark Millard wrote:
> > On 1/8/26 17:05, bob prohaska wrote:
> >> On Thu, Jan 08, 2026 at 10:10:25AM -0800, Mark Millard wrote:
> >>
> >> [snippage of erroneous panic description] 
> >>>
> >>> So: it boots fine but fails to shutdown completely before failing to start
> >>> the boot.
> >>>
> >> Ahh, that clears up much confusion. 
> 
> Your shutdown related panic information is interesting. For example,
> note the examples of "=deadc0de" in:
> 
> QUOTE
> Fatal kernel mode data abort: 'Translation Fault (L1)' on read
> trapframe: 0xd3ab4b28
> FSR=00000005, FAR=deadc0de, spsr=60000013
> r0 =0000003d, r1 =1c5f0c0c, r2 =1c5f0c0c, r3 =00000000
> r4 =d742df00, r5 =00000000, r6 =deadc0de, r7 =deadc0de
> r8 =d7434080, r9 =c091e518, r10=00000000, r11=d3ab4bd8
> r12=d3ab4adc, ssp=d3ab4bb8, slr=c01b3a80, pc =c00d587c
> . . . (more such are persent) . . .
> exception_exit() at exception_exit
>          . . .
>          r6 = 0xdeadc0de  r7 = 0xdeadc0de
> END QUOTE
> 
> looks to me like a "ldr   r2, [r6]" could be the first use of the
> deadc0de (happening to be via r6) adn what is leading to the:
> 
> 'Translation Fault (L1)' on read
> 
> I do not know what lead to the deadc0de values.

Spelling out "deadc0de" in hex seems odd. Is it a
variety of intentional placekeeper, or maybe an 
easter egg of some sort?

Regarding the "sort of panic" you asked about earlier,
I simply meant to say that the newest kernel behaved
the same as the previous one, i.e., it wasn't an improvement.

There already a report of panic on shutdown for amd64,
would something like this warrant a separate bug report, or maybe
a comment to the existing one? It's at
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=291294
To my eye it looks unrelated.

Thanks for writing! 

bob prohaska