On Thu, 22 Jul 2021 23:44:54 +0200
Marcus Glocker <[email protected]> wrote:

> On Thu, 22 Jul 2021 23:07:30 +0200 (CEST)
> Mark Kettenis <[email protected]> wrote:
> 
> > > Date: Thu, 22 Jul 2021 22:59:28 +0200
> > > From: Marcus Glocker <[email protected]>
> > > 
> > > I can install and boot an 6.9 on it (although XHCI fails to
> > > initialize), but with a miniroot69.img snapshot of today, I'm
> > > getting a panic with the RAMDISK kernel before the installation
> > > procedure starts.
> > > 
> > > Is this something known?    
> > 
> > There was a similar report the other day.  No clue; do_el0_error
> > should never happen.  
> 
> I see.  At least when I try to boot a GENERIC kernel, I get a little
> more trace:
> 
> root on sd0a (90e0fb227dbf7852.a) swap on sd0b dump on sd0b
> WARNING: CHECK AND RESET THE DATE!
> gpio0 at bcmgpio0: 58 pins
> bwfm0: address dc:a6:32:da:a4:73
> panic: do_el0_error
> Stopped at      panic+0x160:    cmp     w21, #0x0
>     TID    PID    UID     PRFLAGS     PFLAGS  CPU  COMMAND
> * 12906  15435      0           0          0    0  init
>  195339      1      0         0x2          0    2K init
> db_enter() at panic+0x15c
> panic() at do_el0_error+0x10
> radeon_combios_check_hardcoded_edid() at handle_el0_error+0x74
> handle_el0_error() at 0x1788fadeb0
> address 0x7ffffd0828 is invalid
> https://www.openbsd.org/ddb.html describes the minimum info required
> in bug reports.  Insufficient info makes it difficult to find and fix
> bugs. ddb{0}>  
> 
> radeon_combios_check_hardcoded_edid()?  Need to dig a little deeper
> ...

When I disable bcmpcie(4) the panic is gone and the RPI4 boots up fine.
I also noticed that the bus_space_writes e.g. in bcmpcie_attach() take
very long to complete.  It's like 5-10 seconds for one bus write to
complete.  Any idea what could go wrong in that area?

Reply via email to